GIFT OF
MICHAEL REESE
ASSAYING
BY
SrOTTISWOODE AXD CO., NEW-STREET SQUARE
LONDON"
A MANUAL
OF
PEACTICAL ASSAYING
BY
JOHN MITCHELL, F.C.S.
EDITED BY
WILLIAM CBOOKES, F.E.S., PRES. C.S.
SIXTH EDITION
ILLUSTRATED WITH 201 WOODCUTS
LONDON
LONGMANS, GBEEN, AND CO.
1888
All rights reserved
PBEFACE
TO
THE SIXTH EDITION.
THE last edition of Mitchell's Assaying was published in
1881, and a new edition being required, the opportunity
has been taken of x rewriting some of the descriptions
which the progress of research had made old or ill adapted
to modern requirements. At the same time much new
matter has been introduced, and matter which had become
obsolete has been omitted. The number of woodcuts has
been increased from 188 in the last edition to 201 in the
present edition, whilst the number of pages, notwithstand-
ing numerous excisions, has grown from 809 to 896.
Among the new and important additions to this edition
may be enumerated a description of the ' Automatic
Sampling Machine,' invented by Mr. D. W. Brunton ;
many new gas furnaces and burners for the laboratory,
devised by Mr. Fletcher, Messrs. J. J. Griffin, and others ;
new blowpipe reagents and operations ; new processes,
dosimetric, volumetric, and colorimetric, for the partial
and complete assay of iron ores, iron, steel, spiegeleisen, &c.
In the copper assay a full description is given, for the
first time in this country, of the American system of fire
[vi] PREFACE TO THE SIXTH EDITION.
assay. The system adopted at Swansea is so interwoven
with the customs of the trade that its replacement by a
more accurate process is perhaps not to be expected for
some time, but for those assay ers who are not bound to
this particular process the Lake Superior system can be
strongly recommended, as being quick, inexpensive, and
comparing favourably in accuracy with the wet methods.
In the assay of silver the action of bismuth on the
ductility of this metal a subject hitherto overlooked
has received considerable attention. Much has also been
added on the subject of gold ores, a matter of large
and increasing interest to miners and metallurgists ; and
improved modes of assaying the precious metal and its
detection in poor ores are given. Besides these more
important additions and alterations, minor additions are
to be found in every chapter.
A little more prominence has been given to the English
system of grain weights, as English assayers are more
familiar with these than with the metric system in use on
the Continent ; and, as far as possible, the decimal system
has been adopted, the grain being taken as the unit. So
many valuable memoirs on assaying and metallurgical
subjects are published abroad that it has, however, been
found impracticable to adhere universally to grains.
W. C.
LONDON : April 1888.
EXTRACTS FROM THE
PEEFACE TO THE THIED EDITION.
IN this edition are incorporated all the late important
discoveries in Assaying made in this country and abroad,
and special care is devoted to the very important Volumetric
and Colorimetric Assays, as well as to the Blowpipe Assays.
Most of the chapters are entirely rewritten, whilst the
chapter on Crystallography being a subject only remotely
bearing on Assaying is left out altogether. On the
other hand, in some cases it may seem that by treating
of purely analytical details the limits -of Assaying have
been exceeded. But these departments are so closely
related as to make it impossible to fix the line of demar-
cation between them. Moreover, chemistry is cultivated
by almost all to whom this work is of interest or service,
so that it is hoped these amplifications will add to its value.
The old equivalents are retained, as they are more generally
understood by students of science who do not make
chemistry their chief study.
The Editor is under many obligations to his friend Dr.
Eohrig, M.E., for assistance in revising the manuscript and
incorporating into the work the latest continental improve-
ments, as set forth in Professor Kerl's Probirkunst. The
author of the best work on volumetric analysis which has
[viii] EXTKACTS FROM PEEFACE TO THE THIRD EDITION.
yet appeared in English, Mr. Sutton, F.C.S., has kindly
placed several cuts, &c., at the Editor's disposal, and some
descriptions of German processes have been taken from the
last English Edition of Fresenius's Quantitative Analysis a
work which should be a standard of reference for all
students who desire to carry their chemical researches
further than is possible to be treated of in a work pro-
fessing to deal only with Assaying.
LONDON : September 1868
PBEFACE
TO
THE FIEST EDITION.
WHEN the rank our country holds among nations, as
regards her Mining interest, is taken into consideration, it
must be with all a matter of surprise that no work especially
devoted to the elucidation of the processes to be employed
in ascertaining the richness in rnetal of any sample of ore
(that is, in other terms, its Assay) has of late years appeared
before the British public. Indeed, the only work at pre-
sent known in England is Berthier's ' Traite des Essais par
la Voie Seche,' which, for the mere purpose of inculcating
the principles of Assaying, has many disadvantages not
the least of which is its being written in a foreign tongue ;
and although a knowledge of French is now so very
general, yet many are prevented buying scientific works
in that language on account of the difficulties of finding
equivalents for the technicalities which must necessarily be
employed. It is also a very large work, and one contain-
ing much matter which the assayer does not need matter
relating to the composition of wood and coal ashes, furnace
products, &c., which are more especially adapted for the
metallurgist.
[x] PREFACE TO
These considerations, coupled with the paucity of any
knowledge of Assaying, excepting that confined to a very
limited number of persons, induced the author of the
following pages to turn a considerable amount of his at-
tention to this subject, more especially as much difficulty
was experienced in not having a suitable text book for the
use of his pupils. A portion of the following pages was
drawn up as a Manual for such a purpose ; but, on con-
sideration, it was thought the extension of such a work
was so much needed that it was determined to alter the
original plan as> far as was consistent with the complete
carrying out of the object in view, viz. the production of
a Manual embodying information in every branch of
Assaying, either by the wet or the dry processes.
The following is a sketch of the manner in which this
is accomplished, the author having followed the excellent
arrangement of Berthier as closely as possible, from whose
work also much matter that suited these pages, and which
it would have been useless to rewrite, has been inserted.
First, the Mechanical and Chemical Operations of Assay-
ing are treated in full, inclusive of a description of the
apparatus required, their mode of use, &c. Secondly,
Furnaces, Fuel, and Crucibles, together with a description
of the best Pyrometers, and their applications. Thirdly,
the Fluxes, their properties, preparation, use, &c. Fourthly,
an Essay on the use of the Blowpipe, and all its appurten-
ances ; as Fluxes, Supports, &c. Fifthly, the action of
the Fluxes on some Mineral Substances. Sixthly, a method
of discriminating many Minerals by means of the Blow-
pipe, aided by a few tests by the humid method. Seventhly,
the Humid Analysis of many Mineral Substances, their
composition, locality, &c. (All the minerals mentioned in
the three last heads comprehend such only as generally
THE FIRST EDITION. [xi]
came under the notice of the Assay er.) Eighthly, the
complete Assay of all the common Metals, in addition to
which the Assay of Sulphur, Chromium, Arsenic, Heating
power of Fuel, &c., is fully discussed ; and ninthly, and
lastly, a copious Table drawn up for the purpose of ascer-
taining in Assays of Gold and Silver the precise amount,
in ounces, pennyweights, and grains, of Noble metal con-
tained in a Ton of Ore from the assay of a given quantity
This Table is the most complete and copious yet published
Not only has it been endeavoured to collect all that is
generally known on the subject of Assaying, but many
new facts have been added, and such matter entered into,
that the success of an assay is rendered much more certain ;
and most assays are conducted more rapidly and with
greater exactitude than heretofore.
It has also been endeavoured to introduce a new system,
in which is pointed out the rationale of each process, with
the chemical action taking place between the fluxes and
the ores in course of assay, so that by paying a careful
attention to the matters discussed^ so much of the chemical
nature of all ores that can come under the assay er's hand
may be known, that the practice by * rule of thumb ' (a
rule on which very little dependence is to be placed, ex-
cepting after years of the most laborious practice, and a
rule which cannot be imparted, excepting the pupil pursue
the same unprofitable course) must, it is hoped, be speedily
abandoned when, by knowing the chemical properties of the
body operated on, the necessary fluxes and processes might
be at once indicated, and with a certainty of perfect success.
Having premised thus much, the author must beg to
express his thanks to his friend Mr. F. Field for the kind
[XllJ PREFACE TO THE FIRST EDITIOX.
assistance lie afforded him whilst experimenting on the
various modes of assay described in the body of the work ;
and trusting that any little imperfections which may be
detected will not be harshly criticised, but that it may be
taken into consideration that the author has attempted to
improve a branch of mining knowledge to which unfor-
tunately too little attention has been devoted, and to which,
if he has added anything useful, he is indebted for the first
principles of such knowledge to Berthier's ' Traite des
Essais,' for which, to the talented writer of the above
work, he is under the most lasting obligation.
CONTENTS.
CHAPTER I.
PAGE
Chemical nomenclature Laws of combination, etc. . . 1
Metallic and non-metallic elements .... 2
Oxides ........ 4
Salts ....... .. 4
Binary compounds containing no oxygen . ... 5
Laws of combination ...... 5
Chemical symbols ; their employment and uses . . ,7
CHAPTER II.
Preparation of the sample. Weighing . ... 9
Automatic sampling machine .' . . . .11
Anvil and stand, 14. Hammers, 15. Cold chisel, 16.
Shears, 17.
Pestle and mortar . . . . . . .17
Iron mortar, 17. Wedgwood mortar, 17. Porcelain mortar,
18. Steel mortar, 19. Agate mortar, 20.
The sieve ........ 21
Elutriation . . . . . . .22
Washing, dressing, or vanning . . . . .23
The balance . . . . . . .26
Operation of Aveighing, 26. Bullion balance, 27. Rough
assay balance, 28. Assay balance, 28. Conditions of accu-
racy and delicacy, 29.
The weights . . . . . . .34
Assay weights for silver, 35. Assay weights for gold, 35.
Method of weighing . . . . .36
Mayer's method, 39.
Incinerating precipitates before weighing . . . .41
[xiv] CONTENTS.
CHAPTER III.
PAGE
General preparatory chemical operations . . . .42
Calcination . . . . . . .42
Use of Crucibles, 42.
Roasting . . . . . . . .44
Roasting test, 44. Roasting in crucibles, 45. Roasting in
platinum capsules, 46.
Reduction ...... , . 46
Cementation, 47. Reduction by hydrogen 48.
Fusion ........ 48
Solution ........ 49
Glass and platinum forceps, 50.
Distillation . . . . . . .51
Liquid distillation, 51. Dry distillation, 53. Sublimation,
54.
Scorification Cupellation. . . . . .54
CHAPTER IY.
Production and application of heat . . . .55
Calcining furnace . . . . . . .55
Chimney . . . . . . . .56
Evaporating furnaces . . . . . .57
The hood ........ 57
Fusion furnaces. Wind furnaces. . . . .57
Body of furnace, 58. Ashpit, 57. Bars, 58. Chimney, 59.
Blast furnaces . . . . . . .60
Royal Institution furnace, 60. Sefstrom's blast furnace, 61.
Deville's furnace, 62.
Muffle or cupel furnace . . . . . .62
Aufrey and d'Arcet's furnar ^0. Brick furnace for numbers
of cupellations, 64. Ur .sal furnace, 65.
Furnace operations . . . . .67
Auxiliary a^^aratus . . . . . .67
Poker- or stirring rods, 67. Tongs, 67. Spectacles and look-
ing glass, 68. Curved rods, 68. "Wrought iron ladle, 69.
Ingot moulds, 69.
Fuel for furnaces . . . ... . .69
Coke, 70. Charcoal, 70.
Effects produced by wind and blast furnaces . . .72
Oil and gas blast furnaces . . . . . .74
Oil furnaces . . . . . . .74
Description of the apparatus, 74. Oil lamp furnace, 75.
Oil reservoir, 75. Crucible furnace, 75. Management of
the oil lamp furnace, 76. Power of the oil lamp furnace,
79. Requisite blowing power, 79.
CONTENTS. [xv]
PAGE
Griffin's gas furnace t . . . . .80
Operations in crucibles, 80. Operations in muffles, 89.
Skittle-pots, 83. Flue, 83. Distillation of zinc, 84. Extra
large gas furnace, 84.
Fletcher's universal gas furnace . . . . ,84
Burner, 84. Single jacketed arrangement, 84. Double
jacketed chemical furnace, 84.
Brown's gas assay furnace . , . . . .87
Fletcher's reverberatory gas furnace . . . .90
Fletcher's muffle and draught crucible furnaces . . .92
Fletcher's new melting arrangement . . . .92
Injector gas furnace, 92. The burner, 93. ' Salamander '
crucibles, 93. Injector gas or spirit furnace, 94. Benzo-
line, 95.
Gore's gas furnace . . . . . . .95
Griffin's reverberatory gas furnace . , .100
Melting furnace for lead, tin, antimony, etc. . . .104
Bunsen's gas burner, 105. Mounts for crucibles, 108. Gas
furnace for boiling or evaporating, 110. Solid flame gas
burner, 111. Dirt proof high power burner, 112. Safety
Bunsen burner, 113.
Lutes and cements . . . . . .113
Fire lute, 113. Fat lute, 113. Roman cement, 114. Plas-
ter of Paris, 114. Linseed or almond meal, 114. Lime
and egg lute, 114. White lead with oil, 114. Yellow
wax, 114. Soft cement, 114. Cement for brass or glass,
114. Cement for mending pestles, 115. Adhesive paste,
115. Waterproof cement, 116. Resinous or hard cement,
117. Caoutchouc, 117. Faraday's directions for luting
iron, glass, or earthenware retorts, tubes, etc., 118,
Willis's cement, 119. Iron cemep* , -\J 9. Beal's cement,
120. Boiler cement, 120. Bruyei rement, 120. Oxy-
chloride of zinc cement, 1 20.
Crucibles, cupels, etc. . . . . . V ' 1 20
Hessian, Cornish, Stourbridge, and London clay crucibles, 1 20.
' Salamander ' brand of crucibles and plumbago fittings,
121. Porcelain crucibles, 122. Black-lead crucibles, 123.
Charcoal crucibles, 126. Lined crucibles, 127. Lime
crucibles, 128. Forbes's experiments with lime and char-
coal crucibles, 129. Alumina crucibles, 130. Magnesia
crucibles and bricks, 130. Malleable iron crucibles, 135.
Platinum crucibles, 135. Preserving platinum crucibles,
135, 137, 138. Dexter's lamp for platinum crucibles, 135.
Silver crucibles, 140. Nickel crucibles, 140.
Cupels . . 141
Cupel mould, 142. Powder for cupels, 143. Scorifiers, 144.
Methods of measuring the heat of furnaces . . .144
[xvi] CONTENTS.
PAGE
Wedgwood's pyrometer, 144. Daniel's pyrometer, 144.
Wilson's pyrometer, 147. Siemens's pyrometer, 148.
Comparing the temperatures of two furnaces, 149.
CHAPTER Y.
Fuel, its assay and analysis . . . . .150
External appearance of fuel, its porosity, compactness, fracture,
size, and shape of pieces . . . . .152
Estimation of adhering water . . . . .152
Estimation of specific gravity . . . . .153
Estimation of absolute heating power . . . .154
Pyrometric heating power, 154. Different modes of ascer-
taining absolute heating power, 154. Berthier's method,
" 155. ' lire's method, 157. lire's calorimeter, 158.
Wright's calorimeter, 159.
Estimation of specific heating power . . . .160
Estimation of pyrometric heating power . . . ..160
Estimation of the volatile products of carbonisation . .160
Examination of the coke or charcoal left behind on carbonisation 161
Estimation of the amount of ash . . . . .162
Estimation of the amount of sulphur . . . .162
Examination of other peculiarities of fuel . . .164
Calculation of results . . . . . .164
Assay of coal before the blowpipe . . . . .165
Valuation of coal for production of illuminating gas - . .167
CHAPTER VI.
Reducing agents . . . . . . .169
Hydrogen gas, 169. Carbon, 170. Black-lead or graphite,
170. Estimation of the value of graphite, 170. Anthra-
cite, 171. Coke, 171. Wood charcoal, 171. The fatty
oils, 172. Tallow, 172. Resins, 172. Sugar, 173.
Starch, 173. Gum, 174. Tartaric acid, 174. Oxalic
acid, 174. Ammonium oxalate, 174. Comparative reduc-
ing powers of the above agents, 175. Metallic iron, 175.
Metallic lead, 176.
Oxidising agents . . . . . . .176
Litharge, 176. Ceruse or white lead, 176. Lead silicates
and borates, 177. Potassium and sodium nitrates. 177.
Assay of saltpetre, 178. Lead nitrate, 180. Manganese
peroxide, 181. Copper oxide, 181. Iron peroxide, 181,
The caustic alkalies, potash and soda, 181. Potassium
and sodium carbonates, 181. Lead, copper, and iron sul-
phates, 181. Sodium sulphate, 181.
CONTENTS, [xvii]
PAGE
Desulphurising agents . . . . . .181
Atmospheric oxygen, 182. Charcoal, 182. Iron, 182.
Litharge, 182. Behaviour of litharge with sulphides of
manganese, 183 ; iron, 183 ; copper, 185 ; antimony, 186 ;
zinc, 186 ; lead, 187. Caustic alkalies and their carbonates,
187. Nitre, 188. Lead nitrate, 188. Lead sulphate,
188.
Sulphurising agents . . . . . .189
Sulphur, 189. Cinnabar, 189. Galena, 189. Antimony
sulphide, 189. Iron pyrites, 190. Alkaline persulphides,
190.
Fluxes . 191
Non-metallic fluxes . . . . . .191
Silica, 192. Lime, magnesia, alumina, and their silicates, 192.
Alumina, 193. Borax, 193. Glass, 193. Analysis of
' different kinds of glass, 194. Fluor-spar, 195. Potassium
and sodium carbonates, 195. Potassium nitrate, 196.
Common salt, 196. White flux, black flux, and raw flux,
197. Argol or cream of tartar, 199. Salt of sorrel, or
potassium binoxalate, 199. White and mottled soap, 199.
Reducing power of the various fluxes, 200.
Metallic fluxes . . . . . . .201
Litharge and ceruse, 201. Glass of lead (lead silicate), 201.
Lead borate, 201. Lead sulphate, 201. Copper oxide,
201. The iron oxides, 201.
CHAPTER VII.
The blowpipe and its uses ...... 202
Blowpipe, 203. Trumpet mouthpiece, 204. Nipples, 205.
Lamps and oil, 206. Faraday's directions for using the
blowpipe, 206. Oxidation, 209. Reduction, 209.
Auxiliary blowpipe apparatus . . . . .210
Supports, 210. Charcoal, 210. Platinum, 211. Wire, 211.
Forbes's instruments for preparing charcoal, 211. Plati-
num spoon, 212. Aluminium support, 212. Asbestos
cardboard, 213.
Reagents and fluxes . . . . . .213
Blue litmus paper, 213. Reddened litmus paper, 213.
Brazil-wood paper, 213. Turmeric paper, 213. Nitric
acid, 213. Zinc, 213. Copper, 213. Iron wire, 213.
Potassium cyanide, 214. Sodium carbonate, 214. Re-
duction of metallic oxides, 216. Borax, 218. Sodium
ammonio-phosphate, 219. Potassium nitrate, 220. Potas-
sium bisulphate, 220. Sodium bisulphate, 221. Vitrified
boracic acid, 221. Cobalt nitrate, 221. Nickel oxalate,
a
[xviii] CONTENTS.
PAGE
221. Copper oxide, 221. Silica, 222. Turner's flux, 222.
Ammonium fluoride, 222. Calcium fluoride and sulphate,
223. Bone ashes, 224. Proof lead, 224. Tinfoil, 224.
Dry silver chloride, 224. Tincture of iodine, 229. Dry
silver iodide, 231. Soda-paper, 233. Forbes's soda-paper
apparatus, 233.
General routine of blowpipe operations . . . .234
Size of the assay, 234. Order of blowpipe operations, 234.
Discrimination of minerals . . . . .237
Crystalline form, 237. Mode of fracture, 238. Lustre, 239.
Colour and streak, 239. Hardness, 240. Specific gravity,
241. Sonstadt's solution, 242. Fusibility, 244. Chemical,
characters, 244. Colour of borax bead, 246. Appearance
of reduced bead, 247. Requirements for testing minerals,
248.
Description of minerals . . . . . .248
Quartz and the silicates, 248. Rock crystal, 248. Amethyst,
248. Rose quartz, 250. Cairngorm, 250. False topaz,
250. Chalcedony, 250. Carnelianand sard, 250. Agate, 250.
Onyx or sardonyx, 250. Flint or hornstone, 250. Jasper,
250. Bloodstone, 250. Opal, 250. Talc, 250. Mica, 251.
Chlorite, 251. Serpentine, 251. Meerschaum and ne-
phrite, 251. Augite and horneblende, 252. Chrysolite, or
olivine, 252. Tourmaline, 252. Garnet, 253. Topaz,
253. Beryl or emerald, 254. Zircon or hyacinth, 254.
Felspar, 255. Mica, 255. Zeolites, 255. Corundum or
sapphire, 256. Spinel, 256. Chrysoberyl, 257. Diamond,
257. Graphite or black-lead, 258. Coal, 258. Apatite,
258. Fluor-spar, 259. Calcspar, 259. Magnesite, 260.
Dolomite, 260. Aragonite, 260. Rock salt, 261. Solu-
ble sulphates, 261. Nitre, 261. Gypsum, selenite, ala-
baster, 261. Heavy spar or barytes, 262. Sulphur, 262.
Tin ore, 262. Molybdenite, 263. Bismuth, 263. Anti-
mony sulphide, 263. Arsenic, 264. Arsenic sulphide,
264. Native iron, 265. Iron pyrites, 265. Arsenical
pyrites, 265. Magnetic iron, 266. Specular iron, hema-
tite, or micaceous iron, 266. Red ferric oxide or red
ochre, 266. Brown ferric oxide, or brown ochre, 267.
Titanic iron, 267. Chrome iron, 267. Green-earth, 267.
Iron carbonate, 268. Manganese ores, 268. Arsenical
nickel, 268. Smaltine or tin-white cobalt, 269. Cobalt
bloom, 269. Blende, 269. Zinc carbonate, 270. Zinc
silicate, 270. Galena, 270. Lead carbonate or cerusite,
270. Pyromorphite, 271. Lead sulphate, 271. Cinnabar,
272. Native mercury or quicksilver, 272. Native
copper, 273. Vitreous copper, 273. Copper pyrites, 273.
Grey copper, 274. Black cupric oxide, 274. Red
CONTENTS. [xix]
PAGE
cuprous oxide, 274. Copper carbonates, blue and green,
275. Platinum, 275. Gold, 275. Silver, 276. Silver
sulphide, 276. Antimonial and arsenical silver ores, 276.
Horn silver, 277. Lead and antimony sulphides, 277.
Mercury, lead, silver, or copper selenides, 277. Millerite
or nickel sulphide, 277. White nickel, 278. Rutile, 278.
Sphene, 278. Wolfram, 278. Pitchblende, 278. Mag-
netic iron pyrites, 279.
Determination of minerals . . . . .279
Scheme for the determination of minerals .... 280
Group I. Minerals which have a metallic lustre, and which give
off sulphur . . . . . . .281
Group II. Minerals which have a metallic lustre, and which give
off either an odour of garlic without sulphur, or white
fumes which have not a garlic or sulphurous odour . 283
Group III. Minerals which have a metallic lustre, and which
give off no fumes ...... 284
Group IV. Minerals which possess a non-metallic lustre, a
coloured streak, and which give off fumes or odour when
heated before the blowpipe .... 286
Group V. Minerals which possess a non-metallic lustre, a
coloured streak, but which give off no fumes or odour
before the blowpipe . . . . .287
Group VI. Minerals which have a non-metallic lustre, and which
are scratched by quartz, showing a white streak . . 288
Group VII. Minerals which have a non-metallic lustre, and which
are not scratched by quartz . . . .291
Coloured flames . . . . . . .294
Blue flames, 294. Green flames, 294. Yellow flames, 295.
Red flames, 295. Chlorine, 295. Lead, 295. Arsenic,
295. Selenium and antimony, 295. Bromine, 295. Boracic
acid, 295. Tellurium, 296. Copper, 296. Iodine and
copper, 296. Phosphoric acid, 296. Baryta, 296. Zinc,
296. Soda, 296. Water, 296. Strontia, 296. Lithia,
296. Lime, 297. Potash, 297.
CHAPTER VIII.
Volumetric analysis . . . . . . .298
Reactions of volumetric analysis, 298. Principle of volume-
tric analysis, 298. Standard solutions, 302.
Instruments and apparatus, 302. The burette, 303. Mohr's
burette, 304. Gay-Lussac's burette, 304. Britton's burette,
305.
Modes of operating, 305. The pipette, 306. Measuring
flasks, 307.
a 2
[xx] CONTENTS.
CHAPTER IX.
PAGE
The assay of iron ....... 308
Ores of iron ....... 308
Magnetic iron ore, 308. Red hematite, 308. Brown hema-
tite, 308. Spathic iron ore, 308. Titaniferous iron ore,
308. Franklinite, 308. Clay band and black band iron
stone, 308.
Assay of iron in the dry way ..... 309
Classification of iron ores, 309. Fluxes, 309. Air furnaces, 311.
Crucibles, 311. The charge, 313. Ores of unknown com-
position, 313. Ores previously analysed, 314. The opera-
tion, 315.
Effects produced by the following substances :
Manganese, 317. Titanium, 317. Phosphorus, 317. Sul-
phur, 317. Chromium, 317.
Calculation of results . . . . . .319
Assay of iron and its ores in the wet way . . .321
Dr. Penny's process . . . . . .321
Reduction by zinc, 327. Reduction by stannous chloride, 328.
Reduction by ammonium bisulphite, 328.
Titration of iron with sodium hyposulphite, 330. Oudemans's
method, 330. Haswell's method, 330.
The complete assay of iron ores ..... 332
The mechanical treatment, 332. The chemical treatment,
334. Estimation of phosphoric acid, 334. Sulphur and
iron, 338. Silica, ferric oxide, alumina, manganese, lime,
and magnesia, 341. Nickel, cobalt, and zinc, 344. Esti-
mation of ferrous oxide, 346. Ferrous oxide in insoluble
silicious matter, 347. Alumina, 349. Calculation of the
analysis, 349. Carbonic acid, 349. Water and carbon in
carbonaceous matter, 351. Alkalies, 355. Copper, lead,
arsenic, and antimony, 356 Titanic acid, 358. Estima-
tion of specific gravity, 360.
Gooch's perforated crucible . . . . .361
Estimation of carbon, silicon, phosphorus, &c., in metallic iron and
steel, 363. Boring and sampling, 363. Estimation of total
carbon, 367. Weyl's method, 368. Estimation of gra-
phite, 369. Dr. Eggertz's method, 369. Mr. Tosh's
method, 370. Estimation of combined carbon, 371. O.
Arnold's method, 373. J. Blodgett Britton's method, 374.
Mr. Stead's method of estimating minute quantities of
carbon, 376. A new form of chromometer, 380.
Preparation of inorganic standards for the colorimetric carbon
test . . . . . . . .381
Day standard colours, 382. Night standard colours, 386.
Estimation of sulphur in iron and steel .... 388
CONTENTS. [xxi]
PAGE
Dr. Eggertz's method, 388. J. Wiborgh's method, 393.
Preparation of the cloth, 394.
The colour scale, 398. Details of the process, 401. Estimation
of sulphur in iron ores, 405.
Estimation of silicon in iron and steel . . . .407
Mr. Turner's method . . . . . ,413
Estimation of basic cinder and oxides in manufactured iron . 414
Mr. Bettel's process . . . . . .415
Estimation of phosphorus in iron and steel , . .416
M. Tantin's method, 416. J. B. Mackintosh's method, 417.
J. Lawrence Smith's method, 417. Quantity of iron em-
ployed, 419. Solution, 420. Concentration of the phos-
phorus, 420. Separation of the phosphorus, 421.
Estimation of manganese in iron . . . . .422
E. Riley's method, 424. The direct method, 424. The in-
direct method, 425. W. Kalman and A. Smolka's method,
426. Dr. Peter's method, 427. Mr. Galbraith's method,
429.
Estimation of titanium in iron . ... . . 430
Mr. Riley's method, 430. Mr. Bettell's method, 432.
Turner's table of the hardness of iron and steel 433
CHAPTER X.
The assay of copper ....... 434
Classification of minerals and substances containing copper . 434
Class I. Sulphuretted ores or products with or without sele-
nium, antimony, or arsenic, 434. Copper glance, 434.
Chalcopyrite, 434. Erubescite, 434. Bournonite, 434.
Fahlerz, 434. Covelline, 434. Wolfsbergite, 434. Do-
meykite, 434. Copper regulus, copper speiss, etc., 434.
Class II. Oxidised ores and products, 434. Red copper, 434.
Malachite, 434. Azurite, 434. Cyanosite, 434. Phos-
phate of copper, 434. Arseniates, 434. Chromate, vana-
date, and silicate of copper, slags, etc., 434.
Class III. Copper and its alloys, 434.
Classification of different methods of assaying copper . . 435
A. Assay in the dry way, 435.
a. For rich ores and products of Class I., 435.
English copper assay ...... 435
Moissenet's description, 435. Ticketing in Cornwall, 435.
De la Beche's sketch of the method, 436. Division
adopted, 436.
Section I. Reactions, 437. Two kinds of assays roasted and
raw sample, 437.
[xxii]
CONTENTS.
PAGE.
1. Regulus, 438. Pyrites, 438. Very poor pyrites, 438.
Variegated copper ore, 439. Sulphide of copper, 439.
Carbonated minerals, 439. Native mixture, 439.
2. Calcining, 440.
3. Coarse copper, 440.
4. Washings, 441.
5. Testing, refinery, 442.
6. Slags for prill, 443.
Section II. Manipulations, 443. Crucibles used in Cornwall,
444. Wind furnace, 444. Fusion for regulus, 445. Cal-
cining the matt, 448. Coarse copper, 449. Washings^
450. Testing and refining, 450. Prill, 450.
Section III. Some minerals and substances of a special nature ;
influence of foreign metals, 451. Stanniferous minerals,
451. Antimonial minerals, 451. Zinciferous minerals, 452.
Plumbiferous minerals, 452. Regulus of Chili, 452. Slags
of copper, 452. Old copper, 452.
Section IY. Summary considerations comparison of the
results with the analysis by the wet way, 453. Principal
causes of loss in the Cornish method, 454. Assay of
copper in the dry way, 454.
b. For ores and products of Class II., 455.
Lake Superior fire assay ...... 455
Dr. Peter's description, 455. Sampling, 456. Fluxes, 457.
Furnace, 457.
B. Assays in the wet way, 461.
a. Colorimetric copper assay, 461.
Heine's method, 461. Le Play's method, 46 7. T.O. Cloud's
method, 467. Endemann's method, 467. Carnelly's
method, 470. Standard copper solution, 470. Solution
of, 472. Jacquelain's and Von Hubert's colorimetric
assay, 472.
b. Volumetric copper assays, 476.
Fleck's modification of Mohr's method, 476. E. 0. Brown's
method by sodium hyposulphite, 479.
c. Electrolytic copper assay, 480.
Estimation of copper in the Mansfield ores by Dr. Steinbeck's
process ....... 480^
Extraction of the copper from its ores, 480. Separation of
the copper, 481. Quantitative estimation of the precipi-
tated copper, 482. Special observations on this method,
483.
Estimation of copper in the Mansfield ores by M. C. Luckow's
process . . . . . . : , 487
Roasting the ore, 489. Solution of the roasted product, 489.
Precipitation of the copper, 490. Weighing the copper, 493.
Assay of copper pyrites . . . . . .494
CONTENTS. [xxiii]
PAGE
Detection of traces of copper in iron pyrites and other bodies . 497
Estimation of arsenic in copper . . . . .499
CHAPTER XI.
The assay of lead . . . . . ,502
Classification of minerals and substances containing lead . . 502
Action of reagents on lead sulphide, 502. Oxygen, 502. Me-
tallic iron, 502. Alkalies and alkaline carbonates, 503.
Potassium nitrate, 503. Argol, 503.
Assay of substances of the first class (sulphides, antimonides, etc.) 504
Processes generally adopted :
1. Fusion with potassium carbonate, 504.
2. Fusion with black flux, 510.
3. Fusion with metallic iron, 511.
4. Fusion with sodium carbonate, or black flux and metallic
iron, 514.
5. Roasting and reducing assay, 515. Assay with black
flux and iron, 517. Roasting and reduction assay
with iron, 518. Roasting and fusing with black flux,
518. Level's fusion assay with potassium ferrocyanide
and cyanide, 519.
6. Assay with sulphuric acid, 519.
7. Assay of galena in the wet way, 521.
Assay of substances of the second class . . . .523
Humid assay of ores of the second class . . . .526
Assay of substances of the third class . . . .527
Humid assay of substances of the third class . . . 528
Maxwell Lyte's process, 529. Mascazzini's process, 529.
Jannesay's process, 529.
Alloys of lead (Class IV.) 530
Assay with sulphuric acid, 530.
Level's fusion assay with potassium ferrocyanide and cyanide . 531
Estimation of lead by means of standard solutions . . 531
1. Flores Dumonte's method, 531.
2. Schwartz's method, 532.
3. Buisson's volumetric process, 533.
4. Diehl's volumetric process, 534.
CHAPTER XII.
Assay of tin ..... 537
Tin ores ...... 537
Tin oxide, 537. Crystallised tin oxide, 537. Disseminated
tin oxide, 537. Sandy tin oxide, 537. Concretionary tin
oxide, wood tin, 537.
[xxiv] CONTENTS.
PAGE
Analysis of a sample of tin oxide from Cornwall, 538.
Remarks on tin ore and the minerals which may be mistaken
for it, by Dr. A. Leibius, 538.
Assay of pure tin oxide ...... 539
Method used in Cornwall, 540. Method by means of potas-
sium cyanide, 540.
Assay of tin oxide mixed with silica .... 542
Assay of tin ores containing silica and tin slags . . .542
Assay of tin ores containing arsenic, sulphur, and tungsten . 543
Approximative assay ...... 544
J. H. Talbot's method for assaying tin in the presence of
tungsten, 546.
Estimation of tin by the humid method .... 546
Klaproth's process, 546. J. B. Hallet's process, 548. M.
Moissenet's process, 548.
Assay of tin in gun and bell metal .... 549
E. Burse's method, 550.
Estimation of tin by means of a standard solution . . 550
M. Gaultier de Claubry's process, 550. M. Lenssen's process,
551. M. Stromeyer's process, 551. Method for the analysis
of tin ore, 552. Another method, 553. Decomposition of
tin slags, 553. MM. Mohr and Terreil's method, 554.
Rammelsberg's method, 554.
CHAPTER XIII.
Assay of antimony . . . . . .556
Classification of antimonial substances .... 556
Class I. Native antimony and all antimonial substances con-
taining oxygen or chlorine, but little or no sulphur :
Native antimony, 556. Antimony oxide, 556. Antimo-
nious acid, 556. Antimonic acid, 556.
Class II. Antimony sulphide and all antimonial ores contain-
ing sulphur : Antimony sulphide, 556. Antimony oxy-
sulphide, 556. Haidingerite, 556.
Assay of ores of the first class ..... 556
Assay of ores of the second class . . . . .557
1. Estimation of the pure antimony sulphide (antimonium
crudum) . . . . . . . 557
2. Estimation of regulus of antimony . . . 558
Two methods, 558.
a. By roasting and fusing the oxidised matter with
black flux, 558.
b. By fusing the crude ore with iron, or iron scales,
with or without the addition of black flux, 558.
F. Becker's method, 563. Mr. Button's method, 563.
Detection of antimony in sublimates .... 563
CONTENTS. [XXV]
PAGE
To distinguish arseniuretted hydrogen from antimoniuretted
hydrogen ....... 564
Separation of tin from antimony and arsenic . . . 564
0. Winckler's method, 564.
Assay of alloys of lead and antimony (type metal) . . 565
CHAPTER XIV.
Assay of zinc ....... 567
Classification of bodies containing zinc . . . .567
Class I. Zinc ores in which the metal exists "as oxide not
combined with silica, 567.
Class II. Zinc ores in which the metal exists as oxide, but
partly or wholly combined with silica, 567.
Class III. Zinc ores in which the metal is partly or wholly
combined with sulphur, 567.
Class IV. Alloys, 567.
Assay of ores of the first class . . . . .567
Assay of zinc by the humid process in ores of the first class . 571
Assay of ores of the second class . . . . .572
Wet assay of zinc in ores of the second class . . ,572
Assay of ores of the third class . . . . .573
Humid wet assay of zinc in ores of the third class . . 573
Assay of cupriferous blende ..... 574
Alloys (Class IY.) 574
Volumetric assay of zinc . . . . . .575
Galetti's process, 575. Schaffner's method, modified by C.
Kunzel, 577.
a. Solution of the ore, and preparation of the ammoniacal
solution, 577.
6. Preparation and standardising of the sodium sulphide
solution, 578.
c. Assay of the zinc in the solution of the ore, 579.
d. Further modification of the process, 580.
H. Schwarz's method, 581. Carl Mohr's method, 582. J.
Drewson's method, 583.
Separation of copper from zinc, 584.
CHAPTER XV.
Assay of mercury ....... 587
Assay of mercurial ores ...... 587
Berthier's method, 590. Eschka's method, 591.
Assay for the amount of cinnabar in an ore . . . 592
Electrolytic assay of mercury ..... 594
Volumetric estimation of mercury . . . .595
M. J. Personnel method, 595. M. Rivot's method, 597.
G. Attwood's method, 598.
XXvil CONTENTS.
CHAPTER XVI.
PAGE
Assay of silver ....... 600
Classification of argentiferous substances . . . .600
Class I. Minerals containing silver, 600.
Class II. Metallic silver and alloys, 600.
General observations on the assay of ores and substances of
Class I. . . . . . .600
Fusion with oxidising reagents ..... 603
Litharge, 603.
Special directions for the crucible assay of ores and substances of
Class I. ....... 605
Preliminary assay for dividing all substances of this class
into three sections, 606. Assay of reducing power of
argol, 607. Assay of oxidising power of potassium
nitrate, 607. Assay of litharge for silver, 608.
Assay of ores of the first section ..... 608
Assay of ores of the second section . . . .609
Assay of ores of the third section ..... 609
Scorification . . . . . . .610
The roasting, 612. The fusion, 612. The scorification, 613.
Special instructions for the scorification assay of ores of Class I. . 615
Assay in scorifier, 615.
Assay of substances of Class I. admixed with native or metallic
silver ........ 616
Cupellation . . . . . . .617
Pliers and microscope, 623.
Amalgamation . . . . . . .626
Substances of the second class . . . . .627
Separation of silver from galena . . . . .627
General remarks on the assay of the alloys of silver and copper . 627
Cupellation, 627. D'Arcet's results, 628. Quantity of lead
required, 629. Loss of silver in the assay of coined
alloys, 630.
Special instructions for the assay of the alloys of silver and copper 631
Assay for approximative quantity of alloy, 631. Assay proper
of silver bullion, 631. Assay of alloys of copper and
silver, 632.
Alloys of platinum and silver . . . . .632
Alloy of platinum, silver, and copper .... 633
Assay of native silver, rough silver left on sieve during pulverisa-
tion of silver ores of first class, and native alloys of silver . 633
Dr. W. Dyce's process for separating gold and silver from the
baser metals ....... 633
Assay of silver bullion by the wet method . . .634
CONTENTS. [xxvii]
PAGE
Gay-Lussac's method ...... 634
Measurement of the solution of common salt, 636. Measure
of the normal solution of salt by weight, 636. Preparation
of the decime solution of common salt, 638. Preparation
of the decime solution of silver, 640. Weighing the normal
solution of common salt, 641. Preparation of the normal
solution of common salt when measured by weight, 641.
Preservation of the normal solution of common salt, 646.
Application of the process described in the determination
of the standard of a silver alloy, 647. Correction of the
standard of the normal solution of salt when the tempera-
ture varies, 649. Table of corrections for variations in
temperature of the normal salt solution, 651.
Table for the assay, by the wet method, of an alloy containing any
proportions whatever of silver, by the employment of a con-
stant measure of the normal solution of common salt . 651
Tables for determining the standard of any silver alloy by employ-
ing an amount of alloy always approximatively containing
the same amount of silver . . . . 654, 673
Assay of pure, or nearly pure silver, the temperature of the
normal solution of salt being that at which it was standard-
ised ........ 674
First example, 674. Second example, 675. Third example,
676.
Graduation of the normal solution of salt, the temperature being
different to that at which it was wished to be graduated . 677
First example, 677. Second example, 678.
Approximative determination of the standard of an unknown
alloy ........ 678
Modes of abridging manipulation . . . . .679
Bottles, 679. Stand, 680. Water-bath, 680. Flue, 681.
Agitator, 681. Table, 683. Cleaning the bottles, 684.
Reduction of silver chloride obtained in the assay of alloys by the
wet method . . . . . . . 684
Preparation of pure silver . . . . . .685
Modifications required in the assay of silver alloys containing
mercury ....... 68fr
Method of taking the assay from the ingot, 687. Method of
assaying silver bars adopted in the assay offices of H.M.
Indian mints, 687. The chloride process, 688. Apparatus
and appliances required, 697.
Effect of bismuth on the ductility of silver . . . 699'
Titration o^ silver in presence of other metals . . .711
Copper, 713. Mercury, 714. Palladium, 714.
Determination of silver in galena by Volhard's process . .714
David Forbes on blowpipe assay of silver . . . .715-
Apparatus, 716. Concentration of the silver lead, 717. Cu-
[xxviii] CONTEXTS.
PAGE
.pellation, 720. Estimation of the weight of the silver
globule obtained in cupellation, 723. The scale for this
purpose, 724. Cupellation loss, 727. Table modified by
Plattner, 728.
Classification of argentiferous substances . . . .729
A. Metallic alloys capable of direct cupellation, 730.
a. Consisting chiefly of lead or bismuth, 730.
6. Consisting chiefly of silver, and alloys of silver with
gold and copper, 731.
c. Consisting chiefly of copper, 732.
B. Metallic alloys incapable of direct cupellation, 733.
a. Containing much copper or nickel, with more or less
sulphur, arsenic, etc., 733.
b. Containing tin, 734.
c. Containing antimony, tellurium, or zinc, 735.
d. Containing mercury ; amalgams, 737.
e. Containing much iron, 737.
Alloys of silver and copper, 739.
CHAPTER XVII.
Assay of gold ....... 740
Classification of substances containing gold . . .740
Class I. Ores containing gold, 740. Graphic tellurium, 740.
Folliated tellurium, 740.
Class II. Alloys of gold, 740. Native gold and aurides of
silver, 740. Artificial alloys of gold, 740.
Assay of gold ores . ...... 741
Preparation of the sample, 742. Collection of the gold and
silver, 742. Crucible assay, 742. The charge, 743. Size
of lead button, 743. Preliminary assay of ore, 744.
Roasting the ore, 745. Fusion, 747. Scorification assay,
747. The lead button, 749. Cupellation, 749.
Estimating the weight of minute spheres of gold . . . 753
Oeneral observations on the assay of gold ores . .. .756
Gold and copper, proportion of lead, 756. Examination on
the touchstone, 757. Table for the proportion of lead to
be employed in the cupellation of gold and copper, 759.
Gold, silver, platinum, and copper, 759. Gold alloyed with
silver, 761. Inquartation, 761. Surcharge, 763. Mr.
Seine's method, 763. Makin's method, 765. Aqua regia,
766. Yon Jiiptner's method, 767.
Standard of the alloys of gold . . . . .768
Assay of gold coin and bullion . . . . .769
Preliminary assay, 769. Assay proper, 769. Parting assays,
771
CONTENTS.
PAGE
Assay of pyrites for gold . . . . . .773
Treatment of gold and silver bearing copper ores . . .775
Detection of minute traces of gold in minerals . . . 777
CHAPTER XVIII.
Assay of platinum . . . . . . .781
Platinum in its native state . . . . .781
Analysis of platinum ores . . . . . .782
Bunsen's method, 782. Platinum and palladium, 782. Ru-
thenium, rhodium, and iridium, 785. C. Lea's process,
790. Deville and Debray's process, 795. Sand, 795. Osm-
iridium, 796. Platinum and iridium, 797. Palladium, iron,
and copper, 798. Gold and platinum, 799. Rhodium, 799.
Analysis of platinum ores from various sources, 799.
Guyard's process for extracting metals from platiniferous
residues, 799. Analysis of osm-iridium, 803. Wolcott
Gibbs's process, 804. Nelson Perry's process, 807.
CHAPTER XIX.
Assay of bismuth ....... 809
Varieties of bismuth ores ...... 809
Native bismuth . . . . . .809
Assaying bismuth ores . . . . , .810
Assaying bismuth in ores containing a large amount of copper . 810
Refining crude bismuth . . . . . .813
Purification of the reduced bismuth . . . .814
Purification of bismuth from arsenic, 814. Purification of
bismuth from antimony, 815. Purification of bismuth
from copper, 815. Purification of bismuth from sulphur,
816
Volumetric assay of bismuth . . . . .817
R. W. Pearson's process, 817. M. P. Muir's process, 818.
CHAPTER XX.
Assay of chromium ...... 820
Principal ore of chromium . . . . . 820
Assay of chrome ore ...... 820
Genth's process, 820. O'Neill's process, 822. W.Gibbs's pro-
cess, 823. Clark's process, 824. H. N. Morse and W. C.
Day's process, 825
Estimation of chromium by means of standard solutions . . 827
Estimation of chromium in iron and steel, 827. J. O.
Arnold's process, 827. W. J. Sell's process, 830.
[XXX] CONTENTS.
CHAPTER XXI.
PAGE
Assay of arsenic . . . . . . .831
Minerals from which arsenic is produced . . . .831
Assay for arsenic . . . . . . 831
Approximative method, 832. Mr. Parnell's method, 832.
CHAPTER XXII.
Assay of manganese . . . . . .833
Commercially valuable minerals containing manganese . .833
Valuation of manganese ores . . . . .833
Sherer and Rumpf 's method, 834. Mohr's method, 837. Otto's
method, 837. Bunsen's method, 837. J. Pattinson's
process, 837.
CHAPTER XXIII.
Assay of nickel and cobalt ores . . . . .839
Ores of nickel . . . . . . .839
Ores of cobalt . . . . . . .839
Hadow's process for separating nickel and cobalt . . .839
Decomposition of cobalt speiss . . . . .843
Assay of nickel ores ...... 845
Assay of commercial metallic nickel .... 845
Assay of cobalt ores ...... 846
Liebig's method of separating nickel and cobalt . . . 846
Wolcott Gibbs's improvement . . . . .847
Terreil's method . . . . . . .847
Fleitmann's quantitative assay of small proportions of cobalt in
nickel. ....... 849
Plattner's method for detecting nickel before the blowpipe . 849
Assay of complex nickel and cobalt ores .... 850
A. Olassen's method, 850. Ores containing sulphur, arsenic,
nickel, cobalt, and iron, 851. Alloys of zinc, copper, and
nickel, 853.
Ohl's method for the assay of nickel speiss . . .854
Lead and copper speiss, 855. Speiss containing little or no
lead or antimony, 855. Speiss containing much lead or
antimony, 856.
CHAPTER XXIV.
Assay of sulphur ....... 859
Commercially valuable sulphur minerals . . . .859
Assay of sulphur in iron and copper pyrites . . .859
CONTENTS. [xxxi]
PAGE
Assay of sulphur in the dry way . . . . .859
Assay of sulphur in the wet way . . . . .861
C. R. A. Wright's process, 861. Pearson's method, 862.
Hougeau's process, 864. Deutecom's method, 864.
Holland's method, 865. Breckmann's process, 866.
CHAPTER XXV.
Discrimination of gems and precious stones . . . 867
Explanation and introduction, 867. Principal sources of
recognition, 867.
Colourless stones . . . . . . .867
Diamond, 867. The matrix of the diamond, 868. Quartz,
874. White zircons, 875. White sapphire, 876. White
topaz, 876. Comparative table of the weights of colour-
less stones, 878. Use of the tatye, 878.
Yellow stones . . . . . . .879
Yellow zircon, 879. Yellow sapphire, 879. Cymophane
(chrysoberyl), 879. Yellow topaz, 880. Yellow tourmaline,
880. Yellow emerald, 881. Comparative table of weights
of yellow stones, 883. Yellow quartz, 884.
Brown and flame-coloured stones ..... 884
Zircon (hyacinth), 884. Yermeil garnet, noble garnet, al-
mandine, 884. Comparative table of weights of brown or
flame-coloured stones, 885. Essonite, cinnamon stone, 886.
Tourmaline, 886.
Red and rose-coloured stones ..... 886
Red sapphire, oriental ruby, 886. Deep red garnet, noble
garnet, 886. Spinel ruby, 886. Reddish topaz, 887. Red
tourmaline, 887. Comparative table of weights of red and
rose-coloured stones, 887.
Blue stones ....... 888
Blue sapphire, 888. Disthene, cyanite, 888. Blue topaz, 888.
Blue tourmaline, 889. Blue beryl, 889. Dichroite, water
sapphire, 889. Turquoise, 889. Comparative table of the
weights of blue stones, 890.
Violet stones .890
Violet sapphire, 890. Violet tourmaline, 890. Violet
quartz, amethyst, 890. Comparative table of weights of
violet stones, 891.
Green stones . . . . .891
Green sapphire, 891. Peridot, crysolite, 891. Green tourma-
line, 892. Emerald, 892. Aqua-marine, 892. Chryso-
prase, 892. Comparative table of weights of green stones,
896.
[xxxii]
CONTENTS.
Stones possessing a play of colours (chatoyant) .
Sapphire, 894. Garnet, 894. Cymophane, 894. Antique
emerald, 894. Quartz, 894. Felspar, nacreous felspar,
fish-eye, &c., 895. Comparative table of weights of stones
possessing a play of colours (chatoyant), 895.
Glass and artificial gems ......
Inferior brilliancy, 896. Inferior hardness, 896. Fusibility, 89 6.
PAGE
893
896
APPENDIX.
TABLE I. Showing the quantity of fine gold in 1 oz. of any
alloy to ^ of a carat grain of the mint value
of 1 oz. of each alloy . . ii
TABLES A, B, and C. To convert mint value into bank
value when the standard is expressed in
carats, grains, and eighths ... xx
TABLE II. Table of relative proportions of fine gold and
alloy, with the respective mint values of
1 oz. of each alloy when the standard is
expressed in thousandths . . . xxi-xxxii
TABLE. To convert mint value into bank value when the
standard is expressed in thousandths . . xxxiii
TABLE III. Assay table, showing the amount of gold and
silver, in ounces, pennyweights, and grains,
contained in a ton of ore, &c., from the
weight of metal obtained in an assay of 200
grains of mineral .... xxxiv-xlvii
INDEX
xlix
.
A MANUAL
OF
PEACTICAL ASSAYING,
CHAPTEE I.
CHEMICAL NOMENCLATURE LAWS OF COMBINATION, ETC.
IN a treatise intended to be used principally by the prac-
tical assayer, it is neither necessary nor possible to give
more than a brief outline of the elements of chemical
nomenclature and of chemical combination. A knowledge
of practical chemistry is undoubtedly of great value to
the assayer ; indeed, no one can attain to any degree of
eminence in this branch of industry unless he has had
some amount of practice in the laboratory ; but it will be
beyond the scope of this volume to teach the elements of
chemistry. Such instruction in chemistry must be sought
for in books which are specially devoted to the science.
The student should, above all, endeavour to acquire a
practical knowledge of experimental chemistry by going
through a course of instruction in a laboratory.
CHEMICAL NOMENCLATURE. Every material substance with
which we are acquainted consists of one or more bodies,
termed elements, from the fact that with our present means
of research we are unable to reduce them to a more simple
form. Thus, if a piece of common iron pyrites be exposed
B
L> CHEMICAL NOMENCLATURE
to certain chemical operations, it will be found to consist
of two substances, each physically and chemically distinct
from the other and from the original substance. One
body is sulphur, an opaque yellow substance, fusing at a
very low temperature, igniting readily, and burning with
a peculiar suffocating odour. The other constituent is iron,
a well-known metallic substance, requiring an intense heat
for fusion, and not burning at a red heat. If we perform
any experiment which, in the present state of knowledge,
ingenuity could suggest, we are totally unable to cause
either the sulphur or the iron to assume a more simple or
elementary state of existence. We can with ease cause
either of them to enter into new combinations with other
bodies, and these compounds we can decompose as in the
case of the pyrites and obtain both sulphur and iron again
in their separate forms with all their characteristic proper-
ties ; but nothing more than this can be effected : hence
we are led to the belief that both sulphur and iron are
elements, or bodies containing only one kind of matter.
The following table gives the elements discovered up
to the present time. Those substances whose names are
printed in italics have hitherto been found of no practical
use ; and those marked with an asterisk (*) are often
found native, or unassociated with mineralising elements.
Names of the
Elements
Aluminium
Antimony
Arsenic
Barium
Beryllium
* Bismuth
Boron
Bromine
Cadmium
Ctesium
Calcium
* Carbon
Cerium
Chlorine
Elements.
- -
i
Symbols
Atomic
Weights
Names of the
Elements
Al
27
Chromium
Sb
119-6
Cobalt
As
Ba
74-9
136-9
*Copper
Didymium
Be
9
Erbium .
Bi
207-5
Fluorine .
B
11
Gallium .
Br
80
Germanium
Cd
112
*Gold .
Cs
132-7
Hydrogen .
Ca
40
Indium
C
12
Iodine
Ce
141-2
*Iridium
Cl
35-5
Iron .
Symbols
Cr
Co
Cu
Di
E
F
Ga
Ge
Au
H
In
I
Ir
Fe
A.tomic
Weights
52-5
58-6
63-5
146
166
19
69-9
72-3
196-2
1
113-4
126-5
192-5
56
CHEMICAL NOMENCLATURE.
Elements cont.
"ssas" ! s "" b <">
Atomic
Weights
Names of the
Elements
Symbols
Atomic
Weights
Lanthanum La
138-5
Selenium .
Se
78-9
Lead. . Pb
207
Silicon
Si
28
Lithium . Li
7
*Silver
Ag
108
Magnesium j Mg
24
Sodium
Na
23
Manganese : Mn
54-8
Strontium
Sr
87-3
Mercury . Hg
199-8
* Sulphur
S
32
Molybdenum Mo
95-9
Tantalum
Ta
182
Nickel . ! Ni
58-6
Tellurium
Te
127-7
Niobium (Co
Terbium .
Tb
?
lumbium, Cb) Nb
93-7
Thallium .
Tl
203-7
Nitrogen . N
14
Thorium .
Th
232
Osmium . Os
195
Thulium .
Tm
?
Oxygen . O
16
Tin .
Sn
118
Palladium . j Pd
106-2
Titanium .
Ti
50-2
Phosphorus \ P
31
Uranium .
U
239-8
Platinum . | Pt
194-3
Vanadium
V
51-1
Potassium K
39
Wolfram(Tung
Rhodium . i Rh
104-1
sten)
W
183-6
Rubidium Rb
85-2
Ytterbium
Yb
172-6
Ruthenium Ru
103-5
Yttrium .
Y
89-6
Samarium j Sa
150
Zinc .
Zn
65
Scandium '< Sc
44
Zirconium
Zr
90-4
The first column contains the name of the element ;
the second, the symbol, in which all chemical changes and
decompositions are most readily understood ; and the third,
the atomic weight. These atomic weights are not given
beyond the first place of decimals, to avoid tedious calcu-
lation ; for all practical purposes they may be considered
accurate. Of the compounds of these elements, only those
will be discussed which are likely to fall under the notice
of the assayer.
The principal compound bodies with which the assayer
will have to deal are acids, oxides, salts, and binary, sub-
stances containing no oxygen.
When a body combines in more than one proportion
with oxygen, that compound containing the least oxygen
takes the termination ous^ that containing the most ic\
thus,: sulphurous acid, sulphuric acid ; arsenious acid,
arsenic acid ; ferrous oxide, ferric oxide ; mercurous
u -2
CHEMICAL NOMENCLATURE.
oxide, mercuric oxide ; in only one proportion (or
when they form only one basic oxide) they are distin-
guished by the termination ic, as potassic oxide, aluminic
oxide.
OXIDES are binary oxygen compounds; they may be
divided into three series. The first series comprises those
oxides which do not possess the property of combining with
acids to form salts they are termed indifferent oxides ;
the second contains those capable of uniting with acids to
form salts, and called salifiable oxides or bases ; the third
comprises those oxides which have acid characters, and
form salts by uniting with bases.
When an elementary body combining with oxygen
forms but one oxide, it is simply called the oxide of that
element. Thus we say zinc oxide, potassium oxide, alumi-
nium oxide.
If the body is capable of combining with oxygen in
many proportions, the words proto-, sesqui-, bin-, or per-,
&c., precede the term oxide, to express the progressive
amounts of oxygen. Most metals form one salifiable
oxide, and many of them have two ; these are now gener-
ally distinguished by the terminations ous and ic, in the
same manner as are the acids. Thus we have protoxide of
lead, iron, copper, tin, &c. ; sesquioxide of aluminium, iron,
or chromium, &c. ; binoxide or peroxide of manganese,
copper, mercury, &c. ; and when we speak of them as
salifiable bases, ferrous and ferric oxides ; mercurous and
mercuric oxide. Some metals unite with oxygen in still
higher proportions ; these compounds are almost always
acids, such as chromic acid, stannic acid, antimonic
acid, &c.
SALTS are formed when an acid unites with a base, and
usually the properties of the acid and the base are recipro-
cally neutralised ; thus an acid which before combination
possesses the power of reddening blue litmus, loses it on
combining with the base, and, in like manner, a base which
would at first change reddened litmus paper to blue loses
this property as the acid saturates it. In this case the
acid and base have combined to form a salt.
LAWS OP COMBINATION. 5
The names of salts are governed first by the nature of
the acid ; secondly, by the salifiable nature of the base ;
and, thirdly, by the proportions in which the acid and base
are combined. Acids terminating in ic form salts ending
in ate. Acids terminating in ous form salts terminating in
ite ; and the new names having these terminations are
added to the name of the oxide. Thus sulphuric acid and
iron protoxide form sulphate of iron protoxide, ferrous
sulphate, or, more commonly, iron protosulphate ; arseni-
ous acid and iron protoxide form arsenite of iron prot-
oxide, ferrous arsenite, or iron protarsenite ; nitric acid
and iron sesquioxide form nitrate of iron sesquioxide or
ferric nitrate.
When the salt exists in the neutral state its name is
.formed as above, but if the proportion of acid is greater
than in neutral salts, it is termed an acid salt: thus
potassium bisulphate is sometimes called acid sulphate of
potassium. If, on the other -hand, the base is in excess,
the name is preceded by the words sub or basic ; thus,
lead subacetate or basic- acetate of lead.
Binary compounds containing no oxygen exist very
largely in nature, and it is from them that the greater
part of our copper, lead, silver, &c., is obtained.
When a non-metal combines with a metal to. form
a compound which is neither acid nor basic, its name
is derived from the non-metal by the addition of the
termination uret or ide. The latter term is, however,
gradually displacing the former. Thus the compounds
of sulphur with iron and chlorine with silver are res-
pectively called iron sulphuret or sulphide, and silver
chloride.
If a non-metal combines with a metal in more than
one proportion, the same rule is followed as with the
oxygen compounds : thus we have ironjpn>fo-sulphide, iron
sesqui-sulphide, andiron Si-sulphide (ordinary iron pyrites
or mundic).
Laws of Combination. On examining the compounds
which the same substances afford by their union in dif-
ferent proportions, it has been noticed that the propor-
(5 LAWS OF COMBINATION.
tions of the elements existing in each compound are
definite ; a certain weight of one substance will only com-
bine with a certain weight of another substance, and the
lowest combining weight of any of the elementary bodies
is termed its atomic weight, and is represented by the
numbers in the third column of the table of elementary
substances.
As before stated, all substances combine in fixed or
definite proportions ; thus, if 223 parts of oxide of lead
are analysed, they will be found to consist of 207 parts of
lead and 16 of oxygen. Again, the analysis of 18 parts of
water or oxide of hydrogen would give 2 parts of hydro-
gen and 1 6 of oxygen ; now, taking hydrogen as unity, we
have 207 as the equivalent of lead, and 16 as that of
oxygen. If we follow oxygen further in its combinations,
it will be seen that'
16 parts of oxygen combine with 1 part of hydrogen.
207 lead.
40 calcium.
, 118 tin.
63-5 copper.
The above numbers, therefore, represent the equivalents
of the respective elements.
Again, the equivalent of sulphur is 32, and this
represents the weight of sulphur which will combine with
the above weights of hydrogen, lead, calcium, tin, or
copper to form sulphides of the respective bases. 35*5
parts of chlorine, or 78*9 parts of selenium, also combine
with the same weights, viz. hydrogen 1, lead 207, &c.,
to form chlorides and selenides.
Compounds like these are of the simplest class, and
consist of single equivalents only ; there are, however,
many compounds containing more than two equivalents,
in which case the following laws are followed.
In one class of compounds the quantity of one of the
constituent elements remains constant, while each new
compound is formed by the successive addition of another
CHEMICAL SYMBOLS. 7
atom of the other constituent element ; and it must also
be borne in mind that no element will combine with
another in less than its atomic weight. Another series
will commence with two atoms of an element united with
an uneven number of atoms of another element ; thus
we can have binary compounds in the proportion of 2 to
3, 2 to 5, or 2 to 7.
The atomic weight of a compound body is the sum
of the atomic weights of the elements forming it : thus
sulphuric anhydride is composed of one atom or 32 parts
of sulphur, and 3 atoms, or 48 parts, of oxygen ; its atomic
weight is therefore 80. The atomic weight of any com-
pound body may be ascertained by adding together the
atomic weights of its constituent elements.
Owing to the invariable law of the constancy of chemi-
cal compounds, we are enabled to calculate the reaction
which occurs between two or more bodies when decompo-
sition takes place: thus 174 parts of potassium sulphate
contain 80 parts of sulphuric anhydride and 94 parts
of potassium oxide ; and if it were desired to obtain lead
sulphate by the decomposition of lead nitrate by adding
to it the above quantity of potassium sulphate, the exact
amount of lead nitrate required would be readily found by
adding together the equivalent of the elements forming
nitric acid and lead oxide.
CHEMICAL SYMBOLS : THEIR EMPLOYMENT AND USES The
symbol of an element standing alone signifies one atom
of that element. Thus : S implies not only the element
sulphur, but 32 parts of sulphur ; a small figure on the
right-hand side of the symbol indicates the number of
atoms to be represented ; thus, S 2 is equal to two atoms,
or 64 parts of sulphur.
Two symbols placed thus, FeS, indicate a compound
of equal equivalents of iron and sulphur. Separation
of elements by the sign + or a comma is employed to
show the union of two compound bodies ; thus the com-
pound of silver sulphide and lead sulphide may be thus
written : AgS + PbS, or AgS,PbS. A large figure on the
same line as the symbol, and on its left side, multiplies the
CHEMICAL SYMBOLS.
whole of the symbols to the first comma or + sign : thus,
2AgS,PbS, or 2AgS + PbS, represents a compound of
two atoms of silver sulphide with one of lead sulphide.
If, however, it be thus written, 2(AgS, PbS), it means two
atoms of the whole of the elements which are inclosed in
the brackets.
PREPARATION OP THE SAMPLE.
CHAPTER II.
PREPARATION OF THE SAMPLE WEIGHING.
THE selection and preparation of the sample is the first
and most important 'operation in assaying. It is of little
use for the operator to ascertain with accuracy the per-
centage of every individual constituent in the mineral
operated on, if the sample does not truly represent the
average of the ore. It should be borne in mind that
samples of mineral are generally selected for their richness,
and represent the most favourable portions of the ore ;
and no pains should be spared to secure a sample for
analysis which will truly show the bulk of mineral whose
value is required to be known.
The assayer must always bear in mind the object which
his experiments have in view. If they are to ascertain
the actual percentage of one or more constituents existing
in a certain stone, his labours are comparatively easy, all
that is required being to reduce the whole of the specimen
to the finest possible state of division, and, having well
mixed the powder, to analyse a portion of it.
But if it is desired to find out the composition of a
special mineral or crystal, the greatest possible care must
be taken to remove the whole of the gangue or other im-
purities, and to obtain for analysis those portions only
which represent with greatest accuracy the pure mineral.
To effect this the surrounding rock is first removed as
carefully as possible, and then the specimen is crushed
into coarse pieces on a sheet of clean paper. By means
of a pocket magnifier and a pair of pincers, clean, typical
pieces of the mineral are then to be selected for analysis.
10 PKEPARATIOX OF THE SAMPLE.
If, however, as will most frequently be the case, the
object of the assay be to ascertain the average value of a
mineral lode or heap of ore, then the assayer must proceed
differently. The portion experimented upon must truly
represent, in the respective amounts of its valuable ma-
terial, gangue, quartz, and earthy matters, the whole bulk
of that of which it professes to be a sample ; and this
having been secured, the whole must be carefully pow-
dered and passed through fine sieves, taking care that
every portion of the mineral goes through. If this be not
attended to, it will frequently happen that the few grains
left out are sufficient to vitiate the whole assay ; this is
especially apt to be the case when examining ores the
valuable ingredients of which are of a ductile or malle-
able nature, such as auriferous quartz. In this case it fre-
quently happens that the great, bulk of gold exists in the
form of one or two small pieces, and these being flattened
and beaten out in the operation of powdering will almost
certainly be left upon the sieve. In cases like this it is
better to collect and assay such pieces separately, and esti-
mate their proportion to the whole weight of the sample,
than to attempt to powder and distribute them uniformly!
SAMPLING. The important operation of sampling neces-
sarily precedes any process of assay. Dr. Peters, in his
' Modern American Methods of Copper Smelting,' gives the
following useful description of the operation of sampling
as practised in the great mining centres of the New World.
By sampling we seek to obtain within the compass of a few
ounces a correct representative of the entire quantity of
ore, which may vary in amount from a few pounds to
several thousand tons. As a rule it will lessen the chance
of serious error, in very large transactions, to divide the lot
into parcels of not over fifty tons each, and sample each
of these lots by itself. The utmost care and vigilance
in sampling and assaying should be required at every
smelting works, both in the interest of the works and in
that of the ore-seller.
Until quite recently, it has been customary to sample
lots of ore by quartering them down, rejecting a certain
PREPARATION OF THE SAMPLE. 11
proportional part at each successive operation, and re-
ducing the size of the ore fragments as the quantity to
operate on diminishes. This is a laborious and expensive
method, and in the case of finely pulverised ores may
well be replaced by the use of the ' split shovel,' or one of
the many automatic sampling machines that have been
invented.
But since the establishment of public sampling works
at most of our great mining centres, where the correctness
of the sample is guaranteed by the works, which dis-
tribute packages of each lot of ore to the agents of the
various rival smelting companies for them to assay and
bid upon, the vast quantities of ores handled, and the im-
portance in many instances of retaining the lump form of
the ore as essential to the subsequent metallurgical opera-
tions, have imperatively demanded some method of auto-
matic sampling that shall be rapid, accurate, and equally
applicable to ores in both the pulverised and lump form.
The means hitherto employed all depend upon the
same general principle of cutting or dividing a falling
stream of ore by means of flanges, fingers, ore travelling
buckets, in such a manner as to obtain a certain desired
proportion of it for a sample.
While many of these devices work admirably upon
pulverised ore, free from dampness or foreign obstructing
substances, they are apt to give entirely unsatisfactory
results upon a mixture of fine and coarse ores, while the
presence of strings, chips, rags, &c., usually clogs them
and deranges their working.
Mr. D. W. Brunton, of Denver, Colorado, has invented
an automatic sampling machine that is apparently free
from all the defects enumerated, and which has been
shown by practical trial to be equally applicable to coarse,
fine, or mixed ores, while it cannot be clogged by foreign
bodies of any reasonable size.*
Brunton overcomes these difficulties by deflecting the
entire ore stream to the right or left, while falling through
* See Transactions of the American Institute of Mining Engineers, vol.
xiii. p. 639, for drawings and full description of this sampler.
12 PREPARATION OF THE SAMPLE.
a vertical or inclined spout. By a simple arrangement of
movable pegs, in connection with the driving gear, the
proportion of the ore stream thus deflected into the
sample bin may vary from 10 to 50 per cent. ; the latter
amount only being required in coarse ores of enormous
and very variable richness, while for ordinary lump ores
from 10 to 20 per cent, is the maximum required.
Instead of passing the sample-stream of ore into a bin,
this system may be still further perfected by leading it
directly to a pair of moderately fine rolls, the product of
which is elevated to a second similar sampling-machine
from which the final sample drops into a locked bin.
Six months' constant experience with this sampler has
shown that 10 per cent, of 20 per cent., or 2 per cent, of
the original ore parcel is usually quite sufficient ; though
in exceptional cases, 15 per cent, of 30 per cent., or \\
per cent, of the ore may be required.
The two machines are driven at different speeds, to
prevent any possible error that might arise from isochronal
motion, and by careful tests of this machine in resampling
lots of ore the limit of error has been found less than
one-fourth of 1 per cent. ; while even the best hand-
sampling may vary 2 per cent.
The fact that the division is one of time and not of ore
is one of the most important features of this valuable in-
vention, as it consequently is forced to deflect the exact
proportion of the ore-stream, for which it is set, whether
coarse or fine, wet or dry, Hght or heavy.
MOISTURE. The determination of the moisture present
in any given parcel of ore is also a matter of much im-
portance ; and probably more inaccuracies attend this
apparently simple process than any other of the pre-
liminary operations.
This determination must, of course, take place as nearly
as possible at the same time that the entire ore parcel is
weighed, as otherwise the sample may lose or gain
moisture.
In lump ores, it is difficult to obtain a correct sample
even for moisture, without some preliminary crushing,
PREPARATION OF THE SAMPLE. 13
and to save labour, it is best to use a portion of the regu-
lar assay sample for this purpose ; the accurate weighing
of the entire ore parcel being postponed until just before
or after the sampling, and the portion reserved for the
moisture determination being placed in an open tin vessel,
contained in a covered metal case, having an inch or two
of water on its bottom, in which sample tins stand.
From one-fourth to one-half pound of the sample is
usually weighed out for this determination, and dried
under frequent stirring, and at a temperature not exceeding
212. While it is always important to keep within the
limit of temperature just mentioned, it is especially the
case with certain substances which oxidise easily. Among
these are finely divided sulphides, and, above all, the
pulverulent copper cements obtained from precipitating
copper with metallic iron from a sulphate solution.
Such a sample, containing actually 5J per cent, of
moisture, showed an increase of weight of some 2 per
cent., on being exposed for thirty minutes to a tempera-
ture of about 235 Fahr.
Certain samples of ore especially from the roasting
furnace are quite hygroscopic, and attract water rapidly
after drying.
In such cases the precautions used in analytical work
must be employed, and the covered sample weighed
rapidly in an atmosphere kept dry by the use of strong-
sulphuric acid.
The sampling of the malleable products of smelting,
such as blister copper, metallic bottoms, ingots, &c., can
only be satisfactorily effected by boring a hole deeply
into a certain proportional number of the pieces to be
sampled.
Where such work is only exceptional, an ordinary
ratchet hand-drill will answer, but in most cases a half-
inch drill run by machinery is employed.
The chips and drillings are still further subdivided by
scissors, and, as even then it is difficult to obtain an abso-
lutely perfect mixture, it is best to weigh out and dissolve
a much larger amount than is usually taken for assay,
14 PREPARATION OF THE SAMPLE.
taking a certain proportion of the thoroughly mixed solu-
tion for the final determination.
The ore must always be reduced to a powder, more or
less fine, according to the nature of the chemical operation
or assay to which it is to be subjected. This division is
effected by means of the anvil, hammer, pestle and mortar,
sieve, method of elutriation, or other means generally in
use for the preparation of any fine powder. The actual
process to be adopted must vary according to the nature
of the different bodies under examination. In some cases
simple crushing is sufficient ; in others the ore will have
to be pounded in a mortar ; whilst occasionally it is neces-
sary to reduce it to the very highest degree of fineness by
elutriation.
There are other operations us strictly mechanical as
are the above, viz. washing, dressing, and vanning a
sample of ore, the end and aim of which is to separate,
in a suitable vessel, by means of water and difference of
specific gravity, the earthy or useless and, in some cases,
objectionable portion from the heavier metallic and valu-
able portion. This operation is almost always employed
on the larger scale in dressing ores for the smelter.
The tools and materials employed in preparing the
sample in the assay laboratory are the anvil (and stand),
vice, hammer, files, cold chisel, shears, pestle and mortar,
*t eel-crushing mortar, sieve, &c.
THE ANVIL (fig. 1). The anvil is most useful in size
when it weighs about 28 Ibs. ; but one of 14 Ibs. will
suffice. The anvil recommended is of the shape usually
employed by the blacksmith.
The anvil-stand is constructed of stout wood, about
two inches in thickness, and forms a cube of about two
feet square. It contains three or four drawers, which
serve to hold the hammers, cold chisel, shears, files, &c.,
which are required in an assay office. In the centre the
anvil is fixed, and in one corner: a vice may be also
secured.
In general the anvil and hammer are employed for the
purpose of breaking a small fragment, from a mass of ore
PREPARATION OF THE SAMPLE.
15
for examination, or ascertaining whether the button or
prill of metal produced in an assay be malleable or other-
FlG. 1.
wise. The anvil is also exceedingly useful as a support
for a crucible while breaking it to extract the metallic or
other valuable contents.
THE HAMMERS (figs. 2 and 3), of which two are requisite,
ought to have one end flat and FIG. 2. FIG. 3.
square and the other pick- or
wedge-shaped. The horizontal
wedge end of fig. 2 is useful for
breaking open crucibles and in
detaching small fragments from
a specimen of ore. The flat end
serves for ascertaining the mal-
leability of buttons of metal.
This hammer should weigh about
1 Ib. The larger hammer, fig. 3,
should weigh about 4 Ibs., and is
employed for breaking coke suffi-
ciently fine for the use of the fur-
nace, and detaching fragments
from refractory minerals, in both of which cases either
16 PREPARATION OF THE SAMPLE.
end may be employed, as may seem most serviceable to
the operator. The flat end of this hammer is also used
for driving a cold chisel in separating masses of gold,
silver, copper, lead, &c., for assay. This hammer has a
vertical pick or wedge end.
Very hard and stony materials which have to be broken
on the anvil (and all such ought to be so treated) scatter
many fragments, to the certain loss of a proportion of the
substance, and the probable injury of the operator ; this
can be prevented by wrapping the mineral in a piece of
stout brown paper, or if necessary in several folds. The
fracture can then be safely attempted.
This latter precaution must be specially taken in frac-
turing gold quartz, or hard rock containing metallic silver,
FIG. 4. FIG. 5.
as the loss of a very minute quantity of metal would
involve a considerable error in the result afforded by the
assay.
All minerals, unless very friable, must be reduced to a
moderate size say that of a walnut by means of the
anvil and hammer, before pulverisation ; otherwise, if the
reduction be attempted in a mortar, it is nearly certain to
be injured ; moreover, the operator will find his labours
much abridged by using the anvil for this purpose.
The anvil can also be made very serviceable in repoint-
ing worn or burnt-out tongs &c. It need scarcely be
added that it must be placed as far as possible away from
bottles or other frangible articles, otherwise accidents may
occur by the forcible projection of fragments of crucibles,
stones, &c.
THE COLD CHISEL (fig. 4) is employed for cutting off me-
tallic masses for assay. It should be five or six inches long
and about half an inch wide, which is the best size for
general use. However, for some purposes, as cutting copper
PREPARATION OF THE SAMPLE. 17
and other very tough metals, it is convenient to have a
chisel only a quarter of an inch wide, as these metals are
so much more difficult to cut, and the small chisel meets
with the least resistance.
Small shears (fig. 5) are also exceedingly useful in
cutting off pieces of sheet metal, such as lead, for cupel-
lation, scorification, &c.
THE PESTLE AND MOKTAR. Mortars are made of various
materials, as cast-iron, bronze, porcelain, agate, &c. ; the
assay er requires one of cast-iron, one of porcelain, and
one of agate.
The iron mortar (fig. 6) ought to be of the capacity of
from three to four pints ; the porcelain (Wedgwood ware)
(fig. 7) may contain about two pints. The ease with which
FIG. 6. FIG. 7.
a mortar may be used depends much upon its form, and
opinion is greatly divided on the subject. Faraday* says
that the pestle should be strong, and the size of its upper
part sufficient to allow of its being grasped firmly in the
hand, and below to permit a considerable grinding surface
to come in contact with the mortar. Its diameter in the
lower part may be about one third or one fourth of the
upper diameter of the mortar. The curve at the bottom
should be of shorter radius than the curve of the mortar,
that it may not touch the mortar in more than one part,
whilst at the same time the interval around may gradually
increase, though not too rapidly, towards the upper part
of the pestle.
The bottoms of all mortars ought to be of considerable
* Chemical Manipulation, p. 149.
18 PREPARATION OP THE SAMPLE.
thickness, in order to withstand the smart blows they will
occasionally have to receive.
Berzelius recommended a mass of pumice-stone for
cleansing porcelain mortars. It is used with water as a
pestle, and in course of time will be worn to the shape of
the mortar ; its action will then be more speedy.
Iron mortars can be best cleaned by friction with a
little fine sharp sand, if washing be not sufficient to re-
move the adhering substance. Great care must be taken
to dry mortars perfectly, especially those of iron, other-
wise they will become rusted, and the rust will contaminate
the substances pulverised in them.
The iron mortar is principally of use in the reduction
of the masses of mineral (broken on the anvil, as before
described) to a state of coarse powder, in order to render
the substance more readily capable of pulverisation, strictly
so called. In the use of the iron mortar, all friction with
the pestle ought to be avoided, and the body within it
must be struck repeatedly and lightly, in a vertical direc-
tion, taking care to strike the large pieces, so that all may
be equally reduced. This can be carried on until the
whole is about the size of fine sand. It is transferred to
the porcelain mortar, where direct blows must be carefully
avoided.
The process is now carried on somewhat differently ;
the pestle is to be pressed with a moderate force, and a
circular motion given to it, taking care every now and
then to lessen and then to enlarge the circles so as to pass
over the whole grinding surface of the mortar, and insure
the pulverisation of the mass of mineral submitted to
operation. In general, the finer the state of division to
which a mineral is reduced, the more accurate and ex-
peditious will be its assay ; and in preparing a mineral for
assay by the wet method, no labour ought to be spared
on this point. Pulverisation is rendered much easier by
operating on a small quantity at once, and removing it
very often from the sides and bottom of the mortar by
means of a spatula. The quantity operated on at one
time must be regulated by the hardness and friability of
PREPARATION OF THE SAMPLE. 19
the substance whose pulverisation is to be effected. The
harder it is, the less must be taken, and vice versa.
In the use of the iron mortar fragments are occasionally
projected. This may be prevented by covering the upper
part of the mortar with a cloth. This applies also to the
porcelain mortar, for the dust of some minerals has a
disagreeable taste and smell. Indeed, in some cases the
ambient powder is highly deleterious, as in the pulverisa-
tion of arsenical nickel, cobalt, and other ores. Here the
simple cloth is not a sufficient protection ; it should be
slightly damped with water, and tightly tied round the
mortar, and firmly held round the pestle, when nothing
can escape.
Some minerals can be pulverised with greater ease if
they are ignited and suddenly quenched in cold water.
Amongst them may be named flint, and many other sili-
ceous matters, as gold quartz. In the pulverisation of
charcoal for assays, it will be found useful to heat it, as
hot charcoal is more readily pulverised than cold.
In some instances the powder obtained in the iron or
porcelain mortar is not fine enough ; recourse should then
be had to the agate mortar, in which the mineral, in as
fine a state of division as the larger mortars will give it, is
ground in small portions at
a time, until it is reduced to
an impalpable powder.
When small specimens or
rare minerals are being ope-
rated upon, if it is especially
desirable to avoid loss, it is
advisable to use a steel mor-
tar (fig. 8) for the prepara-
tory reduction of the mineral
to coarse powder. A B, C D,
and E F represent the three
component parts of the mor-
tar ; these may be readily
taken asunder. The substance
to be crushed (having, if practicable, first been broken into
c 2
L>0 PREPARATION OF THE SAMPLE.
small pieces) is placed in the cylindrical chamber E F;
the steel cylinder, which fits somewhat loosely into the
chamber, serves as a pestle. The mortar is placed upon
a solid support, and perpendicular blows are repeatedly
struck upon the pestle with a hammer, until the object in
view is attained. (Fresenius.)
In the selection of agate mortars, they must be examined
to see that they have no palpable flaws in them ; very
slight cracks, however, that cannot be felt, do not render
the mortar useless, although they increase the danger of
its destruction by a chance blow.
THE SIEVE. The operation of sifting is employed when
a very fine powder is required, or when a powder of
uniform size is needed. Sieves of various materials and
different degrees of fineness are necessary. The larger
sieve, for preparing coke for the blast furnace, is made of
stout iron wire, and must have its meshes from 1 inch to
1J inch square. The fine coke, which is sifted from that
which is the proper size for the blast furnace, may be
mixed with that of ordinary size, and employed economi-
cally in the muffle furnace. For the preparation of
minerals a set of three sieves should be provided, each
one finer than the other. The coarsest may contain 40
holes to the linear inch, the finer or medium sieve 60, and
the finest from 80 to 100. The coarsest sieve is used for
preparing galena for assay ; the medium for copper, tin,
iron, and other like ores ; and the finest for gold and
silver ores, or for preparing any substance for the wet
assay, as, in the latter case, the finer the state of division
the substance attains, the more rapid will be its solution
or decomposition by the liquid agents employed.
The sieve fig. 9 is made of wood, over which is strained
in the ordinary manner brass wire-gauze of the necessary
degree of fineness. When in use, j5, fig. 10, is fitted into
the lower part of A (same figure). This contrivance pre-
vents all loss of the fine powder. If the matter to be
sifted be offensive or deleterious to the operator, a sieve
termed the drum or box-sieve may be employed (see fig.
10), where C represents a cover fitting over the sieve. If
SIFTING. 21
small, this may be used in the ordinary way ; but if large,
its method of use is rather peculiar, and requires some
practice to fully develop its powers. One side of the
under edge must be FIG. 9. FIG. 10.
held by one or both
hands according to its
size, whilst the other
rests on a table or a
bench. A semicircu-
lar oscillating motion
must now be com-
municated to it by
moving the hands
up and down at the
same time that they
are being alternately brought into approximation with the
sides of the operator.
In cases of necessity, a sieve may be readily extem-
porised. Place the powder to be sifted in a piece of fine
lawn or muslin, according to the fineness required, tie it
up loosely, and shake or tap the powder, with, its muslin
or other envelope, on a sheet of paper, and the sifting will
be rapidly and easily accomplished.
The sieve is also extremely serviceable in the separation
of some ores from their gangues or vein-stones, especially
if the latter be stony and hard. This point must be par-
ticularly noted, as it is the cause of much variance between
the results of different assayers ; for instance, part of the
same sample of ore might be sent to two assayers, and the
produce made by one would be 8^ per cent., and that by
the other 9 or 9^, or, in some cases, even more. This dis-
crepancy generally arises from the cause above mentioned.
In the one case the workman has rejected part of the
hard gangue, and so rendered the residue richer ; whilst
in the other he has pulverised the whole, making the
produce less, but giving more accurately the amount of
.metal in the substance submitted to assay.
A knowledge of this fact is also very useful from
another point of view. Suppose it were wished to separate
'2*2 SIFTING.
in a speedy manner, as perfectly as possible, any friable
mineral, such as galena or copper pyrites, from its matrix
by mechanical means, it might be accomplished by the
use of the sieve, as follows : Place a small quantity of
the mineral in an iron mortar, and strike, repeatedly,
slight vertical blows. When it is tolerably reduced, sift
it, and it will be found that what passes through is nearly
pure mineral, with only a small quantity of matrix ; repeat
the pounding and sifting operations, until, after a few
repetitions, that which remains in the sieve is nearly pure
gangue.
Native metals, as gold, silver, and copper, are also*
partially separated after the manner above described.
The fine particles of metal, during the process of pounding
and trituration, become flattened, and cannot pass through
the sieve, whilst the more brittle portions pass through
and are separated.
ELUTRIATION. This process can only be employed for
those bodies which are not acted on by water ; and it must
be remembered that many substances which are usually
considered to be insoluble in water are, when in a very
finely divided state, acted upon to a greater or less extent.
The operation is thus effected : The substance is reduced
to the finest possible state of division by any of the fore-
going processes ; it is then shaken up with a quantity of
water in a glass or other vessel, After a few moments'
repose, the supernatant liquid, retaining in suspension the
finer particles of the pulverised substance, is poured off,
and the grosser parts, which have fallen to the bottom of
the vessel, are repulverised, and again treated with water.
By repeating these processes a powder of any required
degree of fineness may be obtained.
It is seldom, however, that a substance is required for
assay by the dry way, in such a minute state of division.
In the humid or wet method it is occasionally very useful.
If the supernatant water is roughly decanted off, where
the powder to be elutriated is light, the least disturbance
of the vessel containing it occasions the distribution of the
portion which has settled, throughout the liquid, and the
ELUTRIATION. 23
consequent mixture of fine and coarse particles. This
can be avoided by the employment of the syphon. The
operation is then thus conducted : The syphon is filled
with water, and the shorter end placed in the liquid whose
transversion is to be effected : the forefinger of the right
hand, which, during this time, has been applied to the
longer end of the instrument, is now removed, when the
water will flow out until it is level with the immersed
end of the syphon. Fresh water can then be added, the
powder stirred up again, and the operation of decantation
by the syphon carried on as long as requisite.
WASHING, DRESSING, OR VANNING. This operation is ex-
ceedingly useful for discovering the approximate quantity
of pure ore, such as galena, copper pyrites, oxide of tin,
native gold or silver, in any sample of earthy matter or
ore in which it may be disseminated.
The theory of the operation about to be described is
easily understood. Bodies left to the action of gravity in
a liquid, in a state of rest, experience a resistance to their
descent which is proportionate to their surface, whatever
may be their volume and density. Hence it follows, firstly,
that of equal volumes the heaviest fall most rapidly;
secondly, that of equal densities those having the largest
size move with the greatest speed ; for in particles of un-
equal size and like form the weight is proportional to the
cube of the dimensions, whilst the surface is only propor-
tional to the square of these dimensions ; hence in small
particles the surface is greater in relation to the weight
than in the large particles. Thirdly, of equal densities and
volumes, particles offering the largest surface (those which
are scaly and laminated, for example) undergo more re-
sistance in their motion than those which, approaching
the spherical form, have less surface. The adhesion of
the liquid to the particles of bodies held in suspension is
also an obstacle to their subsidence. This force, like the
dynamic resistance, is proportional to the surface and in-
dependent of the mass or volume ; whence it follows that,
in a fluid in motion, of bodies having equal volumes, the
least dense acquire the greatest rapidity of movement, and
24 DRESSING OR VANNING.
are deposited at the greatest distance from the point of
departure ; whilst with equal densities the smallest grains
are carried farthest ; and lastly, with equal densities and
volumes, the particles exposing most surface traverse the
greatest space.
It is, therefore, evident that the most advantageous
condition for separating, by washing, two substances of
unequal specific gravity or density is that the heavier
shall be in larger grains than the lighter ; this unfortu-
nately, however, is a condition that can very seldom be
fulfilled, as the heaviest substances are those metallic
minerals whose frangibility is nearly always greater than
the earthy matters accompanying them as gangues. This
being the case, it is very important so to arrange that the
fragments of the various mixed substances shall be nearly
of the same size. This may be effected by very frequently
sifting the mineral during the process of pulverisation,
reducing it also more by blows than by grinding, so as
to get as little fine powder as possible, as that is nearly
certain to be washed away during the process.
The operation of washing or vanning may be performed
by one of two methods. In the first, a small stream of
running water is employed ; in the second, water is added to
the substance to be washed, and poured off as necessary.
In the first process, a vessel somewhat resembling a
banker's gold scoop (but longer in proportion) is employed ;
the mineral to be washed is placed in the upper part, and a
small quantity of water added, with which the mineral is
thoroughly and carefully moistened, and mixed with the
fingers. The scoop must then be so inclined that a fine
stream of water from any convenient source (say a tap) may
fall just above the upper part of the mixture of mineral
and water ; then, firmly holding the larger and upper end
of the scoop with the left hand, and sustaining the lower
part with the right, it is shaken frequently in the direction
of its longitudinal axis. At each shake all the particles
in the scoop are so agitated that they become suspended
in the water, and the current of liquid running from the
tap into the scoop moves them all in its own direction ;
DRESSING OR VANNING. 25
but they are deposited at different distances from the
point at which the water enters, the heaviest being carried
through but a very small space. It is now soon seen that
the mineral assumes a heterogeneous surface ; at the upper
part, the heavy portions are seen nearly pure ; the light
substances, on the other hand, are nearly without mixture
at the lower end, and in the intermediate part the heaviest
portion of the mixture is nearest the upper end. If the
washed matter were now to be divided into horizontal
layers, the heaviest matter would be found at the bottom,
and the lightest on the surface. Things being in this
state, the scoop must be made to oscillate on its axis, so
that the latter remain immovable, and in a slightly in-
clined position. In this manner, the layer of water
running over the surface of the mineral agitates that part
only, and carries off all light substances there deposited in
the previous operation. When necessary, these matters
may be removed by the finger, and made to run into a
vessel placed below the scoop, in which all the water and
matters carried off are received. This operation, however,
must not be hurriedly performed, so as to mix the parts
already separated : each layer must be removed separately,
commencing with the upper one. This being done, the
scoop must be alternately kept in motion by shakings, as
at first, and then on its axis, and the washing off of the
finer particles renewed, and so on until the separation is
effected as far as may be judged necessary.
At the commencement of the operation, the water
carries out of the scoop the lightest particles, as organic
matter, clay, &c, ; at a little later period the water carries
with it a small but definite quantity of the heavier
portion, the proportion of which increases as the operation
proceeds, until at last the greatest possible care is required.
It is always better to rewash the latter portion which
passes off from the scoop ; hence the necessity of allowing
all the wash-water passing from it to collect in a vessel
placed for that purpose.
In the second method of washing, a tin, zinc, or wooden
pan is employed. It should be circular, one or two feet in
20 DRESSING OR VANNING.
diameter and three or four inches deep ; the sides should
descend in a conical manner, so that the bottom is not more
than four inches in diameter, and the angle between it and
the sides as sharp as possible.
The substance to be examined is placed in the washing-
dish, the latter filled with water, and the mineral well mixed
with it until perfectly moistened as before. After a moment
or so the muddy water is poured off, and the operation
repeated until the water passes off clear. When this
happens, only so much water must be placed in the pan as
will leave a slight layer on the mineral. Now, by holding
the pan in one hand, and shaking it with the other, the
greater part of the heavy mineral, gold or otherwise, will
fall below the sand. If now the pan be inclined towards
the hand which is shaking it, the lighter portions, even if
tolerably large, will flow off with the water, leaving the
heavier matters in the angle, from which, with ordinary
care and a little practice, it is difficult to disturb them*
If there be a large quantity of earthy matter, this may be
(after sufficient shaking) removed by the finger, as in the
first-described process. By careful repetitions of these
processes, the whole, or nearly the whole, of the sandy
and earthy matters may be removed, and the gold or
other mineral left nearly pure. This is the plan employed
in prospecting for gold, diamonds, and other gems, and in
some cases for their commercial extraction.
In Cornwall and other mining counties this operation
is very cleverly and carefully performed on the miner's
common shovel, and the richness of any particular sample
of either tin, lead, or copper is thereby determined with a
very near approach to accuracy.
THE BALANCE.
OPERATION OF WEIGHING. At least three balances will
be required in a laboratory where general assays are per-
formed. The first must be capable of carrying three or
four pounds in each pan, and must turn with a quarter of
a grain. This may be of the form of the bankers' or
bullion balance (fig. 11), and may be employed in weigh-
WEIGHING.
ing samples of gold quartz or silver ore containing metallic
grains capable of being separated by the sieve (see p. 21) ;
the second (fig. 12), or rough assay balance, is similar to
FIG. 11.
28 WEIGHING.
the apothecary's scales ; it should take 1,000 grains in
each pan, and turn with one tenth of a grain. This serves
for weighing samples of ore and fluxes for assay, and for
determining the weight of buttons or prills of lead, tin,
iron, copper, &c., obtained in an assay.
The third and most delicate^ or true, assay balance (fig.
13) should carry about 1,000 grains; must turn distinctly
and accurately with the -roW tn of a grain. This is
employed in the assay of gold and silver, bullion, and in
the assay of minerals containing gold and silver ; also for
general analytical purposes. The first two balances may
be used, with ordinary care, by any one ; but the third
balance, in its use and adjustment so as to maintain its
extreme accuracy, requires some particular instructions,
which necessarily involve the principle of the balance.
These have been so admirably given by Faraday, in his
* Chemical Manipulations,' that we can do no better than
transcribe them :
4 The theory of this balance is so simple that the tests
of its accuracy will be easily understood and as easily
practised. It may be considered as a uniform inflexible
THE BALANCE. 29
lever, supported horizontally at the centre of gravity, and
supporting weights at equal distances from the centre by
points in the same horizontal line with the centre of
gravity. If the weights be equal the one will counter-
poise the other ; if not, the heavier will preponderate.
In the balance, as usually constructed, there are certain
departures from the theory as above expressed some
from the impossibility of execution, and others in conse-
quence of their practical utility ; and a good balance may
be said to consist essentially of a beam made as light as is
consistent with that inflexibility which it ought to possess,
divided into two arms of equal weight and length by a
line of support or axis, and also terminated at the end of
each arm by a line of support or axis, intended to sustain
the pans. These three lines of support should be exactly
parallel to each other in the same horizontal plane, and
correctly perpendicular to the length of the beam ; and
the plane in which they lie should be raised more or less
above the centre of gravity of the beam, so that the latter
should be exactly under the middle line of suspension.
It will be unnecessary in this place to speak of the coarse
faults which occur in the ordinary scales these will be
easily understood ; and from what has to be stated of the
examination of the most delicate instrument, the impos-
sibility of avoiding them without incurring an expense
inconsistent with their ordinary use will be as readily
comprehended.'
Two principal things have to be attended to in the
selection of a balance its accuracy and its delicacy. The
accuracy depends upon the following conditions :
1. The arms should be equal to each other in length. The
length of each is accurately the distance from the middle
to the distant knife-edge, all the edges being considered
parallel to each other, and in the same plane. The two
arms should accord perfectly in this respect. This equality
may be ascertained in two or three ways. Suppose the
balance with its pans to vibrate freely, and rest in a
horizontal position, and that after changing the pans from
one end to the other the balance again takes its horizontal
30 WEIGHING.
state of rest in such a case an almost certain proof is
obtained of equality in length of the arms. They may,
however, be equal, and yet this change of the pans from
end to end may occasion a disturbance of equilibrium,
because of the unequal distribution of weight on the beam
and pans ; but to insure an accurate test, restore the pans,
and consequently the equilibrium, to the first state : put
equal, or at least counterpoising, weights into the pans,
loading the balance moderately, and then change the
weights from one pan to the other, and again observe
whether the equilibrium is maintained ; if so, the length
of the arms is equal.
Equality of weight is not so necessary a condition,
although this should be obtained as accurately as possible.
One arm with its pan may be considerably heavier than
the other, but from the disposition of the weight in the
lighter arm towards the extremity, or in the heavier
towards the middle of the beam, the equilibrium may be
perfect, and therefore no inaccuracy be caused thereby in
the use of the balance. Instruments are usually sent home
In equilibrium, and require no further examination as to
this particular point than to ascertain that they really are
in adjustment, and that after vibrating freely they take a
horizontal position.
2. The beam must be of such a form and strength that
it will not bend when loaded with the greatest weight the
balance is intended to carry. All well-made modern
balances are sufficiently rigid in this respect, and may
be safely trusted to carry their full weight without
flexure of the beam. The beam should also be as light as
practicable.
3. The knife-edges supporting the pans, and the centre
one on which the beam vibrates, must be accurately in the
same line.
The delicacy of a balance likewise depends upon
several conditions.
The centre of gravity must be very little below the
fulcrum. If it be considerably depressed, then, upon
trying the oscillations of the balance by giving it a little
THE BALANCE. 31
motion, they will be found to be quick, and the beam
will soon take its ultimate state of rest ; and if weights be
added to one side, so as to make it vibrate, or to bring it
to a certain permanent state of inclination, the quantity
required will be found to be comparatively considerable.
As the centre of gravity is raised the oscillations are
slower, but producible by a much smaller impulse ; the
beam is a longer time before it attains a state of rest, and
it turns with a smaller quantity.
If, however, the centre of gravity coincides with the
fulcrum or centre of oscillation, then the balance is said
to se^ that is, the smallest possible weight will turn the
beam ; the oscillations no longer exist, but one side or
the other preponderates with the slightest force, and the
valuable indication which is furnished by the extent and
velocity of the vibrations is lost.
The case in which the centre of gravity is above the
fulcrum rarely if ever occurs. Such a balance, when
equally weighted, would set on the one side or the other ;
that side which was in the slightest degree lower tending
to descend still further, until obstructed by interposing
obstacles.
In balances intended to carry large quantities (as in
the balance for weighing gold quartz, &c.) it is necessary
to place the centre of gravity lower than in those for
minute quantities, that they may vibrate regularly and
readily. This is one cause why they are inferior in deli-
cacy, for, as a consequence of the arrangement, they will
not turn except with a larger weight.
Balances are also liable to set when overloaded. Thus,
if a balance be equally weighted in each pan, but over-
loaded, it will, if placed exactly horizontal, remain so, .but
the slightest impulse or depression on one side destroys
the equilibrium ; the lower side continues to descend with
an accelerated force, and ultimately remains down, being
to all appearance heavier than the other. Generally
speaking, the more delicate a balance the sooner this
effect takes place ; this is one limit to the weight it can
properly carry.
32 WEIGHING.
The vibrations of a balance vary with the quantity of
matter with which it is loaded : the more the weight in
the pans, the slower the vibrations. These should be
observed, and the appearances retained in the mind, in
consequence of the useful indications they afford in
weighing. A certain amplitude and velocity of vibration
would indicate to a person used to the instrument nearly
the weight required to produce equilibrium ; but the same
extent and velocity, with a weight much larger or smaller,
would not be occasioned by an equal deficiency or redund-
ancy of weight, as in the former case.
The weight also required to effect a certain inclination
of the beam, or to turn it, should be known, both when it
is slightly and when it is heavily loaded. Thus, if the
instrument turns with T ^Vo tn of a grain, with 1,000 grains
in each pan, or with T o~o-g-^- o^h of the weight it carries, it
may be considered perfect.
The friction of the knife-edges must be as slight as
possible.
Most of the faults in the working of a balance, if
ordinarily well made, depend upon imperfections in the
middle knife-edge and the planes upon which it rests.
The edge is made either of agate or ste.el, preferably
the former, and should be formed out of one piece, and
finished at once, every part of the edge being ground on
the same flat surface at the same time. In this way the
existence of the two extreme or bearing parts of the edge
in one line is insured ; but when the two parts which bear
upon the planes are formed separately on the different
ends of a piece of agate or steel, or, what is worse, when
they are formed on separate pieces, and then fixed one on
each side the beam, it is scarcely possible they should be
in the same line : and if not, the beam cannot be correct.
These knife-edges usually rest on planes, or else in curves.
The planes should be perfectly flat and horizontal, and
exactly at the same height ; the curves should be of equal
height, and their axes in the same line. If they are so,
ancl the knife-edge is perfect, then the suspension will be
accurately on the line of the edge, and reversing the beam
THE BALANCE. 33
will produce no change. The balance must always be
kept perfectly level by means of the three screws on
which it stands, and adjusted by the spirit-level or plum-
line with which it is furnished.
The balance should be kept in a well-lighted dry room,
quite away from acid or other vapours. The case should
be kept closed as much as possible, and a glass vessel full
of lumps of good quick-lime should be kept in it. When
the lime falls to powder it should be renewed.
In order to test the accuracy and delicacy of a balance,
remove the pans and their end supports, and notice how
the beam oscillates. When it has been found to oscillate
with regularity, and gradually to attain a horizontal posi-
tion of rest, it should be reversed -that is, taken up and
turned half-way round, so as to make that which before
pointed to the right now point to the left. The beam
should then again be made to oscillate, and if it perform
regularly as before, finally resting in a horizontal position,
it has stood a severe test, and promises well. Then re-
place the pans and repeat the tests, noticing the time
required for each oscillation. When the pans are hung
upon the beam, the balance should of course remain
horizontal. They should be tried by changing, then by
reversing the beam, and afterwards by changing the pans
again. The pans are best suspended by very thin platinum
wire, so as to avoid hygrometrical influence upon them.
Afterwards load the balance with the full weight it
is intended to carry say 1,000 grains in each pan, and
notice if the indications are as rapid upon adding or
subtracting the smallest weight as they were when the
pans were empty.
Tests of this kind are quite sufficient for the purpose
of the assayer, who, having ascertained that his balance,
whether slightly or fully laden, vibrates freely, turns deli-
cately, and has not its indications altered by reversing the
beam or changing counterpoising weights, may be perfectly
satisfied with it.
The irregularities which may be discovered by these
tests are best corrected by a workman ; but as in all the
D
34 WEIGHING.
best balances now made adjusting screws for these pur-
poses are provided, it has been thought advisable to
introduce here such matter as, after careful perusal, will
enable every one to adjust and examine his balance
properly ; so that, in the absence of a skilled workman, it
may without much danger be put into working order by the
assay er himself, if accidentally damaged by rough treatment.
THE WEIGHTS. Various kinds of weights are necessary
for the different balances required by the assay er. For
the larger balance, Troy-weights from 4 Ibs. to ^ grain
will be requisite ; for the second size, weights from 1,000
grains to y^th part of a grain ; and for the assay balance,
weights from 1,000 grains to ToVo tn f a g ram -
The best material adapted for weights is unquestion-
ably platinum. This is, however, too expensive for its
general adoption, and therefore brass weights are almost
invariably employed down to the ten- or twenty-grain
weight, the smaller ones only being of platinum. On the
Continent weights are generally made of silver, and if of
brass are electro-gilt. For the smallest weights of all
(those below 0*10 grain) aluminium is often used, its
lightness, and consequently greater bulk, enabling these
small weights to be made considerably larger than if they
were of platinum. The riders are generally of silver-gilt
wire. The slight tarnish which gradually forms on brass
weights may be disregarded until it becomes thick.
Weights ought never to be touched with the fingers, and
should, when not in use, be kept tightly fastened in their
box, away from all acid fumes. The most convenient series
in which to have the weights is 600, 300, 200, 100, 60, 30,
20, 10, 6, 3, 2, 1, -6, -3, -2, -1, &c. This is preferable
to the series formerly employed, as it admits of the use of
a less number of weights to arrive at any required amount.
According to Deville and Mascart receptacles lined
with velvet are not adapted for the preservation of weights,
as dust is deposited in the velvet and acts upon the weights
when taken out or put in. Small boxes of ivory or smooth
wood are preferable.
Peculiar weights are necessary for the assay of gold
and silver bullion in England (with the exception of assays
GOLD AND SILVER ASSAY POUNDS.
35
for the Bank of England ; see Gold assay), gold being
reported in carats, grains, and eighths, and silver in ozs.
and dwts. The most convenient quantity of either of
the precious metals for assay is 12 grains. The quantity
taken, however, is of no very great consequence ; but
whatever its real weight, it is denominated in England
the assay ' pound' This assay ' pound ' is then subdivided
into aliquot parts, but differing according to the metal.
The silver assay ' pound ' is subdivided, as the real Troy
pound, into 12 ounces, each ounce into 20 pennyweights,
and these again into halves (the lowest report for silver),
so that there are 480 different reports for silver, and
therefore each nominal half-pennyweight weighs ^th part
of a Troy grain, when the ' pound ' is twelve grains.
Assay Weights for Silver.
Assay
grains.
12
11
6 .
3
2
1
0-500
0-250
0-150
0-100
0-050
0-025
The gold assay 'pound is subdivided into 24 carats,
each carat in to 4 assay grains, and each grain into eighths,
so that there are 768 reports for gold; and the assay
* pound ' weighing 12 Troy grains, the lowest report, or -|th
assay grain, equals -g^th Troy grain ; thus
Assay Weights for Gold.
Silver
ozs. dwts. grs.
12
11
600
300
200
100
10
050
030
020
010
12
(
carats.
24
22
12
Joid
grs.
o
eighths.
.
.
o . .
o
Assay
grains.
12
11
. . . .6
3
3
o
o
Iffths
2
1
o
2
1
.
.
. .
o
'. '. '. '. ffths
6 .
- 6 -ths
o
o
3
e?ths
2 .
1
D 2
36 THE WEIGHTS.
In cases where the very smallest weights have to be
employed, great care must be taken in seizing them with
the forceps, as they are apt to spring away and be lost.
In the assay balance (fig. 13) the use of weights less than
-j^thofa grain is avoided by a very ingenious contrivance.
Each side of the beam is equally divided into ten parts,
and over the beam on each side is placed a sliding rod,
as represented in the figure. The object of these rods is
to carry, in the direction of the beam, the small bent piece
of wire (letter c, fig. 13) called a rider, which serves in
lieu of the smallest weights the T ^th and the yo^^h-
These riders are thus employed : one weighing y^th of a
grain is placed on the cross-piece of the extremity of the
sliding rod just mentioned, and the rod thus furnished is
brought gradually along the beam from the centre to the
end, until the rider can be deposited on the division on
the beam marked 10 ; the balance is then loaded on that
side with a weight equal to T V^h of a grain. If now the
rod be advanced to the centre of the balance, and the rider
dropped on the mark 5, the half of T Vth of a grain will be
pressing on that side of the balance, or, in other words,,
05 of a grain ; and when the rider is at the marks 1, 2,
3, 4, respectively, '01, -02, -03, -04 of a grain will be
indicated. With a rider weighing y^^th of a grain thou-
sandths of grains may be indicated : thus the last rider
placed on the marks 1, 2, 3, 4 would equal -01, -0002 y
003, -004 grain, &c.
THE METHOD OF WEIGHING. The operation of weighing is
very simple ; but as in the hands of the assayer it becomes
one of great frequency, the facilities for its performance
require to be mentioned. It should in the first place be as-
certained before every operation that the balance is in order,
so far as relates to its freedom of vibration, and also that
no currents of air are passing through the case, so as to
affect its state of motion or rest, a situation being chosen
where such influence may be avoided. In most cases there
is a small projecting arm on the upper part of the beam,
which, being turned either to the right or left hand side
of the beam as required, serves to establish perfect equili-
WEIGHING 37
brium. Perfect equilibrium is, however, a matter of no
consequence if the assayer observes one or two simple
rules. He should never on any account weigh by the direct
method, that is, he should never obtain the weight of a
substance by putting it at once into one pan and then
counterpoising it by adding weights to the other pan.
This method is only to be relied on when the balance is of
rare perfection, and is used by no one but the assayer him-
self. The plan of weighing by difference should invariably
be adopted. By this means the weight of any body can
be readily ascertained, no matter whether the arms of the
balance are of unequal length or the pans out of equilibrium.
In the first place, it should be a rule that one pan,
preferably the left, be reserved for the substance to be
weighed, and the other pan be set apart for the weights.
Supposing the weight of a portion of mineral is re-
quired. First place a clean watch-glass, or platinum
<capsule, in the left pan, and carefully ascertain its weight.
Let us suppose it weighs 106*347 grains ; now put the
mineral in the watch-glass and ascertain the united weight
of the two. This we will imagine comes to 763*776. By
subtracting the weight of the glass or capsule from this
we find the true weight of the mineral, which is 763*776
-106*347 = 657*429. The substance to be weighed must
never be put direct into the pan. By weighing in this
manner by difference, the errors arising from inequality
in the equilibrium or length of arms are eliminated.
Nothing should ever be weighed until it is perfectly
cold. It is also unadvisable to weigh anything immediately
after it is taken from a cold place to a warmer one, as the
substance in such case will act as a hygroscopic body,
and, by condensing moisture, will appear heavier than it
really is.
Powders are conveniently weighed by filling a small
stoppered tube bottle with them, then weighing the whole,
and, after pouring out the requisite amount of its contents,
reweighing the bottle and powder. The difference gives
the weight of powder used. This is a very convenient
plan if several portions of the same substance are required
.38 WEIGHING.
for different analyses. The tube will require reweighing
each time after the quantities required for each analysis
are shaken out into the receptacles.
A delicate balance is always furnished with means of
supporting the pans independent of the beam ; and the
beam itself is also supported when required by other bear-
ings than its knife-edges, and in such a manner as to admit
of the rapid removal of these extra supports when the
instrument is to be free for vibration. This is done that
the delicate edges of suspension may not be injured by
being constantly subjected to the weight of the beam and
the pans, and that they may suffer no sudden injury from
undue violence or force impressed upon any part of the
balance. When, therefore, a large weight of any kind is
put into or removed from the pans, it should never be
done without previously supporting them by these contriv-
ances ; for the weight, if dropped in, descends with a
force highly injurious to the supporting edges ; also, if a
large weight be taken out without first bringing the pans
to rest, it produces a similarly bad effect.
The weights should not be put into the pan at random,
It is a mistake to suppose that time is saved by such a
plan. The highest probable weight should be added first,
and then the set should be gone through systematically
down to the smallest weight, retaining or removing each
weight in order according as it is too little or too much.
The exact weight of a body will be found in this manner
in far less time than would be required were the weights
added by guesswork.
When a weight is put in which is assumed to be nearly
equal to the substance to be weighed, the balance should
be brought to a state of rest, and should then be liberated
gradually by turning the handle, so as to leave the pans
wholly supported by the beam. The whole being on its
true centres of suspension, it will be observed whether the
weight is sufficient or not ; and the rapidity of ascent or
descent of the pan containing it will enable a judgment to
be formed of the quantity still to be added or removed.
Great care should be observed in recording the weight
WEIGHING. 39
in the notebook. The weight should first be ascertained
from an inspection of the vacancies in the box of weights,
and then verified by an examination of the weights them-
selves. This is conveniently done whilst replacing them
in the box, which should be done immediately after each
weighing.
In some cases, where great accuracy is not of so much
importance as rapidity in getting out approximate results,
a plan may be adopted recommended by Mr. F. F. Mayer
in the American c Journal of Science and Art ' for .1861.
He washes the precipitate thoroughly by decantation,
and then introduces it carefully into a bottle the exact
weight of which when filled with distilled water at a cer-
tain temperature is known. Since the precipitate is heavier
than water, the bottle, when filled again, will weigh more
than without the precipitate, and the difference between
the two weights furnishes the means of calculating the
weight of the precipitate.
In case the precipitate settles but slowly, it may be
collected on a filter, and, together wdth a filter, after wash-
ing, be introduced into the bottle, in which case the weight
of the filter and its specific gravity, supposing any differ-
ence should exist between its own and that of water, is to
be taken into account. Precipitates soluble in or affected
by water may be weighed in some other liquid.
Mr. Mayer applied this principle on a large scale as far
back as 1855.
In that year he was engaged in the manufacture of
lead carbonate from refuse lead sulphate, by treating the
latter, in a pulpy condition, with sodium carbonate. The
lead sulphate used contained very varying proportions of
water and soluble impurities, from which latter it had
first to be freed by washing. It was then in the state of a
thin pulp, and the difficulty was to find the amount of the
dry lead sulphate, as it was a matter of importance to use
as little sodium carbonate, and to obtain as pure a lead car-
bonate and sodium sulphate, as possible. This could only
be done by weighing it in the bulk or in portions ; but as
the drying of a tubful of lead sulphate (from 500 to 1,200
40 WEIGHING MOIST PRECIPITATES.
Ibs.) was impracticable, and sampling not less so, since
the upper strata contained a much larger proportion of
water than the lead at the bottom, the following method
was contrived, which enabled the management of the pro-
cess to be left in the hands of a workman :
A strong oaken pail was taken, weighing 8 Ibs. when
empty, and a black mark was burnt in horizontally around
the inside of the pail two inches below the rim, up to
which mark it held 20 Ibs. of water. The specific gravity
of lead sulphate being 6-3, the pail, if filled up to the
mark, would hold 126 Ibs. of pure lead sulphate. The
specific gravity of water being 5*3 less than that of lead
sulphate, it followed that if there were 1 Ib. of water in the
pailful of moist sulphate, the pail would weigh 5- 3 Ibs. less
than 126 ( + 8, the tare of the pail) =120-7 ( + 8) ; if there
were 2 Ibs. of water present, the weight would be 115-4
( + 8), and so on. This enabled a table to be calculated
giving in one column the actual weight of the pail when
filled with moist sulphate, and opposite, in a second column,
the amount of dry sulphate corresponding to the gross
weight. The weight of dry sulphate was thus found as
accurately as could be desired, although the amounts
varied in practice from 30 to 105 Ibs.
This is nothing but an application of the Archimedean
theorem, that when a solid body is immersed in a liquid
it loses a portion of its weight equal to the weight of the
fluid which it displaces, or to the weight of its own bulk of
the liquid.
Hence the rule, which is of great convenience in
volumetric analysis, that to find the weight of a moist
precipitate which is a compound of known specific gravity,
weigh it in a specific gravity bottle or some other vessel
of known weight when filled with water or any other
liquid at the normal temperature ; again fill it with the
water or other liquid, divide the excess of the new weight
by the specific gravity of the substance, less that of the
water or other liquid (that of water being =1), and add
the quotient to the overweight, which gives the weight of
the precipitate.
INCINERATING PRECIPITATES. 41
INCINERATING PRECIPITATES PREVIOUS TO WEIGHING THEM.
A previous complete drying of the precipitate, as Bunsen
has proved, is in most cases not merely a loss of time but
a disadvantage ; whilst introducing the still moist precipi-
tate into the crucible requires the application of a very
gentle heat at the outset, and thus insures the most
favourable conditions for the easy and complete incinera-
tion of the filter-paper. Precipitates not washed upon the
filter-pump can be readily brought to a sufficient degree
of dryness if laid for a short time upon blotting-paper or
unglazed earthenware.
Dry filters may be also much better incinerated after
previous charring at the lowest possible temperature than
by rapid carbonisation or direct ignition in the flame.
How advantageous it is to char previously very slowly
may best be seen on incinerating filters whose contents
impede the complete combustion of the paper by the old
process, e.g. silicic acid, ammonio-magnesium phosphate,
&c. Charred paper obtained by rapid heating is deep
black and of a silky lustre, whilst if slowly carbonised it
is brownish-black, dull, and smoulders away like tinder.
Charred paper of the first kind appears under the micro-
scope perfectly amorphous, whilst the other displays the
carbonaceous skeleton of the fibre.
A careful removal of precipitates from the filter with
the exception of cases like zinc and cadmium, where
volatile reduction-products may be formed is quite use-
less, since the errors which it was hoped to obviate are
not really avoided. On incineration with the filter, wet
or dry, an error due to reduction may be easily corrected,
e.g. in barium sulphate with sulphuric acid ; in lead sul-
phate with nitric and sulphuric acid ; in iron and copper
oxides with nitric acid ; in silver chloride with nitric and
hydrochloric acids, &c.
CALCINATION.
CHAPTEE III.
GENERAL PEEPAEATORY CHEMICAL OPERATIONS.
CALCINATION. Strictly speaking the term calcination means
the production of an oxide or Calx by combustion, and
it necessarily involves the intervention of atmospheric
oxygen. But in a metallurgical sense the term is re-
stricted to the separation of any volatile matter from a
mineral substance by the aid of heat alone, the atmosphere
being totally or partially excluded ; or the production of
rapid changes of temperature, so as, for instance, to render
minerals more fragile by heating and then quenching in
water, &c.
Thus we speak of the calcination of minerals, as iron
or zinc ores, &c., whose matrices are argillaceous, to
expel water, and also of gypsum to expel water ; calcium,
iron, copper, and lead carbonates are calcined to separate
carbonic acid ; zinc and iron hydro-carbonates, to get
rid of both water and carbonic acid ; cobalt and nickel
ores, &c., to separate arsenic and sulphur. The iron ores
found in the vicinity of collieries are calcined to expel
bituminous matter, and wood and bones to expel volatile
organic matter. Where the operation is accompanied by
combustion, and requires the oxygen of the atmosphere,
it is termed roasting.
Crucibles are conveniently used in calcination, as no
stirring of the mass is required. They may be made of
various materials, as clay, plumbago, platinum, silver, and
iron. Silver must not be employed when sulphur is pre-
sent, and it must not be exposed to a heat greater than
dull redness. The selection of the crucibles must depend
upon the substance under operation ; they must all be
furnished with covers.
CALCINATION. 43
In almost all operations in assaying it is necessary to
estimate the amount of volatile matter lost by calcina-
tion. A very high temperature is seldom required in
calcination ; usually an air-furnace will give enough heat.
When the operation is finished the crucible must be
removed from the fire and allowed to cool gradually.
When completely cold, remove the cover and take out
the contents by means of a spatula. If any adhere, a
small brush will be found very useful for its removal.
The difference in weight before and after calcination will
represent the volatile matter.
When the subject to be calcined is fusible, the crucible
and contents must be weighed before ignition ; the loss of
weight is equal to the quantity of volatile matter expelled ;
in fact, this latter is usually the most satisfactory method
of conducting the experiment.
If the ignited substance be soluble in water, it can be
removed from the crucible by that means, employing heat
if required ; if not, any suitable acid may be used.
If the substance to be calcined decrepitates on heating,
it must be previously pulverised, and heated slowly and
gradually in a well- covered crucible.
Certain substances, as lead carbonate, undergo a
material alteration by contact with the gases given off
during the combustion of the fuel in the heating furnace ;
others, such as carbonaceous matters, are consumed by
the introduction of atmospheric air. All such substances
must be calcined in a closely covered crucible placed in a
second crucible (also covered) for further protection.
In some rare cases, however, these precautions are not
sufficient. In such, either a weighed porcelain or German
glass retort must be employed.
Sometimes earthenware crucibles lined with charcoal
are employed in calcination ; for even if the substance be
fusible, it may generally be collected and weighed without
loss, as very few bodies either penetrate into or adhere
to a charcoal lining. In this way grey cobalt and other
arsenio-sulphides are calcined at a high temperature to '
expel the greatest possible amount of arsenic and sulphur.
44 ROASTING OEES.
The selection and proper management of crucibles will
be given in the next chapter.
BOASTING. In this operation carbon, sulphur, selenium,
antimony, and arsenic are separated from certain metals
with which they were combined. Eoasting differs from
calcination in this particular : the latter is carried on in
close vessels, independent of the atmosphere ; the former,
in open vessels by the aid of the atmosphere. It is thus
we are enabled to separate the bodies just mentioned by
this process ; for the oxygen of the air, by combining
with them, forms a volatile substance which the heat
expels. Thus, in roasting copper and iron sulphide
(copper pyrites), 'the sulphur, copper, and iron mutually
combine with oxygen' to form sulphurous anhydride (vola-
tile), copper protoxide, and iron peroxide, thus :
2(FeS + CuS) + 130 = Fe 2 3 + 2(CuO) + 4(SO S ).
This is the final change in this case. During the process,
however, some copper and iron sulphates and sub-sul-
phates are formed. This change will be explained under
the head of Copper Assay.
When carbonaceous matters are roasted, the operation
also takes the name combustion, or incineration ; because
the object of roasting a fuel, for instance, is generally to
ascertain the amount of ash left.
In roasting, in the ordinary acceptation of the term,
the body must not be fused, but kept in a pulverulent
state ; there are, however, some cases in which fusion is
allowable, as in cupellation and .scarification.
The process of roasting is performed in different ways.
In one, a small flat vessel, called a roasting-dish (fig. 14),
is employed, made of the same material as the earthen
crucibles, and similar to a saucer. It is most conveniently
heated in muffle. The substance to be roasted must be
finely pulverised, placed in the roasting-dish, and con-
stantly stirred with an iron or glass rod until no fumes
are given off, or until it ceases to evolve the odour of
sulphurous acid if sulphur is one of the constituents to be
eliminated.
ROASTING ORES. 45
The operation may also be performed in a crucible,
in which case it must be inclined to the operator, so
that the draught of air passing to the furnace flue may
impinge as much as possible on the substance under
manipulation.
During roasting the heat must be carefully regulated
for some time. At first it ought only to be the dullest
red ; and the substance must be assiduously stirred in
order to present the largest possible surface to the action
of the atmosphere and prevent fusion, for some assays,
when roasting, will fuse readily at a low temperature
unless the surface be continually renewed. Even by
paying the utmost atten- FlG 14
tion to this point it can-
not be always prevented,
as, for instance, when
antimony sulphide is
being roasted. In these
cases the assay must
be mixed with its own
weight of powdered quartz or fine white sand (silver
sand) ; the operation will then proceed steadily.
If the assay at all agglutinates, it must be taken from
the fire and rejected if the substance be plentiful ; if not,
the fused mass must be carefully removed from the cru-
cible or dish, pulverised, and the roasting recommenced.
In this case, however, the operation is always very tedious,
and the final result less exact, so that great care ought to
be taken at the commencement of the roasting.
When the assay has been kept at a dull red heat for
some time, and shows no signs of agglutination, the heat
may be slightly increased ; at the same time stirring must
be diligently pursued. After the heat has arrived at full
redness there is little fear of fusion ; and as the operation
proceeds more rapidly at a high temperature than at a low
one, it is well now to increase the heat to a yellowish red,
and even in certain cases to nearly a white heat. If the
stirring of the assay has been constant during the various
gradations of heat, the roasting at this point will be
46 ROASTING OEES.
accomplished ; and the remaining operations of the assay
may be proceeded with.
This is the general plan of operation, but different sub-
stances require for roasting a different degree of heat ; for
instance, copper pyrites require a higher temperature than
grey copper ore, and the heat employed must in every
case be adapted to the substance to be roasted. Some
substances, for instance, arseniates, lead sulphate, &c.,
cannot be roasted by heat alone. These require the
addition of a carbonaceous body to remove the combined
oxygen and allow the arsenic, sulphur, &c., to be com-
pletely roasted off. Ammonium carbonate in some cases
is also added to the mixture to decompose the sulphates
formed during the roasting of sulphides.
In cases where the metallic bases of the sulphides are
volatile, either as such or as oxides, as for instance galena,
antimony sulphide, &c., a loss of metal will always result
during the roasting process.
It may be as well to mention here that platinum cap-
sules are useful in certain roasting operations. Copper,
iron, and molybdenum sulphides are conveniently oxidised
in this kind of vessel, without much fear of injury, pro-
vided fusion of the roasting substance be carefully avoided.
Platinum vessels should also be used in ascertaining the
amount of ash in coal.
REDUCTION.- The process of reduction consists in re-
moving oxygen or an analogous element from any body con-
taining it, usually by means of either carbonaceous matter,
hydrogen, or a body containing both these elements, and
leaving the metal behind, usually in the form of a melted
button. The rationale of the operation is as follows, when
lead oxide is reduced with carbon :
2(PbO) + C = 2Pb + C0 2 .
In this case we start with lead oxide and carbon, and as
a result we obtain metallic lead and carbonic acid.
The reaction between nickel oxide and hydrogen is
thus expressed :
MO + 2H = Ni + H 2 0.
REDUCTION. 47
Here we have at the commencement nickel oxide and
hydrogen ; and after the conclusion of the operation there
remains metallic nickel, and water which has volatilised. If
the reducing substance contain both carbon and hydrogen
the action will be thus, when a metal (e.g. lead) is reduced
from its oxide, carbonic acid and water being formed :
3(PbO) + CH 2 - 3Pb + C0 2 + H 2
In the operation of reduction by the aid of carbona^
ceous matters two methods are employed : in the one,
charcoal, coal, sugar, starch, or any carbonaceous or
hydro-carbonaceous body, as argol, is mixed with the
substance to be reduced ; in the other, the process of
cementation is employed. Where sulphides are to be
reduced, metallic lead or iron is usually employed to
remove the sulphur. Generally, however, the sulphides are
previously converted into oxides by the operation of
roasting, and the reduction is then effected by means of
carbonaceous matter.
The process of cementation is conducted by placing
the oxide to be reduced in a crucible lined with charcoal,
and covering it closely while it is in the furnace ; the re-
duction proceeds gradually from the outside of the oxide
to the centre of the mass. The time requisite for this opera-
tion depends on three circumstances viz. the nature of the
oxide, the degree of temperature, and the mass acted on.
Some oxides treated this way are reduced very readily ;
others, again, take a considerable time ; while certain
of them do not appear to be acted on beyond the outer-
most layer. Of the first class is nickel oxide ; of the
second, manganese oxide ; and of the third and last, chro-
mium oxide.
Each of these classes of reduction has its advantages.
The former, or reduction by mixture with carbonaceous
matter, takes place very quickly and completely, but the
reduced metal is often mixed with carbon ; in the latter
process the residue is comparatively pure, but it is not
' generally preferred, on account of the time and high tem-
perature necessary.
48 REDUCTION.
Eeduction by hydrogen gas is very seldom employed ;
it is, however, necessary in some cases, as for instance in
the estimation of the percentage of cobalt or nickel in
a sample, where perfect accuracy is desirable. The opera-
tion is carried on in a tube of hard German glass, having
a bulb blown in its centre, which is heated either by a
spirit or gas lamp. Attached to it is a tube full of dried
calcium chloride, through which the hydrogen gas effect-
ing the reduction passes to perfectly dry it.
The bulb tube is weighed and the oxide introduced into
it ; it is again weighed, and the apparatus united by caout-
chouc tubes ; hydrogen gas (see Eeducing Agents) is then
passed through it until the whole of the atmospheric air
is expelled. Heat is afterwards applied till the bulb is
bright red, and the current of gas continued until no more
water from the decomposition of the oxide is formed ; the
source of heat is then removed, and the current of gas
continued until the apparatus is cold. The bulb tube,
with the reduced metal, is then weighed, and the amount
which it has lost represents the oxygen which the hy-
drogen has removed. By subtracting this oxygen from
the original weight of the substance, the difference gives
the amount of metal in the amount of oxide operated on.
FUSION. This operation is sufficiently simple, and is
employed in all assays by the dry way, in order to obtain,
in conjunction with the last process, a button or prill, as
it is termed, of the metal whose assay is in progress. It is
also a necessary step in the granulation of metals, the
preparation of certain fluxes and alloys, also lead for the
assay for silver, in order that a homogeneous ingot may
be obtained. Some ores, such as those of copper, are
melted instead of being roasted or calcined, in order to
prepare them for reduction. Minerals are also melted per
se, or with the addition of borax or sodium carbonate, in
order to ascertain the best treatment to be adopted in a
subsequent operation. Metals too are frequently melted
to drive off other volatile metals ; in this case the heat
should be continued for some time, and should be very
high, as it is difficult to remove the last traces of volatile
SOLUTION. 49
metals. Thus, in melting the spongy gold left behind in
the retort after the distillation of gold amalgam, the ingot
of gold almost always retains mercury, which can only be
removed by repeated meltings at a very high temperature.
In some cases the fusion is intended to be only partial,
the object being to melt out an easily fusible part of the
mineral for instance, in assaying grey antimony ore and
different bismuth ores.
SOLUTION. In all cases where analysis in the wet way
is required, the mineral must be either wholly or partially
brought into the state of solution. The choice of a solvent
necessarily depends upon the nature of the material under
treatment. In some few cases water will be sufficient ; but
in the majority acids are required. Sometimes advantage
will be derived by first extracting all that water will dis-
solve, and then applying acids to the residue. In speaking
of the minerals, &c., which require solution for their assay,
the most appropriate solvents will be pointed out. In all
cases heat promotes solution.
Solution is best effected in glass flasks ; clean Florence
oil flasks are very appropriate for most purposes. They
may be supported on a hot sand bath, or on a metal ring
or coarse wire gauze, over the naked gas or spirit-flame.
The flask should then be placed in a sloping position, so
that when the liquid boils or effervesces from the escape of
gas, the drops spirted up may strike against the sloping
side, and run back into the liquid instead of being thrown
out of the mouth.
A porcelain dish may also be used, although from the
great surface exposed these vessels are more appropriate
for evaporation than solution. Beakers may likewise be
employed, but they should be covered over with an in-
verted funnel sufficiently large to rest within the top edge
without slipping down more than about half an inch ; or
a large watch-glass or dial-plate turned concave side up-
wards may be used as a cover. Both the funnel and dial-
plate serve the double object of keeping out dust and
preventing loss of the liquid by projection of fine drops
during ebullition.
E
50
GLASS AND PLATINUM FORCEPS.
In many cases solution of the whole or part of a
mineral must be preceded by its fusion at a high tempera-
ture with sodium carbonate, nitre, or some other flux.
The fused mass must then be well extracted by boiling
water, when the residue will usually be found soluble in
hydrochloric or other acid. Special instructions in this
FlG 15 F IG> 16> method of effecting solu-
tions will be given in those
cases where it is necessary.
Where it is necessary
to manipulate in acid or
other solutions, the glass
and platinum forceps de-
scribed by David Forbes,
F.E.S., in the 'Chemical
News' for October 2, 1868,
will be found useful.
The accompanying
woodcut, fig. 15, shows
them in front and side
view, and will require but
little explanation. They
are made as follows : an
ordinary pair of strong
surgical forceps are taken
and the points cut off; a
small piece of sheet brass,
bent into a cylinder, is
then soldered to each arm
as shown at a ; these cylin-
ders being formed by merely bending the brass round, so
as to leave an open slit about one twentieth of an inch
wide in front. Two glass rods, such as are used for
stirrers, as long as the glass arms of the forceps are in-
tended to be, and just as thick as will enter these brass
cylinders when pressed with some force, are rounded by
the blow-pipe at the one end, whilst the other, when
softened, is somewhat flattened between the glassblower's
pliers, as seen in the woodcut. In order to complete the
DISTILLATION. 51
forceps it is now only necessary to push each of these rods
into its corresponding brass cylinder or socket, the longitu-
dinal slits of which, by imparting a certain amount of
elasticity to the sockets, cause them to grasp the glass rods
firmly, and retain them without any cement or other fixing.
The relative lengths of the arms are easily adjusted by
slipping one rod more or less forward, whilst the points
can be made to hold and meet accurately, by rubbing
them down on a piece of sandstone.
Such forceps may, of course, be made to any convenient
size ; the one figured in the woodcut is drawn to exactly
half-size, and is found to be of very useful dimensions for
general analytical work, especially when manipulating in
nitro-hydrochloric acid, nitrate of silver, and other solutions
which would have acted upon metals, horn, ivory, &c.
Fig. 16 represents another convenient form of forceps,
also drawn to one half the real size, with long platinum
points soldered to the steel body at a\ these have been
also found of great service in general laboratory operations,
especially when hydrofluoric acid is in question.
DISTILLATION. There are two distinct classes of this
operation : in the one, liquids are submitted to experiment
with the object generally of separating them from sub-
stances which are non-volatile, and will consequently be
left behind when the liquid comes over. Belonging to
this class may be mentioned the distillation of nitric acid,
the preparation of distilled water, and the separation of
mercury from gold and silver amalgam. In the other
kind of distillation, which goes by the name of dry distil-
lation^ solid bodies, as wood, coal, &c., are subjected to
heat in order generally to ascertain the amount of gas or
other volatile matter given off in the course of an experi-
ment, from a certain quantity of the coal or other sub-
stance operated upon.
In liquid distillation (as in the purification of nitric
acid, &c.), retorts are used. The best form for general use
is that which is furnished with a stopper at the upper
part of the body, a (fig. 17), through which the liquid is
introduced ; the neck of the retort is then placed in that
E 2
52
DISTILLATION
FIG. 17.
of a receiver, b, over which a piece of wet cotton or
woollen cloth is placed, and which must be kept cold
by means of a stream of water
from a funnel, c, the shaft of
which is partially plugged up
with cotton wool. Heat is then
applied to the retort, and as
much of the liquid as is desired
is distilled over into the re-
ceiver. It is advisable not to
fill the retort more than two
thirds full, and to apply the
heat at first very gently, other-
wise there is a risk of breaking
the vessel.
A more convenient form of apparatus for distillation
and condensation is shown at fig. 18, in which a Liebig's
FIG. 18.
condenser is attached to the retort. Fig. 19 will show
the construction of the condensing apparatus. The cold
water passes into the funnel above, is conveyed at once
to the lowest end of the condenser, whilst the heated water
passes off by the upper tube.
Distilled water is a most important agent in the labora-
tory ; and, as much is needed, it is better to have a still
DISTILLATION.
53
specially adapted for its production. Such a one is de-
picted at fig. 20, where A is the body of the still ; B the
furnace in which it is
, /,i ,.-,-. ! , FIG. 19.
set (the still may also be
placed in the portable
furnace, fig. 25, p. 65) ;
C the still head ; D E
the neck ; F the worm ;
I J K L the worm-tub
containing cold water to condense the steam generated in
the still ; M N the pipe to lead fresh cold water to the
FIG. 20.
bottom of the worm-tub, while the warm water runs off
.at the top, as in Liebig's condenser ; and P the vessel in
which the distilled water is received.
In the dry distillation of bodies, earthenware, glass, or
iron retorts are employed ; but for small operations a tube
of wrought-iron, about one inch internal diameter, and
plugged at one end, is found to be a convenient form of
apparatus. It is placed with the substance contained in it
in a furnace, and a small tube, either of glass or pewter,
is fixed by means of a perforated cork to the open end of
the large tube. The gas given off during the operation
may be collected by the aid of a pneumatic trough.
64 SUBLIMATION.
SUBLIMATION. This operation is a kind of distillation-
in which the product is obtained in the solid form. The
apparatus which may be employed for this purpose are
tubes, flasks, capsules, or crucibles. Florence flasks are
exceedingly useful ; they may be sunk in a sand bath, and
the sublimed substance received directly into another flask,
or by passing through an intermediate tube. Sometimes,
however, it is difficult to entirely remove the sublimed sub-
stance ; and in order to avoid this inconvenience, Dr. Ure
has proposed the following very excellent subliming appa-
ratus : It consists of two metallic or other vessels, one of
which is flatter and larger than the other. The substance
to be sublimed is placed in the smaller vessel, and its
opening is covered by the larger filled with cold water,
which may be replaced from time to time as it becomes
hot. The sublimed substance is formed on the lower part
of the upper vessel. A large platinum crucible, filled with
cold water, and placed on the top of a smaller one, answers
the purpose of the before-mentioned apparatus very well.
SCORIFICATION CuFELLATiON. These operations will be
described under the head of Silver Assay.
FURNACES. 55
CHAPTEE IV.
PRODUCTION AND APPLICATION OF HEAT.
FURNACES for assay purposes may be heated either by solid
fuel, oil, or gas, and they may be divided into wind and
blast furnaces. In the former the fire is urged by the or-
dinary draught of a chimney, and in the latter by means
of bellows or artificial blast. We shall commence with the
former, as they are in most common use. They are of
various kinds, according to the purposes for which they are
required. The three principal kinds are those for fusion,
calcination, and cupellation. Coal, coke, and charcoal
are the fuels employed, and the merits of each will be par-
ticularly discussed. Blast furnaces are only employed for
the purpose of fusion, although their forms are various ;
charcoal and coke are the fuels most in use, but oil and gas
blast furnaces are used in small laboratory operations, and
for many purposes they are preferable to other furnaces,
on account of their freedom from dust and dirt, and the
perfect control the operator possesses over the heat.
Furnaces consist of certain essential parts viz. first,
the ash-pit, or part destined to contain the refuse of the
combustible employed ; secondly, the bars on which the
fuel rests ; these are sometimes made movable, or are fixed
to a frame ; the former arrangement is more convenient,
as it allows clinker and other refuse matters to be readily
removed ; thirdly, the body of the furnace in which the
heat is produced ; and lastly, in wind furnaces, the chimney
by which the heated air and gaseous products of combus-
tion are carried off.
CALCINING FUKNACE. Calcining furnaces are small and
shallow, because a high temperature is not required. They
56 CALCINING FURNACES.
may be made square or circular ; the former are most
readily constructed, and, where many crucibles are to be
heated at once, they are preferable to the circular ; but the
latter give the greatest degree of heat with the least pos-
sible consumption of fuel, and are to be preferred on that
account where one crucible only is to be ignited.
The body of the furnace is best made with good bricks,
lined with Welsh lump, fire-bricks, or a mixture of Stour-
bridge clay and sand. It is also desirable that a plate of
iron with a ledge be placed over the upper part of the
furnace to protect the brickwork from blows with crucible
tongs, &c., and to keep it in its place when disturbed by
sudden alterations of temperature. The bars of the fur-
nace may be either in one single piece, or made up of
several bars of iron fastened to a frame. They ought to
be as far as practicable from each other, and must not be
too large, although large enough not to bend under the
weight of the fuel and crucibles when they become hot,
and they must not be so far removed from each other
as to allow the coke or charcoal to fall through easily.
Lastly, the more readily the air can find access to the
centre of the fuel, the higher will be the temperature pro-
duced in the furnace ; very simple assays occasionally fail,
only because the bars are either too large or too close
together.
CHIMNEY. Calcining furnaces generally have no fixed
chimney, but are covered with a movable one when a
greater degree of heat is required. This chimney may be
about five feet high, the diameter of the furnace at the
bottom, and tapering off to about two thirds of that dia-
meter at the top. It is made of strong plate iron, fur-
nished with a wooden handle. The lower part is provided
with a door, by means of which the interior of the furnace
may be examined without disturbing the whole arrange-
ment of the chimney, and consequent cooling of the con-
tents of the furnace.
If, during the course of any experiment, noxious or
offensive vapours are expected to be given off, the furnace
must be so arranged that they may be introduced into a
WIND FUKflACE. 57
flue, by fastening a piece of iron plate pipe, furnished with
an elbow joint, on to the movable chimney before spoken
of.
EVAPORATING FURNACES. The furnaces just described
answer exceedingly well in the absence of gas, for heating
small flasks, evaporating basins, &c., when surmounted by
a tripod stand or sand bath. This is necessary, as many
assays by the dry way are preceded and followed by cer-
tain operations in the wet way.
THE HOOD. In order to prevent certain gases or va-
pours from fires, evaporating basins, &c., from entering
into the laboratory, a large metal covering, termed a hood,
is employed, terminating in a chimney having a good
draught. It is best made of sheet or galvanised iron.
FUSION FURNACE WIND FURNACE. The wind furnace,
properly so called, is a furnace provided with a chimney,
and capable of producing a very high temperature.
Wind furnaces are generally square, but, if more than
four crucibles are to be heated at one time, they may be
made rectangular, the chimney being placed at one of the
long sides. When the furnace is required to hold but one
pot, it may, however, be made circular.
The body of the furnace ought to be made of good
bricks, solidly cemented with clay, and bound by strong
iron bands. The bricks must be very refractory, and
capable of sustaining changes of temperature without
cracking. They are ordinarily made with the clay used
in the manufacture of crucibles. In some cases bricks are
not used for the lining of this kind of furnace ; for in-
stance, a mould of wood is placed in the centre, and the
open space between the surface of that and the outer brick-
work is filled with a paste of very refractory clay, each
layer being well beaten down. When the space is filled the
case is withdrawn, and the crust of clay dried with much
precaution, every crack that may be caused by unequal
desiccation being filled up as fast as formed. This method
of manufacture is very applicable to circular furnaces. In
every case, however, it is necessary to border the edge
with a band of iron to prevent injuries from tongs, or pots.
58 WIND FURNACE.
By using a mixture of 1 part of refractory clay and 3 to
4 parts of sifted quartz sand, no cracks are formed during
desiccation. This mixture is used on the Continent for the
interior fittings of Sefstrom's blast furnace, as well as for
larger blast furnaces for manufacturing purposes. It is
said to stand a high temperature exceedingly well.
Makins * recommends for small furnaces the second
kind of bricks, known as Windsor, or in the trade P.P.
bricks. ' These are of a red colour, very siliceous, but soft,
easily cut and shaped, and yet standing heat very well
The best method of cutting them is by a piece of zinc
roughly notched out as a saw, and then the more accurate
figure required may readily be given them by grinding
upon a rough flat stone. In this way the small circular
furnace formerly made by Newman, and sold by him as
his " universal furnace," is lined by cutting the bricks with
care to the radii of the circle they are to form, when they
key in, like an arch, and so need no lining whatever.'
THE ASH-PIT is an open space under the bars, which
serves as a receptacle for ashes, clinkers, &c., produced
during the time the furnace is in use. It should have the
same area as the furnace, and be completely open in front,
so that the air may have free access ; it is well, however,
for the sake of economy, to furnish this opening with a
hinged door, having a register plate fixed in it, so that the
draught may be reduced, or entirely shut off, in order that
the fire may be extinguished when desirable, and fuel
saved which would otherwise be burnt in waste.
On the one hand, it is well to have the power of cut-
ting off access of air into the body of the furnace by the
lower part, either to put out the fire entirely, or to dead en
it whilst putting in a pot ; and, on the other, to attain the
maximum of temperature, we must have the means of
allowing the air to pass with the greatest possible facility
into the furnace. In order to do this it is necessary to
furnish the ash-pit with doors, or valves, whereby the
quantity admitted may be regulated as desired. It is ad-
vantageous to lead the air to the ash-pit from a deep and
* Makiris's Metallurgy, p. 88.
WIND FURNACE. 50
cold place, by means of a wide pipe. A chimney of less
height will then be required.
THE BARS are made in one piece, or are made up of
movable pieces of metal; the latter arrangement is the
most convenient. Wherever a wind furnace is in use, the
upper opening is closed by a cover made of a fire-tile r
encircled with iron.
THE CHIMNEY is a very essential part of a wind furnace :
it is on its height and size that the draught depends, and,
in consequence, the degree of heat produced within the
furnace. In general, the higher and larger the chimney r
FIG. 21.
the stronger is the draught ; so that, by giving it a great
elevation, exceedingly high temperatures may be obtained.
But there is a limit which it is useless to pass in a furnace
destined for operations by the dry way ; and, besides this,
the building a very high chimney presents many difficulties
and much expense, so that in laboratory operations, where
a very strong current of air is required, recourse is had to
a pair of double bellows. A temperature can be produced
in a wind furnace sufficiently strong to soften the most
60 WIND FURNACE.
refractory crucibles, by means of a chimney from
six to forty feet high.
Chimneys are generally made square or rectangular,
and have interiorly the same dimensions as the body of
the furnace. About two feet above the upper part of the
furnace they are furnished with a register or damper, by
means of which the current, of air may be regulated or
entirely stopped at will. The damper is a plate of iron
sliding into a small opening across the chimney.
A wind furnace of the kind above described is repre-
sented by fig. 21.
The left-hand figure in 21 is the plan, the middle the
elevation, and the right is a sectional view. A the body
of the furnace in which the crucibles to be heated are
placed, G the bars, and P the ash-pit ; the cover is formed
of a thick fire-tile of the requisite size, firmly encircled by
a, stout iron band, and furnished with a handle for con-
venience in moving it ; B the flue, C the chimney, E the
damper ; H a hood over the furnace, supported by iron
bands h h h ; M the handle of a ventilator T 7 , which serves
to carry off hot air and fumes from furnace when open ;
and finally, /S, a small sand bath, in which to set the red-
hot crucibles when taken from the fire ; one foot square
inside the fireplace of the furnace is a good and convenient
size ; the remainder will then be in proportion.
BLAST FURNACES. In this species of furnace the air
necessary to keep up the combustion is forced through the
fuel by means of a blowing apparatus, instead of being-
introduced by the draught of a chimney as in the wind
furnace.
The most convenient apparatus for forcing air into a
furnace is a double bellows ; a fan may be used, but it is
not so powerful.
The quantity of air passing into a furnace varies with
the length of the assay, and ought to increase gradually
as the temperature becomes higher.
The following is the description of a most excellent
blast furnace which has been in use for some years in the
laboratory of the Royal Institution : The temperature
BLAST FURNACE. 61
produced by it is extraordinary, considering the small
amount of time and fuel employed. It is sufficiently
powerful to melt pure iron in a crucible in ten or fifteen
minutes, the fire having been previously lighted. It will
effect the fusion of rhodium, and even pieces of pure
platinum have sunk together into one button in a crucible
subjected to its heat.* All kinds of crucibles, including
the Cornish and Hessian, soften, fuse, and become frothy
in it ; and it is the want of vessels which has hitherto put
a limit to its application. The exterior (fig. 22) consists
of a black-lead pot, eighteen inches in height, and thir-
teen inches in external diameter at the top ; a small blue-
pot of seven and a half inches external diameter at the top
has the lower part cut off so as to leave an aperture of
five inches. This, when put into the larger part, rests
upon its lower external edge, the tops of the two being
level. The interval between them, which gradually in-
creases from the lower to the upper part, is filled with
pulverised glass-blowers' pots, to which p IG< 2 2.
enough water has been added to moisten
the powder, which is pressed down by
sticks so as to make the whole a com-
pact mass. A round grate is then
dropped into the furnace, of such a size
that it rests about an inch above the
lower edge of the inner pot ; the space
beneath it, therefore, constitutes the air-
chamber, and the part above, the body
of the furnace. The former is 7-J inches from the grate
to the bottom, and the latter 7-^ inches from the grate to
the top. Finally, a longitudinal hole, conical in form, and
1^ inch in diameter in the exterior, is cut through the
outer pot, forming an opening in the air-chamber at the
lower part, its use being to receive the nozzle of the
bellows by which the draught is thrown in.
Sefstrom's blast furnace, obtainable at most chemical-
instrument makers, is also very powerful and convenient ;
it consists of a double furnace. It is made of stout sheet-
* Faraday.
02 CUPEL FURNACE.
iron, lined with fire-clay, and is used with coke, or char-
coal and coke, broken into pieces of about a cubic inch in
size. The blast of air is supplied by a powerful blowing-
machine. It will readily produce a white heat. Indeed
the limit to its power seems to be the difficulty of finding
crucibles or interior furnace fittings which will stand the
temperatures produced in it without softening. Kersten
states that he increases the heat in Sefstrom's blast furnace
by using a hot blast.
H. Ste.-Claire Deville has employed for melting platinum
a furnace 12 inches high, and 11 inches wide, which
rests upon a cast-iron plate full of holes. This is connected
with a forge bellows. After blowing for a few minutes,
the temperature of the furnace will have reached the
highest possible degree, but this zone of maximum heat
only extends to a small height above the bottom of the
furnace. Above this point a considerable quantity of car-
bonic oxide gas is formed, which burns with a very long
flame. The heat produced in this furnace is so high that
the best crucibles melt, and only crucibles made of good
and well-burned lime can be used.
THE MUFFLE OR CUPEL FURNACE is a furnace in the centre
of which is placed a small semi-cylindrical oven, which is
termed the muffle. These furnaces were in use as early
as the thirteenth century. Their construction and dimen-
sions depend
1. On the temperature which the furnace is intended
to produce ;
2. On the number of cupellations required to be per-
formed at one time ; and
3. On the kind of fuel to be used.
The muffles are mostly made of refractory clay, gener-
ally of one piece, but it is best to make them of two pieces ;
the bottom being one, and the cover or vault the other.
Muffles are sometimes made of cast-iron, cast in one
piece. They are useful in small furnaces intended for
cupellations only.
Muffle furnaces must always be provided with a good
chimney to carry off the hot gases.
CUPEL FURNACE.
The muffle, being completely surrounded by ignited
fuel, acquires a very high temperature, and in its interior
all operations requiring the presence of air, and which
cannot be carried on in contact with carbonaceous matters,
may be performed such as roastings, scorifications, and
cupellations.
When from ten to twenty cupellations ha.ve to be
effected at one time, large brick furnaces are employed ;
and, in consequence, much fuel is consumed to waste in
these when only a few cupellations are required. This
has occasioned many persons to endeavour to form small
furnaces, where one or two cupellations may be carried
on with the smallest possible quantity of fuel.
MM. Aufrye and d'Arcet have contrived a furnace
which is capable of fulfilling all these conditions.
The furnace is elliptical, about 7 inches wide and 18
high; its ash-pit has but one circular opening, and its
height is such that, when the furnace is placed upon it,
and .the whole upon a table, the assayer can, when seated,
readily observe the course of the assay within the muffle.
The hearth has five openings, in one of which the muffle
is placed ; in another a brick to support it ; a third is for
the purpose of introducing a poker to stir the ashes, and
make them fall through the grate-holes : this can be closed
with a small earthen plug ; and lastly, there are two round
holes, placed in its largest diameter, to facilitate the in-
troduction of air, either by draught or a pair of bellows,
as the case may require. The support for the fuel is
generally a plate of earthenware, pierced with holes, and
bound round with iron wire to keep it together in case it
cracks by changes of temperature ; but it is better to use
an iron grating.
The dome of the furnace has a circular opening, which
can be closed by a plug of earthenware ; this opening
serves for the introduction of the fuel. A chimney is
necessary to increase the draught; it is made of sheet-iron,
and may be from 1^ to 2 feet in height, and ought to fit
the upper part of the dome very exactly. At its base
there is a small gallery, also of sheet-iron, in which it is
64 CUPEL FURNACE.
intended to place the new cupels, so that they may be
strongly heated before introduction to the muffle. This
saves many of them from fracture.
MM. Aufrye and d'Arcet have estimated the quantity
of charcoal necessary to heat this furnace. The following
are comparative experiments :
Silver Lead Time Standard Charcoal
employed, employed, of assay, used,
No. grains grains minutes thousandths grains
1 1 4 12 947 173
2 1 4 11 950 86
1 4 13 949 93
1 4 10 949 60
Coke or charcoal may be used in this furnace, but the
fire must be lighted by means of charcoal alone, as coke is
very difficult to inflame in a cold furnace. When it is red-
hot it may be fed with coke, or, better still, a mixture of
coke and charcoal.
Where great numbers of cupellations have to be made
at once, the following form of brick furnace is requisite :
Fig. 23 shows an elevation of the furnace ; fig. 24
shows a section. The interior of the furnace is of fire-
brick ; the exterior, of common brick. The upper part
is protected by a plate of iron, and the upper opening,,
through which the fuel is introduced, is covered, when
necessary, by a large fire-tile strongly encircled with an
iron band, to which are attached two handles, by which
the whole can be moved.
The muffle opening, as seen partially open in the dia-
gram, can be entirely closed by means of two sliding doors y
made of sheet iron, running in a stout wrought-iron frame,
built into the brickwork. Immediately below the muffle
entrance are two movable bricks ; these close the openings
through which the fire-bars are introduced ; and still
lower down is the ash-pit door, furnished with a register
for the better regulation of the current of air required by
the furnace. In fig. 24 is shown a brick built into the
back of the furnace, on which the close end of the muffle
is supported. This brick may, however, be replaced by a
crucible or fire-brick standing on the bars of the furnace:
A very useful furnace for small operations is one which
UNIVERSAL FURNACE.
has been aptly termed the ; universal furnace,' as it is
capable of performing all that is required of any furnace
FIG. 23.
FIG. 24.
in an assay (see figs. 25 and 26,^elevation and section). Tt
is much to be recommended for its durability, cheapness,
FIG. 25.
FIG. 26.
and its small size compared with the heat it can produce.
It is constructed externally of sheet iron, very stout, and
F
66 FURNACE OPERATIONS.
is lined with fire-brick, not cemented together, but ground
and keyed as an arch, so that it can never fall out till it
is completely useless. Its height is about 2J feet and
diameter 1 foot ; internal diameter 8 inches and depth of
fireplace 1J feet. It is furnished with five doors, one in
the ash-pit and four in the body of the furnace, two in the
front, one above the other, and two opposite each other,
at the sides. The cover serves as a sand-bath, and when
that is taken off there is a series of cast-iron rings, fitting
the top of the furnace, where basins can be placed either
for the purpose of evaporation, calcination, or roasting.
The two opposite holes serve for the introduction of a
tube in operations where it is requisite to pass a gas over
any body at a red heat. In the lower hole in front can be
placed a muffle for roastings and cupellations, introducing
fuel and crucibles by the upper one ; it also serves as an
opening through which the state of the furnace can be
seen, or the progress of any assay observed.
Iron, manganese, nickel, and cobalt can be fused in
this furnace when it has a flue of about thirty feet in
height attached to it, and by closing the ash-pit door the
dullest red heat, for gentle roastings, can be obtained.
FURNACE OPERATIONS.
Crucibles must be carefully supported in the fire, and
must always be covered. They must stand solidly, and
be at equal distances from the sides and bottom of the
furnace, so as to receive a like share of heat, and they
must be completely surrounded with fuel. If a crucible
is supported on the grate-bars of a furnace, the draught
of cold air will prevent the bottom from getting hot. If
it is supported on the fuel, the bottom gets heated quickly,
but the fuel on burning away allows the crucible to fall
down, and may cause the loss of the contents. For these
reasons a crucible should always be supported on a piece
of fire-brick about three or four inches high. In many
cases an old crucible inverted will serve as a convenient
support. The fire must be got up gradually, so as to
FURNACE APPARATUS. 67
prevent the sides of the furnace and the crucibles within
from cracking from the sudden increase of heat. No time
is saved by urging the fire strongly at first, for crucibles
are bad conductors of heat, and a high temperature at
the commencement scarcely causes the heat to penetrate
to the interior faster than a moderate redness. After the
furnace has arrived at a full red heat, more air may be
given, and in from about twenty minutes to one hour
the assay will be finished. During the time that the fur-
nace is in full action, the cover must be occasionally re-
moved to add more fuel, if any open spaces occur round
the crucibles, also to press the fuel close to the pots.
When the pots are taken out they may be placed on the
anvil or in a sand-bath, and allowed to cool gradually
before they are broken to examine their contents.
In commencing a second assay immediately in the same
furnace, certain precautions must be taken to insure
success. In the first place, all ash and clinker must be
removed from the grate by means of a crooked poker ;
secondly, the fuel must be pressed down firmly ; and
lastly, a layer of fresh combustible must be placed on
the fire, and before that is ignited the crucibles must be
' O
arranged upon the support and the spaces about them be
filled with coke or charcoal, as the case may be, and the
assay be proceeded with as before.
In executing many assays, one after the other, a great
saving of fuel is effected, for the furnace is not allowed
time to cool.
AUXILIARY APPARATUS. Ordinary assay furnaces require
very few instruments ; they are, firstly, pokers or stirring
rods, made of stout bar-iron ; these may be straight, as for
stirring the fuel from the top of the furnace, so as to fill
up cavities formed by uneven combustion; or curved, for
clearing the bars from below from clinker and ashes.
Straight and curved tongs are also required ; for small
crucibles the blacksmith's common forge tongs are the
most suitable ; tongs with semicircular ends (see fig. 27)
are very serviceable for larger crucibles. The tongs a are
particularly adapted for removing large cupels or calcining
F 2
68
FURNACE TOXGS.
dishes from the muffle ; the tongs b and c are used for
lifting heavy crucibles from the wind or blast furnace. In
case the eyes of the operator are weak, it is advisable to
make use of a pair of deep neutral-tint spectacles. Most
of the radiant heat from the interior of a furnace may be
FIG. 27.
cut off by holding before the face a large sheet of window
glass'; or the operator may look at the reflected image in
a looking-glass, instead of looking direct into the furnace
itself. Some assayers recommend the use of masks for the
face and gloves for the hands, but these are not needed.
In cupel furnaces, both curved and straight pokers or
stirring rods are required ; also a curved rod made of
lighter iron, to be used in closing the sliding doors, slightly
moving cupels, &c. The tongs used vary in form (see
fig. 28). a represents very light elastic tongs or pincers
FIG. 28.
employed in the introduction of lead and other matters to
the cupel ; b shows the tongs for holding the scorifier ;
the curved part fits the lower part of the scorifier, and the
upper or single part passes over the upper part of the
scorifier, so that its contents may be emptied into the
proper mould without fear of its slipping from the opera-
INGOT MOULDS.
tor's grasp ; c represents the tongs used in moving cupels ;
they are slightly curved, so that the cupels from the back
part of the muffle may be removed without disturbing those
FIG. 29.
in front. Pig. 29 shows the plan and section of the ingot
mould, into which the contents of the scorifiers are poured ;
it is made of thin sheet iron, and the depressions for the
reception of the fused lead slag^ and ore are hammered out.
FIG. 30.
Fig. 30 is a wrought-iron ladle, in which lead clippings,
&c., are melted, in order to obtain a fair average of a large
Fio 31 , quantity ; and fig. 31 represents
the ingot mould into which the
fused lead or other metal is
poured. Other special apparatus
will be described under the assay
in which they are required.
Furnaces are heated with
anthracite, coke, and charcoal, and sometimes with a
70 FUEL FOR FURNACE.
mixture of the two latter ; coal is very seldom employed,,
and therefore will not be much spoken of; coke is the
principal combustible used in assaying. Calcining fur-
naces ought to be heated with charcoal alone, because
coke employed in small quantities ignites and burns with
too much difficulty. All fuel, contains certain fixed matters
which remain after combustion, and which constitute the
ash. This ash fuses or agglutinates together, and when a
certain quantity is formed, if it be not removed, the fire
will decrease in intensity, and finally die out. As all com-
bustibles do not contain the same amount of ash, they
should be carefully selected ; those containing the least
are to be preferred, in the first place because, weight for
weight, they contain more available fuel, and, secondly,,
because they can be used in a furnace a longer time with-
out the formation of so much clinker. The composition
of the ash likewise merits much attention.
Charcoal contains, in general, from 3 to 4 per cent, of
ash, the chief components of which are lime and potash
carbonates. Certain other matters are also present, as
phosphoric acid, iron oxide, manganese, &c., but these
are in very minute proportions. The ash is not fusible
per se, and, if it does not meet with any substance capable
of combining with it, it passes through the bars as a white
powder ; but when the potash predominates, it exercises
a corrosive action on the bricks with which the furnace is
lined, as also on crucibles, lutes, &c., by the formation of
a fusible potassium silicate, which in course of time runs
down the sides of the furnace, and chokes the bars.
Whenever the ash is in very small proportion to the
charcoal, its fusion is rather useful than otherwise, be-
cause it forms a species of varnish, which, penetrating the
surface of the bricks and lutes, gives them solidity by
binding them together with a cement, forming part of
their substance.
The proportion of ash which coke contains is very
variable; some commercial samples contain from 8 to 10
per cent., while others, made from very pure coal, give
but 2 to 3 per cent. ; so that this fuel also ought to be
CHARCOAL COKE. 71
carefully chosen. The nature of this ash is different from
that of charcoal ; it consists principally of iron oxide and
clay. The former is produced from the pyrites which
coal generally contains. The clay is similar to the car-
bonaceous schists, not very fusible by itself, but neverthe-
less capable of softening, When pure, it forms a slag,
which attacks neither the bricks nor crucibles. This
happens very rarely ; it is more often that iron oxide
predominates, and this by contact with the carbonaceous
matter becomes reduced to the state of protoxide, and is
then not only very fusible, but exercises on all argillaceous
matters a very corrosive action, so that crucibles are very
seriously injured, and the sides of the furnace require
frequent repairs.
Weight for weight, coke and charcoal give out nearly
the same quantity of heat ; but in equal bulks, the former
develops much more heat, because its density is greater.
From this difference in the calorific power of coke and
charcoal, it results that in the same furnace the former
produces a greater degree of heat than the latter ; and at
high temperatures the difference has been proved to be
nearly 10 per cent. In order to account for this, we must
consider, firstly, that in a given space the quantity of heat
produced in a certain time (and, in consequence, the tem-
perature) depends on the amount of fuel burnt, and in-
creases with its weight ; secondly, that combustion takes
place but at the surface of the masses, whatever may be
the nature of the fuel ; from which may be deduced that
the weight of fuel burnt in an unit of time ought to be
exactly proportionate to its density ; and in consequence,
the densest fuels, furnishing the most food for combustion,
ought to give out the greatest heat. But, as for the same
reason they consume a larger proportion of oxygen, they
require, in order to produce the maximum effect, a more
rapid and stronger current of air.
It is clear, from what has been stated concerning the
relative properties of coke and charcoal, that when the
former can be procured of good quality, and especially
when the ash contains but little oxide of iron, it ought to
72 EFFECTS OF FURNACES.
be preferred to charcoal, for assays requiring a high tem-
perature.
This being an important subject, it has been thought
advisable to devote a special chapter to the assay of
fuel.
A very essential condition in obtaining the maximum
effect of a furnace, the importance of which can alone be
appreciated by experience, is to choose pieces of fuel of a
suitable size. If, on the one hand, a shovelful of coke or
charcoal be taken at random, it generally contains the
dust and dirt found in most fuel, and which, by filling the
interstices, prevent the air from passing as required, and
consequently render the combustion slow. On the other
hand, if a furnace be filled with large pieces, considerable
spaces are left between them, so that but a comparatively
small surface is exposed to the action of the atmospheric
oxygen, and a correspondingly small quantity of fuel is
consumed in a given time ; so that the maximum heat
can never be obtained. In order to produce the desired
result, it is necessary that the pieces shall have a certain
mean size, and experience has proved that pieces about
1 inch to H inch diameter produce the best effect.
These may be selected by sifting the coke through two
strong wire sieves, one of which has meshes about 1 J
inch square, and the other about 1 inch square. The
coke which passes through the larger one, but will not
go through the smaller sieve, will be the right size for the
furnace.
THE EFFECTS PRODUCED BY WIND AND BLAST FURNACES.
Assays by the dry way can be made either in wind or
blast furnaces. In either of them, the degree of heat
depends upon the volume of air which passes through
the fuel in the same time ; but, cceteris paribus, large
furnaces produce more heat than small ones, because
comparatively less heat is lost by radiation in the former
than the latter.
In a wind furnace, the maximum of heat is limited by
the size of the chimney, and in a blast furnace by the
dimensions of the bellows ; but by weighting the latter,
FOCUS OF MAXIMUM TEMPERATURE. 73
more or less, the force of the blast and, in consequence, the
temperature can be increased, to a considerable extent. In
this respect blast have the advantage over wind furnaces.
In the latter, the draught increases in proportion as
the heat becomes more intense in the furnace, so that
the temperature producible increases progressively. In a
blast furnace, the bellows can be weighted and worked
as heavily as possible at once, and, by opening all the
apertures for receiving air, the maximum temperature can
be produced more rapidly than in a wind furnace ; but
this is of little use, because, as heat passes very slowly
through the substance of a crucible, when the object is
to fuse its contents it must be heated gradually, so as
to avoid running the risk of softening the crucible before
its contents are acted upon, or even scarcely made warm.
Wind furnaces are, however, much more serviceable
and economical than blast, because they work themselves,
and do not require the service of a man to attend to the
bellows. A blast furnace is used in a laboratory, in cer-
tain cases ; for instance, when a single crucible has to be
submitted to an intense heat, and when the furnace is
small, and the bellows large, in which case the operation
resembles a blow-pipe assay.
In whatever manner the air is introduced into any
kind of furnace, whether wind or blast, it is evident that
the quantity of heat developed in equal-sized furnaces
depends upon the quantity of air introduced in the same
time ; but the degree of temperature is not the same in
different parts of the furnace, and the distribution of heat
varies according to the manner in which the air is intro-
duced into the midst of the fuel. The side over which
the air passes is kept cold by the current, on which
account fire-bars last a long time without becoming oxi-
dised, but the heat rapidly augments up to a certain
distance from the bars, at which place it arrives at its
maximum ; above that it diminishes rapidly, because the
air is nearly deprived of its oxygen. Experiment has
proved that this maximum is about 2J to 3 inches above
the bars or tuyeres.
74 OIL AND GAS FURNACES.
In common wind furnaces the air enters through the
spaces between the horizontal bars which form the bottom
of the furnace, and the crucibles are placed on a stand
which rests on these bars. By this means the lower and
centre part of the crucibles, in which parts the matter to
be fused is placed, are exactly situated in the maximum
of heat ; but the stand being constantly kept cold, by the
contact of a current of air, establishes a continual draining
or carrying away of heat from the interior of the crucibles
outwards, so that the substance submitted to assay can
only arrive at the maximum temperature after a length of
time, and the maximum then is always inferior to that in
the mass of fuel. It is on this account that assays in a
blast or wind furnace generally occupy from one hour to
two hours. The author has found that the time may be
reduced to half that just stated, if a good solid foundation
of fuel be made, and the crucible placed on that, and well
surrounded by coke, constantly kept close to the pot and
the sides of the furnace ; in this manner the cooling effect
of the stand is removed, and the consequent maximum
effect of the furnace is produced, but then there is danger
of the supporting fuel being burnt away from the crucible
and the latter getting upset.
OIL AND GAS BLAST FURNACES.
It sometimes happens that metallurgists and assay era
have occasion to melt metals at a white heat, but do not
wish to heat a large furnace for the purpose. In these
cases either the gas or oil furnaces now to be described
will prove very useful.
OIL FURNACES. Mr. Charles Griffin, the well-known
chemical instrument maker, has described an oil lamp,
which is not only as powerful in action as the best gas
furnaces, but almost rivals them in handiness and
economy.
DESCRIPTION OF THE APPARATUS. The oil-lamp furnace
is represented in perspective by fig. 32, and in section by
fis. 33. It consists of a wick-holder, an oil-reservoir, and
GRIFFINS OIL BLAST FURNACE.
75
a fire-clay furnace ; to these must be added a blowing-
machine for the supply of atmospheric air.
The oil-reservoir is represented at a, fig. 32 ; it is
made of japanned tin-plate, mounted on iron legs, and
fitted with a brass stop-cock and delivery-tube. Its
capacity is a little more than a quart. The wick-holder
is represented at b, and the upper surface of it by the
separate figure <?, fig. 34. The wick-holder and the oil-
reservoir are consequently detached, d is a tube which
brings oil from the funnel 0, and / is a tube to be placed
FIG. 32.
in connection with the blowing-apparatus. The wick-
holder contains three concentric wicks, placed round the
multiple blow-pipe c, which is in communication with the
blowing-tube.
The crucible furnace consists of the following parts y
shown in figs. 32 and 33 : g is an iron tripod ; h is a flue
for collecting and directing the flame. The flue is of such
a width that when the wick-holder, &, is pushed up into
it until the top of the wick is level with the top of the
clay cone, there remains a clear air-space of about |- inch
all round between the wick-holder and the cylindrical
walls of the flue, i represents a fire-clay grate having
76
DETAILS OF THE MANAGEMENT OF THE
three tongues, shown by i (fig. 34), on its upper surface.
These tongues support the crucible, without stopping the
rising flame, k is a fire-clay cylinder which rests upon
the grate i, and incloses the crucible, forming, in fact, the
body of the furnace. Of this piece there are three sizes :
the smallest is of 3 inches bore, and works with crucibles
that do not exceed 2f inches diameter ; a middle size,
4 inches bore, for crucibles not exceeding 3| inches
diameter; the largest size, 5 inches bore, for crucibles
not exceeding 4f inches diameter. This piece, being
heavy, is provided with handles, as represented at p,
FIG. 33.
FIG. 34.
fig. 34. The walls of the cylinders are from 1 inch to
1-| inch thick. I is a fiat plate of fire-clay with a hole in
the centre, used to cover the cylinder k, so as to act like
a reverberatory dome ; m is a cover which prevents loss
of heat from the crucible by radiation, but gives egress to
the gaseous products of the combustion of the oil ; n is an
extinguisher to put over the wick-holder when an opera-
tion is ended ; and o is a support for the wick-holder. No
chimney is required.
MANAGEMENT OF THE OIL-LAMP FURNACE. The apparatus
is to be arranged for use as it is represented by fig. 33.
The cylinder, , is to be selected to fit the crucible, and
OIL BLAST FURNACE. 77
the crucible of a size to suit the quantity of metal that is
to be melted : 1 Ib. of iron requires the smallest of the
three cylinders, described above ; 1^ Ib. the middle size ;
5 Ibs. the largest size. The air-way between the crucible
and the inner walls of the cylinder should never exceed
J inch nor be less than -|- inch.
The cotton wicks must be clean, and be trimmed a
little below the level of the blowpipe c. If properly
managed, they do not readily burn away, but can be used
for several fusions. The reservoir should be filled with
oil for each operation. The proper sort of oil for use is
the more volatile kind of mineral oil, of the specific
gravity of *750, which is now easily procurable at about
three shillings per gallon. The variety known by the
commercial name of turpenzine answers well. The com-
bustion of a quart of this oil, costing ninepence, gives
heat sufficient to melt 5 Ibs. of cast iron. Probably the
lighter kinds of paraffin oil may be suitable. Liquids of
the alcoholic class, spirits of wine, and pyroxylic spirit
can be -used ; but they are less effective and more expen-
sive than turpenzine. Care must be taken not to spill
the oil on the table or floor, and not to decant it carelessly
in the neighbourhood of a light, because atmospheric air
strongly charged with the vapour of these light oils is
explosive. -When the oil is burnt in the furnace in the
manner described below, there is no danger. During an
operation, a wooden screen, as represented by the dotted
lines in fig. 32, should be placed between the oil-reservoir
and the furnace, to prevent the vaporisation of the oil by
radiant heat. As the wick-holder b, and the supply pipe
d) contain only about one fluid ounce of oil, the oil must
run continuously during a fusion, from the reservoir #,
into the funnel 0, in order that the cotton may be always
flooded. The success of the fusion depends upon the due
supply of oil, to which point the operator must pay atten-
tion. At the commencement of a fusion, the oil must be
run from the reservoir until the surface of the oil in the
funnel has a diameter of about an inch. The wicks will
then be flooded, and a lidit may be applied, and a gentle
78 THE OIL BLAST FURNACE.
blast of air then turned on. The oil immediately sinks in
the funnel, and the stop-cock must be opened and so regu-
lated as to keep the oil barely visible at the bottom of the
funnel. If too much oil is supplied it immediately rises
in the funnel, and simultaneously overflows the wick-
holder. Too much vapour is then thrown into the fur-
nace, and the heat is immediately lowered, especially at
the beginning of an operation, before the fire-clay portions
of the furnace are well heated. If, on the contrary, too
little oil is supplied, the wicks burn, and the operation is
spoilt. The demand of the wick-holder for oil depends
upon the condition of the furnace and the character of the
fusion in progress. When the lamp is newly lighted and
the furnace cold, the oil should be passed slowly in distinct
drops ; but as the furnace becomes heated the rapidity of
the supply of drops should be increased ; and finally, when
the furnace is at a white heat, the oil should be supplied
in a thin continuous stream. When the fusion to be
effected is that of only a small quantity of metal, such as
1 Ib. of iron, a rapid supply of drops of oil is sufficient
even to the close of the operation. At that rate the
burner consumes about 1 J pint of oil in an hour. When
the fusion to be effected is that of 4 Ibs. or 5 Ibs. of iron
and the large furnace is in action and has been brought
to a white heat, the supply of oil must, as stated above, be
in a thin continuous stream, and the operation will then
consume two pints of oil in an hour. And here it requires
remark that, with that continuous supply, when the fur-
nace is large and is at a white heat, the oil does not rise
in the funnel, being instantaneously converted into gas at
the mouth of the burner, and thrown up in that state into
the furnace for combustion. The operation, indeed, con-
sists at that point of a rapid distillation of oil-gas, which
is immediately burnt, in the presence of air supplied at a
suitable pressure by a dozen blowpipes, in effective contact
with the crucible to be heated.
The flame produced in this furnace is as clear as that
produced by an explosive mixture of air and coal-gas. It
is perfectly free from smoke, and the consumed vapours
POWER OF THE FURNACE. 79
which occasionally escape with gaseous products of a com-
bustion are even less unpleasant to smell and to breathe
in than are those which are usually disengaged by a blast
gas furnace, or by an ordinary lamp, fed with pyroxylic
spirit.
The contents of a crucible under ignition in this furnace
can at any moment be readily examined, it being only
necessary to remove the pieces I and m with tongs, and
to lift the cover of the crucible, during which the action
of the furnace need not be interrupted.
When the operation is finished, the blast is stopped,
the stop-cock is turned off, the oil-reservoir is removed,
the wick-holder is lowered on the support 0, withdrawn
from the furnace, and covered with the extinguisher n.
The quantity of oil which then remains in the lamp is about
one fluid ounce.
POWER OF THE OIL-LAMP FURNACE. The furnace being
cold when an operation is commenced, it will melt 1 Ib. of
cast iron in 25 minutes, 1^ Ib. in 30 minutes, 4 Ibs. in 45
minutes, and 5 Ibs. in 60 minutes. When the furnace is
hot, such fusions can be effected in much less time ; for
example, 1 Ib. of iron in 15 minutes. It need scarcely be
added that small quantities of gold, silver, copper, brass,
German silver, &c., can be melted with great ease, and
that all the metallurgical and chemical processes that are
commonly effected in platinum and porcelain crucibles can
be promptly accomplished in the smallest cylinder of this
furnace ; and in the case of platinum vessels, with this
special advantage, that the oil-gas is free from those sul-
phurous compounds the presence of which in coal-gas fre-
quently causes damage to the crucibles.
REQUISITE BLOWING POWER. The size of the blowing
machine required to develop the fusing power of this oil-
lamp furnace depends upon the amount of heat required
or the weight of metal that is to be fused. For ordin-
ary chemical operations with platinum and porcelain
crucibles, and even for the fusion of 1 Ib. of cast iron in
clay or plumbago crucibles, a blowing power equal to that
of a glass-blower's table is sufficient, provided the blast it
80 GAS FURNACE.
gives is uniform and constant. But the fusion of masses
of iron weighing 4 or 5 Ibs. can be effected by the gas
which this oil-lamp is capable of supplying, provided a
sufficiently powerful blowing machine supplies the requi-
site quantity of air. When more than a quart of oil is to
be rapidly distilled into gas, and the whole of that gas is
to be burned with oxygen, it is evident that effective work
demands a large and prompt supply of air.
As in all practical matters of this sort the cost is a
main question, it may be useful to state that the price of
this apparatus complete, without the blowing machine, but
including every other portion necessary for heating cruci-
bles up to the size sufficient to fuse 1 Ib. of cast iron, is one
guinea ; and that with the extra furnace-pieces for cruci
bles suitable for 5 Ibs. of iron, or any intermediate quantity,
the cost is one guinea and a half.
Mr. Griffin has described before the Chemical Society
an improved gas furnace for chemical operations at a white
heat without the aid of a blowing machine, and a new
method of supporting crucibles in gas furnaces. The fol-
lowing extracts from his paper are taken from the Journal
of the Chemical Society :
4 On a former occasion I introduced to the notice of
the Chemical Society a gas furnace for operations at a
white heat in crucibles, or a copper-melting heat in muffles.
A detailed description of that furnace is given in the
Journal of the Society for August 1870. The crucibles
are either suspended in a pierced plumbago cylinder or
supported on a trivet grate, both of which are liable
to break when white hot, and therefore cause trouble
and expense. Crucibles vary so much in form and size
that they are often not suspended from these cylinders
exactly in the focus of the heating power. Trivet grates
have the disadvantage that they interfere with the direct
action of the flame upon the crucible, and if made slightly
they break when heated to whiteness. I desire now to
place before you a new form of burner by which these
defects are remedied. In the new burner the circle of gas-
jets is enlarged so as to leave a space round the central
GAS FUKNACE. 81
jet. An atmopyre similar to those used in Hofmann's
combustion furnace, but of greater bulk and strength, is
dropped over this central jet, and forms a solid support
for the crucible. This support does not readily break,
but, should an accident happen, it can be replaced at the
cost of a few pence. It brings the bottom of the crucible
exactly into the focus of heat, and itself supplies a portion
of the heating power of the burner. It also enables one to
use any crucible at hand, independent of its form or size.
A strong lateral arm cast on the body of the burner sup-
ports an upright iron rod, which carries the chimney of
the furnace. By prolonging the legs of the burner up-
wards they are made to carry the clay furnace, and thus,
by doing away with a stool or other support, the con-
struction is simplified and the cost lessened. A plumbago
cylinder, to deflect the flame and entrap the heat, is placed
round the crucible, and is covered with an ordinary
crucible cover, by removing which the crucible can be
inspected. These fittings, however, present nothing new,
being adapted from Griffin's gas blast furnace, which was
introduced sixteen years ago. Access to the crucible in
the furnace is gained by turning aside the chimney and
lifting the top plate of the furnace, which is provided with
handles for this purpose. These handles do not become
very hot, even when the furnace is at a white heat. The
power of these new burners is very remarkable, one of
small size consuming only 20 feet of gas per hour, and
having a chimney 4 feet high, being capable of fusing half
a pound of cast iron in 35 minutes from the time of lighting
the gas ; or of melting gold, silver, or copper in crucibles
placed within a muffle measuring 5 inches long by 3 wide.
If a chimney 6 feet high be employed, cast iron can be
melted in crucibles placed within the muffle. A burner of
larger size, consuming 40 feet of gas per hour, will melt
cast iron in crucibles placed within a large muffle measuring
8 inches long by 4 inches wide. In the crucible furnace
it will melt 1 Ib. of cast-iron in 35 minutes, 2 Ibs. in
45 minutes, 3 Ibs. in 55 minutes, and 4 Ibs. in 65
minutes, from the time of lighting the gas. It is thus
G
82
GAS FURNACE.
seen that, when a white heat has been once obtained,
10 minutes' time is required for the fusion of every
additional pound of iron. These results, attainable with
certainty and rapidity, are, I believe, the highest that have
hitherto been placed at the command of the chemist. As in
my former furnace, the proper admixture of gas and air
is judged of from the colour and quantity of flame which
passes up the chimney. To enable the operator to see
this flame, three small holes are bored in the chimney.
FIG. 35.
FIG. 36.
The flame is not seen at the upper hole, unless the supply
of gas is too large, but it is always visible at both the
lower holes.'
In the above figure the muffle is provided with a small
draught flue, having a regulating cap on its upper end.
In the small furnaces this is omitted, and the muffle is
slotted in the usual manner. The cover of the furnace is
now made without the zigzag opening in the roof. The
burner of the muffle furnace is the same as that used in
the crucible furnace, fig. 35.
GAS FURNACE. 83
Skittle pots up to 8 inches can be used in these furnaces
for collecting and burning waste with fluxes, &c., and in
much lees time than is required by a coke fire, An 8-inch
pot can be worked in half an hour from lighting the gas.
Two of the outer cylinders are used, placed one on the
top of the other.
FIG. 37.
With a 4-foot flue the muffle furnace melts gold, silver,
and copper ; with a 6-foot flue it melts cast iron placed i
crucibles within the muffle. The consumption of gas is 20
cubic feet per hour.
84 UNIVERSAL GAS FURNACE.
The Distillation of Pure Zinc, per descenswn, can be
performed in one of these gas furnaces by passing a tube
from the top of the crucible downwards through the
burner to the table.
Fig. 37 shows one of Griffin's extra large gas fur-
naces, capable of raising a No. 12 plumbago pot, measur-
ing 8 inches high by 6 inches wide, to a white heat. The
cover of this furnace is let into the body, which rises higher
.than in the smaller patterns, and from which the flue passes
off laterally to a standing flue or other house chimney.
UNIVERSAL GAS FURNACE. Mr. Thomas Fletcher has
devised what he terms a universal gas furnace ; it works
without blast for crucible operations, up to a clear white
heat.
The specialty of this furnace is the burner. It is as
simple and easy to use as an ordinary Bunsen's burner, but
the flame is solid to the centre, unlike the flame of every
heating burner which has previously been made. The
open flame will readily fuse a coil of thick copper wire,
and to make a crucible furnace it simply requires a support
for the crucible, and a fire-clay jacket to prevent radia-
tion, as the flame is in itself perfect, and requires no
blowing or attention in any way. The furnace is so small
and light that it can be used on the work-bench, and put
away on a shelf, and can be used on a sitting-room table
without the slightest dirt or inconvenience. The body of
the furnace is only 6 inches in diameter.
THE SINGLE-JACKETED ARRANGEMENT is capable of melting
5 or 6 oz. of gold in 15 minutes with a 10-inch chimney,
or in 10 minutes with a 2-ft. chimney. With the ladle-
holder it will melt 8 Ibs. of zinc in 15 minutes, without
chimney, in an ordinary iron ladle, and lead, tin, &c., in
a proportionately shorter time.
THE DOUBLE-JACKETED CHEMICAL FURNACE, having the same
burner as above, and requiring no more gas, will melt 3
or 4 oz. of c.ast iron in 35 minutes, if used with a 3-ft.
chimney, or, if with a longer chimney, in a proportion-
ately shorter time, and will give any required tempera-
ture in proportion to the length of chimney used, provided
FLETCHER'S GAS FURNACE.
the gas is turned on sufficiently just to cover the crucible
with flame when the chimney is in its place.
' In case of the fusion of a crucible, or its penetration
by fluxes when the furnace is used for extremely high
temperatures, no damage can be done to the furnace,
except perhaps the destruction of a few of the burner
tubes, which can be replaced at a trifling expense. No
part of the furnace is liable to injury with constant
use.
The lower part of the burner is a chamber 6 in. x
3 in., open at the bottom, in which the gas is partially
mixed with air. This mixture is conducted to the top of
the burner through a mass of fine tubes,. 55 in number,
with an arrangement to supply between each exactly the
amount of air necessary to consume it instantly. A flame
produced by this means, consuming 20 feet of gas per
hour, is only about 2 inches high, and almost colourless.
The whole of the available heat is generated in the proper-
place, viz. below the object to be heated, which, there-
fore, is not also cooled by the passage of unburnt gas and
air. The flame is very similar in appearance to a mass of
blowpipe jets, and has the same heating power. The point
of greatest heat commences, as with a blowpipe, at the
point of the blue cones, about ^ in. or f in. above the
tubes.
The Double-jacketed Chemical Furnace is made to
take crucibles not exceeding in size No. 00 of the Plum-
bago Crucible Co., which are the most convenient for
melting quantities not exceeding 5 or 6 oz. of gold, &c.
The gas supply tap and pipe must be large and clear,
so as to give as great a pressure of gas as possible at the
burner nozzle, although the actual consumption of gas is
small. The indiarubber tubing used must of necessity
be perfectly smooth inside. The tubing made on wire,
whether the wire is removed or not, will not work these
burners satisfactorily. The gas supply specified is re-
quired to work each furnace at its full power, and the
flame must be visible in the chimney. If the gas supply
is deficient, the furnaces can be worked at a lower heat
o FLETCHER'S GAS FURNACE.
by partially closing the top of the chimney until the flame
becomes visible, or by working without the chimney. If
the burner plate becomes red hot, it is a sign that the
gas supply is deficient. If the supply of gas is not suffi-
cient, it will be found necessary to examine the tap which
supplies the gas, many of which, as supplied by gasfitters,
are exceedingly faulty, and partially stopped up. If this
be the case, the tap should be replaced by a better one.
The furnace requires supply from a f or T 7 in. pipe. If
a larger crucible be used than the one supplied, an extra
fire-clay ring must, be obtained of the same proportionate
width, and half an inch higher than the top of the cru-
cible.
Where the gas supply is more than necessary, a longer
chimney may be used with advantage, but, as the furnace
itself is too small and light to form a steady support for a
longer chimney, it will be necessary to suspend the upper
part with a wire, or to support it with a bracket from
the wall. The furnace requires the following supply
of gas : Without chimney, 18 cubic feet ; with 10-inch
do., 22 cubic ft. ; with 36-inch do., 28 to 30 cubic ft. per
hour.
When a white heat is required, the crucible must be
covered with an inner perforated plumbago dome, and
which forms the inner jacket, with the perforations in
such a position that the crucible can be seen through the
hole in the lower part of the chimney ; a chimney not
less than 3 feet high must be used. If a greater heating
power is required, a longer chimney will enable it to be
obtained in the same or a shorter time, but under all
circumstances the flame must just cover the crucible to
obtain proper results.
The addition of a small muffle for high temperatures
makes this furnace complete for all purposes. The clear
working space, which is equally heated in every part,
measures 2^ x 2f x 2J in. The muffle is not in contact
with the outer casing in any part, and therefore all parts
are equally heated. The temperature obtained depends
on the length of chimney used, provided the gas supply is
FLETCHER'S GAS FURNACE. 87
sufficient to cover the muffle with flame when the chimney
is on. For silver assays about 4 feet of chimney should
be used, which gives the melting point of fine silver in 16
minutes from the time the gas is lighted; the same tem-
perature being always obtained in exactly the same time,
within a few seconds ; and after the first trial the opera-
tion does not require watching. With this furnace there
will be found no variation in a number of assays done at
different times.
A chimney of about 6 or 8 ft. gives nearly a clear
white heat in 30 minutes, and for special operations
requiring very high temperatures a longer chimney may
be used as required. A chimney or stove pipe 10 or 12
ft. long may be used as a fixture, and the draught par-
tially stopped with a damper or slide when lower tempe-
ratures are required, the gas being turned down in pro-
portion ; the guide for proper adjustment being, that
under all circumstances the flame must just cover the
muffle, but must not extend into the chimney so as to
make it red hot.
It will not be found necessary in practice to partially
close the chimney for lower temperatures, as the same
effect may be produced equally well by simply turning the
gas lower ; but it is more economical to partially stop the
draught, so as to prevent excess of cold air being drawn
in. When the flame covers the muffle, the gas is doing
its extreme duty under the most favourable circumstances,
without waste. The same rule applies also to the crucible
arrangement. Particles of flux should not be allowed to
fall on the fire-clay casing, where the parts touch each
other ; and the power of the furnace should not be urged
too far by the use of very long chimneys, as there is
danger of the fusion of the fire-clay parts together so that
they cannot be separated. Fire-clay fittings, as a rule,
cannot be safely used for temperatures much exceeding
the fusing point of cast iron.
An excellent gas assay furnace, slightly modified from
one of Mr. Fletcher's, has been described in the Chemical
News by Mr. Walter Lee Brown, and is shown in fig. 38.
88 FLETCHER'S GAS FURNACE.
As shown, the form is almost that of the reverberatory
furnace, the movable bricks when in place forming the
roof. From another point of view, it may be described
as the muffle of an ordinary furnace, but having the flame
as well as the heat inside. The exterior dimensions are
as follows : 20 inches long, 7 inches wide, and 5i inches
deep. The nozzle of the burner is connected with a
| inch tap by means of a full ^ inch rubber tube. A
3-inch stove-pipe, tightly fitting, is connected with a flue.
In the interior, upon the floor, rest four little wedge-
shaped pieces of fire-clay which are movable, and upon
FIG. 38.
them rests a false floor, also movable. The latter (not
shown in the cut) corresponds to the mutfle bottom of an
ordinary furnace, and upon it is done all the work. It
is 3^ inches wide by 7^ inches long, and will accommo-
date three 2f inch or four 2J inch scorifiers, or eighteen
1 J inch cupels at once ; but like any other furnace, it is
better not to crowd it.
The manner of operation is simple. The covering
bricks are removed, the milled wheel at the gas entrance
to the burner is turned back so as to allow a full flow of
gas, the handle at the supply tap turned full on, the
gas lighted, and the bricks put into place. The flame,
FLETCHER'S GAS FUKNACE. 89
when the full amount of gas is used, will be highly car-
buretted, and of course strongly reducing, but intensely hot.
In from fifteen to twenty minutes the interior will be hot
enough for work. The bricks are removed, the charged
scorifiers are placed within, the bricks set back, the excess
of gas turned off at the milled wheel, and the furnace kept
closed till the charges are melted. After this has been
effected, the bricks are separated or slid aside more or
less to admit of air for scorification. Proper regulation is
made by the milled wheel, by which the gas may be
turned partially off as required, always leaving the supply
tap fully open. The action of the air is also controlled by
the damper in the pipe leading to the chimney.
For cupelling, the gas supply is turned down more than
in scorifying. The time of performing either scorification
or cupellation varies according to the nature of the ore
and other circumstances, but is about the same as in the
use of a coke furnace.
This furnace does well for small crucible fusions, by
removing the false floor and its supports.
The advantages of this furnace are many : convenience
of operating whereby the assayer sees every step and
stage of the operation, and so can tell when and where to
change or improve ; perfect control of the source of heat,
so that a higher or lower temperature, a reducing or
oxidising effect, may be produced in an instant ; entire
noiselessnesss, in which characteristics it is the superior
of all blast assay furnaces ; saving of time, which in
furnaces employing coke, charcoal, or coal is spent in
4 bedding down,' feeding, &c. ; freedom from the annoy-
ances of dust and ashes ; absence of waste, for when the
work is performed the gas is at once shut off; comfort in
manipulation, for it does not heat up a room as do most
furnaces (quite a desideratum in the summer time) ;
finally, its remaining qualifications, which need not be
dwelt upon, are simplicity of construction, durability, and
portability.
The consumption of gas is not far from 30 cubic feet
per hour.
90 FLETCHER'S GAS FURNACE.
Fig. 39 shows one of Fletcher's Eeverberatory Gas Fur-
naces for crucibles, muffles, cupels, &c.
One of these furnaces will do most of the general work
of an ordinary laboratory. They work perfectly with
chimney draught to a bright red about the fusing point
of fine copper and fine silver. With a blast they will
work up to the fusing point of cast iron. The furnaces can
be made to take either two muffles at once, a number of
crucibles, trays of cupels, or one muffle and crucibles or
cupels at the same time.
The opening may be either at the side or the top, the
furnace working either way equally well. The burner is
FIG. 39.
at one end, out of the way of injury in case of accident
to a crucible. Crucibles, cupels, &c., stand on the solid
bottom of the furnace, perfectly steady and firm. When
a blast burner is used a clay collar fits into the larger
opening necessary for a draught burner, and the instruc-
tions for both draught and blast for the ordinary furnaces
apply" equally well to this, the burners being identical in
principal with those of the previous patterns.
When used with the draught burner the blue cones
of flames must be clearly seen on the burner, or if they
disappear the gas supply must be increased, or the slide
over the burner air tube closed until they reappear. In
the latter case the furnace works with a smaller gas supply
at a lower temperature, and by closing this slide and
GORES GAS FURNACE.
91
reducing the gas supply any temperature required can
be obtained. If the adjustment of gas and air is neglected
the burner grid becomes red hot and is quickly rendered
useless. The grids will last for years if properly used.
A good
muffle fur-
nace to work with gas
has long been a deside-
ratum in the assay labo-
ratory. Mr. Fletcher
has supplied this want
in the furnace shown in
fig. 40.
The gas required for
this furnace is 70 cubic
feet per hour. There
FIG. 40.
be
-| inch
clear
the
gas
must
bore through
pipes and tap.
Fig. 41 shows one of
Fletcher's Draught Crucible Furnaces. This will melt
brass, silver, copper, and gold, but is not suitable for cast
FIG. 41.
<* It takes about 25 cubic feet of gas per hour, and
requires a i inch pipe and tap. The largest size crucible
iron.
92
GORES GAS FURNACE.
it will take is 4x3^ inches, and it will melt 6 Ibs. of
brass.
Mr. Fletcher has recently brought out a new melting
arrangement for melting up to 3 oz. of gold or silver,
FIG. 42. rapidly, without the use
of a furnace.
The figure (42). is
slightly under half-size.
In this arrangement the
two parts of the ingot-
inould slide on each other ,
to enable ingots of any
width to be cast, And
the blowpipe is part of
the rockingr-stand. Con-
nect the blower to the
upper tube, and the gas
to the lower. When the
metal is melted in the
shallow crucible of com-
pressed charcoal, tilt the
whole apparatus over so as to fill the ingot-mould. A
sound ingot can be obtained in about 2 minutes. Very
bulky scrap should be run into a mass in one of the
moulded carbon blocks before being placed in the cru-
cible. No flux must be used with the charcoal crucibles.
With a larger- sized melting arrangement similar to the
above, as much as 14 ounces of fine silver, or 20 ounces
18-carat gold, can be melted and cast in an ingot in 5 or 6
minutes. This size requires a ^ inch gas supply and a foot-
blower.
Mr. Fletcher has also devised an injector gas furnace
(fig. 43) for metallurgists, jewellers, chemists, iron, brass,
and nickel founders, manufacturers of artificial gems, and
other purposes where an ordinary furnace is useless or
unreliable.
It has been found that in working at extremely high
temperatures, the ring which holds the gauze is liable
to be fused. To prevent this, a new burner has been
GORE'S GAS FURNACE. 90
designed, in which the ring is entirely dispensed with, and
the gauze cap is pushed up from the back of the burner
against a small shoulder inside the nozzle of the burner.
The burner is in one
casting, and therefore FlG - 43 -
there is no tendency
for the nozzle to get-
hot, as in the former
pattern. See that the
gauze is pushed up AIR CHECK
from behind to within AIR
about J inch of the
nozzle. The power
and speed of working
are practically without limit, depending only on the gas
and air supply, and are under perfect control. With
^ inch gas pipe and the smallest foot-blower, the small
furnace will melt .a crucible full of cast-iron scrap in 7
minutes, and steel in 12 minutes, starting with all cold.
To adjust a new furnace to its highest power, put
the nozzle of the burner tight up against the hole in
the side of the casing, turn on the full gas supply,
with the air-way full open. If the flame comes out of
the lid about 2 inches, the adjustment is right. If the
flame is longer, enlarge the hole in the air-jet until the
proper flame is obtained, or reduce the gas- supply ; if
smaller, or not visible, turn the air-check until the flame
appears.
Keep all fluxes away from the furnace jacket. Before
stopping the blower, draw the burner back from the
hole. Commence blowing before the lid is put on the
furnace.
The old pattern blower is liable to pick up dirt from
the floor, throwing it against the gauze of the burner, and
stopping the proper working of the furnace until cleared
away. A thin layer of silver sand on the bottom will
prevent crucibles adhering when at a white or blue heat.
Crucibles must be heated very slowly the first time they
are used, unless of the ' Salamander ' brand.
94
GORE S GAS FURNACE.
In cases where gas cannot be obtained, Mr. Fletcher has
devised a simple furnace (fig. 44) for high temperatures,
working with either
FIG. 44. fo . .,
gas or spirit-petroleum
without alteration, and
giving perfect results
with either fuel. This
furnace is supplied
with a small, simple,
and perfectly safe ar-
rangement for burning
the vapour of gasoline
or benzoline, giving a
power and efficiency
fully equal to that
which can be obtained
by a larger gas supply.
The apparatus is in
every way as simple as- when gas is used, requiring no more
trouble or attention. '
It equals a gas furnace in every respect, and, in
addition, gives a heat of absolute purity, fitting it for the
most delicate chemical operations where gas cannot be
used owing to the presence of sulphur and other matters.
The ordinary pattern of injector furnace is used in
precisely the same way as with gas, the only difference
being that a branch pipe is taken out of the air supply
and connected to the lower tap A on the generator, and a
tube is carried from the upper tap B to the side tube of
the injector burner marked 'gas.' The quantity of vapour
required is adjusted by the lower tap A when the furnace
is working, and the flame must be just visible at the hole
in the lid, exactly as when gas is used, the instructions
being precisely the same 'for both fuels.
To charge the generator, pour benzoline or gasoline in
the top hole until it overflows at the small tap C in the
side, replace the cork firmly, and close the overflow tap.
It will then work for about 10 to 12 hours at the full .
power of the furnace. ' ' ! o 1
GORES GAS FURNACE. 95
Benzoline varies much in quality ; it must, when a
few drops are poured on a plate or the hand, evaporate
quickly and completely, leaving no greasy stain, and if
good will produce more vapour than the furnace can burn
at its maximum power. All the tubing used must be
perfectly smooth inside, or the power of the furnace will
be greatly reduced.
At the conclusion of an operation, close both taps on
the generator. It can then be left for any length of time
ready for instant use. For ordinary meltings, the gene-
rator can be used about thirty or forty times without
refilling.
The No. 3 size will refine and perfectly fuse 6 Ibs. of
chemically pure nickel so that it can be poured clean,
using an open crucible, a feat beyond the capabilities of
any other known furnace.
Benzoline often contains heavy oils. If the generator
works badly, empty it and refill with fresh.
G. Gore, Esq., F.E.S., has devised a gas furnace which
will fuse cast iron, &c., and still allow the melted sub-
stances to be perfectly accessible to be manipulated upon
for a continuous and lengthened period of time, without
contact with impurities or with the atmosphere, and with-
out lowering their temperature sufficient to cause them
to solidify. These conditions Mr. Gore has obtained by
means of ordinary coal-gas and atmospheric air, without
the use of a bellows or lofty chimney, or of regenerators
or valves requiring frequent attention. The arrangement
is as follows : A (figs. 45 and 46) is a cylinder of fire-clay
about 9 inches high and 6 inches diameter, open at both
ends, with a hole in its side near the bottom to lead into
the chimney ; it is covered by a movable plate of fire-clay,
B, with a hole in its centre for the introduction or removal
of the crucible, c. ; this hole is closed by a perforated plug
of clay, C, for access to the contents of the crucible, and
that again is closed by another clay stopper D ; E is a
chimney of sheet iron about 5 or 6 feet high, kept
upright by a ring of iron F attached to the top of the
furnace. The fire-clay cylinder is enclosed in a sheet of
90
GORES GAS FURNACE.
iron casing with a bottom of iron, to which are fixed
three iron legs G. An iron tube H, with a prolongation
/, supports by means of the screw / the burner K and its
tube jL, which is open at both ends. Gas is supplied to
the burner by means of the tap M, which has a small
index A 7 attached to it for assistance in adjusting the gas.
Inside the large cylinder is another fire-clay cylinder or
cupola 0, with open ends, and with three projections of
FIG. 45.
FIG. 46.
fire-clay P, for supporting the crucible Q ; it is kept
steady by means of three clay wedges R ; S is an air-valve
for closing the bottom of the tube L. The gas-burner is
a thin metal cylinder, deeply corrugated at its upper end,
with the corrugations diminishing to nothing at its lower
end, as shown in the engravings. The action of this
furnace is as follows : Gas is admitted to the open tube L
by the tap M ; it there mixes with air to form a nearly
combustible mixture, which ascends through the burner,
GORES GAS FURNACE. 97
and burns in the clay cylinder 0, being supplied with the
remainder of air necessary for combustion through the
tube H to the outer surface of the flame ; the products of
combustion pass up through the cylinder 0, and then
downwards outside of it to the chimney, the point of
greatest heat being at Q.
It is important in using this furnace that the burner be
placed quite in the centre of the bottom of the tube ;
also that a crucible of not too large nor too small dimen-
sions be selected. The most suitable way of supporting a
smaller crucible is by placing it in a larger one that has
had its upper part broken off. If desirable, a little clay
luting may be placed round the top edge of the iron
casing to exclude air entering between it and the cylinder ;
also a little thin clay luting upon the part of the bottom of
the furnace where the inner cylinder rests.
In lighting the furnace, the plugs C and D are removed,
a light held inside the opening, and the gas turned on full.
Should the flame blow down to the bottom of the tube L
on lighting (which, however, rarely occurs unless . the
furnace is already hot), the gas must be turned off, and
the bottom of L momentarily closed whilst lighting the gas
as before. Should the flame not burn down to the burner,
but only burn to the orifice in the clay plate B, it must at
once be extinguished and relighted, otherwise some of the
gaseous mixture will pass into the chimney unburned, and
subsequently ignite and cause an explosion. A large
flame now issues from the top orifice, and is white if too
much gas is on. and chiefly violet or red with the proper
quantity ; it should now be coarsely adjusted until these
appearances are represented. The annular plug C should
now be inserted, which will compel it to pass downwards
to the chimney, and as soon as the small remaining flame
now issuing disappears, or nearly disappears, as it will in
a few seconds, the smaller stopper D should also be
inserted. In lieu of this, the large flame may be deflected
against the chimney by means of a piece of sheet iron until
it withdraws inwards as before mentioned ; the two plugs
may then be reinserted. The gas tap may now be partly
H
98 -GORES GAS FURNACE.
adjusted. The crucible should be placed in the furnace
after the act of lighting the gas, but not immediately after
if the furnace is cold, or explosions may occur by un-
burned gaseous mixture passing the crucible into the
chimney, and igniting afterwards.
After about five minutes the gas should be slowly ad-
justed, until a sound is heard inside like a series of small
explosions. This sound is sometimes not very distinct,
especially at high temperatures, and therefore requires a
little experience in the use of the furnace in order to be
detected. It is, however, a chief guide in determining the
proper amount of gas, and should therefore be carefully
studied. To assist in adjusting the gas it will be found
very useful to place a small piece of looking-glass beneath
the tube Z, and to adjust the gas tap until the flame
between the -burner and crucible appears wholly violet or
slightly white ; but this test is liable to fallacy if em-
ployed when the gas is first lighted, because the coldness
of the parts makes the flame much whiter than it other-
wise would appear. It is also fallacious, by the flame
appearing whiter than it really is when the crucible is
very hot. It is, however, of great assistance, especially at
intermediate temperatures. A rough deposit upon the
outer edge of the crucible indicates an excess of gas ; the
deposit is carbon. Less gas is required with a crucible in
the furnace than without one : also less is required when
the small hole at the top of the furnace is open than when
it is closed ; and less is also required when the furnace is
cold than after it has been lighted some time, because the
draught gradually increases and draws in more air. After
having accurately adjusted the gas, no further attention
to the furnace is requisite.
Having once found the proper adjustment of gas under
certain known conditions, it is well to notice the position
of the index pointer N, in order to be able at once to
adjust it to about the right point on other occasions.
Under ordinary circumstances, during daylight it is best
to set the gas nearly full on at first, and fully on at about
five minutes afterwards when the draught has become more
GORE'S GAS FURNACE. 99
powerful ; but during twilight, when the supply of gas
from the gas works is more free, the index pointer may be
set at the numbers 2-J or 3. The gas should be supplied
by a pipe of not less than f-inch bore, with a main pipe of
^ an inch ; but all depends on the pressure of gas at the
particular locality, which is very variable. The consump-
tion of gas varies from 30 to 40 cubic feet per hour, the
value of which is about twopence.
The top of the chimney should be placed in a position
where the products of combustion can pass freely away.
If it is placed in an opening or pipe leading to another
chimney, care must be taken not to have the draught too
powerful, otherwise the heat will be drawn more into the
chimney, and the supply of gas in the daytime may be
found rather deficient. The furnace will act satisfactorily,
though less powerfully, with the chimney standing in an
open room without any special outlet for the products of
combustion, provided the full height (6 feet) of chimney is
employed. Under other circumstances a chimney 4^ or 5
feet high may be used.
This furnace will produce what is generally called a
white heat; it will readily melt half a pound of copper, or
six ounces of cast iron ; it will melt as large a quantity of
those substances as the largest-sized crucible that can be in-
troduced into it will contain, sufficient space being reserved
around the crucible for draught. It requires from 20 to 30
minutes to acquire its highest temperature, and then the
entrance part of the chimney exhibits a faint red heat in
daylight. If it exhibits much more than this, the draught
is too powerful, and, if less, there is not sufficient gas.
With one dunce of copper put into the cold furnace,
and the gas lighted and properly adjusted, the copper
generally begins to melt at about the tenth or twelfth
minute, and is completely melted by the fifteenth. With
the heat well up, 1 ounce of copper has been melted in it
in 2J minutes, 1 ounce of cast iron in o minutes, 5 ounces
of copper in 4^ minutes, and 3 ounces of cast iron in 5
minutes. With the smaller hole in the top of the furnace
open, 1 ounce of copper has been melted in 3J minutes,
H 2
loo GORE'S GAS FURNACE.
and several ounces of copper have been kept in fusion
for upwards of half an hour, and may be kept so for any
length of time ; cast iron has also been fused and kept
melted under the same conditions. These various effects-
have also been obtained in a somewhat diminished degree
with the chimney standing in an open room.
When the small hole D is open some air is drawn in
that way, and less air passes up with the gas through the
tube 0, but the cold air does not much diminish the
temperature of the crucible, because it combines with the
excess of gas now passing over the edge of the inner
cylinder ; it, however, renders the flame round the crucible
white by deficiency of air, and this should be partly cor-
rected by lessening the gas. An excess of either gas or
air renders the surface of melted copper dull.
When it is desirable to perfectly avoid contact of air
with the fused substance during manipulation, a narrow
crucible should be employed, and a thin narrow ring of
fire-clay should be placed upon the top of the tube to
contract its opening ; the flame then closes completely
over the top of the crucible and prevents access of air ;
a proper adjustment of gas, together with exclusion of air
in this manner, enables a perfectly bright surface of
melted copper or even tin to be continuously maintained
from which the images of parts above are clearly reflected *
The clay ring may be withdrawn by lifting the plate B.
A less perfect exclusion of air may be obtained by em^
ploying a narrow crucible placed rather low down in its
support. A small iron dish should be placed beneath the
tube Z, to receive any melted substance that may fall.
The chief conditions of success in the use of this furnace
are sufficient gas, a suitable degree of draught, and proper
regulation of gas to air.
Mr. Griffin has devised what he calls a Reverberatory
Gas Furnace, which also produces a high temperature
without the use of a blowing machine. It is especially
suitable for assay purposes on a small scale, and for the
decomposition of siliceous minerals by fusion with alkaline
carbonates in platinum crucibles, being capable of fusing
REVERBERATORY GAS FURNACE. 101
1,000 grains of anhydrous sodium carbonate in ten
minutes.
The different parts of this furnace are also arranged in
a very convenient manner, so as to admit of its being
employed for various purposes in a chemical or assay labo-
ratory. It is based upon a new form of gas-burner which,
aided by suitable bellows, can be used as a convenient
source of heat for most operations of the chemical laboratory
and lecture table. It will boil a quantity of liquid exceed-
ing two gallons, at once ; it will raise a 4|~-inch fire-clay
crucible to full redness ; it will fuse anhydrous sodium car-
bonate in greater quantity than is required for the analysis
of a siliceous mineral ; and it will melt small quantities
of sterling silver. This amount of power is sufficient for
most chemical and many metallurgical operations.
Fig. 47 represents the gas-burner of this apparatus.
The gas is supplied by the horizontal tube, whence it
passes through a set of small holes into the box a, in
which it mixes with atmospheric air that enters freely by
the holes shown in the sketch. The gaseous mixture
passes up the vertical tube b, and is inflamed at the top,
where it burns with a single tall blue flame, which gives
no smoke, very little light, but much heat. In this
condition the apparatus differs from 'Bunsen's gas-burner'
only in size, c represents a thin brass cap, which fits the
air-box a, but moves easily round it ; d is a flat cast-iron
box with many holes around the margin, and a few small
ones on the top. This box fits loosely on the upper part
of the tube , and when it is placed on it and the gas is
lighted the flame produced consists of a series of radiating
jets, forming a horizontal circular flame of about seven
inches in diameter. If b gives a single vertical flame.
The ring of flame is suited to the purposes of boiling and
evaporation ; the single flame, to ignition and fusion.
The height of the apparatus represented by fig. 47 is
twelve inches ; the bore of the tube b is one inch ; and
the diameter of the fire-box d is four inches.
When a large crucible is to be heated to redness, the
gas-burner is to be used without the rose, and is to be
102
EEVERBERATORY GAS FURNACE.
FIG. 47.
arranged with the furnace fittings that are represented in
perspective by fig. 48, and in section by fig. 49, and the
lower part of fig. 50, a, , c, d. Letter a represents the
gas-burner ; fig. 48 b is a tall iron stool ; c a chimney
which collects atmospheric air to feed the flame, and leads
it up close to the vertical tube of a, by which contrivance
the air is warmed and the tube cooled ; d is a furnace-sole
or plate of fire-clay ; f is a reverberatory
dome, the interior of which is best shown in
the section fig. 49 ; c is a cast-iron ring or
trivet, represented more clearly in fig. 51 ;
g is an iron chimney 24 inches long and
3|- inches wide ; and h a damper to lessen
the draught when small crucibles are to be
heated. The height of this apparatus from
FIG. 49.
a to the top of / is 24 inches ; and the external diameter
of the dome f is about 8 inches. The crucible, which
REVERBERATORY GAS FURNACE. 103
may be from 4J to 4j- inches in height, is placed on the
iron ring -0, fig. 50, and that on the clay sole d, and it is
then covered by the dome /. The gas should be lighted
after the crucible is placed in its position and before the
dome is put on. The dome and the chimney are then to
be added and the operation allowed to proceed. With a
crucible of the above size the damper h is not required ;
but it must be used when the crucible is under 4 inches in
height, otherwise the draught occasioned by the extra
space within the dome causes the flame to blow down.
The damper must be put on the chimney before the
chimney is put on the dome. The iron ring (fig. 51) suits
crucibles of different sizes, according to which side, of it is
turned uppermost.
The figures show that a crucible mounted in this
FIG. 50.
FIG. 51.
furnace can lose very little heat by radiation or conduc-
tion, and hence it is that a small gas flame produces a
powerful effect. In half an hour a 4f -inch clay crucible,
filled and covered, can be heated to full redness. The
progress of the ignition can be easily examined by lifting
up the chimney g and the dome / by their respective
wooden handles. But the action of the furnace can also
be judged of by a peculiar roaring noise which it pro-
duces. If the gas and air are mixed in due proportions,
the roar is regular and continuous ; if there is too much
gas the roar is lessened, if too much air the roar is
increased, but is rendered irregular and intermittent'.
The greater the noise, the greater the heat in the furnace.
And when the roar becomes spasmodic the flame is on the
104
REVERBERATORY GAS FURNACE.
FIG 52.
point of blowing down. To prevent that occurrence, the
proportion of air must be lessened or that of gas increased.
The following arrangement is convenient when small
crucibles are to be strongly heated : anhydrous carbonate
of soda in quantities exceeding 1,000 grains can be thus
readily fused in a platinum crucible, and sterling silver can
be melted in a clay crucible. It is also available for igni-
tions or fusions in small porcelain crucibles. Fig. 49 re-
presents the arrangement of apparatus, as seen in section :
a is the gas-burner ; b the stool ; c the air chimney, and d
the furnace-sole, as already explained ; i is a cylinder of
fire-clay, 4 inches high, and 4^ inches diameter ; k is a
fire-clay furnace, in which is placed a small cast-iron ring
about 2 inches in diameter, similar in form to that repre-
sented by fig. 51, and on this ring the platinum crucible is
adjusted ; / is a fire-clay or plumbago reverberatory dome ;
and g is the chimney that forms part of the furnace repre-
sented by fig. 48. The crucible being adjusted, the gas
lighted, and the dome and chimney
put on, the lapse of twelve or fifteen
minutes, according to the quality and
pressure of the gas, suffices for the
fusion of 1,000 grains of carbonate of
soda in a platinum crucible. At the
heat which this furnace produces, the
cast-iron ring does not melt nor alloy
with the platinum crucible placed
upon it.
By a modification of these ar-
rangements, Mr. Griffin has made a
gas furnace for melting quantities of
lead, zinc, antimony, &c. This is re-
presented by fig. 52. The iron crucible
will contain nearly 30 Ibs. of lead and
about 24 Ibs. of zinc. The burner
readily melts these quantities, and then,
with a diminished quantity of gas, will
keep the metals fluid. The metals being protected from
the air suffer little loss by oxidation. Such operations as the
PRINCIPLES OF HEATING BY GAS. 105
granulation of zinc are performed with this apparatus with
great facility; it serves also for baths of fused metal. In a
large furnace of this kind, made for a special operation,
60 Ibs. of zinc have been melted with ease, and the inventor
believes that, used in this manner, the burner is powerful
enough to melt a hundredweight of zinc.
The principles of heating by gas, which have led Mr.
Griffin to the construction of these gas furnaces, may be
summed up as follows. When a crucible or other solid
body is to be heated, it is to be wrapped in a single flame
at the point of maximum heat, and loss of heat by radiation
and conduction is to be prevented by the interposition
of non-conducting materials (plumbago or fire-clay) ; and
when liquids are to be boiled or evaporated, particularly
when they are contained in vessels of glass or porcelain,
the flame is to be broken up into numerous horizontal jets,
and these are to be made to supply a large and regular
current of highly heated air, by which alone, and not by
the direct application of the flame, the vessel that contains
the liquid is to be heated. In both cases provision must
be made to secure a sufficient draught of air through the
furnace, because every cubic foot of gas requires for com-
bustion 10 or 12 cubic feet of air, and the gases which
have done their duty must be rapidly carried away from
the focus of heat. If the steam, the carbonic acid gas, and
the free nitrogen which constitute the used-up gases are not
promptly expelled, fresh gaseous mixture in the act of
producing additional heat by combustion cannot get near
the object that is to be heated, and the heat so produced
out of place is wasted.
Bunsen's gas-burner, whatever its size, is subject to
two defects : sometimes the flame burns white and smoky,
and sometimes it blows down, the gaseous mixture ex-
plodes, and the gas then burns with a smoky flame in the
tube. The remedies for these defects are as follows : If
the flame is white only when the gas is turned on very
full, the remedy is to lessen the supply of gas ; but if the
flame continues to burn white at the top when the gas is
gradually turned off and the mass of flame slowly sinks,
loo BUNSEN'S GAS BURNER.
then the holes which deliver the gas from the supply pipe
into the air-box a (fig. 47) are too large, and are placed
too directly under the centre of the vertical tube b (fig. 47),
and these defects must be corrected in the instrument.
Finally, when the flame blows down, it is because the
supply of atmospheric air is too large in proportion to
the supply of gas, and their relative proportions must be
altered. To effect this alteration the cap c is to be turned
round on the air-box a so as partially to close the holes,
and thus lessen the supply of air. If, when the gas is
alight, the fiaine needs to be lowered, first the supply of
air is to be lessened and then the supply of gas. If the
flame is to be enlarged, first the supply of gas must be
increased and then the supply of air. In short, to prevent
the fiame blowing down, the gas must always be placed in
excess, and then have the proper quantity of air adjusted
to suit it by means of the regulator c. When gas-burners
of this description have to be used in a locality where the
pressure of the gas is slight, especially in the daytime,
there is a constant tendency in the flame to blow down.
The best way to prevent that occurrence is to supply the
gas by a pretty wide tube, and to see that the current of
gas is not checked by a very narrow bore in the plug of
an intervening stop-cock, which is often the unsuspected
cause of want of pressure in the supply of gas. If this
does not suffice to prevent the blowing down of the gas,v
the holes which admit the gas from the supply pipe into
the box a of the burner should be enlarged, more or less
according to necessity. A large supply of gas compen-
sates, to some extent, for want of pressure.
When a steady and long-continued heat is desired from
a Bunsen's burner, it is proper to use two stop-cocks and
a length of caoutchouc tube between them. One of these
stop-cocks is to be affixed to the burner, and the other to
the supply pipe. The latter is to be opened wider than is
necessary to supply the required quantity of gas, and the
former is to be used to regulate the supply to the burner
exactly ; under these circumstances, if another stop-cock
is opened and gas burnt in the immediate neighbourhood,
METHOD OF MOUNTING CRUCIBLES. 10T
the flame does not so readily blow down in the regulated
burner as it does when only the stop-cock on the supply
pipe is used.
When a crucible is suspended by wires or by a ring
over the flame of a spirit lamp or gas-burner, the flame
and the hot air supplied by the flame strike the crucible
for an instant and then pass away and do no more good.
At the same time, the effect of the heating power on the
crucible is lessened by other circumstances namely, by
radiation on all sides, by a mass of cold air which con-
stantly rises around and in contact with it, and by the con-
ducting power of the metallic apparatus which supports
both the crucible and the lamp. These losses are avoided
if the crucible is inclosed in a furnace made of a non-
conducting material, such as fire-clay, which can absorb
and retain heat. In the description of the gas furnaces,
and in that of Mr. Charles Griffin's oil-lamp furnace, several
methods of mounting crucibles in fire-clay jackets have
been shown ; and we will now describe some of Mr. Griffin's
fittings that may be used to construct temporary table fur-
naces for crucibles that are to be exposed to the flame
produced by gas, oils, or spirit up to a temperature close
upon, but not quite up to, a white heat; that is to say,
up to a heat that will readily melt anhydrous carbonate of
soda and small quantities of silver, and so be fit for many
metallurgical operations, but which will not melt copper
nor cast iron.
Fig. 53 represents sections of cylinders of fire-clay
which are drawn on a scale of 1 to 8, and have the re-
lative heights and bores represented in the figures. The
clay pieces that is to say, as many of them as are necessary
for a given purpose can be adjusted over a gas flame by
means of a tripod (fig. 48) or a clay support.
The crucible to be operated upon is to be supported on
a toothed ring made either of cast iron or fire-clay, such
as are represented by figs. 51 and 54. Fig. 51 is a ring
of east iron, h representing it in section and i as seen from
above. It is about two inches in diameter, and has three
teeth projecting towards the middle of the ring. This
^08
MOUNTS FOR CEUCIBLES.
FIG. 53.
ring can be supported by any of the clay cylinders whose
bore does not exceed two inches. Fig. 54 is a ring of
fire-clay of 4 inches ex-
ternal diameter, and 1
inch in thickness, pro-
vided with three teeth
that project inwards,
and upon which a cru-
cible can be supported
without injuring the
draught of the gas fur-
nace.
Both these grates
will support crucibles at
the highest temperature
which can be produced
by spirit, oil, or gas,
without a blast of air ;
but at a white heat pro-
duced by any of these
fuels with a blast of air,
the iron ring melts, and,
if the heat is long con-
tinued, those of fire-clay
soften and partially give
way. When the fire-
clay grate (fig. 54) is
required to sustain a
very high temperature for a considerable time, it is proper
to have it made of 6 inches diameter, as represented by
fig. 53 j9, the air-way in which is the same
as that of the small grate, but the clay ring
is much stronger.
The grate is fixed above the flame at
a distance which is found by trial to place
the crucible on the point of greatest heat.
Commonly a 4-inch cylinder (53, h or g)
placed upon a suitable support serves the purpose. The
bore of the cylinders at the bottom must be wider than
FIG. 54.
MOUNTS FOR CRUCIBLES. 109
the burner, to allow of a considerable influx of atmo-
spheric air around the flame. The grate is placed on this
cylinder, the crucible on the grate, and then another cylin-
der around the crucible. The choice of this upper cylinder
depends entirely upon the size of the crucible that is to be
heated. Whatever the size of the crucible, the cylinder
must be so chosen as to fit the crucible as accurately as
possible, leaving between it and the furnace walls an open
space of not less than ^ inch, nor more than ^ inch all
round. If the upper cylinder is not contracted at the top
like 53 efg, then a cylinder of narrow bore, such as 53 a
or , must be put upon it, in order to deflect the flame and
the rising current of hot air upon the top of the crucible,,
and thus produce a reverberatory furnace. Finally, an iron
chimney, 2 or 3 feet long, must be put upon the furnace,,
to force up a draught of air sufficient to feed the flame.
Suppose a small rose gas-burner is to be arranged for
an ignition, with the use of a fire-clay support, the com-
bination of pieces necessary for the purpose may be those
represented by figure 55, where A is the fire-clay support,
and the rest of the pieces are those which are shown at
fig. 53, and described at the letters placed against each of
them in this figure. It is evident that the application of
this furnace to crucibles of different sizes depends upon
the proper choice of the cylinders here marked i and e.
Of course there is only a limited choice of crucibles suit-
able for such operations. Three inches is the extreme
width between the furnace walls of any of the pieces in
fig. 53, from a to </, and though larger cylinders^ could be
used, such as i to 0, it must be remembered that the flame
of a lamp without blast has only a limited power, and that
although a given flame will fuse 1,000 grains of carbonate
of soda in a platinum crucible, it may only heat to a
moderate redness a large clay crucible. Yet, considering
that low degrees of heat are suitable for many purposes, it
is convenient to have the power of readily adjusting a
temporary furnace to the bulk of any crucible which it
is desired to heat.
The clay pieces (fig. 53 i to p) are those that have been
110
MOUNTS FOE CRUCIBLES.
expressly designed for the blast oil furnace already
described ; but these can also be used for spirit and gas
furnaces, the respective sizes being chosen in each case
according to the size of the crucible that is to be ignited.
In respect to the means of supporting a crucible, it has
been shown that clay trivets with a wide flange namely,
the 6-inch trivets fig. 53 p will support a crucible contain-
ing 5 Ibs. of iron until that quantity of iron is melted, even
under the operations of a blast ; so that it is evident that
this method of supporting a crucible in a gas flame may
be always depended upon when no blowing- machine is
employed.
Fig. 56 represents the gas furnace arranged for boiling
or evaporating : a is the gas-burner ; b an iron stool with
FIG. 55.
FIG. 56.
three legs; c .a furnace body or iron
jacket lined with plumbago or fire-clay.
This furnace may be 14 inches high and
9 inches in diameter. The three brackets
fixed on the upper part of the jacket
serve to support the vessel that contains
the liquid that is to be boiled or evapo-
rated. A porcelain basin of 16 or 18
inches in diameter can be thus supported. It is important
to allow -between the jacket c and the evaporating basin
plenty of space for the escape of the heated air, which
ascends from the interior of the furpace. When the
evaporating basin is of small diameter, it may be sup-
TEMPORARY GAS FURNACES. Ill
ported on iron triangles, placed in the furnace c. The
section shows that around the vertical tube of the gas-
burner a there is in the bottom of the furnace c a circular
opening which is of 2 inches diameter, and through which
air passes freely, partly to feed the flame and partly to be
heated by the flame and be directed upwards in a con-
tinuous current upon the lower surface of the basin that
is to be heated. The flame within the furnace burns
steadily. No side currents of air agitate it. No part of
it touches the basin, which should receive its heat solely
from the mass of ascending hot air. The gas-burner. thus
arranged and supplied by a gas-pipe of J-inch bore, burns
about 3 cubic feet of gas in an hour, and the flame which
it produces, acting upon water contained in an open porce-
lain evaporating basin, will heat from 60 to 212 F.
1 quart in 5 minutes
1 gallon in 15 .
2 gallons in 30
and when the water boils it is driven off in steam at the
rate of more than a gallon of water per hour. The method
is consequently applicable to distillation on a small scale,
and to numerous other laboratory operations.
An excellent gas-burner for general laboratory use is
Mr. Fletcher's solid-flame burner, shown in fig. 57. This
is one of the very best heating burners which has yet been
made. The flame is FIG. 57.
of a brilliant green
colour, solid, and of
the same tempera-
ture throughout ; the
usual heating bur-
ners having a flame
with a hollow centre
of unconsumed gas.
The pattern shown
in the figure mea-
sures only 3-^ inches
in height, and it will melt half a hundredweight of lead in
an iron pot. It will boil half a gallon of water in a
112 SOLID-FLAME BURNER.
flat copper kettle in five minutes, and will melt 6 Ibs. of
lead or solder in an iron ladle in seven minutes. The
burner can be adapted for a blast where very high power
is required in a small burner. With a blast it can be
made to consume any gas supply up to 200 cubic feet per
hour, giving a very high duty. One advantage which
this burner possesses over those of the ordinary kind with
wire-gauze tops is that it cannot be spoiled by any acci-
dent. In case a solution boils over, and chokes the holes
in the perforated copper dome, the latter can be lifted
off (when the burner is warm) and cleaned. The whole
burner is designed to stand the roughest and heaviest
work without injury.
For laboratory purposes, where there is always a
liability of liquids falling on the burners, Mr. Fletcher has
introduced drip-proof high-power burners with pure solid
nickel flame surfaces. These are shown at fig. 58.
These are undamaged by the dirtiest work, and will
burn perfectly under a constant drip. The nickel flame
FIG. 58.
DRIP PROOF
HIGH POWER BURNER
surface adds considerably to the first cost of the burner,
but it is practically everlasting, and will neither rust nor
burn away.
These burners are generally used under vessels either
fixed or supported on wrought-iron stands. The burners
themselves are very small in proportion to the power
and the size of vessel they will heat. The bottom of
the vessel should be about 1-J inch clear above the top
of the burner.
LUTES AND CEMENTS. 113
An improved pattern of Safety Bunsen Burner, also
introduced by Mr. Fletcher, is shown in fig. 59.
This will be found FIG 59
as perfect as any up-
right tube burner can
possibly be made, of
the highest possible
power for size, can be
turned down to the
merest flicker without
lighting back, and can
be mounted on tubes
in any form or number
when very high powers
are required.
The india-rubber
tubing for these burners
must be of good size,
and smooth inside, made
without wire.
LUTES AND CEMENTS.
It may be as well
to mention in this part
of the work the various
lutes and cements which
may be employed, either in fire operations or in making
good joints in experiments with gases or liquids. The
following are the principal kinds.
FIRE LUTE is composed of good clay two parts, sharp
washed sand eight parts, horse-dung one part. These
materials are to be intimately mixed ; and afterwards the
whole is to be thoroughly tempered like mortar. Mr.
Watt's fire lute is an excellent one, but is more expensive.
It is made of fine powdered Cornish (porcelain) clay
mixed to the consistence of thick paste, with a solution of
borax in the proportion of 2 ounces of borax to a pint of
hot water,
FAT LUTE is prepared by mixing fine clay, in a fine
I
114 LUTES AND CEMENTS.
powder, with drying oil, so that the mixture may form a,
ductile paste. When this paste is used, the part to which
it is applied ought to be very clean and dry, otherwise it
will not adhere. Glazier's putty is very similar to this. .
ROMAN CEMENT. This must be kept in well-closed
vessels, and not moistened until it is required for use.
PLASTER OF PARIS. This is mixed with water, milk, or
weak glue, or starch water.
These three lutes stand a dull red heat : the two latter
may be rendered perfectly impermeable to gaseous bodies
by being smeared over with oil, or a mixture of oil and
wax.
LINSEED OR ALMOND MEAL, mixed to the consistence of a
paste with water, milk, lime water, or starch paste. This
lute is very manageable and impermeable, but does not
withstand a heat greater than about 500 P.
LIME AND EGG LUTE If just the sufficient quantity of
water be added to quick lime to reduce it to a dry powder,
and that is mixed well and rapidly with white of egg
.diluted with its own volume of water, and the mixture
spread immediately on strips of linen and applied to the
part, then powdered with quick lime, it forms a good
cement. Instead of white of egg, lime and cheese may
be used, or lime with weak glue water. This lute dries
very rapidly, becoming very hard and adhering strongly
to glass ; but its great inconvenience is the want of
flexibility.
WHITE LEAD MIXED WITH OIL. If this mixture be spread
upon strips of linen, or bundles of tow, it acts much in the
same manner as the lime lutes.
YELLOW WAX is often used as a lute, but it becomes
very brittle at a low temperature. It may be rendered
less brittle, and at the same time more fusible, by an
admixture of one eighth crude turpentine.
SOFT CEMENT is prepared by fusing yellow wax with
half its weight of crude turpentine and a little Venetian
red in order to colour it. It is very flexible, and takes
any desired form under the pressure of the fingers.
CEMENT FOR BRASS ON GLASS (for instance, petroleum
LUTES AND CEMENTS, 115
lamps). 1 part of caustic soda, 3 parts of resin, arid 5
parts of water are boiled together till solution is effected,
when it is. intimately mixed with one-half of its weight in
plaster of Paris. The mixture hardens within one hour,
and is impermeable for coal oil.
CEMENT FOR MENDING PESTLES &c. One of the strongest
cements and one that can be very readily made is ob-
tained when equal quantities of gutta-percha and shellac
are melted together and well stirred. This is best done in
an iron capsule placed on a sand-bath, and heated either
over a gas furnace or on the top of a stove. It is a com-
bination possessing both hardness and toughness, qualities
that make it particularly desirable in mending pestles and
mortars. It is very useful for securing the handles to the
wedgwood ware, and some old ones that were much
chipped and split, when thus mended, have been quite as
useful as new ones, and have stood several months' wear
without any apparent change. When this cement is used
the articles to be mended should be warmed to about the
melting point of the mixture and then retained in proper
position until cool, when they are ready for use.
ADHESIVE PASTES. 1. Tragacanth, 1 oz. ; gum arabic,
4 ozs. ; water, 1 pint. Dissolve, strain, and add thymol, 14
grains ; glycerin, 14 ozs. ; and water to make 2 pints.
Shake or stir before using it.
2. Eye-flour, 4 ozs. ; alum, -J oz. ; water, 8 ozs. Eub to
a smooth paste, pour into a pint of boiling water, heat until
thick, and finally add glycerin, 1 oz. and oil of cloves, 30
drops.
3. Eye-flour, 4 ozs. ; water, 1 pint. Mix, strain, add
nitric acid, 1 drachm ; heat until thickened, and finally add
carbolic acid 10 minims, oil of cloves, 10 minims, and
glycerin, 1 oz.
4. Dextrin, 8 parts ; water, lOparts ; acetic acid, 2 parts.
Mix to a smooth paste and add alcohol, 2 parts. This is
suitable for bottles or wood, but not for tin, for which,
however, the first three are adapted.
5. The paste used by the United States Government
for gumming postage stamps is made by the following
i 2 -
116 LUTES AND CEMENTS.
formula. It has the properties of being very adhesive, does
not become brittle or scale off, and is well adapted for
sticking paper labels to tin and other metals. Take of
starch, 2 drachms ; white sugar, 1 ounce ; gum arabic, 2
drachms ; water, q.s. Dissolve the gum, add the sugar,
and boil until the starch is cooked.
WATERPROOF CEMENT. Mr. Edmund Davy, F.K.S., de-
scribed a cement made by melting in a saucepan two
parts by weight of common pitch, and adding to it one
part by weight of gutta-percha, stirring and mixing them
well together until they were completely incorporated
with or united with each other. The mixture then forms
a homogeneous fluid which may be used in this state for
many purposes, and is remarkable on account of the
facility and tenacity with which it adheres to metals,
stones, and glass. It may be poured into a large basin of
cold water, in a thinner or thicker stream, or as a cake.
In this state, while warm, it is quite soft, but may be soon
taken up out of the water and drawn out into longer or
pressed into shorter pieces, or cut or twisted into frag-
ments, which may again be readily reunited by pressure.
When the cement is cold, or before, it may be removed
from the water and wiped dry, when it is fit for use. It
is of a black colour ; when cold, it is hard. It is not
brittle, but has some degree of elasticity, which is in-
creased by a slight increase of heat. It appears to be not
so tough as gutta-percha, but more elastic. Its tenacity
is very considerable, but inferior to gutta-percha. It
softens when put into water at about 100 F. ; when
heated to above 100 F. it becomes a thin fluid ; and if
the heat is gradually increased, it passes through inter-
mediate stages of softness, becomes viscous like bird-lime,
and may be extended into threads of indefinite length : it
remains in this state even when exposed for some time in
a crucible to the heat, of boiling water, 202 F. Water
appears to have no other action upon it but that of soften-
ing it when warm or hot, and slowly hardening it when
cold* The cement adheres strongly, if pressed on metal
or other surfaces, though water be present, provided such
LUTES AND CEMENTS. 117
surfaces be warm. This cement is applicable to many
useful purposes. It adheres with great tenacity to metals,
wood, stones, glass, porcelain, ivory, leather, parchment,
paper, hair, feathers, silk, woollen, cotton, linen fabrics,
&c. It is well adapted for glazing windows, or as a
cement for aquariums. This cement does not appear to
affect water, and it will apparently be found applicable
for .coating metal tanks ; to secure the joints of stone
tanks ; to make a glue for joining wood, which will not
be affected by damp ; and to prevent the depredations of
insects on wood.
RESINOUS OR HARD CEMENT is made by fusing together
at the lowest possible temperature one part of yellow wax
and five or six of resin, and then adding gradually one part
of red ochre or finely powdered brickdust (plaster of Paris
succeeds very well), and then raising the temperature to
212 at least, until no more froth arises or agitation takes
place, and stirring it continually until cold. This cement
is employed in a hot state. This lute is much used for
fixing brass caps, &c., to air-jars.
CAOUTCHOUC. Tubes of vulcanised caoutchouc form a
very ready means of attaching one piece of apparatus to
another, and they possess the peculiar advantage of flexi-
bility, which allows the various parts of the apparatus
which they connect to move in different directions to
a slight extent, so that the whole is not so likely to be
fractured as when connected in an inflexible manner.
Caoutchouc is also less acted upon by gases and vapours
than almost any other substance w^e know ; even chlorine
attacks it but slowly, and when unvulcanised it possesses
the valuable property of forming a perfect joint when
freshly cut edges are brought and pressed together, hence
the facility with which it is manufactured into tubes. The
mode of manufacturing small connecting tubes, which are
< .
often required to be of unvulcanised caoutchouc, is as
follows : Take a piece of the sheet caoutchouc of the
required size, and warm it either in the hand or before a
fire, until it is perfectly soft ; then place it around a glass
rod of the requisite size, pressing the edges close together
118 LUTES AND CEMENTS.
with the fingers ; when close together cut off the super-
abundance with a sharp pair of scissors, and the newly
cut edges will unite by simple pressure of the nail. When
well executed the joint is scarcely apparent. In order
to prevent the caoutchouc from adhering to the rods on
which the tube is formed, a little moisture or dry starch
may be employed. When caoutchouc is not at hand,
oiled paper may be substituted, the joint being made of
wax.
Faraday gives the following directions for luting iron,
glass, or earthenware retorts, tubes, &c. ; for furnace opera-
tions. When the lute has to withstand a very high tem-
perature it should be made of the best Stourbridge clay,
which is to be made into a paste varying in thickness
according to the opinion of the operator. The paste
should be beaten until it is perfectly ductile and uniform,
and a portion should then be flattened out into a cake of
the required thickness, and of such a size as shall be most
manageable with the vessel to be coated. If the vessel be
a retort or flask, it should be placed in the middle of the
cake, and the edges of the latter raised on all sides and
gradually moulded and applied to the glass ; if it be a
tube, it should be laid on one edge of the plate, and then
applied by rolling the tube forward. In all cases the
surface to be coated should be rubbed over with a piece
of the lute dipped in water for the purpose of slightly
moistening and leaving a little of the earth upon it; if
any part of the surface becomes dry before the lute is
applied, it should be re-moistened. The lute should be
pressed and rubbed down upon the glass successively
from the part where contact was first made to the edges,
for which purpose it is better to make them thin by
pressure and also somewhat irregular in form, and if
at all dry they should be moistened with a little soft
lute. The general thickness may be about J to ^ oi an
inch.
Being thus luted, the vessels are afterwards to be
placed in a warm situation, over the sand-bath or near
the ash-pit, or in the sun's rays. They should not be
LUTES AND CEMENTS. 119
allowed to dry rapidly or irregularly, and should be
moved now and then to change their positions. To pre-
vent cracking during desiccation, and the consequent
separation of the coat from the vessel, some chemists
recommend the introduction of fibrous substances into
the lute, so as mechanically to increase the tenacity of its
parts. Horse-dung, chopped hay and straw, horse- and
cow-hair, and tow cut short, are amongst the number.
When these are used, they should be added in small
quantity, and it is generally necessary to add more
water than with simple lute, and employ more labour
to insure a uniform mixture. It is best to mix the
chopped material with the clay before the water is put-
to it, and, when adding the latter, to mix at first by
stirring up the mass lightly with a pointed stick or fork ;
it will then be found easy, by a little management, to
obtain a good mixture without making it very moist.
The luting ought to be made as dry as possible con-
sistent with facility in working it. The wetter it is, the
more liable to crack in drying, and vice versa.
Mr. Willis recommends, when earthenware retorts,
&c., are to be rendered impervious to air, the following
(-oating : One ounce of borax is to be dissolved in half a
pint of boiling water, and as much slaked lime added as
will make a thin paste. This composition is to be spread
over the vessel with a brush, and, when dry, a coating of
slaked lime and linseed oil is to be applied, This will dry
sufficiently in a day or two, and is then fit for use.
IRON CEMENT. This mixture is used for making per-
manent joints generally between surfaces of iron. Clean
iron borings or turnings are to be slightly pounded, so as
to be broken but not pulverised : the result is to be sifted
coarsely, mixed with powdered sal-ammoniac and sulphur,
and enough water to moisten the whole slightly. The
proportions are 1 sulphur, 2 sal-ammoniac, and 80 iron.
No more should be mixed than can be used at one time.
Mr. Cooley states that he is informed by one of the first
engineers in London that the strongest cement is made
without sulphur, and with only one or two parts of sal-
120 PLUMBAGO AND CLAY CEUCIBLES.
ammoniac to 100 of iron borings ; but that when the
work is required to dry rapidly, as for the steam joints of
machinery wanted in haste, the quantity of sal-ammoniac
is increased a little, and a very small quantity of sulphur
is added. This addition makes it set quicker, but reduces
its strength.
Several excellent cements are described in Cooley's
' Cyclopedia of Practical Eeceipts.' From these the fol-
lowing are selected :
BEALE'S CEMENT. Chalk, 60 parts ; lime and salt, ol
each 20 parts ; Barnsey sand, 10 parts ; iron filings or
dust, and blue or red clay, of each 5 parts. Grind together
and calcine. This is patented as a fire-proof cement.
BOILER CEMENT. Dried clay in powder, 6 Ibs. ; iron
filings, 1 lb., made into a paste with boiled linseed oil.
This is used to stop leaks and cracks in iron boilers,
stoves, &c.
BRUYERE'S CEMENT. Clay, 3 parts ; slaked lime, 1 part ;
mix and expose them to a full red heat for 3 hours, then
grind to powder.
This makes a good hydraulic cement.
OXYCHLORIDE OP ZINC CEMENT. In solution of zinc chlo-
ride, of 1-49 to 1-65 specific gravity, dissolve 3 per cent,
of borax or sal-ammoniac, and then add zinc oxide, which
has been heated to redness, until the mass is of a proper
consistency. This cement becomes as hard as marble. It
may be cast in moulds like plaster of Paris.
CRUCIBLES, CUPELS, ETC.
The crucibles best known in commerce are the Hessian,
the Cornish, the Stourbridge, and the London clay cru-
cibles ; charcoal, porcelain, plumbago, platinum, iron,
nickel, silver, and gold crucibles are also required in small
operations. Of the clay crucibles, the London pots are
much to be preferred, on account of their very refractory
nature. They resist the action of fused oxide of lead
better than most clay crucibles, and they are also
better made than the two other kinds, being much
PLUMBAGO AND CLAY CRUCIBLES. 121
smoother and more regularly formed. They have the
form of a triangular pyramid (see fig. 60, crucibles and
cover), and are made in such sizes that they fit one into
the other, forming nests. The triangular form is very
convenient, because there are three spouts, from either of
which can be poured the fused contents of the pot. The
Cornish crucibles are circular, and do not stand changes
of temperature so well as the London pots, neither can
they endure such an extreme FIG. 60.
heat, for they agglutinate
and run together at a tempe-
rature which does not touch
the others. Dr. Percy says
they are more generally useful
than any other crucible. The
Hessian pots are the worst of
all ; they do not stand mode-
rate change of temperature
without risk of fracture, so
that they require to be very
carefully used. There is also another kind of pot in use,
made of the same material as the London crucibles, termed
a ' skittle pot,' from its resemblance to the ordinary wooden
skittle or ninepin. They are exceedingly useful for the
fusion of large masses of matter, or such substances as
boil or bubble much when heated. Plumbago or graphite
crucibles are rapidly superseding all other kinds when
metals have to be melted. They possess many advan-
tages over clay crucibles. Their surface is very smooth ;
they are not liable to crack, however violent the changes
of temperature may be to which they are subjected ; they
bear the highest heat without softening, and can be used
repeatedly. Owing to the reducing property of the carbon
they contain, they must not be employed when oxidising
actions are required.
Crucibles and all plumbago fittings for furnaces should
be of the ' Salamander ' brand. These require no anneal-
ing or care in heating up, and stand strong fluxes better
than the ordinary make.
2-2 PLUMBAGO AND CLAY CRUCIBLES.
The Patent Plumbago Crucible Company have re-
cently introduced a very excellent fluxing crucible. It is
made of fine white china clay, is perfectly smooth inside
and out, and will stand very high temperatures without
softening.
Stou'rbridge clay crucibles are not much used. They
require the greatest care in using them, and are spoilt
after the first operation.
Porcelain crucibles are not used in large assaying or
metallurgical operations, but they are invaluable in small
laboratory experiments. They are practically infusible,
are little liable to crack, and are almost unacted on by
reagents and fluxes. In many cases they will replace the
more expensive platinum crucibles, and where easily re-
ducible metals are under treatment they must be used in
preference to platinum.
Crucibles, in order to be perfect and capable of being
used indifferently for any operation, ought to possess the
four following qualities : first, not to break or split when
exposed to sudden changes of temperature ; secondly, to
be infusible ; thirdly, to be only slightly attacked by the
fused substances they may contain ; fourthly and lastly, to
be impermeable, or nearly so, to liquids and gases. But
as it is very difficult to unite all these qualifications,
various kinds of pots are made to fulfil one or more of
them.
In order to render crucibles capable of withstanding
changes of temperature without breaking, a certain pro-
portion of substances infusible by themselves is mixed
with the pasty clay; sand, flint, fragments of old crucibles,
black-lead, and coke are used for this purpose. They are
reduced to a state of division more or less fine, according
to the grain of the clay paste. For ordinary pots the
powder ought not to be very fine ; but for porcelain cru-
cibles it ought to be as fine as flour. The choice of these
various bodies depends upon the use for which the crucible
is intended.
The most refractory crucibles are those made with the
pure clays, or such as contain little or no oxide of iron r
PLUMBAGO AND CLAY CRUCIBLES. 123
and especially those free from calcareous matters. Amongst
clays, the best are those which contain most silica ; never-
theless, these are not absolutely infusible, and in the high
temperature of a wind furnace they sometimes soften so
much as actually to fall into a shapeless mass. This defect,
as before stated, can be in some measure diminished by
mixing with the clay a quantity of graphite or coke ;
either of these substances forms a kind of solid skeleton,
which retains the softened clay, and prevents its falling
out of shape.
Coke and black-lead are more efficacious than sand,
because they have no action on clay, whilst sand forms a
fusible compound with it. If too large a quantity of
black-lead or coke be employed, it gradually consumes in
the fire, and the pots become porous, and break at the
least movement. Wood charcoal can be used instead of
black-lead or coke, but is not so good, as it burns more
readily.
Black-lead crucibles are generally composed of 1
part of refractory clay, and from 2 to 3 of black-lead.
These pots withstand all possible changes of tempera-
ture without cracking, and their form is rarely changed
by the heat ; not because they are absolutely infusible,
but because they are supported by the skeleton of
graphite.
Crucibles into whose composition carbonaceous matters
enter, reduce any oxides that may be heated in them, and
hence are inconvenient in certain cases. They can, never-
theless, be employed in all cases by giving them a lining
of clay, which must be tolerably thick, and well dried
before use.
Earthen crucibles which have not been baked at a
white heat are more or less permeable to liquids and
gases, according to the grain. In order to render them
impermeable to liquids, they must be heated to such a
temperature as will suffice to fuse the outside. When
treated in this way, however, they are very liable to crack
with sudden changes of temperature : the best method,
therefore, of rendering them capable of containing water,
124 GOOD AND BAD CRUCIBLES.
<fcc., is to coat them with the mixture of borax and lime
described as Willis's lute.
In order that crucibles may resist the corrosive action
of the fused substances contained within them, they must
be as compact as possible, and the substance of which they
are made must have little or no tendency to combine with
the fused contents. The metals and their non-oxidised
compounds attack neither clay nor black-lead ; but there
are, nevertheless, some metallic substances galena, for
instance which, without exercising any chemical action
on earthy matters, have the property of filtering through
their pores.
The readily reducible oxides gradually corrode black-
lead crucibles and those pots into the composition of which
coke enters, by burning the carbonaceous matter. The
greater number of these oxides, the alkalies, earths, and
glasses (which are fusible silicates, borates, &c.), act more
or less powerfully on the earthy base of all crucibles ; so
that these substances are most difficult to keep in fusion
for any length of time. They attack the crucible layer
by layer, dissolving the substance of which it is composed,
and after a lapse of time rendering it so thin that it can-
not withstand the pressure of the molten mass within it ;
and the fracture of the pot, and consequent loss of con-
tents, is inevitable.
Under tlie same circumstances, all those crucibles
whose texture is loose are more readily corroded than
those with a firm, compact body ; because the corrosive
substance filters to a certain depth in the former crucibles,
and in conseqiience has a larger surface to act upon than
when it is contained in a compact pot.
Earthen crucibles may be assayed by noticing the time
they will contain fused litharge, which exercises a very
corrosive action on them, honeycombing them in all di-
rections ; and those pots which contain it longest without
undergoing much damage, may be considered the best.
However, this method of assay is not exact, even by taking
into account the thickness of the pot, for litharge runs
through crucibles : first, because it is very fusible, and
GOOD AND BAD CRUCIBLES. 125
easily filters through their pores ; and secondly, it has the
property of forming fusible compounds with all the sili-
cates by combining with them. From these remarks it
will be evident that a crucible whose grain is loose will
readily allow litharge to pass through it, however slightly
its substance may be fusible or acted on ; or, on the con-
trary, it may be very easily acted on (even when infusible)
when it has an extremely fine grain ; so that the prompti-
tude with which a crucible is pierced by litharge bears
no relation to its fusibility. A crucible of pure quartz
will be very readily attacked by litharge, because the
latter has much affinity for silica, and the simple silicates
of lead are all very fusible ; whilst a crucible composed of
silica, alumina, and lime, which by itself is very fusible,
would be corroded less rapidly, because the oxide of lead
has much less affinity for the earths than it has for the silica;
moreover, it forms less fusible compounds with the earths
than with silica alone. The assay of crucibles with li-
tharge, if not of use in ascertaining their degree of fusi-
bility, fulfils perfectly its object when it is wished to prove
the resistance a crucible has to the corrosive action of
various bodies in a state of fusion ; for of all fusible sub-
stances, none exercises such a powerful action on earthy
matters as litharge.
Crucibles ought not only to resist the corrosive action
of those bodies they may contain, but also that of the ash
produced by the combustion of the fuel in which they
may be placed. These ashes being often calcareous, alka-
line, or ferruginous, act on the clayey part of the crucibles
exactly as the fluxes. Whence it follows, that those cru-
cibles which contain litharge longest will also resist the
action of the fluxes best.
In order to ascertain the fusibility of a crucible, a
direct experiment must be made, either by heating a piece
in a crucible lined with charcoal, and ascertaining if its
angles be rounded, if its substance has become translucent,
&c. ; or, better still, by heating the crucible to be assayed
with another whose properties are well known.
As to permeability, it may approximately be ascer-
126 CHARCOAL CRUCIBLES.
tained by filling two crucibles with water, and noting what
length of time is required to empty them completely ; the
crucible which contains it longest being, of course, the
least permeable.
To prove if a crucible be able to sustain great changes
of temperature without breaking, introduce it, perfectly
<>old, into a furnace full of lighted coal : take it out when
reddish white, and expose it to a current of cold air pro-
duced by a bellows or otherwise : if it stand these trials,
it may be heated afresh and plunged red-hot into water,
and, if it be not broken, placed immediately in the fire.
The best pots support all these operations without break-
ing ; but it often happens that they are filled with innu-
merable small fissures, through which fused matters can
pass. This can be ascertained by fusing rapidly in the
assay pot a quantity of litharge : if these fissures be pre-
sent, the fused oxide will readily filter through them.
CHARCOAL CRUCIBLES. As all oxidised matters act readily
on clay pots, and a great number of the metals and their
compounds adhere to them, they have long since been
replaced, under certain circumstances, by charcoal cru-
cibles, which do not possess these disadvantages. The
older assayers used merely a piece of charcoal, with a
hole made in it, and then bound round with iron or other
wire. The use of these has, however, been abandoned for
some time, and earthenware crucibles lined with charcoal
have been substituted (see fig. 61, a, 6, and c). These may
be considered as charcoal pots enveloped with refractory
clay ; they are solid, always free from cracks, and easy of
preparation, and they have the same properties as the
solid charcoal crucibles without their inconveniences.
In order to prepare these crucibles, the charcoal must be
chosen carefully, so as to contain no foreign substances ; it
must be pulverised and passed through a sieve ; the powder
moistened with water or treacle, mixed with a spatula, and
then kneaded with the fingers until it just adheres, and
forms adhesive lumps without being sufficiently wet to
adhere to the hand. Some advise the addition of gum to
the water with which the charcoal is moistened. The
LINED CRUCIBLES. 127
crucible is moistened slightly by being plunged into water,
and withdrawn as speedily as possible, and about half an
inch in depth of the charcoal paste, prepared as above, is
placed in it ; the paste is then pressed firmly down, by
means of a wooden pestle : the blows are to be slight at
first, and then increase in force until it is as firm as possible:
another layer is then applied and pressed as before, and
the process repeated until the crucible is quite full, taking
great care to render all as firm as possible, especially at
the sides. In order to make each layer adhere firmly to
the other, the surface must be scratched rather deeply
with the point of a knife before a new layer is applied.
When the crucible is completely filled, a hole is to be
scooped in the charcoal of about the form of the pot. The
sides are then rendered smooth by friction w r ith a glass
rod. This is absolutely necessary, so that the metallic
globules produced in an assay may not be retained by the
asperities of the lining, but may be readily enabled to unite
into one button. When a lined charcoal pot is well made,
its sides are very smooth and shining. For ordinary use,
the lining may be f ths of an inch thick at the bottom and
^th or so at the sides : but in some cases, for instance,
when the substance to be fused is capable of filtering
through the lining and attacking the pot as a flux, it must
be at least twice the above thickness in every part.
Lined crucibles have many advantages over plain
128 LIME CRUCIBLES.
crucibles. The lining gives them greater solidity, and
prevents a loss of shape when softened ; for plain crucibles
are always three-fourths empty when their contents are
fused, on account of contraction in volume : the pots
then have nothing to sustain their sides when they soften
towards the end of the assay, at which period the highest
temperature is employed. Besides, vitreous matters do
not penetrate the lining, and, exercising no action on it,
can be obtained in a state of purity, and the exact weight
determined : if fused in a plain pot, the weight could not
be ascertained, because a portion would adhere to the
sides, and the resulting mass would not be pure, having
taken up a portion of the crucible in which the fusion was
effected.
The lining, too, effects the reduction of certain metallic
oxides by cementation, and does away with the necessity
of adding powdered charcoal to the body to be reduced.
This property is very valuable, because, when an oxide is
reduced by mixing it with charcoal, an excess must always
be employed, and this excess remains with the metal, and
prevents its exact weight from being ascertained. No
oxidising substances or bodies which readily part with
oxygen (oxide of copper, for examp]e) must be calcined in
a plumbago or charcoal-lined crucible, unless indeed the
chemical union of the charcoal with the oxygen is desired.
LIME CRUCIBLES. Some years ago Deville proposed the
use of crucibles cut out of solid blocks of pure lime, in
order to prevent the introduction of carbon and silicon
into metals and alloys during the process of fusion.
The results of experiments made with such crucibles
were found to be extremely satisfactory, and metals fused
therein, as iron, manganese, nickel, cobalt, &c., were
obtained far purer and more malleable and ductile than
when fused in the usual clay or brasqued crucibles. Un-
less, however, the crucibles required were of very small
size, it was found difficult to obtain blocks of lime for
shaping them sufficiently large and free from flaws ; and
experiment showed a considerable loss, both by breakage
when shaping them, and by their cracking when in the
\v
TY }}
IJME CRUCIBLES. 129
furnace. In order to obviate this, trials were made with
clay crucibles lined with lime, but ineffectually, as these
crucibles invariably melted down before the requisite heat
was arrived at a result due to the action of the lime
itself upon the outer clay crucible.
Mr. David Forbes, F.R.S.,has published in the c Chemi-
cal News ' the result of some very valuable experiments
on this subject. The arrangement he proposes fully
answers the purpose, the crucibles being capable of stand-
ing the heat of melted wrought iron or cobalt without
fusion or cracking, as well as of being made of any reason-
able size.
A clay crucible of somewhat larger capacity than the
desired lime one is filled with common lamp-black, com-
pressing the same by stamping it well down. The centre
is then cut out with a knife until a shell or lining of lamp-
black is left firmly adherent to the sides of the crucible,
and about half an inch or less in thickness, according to
the size of the crucibles ; this lining is now well rubbed
down with a thick glass rod until its surface takes a fine
glaze or polish, and the whole cavity is then filled up with
finely powdered caustic lime, thoroughly pressed down,
and a central cavity cut out as before ; or the lime powder
may be at once rammed down round a central core of the
dimensions of the intended lime crucible.
This lime lining is naturally rather soft before being
placed in the furnace, but upon heating, it agglutinates, and
forms a strong and compact crucible, which is prevented
from acting upon the outer one by the interposed thin layer
of lamp-black, and at the end of the experiment it generally
turns out as solid and compact as those made in the lathe.
Experiments made with such crucibles, even up to
dimensions containing several pounds of metal, have proved
them extremely well suited for these operations, and doubt-
less similar crucibles could be made, lined with magnesia
or alumina as required. In some cases ordinary black-
lead crucibles, lined with powdered lime, magnesia, or
alumina, might possibly be found to answer.
Having frequently used lime crucibles in metallurgical
K
130 ALUMINA CRUCIBLES.
operations, and having met with the inconvenience pointed
out by Mr. Forbes, the editor can appreciate the great
value of his improvement. It is one which cannot fail to
be extensively adopted in metallurgical laboratories.
In certain particular experiments, crucibles are lined
with other bodies besides charcoal and lime, such as silica,
alumina, magnesia, or chalk, by merely moistening their
respective powders with water, and applying the paste as
above described for the charcoal. A slight layer of chalk
lessens the liability of attack from fused litharge.
ALUMINA CRUCIBLES are strongly to be recommended in
many metallurgical operations. They are made in the
following manner. Ammonia alum is ignited at a full
white heat, when it leaves behind pure alumina in a
dense compact form : this is to be finely powdered. To a
solution of another portion of ammonia alum in water,
ammonia is added, when alumina is precipitated in a gela-
tinous state : this is to be washed until free from sulphate
of ammonia. The dense alumina is then mixed with water
and worked up into a paste, the precipitated gelatinous
alumina being kneaded in from time to time ;' this gives
coherency : and when sufficient has been added (which
must be ascertained experimentally), the mass may be
moulded into shape. These crucibles require slow and
careful drying ; but they well repay all the care which is
bestowed on them, for they do not readily crack, are
attacked by very few fluxes, give out no impurities to
metals which are melted in them, and are infusible at the
highest heat of the furnace.
MAGNESIA CRUCIBLES AND BRICKS. M. H. Caron has
pointed out the advantages which would accrue to metal-
lurgy from the employment of magnesia as a refractory
material. Formerly the high price of this earth appeared
likely to confine it to the laboratory. Now, circumstances
have happily changed ; recent modifications introduced
into the manufacture of cast steel, and especially the
employment of Siemens's furnace and Martin's process,
absolutely demand more refractory bricks than those at
present in use, irrespective of price. On the other hand,
MAGNESIA CRUCIBLES AND BRICKS. 131
native carbonate of magnesia, which formerly cost 10/.
the ton, may now be obtained at the price of 21. 15s.
delivered at Marseilles or Dunkirk. Calcination at the
place where the carbonate is obtained may still further
reduce its price. 1 The following is M. Caron's process
for its agglomeration, which may be employed by the
chemist for the ready preparation of refractory vessels
of all forms ; by the physicist to obtain pencils for oxy-
hydrogen lighting purposes ; and also by the manufac.
turer, to replace, in some cases, fire-bricks which have
become unsuitable for carrying out different processes
of heating.
The magnesia which he employs comes from the island
of Elba, where it is found in considerable quantities as a
native carbonate, white, very compact, and of great hard-
ness. This carbonate contains traces of lime, silica, and
iron ; it is, besides, interspersed sometimes with serpentine
and large plates of silica, which would diminish the in-
fusibility of the substance, and render it especially unfit
for oxyhydrogen illumination if their removal is neglected.
These plates are, however, easily recognised, and may
be readily separated, even in working on the large scale.
In the case of refractory bricks, the presence of a small
quantity of these foreign bodies would, at the most, give
rise, under the influence of the highest temperatures, to a
slight vitrification, offering no serious inconvenience.
Before powdering the carbonate, it is advisable to bake
it at the temperature necessary for the expulsion of the
carbonic acid ; the material then becomes very friable,
and may be pulverised more easily. It is then possible to
separate the serpentine and silica, which do not become
friable under the influence of heat. This preliminary
treatment does not permit of the agglomeration of the
magnesia, and, even were this difficulty to be overcome, a
temperature higher than that of the original calcination
would cause an enormous contraction, producing fissures
and alterations in shape which would interfere with the
1 This preliminary calcination requires less heat than burning lime, and
diminishes the weight of the carbonate one-half.
K 2
132 MAGNESIA CRUCIBLES AND BRICKS.
use of this substance. It is therefore indispensable, before
moulding the magnesia, to submit it to a very intense heat,
at least equal to that which it is intended to support sub-
sequently.
Thus calcined it is not plastic, its appearance is sandy,
and compression does not cause it to acquire any cohesion ;
a mixture of magnesia, less calcined, imparts to it this
quality. The quantity of the latter to be added necessarily
varies with the degree of calcination of the two magnesias ;
it is scarcely one-sixth of the weight of that which has
been exposed to the temperature of melting steel. It only
now remains to moisten it with 10 or 15 per cent, of
its weight of water, and strongly compress it in iron
moulds, as adopted in making artificial fuel. The brick
produced in this operation hardens on drying in the air,
and becomes still more resisting when it is subsequently
calcined at a red heat. The same process would appear
practicable, varying the form of the moulds, for obtaining
crucibles of great capacity ; but compression is difficult in
large masses, as well as when the moulds have a large
surface, as the magnesia adheres strongly to the sides.
Although M. Caron has been able to obtain small crucibles
for the laboratory, he does not consider this process
adapted to make the large crucibles employed in steel
melting. In this and other cases it is preferable to ag-
glomerate the magnesia in the wet way.
To endow magnesia with a sort of plasticity, advantage
is taken of a property of this earth pointed out in ' Ber-
zelius's Chemistry.' When strongly calcined and then
moistened, it hardens in drying. This fact is doubtless
due to a hydration which takes place without sensible in-
crease of temperature. When solidified in this manner
the magnesia only loses the assimilated water at a high
temperature. Then calcination not only does not disin-
tegrate it, but, on the contrary, confers upon it a hardness
and resistance comparable to those of ordinary crucibles
after their baking. This being understood, the appli-
cation of this fact is obvious. Thus, the magnesia to be
employed in the manufacture of crucibles should only be
PLATINUM CRUCIBLES. 133
moistened, moulded into shape, dried, and then ignited.
For the construction of steel-melting furnaces, a paste of
moistened magnesia should be plastered over the walls ; it
will become ignited in due course without any particular
precautions being taken. It sometimes happens, however,
either owing to the magnesia being too much or too little
hydrated, or owing to its containing siliceous matter, that
the vessels before or after firing do not possess quite the
desirable solidity ; they should then, to acquire this, be
simply moistened in a cold saturated solution of boracic
acid, dried, and then fired as before. This operation does
not render the magnesia more fusible ; it only causes the
grains of the substance to cohere more strongly together.
Very pure, strongly calcined, and finely pulverised
magnesia may be employed in the form of paste (barbotine),
and yields the most delicate and translucent crucibles, as
well as the sharpest and most complicated impressions.
MALLEABLE IRON CRUCIBLES are often very serviceable in
assays of fusibility, and of certain selenides and sulphides,
also in assays of galena or ordinary lead ore. They are
either made of hammered sheet iron, or by plugging up
small iron tubes, such as gun-barrels, &c. The latter are pre-
ferable, because thick solid crucibles can be used a number
of times, whilst the others are necessarily very thin and
can be used only once. Whenever iron crucibles are em-
ployed at a very high temperature, they must be placed
in those of earthenware, which protect them from the
oxidising action of the air ; but when they are not heated
above the temperature of a copper assay, they may be used
naked, if tolerably thick.
For assays at the above temperature, cast-iron crucibles
may be employed with advantage, instead of wrought-iron,
because they are very nearly as good, and much less
expensive.
PLATINUM CRUCIBLES. Platinum crucibles are invalu-
able in a laboratory. Few pieces of apparatus are used
so frequently by the chemist. Their chief use is in the
ignition of precipitates and the decomposition of siliceous
minerals by fusion with alkaline carbonates. They are
134 PLATINUM CRUCIBLES.
preferable to porcelain, as not being fragile and being
more readily heated to redness over the gas or spirit flame.
Their most convenient size is 1^ inch high and 1J inch
wide at the top.
In employing a crucible for the incineration of filters
in quantitative assays by the wet way, it sometimes happens
(as, for instance, with chloride of silver or sulphate of lead)
that the employment of platinum is inadmissible. In these
cases thin porcelain crucibles must be used. The analyst
will, however, frequently experience difficulty, owing to
the extreme slowness w^ith which, in many cases, the last
portions of the carbon of a filter are consumed when
ignited in a porcelain crucible. It does not appear, how-
ever, that the following simple method of obviating the
difficulty, as practised in the laboratory of Professor
Sheerer, in Freiburg, has ever received the publicity which
it deserves. Whenever a filter upon which a substance
capable of injuring platinum has been collected, has to be
incinerated, the porcelain crucible or capsule in which the
process is to be conducted should be placed within a vessel
of platinum of similar form, and the whole ignited in the
usual way. Whether the greatly accelerated rapidity of
combustion of the carbon which ensues depends upon a
more equal distribution of heat brought about by the
greater conducting power of the metal an explanation
which is current for the somewhat analogous case of copper-
coated glass flasks or whether, as seems probable, the
power of the porcelain vessel to absorb heat be really
increased by the interposition of the platinum ; whether
both these causes be of influence, or the result depends
upon another less apparent reason ; or, finally, whether
vessels of some other metal would not be preferable to
those of platinum, are questions which are open to dis-
cussion.
Fresenius gives the following excellent directions as to
the preservation of platinum crucibles. The analyst should ;
acquire the habit of cleaning and polishing the platinum
crucible always after using it. This should be done, as
recommended by Berzelius, by friction with moist sea-sand,
PLATINUM CRUCIBLES; 135
whose grains are all round, and do not scratch. The
writer has found this method to answer extremely well.
The sand is rubbed on with the finger, and the desired
effect is produced in a few minutes. The adoption of this
habit is attended with the pleasure of always working with
a bright crucible, and the profit of prolonging its existence.
This mode of cleaning is all the more necessary when one
ignites over gas-lamps, since at this high temperature
crucibles soon acquire a grey coating, which arises from
a superficial loosening of the platinum. A little scouring
with sea-sand readily removes the appearance in question,
without causing any notable diminution in the weight of
the crucible.
The ordinary Bunsen burner is known to act upon the
surface of platinum vessels brought into contact with the
inner line of the flame ; the metal loses its polish, becom-
ing superficially porous and spongy, and requires the use
of the burnisher to bring it back to its original state. This
alteration of the surface Mr. Dexter has found to be at-
tended with a change of weight, so that for some years he
has used a lamp of different construction for the heating
of platinum crucibles in analytical operations. Such a
lamp may be made by removing the air-tube of a common
Bunsen lamp, and putting in its place a somewhat longer
one of glass or iron of about 12 millimetres internal dia-
meter. The gas jet should have a single circular aperture,
and be in proper proportion to the diameter of the tube,
which may be held in any of the ordinary clamp supports.
The tube being raised sufficiently above the jet to allow
free entrance of air, and a full stream of gas let on, a ' roar-
ing ' flame is produced, of which the interior blue cone is
pointed, sharply defined, and extends only about half an
inch from the top of the tube. A polished platinum sur-
face is not acted upon by this flame, provided it be not
brought into contact with the interior cone. In the
Bunsen burner, as usually made, the supply of air depends
upon the diameter of the tube, the holes at its base being
more than sufficient to supply the draught. With the
wider tube it is necessary to limit the admission of air by
130 -PLATINUM CRUCIBLES.
depressing the tube upon the lamp when the force of
the gas is diminished. Otherwise the proportion becomes
such that an explosive mixture is formed ; for this reason
it is more convenient to use an arrangement in which the
access of air can be regulated by an exterior tube sliding
obliquely downward over the air-apertures. The gas jet
should be on a level with the top of these apertures, which
must be much larger than those of the ordinary Bunsen's
burner. On account of the liability to explode and burn
at the jet inside, the lamp is not well adapted for ordinary
use ; but for ignition of crucibles, working of glass, &c.,
it has proved efficient and practical.
In connection with some sensible remarks upon the
before-mentioned use of sand in cleaning platinum cru-
cibles, Erdmann explains in the following way the cause
of this grey coating which forms upon platinum crucibles
whenever they are ignited in the flame of Bunsen's gas-
burner. This coating has given rise to much annoyance
and solicitude among chemists. Indeed, it has often been
asserted that the use of Bunsen's burner is unadvisable in
quantitative analysis, since by means of it the weight of
platinum crucibles is altered and the crucibles themselves
injured. The coating is produced most rapidly when the
crucible is placed in the inner cone of the flame, and the
more readily in proportion as the pressure under which
the gas is burned is higher. Having found it advantageous
to maintain, by means of a special small gas-holder, a
pressure of four or five inches upon the gas used in his own
laboratory, Erdmann observed that the strong gas-flame
thus afforded immediately occasioned the formation of a
dull ring upon the polished metal placed in the inner
flame, this ring being especially conspicuous when the
crucible becomes red-hot ; it increased continually, so
that after long-continued ignition the whole of the bottom
of the crucible was found to be grey and with its lustre
dimmed.
The ring is caused neither by sulphur, as some have
believed, nor by a coating of inorganic matter, but is simply
a superficial loosening of the texture of the platinum,
PRESERVATION OP PLATINUM CRUCIBLES. 137
in consequence of the strong heat, whence it first of all
appears in the hottest part of the flame. In consequence
of the serious damage which the gas furnace causes, many
chemists now discard gas and ignite platinum crucibles
over specially constructed spirit lamps.
In conjunction with Pettenkofer, Erdmann instituted
several experiments, which have left but little doubt that
the phenomenon depends upon a molecular alteration of
the surface of the metal. If a weighed polished crucible
be ignited for a long time over Bunsen's lamp, the position
of the crucible being changed from time to time, in order-
that the greatest possible portion of its surface shall be
covered with the grey coating, and its weight be then de-
termined anew, it will be found that this has not increased.
The coating cannot be removed either by melting with
bisulphate of potash or with carbonate of soda. It dis-
appears, however, when the metal is polished with sand ;
the loss of weight which the crucible undergoes being
exceedingly insignificant, a crucible weighing 25 grammes
having lost hardly half a milligramme. When the grey
coating of the crucible is examined under the microscope,
it may be clearly seen that the metal has acquired a rough,
almost warty, surface, which disappears when it is polished
with sand. Platinum wires, which are frequently ignited
in the gas-flame for example, the triangles which are
used to support crucibles become, as it is known, grey
and brittle. Under the microscope they exhibit a mul-
titude of fine longitudinal cracks, which, as the original
superficial alteration penetrates deeper, become more open,
or, as it were, spongy, until finally the wire breaks.
If such wire is strongly and perseveringly rubbed with
sand, the cracks disappear, ana the wire becomes smooth
and polished ; for the grains of sand, acting like burnishers,
restore the original tenacity of the metal, very little of its
substance being rubbed off meanwhile. The loosening
effect of a strong heat upon metals is beautifully exhibited
when silver is ignited in the gas-flame, a thick polished
sheet of silver immediately becoming dull white when thus
heated. Under the microscope the metal appears swollen
138 CLEANING PLATINUM CRUCIBLES.
and warty. Where it has been exposed to the action of
the inner flame along its circumference, this warty con-
dition is visible to the naked eye. A stroke with the
burnishing stone, however, presses down the loosened par-
ticles, and reproduces the original polish. This peculiar
condition which the surface of silver assumes when it is
ignited, is well known to silversmiths ; it cannot be re-
placed by any etching with acids, and it must be remem-
bered that what is dull white in silver appears grey in
platinum.
If each commencement of this loosening is again de-
stroyed, the crucibles will be preserved unaltered, other-
wise they must gradually become brittle. Crucibles of
the alloy of platinum and iridium are altered, like those of
platinum, when they are ignited. It is, however, some-
what more difficult to reproduce the original polish of the
metal by means of sand, as might be expected, from the
greater hardness of the alloy.
The sand used should be well worn. When examined
under the microscope no grain of it should exhibit sharp
edges or corners ; all the angles should be obtuse.
If there are spots on the platinum crucibles which
cannot be removed by the sand without wearing away too
much of the metal, a little potassium bisulphate is fused
in the crucible, the fluid mass shaken about inside, allowed
to cool, and the crucible finally boiled with water. There
are two ways of cleaning crucibles soiled outside ; either
the crucible is placed in a larger one, and the interspace
filled with potassium bisulphate, which is then heated to
fusion, or the crucible is placed on a platinum-wire triangle
heated to redness, and then sprinkled over with powdered
potassium bisulphate. Instead of the bisulphate, borax
may be used. Never forget at last to polish the crucible
with sea-sand again.
A remarkably rapid and perfect method of cleaning
platinum apparatus consists in gently rubbing upon the
dirty metal a small lump of sodium-amalgam. Sodium has
the curious property of lending to mercury the power of
' wetting ' platinum in so complete a manner that the
CLEANING PLATINUM CRUCIBLES. 139
positive capillarity between platinum and an amalgam con-
taining even only one per cent, of sodium appears to be
as great as that between mercury and zinc, with this
important difference, however in the former case, the
4 wetted ' metal does not suffer the least trace of amalga-
mation. Even when foreign metals, such as lead, tin, zinc r
silver, are purposely added to the sodium -amalgam, the
platinum surface suffers no disintegration.
When the amalgam has been rubbed on with a cloth
until the whole surface is brilliantly metallic, water is ap-
plied which oxidises the sodium and allows the cohesion
of the mercury to assert itself. On wiping the mercury
off, the platinum surface is left in admirable condition for
the burnisher.
When the crucible is clean it is placed upon a clear
platinum-wire triangle, ignited, allowed to cool in the
desiccator, and weighed. This operation, though not in--,
dispensable, is still always advisable, that the weighing of;
the empty and the tilled crucible may be performed under
as nearly as possible the same circumstances.
In using platinum crucibles, it must be remembered
that certain substances must not be ignited in them.
Berzelius says that ' it is improper to ignite in platinum
vessels the caustic alkalies or the nitrates of any alkaline
base, such as lime, baryta, or strontia, because the affinity
of the alkali for platinum oxide causes a very considerable
oxidation of the metal ; and after the saline matter is re-
moved, the surface of the metal is found to be honey-
combed.'
The alkaline sulphides or the alkaline sulphates mixed
with charcoal are inadmissible, because the sulphides so-
formed attack platinum even more energetically than the
caustic alkalies ; so are metals whose fusing-point is lower
than that of platinum, because an alloy would be formed.
Gold, silver, and copper may be heated to dull redness in
platinurn vessels without danger ; but fused lead cannot
come in contact with platinum without destroying it. A
drop of fused lead, tin, zinc, or bismuth, placed on red-hot
platinum, always produces a hole. Neither can a phosphide
140 SILVER CRUCIBLES NICKEL CRUCIBLES.
or phosphoric acid mixed with charcoal be ignited in
vessels of platinum, because a platinum phosphide is pro-
duced, which is an exceedingly brittle compound.
In analyses by the wet method, nitro-hydrochloric
acid (aqua regia), even when very dilute, must not be
allowed to come in contact with platinum, and, as a
general rule, liquids containing either free chlorine, bro-
mine, or iodine must not be boiled in platinum capsules.
SILVER CRUCIBLES can only be used at temperatures
below full redness. They are not affected by caustic
alkalies, but must not come in contact with sulphur or
be heated over coke, coal, gas, or other fuel containing
sulphur.
NICKEL CRUCIBLES. Recently nickel has been intro-
duced as a material for crucibles. It is well known that
pure nickel is one of the toughest of all the metals,
.and that it fuses only at very high temperatures. It lias
a fine grain, takes a high polish, and is very compact
and unalterable. These qualities have led to its being
employed for crucibles and evaporating dishes. Mr.
Wanklyn has published some notes on the behaviour of
these vessels in his laboratory. He finds that for many
purposes crucibles of pure nickel are quite as serviceable
as platinum crucibles, and they are much cheaper, costing
only about one-tenth as much as platinum. They stand
the action of alkalies remarkably well there was no
alteration in the weight of the crucible after caustic pot-
ash had been fused in it. Hydrochloric acid in the cold,
whether dilute or concentrated, may be used to clean out
these crucibles and no alteration in weight is the result.
Cold oil of vitriol is likewise without action ; but con-
centrated nitric acid attacks them, causing rapid loss of
weight. Nickel dishes are especially useful for taking
water-residues and milk- solids, and indeed for these pur-
poses nickel is not inferior to platinum.
Mr. Bertram Blount considers that the uses of nickel
for chemical purposes are confined to
(i.) Dry ignitions (i.e. not fusions) in an oxidising
flame, provided the heat be not too intense.
CUPELS. 141
(ii.) Fusions with caustic alkalies.
From which may probably be deduced
(iii.) Fusions with barium hydrate.
(iv.) Fusions with alkaline carbonates.
The inability of nickel to stand high temperatures, in
contact with a reducing flame, limits its employment con-
siderably, and demands caution in dealing with it.
CUPELS. These are vessels in which the operation
termed cupellation is carried on. They must be made of
such substances as are not acted upon by certain fused
oxides, as those of lead or bismuth, and their texture has
to be sufficiently loose to allow of the oxides penetrating
their substance readily, and yet be sufficiently strong to
bear handling without breaking.
There are several substances of which cupels might be
made, which will fulfil all these conditions, but only one
is in general use, viz. the ash of burnt bones. This consists
principally of calcium phosphate, with a little calcium car-
bonate and fluoride. Berzelius found that bones of oxen
contained 57 parts of calcium phosphate for every 3'8
parts of calcium carbonate, whilst, according to Barros,
sheep bones contain 80 parts of calcium phosphate to 19
parts of calcium carbonate. When bones are burnt whole
they likewise contain mineral matter derived from the
cartilage, such as alkaline sulphates and carbonates. The
greater part of the calcium carbonate is likewise converted
into caustic lime.
The bones of sheep and horses are best for cupels. In
getting rid of the organic matter, it is advisable to boil
them repeatedly in water before burning them. This dis-
solves a great part of the organic matter. If the bones
are not rendered quite white by the first ignition, but con-
tain a little carbon, they should be ground up, moulded
into shape, and burned again.
Care should be taken not to heat the bone earth too
strongly. In this case the bones will have a smooth, glassy
fracture, and will not be sufficiently spongy or absorbent
to make good cupels.
When the bones are burnt white throughout, they must
142 MANUFACTURE OP CUPELS.
be finely ground, sifted, and washed several times with
boiling distilled water till all soluble salts are removed.
The finest particles of the powdered bone earth will remain
longest suspended in the washing waters. This must be
allowed to settle separately, and should be reserved for
giving a final coating to the surface of the cupels ; this
coating acts, to a certain extent, like a fine filter, and may
be applied to all cupels, although the body of the cupel is
made of different materials.
For the body of the cupels, the bone-ash should be
-about as fine as wheat flour. If too coarse, litharge con-
taining silver will be absorbed into its pores, and will occa-
sion a loss of silver.
Cupels must neither crack nor alter in texture at a
white heat. It is very important that they should not
contain carbon, and therefore, in making them, the bone
earth must not, as sometimes recommended, be mixed with
beer, or water containing adhesive substances. Nothing
but pure water should be used, and the mixture should be
just sufficiently moist to adhere strongly when well pressed,
but not so moist as to adhere to the finger or the mould
employed to fashion the cupels. The mould (fig. 62) con-
FIG. 62. sists of three pieces : one a ring, b, having a
conical opening ; another a pestle, a, having
a hemispherical end fitting the larger opening
of the ring ; and the third, c, a piece of turned
metal, into which b fits ; c serves to form an
even bottom to the cupel. In order to mould
the cupels, proceed as follows : Place the ring
on the lower piece e, and fill it with the com-
position : then place the pestle upon it, and
force it down as much as possible : by this means the
moistened bone ash will become hardened, and take the
form of the pestle ; the latter must then be driven as
much as possible by repeated blows from a hammer,
until quite home. The surface of the cupel may then
have sifted over it a little of the very fine levigated bone-
ash, and the pestle hammered again on it. It is then to
be turned lightly round, so as to smooth the inner sur-
MANUFACTURE OF CUPELS.
143
face of the cupel, and withdrawn : the cupel is removed
from the mould by a gentle pressure on the narrowest end.
When in this state, the cupel must be dried gently by a
stove ; and lastly, ignited in a muffle, to expel all moisture.
It is then ready for use.
There are two or three points to attend to in manufac-
turing the best cupels. First, the powdered bone-ash must
be of a certain degree of fineness ; secondly, the paste must
be neither too soft nor too dry ; and thirdly, the pressure
must be made with a certain degree of force. A coarse
powder, only slightly moistened and compressed, furnishes
cupels which are very porous, break on the least pressure,
and, as before mentioned, allow small globules of metal to
enter into their pores.
When, on the contrary, the powder is very fine, the
paste moist, and compressed strongly, the cupels have
much solidity, and are less porous ; the fine metal cannot
penetrate them, but the operation proceeds very slowly :
besides, the assay is likely to become dulled, and incapable
of proceeding without a much higher degree of tempera-
ture being employed.
Cupels for assaying silver bullion are sometimes made
of equal parts of bone-ash and beechwood-ash ; and for
assaying gold, 2 parts of beechwood-ash and 1 part of
bone-ash are used. The hemispherical cavity of both
these kinds are coated with the fine levigated powder of
bone-ash .
Beechwood-ash is preferred for the manufacture of
cupels on account of the larger proportion of phosphoric
acid it contains.
According to Hertwig, beechwood-ash contains in 100
parts :
Potassium carbonate
11-72
Sodium carbonate
12-37
Potassium sulphate
.
3-49
Calcium carbonate
49-54
Magnesium carbonat
:*
7-79
Calcium phosphate
3-32
Magnesium
2-92
Iron
0-76
Aluminium ,,
1-51
Manganese
Silica
1-59
2-46
144 PYROMETRY.
SCORIFIERS. A scorifier (fig. 63) is a vessel made much
in the shape of a cupel, but of crucible earth. The proper
use both of cupels and scorifiers will be explained under
the head of silver- assaying.
METHODS OF MEASURING THE HEAT OF FURNACES. As
much of the accuracy of an assay depends on the tempera-
ture at which it is made, and as the temperature required
FIG. 63.
varies with each metal, it is very desirable to possess some
means of ascertaining the heat of the furnace more accu-
rately than by the eye. Many persons have devised
instruments, called pyrometers, for this purpose ; the
earliest being those of Mr. Wedgwood and the late Pro-
fessor Daniel, of King's College.
We shall not give a description of Wedgwood's pyro-
meter, as its indications are inaccurate, from the fact that
the clay cylinders, whose contraction serves to measure
the temperature, will contract as much as by the long con-
tinuance of a low heat as by the short continuance of a
high one. Hence the degrees of heat measured by Wedg-
wood's pyrometer have been enormously exaggerated. It
was long since noticed that it did not produce comparable
effects; and this was supposed to proceed wholly from
the impossibility of obtaining clay perfectly alike for each
experiment.
Daniel's pyrometer is composed of a rod of platinum
simply laid in a groove made of refractory clay, and baked
in the highest degree of heat. This rod rests at one end
DANIELS PYROMETER.
145
on the edge which terminates the groove, and at the other
on a lever with two arms, the larger of which forms a
needle on a graduated arc of a circle ; so that the removal
of this needle from its position marks the additional length
which the metal acquires by the heat. The following is
Daniel's description of his pyrometer : ' It consists of two
parts (see fig. 64), which may be distinguished as the
The register is a solid bar of black-
register and the scale.
FIG. 64.
J
lead or earthenware highly baked. In this a hole is drilled,
into which a bar of any metal, a, six inches long, may be
dropped, and which will then rest upon its solid end. A
cylindrical piece of porcelain, , called the index, is then
placed upon the top of the bar, and confined in its place
by a ring or strap of platinum passing round the top of
the register, which is partly cut away at the top, and
tightened by a wedge of porcelain. When such an arrange-
ment is exposed to a high temperature, it is obvious that
the expansion of the metallic bar will force the index for-
ward to the amount of the excess of its expansion over that
of the black-lead, and that when again cool it will be left
at the point of greatest elongation. What is now required
is the measurement of the distance which the index has
been thrust forward from its first position, and this,
L
146 DANIEL S PYROMETER.
though in any case but small, may be effected with great
precision by means of the scale c.' *
This is independent of the register, and consists of two
rules of brass accurately joined together at a right angle
by their edges, and fitting square upon the two sides of
the black-lead bar. At one end of this double rule a small
plate of brass projects at a right angle, which may be
brought down upon the shoulder of the register formed by
the notch cut away for the reception of the index. A
movable arm is attached to this frame, turning at its fixed
extremity on a centre, and at its other carrying the arc of
a circle, whose radius is exactly five inches, accurately
divided into degrees, and thirds of a degree. Upon this
arm, at the centre of the circle, another lighter arm is made
to turn, one end of which carries a vernier with it, which
moves upon the face of the arc, and subdivides the former
graduation into minutes of a degree ; the other end crosses
the centre, and terminates in an obtuse steel point, turned
inwards at a right angle. When an observation is to be
made, a bar of platinum or malleable iron is placed in the
cavity of the register ; the index is to be pressed down upon
it, and firmly fixed in its place by the platinum strap and
porcelain wedge. The scale is then to be applied by care-
fully adjusting the brass rule to the sides of the register,
and fixing it by pressing the cross piece upon the shoulder,
and placing the movable arm so that the steel part of the
radius may drop into a small cavity made for its reception,
and coinciding with the axis of the metallic bar. The minute
of the degree must then be noted which the vernier indi-
cates upon the arc. A similar observation must be made
after the register has been exposed to the increased tem-
perature which it is designed to measure, and again cooled,
and it will be found that the vernier has been moved
forward a certain number of degrees or minutes. The
scale of this pyrometer is readily connected with that of
the thermometer by immersing the register in boiling
mercury, whose temperature is as constant as that of boil-
ing water, and has been accurately determined by the
* Daniel's ' Chemical Philosophy,' p. 111.
DANIEL S PYROMETER,
147
thermometer. The amount of expansion for a known
number of degrees is thus determined, and the value of
all other expansions may be considered as proportionate.
By Daniel's pyrometer the melting-point of cast iron
has been ascertained to be 2,786, and the highest tem-
perature of a good wind furnace 3,300 Fahrenheit
points which were estimated by Mr. Wedgwood at 17,977
and 21,877 respectively.
The following is a list of the melting-points of some of
the metals as ascertained by Professor Daniel ; and it is
obvious that in an assay of each particular metal the tem-
perature employed must exceed by a considerable number
of degrees its melting-point. The table is, therefore, very
useful.
Fahr.
Tin melts at - - - - - - - 4-22
Cadmium
442
Bismuth
497
Lead .
612
Zinc .
773
Silver
1860
Copper
1996
Gold .
2016
Cast iron
2786
Cobalt and nickel are rather less fusible than iron.
Mr. S. Wilson* has described an ingenious process of
measuring high temperatures. He exposes a given weight
of platinum or Stourbridge clay to the action of the heat
which is to be measured, and then quenches it in a definite
weight of water at a certain temperature. Thus, if the
piece of platinum weigh 1,000 grains and the water 2,000
grains at 60 F., and should the heated platinum when
dropped into the water raise its temperature to 90, then
90- 60 =30 ; which, multiplied by 2 (because the weight
of the water is twice that of the platinum), gives 60
the temperature to which a weight of water equal to the
platinum would have been raised. To convert this into
Fahrenheit degrees we must multiply by 31J, which is the
specific heat of water as compared with platinum, that of
the latter being 1. Therefore 60 x 31 J 1875, which
will be the temperature of the furnace.
Philosophical Magazine,' ser. iv. vol. iv. p. 157.
L2
148
COMPARISON OF HIGH TEMPERATURES.
FIG. 65.
One or two other methods of measuring high tempera-
tures applicable to special cases may here be mentioned.
Mr. C. W. Siemens, C.E., F.E.S., has invented an inge-
nious pyrometer, the principle of which is, that as the
electrical conductivity of platinum, iron, and
other metals decreases as they rise in tem-
perature, their increase of resistance to the
passage of the current is a measure of the heat
to which the metals are subjected.
The principle of construction may be ex-
plained by the aid of fig. 65, in which F A B
is a tube of pipe-clay, and the length between
the projections A and B has a screw-shaped
spiral groove cut on its outer surface ; the
length of this part of the tube is about 3 inches.
A spiral of fine platinum wire lies in the groove,
each turn of the platinum spiral being thus pro-
tected from lying in contact with its neighbour
by the projecting edges of the groove, by which
plan of insulation the current is forced to pass
through the whole length of the fine wire. D
is a little platinum clam, connected with one
pole of the battery, and the position of this
clam on the spiral regulates the length of pla-
tinum wire through which the current shall
pass. By this plan of adjustment all the pyro-
meters constructed by Mr. Siemens are made
to agree with each other.
At F the ends of the thin platinum wire are connected
with very thick platinum wire, and higher up, near E r
where the heat of the furnace is less felt, the thick
platinum wires are connected with thick copper wires,
shown at P ; from E to F these connecting wires are pro-
tected by clay pipes, as shown in the cut.
When this arrangement has to be used, the whole of it
is dropped into a thick metal pipe made of iron, copper,
or platinum, according to the heat of the furnace to be
tested. The lower end of this outer pipe is shown at KM,
and when it is used the spiral A B lies inside it at N M. At
COMPARISON OF HIGH TEMPERATURES. 149
R there is a very thick collar of metal in which the heat
accumulates, and this prevents the cooling action of the
length K R (most of which does not enter the furnace) from
interfering with the accuracy of the indications. The
ends of the wires p are connected with a suitable and very
delicate electrical apparatus, by which the increasing elec-
trical resistance of the hot spiral is measured.
A good plan for comparing the temperatures of two
furnaces is to prepare alloys of platinum and gold, con-
taining definite quantities, say 5, 10, 15, 20 per cent. &c.,
of gold. These fuse at intermediate temperatures between
gold and platinum. By placing small angular chips of
these alloys separately in muffles, and noticing which are
melted, which softened only, and which resist the action
of the heat, an idea of the power of the furnace is obtained.
In this way the amount of heat required to perform any
operation may be registered for future reference, by sim-
ply recording that it was sufficient just to melt, say, a 20
gold 80 platinum alloy.
150
CHAPTEE V.
FUEL : ITS ASSAY AND ANALYSIS.
BEFORE treating of the assay of metals and metalliferous-
ores, it is advisable to devote some space to the important
subject of fuel. The substances employed as fuel, although
all of vegetable origin, are derived either from the veget-
able kingdom (wood), or from the mineral kingdom (peat,
brown coal, coal, anthracite). These natural fuels can
be converted into artificial fuels by heating them more or
less out of contact with the air (charcoal, turf-charcoal,
coke).
The essential elements of combustible matters are car-
bon, oxygen, and hydrogen ; nitrogen being present some-
times, but only in small proportions. These constitute
the organic part ; various salts and silica constitute the
inorganic part, or ash. The valuable constituents of fuel,
on which its calorific and reducing powers depend, are
the carbon and hydrogen, and it is upon the combustion
or union of these elements with oxygen to form carbonic
acid and water that the heating effect of the fuel depends.
The more oxygen a fuel contains, the less carbon and
combustible gases it will yield, and the more hydrogen,
the more combustible gases.
The proportion of hydrogen to oxygen in wood . is 1
turf . 1
fossil wood 1
coal . 1
anthracite 1
6
4
2-3
1
The more oxygen, the less carbon the fuel contains,
thus :
Anthracite contains about 90 per cent, carbon
Coal 80
Brown coal 70
Fossil wood and turf 60
Wood 60
ASSAY OF FUEL. 151
The more carbon a fuel contains, the greater heat it pro-
duces, and the more difficult it is to ignite.
The greater the amount of hydrogen in a fuel, the
more inflammable it will be, and the larger flame it gives,
the hydrogen being evolved below a red heat. But the more
carbon present the less flame. These differences are shown
in a blazing fire and a glowing fire. In a flame the hottest
part is at the periphery, whilst in a glowing fire the greatest
heat is in the immediate contact of the burning surface.
An elementary analysis of coal teaches little with
regard to the nature or practical value of the combustible.
A proximate analysis, on the contrary, enables us to learn
something in regard to the real nature of the coal. The
moisture and ash are not only diluents of the fuel, but
are in themselves obstacles to its effectiveness ; the vapo-
risation of the moisture causes a serious loss of heat, whilst
the ashes, by hindering complete combustion and by the
heat they contain when dropped through the grate, con-
stitute another loss. By furthermore determining the
total amount of volatile matter we learn both the per-
centage of coke in the fuel and the amount of carbon
(fixed combustible) and bitumen (volatile combustible
matter). Although neither of these two products can
be considered as simple chemical . compounds, it is never-
theless of the utmost practical importance to know these
two quantities, because of the great value of coke and gas
in manufactures.
The assay of fuel comprises the following examina-
tions :
1. The examination of the external appearance of the
fuel, its porosity or compactness, its fracture, the size and
shape of the pieces composing it.
2. Estimation of the adhering water.
3. Estimation of the specific gravity.
4. Estimation of the absolute heating power.
5. Estimation of the specific heating power.
6. Estimation of the pyrometric heating power.
7. Estimation of the volatile products of carbonisa-
tion.
:152 ASSAY OF FUEL.
8. Examination of the coke or charcoal left behind on
carbonisation, both with regard to quality and quantity.
9. Estimation of the amount of ash, and its composi-
tion.
10. Estimation of the amount of sulphur.
11. Examination of any other peculiarity which may
be noticed during the burning or carbonisation of the
fuel.
1. EXTERNAL APPEARANCE OF THE FUEL, ITS POROSITY,
COMPACTNESS, FRACTURE, SIZE, AND SHAPE OF PIECES. From
the outward appearance of a fuel, its cleavage, and an
examination of the embedded earthy matter, iron pyrites,
gypsum, &c., its applicability to any special purpose may
be judged. Its degree of inflammability, together with
the pressure of blast which it will bear in the furnace,
partly depend on the more or less compactness of the fuel.
iThe amount of loss which it will suffer in transport de-
pends upon its friability. Playfair and De la Beche *
estimated the amount of this loss in coal by rotating
in a barrel different qualities of coal for the same time.
The powder produced was separated and weighed, and in
this way the friability or cohesion of a fuel could be ex-
pressed in percentages. Schrotter f made the same experi-
ments with brown coal.
The size and form of the pieces composing the fuel is
important, as on this depends the space occupied in its
stowage an important point for steam-vessels. This
space cannot be calculated from its specific gravity, but
must be ascertained by direct measurement. The space
occupied will be smallest when the form of the lumps is
cubical.
2. ESTIMATION OF THE ADHERING WATER. The water
contained in a fuel exerts great influence on its heating
power. It not only increases its bulk, but it acts in-
juriously by abstracting a certain quantity of heat required
for its evaporation, and it also causes imperfect combus-
* Dingl. ex. 212, 262; cxiv. 346. Liebig's * Jahresber.,' 1847-1848,
p. 1117; 1849, p. 708.
t Wien. Akad. Ber. 1849, Nov. and Dec. p. 240. Liebig's ' Jahresber.,
1849, p. 709.
ASSAY OF FUEL. 15U
tion. For this reason, wood, turf, and brown coal never
give so high a temperature as coal, anthracite, and coke.
The estimation of the adhering water is effected by
drying a certain weight of the pounded fuel in a water-
bath at 212 F., or in an air-bath at 220. It may also be
ascertained by placing a certain weight of the powdered
fuel in a glass tube, heating to 212, and passing over it
air dried by means of chloride of calcium, till the fuel no
longer loses weight. The amount of water which the
dried fuel will absorb from the atmosphere in twenty-four
hours should also be estimated, in order to ascertain its
hygroscopic qualities.
3. ESTIMATION OF THE SPECIFIC GRAVITY. The specific
gravity of a fuel depends on its density and the amount
of ash, and it appears also to be in proportion to its greater
or less inflammability. Of two equal volumes of carbonised
fuel, the one will produce the greatest heating effect which
has the greatest specific gravity, provided the density is
not produced by mineral constituents.
The estimation of the specific gravity is difficult, and
sometimes uncertain, owing to the cleavage of the fuel,
and the entanglement of air in its pores. The best way
of obviating this difficulty is as follows :
Coarse fragments, freed by means of a sieve from all
small particles, and averaging l-10th c.c. in volume, are
introduced into a fifty-gramme flask provided with a
thermometer stopper. The constants for this flask for
temperatures varying from 50 to 80 F. are previously
carefully estimated.
The true sp. gr. corresponds to the coal perfectly
soaked, so that all its pores are filled with water. That
requires, on the average, 12 hours, permitting two es-
timations per day, one in the morning, another in the
evening.
That this precaution is important may be seen from
the following example : a sample of coal gave the sp. gr.
1-309 at 64 F., immediately after filling the flask with
water : after about 12 hours' soaking, the sp. gr. had
increased to 1-328, for the same temperature. According
154 ASSAY OF FUEL.
to this latter estimation, a cubic foot of this coal would
weigh 82-76 Ibs. ; according to the former, only 81*58, or
1-18 Ibs. less. This shows a considerable degree of po-
rosity of the coal, and indicates the absurdity of giving
the weight in pounds of a cubic foot of coal with four
decimals, although no statement in regard to temperature
or time of weighing is made.
4. ESTIMATION OF THE ABSOLUTE HEATING POWER. The
value of a fuel for any purpose depends chiefly on its
price and the quantity required for that purpose. The
quantity required depends on the heating power possessed
by a certain weight of fuel (its absolute heating power) or
that possessed by a certain volume (its specific heating
power).
The less oxygen, ash, and water the fuel contains, the
greater its heating power will be, and this will also increase
in proportion to the carbon and hydrogen present.
Whether the combustion is effected quickly or slowly,
the amount of heat produced will be the same, but the
degree of temperature attained will be very difficult. This
latter constitutes the pyrometric heating power.
The estimation of the absolute heating power of a fuel
may be effected
a. By heating a definite quantity of water from 32 F.
to 212 ;
. b. By ascertaining how much fuel is required to melt
a known weight of ice ;
c. By ascertaining how much water may be evaporated
by 1 Ib. of different kinds of fuel ;
d. By ascertaining how much the temperature of a
room increases by burning a certain weight of a fuel in a
stove.
e. By ascertaining the elementary composition of the
fuel, and calculating how much oxygen will be required
to convert the carbon and hydrogen into carbonic acid
and. water ; the quantity of heat produced will be in pro-
portion to the amount of oxygen consumed.
/. By Berthier's method.
g. By lire's method.
BERTHIER'S METHOD FOE ASSAY OF FUEL. 155
According to Berthier, the most convenient method
for ascertaining the comparative calorific power of any
combustible matter is by means of litharge. He says : It
has been proved by the experiments of many philosophers
that the quantities of heat emitted by combustible sub-
stances are exactly proportioned to the amounts of oxygen
required for their complete combustion. Whence, after
the elementary constitution of any combustible is known,
its calorific power is easily estimated by calculation.
For instance, it is only necessary to ascertain the quantity
of oxygen absorbed in the conversion of all its carbon
into carbonic acid, and all its hydrogen into water, and
compare that quantity with that which is consumed in
burning a fuel whose calorific power is well ascertained.
Such a fuel is pure charcoal.
By adopting the principle just pointed out, it may be
conceived that, without knowing the composition of a
fuel, its heating power may be ascertained by estimating
the amount of oxygen it absorbs in burning. This can be
done in a very simple and expeditious manner, if not
exactly, at least with sufficient exactitude to afford very
useful results in practice. It is as follows : many metallic
oxides are reduced with such facility that when heated
with a combustible body, the latter burns completely,
without any of its elements escaping the action of the
oxygen of the oxide, if the operation be suitably per-
formed. The composition of the oxide being well known>
if the weight of the part reduced to the metallic state be
taken, the quantity of oxygen employed in the combustion
can be ascertained. In order to collect the metal and
separate it from the non-reduced mass, it, as well as its
oxide, must be fusible. Litharge fulfils these conditions, and
experiment has proved that it completely burns the greater
part of all ordinary fuels ; the only exceptions are some
very bituminous matters containing a large proportion of
volatile elements, a portion of which escapes before the
temperature is sufficiently high to allow the reduction to
take place. The experiment is made as follows : 10 grains
of the finely powdered or otherwise divided fuel is mixed
156 BERTHIER'S METHOD FOR ASSAY OF FUEL.
with about 400 grains of litharge. The mixture is care-
fully placed in an earthen crucible, and covered with 200
grains more litharge. The crucible is then placed in the
fire and gradually heated. When the fusion is perfect, the
heat is urged for about ten minutes, in order that all the
lead may collect into a single button. The crucible is
then taken from the fire, cooled, broken, and the button
of lead weighed. Sometimes the button is livid, leafy,
and only slightly ductile ; in/ which case it has absorbed
a little litharge. This can be partially prevented by fusing
slowly, and adding a little borax.
Two assays, at least, ought to be made, and those re-
sults which differ more than a grain or two ought not to
be relied on. The purer the litharge, the better the re-
sult ; it ought to contain as little minium as possible. It
is an excellent plan to mix up the litharge of commerce
with one or two thousandths of its weight of charcoal, and
fuse the whole in a pot ; when cold, pulverise the litharge,
which will now be deprived of minium.
Pure carbon produces, with pure litharge, 34 times
its weight of lead, whilst hydrogen gives 103 '7 times its
weight of lead ; that is to say, a little more than three
times as much carbon. We can, therefore, from these
data, find the equivalent of any fuel, either in carbon or
hydrogen.
When a fuel contains volatile matters, the quantity
can be ascertained, as before pointed out, by ignition in a
close tube or crucible. If, further, we ascertain the pro-
portion of lead it gives with litharge, it is easy to calculate
the equivalent in carbon of the volatile matters, and, in
consequence, to ascertain its calorific value.
Supposing that a substance gives by distillation C parts
of coke, or carbon, having deducted the weight of the ash
and of volatile substances, and that it produces P parts of
lead with litharge. The quantity C of carbon would give
34 x C of lead ; the quantity of volatile matter would give
but P 34 x C ; it would be equivalent to * |^ of carbon :
whence it follows that the quantity of heat developed by
the charcoal, the volatile matter, and the unaltered com-
DR. URK ON BERTH IERS METHOD. 157
bnstihle, will be to each other as the numbers 34x0,.
P-34xC, and P.
Dr. Ure* says, speaking of the above method of assay,.
' On subjecting this theory to the touchstone of experi-
ment, I have found it to be entirely fallacious. Having
mixed very intimately 10 grains of recently calcined char-
coal with 1,000 parts of litharge, both in line powder, I
placed the mixture in a crucible, which was so carefully
covered as to be protected from all fuliginous fumes, and
exposed it to distinct ignition.
' No less than 603 grains of lead were obtained, whereas,
by Berthier's rule, only 340 or 346*6 were possible. On
igniting a mixture of 10 grains of pulverised anthracite
with 500 grains of pure litharge previously fused and
pulverised, I obtained 380 grains of metallic lead. In a
second experiment, with the same anthracite and the same
litharge, I obtained 450 grains of lead ; and in a third,,
only 350 grains. It is therefore obvious that this method
of Berthier's is altogether nugatory for ascertaining the
quantity of carbon in coals, and is worse than useless
in judging of the calorific qualities of different kinds of
fuel.'
This discrepancy in the results obtained by Dr. Ure is
very perplexing, and does not at all accord with Berthier's
experience, as shown by his experiments, or by the author's
on the subject. The latter never had a difference of more
than 50 grains, and in general only two or three, which
latter result is satisfactory. The only precaution he found
necessary was to heat very gradually until the mixture
was fully fused, and then to increase the fire to bright
redness for a few minutes.
Further experiments have been made by the author
on this subject, and he has succeeded most perfectly in
estimating the value of a fuel. With the litharge of com-
merce, which contains much minium, the process is never
exact : results have been obtained differing as much as 40
or 50 grains when the litharge employed had not been
purified, and to purify it completely is a troublesome
* ' Supplement to the Dictionary of Arts, Mines, and Manufactures.'
158
UEE'S CALORIMETER,
process. This difficulty may be completely obviated,
however, by substituting for litharge, white-lead, using
for each 10 grains of fuel 700 grains of white-lead, which
FIG. 66.
DIMENSIONS.
A B=3J in. ; diameter f in.
C D=2 in.
C B, socket for AB.
Diameter across D=4 in.
E F=6 in.
F G=5 in. ; diameter H G = l| in.
A B weighs 39 grammes.
The remainder of the apparatus, including
the stopcock, weighs 391 grammes.
SCALE OF 12 INCHES.
12
are well mixed with it, and 300 grains of pure white-lead
to cover the mixture.
When the whole is heated, the carbonate of lead de-
composes, forming pure lead oxide, which is then reduced,
as in the former case. By this process the results corre-
spond to 1 grain in the quantity of lead produced from a
WRIGHT'S CALORIMETER. 159
given sample of fuel. Of course great care must be taken
that the white-lead is genuine.
Commercial samples are frequently adulterated with
lead sulphate and barium sulphate, lead oxychloride, zinc
oxide, &c. This is a serious drawback to this otherwise
excellent modification.
WRIGHT'S CALORIMETER. The instrument known as
Wright's calorimeter gives very accurate results, and is
the one most generally used now in experiments on the
heating power of fuel, in all but the most refined investi-
gations. It is shown in the accompanying figure (fig. 66).
The copper cylinder A B is filled with a mixture of 20
grains of the combustible, and 240 of the deflagrating com-
pound, which is composed of three parts of potassium
chlorate and one of potassium nitrate. A little piece of
cotton soaked in potassium chlorate is placed partly in the
mixture, the other end projecting above the top of the
cylinder ; this is ignited, quickly covered with the bell-
shaped part of the apparatus, and immersed in a measured
quantity of water. As constructed, the whole metallic
apparatus weighs 6,642'7 grains, and with this weight
290-1 grains of water are used. The temperature is re-
corded before and after making the experiment. During
the deflagration the stopcock is closed ; it is, however,
opened before taking the temperature the second time. A
tenth of the temperature that the water is raised by the
combustion is added for errors that are incidental to the
use of the instrument.
If the instrument is made of the weight above given,
the result is obtained by a very simple calculation. Each
Fahrenheit degree by which the temperature of the water
has been augmented corresponds to a pound of water
converted into steam.
EXAMPLE.
Fahr.
Temp, of water before making experiment = 56
after the combustion . = 65
~9 + Ath
produced by the combustion = 9'1
One pound of the coal will convert into steam (maximum effect) 9'1 Ibs. of
water at 212 F.
100 PYROMETKIC EXAMINATION OF FUEL.
5. ESTIMATION OF THE SPECIFIC HEATING POWER. This
represents the heat produced from a certain volume of
fuel. It may be ascertained by multiplying the absolute
heating power by the specific gravity.
6. ESTIMATION OF THE PYROMETRIC HEATING POWER. By
pyrometric heating power is meant the degree of tem-
perature which may be obtained by completely burning
the fuel. This heating power not only depends upon the
composition of the fuel, but chiefly on the time required
for its combustion, and this again depends on the looseness
and inflammability of the fuel. The absolute heating
power of hydrogen is greater than that of carbon, but with
regard to the pyrometric heating power it will be found
that the reverse is the case.
Carbon burned in contact with the air to carbonic
acid will produce a heat of 2,558 C. ; if burned to car-
bonic oxide it only produces 1,310 ; hydrogen burning to
water will produce a heat of 2,080. From this we learn
that fuel rich in carbon, such as anthracite, coal, and coke,
will produce a greater pyrometric effect than fuel rich in
hydrogen, as wood, &c.
Density is an essential quality of fuel required to pro-
duce great pyrometric effect. This is proved in the follow-
ing way.
When atmospheric air first acts on the carbon con-
tained in fuel, carbonic acid is formed, and the tempera-
ture rises to a certain degree, but on passing over glowing
coal, carbonic acid becomes converted into carbonic oxide,
and this causes a portion of the heat at first produced to
become latent. This conversion of carbonic acid into car-
bonic oxide is more easy and complete as the fuel used is
more inflammable ; and as v a greater quantity of heat is
thereby rendered latent, it follows that the heating power
of such a fuel is inferior. This accords with general ex-
perience ; for it is well known that coke is able to produce
a greater heat than charcoal.
Good methods for estimating pyrometric heating power
were given in the last chapter.
7. ESTIMATION OF THE VOLATILE PRODUCTS OF CARBONI-
ASSAY OF FUEL. 161
SATION. The amount of volatile matter yielded on car-
bonising a fuel depends partly on the composition of the
fuel, and partly on the temperature employed. If a fuel
rich in oxygen and hydrogen is quickly heated, it will yield
the greatest amount of volatile products. These are partly
liquid (tar, naphtha, and acetic acid or ammoniacal water),
and partly gaseous (carbonic oxide, carbonic acid, and
light and heavy carburet ted hydrogen). The more oxy-
gen a fuel contains, the more carbonic acid and carbonic
oxide it will produce ; the more hydrogen it contains, the
more illuminating gas it yields. The applicability of a
sample of coal to the production of illuminating gas
depends on these conditions.
Coal distilled at a low temperature yields much tar and
comparatively little gas, and when a very high temperature
has been used, less tar and more gas is produced, but the
great heat will have reacted on the gas and injured its
illuminating qualities. If the coal contains pyrites, the
gas will contain sulphur compounds. The amount of
water produced is generally larger than that of the tar.
In order to estimate the amount of volatile matter given
off from any particular sample of coal, proceed in the fol-
lowing manner : Place a given weight, say 200 grains, of
the coal in an iron tube closed at one end, to the other
end of which adapt, by means of a cork, a glass or other
tube, which must be led into an inverted jar full of
water standing in the pneumatic trough. Eaise the tem-
perature very gradually to redness, and continue the heat
until no more gas is given off, then ascertain its quantity
in cubic inches, with due correction for temperature and
pressure.
8. EXAMINATION OF THE COKE OR CHARCOAL LEFT BEHIND ON
CARBONISATION. The amount of coke or charcoal yielded
by a sample of fuel is found by the last operation. This
residue is the amount of coke which that particular sample
of coal produces ; and its weight, divided by two, gives
the percentage of coke.
The process of coking, charring, or carbonising fuel,
whilst it drives off some of the valuable hydrocarbon con-
M
162 EXAMINATION OF COKE OF FUEL.
stituents, also gets rid of all the aqueous elements. And
therefore the coke or charcoal which is left behind has
its value greatly increased when high temperatures are
required, although, from the absence of flame-yielding
constituents, it is much more difficult to ignite.
The degree of inflammability of coke or charcoal is
relatively the same as that of the raw fuel from which they
were produced. The more inflammable a fuel has been,
the more inflammable will be the coke or charcoal pro-
duced from it.
The temperature employed in the carbonisation, as has
been already explained, exerts great influence on the yield
of coke.
If the fuel contains iron pyrites, part of the sulphur
goes off in the volatile portion, but from one-fourth to
one-half is retained in the form of iron sulphide.
9. ESTIMATION OF THE AMOUNT OF ASH. In order to
ascertain the amount of ash : Fully ignite about 50 grains
of coal in a platinum capsule, allowing the air to have
free access all the time until nothing but ash is left. Its
amount may then be ascertained by weighing : good fuel
should contain little ash. It may vary from 1 to 10 per
cent., but if it exceeds 5 per cent, it becomes deleterious.
The chemical composition of the ash also influences the
quality of the fuel to some extent.
10. ESTIMATION OF THE AMOUNT OF SULPHUR. This is
an important operation in the assay, as a coal containing
sulphur cannot be employed for particular operations, and,
indeed, those which contain much sulphur ought only to
be used for the commonest purposes. This assay is most
important to ironmasters as well as to steamboat and other
companies, who consume fuel under steam boilers ; and
the coal they purchase should always be subjected to this
particular test, as sulphur has a corroding and destroying
action on iron and copper. Where sulphurous coals are
continually burnt under boilers, the metal of the latter
becomes deteriorated, and the boiler is rapidly rendered
useless. Sulphur exists in coal in the form of iron pyrites ;
this can generally be detected by its brassy colour. Some
DETERMINATION OF SULPHUR IN FUEL. 1G:>>
coals and lignites also contain calcium sulphate, and in
rare cases barium sulphate.
The process for the estimation of the amount of
sulphur in coal is not difficult. 1 part of the coal is to be
finely pulverised, and then mixed with 7 or 8 parts of
potassium nitrate, 16 parts of common salt, and four parts
of potassium carbonate, all of which must be perfectly
pure ; the mixture is then placed in a platinum crucible
and gently heated. At a certain temperature the whole
ignites and burns quietly. The heat is then increased until
the mass is fused : the operation is finished when the mass
is white. It must, when cold, be dissolved in water, the
solution slightly acidulated by means of hydrochloric acid,
and barium chloride added to it as long as a white preci-
pitate forms. This precipitate is barium sulphate, which
must be collected on a filter, washed, dried, ignited, the
filter burnt away, and the remaining barium sulphate
weighed : every 116 parts of it indicate 16 of sulphur.
Dr. Price has drawn attention to a source of error
which lias hitherto escaped notice in the estimation of sul-
phur, where fusion of the substance with nitre is the pro-
cess employed. This author has found that unless great
care be taken to prevent the fused mass passing over to
the outside of the vessel, and so coming in contact with
the flame or products of combustion, an appreciable and,
in some cases, serious error will arise, owing to the sul- .
phuric acid produced from the sulphurous acid in the
flame a product of the oxidation of the sulphide of car-
bon contained in the gas combining with the potassium
of the fused salt. Several experiments have been made to
ascertain the amount of error that may be occasioned from
the above cause. In one instance the flame issuing from
a Bunsen's burner was made to strike against a little fused
nitre on the under side of a small platinum dish, when, in
three quarters of an hour, as much sulphuric acid was
obtained as is equivalent to 12 milligrammes of sulphur.
As a check on these experiments, nitre was fused by the
flame of the spirit lamp, when, as was to be anticipated,
not a trace of sulphuric acid could be detected upon the
IT 2
164 DETERMINATION OF SULPHUR IN FUEL.
addition of a barium salt to the aqueous solution of this
fused mass, rendered acid by hydrochloric acid. In es-
timations of sulphur in coke or coal, great care should,
therefore, be taken to prevent any of the fused saline con-
tents of the crucible from getting on to the outside. In
fusing pig iron with nitre, a process recommended by some
for the estimation of the sulphur it contains, the mass,
especially if the iron be rich in manganese, invariably
creeps over to the outer wall of the crucible ; and it is,
therefore, impossible to obtain correct results when the
operation is conducted over the gas flame. The assayer
should for these reasons always employ a spirit flame in
preference to gas in sulphur estimations.
Mr. Teikichi Makamura recommends the following pro-
cedure : 3 or 4 parts of the mixed alkaline carbonates, or
of sodium carbonate, are intimately mixed with one part of
coal in very fine powder in a large platinum dish. The
mixture is heated at first very gently, a spirit lamp being
used instead of a Bunsen, to prevent possible absorption
of sulphur ; the heat is then raised slowly without attain-
ing that of visible redness, until the surface becomes
only faintly grey. No smoke or odorous gases should
escape during the whole of the oxidation. The tem-
perature is now raised to a faintly red heat for sixty
minutes, at the end of which time the mass is perfectly
white, or reddish if iron be present. The mass is not
to be stirred during the ignition. The residue is heated
with water, filtered, and the sulphates estimated in the
ordinary way.
11. EXAMINATION OF OTHER PECULIARITIES OF FUEL.
Besides the above-named examinations, the assayer should
notice the degree of inflammability of the fuel, and whether
any particular smell is evolved during combustion ; whether
the coal is good for coking purposes ; whether it burns
with a large smoky flame or a luminous flame ; whether it
burns quietly or with decrepitation ; and whether the ash
is dusty or fusible, and likely to accumulate and clog up
the grate-bars.
32. CALCULATION OF RESULTS. It may be sufficient here
ASSAY OF FUEL. 165
to state that, beside the percentage composition of the
-coal, it is proper to reduce the composition to the com-
bustible =100, in order to obtain a comparative estimate
of the character of the fuel itself (in regard to the propor-
tion of bitumen and carbon), and of the amount and
quality of the impurities (ash and moisture). It has also
been shown that, for considerable areas of the coal-field,
the sum of the constituents on the scale of combustible =
100 is the proper calorific equivalent, and that the per-
centage of the combustible in the fuel gives a proper
estimate of its value.
ASSAY OP COAL BEFORE THE BLOWPIPE. The blowpipe
method is well adapted to the assaying of .coal. Not only
does the portableness of the apparatus make it very con-
venient for use away from home, wherever the balance
can be set up ; but its use at home is quite as satisfactory
on the score of exactness as the assay with the muffle or
retort, or large platinum crucible, and large balance. Mr.
B. S. Lyman gives the following directions for carrying
out this assay :
Besides the ordinary pieces of the blowpipe apparatus,
as made at Freiburg, all that needs to be made expressly
for the coal assay is a small covered platinum crucible of
the same size and shape as the clay crucibles of that appa-
ratus ; and there must be a little ring of German silver,
for the crucible to stand on, about three-eighths of an inch
across and half that in height. Such a crucible cover and
ring weigh about 40 grains more than the ordinary
metallic cup that rests on the pan of the balance ; the
crucible and ring without the cover weigh less than 30
grains more than the cup. If it be desired to estimate the
amount of hygroscopic moisture in the coal, a small dry-
ing bath must be made too ; but the hygroscopic water
in ordinarily well-dried coals (not brown coals) is of little
importance.
The size of the crucible allows the coking of 3 to 10
grains of coal, according to the dryness of the coal and
the extent of its swelling up when heated ; and as the
blowpipe balance weighs within l-l,000th of a grain, it
16G ASSAY OF FUEL.
is easy to weigh within much less than l-10th of one per
cent, of the amount of coal assayed much nearer, in fact,
than the exactness of the coke assay in other respects.
On this point, indeed, the blowpipe assay is quite as good
as the assay with the larger balance, especially the muffle
assay, where the coal must be brushed into a clay recep-
tacle after weighing, and the coke or ashes brushed off
from it before weighing ; while here the crucible is
weighed each time without removal of its contents, and
without danger, therefore, of losing anything or adding
any dust. It may be objected that the smallness of the
amount of coal that can be assayed with the blowpipe
makes it a less trustworthy indicator of the general com-
position of the coal than a larger assay ; but the size of
the lumps or powder assayed may be made finer accord-
ingly, so that, when mixed up, an equally just sample of
the whole mass would be got for the small assay as for
the large.
Any one who has had a little experience, both in the
use of the blowpipe and in the ordinary muffle assay of
coal, will scarcely need any further teaching for the coal
assay with the blowpipe. For others it is worth while to
say that the coal may be assayed either in a fine powder
or in little lumps, and either with a slowly increasing or
with a quickly increasing heat. A quick heat will give
less coke by several per cents., but will often make a dry
coal cake together that would not cake with a slow heat.
The cover of the crucible should be left open a little, for
the easy escape of the gas, but covered enough to prevent
any flying off of solid material. The heat should increase
to redness, and as soon as the escaping gas stops burning
the heat should be stopped. As some coals part with their
gas more quickly than others, of course no definite time
can be fixed for heating all coals ; but the burning of the
gas is a good enough sign. Care should be taken not to
let the coke take up moisture from the air before weighing,
as it will quickly do if it has a chance. Of course, owing
to the different effects of quick or slow heating, a certain
uniformity of result, even with perfectly uniform samples
ASSAY OF FUEL. 167
of coal, can only be got without error by practice and by
mechanical skill, by reproducing with nicety the same
conditions in successive assays.
After the coke has been weighed, it can be heated
again with very free access of air, say with the crucible
tilted to one side, and the cover off, until everything is
thoroughly burnt to ashes ; and these should be re-heated
until no change for the less is made in the weight. With
free-burning, soft (semi-bituminous) coals this burning to
ashes is very slow, so that it is very fatiguing or even im-
possible to carry it out with a blowpipe ; but in that case
the crucible may be heated over a Bunsen gas-burner or
an alcohol lamp, and left to glow for hour after hour.
The coking is far more conveniently done in the same way
than by blowing with the mouth.
As an illustration of the degree of accuracy which this
method may be expected to give, the author adduces a
pair of blowpipe assays, made five years ago, of some West
Virginia asphaltum, that seemed itself to be much more
uniform in composition than coal from different benches in
one bed is apt to be :
Volatile Matter Coke Ashes
No. 1 . . 47-29 per cent. 52-71 per cent. 1-65 per cent.
No. 2 . . 46-93 53-07 1-81
Mean . . 47-11 52'89 1'73
VALUATION OF COAL FOR THE PRODUCTION OF ILLUMINATING
GAS. Take 100 grains of the coal in small lumps, so that
they may be readily introduced into a rather wide com-
bustion tube. This is drawn out at its open end (after the
coal has been put in) so as to form a narrow tube, which
is to be bent at right angles ; this narrower open end is to
be placed in a wider glass tube, fitted tight into a cork
fastened into the neck of a somewhat wide-mouthed bottle
serving as tar vessel (hydraulic main of the gasworks).
The cork alluded to is perforated with another opening,
wherein is fixed a glass tube, bent at right angles for con-
veying the gas first through a chloride of .calcium tube,
next through. Liebig's potash bulbs containing a solution
of caustic potash, having lead oxide dissolved in it. Next
168 ASSAY OP FUEL.
follows another tube, partly filled with dry caustic potash,
and partly with calcium chloride ; from this last tube a
gas-delivery tube leads to a graduated glass jar standing
over a pneumatic trough, and acting as gas-holder. Before
the ignition of the tube containing the coal is proceeded
with, all the portions of the apparatus are carefully weighed
and next joined by means of india-rubber tubing. After
the combustion is finished, which should be carefully con-
ducted, so as to prevent the bursting or blowing out of
the tube, the different pieces of the apparatus are discon-
nected and weighed again. The combustion-tube has to
be weighed with the coal after it has been drawn out at
its open end, and with the coke after the end of the com-
bustion when it is again cold ; and for that reason care is
required in managing it. We thus get the quantity of
tar, ammoniacal water, carbonic acid, and sulphuretted
hydrogen (as lead sulphide) ; and the gas is measured by
immersing the jar in water, causing it to be at the same
level inside and out. Empty the Liebig's bulbs into a
beaker, and separate the lead sulphide by filtration, wash
carefully, dry at 212 F., and weigh. From the lead sul-
phide the sulphuretted hydrogen present is calculated.
This process, devised by the late Dr. T. Eichardson, of
Newcastle-on-Tyne, was found by him to yield very trust-
worthy results, so as to be suitable for stating what
quantity of gas a ton of coal thus analysed would yield.
109
CHAPTEE VI.
SEDUCING AND OXIDISING AGENTS FLUXES, ETC.
IN some assay operations in the dry way, bodies are heated
in suitable vessels per se ; but it is more often necessary
to add to the bodies submitted to assay other substances,
which are varied according to the nature of the change to
be effected. These substances may be divided into five
classes : I. reducing agents ; IE. oxidising agents ; III.
desulphurising agents ; IV. sulphurising agents ; and lastly,
V. fluxes properly so called.
I. EEDUCING AGENTS.
The substances belonging to this class have the power
of removing oxygen from those bodies with which it may
be combined. In assaying, the substance under examina-
tion is generally fully oxidised either naturally or artifi-
cially before reduction is required to be effected. The
most common reducing agents are as follows :
1. Hydrogen gas.
2. Carbon.
3. The fatty oils, tallow, pitch, and resins.
4. Sugar, starch, and gum.
5. Tartaric acid.
6. Oxalic acid.
7. Metallic iron and lead.
HYDROGEN GAS The most common method of prepar-
ing this gas consists in dissolving zinc in dilute sulphuric
acid. But as this plan gives the gas in the moist state, it
must be dried by being allowed to bubble through oil of
170 EEDUCING AGENTS.
vitriol or by being passed through a bottle containing
fragments of dried calcium chloride before it is used for
assaying purposes. This gas will only be required in very
accurate assays, which are generally performed where
there are ample conveniences for generating pure hydro-
gen gas. The gas is colourless, invisible, and inodorous
when absolutely pure. It is a most powerful reducing
agent, and reduces a great number of metallic oxides at a
red or white heat, viz. the oxides of lead, bismuth, copper,
antimony, iron, cobalt, nickel, tungsten, molybdenum, and
uranium. When any metal is required in a state of abso-
lute purity, this is the only reducing agent admissible, as
others give the metal combined with a certain proportion
of carbon.
CARBON Found in large quantities in the mineral
kingdom, but generally combined with other bodies. In
a state of purity it constitutes the diamond. The diamond,
like all other species of carbon, is unacted on by the highest
possible temperature when in close vessels. It burns in
atmospheric air and oxygen gas, but requires for combus-
tion a higher temperature than ordinary charcoal. After
the diamond, the varieties of carbon found in nature or
artificially prepared are
1. BLACK-LEAD, or GRAPHITE. This is a mineral found
in beds in the primitive formations, principally in granite
and mica-schist. It is generally mixed with earthy sub-
stances, and rarely yields less than 10 per cent, of ash.
Before employing it for reduction purposes it should be
purified. Lowe * has given an excellent plan for effecting
this object.
ESTIMATION OP THE VALUE OP GRAPHITE. G. C. Witt-
stein f heats 15 grains of the sample to dull redness, and
takes the loss as water. The residue is ground up with
45 grains of a mixture of equal equivalents of potassium
and sodium carbonates, introduced into a platinum cru-
cible, covered with 15 grains of potash or soda, and slowly
heated to redness. The crust which is formed must be
* Polyt. Centr. 1855, p. 1404. f Dingler's Pqlyt. Journal, 216, 45.
REDUCING AGENTS. 171
from time to time pushed down with a platinum wire.
After fusion for half an hour it is let cool, softened with
water, heated almost to a boil for a quarter of an hour,
filtered, washed well, and the liquid set aside.
The residue on the filter is dried, placed in a small
flask, the ashes of the filter added, and about 50 minims
of hydrochloric acid (sp. gr. 1-12) poured in. On the
mixture becoming slightly gelatinous a few more drops of
the acid are added, the digestion is continued for an hour,
the mixture diluted with water, filtered, and washed. Pure
graphitic carbon remains on the filter, and is dried, slightly
ignited, and weighed. The acid filtrate mixed with the
former one of an alkaline character is evaporated to
dryness, and in it silica, alumina, ferric oxide, &c., are
estimated in the usual manner.
2. ANTHRACITE. This is another species of fossil carbon
much resembling ordinary coal, but differing from it by
burning with neither smell, smoke, nor flame.
3. COKE. This is the residue of coal after all the
volatile matter is expelled. It is generally iron-black,
and has nearly a metallic lustre ; it is difficult to inflame,
and burns well only in small pieces, but gives a very in-
tense heat. Oven or furnace coke is preferable, as it is
harder, lasts longer, and is more economical in use.
4. WOOD CHARCOAL. This is obtained by burning the
woody part of plants with a limited supply of air, so as
to drive off all their volatile matters, and leave merely
their carbon. It is this kind that is generally employed
in assays. It ought to be chosen with care, well pulverised,
passed through a sieve, and preserved in well-stopped
vessels. Wood charcoal is never perfectly pure, generally
containing, besides ash, a proportion of hydrogen and
watery vapour ; these bodies are not generally prejudicial,
but in some experiments ash must not be present ; in that
case pure charcoal may be procured by heating sugar to
redness in a closed crucible.
The advantage of carbon as a reducing agent consists
in its great affinity for oxygen, which at a red heat sur-
passes that of most other substances. Charcoal by itself
172 REDUCING AGENTS.
possesses two inconveniences : first, it lias the property of
combining with many metals ; and in the second place, it
is infusible, and cannot combine with vitreous substances.
The property it possesses of combining with iron, nickel,
cobalt, &c., is of no consequence to the assay er, for the
increase of weight it gives is not material, excepting under
the circumstances to be hereafter pointed out ; but its
infusibility and inability to combine with fluxes is a very
serious inconvenience ; for after the reduction, that portion
which has not been consumed remains disseminated with
the grains of metal in the fused slag, and prevents the
separation of all the metal, and the consequent formation
of a good button ; a large quantity of charcoal can thus
irreparably injure an assay. This inconvenience does not
happen, however, when an oxide is reduced by cementa-
tion in a lined crucible, but there are some cases in which
this mode of reduction is inadmissible.
Coke should never be used as a reducing agent in
assays when it is possible to avoid it. It often contains
a very large proportion of earthy and other extraneous
matters (more particularly sulphur, which is very in-
jurious). Coke is never so good as wood charcoal as a
reducing agent, because it burns more slowly. When it
is used, the temperature employed for an assay must be
much increased.
Coal is nearly always inconvenient, because it swells
by heat ; nevertheless, as it is not required in very large
quantities, it is sometimes employed, being finely powdered
and sifted previous to use.
THE FATTY OILS. The name oil is generally given to
those bodies that are fat and unctuous to the touch, more
Or less fluid, insoluble in water, and combustible. They
all become solid at various degrees of temperature.
There are some which, at the temperature of our climate,
have constantly a solid form, as butter, palm oil, cocoa-
nut oil, &c.
; TALLOW is an animal product analogous to the fatty oils
in properties.
EESINS. The greater part of the resins are solid, but
REDUCING AGENTS. 173
some are soft. They are brittle, with a vitreous and
shining fracture, and are often transparent. They are
very fusible, but cannot be raised to their boiling-point
without partial decomposition.
Although all the bodies just mentioned consume in
their combustion a large quantity of oxygen, they cannot
generally effect the total reduction of an oxide on account
of their volatility ; so that, before the temperature at
which the reduction takes place can be attained, the
greater part of the reducing agent has been expelled.
They generally act only by virtue of the small carbonaceous
residue produced by the action of heat ; so that their use
is very limited and uncertain. Whenever they are em-
ployed as reducing agents, without covering the substance,
a loss is experienced, on account of the bubbling and boil-
ing caused by their decomposition; this will always take
place unless the contents of the crucible be covered with
charcoal powder. Oils are very serviceable in the reduc-
tion of a large mass of oxide by cementation ; in this case,
after the oxide has been placed in the crucible, as much
oil is added as the oxide and the lining of the crucible will
soak up. Fat or resin is also used to prevent the oxida-
tion of the surface of a metallic bath (as in the fusion of
bar-lead samples), by coating the metal, and preventing
the action of the atmospheric oxygen.
SUGAR in its decomposition by heat leaves a much larger
proportion of carbon than the oils, fats, or resins ; so that
it would appear serviceable as a reducing agent. There
are some cases in which it may be used with advantage,
but it undergoes a great increase in volume when heated ;
so that losses in an assay may occur by the use of this
agent. To purify sugar from mineral ingredients it should
be re-crystallised from alcohol. It then may be used
as such, or after carbonisation. It yields about 14 per
cent, of charcoal : this is pure carbon, and leaves no resi-
due when burnt ; it is, therefore, preferable to wood char-
coal in cases where no foreign matter should be introduced
into the assay.
STARCH, well dried, and, better still, terrified, is em-
174 KEDUCIXG AGENTS.
ployed with advantage as a reducing agent, and is better
than sugar, as it neither fuses, swells up, nor spirts, and in
many cases is even preferable to charcoal, because it is in
such a fine state of division that it can be more readily and
intimately mixed with the substance to be reduced. Wheat
and rye flour have nearly the same qualities as starch.
They are sometimes used.
GUM decrepitates slightly by heat, softens, agglomerates,
and boils, without spirting. The gums can be employed as
reducing agents under the same circumstances as sugar and
starch, but the two latter are preferable, because they con-
tain no earthy substances.
TARTARIC ACID is the reducing agent in cream of tartar,
or argol, of which so frequent use is made ; but the acid is
never employed by itself. When heated in close vessels
it fuses and decomposes, giving off combustible gases, leav-
ing a little charcoal. It burns when heated in contact
with air, giving rise to a peculiar and not unpleasant
odour.
OXALIC ACID fuses at a temperature of 208 without
decomposing, but when heated to 230 it is decomposed,
giving rise to carbonic acid, carbonic oxide, and a little
formic acid vapour ; and when heated strongly some por-
tions volatilise without decomposition : it does not leave a
carbonaceous residue.
The property which oxalic acid possesses of not leaving
a residue would render it remarkably valuable for the re-
duction of the metallic oxides in cases where the slightest
trace of carbon is to be avoided, if its reducing power were
greater ; but it decomposes at a low temperature, and in
burning absorbs but a small quantity of oxygen, especially
when it has not been dried ; so that even for the most
easily reducible oxides a large proportion must be em-
ployed. When it is combined with a base, as potash in
potassium binoxalate, its reducing power is much aug-
mented.
AMMONIUM OXALATE, when heated in close vessels, is
decomposed. Its reducing power is nearly double that
of oxalic acid.
COMPARATIVE REDUCING POWER.
175
COMPARATIVE REDUCING POWER OP THE ABOVE AGENTS. In
order to give an idea of the comparative reducing power
of the agents just described, the result of some assays
made on them by Berthier, by means of litharge, are given
below.
By heating the same weight of each reducing agent
with an excess of litharge, buttons of leads were obtained
whose weights were proportional to the quantity of oxygen
absorbed, and by comparing them with each other the re-
ducing power of each flux is given ; by taking for unity
the weight of the reagent, calculation has proved that
1 part of pure carbon reduces from litharge 34-31 of
lead. The following are the results of Berthier's experi-
ments :
Hydrogen.
Pure carbon
Calcined wood charcoal
Amber resin
Ordinary wood charcoal
Animal oil
Tallow
Resin
Sugar
Torrified starch
Common starch
Gum Arabic
Tartaric acid
Ammonium oxalate .
Oxalic acid
104-00
34-31
31-81
30-00
28-00
17-40
15-20
14-50
14-50
13-00
11-50
11-00
6-00
1-70
90
It must be borne in mind that these numbers do not
represent the quantities of oxygen each reagent would ab-
sorb in complete combustion ; but that it only indicates
the quantity of metal produced by equal weights of the
reagents.
In assaying, however, it is rarely that these agents are
used by themselves ; they are generally mixed with a flux
properly so called, which will be more particularly de-
scribed under the head of Fluxes.
METALLIC IRON removes oxygen from the oxides of lead,
bismuth, copper, &c., but is rarely added for that especial
purpose ; and when it does produce this effect it is gener-
ally secondary, because it previously existed in the matter
subjected to assay, or was added for some other purpose.
176 OXIDISING AGENTS.
METALLIC LEAD reduces but a very small number of
oxides, but it reduces many to the minimum of oxidation ;
it also decomposes some sulphates and arseniates.
II. OXIDISING AGENTS.
The oxidising agents in general use are as follows : -
1. Oxygen, atmospheric or combined.
2. Litharge and ceruse.
3. Lead silicates and borates.
4. Potassium nitrate.
5. Lead nitrate.
6. Manganese peroxide.
7. Copper oxide.
8. Iron peroxide.
9. The caustic alkalies.
10. The alkaline carbonates.
11. Lead, copper, and iron sulphates.-
12. Sodium sulphate.
LITHARGE is a fused protoxide of lead, and is generally
obtained from the silver-lead works. When melted, it
oxidises nearly all the metals, except mercury, silver, gold,
palladium, platinum, &c., and generally forms very fusible
compounds with the oxides. These two properties cause
it to be a very valuable agent in separating silver and
gold from all the substances with which they may be
mixed.
Litharge is occasionally mixed with a little of the red
oxide of lead ; the presence of this in large quantities be-
comes injurious, as it has the property of oxidising silver.
Ordinary litharge can be easily freed from this oxide by
fusing it and pouring it- into a cold ingot mould, then
pulverising, and carefully keeping it from contact with air,
as it readily absorbs oxygen, and if it be allowed to cool in
the atmosphere it will nearly all be converted into the red
oxide.
CEKUSE, or WHITE-LEAD, is a carbonate of lead protoxide.
OXIDISING FLUXES. 177
As it does not contain the slightest traces of red oxide, it
may be used where the presence of that substance may be
inconvenient ; but it is troublesome to use, as it is much
less dense than litharge ; large vessels must be employed
in consequence ; besides, it generally contains a small
quantity of lead acetate or subacetate, and sometimes
metallic lead separates from it on ignition, which is, in
some cases, disastrous to the result of an experiment.
When ceruse is employed, a certain quantity must be fused,
to ascertain if any metallic lead be produced ; and, on the
other hand, it must be examined to ascertain if it be adul-
terated with barium sulphate. When it is pure it dissolves
completely in acetic or nitric acid.
LEAD SILICATES AND BORATES behave as litharge, but they
oxidise less rapidly.
They may be prepared by fusing together 1 part of
silica or boracic acid with 1 part of litharge. The borates
are more fusible than the silicates, but their use is attended
with inconvenience, as they swell very much in fusing.
POTASSIUM AND SODIUM NITRATES fuse at a temperature
below redness without alteration, but when heated more
strongly they give up oxygen. The action of these salts,
when fused, is very energetic, because they have a great
tendency to decompose, and because they contain a large
quantity of oxygen. They are used as oxidising agents in
the purification of the noble metals, and for ' preparing
some fluxes. They ought always to be employed in a
state of purity.
Saltpetre often contains impurities. On this account
an estimation of the real amount of potassium nitrate
often becomes necessary, not only in cases where saltpetre
is to be used for docimetric purposes, but also when used
in certain technical operations, viz. the manufacture of
gunpowder, enamel, &c.
If saltpetre is very impure, it may easily be purified
by recrystallisation to such a degree that it will only con-
tain 2 to 3 per cent, foreign substances (chiefly sodium
chloride).
An exact assay of saltpetre is most difficult, and the
178 IMPURITIES IN SALTPETRE.
different modes in use are not quite accurate, on ac-
count of the nitrate of soda generally present as an
impurity in saltpetre, and not easy to estimate by means
of reagents.
Manufacturers often intentionally mix the raw salt-
petre with soda-saltpetre, and it is also often manufac-
tured from a mixture of soda-saltpetre and potassium
carbonate.
ASSAY OF SALTPETEES. The following are the different
modes of assaying saltpetre.
a. To estimate the nitric acid directly, 150 grains of
the well-ground sample are mixed with six times its weight
of well-powdered silica, placed in a platinum capsule and
dried in the water-bath till there is no further loss of
weight. The temperature is then raised to dull redness
for about half an hour, by which time the whole of the
nitric acid is expelled. The difference in weight gives the
nitric acid.
Sodium nitrate may be treated in the same manner.
The most usual, though not the most satisfactory, pro-
cess for the assay of potassium and sodium nitrates is to
estimate the total impurities, viz. water, insoluble matter,
alkaline chlorides and sulphates, and the residue being
taken as pure nitrate.
b. Gay-Lussac's mode of assaying saltpetre consists in
converting the potassium nitrate into potassium carbonate,
and in estimating its amount volu metrically by means of
standard sulphuric acid. 2-639 grains of saltpetre are
mixed with 1 grain of ignited pine-root, and 12 grains of
ignited and finely pulverised sodium chloride (the latter
is added in order to moderate the combustion), and this
mixture is heated in a platinum crucible. After cooling,
the mass is extracted by water, and either a standard solu-
tion of sulphuric acid or oxalic acid is added to the solu-
tion. The sulphuric acid is prepared by mixing 70 grains
of sulphuric acid, sp. r gr. 1*84, with 600 grains of water,
and to this mixture so much water is added again that
100 measures of it will saturate 6-487 grains of potassium
carbonate. The number of measures used for saturation
ASSAY OF SALTPETRE. 179
will then indicate directly the percentage of potassium
carbonate.
The following foreign substances in raw saltpetre
should be estimated.
Water. 150 to 300 grains of air-dried, '.'finely pul-
verised saltpetre are heated in a porcelain crucible to
120 C. 5 and the resulting loss is calculated as water.
Mechanically mixed Impurities. The substance ob-
tained in the former assay is dissolved in hot water, and
filtered through a dried and weighed filter. The residue
is well washed with hot water, dried on the filter at
250 F., and weighed. On deducting the weight of the
filter, there will be left the weight of the mechanically
mixed impurities (alumina, silica, calcium carbonate, iron
peroxide, &c.), which usually amount to 2 to 5 per cent.
Lime and Magnesia. These substances are precipitated
as carbonates by sodium carbonate, in the former filtered
solution raised to the boiling-point ; the carbonates are
then dissolved in hydrochloric acid, and neutralised with
ammonia. The lime can be precipitated by oxalic acid,
and filtered off; the magnesia which remains in solution
may then be precipitated by sodium phosphate.
The amount of lime in East Indian raw saltpetre which
has been once crystallised varies between 0*21 and 0-26
per cent., the amount of magnesia between 0-26 and 0'28
per cent.
Chlorine. Thirty or 40 grains of raw saltpetre are dis-
solved in about an ounce of pure warm water in a flask
furnished with a tight-fitting stopper, and the amount of
chlorine is estimated by a standard solution of nitrate of
silver. The solution, after being warmed and acidulated
with nitric acid, is mixed gradually with the solution of
silver ; after each addition of the latter it is to be shaken
and then allowed to rest. (For particulars see the chapter
on the assay of silver in the wet way.)
The amount of chlorine found by this assay is cal-
culated as being derived from f potassium chloride and
^ sodium chloride, so that 1 part of chlorine corresponds
to 1-927 part of metal (1-285 potassium, 0-642 sodium).
N 2
180 ASSAY OF SALTPETRE.
Experience has proved that East Indian saltpetre contains
potassium and sodium chlorides in these proportions.
Sulphuric Acid. 100 to 120 grains of raw saltpetre
are dissolved in six ounces of water, and from this solu-
tion, heated to the boiling-point, the sulphuric acid is
precipitated by means of a solution of barium nitrate.
The precipitated barium sulphate is filtered off, washed ,
ignited, and weighed. The amount of sulphuric acid in
East Indian raw saltpetre varies between 0-05 and (Ml
per cent.
Sodium Nitrate. This estimation is most difficult, arid
the following modes are recommended.
a. Longchamps's mode is based upon the decomposition
of soda-saltpetre by potassium chloride, producing sodium
chloride and potassium nitrate. The saltpetre is mixed
with potassium chloride, and the solution evaporated
down. By this operation, first sodium chloride and after-
wards saltpetre become separated. The latter is washed,
dried at 150 C., and weighed. Werther has recommended
a similar mode.
3. If the saltpetre does not contain certain oxides,
such as alumina, lime, &c. (or if, previously present, they
have been precipitated), a solution of potassium antimo-
niate will precipitate the soda contained in the saltpetre
solution. The precipitate consists of sodium antimoniate,
100 parts of which contain 84*39 antimonious acid, and
15-61 soda.
7. The presence of soda is also to be ascertained by
washing saltpetre with a saturated solution of pure potash-
saltpetre. This saturated solution will then contain a pro-
portionally large amount of sodium nitrate. If a small
quantity of the solution is made to crystallise upon a
watch-glass, soda-saltpetre, showing a rhombohedricform,
may be detected by means of a microscope, while potash-
saltpetre crystallises in prisms, and sodium and potassium
chlorides in cubes arranged in the form of steps.
LEAD NITRATE acts in a similar way to the two last-
mentioned salts. It is prepared by dissolving litharge in
nitric acid, and crystallising the solution.
DESULPHURISING REAGENTS. 181
MANGANESE PEROXIDE is easily reduced to the state of
protoxide by many metals, and is a very powerful oxi-
dising agent ; but is rarely employed, because its com-
pounds are very infusible. It is employed occasionally
in the purification of gold and silver.
COPPER OXIDE is not much employed as a Hux, but is
often contained in substances submitted to assay ; it then
acts as an oxidising agent. A great number of metals,
ven silver, reduce it to the minimum of oxidation ; and
other metals, as iron, for instance, totally reduce it.
IRON PEROXIDE. This, like copper oxide, sometimes acts
incidentally as an oxidising agent.
THE CAUSTIC ALKALIES, POTASH AND SODA, fuse below a
red heat, and volatilise sensibly at a higher temperature.
Charcoal, at a high temperature, decomposes the water
combined with the hydrates of potash and soda, convert-
ing them into carbonates, but an excess at a white heat
decomposes the carbonate, and potassium or sodium is
the product.
POTASSIUM AND SODIUM CARBONATES are very much em-
ployed as agents in the assay by the dry way. They
have the power of oxidising many metals, as iron, zinc, and
tin, by the action of the carbonic acid they contain ; part of
it being decomposed, with the formation of carbonic oxide.
LEAD, COPPER, AND IRON SULPHATES. These three salts
at a high temperature oxidise the greater number of the
metals, even silver, the sulphuric acid giving off oxygen
and sulphurous acid. They are used in the assay of gold.
SODIUM SULPHATE is not used by itself as a reagent, but
is a product in many operations ; it is either formed in
the course of an assay, or is contained as an impurity in
some of the bodies used.
III. DESULPHURISING EEAGENTS.
1. The oxygen of the atmosphere.
2. Charcoal.
3. Metallic iron.
4. Litharge.
5. The caustic alkalies.
182 LITHARGE.
6. The alkaline carbonates.
7. Nitre.
8. Lead nitrate.
9. Lead sulphate.
1. THE OXYGEN OF THE ATMOSPHERE acts as a desulphuris-
ing agent in roasting, combining with the sulphur present,,
forming sulphurous acid or sulphuric acid, sometimes
both.
2. CHARCOAL decomposes many sulphides by taking
their sulphur to form sulphide of carbon. It acts in this
manner with the mercury, antimony, and zinc sulphides.
It is only employed as an auxiliary to the desulphurising
power of the alkalies and their carbonates.
3. IRON separates sulphur from lead, silver, mercury,
bismuth, zinc, antimony, and tin, but only partially decom-
poses copper sulphide. It is generally used in the state
of filings, or nails ; the latter are preferable, and ought to
be kept free from rust. Oxide of iron may be used if it
be mixed with the requisite quantity of charcoal to reduce
it. Cast iron must not be employed, as it has very little
affinity for sulphur.
4. LITHARGE exercises a very energetic action on sul-
phides, even at a low temperature. If it be employed
in sufficient proportion, the sulphide acted on is wholly
decomposed. The sulphur is often disengaged as sulphur-
ous acid, and the metal remains alloyed with the lead
proceeding from the reduction of a portion of the litharge,
or combines as oxide with that portion of the litharge
which is not reduced. The quantity of litharge requisite
for the decomposition of a sulphide is considerable, and
varies according to its nature ; some sulphides require
34 times their weight. When less than the requisite
quantity is used, only a portion of the sulphide is decom-
posed, and a corresponding quantity only of lead reduced,
whilst the remainder of the sulphide forms, with the
litharge and the metallic oxide which can be produced,
a compound belonging to the class of oxysulphides, which
is generally very fusible.
ACTION OF LITHARGE ON SULPHIDES. 183
When the sulphides have a very strong base, as an
alkali or alkaline earth, no sulphurous acid is given off by
the action of litharge, but all the sulphur is converted into
sulphuric acid.
Litharge is a very valuable reagent, and its use is
nearly exclusively confined to the assay of sulphides
containing the noble metals, as these metals are thus
obtained as alloys of lead, which are afterwards assayed
by cupellation.
The following is an account of the behaviour of this
reagent with the ordinary sulphides.
Manganese Sulphide requires at least six times its
weight of litharge to produce a fusible compound, and
thirty times its weight to desulphurise it completely. The
sulphur and metal oxidise simultaneously, and a manganese
protoxide is formed, which partly peroxidises, taking a
brownish tint in contact with the atmosphere. Berthier
assayed the four following mixtures :
Manganese sulphide . . 5 5 5 5
Litharge . ... 20 30 100 150
The first produced an infusible, greyish-black, scori-
forin mass, in which small plates, having the look of
galena, could be discovered. It was composed of the
sulphides and oxides of manganese and lead. Much
sulphurous acid was given off during the operation.
The second fused to a soft paste, and gave 17' 5 of
lead, and a compact, vitreous, opaque slag, of a very deep
brown colour. The slag contained about half its weight
of manganese sulphide.
The third fused readily, and produced 31-5 of ductile
lead, and a transparent, vitreous slag, of a deep hyacinth-
red.
The fourth produced 33'T of lead, exceedingly ductile,
and the desulphurisation was complete.
Iron Sulphide. -Thirty; parts of litharge are sufficient
to scorify iron protosulphide ; the metal is converted into
the protoxide.
184 ACTION OF LITHARGE ON SULPHIDES.
The four following mixtures
Iron protosulphide ... 10 10 10 10
Litharge 60 125 250 300
gave, the first a pasty, scoriform mass, colour metallic
grey, and very magnetic. It was composed of the sul-
phides and protoxides of iron and lead.
The second gave a very fluid metallic black slag, very
magnetic, opaque, and possessing great lustre, and 36 of
lead.
The third gave a compact vitreous transparent slag of
a fine resin-red, and 67 of lead.
The last yielded a similar slag to the former, but con-
taining no sulphur, and 70 of lead.
Native iron pyrites was treated with the following
proportions of litharge :
Iron pyrites . . 10 10 10 10 10 10
Litharge. . . 60 125 200 300 400 500
The mixtures fused very readily with an abundant
disengagement of sulphurous acid.
The first produced only a metallic button, divisible
into two parts : the lower was the larger, and was a lead
subsulphide ; the other looked like compact galena, but
was magnetic ; it was composed essentially of the sulphides
of iron and lead, but probably contained a small quantity
of their oxides.
The second and third gave black vitreous opaque
slags which stained the crucibles brown, together with
lead, having a granular fracture and a deep grey colour ;
the first button weighed 35, and the second 40. Both
samples of lead were contaminated with a small quantity
of slag, and contained from y^Vo ^ Tiro f sulphur, and
a small quantity of iron.
The slags from the last three mixtures were vitreous,
transparent, and of a fine resin-red colour : the buttons of
lead equalled 45-4, 54*8, and 86 parts. A much larger
proportion of litharge does not produce more than 86 of
lead, proving that 50 parts of litharge completely effect
the desulphurisation of iron pyrites.
ACTION OP LITHARGE OJS T SULPHIDES. 185
Copper Sulphide. The following mixtures of sulphide
of copper and litharge
Copper sulphide . .10 10 10 10 10
Litharge .... 20 30 50 100 250
fuse very readily, giving off an abundance of sulphurous
acid.
The slags formed were compact, vitreous, opaque,
or translucid, and more or less bright red. The copper
which they contained was at the minimum of oxidation.
The first three mixtures gave metallic buttons com-
posed of uncombined lead and sulphide of copper.
The fourth gave 28 of lead, with a little adhering
sulphide of copper.
The fifth gave 38*5 of pure ductile lead, the exact
quantity that ought to be reduced from litharge by the
transformation of the above quantity of copper sulphide
into suboxide and sulphurous acid.
Copper sulphide does not combine with litharge ; this
is an exception to the general rule. It requires about
twenty-five times its weight of litharge to decompose it
completely. When litharge is combined with a certain
quantity of copper protoxide, it -has no action on the
sulphide of that metal.
The desulphurisation of copper pyrites requires about
30 parts of litharge.
Copper pyrites .... 10 10 10 10
Litharge ..... 50 100 200 300
were fused together.
In the first assay the fusion was accompanied with
much ebullition, and the mass remained pasty : 6 parts of
ductile lead were produced, and a matte similar to galena,
but deep grey, with small facets, and a brownish-black
vitreous slag.
In the second, much ebullition and swelling up took
place : 35 of lead, 45 of matte, and a deep brown vitreous
slag were produced.
In the third assay, 49 of lead was the result. It was
covered by a thin layer of matte, and a very shining, dee])
brown, vitreous, translucid slag.
18G ACTION OF LITHARGE ON SULPHIDES.
The last mixture fused readily, almost without ebul-
lition, and gave 72 of lead, and a compact shining slag, of
a bright grey, and without the least trace of matte ; the
desulphurisation was complete.
Antimony Sulphide has a great tendency to combine
with litharge, and it must be heated with at least 25 parts
to effect its desulphurisation. By mixing these two sub-
stances in the following proportions
Antimony sulphide .10 10 10 10 10
Litharge . . .38 60 100 140 250
the first three mixtures afforded very fluid slags, compact,
deep black, and slightly metallic, and. buttons of ductile
lead, weighing 2, 9, and 26 parts. These slags resemble
the black litharge produced at the commencement of a
cupellation.
The fourth mixture gave a transparent compact slag,
vitreous and shining, having a splendid hyacinth-red
colour, and 50 parts of lead.
The last produced 57 of lead, proving the desulphuri-
sation to be complete. The antimony, in this case, exists
as protoxide in the slag.
M. Fournet has observed that antimony sulphide has
the property of carrying copper sulphide, and even silver
sulphide, into the compounds formed with litharge. In
one of the experiments which he made, a double sulphide,
composed of equal parts of silver sulphide and antimony
sulphide, was fused with three times its weight of litharge,
and gave, first, a button of lead mixed with silver ;
secondly, a matte like galena ; and thirdly, a black slag.
This slag was analysed, and found to contain from 8 to 9
per cent, of silver.
It is probable that all the sulphides, having a strong
tendency to combine with lead oxide, have, like antimony
sulphide, the property of determining the scorification
of a certain quantity of silver sulphide ; like all the sul-
phides, which, in a state of purity, are completely decom-
posed by lead oxide.
Zinc Sulphide must be fused with twenty-five times its
ACTION OF LITHARGE OX SULPHIDES. 187
weight of litharge to be decomposed. The following mix-
tures were heated together :
Blende .... 24-08 12-08 10 10
Litharge ... . 55-78 83-68 100 250
However strongly the first mixture was heated, it
always remained pasty ; 29*2 of a greyish-black lead were
produced, which contained -018 of sulphur and -008 of
zinc. The button was covered by a metallic-looking black
substance, intermediate between a matte and a slag ; it
was composed of zinc and lead sulphides and oxides.
The second mixture gave 35 -5 of lead and a fluid slag,
which was compact, opaque, and black.
The third gave 43 of lead, and a deep grey slag.
The last produced 65 of pure lead, and a vitreous slag,
of an olive colour, and translucid on the edges.
Lead Sulphide. Galena and litharge, at a heat just
sufficient to fuse them, combine and form an oxysulphide;
but if the temperature be increased, the two bodies react
on each other, and are mutually decomposed. If 2,789
parts of litharge be employed to 1,496 of lead, or 1,865 of
litharge to 1,000 of galena, nothing but pure lead is
obtained. If more litharge be employed, a portion is not
decomposed, and covers the lead. If less be employed,
the galena is not completely decomposed, and the lead is
covered by a matte of subsulphide.
But when litharge is combined with a certain prp-
portion of sulphides or metallic oxides, it completely loses
its oxidising power on galena, even at a white heat ; so
that it can be combined with this substance as with the
other sulphides, without effecting its total decomposition.
5, 6. CAUSTIC ALKALIES AND THEIR CARBONATES. All the
sulphides are decomposed by caustic alkalies and their
caroonates; but in the latter case carbonaceous matter
must be present. In the absence of charcoal there are
some sulphides, as of copper, on which they have no
action. In these decompositions alkaline sulphides are
formed, and combine with and retain a certain quantity
of the sulphide submitted to experiment. The proportion
188 CAUSTIC ALKALIES AND THEIR CARBONATES.
of the sulphide which remains in combination with the
alkaline sulphides depends on many circumstances. It is
always less when a large proportion of alkali or carbonate
has been employed, as it is also when a high degree of
temperature has been employed ; and the presence of
charcoal always much diminishes the proportion. When
the metal of a sulphide is very volatile, as mercury or
zinc, the decomposition may be perfect-
Potash, as it is sold in commerce, generally contains
foreign substances, viz. silica, peroxide of iron, potassium
sulphate, chloride, phosphate, and silicate, soda-salts, &c.,
and also water.
A partial purification of the potash may be effected by
dissolving it in water, which will not dissolve some of the
above-named foreign substances.
Soda also is never free from foreign substances.
Ammonium carbonate is used for decomposing metallic
sulphates which are formed during the roasting process of
several sulphur minerals. Ammonium sulphate is then
formed, which is volatile when slightly heated.
7. NITRE, SALTPETRE, OR POTASSIUM NITRATE has a very
powerful action on the sulphides ; in fact, if not modified
by the addition of some inert substance, as an alkaline
carbonate or sulphate, explosion may take place, and a
portion of the contents of the crucible be thrown out.
Where an excess of nitre is used, all the sulphur is con-
verted into sulphuric acid, and every metal but gold and
silver is oxidised. When only the exact quantity of nitre
is employed that is to say, just as much as is sufficient
to burn all the sulphur in the sulphides of those metals
which are not very oxidisable, as those of copper, silver,
and lead, the metal is obtained in a state of purity, and
the whole of the sulphur converted into sulphuric acid ;
but with the sulphides of the very oxidisable metals the
oxygen of the nitre is divided between the sulphur and
the metal.
8. LEAD NITRATE possesses the combined properties of
nitre and litharge. It is not much used.
9. LEAD SULPHATE is not used as a reagent, but is often
SULPHURISING REAGENTS. 189
formed in the assay of lead ores. It decomposes lead
sulphide by burning the sulphur. It acts on many other
sulphides in a similar manner.
IV. SULPHURISING REAGENTS.
1. Sulphur.
2. Cinnabar, or mercury sulphide.
3. Galena.
4. Antimony sulphide.
5. Iron pyrites.
6. The alkaline persulphides.
1. SULPHUR fuses at 226, and at 284 is very liquid.
It has very powerful affinities, and combines with the
greater number of the metals. That kind generally known
as flowers of sulphur ought to be employed ; and before
use, the presence or absence of earthy matters should be
ascertained by exposing it to a dull red heat in a crucible.
The sulphur will go off, and the earthy impurities will be
left behind.
Sulphur is principally used in the preparation of the
alkaline sulphides and in the assay of some of the noble
metals.
2. CINNABAR is decomposed by many of the metals, and
is a better sulphurising agent than sulphur itself, as it is
less volatile.
3. GALENA. Many metals, as iron, copper, &c., separate
sulphur from lead, while some others, as silver, gold, &c.,
do not ; so that if galena be heated with an alloy of various
metals, some of which decompose it, and some do not, the
former are transformed into sulphides, and the latter com-
bine with the metallic lead which is produced. It is often
employed for this purpose. It is a common ore, and
readily procured.
The samples employed must contain no antimony sul-
phide, and alHhe matrix must be carefully separated by
sifting and washing.
4. ANTIMONY SULPHIDE yields its sulphur to many of
the metals, but it is only used in the separation of gold
190 SULPHURISING REAGENTS.
from some alloys. In this operation the sulphur combines
with the alloyed metals, and the antimony with the gold,
for which it has much affinity. ;
5. IRON PYRITES is a persulphide which loses half its
sulphur at a white heat. It is much employed in metal-
lurgical operations, but not in assaying.
6. ALKALINE PERSULPHIDES can support a tolerably
elevated temperature without losing sulphur, but they
have a great tendency to do so, and to this is due their
sulphurising power. By their means almost every metal
can be made to combine with sulphur. When an alkaline
persulphide is heated with a metal, or an oxide of a metal
mixed with charcoal, a fused compound, a mixture of
the sulphide of the metal and an alkaline sulphide, is
obtained.
When they are in combination they are held together
by very feeble affinities, and their decomposition is gener-
ally eifected by the mere action ; of water, which dissolves
the alkaline sulphide and leaves the other perfectly pure.
But with gold, molybdenum, tungsten, antimony, &c., the
compound is stable and soluble in water ; and it is from
this fact that the alkaline sulphides are sometimes em-
ployed in the assay of auriferous substances.
In order to effect a sulphurisation by means of the
alkaline sulphides, it is much better to use equivalent
mixtures of sulphur and alkaline carbonates than to pre-
pare them beforehand. To obtain potassium persulghide,
46 parts of potassium carbonate, and 54 of flowers of
sulphur, must be fused together ; and for sodium persul-
phide, 40 parts of dry sodium carbonate must be heated
with 60 parts of sulphur. >
When the mixture is fused in a plain crucible, potas-
sium sulphate or sodium sulphate is formed, because part
of the alkali gives up its oxygen to a portion of the
sulphur, which becomes sulphuric acid ; but when char-
coal-lined crucibles are used, the carbon combines with
the oxygen of the alkali, and no sulphate is produced.
FLUXES. 191
V. FLUXES.
Fluxes are used in the following cases :
1st. To cause the fusion of a body either difficultly
fusible or infusible by itself.
2ndly. To fuse foreign substances mixed with a metal,
in order to allow the latter to separate by its difference of
specific gravity.
ordly. To destroy a compound into which an oxide
enters, and which prevents the oxide being reduced by
charcoal. Zinc silicate, for instance, yields no metallic
zinc with charcoal, unless it be mixed with a flux capable
of combining with the silica.
4thly. To prevent the formation of alloys of some
metals with others, as, for instance, in the case of a mix-
ture of the manganese and iron oxides ; when a suitable
flux is employed, the iron is obtained in a state of purity,
whereas if no flux had been added an alloy would have
been obtained. Gold and silver can be separated from
many other metals by means of a flux.
5thly. To scorify some of the metals contained in the
substance to be assayed, and obtain the others alloyed
with a metal contained in the flux, as gold or silver with
lead.
6thly. A flux may be employed to obtain a single
button of metal, which otherwise would be obtained in
globules.
Fluxes are divided into non-metallic and metallic ; the
non-metallic fluxes are
1. Silica.
2. Lime.
3. Magnesia.
4. Alumina.
5. Calcium and aluminium silicates.
6. Glass.
7. Borax (sodium biborate).
8. Fluor-spar (calcium fluoride).
9. Potassium carbonate.
192 SILICA, LIME.
10. Sodium carbonate.
11. Nitre (potassium nitrate).
12. Common salt (sodium chloride).
13. Black flux and its equivalents.
14. Argol (potassium bitartrate).
15. Salt of sorrel (potassium binoxalate).
16. Soap.
The metallic fluxes are
17. Litharge (lead oxide) and ceruse (lead carbonate).
18. Glass of lead (lead silicate).
19. Lead borate.
20. Lead sulphate.
21. Copper oxide.
22. Iron oxides.
1. SILICA is employed frequently to cause the fusion of
some gangues in assays made at an elevated temperature.
Silica combines with all the bases, and forms with them
silicates, which are more or less fusible.
Quartz is the beet form of silica to use. For that pur-
pose it must be strongly heated, and then quenched in cold
water. It can then be easily pulverised. In case the
quartz takes a yellow or reddish colour on ignition, it
must be digested with hydrochloric acid.
2, 3, 4, 5. LIME, MAGNESIA, ALUMINA, AND THEIR SILICATES.
No simple silicate is readily fusible, so that lime, magnesia,
or alumina are employed, according to circumstances, to
reduce a simple silicate to such a condition that it will
readily fuse in an assay furnace. Sometimes it may be
requisite to use all the above-mentioned earths.
Pure lime, when exposed to atmospheric air, attracts
carbonic acid and water so quickly that, in practice, pure
calcium carbonate is used in the form of chalk, calcareous
spar, or marble, if they are pure. Calcium carbonate fre-
quently contains foreign substances, viz. iron, manganese,
alumina, silica, and also magnesium carbonate. A certain
quantity of magnesium carbonate is, in many cases, advan-
tageous, and alumina and silica are not disadvantageous.
GLASS, BORAX. 193
Alumina is never used. in the pure state. Washed
china-clay which, on burning, becomes white, is used
instead. Clay generally contains from 20 to about 40 per
cent, alumina, and, if it is used for the formation of sili-
cates, a quantitative analysis of its components should
first be performed.
6. GLASS is a very useful flux in certain assays, and,
being a saturated silicate, it will serve by itself either as a
slag or merely as a covering. The kind employed must
contain no easily reducible metallic oxides, and it must
especially be free from arsenious acid and lead oxide.
The subjoined analyses of glass from Bodemann KerVs
ProUerkunst will be found useful. (See p. 194.)
7. BORAX. That kind with 10 atoms .or 47*1 per cent,
water effloresces when exposed to atmospheric air ; and the
other kind, with 5 atoms or 30 per cent, water, does not
effloresce, and crystallises in octahedrons. This difference
is immaterial for assaying purposes, but it is of importance
in purchasing borax.
When borax is heated, it loses its water of crystalli-
sation and undergoes an enormous increase of volume ; at
a higher temperature it fuses and forms a transparent
glass, which becomes dull on the surface by exposure to
air. Only the fused vitrified borax ought to be used in
assays. It must be reduced to powder, and kept in well-
closed vessels.
Borax may be regarded as containing free boracic
acid ; it is an excellent and nearly universal flux : it has
the property of forming, like boracic acid, fusible com-
pounds with silica and nearly all the bases, and is prefer-
able to boracic acid because it is much less volatile. It may
be used at a high or a low temperature. It is employed
in the assay of gold and silver, because it fuses and com-
bines with most metallic oxides, or in obtaining a regulus
that is to say, to separate the metals, their arsenides and
sulphides, from any stony matter with which they may be
mixed ; because this salt is neither oxidising nor desul-
phurising. It is also employed in the assay of iron and
tin ores, as in the presence of charcoal it retains but traces
o
194
ANALYSES OP GLASS.
CQ
ft
H
FH
^5
P^3
.
W
e
p
fe
o
QQ
w
S
<j
&
-<
I
? 9 > *? 9 9 T 1
XO "tf O rH rH C- O
rH
. CO
:6
.
: |
co
6
jt
$
O
o 6
. . C5 C5 . .0
: : TH o : ' o
cT
fl
rH
O 00 CM CO
OCMrH
OTH
03
i
6
CM
10 Tin 00 CO . t- CO CM
6 CM rH CM C CM CO rH
CO rH T^ rH O rH <3 rH
o
.CM" .cp
. rH - O
. CM
:6
%
. . CM .
: CM :
O O
OOCM
,
.
.
.
. O CM
: CM 6
rH
I 1
CM CO CO ^
(jq (jq rH rH
O C5 t- O cp Oi t>- .
rH rH CM rH ^
CO
rH
CO
t-
rH CO TH CO rH 00 O
o t*
rH
t"
00
rH rH rH
1
: cb :
rH
rH
00 CM
CO O t-
rH
: CM
CO t^
. rH O O .
: cb ob cb "
CM
9
CO
s
00 CM O O CO <* rH
t-
(M O t^ CO rH CM X
8
00 C
rt< . O O t- O O t-
h> '. O OD rH CM >b CM
rH rH T 1
cO
o t- o
rH O Tt<
CM rH rH
C5
2o?
rH rH
$
S
OOrH
CM
I-H
. . . O CO t- Ci
: : : CO rH rH CO
o"
S
So o cp
W3 O CO.C-
99
t-
ft
CO
rHO
CM
gcOO^t-OCO
t>- 1>
t- t-
CO t>
L-
L
C~
CO
.
Si^^^^g
i
L
y wine (Warrington) -j
^-y-^
I
'Q
by acids (Struve)
(Peligot)
inalysis (Rowneyand f
0) *
If
ii
is
r ^.
11
11
s^
ories in Paris (Salvetat)
. .
)
ble
wine-glass
ss (Peli
decomp
white hollow
plate (mirror) gl
glass goblet, not
bes for organic
on
mbus
Otto
Gla goblet from Petersburg, of which 8
n extracted by muriatic acid at the
French hollow wine-glass ...
Frenchglasa for the apparatus of the labor
gla
las
las
bee
l
French window
French mirror-g
g
irr
ir
English m
CMrj5
rH rH
CO
COt-OO
rH rH rH . CM '
FLUXES FLUOR-SPAR, ALKALINE CARBONATES. 19o
of their oxides, and, indeed, much less than generally
remains with the silicates.
8. FLUOR-SPAR or CALCIUM FLUORIDE is rarely employed
in assays, but in certain cases is an excellent flux, as will
be hereafter shown.
9, 10. POTASSIUM CARBONATE and SODIUM CARBONATE. It
has been already shown that they possess oxidising and
desulphurising power ; they will now be considered as
fluxes.
They are decomposed in the dry way by silica and the
silicates, with the separation of carbonic acid. The presence
of charcoal much facilitates this decomposition.
They form fusible compounds with many metallic
oxides. In these combinations the oxide replaces a cer-
tain quantity of carbonic acid ; but they are not stable
they are decomposed by carbon, which reduces the oxides,
or by water, which dissolves the alkali.
On account of their great fusibility, the alkaline car-
bonates can retain in suspension, without losing their
fluidity, a large proportion of pulverised infusible sub-
stances, as an earth, charcoal, &c.
The alkaline carbonates ought to be deprived of their
water of crystallisation for assaying purposes ; in fact, it
would be better to fuse them before use. They must in
all cases be kept in well-stopped vessels.
They may be used indifferently, but sodium carbonate
is to be preferred, as it does not deliquesce, and is gener-
ally much cheaper.
The alkaline carbonates of commerce always contain
sulphates and chlorides. In some cases this causes no in-
convenience, but there are many circumstances in which
the presence of sulphuric acid would be injurious.
Potassium carbonate can readily be procured free from
sulphate and chloride by means of pure nitre and charcoal,
as follows : Pulverise, roughly, 6 parts of pure nitre, and
mix with 1 part of charcoal ; then project the mixture,
spoonful by spoonful, into a red-hot iron crucible. The
projection of each spoonful is accompanied by a vivid
deflagration, and potassium carbonate is found in a fused
o 2
190 FLUXES XITRE, COMMOX SALT.
state at the bottom of the crucible. It must be pulverised,
separated from excess of charcoal, and kept in a dry state
for use.
Sodium carbonate may be obtained in much the same
way, substituting sodium nitrate for potassium nitrate.
Either carbonate may also be obtained in a sufficient state
of purity by repeatedly crystallising the commercial car-
bonates.
11. POTASSIUM NITRATE. Its properties have already
been pointed out. The presence of silica or of silicates
much assists its decomposition.
12. COMMON SALT (SODIUM CHLORIDE) is recommended
either mixed with flux, or placed above it, for the purpose
of preserving the substance beneath from the action of
the atmosphere, or to moderate the action of such bodies
as cause much ebullition. It is very useful in lead assays,
and is much used in the assay of silver by the wet way.
It must be previously pounded, and heated to dull redness
in a crucible, to prevent its decrepitation.
Common salt, though containing calcium and magne-
sium sulphates and chlorides, is in most cases sufficiently
pure for assaying purposes. If intended for copper assays
it must be previously purified from sulphates.
Plattner * has examined the influence of common salt
upon different oxides and sulphates. It does not act upon
uncombined lead and zinc oxides. Lead sulphate, when
melted with it at a dull red heat, becomes liquid, and
evolves vapour of lead chloride. By raising the tempera-
ture, and by giving more draught of air, the evolution of
such vapour is increased. Common salt acts upon zinc
sulphate in the same way. Antimony oxide and antimo-
nious acid heated with it at a dull red heat evolve vapour
of antimony chloride, though not in a great quantity.
Copper sulphate melted with salt at a red heat becomes
converted into copper chloride and sodium sulphate.
Copper chloride becomes vaporised if air is admitted, and
it becomes converted into copper subchloride by raising
the temperature a little, chlorine being then evolved.
* <B. u. h. Ztg.' 1854, p. 126.
WHITE, BLACK, AND RAW FLUX. 197
13. WHITE FLUX, BLACK FLUX, and RAW FLUX White flux
Is produced by deflagrating together equal parts of salt-
petre and argol (crude potassium bitartrate) ; black flux,
by deflagrating one part of saltpetre with two or three or
more parts of argol. Generally one part of saltpetre and
two and a half parts of argol are taken. The finely pul-
verised and intimate mixture for either flux, before it is
deflagrated, is called raw flux.
After the saltpetre and argol have been finely pulver-
ised and sifted separately, they are intimately rubbed to-
gether, and then deflagrated by throwing the mixture little
by little into a low-red-hot crucible, which after each
addition is lightly covered over. The deflagration may
also be conducted, though less advantageously, by filling
the crucible about two-thirds full of the raw flux and then
touching it with a red-hot coal or iron. It can only be
performed in the open air or under a flue with a strong
draught, as the tartaric acid evolves various empyreumatic
volatile matters in considerable quantity during its decom-
position.
With white flux the saltpetre suffices to burn all the
charcoal produced by the carbonisation of the tartaric
acid, and the result is therefore almost pure potassium
carbonate, if pure saltpetre and pure argol have been used.
If the latter was impure, the resulting neutral potassium
carbonate may contain much, perhaps 10 per cent., of cal-
cium carbonate. White flux works like ordinary potas-
sium carbonate, which is therefore almost always preferred
to the far more expensive flux.
With the black flux the quantity of saltpetre is not
sufficient to burn all the carbon from the argol, and there
remain therefore in the black flux, according as two, two
and a half, or three parts of argol were taken, about 5,
8, or 12 per cent, of free carbon, which is mixed in the
most intimate manner with the resulting neutral potassium
carbonate more intimately, indeed, than would be possible
by any mechanical means. This charcoal does not hinder
the fusing of the assay when the flux is used, and effects or <
promotes the reduction of the metallic oxides.
1 ( J8 WHITE, BLACK, AND RAW FLUX.
Fusion and reduction, sometimes also desulphurisation..
are the purposes for winch black flux is used, and, accord-
ing to the special character of the assay, a greater or a less
proportion of charcoal to the carbonate of potash may be
desirable, and this is to determine whether two, two and
a half, three, or more parts of argol are to be used to one
of saltpetre. As a general rule it may be stated, the more
difficultly fusible is the assay, the more potash must be
present ; and the more metallic oxide is to be reduced,
the more charcoal ; and the more also of the latter, the
more oxygen the oxide contains.
In many cases, instead of black flux, a mixture of po-
tassium carbonate and powdered charcoal, in a suitable
ratio to each other, suffices, especially if the mixture, before
use, is passed through a sieve, or otherwise very intimately
mingled. Instead of the powdered charcoal, also, a corre-
sponding (about twice or four times as large) quantity of
flour, sugar, or starch may be mixed with the potassium
carbonate. Lamp-black is, however, the best form of car-
bon. The three following fluxes are very useful :
Sodium carbonate 94 88 816
Charcoal . . . . . .6 12 184
The second is very nearly equivalent to sodium and
carbonic acid, and the third to sodium and carbonic oxide ;
but it must be observed that, whatever precautions be
taken, these mixtures never become so liquid as black
flux, because the charcoal tends very much to separate and
rise to the surface.
A mixture of 100 parts .of pure potassium carbonate
and 10 to 15 parts of wheat or rye flour is to be preferred
to black flux in case the argol contains gypsum, or the
saltpetre, sulphates, which in many cases might work in-
juriously upon the assay. If this is the case, then, in the
presence of a reducing flux, sodium sulphide is apt to form,
which, for example in the copper assay, occasions the
slagging of copper. -
Cream of tartar, carbonised by a semi-combustion until
it is reduced to half its weight, is a very good substitute
for black flux; it contains about 10 per cent, of charcoal.
FLUXES CREAM OF TARTAR.
As a perfectly general rule for the use of black flux,
and of mixtures similar to it, it is to be observed that the
crucible should never be more than two-thirds filled, as
the assay always intumesces, i.e. evolves gaseous matters,
when free carbon is present.
14. ARGOL, CREAM OF TARTAR, or POTASSIUM BITARTRATE.
When potassium bitartrate is heated in a covered crucible,
a rapid decomposition takes place, accompanied by a dis-
engagement of inflammable gases ; the substance agglo-
merates, but without fusing or boiling up. The residue
is black and friable, and contains 15 per cent, of carbon
when produced from rough tartar or argol, and 7 per
cent, from cream of tartar,
These reagents produce the same effects as black flux,
and possess more reducing power, because they contain
more combustible matter : but this is an inconvenience,
for the excess prevents their entering into full fusion when
the substance to be assayed requires but a small propor-
tion of a reducing agent. They can be used with success
in assays requiring much carbonaceous matter.
15. SALT OF SORREL, or POTASSIUM BINOXALATE, when
heated, is decomposed. It decrepitates feebly, and during
its decomposition is covered with a blue flame ; it at first
softens, and when fully fused is wholly converted into car-
bonate. When the oxalate is very pure, the resulting car-
bonate is perfectly white, and free from charcoal ; but
very often it is spotted with blackish marks. It has no
very great reducing power.
16. WHITE or MOTTLED SOAP is a compound of soda
with a fatty acid. When heated in closed vessels it fuses,
boiling up considerably, and during its decomposition
gives off smoke and combustible gases, and leaves a residue
composed of sodium carbonate with about 5 per cent, of
charcoal. Of all reducing agents, soap absorbs the great-
est quantity of oxygen ; and as the residue of its decom-
position by heat affords but little charcoal, it has the
property of forming very fluid slags. Nevertheless it is
rarely employed, because certain inconveniences outweigh
its advantages. These inconveniences are, its bubbling up,
200 REDUCING TOWER OF FLUXES.
and its extreme lightness. It also requires to be rasped, in
order to mix it perfectly with the substances it is to decom-
pose, and it then occupies a very large volume, and requires
correspondingly large crucibles. By mixing rasped soap
with potassium binoxalate or sodium carbonate excellent
reducing fluxes may be made.
Reducing Power of the Various Fluxes. By fusing
equal weights of each of the above-mentioned reducing
fluxes with an excess of litharge, the following quantities
of lead were yielded :
Common black flux, made with two parts of tartar . . 1'40
Ditto, with 2 of tartar 1'90
Ditto, with 3 of tartar . . . >>. 3 ' 80
Sodium carbonate
Charcoal .
Sodium carbonate u L
Charcoal . *" *
Sodium carbonate c/v i
Sugar . 10 /
Sodium carbonate 80 1
Sugar . . ^ 2 ' 8(
Sodium carbonate
Starch
Sodium carbonate 80 ) aA
Starch > ' ' ' ' ' ' 2<d
Crude tartar, argol
Cream of tartar .
5-60
4-50
Salt of sorrel . 851 or
Soap ... 15 / ' ' ' ' '
Sodium carbonate 85 1
Soap ... 15/
Cream of tartar, carbonised ...... 3*10
Ditto, ditto, calcined 2-20
Potassium binoxalate '90
White soda soap . . 16-00
All fluxes containing alkaline and carbonaceous sub-
stances are reducing and desulphurising, besides acting as
fluxes properly so called. They also produce another
effect which it is useful to know, viz. they 'have the pro-
perty of introducing a certain quantity of potassium or
sodium into the reduced metal. This was first pointed out
by M. Vauquelin.* He found that when antimony, bis-
muth, or lead oxide was fused with an excess of tartar, the
metals obtained possessed some peculiar characters, which
they owed to the presence of potassium.
* ' Annales des Mines.'
METALLIC FLUXES. 201
METALLIC FLUXES.
IT. LITHARGE AND CERUSE. These bodies always act as
fluxes, but at the same time often produce an alloy with
the metal contained in the ore to be assayed. Ceruse
produces the same fluxing effect as litharge. The former
is the better flux, and is very useful in a great number of
assays.
18. GLASS OF LEAD (LEAD SILICATE). Lead silicates are
preferable to litharge in the treatment of substances con-
taining no silica, or which contain earths or oxides not
capable of forming a compound with lead oxide excepting
by the aid of silica. It may be made by fusing 1 part of
sand with four parts of litharge : if required more fusible,
a larger proportion of litharge must be added.
19. LEAD BORATE. The lead borates are better fluxes
than the silicates when the substance to be assayed con-
tains free earths ; but in order to prevent them swelling
up much when fused, they must contain an excess of lead
oxide. The lead borate containing 90*56 of lead oxide
and 9 -44 of boracic acid is very good. Instead of lead
borate, a mixture of fused borax and litharge may be
employed ; it is equally serviceable.
20. LEAD SULPHATE is decomposed by all siliceous
matters, and by lime, so that when these substances are
present litharge is produced, which fluxes them.
21. COPPER OXIDE is rarely used as a flux for oxidised
matters, but is sometimes employed in the assays of gold
and silver, to form an alloy with those metals. In this
case a reducing flux must be mixed with the oxide.
Metallic copper may be used, but is not so useful, as it
cannot be so intimately mixed with the assay.
22. THE IRON OXIDES are good fluxes for silica and the
silicates. They are, however, rarely employed for that
purpose ; they are more often used to introduce metallic
iron into an alloy to collect an infusible, or nearly infusible,
metal, by alloying it with iron ; such as manganese,
tungsten, or molybdenum.
202
CHAPTEE VII.
THE BLOWPIPE AND ITS USE.
THE blowpipe was formerly only used by jewellers and
workers of metal for producing sufficient heat for solder-
ing certain small portions of their work ; and it was not
till about the year 1733 that Anton Swab applied it to
the analysis of mineral substances. Cronstedt used the
blowpipe to ascertain the difference between various mine-
ral substances as to fusibility, &c. In 1765 Yon Enge-
strom published Cronstedt's ' System of Mineralogy,' and
added to it a ' Treatise on the Blowpipe,' in which he
pointed out the process of Cronstedt.
This work attracted the attention of philosophers to
this valuable instrument, and its use became more general,
and was further developed by Bergman and Gahn.
Berzelius, after Gahn, was particularly famed for his
skill with the blowpipe, and for his improvements in the
form of apparatus ; and from his excellent work on this
subject some of the following descriptive part of Blow-
pipes, Lamps, Tongs, &c., is derived.
The common blowpipe of gasfitters, jewellers, &c., is
a tube of brass, tapering towards one end, and curved at
that extremity, which has an opening as fine as that made
by the finest needle; it is this opening which is held
against the flame of the lamp, and air is blown to it to
increase the amount of heat. In all ordinary operations
the blast is required to be kept up not more than a
minute, so that the quantity of moisture exhaled from the
lungs produces no inconvenience by stopping up the tube.
But in certain chemical operations this is exceedingly
troublesome, as a continuous blast is required, and a large
THE BLOWPIPE AND ITS USE. 203
quantity of water collects in consequence, generally suffi-
cient to mar the success of an experiment. In order to
obviate this, Cronstedt placed in the centre of his blow-
pipe a bulb, in which the greater part of the water col-
lected. This form was, however, inconvenient, because
if the jet of the blowpipe were at all inclined, even for an
instant, the water ran from the bulb and filled it. In a
series of articles communicated to the ; Chemical News,'
Mr. David Forbes, F.K.S., has given directions which are
invaluable to all who practise with this instrument. From
these we quote the following.
' BLOWPIPE. The form adopted long ago by Gahn is con-
sidered as the most convenient. FIG 67
Fig. 67 shows an improvement
made by the author upon this
form.
' In this figure it will be seen
that the arm of the jet is
double, turning upon a central
hollow axis, which allows the
blast to be directed at will
through either half of the arm,
merely by rotating the arm it-
self half round ; by having con-
sequently the two holes with
respectively a large and small
orifice, a corresponding blast
may be obtained at pleasure,
without suspending the opera-
tion.
' As a more steady and
long-continued blast is required
in quantitative operations than
could be kept up by using a
blowpipe provided with an
ordinary mouthpiece held be-
tween the lips without seriously
distressing the muscles of the
cheeks, it is quite essential that the trumpet mouthpiece
204 MR. FORBES'S BLOWPIPE:
shown in fig. 67 be adopted ; for the same reasons also
the mode of holding the blowpipe represented in fig. 68 is
FlG 68 recommended, as securing the greatest
steadiness from motion, and as greatly
assisting the muscles of the cheeks
by the external support afforded
them by the position of the thumb
pressing against the trumpet mouth-
piece.
' The nipples are turned, and bored
of three different sizes, and are made both of platinum
and of brass. The first, of platinum, contains the smallest
.aperture, and is employed for qualitative analysis ; the
second, of brass, is used for such qualitative experiments
as require a strong oxidising flame, and for heating silver,
gold, and copper, in quantitative assay ; also for roasting
copper, lead, and tin ores, the metallic contents of which
are to be accurately determined ; and the third, which is
.also manufactured of brass, has the largest bore, and is
used for the quantitative estimation of lead and tin.
' Platinum nipples are, however, always preferable to
those of brass, because by exposure to a moderate red
heat on charcoal before the blowpipe they are more easily
cleaned from the sooty particles which obstruct the aper-
ture. This method of cleansing cannot be applied to brass
nipples, owing to their rapid oxidisation ; to clean these
the operator must adapt in the opening a sharp-pointed
fragment of horn, or a small needle, ground along one-
half of its length ; by this means the aperture throng] i
which the air passes may be readily cleansed.'
The power and perfection of the blowpipe flame greatly
depends upon the internal construction of these jets or
nipples. The current of air for at least three-quarters of
an inch of the orifice should meet with no obstruction or
roughness such as a screw thread or angle so frequently
met with in blowpipes having removable nozzles. The
most perfect form for the blowpipe jet is that obtained by
slightly thickening and drawing down a piece of glass
tubing of about 5 millims. internal diameter to the required
BLOWPIPE JETS. 205
size for the jet, then cutting the contracted part cleanly
across at a point about half an inch from one of the
shoulders. We then have a jet of the shape shown in
FIG 69.
fig. 69, with which the most beautiful flame can be pro-
duced, and almost any desired pressure of air used without
hissing.
It is needless, however, to remark that the liability of
glass jets to crack and fuse renders their use impracticable
except for an oxidising flame, where the jet need not be
inserted in the flame of the lamp. Metal, therefore, must
be resorted to, and, as the superiority of the glass jet?
solely depends on their internal shape and smoothness, the
nearer the metal jets can be made to approach them the
greater will be the satisfaction in using the blowpipe.
The jets should be at least half an inch long and
FIG. 70.
coned into the blowpipes, as shown in fig. 70, not screw r ed
as is generally the case.
To obtain, therefore, the necessary internal shape for
satisfactory blowpipe jets, too much care cannot be taken
over the work of drilling, which operation can only be
successfully performed by means of a specially constructed
drill.
Any kind of flame may be used for the blowpipe, pro-
vided it be not too small ; a candle, a lamp, or gas may
be employed ; Engestroni and Bergman used common
candles in preference. Berzelius employs a lamp, which
is certainly much preferable to a candle. I have occasion-
ally employed the flame of coal gas, which answers very
well, but is not so good as that of a lamp. Berzelius says
on this subject, ' Lamps have doubtless many advantages
206 USING THE BLOWPIPE.
over candles, but are not so convenient in travelling, on
account of the escape of oil. The oil employed ought to
be the best olive or salad oil.'
The lamp recommended by Mr. Forbes has the ad-
vantage of being portable, and closes in such a manner
that no oil can escape. It is made of japanned tin-plate,
and is about 4 inches long and 1 inch wide, furnished at
one end with a wick-holder, capable of being completely
closed by a screw, and at the other with a ring of tin-plate,
which passes over the upright end of a support. It may
be mentioned that the screw cap is furnished with a
leather washer, by the aid of which it can be rendered
much tighter, and the escape of oil entirely prevented.
Mr. Forbes says that olive oil, burnt in the usual Ber-
zelius blowpipe lamp, is probably superior to any other.
Gas is not to be recommended, as it is difficult to obtain a
good reducing flame when using it. For cupellation and
such other operations, however, which only require an
oxidising flame, it is excellent.
A spirit lamp may sometimes be used in blowpipe
assays, particularly when glass tubes are employed, as in
the detection of volatile substances. In these cases it is
much more convenient ; as an oil lamp, in the first place,
blackens the tube ; and secondly, does not yield sufficient
heat, except when the blowpipe blast is employed.
It is very difficult to describe in writing a method
whereby a student may acquire the practice of using the
mouth-blowpipe; that given by Faraday * is perhaps the
clearest and most concise. He says, ' The practice neces-
sary, in the first place, is that of making the mouth replace
the lungs for a short time, by using no other air for the
blowpipe than that contained in it.' This practice is sim-
ple in itself, and easy to acquire, but, as before stated,
difficult to describe. Let the student first observe that it
is easy after having closed the lips to fill the mouth with
air, and to retain it so, at the same time respiration may
be carried on ; and if, while the mouth is in this state, a
blowpipe be introduced between the lips, it will be found
* ' Chemical Manipulation.'
USING THE BLOWPIPE. 207
that the small quantity of air which its jet allows to pass
through it will be amply supplied for ten or fifteen
seconds by the quantity contained in the mouth ; and this
being repeated a few times, a ready facility for using the
blowpipe, independent of the lungs, will be acquired.
This step being taken, the next is to combine this pro-
cess with the ordinary one of propelling air directly from
the lungs through the mouth, in such a way that when
the action of the lungs is suspended during inspiration,
the blast may be continued by the action of the mouth
itself, from the air contained within it. The time of four-
teen or fifteen seconds, during which the mouth can supply
air independently of the lungs, is far more than that re-
quired for one or even many more inspirations ; and all
that is required to acquire the necessary habit is the power
of opening and closing the communication between the
mouth and the lungs, and between the air and the lungs,
at pleasure.
The capability of closing the passages to the nostrils is
very readily proved : every one possesses and uses it when
he blows from the mouth, and that of closing or opening
the mouth to the lungs may be acquired with equal readi-
ness. Applying the blowpipe to the lips as before, use the
air in the mouth to produce a current, and, when it is
about half expended, open the lungs to the mouth, so as
to replace the air which has passed through the blowpipe ;
again cut off the supply, as at first, but continue to send
a current through the instrument, and, when the second
mouthful of air is gone, renew it as before from the lungs.
To some this may be difficult ; b ut if the preceding
instructions be followed and persevered in for a short time,
the learner will soon find that he can keep up a continuous
blast from ten minutes to a quarter of an hour, without
any other inconvenience than the mere lassitude of the
lips caused by compressing the mouthpiece of the instru-
ment, and this may be avoided by using the trumpet
mouth-piece as recommended by Mr. Forbes.
After having conquered the difficulty of keeping up a
continuous blast, the student must learn how to attain the
208 REDUCTION AND OXIDATION.
maximum of heat with the least exertion to himself. The
chief points to be observed are, neither to blow too
fiercely nor too gently : in the first case, the force of the
blast would carry away heat by the quantity of cold air
thrown into the flame ; and, in the second, a sufficient
amount of heat would not be obtained, because a less
amount of air would pass into the flame than that required
for perfect combustion.
The highest degree of temperature is required in test-
ing the fusibility of many bodies, as also in the reduction
of certain oxides, as those of iron, tin, &c. We have yet
another class of phenomena to describe, which do not
essentially depend on a high temperature ; these are the
processes of reduction and oxidation. In order to explain
and point out the best methods of effecting these two
objects, it will be necessary to enter somewhat into the
nature of flame ; this will be done as briefly as is consis-
tent with perspicuity. The species of flame examined
will be that of a candle, as it is a similar one to that
with which the blowpipe operator will have to experi-
ment.
On careful examination it will be found that the flame
of a candle or lamp may be divided into four distinct por-
tions : first, a deep blue ring at the base ; this consists of
the vapour of the combustible, which can hardly burn
because it has not acquired a sufficient temperature ;
secondly, a dark cone in the centre ; this is also the vapour,
but heated intensely, not, however, in a state of combus-
tion, on account of the absence of air ; thirdly, of a very
brilliant envelope, which surrounds the dark parts just
mentioned ; this is the partially consumed vapour at a very
high temperature ; the luminous property it possesses is
due to the precipitation and subsequent ignition of par-
ticles of solid carbon ; and fourthly, of an almost invisible
envelope which surrounds the luminous portion ; this is
the substance of the combustible in full ignition, it here
mingles with the atmospheric oxygen, and is consumed.
The highest degree of temperature in the whole flame is to
be found at the point of contact between the luminous and
KEDUCTIOX AXD OXIDATION. 209
this part. It must be particularly borne in mind that the
inner portions of the flame have an excess of carbonaceous
matters, and the outer an excess of oxygenated matters.
Having premised this much, we will examine the nature
of the flame of a candle when acted on by the blowpipe
blast, and ascertain how far it is altered, and what are the
properties of its separate parts in relation to their oxidis-
ing and reducing powers. Supposing the lighted lamp or
candle be ready and neatly snuffed, place the nozzle of the
blowpipe just in the edge of the flame, and about the six-
teenth of an inch above the level of the wick : when
things are in this state, blow gently and evenly through
the blowpipe, and a conical jet or dart of flame will be
produced, which, when formed in a steady atmosphere,
free from accidental draughts and currents, will be found
to consist of two essential parts the inner cone, blue,
small, and well defined ; the outer, brownish and vague.
The greatest intensity of heat is found a little beyond the
apex of the blue flame ; it is there, also, reduction takes
place. The outer flame is formed by the complete com-
bustion of the combustible matter of the inner ; and at that
place, and just beyond it, oxidation takes place.
Oxidation^ as before stated, takes place at the extremity
of the outer flame, hence it is termed the oxidising flame ;
in it all the combustible portions are supersaturated with
oxygen. In general the further the substance to be oxidised
can be placed from the extremity of the flame, the better
the operation proceeds, provided always that the necessary
temperature be maintained. Dull redness is the best suited
for oxidation.
Reduction. In this operation the jet of the blowpipe
must be introduced into the body of the flame, so as only
to produce a small dart ; and a jet having a smaller hole
than that used for oxidation ought to be employed. By
operating thus a more brilliant flame than the last is pro-
duced ; it is the result of a less perfect combustion, and
therefore contains a large amount of carbonaceous matter,
fitting it more especially for the purpose of separating
oxygen from all metallic bodies.
p
210 AUXILIARY BLOWPIPE APPARATUS.
Berzelius says, ' The most important point in blowpipe
assays is the power of producing oxidation and reduction
at will.' Oxidation is so easy, that to do it requires only
to read a description of it ; but reduction requires some
practice, and a certain knowledge of producing various
kinds of blasts. One of the best methods of exercise in
this operation is to take a small grain of tin, and place it
on charcoal ; then direct the blowpipe dart upon it it
will soon fuse ; and if the operator has not produced a
good reducing flame, it will become covered w r ith a crust
of oxide ; so that it becomes a witness against him each
time this happens. The nature of the flame must be altered
until, by observation, the proper kind is produced at will.
The longer the button of tin is kept bright, the better and
more expert the operator.
AUXILIARY BLOWPIPE APPARATUS, ETC.
Supports. The support is the substance destined to
hold the material to be assayed whilst under the influence
of heat. From this it will be seen that an exceedingly
refractory body must necessarily be employed, so as not
to give way under the excessive heat ; and also (with the
exception of charcoal) it ought to have no chemical action
on the substances placed in contact with it. Supports
may be divided into combustible and incombustible :
the former is charcoal ; and for the latter, metal, glass,
and earthenware, and in some cases certain minerals, have
been employed.
Charcoal. Mr. Forbes gives the following excellent
description of the preparation of charcoal for blowpipe
purposes : c lt is extremely difficult to obtain, in England,
charcoal fit for blowpipe operations without special pre-
paration. The charcoal sold is generally of hard wood,
badly burnt, full of cracks, and decrepitating upon appli-
cation of heat. Good charcoal should be soft, yet com-
pact, and without cracks, and is best made from fir or
pine. Where good charcoal cannot be obtained, it can be
made artificially by moulding charcoal powder agglutinated
CHAKCOAL APPARATUS. 211
by some starch paste, and, after desiccation, burning the
pieces in a crucible filled with sand.
' For the preparation of the charcoal used as a support
for the assays, the instruments represented in fig. 71 are
required, all of which are fitted in the universal handle a,
which is shown in this figure p IG> 71<
as holding the largest charcoal
borer, a section and plan of
which are shown in b. This
large borer is employed for
forming the deep holes in the
charcoal used in the blowpipe
furnace, and which serve to
contain the clay crucibles or
capsules in which the assays
are fused. The blast-holes in
the charcoal inside the blowpipe
furnace are bored out by the
gouge-shaped borer d, which
also serves for making small
holes or grooves in charcoal for
general purposes. The smaller
borer c is most useful, particularly in boring out the holes
for receiving the soda paper cornets containing the assay
for reduction. The saw-knife e also fits into the same
handle, and is used for trimming and sawing across the
charcoal pieces, having coarse saw-teeth in front, whilst
the back presents a sharp knife-edge. The figures are
all drawn to one-half of the real size.'
Platinum. This metal is much employed as a support
in cases where charcoal would be injurious by its reducing
power. It is used in three forms, viz. wire, foil, and as a
spoon or small capsule.
Wire. A moderately strong wire of platinum, about 2
inches long, and curved at one end, is used with great
advantage in many quantitative examinations. The curve
serves as a support in all experiments on tests of oxidation
and reduction, where alteration of colour only is to be
observed. This support can be relied on, for it is totally
p 2
212 CHARCOAL APPARATUS.
free from the false varieties of colour which are too often
perceptible when the assay rests on charcoal. In the
treatment of metals, or in reduction tests, where an easily
melted body is to be operated upon, charcoal must, how-
ever, be used. It is necessary to have at hand several
platinum wires, so as to proceed to another experiment
without being obliged to forcibly remove the adhering
borax glass, or to wait for its solution in hydrochloric
acid, which is the better mode. If the platinum loop
melts with the reagent, it must be cut away, and a new
one formed. A wire can be used for a very long time,
and when it becomes too short to be held between the
fingers the straight end may be fastened into a cork, or a
piece of glass tubing.
The platinum spoon (see fig. 72) and foil are used in
much the same way ; but as charcoal and the platinum wire
FIG. 72. answer every purpose, it
will be unnecessary to de-
scribe their use further :
small iron spoons of the
above form are also made,
and are very useful in
cases where the presence of
iron is not objectionable.
Other instruments, as forceps, hammer, anvil, agate
mortar, scissors, &c., are sufficiently familiar to every-
body not to require description. Special apparatus
required for any operation will be described in the course
of the processes.
Colonel Eoss recommends as a support in certain cases
a thin rectangular slip of aluminium plate, 4 inches by 2,
half an inch of the lower end of which is turned up at an
angle of 80 as a rest for the assay. A fragment of the
substance of the size of a pea is laid upon the edge close
to the angle, and heated very slightly about half an inch
from the top of a pure blue flame. The sublimate obtained
should be examined from time to time, increasing the heat
after each examination, till nothing more is obtained.
For a full description of this method, by means of which
BLOWPIPE EEAGENTS AND FLUXES. 213
substances existing in mixture or combination may be
usefully separated by taking advantage of their different
degrees of volatility, the reader is referred to 'Alphabetical
Manual of Blowpipe Analysis ' (Triibner & Co.)
Asbestos cardboard, according to Mr. W. M. Hartley,
may also serve for supports. This substance resembles
greyish cardboard, but has a soapy feel, like steatite ; it
can be used for making crucible supports, sand-baths,
muffles, retort supports, &c. ; it can be cut with cork
borers or scissors ; by moistening with water it can be
moulded to any shape. After moistening it should be
gradually dried and ignited, to get rid of organic matter.
It stands the ordinary wear and tear of the laboratory
well. It is formed principally of asbestos fibres. It can
be obtained from the manufactory, 31 St. Vincent Place,
Glasgow, at 4s. a pound.
REAGENTS AND FLUXES.
BLUE LITMUS PAPER is used for detecting free acid in
solution, its colour being changed to red.
REDDENED LITMUS PAPER is used for the detection of free
alkali, its colour being restored to blue.
BRAZIL-WOOD PAPER is used for detecting hydrofluoric
acid, being tinged straw-yellow when immersed in a very
dilute solution of this acid.
TURMERIC PAPER is used for detecting free alkalies : the
change produced is very characteristic, its bright yellow
colour becoming dark brown.
NITRIC ACID is employed in the solution of various
metals, alloys, and ores, and for the discrimination of
certain precipitates ; also as an oxidising agent.
ZINC is principally employed for the reduction of
antimony and tin.
COPPER is used for the reduction of mercurial salts and
for the detection of arsenious acid. It is also used to
detect the presence of nitric acid.
IRON WIRE is employed to precipitate many metals, and
in the separation of sulphur and the fixed acids from any
214 BLOWPIPE REAGENTS AND FLUXES.
substance with which they may be combined. The metals
which can thus be precipitated, or deprived of sulphur,
are copper, lead, nickel, and antimony. For instance, if a
small piece of iron (pianoforte) wire be placed in a sub-
stance in fusion, and acted upon by the blowpipe, it
becomes covered with the reduced metal. The latter
sometimes appears as small globules.
Iron has the property of reducing phosphorus from
phosphoric acid or the phosphates, giving rise to a phos-
phide of iron, which forms on fusion a white, brittle,
metallic globule.
POTASSIUM CYANIDE. This is a most useful flux. MM.
Haidlen and Fresenius say, ' We have examined its action
on many oxides, sulphides, salts, &c., in reference to its
use as a reagent combined with the blowpipe. We prefer,
in general, a mixture of equal parts of anhydrous soda
and potassium cyanide. This mixture was employed on
account of the great facility with which the pure cyanide
fuses. In acts, in general, so very similarly to pure soda,
that it would be superfluous to describe singly the changes
which each individual body appeared to undergo when
exposed to its action. We cannot, however, pass over the
following especial advantages which it possesses as com-
pared with soda. First, reductions are obtained with such
great facility that the least practised operator may execute
reductions which would otherwise be very difficult : for
instance, the reduction of tin from either its oxide or
sulphide ; and secondly, that the fused mixture of potas-
sium cyanide with soda is so easily absorbed by the char-
coal, that the grains of reduced metal can always be most
distinctly perceived, and may be most easily separated
therefrom for further examination. 7
SODIUM CARBONATE, called, for the sake of brevity, Soda.
The plain carbonate or the bicarbonate may be in-
differently employed ; but in either case it is absolutely
necessary that it be free from sulphates.
There are two objects in view in the employment of
soda as an auxiliary to the blowpipe : first, to ascertain if
the substances combining with this body be fusible or
REAGENTS IN THE DRY WAY. 215
infusible ; and secondly, to facilitate the reduction of
certain metallic oxides.
In the fusion of substances with soda there are many
things to observe. The necessary quantity must be taken
on the moistened point of a knife, and kneaded in the palm
of the hand, so that it may form a coherent mass. If the
body under examination be pulverulent it must be incor-
porated with it, but if in lump it must be placed upon it,
forcing it slightly into the moistened soda ; then carefully
heated on the charcoal with a gentle flame, until thoroughly
dry ; and lastly it may be fused. It generally happens
that the soda, at the instant of fusion, is absorbed by the
charcoal : but this does not hinder its action on the assay ;
for if it be fusible with soda, the latter comes to the surface
and attacks it, finally forming a liquid globule. If the
substance be infusible in soda, but decomposable by it, it
alters its appearance without entering into fusion. But,
however, before pronouncing any substance to be infusible
by soda, the flux ought to be mixed with the pulverised
substance. If in these trials too little soda be taken, a
portion of the substance remains solid, and the rest forms
a covering of transparent glass ; if too much, the bead of
glass becomes opaque on cooling. - It sometimes happens
that the assay contains a substance which, being insoluble
in the glass of soda, prevents it becoming transparent.
Then, in order that we may fall into no error respecting
the nature of the glass, it becomes necessary in the two
last-mentioned cases to add a new quantity of the body
under examination, and then ascertain if a limpid globule
cannot be obtained. In general it is the best method to
add the soda by successive small doses, and note the
changes produced by each addition. It sometimes happens,
in this kind of assay, that the glass becomes coloured at
the moment of cooling, and finally takes a yellow or deep
hyacinth-red ; it even becomes occasionally opaque and
yellowish-brown. These phenomena indicate the presence
of sulphur, either in the assay or the soda employed. If
the same colour be constantly produced by the same soda,
it is a proof that it contains sulphate of soda ; it must then
2K5 SEDUCTION BEFORE THE BLOWPIPE.
be discarded ; but if it gives generally a colourless glass, it
is the substance under assay that contains sulphur or sul-
phuric acid.
Reduction of Metallic Oxides. This species of assay,
by which quantities of reducible metals, so small as to
escape ordinary wet analyses, can be detected, is the most
important discovery Gahn made in the application of the
blowpipe.
If a small quantity of native or artificial oxide of tin
be placed on charcoal, it requires a long blast and a skil-
ful operator to produce a grain of metallic tin ; but if a
small quantity of soda be added, the reduction takes place-
readily, and so completely with pure oxide, that the whole
is transformed into a button of tin. From this it is
certain that the presence of soda favours the decom-
position.
The action, however, can be explained thus, as Berzelius
himself hints : the red-hot charcoal reacts upon the sodium
carbonate, producing by its reduction a certain amount of
sodium, which by its strong attraction for oxygen seizes on
that contained by the metallic oxide which is required to
be reduced. If the metallic oxide contain an unreducible
substance, the reduction of the former becomes difficult ;
but if a little borax be added the reduction takes place as
usual.
This assay is very easy of execution, and the metal is
readily recognised, as by previous assays the nature of it is
somewhat ascertained, and the reduction but confirms the
previous idea.
If the metallic oxide be mixed with such a quantity of
non-reducible substances that its nature cannot be ascer-
tained by previous experiment, how can it be proved that
a reducible metal is present ?
Gahn has solved this question in a very simple manner.
' After having pulverised the substance to be assayed, it
is kneaded in the palm of the hand with moistened soda,
and the mixture placed on charcoal and exposed to a good
reducing flame ; a little more soda is then added, and the
blast recommenced. As long as any portion of the sub-
REDUCTION BEFORE THE BLOWPIPE. 217
stance remains on the charcoal, soda is added in small por-
tions, and the blast continued until the charcoal has ab-
sorbed the whole of the mass. The first quantities of soda
serve to collect the metallic particles scattered in the sub-
stance to be assayed, and the final absorption of the latter
completes the reduction of any that may remain in the state
of oxide.
' This done, the burning charcoal is extinguished with
a few drops of water ; then having cut out the part which
absorbed the soda and assay, grind it to a very fine powder
in an agate mortar. This powder is then washed with
'water to carry away the finest portion of the charcoal.
The grinding and washing are repeated until all the char-
coal is washed away. If the substance contained no
metallic body, nothing will remain in the mortar after this
last washing. But if it contained the smallest quantity of
reducible matter, it is found at the bottom of the mortar,
as small brilliant plates if it be malleable, or as a fine pow-
der if it be brittle or not fusible. In either case the bottom
of the mortar is covered by metallic traces, resulting from
the friction of the particles of metal against its sides (pro-
vided that the quantity of metal contained in the sample
be not too small). The flattening of almost imperceptible
globules of any malleable metal converts them into shining
discs of a perceptible diameter. In this manner may be
discovered by the blowpipe, in an assay of ordinary size,
less than a half per cent, of tin, and even less than that of
copper.'
The following points in this class of assay ought to be
particularly attended to. First, to produce the strongest
possible flame, taking care that it covers every part of
the assay < Secondly, to leave none of the metal in the
charcoal, or lose the smallest quantity in the collection.
Thirdly, to well grind the carbonaceous mass. Fourthly,
to decant very slowly, so that only the lighter parts may
be carried away by the water. Fifthly, not to judge of the
result until the whole of the charcoal has been removed,
for a small quantity remaining suffices to hide the metallic
particles ; and, moreover, the particles of charcoal, viewed
218 BORAX,
in a certain light, have themselves a metallic lustre, which
will deceive an inexperienced eye. Sixthly and lastly,
not to trust to the naked eye, however plain the sample
may be, but always examine by the aid of a good micro-
scope.
The metals reducible by this process are (besides the
noble metals) molybdenum, tungsten, antimony, tellurium,
bismuth, tin, lead, thallium, copper, nickel, cobalt, and
iron. Amongst these, antimony, bismuth, thallium, and
tellurium volatilise easily when they are exposed to a
strong heat. Selenium, arsenic, cadmium, zinc, and mer-
cury volatilise so completely that they cannot be collected
except by means of a small subliming apparatus.
The reduction can always be effected the first time
when the assay contains from 8 to 10 per cent, of metal ;
but in proportion as the standard decreases more attention
and care must be paid to the washing and recognition of
the reduced metal in the mortar. A good system of prac-
tice in this experiment is to employ any cupreous sub-
stance, and make on it a great number of experiments,
taking care to mix it each time with a substance contain-
ing no copper ; thus the metallic value will diminish at
each new assay, until at last no copper can be found.
If the substance to be assayed contains several metals,
the reduction of their oxides must be made in globo, and
a metallic alloy obtained. Some, small in number, are
reduced separately. For instance, copper and iron give a
regulus of each metal ; copper and zinc, the first gives a
regulus of copper, whilst the latter volatilises. But when
the result of the operation is an alloy, recourse must be
had to the reactions produced by other fluxes to ascertain
its constituents.
BORAX (SODIUM BIBORATE). The borax of commerce
must be dissolved in hot water and recrystallised before it
can be used in blowpipe analysis.
Borax may be employed either in crystals, the requisite
size for an assay, or in a pulverulent form ; in this case it
may be taken up on the moistened point of a knife. Some
operators prefer fusing the borax before use, in order to
FUSION WITH BORAX. 210
V y-> , . "~,-\K\\Pv .
drive off its water of crystallisation, and thus
tumefaction of the crystal on charcoal.
Borax is employed in the solution or fusion of a variety
of substances. It is best to commence by acting upon a
scale of the substance to be examined ; because if a pow-
der be employed the resulting action cannot be so well
ascertained. The following phenomena are to be carefully
watched, for in treating any substance with borax it must
be particularly noted whether the fusion takes place rapidly
or otherwise ; without motion or with effervescence ; if
the glass resulting from the fusion is coloured, and if that
colour changes in the oxidising or reducing flame ; and,
lastly, if the colour diminishes or increases on cooling, and
if, under the same circumstances, it loses or retains its
transparency.
Some substances possess the property of forming a
limpid glass with borax, which preserves its transparency
on cooling, but which, if slightly heated in the exterior
(oxidising) flame, becomes opaque and milk-white, or
coloured when the flame strikes it in an unequal or inter-
mittent manner. The alkaline earths, as yttria, glucina,
zirconia ; the oxides of cerium, tantalum, titanium, &c.,
belong to this class. In order to be certain of this result
we must assure ourselves that the glass is saturated to
a certain point with either of the above class of bodies.
The same thing, however, does not happen with silica,
alumina, iron, manganese, &c., oxides, and the presence of
silica prevents the production of this phenomenon with the
earths ; so that alone they present this peculiar appearance
with borax ; but when combined with silica (as natural
silicates, for instance) no such effect is produced. This
operation has received the name of flaming r , and any
substance thus acted upon is said to become opaque by
flaming.
SODIUM AMMONIA-PHOSPHATE (MICROCOSMIC SALT) is obtained
by dissolving 16 parts of sal-ammoniac in a very small
quantity of boiling water, and mixing with it 100 parts of
crystallised sodium phosphate, dissolving the whole with
heat, filtering the boiling liquid ; during cooling the double
220 FUSION WITH MICROCOSMIC SALTS.
salt crystallises. When microcosmic salt is not pure it
forms a glass which becomes opaque by cooling. It is then
necessary to dissolve it in a small quantity of w r ater and
recrystallise it.
It may be collected in large crystals, or in a pulverulent
state. The crystals are in general of a suitable size for
ordinary assays. Placed on charcoal, and submitted to the
blowpipe flame, it bubbles and swells up, giving off am-
monia ; that which remains after this treatment is an acid
sodium phosphate, which fuses readily, and forms on cool-
ing a transparent and colourless glass. As a reagent, it
acts principally by its free phosphoric acid ; and if the salt
be employed in preference to the acid, it is because it is
less deliquescent, costs less, and passes readily into the
charcoal. By means of microcosmic salt we ascertain
the action of free acids on any substance we may wish to
assay. The excess of acid it contains combines with all
bases, and forms a class of double salts, more or less fusible,
which are examined as to their transparency and colour.
In consequence this flux is used more particularly in the
detection of the metallic oxides, most of which impart to
it very characteristic colours. This flux exercises on acids
a repulsive action. Those which are volatile, sublime ; and
those which are fixed remain in the mass, dividing the base
with the phosphoric acid, or yielding it up entirely ; in
w r hich case they are suspended in the glass without being
dissolved. In this respect microcosmic salt is a good test
for silicates ; for by its aid silica is liberated, and appears
in the glass as a gelatinous mass.
POTASSIUM NITRATE (NITRE), in long and thin crystals, is
employed in hastening the oxidation of those substances
which do not readily combine with oxygen in the exterior
flame. It is used as follows : The point of a crystal is thrust
into the fused bead ; but in order to prevent the cooling
of the latter the crystal is held by a pair of pliers, so that
when the bead begins to cool it may be withdrawn, the
bead reheated, and the crystal employed as before, until
the desired effect is produced.
POTASSIUM BISULPHATE is employed in the detection of
FUSION WITH BISULPHATES. 221
lithia, boracic acid, nitric acid, hydrofluoric acid, bromine,
and iodine. It separates baryta and strontia from the earths
and metallic oxides.
SODIUM BISULPHATE. Professor J. Lawrence Smith* 1 has
suggested the use of sodium bisulphate as a substitute for
the potassium bisulphate in the decomposition of minerals,
especially the aluminous minerals. He finds that ' while
the soda salt gives a decomposition at least as complete as
the potash salt, the melted mass is very soluble in water,
and in the future operations of the analyses there is no
embarrassment from a deposit of alum.
6 The ordinary commercial article is not sufficiently
pure for use, and he prepares it from pure sodium carbo-
nate, or sodium sulphate that has been purified by recrys-
tallisation. In either instance pure sulphuric acid is
added in excess to the salt in a large platinum capsule, and
heated over a flame, until the melted mass, when taken
up on the end of a glass rod, solidifies quite firmly. The
mass is then allowed to cool ; moving it over the sides of
the capsule will facilitate this operation. When cool it is
readily detached from the capsule, is then broken up, and
put into a glass stoppered bottle. In almost every instance
where we have been in the habit of using potassium bisul-
phate the sodium bisulphate can be substituted.'
VITRIFIED BORACIC ACID is used to ascertain the pre-
sence of phosphoric acid and small portions of copper in
lead alloys. For quantitative analysis it is generally used
to ascertain the quantity of copper contained in a lead ore,
and also the amount of copper united with various metals.
COBALT NITRATE in solution ought to be free from
arsenic and nickel, and the solution must be moderately
strong. It is used as a test for alumina, magnesia, tin, and
zinc, by the blowpipe.
NICKEL OXALATE is used in qualitative examinations for
the detection of potash in a salt which also contains soda
and lithia.
COPPER OXIDE is employed to detect the presence of
hydrochloric acid and chlorine.
* ' American Journal of Science and Arts.'
222 AMMONIUM FLUOEIDE.
SILICA is, with soda, an excellent test for the presence
of sulphuric acid ; and when in combination with borax
or soda, separates tin from copper.
TURNER'S FLUX. This is a mixture of potassium bisul-
phate and fluor-spar. It is used for producing coloured
flames before the blowpipe.
AMMONIUM FLUORIDE has been proposed as a blowpipe
reagent by Professor N. W. Lord, of the Ohio State Uni-
versity. The use of bisulphate of potassium and fluor-
. spar as a reagent for developing the flame coloration of
boron is well known ; but the alkali present prevents the
application of the method for liberating some other bodies
in the same way. Fluoride of ammonium, on the contrary,
having all the value of fluor-spar as a source of fluorine,
admits of much easier application, and is a most useful
reagent for detecting the alkalies, boron, and other similar
bodies in their mineral combinations. The method of using
the reagent is simple. For testing felspar, or similar silicates,.
a little of the powdered mineral is mixed with this reagent,
then placed on a piece of platinum and moistened with
sulphuric acid ; the mixture allowed to stand a few
moments, or else gently warmed, taken upon a loop of
platinum wire, and tested either in the blowpipe flame or
in a Bunsen burner, being dried a little on the wire first.
The alkali flame is nearly as well shown as with the pure
salts. As the fluoride of ammonium is permanent, is
easily obtained free from alkalies and boron, and can be
kept indefinitely in a small wooden box, it is always easy
to use.
As a test for boron the reaction is of surprising deli-
cacy. The fact that the fluoride of boron is volatile at a
temperature far below that required for alkalies, permits
thus its detection in borax or any alkaline compound..
To a drop of sulphuric acid placed on a platinum crucible
cover, a few grains of the fluoride should be added, and
then the mineral (powdered) to form a paste. This is
taken as before described on a platinum loop. It should
be heated gently until it stops ' sputtering ' from escape
of free acid and water (but on no account heated to
CALCIUM FLUORIDE. 223
redness), then brought not in, but near the lower part of
the flame of a Bunsen burner or a good blowpipe flame.
A bright green coloration is at once given, untinged by
soda-yellow. The coloration is, of course, evanescent,
and disappears before the assay is red-hot. A little prac-
tice is needed to find the right part of the flame, to get
the right heat, and at the same time to draw the volatile
boron compound into the heated zone.
This reaction shows boron very strongly in all speci-
mens of tourmaline. With a hand spectroscope the appli-
cation of this method gives instant proof of the existence-
of boron, potassium, sodium, and lithium, even in very
small traces in rocks.
Borax treated in this way gives a bright green flame,,
almost like copper.
CALCIUM FLUORIDE (FLUOR-SPAR) AND CALCIUM SULPHATE
(GYPSUM). These two bodies (deprived of water) are used
to indicate the presence of each other. If a small piece
of gypsum be ignited in contact with a similar piece of
fluor-spar, they soon liquefy at their points of contact ;
they then combine, and form, by fusing, a colourless and
transparent bead of glass, which becomes enamel white on
cooling. Calcium fluoride is thus employed as a test for
gypsum, and vice versd.
Mr. S. D. Poole prefers a mixture of calcium sulphate
(selenite) and fluor-spar (in the proportions of two-
parts of the former to one of the latter) to Turner's flux
described above. It forms an easily fusible bead, which
by itself gives only a very faint dull red tint (calcium) to-
the flame, but which renders the presence of many elements
which give colour most beautifully evident. Thus small
portions of lepidolite, petalite, &c., mixed with this flux
impart the fine carmine tint of lithium ; copper and stron-
tium show their well-known colours, especially after con-
tinued blowing. Potassium and sodium minerals (felspar
and albite) are at once distinguished.
This flux is of more limited applicability than that of
Turner, because there is no provision in it for the libera-
tion of hydrofluoric acid.
224 BONE ASH.
It serves, when mixed with bisulphate of potash, to
detect lithia and boracic acid in their various combina-
tions.
BONE ASHES are employed in the cupel lation of gold and
silver. Harkort reduced them to many states of minute
division by the processes of sifting and washing. The bones
are burned until they become perfectly white, and then
freed from any carbonaceous matter that may have ad-
hered to them. This being done, they are pulverised in
a mortar, and the finer portions separated by a sieve. The
remaining powder is then thrown upon a filter, and treated
with boiling water, which extracts the soluble matter. The
washing, which is then resorted to, is for procuring the
bone ashes of a more uniform degree of fineness. The
mass from the filter is mixed with water in a cylindrical
glass, allowed to settle for a few minutes, and then de-
canted ; the coarser powder is deposited at the bottom of
the vessel, while the finer passes over suspended in the
water. By repeated decantations in this way deposits are
obtained of different degrees of fineness ; the last, or that
which remains longest floating through the liquid, being
the finest. The resulting powders must be kept in sepa-
rate bottles. The coarser ashes are used for the cupella-
tion of rich silver ores, and the finer for assaying ores
in which only a minute quantity of gold or silver is
present.
PROOF LEAD is made use of in cupelling argentiferous or
auriferous substances ; it must be free from silver. Dumas
states that the best method of obtaining lead in this desir-
able state is to decompose the best white-lead by means
of charcoal, a it is then impossible for it to contain any
other metal.
TINFOIL is employed to reduce certain peroxides to the
state of protoxide. When it is used, a small roll, about a
quarter of an inch long, is plunged into the fused button,
and heated strongly in the reducing flame : the desired
effect is then produced.
DRY SILVER CHLORIDE. Herr H. Gericke proposes the
employment of this compound in qualitative blowpipe
SILVER CHLORIDE IX BLOWPIPE ASSAYS. 225
assays. In an elaborate paper on this subject, communi-
cated to the 'Chemical Gazette' (vol. xiii. p. 189), he
says :
' Amongst the phenomena which characterise different
bodies before the blowpipe, and serve for their distinction,
the colour of the flame is of no small importance. This
power of colouring the blowpipe flame is not, however,
exhibited by all bodies with sufficient intensity to enable
them to be distinguished by it with certainty ; and certain
substances are consequently usually employed, such as
muriatic acid with baryta, strontia and lime, or sulphuric
acid, partly to form and partly to set free volatile com-
pounds. By this means, however, although the intensity
of the coloration is heightened, its duration is not in-
creased, as these acids, and particularly muriatic acid,
evaporate for the most part before they have acted suffi-
ciently, so that the coloration lasts only for a few mo-
ments. This defect may be got over by the employment,
instead of the volatile muriatic acid, of a chloride, which
will retain the chlorine at a high temperature, so that it
may only be set free by degrees in small quantities, while
the body forming its base may be without action upon the
colouring power of the body under investigation. For this
purpose chloride of silver appears to be the best, espe-
cially as it may readily be prepared in a state of purity.
The best plan is to stir it with water into a thick paste,
and keep it in a bottle.
' In regard to the action of chloride of silver upon the
coloration of the blowpipe flame, I have investigated
several compounds of potash, soda, lithia, lime, baryta,
strontia, copper, molybdenum, arsenic, antimony, and lead,
and mixtures of these substances. Chloride of silver, of
course, has no action upon borates and phosphates, both
of these acids being amongst those which offer the most
resistance to the action of heat.
' For a support, I employed first of all platinum wire,
but this is soon alloyed by the metallic silver which sepa-
rates, and thus rendered useless in testing metals. Silver
wire is too readily fusible, and also difficult to obtain free
Q
226 USE OF CHLORIDE OF SILVER
from copper, which may give rise to errors when in con-
tact with chloride of silver. For these reasons, iron wire
is best fitted for experiments with chloride of silver, as
from its cheapness a new piece may be employed for each
experiment, while the silver may readily be obtained in
the form of chloride from the broken pieces, If the size
of the fragment under examination be sufficient, the plati-
num forceps may be employed.
' The results at which I arrived, by the employment
of chloride of silver, in comparison with those obtained
without this reagent, are as follows :
' With potash compounds, such as saltpetre, potashes,
&c., the flame is decidedly of a darker colour with chloride
of silver ; and even in ferrocyanide of potassium, which,
when treated by itself with the blowpipe, colours the
flame blue, the addition of chloride of silver produces a
distinct potash coloration.
' The action of chloride of silver upon soda salts is not
so favourable ; for although with some, as nitrate of soda,
common soda, and labradorite, the flame acquires a more
intense yellow colour by the addition of chloride of silver,
this reagent produces no observable difference with other
soda compounds, such as sulphate of soda and analcime.
This also applies to the compounds of lithia, some of
which give a finer purple-red colour on the addition of
chloride of silver, whilst upon others it has no such effect.
' With lime compounds chloride of silver acts favour-
ably upon the colouring power. Thus the addition of
chloride of silver to calcareous spar or gypsum (in the
reduction flame) gives the flame a more distinct yellowish-
red colour, but stilbite gives no coloration either with or
without chloride of silver. With fluor-spar the coloration
cannot well be observed, as this decrepitates too violently
under the blowpipe.
' The action of chloride of silver upon compounds of
baryta and strontia is decidedly advantageous, as both the
intensity of the coloration and its duration leave nothing
to be desired. Siliceous celestine, which, when heated by
itself in the forceps, scarcely coloured the flame, imme-
IN BLOWPIPE ASSAYS. 227
diately produced a permanent red coloration, when heated
with chloride of silver.
4 Although it appears from the preceding statements,
that the employment of chloride of silver presents no
advantage with some substances, it may be used with
good results in the treatment of mixtures of alkalies and
earths.
4 Thus, with petalite alone the lithia coloration is first
produced, and a slight soda coloration is afterwards
obtained ; whilst with chloride of silver the soda coloration
appears very distinctly after that of the lithia. With
lithion-mica alone a very distinct lithia coloration is
presented ; but in the presence of chloride of silver a
colour is first produced which may lead to the conclusion
that potash is present, but the lithia coloration is weak-
ened. Ryacolite, heated by itself in the blowpipe flame,
only gives a distinct soda coloration ; but with chloride of
silver a slight potash coloration is first produced, and the
colour of soda then appears very distinctly ; the lime
contained in it cannot, however, be detected by the colora-
tion of the flame.
4 Chloride of silver may be employed with still greater
advantage with the following metals, but in these cases it
is particularly necessary that the operator should become
familiar with the colour produced by each individual
substance.
4 With copper compounds, such as red copper ore,
malachite, copper pyrites, sulphate of copper, &c., when
contained in other minerals so as to be unrecognisable by
the eye, the employment of chloride of silver may be of
the greatest service, as the smallest quantities of copper,
when treated with chloride of silver under the blowpipe,
give a continuous and beautiful blue colour to the flame.
With chloride of silver the presence of copper may be
distinctly ascertained by the blowpipe, even in a solution
which is no longer coloured blue by the addition of
ammonia.
4 The employment of chloride of silver will be equally
advantageous with molybdenum, as in this case also the
Q 2
228 USE OF CHLORIDE OF SILVER
flame gains greatly in intensity. Arsenic, lead, and anti-
mony are already sufficiently characterised, the former by
its odour, the two latter by their fumes ; but even with
these metals chloride of silver may be employed with ad-
vantage to render their reactions still more distinct. It is
only necessary to observe, that the greenish -blue flame of
antimony appears greener and more like that of molyb-
denum under the influence of chloride of silver.
'Chloride of silver may also be employed with com-
pounds containing several of the above-mentioned metals.
' If bournonite be heated in the oxidation flame of the
blowpipe, a fine blue flame is first produced, which in-
dicates lead with certainty ; if chloride of silver be now
applied, copper is also readily shown. The antimony
contained in bournonite cannot be ascertained by the
coloration of the flame ; but this may easily be detected
upon charcoal, or in a glass tube open at both ends.
' Native molybdate of lead, without chloride of silver,
only gives a blue colour to the blowpipe flame ; with
chloride of silver this blue coloration of the lead comes
out more distinctly, but at the same time the tip of the
flame, particularly when the reduction flame is employed,
appears of a beautiful yellowish-green colour from molyb-
denum.
' With mixtures of arsenic and copper, or antimony
and copper, the flame first acquires a greyish-blue or
greenish-blue colour from the oxidation of the arsenic or
antimony ; the copper may then be very easily detected
by chloride of silver. This applies also to mixtures of
arsenic and molybdenum, or antimony and molybdenum ;
with .chloride of silver the yellowish-green flame of
molybdenum appears distinctly. It will be more difficult
to analyse mixtures of arsenic and lead, or antimony and
lead, in this manner ; and if a compound contain both
arsenic and antimony, these two bodies are not to be
distinguished with chloride of silver under the blowpipe.
' From these experiments it appears that in blowpipe
testing it is advantageous to employ chloride of silver
instead of muriatic acid.
IN BLOWPIPE ASSAYS. 229
' Chloride of silver is particularly to be recommended
in testing metallic alloys for copper. Thus, to test silver
for copper, chloride of silver may be applied to the ends
of silver wires, and on the application of heat the smallest
quantity of copper will furnish the most distinct reaction.
This is as sensitive as any of the known copper reactions,
and may be performed quickly and easily. In testing
metallic alloys for traces of copper, it may be advisable to
submit those which contain antimony, zinc, lead, and other
volatile metals to roasting^ so as to drive off these metals
before the addition of chloride of silver.'
TINCTURE OF IODINE. Messrs. Wheeler and Ludeking
have given, in the ' Transactions of the Royal Society of
Canada,' a series of experiments proving that tincture of
iodine can be made a very valuable blowpipe reagent.
The tincture of iodine is prepared by making a saturated
alcoholic solution of the element, which dissolves very
readily to a dark red liquid in this menstruum. The
usual blowpipe charcoal support is replaced by long thin
tablets of plaster-of-Paris, in order to develop the true
colours of these varicoloured iodides on a white back-
ground. These tablets are prepared by mixing plaster-of-
Paris with water to a thin fluid paste, which is poured
over a smooth flat surface (as a plate of glass) that has
been previously oiled to prevent its adhesion. In a few
minutes it will set into a hard cake ; but before this takes
place, when it has become stiff, it is divided by a knife or
spatula into pieces about 4 by 1-^ inches for use. These
tablets are the supports on which these iodide reactions
are made by putting the substance on one end, then
moistening with the tincture, and blowing with the blue
flame, when the volatilised iodides are deposited on its cold
surface when suitably inclined. The oxidising flame must
be employed in order to prevent the deposition of soot,
which tends to interfere by its black film.
Description of the Iodide Coats.
Arsenic. A reddish-orange coating.
Lead. A chrome-yellow coating.
230 TINCTURE OF IODINE.
Tin. A brownish- orange coating.
Silver. A faint greyish : yellow white cold, bright
yellow when hot ; close to assay.
Antimony. An orange-red coating.
Mercury. A yellow and scarlet, the yellow changing
completely to scarlet on standing.
Selenium. A reddish-brown coating.
Tellurium. A purplish-brown coating.
Bismuth. A chocolate brown, fringed with red near
the assay.
Cobalt. A greenish brown, edged with green ; brown
coat evanescent, changing into faint green, especially when
breathed upon.
Molybdenum. A deep ultramaririe-blue coating ; close
to assay, that is, a permanent oxide.
Tungsten A faint greenish blue near assay ; that is,
a permanent oxide (brought out stronger by dropping
more tincture on tablet after the operation).
Copper. A white coating.
Cadmium,. A white coating ; becomes bright golden-
yellow on blowing ammonium sulphide vapours over it.
Zinc. A white coating that soon disappears.
As the copper, cadmium, and zinc iodides are white,
the tablets should first be coated with a film of soot
(obtained by holding the tablet in a smoky flame) or else
charcoal should be used in order to give a black back-
ground to the white coats. If to the peculiar velvety
chocolate-brown coating of bismuth a drop of dilute
ammonia be added, or ammonia vapour be blown over
it, the brown disappears, leaving a brilliant red coating.
Many of these coatings are more or less evanescent,
disappearing on prolonged exposure at ordinary tem-
peratures.
In studying these coats it will be observed that we can
now detect tin in the presence of zinc, which has hitherto
been impossible with the blowpipe, and that we have a
very striking and characteristic reaction for molybdenum.
The other coatings are more or less characteristic indivi-
dually, and will find favour as confirmatory tests. For
DRY SILVER IODIDE. 231
mixtures and complicated cases the iodide reactions will
not supersede the standard methods.
DRY SILVER IODIDE. Mr. P. Casamajor has introduced
silver iodide into blowpipe work with marked advantages.
It gives characteristic sublimates of iodide very beauti-
fully and quickly, and it has the advantage over tincture
of iodine that it is a dry powder, easily kept in bottles
which need not close very perfectly.
The iodide coatings of mercury, bismuth, and lead are
familiar to all chemists who use a blowpipe. They are
obtained by V. Kobel's mixture of equal parts of sulphur
and of potassium iodide. Silver iodide has over this
mixture the advantage that there is no sulphur to give
deposits when operations are carried on in glass tubes,
and no fumes of sulphur dioxide. It requires less time to
obtain the coatings, and they have a more distinct appear-
ance.
In experimenting with silver iodide, mixed with various
metallic compounds, the iodide coatings are best deposited
in open glass tubes of about 4 inches in length and J inch
in diameter. The substance to be tested is mixed into a
paste with the silver iodide. A small portion of this
mixture is placed at one end of the open tube, and the
blowpipe flame is blown on it for a short time. The
iodide coatings immediately appear and are seen through
the glass. The glass tube may be held by a tongs or
simply by a piece of paper as the blowing is not suffi-
ciently prolonged to heat the glass tube beyond what
paper will stand.
A small quantity of powdered charcoal or lampblack,
mixed with silver iodide and the substance to be tested,
gives the characteristic coatings more quickly than dis-
tinctly.
The following metals have given iodide coatings in
glass tubes :
Mercury. In tubes, as on other supports, the yellow
and red iodides are produced simultaneously, and streaks
of bright red are seen on a yellow ground.
Bismuth. Yellowish red near the end of the tube and
thick brown coating beyond.
DRY SILVER IODIDE.
Lead and Tin. Both these metals give bright yellow
deposits which retain their colour when cold. These de-
posits cannot be distinguished one from the other, both
being equally bright. In the case of tin a very strong
smell of iodine is given off, which is possibly due to stannic
iodide.
Arsenic. Near the end of the tube to which the flame
is applied there is a yellow deposit ; beyond this a white
coating of arsenious acid. The yellow portion turns white
on cooling, but becomes yellow again when the tube is
heated over a flame.
Antimony. The orange-red coating given by this metal
becomes quite faint on cooling, but the colour becomes
bright again when the tube is heated.
Zinc. The deposit is white both when cold and when
hot ; the fumes are not very abundant, much less so than
those due to lead or tin.
Iron makes a deposit which may be considered as
characteristic from the fact that the yellow coating in the
tube is streaked with distinct dashes of brown. The
yellow portion becomes white on cooling, but the brown
streaks do not change.
Thallium. A yellow coating is deposited as with most
metals. After this has taken place, if a reducing flame
touches the deposit, this fades, leaving a grey tinge with
an edge of purple. This seems to be characteristic of
thallium.
Ccesium, Rubidium, and Lithium have not given de-
posits which can be called characteristic. The deposit
from caesium differs from those of the other two metals in
being less volatile. The caesium deposit does not extend
far beyond the heated end of the tube. By increasing the
heat it melts, but does not move forward.
Chromium gives a white coating which remains at the
hot end of the tube. The portion nearest to this end by
further heating becomes of a pale reddish brown.
Manganese. Yellow hot, but white when cold, like
deposits from many other metals.
Molybdenum. Beyond the yellow coating, which turns
SODA-PAPER.
233
white on cooling, are distinct blue streaks, which are very
characteristic. These were first observed by Messrs.
Wheeler and Ludeking, by treatment with tincture of
iodine on tablets of plaster-of-Paris. With a glass tube, the
blue streaks extend through the whole length of the tube.
Manganese and Uranium give deposits which are
yellow when hot, and white when cold. These are too
common to be characteristic.
Deposits have been obtained on charcoal and on thin
sheets of iron, either on the metallic surface or on a coat
of soot, by the use of silver iodide and metallic compounds.
Some of these deposits are very good, but they are not so
uniform for the same metal as deposits obtained in glass
tubes.
SODA-PAPER. Mr. Forbes writes as follows in the
' Chemical News ' : ' As it would be impossible to submit
any powdered substance to the direct action of the blow-
pipe flame without its suffering mechanical loss, some
means must be employed for holding the particles together
until they are so agglutinated by the heat that no such
loss need be apprehended ; this is secured by the use of
the soda-paper envelope or cornet, as devised by Harkort.
For this purpose slips of thin slightly sized writing paper,
about 1 J inch long by 1
inch broad, are steeped in a
solution of one part crystal-
lised pure carbonate of soda
(free from sulphate) in two
parts of water. When dried
these are used for forming
small cylindrical cornets, by
rolling them round the ivory
cylinder, fig. 73 d, previously
described. A bottom is formed
to them by folding down a
portion of their length on to
the end of the cylinder, which
is then pressed firmly into the corresponding mould in the
blowpipe anvil, and which, upon the withdrawal of the
F IG . 73.
234 SODA-PAPER.
cylinder, serves as a support until they are filled with the
assay from the scoop in which the assay and flux have
been mixed. After pressing the assay down, the super-
fluous paper is cut off, leaving only sufficient when folded
down upon the contents of the cornet, to form a paper
cover to the top similar to the hollow of the cornet. The
assay is then ready for placing in a bore in the charcoal,
formed by the charcoal borer c, fig. 71, and is then
submitted to a reducing fusion.'
GENERAL ROUTINE OF BLOWPIPE OPERATIONS.
Size of the Assay. The morsel operated on is suffi-
ciently large when the effect of the heat and the fluxes added
can be distinctly discerned. The size of the assay-piece
generally recommended is much too large ; its size ought
to be about that of a mustard seed ; that of the flux added,
about the size of a hemp-seed. It should in general be
previously reduced to fine powder.
When a large piece is employed, the experiment con-
sumes much more time and requires much more labour
than a smaller piece. It is only in reductions that a
larger piece may be successfully employed, because in
that case the more metal produced, the more readily can
its nature be ascertained. Having thus endeavoured to
fix the size of the assay, we will now lead our readers to
the operations necessary in blowpipe analysis, and in the
order in which they are to be performed.
First. The substance is heated in the closed tube, or
mattrass, over a spirit lamp. It may, by this treatment,
decrepitate or give off water, or some other volatile sub-
stance.
Secondly. It is heated gently on charcoal, by aid of
the blowpipe ; and, as soon as warm, withdrawn from the
heat, and the odour given off ascertained : volatile acids,
.arsenic, selenium, or sulphur, may be present. The odour
thus produced by the oxidising flame must be compared
with that produced by the reducing flame ; if there is any
difference, it must be carefully noted. Sulphur, selenium,
OPERATIONS IN BLOWPIPE ANALYSIS. 235
<&c., are best detected in the oxidising flame, and arsenic
in the reducing flame.
Thirdly. The substance is examined as regards its
fusibility. If it be in grains, it is better acted upon on
charcoal, notwithstanding its liability to escape on the first
insufflation, when not very fusible. But if we can choose
the form, it is better to knock off a small splinter, by means
of the hammer, and hold it in the flame by the platinum-
pointed pincers. A fragment with the most pointed and
the thinnest edges ought to be selected. By thus acting,
we can always ascertain at a glance if the substance be
fusible or not. Infusible substances retain their sharp
points and angles, which can be ascertained immediately by
means of a microscope. The points are merely rounded in
bodies of difficult fusibility, and in substances of easy
fusion are rendered globular.
Certain substances, and particularly some minerals,
change both aspect and form when exposed to the
blowpipe flame, without entering into fusion ; some swell
up like borax ; some of them fuse after tumefaction ;
others keep in that state without fusion. Some minerals
give off a sort of foam on fusing, giving rise to a kind of
blebby glass, which, although transparent itself, does not
appear so, on account of the multitude of air-bubbles it
contains.
This bubbling and tumefaction takes place in the
greater part of the minerals only at that temperature at
which all the water is disengaged ; and these ramifications
appear to proceed from a new molecular arrangement,
produced by the action of heat on the constituent parts of
the substance. It cannot be said that the expansion of a
particular part of the substance or its formation into gas,
gives rise to this, because it most often happens in those
substances which contain no such substance. The minerals
which generally give these indications are the double
silicates of lime or alkali, and alumina. It sometimes
disappears after a few instants, and occasionally lasts as
long as the substance is kept in fusion. In the latter
case, it seems that the assay takes carbonic acid from the
236 PRELIMINARY BLOWPIPE OPERATIONS.
flame, which carbonic acid is transformed by the charcoal
into carbonic oxide, and it is that gas which causes the
bubbles.
In the employment of fluxes, it is necessary to continue
the blast for a sufficiently long time, because some sub-
stances appear infusible at the commencement of the oper-
ation, and gradually yield to the influence of the flux, and
in about two minutes enter into full fusion. The substance
is best added in small quantities, and no new dose must be
introduced until the former one is acted upon, so that at
last the glass arrives at that degree of saturation that it can
dissolve no more : it is at this particular point that the
reactions are most vivid and plain. Beads of glass, not so
saturated, do not give such decided indications.
Occasionally, in operating with a flux in the reducing
flame, it happens that the assay-bead reoxidises during the
cooling of the charcoal, and thus the labour of a preceding
operation is lost. In order to obviate this inconvenience,
the charcoal is turned over, so that the bead may fall in a
yet liquid state on some cold body, as a porcelain plate.
When the colour of the bead is so intense that it ap-
pears opaque, its transparency can be proved by holding
it opposite to the flame of a lamp ; the reversed image of
the flame can then be seen in the bead, tinged with the
colour imparted to the flux by the body under experiment.
The globule may also be flattened by a pair of pliers before
it cools, or it may be drawn into a thin thread. In either
of the last-mentioned cases its colour can readily be ascer-
tained.
Minerals exposed to the exterior and interior flame,
either with or without fluxes, present a variety of pheno-
mena, which ought to be carefully noted, and which,
collectively, form the result of the assay. The smallest
circumstance must not be overlooked, because it may lead
us to ascertain the presence of a substance not suspected.
It is necessary, in all cases, to make two assays, and com-
pare the separate results ; because it sometimes happens
that an apparently trivial fact had been overlooked in the
first series of operations, which materially conduces to the
good result of the experiment.
DISCRIMINATION OF MINERALS. 237
DISCRIMINATION OF MINERALS.*
Selecting those minerals which are remarkable for
their wide distribution, and at the same time those which
are of value, whether rare or frequent, the number which
it is advisable for the explorer to know is reduced to about
seventy-five, and it is only to this limited number that
reference will be made in this chapter.
The distinctions in minerals lie in their chemical and
physical characters ; but as the appearance of a mineral
may afford little or no indication of what these are, they
have to be discovered by testing.
These characters will be described in detail, and the
modes of testing will be indicated, where necessary.
The characters include crystalline form, mode of
fracture, colour, lustre, transparency, taste, odour, feel,
hardness, specific gravity, fusibility, and chemical com-
position.
Crystalline Form.
The crystalline form is an important guide in many
cases, but crystals are subject to so many modifications,
that this character ceases to be a practical one except to
those persons who are familiar with its geometrical laws.
Stil] it is very useful to know a few common primary
forms such as the following :
CUBE. 6 sides or faces, each square (fig. 74).
OCTAHEDRON. 8 sides, each an equal-sided triangle
(fig. 75).
DODECAHEDRON. 12 sides, each rhombic, (shaped like
the diamond of a pack of cards) (fig. 76).
TETRAHEDRON. 4 sides, each triangular (fig. 77).
* This method of distinguishing the various minerals, ores, etc. which are
likely to be met with during an exploration is partly condensed from a very
valuable pamphlet by Alexander M. Thompson, D.Sc., Professor of Geology at
the University of Sydney, entitled ' Guide to Mineral Explorers in Distin-
guishing Minerals, Ores, and Gems.' The instructions given will not be found
to involve greater difficulties than can be overcome by a little practice and
perseverance. The appliances which are needed may be bought for a few
shillings, and will be found too trifling in weight and bulk to incommode the
traveller.
238
DISCRIMINATION OF MINERALS.
EHOMBOHEDRON. 6 sides, each rhombic (fig. 78).
PRISM. Any column with three or more sides ; when
FIG. 74.
FIG. 75.
FIG. 76.
FIG. 77.
FIG. 80.
FIG. 79.
Fm. 81.
placed on its base, it may stand straight or oblique ; it may
terminate abruptly with a flat face, or come off to a point,
blunt or sharp, like a pyramid (figs. 79, 80, 81).
Mode of Fracture.
The mode in which a mineral breaks when smartly
struck with a hammer, or pressed with the point of a knife,
is a character of importance. Many can only be broken
in certain directions ; for instance, a crystal of calc-spar
can only be split parallel to the faces of a rhombohedron ;
many crystals break more easily in one direction than in
DISCRIMINATION OF MINERALS. 239
others. Whenever a mineral breaks with a smooth, flat,
even surface, it is said to exhibit cleavage. Cleavage
always depends upon the crystalline form. But minerals
often break in irregular directions, having no connection
whatever with the crystalline form, and this kind of
breaking is called fracture ; the broken surfaces are usually
irregular or conchoidal (i.e. with concave and convex out-
lines, like shells).
Lustre.
Some minerals have a brilliant lustre like that of
metals ; in others the lustre resembles that of glass, silk,
resin, or wax ; while others are dull, or destitute of lustre.
The lustre of the diamond is called adamantine.
Colour and Streak.
Minerals may be colourless, white, black, or of any
colour, either dull or brilliant. The same mineral may
present a variety of similar tints, or even distinct colours.
It often happens that a mineral, which when viewed in a
solid mass possesses a distinct colour, affords a powder
which has a colour different from that of the solid mass,
or is even destitute of colour, that is to say, white or
nearly so. The dust formed on scratching a mineral with
a knife, or by a splinter of quartz, or by a diamond, is
termed the ' streak ' : it has usually the same colour as the
powder. The colour of a mineral and its streak may cor-
respond ; or the mineral and its streak may possess diffe-
rent colours ; or the mineral may be coloured, while its
streak is colourless.
For instance, cinnabar has both a red colour and a red
streak.
Specular iron has a black colour, but a red streak.
Sapphire has a blue colour, but a white or colourless
streak.
The streak of most minerals is dull and pulverulent,
but a few minerals exhibit a shining streak, like that
formed on scratching a piece of lead or copper. This
kind of streak is distinguished by the name of Metallic.
240 DISCRIMINATION OF MINERALS.
In judging the colour of a mineral a surface quite free
from tarnish should be chosen.
The colour of a transparent gem can be seen to best
advantage by immersing it in water, about half an inch
below the surface.
Hardness.
This character is of great importance in distinguishing
minerals ; it implies the degree of facility with which the
particles may be separated by cutting or scratching. The
diamond is the hardest substance known, as it will scratch
all others. Talc is one of the softest minerals. Other
minerals possess intermediate degrees of hardness. To
express how hard any mineral is, it becomes necessary to
compare it with some known standard. Ten standards of
different degrees have been chosen, and are given in order
in the following scale :
1. Talc. 6. Felspar.
2. Gypsum. 7. Quartz.
3. Calc Spar. 8. Topaz,
4. Fluor Spar. 9. Corundum, or Sapphire.
5. Apatite. 10. Diamond.
The hardness of a mineral may often be found by
drawing the point of a steel knife across it. For instance,
the slightest pressure will suffice to scratch talc ; fluor-spar
is not so easily scratched as calcite ; the greatest pressure
is needed to scratch felspar ; and quartz does not yield to
the knife at all.
The hardness may also be found by scratching one
mineral with another ; thus, the diamond will scratch all
other minerals ; corundum scratches topaz, topaz scratches
quartz, quartz scratches felspar ; and so on.
If on drawing a knife across a mineral it is impressed
as easily as calcite, its hardness is said to be 3. If a
mineral scratches quartz, but is itself scratched by topaz,
its hardness is between 7 and 8.
In trying the hardness of a mineral a little judgment
is necessary ; for instance, a sound portion of the mineral
DISCRIMINATION OF MINERALS. 241
must be chosen ; a sharp angle used in trying to scratch :
a streak of dust on scratching one mineral with another
may come from the waste of either, and it cannot be
determined which is the softer, until after wiping off the
dust, and viewing with a lens.
The use of the above scale, however, implies an ac-
quaintance with the ten standard minerals. It is very
desirable to have specimens of these minerals for reference.
But a collection of the following common minerals, which
can easily be procured, will be found quite enough for
ordinary purposes viz. : calc-spar, felspar, quartz, topaz
(white or yellow), and corundum or sapphire.
By the test of hardness clear distinctions may be drawn
between minerals which resemble each other ; for instance,
iron pyrites and copper pyrites are similar in appearance,
but copper pyrites can easily be scratched with a knife,
whereas iron pyrites is nearly as hard as quartz and cannot
be impressed with a knife at all.
In determining the hardness of small stones, it is most
convenient to fasten them by heating upon a stick of seal-
ing wax.
Specific Gravity.
By specific gravity is meant the comparative weight of
equal bulks. Water is taken as the standard of compari-
son ; the specific gravity of a mineral is a number showing
how many times it is, bulk for bulk, heavier than water.
The specific gravity of water is called 1, of gold 19,
implying that, if equal bulks of gold and water were taken,
the gold would weigh 19 times as heavy as the water. The
specific gravity of a mineral can be found by weighing it
first in air in the usual manner, and then observing how
much of its weight it loses, when suspended from the arm
or pan of a balance, and allowed to hang freely in water.
If a piece of quartz weighing 26 grains is attached by a
hair or thin cotton to the scales, and weighed whilst hang-
ing in water, it will be found to weigh only 16 grains ; it
thus loses 10 grains, or if of its entire weight. Similarly
gold would lose -fa of its weight.
R
242 DISCRIMINATION OF MINERALS.
Minerals differ very widely in the proportion of weight
which they lose in water, but the same mineral invariably
loses the same proportion ; for instance :
Quartz loses -J-g- of its weight ; topaz ^g- ; sapphire ^ ;
zircon i-g. ; tin ore f [}.
These proportions depend upon the specific gravity of
these minerals. The specific gravity of water is called 1,
.of quartz 2*6, of topaz 3*5, of sapphire 4*0, of gold 19.
In estimating how much weight a mineral loses in
water, a very delicate balance is required when the weight
in air is under 10 grains ; but for portions weighing heavier
than this, a common balance turning readily to a grain
may be used for practical purposes. The mineral must be
sound throughout, and free from any pores or cracks, and
its surface should be rubbed over with water, before im-
mersing it, to prevent bubbles of air adhering, which would
falsify the result. In careful trials distilled or rain water
should be used. A trial .of specific gravity can have no
value, unless it is made on a pure portion of a mineral, quite
free from any adhering foreign matter.
A rough estimate of specific gravity can be formed from
the feeling of pressure in shaking any mass loosely in the
palm of the hand : in this way it can be judged whether
the specific gravity is high or low.
EULE. The rule for finding the specific gravity is to
divide the weight of the mineral in air by its loss of weight
in water. Example: A piece of quartz weighed 1,398
grains in air, and 862 grains in water ; here the loss of
weight is 536, and the weight in air divided by this number
is 2 -6, which is the specific gravity of quartz.
For separating from each other minerals of different
specific gravities R. Breon proposes the use of melted mix-
tures of lead and zinc chlorides in different proportions, so
as to yield liquids ranging from 2' 4 to 5*0 in specific
gravity.
Sonstadt* has devised a very ingenious and useful
method of taking specific gravities of minerals, when a
rough and ready method of discriminating between bodies
* ' Chemical News,' March 20, 1874, vol. xxix. p. 127.
DISCRIMINATION OP MINERALS. 243
of similar appearance but of different specific gravities is
all that is needed. He takes a solution of potassium
iodide saturated at the common temperature, and dissolves
in it as much iodide of mercury as possible. It will
then dissolve more potassium iodide, then more mercuric
iodide, and so forth. The iodides dissolve very slowly at
the last, and as it is best not to accelerate the solution by
the application of heat, considerable time must be allowed
when a liquid of maximum strength is required. The
solution, after filtering, is fit for use, and it may be
obtained without danger of its crystallising above zero,
of sp. gr. 3*085 at 12 C. This liquid is transparent, very
mobile, filters easily', is of an amber colour, gives no preci-
pitate on addition of water, and does not readily lose or gain
water on long exposure to the atmosphere. The solution
undergoes no perceptible chemical change by free exposure
to the air for many months. It may be diluted to any
extent, and then concentrated by heat without injury.
When this liquid is diluted so that quartz will just
float in it, any mineral heavier than quartz will of course
sink. As many minerals closely resemble quartz in external
characters, and can be distinguished more readily by their
specific gravity than in any other way, this method of
testing the specific gravity furnishes a very quick and
certain means of discrimination. A small bottle of the
liquid, containing say half an ounce, and carried in the
waistcoat pocket, will enable the mineralogist to make a
great many examinations in his walks as to whether the
specimens examined are above or below a certain specific
gravity. The smallest distinctly visible particle of a
mineral is enough for an estimation. After the bottle
becomes clogged with specimens, the liquid may be poured
out for further use, and if the washings are added and the
whole concentrated there need be no sensible waste.
Mr. Hardman suggests that with this liquid it would
be possible to separate completely the three constituent
minerals of granite mica, felspar, and quartz weigh
them, and estimate absolutely their percentage.
Professor Church has used Sonstadt's solution exten-
R 2
244 DISCRIMINATION OF MINERALS.
sively for years in the separation of rare minerals from
their siliceous gangue. He says he has often identified
the minerals of a crushed fragment of rock by introducing
a drop of the solution into the ' live box ' under the
microscope. By using solutions of increasing strength, or
adding the solid salt, the felspars, the quartz, and then
some of the heavier minerals may be successively brought
to the surface and thus into focus.
Fusibility.
Some minerals can be easily fused ; others only with
difficulty, while others resist the highest heat which can
be applied to them. There are such wide differences
between the various degrees of fusibility of minerals, that
this character helps greatly in distinguishing them. The
fusibility is most readily tested by holding a small splinter
of the mineral with a forceps in a candle flame, urged
by the blowpipe ; or the mineral may be laid upon a piece
of charcoal, and the flame directed upon it by the blow-
pipe. Some minerals fly to pieces when heated, others
swell up, or give off peculiar and characteristic odours.
Chemical Characters.
The chemical composition of a mineral is one of the
most important facts to know respecting it; the various
processes for estimating this form a science by them-
selves ; but only the tests depending upon the use of the
blowpipe will be described here. A great deal can be
learnt respecting a mineral by a few simple trials with the
blowpipe. The only requirements are a common blow-
pipe, a candle, a forceps or pliers, a piece of No. 22 plati-
num wire, two inches long, dried sodium carbonate, dried
borax, and potassium cyanide. The charcoal selected for
these experiments should be free from cracks and openings.
By dry sodium carbonate is meant not merely dry to the
touch, but quite free from water ; this may be prepared
from recrystallised common washing soda, by expelling
the water which it contains ; the washing soda is put in a
shallow clean iron dish, and placed over a clear fire until
a white dry powder is formed ; too strong a heat might
DISCRIMINATION OF MINERALS. 245
fuse the dry powder. A quarter of an ounce may be kept
in a well-corked bottle or tube for use. Sodium bicarbon-
ate may be used instead without previous heating ; or if
the bicarbonate be moderately heated, it loses weight, and
becomes sodium carbonate, quite free from water like the
above.
The borax is to be dried in the same way ; a quarter
of an ounce will be enough ; it is convenient to keep the
platinum wire in the same tube. Unless these tubes are
well corked these chemicals reabsorb moisture. For test-
ing tin ore it is useful to have a little potassium cyanide
kept in a bottle with the cork and rim well covered
with melted bees'-wax ; it would otherwise liquefy, by
-absorption of moisture, and become useless. It is a
most dangerous poison, and the greatest caution must
be observed in its use.
Three kinds of effect can be produced by the blowpipe :
first, it can be used as a source of heat for testing fusibility ;
secondly, it can add oxygen to a mineral ; thirdly, it can
take oxygen away.
The three principal means of chemically testing
minerals before the blowpipe are (1) with borax; (2) on
charcoal, usually with the addition of sodium carbonate ;
(3) by holding in the point of oxidation.
'. EXPERIMENT No. 1. Many metals impart a colour to
fused borax, by which their presence can be recognised.
To try this experiment a bead of fused borax must first
be obtained on the platinum wire. The end of the wire
is bent into a loop or ring about the twelfth part of an
inch in diameter. The wire is then heated in the blow-
pipe flame, and dipped whilst hot into the borax : the por-
tion of borax that adheres is then fused on to the wire in
the blowpipe flame, and the hot wire is again dipped ; this
is repeated until the loop contains a glass-like bead of fused
borax. If the bead has become cloudy, the soot causing
this may be burned off in the oxidising point of the flame.
Having thus obtained a clear, colourless, transparent bead,
-the next step is to add to it a minute portion of the mineral
which is to be tested. By touching a little of the finely
240 DISCRIMINATION OF MINERALS.
pulverised mineral with the borax bead, while softened by
heat, enough will adhere to the bead for a first trial. The
bead is then kept at a white heat in the oxidising point of
the flame for a few seconds, and on removal its colour is
noted, both whilst hot and when cold. If no colour is
imparted a fresh trial may be made with a larger quantity
of the powder ; but if the bead is opaque owing to the
depth of colour, as is often the case, a fresh experiment
must be made, using a still smaller quantity of the powder.
The colour can only be fairly judged in a perfectly trans-
parent bead. If no colour can be obtained in the oxidising
flame, further experiment with the borax bead is needless ;
but if a colour is obtained, it is then advisable to try the
effect of the reducing flame upon the same bead. The
following observations and inferences may result from this
test :
Colour of Bead
in
Showing the
OXIDISING FLAME. EEDUCING FLAME.
presence of
Green (hot) ; Blue (cold).
Red.
COPPER.
Blue (hot and cold).
Blue.
COBALT.
Amethyst.
Colourless
MANGANESE.
Green.
Green.
CHROMIUM.
Bed or Yellow (hot). 1
Yellow or colourless (cold). f
Bottle Green.
IRON.
Violet (hot) ; Ked-Brown (cold).
Grey and Turbid,
NICKEL.
difficult to ob-
tain.
This mode of testing may often be used to prove the
presence of the above-mentioned metals.
It requires some practice before trustworthy results
can be obtained in reducing ; the reduced bead, if brought
out of the flame at a white heat into the air, may at once
oxidise ; but this may be prevented by placing it inside
the dark inner cone of an ordinary candle flame, and
allowing it to cool partially there.
EXPERIMENT No. 2. The mode of testing with sodium
carbonate on charcoal is performed as follows : A sound
piece of charcoal half an inch square is chosen, and a neat
cavity is scooped out on its surface, into which is placed
a mixture containing the pulverised mineral to be tested,
with three or four parts of sodium carbonate, the whole
not exceeding the bulk of a pea ; after lightly pressing
DISCRIMINATION OF MINERALS. 247
the mixture into the cavity, the blowpipe flame may be
cautiously applied to it ; and afterwards, when the mix-
ture no longer shows a tendency to fly off, the charcoal
may be advanced nearer to the blowpipe, and finally be
kept at as high a temperature as possible in the reducing
part of the flame.
In testing for tin ore, a piece of potassium cyanide
about the size of a pea may be placed upon the mixture
after the first application of heat, and the further applica-
tion of heat may then be continued.
This treatment is designed to extract metals from
minerals ; it favours in the highest degree the removal of
oxygen. But like the borax test, jt is limited in its appli-
cation ; it can only be used to detect certain metals. The
failure of the test, in any case, must not be looked upon as
a conclusive proof of the absence of the particular metal
sought : for instance, copper can be easily abstracted from
copper carbonate by this test, but not from copper pyrites.
Still the test is a most valuable and indispensable one to
the mineralogist.
The test is complete when the metal is obtained as
a globule, in the cavity of the charcoal ; in many cases
the globule will be found surrounded with the oxide
of the metal, forming an incrustation on the charcoal ;
and the colour of such incrustation should be carefully
noted, both at the moment of removal, from the flame,
and after cooling. By pressing the globule between
smooth and hard surfaces, it can be seen whether the
metal is flattened out (or malleable), or crushed to pieces
(brittle).
The following observations and inferences may result
from this test:
GLOBULE. INCRUSTATION. Presence of
Yellow (malleable). None. GOLD.
White do. Do. SILVER.
Ked do. Do. COPPER.
White do. White. TIN.
Do. do. Bed (hot) ; Yellow (cold). LEAD.
Do. (brittle). Do. do. BISMUTH.
None. Yellow (hot) ; White (cold). ZINC.
White, brittle, giving off White. ANTIMONY.
fumes when removed
from the flame.
248 DISCRIMINATION OF MINERALS.
EXPERIMENT No. 3.- In addition to these substances
there are others which occur abundantly in minerals, and
which may be recognised by the blowpipe with the greatest
ease ; for instance, sulphur and arsenic. These may be
discovered by heating a fragment of the mineral, supported
on a piece of charcoal, or held in a forceps, in the oxidising
point of the flame, and noticing the odour which is given
off; a smell of burning sulphur indicates that the mineral
contains that substance, and white fumes having a garlic
odour indicate the presence of arsenic.
Mercury, antimony, and other substances may escape
as fumes when heated in this manner.
Requirements for Testing Minerals.
The requirements already mentioned are blowpipe,
candle, forceps, platinum wire, dried borax, dried sodium
carbonate, potassium cyanide, charcoal ; also the minerals
for comparisons of hardness namely, calc-spar, felspar,
quartz, topaz, sapphire. In travelling it is well to dis-
pense with the grain scales and weights for taking specific
gravities. It would be dangerous to attempt to travel
with nitric acid. In addition to these it will be found
useful to have a steel pocket-knife, one blade of which
should be kept magnetised, which may be easily done by
touching it occasionally with a strong magnet ; a small
iron spoon for heating minerals, such as cinnabar, over a
candle flame ; also a small pocket lens.
DESCRIPTION OF MINERALS.
QUARTZ AND THE SILICATES.
Perhaps no other mineral presents such a great variety
of forms and colours as quartz, and no mineral occurs in
greater abundance. When pure it consists of silica only,
but it is usually contaminated with other ingredients,
principally alumina, iron, and clay. The impure varieties
of quartz compose most of the pebbles and sand of the
soil.
DISCRIMINATION OF MINERALS. 249
Out, of the six hundred minerals at present known, at
least two hundred and fifty contain silica in greater or less
proportion, and are hence termed silicates.
Silicates are indispensable in the manufacture of glass,
porcelain, earthenware, and for other purposes ; but they
exist in such profusion that their economic value is ex-
ceedingly trifling. The great majority of the silicates are
purely objects of scientific interest ; a few forms of silica
&re esteemed as gems, such as the precious opal, and some
varieties of quartz ; also the silicates, topaz, emerald,
zircon or hyacinth, garnet and carbuncle, the tourmaline,
.and some others. Handsome varieties of serpentine are
often used in ornamental stonework.
1. QUARTZ.
Chemical Composition. Silica.
Crystallised in six-sized prisms, terminated by pyramids.
Sides of the crystal often marked across with fine parallel
lines. Transparent or opaque. Colourless, or of various
colours. Glassy lustre. Fracture irregular, conchoidal.
Specific gravity, 2'6. Hardness, 7. Cannot be scratched
with a knife ; scratched by topaz, zircon, sapphire, and
diamond, and thus easily distinguished from these gems.
Quartz scratches glass with facility ; felspar and many
other minerals can be scratched by quartz. The irregular
fracture, the fine parallel markings, and the hardness gene-
rally suffice to distinguish quartz.
Infusible in the blowpipe flame.
The following are the chief varieties of quartz ; the
differences are due either to their mode of formation or
the presence of impurities. They have the same general
characters as pure quartz,
1. TRANSPARENT VARIETIES :
BOCK CRYSTAL. Pure, transparent, colourless quartz.
Used for spectacle glasses and ornaments.
AMETHYST. Transparent. Of a rich violet or purple
colour. Used as a gem.
250 DISCRIMINATION OF MINERALS.
EOSE QUARTZ. Seldom perfectly transparent. Of a
rosy tint.
CAIRNGORM, or SMOKY QUARTZ, is transparent, with a
smoky tinge.
FALSE TOPAZ has a yellow pellucid colour, distinguished
from topaz by its inferior hardness.
2. SEMI-TRANSPARENT VARIETIES :
CHALCEDONY. Pale colour and waxy lustre. Kesem-
bling icicles in some instances, the frothy surface of a
liquid in others.
CORNELIAN and SARD have red tints.
AGATE exhibits cloudy or moss-like patches, or a
number of lines arranged in circular and angular patterns.
ONYX or SARDONYX is made up of regular layers one
above another of different colours, often black, white, and
red. It is much used for cameos.
FLINT or HORNSTONE. Common dull varieties.
3. OPAQUE VARIETIES :
JASPER is quartz rendered opaque by clay, iron, and
other impurities ; it is of a red, yellow, or green colour ;
sometimes the colours are arranged in ribands, or in
other fantastic forms. It is used for ornamental work.
BLOODSTONE is green jasper with splashes of red re-
sembling blood-spots.
2. OPAL.
Chemical Composition. Silica and water.
Never crystallised. Fracture conchoidal. Specific
gravity, 2'2. Hardness, 6. Can be scratched by quartz,
and thus distinguished from it.
Infusible. It is generally milk-white.
Precious or Noble Opal exhibits a beautiful play of
colours, and is a valuable and rare gem. The common
varieties are of no value.
3. TALC.
Chemical Composition. Silica, magnesia, water (pro-
toxide of iron).
DISCRIMINATION OF MINERALS. 251
Usually in irregular layers. Nearly opaque. White
or green ; pearly lustre ; greasy feel. Specific gravity,
2-7. Hardness, 1. Easily impressed by the nail. But
impure varieties are much harder.
Infusible. Yields no water when heated in a glass
tube. Is not attacked by boiling sulphuric acid.
Its greasy feel and pearly lustre readily distinguish it.
Mica, which is often confounded with it, is not so soft,
has not a greasy feel, and can be split into very thin trans-
parent layers. Steatite is a variety often applied to useful
purposes.
4. CHLORITE.
Chemical Composition. Silica, magnesia, alumina, pro-
toxide of iron, water.
Often forming rocks. Opaque. Green of various
shades. Lustre pearly or dull. Hardness, 1 to 2. Very
easily cut with a knife.
Infusible. In glass tube yields water. Boiling sul-
phuric acid extracts from it magnesia, alumina, and fer-
rous oxide. '
Abundant. Of no value.
5. SERPENTINE.
Chemical Composition. Silica, magnesia, water (ferrous
oxide).
Often 'forming rocks. Never crystallised. Opaque.
Green. Lustre resinous or dull. Streak white. Hard-
ness, 3. Can be cut with a knife. Specific gravity, 2'5.
Infusible except in thin edges ; turns white in blowpipe
flame. Powder decomposed by sulphuric acid like chlorite.
Gives off water in glass tube.
Some varieties form handsome stone for slabs and
ornamental work.
Meerschaum, which is soft and earthy, and nephrite
(the New Zealand Maori greenstone), which is as hard as
quartz, both contain silica and magnesia.
252 DISCRIMINATION OF MINERALS,
0. AUGITE AND HORNBLENDE.
Chemical Composition. Silica, lime, magnesia, ferrous
oxide (occasionally alumina and fluorine).
In four, six, or eight-sided prisms, exhibiting cleavage
in some directions. Usually opaque. Green, black, or
white. Glassy, pearly, or resinous lustre. Specific gravity,
3 to 3-5. Hardness, 5 to 6. Can be scratched with a
knife, using pressure. Scratched by quartz.
Dark varieties fusible.
Augite usually occurs in stout six or eight- sided prisms
with roof- like terminations ; hornblende in long slender
prisms. Asbestos is a variety composed of separable fibres
like flax.
Abundant in igneous and other rocks. Of no value.
Asbestos is used as a fireproof material, and the long silky
variety is woven into fireproof fabrics.
7. CHRYSOLITE, OR OLIVINE.
Chemical Composition. Silica, magnesia, protoxide of
iron.
In prisms ; but usually in grains or lumps resembling
bottle glass. Transparent or opaque. Green, yellow, or
black. Glassy lustre. Fracture conchoidal. Hardness,
6 to 7. Cannot be scratched with a knife. Specific
gravity, 3-4.
Infusible.
In appearance, hardness, and infusibility may resemble
tin ore, but is very much lighter in weight. Differs from
obsidian (volcanic glass) in its infusibility and superior
hardness.
Occurs in basalt and lava. Of no value.
8. TOURMALINE. .
Chemical Composition. Silica, alumina, magnesia,
boracic acid, fluorine, oxides of iron (lime and alkalies).
In prisms, with three, six, nine, or more sides, fur-
rowed lengthwise, terminating in low pyramids. Com-
DISCRIMINATION OF MINERALS. 25$
monly black and opaque ; rarely transparent, and of a rich
red, yellow, or green colour. Glassy lustre. Fracture
uneven. Specific gravity, 3vl. Hardness, 7 to 8. Can-
not be scratched with a knife. Not scratched by quartz.
Infusible.
When the smooth side of a prism is rubbed on cloth it
becomes electric and can attract a small piece of paper ; if
the prism is as wide as a pipe-stem, when one side is heated
for a moment in the blowpipe flame the opposite side be-
comes electric and can attract paper until the heat spreads
uniformly through the crystal. (On this account tour-
maline is said to be pyro- electric.)
Occurs in granite and slate. Of no value ; except the
fine-coloured transparent varieties, which are used as gems
and for optical purposes.
9. GARNET.
Chemical Composition. Silica, alumina, lime, iron, mag-
nesia, manganese.
Crystallised in dodecahedrons, never in prisms. Trans-
parent or opaque. Generally red ; also brown, green,
yellow, black, white. Glassy or resinous lustre. Fracture
conchoidal or uneven. Specific gravity, 3'5 to 4-3. Hard-
ness, 6-5 to 7'5. Cannot be scratched with a knife.
Fusible with more or less difficulty. Eed varieties
impart a green colour to borax bead owing to presence of
chromium.
Common in gneiss and schists. Fine-coloured trans-
parent varieties (carbuncle, cinnamon stone, almandine)
are used in jewellery.
10. TOPAZ.
Chemical Composition. Silica, alumina, fluorine.
In prisms, ''sometimes furrowed lengthwise, variously
terminated, breaking easily across with smooth brilliant
cleavage.
Transparent or semi-transparent. White, yellow,
greenish, bluish, pink. Glassy lustre. Specific gravity,
254 DISCRIMINATION OF MINERALS.
3'5. Hardness, 8. Scratches quartz. Is scratched "by
sapphire.
Infusible, but often blistered and altered in colour by
heat. When smooth surfaces are rubbed on cloth they
become strongly electric and can attract small pieces of
paper, but rough surfaces do not show this. The brilliant
cleavage of topaz distinguishes it from tourmaline and
other minerals.
Occurs in granite. Used in jewellery. The topaz be-
comes electric by friction much easier than other gems,
such as the balas ruby, which it may resemble. The white
topaz resembles the diamond ; but, unlike diamond, it can
be scratched by sapphire.
11. BERYL, OR EMERALD.
Chemical Composition. Silica, alumina, glucina.
In six-sided prisms. Usually green. Transparent or
opaque. Glassy lustre. Fracture uneven. Specific gravity,
2'7. Hardness, 7 to 8. Scratches quartz.
Infusible, or nearly so, but becomes clouded by heating.
Occurs in granite and schist. Valuable for jewellery
when transparent and rich glass green (emerald), or sea
green (aquamarine). Opaque crystals of large size, exceed-
ing a ton in weight, have been found in North America.
12, ZIRCON, OR HYACINTH.
Chemical Composition. Silica, zirconia.
In square prisms, terminated by pyramids, and in octa-
hedrons. Often found in pebbles and grains. Transparent
or opaque. Wine or brownish red, grey, yellow, white.
Glassy lustre. Fracture usually irregular, but in one
direction it can be split so as to exhibit a smooth, even
cleavage face having an adamantine lustre like the diamond.
Specific gravity, 4'0 to 5'0. Hardness, 7'5. Scratches
quartz ; is scratched by topaz.
Infusible ; the red varieties, when heated before the
blowpipe, emit a bluish phosphorescent light, and become
permanently colourless.
DISCRIMINATION OF MINERALS. 255
Occurs in syenite, granite, basalt. Clear crystals used
in jewellery, in jewelling watches, and in imitation of
diamond. It may be distinguished from diamond by its
inferior hardness, and in not becoming so readily electric
by friction.
13. FELSPAR.
Chemical Composition. Silica, alumina, potash or soda
(lime).
Crystallised or in irregular masses. Opaque. Usually
flesh-red, or white, or of various dull tints. Glassy or
pearly lustre. Fracture irregular ; but in some directions
it splits with an even, glimmering, cleavage face. Specific
gravity, 2-3 to 2-8. Hardness, 6. Easily scratched by-
quartz. Cannot be scratched with a knife without greatest
pressure.
In thin edges fusible with difficulty.
Abundant in granitic and porphyritic rocks. No value.
By its decomposition it forms porcelain clay or kaolin.
14. MICA.
Chemical Composition. Silica, alumina, magnesia,
potash, iron.
Always crystallised in thin plates, which may be split
into extremely thin flexible layers. Transparent in thin
layers. Brown or black. Lustre glassy, pearly, or
metallic. Streak white. Specific gravity, 2 -7 to 3*1.
Hardness, 2 to 2- 5. Very easily scratched with a knife.
Infusible. Differs from talc in not having a greasy
feel, in being harder, flexible, and affording thinner layers
perfectly transparent.
Abundant in granite and schist ; fine particles common
in sandstone. Applied to various uses when in large plates,
otherwise of no value. Was formerly used instead of glass
for windows.
15. ZEOLITES.
This name is used for a large class of silicates, com-
prising from fifty to a hundred different minerals, which
all contain water as an essential ingredient, and which
256 DISCRIMINATION OF MINERALS.
melt readily before the blowpipe, and boil up owing to
the disengagement of steam. They occur filling pores
and cavities in basalt, lava, and other rocks. They are
usually white and well crystallised. Can be scratched with
a knife. Of no value. Silica, alumina, lime, soda, potash,
and water are the principal ingredients.
16. CORUNDUM, OR SAPPHIRE.
Chemical Composition. Alumina.
In six-sided prisms, often irregularly shaped. Some-
times in granular masses. Transparent or opaque. Blue,
black ; also red, green, yellow, white. Glassy lustre, some-
times pearly. Fracture uneven or conchoidal. Specific
gravity, 3-9 to 4'2. Hardness, 9. Easily scratches topaz
and quartz. In hardness it is only inferior to the diamond.
Infusible.
Occurs in river sands ; in granite, felspar, magnetic iron,
basalt. As a gem it stands next in value to the diamond r
but its tint must be brilliant and clear. The blue variety
is called Sapphire, the most esteemed shade being deep
velvet blue ; the blood-red variety is the Oriental ruby,
which can be easily distinguished from other red gems by
its superior hardness ; the bright yellow variety is the
Oriental topaz, distinguished by its hardness from the topaz,
yellow tourmaline, and false topaz ; the bright green is the
Oriental emerald ; the bright violet, Oriental amethyst ;
these varieties readily scratch the emerald (No. 11) and
amethyst (see quartz, No. 1) ; one variety exhibits a six-
rayed star inside the prism, and it is called the Asterias.
Dull crystals are called corundum, and grey or black
granular varieties emery ; these two kinds are used for
polishing powder. The ruby is the most highly prized
form of this mineral.
17. SPINEL.
Chemical Composition. Alumina, magnesia.
In octahedrons, occurring only crystallised. Usually
red and transparent ; also white, blue, green, yellow, brown,
black; the dark shades usually opaque. Lustre glassy.
DISCRIMINATION OF MINERALS. 257
Fracture conchoidal. Specific gravity, 3*5 to 4*0. Hard-
ness, 8. Scratches quartz.
Infusible, and thus distinguished from garnet, which it
may resemble. Colour altered transiently by heat. Dis-
tinguished from zircon by its superior hardness and inferior
specific gravity.
Occurs in river sand ; in igneous rocks, gneiss, lime-
stone. The bright transparent varieties are used in jewel-
lery. When red it forms the common spinel, or balas-
ruby, which is distinguished from the Oriental ruby by
its inferior hardness ; bright green, chlorospinel ; orange,
rubicelte ; violet, almandine-ruby ; black, pleonast.
18. CHRYSOBERYL.
Chemical Composition. Alumina, glucina.
In prisms or tables. Transparent or semi-transparent.
Green. Lustre glassy. Fracture conchoidal ; imperfect
cleavage. Specific gravity, 3*5 to 3-8. Hardness, 8'5.
Infusible and unaltered before the blowpipe.
Distinguished from beryl by its specific gravity, its
tabular crystallisation, and its entire infusibility.
Occurs with beryl in river sand, gneiss, and granite.
Pellucid and fine opalescent varieties are used as gems.
19. DIAMOND.
Chemical Co mposition. Carbon .
In octahedrons, tetrahedrons, dodecahedrons, and forms
related to these ; the faces of the crystal sometimes curved.
Transparent. Colourless, yellow, red, green, blue, white,
brown, or black. Lustre adamantine. Breaks with smooth
cleavage planes parallel to the octahedral faces. Specific
gravity, 3*5 ; loses I0-35ths of its weight in water. Hard-
ness, 10. It is the hardest substance known, and scratches
all other minerals and gems.
Infusible. It burns and is consumed at a high tempe-
rature.
Becomes strongly electric when rubbed, and can then
attract light objects ; other gems do not exhibit this pro-
s
258 DISCRIMINATION OF MINERALS.
perty unless polished. Some varieties, after exposure to
the sun, are said to give out light when placed in the
dark.
The distinguishing characters of diamond are, its hard-
ness, crystalline form, clean fracture, its brilliant reflection
and adamantine lustre, the facility with which it may be
electrified by friction, and a peculiar sound on rubbing
together.
Diamonds have been found in quartz- conglomerate and
micaceous sandstone, but are mostly obtained in river
beds.
20. GRAPHITE, OR BLACKLEAD.
Chemical Composition. Carbon .
In six-sided prisms ; but usually in uncrystallised wavy
layers. Opaque. Black. Lustre metallic. Specific gravity,
2. Hardness, 1 to 2. Very easily cut with a knife. Has
a greasy feel ; marks paper like a lead pencil.
Infusible. Burns slowly away.
Molybdenite and foliated tellurium resemble graphite ;
the former has a paler colour than graphite, and the latter
is very easily fusible.
Occurs in gneiss and slate. Valuable for lead pencils
and crucibles.
21. COAL.
Coal and carbonaceous shale differ from all the minerals
which they may resemble by burning before the blowpipe
and leaving a white or brown ash. The quantity of ash
affords an estimate of the purity of the coal.
22. APATITE.
Chemical Composition. Phosphoric acid, lime, fluorine.
In six-sided prisms. Also in masses. Transparent or
opaque. Usually green. Sometimes white, yellow, blue,
red, brown. Lustre resinous. Fracture conchoidal or
uneven. Specific gravity, 3*2. Hardness, 5. Can be
scratched with a knife, using pressure.
Infusible, except on very thin edges. Some kinds
DISCRIMINATION OF MINERALS. 259
phosphoresce when heated. The pure mineral in powder
dissolves slowly in nitric acid without effervescence. The
crystals may resemble beryl, which, however, is too hard
to be scratched with a knife.
Occurs in gneiss, slate, limestone. Of value from its
use in the manufacture of artificial manures.
23. FLUOR-SPAR.
Chemical Composition. Fluorine, calcium.
In cubes or octahedrons. Also in masses. Transparent
or opaque. White or light violet, blue, green, or yellow;
sometimes layers of different tints in the same piece. Lustre
glassy. Breaks with smooth cleavage planes parallel to
the octahedral faces. Specific gravity, 3'0 to 3*2. Hard-
ness, 4. Can be scratched with a knife, but not so easily as
calcite.
Fusible with difficulty ; generally flies to pieces when
heated. Some varieties phosphoresce.
Occurs in veins with lead and silver ores. Used in
etching glass, and as a flux in smelting ; sometimes for
ornaments, but is too brittle. Abundant in many countries,
and of little value.
24. CALC-SPAR, OR CALCITE.
Chemical Composition. Carbonic acid, lime.
In rhombohedrons and other crystalline forms. Also
massive, earthy, or fibrous. Transparent or opaque. White
when pure ; often tinted. Lustre glassy, or dull. Breaks
with smooth cleavage planes parallel to the rhombohedral
faces. Specific gravity, 2-5 to 2*8. Hardness, 3. Easily
scratched with a knife, streak white.
Infusible before the blowpipe, but emits a strong light.
When burned, as in a kiln, it forms quicklime. Effervesces
when vinegar is poured upon it. It completely dissolves
in nitric acid with rapid effervescence.
Calc-spar is one of the most abundant minerals ; it
occurs in cavities and veins of all kinds of rock. The term
8 2
260 DISCRIMINATION OF MINERALS.
calc-spar or calcite is restricted to the crystallised variety.
In an imperfectly crystallised and compact form it exists in
large rocky masses and beds ; all marbles and limestones
consist of it, mixed more or less with impurities. Chalk
and stalactites are nearly pure calcium carbonate. All
varieties of calcium carbonate may be easily distinguished
by being scratched with a knife, giving a white streak
whatever the colour of the mass may be, by effervescing
with an acid, and by being infusible.
25. MAGNESITE.
Chemical Composition. Carbonic acid, magnesia.
In rhombohedrons. Also globular, compact, earthy.
Transparent or opaque. White, yellow, brown. Lustre
glassy. Cleavage parallel to rhombohedral faces. Specific
gravity, 2-8 to 3*0. Hardness, 4 to 5.
Infusible. Dissolves in nitric acid, with very slow
effervescence. Does not yield quicklime when burnt.
Occurs with serpentine and limestone. Used in pre-
paring Epsom salts.
26. DOLOMITE.
Chemical Composition. Carbonic acid, lime, magnesia.
In rhombohedrons, faces often curved. Often granular
or massive. White or dull- tinted. Glassy or pearly.
Specific gravity, 2-8 to 2-9. Hardness, 3-5 to 4.
Infusible. Effervesces in nitric acid, and dissolves
more slowly than calc-spar. Yields quicklime when burnt
Occurs in extensive beds of various ages like limestone.
Used as a building stone, and in the manufacture of Epsom
salts. Difficult to distinguish from calcite without chemical
analysis.
27. ARAGONITE.
Chemical Composition. Same as calc-spar.
It differs from calc-spar in its crystalline form, which
is usually difficult to discern. It often occurs in fibrous
DISCRIMINATION OF MINERALS. 261
clusters or in tangled branches. Specific gravity, 2'9.
Hardness, 3-5 to 4-0.
It has the general characters of calc-spar, but may be
distinguished from it by falling to powder in the blowpipe
flame, as well as by its superior hardness.
28. ROCK SALT.
Chemical Composition. Sodium chloride.
Has the characters of ordinary table salt, but is more
or less impure. Occurs in beds interstratified with sand-
stones and clays, which are usually of a red colour and
associated with gypsum. In the county of Cheshire,
where salt mines are worked, the surface indications are
brine springs supporting a vegetation like that near the
sea-coast ; also occasional sinking of the soil caused by the
removal of the subterranean bed of salt, by spring water
in some cases, and by mining operations in others. Small
and unimportant quantities of salt are often found encrust-
ing various rocks in dry weather.
29. SOLUBLE SULPHATES.
Aluminium sulphate alone or combined with potassium
(alum), and magnesium sulphate (Epsom salts), are often
found encrusting rocks. They are soluble in water, and
easily distinguished by their taste.
30. NITRE.
Chemical Composition. Potassium nitrate.
Nitre or saltpetre is another soluble mineral. It has a
cooling taste. It can be easily distinguished by the vivid
manner in which it burns on red-hot charcoal.
31. GYPSUM, SELENITE, OR ALABASTER.
Chemical Composition. Sulphuric acid, lime, water.
In prisms with oblique terminations ; sometimes resem-
bling an arrowhead. Transparent or opaque. White or
262 DISCRIMINATION OF MINERALS.
dull-tinted. Glassy, pearly, or satin lustre. Cleavage
occurs easily in one direction. Specific gravity, 2-3. Hard-
ness, 2. Very easily cut with a knife.
Fusible with difficulty. In the blowpipe flame it be-
comes white and opaque without fusing, and can then be
easily crumbled between the fingers. Nitric acid does not
cause effervescence.
Occurs in fissures and in stratified rocks, often forming
extensive beds. When burnt it forms plaster-of-Paris ; it
is also used for ornaments, and as a manure.
32. HEAVY SPAK, OR BARYTES.
Chemical Composition. Sulphuric acid, baryta.
In tabular glassy crystals. Also in dull masses. Trans-
parent or opaque. White or tinted. Specific gravity, 4-3
to 4*8 ; its great comparative weight readily distinguishes
it. Hardness, 2-5 to 3 -5.
Splinters fly off the crystals when heated in the blow-
pipe flame. Fusible with difficulty. Not acted upon by
acids.
Occurs with various ores. Used as a white paint.
33. SULPHUR.
Crystallised or massive. Yellow. Eesinous lustre.
Specific gravity, 2-1. Hardness, T 5 to 2-5. Fusible;
burns with a blue flame and well-known odour. Occurs in
volcanic regions, and in beds of gypsum.
34. TIN ORE.
Chemical Composition. Tin, 78'4 ; oxygen, 21-6.
In four-faced prisms and pyramids, having an ada-
mantine lustre. Also in masses and grains (stream tin)
usually dull; sometimes resembling wood (wood tin).
Semi-transparent or opaque. Brown or black ; streak and
powder pale brown. Fracture uneven. Specific gravity,
6*8 to 7*0 ; the great comparative weight is an important
character to observe in distinguishing tin ore from other
. DISCRIMINATION OF MINERALS. 263
minerals. Hardness, 6'0 to 7*0. Cannot be scratched
with a knife, and may thus be distinguished from blende,
which it resembles in lustre and infusibility.
Infusible. When mixed in powder with sodium carbon-
ate, placed on charcoal and covered with a small piece of
potassium cyanide, and then heated in the blowpipe flame,
a malleable globule of metallic tin is obtained.
Occurs in veins, and disseminated in granite, schist,
slate, and porphyry ; and in alluvial deposits. It is a
valuable ore, and the sole commercial source of the
metal.
35. MOLYBDENITE.
Chemical Composition. Molybdenum, 58*9 ; sulphur,
41-1.
In thin plates, like graphite (No. 20). Lustre metallic.
Colour, lead grey. Specific gravity, 4*5 to 4-6. Hardness,
1*0 to 1*5. Easily scratched by the nail.
Infusible. Tinges blowpipe flame faint green. Heated
on charcoal for a long time, it gives off a faint sulphurous
odour, and becomes encrusted white.
Occurs in granite, syenite, and chlorite schist. Not
applied to any particular uses.
36. BISMUTH.
Chemical Composition. Metallic bismuth.
Sometimes crystallised in rhombohedrons closely re-
sembling cubes, but generally massive. Lustre metallic.
White, with a tinge of red, liable to tarnish. Brittle.
Specific gravity, 9-6 to 9-8. Hardness, 2*0 to 2-5. Easily
scratched with a knife.
Easily fusible. Sometimes gives off an odour of garlic,
owing to admixture of arsenic.
Occurs with cobalt, silver, and tin ores in granite and
'slate rocks. Bismuth is a very valuable metal.
37. ANTIMONY SULPHIDE.
Chemical Composition. Antimony, 72*9 ; sulphur, 27*1.
Usually in long columnar or fibrous crystals. Also
264 DISCRIMINATION OF MINERALS.
massive and granular. Lustre metallic. Lead colour.
Often tarnished. Streak metallic. Specific gravity, 4'6 to
4- 7. Hardness, 2. Very easily scratched with a knife.
Easily fusible. Before blowpipe gives off white vapours
and an odour of sulphur, and is entirely volatilised. When
the corner of a large piece of ore is fused, the border of
the fused part is often tinted red. When heated on char-
coal with cyanide of potassium, it gives a globule of metallic
antimony, which is brittle, has a crystalline surface, burns
when strongly heated, emitting white fumes, and can be
entirely volatilised.
Occurs in veins in granite and slate, alone, or with
ores of silver, lead, and other metals. This ore is the
principal commercial source of the metal.
38. ARSENIC.
Chemical Composition. Metallic arsenic.
Seldom distinctly crystallised. Usually in fine granular
or spherical masses. Colour white, usually with a black
tarnish. Streak white, metallic. Brittle. Specific gravity,
5-7 to 5-8. Hardness, 3-5.
Before the blowpipe it quickly volatilises without
fusing, giving off white fumes having an odour of garlic.
Occurs in veins with lead and silver ores.
39. ARSENIC SULPHIDE.
Chemical Composition. Sulphur, arsenic.
In crystals, or massive. Yellow or red. Semi-transparent
or opaque. Eesinous or glassy lustre. Specific gravity, 3*5.
Hardness, 1*5. Volatilised before the blowpipe with a
blue flame.
Occurs in veins with arsenical ores ; in beds of clay,
limestone, and gypsum, and has been observed in lava.
Used as a pigment and in making fireworks, but objection-
able for each of these uses, owing to its highly poisonous
character.
DISCRIMINATION OF MINERALS. 2(35
40. NATIVE IRON.
Chemical Composition. Iron, with a small percentage
of nickel.
Occurs in meteorites. Resembles ordinary iron. Mal-
leable. Is attracted by a magnet. Specific gravity, 7*0
to 7-8.
41. IRON PYRITES.
Chemical Composition. Iron, 46 '7 ; sulphur, 53'3.
In cubes and allied forms ; sides often marked by fine
parallel lines. Also massive. Brass yellow. Lustre metallic.
Fracture irregular. Specific gravity, 4*8 to 5*1. Hardness,
6 to 6- 5. Cannot be scratched with a knife ; scratched by
quartz ; scratches glass with great facility. Strikes fire
with steel (the origin of the term pyrites).
Before the blowpipe it burns with a blue flame, giving
off an odour of sulphur, and ultimately fuses into a, black
magnetic globule.
Abundant. Used as a source of sulphur and sulphuric
acid ; occasionally auriferous. This ore and arsenical
pyrites form the ' mundic ' of miners. It is easily dis-
tinguished from copper pyrites by its hardness ; copper
pyrites being easily cut with a knife. Distinguished from
gold by its hardness and in not being malleable, and in
giving off* sulphurous odours in the blowpipe flame.
42. ARSENICAL PYRITES (MISPICKEL).
Chemical Composition. Iron, 34'4 ; arsenic, 19-6 ;
sulphur, 46 -0.
In flattened prisms. Also massive. White. Lustre
metallic. Streak grey. Fracture uneven. Specific gravity,
6-0 to 6*3. Hardness, 5- 5. Cannot be scratched with a
knife ; scratched by quartz.
Heated before the blowpipe it gives off white arsenical
fumes of a garlic odour, and ultimately fuses into a black
globule.
Abundant in mining districts ; sometimes auriferous.
This ore and iron pyrites form the ' mundic ' of miners.
DISCRIMINATION OF MINERALS.
43. MAGNETIC IRON.
Chemical Composition. Iron, 72-4 ; oxygen, 27*6.
In octahedrons and dodecahedrons. Also in masses
(lodestone) and in grains. Black. Lustre metallic. Streak
or powder black. Fracture irregular. Specific gravity,
5-0 to 5-2. Hardness, 5-5 to 6 -5. Not scratched with a
knife. Magnetic ; it can attract iron filings. Is itself
attracted by a magnet.
Infusible. With borax bead gives the indications of
iron.
Occurs in many rocks, sometimes in beds, or forming
mountainous masses ; common in river sands. Used as an
ore of iron.
44. SPECULAR IRON, HEMATITE, OR MICACEOUS IRON.
Chemical Composition. Iron, 70 ; oxygen, 30.
In tabular crystals or scales. Also fibrous, massive,
granular, earthy. Colour black. Streak or powder, red.
Lustre, metallic or dull. Specific gravity, 4*5 to 5-3.
Hardness of crystals, 5-5 to 6*5. Not scratched with a
knife. Earthy varieties softer, and can be scratched with
a knife.
Infusible. With borax bead gives the indications of
iron. An abundant ore of iron. Oft en gradually changes
into red or brown ochre.
45. RED FERRIC OXIDE, OR RED OCHRE.
Chemical Composition. Iron oxide, and more or less
water.
An uncrystalline earthy variety of the preceding, often
mixed with clay. Colour, bright or dull red. Can generally
be scratched with a knife.
Blackens when heated, but regains its red colour on
cooling. With borax bead gives the indications of iron.
Abundant ore of iron.
DISCRIMINATION OF MINERALS. 267
4G. BROWN FERRIC OXIDE (YELLOW OR BROWN OCHRE).
Chemical Composition. Iron oxide, water.
Like the last, but of a brown, yellow, or black colour.
Eartliy, fibrous, stalactitic. Scratched with a knife.
Blackens before the blowpipe. With borax bead gives
the indications of iron. An abundant ore of iron.
47, TITANIC IRON.
Chemical Composition. Iron oxides and titanic acid in
variable proportions.
In octahedrons or in tabular plates. Also in grains.
Black. Lustre metallic. Streak or powder black. Specific
gravity, 4- 5 to 5*3. Hardness, 5 to 6-5. Not scratched
with a knife.
Infusible. With borax gives the indications of iron.
With microcosmic salt, which is often used instead of borax
in an exactly similar way, it gives a red bead in the
reducing part of the flame, but rather a large quantity
of the mineral must be used to obtain this result.
It is sometimes magnetic.
Its black streak or powder distinguishes it from
specular iron, which it often resembles. Common in some
river sand.
48. CHROMIC IRON.
Chemical Composition. Chromium sesquipxide, ferrous
oxide (alumina, magnesia).
In octahedrons. Usually massive. Black. Lustre
faintly metallic. Streak or powder dark brown. Fracture
irregular. Specific gravity, 4-4 to 4'6. Hardness, 5'5.
Not scratched with a knife.
Infusible. With borax bead gives the characteristic
indications of chromium.
Occurs in serpentine. Used in the preparation of
chromium colours. *
49. GREENEARTH.
Chemical Composition. Iron silicate, and other ingre-
dients. Has a green earthy appearance, often resembling
268 DISCRIMINATION OF MINERALS.
an ore of copper, but is readily distinguished from copper
by blowpipe tests, and by not forming a blue solution in
nitric acid.
50. IRON CARBONATE.
Chemical Composition. Carbonic acid, 37 '9; ferrous
oxide, 62-1.
In rhombohedrons ; faces often curved. Usually
massive, globular, fibrous, or encrusting. Light or dark
brown. Glassy or pearly lustre. Streak white. Specific
gravity, 3 -7 to 3 -9. Hardness, 3*5 to 4-5. Scratched with
a knife.
Infusible. Blackens when heated. With borax bead
gives the indications of iron. Dissolves in nitric acid with
effervescence when heated.
Occurs in beds and nodules in stratified rocks ; in veins
and cavities. It is often mixed with clay (clay ironstone).
Abundant ore of iron.
61. MANGANESE ORES.
Chemical Composition. Various manganese oxides.
Crystallised or massive. Black. Lustre unmetallic ;
dull or shining. Powder or streak brown or black. Spe-
cific gravity, 4 to 5. Hardness generally below 3. Very
easily scratched with a knife.
Infusible. With borax bead gives the characteristic
indications of manganese. Widely distributed. Used in
chemical manufactures.
52. ARSENICAL NICKEL.
Chemical Composition. Nickel, 44 ; arsenic, 56.
Usually in masses of a pale copper colour and metallic
lustre. Specific gravity, 7*2 to 7*8. Hardness, 5 to 5-5.
Scratched with a knife, using pressure.
Before the blowpipe on charcoal it melts, giving out
white arsenical fumes having a garlic odour. It is readily
distinguished by its pale copper red colour and its arsenical
fumes when heated.
DISCRIMINATION OF MINERALS. 26$
Occurs in veins in granite and slate, with ores of cobalt,
silver, copper, bismuth, lead. A valuable source of metallic
nickel.
53. SMALTINE (TIN-WHITE COBALT).
Chemical Composition. Cobalt up to 24 per cent.,
arsenic.
In octahedrons, cubes, dodecahedrons, and allied forms.
Also massive. Tin-white or steel-grey. Lustre metallic.
Streak greyish -black. Fracture uneven. Specific gravity,
6-3 to 6-6. Hardness, 5-5
Fusible. In the blowpipe flame gives off arsenical
fumes (odour of garlic). With borax bead gives the
characteristic indications of cobalt. In nitric acid forms
a pink solution. Eesembles mispickel and iron pyrites,
but is at once distinguished by the test with borax bead.
Its arsenical fumes distinguish it from iron pyrites, and its
crystalline form from mispickel.
Occurs in veins in slate and gneiss. A valuable ore of
cobalt.
54. COBALT BLOOM.
Chemical Composition. Cobalt oxide, 37*6 ; arsenic
acid, 384 ; water, 24'0,
In oblique crystals, with a highly perfect cleavage like
mica. Also in incrustations. Eed or pink, grey, green.
Lustre brilliant pearly. Transparent or opaque. Specific
gravity, 2-9 to 3-1, Hardness, 1-5 to 2. Very easily cut
with a knife.
Fusible in blowpipe flame, evolving arsenical fumes.
When heated on charcoal it gives off an odour of arsenic.
With borax bead gives indication of cobalt.
Occurs in beds and veins with other ores of cobalt. A
valuable ore of cobalt.
55. BLENDE.
Chemical Composition. Zinc, 66*7 ; sulphur, 33 -3.
In dodecahedrons, octahedrons, and allied forms. Also
massive. Yellow, red, brown, black. Lustre adamantine,
270 DISCRIMINATION OP MINERALS.
resinous, or waxy. Transparent or opaque. Breaks with
brilliant cleavage faces in some directions. Specific
gravity, 4*0 to 4*1. Hardness, 3 -5 to 4*0. Easily scratched
with a knife.
Infusible. Emits a strong light when heated, but no
odour of sulphur is perceptible. It is easily distinguished
by its waxy lustre, softness, infusibility, and perfect cleav-
age. It dissolves at once in nitric acid.
Occurs with lead and copper ores. It is the 'black
jack' of miners. An ore of zinc, but more difficult to
smelt than the carbonate and silicate.
56. ZINC CARBONATE (CALAMINE).
Contains 52 per cent, of zinc.
Usually in crusts or masses. White, green, or brown.
Opaque. Pearly or glassy. Specific gravity, 4*1 to 4'5.
Hardness, 5. Can be scratched with a knife, using a little
pressure.
Infusible. On charcoal becomes yellow whilst hot,
white on cooling. Dissolves rapidly with effervescence
when- heated with nitric acid.
Occurs with galena and blende. A valuable zinc ore.
57. ZINC SILICATE (SMITHSONITE).
Contains 53 per cent, of zinc.
In prisms, or massive. White, greenish, bluish, or
brownish. Glassy lustre. Transparent or opaque. Specific
gravity, 3- 3 to 3*5. Hardness, 5.
Infusible. Shines with a green light in the blowpipe
flame. Does not effervesce with nitric acid, but dissolves,
leaving a jelly of silica.
Occurs with zinc carbonate. A valuable zinc ore.
58. GALENA.
Chemical Composition. Lead, 86*6 ; sulphur, 13*4.
In cubes. Also granular, massive. Lead colour.
Metallic lustre. Streak metallic. Breaks into cubical
fragments with bright cleavage faces. Specific gravity,
DISCRIMINATION OF MINERALS. 271
7*4 to 7'7. Hardness, 2-5. Very easily scratched with a
knife.
Easily fusible. Before the blowpipe on charcoal is
reduced to a metallic globule of lead, giving off an odour
of burning sulphur.
Occurs in granite and stratified rocks. Often associated
with copper and other ores. The principal ore of lead.
It usually contains a small quantity of silver.
59. LEAD CAKBONATE (CERUSITE).
Contains 77 per cent, of lead.
In prisms, sometimes united in four or six-rayed crosses.
White or grey'." Transparent or opaque. Lustre glassy.
Specific gravity, 6-4 to 6 -6. Hardness, 3- 5.
Flies violently to pieces in the blowpipe flame. If
placed in a cavity on charcoal, and covered with sodium
carbonate, then carefully fused by the flame, it yields a
globule of metallic lead. In nitric acid it dissolves with
effervescence.
Usually occurs with galena. It is a valuable lead ore.
60. LEAD SULPHATE (AJSTGLESITE).
Contains 68 per cent, of lead.
In slender brilliant crystals upon galena. Also massive.
White or grey. Transparent or opaque. Specific gravity,
6-3. Hardness, 3.
Before the blowpipe fusible, but apt to decrepitate on
charcoal ; with sodium carbonate yields a globule of me-
tallic lead. Differs from lead carbonate in not dissolving
with effervescence in nitric acid.
Usually occurs with galena, and results from its decom-
position.
61. PYROMORPHITE (LEAD PHOSPHATE).
Chemical Composition. Lead oxide, 74'0 ; phosphoric
acid, 15-8 ; lead chloride 10-2.
In stout prisms, grouped together. Also massive.
Bright green or brown. Opaque or semi transparent.
272 DISCRIMINATION OF MINERALS.
Lustre resinous. Streak white. Fracture irregular.
Specific gravity, 6*9 to 7*1. Hardness, 3-5 to 4-0. Easily
scratched with a knife.
Easily fusible. With sodium carbonate on charcoal
the lead is reduced. Soluble in nitric acid.
Occurs sparingly in veins with galena.
62. CINNABAR.
Chemical Composition. Mercury, 86'2 ; sulphur, 13 -8.
In granular, compact, and earthy masses. Sometimes
in crystals, exhibiting adamantine cleavage faces. Opaque
or semi-transparent. Vermilion or brownish red. Specific
gravity, 8-0 to 8-2. Hardness, 2-5. Very easily scratched
with a knife.
Before the blowpipe it volatilises, giving off a strong
odour of burning sulphur. Mixed with dried sodium car-
bonate, and heated over a candle-flame, in an iron spoon,
it gives off vapours of mercury, which may be condensed
on a gold coin held half an inch above the mixture. The
surface of the coin appears whitish at first, but when
rubbed between the fingers becomes brilliantly amalga-
mated. With care, this test easily detects one per cent,
of cinnabar in an ore. The mercury is removed from the
gold coin by gentle heating. The blowpipe tests distin-
guish it at once from red oxide of iron and all other red
minerals.
Occurs in talcose and argillaceous rocks. It is the
principal source of the mercury of commerce.
63. NATIVE MERCURY, OR QUICKSILVER.
The metal in a pure state is rarely found. It occurs
disseminated in liquid globules through sandstone and
other rocks, in cavities of which it may accumulate. It
is easily recognised. A rock suspected to contain mer-
cury may be tested by simply heating it as described
under cinnabar ; but without the addition of carbonate
of soda.
DISCRIMINATION OF MINERALS. 273
64. NATIVE COPPER.
Usually in strings, plates, or irregular masses : some-
times crystalline. Like ordinary copper, but often tar-
nished. Specific gravity, 8-9. Easily scratched with a
knife. Malleable, can be flattened out under a hammer.
Occurs with copper ores.
65. VITREOUS COPPER.
Chemical Composition. Copper, 79*8 ; sulphur, 20*2.
Sometimes in prisms, but usually massive. Blackish lead
grey, tarnished. Streak metallic. Specific gravity, 5'5 to
5 -8. Hardness, 2-5 to 3. Very easily scratched with a knife.
Fusible. Before the blowpipe gives off an odour of
sulphur. When heated on charcoal, a malleable globule
of metallic copper remains, tarnished black, but rendered
evident on flattening under a hammer. With borax bead
gives the indications of copper. Dissolves in nitric acid,
forming a blue solution. (These tests distinguish it from
sulphide of silver.)
Occurs with other copper ores. A valuable ore of copper.
66. COPPER PYRITES.
Chemical Composition. Copper, 34*6 ; iron, 30'9 ; sul-
phur, 34-9.
In tetrahedrons or octahedrons. Usually massive.
Brass yellow, often tarnished. Lustre metallic. Streak
unmetallic, blackish green. Fracture uneven. Specific
gravity, 4*1 to 4*3. Hardness, 3'5 to 4'0. Easily scratched
with a knife.
Fusible. Gives off an odour of sulphur before blow-
pipe. Does not give the indications of copper with borax
bead, or when heated upon charcoal with sodium carbon-
ate. Dissolves in nitric acid, forming a blue solution.
Distinguished from iron pyrites by being easily cut with a
knife ; and from gold by not flattening under a hammer,
and by its greenish powdery streak.
Occurs in granite and slate in lodes or veins. Valuable
ore of copper.
T
274 DISCRIMINATION OF MINERALS.
07. GREY COPPER.
This term includes a variety of ores having a common
crystalline form, generally the tetrahedron also ; a definite
chemical formula, though the ingredients are numerous
and may be variously combined within certain limits.
Sulphur is an invariable ingredient ; and arsenic or anti-
mony, one or both, must be present ; the other ingredients
are copper, iron, zinc, lead, silver, or mercury, in variable
proportions. The copper ranges up to 40 per cent. ; and
in some kinds as much as 30 per cent, of silver has been
found. It also occurs massive. Steel grey to iron black.
Lustre metallic. Streak black, or dark red when zinc is
present. Fracture uneven. Specific gravity, 4'5 to 5-2.
Hardness, 3 to 4. Can be easily scratched with a knife.
Fusible. Before the blowpipe gives off an odour of
sulphur, also white inodorous fumes of antimony, and
occasionally arsenic.
Copper cannot be detected by the blowpipe tests. It
dissolves in nitric acid, forming a greenish brown solution.
Occurs with copper pyrites, galena, and blende. This
ore is wrought for copper, and occasionally for silver.
68. BLACK CUPRIC OXIDE.
Heavy black powder or mass. Soft. Easily distin-
guished from manganese by affording the indications of
copper by the blowpipe tests. It results from the waste
of various copper ores. Valuable as a source of the
metal.
69. RED CUPROUS OXIDE.
Chemical Composition. Copper, 88-8 ; oxygen, 11*2.
In octahedrons and dodecahedrons. Also in granular
and earthy masses. Eed. Lustre adamantine, metallic,
or earthy. Streak red. Semi-transparent or opaque.
Exhibits cleavage parallel with octahedral faces. Specific
gravity, 6. Hardness, 3'5 to 4-0. Can be scratched
with a knife.
Before the blowpipe on charcoal it yields a globule of
metallic copper. With borax bead gives the indications
DISCRIMINATION OF MINERALS. 275
of copper. Forms a blue solution in nitric acid. These
tests distinguish it from red ferric oxide.
Occurs in granite and slate, with copper ores and
galena. Valuable source of the metal.
70. COPPER CARBONATES, BLUE AND GREEN.
Chemical Composition. Oxide of copper, carbonic acid,
water ; the percentage of metallic copper about 56.
In crystals, but usually in fibrous, silky, globular,
incrusting masses. Blue or green. Opaque. Glassy, silky,
or dull. Specific gravity, 3-7 to 4'0. Hardness, 3'5 to
4*0 ; can be scratched with a knife.
Blacken when heated. On charcoal are reduced to a
globule of pure copper. Give the indications of copper
with borax bead. Soluble in nitric acid with effervescence,
forming a blue solution.
Copper Silicate resembles the carbonate, and is distin-
guished by dissolving in nitric acid without effervescecce.
Occur with copper ores, and result from their decom-
position. Valuable sources of the metal.
71. PLATINUM.
In flattened or angular grains or nuggets, which are
malleable. Steel-grey. Lustre metallic. Specific gravity,
IT to 19. As heavy as gold, and, therefore, easily distin-
guished and separated from lighter materials.
Infusible. Insoluble in nitric acid.
Occurs in quartz veins, but principally in alluvial
deposits with gold. Used chiefly for chemical apparatus.
Of great value.
72. GOLD.
In dust, grains, or nuggets in river sand ; or in wiry,
branching, and irregular forms in quartz. Pale or deep
yellow. Malleable. Specific gravity, 15 to 19. Hard-
ness, 2-5 to 3-0.
Fusible without blackening, and without giving off
any odour. Imparts no colour to boiling nitric acid.
T 2
270 DISCRIMINATION OF MINERALS.
The minerals commonly accompanying gold are iron
pyrites, arsenical iron, iron and manganese oxides, galena.,
zinc blende, and copper pyrites in quartz veins ; and
magnetic iron, titanic iron, chromic iron, tin ore, quartz,
zircon, topaz, corundum, diamond, in alluvial deposits.
73. SILVER.
In strings, plates, and branching forms penetrating
quartz, porphyry, slate, granite. Silver white, but usually
tarnished black. Malleable. Specific gravity, about 1O5,
Hardness, 2-5 to 3v
Fusible without giving off any odour. Soluble in
nitric acid, and on adding salt to the solution a white
curd is thrown down which blackens on exposure to sun-
light.
74. SULPHIDE OP SILVER.
Chemical Composition. Silver, 87 ; sulphur, 13.
In dodecahedrons or allied forms. Also massive.
Black. Opaque. Lustre metallic. Streak shining.
Specific gravity,* 7*2. Hardness, 2-0 to 2'5. Very easily
cut with a knife.
Very fusible, giving off an odour of sulphur when
heated. Before the blowpipe on charcoal, with or with-
out sodium carbonate, it yields a white globule of metallic
silver which can be flattened under a hammer.
The ore .is soluble in nitric acid, and on adding salt to
the solution a white curdy precipitate is thrown down y
which blackens on exposure to sunlight.
Occurs in veins in granite, porphyry, and slate, with
arsenic, silver, and lead ores.
75. ANTIMONIAL AND ARSENICAL SILVER ORES.
Several ores of silver contain arsenic and antimony,
as well as sulphur ; the percentage of silver in these ores
varies from 12 to 68. Bed, grey, or black. Lustre ada-
mantine or metallic. Red streak. Specific gravity, 5 to
0. Hardness, 2 to 3. Easily scratched with a knife.
DISCRIMINATION OF MINERALS. i>77
Fusible. Before the blowpipe give off an odour of
sulphur, or arsenical fumes of a garlic odour, or fumes of
antimony. Heated on charcoal with sodium carbonate,
afford a globule of metallic silver.
Nitric acid extracts the silver from these ores, forming
a solution in which salt throws down a white curd, black-
ening on exposure to sunlight.
76. HORN SILVER.
Chemical Composition. Silver, 75-3 ; chlorine, 24*7.
In veins with silver ores. Greenish. Waxy lustre. Hard-
ness, I'O to 1'5. Cuts like wax or horn.
Very easily fusible. Heated with sodium carbonate,
'On charcoal, it yields a globule of metallic silver.
77. LEAD AND ANTIMONY SULPHIDES.
Lead- or steel-grey. Lustre metallic. Hardness not
exceeding 3. Specific gravity about 6.
Fusible. Give off an odour of sulphur and white
fumes of antimony before the blowpipe ; and with sodium
carbonate on charcoal afford a bead of metallic lead.
78. MERCURY, LEAD, SILVER, OR COPPER SELEXIDE.
Gives off a strong odour of horse-radish, due to sele-
nium, when heated before the blowpipe. The metal may
be found in some cases by sodium carbonate on charcoal,
in others by the use of nitric acid.
79. MILLERITE (NICKEL SULPHIDE).
Chemical Composition. Nickel, 64- 9 ; sulphur, 35*1.
In delicate needles. Brass-yellow. Metallic lustre.
Fusible. Hardness, 3 to 5. Imparts to borax the
-colour of nickel.
278 DISCRIMINATION OF MINERALS.
80. WHITE NICKEL.
Chemical Composition. Nickel, 28-3 ; arsenic, 71*7.
Same colour, hardness, specific gravity, and crystalline
form as smaltine, No. 53. Often contains cobalt, and
graduates into smaltine, with which it is usually found.
81. RUTILE.
Chemical Composition. Titanium and oxygen.
In crystals and masses. Eed brown. Streak paler.
Lustre sub-metallic. Hardness, 6'0 to 6-5. Specific
gravity, 4-2 to 4-3.
Infusible. With borax bead, yellowish-green or colour-
less (oxidising), dirty violet (reducing).
8-2. SPHENE.
Chemical Composition. Titanic acid, silica, lime.
In thin wedge-shaped crystals. Yellow, green, brown.
Transparent or opaque. Lustre adamantine. Hardness.
5 to 5*5. Specific gravity, 3-2 to 3*6. Transparent yellow,
bead with borax when hot.
Fusible with difficulty on edges.
83. WOLFRAM.
Chemical Composition. Tungstic acid, iron, manga-
nese.
Crystals or masses. Brownish black. Lustre shining
or dull. Opaque. Hardness, 55. Specific gravity, 7
to 8.
Fusible with difficulty. With borax gives the colour
of iron. Characterised by its great weight. Found often
with tin ores.
84. PITCHBLENDE.
'Chemical Composition. Uranium, 84'S ; oxygen, 15*2..
Massive. Black, opaque. Resinous lustre. Streak
black. Hardness, 5-5. Specific gravity, 6*4 to 6*7.
Infusible. With borax gives yellow bead. Dissolves
in hot nitric acid, forming a yellow solution.
DISCRIMINATION OF MINERALS. 279
85. MAGNETIC IRON PYRITES.
Chemical Composition. Iron, 60'5 ; sulphur, 39-5.
Usually massive. Bronze yellow, tarnished. Metallic
lustre. Feebly magnetic. Hardness, 3*5 to 4-5.
Fusible.
DETERMINATION OF MINERALS.
The following scheme* will show how a mineral may
be determined by means of the tests and characters
already indicated.
The scheme first separates minerals into two divisions
namely, those which possess a metallic lustre, and those
which do not ; and these are further subdivided into
groups by means of other characters.
The various members of each group are enumerated
in the tables, which also give the mode of discrimination.
In most cases it is easy to decide whether a mineral
possesses a lustre like a polished metal, but where the
lustre is less decided there is room for a difference of
opinion ; besides, a mineral may possess a metallic lustre
in some specimens, and not in others.
To provide against this, any mineral about which
difficulty may arise is placed in both divisions, and even in
several groups if necessary. The mineral blende affords
an instance.
From inspection of the scheme it will be observed that
valuable metalliferous minerals possess, as a rule, a metallic
lustre, are heavy in proportion ta their bulk, and are easily
scratched with a knife. The chief exception is tin ore,
which has not a metallic lustre, and which cannot be
impressed with a knife.
In testing it is necessary to bear in mind that a piece
of ore may consist of a single mineral only, either pure, or
mixed with a readily distinguishable gangue ; or it may be
composed of finely disseminated particles of two or more
metalliferous minerals. But by search in localities where
such mixtures occur, pure and isolated specimens of each
* Drawn up by Dr. A. M. Thompson.
280 DETERMINATION OF MINERALS
kind may generally be found. Moreover, after a little
practice the complication of characters which mixtures
present will not interfere with the recognition of their
components.
The scheme is not expected to prove an infallible means
of determining minerals in all possible instances, though
previous experience has shown that it is of use to those
who have learned to apply the tests recommended in the
former part of this chapter. To avoid error, it is necessary
in all cases, before coming to a decision, for the observer
to be convinced that the mineral under examination ac-
tually possesses all the characters of the mineral which it
is supposed to be.
Serious mistakes may arise by drawing inferences from
the scheme only, and neglecting to confirm all the other
characters. In performing the experiments recommended,
the greatest care and accuracy are needful ; to the careless
and superficial observer, or whoever is not proof against
self-deception, the scheme can be of little use.
The characters of each mineral have been described in
the preceding pages.
SCHEME FOR THE DETERMINATION OF MINERALS.
DIVISION I. Minerals possessing a METALLIC lustre.
Experiment. Heat a fragment of the mineral supported
on charcoal, or held with a forceps, in the oxidising point of
the blowpipe flame.
1. An odour of burning sulphur is given off. (See
Group I.)
2. An odour of garlic is given off, or white fumes not
having an odour of sulphur. (See Group II.)
3. No fumes or odour given off. (See Group III.)
Note. An odour of horse-radish indicates Selenium (78).
DIVISION II. Minerals which do NOT poss'ess a metallic
lustre.
Experiment. Scratch the mineral with a knife or sharp
BEFORE THE BLOWPIPE. 281
fragment of quartz or sapphire, according to its hardness,
and observe the colour of the streak.
Note. If the streak is metallic the unmetallic lustre is due to tarnish. See
Division I.
A. The streak is black or coloured.
Experiment. Hold a fragment of the mineral on char-
coal, or with a forceps, in the oxidising point of the blowpipe
flame.
1. Fumes or odour given off. (See Group IV.)
2. No fumes or odour given off. (See Group V.)
B. The streak is uncoloured that is to say, white, or
nearly so.
Experiment. Try to scratch a smooth surface of the
mineral with a splinter of quartz.
3. The mineral can be scratched by quartz. (See
Group VI.)
4. The mineral cannot be scratched bv quartz. (See
Group VII.)
Note 1. The determination must always be made upon pure minerals
free from any adhering foreign matter.
Note 2. Coal and carbonaceous shale are distinguished from minerals
which they may resemble by burning in the blowpipe flame, leaving a more
or less copious ash.
GROUP I.
MINERALS WHICH HAVE A METALLIC LUSTRE, AND WHICH GIVE OFF
SULPHUR.
Experiment. Try to scratch the mineral with a knife,
and observe the streak.
I. Not scratched with a knife ; scratches glass witli
great facility.
Pale brass-yellow or white. Iron pyrites , 41.
II. Scratched with a knife ; does not scratch glass
easily.
* Streak unmetallic.
282 DETERMINATION OF MINERALS
a. Brass-yellow. Solution in nitric acid blue. Copper
pyrites, 66.
b. Black or red. Solution in nitric acid colourless, but
on adding salt water a white curd of precipitate is thrown
down which blackens on exposure to sunlight. Mineral
gives off white fumes before the blowpipe. Antimonial or
arsenical silver ores, 75.
c. Lead- or steel-grey. Abundant white fumes before
the blowpipe. On charcoal with sodium carbonate yields
a globule of metallic lead. Lead and antimony sulphides, 77.
d. Eed. Wholly volatile before the blowpipe. Mixed
with carbonate of soda and heated in an iron spoon over a
candle flame, white vapours are given off which may be
condensed on a gold coin held a little above the mixture,
covering its surface with a brilliant amalgam when rubbed.
Cinnabar i 62.
Note. The streak of Grey Copper (see below) may be unmetallic, often
red.
** Streak metallic.
Experiment. Try fusibility in bio wpipe flame.
A. Infusible :
e. Lead- grey. Very soft. In thin leaves. Tinges the
blowpipe flame faint green. Molybdenite, 35.
Note. Magnetic Iron Pyrites and Vitreous Copper are fusible with diffi-
culty.
B. Fusible : (Easily, except k and m.)
Experiment. Heat a small piece (size of linseed), free
from gangue, on charcoal without adding sodium carbonate,,
hold first in oxidising point and finish in reducing point.
f. Wholly volatile, emitting abundant white fumes.
Mineral has a lead-colour. Crystallised in long prisms,
Antimony sulphide, 37.
g. Globule of metallic lead remains. Mineral has a
lead-colour. Cubical cleavage. Galena, 58.
h. Abundant white fumes given off. Mineral has a lead
or steel colour ; and when heated on charcoal with sodium
carbonate affords a globule of metallic lead. Lead and
antimony sulphides, 77.
BEFORE THE BLOWPIPE. 283
i. Globule of metallic silver remains (facilitated by
adding a little sodium carbonate). Mineral has a black
colour. Is soluble in nitric acid ; on adding salt water
to the solution a white curd is thrown down which blackens
on exposure to sunlight. Silver sulphide, 74.
k. Globule of metallic copper remains (facilitated by
adding a little sodium carbonate) ; discovered by crushing
the residue and flattening it under a hammer. Mineral ha?
a black colour. Does not give off copious white fumes
before the blowpipe. It forms a blue solution with nitric
acid. Vitreous copper, 65.
I. Slaggy globule remains, usually attracted by a
magnet, not yielding copper when crushed. Mineral has
a black or steel colour. It gives off copious white fumes
before the blowpipe. It forms a greenish-brown solution
with nitric acid. Grey copper + 67.
m. Greyish-black magnetic bead remains. Mineral
has a bronze or copper colour. Can be scratched with a
knife ; is attracted by a magnet ; does not form a blue or
green solution with nitric acid. Magnetic iron pyrites, 85.
n. Black magnetic globule. Mineral has the form of
brass yellow needles. Nickel sulphide, 79.
N.B. Before determining the name of a mineral, it is necessary to com-
pare it with the description to which the number refers.
GBOUP II.
MINERALS WHICH HAVE A METALLIC LUSTRE AND WHICH GIVE OFF
EITHER AN ODOUR OF GARLIC WITHOUT SULPHUR, OR WHITE FUMES
WHICH HAVE NOT A GARLIC OR SULPHUROUS ODOUR.
1. The 1 fumes evolved have an odour of garlic
(arsenic).
* Not scratched with a knife. Scratches glass.
a. White metallic. Eesidue after roasting gives indi-
cations of iron, with borax bead. Crystallised in flattened
prisms. Mispickel, 42.
b. White metallic. Eesidue after roasting gives indi-
284 DETERMINATION OF MINERALS
cations of cobalt, with borax bead. Crystallised in octa-
hedrons. Smaltine, 53.
Note. White Nickel resembles Smaltine, and is found with it.
** Scratched with a knife. Does not scratch glass.
c. Pale copper red. Arsenical nickel, 52.
d. Carmine red. Arsenical silver ore, 75.
e. White (as streak shows), tarnished black ; wholly
volatilised by heat. Arsenic, 38.
Note. Metallic Bismuth is frequently associated with Arsenic. Possibly
Grey Copper and Arsenical Silver may be looked for here ; for their charac-
ters see Group I.
2. The furnes evolved have not a garlic odour.
Possibly grey copper, antimonial silver ore, lead and
antimony sulphides, or cinnabar : but these ores generally
give off a faint odour of sulphur in addition to white
fumes. For their characters, see Group I.
N.B. Before determining the name of a mineral, it is necessary to com-
pare it with the description to which the number refers.
GROUP III.
MINERALS WHICH HAVE A METALLIC LUSTRE, AND WHICH GIVE OFF
NO FUMES.
I. The mineral is malleable (can be flattened under a
hammer).
a. Yellow. Gold, 72.
b. Eed. Copper, 64.
c. White, rusty surface ; strongly attracted by a
magnet. Iron, 40.
d. White, feebly or not attracted by a magnet. In-
fusible. Insoluble in nitric acid. Platinum, 71.
e. White, often tarnished ; not attracted by a magnet.
Fusible. Soluble in nitric acid ; on adding salt water to
the solution, a white curd is thrown down, which blackens
on exposure to sunlight. Silver, 73.
II. The mineral is brittle (breaks to pieces under a
hammer).
BEFORE THE BLOWPIPE. 285
Experiment. Observe the colour imparted to a borax
bead in oxidising and in reducing.
1. Violet in oxidising, colourless in reducing.
a. Manganese ores, 51.
2. Red (hot), yellow (cold), in oxidising, bottle-green in
reducing.
Scratch the mineral with quartz and observe the colour of the streak.
b. Brown streak ; mineral not attracted by a magnet.
Brown iron oxide, 46.
c. Eed streak ; not attracted by a magnet. Specular
iron, 44.
d. Black streak ; not attracted by a magnet. Titanic
iron, 47.
e. Black or brown streak ; strongly attracted by a
magnet. Magnetic iron, 43.
Note. Possibly Wolfram (83) ; non -magnetic ; streak reddish-brown or
black ; remarkable for its great weight.
3. Green (hot], blue (cold), in oxidising, m^in reducing,
Red, cuprous oxide, 69.
4. Green in oxidising, green in reducing.
/. Chromic iron, 48.
5. Colourless.
* Infusible.
Minerals HARDER than calc-spar.
g. Eeadily soluble in nitric acid. Scratched with a
knife. Blende, 55.
h. Insoluble in nitric acid. Not scratched with a
knife. With potassium cyanide upon charcoal, yields a
globule of metallic tin. Tin ore, 34.
Minerals SOFTER than calc-spar.
i. Black, like black-lead. Specific gravity about 2.
Graphite, 20.
286 DETERMINATION OF MINERALS
k. Lead-colour. Specific gravity about 4:5. linger
blowpipe flame pale green. Molybdenite,
** Fusible.
35.
Heated on charcoal forms an incrustation which is red
whilst hot, yellow when cold. Bismuth, 36.
6. Note. Green in oxidising, dirty violet in reducing, indicates Entile
(81) ; yellow in oxidising, green in reducing (difficult to obtain in a candle
flame), afforded by a black mineral, soluble in boiling nitric acid, forming a
yellow solution, indicates Pitchblende (84).
Note. Mica may possibly be sought for in this group. It is distinguished
by its lightness of weight, and its capability of being split into very thin, trans-
parent, flexible layers.
N.B. Before determining the name of a mineral, it is necessary to com-
pare it with the description to which the number refers.
GROUP IV.
MINERALS WHICH POSSESS AN UNMETALLIC LUSTRE, A COLOURED STREAK,
AND WHICH GIVE OFF FUMES OR OB-OUR WHEN HEATED BEFORE THE
BLOWPIPE.
a. Colour and streak yellow, entirely volatile, burning
with a blue flame and sulphurous odour. Sulphur, 33.
b. Colour red, orange, or yellow, entirely volatile, burns
on charcoal with a blue flame and garlic odour. Arsenic
sulphide, 39.
c. Streak red. Strong odour of garlic or white fumes
(of antimony) when heated. Dissolved in nitric acid ; on
adding salt to the solution a white curd is thrown down
which blackens on exposure to sunlight. Arsenical or
antimonial silver ore, 75.
d. Streak red. Wholly volatile before blowpipe.
Mixed with sodium carbonate and heated in an iron spoon
over a candle flame, vapours are given off which may be
condensed on a gold coin held a little above the mixture,
covering its surface with a brilliant amalgam when rubbed.
Cinnabar, 62.
Sfjj. Before determining the name of a mineral, it is necessary to com-
pare it with the description to which the number refers.
BEFORE THE BLOWPIPE. 287
GEOUP V.
MINERALS WHICH POSSESS AN UNMETALLIC LUSTKE, A COLOURED STREAK,
BUT WHICH GIVE OFF NO FUMES OR ODOUR BEFORE THE BLOWPIPE.
Experiment Observe the colour imparted by the mineral
to a borax bead, both in oxidising and reducing.
1. Green (hot), blue (cold), in oxidising ; red in reduc-
ing (difficult to obtain).
a. Colour of mineral blue or green. Copper carbonate ,
70.
b. Colour of mineral red. Red cuprous oxide , 69.
c. Colour of mineral black. Black cuprous oxide, 68.
2. Green in oxidising, green in reducing.
d. Colour of mineral black. Chromic iron, 48.
3. Red (hot), yellow (cold), in oxidising : bottle-green in
reducing.
e. Colour of mineral green. Greenearth, 40.
/. Colour of mineral brown ; blackened during heating.
Brown ferric oxide, 46.
g. Colour of mineral red ; blackened during heating.
Red ferric oxide, 45.
Note. Iron Arseniate is green or yellow, Iron Phosphate is blue.
4. Amethyst in oxidising, colourless in reducing.
h. Colour of mineral brown or black. Manganese
-ores, 51.
Note. Manganese Carbonate is rose red or brownish, streak white.
5. Colourless in both oxidising and reducing.
i. Easily scratched with a knife. Cleavage. Soluble
in nitric acid. Blende, 55.
k. Not scratched with a knife. No cleavage. Heated
with potassium cyanide upon charcoal yields a malleable
globule of metallic tin. Tin ores, 34.
Note. Cobalt Bloom (54) has a red colour and streak ; it gives the indi-
cations of Cobalt with borax bead. Entile, Wolfram, and Pitchblende might
be considered to have an unmetallic lustre. See under Group III.
Njg. Before determining the name of a mineral, it is necessary to com-
pare it with the description to which the number refers.
283 DETERMINATION OF MINERALS
GEOUP VI.
MINERALS WHICH HAVE AN UNMETALLIC LUSTRE, AND WHICH ARE
SCRATCHED BY QUARTZ, SHOWING A WHITE STREAK.
I. Minerals soluble in water (having a taste).
a. Taste of common table salt. Salt, 28.
b. Sweetish astringent taste. Alum, 29.
c. Bitter taste of Epsom salts. Epsomite, 29.
d. Cooling taste. Causes vivid combustion when
thrown on a piece of red-hot charcoal. Nitre, 30.
Note. Zinc (white), Iron (green), and Copper (blue) sulphates have a
nauseous metallic taste.
II. Minerals insoluble in water ; but soluble in nitric
acid (best determined by placing a very little of the
powder in a test tube, pouring on it a few drops of the
acid, and heating if requisite).
N.B. All of these can be scratched witli a knife.
* Easily fusible.
a. Green, yellow, or brown. Hardness, 3 to 4. Glo-
bule of lead, on charcoal with sodium carbonate. No
effervescence in nitric acid. Pyromorphite, 61.
b. White or grey. Hardness, 3 to 4. Flies violently
to pieces in the blowpipe flame. Globule of lead on char-
coal with sodium carbonate. Effervesces in nitric acid.
Cerusite, 59.
Note. White or grey. Effervesces in nitric acid. Specific gravity 4'3.
Does not fly to pieces when heated ; yields no metal on charcoal, WITHERITE
BARIUM CARBONATE.
** Infusible.
c. Hardness, 3. Dissolves with very brisk efferves-
cence in cold nitric acid. Does not crumble to powder in
the blowpipe flame. Calc-spar, 24.
d. Hardness, 3- 5 to 4. Dissolves like calc-spar ; but
crumbles to powder in the blowpipe flame. Aragonite, 6.
e. Hardness, 3 -5 to 4. Dissolves much slower than
calc-spar ; but difficult to distinguish without chemical
analysis. Dolomite, 26.
BEFOEE THE BLOWPIPE. 289
f. Hardness, 4 to 5. No effervescence in cold nitric
acid, very slight in hot. Magnesite, 25.
g. Hardness, 3'5 to 4*5. Effervescence in hot nitric
acid. The mineral blackens when heated in the blowpipe
flame ; it gives the indication of iron with borax. Iron
carbonate, 50.
h. Hardness, 5. Effervescence in hot nitric acid.
Heated on charcoal w r ith sodium carbonate it forms a
crust which is yellow whilst hot. Ckdamine, 56.
i. Hardness, 5. Dissolves in nitric acid without effer-
vescence, leaving a jelly of silica. In blowpipe flame it
shines with a green light. Zinc silicate. 57.
j. Hardness, 3-5. Effervesces in hot nitric acid. Flies
violently to pieces in the blowpipe flame. On charcoal
with sodium carbonate yields a globule of lead. Cerusite,
59.
Jc. Hardness, 3- 5 to 4. Brilliant cleavage. Eapidly
soluble in nitric acid with effervescence. Heavy (specific
gravity, 4). Blende, 55.
1. Hardness, 4- 5 to 5. Slowly soluble in nitric acid
without effervescence. Apatite, 22.
Note. Hardness, 3'5. Effervesces in hot nitric acid. Tinges the blow-
pipe flame crimson. Strontianite (Strontium carbonate).
HI. Minerals insoluble in both water and nitric acid.
* Easily fusible.
HARDNESS.
TO to 1*5. Like wax. On charcoal with sodium carbon-
ate yields a globule of silver. Horn silver,
76.
2-5 to 3-0. White or grey. On charcoal with sodium
carbonate yields a globule of lead. Angle-
site, 60.
3-5 to 4-0. Green or brown. On charcoal with sodium
carbonate yields a globule of lead. Pyro-
morphite, 61.
3*5 to 6'5. Usually white. Swell up before the blowpipe.
Occur in cavities of rocks. Zeolites, 15.
5-0 to 6-0. Black, green, or white. Prisms, fibrous or
flax-like. Hornblende and augite, 6.
u
290 DETERMINATION OF MINERALS
HARDNESS.
6 -5 to 7 *5. Usually red. Crystallised in dodecahedrons.
Garnet, 9.
Note. Glass (artificial) may be included here ; easily distin-
guished by its appearance, easy fusibility, and inferior hard
ness, 4*5 to 5*5.
** Infusible, or not easily fusible.
rO to 1-5. Light green. In layers. Unctuous. Talc, 3.
1-5. Dark green. In layers or granular. Some-
times fusible to a black glassy bead. Chlo-
rite , 4.
2'0 to 3*5. Green. Turns white in blowpipe flame. Feels
often greasy. Serpentine, 5.
2-0. Usually white. After heating becomes opaque,
and can be rubbed to powder between the
fingers. Fusible with difficulty. Gypsum, 31.
2-0 to 2-5. Grey or black. Can be split into extremely
thin semi-transparent flexible layers. Mica,
14.
3-0 to 3-5. White or grey. Tabular crystals or massive.
Heavy (specific gravity, 4-3 to 4-8). Splinters
fly off crystals when heated. Fusible with
difficulty. Heavy spar, 32.
4-0. Cubes or massive. Octahedral cleavage. Flies
to pieces when heated. Specific gravity, 3
to 3-2. Fusible with difficulty. Fluor-spar,
23.
5-0. Six-sided prisms, massive. Green or white.
Specific gravity, 3 to 3 -3. Generally flies
to pieces when heated. In fine powder,
slowly soluble in nitric acid. Apatite, 22.
5-0 to 6-0. Prisms, fibrous or flax-like. Cleavage. Green,
black, or white. Generally fusible. Specific
gravity, 2-9 to 3-5. Hornblende and augite, 6.
Note. Possibly Epidote. Yellowish green. Hardness, 6'0
to 7-0.
6-0. Prisms or massive. Flesh-colour, white, or
tinted. Cleavage. Specific gravity, 2-3 to
2-8. Felspar, 13.
BEFORE THE BLOWPIPE. 291
HARDNESS.
() to 7*0. Masses or grains like glass. Black or green.
Specific gravity, 3'3 to 3'6. Chrysolite, 7.
6-5 to 7' 5. Dodecahedrons. Usually red. Generally fus-
ible. Eed infusible varieties impart a green
colour to borax bead. Specific gravity, 3*5
to 4-3. Garnet, 9.
5-5 to 6'5. Never crystallised. White, tinted, or chatoyant.
Specific gravity, 1-9 to 2-3. Opal, 2.
7'0. Six-sided prisms. Specific gravity, 2 -6.
Quartz, 1.
Note. Spliene (82) occurs in acute thin crystals. Yellow,
green, or brown. Specific gravity, 3'2 to 3'6.
Chiastolite and other Aluminium Silicates occur in
prisms. Brown or white. Opaque. Hardness, 6 to 7'5.
Infusible. Specific gravity, 3 to 3*7.
N.B. Before determining the name of a mineral, it is neces-
sary to compare it with the description to which the num-
ber refers.
GROUP VIL Including the Gems.
MINERALS WHICH HAVE AN UNMETALLIC LUSTRE, AND WHICH ARE
NOT SCRATCHED BY QUARTZ.
Note. Tin ore (34) may occur here. Hardness, 6 to 7. Bemarkable for
its great weight. Specific gravity, 6'8 to 7'0. Brown or black. With cya-
nide of potassium, on charcoal, yields a globule of Tin.
GEMS.
If crystallised, the determination is greatly facilitated,
but if the crystallisation is not evident, the hardness will
be found a sufficiently near indication in the following
scheme. The requisites in determining the hardness of
gems are good specimens of sapphire, topaz, and quartz
possessing smooth bright surfaces, as well as sharp points
or corners. The hardness is determined by drawing the
sharp points of these three test-minerals over smooth bright
surfaces of the mineral under trial. (See hardness.) Or
it may be ascertained with great certainty by means of a
small angular fragment broken off the mineral to be tried.
For convenience of holding, it should be mounted on a
stick of sealing-wax, which is best done by previously
heating the fragment, held with a forceps over a candle
flame, and applying it, whilst hot, to the wax ; the hard
U 2
292 DETERMINATION OF MINERALS
ness may then be ascertained by drawing the fragment
thus mounted across smooth surfaces of the sapphire,
topaz, and quartz, and observing whether a scratch is
thus produced. In this way the hardness of a mineral not
larger than a grain of sand may be determined.
The facility of cleavage is an important character in
the topaz and the diamond.
The electrical properties are characteristic in the dia-
mond, topaz, and tourmaline.
The specific gravity is a most important aid in distin-
guishing gems, but it cannot be attempted with an ordinary
pair of grain scales in the case of minerals weighing under
ten grains at the very least.
Many varieties of zircon and spinel, and in some cases
the topaz and emerald, can readily be distinguished by the
effect of heat, which is applied most conveniently by the
blowpipe.
The diamond may be easily distinguished by the use of
a small writing or scratching diamond, which fails to mark
the faces of a real diamond when drawn lightly across, but
scratches all other gems with facility. After practice, the
sound caused by tapping two diamonds together becomes
characteristic.
1. CRYSTALLISED IN PRISMS.
HARDNESS.
7*0. Six-sided prisms or pyramids ; sides of prism
finely marked across. No cleavage. Frac-
ture fconchoidal. Infusible. When two
pieces are Drubbed together in the dark,
they emit a phosphorescent light and a
peculiar odour. Specific gravity, 2 -5 to 2*8.
Quartz, 1.
7*0 to 8*0. Prisms, three, six, nine, or more sided ; fur-
rowed lengthwise. Black or coloured.
Opaque or transparent. Smooth sides of
prisms become electric by friction. No
cleavage. Infusible or nearly so. Specific
gravity, 3'0 to 3- 3. Tourmaline, 8.
BEFORE THE BLOWPIPE. 293
FARDNESS.
7*5. Prisms, four-sided, not furrowed. Wine-red,
brown, or white. Decolourised permanently
by heat. Adamantine cleavage, but rather
difficult to obtain. Infusible. Specific
gravity, 4 to 5. Zircon, 12,
7 to 8. Prisms, six or more sided. Usually green.
Imperfect cleavage. Fusible with difficulty
in thin edges ; becoming clouded by heat.
Specific gravity, 2'7 ; its low specific gravity
combined with its hardness is characteristic.
Beryl or Emerald, 11.
7'5 to 8. Sides of prisms often finely marked lengthwise.
Cleavage across the prisms brilliant and
easily obtained. Smooth surfaces become
strongly electric by friction. Infusible, but
sometimes blistered and altered in colour by
heat. Specific gravity, 3-5. Topaz, 10.
8'5. Prisms or tables. Green. Infusible and un-
altered by heat. Specific gravity, 3-5 to 3 '8.
Chrysoberyl, 18.
9'0. Prisms, six-sided or irregular. Cleavage across
the prisms, but difficult to obtain. Fracture
irregular. Blue, green, black, red, yellow,
brown, or white. Infusible. Easily distin-
guished by its great hardness, scratching
all other gems except diamond. Specific
gravity, 3*9 to 4- 2. Sapphire, ruby, or co-
rundum, 16.
2. CRYSTALLISED IN OCTAHEDRONS OR DODECAHEDRONS.
HARDNESS.
7'5. See characters above. Zircon, 12.
'6-5 to 7-5. Dodecahedrons. Usually red. Most varieties
are easily fusible ; red infusible varieties
impart a green colour (due to chromium)
to borax bead. Specific gravity, 3-1 to 4-3.
Garnet, 9.
294
COLOURED FLAMES.
HARDNESS.
8-0. Octahedrons. Usually red or black, also blue,,
green, yellow, and brown. Some red varie-
ties become opaque- and black when heated ;
rose-red varieties become green when heated,,
but regain their original colour on cooling.
Specific gravity, 3 -5 to 4'0. Spinel, 17.
10-0. Octahedrons, dodecahedrons, or modifications
of these forms. Crystalline facets often
curved. Cleavage perfect. Lustre, brilliant
adamantine. Usually colourless or straw-
coloured. Not water-worn. Strongly elec-
tric by the slightest friction. Specific gravity,
3-5. Diamond, 19.
N.B. Before determining the name of a mineral, it is neces-
sary to compare it with the description to which the number
refers.
For further particulars respecting gems and precious
stones see the concluding chapter of this volume.
COLOURED FLAMES. There are a great number of sub-
stances best detected by the colours they impart to the
flame of the blowpipe. Indeed, so important is this point
that it has been thought advisable to collect all the facts
known on this subject into one place, rather than scatter
them over the work. These experiments are best made
in a dark room, and with a very small flame.*
BLUE FLAMES.
Large intense blue . Copper chloride.
Pale clear blue . . Lead.
Light blue . . Arsenic.
Blue ... . Selenium.
Greenish blue . . Antimony.
Blue mixed with green Copper bromide.
GREEN FLAMES.
Intense emerald green . Thallium.
Very dark green, feeble . Ammonia.
Dark green . . . Boracic acid.
Full green . . . Tellurium.
Full green Copper.
Emerald green mixed with blue . \ PP er j did ?-
( Copper bromide.
Pale green . . . . Phosphoric acid.
Very pale apple -green . . . Barium.
Intense whitish green . . . Zinc.
Bluish green Tin binoxide.
* Griffin's ' Blowpipe Analysis,' p. 148.
COLOURED FLAMES. 295
YELLOW FLAME.
Intense greenish yellow . . . Sodium.
RED FLAMES.
Intense crimson .... Lithium.
Eed Strontium.
Keddish purple Calcium.
Violet Potassium.
Chlorine, combined with copper, gives an intense blue
flame. This phenomenon may be produced as follows :
Take a piece of thin brass wire, and bend one end of it
several times upon itself; place upon this some microcosmic
salt, and fuse it until it has acquired a green colour. Then
add to it the substance suspected to contain chlorine, and
place it in the oxidising flame, just at the point of the blue
flame ; if any chloride be present a splendid blue colour
will be produced.
Lead. The blue colour produced by this metal is
readily obtained. Fragments of a mineral must be held
in the tongs, and powder may be assayed on charcoal.
Arsenic, in the metallic state, gives rise to a light blue
flame.
Selenium and Antimony, when treated in the same
manner, afford characteristic flames.
Bromine. If any substance containing bromine be
placed in a bead of fused microcosmic salt on the brass
wire, and then in the oxidising flame, a bright blue flame
with emerald-green edges will be produced.
Boracic Acid. The following is Dr. Turner's process
for the detection of boracic acid. The substance is to be
mixed with a flux composed of 1 part of fluor-spar and 4J
parts of potassium bisulphate. This mixture is to be
made to adhere to the moistened end of a platinum wire,
and held at the point of the blue flame ; at the instant of
fusion a dark green flame will be produced. It may also
be produced by merely dipping the mineral in sulphuric
acid, and exposing it to the blowpipe blast. In case a
very small quantity of boracic acid is contained in a
mineral, the following process may be employed : The
substance must be fused with potassium carbonate on
296 COLOURED FLAMES.
charcoal, moistened with a drop or two of sulphuric acid,
and then a few drops of alcohol ; the latter will burn
with a green flame when exposed to the flame of the
blowpipe.
Tellurium. The peculiar flame given by this metal is
produced by heating a portion of its oxide on charcoal in
the reducing flame.
Copper. All the compounds of copper, except those
in which bromine and chlorine enter, give a beautiful
green flame. The soluble salts give it per se, but the
insoluble require moistening with sulphuric acid.
Iodine and Copper. To the bead of microcosmic salt
on the brass wire add any compound containing iodine,
and a bright green flame will be produced when the mass
is heated in the oxidising flame.
Phosphoric Acid. The phosphates, when moistened
with sulphuric acid, give a light green tint to the outer
flame.
Barium. The soluble salts of barium give a light green
colour to the outer flame when moistened with water.
Zinc, when exposed to the blowpipe flame, burns with
an intense whitish-green light.
Sodium. Any salt of soda, or substance containing soda,
being exposed to the outer flame, gives a brush of intensely
coloured flame, of a fine amber or greenish-yellow.
Water. Certain minerals containing water give a feeble
yellowish tint to the flame due to the presence of a little
sodium.
Strontium. All the salts of this substance which are
soluble in water give a crimson tint to the flame, which
does not endure after the substance is fused. Strontium
carbonate must be moistened with hydrochloric acid, and
strontium sulphate must be reduced to a state of sulphide
by ignition with charcoal ; it must then be moistened with
hydrochloric acid, after which treatment it will exhibit
the characteristic flame.
Lithium. All that has been said of strontium applies to
lithium, but the coloured flame given by lithium is perma-
nent, whilst that afforded by strontium is evanescent.
COLOURED FLAMES, 297
Calcium acts as strontium.
Potassium, treated as sodium, gives a purplish light.
This is, however, very liable to be masked by the intense
yellow communicated by small quantities of sodium, which
are almost always present with potassium. The potassium
flame can be seen even in presence of sodium by looking
through cobalt-coloured glass.
CHAPTER VIII.
VOLUMETRIC ANALYSIS.
THE main feature of volumetry is not so much analysis in
the proper sense of the term, as the quantitative estima-
tion of one principal constituent of a substance.
This estimation is done by means of solutions, con-
taining a certain quantity of reagents in a certain volume,
which is called a standard solution, the quantity used of
such solution being measured by graduated tubes (burettes,
pipettes, &c.)
The reaction of a volumetric analysis can be of three
different kinds, according to the reagent used and to the
substance to be estimated.
1. The substance to be analysed being an acid or a
base, it can be saturated by a suitable standard solution
(saturation-analysis, used for acids, potash, soda, &c.)
2. The substance to be assayed maybe precipitated by
the standard solution, and the completion of the process is
observed when no further precipitate occurs (precipitation-
analysis, e.g. Pelouze's copper assay, Gay-Lussac's silver
assay).
3. The substance to be estimated becomes, by the
standard solution, either oxidised or reduced, and by the
performance of this process certain colours will appear or
disappear, from which the completion of the process is to
be observed (oxidation or reduction-analysis, e.g. Schwarz's
copper assay).
These processes of volumetric analysis are frequently
used in assaying.
The principle of volumetric analysis may be fully ex~
plained by the following examples given by Fresenius.*
* Fresenius's ' Quantitative Analysis,' fourth edition, p. 76.
PEINCIPLES O'F VOLUMETRIC ANALYSIS. 299
' Suppose we have prepared a solution of chloride of
sodium of such a strength that 100 c.c. will exactly pre-
cipitate 1 grm. silver from its solution in nitric acid, we
can use it to estimate unknown quantities of silver. Let
us imagine, for instance, we have an alloy of silver and
copper in unknown proportion, we dissolve 1 grm. in nitric
acid, and add to the solution our solution of chloride of
sodium, drop by drop, until the whole of the silver is
thrown down, and an additional drop fails to produce a
further precipitate. The amount of silver present may
now be calculated from the amount of solution of chloride
of sodium used. Thus, supposing we have used 80 c.c.,
the amount of silver present in the alloy is 80 per cent. ;
since, as 100 c.c. of the solution of chloride of sodium will
throw down 1 grm. of pure silver (i.e. 100 per cent.), it
follows that every c.c. of the chloride of sodium solution
corresponds to 1 per cent, of silver.
' Another example. It is well known that iodine and
sulphuretted hydrogen cannot exist together ; whenever
these two substances are brought in contact decomposi-
tion immediately ensues, the hydrogen separating from the
sulphur and combining with the iodine (I + HS = HI + S).
Hydriodic acid exercises no action on starch paste, whereas
the least trace of free iodine colours it blue. Now, if we
prepare a solution of iodine (in iodide of potassium) con-
taining in 100 c.c. 0-7470 grm. iodine, we may with this
decompose exactly O'l grm. sulphuretted hydrogen ; for
17 : 127:: 0-1 : 0-7470.
' Let us suppose, then, we have before us a fluid con-
taining an unknown amount of sulphuretted hydrogen,
which it is our intention to estimate. We add to it a
little starch paste, and then, drop by drop, our solution
of iodine, until a persistent blue coloration of the fluid
indicates the formation of iodide of starch, and hence the
complete decomposition of the sulphuretted hydrogen.
The amount of the latter originally present in the fluid
may now be readily calculated from the amount of solution
of iodine used. Say, for instance, we have used 50 c.c.
of iodine solution, the fluid contained originally 0-05 sul-
300 PRINCIPLES OF VOLUMETRIC ANALYSIS.
phuretted hydrogen ; since, as we have seen, 100 c.c. of
our iodine solution will decompose exactly O'l grm. of
that body.
' Solutions of accurately known composition or strength,
used for the purposes of volumetric analysis, are called
standard solutions. They may be prepared in two ways,
viz. (a) by dissolving a weighed quantity of a substance
in a definite volume of fluid ; or (b) by first preparing a
suitably concentrated solution of the reagent required,
and then estimating its exact strength by a series of
experiments made with it upon weighed quantities of the
body for the estimation of which it is intended to be
used.
' In the preparation of standard solutions by method a,
a certain definite strength is adopted once for all, which is
usually based upon the principle of an exact correspond-
ence between the number of grammes of the reagent
contained in a litre of the fluid, and the equivalent number
of the reagent (H=l). In the case of standard solutions
prepared by method b, this may also be easily done, by
diluting to the required degree the still somewhat too
concentrated solution, after having accurately estimated
its strength ; however, as a rule, this latter process is only
resorted to in technical analyses, where it is desirable to
avoid all calculation. Fluids which contain the equivalent
number of grammes of a substance in 1 litre are called
normal solutions ; those which contain y 1 ^ of this quantity,
decinormal solutions.
' The estimation of a standard solution intended to
be used for volumetric analysis is obviously a most im-
portant operation ; since any error in this will, of course,
necessarily falsify every analysis made with it. In scien-
tific and accurate researches it is, therefore, always ad-
visable, whenever practicable, to examine the standard
solution no matter whether prepared by method a or
by method />, with subsequent dilution to the required
degree by experim en ting with it upon accurately weighed
quantities of the body for the estimation of which it is
to be used.
PRINCIPLES OF VOLUMETRIC ANALYSIS. 301
' In the previous remarks no difference has been made
between fluids of known composition and those of known
power ; and this has hitherto been usual. But by accept-
ing the two expressions as synonymous, we take for granted
that a fluid exercises a chemical action exactly correspond-
ing to the amount of dissolved substance it contains ; that,
for instance, a solution of chloride of sodium containing
1 eq. NaCl will precipitate exactly 1 eq. silver. This pre-
sumption, however, is very often not absolutely correct.
In such cases, of course, it is not merely advisable, but even
absolutely necessary, to estimate the strength of the fluid
by experiment, although the amount of the reagent it
contains may be exactly known, for the power of the fluid
can be inferred from its composition only approximately,,
and not with perfect exactness. If a standard solution
keeps unaltered, this is a great advantage, as it dispenses
with the necessity of estimating its strength before every
fresh analysis.
' That particular change in the fluid operated upon by
means of a standard solution, which marks the completion
of the intended decomposition, is termed the FINAL REACTION.
This consists either in a change of colour ', as is the case
when a solution of permanganate of potash acts upon an
acidified solution of protoxide of iron, or a solution of
iodine upon a solution of sulphuretted hydrogen mixed
with starch paste ; or in the cessation of the formation of a
precipitate upon further addition of the standard solution,
as is the case when a standard solution of chloride of sodium
is used to precipitate silver from its solution in nitric acid ;
or in incipient precipitation , as is the case when a standard
solution of silver is added to a solution of hydrocyanic
acid mixed with an alkali ; or in a change in the action of
the examined fluid upon a particular reagent, as is the case
when a solution of arsenite of soda is added, drop by drop,
to a solution of chloride of lime, until the mixture no longer
imparts a blue tint to paper moistened with iodide of
potassium and starch paste, &c.'
The only condition on which the volumetric system of
analysis can be carried on successfully is, that the greatest
302 APPARATUS FOR VOLUMETRIC ANALYSIS.
care is exercised with respect to the graduation of the
measuring instruments, and the strength and purity of the
standard solutions. A very slight error in the analytical
process becomes considerably magnified when calculated
for pounds, hundredweights, or tons of the substance
tested. The end of the operation in this method of analysis
is in all cases made apparent to the eye. (Button.)
STANDARD SOLUTIONS.
A very useful form of bottle for their preservation is
the ordinary wash-bottle, or any common bottle fitted with
the same arrangement of tubes. The mouth end of the
blowing tube should be furnished with a tightly fitting-
india-rubber cap to prevent alteration of the standard by
evaporation. A similar cap over the point will be useful,
although not absolutely necessary, or it may be closed by
a small cork fitting over it. Burettes can then be filled
with the solution without its frothing, and if the tube
which enters the liquid does not reach the bottom of the
bottle, the sediment, if any, is not disturbed ; another ad-
vantage is that the solution does not come into contact
with the cork, nor can any dust enter.
The Instruments and Apparatus.
The Burette, or graduated tube for delivering the stan-
dard solution, may be obtained in a great many forms
under the names of their respective inventors, such as Mohr,
Gay-Lussac, Binks, &c. ; but as some of these possess a de-
cided superiority over others, it is not quite a matter of in-
difference which is used, and therefore a slight description
of them may not be out of place here. The burette, with
india-rubber tube and clip, contrived by Dr. Frederic Mohr
of Coblentz, shown in figs. 82 and 83, has the preference
over all others for general purposes.
The advantages possessed by this instrument are, that
its constant upright position enables the operator at once
APPARATUS FOR VOLUMETRIC ANALYSIS.
303
to read off the number of degrees of test solution used for
any analysis. The quantity of fluid to be delivered can
be regulated to the greatest nicety by the pressure of the
thumb and finger on the spring clip* and the instrument
FIG. 82.
FIG. 83.
not being held in the hand, there is no chance of increas-
ing the bulk of the fluid by the heat of the body, and thus
leading to incorrect measurement, as is the case with Binks's
or Gay-Lussac's form of instrument. The principal disad-
vantage of these two latter forms of burette is, that a correct
reading can only be obtained by placing them in an up-
right position, and allowing the fluid to find its perfect level.
304
THE BURETTE.
FIG. 84.
FIG. 85.
The preference should, therefore, unhesitatingly be given to
Dr. Mohr's burette wherever it can be used ; the greatest
drawback to it is that it cannot be used for potassium
permanganate, in consequence of its india-rubber tube,
which decomposes the solution.
We are again indebted to Dr. Mohr for another form
of instrument to overcome this difficulty, viz. the foot
burette, with india-rubber
ball, shown in fig. 84.
The flow of liquid from
the exit tube can be regu-
lated to a great nicety by
pressure upon the elastic
ball, which is of the ordi-
nary kind sold for children,
and has two openings, one
cemented to the tube with
shellac, and the other at the
side, over which the thumb
is placed when pressed, and
on the removal of which it
refills itself with air.
Gay-Lussac's burette,
supported in a wooden
foot, may be used instead
of the above form, by in-
serting a good fitting cork
into the open end, through
which a small tube bent at right angles is passed. If the
burette is held in the right hand, slightly inclined towards
the beaker or flask into which the fluid is to be measured,
and the mouth applied to the tube, any portion of the solu-
tion may be emptied out by the pressure of the breath, and
the disadvantage of holding the instrument in a horizontal
position, to the great danger of spilling the contents, is
avoided ; at the same time the beaker or flask can be held
in the left hand and shaken so as to mix the fluids, and by
this means the end of the operation is more accurately
determined .
THE BUEETTE.
305
FIG. 86. FIG. 87.
Fig. 85 will show the arrangement here described.
Mr. J. Blodget Britton has described, in the ' Chemical
News ' for August 5, 1870, a burette for use in determining
iron in metals and ores.
Figs. 86 and .87 represent two of the kind, but of
different patterns, mounted on walnut-wood stands; the
former is for metals, and
the latter, of smaller capa-
city, for ores.
Securely fastened to the
upright B is a graduated
tube A, having its lower
part drawn out in the
usual manner, but bent
outwards at an angle of
about 25. E is a piece of
cork riveted into a sheet
steel spring D, which presses
tightly, by means of the
latter, against the vent of
the tube, c is a thumb-
screw passing through the
frame and bearing against
the spring.
The tube of fig. 87 has
a capacity of 100 c.c., and
is graduated into tenths, or
1000 ; but that of fig.. 86
has a capacity of 150 c.c.,
and is graduated into twen-
tieths, though only at its lower and narrow part.
Modes of operating. Place a small narrow-necked
funnel in the tube, as shown by the figures ; pour in the
solution to be used until it quite reaches the funnel, and
then remove the latter to carry away any floating bubbles ;
turn the thumbscrew and bring the top line of the solution
exactly to the zero line of the scale ; stop the flow, and after-
wards touch the point of the cork with a glass rod to take
from it any adhering drop. The instrument is then ready
x
306
THE PIPETTE.
FIG. 88.
50CC
10 CC
for use. By means of the thumbscrew the dropping may be
controlled with extreme nicety or instantly stopped. Cork,
after a little use, becomes quite inert towards potassium
permanganate. The occasional application of some pure
tallow to the end of the tube and cork will be quite effectual
in preventing any of the fluid from
running upwards by capillary at-
traction.
For everyday use in the labo-
ratory, as well as for very accurate
determinations, this burette will
find favour.
The Pipette. The pipettes used
in volumetric analysis are of two
kinds, viz, those which deliver
one certain quantity ,only, and
those which are graduated so as to
deliver various quantities at the
discretion of the analyst. In the
former kind, or whole pipette, the
graduation may be of three kinds ;
namely, 1st, in which the fluid is
suffered to run out by its own
momentum only. 2nd, in which
it is blown out by the breath.
3rd, in which it is allowed to run
out to a definite mark. Of these
methods the last is preferable in
point of accuracy, and should
therefore be adopted if possible.
The next best form is that in
which the liquid flows out by its
own momentum, but in this case
the last few drops empty themselves very slowly ; but if
the lower end of the pipette be touched against the beaker
or other vessel into which the fluid is poured, the flow is
hastened considerably, and in graduating the pipette it is
preferable to do it on this plan.
In both the whole and graduated pipettes the upper
COLOEIMETRIC ANALYSIS. 307
end is narrowed to about J inch, so that the pressure of
the moistened finger is sufficient to arrest the flow at any
point.
Fig. 88 shows two whole pipettes, one of small and the
other of large capacity, and also a graduated pipette of
medium size.
The Measuring Flasks. These indispensable instru-
ments are made of various capacities ; they serve to mix
up standard solutions to a given volume, and also for the
subdivision of the substance to be tested by means of the
pipettes, and are in many ways most convenient. They
should be tolerably wide at the mouth, and have a well -
ground glass stopper, and the graduation line should fall
just below the middle of the neck, so as to allow room for
vshaking up the fluid.
Colorimetric Analysis is also used in assaying. It is
based upon the fact that a coloured solution appears the
more intense the more of the colouring substance it con-
tains.
If, therefore, a solution containing a certain amount of
a substance, and being in consequence of a certain intensity
of colour, is prepared, it will be possible to obtain the solu-
tion under assay of an equal intensity of colour by appro-
priate dilution.
By measuring the volume of the assay solution and
taking into consideration the amount of the standard solu-
tion, the quantity of the substance contained in the assay
solution may be readily calculated.
308
CHAPTEE IX.
THE ASSAY OF IRON.
THE ores of iron, properly so called, always contain the
metal in the oxidised state, and in various degrees of
purity.
The oxides and carbonates are the only minerals of
iron which can be used as ores in the blast-furnace. These
are associated with different impurities or foreign materials
in greater or less proportion. The following is a list of the
principal ores of iron, with the maximum percentage of
metallic iron that could occur in each if it were absolutely
pure, as in its formula :
MAGNETIC IRON ORE, Fe 3 4 =Fe 2 3 + FeO, 72-41.
RED HEMATITE and specular ore (anhydrous iron, ses-
quioxide or ferric oxide), Fe 2 3 , 70-00.
BROWN HEMATITE, limonite (brown iron ore), 2Fe 2 3 ,3H 2 0,
59-92.
SPATHIC IRON ORE (iron carbonate), FeO,C0 2 , 48-22.
TITANIFEROUS ORE (ilmenite), FeO,Ti0 2 , + nFe 2 3 .
FRANKLINITE, 3(FeO,ZnO,MnO) + (Fe 2 3 ,Mn 2 3 ), 45-16.
Besides these may be mentioned the varieties of impure
iron carbonates, known as clay or clay-band ironstone and
black-band ironstone. Clay-band ironstone sometimes re-
sembles compact limestone, sometimes greyish hardened
clay. Its great specific gravity, its effervescing on the addi-
tion of an acid, and acquiring a brown-red colour on roast-
ing, are sufficient means of identifying it.
The following is the result of an analysis of this class
of ore by the author ; the specimen was from Ireland,
county Leitrim :
THE ASSAY OF IRON. 309
Ferrous oxide . 51-653
Ferric oxide 3'742
Manganese oxide -976
Alumina ....... 1-849
Magnesia -284
Lime -410
Potash -274
Soda -372
Sulphur . -214
Phosphoric acid -284
Carbonic acid . 31-142
Silica 6-640
Carbonaceous matter and loss . . . 2-160
100-000
Black-band is a combustible schistose variety of this ore.
The following analysis is also by the author :
Ferrous oxide ....... 20-924
Ferric oxide ....... -741
Manganese oxide . . . . . . 1-742
Alumina ........ 14*974
Magnesia . . . . . . -987
Lime . . . . . . . -881
Phosphoric acid ...... '114
Silica ..... "'"-. . . 26-179
Sulphur ........ -098
Carbonic acid . ..... . 14-000
Carbonaceous matter ..... 16*940
Water and loss 2-420
100-000
Besides these iron ores the following substances, con-
taining iron and used as fluxes, require assaying : Granite,
Chlorite, Basalt, Pyroxene; Amphibole, and also some kind
of slags (finery cinder, tap cinder, &c.)
A. THE ASSAY OF IRON IN THE DRY WAY.
Iron ores very seldom occur in a pure state, and the
ores may be arranged for their assay in the dry way (and
also for smelting) into five classes.
1. Iron ores containing silica, lime, and another base,
which ores are fusible per se.
2. Iron ores containing predominantly silica.
3. lime.
3. alumina.
310 THE ASSAY OF IRON IN THE DRY WAY.
5. Iron ores containing a large amount of magnesia ;.
these ores are most difficultly rendered fluid.
The flux used for assaying (and also melting) varies
according to the nature of the predominant compound, and
the quantity used according to the amount of that com-
pound.
If the composition of the ore is known, it is easy to
ascertain the amount of flux necessary to form a slag with
the bases or silica present ; in most cases an extra quantity
of the flux should be added, in order to produce a sufficient
volume of slag to cover the button.
According to Dr. Percy,* blast-furnace cinder of the
following formula may be taken as a type of the kind of
slag desirable :
Al 2 3 ,Si0 3 2(3CaO,Si0 3 ).
Its approximate composition per cent, is as under :
Silica ..... 38"| T2 parts
Alumina . . . 15 > or about < 1 part
Lime .... 47; L3 parts
The following mixtures of various fluxes, when fused,,
produce a slag which may be regarded as approximating
to the above composition :
Quartz
1
1
^|
T36-5 per cent.
China Clay
".
2
f Silica .
\Alumina
0-92 /
0-82
0-82 f
J
= < 15*5
Lime .
.
2
.
.
2-5J
148-0 '
Glass .
2*
I Silica .
( Materials = Alumina
1-75^
tO-75 \-
r35
= 4 16
Lime .
,
2^
2-5 J
L50
Shale or fireclay
Lime .
3
f Silica .
\Alumina . .
wn
0-9 ^>
2-5 J
f35
- J 17
Us
>
According to Bodemann a compound of 56 % silica,
30 % lime, and 14 % alumina forms a slag most easily ren-
dered fluid, but as it is found that this slag itself is not
sufficiently fusible in a small assaying furnace (air-furnace),
an addition of fluor-spar is made to the mixture, and in
some cases (the iron ore being very difficultly rendered
* Percy's ' Metallurgy,' p. 240.
t 30 % say of alkalies, liine, &c., on account of their fusibility, are taken as?
equivalent to so much alumina.
THE ASSAY OF IKO^ IN THE DEY WAY. 311
fluid) some borax is added, or a mixture of borax and fluor-
spar.
An exact knowledge of the mineralogical properties
of the iron ores, and a due experience, will enable the
assay er to properly adjust the fluxes without resorting to
an analysis to find what amount of silica and bases are
present.
In some iron-works of Germany the following propor-
tions of fluxes are used :
For magnetic ore, red hematite ~) r on n/ t, n r /a
(very rich) ' ' c " a ^ anc * 25 %fluor-spar
argillaceous brown iron ore 20 to 40 30 to 40
bog iron ore ... 50 50
spathose iron ore . . 10 to 15 20 to 25
finery cinder . . . 20 to 25 20 to 25
Air-furnaces are best adapted for assaying iron ores
where many assays are required.
The furnace should have a cross section of 18 in. x 18
in., and depth to grate-bars 21 inches. Flue 7 in. x 7 in.
Anthracite is the best fuel to use ; then coke.
In the assay of an iron ore it is required to reduce the
oxide of iron to the metallic state, or rather to that of cast
iron, to collect it in a button, and to form with the foreign
materials of the ore by means of fluxes a fusible slag;
that will not retain any of the iron in combination or in
the form of pellets. ;
Naked crucibles, either* of clay or black-lead, or cru-
cibles lined with charcoal, are employed ; the latter are
preferable. The button of metal does ,not adhere to naked
pots, but the slag adheres very strongly ; so much so that
it cannot be detached with any degree. of accuracy for
weighing (which in some of M. Berthier's processes is of
importance). Black-lead pots allow neither the slag nor
button to adhere, but the former dissolves much argilla-
ceous matter from the pot, so that its weight is greatly"
increased, and the assay cannot be verified. In naked
crucibles charcoal must always be added to the assay, to
reduce the iron oxide ; in which case, if an excess be.added,
it prevents the button from completely forming, so that
312 THE ASSAY OF IRON IN THE DRY WAY.
globules remain in the slag (with care this may, however,
be avoided). Neither do naked crucibles resist the fire as
well as those lined with charcoal, because the lining sup-
ports the sides when they soften. The charcoal lining also
allows the assay to be finished without adding any reagent
to the ore ; the button can be readily taken out, because
it does not adhere to the charcoal ; and lastly, the earthy
matters in the ore, which have formed a slag, may be col-
lected and weighed. If we have added any flux to the ore,
the total weight can also be ascertained.
The method of lining crucibles with charcoal brasque
and conducting the assay is given as follows by Mr.
Blossom : *
The brasque has a composition of four parts of finely
pulverised charcoal to one part of molasses. This must
be thoroughly kneaded until a ball of it made in the hands
resists, to a sensible degree, an attempt to pull it apart.
The crucibles are packed full by driving the brasque in
with a mallet ; a conical cavity, of sufficient size for the
charge, is then cut out, and the brasque dried in an oven.
Care must be taken not to burn the molasses, for the
brasque would in that case crumble, and be useless.
These are the best crucibles for iron assays, because
they combine the following advantages :
Being lined with charcoal, none need be mixed with
the charge to reduce the oxide, whereas in naked crucibles
charcoal must be added, and is liable to prevent the com-
plete collection of the iron in a button by holding little
pellets in suspension.
The slag neither adheres to a charcoal lining nor takes
up any material from it, while it does adhere to ordinary
naked crucibles, and dissolves argillaceous matter from
black-lead crucibles. In the former case, then, the slag
may be weighed as a verification of the assay, while in the
latter this is, of course, impossible.
The lining serves as a support for the crucible, which,
under the high heat employed, is very apt to be softened
and crushed beneath the weight of the fuel.
* ' Chemical News,' April 5, 1872.
THE ASSAY OF IRON IN THE DRY WAY. 313
The Charge.
In making up the charge it is only necessary to con-
sider the materials required as fluxes for the foreign
matters of the ores. It may be well to sprinkle a little
charcoal into the charge as a precaution, but none is abso-
lutely required. Two cases may arise, in which we have
(1) ores of unknown composition, and (2) ores previously
analysed. The assay in both cases gives us a clue to the
nature of the iron that may be obtained from the ore, and
to the character and proportion of the fluxes to be added in
the blast-furnace, in order that we may produce a fusible
slag free from iron. In the former case we obtain the
additional information of the approximate percentage of
iron, though the iron assay is seldom, if ever, made for
this purpose. Eecourse is had to the more accurate
chemical analysis, which gives us the exact proportions of
the substances which affect the iron injuriously or other-
wise. In all the assays a constant weight of ore, 300
grains, is taken.
1. Ores of Unknown Composition.
In the assay of an ore the composition of which is
unknown, we employ one or more preliminary assays,
and, if satisfactory results be not obtained from either,
we make another assay with a charge modified according
to the indications of the preliminary assay. The follow-
ing charges may be used to advantage in the preliminary
assay :
Preliminary Assay Charges.
i. Ii. m. iv.
Silica .... 75 30 120 75 grains.
Lime .... 75 120 45 75
Ore . . . . 300 300 300 300
1. The first charge is employed for the purer ores, those
containing very little gangue such as some varieties of
magnetic ore, red and brown hematites, specular and
micaceous ores.
314 THE ASSAY OF IRON IN THE DRY WAY.
2. Ores containing silica ; some varieties of brown
hematite, magnetic ore, &c.
3. Ores containing lime, magnesia, or protoxide of man-
ganese carbonates, &c. ; calcareous hematites, spathic iron.
4. Ores containing silica and alumina ; clay ironstones,
black-band, &c.
The principle involved in all the charges is that of
furnishing for a base, lime ; for an acid, silica ; and vice
versa.
The choice of a charge, therefore, depends on the
acid or basic nature of the gangue of the ore. The mate-
rials of the gangue might possibly be associated in such
proportions as to flux themselves, but this would happen
rarely.
Ores containing titanium require the addition of fluor-
spar to the charge, in quantity varying from 15 to 300
grains, according to the amount of titanium present.
2. Ores previously Analysed.
When we know the percentage composition of an ore r
it is a very simple matter to calculate a charge for the dry
assay. Good results are obtained from a charge so propor-
tioned as to yield a slag corresponding to the following
formula of a blast-furnace cinder, as given by Percy :
H 2 3 represents alumina, and BD, lime, magnesia, and other
bases. Its approximate percentage composition is as fol-
lows :
Silica ...... 38 "| r 2^ parts
E 2 O 3 (alumina) . . . . 15 ^> or about < 1 part
KO (lime, magnesia, &c.) . . . 47 J [9 parts
We have, then, only to establish the latter relation
between the component materials of the gangue, to obtain,.
on fusion, the above slag.
Let us take the following example :
300 grains Ore Difference
The Ore contains Per cent. contain Required to be added
Silica .:....' 165 - 4'95 25' 20-05 grains.
Alumina . . 1-94 5-82 10- 4-18
Lime, MgO, &c. . 4'51 13-53 30- 6-47
THE ASSAY OF IKON IN THE DRY WAY. 315
Silica is supplied by ground quartz. For the bases
EO represented in the furnace slag and in the ore by lime,
magnesia, &c., we add pure unslaked lime. The alumina
is added in the form of kaolin, which may be assumed to
contain equal parts of alumina and silica. Allowance must
be made in adding silica for that introduced with the
kaolin.
It happens sometimes that the ore contains more than
is required of one of the ingredients of the slags, or the
silica introduced with the kaolin may, when added to that
already present, increase the quantity beyond the re-
quirement. In either case make up a new slag with the
excess.
The charge having been weighed out, must be tho-
roughly mixed on glazed paper ; after placing it in the
crucible, the conical cavity is closed with a piece of char-
coal, and the whole top of the crucible is covered with a
luting of fireclay. The latter is mixed with ^ ^ part
fine sand, and is made plastic with borax water. Hair is
sometimes employed to prevent the luting from cracking
off when dry ; but no trouble is experienced from this
source if the luting be properly made and applied. It
should not be put on too thick, should be lapped over the
edges of the crucible, and thoroughly dried before placing-
the crucible in the furnace. ..
Four crucibles are introduced at one time, and rest
upon two fire-bricks placed one upon the other, to keep the
crucibles in the very midst of the glowing coals. If the
crucibles do not rest steadily on the bricks, it is well to
support them with a little luting, to prevent their being
knocked over in the fire. A low fire may be kindled before,
the introduction of the crucibles, or it may be kindled
around them. The fuel is added gradually until it fills
the furnace above the tops of the crucibles ; the fire is
then maintained at its maximum temperature for 2-J 3-|
hours, according to the refractory nature of the ore. Ores
containing much titanium may even require 4 hours,
while carbonates containing manganese may fuse well in 2-|
hours, or in less time. Three hours will generally be suifi-
316 THE ASSAY OF IRON IN THE DRY WAY.
cient for ores that do not contain much titanium. When
the fire has burned out, the bricks and crucibles are
removed in one mass, cemented together by the slag of
the fuel. The crucibles are detached, and the exteriors
broken with a hammer ; on inverting and tapping the
brasque lining, the slag and the button of cast iron will
fall into the hand, when, if they adhere together, a slight
tap will suffice to separate them. Before separation, how-
ever, they ' should be carefully cleansed and weighed ; if
necessary, the slag may then be broken, and any particles
of iron it retains mechanically may be extracted with a
magnet. The weight of the iron being deducted from the
weight of the slag and button, we obtain the weight of the
slag. This ought to approximate closely to the weight of
the fluxes introduced and the corresponding material of
the ore. If a large amount of iron has combined with the
slag it will be indicated by the excess in weight. Titanium
and manganese enter the slag almost completely ; hence
if they are present, allowance must be made for them.
Duplicate assays are made, and the two results should
not differ more than 0'3 0-4 of one per cent. The slag
ought to be well fused, colourless, transparent, and vitreous,
or white, light-grey, bluish-grey , opaque, and semi-vitreous,
resembling porcelain or enamel.
A good button will be well formed, and will separate
completely from the slag.
If the metal be of good quality, the button, when.
wrapped in a piece of thin tin-plate, and struck on the
anvil, will flatten slightly before breaking. It ought to be
grey or greyish-white, and the grain fine, or tolerably fine.
A button of bad iron breaks readily without changing
form, sometimes even pulverising : the metal is generally
white and crystalline on the broken surface.
The following are some of the characters that may be
observed in slags, and their indications with reference to
the charge : A perfectly transparent slag of greenish tint
indicates an excess of silica. A stony rough slag, or one
that is crystalline in fracture and dull in lustre, indicates
an excess of bases.
ASSAY OF IRON IN THE DRY WAY. 317
If the product, instead of being melted, is only fritted,
and contains the reduced iron interspersed as a fine grey
powder, both silica and alumina are deficient in the flux,
lime and magnesia being in excess. The latter is one of
the most refractory substances found in iron ores, and,
when present in quantity, requires an addition of both
silica and lime. A vesicular slag, with the iron interspersed
in malleable scales, indicates the presence in the ore, of iron
and manganese silicates, or an excess of silica, which react
on the carburetted iron as it forms, producing malleable
iron and carbonic acid. This trouble is corrected by the
addition of lime.
Manganese in small quantity gives an amethystine tint
to the slag ; in larger proportion it produces a yellow,
green, or brown colour.
Titanium often produces a resinous, black, and scoria-
ceous slag, sometimes curiously wrinkled on the outside.
It is covered on the outside with a metallic pellicle of the
cyano-nitride of titanium with its characteristic copper
colour ; sometimes the slag is vitreous and of a bluish tint.
Chromium gives a dark resinous slag, sometimes surrounded
by a thin metallic coating.
The following are some characters of the button depend-
ent on the substances named :
Phosphorus. A hard, brittle, white metal, known as
cold-short.
Sulphur. A strong, reticulated, mottled structure, and
red-short iron.
Manganese. A button smooth exteriorly, hard and
non-graphitic : it breaks under the hammer, and presents a
white crystalline fracture.
Titanium. The button is smooth on the outside, and
breaks under the hammer with a deep grey fracture, dull
or crystalline. It adheres strongly to the slag. The button
is covered sometimes with titanium cyano-nitride with its
characteristic copper colour. Titanium is said to increase
the strength of the metal. It may be present to the extent
of one per cent.
Chromium. Sometimes the button is smooth, well
318 THE ASSAY OF IRON IN THE DRY WAY.
fused, with a brilliant crystalline fracture and tin- white
colour ; at other times it is white, only half fused, or it
may even form a spongy mass of a clear grey colour,
according to the quantity of chromium contained in the
iron. Many alloys of iron and chromium will scratch
glass.
Berthier recommends the following method for esti-
mating the other, chiefly slag-forming, components of iron
ores. The operations of this method are comprised in
roasting or calcining, to drive off any volatile or combust-
ible matters, and in treating the ore with certain acids,
the object of which is to ascertain the amount of insoluble
matter, by difference of weight, before and after the action
has taken place.
The hydrated ores are calcined to estimate water ;
and those containing manganese, to reduce it to a fixed and
known state of oxidation (sesquioxide). The carbonates
are roasted to expel carbonic acid, and the ores from the
coal formations to burn the combustible matter with which
they are mixed.
Slags and dross are also roasted to free them from
charcoal. A simple calcination sometimes is sufficient, as in
the case of carbonates ; but where mixtures of ferric and
ferrous oxides are to be assayed, they must be subjected
to a long roasting, in order to convert all the contained
ferrous oxide into ferric oxide.
Diluted and cold nitric or acetic acid is employed
for minerals whose matrix is purely calcareous or magne-
sian, as these acids dissolve the earthy carbonates, without
attacking either stones, clay, or the iron oxides. The
residue is to be well washed, dried, and weighed, and the
amount of carbonates calculated by the difference. It is
now to be treated with boiling hydrochloric acid, or, what
is preferable, aqua regia. The ores which contain sub-
stances insoluble in these acids are generally of a clayey or
flinty nature. These are to be weighed, and according to
their weight that of the flux to be added in the assay is
estimated, as will be shown hereafter.
It must be borne in mind, however, that the clays
THE ASSAY OF IRON IN THE DRY WAY. 319
are not absolutely insoluble in hydrochloric acid, for a
certain quantity of alumina is always dissolved, which is
generally greater in proportion to the amount present in
the iron ore.
The ores containing titanium are boiled with concen-
trated sulphuric acid, after they have been reduced to the
finest possible state of division. All the iron, titanium, and
manganese oxides are dissolved, and the stony gangues
which resist the action of this acid can be estimated.
The utility of this estimation will be pointed out as we
proceed.
When all the operations necessary for each particular
case have been completed, we know the proportion of vola-
tile substances, of substances soluble in acetic acid, and
those insoluble in hydrochloric and sulphuric acids, con-
tained in the substance under assay.
Let A be the weight of the rough or non-calcined ore ;
B the weight of the same calcined ; C the weight of the
fluxes in a rough state ; D the weight of the same calcined ;
P the weight of matter insoluble in hydrochloric or sul-
phuric acid ; E the weight of the fixed substances soluble
in acetic or nitric acid a weight which can be readily
calculated when we know the loss which the ore, not
treated by acids, suffers by calcination, and the residue of
the treatment of this substance by acetic or nitric acid ; M
the weight of the button of metal and scattered globules ;
S the weight of the slag ; and the loss of weight in the
assay, which represents the quantity of oxygen disengaged
during the reduction.
The following is the disposition of the data from which,
at one view, all the useful results of the assay can be de-
termined.
In the assay has been employed
A, rough ore = calcined ore . . . . B
B, of rough fluxes added = fixed flux . . D
Total of fixed matter B + D
The result has been
320 THE ASSAY OF IRON IX THE DRY WAY.
Metal M\ T , i M
Slag-S / T tal M +
Loss
Fluxes D
Verifiable matters S D
Substances insoluble in hydrochloric acid, &c. . T
Substances soluble in hydrochloric acid, &c. . S D T
Substances soluble in acetic acid ... R,
Substances insoluble in acetic acid, and solu-
ble in hydrochloric acid .... S D T R
When the iron in the substance assayed is in a known
degree of oxidation, and when but little manganese is
present, the quantity of oxygen ought to correspond
very nearly with the quantity of metal M produced : if it
does, the assay must be correct.
A rigorous correspondence between the two numbers,
however, cannot always be obtained, because the iron is
not pure, but always contains carbon, so that in ordinary
assays the ferric oxide loses but from twenty-eight to
twenty-nine per cent, of oxygen.
On the other hand, the quantity of iron remaining in
the slag makes up in part for the carbon combined with
the metal reduced ; but when the assay has been made
with a suitable flux, the quantity of oxide remaining is
very small, and never exceeds one per cent, of the weight
of the slag. When the iron is in an unknown degree
of oxidation, the loss produced in the assay gives the
degree, if it has been made without accident ; but if there
is any doubt, and the result is of importance, the assay
must be recommenced for verification. If the ferruginous
matter contain manganese, and if that metal be in the
state of protoxide, the verification just described can be
made without modification, because the manganese dis-
solved in the slag is always at the minimum of oxidation ;
and when a sufficient quantity of flux is employed, the
amount reduced is of no consequence. But when the
manganese is in the state of red oxide, it parts with a
certain quantity of oxygen on being reduced to the
minimum of oxidation f quantity estimated in the loss
0), so that a perfectly accurate verification cannot
be made. Nevertheless, the difference between the loss
THE ASSAY OF IKON IN THE WET WAY. 821
0, and the quantity of oxygen calculated from the metal
M, cannot be very great, because the red oxide of man-
ganese loses but *068 of oxygen in its transformation to
protoxide.
If the assay has been made with care, the loss of
oxygen indicates the amount of iron in a very approxi-
mate manner, and nearly always with an exactitude
which is surprising to those not accustomed to this kind
of operation.
Titanic acid behaves in iron assays exactly as the
oxides of manganese ; it disengages at most but -06 of
oxygen when dissolved in the earthy glasses in contact
with charcoal.
B. THE ASSAY OF IRON AND ITS OEES IN THE WET WAY.
This will be subdivided into the following sections :
a. Assay of iron ores, pig iron, and steel, for the
metallic iron they contain.
$. Complete assay of iron ores.
7. Estimation of carbon, sulphur, phosphorus, silicon,
&c., or metallic iron and steel.
a. Dr. Penny's Process. The following method of
estimating the amount of iron in a sample by means of a
normal solution has been contrived by Dr. F. Penny, who
was led to substitute potassium bichromate for potassium
permanganate, as recommended by Marguerite. The
reason of employing the bichromate is that it is an
unchangeable salt, whilst the permanganate sometimes
undergoes decomposition, so that its strength is variable,
and each series of experiments made with it requires a
separate verification by means of a weighed quantity of
pure iron. This inconvenience is avoided in Dr. Penny's
method, which is described in his own words as under :
. ' I shall proceed to describe the method of employ-
ing the potassium bichromate for the determination of the
amount of iron in clay-band and black-band ironstone. I
shall be purposely minute, as I particularly desire that
the process may be serviceable to those who, from their
Y
322 THE ASSAY OF IRON IN THE WET WAY.
pursuits in life, are interested in the value and quality of
ironstone, and who may be imperfectly acquainted with
analytical operations.
4 A convenient quantity of the specimen is reduced to
coarse powder, and one-half at least of this still further
pulverised, until it is no longer gritty between the fingers.
The test solution of potassium bichromate is next prepared.
44*4 grains of the salt in fine powder are weighed out, and
put into an alkalimeter (graduated into 100 equal divisions),
and tepid distilled water afterwards poured in until the
instrument is filled to 0. The palm of the hand is then
securely placed on the top, and the contents agitated by
repeatedly inverting the instrument, until the salt is dis-
solved and the solution rendered of uniform density
throughout. It is obvious that each division of the
solution thus prepared contains 0-444 grain of bichromate,
which corresponds to ^ a grain of metallic iron. The
potassium bichromate used for this process must of course
be purchased pure, or made so by repeated crystallisation,
and it should be thoroughly dried by being heated to
incipient fusion.
' 100 grains of the pulverised ironstone are now intro-
duced into a Florence flask, with 1^ oz. by measure of
strong hydrochloric acid, and \ an ounce of distilled
water. Heat is cautiously applied, and the mixture occa-
sionally agitated, until the effervescence caused by the
escape of the carbonic acid ceases ; the heat is then
increased, and the mixture made to boil, and kept at
moderate ebullition for ten minutes or a quarter of an
hour. During these operations it will be advisable to
incline the flask, in order to avoid the projection, and
consequent loss, of any portion of the liquid by spirting.
About 6 oz. of water are next added, and mixed with the
contents of the flask, and the whole rapidly transferred to
an evaporating basin. The flask is rinsed several times
with water, to remove all adhering solution.
'.Several small portions of a weak solution of pure red
potassium prussiate (containing one part of the salt to 40
of water) are now dropped upon a white porcelain slab,
DE. PENNY'S PROCESS. 323
which is conveniently placed for testing the solution in
the basin during the next operation.
' The prepared solution of potassium bichromate in the
alkalimeter is then added very cautiously to the solution of
iron, which must be repeatedly stirred, and as soon as it
assumes a dark greenish shade it should be occasionally
tested with the red potassium prussiate. This may be
-easily done by taking out a small quantity on the top of a
glass rod, and mixing it with a drop of the solution on a
porcelain slab. When it is noticed that the last drop
communicates a distinct red tinge the operation is ter-
minated. The alkalimeter is allowed to drain for a few
minutes, and the number of divisions in the test-liquor
consumed read off. This number multiplied by two gives
the amount of iron per cent, in the specimen of ironstone,
assuming that, as directed, 100 grains have been used for
the experiment. The necessary calculation for ascertain-
ing the corresponding quantity of protoxide is obvious.
' When black-band ironstone is the subject of ana-
lysis, or when the ore affords indication, by its appear-
ance or during the treatment with hydrochloric acid, that
it contains an appreciable quantity of carbonaceous
matter, then the hydrochloric acid solution must be
filtered before being transferred to the basin, and the
filter, with the insoluble ingredients, must be washed in
the usual way with warm distilled water, slightly acidu-
lated with hydrochloric acid until the filtrate ceases to
give a blue colour with the red potassium prussiate.
In those cases, also, where the presence of iron pyrites in
the ironstone is suspected, it will be necessary to remove
the insoluble matter by filtering before applying the
bichromate solution ; but with ironstones in which the
insoluble ingredients are merely clay and silica, filtration
is not essential.
6 Now it is evident" that the foregoing process, so far as
I have described it, serves for the determination of that
portion of iron only which exists in the ore in the state
of protoxide. But many specimens of the common iron-
stone of this country contain appreciable quantities of
T 2
324 THE ASSAY OF IRON IN THE WET WAY.
ferric oxide, which, being unacted upon by the potas-
sium bichromate, would escape estimation by the present
method. By an additional and easy operation, however,
the amount of metallic iron in the ingredient may be
likewise determined. It is only necessary to reduce it
to the minimum state of oxidation and then to add the
bichromate as previously directed.
' The best and most convenient agent for effecting the
reduction of the ferric oxide is sodium sulphite. The
only precaution to be observed is to use it in sufficient
quantity, and at the same time to take care that the iron
solution contains excess of acid. When the reduction is
complete, a few minutes' ebullition suffices to decompose
the excess of sodium sulphite, and effectually to expel
every trace of sulphurous acid.
' In order to test the exactness of this mode of
estimating the iron of the peroxide, I made several ex-
periments with peroxide prepared from known quantities
of pure iron wire. The peroxide was thoroughly washed,
dissolved in hydrochloric acid, reduced with sulphite
of soda, and after complete expulsion of the excess of
sulphurous acid, the solution was diluted with water and
treated with potassium bichromate. I select three of the
experiments :
' Exp. I. 10 grs. of iron consumed 8-87 of bichromate.
II. 18 15-94
III. 25 22-15
; The mean of all my experiments on this point gives
the ratio of 100 of iron to 88'6 of bichromate.
' Whenever, therefore, the ore of iron contains ferric
oxide it will be necessary to add sodium sulphite to the
hydrochloric acid solution before the addition of the test-
liquor from the alkalimeter. The sulphite should be dis-
solved in distilled water, and added to the solution of iron
in small successive portions, until a drop of the liquor
gives merely a rose-pink colour with potassium sulpho-
cyanide, which indicates that the reduction of the ferric
salt is sufficiently perfect. The liquor is now heated till
the odour of sulphurous acid is no longer perceptible.
DR. PENNY S PROCESS. 325
These operations should be performed while the solution
is in the flask, and before it is filtered or transmitted to
the basin.
' I will here mention, for the guidance of those who
may not be fully aware of the reactions of the oxides of
iron, that the existence of an appreciable quantity of
peroxide in the ironstone may be readily discovered by
dissolving (as directed in the process) 39 or 40 grs. of the
ore in hydrochloric acid, diluting with about 8 oz. of
water, filtering, and testing a portion of the solution with
potassium sulphocyanide. If a decided dark blood-red
colour is produced, the quantity of ferric oxide in the
stone must be determined ; but if the colour is only light
red or rose-pink, the proportion is exceedingly small, and
for practical purposes not worth estimating. Of course,
when the specimen of ironstone has an ochrey or a reddish
appearance on the surface or in the fracture, the presence
of a large proportion of ferric oxide is indicated, and its
exact quantity must be determined.'
The details of Dr. Penny's process have been carefully
examined by Mr. R. W. Atkinson, with the result of elimi-
nating several small sources of error which interfered
with accuracy. In all cases where volumetric methods
are employed, the first and most important point to be
considered is how to obtain the standard solution invari-
ably of an accurately known strength, and it is from a
variation in the methods employed for this end that a large
amount of the variation in the results is due. Presuming
that potassium bichromate is the salt almost universally
employed in England, it seems simple enough to weigh out
the exact amount of salt and dissolve it in a known volume
of pure distilled water. Dr. Penny recommends the fusion
of the bichromate for the purpose of driving off com-
pletely the water entangled in the crystals ; but if this be
done, however carefully the heat be regulated, on dis-
solving in water, and allowing to settle, a green deposit of
chromic oxide will be found at the bottom of the flask or
bottle, showing that a small amount of decomposition
takes place during fusion, and that the value of the stan-
326 THE ASSAY OF IKON IN THE WET WAY.
dard must be lower than the calculated value. Instead of
fusing the salt, it is much better to grind up the very
purest crystals and to dry them in a steam oven for
several hours before weighing out. But, however the
solution be made up, it is always safer to standardise it
before use.
The great difficulty in the use of salts of iron for stan-
dardising is to obtain one which will not alter in composi-
tion by keeping for two or three months. Ferrous sulphate
is out of the question ; and most specimens of the iron and
ammonium sulphate are liable to oxidation, though in a
less degree than the ferrous sulphate. We have been for-
tunate enough to get from Messrs. Hopkin and Williams
a sample of granulated iron and ammonium sulphate,
which after several months' use still gives the theoretical
percentage of iron, ammonia, and sulphuric acid. Pre-
suming that the strength of the bichromate solution in
terms of iron is accurately known, a fair sample of the
ore having been ground sufficiently fine to pass com-
pletely through a sieve of 120 meshes to the linear inch,
part of it is dried at 212, and when cold portions of about
fifteen or sixteen grains are weighed out for analysis.
Unless the balance is very rapid in its action, it will be
found necessary to re-dry the portions first weighed out
if accurate and concordant results are to be expected..
The ore is in such a very fine state of division that it
greedily absorbs moisture from the air during the opera-
tion of weighing out : this applies most forcibly to the
soft red ores, like Campanil and Vena Dulce, but it is also
true of the harder Eubio and other brown ores. The
weight of the portion being accurately known, it is trans-
ferred to a conical flask, and digested at a gentle heat on
a hot plate, with from 150 to 230 grains strong hydro-
chloric acid, the flask being closed with a watch-glass.
It has been asserted by K. F. Fohr that ferric chloride
is volatile at about 212, but this is contrary to experience.
It is true that yellow drops of ferric chloride may some-
times be seen depending from the under surface of the
watch-glass, but this is only the case when the solution is
REDUCTION BY ZINC. 327
accompanied by a boiling of the liquid, and is doubtless
due to spirting. If the digestion is carried on so as to
avoid bubbling, the drops on the cover are always colour-
less, and no loss of iron from volatilisation need be feared.
When the ore has been completely dissolved, the next
step is the reduction from the ferric to the ferrous state,
to effect which there are three reducing agents mainly em-
ployed, viz. zinc, stannous chloride, and sulphurous acid
in the form of one of its salts. These will be dealt with in
the order given.
Eeduction by means of zinc is capable of giving
trustworthy results, provided that pure zinc free from
iron is employed ; or if the zinc contains iron, provided
that the amount thus introduced be allowed for. But the
use of zinc is open to other objections : in the first place,
it dissolves only slowly, and thus unduly retards the ope-
ration, which, when a number of analyses are to be carried
out, is a matter of no small moment. In the second place,
the titration is further retarded by the slowness with
which the blue colour with potassium ferricyanide is deve-
loped, in consequence of the presence of zinc chloride in
the solution. And lastly, the colour towards the end of
the titration becomes so faint, even when fully developed,
that it is impossible to distinguish the presence of an
amount of iron less than one or two tenths per cent, of
the iron contained in the ore. Consequently, although
reduction by means of zinc permits the analyst to obtain
uniformly concordant results, without the risk of error, its
action is too slow, and its indications are not sufficiently
delicate for the most accurate work.
But, however much rapidity of work may be an object
to be kept in view, there can be no doubt but that reduc-
tion by zinc is greatly to be preferred to the second
method, viz. reduction by means of stannous chloride.
Indeed, the latter method, as usually performed, is open
to the grossest abuses, and ought to be prohibited by any
authority which may seek to reform the present methods
of analysis.
One reason why stannous chloride has become a favour-
3-28 THE ASSAY OF IRON IN THE WET WAY.
ite reducing agent is the rapidity with which the analysis
can be made. Starting with a sample of ore the percentage
of moisture and iron in the ore can be found by this
method (with the above qualifications as to accuracy) in
from two to three hours ; but it is contrary to the best
interests of a chemist to seek rapidity of work at the
expense of accuracy, and the abandonment of this method
of reduction is strongly recommended.
Of all methods, reduction of the ferric salt by the use
of a concentrated solution of ammonium bisulphate is the
most accurate and trustworthy. Sodium bisulphite is some-
times used, but is not nearly so satisfactory as the am-
monium salt, as it is more difficult to separate the last traces
of sulphurous acid from the former than from the latter.
The mode of manipulation is as follows : The ore having
been dissolved in hydrochloric acid in the conical vessel,
the solution is diluted with acidified water and filtered
into pear-shaped flasks, the filters being thoroughly washed
with hot acid water. The filtered ferric chloride is next
carefully neutralised with ammonia, strong at first and
afterwards dilute, until a faint reddish precipitate remains
permanent. Two or three drops of strong hydrochloric
acid are washed round the inner neck of the flask, and as
the acid flows down it spreads out, dissolving any par-
ticles of ferric hydrate which may have remained on the
sides of the flask. When the solution is quite clear, and
of a faint reddish colour, 75 to 90 grains of a strong solu-
tion of ammonium bisulphite (sp. gr. 1-06) are added, the
flask shaken, and boiling water added. On shaking the
flask the colour entirely disappears, and the flask is then
put over a burner. A small piece of thick platinum
wire is introduced to assist the boiling, and about 25 or
30 grains of dilute sulphuric acid (1 acid to 6 water) are
added to acidify the solution and to assist in the expulsion
of the excess of sulphurous acid. After the liquid is once
in a state of ebullition it is kept boiling briskly for thirty
minutes (less time is sufficient, but it is always well to err
on the safe side), during which time nearly the required
amount of bichromate is run out into the dish. At the
REDUCTION BY SULPHUROUS ACID. .320
end of the half-hour the boiled solution is added to the
potassium bichromate, and the titration is carried out as
usual.
By proceeding as above there is only one loophole for
the introduction of error, viz. in the length of time allowed
for boiling off the sulphurous acid ; but if the conditions
as given above are fulfilled, constant and accurate results
may be relied upon. The above method has this advantage,
viz. that the solution is practically one of ammonio-ferrous
sulphate, a salt which is one of the most stable of all the
ferrous salts. It is, therefore, less liable to become oxi-
dised by exposure to the air in transferring to the basin
than the acid solution of ferrous chloride obtained by the
two previous modes of reduction. A further advantage
lies in the fact that the end reaction with potassium ferri-
cyanide is beautifully clear and delicate ; so that -there is no
difficulty in distinguishing the addition of ^ c.c. of bichro-
mate (strength 1 c.c. = O077 grain iron), equivalent to 0-0038
iron. Numerous experiments have shown that perfectly
constant results can be obtained by this method, the same
percentages of iron in a given ore having been found with
different standard bichromate solutions after the lapse of
several months. It is rarely the case that three experi-
ments carried out simultaneously give percentages of iron
differing by more than 0' 05 to 0-07 per cent, of the ore ;
but the main advantage which the method of reduction
by ammonium bisulphite possesses is that results are found
which are perfectly independent of any unconscious bias
on the part of the operator, and the author feels convinced
that were this method constantly and generally used, we
should hear less of differences of two per cent, and three
per cent, beween two chemists' analyses of the same sample
of ore. The existence of ' sellers ' and ' buyers ' chemists
is a disgrace to the profession, and anything which is likely
to put an end to the scandal, even in one trade, ought to
be welcomed.
One other point in the volumetric estimation of iron
remains to be noticed. The solution of potassium ferri-
cyanide slowly decomposes when the bottle containing it
330 TITRATION OF IRON
is exposed to diffused daylight, and a yellowish sediment
is deposited. This change is very greatly retarded, if not
entirely prevented, by protecting the solution from light
by covering the bottle with an inverted tin canister. Con-
nected with this decomposition of the ferricyanide solution
is the fact, already well known, that when a mixture of
that solution with one of ferric chloride is exposed to day-
light reduction takes place, and the solution turns blue.
But it is not so generally known how rapidly this takes
place, and that if the drops of ferricyanide to which the
completely oxidised iron solution has been added be allowed
to remain exposed to the light (protected from dust by
means of a glass plate) for ten or fifteen minutes, a distinct
blue tinge will be developed. It is important to remember
this, for in titrating the blue colour requires two or three
minutes to become fully developed when the amount of
ferrous salt remaining in the solution is very small, and in
order to prevent the drop turning blue by reduction under
the influence of daylight, it is advisable to keep the slab
covered with a black cloth or with a flat tin cover. By
this means it is possible so to protect the mixture from
change that no blueness is perceptible after the lapse of
an hour or more, provided that all the iron in the solution
to be titrated has been fully oxidised.
TITRATION OF IRON WITH SODIUM HYPOSULPHITE.
A. C. Oudemans, jun., has proposed to estimate iron
in the acid solution of the chloride to which a little solu-
tion of copper sulphate and potassium sulpho-cyanide has
been added by dropping in a solution of sodium hyposul-
phite of known strength until the red colour of the iron
sulpho-cyanide has disappeared, and estimating the excess
of hyposulphite by titrating back with iodine solution.
A. E. Haswell modifies this method so as to avoid the
possibly disturbing separation of copper sulpho-cyanide,
and to dispense with the back titration with iodine. Ac-
cording to his experiments, the iodine-starch reaction often
takes place too early, before all the hyposulphite has been
converted into sodium tetrathionate. He explains this
WITH SODIUM HYPOSULPHITE. 331
occurrence by the tendency of cupric iodide to split up
into cuprous iodide and free iodine, and thus produce a
premature blue coloration which after a time disappears
again as the cuprous iodide recornbines wdth the free iodine
to form the cupric iodide.
Haswell mixes the moderately acid solution of ferric
chloride in presence of a cupric salt with a few drops of
a dilute solution of sodium salicylate, and then reduces
with sodium hyposulphite. The deep violet colour of the
solution fades gradually and becomes colourless in presence
of a very slight excess of the reducing agent. The excess
of sodium hyposulphite is then oxidised with a dilute solu-
tion of sodium bichromate. The limit of the reduction is
sharply marked by the faint violet colour which indicates
the oxidation of a trace of the iron. It must be remem-
bered that strong hydrochloric acid destroys the colour
produced by salicylic acid in ferric chloride, which, how-
ever, is restored on moderate dilution with water.
For the execution of the method there are required a
solution of sodium hyposulphite, standardised by means of
a solution of ferric chloride of known strength ; a solution
of potassium bichromate about half the strength of the
sodium hyposulphite ; a solution of copper, prepared by
dissolving two grammes cupric ammonium chloride in
100 c.c. water ; and a solution of sodium salicylate con-
taining about 5 grammes of the salt per litre.
Five or ten c.c. of the iron solution are measured into
a small flask, slightly acidulated with hydrochloric acid,
and mixed with 1 to 2 c.c. of the copper solution, and a
few drops of the sodium salicylate. If the colour resulting
is not a pure violet, but an olive brown, the liquid is
diluted with water and the hyposulphite is added until the
liquid appears perfectly colourless on standing with the
back to the window and looking through the flask at a
sheet of white paper. It often happens that on adding
more sodium salicylate a faint coloration reappears, but
it is removed by a drop of .hyposulphite. It is then titrated
back with the bichromate until a faint viftlet coloration
appears.
332 COMPLETE ASSAY OF IRON ORES.
Complete Assay of Iron Ores. The methods employed
in the analysis of iron ores have been thoroughly investi-
gated by Mr. A. A. Blair. The following account of the
method he recommends is taken from ' Mining Industries
of the United States.' The treatment of samples of iron
ore naturally divides itself into two parts, the mechanical
and the chemical, and it will be described under these
heads. The care with which the identity of every sample
is preserved throughout will be shown, and the methods by
which the estimation of the different elements is rendered
as accurate as the state of chemical knowledge would allow
will be given in detail.
THE MECHANICAL TREATMENT.
The sample is first taken from the bag and placed
upon a large piece of clean strong paper, and the label
removed from the box and put in a little card-rack fixed
over the bench. Half a dozen small chips, representing,
as nearly as possible, the different varieties of the ore, are
put aside and carefully labelled, to be used for specific
gravity estimations and for thin sections for the micro-
scope.
The large steel mortar, fig. 89, is half filled with the ore,
the leather cover adjusted, and the machinery started-
This mortar was cast at the Mid vale Steel Works of Phila-
delphia, and was made of an exceptionally fine quality of
propeller steel, containing over one per cent, of carbon and
about fifteen-thousandths of one per cent, of phosphorus.
The pestle was forged and hardened, and the wear, after
crushing between 15,000 and 20,000 pounds of ore, was
scarcely perceptible in either mortar or pestle. The mortar
weighs about 70 pounds, and the pestle with the stem and
weight about 25 pounds. The tappets A are faced with
raw hide, which stand the wear of the cams, H, remarkably
well, much better than either hard or soft iron or steel,
the dust from the ore causing the latter to cut very fast.
The shaft makes about 90 revolutions per minute, and the
ore, when the mortar is about half filled, feeds itself, so
THE MECHANICAL TREATMENT.
333
that without any attention 25 pounds of hard ore in lumps
are reduced almost to powder in .about one and a half
hours.
The stem B is hooked up, so that the pestle clears the
FIG. 89.
top of the mortar A. The pulley D raises the mortar
clear of the block, and by means of the traveller E the
mortar is emptied on the chilled iron plate F, as figured
in the sketch. The ore is ground to a fine powder on this
chilled plate by the hardened steel muller c, any of it
falling off the plate being caught in the sheet-iron
troughs G.
It is thoroughly mixed and quartered, and the result-
ing sample, reduced finally to about 6 to 8 ounces, is
placed in a clean dry bottle. This bottle has the number
of the sample etched upon it, the same number in figures
half an inch high pasted on the neck, and the label which
came in the sample bag pasted on its side. The bottle
is not taken into the grinding-room until the sample is
being ground, and is always previously numbered, so
that when it is brought upstairs with the label the num-
bers can be compared and the label pasted on. A por-
334
COMPLETE ASSAY OF IRON ORES.
tion of the sample is ground in the agate mortar A (fig.
90), with an agate pestle B, fitted in a flexible shaft c, and
FIG. 90.
CEILINC LINE
revolving at the rate of 700 times a minute, transferred to
a ground-glass stoppered bottle, and dried at 100 C. It
is then ready for analysis.
THE CHEMICAL TREATMENT.
Estimation of Phosphoric Acid.
For ores low in phosphoric acid 10 grms. are dissolved
in hydrochloric acid ; the solution is evaporated to dry-
ness on the sand bath, re-dissolved in dilute hydrochloric
acid (one part of acid to two of water), filtered, the filtrate
treated with 10 c.c. ammonium sulphite * to reduce
the iron to the ferrous condition, ammonia added until
the solution is nearly neutral, and then heated until it has
decolourised. Five c.c. of strong hydrochloric acid are
* Made by saturating ammonia with sulphurous acid, generated by heating
powdered charcoal and strong sulphuric acid in a flask. The mixture is
made of the consistency of cream, and the gas passed first through a wash-
b ottle containing water.
ESTIMATION OF PHOSPHORIC ACID. 335
added to decompose any excess of ammonium sulphite
and the sulphurous acid driven off by passing carbonic acid
through the boiling solution. When the last trace of
sulphurous acid is driven off, sulphuretted hydrogen is
passed through the boiling solution to precipitate any
arsenic, the sulphide of arsenic filtered off, the beaker
placed in cold water, and when thoroughly cooled, dilute
ammonia added until a slight permanent green precipitate
of ferrous hydrate remained after stirring. Acetic acid
is added until the precipitate dissolved (a few drops
should be sufficient) the solution, diluted with hot water
to about 800 or 900 c.c. And if the precipitate is white,
a dilute solution of ferric chloride or bromine water is
added until it becomes red. If it is necessary to add
much ferric chloride a few drops of ammonic acetate are
also added to insure the decomposition of all the former
salt. The solution is then heated to boiling, boiled a
few minutes, filtered rapidly, and washed once with hot
water. The filtrate should pass through the filter perfectly
clear, and if the precipitate is red, any subsequent cloudi-
ness of the filtrate is of no consequence. If, however,
the filtrate passes the filter cloudy, it should be returned
to the main portion, a few drops more of ferric chloride
added, and the solution again boiled and filtered. The
precipitate of ferric phosphate, hydrate, and basic acetate
is dissolved in hot dilute hydrochloric acid (1 1) on
the filter, and the large beaker cleared of any adhering
precipitate with hydrochloric acid, received in a small
beaker, and evaporated nearly to dryness. Five to ten
grms. of citric acid are dissolved in the least possible
quantity of hot water, and filtered into a small beaker into
which are also filtered about 5 c.c. of magnesium mix-
ture,* and the whole is added to the solution of the
ferric phosphate. This solution is then neutralised by
ammonia and cooled. When perfectly cold, from one-
third to one-half its bulk of strong ammonia is added,
* Made by dissolving equal weights of sulphurous acid and ammonium
chloride in the least possible quantity of water, filtering, adding bulk of
strong ammonia, stirring occasionally, and allowing it to stand for some days
before using.
336 COMPLETE ASSAY OF IRON ORES.
and the solution is stirred until the precipitate of ammo-
nium-magnesium phosphate begins to form. After standing
for some time the solution is stirred again, and the
stirring is repeated at intervals for an hour. It is allowed
to settle overnight, filtered on the asbestos felt in Dr.
Gooch's pierced crucible, washed with dilute ammonia
(1 3), dried on the pump, moistened with a drop or two
of ammonium nitrate in ammonia, dried and ignited until
the glow passes over the precipitate, cooled in a desiccator,
and weighed as magnesium phosphate. This precipitate
is then dissolved in dilute hot hydrochloric acid, the felt
is washed on the pump, and the crucible is heated to
redness and re-weighed. This weight, unless the precipitate
of magnesium phosphate contained silica, agrees perfectly
with the original weight. In the latter case, however, the
last weight of the crucible is subtracted from the weight
of the crucible with the precipitate to obtain the weight of
magnesium phosphate. The following table is used for cal-
culating the percentage of phosphoric acid or phosphorus,
instead of the factors 0-6396 or 0-2793.
When the amount of magnesium-ammonium phosphate
is large it is dissolved, boiled, and re -precipitated, as a
small amount of magnesia is liable to be carried down
mechanically with it. It is found necessary to avoid
heating the magnesium phosphate after the glow has
passed, as this salt is liable to attack the asbestos slightly
at high temperatures, causing a small loss in weight of
the felt on the subsequent solution of the precipitate in
dilute hydrochloric acid. For ores containing titanic acid
it is necessary to modify this process in several particu-
lars. It is found that under certain circumstances phos-
phoric acid combines with titanic acid, forming a salt
(possibly a phospho-titanate) very insoluble in hydrochloric
acid, so that in almost every instance phosphoric acid is
found in the insoluble silicious residue when this contains
titanic acid. Also, upon neutralising the main solution
after adding ammonium sulphide in ores containing titanic
acid, a fine white precipitate resembling barium sulphate
is formed, which usually remains after the subsequent
ESTIMATION OF PHOSPHORIC ACID.
337
TABLE FOR PHOSPHORUS AND PHOSPHORIC ACID.
Mg
p
P 2 5
Mg
P
PA
Mg
p
P 3 5
1
0-003
0-006
35
0-098
0-224
68
0-190
0-434
2
0-005
0-013
36
0-100
0-230
69
0-193
0-441
3
0-008
0-019
37
0-103
0-237
70
0-195
0-448
4
0-011
0-026
38
0-106
0-243
71
0-198
0-454
5
0-014
0-032
39
0-109
0-249
72
0-201
0-460
6
0-017
0-038
40
0-112
0-256
73
0-204
0-467
7
0-019
0-045
41
0-114
0-262
74
0-207
0-473
8
0-022
0-051
42
0-117
0-269
75
0-209
0-479
9
0-025
0-057
43
0-120
0-275
76
0-212
0-486
10
0-028
0-064
44
0-123
0-281
77
0-215
0-492
11
0-031
0-070
45
0-126
0-287
78
0-218
0-499
12
0-033
0-077
46
0-128
0-294
79
0-221
0-505
13
0-036
0-083
47
0-131
0-300
80
0-223
0-512
14
0-039
0-089
48
0-134
0-307
81
0-226
0-518
15 0-042
0-096
49
0-137
0-313
82
0-229
0-524
16
0-045
0-102
50
0-139
0-319
83
0-232
0-531
17
0-047
0-108
51
0-142
0-326
84
0-235
0-537
18
0-050
0-115
52
0-145
0-332
85
0-237
0-544
19
0-053
0-121
53
0-148
0-339
86
0-240
0-550
20
0-056
0-128
54
0-151
0-345
87
0-243
0-556
21
0-059
0-134
55
0-154
0-352
88
0-246
0-563
22
0-061
0-141
56
0-156
0-358
89
0-248
0-569
23
0-064
0-147
57
0-159
0-364
90
0-251
0-576
24
0-067
0-153
58
0-162
0-371
91
0-254
0-582
25
0-070
0-159
59
0-165
0-377
92
0-257
0-588
26
0-073
0-166
60
0-167
0-384
93
0-259
0-595
27
0-075
0-173
61
0-170
0-390
94
0-262
0-601
28
0-078
0-179
62
0-173
0-396
95
0-265
0-607
29
0-081
0-185
63
0-176
0-403
96
0-268
0-614
30
0-084
0-192
64
0-179
0-409
97
0-271
0-620
31
0-086
0-198
65
0-181
0-416
98
0-274
0-627
32
0-089
0-204
66
0-184
0-422
99
0-276
0-633
33
0-092
0-211
67
0-187
0-428
100
0-279
0-639
34
0-095
0-217
addition of hydrochloric acid, and consists of titanic
acid with some phosphoric acid. If, after dissolving
the precipitate of ferric phosphate, &c., thrown down by
ammonic acetate, the solution is allowed to run to dry-
ness, there is found after re-solution in hydrochloric
acid a granular white or yellowish residue, absolutely
insoluble in hydrochloric acid, which consists essentially
of titanic and phosphoric acid. This reaction affords a
very delicate test for titanic acid, for if the residue obtained
in this latter case be collected on a small filter, dried,
ignited, and fused with sodium carbonate, and the fusion
treated with hot water, a residue of sodium titanate
will remain, which is insoluble in water. This residue
z
338 COMPLETE ASSAY OF IRON ORES.
collected on a small filter, dissolved in a little hot dilute
hydrochloric acid, and treated with zinc in a test-tube,
gives the very characteristic violet colour due to titanic
oxide. When titanic oxide is found it is necessary to
fuse the insoluble silicious residue, the residue left in the
filter from the solution of the precipitate by ammonic
acetate, and the residue left from the re-solution of this
latter, after it has run to dryness, with sodium carbonate.
The fused mass is treated with hot water and filtered ; the
filtrate containing, besides silica and alumina and sodium
carbonate, all the phosphoric acid, while the titanic acid
remains on the filter as sodic titanate, insoluble in water.
Potassium salts should not be used, as potassic titanate is
decomposed by water. The filtrate is acidulated with
hydrochloric acid, .evaporated to dryness, dissolved in
water with a little hydrochloric acid, filtered, a few drops
of solution of ferric chloride added, and the ferric phos-
phate precipitated by ammonic acetate. This precipitate
is filtered, washed, dissolved in hydrochloric acid, and
the phosphoric acid estimated by magnesium mixture
as usual.
Sulphur and Iron.
One grm. of ore is fused with 10 to 12 grms. of
sodium carbonate and a little nitre, the fused mass is
run well up on the sides of the crucible, and the cru-
cible is chilled. The mass is then detached from the
crucible and transferred to a tall beaker, or the crucible
with its contents is placed in the beaker and treated
with boiling water. When the fused mass is entirely
disintegrated (the crucible, if placed in the beaker, having
been washed off and removed) the ferric oxide is allowed
to settle. If the solution is coloured by alkaline man-
ganate a few drops of alcohol are added, and the solution
is allowed to stand until the colour disappears. If the
solution after the disappearance of the colour due to the
manganate is yellow, it is an indication of the presence
of chromium in the ore. It is then decanted through a
filter and washed twice with hot water by decantation.
SULPHUR AND IRON. 339
The filter is washed once or twice with hot water,
and the filtrate is acidulated with hydrochloric acid,
evaporated to dryness, re-dissolved in hot water with
a few drops of hydrochloric acid, filtered from the
silica, and the sulphuric acid precipitated by barium
chloride in the boiling filtrate, After standing over-
night in a warm place the barium sulphate is fil-
tered on the felt, washed thoroughly with hot water
(and ammonium acetate if it is large), and weighed
as barium sulphate, which multiplied by 0-1373 gives
the sulphur.
The crucible in which the fusion was made is treated
with hydrochloric acid, and when the adhering ferric oxide
is dissolved a little hot water is added, and the whole
is poured on the filter to dissolve any ferric oxide which
might have been in suspension in the solution of the fused
mass. This is allowed to run into the beaker containing
the ferric oxide, the crucible is rinsed, and the filter
is washed with hot water. The whole is evaporated
to dryness to render silicic acid insoluble, re-dissolved
in hydrochloric acid, and transferred to a small flask
of about 50 c.c. capacity. A small funnel is put in
the neck of the flask, and 3 grms. of granulated zinc in
small lumps are added carefully to the solution. By
heating the solution and adding a few drops of hydro-
chloric acid from time to time the ferric chloride is soon
reduced to ferrous, which is shown by the solution be-
coming colourless. After washing down the neck of the
flask and the funnel with a fine jet of water, if the addition
of a few drops of hydrochloric acid fails to impart any
colour to the solution, the reduction is considered com-
plete. Fifteen c.c. of dilute sulphuric acid (1 1) are
added little by little to the solution, and when the zinc
has all dissolved the neck of the flask is filled nearly
to the top with hot water. The flask is then placed
in cold water, and when the solution is thoroughly
cooled it is washed out into an oblong white dish of
about 1,500 c.c. capacity and diluted to about one
litre. A solution of potassium permanganate is run
z 2
340
COMPLETE ASSAY OF IRON ORES.
FIG. 91.
in from a burette, the representation of which is given in
fig. 91.
This form of burette, the invention of Mr. Thomas H.
Garrett, is the most satis-
factory one I have ever used.
The solution of permanga-
nate is carefully standard-
ised by means of ferric chlo-
ride of known strength, made
by dissolving wrought iron
of known composition in ni-
tric acid, driving off the nitric
acid by repeated evapora-
tions with hydrochloric acid,
diluting, filtering into a clean
glass-stoppered bottle, and
estimating the ferric oxide
gravimetrically in a weighed
amount of the solution. The
strength of the ferric chlo-
ride solution being thus ac-
curately known, a portion
is weighed out into one of
the little flasks previously
mentioned, reduced with zinc,
sulphuric acid added, and
the strength of the perman-
ganate solution necessary to
colour 3 grms. of zinc treated
with hydrochloric acid and
sulphuric acid, and diluted to
the same volume as the solution of the ore, is subtracted
before calculating the value of the permanganate solution ,
and also from the amount of permanganate solution re-
quired for the ore before calculating the percentage of
iron. By using sulphuric acid in addition to the hydro-
chloric acid as described above it is found that the end
reaction with potassium permanganate is as sharp as if no
hydrochloric acid is used, and the results obtained are
SILICA, FERRIC OXIDE, ALUMINA, ETC. 341
in all cases as accurate as could be desired. For many
ores it is possible to dissolve the finely ground sample
directly in the flask in hydrochloric acid without previous
fusion, but for those containing ferrous silicates or pyrites
it is generally necessary to resort to fusion for the accu-
rate estimation of the total iron.
Silica, Ferric Oxide, Alumina, Manganese, Lime, and
Magnesia.
One gramme is dissolved in hydrochloric acid, evapo-
rated to dryness, re-dissolved in dilute hydrochloric acid,
and evaporated a second time to render the silica in-
soluble. It is then re-dissolved in hydrochloric acid and
water (1 3), filtered on a small ashless filter, washed,
dried, ignited, and weighed as insoluble silicious matter.
This is fused with five times its weight of dry sodium
carbonate, treated with hot water, and washed into a
platinum dish, the crucible treated with acid and carefully
rinsed into the dish, the whole acidulated with hydro-
chloric acid, evaporated to dryness, treated with water and
a little hydrochloric acid, and evaporated to dryness.
The mass is then treated with hydrochloric acid and
water (] 5), heated to boiling, and filtered on a small
ashless filter. The filtrate is dried, ignited, and weighed.
The silica is treated in the crucible with hydrofluoric
acid * and sulphuric acid, evaporated to dryness, ignited,
and weighed. The difference between this and the weight
of the precipitate is silica. Any residue obtained by
the treatment with hydrofluoric acid and sulphuric acid
is examined. It might consist of barium sulphate or
titanic acid if either of these exists in the ore, or of a
little alumina or ferric oxide or of a small amount of
sodium sulphate if the silica is not carefully washed.
The filtrate from the silica is treated in a platinum dish
with ammonia, boiled until it smells but faintly of
* The hydrofluoric acid is re-distilled with the addition of a little sul-
phuric acid and potassium permanganate from a platinum still, and collected in
a platinum bottle, as the crude hydrofluoric acid always contains ferric oxide,
besides various sulphur compounds.
342 COMPLETE ASSAY OF IRON ORES.
ammonia, filtered, washed, dried, ignited, and weighed as
alumina with or without a tinge of ferric oxide. To
the filtrate, ammonium oxalate is added, the solution is
boiled and allowed to stand overnight, filtered, washed,
and ignited at a high temperature, and weighed as lime.
This filtrate is evaporated down, sodium and am-
monium phosphate are added, and the solution is well
stirred to precipitate the magnesium-ammonium phos-
phate. After standing overnight it is filtered, washed,
ignited, and weighed as magnesium phosphate. Whence
111 : 40 = weight of magnesium phosphate to weight of
magnesia. The sum of the weights of the silica, alumina,
lime, and magnesia should about equal that of the in-
soluble silicious matter. When there is a deficiency,
alkalies are looked for in another portion. If there is
an excess, and the precipitated alumina is red, iron in
insoluble matter is estimated.
The filtrate from the insoluble silicious matter is nearly
neutralised with sodium carbonate solution, 2 grammes
of sodium acetate are added, and the whole after being
diluted to about 700 c.c. with hot water is boiled, and
the precipitate ferric oxide, &c., is filtered on a washed
filter. The precipitate is washed two or three times on
the filter, and then transferred back to the beaker with
a platinum spatula, the filter is washed with hydro-
chloric acid and finally with water, the whole being re-
ceived in the beaker with the mass of the precipitate.
Sufficient acid is added to dissolve the precipitate, and
the operation is repeated, the filtrates being added to-
gether. The solution of the precipitated ferric oxide, &c.,,
is evaporated to dryness to render insoluble any silica
from the reagents ; re-dissolved in dilute acid, and filtered
into a large platinum dish ; the solution is boiled, am-
monia is added, and the precipitate is collected on an
ashless filter, washed thoroughly on the pump with hot
water, dried, ignited (finally over the blast), and weighed
as iron and aluminium phosphate. The filtrates from the
acetate precipitations are evaporated down to about 300
c.c., 2 or 3 grammes of sodic acetate are added, the-
SILICA, FEEKIC OXIDE, ALUMIXA, ETC. 343
solution is heated to boiling, and sulphuretted hydrogen
is passed through to precipitate any copper, nickel,
cobalt, and zinc.
The precipitate by sulphuretted hydrogen is filtered
off', and after all smell of sulphuretted hydrogen has
been boiled off, bromine water is added to the solution.
When the precipitated manganic oxide has collected at
a gentle heat, and while the solution is still coloured by
bromine, it is boiled until colourless, filtered, washed
several times, and the manganic oxide is dissolved on
the filter in hydrochloric acid with the addition of solu-
tion of sulphurous acid, which causes its very rapid solu-
tion. This solution is evaporated to dryness, re-dissolved
in hydrochloric acid and water ; a slight excess of am-
monia is added, nearly all smell of ammonia is boiled
off, the solution is filtered, any slight precipitate of
ferric oxide, &c., is re-dissolved and re-precipitated and
filtered as before. The filtrates are added together, and
excess of sodium- ammonia phosphate is added with
enough hydrochloric acid to render the solution decidedly
acid, and after boiling for some time an excess of am-
monia is added to precipitate the manganese-ammonium
phosphate. This is boiled until the precipitate becomes
crystalline and the solution smells but slightly of am-
monia, when it is filtered, washed with cold water, ignited,
and weighed as manganese phosphate, which multiplied
by 0-5 gives oxide of manganese.
To the filtrate from the precipitate of manganic oxide
by bromine is added the ammoniacal filtrate from the
final precipitation of the iron, alumina, and phosphoric
acid ; the whole is evaporated down to about 400 c.c.,
and the lime and magnesia are precipitated as described
in the analysis of the insoluble silicious matter. There
are several sources of error which it is found very
necessary to guard against in the above analysis namely,
the contamination of the distilled water by silica when it
is boiled in glass, the strong action of ammoniacal solu-
tions on the beakers, and the presence of silica, alumina,
ferric oxide, lime, and magnesia in many reagents, and
344 COMPLETE ASSAY OF IRON OKES.
especially in sodium carbonate. Distilled water which has
been heated overnight in a Bohemian flask on the sand-
bath is found to contain 52 milligrammes of solid residue
to the litre, 26 milligrammes of which is silica ; and
ferric oxide precipitated from a solution to which water
boiled in a flask from two to six hours has been added,
contains after ignition as much as 8 per cent, of silica.
To avoid this tin-lined copper flasks are used for heat-
ing distilled water, and to avoid the error due to the
action of ammonia salts and ammoniacal solutions on the
beakers, all the precipitations and evaporations are made
in platinum. We were fortunate enough to obtain some
remarkably pure sodium carbonate, containing only 2
milligrammes of silica, 1^ milligramme of alumina, 2
milligrammes of lime, and 2 milligrammes of magnesia to
the 100 grammes from Messrs. Powers and Weightman,
Philadelphia.
Nickel, Cobalt, and Zinc.
Three grammes of ore are treated as if for the esti-
mation of manganese, and the sulphides of copper,
nickel, cobalt, and zinc are precipitated from the boiling
solution of the acetates by sulphuretted hydrogen. This
precipitate is collected on a small filter, washed with
sulphuretted hydrogen water containing a little acetic
acid, dried, ignited in a porcelain crucible, and transferred
to small beaker. It is then treated with hydrochloric
acid and a little nitric acid, evaporated to dryness, re-
dissolved in hydrochloric acid, diluted, boiled, and the
copper precipitated by sulphuretted hydrogen. The
copper sulphide is filtered off and washed with hot
water, the filtrate containing the cobalt, nickel, and zinc
is evaporated to dryness, re-dissolved in a few drops
of water containing not more than two or three drops of
hydrochloric acid, and an excess of potassium nitrate
acidulated with acetic acid added. After standing for a
day or two the potassic-cobaltic nitrite is filtered off,
washed once or twice with a strong solution of potassic
acetate, and then, after removing the filtrate, with alcohol
NICKEL, COBALT, AND ZINC. 345
to displace the alkaline salts. The precipitate is then
ignited very carefully in a porcelain crucible, treated with
sulphuric acid to decompose all the nitrate, made alkaline
with ammonia, filtered, and the cobalt is precipitated by
the battery ; or the ignited precipitate is dissolved in a few
drops of hydrochloric acid, transferred to a small beaker,
dilated, and any alumina and iron present is precipi-
tated by boiling for several hours with an excess of sodic
acetate. The precipitate of alumina and iron is filtered
off, the filtrate is made alkaline with ammonia, and the
cobalt is precipitated as sulphide by ammonium sul-
phide. The precipitate is allowed to settle, filtered off
on a small ashless filter, washed, dried, ignited, and weighed
as cobaltous sulphide, or treated with sulphuric acid and
zinc, and weighed as cobaltic sulphide, which multiplied
by 0*4872 gives cobalt oxide.
The filtrate from the potassium- cobaltic nitrite is
acidulated strongly with hydrochloric acid and heated to
decompose all the nitrite, and the nickel oxide is preci-
pitated by an excess of soda or potash, filtered, and the
filtrate tested for zinc oxide with ammonium sulphide.
If any zinc sulphide is found, the nickel oxide is re-
dissolved and precipitated as before, filtered, and the
filtrate and washings are added to the first filtrate. The
zinc sulphide is allowed to settle, filtered, washed with
water containing sulphide of ammonium, re-dissolved in
hydrochloric acid, and evaporated to dryness. This
is treated with dilute hydrochloric acid, filtered, and
the zinc is precipitated by solution of sodium carbonate,
filtered, washed, dried, ignited, and weighed as zinc oxide.
The precipitate of nickel oxide is dissolved on the
filter in hydrochloric acid, the filtrate run to dryness with
sulphuric acid, diluted, excess of ammonia is added,
filtered, and the nickel precipitated by the battery ; or
the hydrochloric acid solution is evaporated to dryness,
dissolved in a drop or two of hydrochloric acid, diluted,
and boiled with an excess of sodium acetate. Any pre-
cipitate of alumina, &c., is filtered off, excess of sulphide
of ammonium is added, then an excess of acetic acid
346
COMPLETE ASSAY OF IRON ORES.
and sulphuretted hydrogen is passed through the boil-
ing solution. The precipitated nickelic sulphide and
sulphur are collected on a small ashless filter, dried,
ignited, heated with a little ammonic carbonate, and
weighed as nickelous sulphide.
Estimation of Ferrous Oxide.
One gramme of finely ground ore is weighed into a
flask A (fig. 92) of about 100 c.c. capacity, fitted with a
FIG. 92.
rubber stopper, through which pass two glass tubes as
shown in the cut. Dry carbonic acid is passed in through
the tube B to expel the air through the tubes c and D,
the latter dipping beneath the surface of the water in
the beaker E, and when this is accomplished the stop-
per is removed, 15 c.c. of strong hydrochloric acid is
quickly poured in, the stopper is replaced, and the ore
is dissolved with the aid of heat, the current of carbonic
acid being continued uninterruptedly. When the solution
ESTIMATION OF FEEEOUS OXIDE. 347
of the ore is accomplished the source of heat is re-
moved, and, the current of carbonic acid being temporarily
stopped, the water in the beaker B is allowed to run
back into the flask A until the latter is nearly filled,
when the current of carbonic acid is turned on again,
and allowed to continue until the solution in the flask is
thoroughly cooled. This is accomplished by removing
the tripod c, placing a dish nearly filled with cold water
under the flask, and replacing the tripod. The solution
is then washed out into the dish used for titrating, into
which 3 grms. of zinc dissolved in 15 c.c. of the dilute
sulphuric acid (1 1) are previously poured, and the whole
is diluted to 1 litre. The amount of ferrous oxide is
then estimated by the potassic permanganate solution.
The amount of solution required to colour about 1 grm.
of ferric chloride diluted to the same bulk, and containing
the same amount of sulphuric acid, is subtracted from
the amount required for the titration before calculating
the amount of ferrous oxide.
Ferrous Oxide in Insoluble Silicious Matter.
When the insoluble silicious matter contains iron in
the ferrous condition for instance, in the form of epidote
1 grm. is treated with hydrochloric acid diluted, and
the insoluble matter is collected on the felt in a pierced
crucible, carefully dried, transferred with the felt to an
ordinary platinum crucible, and treated in the apparatus
of fig. 93, with hydrofluoric acid and hydrochloric acid in a
current of carbonic acid. When entirely decomposed it
is allowed to cool, the current of carbonic acid being
kept up, and then transferred quickly to a dish containing
about 1 litre of water and 15 c.c. of the dilute sulphuric
acid and zinc ; the crucible is washed out, and the amount
of ferrous oxide is estimated by the standard solution
of potassium permanganate.
348
COMPLETE ASSAY OF IRON ORES.
Sulphuric Acid.
Sulphuric acid may exist in the ore in the form of
barium sulphate, calcium sulphate, or as sulphates of iron,
FIG. 93.
&c., formed in the roasting or weathering of an ore con-
taining pyrites. When it exists as barium sulphate it is
found in the analysis of the insoluble silicious matter in
the residue remaining after the treatment of the silica with
hydrofluoric acid. In this case the residue is fused with
sodium carbonate, treated with hot water, and filtered.
The filtrate is acidulated with hydrochloric acid, and
the sulphuric acid is precipitated by barium chloride
and estimated as barium sulphate, which multiplied by
0*3433 gives sulphuric anhydride.
The barium carbonate on the filter is dissolved in
dilute hydrochloric acid, sulphuric acid is added, and
the barium sulphate is finally weighed, which multiplied
SULPHUEIC ACID. 34$
by 0-6567 gives baryta. In the other cases 10 grammes
of the ore finely ground are treated with water contain-
ing a little hydrochloric acid, filtered, the sulphuric acid
is precipitated in the filtrate as barium sulphate, and the
weight is determined with the usual precautions.
Alumina.
The total amount of the soluble ferric oxide, alumina,
and phosphoric acid is added to the ferric oxide and
alumina found in the insoluble silicious matter, and from
this is subtracted the phosphoric acid which gives the
total iron and alumina. From this is subtracted the
ferric oxide found by titration ; the difference is alumina.
The accuracy of this result depends of course upon that
of the volumetric estimation of the ferric oxide. The
comparison of many analyses shows that the error due
to this need never exceed a few hundredths of 1 per cent.,
and no direct method is found to yield results equally
accurate or concordant.
Calculation of the Analysis.
The sulphur in the sulphuric acid found as such is
subtracted from the total sulphur ; so much as is neces-
sary to form the sulphides of copper, nickel, cobalt, zinc,
and antimony with the amounts of these elements in the
ore, and supposed to exist in this condition, is sub-
tracted from this, and the remaining sulphur is calcu-
lated to iron disulphide. The amount of iron required
for this is calculated to ferric oxide, and subtracted from
the ferric oxide found by titration. From this is taken
the amount of ferric oxide calculated from the ferrous
oxide found in the estimation of ferrous oxide, and the
result is the amount of ferric oxide existing in this state
in the ore.
Carbonic Acid.
Three grammes of ore are weighed into the flask A
of the apparatus (fig. 94), and the connections are made
as there shown ; 10 c.c. of strong sulphuric acid are
350
COMPLETE ASSAY OF IRON ORES.
added to 65 c.c. of water and poured into the bulb-tube
B, the stopcock c having first been closed. After the
potash bulb and the drying-tube are weighed they are
attached to the apparatus, the tube s is filled with fused
calcium chloride, being added to prevent the drying-tube
R from, absorbing moisture from the atmosphere. The
WATEE AND CARBON. 351
TJ-tube o is empty, p is filled with pumice saturated
with anhydrous copper sulphate, and Q with dried calcium
chloride. When the connections are all proved to be
tight, N is fitted into the neck of the bulb-tube with a
piece of rubber tubing, and the acid in the bulb is
allowed to run into the flask very slowly, and when it is
all in, a slow current of air is forced through the appa-
ratus by means of the bottles L L. The air is freed from
carbonic acid by potassic hydrate in the tube M. As soon
as the current of air is started, the flask A is heated
gradually, and finally the solution is boiled until the
bend of the tube o is filled with condensed water. It
is then allowed to cool while the current of air is con-
tinued. The potash bulb and the drying-tube are de-
tached and weighed with the usual precautions ; the
increase of weight is of course the carbonic acid due to
the carbonates contained in the ore.
Water and Carbon in Carbonaceous Matter.
Many ores besides black-bands contain carbon ; for in-
stance, several ores from New Jersey contain graphite,
and nearly all limonites and magnetites contain carbon in
organic matter, probably from the organic acids which
originally held the iron in solution. As pyrites is also
of common occurrence in such ores it is necessary to
devise some method by which the water of composition
and the carbon could be estimated in the presence of
pyrites. In attempting to estimate the water by heating
the ore in a current of air, some sulphuric acid is always
formed and driven over into the drying-tube, and some
organic matter is certain to be dissolved if the ore is
treated with acid for the estimation of the carbon. It
is found by careful experiments that when a carbonate
of any kind is fused with potassium anhydrous chromate in
excess the carbonic acid is all expelled together with
any water present, while pyrites treated in the same way
is oxidised with formation of potassium sulphate, which
is not decomposed even at a bright red heat.
352
COMPLETE ASSAY OF IRON OEES.
A very practical and easy way to apply this in the
treatment of ores is afforded by the use of Dr. Gooch's
tubulated crucible* (fig. 95), and the process is as
follows :
For ores containing much water or carbonic acid I
FIG. 95.
HES
gramme, and for others 3 grammes are weighed into the
crucible A of fig. 96, and carefully mixed, by means of a
rod or wire, with 7 grammes of potassium bichromate,
which has been previously fused and powdered, the cap
B is adjusted, and the whole is placed in the air-bath
and heated to 100 for a short time. The crucible is
then placed on the platinum triangle c, and connected by
means of a cork with the weighed tube D, containing
dried calcium chloride, and the weighed potash bulb E
and drying-tube F are attached, the latter guarded by
the calcium chloride tube G, as seen in the sketch.
The cap for the crucible consists of a conical cover H
drawn out vertically into a tube i, into which is burned
a horizontal tube J, of the same diameter. Through the
top of the tube I passes the tube K to a distance of half an
inch below the bottom of the cap, the end being slightly
bent ; K is burned into I at its point of entry a, sealing i
* ' Chemical News,' vol. xiii. p. 326.
WATER AND CARBON.
353
at this point. A straight glass tube M, drawn tapering,
is fused to the platinum tube j at b and another piece, N,
is bent at a right angle to the platinum K at c. A piece
of rubber tubing, closed with a piece of glass rod at one
end, is drawn over the end of the tube N, the space
around the cap in the flange d is filled with small
A A
354 COMPLETE ASSAY OF IRON ORES.
pieces of neutral anhydrous -sodic tungstate, which are
fused by means of a blowpipe flame, making an air-tight
joint. The mixture in the crucible is kept cool during
this operation by dipping the end in a beaker of cold
water. The expansion of the air in the apparatus during
the heating causes it to bubble through the potash bulb,
and the reflux of the solution in the bulb as the apparatus
cools is a good index of the tightness of the joints.
When the joints are shown to be all tight the cap is
removed from N, and the bottle L is connected with N,
as shown in the sketch. A slow current of air, freed from
carbonic acid and water by passing through the tube Q,
containing potash and calcium chloride, is then started
through the apparatus, and the crucible is heated very
gradually and cautiously by the burner o. As the steam
is gradually driven into the drying-tube it is not allowed
to condense at y, but is driven forward into the calcium
chloride by the heat of a small alcohol lamp, applied
to the drying-tube at this point. When the greater part
of the water has been thus driven over, the crucible
is heated by a horizontal flame from the blast-lamp p,
above the level of the mixture to prevent the latter from
frothing up and stopping the end of the tube K. The
bottom of the crucible is gradually heated to a dull red,
and allowed to remain at this temperature for about fifteen
minutes, when the lamps are turned out and the whole
is allowed to cool in the current of dried air from the
bottle L. The tubes are then re-weighed ; the increase
in the weight of D is the weight of the water of com-
position in the ore, and that of the potash bulb and drying-
tube the weight of the total carbonic acid. This latter, of
course, is the sum of the carbonic acid, due to the car-
bon of the carbonaceous matter or graphite, and of that
due to the carbonic acid in the carbonates. The amount
of the latter (previously estimated) is subtracted from
the total carbonic acid, and the difference multiplied by
0-2626 gives the carbon in carbonaceous matter or the
graphite.
ALKALIES. 355
Alkalies.
The alkalies are generally confined to the insoluble
silicious matter in the ores, but are occasionally found
in the portion soluble in hydrochloric acid. In the former
case 5 or 10 grammes of the ore are digested in hydro-
chloric acid until only the silicates remained undissolved,
diluted, and filtered. The insoluble residue is washed,
dried, ignited, and treated in the crucible with hydrofluoric
and sulphuric acids until it is entirely decomposed ; it
is then evaporated down until all the hydrofluoric acid
and nearly all the sulphuric acid has volatilised. It is then
treated with hot water and a little hydrochloric acid if
necessary ; the clear solution is transferred to a small
platinum dish, the alumina and ferric oxide are pre-
cipitated by ammonia, and the solution is boiled until
nearly all ammonia is driven off. It is then filtered into
another platinum dish, evaporated to dryness, and heated
until all the ammonium salts are volatilised. The residue
is treated with a little water, a few drops of ammonium
oxalate, and excess of ammonia, and the solution is
boiled and filtered. The filtrate is evaporated to dry-
ness, and the residue is heated to dull redness. It is then
treated with water, filtered, and an excess of barium
hydrate * is added to precipitate the sulphuric acid and
magnesia. The solution is boiled and filtered from the
barium sulphate, and the filtrate is evaporated to dry-
ness after the addition of a little ammonium carbonate
and hydrate. The residue is treated with a little water,
filtered to get rid of the barium carbonate, and the filtrate
is evaporated to dryness and heated carefully to vola-
tilise any ammonium salts. The residue is treated with
water, filtered into a small weighed platinum dish ; the
filtrate is acidulated with hydrochloric acid, evaporated
to dryness, heated carefully to a very low red, and weighed
as quickly as possible as alkaline chlorides. The alkaline
chlorides are then dissolved in water, any residue is
* Made by treating ordinary barium carbonate with tolerably strong nitric
acid, and washing the barium nitrate with nitric acid. The barium nitrate is
dried and fused until all the nitrous acid is driven off.
A A 2
356 COMPLETE ASSAY OF IRON ORES.
filtered off, weighed, and its weight subtracted from the
weight of the chloride, an excess of platinum chloride is
added, and the solution is evaporated on the water-bath
until the syrupy liquid solidifies on cooling. The residue
is treated with alcohol, 95 per cent, filtered on the felt,
washed with 95 per cent, alcohol, dried at 120 C., and
weighed as platinum-potassium chloride, which multiplied
by 0*1931 gives potassium oxide. The weight of the
platinum-potassium chloride, multiplied by 0-3056 to re-
duce to potassium chloride, is subtracted from the weight
of the alkaline chlorides, and the difference (sodium
chloride), multiplied by 0-5299, gives sodium oxide.
The filtrate from the platinum-potassium chloride is
evaporated to dryness in a platinum dish with the addition
of a little oxalic acid to decompose all the platinum
chloride ; the residue is treated with water, filtered, and
any magnesia is precipitated by microcosmic salt and
weighed as magnesium phosphate, which multiplied by
0- 4*2 8 gives the amount of magnesium chloride to be sub-
tracted from the weight of chlorides to get the amount of
alkaline chlorides.
For the estimation of alkalies in the portion of the ore
soluble in hydrochloric acid 1 gramme is treated with
hydrochloric acid, diluted, filtered into a platinum dish,
and heated to boiling ; a slight excess of ammonia is
added, and the whole is evaporated to dryness to render
the ferric oxide very granular and easy to wash. It is
then treated with water, a drop or two of ammonia is
added, and the ferric oxide, &c., filtered off and washed.
The filtrate, after the addition of a drop or two of sulphuric
acid, is evaporated to dryness, and the ammoniacal salts
are driven off by ignition. The residue is treated with
water, a little ammonium oxalate is added to precipitate
the lime, and the alkalies are estimated, as in the former
case, after the precipitation of the calcium oxalate.
Copper, Lead, Arsenic, and Antimony.
Ten grammes of the very finely ground ore are treated
with hydrochloric acid with the addition of potassium
COPPER, LEAD, ARSENIC, AND ANTIMONY.
357
chlorate until only the silicious residue remains unacted
upon. It is then diluted, filtered, reduced with ammo-
nium sulphite, all the excess of sulphurous acid is boiled
off, and sulphuretted hydrogen is passed through to satu-
ration. The precipitate is allowed to settle in a warm
place, and when the smell of sulphuretted hydrogen has
nearly passed off, it is filtered on the pump, washed,
and treated on the filter with potassium sulphide to dis-
solve out the sulphides of arsenic and antimony. The
filter containing the sulphides of copper and lead is
dried and ignited in a porcelain crucible. The ignited
sulphides are transferred to a small beaker and treated
with hydrochloric acid and nitric acid, excess of sulphuric
acid added, and the whole is evaporated down until sul-
phuric acid begins to volatilise. Water is then added,
and, if any lead sulphate separates out, an equal bulk of
alcohol, and the whole is allowed to stand some hours,
when the lead sulphate is filtered off, dissolved in basic
ammonium tartrate or acetate, and the lead sulphide pre-
cipitated by sulphuretted hydrogen is filtered, washed,
and ignited. It is treated in the crucible (porcelain)
with nitric and sulphuric acids, and finally ignited and
weighed as lead sulphate, which multiplied by 0-6832 gives
lead. The solution containing the copper is evaporated
down to drive off the alcohol, washed
out into a platinum crucible, and the
copper is precipitated on the cru-
cible by the battery in the apparatus
shown in fig. 97, washed with water
and alcohol, and weighed as copper.
The potassium sulphide solution
containing the sulphides of arsenic
and antimony is acidulated with
hydrochloric acid, and allowed to
stand in a warm place until the sul-
phides and free sulphur have collected
and the solution smells but faintly
of sulphuretted hydrogen. It is then
filtered on the felt, and the precipitate, if it contains
FIG. 97.
358 COMPLETE ASSAY OF IKON OKES.
much free sulphur, is treated with carbon disulphide r
transferred with the felt to a small beaker, and treated
with hydrochloric acid and potassium chlorate, or aqua
regia, which dissolves the arsenic and antimony very
readily.
A little tartaric acid is added to keep the antimony
in solution, the whole is diluted and filtered, excess of
ammonia is added, and then magnesium mixture, and
the arsenic precipitated as ammonio-magnesium arseniate
with the same precautions that were used in the precipi-
tation of phosphoric acid. The precipitate is filtered on
the felt, washed with dilute ammonia, dried at 100 C., and
weighed as ammonio-magnesium arseniate.
It is then heated very gradually at first, and finally
to a full red, and weighed as magnesium arseniate. The
first weight obtained multiplied by 0*3947, and the second
by 0-4839, gives arsenic. If the heat is applied carefully
and slowly enough at first the latter result is most apt
to be -the correct one. The filtrate which contains the
antimony is acidulated with hydrochloric acid, and the
antimony is precipitated by sulphuretted hydrogen with
the usual precautions, filtered on a small disc of paper on
the bottom of the perforated crucible, dried, and the pre-
cipitate and paper are treated in a porcelain crucible
with fuming nitric acid, evaporated to dryness, ignited, and
weighed as antimony oxide, which multiplied by 0'7922,
gives antimony.
Titanic Acid.
As titanic acid is found to interfere with the estima-
tion of phosphoric acid, so the latter is found in many
cases to absolutely prevent the precipitation of the
former. When the ore contains only a small amount of
titanic acid 5 grammes are treated with hydrochloric
acid, evaporated to dryness, re-dissolved in dilute hydro-
chloric acid, filtered, the insoluble residue dried, ignited,
and treated in the crucible with hydrofluoric and sulphuric
acids. The solution in the crucible is evaporated down
until all the hydrofluoric and sulphuric acids are driven
TITANIC ACID. 359
off, ignited, and the residue is fused with sodium car-
bonate. The fused mass is treated with hot water, which
dissolves the sodium phosphate, aluminate, and excess of
sodium carbonate,* leaving insoluble the sodium titanate
and any ferric oxide, calcium carbonate, barium carbonate,
&c. The insoluble portion is filtered off, washed, dried,
and put aside to be added to the titanic acid in the soluble
portion. The filtrate obtained after the first treatment of
the ore with hydrochloric acid, which contains the great
mass of the iron with the titanic acid which dissolves with
it, is deoxidised by ammonium sulphite, the excess of sul-
phurous acid driven off, and the ferric phosphate, titanate,
and small excess of ferric oxide are precipitated by am-
monium acetate exactly as in the estimation of the phos
phoric acid. This precipitate is filtered off, washed;
dried, and fused with sodium carbonate, the fused mass
is treated with water, filtered, and the insoluble portion
containing ferric oxide and sodic titanate is washed and
dried. The two filters containing the sodic titanate are
separated as carefully as possible from the adhering pre-
cipitates, ignited in a platinum crucible, the precipitates
are added, and the whole is fused at a low heat with
15-20 times the weight of acid potassium sulphate. The
heat is gradually increased to a dull red, and kept at
this point until the fusion was perfectly clear and of a
reddish-brown colour. It is then cooled and treated
with a large excess of cold water, containing a few drops
of sulphuric acid, with constant stirring until nothing re-
mains undissolved, except, perhaps, a little silica. It
is then filtered, washed with cold water, the filtrate is
diluted to about 500 c.c., and boiled in a large platinum
dish or beaker with the constant addition of strong aque-
ous solution of sulphurous acid, until all the titanic acid
is precipitated. The precipitate is allowed to settle, the
clear solution is decanted through a filter, the pre-
cipitate is boiled up with water, thrown on the filter,
and washed as titanic acid. The filtrate is boiled again
with the addition of sulphurous acid, and any further pre-
* Any tungsten in the ore would be in this solution as sodium fcungstate.
360
COMPLETE ASSAY OF IRON ORES.
cipitate filtered off and added to the first. It is found
.that when the solution is kept at the boiling-point during
the filtration, and the precipitate washed with boiling
water, there is no tendency to pass the filter ; whereas,
if the solution is allowed to get cold, the filtrate is
nearly always cloudy. To prevent the precipitation of
ferric oxide with the titanic acid, it is necessary to keep
an excess of sulphurous acid in the solution ; but if in
spite of this precaution the ignited titanic acid is
coloured with ferric oxide, it is fused again with potas-
sium sulphate and re-precipitated. In the analysis of ores
containing much titanic acid it is preferable to fuse the
ore at once with sodium carbonate, treat with water, filter,
wash and dry the residue, separate and burn the filtrate,
fuse the residue and filter ash with potassium sulphate, and
estimate the titanic acid as before.
> 98<
Estimation of Specific Gravity.
A great many experiments were made both on lumps
and on powdered ore, and it was decided that in commer-
cial samples good results could
be obtained only by using the
powdered material. The work
was done and the flask herein
described was made by Mr. James
Hogarth. About 20 grammes
of the powdered sample are
weighed out and transferred to
the specific gravity flask, the con-
struction of which is shown in
fig. 98. This flask is made
with the view of overcoming two
difficulties which occur in using
the common flask, viz. the ex-
pansion and overflow consequent
on transferring the flask at 60
P. to the higher temperature of
SPECIFIC. BRAVITV FLASK
the balance-case, and the necessity for waiting until the
SPECIFIC GRAVITY. 361
finely divided mineral has settled before the stopper
could be inserted without loss of material. These ends
are successfully met by melting on a capillary tubulus to
the lower part of the neck, and grinding in a stopper
having a small bulb above the capillary to allow for
expansion.
Having transferred the weighed quantity of ore to the
flask, enough water is added to cover it, and the tem-
perature is raised almost to the boiling-point by placing
the flask in a water-bath. To insure complete expulsion
of air the flask is now placed under a bell-jar, connected
with the aspirator, and allowed to boil for a few moments
at the reduced pressure. It is then filled up with water
almost to the tubulus, cooled, tlie stopper is inserted,
and by suction it is filled up slightly above the mark on
the capillary part of the stopper. When the thermo-
meter is stationary at 60 F. the volume is adjusted by
touching the capillary end of the tubulus with blotting-
paper, or by bringing a drop of water to it as the case
might require. The flask is then dried, transferred to
the balance-case, and, after a sufficient time has elapsed
to allow it to take the temperature of the room, weighed.
If W be the weight of ore taken, W the weight of ore
4- water at 60, and K the weight of the flask + water at
60 then
Q
Sp.gr- = =
To obtain K the flask is nearly filled with boiled water and
treated exactly as described above.
The Perforated Crucible.
In processes in which the use of filter-paper is delete-
rious or inconvenient, nitrations and ignitions have been
made with the aid of the asbestos filter devised by Mr.
Gooch, which consists essentially of a felt of asbestos
deposited upon a perforated surface of platinum.*
* For a full description of the process see ' Chemical News,' xxxvii.
D. 181.
362
COMPLETE ASSAY OF IRON ORES.
FIG. 99. PERFORATED CRUCIBLE.
The mode of preparing the filter is as follows:
Asbestos of fine silky flexible fibre is scraped longitudi-
nally (not cut) to a
fine short down,
which is purified by
boiling in strong hy-
drochloric acid and
washing by decanta-
tion. A platinum
crucible with bottom
perforated with nu-
merous fine holes is
fitted to the upturn-
ing end of a piece of
soft rubber tubing,
the other end of
which is stretched
over the top of a
' ' f l funnel as shown in
fig. 99, and the neck of the latter passes through the
stopper of a vacuum flask in the usual way. The vacuum
pump having been put in action, a little of the prepared
asbestos suspended in water is poured into the crucible,
and the pump is attached, when the asbestos at once
assumes the condition of a firm compact layer, which is
washed with ease under the pressure of the pump.
After washing the felt, the crucible with the felt ad-
hering is removed from the funnel, ignited, cooled, and
weighed as usual ; then set again in place in the rubber-
holder, taking care that the pump is working, and the liquid
to be filtered is poured into the crucible. The crucible,
felt, and adhering precipitates are, after washing, ignited,
or merely dried, as the case demands, with no further care
than the nature of the precipitate itself necessitates.
During the ignition of precipitates it is frequently
necessary to prevent the contact of reducing gases with
the perforated bottom, and this is accomplished by placing
the crucible either upon a platinum crucible cover or
within a second crucible, or better by using a form of
CARBON ETC. IN IRON AND STEEL. 363
crucible which is conical in shape and provided with a
cap removable at pleasure, to cover the bottom during
ignitions. When a larger filtering surface than the bottom
of a crucible affords is needed a cone of platinum, per-
forated and fitted to a rubber seat, is substituted for the
crucible.
7. Estimation of Carbon, Sulphur, Silicon, Phosphorus,
fyc., in Metallic Iron and Steel.
Boring and Sampling. Samples of metallic iron for ana-
lysis are always obtained by boring, and the proper sam-
pling of these borings is as difficult a task as the corresponding
one of sampling a heap of ore. As is well known, cast-iron
borings are a mixture of small particles of iron with more
or less of finely divided graphite separated from the sur-
faces of these small particles during the process of boring.
The amount of graphite thus mechanically mixed is in
all cases quite large enough to cause serious difficulty
when the problem is to obtain an average sample of a
given lot of borings for carbon estimation. This diffi-
culty arises from the necessity of obtaining a uniform
mixture of the heavy and comparatively coarse borings
with the finely divided and light graphite, and of remov-
ing a sample for analysis without disturbing the uniformity
of this mixture.
The difficulties here presented have been investigated
by Mr. Porter W. Shinier. It was first sought to over-
come the difficulty by having the borings made very fine,
so that they might have more nearly the size of the par-
ticles of graphite. It was not practicable, however, to
secure this uniformity of size, since much of the graphite
is in the form of the finest dust. At all events, the dupli-
cate estimations of carbon in these fine borings, made
with every precaution by combustion in oxygen, frequently
showed differences too large to be accounted for in any
other way than by imperfect sampling.
Very coarse borings were also tried, with the idea that
the graphite separated from these large and heavy pieces
364 ASSAY OP IRON ORES IN THE WET WAY.
would perhaps be inappreciable ; but it was found that
quite enough was separated to vitiate the results.
Various methods of mixing were tried, both in the
bottle and on glazed paper ; but in all these mixtures it
was found that duplicate samples seldom contained the
same proportion of the mechanically mixed graphite ; and
when they did there could be no certainty that it was the
true proportion. All this was shown many times over by the
failure of duplicates to agree closely, the difference being
sometimes as much as O2 per cent. It was finally clear
that nothing was to be hoped from these methods. Even
if we could secure a perfect mixture containing in every
part its due proportion of graphite, it would run a great
risk of being destroyed the moment a spatula was inserted
to remove a sample ; for the least agitation causes some
of the graphite to fall through between the coarser bor-
ings.
The sampling difficulty was finally overcome in this
way : The borings were carefully poured out into a large
porcelain crucible or dish, and enough alcohol is added to
merely moisten them (2 c.c. alcohol to 450 grains). Mix
thoroughly for about five minutes. Bernove with a spatula
as many samples of the moist borings as are needed, weigh-
ing them out roughly so as to get approximately the weights
desired. Dry off the alcohol and weigh accurately* The
samples thus obtained perfectly represent the average of
the original sample, for the alcohol simply serves tempo-
rarily to hold the graphite where it belongs, upon the
surface of the borings.
As the alcohol evaporates from the original sample
the graphite again falls away from the borings, and the
moistening must be repeated when new samples are to be
removed.
The following total carbon results show the im-
provement :
i. ii.
Mixed dry in the bottle . > . > . 3'68 3'84 p.c.
Moistened with alcohol and mixed . . 3'97 3-99
The first results are an extreme case of unsatisfactory
duplicates ; and they also show the tendency of the gra-
CARBON, SULPHUR, SILICON, PHOSPHORUS, ETC. 365
phite to go to the bottom of the bottle when mixed
dry. The mechanically mixed graphite, approximately
estimated by removing the iron by means of a magnet from
samples taken moist, was 0-76 and 0-79 per cent, in dupli-
cate estimations.
After shaking up the dry borings a few times, and re-
moving nearly all from the bottle, the remaining sample
contained 1*29 per cent, mechanically mixed graphite.
The other estimations, such as phosphorus and silicon
(iron excepted), are perhaps not appreciably affected
when samples are taken dry. But whenever samples
are thus taken, the borings which remain are for ever
vitiated for carbon estimations which shall truthfully
represent the original sample. In very careful work,
therefore, and especially when there is any prospect of
subsequent carbon estimations to be made, it is neces-
sary to moisten the borings when weighing out samples
for phosphorus and silicon, or to have two sets of borings
made, one of which is to be reserved for carbon esti-
mations. For iron estimations the samples should always
be taken moist.
It was suggested by Mr. Frank Firmstone, Superin-
tendent of the Glendon Iron Works, to determine whether
or not there is any appreciable loss of graphitic dust while
making cast-iron turnings according to the method devised
by him, in which the turnings are taken equally from the
whole section of a piece of pig iron, the outside being first
removed. Through the kindness of Mr. Firmstone, Mr.
Shinier was enabled to procure from the same piece of
pig iron a sample of turnings taken dry, a second sample
kept moist with alcohol to prevent loss of dust, and several
sections -^ inch in thickness.
The following are the total carbon results from these
samples :
I. Turnings taken dry, but sampled moist.
II. Turnings taken moist.
III. Thin section.
i. ii. in.
4-132 4-122 4-031
4-123 4-123 4-038
4-111 4-103 4-108
3G6 ASSAY OF IRON ORES IN THE WET WAY.
The results under I. and II. show that, with care in
catching the turnings, and avoidance of draughts of air,
there is no appreciable loss of graphitic dust. The first
two results on the thin section are low, because of in-
complete decomposition of the iron, when, after forty
hours, hydrochloric acid was added to complete the solu-
tion. The third result is about right ; but complete decom-
position was obtained in this case by frequent and long-
continued stirring. While, therefore, it is possible to
obtain the correct total carbon by use of a thin section, it
offers no advantage in point of time, since borings may
be dissolved in neutral and cold copper and ammonium
chloride solution in fifteen to twenty minutes by constant
stirring.
Some chemists, with a keen appreciation of the sam-
pling difficulty, have proposed to overcome it by using
clippings of the iron from -J to y 1 ^ inch in thickness.
While it is no doubt possible, in most cases, to obtain the
true result by this means, it is open to the objection of
being very slow when these clippings are to be dissolved
in neutral copper and ammonium chloride solution. Fur-
thermore, the weight of iron taken for carbon estima-
tion in cast iron being limited by convenience to about
50 grains, a very few clippings make up this weight,
and these represent only those parts of the sample from
which they happen to come. While there may not be an
appreciable difference in the composition or in the condi-
tion of the carbon of the different parts of a section of pig
iron, it is yet possible that there may be, and in this case
it is easy to be on the safe side.
When turnings are taken equally from the whole sec-
tion of a piece of pig iron, and are then moistened with
alcohol and thoroughly mixed, every part of the section
has its due representation in a sample of 50 grains.
It may not be out of place to add that, in sampling ores,
and, in fact, any similar mixture of particles of different
size, composition, and specific gravity, it is advisable to
moisten with water before mixing. This secures a uniform
mixture of coarse and fine parts from which a true sample
ESTIMATION OF THE TOTAL CARBON. 367
can easily be removed, prevents loss of dust, and is in every
way more satisfactory than mixing dry.
In pulverised ores it is very common for the finest part
to consist principally of lighter gangue material ; and the
problem of securing a true sample of such a mixture is
similar to that of obtaining a true sample of a mixture of
cast-iron borings and graphite.
Carbon. This is a problem of considerable difficulty,
and to secure accurate results many special precautions are
necessary, owing to the large preponderance of the iron
over the carbon present. The carbon may be present in
two forms as combined carbon, and as free or graphitic
carbon. The estimation may be of the total carbon
present, or of either the combined or graphitic separately,
and the method of analysis adopted will have to be selected
accordingly. The following is a description of the most
satisfactory processes which have been devised for these
estimations.
A. Estimation of the Total Carbon. Mr. Addison B.
Clemence employs the following method for estimating
carbon in steels. The cut (fig. 100) shows the form of
apparatus adopted. It is made of platinum.
The following is the process :
Dissolve from 50 to 80 grains of the borings in double
chloride of copper and ammonium, using 560 grains of
the salt to 40 ounces of water for 46 grains of steel. After
the separated copper is completely dissolved, filter on to
a plug of asbestos placed at 5, and wash thoroughly with
hot water. Any carbon that adheres to the sides of the
tube may be swept down by moistened asbestos. The tube
is then placed in an air-bath, and dried at a temperature
of from 350 to 380 F. for about one hour. A hard
rubber cork through which is passed a glass tube is in-
serted at e, the oxygen gas passing from c to a. Around
the tube at c is a single thickness of filter-paper, about
two inches wide, kept wet by a stream of water supplied
by a reservoir on a shelf above. Heat is applied at b for
one half-hour, at the end of which time the potash bulb
is ready for the balance. One Bunsen burner is sufficient
368
ASSAY OF IKON ORES IN THE WET WAY.
for the combustion. The same precautions are taken
to dry the gas before entering the platinum tube, as
well as before entering the potash bulb, as in the case
with the porcelain tube method.
; Six burners of a combustion furnace
will consume 14 ft. of gas in one half-
hour, while one Bunsen burner will con-
sume 2 ft. in the same time.
The following table shows some re-
sults obtained by the " platinum tube "
process :
A is a Swedish Bessemer iron contain-
ing 0*10 per cent, of carbon ;
B, an American Bessemer, with 0-18
per cent. ; and
C, an American Bessemer, with 0'50
per cent. ;
all obtained by the "porcelain tube"
process.
<----*
A.
0-09
0-10
0-10
0-10
B.
0-18
0-18
0-16
0-17
c.
0-49
0-51
0-49
0-44
Weyl's very ingenious method for the
estimation of the total carbon is founded
upon the fact that a piece of iron at-
tached to the positive pole of a galvanic
battery, and suspended in hydrochloric
acid, is dissolved, while the hydrogen is
given off at the negative pole. The for-
mation of hydrocarbons, and a consequent
loss, is in this manner prevented. One
great advantage in this method is that
the iron does not require to be in pow-
der. A piece of iron 2 to 4 grammes in
weight, attached to the positive pole of
** a Bunsen's cell, is suspended in dilute
hydrochloric acid, just below the surface of the liquid.
From the negative pole hydrogen passes off, while the
ESTIMATION OF GRAPHITE. 369
iron dissolves quite quietly, and the strong solution of
ferrous chloride formed may be seen falling in a regular
stream through the lighter liquid. The iron is dissolved
in about twenty-four hours, and the carbon is left behind
in the same shape as the piece of metal from which it
was derived. In Weyl's earlier experiments it was found
that some of the liberated carbon at the positive pole
was carried over to the negative pole by the mechanical
working -of the stream. To prevent this, a diaphragm
of bladder or parchment paper is interposed between
the two, which entirely obviates the possibility of loss
in this way.
B. Estimation of the Graphite. Dr. Eggertz reduces
1 gramme of iron to small pieces, mixes with 5 grammes
of pure iodine and 5 c.c. of water in a small flask, covered
with a watch-glass, and placed in ice-cold water before
adding the iron. It is to be kept for twenty-four hours at
C., and frequently stirred meanwhile. By keeping the
liquid cold, no carburetted hydrogen is produced. The
greater the amount of silicon in the iron the greater is
the tendency to the production of carburetted hydrogen.
The residue of carbon and silica left after the iron is dis-
solved is collected on a filter of known weight, when it is
dried at from 95 to 100 C., and washed thoroughly with
hot water. After twelve hours, it is to be washed with
a mixture of hydrochloric acid and twice its volume of
water, heated to 70 or 80 C., until the filtrate ceases to
give a blue colour with ferrocyanide solution. The object
of leaving the residue for twelve hours is to allow any
small particles of iron remaining undissolved by the iodine
being oxidised by atmospheric air, and prevent disengage-
ment of hydrogen when the hydrochloric acid is added.
After the hydrochloric acid is washed out of the filter, it
is dried, with its contents, at 95 to 100 C., until constant
in weight. This weighing gives the amount of the carbo-
naceous residue and silica (but not the whole of the silica,
because some part of it would have been dissolved), and
by burning the carbon away and weighing the silica the
weight of the carbonaceous residue may be ascertained.
B B
370 ASSAY OF IRON ORES IN THE WET WAY.
The carbonaceous residue may consist either of graphite
or of the compound of carbon, iodine, and water already
mentioned, if the carbon was combined with the iron. To
ascertain which is the case, 1 gramme of iron is dissolved in
15 c.c. of hydrochloric acid (1*12 density) in a flask covered
with a watch-glass, and, when the iron is dissolved, the solu-
tion is boiled for half an hour. All the carbon combined with
the iron is disengaged in the form of carburetted hydrogen
gas, while the graphite and silica remain. If the carbona-
ceous residue left after dissolving the iron comes in con-
tact with atmospheric air before the liquid is boiled, it is so
altered that it is not dissolved and disengaged as gas. The
graphite that remains after boiling the liquid is collected on
a filter of known weight, washed, dried, and weighed. It is
then burnt, and the residual silica weighed to ascertain the
quantity of graphite. Very satisfactory results have been
obtained by this method. The differences do not amount
to more than Ol per cent. When the quantity of carbon
to be estimated is very small, more than 1 gramme of iron
must be used in the analysis.
Mr. Tosh gives the following process for the estimation
of graphite. Two to three grammes of iron are treated
with dilute hydrochloric acid, and when the solution
approaches completion a considerable quantity of strong
acid is added to separate the last portions of iron and
manganese. The insoluble matter, consisting mostly of
graphite, is collected on a carefully weighed filter, washed
with hot water, dilute hydrochloric acid, solution of caustic
soda, and hot water again, successively, and lastly with
alcohol and ether to remove oily hydrocarbons. (By wash-
ing with dilute acid and with alkali, the iron and silica or
oxide of silicon are separated.) After drying at 120 C.,
the filter and graphite are weighed, and burned away.
The small residue (a mere trace of silica or titanic acid) is
weighed, and this weight subtracted from the first gives
the amount of graphite. The results obtained agree very
closely.
In washing the graphite with solution of soda, there is
always a brisk effervescence, due to the oxidation of oxide
ESTIMATION OF COMBINED CARBON. 371
of silicon to silicic acid, by decomposition of water, with
consequent liberation of hydrogen.
C. Estimation of Combined Carbon. When steel, or
pig-iron containing carbon in chemical combination, is dis-
solved in nitric acid, a soluble brown colouring matter is
formed, whose colouring power is very intense, and the
solution assumes a tint which is dark in proportion to the
quantity of the chemically combined carbon. Iron and
graphite (or free carbon) do not influence this colouration ;
for the solution of nitrate of iron is only slightly greenish,
unless extremely concentrated, and graphite is insoluble in
nitric acid.
Thus, in dissolving two pieces of different steels of the
same weight in nitric acid, taking care to dilute the darker
solution until the two liquids present exactly the same
colour, it is very evident that the more highly carburetted
steel will furnish the larger quantity of liquid, and that the
proportion of the volumes will indicate the relative pro-
portion of colour in the two steels. If now the composi-
tion and content of carbon of one of the steels is known,
the absolute percentage of carbon in the other steel may
be immediately deduced. Dr. Eggertz has applied these
reactions to a method of estimating the combined carbon.
To obtain trustworthy results certain precautions must be
taken.
In a cylindrical test-tube, dissolve gradually in the
cold 10 centigrammes of wrought-iron, steel, or cast iron,
reduced to a fine powder, in 1-J to 5 c.c. of nitric acid of
1*2 sp. gr. The use of nitric acid containing hydrochloric
acid must be avoided, because the solution of iron would
have a yellow tint. In proportion as the metal contains more
carbon, more nitric acid must be used. After some time,
when the chief part of the metal appears to be attacked,
place the tube in a water bath, and warm it to 80 C., in
such a position that only the lower part of the tube is in
contact with the warm water ; a movement takes place in
the acid, which favours its reaction upon the metal; a
slight disengagement of carbonic acid from all the particles
of carbon may be observed. The operation should always
B 2
372 EGGERTZ'S COLOUR TEST FOR THE
be conducted under the same circumstances as to heat and
length of time. The evolution of gas having ceased (in
operating upon steel, the reaction must continue two or
three hours), place the tube in a large vessel filled with
water, to bring the solution always to the same temperature.
This precaution is indispensable, because the same liquid is
darker when warm than when cold. Afterwards, pour off,
as exactly as possible, the clear liquid into a graduated
burette. Upon the black residue remaining in the tube
pour some drops of nitric acid, and heat carefully over a
lamp. If there is no further liberation of gas, the residue
consists of nothing but graphite or silica. Cool the new
solution, and add it to that which is already in the
burette.
The liquid is then diluted with water until its colour
corresponds exactly with that of the normal liquid, which
latter should be of such a degree of concentration that
each c.c. represents O'OOOl grm. of carbon. If, for instance,
this normal liquid is prepared from cast steel containing
exactly 0*85 per cent, of carbon, 1 decigramme of that
steel must be dissolved in 8*5 c.c. of nitric acid ; lOOgrms.
of steel containing 85 centigrammes of carbon would thus
be dissolved in 8500 c.c. of the normal solution, 100 c.c.
of that solution would represent 1 centigramme of carbon,
and, consequently, 1 c.c. of the normal solution would
represent O'OOOl grm. of carbon. To compare the normal
solution with the solution of iron under examination, it
should be contained in a tube of the same kind, and when
the two tubes are held together by daylight before a thin
sheet of paper, the colour should be exactly the same in
both of them. For the preparation of normal colour solu-
tions see pp. 382-388.
The Eggertz colour test for estimating carbon in steel
is perhaps used more extensively commercially than any
other analytical process. This is particularly the case in
Bessemer steel works, where every blow is tested to ascer-
tain its percentage of carbon. The number of estimations
thus amounts even in a small works to 25 daily.
In carrying out the analytical process, which is too
ESTIMATION OF CARBON IN STEEL.
373
FIG. 101.
well known to require description, it is of course necessary
to preserve the identity of each blow-number ; and this is
effected by dissolving the steels in a square shallow copper
water-bath provided with a lid perforated with holes in
which the tubes are inserted in a standing position, the
first hole indicating the starting-point of the day's blows.
This form of bath is objectionable from several points
of view.
Mr. J. Oliver Arnold has consequently designed a
bath which two years' constant use has proved to have
none of the disadvantages
connected with the old form
of bath.
The bath consists of a
cylinder of glass 5 inches high
by 5 inches in diameter, closed
at one end to contain the
water ; this is heated on an
iron pi ate or sand-bath nearly
to boiling-point (90-95).
On the top is a perforated
cover of glazed white earth-
enware ; this cover contains
17 holes, numbered 1 to 15
in burnt-in black figures,
and the other two holes are
marked S for the standard
steels.
The tubes containing the
steels and nitric acid are suspended in the water to the
proper depth by slipping on the tube a rather tightly
fitting cylinder of india-rubber f inch long ; this india-rubber
serves the double purpose of keeping the ends of the tubes
off the bottom of the bath, bumping thus being avoided,
and of preventing the bobbing up of the tubes by flota-
tion (see figure).
The apparatus, being of glass and glazed earthenware,
can be washed perfectly clean every day. The tubes are
suspended vertically in the water to any desired depth.
374 ESTIMATION OF CAEBON IN STEEL.
The clear numbering enables the selection of any given
blow about which 4 a hastener ' has been received with-
out chance of error, and the progress of the solution of
the steels in the tubes can be watched without removing
them.
Except from mechanical breakage the apparatus is
practically indestructible. In cases where the daily car-
bons exceed 15, a second and third bath may be employed,
the numbering following on, or one large bath could be
made ; this, however, might prove inconveniently large.
In cases where only a few odd carbons are required a
smaller bath perforated for 6 steels and a standard would
suffice.
Several modifications of Eggertz's method have been
proposed. The most successful of them, affording exceed-
ingly accurate results, was communicated to the ' Journal
of the Franklin Institute ' for May 1870 by J. Blodget
Britton, and has been tested for a considerable time. In-
stead of a single tube, containing a standard solution for
comparison, as suggested by Eggertz, a number of tubes-
FIG. 102.
A
8aJJjJLIlLttl
having their solutions differently standardised, one from
the other, are employed. These are arranged securely in
a wooden frame, with spaces between for placing the tube
containing the solution to be tested, and forming together
a convenient portable instrument called a colorimeter a
representation of such an instrument being shown in the
annexed cut. The position of the tube containing the
solution to be tested is shown at A. The tubes are fths
of an inch in diameter, and 3-| inches in length, filled with
water and alcohol coloured with roasted coffee, and herme-
tically sealed. The solution in the tube to be left has its
colour to correspond exactly with one produced by 1
BRITTOIs T S METHOD. 375
gramme of iron, containing 0'02 per cent, of combined
carbon, dissolved in 15 c.c. of nitric acid. The solution
in the tube next to it has its colour to correspond with
one produced by the same quantity of iron, but containing
0-04 per cent, of combined carbon ; and so with each of
the other tubes, increasing 0*02 per cent, of carbon in
regular succession to the right, the last reaching 0*3 per
cent., as indicated by the figures on the upper rail of the
instrument. On the back of the instrument, and for
the purpose of partially screening the light and allowing
the different shades of colour to be distinctly discerned,
there is tightly stretched between the rails some fine white
parchment paper. This screen is not shown by the cut,
but it serves a very important purpose. The process is
conducted as follows : 1 gramme of the finely divided
metal is put into a tube of about 1-J inch in diameter
and 10 inches long, and digested for fifteen or twenty
minutes in 10 c.c. of nitric acid of a little more than 1-20
specific gravity, free from chlorine. The solution is then
cautiously poured into a beaker, and a small portion of
metal which remains undissolved and adheres to the bottom
of the tube is treated with 5 c.c. of fresh acid, exposed to
a gentle heat till completely dissolved, and added to the
other. The contents of the beaker, when sufficiently cool,
are filtered through two thicknesses of German paper (not
previously moistened, and of a diameter not exceeding 4^
inches) into a tube about 5 inches long and of precisely
the same diameter as those in the instrument. After the
filtered solution has remained for some minutes at the tem-
perature of the atmosphere, and its colour become fixed,
the tube is placed in the instrument and the carbon esti-
mated by a comparison of shades : the estimation may
be made readily as close as O'Ol per cent. Heat should
not be applied in the first instance to facilitate the solution
of the metal, because a high temperature is apt to cause a
slight loss of colour. Two thicknesses of paper are taken,
because one alone is liable to break ; and the paper should
be used dry, for, if previously wetted, the water will
weaken the colour of the solution ; and it ought to be cut
376 ESTIMATION OF CARBON IN STEEL.
to a size not exceeding 4-^ inches, to prevent undue absorp-
tion. If the metal to be examined contains more than
0-3 per cent, of carbon, O5 gramme, or less of it, may be
taken, or the solution may be diluted with an equal volume
or more of water and the proper allowance made ; or an
instrument of higher range may be used. On the other
hand, if the metal contains a very small percentage of
carbon, 2 grammes of it may be taken. For preparing the
standard solutions (the normal ones begin to lose colour
after some hours), caramel dissolved in equal parts of
water and alcohol, as suggested by Eggertz, answers well ;
but with roasted coffee as the colouring matter the true
shades may be obtained. (See also pp. 382-88.)
An important paper by Mr. J. E. Stead was read in
1883 before the Iron and Steel Institute on a new method
of estimating minute quantities of carbon in iron and steel,
and on a new form of chromometer.
The following a,re extracts from Mr. Stead's paper :
1. ESTIMATION OF MINUTE QUANTITIES OF CARBON. As is
well known, it is impossible to estimate with accuracy
minute quantities of carbon by the ordinary colour method,
owing to the colour of the nitrate of iron present, which
interferes so as to make it impossible to judge of the colour
due to carbon.
Having been engaged in some careful investigations on
the nature of the colouring matter which is produced by
the action of dilute nitric acid upon white iron and steel,
it was found it had the property of being soluble in
potash and soda solutions, and that the alkaline solution
had about two and a half times the depth of colour pos-
sessed by the acid solution. This being so, it was clear
that the colouring matter might readily be separated from
the iron, and be obtained in an alkaline solution, by
simply adding an excess of sodium hydrate to the nitric
acid solution of iron, and that the colour solution thus
obtained might be used as a means of estimating the
amount of carbon present. Upon trial this was found to
be the case, and that as small a quantity as 0-03 per cent,
carbon could be readily estimated.
STEzVDS METHOD. 077
The method as now in use is conducted as follows :
Standard solution of nitric acid, T20 sp. gr.
Standard solution of sodium hydrate, 1*27 sp. gr.
One gramme of the steel or iron to be tested is weighed
off and placed in a 200 c.c. beaker, and, after covering
with a watch-glass, 12 c.c. of standard nitric acid are
added. The beaker and contents are then placed on a
warm plate, heated to about 90 or 100 C., and then
allowed to remain until dissolved, which does not usually
take more than ten minutes. At the same time a stan-
dard iron containing a known quantity of carbon is treated
in exactly the same way, and when both are dissolved 30
c.c. of hot water are added to each, and 13 c.c. soda
solution.
The contents are now to be well shaken, and then
poured into a glass measuring-jar and diluted till they
occupy a bulk of 60 c.c. After again well mixing and
allowing to stand for ten minutes in a warm place, they
are filtered through dry filters, and the filtrates, only a
portion of which is used, are compared. This may be
done by pouring the two liquids into two separate measur-
ing-tubes in such quantity or proportion that upon looking
down the tubes the colours appear to be equal.
Thus if 50 m.m. of the standard solution is poured
into one tube, and if the steel to be tested contains, say,
half as much as the standard, there will be 100 m.m. of
its colour solution required to give the same tint. The
carbon is therefore inversely proportional to the bulk
compared with the standard, and in the above-assumed
case, if the standard steel contained 0-05 percent, carbon,
the following simple equation would give the carbon in
the sample tested :
Q-" = 0-025 per cent.
100
Experiments were made upon a steel which contained
0-11 per cent, carbon, to ascertain what the influence
would be of heating the nitric acid solution for an increas-
ing length of time after dissolving on the bath, and it
378 ESTIMATION OF CARBON IN STEEL.
was found that the carbon colour was not materially
affected by heating the acid solution twice as long as was
necessary for completely dissolving the carbon compound,
and that although the iron is dissolved in five minutes
it is evident that some of the carbon compound at first
formed escapes solution in that period.
The next point was to ascertain what effect the use of
an excess of nitric acid in dissolving the steel w^ould have
on the colouring matter. It was seen that 6 c.c. acid in
excess does not materially affect the estimation, but when
this is exceeded the colour is reduced in quantity.
It now became important to know if a greater or less
quantity of soda solution would have a different solvent
power on the colouring matter.
To ascertain this, four separate portions of the soft
steel were treated alike in dissolving, but to the solutions
different quantities of soda solution were added. It was
here seen that, as before stated, 13 c.c. sodium hydrate solu-
tion is capable of effecting solution of the colouring matter.
By using a less amount, however, by experiment it was
found that the colour is precipitated with the iron oxide.
It is very well known that, in the old colour
method, very slight traces of hydrochloric acid, if present,
alter the character of the colour to such an extent as
to make the colour estimation unreliable. It therefore
was of interest to ascertain if the same would occur in
the alkaline method. Four portions of steel were treated
as usual, excepting that to one portion a single drop of
hydrochloric acid was added when being dissolved, to a
second five drops, and to a third ten drops, but to the
last portion no hydrochloric acid was added.
Th.e following are the results obtained, viz. :
Hydrochloric Acid.
1 drop 5 drops 10 drops None
Carbon . 0-105 p. c. 0-090 p. c. 0-078 p. c. 0-110 p. c.
Second Test.
1 drop 5 drops 10 drops None
Carbon . 0-356 p. c. 0*338 p. c. 0-324 p. c. 0-410 p. c.
The colour in each case, and even in that in which the
larger quantity of hydrochloric acid was added, was the
STEAD S METHOD.
379
same in quality although differing in quantity, showing
(1) that the presence of chlorides is harmless, and (2)
that nitro-hydrochloric acid, even in small quantities, pre-
vents the formation of the full amount of colouring matter
capable of being produced by nitric acid alone.
A large number of samples of low carbon iron have
been examined by the alkaline method, including samples
of iron taken from the Bessemer converter at the end of
the blow before any addition of spiegeleisen. The results
are given here
Blown Iron taken from the Bessemer Converter.
No. 1 0-040 p. c. carbon.
No. 2
No. 3
No. 4
No. 5
No. 6
Average
Average by combustion
Standard soft steel .
Pure iron wire
Cleveland iron ship plates
Ditto ditto
The colour solutions from these low carbon irons are
different in tint from those obtained from the higher car-
bon steels ; and it is important that a low carbon iron be
used as a standard for comparison.
When high carbon steel is heated to redness and chilled,
it is well known that the colour from the chilled steel is
very much less in quantity than that from the same steel
before hardening. The difference, however, is not nearly
so marked when there is little carbon present in the steel,
as was proved by the following results, viz. :
Several samples of iron and steel after being drilled
were heated to redness and chilled in water, the results
before and after being as follows :
0-036
0-045
0-039
0-061
0-048
0-045
0-048
New Colour
Method Combustion
0-120 p. c.
carbon. 0*122 p. c.
0-038
,
0-055
>
0-035
Soft steel = soft 0-168 p.
chilled in cold water 0-158
chiUed in hot water . 0-168
Staffordshire square iron bar . O'llO ,
chilled 0-100 ,
flat . 0-069 .
chilled 0-069 ,
Soft steel 0-077 :
chilled. . . 0-071 ,
Difference
>0-010 p. c.
0-010
None.
0-006
380 ESTIMATION OF CARBON IN STEEL.
It is not often that soft iron or steel is chilled before
being placed in the hands of the analyst, but it is satisfac-
tory to know that even if they were the results by colour
would not be rendered useless.
When using the new method Mr. Stead found that
some steels give a much yellower colour than others, and
in course of investigation he discovered that there are
present in all nitric acid steel solutions two distinct colour-
ing matters, which have been separated arid obtained in a
nearly pure state, one of which is bright yellow, resembling
potassium chromate, the other being of a dark brown-red
colour. In some steel solutions the yellow colour pre-
ponderates, and in others the brown.
2. A NEW FORM OP CHROMOMETER. In comparing
colour solutions there are two methods of procedure.
The first is that generally adopted in making estima-
tions of the carbon by the acid colour process, in which
the darker solution is diluted with water until the
colours of the two solutions are equal in density that is
to say, until the colour is equal per cubic centimetre.
The diluted volume is then noted, and the amount of
carbon read off in 0*1 per cent. c.c. In the alkaline
method it is better to use the second plan, already
described, of comparing directly the relative density of
the colour solution without dilution, and ascertaining
the lengths of the two columns of liquids, which, when
examined from the surface, give the same depth of colour.
The carbon in this process is, as compared with the stan-
dard, inversely proportional to the length of the liquid
column.
If a fixed length of liquid column be used of the solu-
tions of carbon and a variable standard column, then, by
using a suitable standard solution, the carbon may be de-
duced from the length of the latter required to make a
colour column equal in depth to the former and the per-
centage read directly from a graduated scale. The instru-
ment arranged by Mr. Stead is made on this principle; it is
extremely simple and easily constructed. It consists of
two parallel tubes, which may be of any suitable diameter,
STEAD'S CHROMOMETER. 381
one of which is contracted at a point 9 inches from the
top, and is open at both ends. The lower end passes
through an india-rubber cork to the bottom of a 4-oz.
bottle, which contains the standard colour solution. A
second tube of smaller diameter also passes through the
cork into the bottle, the outside end of which is in com-
munication with a large syringe.
Just above the contracted part of the first-mentioned
tube a small glazed cylinder of china clay rests. By
pressing the syringe the liquid can be forced from the
bottle below up this tube. The second tube is about 9
inches long, and is closed at the lower end. At this end a
small glazed clay cylinder is also placed. When this tube
is placed parallel to the first, the length from the open
upper ends to the flat surfaces of the clay prisms is equal
in each. A small looking-glass at an angle of 45 is fixed
above the open ends of the tubes, and the standard tube
is graduated into O'Ol part to O15 part.
The method of working with the apparatus is very
simple. The colour solution to be compared is placed in
the second tube, with which it is filled up to a certain
fixed mark.
It is only now necessary to squeeze the syringe and
force the liquid up the first tube until the colours in the
two columns are equal, as can be seen by looking into the
mirror above. The height of the standard solution is read
off on the graduated scale, which will be the percentage
of carbon in the steel or iron under examination.
The preparation of Inorganic Standards for the
Colorimetric Carbon Test.
Wherever the amount of work renders it practicable
the plan of using permanent standard solutions in con-
nection with the colorimetric carbon test affords such
manifold advantages that it is to be strongly recommended.
That it has not attained a wider and more general applica-
tion is mainly due to the fact that, as the method is
generally employed, the difficulty attending the produc-
382 ESTIMATION OF CARBON IN STEEL.
tion of the colours, and their doubtful stability when
produced, have been only too evident to those who have
undertaken their composition. The old method of using
standard solutions made from sugar, coffee, &c., has
always been hampered by the tendency of the colours to
change upon exposure to the light, and the consequent
liability to serious error unless they are closely watched.
Appreciating these difficulties while recognising the de-
sirability of permanent colours, Mr. Magnus Troilius, at
the suggestion of Professor F. L. Erckmann, investigated
the properties of cobalt, copper, and ferric chlorides with
the view of determining their efficacy in the production
of standard solutions. His results seem to have been
satisfactory, but his treatment failed to eliminate many
of the objections that have rendered the use of organic
colours undesirable. The following mode of procedure
in the making-up of inorganic standards has been found
by Mr. W. Eobinson, of Joliet, Illinois, to be most effi-
cient in point of accuracy, ease, and celerity of production
and stability.
DAY STANDARD COLOURS.
Mr. Eobinson uses as a range in illustration the car-
bon-contents included between the extremes 0*5 per
cent, arid 0-3 per cent, as representatives of Bessemer
rail steel that branch of the industry in which the uni-
formity and continuity of the product renders the use of
permanent standard solutions specially advisable. The
principles formulated below will evidently apply to either a
higher or lower carbon range. The salts employed are the
same as those used by Mr. Troilius, viz. the neutral chlorides
of cobalt, copper, and iron. Solutions of these are made
as follows :
Cobalt Chloride. Dissolve in water containing 1 per
cent, of free hydrochloric acid (1*12 specific gravity) in
the ratio of 1 gramme salt to 1 c.c. water.
Copper Chloride. Dissolve in water containing 1 per
cent, hydrochloric acid (1-12 specific gravity) in the ratio
of 1 gramme of the salt to 1^ c.c. water.
PREPARATION OF STANDARD COLOURS. 383
Ferric Chloride. Dissolve in water containing 2 per
cent, hydrochloric acid (1-12 specific gravity) in the ratio
of 1 gramme salt to 1 c.c. water.
The solution in all cases is best effected by the appli-
cation of a gentle heat ; and filtering through a ribbed
filter assures freedom from turbidity. Free acid is used
to prevent the otherwise enhanced liability to decomposi-
tion of the chlorides. Should any difficulty be encoun-
tered in obtaining a clear solution of the ferric chloride,
subsidence of the liquid for a few hours, and then decan-
tation, will obviate the trouble.
Mr. Eobinson proceeds : ' Our purpose is now to
make up a set of colours that shall correspond to those
made by the nitric acid solution of steels containing in car-
bon every alternate point between the extremes of 0-5 and
0-3 per cent. It is clearly impracticable to have steel
standards with this gradation of carbon ; while, on the
other hand, to make a single colour comparable with
some one steel, and then by dilution make up others,
corresponding to all the desired percentages, is liable to
introduce an error. I have two steel bars that are kept
as standards, the carbon composition of which, as shown
by repeated combustions, is 0*501 per cent, and 0-304
per cent, respectively. As the limit of error in the
colorimetric test can be considered as about 0-01 per cent,
for this grade of steel, I am safe in calling the first a 0-5
and the second a 0-3 per cent, metal. From the 0-5 and
0-3 standard steels are now made up nitric acid solutions
under the same conditions, as regards amount of steel
taken, acid used, time of heating and temperature, as those
to which the regular tests are to be subjected. I am in
the habit of dissolving 0-5 grammes of drillings in 12 c.c.
nitric acid (1-20 specific gravity), and subjecting it in a
steam-bath to a temperature of 98 to 100 C. for fifteen
minutes.
'The next step is to make up from the chlorides a
colour which shall exactly match the 0'5 per cent, steel
solution, and another which, diluted to J- ths of its strength,
shall precisely correspond to the 0'3 per cent, steel. The
384 ESTIMATION OF CAEBON IN STEEL.
former we will call a brown 0-5 colour, and the latter a
green 0-5 colour. I use the term " green 0'5 " for the
latter, because it will be found to be a colour which,
though of a greener cast, corresponds in intensity, before
diluting, to the solution of the 0-5 steel. These are pre-
pared as follows :
4 For the brown 0*5 colour, mix the chloride solutions
and water containing 0-8 per cent. (1-12 specific gravity)
hydrochloric acid in the ratio of 87 '5 c.c. water, 5*7 c.c.
cobalt chloride, 25 c.c. copper chloride, and 43 c.c. ferric
chloride. For the green 0'5 colour take 82-6 c.c. water
containing 0'8 per cent, hydrochloric (1-12 specific gravity),
5-4 c.c. cobalt, 5 c.c. copper, and 7 c.c. ferric chloride
solutions. The first of the colours so produced will ap-
proximately match the 0-5 steel solution, while the second,
diluted to J-ths of its strength with water containing
0*8 per cent, hydrochloric acid (1*12 specific gravity), will
approximately correspond to the 0-3 steel solution. The
absolute identity of the colours may be obtained by the
addition of the chloride or chlorides, the colour of which
corresponds to the shade that is wanting. The compari-
son of the green 0*5 colour with the 0-3 steel is readily
accomplished by the aid of a reading-tube suitably cali-
brated 6 c.c. and 10 c.c., for instance In the final match-
ing care should be exercised that the same light is used
as is to be regularly worked with the same window if
possible as this assures similarity of conditions as regards
transmitted and reflected light. With correct brown and
green 0*5 colours so obtained the alternate intermediate
points are easily acquired by mixing the two and diluting
with water containing 0'8 per cent, hydrochloric acid
(1-12 specific gravity) by the aid of a burette, as per
following table :
DAY STANDARD COLOURS.
385
Per cent. Carbon
No. c.c. Brown
0'5 colour
No. c.c. Green
0'5 colour
No. c.c. Water
0-5 10
0-48
8-64
0-96
0-4
0-46
7-36
1-84
0-8
0-44
6-16
2-64
1-2
0-42
5-04
3-36
1-6
0-4
4
4
2
0-38
3-04
4-56
2-4
0-36
2-16
5-04
2-8
0-34
1-36
5-44
3-2
0-32
0-64
5-76
3-6
0-3
6
4
This table is based on the following considerations :
1. That theoretically, from a given solution represent-
ing carbon ol known value, other solutions representing
definite lower carbon percentages may be made by dilut-
ing with water.
2. That practically this will give colours of shades at
variance with those aimed at.
The explanation is clear. We have first to decide upon
some definite bulk as a suitable amount of the solutions
to make up. With the 5 inch by -| inch reading-tubes
adopted here, 10 c.c. has been found convenient. To
make the O48 standard, for instance, we must have such
a 0*5 colour that the shade produced upon diluting with
the proper amount of water will match that made from a
0-48 steel. This neither the green nor the brown 0-5
alone will do. The former is too green, the latter too
brown. The operation of determining in the mixture the
correct proportion of the two requisite to give the desired
colour (for the 0'48 colour as an example) is purely a
mathematical one. The computation is as follows :
Amount of green : amount of brown : : 0'02 : 0'18 ; 0-02
and 0-18 being the relative distances of the 0*48 colour
from the extremes, 0-5 and 0'3. The proper ratio of the
brown and green 0*5 colours being found, the amount of
the mixed chlorides and water to make a total of 10 c.c.
is easily calculated thus :
0-5 : 0-48 : : 10 : x
10-9-6 = 0-4
CC
386 ESTIMATION OF CARBON IN STEEL.
Hence 9'6 will be the number of cubic centimetres of
the combined chloride solutions and 0-4 the number of
water. Any of the other gradations in this or any range
is found in a like manner. If it is desirable to make
either a higher or lower, all that is necessary is to accur-
ately know the carbon-contents of two suitable steels and
proceed as above. It is always well to keep the colours
from, exposure to the light as much as possible. When it
is necessary to use the colours constantly, as in testing
every heat at a Bessemer plant, a reading-rack capable of
being closed is to be recommended. This avoids the neces-
sity of removing to a dark place. Such a rack, provided
with a front and back door, swung on hinges fastened to
the bottom, and arranged with a thin plate of fine ground
glass for a reading background, has proved very effective.
For this purpose I have found the glass to possess advan-
tages over the mediums ordinarily employed. The solu-
tions must of course be corked tight. With due regard
to the precautions and directions here laid down, a set of
reliable inorganic standard solutions can be made from
the above chlorides that will be found to be constant for
four or five months of continued use, and often for even
a longer period. When decomposition does set in, the
colours gradually grow dark.
NIGHT STANDARD COLOURS.
As the method is commonly conducted, standards are
weighed from the steel direct. When tests have to be
made after every blow, this is obviously inconvenient, and
renders it unadvisable to leave the work in unskilled hands.
With the purpose of placing the night-tests as nearly as
possible on an equal footing with the day-tests, the follow-
ing investigation was undertaken by the writer. The day
chloride standards, when compared by both kerosene and
gaslight, were found to appear much lighter than the
corresponding steel solutions. The arc light gave the
same result. Monochromatic light as obtained by the
soda bead and various coloured glasses was tried without
NIGHT STAND AED COLOURS. 387
success. The magnesium light gave the same reactions as
daylight, but was laid aside as impracticable, and the idea
of using the day-standards for night work was then aban-
doned. Night-standards that proved entirely satisfactory
were finally made up from the same solutions and on ex-
actly the same principle as the day-standards, with the
simple modification that the brown 0*5 and green 0'5 were
matched with the (>5 and O'o steel solutions in a dark
room, by the aid of kerosene light. To make the brown
0-5 night-standard, mix the solutions in the following
ratio: 7*4 c.c. cobalt, 1-5 c.c. copper, and 4*3 c.c, ferric
chloride solution with 86*8 c.c. water containing 0*8 per
cent, hydrochloric acid (1-12 specific gravity). For the
green 0*5 night-standard, mix in the ratio of 7*7 c.c. cobalt,
2*2 c.c. copper, and 4'5 c.c. ferric chloride solution with
85-6 c.c. water containing O8 per cent, hydrochloric acid
(1-12 specific gravity). The extremes being made up, the
intermediate colours are produced by the same table as
was used for the day -standards. The colour and intensity
of the brown and green 0*5 thus prepared will be approxi-
mately correct. The final shading must be accomplished
in the same manner as for the day-solutions. Colours so
produced will be found to be considerably darker than
the corresponding day- colours, when compared by day-
light.
For this work and for the regular night-tests I use a
rectangular camera, made of sheet iron, arranged on the
back with a door, open in front, and so cut, about midway
between -the ends, as to admit a light wooden rack holding
the colours. Three medium -sized kerosene-lamps, placed
on a stand in front, furnish the light. The rack is pro-
vided with a plate of ground glass as a reading back-
ground, and is so adjusted that a slide between the glass
and the lamps protects the standards from the light when
not in use, while the door on the back of the camera
serves to darken the other end. With standards made as
above, I have found no difficulty in turning out as accur-
ate work by night as by day, -and with the requisition of
no greater skill and ability than is required of any day
c c 2
388 ESTIMATION OF SULPHUR IN IRON AND STEEL.
ca,rbon-boy. In these night-standards it will be noticed
that there does not exist that sharp contrast in shade
between the 0'3 colours produced by diluting the brown
and the green 0*5 as is found in the day-standards. It is
almost entirely a case of comparative intensity, when, un-
like the day-standards, the shade is of but little moment.
The kerosene light seems to eliminate to a great degree
the power of distinguishing variations in the green tint ;
and for this reason a very good set of night-colours could
be made by simply diluting the brown extreme with water
alone. It will not be as accurate a match, however, as if
the green is used also. It takes a little time to familiarise
the eye to night-reading ; but when once accustomed it
has been my experience that as close, if not closer, discri-
minations can be made between the colours by night as by
day. In comparing by dilution, the failure to take into
consideration the distinction of shade as well as intensity
has often been the source of considerable error. With
standard solutions made as directed, this liability is reduced
to a minimum ; and this, with the manifest advantages of
ease and celerity, renders inorganic standards a desidera-
tum for any laboratory dealing with continuous work of
this kind.
Estimation of Sulphur in Iron and Steel. The plan
usually adopted is to dissolve a weighed quantity (about
3 grms.) of the metal in strong nitric acid, adding a little
hydrochloric acid occasionally, and evaporating the solu-
tion to dryness. Dissolve the residue in very dilute warm
hydrochloric acid, and precipitate the sulphuric acid in
the solution by means of chloride of barium. If the pre-
cipitated barium sulphate, after washing once or twice by
decantation, has a yellow or brown colour, owing to the
presence of iron mechanically carried down, heat it, before
filtering, with dilute hydrochloric acid.
Dr. Eggertz, to whom analytical chemistry is indebted
for the colorimetric process of estimating combined carbon
in iron and steel, has devised a very expeditious plan
for estimating the sulphur. He takes one decigramme of
cast iron, wrought iron, or steel, cut up or pulverised, and
EGGERTZS METHOD. 889
passed through a sieve with holes not larger than 0*6 m.m.,
and introduces it, by means of a glass or glazed-paper
funnel, into a flask about 0-15 metre high and 5 centimetres
diameter, previously containing 1 grm. of water and 0*5
grm. of concentrated sulphuric acid ; or, in preference,
1-5 grin, of sulphuric acid, sp. gr. 1-25, and whose volume
(1-5 c.c.) has been marked on the flask. A piece of
polished silver plate (18 m.m. long, 7*5 m.m. wide, and 1
m.m. thick, with a hole at one end), composed of 75 per cent,
of silver, 25 per cent, of copper, and attached to a thin
platinum or silver wire, is quickly introduced into the
flask, so that it may be a little below the neck ; a cork is
put in so as to hold the wire without completely closing it.
It is allowed to stand fifteen minutes at the ordinary tem-
perature, and the silver plate is then removed. If the iron
contains sulphur, the plate is coloured by the sulphuretted
hydrogen gas disengaged during the solution of the iron
in the dilute sulphuric acid ; and, according to the amount
of sulphur present, the coloration of the plate passes to
a coppery yellow, a bronze brown, a bluish brown, or a
blue. These colorations are estimated with the greatest
accuracy especially that of the silver plate alone, No. 1 ;
that of coppery yellow, No. 2 ; that of bronze brown, No.
3 ; that of blue, No. 4. The intermediate degrees may be
represented by decimals, thus : 2 ! 5 if the coloration is
between 2 and 3; 3 p l if the plate is one-tenth towards the
blue ; 3*5 if it is as much blue as brown ; 3*9 if the brown
coloration is feeble.
As the normal coloration, No. 2, Dr. Eggertz has
adopted that of the bronze called yellow metal, newly rubbed
with fine sand on leather. (This metal consists of 60 parts
of copper and 40 of tin.) For the coloration No. 3, a
convenient alloy has not yet been found. A bronze, con-
sisting of 85 parts of copper and 15 parts of zinc, does not
quite represent the colour which should be obtained, for
when freshly cleaned it is too bright, and at last takes a
bluish coloration'. For the colour in question it is better
to use a plate of silver which remains in the flask during
the solution of the iron, until it has become as brown as
390 ESTIMATION OF. SULPHUR IN IRON AND STEEL.
possible,- and a slight bluish colour begins to be perceived ;,
the plate is then removed and preserved in a well-closed
tube. The colour No. 4 resembles that of a watch-spring.
If the amount of sulphur is very considerable, this colora-
tion passes to a clear bluish grey. By passing the plate of
silver over a bottle of sulphide of ammonium the desired
number can be easily obtained. To obtain in these assays
for sulphur the proper tint on a silver plate, it is necessary
to take certain precautions. The plate is to be held in
pincers and cleaned as well as possible by rubbing it with a
soft leather on which is placed a little fine rotten-stone.
Contact with the fingers is avoided by means of a piece of
paper, and the plate is to be carefully dried with a piece of
filtering-paper. If the plate, by cleaning or by the action
of the burnisher, has been purified on its surface, this should
be carefully removed by again rubbing with the leather, for
pure silver is less sensitive to the action of the gas than that
of the given standard. Thus it has sometimes been found
that the silver employed for coinage furnishes less homoge-
neous plates, of which those parts richest in copper assume
more quickly the blue coloration. On this account, the
plates should be compared between themselves, by intro-
ducing them at the end of a wire into a flask in which iron
is dissolving containing from 0-05 to 0-08 per cent, of sul-
phur. On introducing the plate, care must be taken not to
turn the side but the edge against the strongest current of
gas, whicn would otherwise colour one face of the plate
stronger than the other. The plate should be rapidly in-
troduced into the flask after the introduction of the iron, as
then a very strong disengagement of sulphuretted hydrogen,
immediately takes place. After a first experiment the flask
is to be filled with water several times, so as to get rid of
the odour of sulphuretted hydrogen. If a steel mortar is
employed to pulverise the iron, the whole of the piece
selected should be reduced to a very fine powder. The
mortar should be well cleaned each time, taking care to
remove the disc from the bottom. Changes in temperature
between 15 and 25C. seem to have no sensible influence on
the coloration of the metal ; if the temperature exceeds.
EGGERTZS METHOD. 391
40 the plate becomes moist and gives false indications.
Some practice is required to judge of the coloration of
the plate, but it may be easily acquired. Generally the
best plan is to place the standard plates of tints 1, 2, 3,
&c., on a sheet of white paper by the side of the plate
under experiment, exposing them to a good light near a
window (but not sunlight), and to examine them with a
lens. The colorations between 2 and 4 are the most
difficult to recognise ; but with a little experience none
will vary more than Ol ; so that, for instance, the colora-
tion may be estimated between 3'5 and 3 '6.
The following is somewhat an approximation between
the different colorations upon the silver plate and the
amount of sulphur in a great number of different samples
of iron :
Number of Percentage
for Coloration of Sulphur
1-0 0-00
1-2 0-01
2-0 0-02
2-5 0-03
3-0 0-04
3-1 0-05
3-2 0-06
3-3 0-07
3-5 0-08
3-6 0-09
3-7 0-10
3-8 0-12
3-9 0-15
4-0 0-20
It is evident that in this way the exact quantity of the
sulphur is not estimated ; but several years' experience
has shown that if these experiments are made with care,
and the quantity of sulphur does not exceed O'l percent.,
the results are near enough for all practical purposes.
Iron which does not colour the silver plate will sometimes
produce a coloration if we double the quantities of iron
and acid. With half the quantities of acid and sulphurised
iron, silver generally gives a little more than half the real
quantity of the sulphur which is present.
Amongst experiments on the estimation of sulphur in
iron, the following deserve mention : 1. The quantity of
sulphur in wrought iron is often so small that it produces
S92 ESTIMATION OF SULPHUR IN IRON AND STEEL.
no coloration on the silver plate ; this iron, therefore, not
being red-short, may be employed for all kinds of uses. It
must not, however, be forgotten that the quantity of sul-
phur is not equally distributed throughout a piece of iron,
but that it may vary considerably in different places. On
experimenting with the turnings obtained from a portion
of an iron bar which was visibly red-short, a stronger tint
is often obtained on the silver plate than when using other
parts of the bar. The fragments obtained from red-short
iron in boring a horse-shoe do not often give on the silver
plate a deeper coloration than 2, and it appears to follow
that ordinary wrought iron which contains 0'02 per cent,
of sulphur in certain parts cannot conveniently be em-
ployed for this purpose. If the red-short iron gives to the
plate a slighter and more feeble coloration than 2, it may
be supposed that the breaking is due less to sulphur than
to an insufficient working of the cast iron, the crude pieces
in wrought iron entirely free from sulphur often acting as
if they were red-short. In general it appears certain that
the quantity of sulphur in iron is more injurious when the
iron has been badly worked. In a hard iron melted in a
steel crucible we may, in spite of its containing O04 per
cent, of sulphur, make holes like those in a horse-shoe
without any trace of cracks, which may undoubtedly be
attributed to the homogeneity and good working of this
iron ; the quantity of phosphorus being only 0'3 per
cent. The lower portion of an English rolled rail, without
fault, contained 0*11 per cent, of sulphur and O03 per cent,
of phosphorus ; a portion was cut off which was so red-
short that it could not be made use of.
2. The amount, in sulphur, of steel of the best quality
is such that the colorations on the silver plate vary only
between 1 and 1'5. As in the case of wrought iron, the
quantity of sulphur often varies in different parts of the
same piece of steel, and that also appears to be the casein
a little less decided manner in cast steel.
3. The quantity of sulphur in cast iron is rarely so little
as not to colour the plate. In the greater number of
Swedish cast irons, this quantity is such that the silver
plate varies in coloration between 2 and 3. In iron for
NEW COLORIMETE1CAL PROCESS. 30:}
gun-castings it is between 3'3 and .3*7, and sometimes
more. In cast iron the quantity of sulphur is often dis-
tributed unequally ; there is generally more on the surface
than is met with below. If the coloration of the silver-
plate does not- exceed 3, we can assume that the cast iron
refined in the ordinary manner will not give red-short
iron, especially if the refining is done carefully. But as in
different methods of refining, different quantities of sulphur
may be removed from the iron, and in general more if the
iron is the result of a light charge of the blast furnace, it
cannot be said beforehand that all cast iron which commu-
nicates a bluish coloration to the plate will necessarily give
red-short iron. This will be the case, however, with cast
iron which colours the plate as deep a blue as that of
a watch-spring. In cast iron, which gives a red-short
wrought iron, without rendering the silver plate more
than brown, it is probable (the iron having been well re-
fined) that the cause is owing to the presence of other
substances than sulphur ; but this occurrence is very rare.
Many circumstances appear to show that the quantity
of sulphur in iron diminishes with time, at least on the
surface, and under favourable conditions.
New Colorimetrical Estimation of Sulphur in Iron.
The above-described colorimetrical method of esti-
mating the amount of sulphur in iron, worked out by Prof,
v. Eggertz, has been, and still is, indisputably of great
utility to the Swedish iron manufacture. The method
since then has come into general use in our ironworks.
Mr. J. Wiborgh sought to perfect such a colorimetrical
process as would be applicable also for greater amounts
of sulphur than that for which the Eggertz method is in-
tended. This new colorimetrical sulphur test fulfils the
demands that in general may be put upon the practical
requirements intended in the method namely, it is quick
and easy to execute ; also, it has sufficient accuracy to give
the amount of sulphur, whether in cast iron, steel, or
wrought iron. The following is Mr. Wiborgh's description
of the method.
394 ESTIMATION OF SULPHUR IN IKON AND STEEL.
Basis of the Method.
Iron dissolves completely in diluted sulphuric acid
or hydrochloric acid, and the gases evolved hydrogen,
carburetted hydrogen, and sulphuretted hydrogen pass
through a cloth impregnated with a metallic salt, and
through the action of sulphuretted hydrogen a metallic
sulphide is formed, which colours the cloth. From the
intensity of the colour afterwards the amount of sulphur
in the iron is decided.
I proceed here upon the assumption that a given sur-
face is always coloured equally strongly by a fixed quantity
of sulphur. But to obtain an equal quantity of sulphur
from two specimens of iron which have unequal amounts
of sulphur, evidently the amounts of iron weighed out
must be inversely proportionate to the amounts of sulphur.
Therefore a w r eight W of an iron with sulphur amount S
gives the same colour as a weight W of another iron with
sulphur amount S' ; so must
WS = WS';
and as S 7 is the amount of sulphur sought,
WS
S'-'w 7 '-
If you have thus an iron (normal iron) with the amount
of sulphur accurately known, you can, by varying the
amounts of it weighed out, produce a colour-series in
which for every colour you know the product W S,
This colour-series constitutes a scale with the aid of which
the unknown amount of sulphur S' in another iron can be
estimated by dividing the same colour's known product,
W S, by the weight of the iron used for the assay.
The apparatus shown in the accompanying figure con-
sists of a small boiling flask, A, provided with a close-fitting
caoutchouc stopper, m, in which are placed a glass cylin-
der, R, with one end drawn out to a tube, p, and the other
end having a flat polished flange, G, also a funnel tube, T,
to introduce the acid. This latter consists properly of
two, T and T', united with a caoutchouc tube, over which
NEW COLOR IMETRICAL PROCESS.
395
FIG. 103.
is placed a nipper tap, K, with a screw, so that the quan-
tity of acid to be passed into the flask may be accurately
regulated. Upon the cylinder flange is placed a caoutchouc
ring, N, and upon this the prepared cloth o. That the
ring may close tightly against the flange and cloth, lay
upon the cloth another caoutchouc ring, N', of the same
size as the former, and outermost a wooden ring, s, which
is pressed against the flange by the spring B.
These caoutchouc rings ought to have a precise inside
diameter, for upon the size depends how much of the sur-
face of the cloth will be coloured ;
as showing, by the drawing, it is
made less than the opening of
the cylinder, because it is easier
to obtain such rings of a precise
size than the glass cylinder. In
the apparatus represented the
diameter of the ring o is 55, and
the opening of the cylinder 58
millimetres.
In order that no steam may
condense upon the ring s and
run down to the cloth, it is best
to make this ring of wood rather
than of glass or metal, of the
size and form shown in the
drawing.
By the arrangement now
described neither gas nor steam
can come from the apparatus
without first passing through the cloth.
To heat the flask to boiling it is placed upon a sand-
bath, E, which rests upon the stand D, and is heated by
means of a gas or spirit lamp, p.
Preparation of the Cloth.
In the beginning I used, instead of cloth, unglazed
paper, such as filter-paper ; but I soon found that cloth
396 ESTIMATION OF SULPHUR IN 'IRON AND STEEL.
was much more serviceable. Such paper certainly per-
mits steam and gas to pass ; but it is tender, and easily
breaks asunder with the least incaution while boiling.
The cloth used is common, fine, white cotton calico.
Linen is less suitable, for it is thin, and therefore does not
absorb so easily all the sulphuretted hydrogen as evolved.
The preparation of the cloth is simply to moisten it with
a solution of a metal salt. For this may be employed
either lead, silver, copper, cadmium, or antimony salts ;
but of all these I have found cadmium salts the most
serviceable.
Silver and lead salts are certainly very sensitive to
sulphuretted hydrogen ; but the combinations of these
metals with sulphur are black, and colour the cloth too
strongly, so that it will be necessary to employ small
weighings of the sample or extravagantly large apparatus.
Copper salts certainly give brownish and considerably
softer colours ; but, in consequence of the property of the
sulphide to enter into variable combinations with the oxide,
they are easily oxidised, and not permanent.
Antimony has few soluble salts ; and, besides, in one case
these are seen to be less sensitive to sulphuretted hydro-
gen : they are little fit for this purpose. A sufficiently
dark colour is obtained by mixing the salts as, for ex-
ample, cadmium and lead salts but has not led to any
good results, as the colour comes out uneven, darker upon
some places than upon others, so that the cloth has a
more or less flannel-like aspect, varying between yellow
and black.
Eor these reasons I have selected cadmium salt alone
for the preparation of the cloth, and this the rather as the
sulphur combination of cadmium is a particularly beau-
tiful and constant yellow colour ; also, that the affinity of
this metal for sulphur is so great that it surpasses that of
lead.
When the cloth is impregnated with a mixture of cad-
mium and lead acetates, equal parts of each, and then
exposed to a small quantity of sulphuretted hydrogen,
the cloth at first will be almost exclusively yellow, and
NEW COLORIMETRICAL PROCESS. 397
by the passage of more sulphuretted hydrogen it becomes
dark, which shows well that the gas is sooner decom-
posed by cadmium than by lead. Again, concerning which
of the cadmium salts ought to be selected for impregna-
tion of the cloth it appears to be a matter of indifference
which is employed, although the colour tint will be some-
what different as one or other salt is used.
Cadmium nitrate gives a singularly high and strikingly
beautiful orange colour ; the sulphate is somewhat weaker,
with brownish yellow tint, while the chloride and acetate
give lighter colours.
The colours occasioned by the varying amounts of sul-
phur may perhaps be most easily judged separately when
the cloth is prepared with cadmium nitrate ; but not-
withstanding I have likewise used the acetate, because
I think it likely that this salt, which has the weakest
and most volatile acid, ought to give colours that would
be least changeable, which property is of considerable
weight.
Tn order that all the sulphuretted hydrogen may be
absorbed by the cloth, it must contain a certain amount
of cadmium salt in proportion to the largest quantity of
sulphur which may possibly be present ; otherwise a por-
tion of the sulphuretted hydrogen will pass through the
cloth, colouring yellow not only the under side but also
the upper side.
A solution of 5 grms. crystallised cadmium nitrate in
100 c.c. distilled water is of suitable strength.
Cloth of the proper fineness, prepared with such
solution, allows not a trace of sulphuretted hydrogen to
pass through, for the cloth was coloured only on the
under side, and when double folds were used the upper
had not the faintest colour.
The general influence of the degree of concentration of
the solution is that the colours from strong solutions lie
more upon the surface of the cloth, and that of weaker
solutions penetrates deeper into the cloth ; and this
causes a certain unevenness in the aspect of the colour,
although the amount of sulphide of cadmium in both
398 ESTIMATION OF SULPHUR IN IRON AND STEEL.
cases is alike. The influence of a smaller variation in
the degree of concentration is not perceptible.
The cloth is prepared by cutting, on a round model,
several folds to about 80 m.m. diameter. These are laid
in the solution of cadmium acetate, care being taken that
each circle is thoroughly saturated by the solution. After
some minutes take out the pieces, and spread them out
on a clean cloth until completely dried ; then place them
for safety in a suitable box.
The Colour Scale.
According to the quantity of sulphuretted hydrogen
which reacts upon the prepared cloth, so is it covered with
different amounts of cadmium sulphide, and receives a
more or less yellow colour. The sensibility in this respect
is so great that indeed T oVo tn P art f a ni.grm. of sulphur
is able to communicate to the surface of a square centi-
metre of the white cloth a certainly weak but yet very
manifest yellow colour. Increase successively the amount
of sulphur by T^OO^ P art f a m -g rm - per square centi-
metre ; you obtain by this small addition of sulphur from
the first a clear distinction in the intensity of the. colour ;
but in proportion as the strength of the colour increases
with the augmented amounts of sulphur, the difference
will be all the more difficult to observe. When the amount
of sulphur increases to about 0'02 m.grm. per square c.m.,
the cloth is now strongly coloured, and to produce a
manifestly distinct difference of colour intensity it requires
two or three times greater excess of sulphur than with the
lower colour estimations. Consequently you cannot with
advantage make use of very strong colours, because the
difficulty of their comparison is considerably increased.
In order to estimate the amount of sulphur in an iron,
you must first have a colour series, or scale of colours,
whereon every colour number represents a certain amount
of sulphur per cent., presupposing that a certain quantity
of iron has been weighed out for assay.
This colour series can be easily procured by the help of
NEW COLORIMETRICAL PROCESS. 399
a normal iron in which the amount of sulphur has been
estimated beforehand with the greatest accuracy. How
great the amount of sulphur this normal iron may con-
tain is quite immaterial, but use the most trustworthy
among the known methods of estimating sulphur in iron
by oxidation of the sulphur, and thereafter precipitating
by chloride of barium. With small amounts of sulphur
this is to a certain degree unsafe, so it is better to use for
the scale an iron with a pretty high amount of sulphur,
such as about O'l per cent., because the influence of
a small error in this estimation of the sulphur will be
less.
If you have a known amount of sulphur S, also an iron
of which 0'4 grm. is weighed out, and its amount of sul-
phur S', to obtain the same colour from both of the irons
must be
w x s = 0-4 x S'
Or
w _0'4xS / .
In this formula place instead of S' the successive
amounts of sulphur per cent., as 0*005, Q-01, 0-02, &c. As
you know how much of the normal iron in every case
ought to be weighed, so you have a series of colours that
are entirely the same as those which will be obtained if
you had different irons with their respective amounts of
sulphur and 0'4 grm. of each weighed ; you get, in other
words, a colour scale, where colours give directly the
amount of sulphur the iron contains.
Make up such a scale for the apparatus shown in the
sketch, with the inner ring 55 m.m. diameter, under the
supposition that 0'4 grm. has been weighed for testing.
It is not desirable to go farther than this, so that the
highest colour number corresponds to 0*1 per cent, sul-
phur, because otherwise the colours will be too strong.
Nevertheless, you must not take between the high colour
numbers a greater difference in the amount of sulphur
than between the lower, if the different colour intensities
are to be clearly separated from each other.
400 ESTIMATION OF SULPHUR IN IRON AND STEEL.
For example, you can allow the scale to be composed
of seven colour numbers, taken thus
No. 1 corresponding to 0*005 per cent sulphur.
2 0-01
3 0-02
4 0-03
5 0-05
6 0-07
7 0-10
If thus 0-4 gramme iron is weighed for assay, one can,
with the help of the above scale, estimate the amount of
sulphur up to O'l per cent. The accuracy wherewith the
estimations are done is, for the lower of amounts of sul-
phur at least, equal to 0*005 per cent., and for higher O'Ol
per cent. ; for the differences between the colours of the
scale are here so great that one with the greatest ease can
estimate a colour lying between two colour numbers.
The same scale may, however, be used to estimate
quickly whatever amount of sulphur is desired, if only the
weight of iron used for assay be varied. It is obvious
that absolutely the same colour which an iron gives from
0'4 gramme must be got from another iron with half as
much sulphur if 0-8 gramme has been weighed ; and that,
in general, if 0'4 n is weighed, the colour scale comes to
represent sulphur amounts which are J of those that
answer to the weight 0'4 gramme.
One ought therefore to use in general greater amounts
for assay, when low amounts of sulphur are to be esti-
mated with great accuracy, also lesser quantities for high
amounts of sulphur.
In case the colour comes out too strongly lying upon
the margin or beyond the greatest amount of sulphur
in the scale, the assay ought to be made with a lesser
weight.
To avoid calculation one can, under every colour num-
ber, place not only the amount of sulphur which corre-
sponds to that weighed out in making the scale, but also
that which corresponds to some other weighings which
possibly may occur. Note the following weighings, W,
also the corresponding amounts of sulphur, S, per
cent. :
NEW COLORIMETKICAL PROCESS. 401
0-8 gramme. 0-005 per cent.
0-4 , 0-01
0-2
0-1
0-08
0-04
0-02
0-02
0-04
0-05
0-1
0-2
With these different weighings are thus given the colour
numbers of the amounts of sulphur from 0-005 per cent,
to 0-2 per cent.
Having by the above method prepared the different
colour numbers which form the scale, they may be arranged
in order upon small white drawing-paper, bound, and pre-
served in a suitable portfolio.
When cadmium acetate is used for the preparation of
the cloth, the colours will be very constant. They have
undergone no sensible change for several months, though
preserved only in the above manner. One such scale
ought to be made for every apparatus, and observe that
the cloth used for the preparation of the scale must be the
same sort as that afterwards used for testing.
Details of the Process.
Rinse every part of the apparatus with water, so that
no acid remains from the former assay. The boiling-flask
is half filled with distilled water ; then fit the apparatus
together and place on the sand-bath, which is heated by
means of a gas- or spirit-lamp, so that the water in the
flask comes to gentle boiling. While heating, weigh out
the iron to be tested. This may certainly be in the form
of small pieces, but, that the solution may not proceed too
slowly, it is best for the iron to be in the form of filings,
turnings, or powder. Difficultly soluble iron, such as white
iron high in silicon, phosphorus chrome iron, &c., ought
always to be finely pulverised.
The weighed sample is brought, by means of a small
funnel and hair-pencil, into the test-tube r. This test-
tube has the mouth widened. It is set in a loop made on
the end of a platinum wire w, which is bent over the
edge of the test-tube as shown in the figure, so that the
platinum wire may not slip over the tube and rest on the
D D
402 ESTIMATION OF SULPHUR IN IKON AND STEEL.
bottom of the flask ; this ought to be avoided, because
the test-tube may take such a position that the access of
the acid to the iron maybe made difficult. The platinum
with which the test-tube is thus fastened ought to be
rather stiff about 0'3 m.m. diameter.
After the water has boiled two minutes, and the air
has been expelled, take out the stopper with the cylinder
attached. The test-tube with the sample is lowered into
the flask, where it rests on the bottom, being held upright
by the platinum wire.
Put the apparatus together again, and upon the
cylinder flange lay the above-mentioned caoutchouc ring,
55 m.m. in diameter, and over it the prepared cloth,
another caoutchouc ring, and lastly the wooden ring, all
of which are pressed together by the spring clamp.
As soon as the cloth is laid on screw up the clamp K on
the funnel-tube, and steam must now pass through the cloth.
Partly to get the air as much as possible driven out of
the apparatus, and partly to moisten the cloth, maintain
the water in gentle boiling for about eight to ten minutes
before any acid is introduced. Then fill the funnel-tube
with diluted sulphuric acid (for example, J volume sul-
phuric acid, 1*83 sp. gr., and |- volume water). Open care-
fully the screw-clamp K, and allow the acid to drop into the
flask. For 0-4 gramme iron use about 10 c.c. dilute acid.
As soon as acid comes down, the solution of the iron
begins ; steam and gases pass through the cloth, and in
proportion as the solution proceeds, also according to the
amount of sulphur in the iron, the under side of the cloth
is more strongly coloured yellow.
After all the iron is dissolved, boil the solution further
from five to ten minutes, to drive out the sulphuretted
hydrogen which yet remains in the apparatus. Then open
the spring clamp, remove the bottom ring, and lay the
cloth upon a piece of filter-paper to dry completely. It
only remains now to estimate the amount of sulphur by
comparing the colour of the cloth with the colour scale.
That, the liquid in the flask during the whole period be
maintained at an even but gentle boiling is very essential
NEW COLORIMETRICAL PROCESS. 403
for this process. The boiling ought to be so strong that
the steam is always seen to pass through the cloth,
but by no means so violent that the cloth becomes
stretched by the boiling. For when it happens that the
cloth, in consequence partly of this stretching and partly
from the condensed steam, becomes more and more close,
and takes a strongly convex form, then the pressure in the
apparatus increases, so that on the introduction of the acid
the gases formed rush through the funnel-tube. With
cautious boiling one need never fear such an occurrence.
In order to avoid oxidation of the sulphuretted hydro-
gen formed during the solution of the iron, it is of
importance that before the acid is introduced into the
flask the air be as completely as possible expelled from the
water and the apparatus, also that the boiling be so strong
that plenty of steam always accompanies the gases which
pass through the cloth.
The cloth in the assay ought to be coloured evenly, for
if the colour is uneven it very considerably increases the
difficulty of estimating the colour strength.
Whether the colour comes out even or not depends
chiefly upon the construction of the glass cylinder. It
ought to be so accurately made that the centre line of the
tube coincides with that of the cylinder. Further, the
tube ought to be short and of a conical form, 7 to 8 m.m.
diameter at the lower end.
If the tube be too wide the cloth is always coloured un-
evenly ; and again, if it be too small, drops of water con-
densed in the tube are cast up on the cloth, which give
it a spotted aspect. The drawing represents the cylinder
so set in the caoutchouc stopper that the mouth of the
tube is even with the under side of the stopper ; but it is
still better to allow the end of the tube to be 5 to 10 m.m.
under the stopper. It is certainly now more difficult for
the water to leave the tube, and causes especially at
first, before the cylinder becomes warm a weak bubbling ;
but this has no hurtful influence, but contributes to the
more even coloration of the cloth.
Observe, further, that the glass cylinder be placed by
D D 2
404
ESTIMATION OF SULPHUR IN IRON AND STEEL.
the eye as vertical as possible, also that the apparatus be
riot placed in a draught.
The time required for a sample is from half to three-
quarters of an hour, according as the iron is more or less
soluble.
This new method, in many special experiments, has given
most satisfactory results, in that it is independent of the
amounts of carbon and silicon in the iron, indicating high
as well as low amounts of sulphur in iron, as seen in the
accompanying table, with sulphur estimations in different
sorts of iron. These have been performed by this method,
and controlled by accurate estimations of the sulphur by
the wet method.
No.
Sorts of Iron tested for Sulphur by
Wiborgh's Colorimetric Method and by
Chloride of Barium
By
Wiborgh's
Method.
Sulphur,
p.c.
By Wet Method with Chloride
of Barium
.A.
Sulphur,
p.c.
Estimated by
1
White charcoal cast iron
0-005
0-005
A. Tamm.
2
Spiegeleisenfrom Siegen,with
0*05 per cent, copper .
0-005
0-006
J. W T iborgh.
3
Bar iron, with 0*076 per cent.
arsenic . . . .
0-007
0-008
Y. Lagervall.
4
Casting
0-0075
0-005
A. Tamm.
5
Grey charcoal cast iron
0-0075
0-005
A. Tamm.
6
White
0-012
0-01
J. Wiborgh.
7
Casting .....
0-012
0-014
A. Tamm.
8
Half- white charcoal cast iron
0-018
0-018
J. Wiborgh.
9
Malleable casting .
0-015
0-013
N. Lagerfelt.
10
White charcoal cast iron,with
0*071 per cent, arsenic
0-02
0-025
Y. Lagervall.
11
White charcoal cast iron
0-02
0-02
J. Wiborgh.
12
Malleable casting .
0-023
0-024
E. V. Zweigbergk.
13
White charcoal cast iron
0-023
0-024
Y. Lagervall.
14
?> >
0-025
0-022
A. Tamm.
15
> J> JJ
0-028
0-029
J. Wiborgh.
16
Cast iron melted with copper,
containing 1'55 p.c. Cu
0-039
0-038
5
17
Siemens-Martin iron
0-04
0-037
>5
18
5> >5
0-05
0-047
}>
19
White charcoal cast iron,with
0-015 per cent. Cu
0-06
0-061
JJ
20
Steel
0-07
0-068
21
Siemens-Martin iron .
0-1
0-093
1J
22
Grey cast iron
0-135
0-134
Y. Lagervall.
23
Cannon cast iron .
0-15
0-145
J. Wiborgh.
24
White cast iron from Horde,
with 1-88 per cent. P .
0-21
0-19
E. Aquilon.
25
Malleable casting .
0-35
0-34
J. Jungner.
26
WTiite cast iron from coke .
0-45
0-46
J. Wiborgh.
27
"
0-7
0-66
P. G. Lidner.
KEW COLORIMETRICAL PROCESS. 405
I have also tested if some impurities in iron such as
copper and arsenic had any influence upon this method.
For this purpose several sorts of iron containing copper
and arsenic were specially selected from the collection of
the Mining School ; and in the table it is found
I. In the samples 2, 16, and 19 copper has no action ;
also
II. In samples 3 and 10 the same is probably the case
with arsenic, for the certainly somewhat large difference
in the sulphur estimation which occurs in sample 10
must be sought in other circumstances, as sample 3 even
with its small amount of sulphur and high amount of
arsenic shows no noticeable difference.
The apparatus may be made of whatever size is
wished, but in relation to use in Sweden it is seldom
necessary to examine iron with a greater amount of sul-
phur than O'l per cent. I regard an apparatus with an
inner ring of 55 in.m. as diameter of a suitable size. On
the other hand, if one has in general to test iron with
high amounts of sulphur, as cast iron made with coke,
&c., I would recommend that the apparatus be made
somewhat larger.
Estimation of Sulphur in Iron Ores. Five grammes
of the mineral, ground as finely as possible in an agate
mortar, are treated with potassium chlorate and hydro-
chloric acid. After desiccation and extraction with hydro-
chloric acid and water, the insoluble substances may be
lead, calcium, barium, and strontium sulphates, silica, and
undecomposed mineral. By stirring well and filtering the
liquid whilst warm, the two former salts may generally,
however, be dissolved. The filtration should be performed
through a double filter, to prevent the pulverised mineral
from passing through. When the clear portion of the
solution has been poured upon the filter, add to the in-
soluble matter 5 c.c. of hydrochloric acid and 5 c.c. of
water ; then leave the mixture for two hours on the water-
bath at a boiling temperature, when, if care be taken to
stir well, the calcium sulphate will be completely dissolved.
Wash the insoluble portion in warm water, and pour it on
40G ESTIMATION OF SULPHUR IN IRON ORES.
a filter, taking care to place below a flask specially to^
receive that portion of mineral which may have passed
through the filter. The filtered liquid, whose volume is
about 50 c.c., should be rapidly boiled and mixed with
2 c.c. of a saturated solution of barium chloride. (This
amount is sufficient to precipitate the sulphuric acid
formed from 0*1 grm. of sulphur.) After cooling add to-
the mixture 5 c.c. of ammonia, sp. gr. 0-95, then stir well
with a glass rod, and leave the whole to rest at the ordi-
nary temperature for twenty-four hours. The clear solu-
tion should be decanted as completely as possible on to a
strong filter, and the precipitate stirred up with about 20-
c.c. of cold water, and then left to itself until it has quite
settled. If warm water is used without having added a
few drops of hydrochloric acid, a little oxide of iron will be
precipitated. The clear liquid is likewise thrown on to the
filter, and the operation is repeated several times with cold
water, and then two or three times with boiling water,,
without which precaution the barium sulphate will pass
through the filter. Finally collect the precipitate, and
wash it with warm water. The last drops of this water,
on being evaporated on a watch-glass, ought not to leave
more than a scarcely visible white ring. The precipitate
is then dried, heated to redness, and weighed. If it is
coloured with iron oxide it must be washed with a little
hydrochloric acid, dried in the water-bath, and taken up
with a few drops of acid and w r ater, and then the pre-
ceding operation repeated (w r ashing, drying, heating, and.
weighing). If the precipitate has only a feeble red colour,
which is often the case, this latter operation will be un-
necessary.
To dissolve the lead sulphate which may occur in the
insoluble portion, remove this from the filter with the end
of a feather, introduce it into a flask, and pour over it
10 c.c. of concentrated ammonium acetate. The solution,
after having been strongly agitated and heated in the'
water-bath, is carefully poured on to a filter. Then wash
the residue with a little ammonium acetate, and repeat
the treatment until a few c.c. of solution, acidulated with
ESTIMATION OF SILICON IN IRON AND STEEL. 407
a little acetic or hydrochloric acid, is not rendered turbid
when warmed with barium chloride solution. The ni-
trate is then diluted, slightly acidified, and the sulphuric
acid precipitated by means of barium chloride. After the
lead sulphate has been removed, there may still occur
barium and strontium sulphates in the insoluble portion.
To decompose these salts the residues must be dried,
heated to redness, and weighed, and then fused with five
times their weight of pure dry sodium carbonate. The mass
is digested with water over a water-bath, the liquid filtered,
and the residue washed with warm water. The silicic acid
is separated from the solution, which contains the sodium
silicate, carbonate, and sulphate, by adding hydro-
chloric acid and drying on the water-bath. After filter-
ing, precipitate the solution with barium chloride. To
ascertain if the iron mineral contains gypsum or other
soluble sulphates, take 5 grammes, place them in 20 c.c,
of hydrochloric acid and 60 c.c. of distilled water, and
digest them for three hours on the water-bath, with
frequent agitation. The filtered solution is mixed with
barium chloride and 15 c.c. of ammonia ; then proceed as
already described. If there be present in the mineral
grains of iron- or copper-pyrites, or galena, they will only
give traces of sulphuric acid in this operation.
Estimation of Silicon in Iron and Steel. All who are
occupied in the analysis of iron and steel are aware how
very uncertain the estimation of silicon is when the
method hitherto used for its separation is followed, because
cast iron, and even bar iron and steel, are never found
absolutely free from intermingled slag. This slag is de-
composed by the ordinary method of dissolving the iron
in acids, and its silica then augments the amount of silica
formed from the silicon contained in the iron or steel*
This cannot be said of every sort of cast iron, but these
sometimes contain blast-furnace slag, although pig iron
containing slag may be considered as rare. It ought also
to be mentioned that crystallised silicon has been found
in crystallised cast iron from Krain, in the form of small
silvery plates, which are neither acted upon by boiling
408 ESTIMATION OF SILICON IN IRON AND STEEL.
aqua regia nor by ignition in oxygen gas ; but they are
converted into silica by fusing with potassium and sodium
carbonates.
Crystallised silicon is insoluble in hot solutions of
sodium carbonate, but is soluble, with development of
hydrogen, in hot solutions of caustic potash, and also in
hot hydrofluoric acid. The accurate estimation of the
silicon in iron and steel has been effected by Dr. Eggertz,
who, after fruitless efforts to dissolve iron in highly diluted
organic or inorganic acids, which should have no effect on
the refinery slag, finally discovered such a solvent in
bromine, which, when mixed with water, dissolves the
iron without the slightest action on the accompanying slag.
But as experimenting with bromine in large quantities is
very disagreeable, trials were made to use iodine instead ;
and this, like bromine, has been proved to have no effect
on the slag, nor on iron oxide or proto-sesquioxide, or
manganese proto-sesquioxide. At the same time, bromine
dissolves iron quicker than iodine does, and is, perhaps,
more easily obtainable in the requisite state of purity.
Moreover, as continued experiments have shown that a
solution of sodium carbonate can separate finery slag from
the silica which has been formed by the use of iodine or
bromine on the silicon contained in the iron, the follow-
ing method for the estimation of silicon and slag in bar
iron or steel has been used and considered successful ; the
same method may be employed for cast iron, because
blast-furnace slag, when such is found, is not perceptibly
changed by iodine or bromine, nor by solutions of sodium
carbonate. Three grammes of bar iron or steel which
have passed through a sieve of 0*2 of a line are taken for
analysis. Fifteen grammes of iodine are added in small
portions at a time to 15 c.c. of water in a beaker of 100
c.c. capacity. The water must be previously boiled to
expel the air, which would otherwise oxidise the iron.
The iodine is stirred in the water with a glass rod, in order
to get rid of the air which has accompanied it, and the
floating iodine particles are allowed to sink. The beaker
with the iodine and water, which is kept covered with a
ESTIMATION OP SILICON IN IRON AND STEEL. 409
watch- glass, is cooled in ice before the iron is put in, and
during the solution it is kept at the temperature of C.
For the first few hours it must be well stirred every hour,
or oftener, with a glass rod, but afterwards not so fre-
quently. In consequence of the low temperature and the
careful admixture of the iron (by which heat is prevented),
the solution may be performed without the least de-
velopment of gas, and the iron has less tendency to
become oxidised by the air when at this low tempera-
ture. By pressure, and by agitating with the glass
rod, the solution of the iron particles which collect
at the bottom of the beaker is much facilitated ; if no
lumps of the particles are visible, the beaker may be
kept at the ordinary temperature or, preferably, in ice
water. If some of the solution has dried on the sides of
the beaker or on the glass rod, it must be well moistened
with the same solution before water is added. About 30
c.c. of water, which should be very cold in order to pre-
vent the formation of basic salts, are added to the solution ;
it is then well stirred, left to settle, and the fluid with the
lighter particles of graphite is poured on to a filter of 2
in. diameter ; the filtration is kept up without interruption
until there remains only a somewhat heavy dark powder
of slag, &c., at the bottom of the beaker ; about 5 c.c. of
water, with a few drops of hydrochloric acid, are now
poured in and stirred with the glass rod ; if hydrogen is
given off, it is an indication that there is still some metallic
iron undissolved. The acidified water is quickly poured
on the filter, in order not to act on the slag. If a de-
velopment of gas is perceived, a little iodine, with sodium
carbonate and water, is added for the complete solution
of the iron, and the residue is thrown on the filter and
washed with cold water, until a drop of the filtrate gives
no reaction with a solution of 0-2 per cent, of potassium
ferrocyanide contained in a small porcelain crucible.
Solutions containing 0- 00001 gramme of ferric oxide per
c.c. show in this way very distinct reactions, particularly
if a drop of dilute nitric or hydrochloric acid be added.
The filtrate is evaporated to dryness, in which operation
410 ESTIMATION OF SILICON IN IRON AND STEEL.
some of the iodine is sublimed away. Thirty c.c. of
hydrochloric acid, 1*12 sp. gr., are then added, and it
is again evaporated in order to obtain the silica which
may be dissolved in it. The filter, previously dried and
weighed, is again dried and weighed when containing the
precipitate. It is then ignited, and the residue weighed.
After ignition, the residue is boiled in a solution of soda,
in order to extract the silica, and weighed. It should be
observed that some part of the silica which has been
formed from the silicon in the iron may possibly unite
with the slag during the drying and ignition. In conse-
quence of this, it is difficult to extract it by means of a
soda solution, whence this method is not to be recom-
mended in exact estimations of silicon.
When using bromine as a solvent 6 c.c. must be taken
to 8 grammes of finely powdered iron or steel, with 60 c.c.
of water, which has been previously boiled and cooled to
C. ; and this temperature must be preserved by placing
the beaker in ice water until the solution is complete,
which usually takes place in two or three hours ; it is
cautiously stirred once or twice with a glass rod ; if stirred
hastily, the solution proceeds too violently. The further
operations are conducted in the same manner as when
using iodine. The bromine is most conveniently preserved
under water, and is taken up by a pipette, which is in-
troduced into the bottle, the upper end being kept closed
by the finger.
When it is preferred to dissolve iron or steel in lumps,
instead of in powder, this may be done ; but it is not then
necessary to place the beaker in ice water, as the metal is
less violently acted upon in this form. Several days are
required for the solution ; the iron, and particularly the
pieces of steel, must be occasionally cleaned from the
graphite which adheres to their surface.
In order to estimate the silica (formed from the silicon
in the iron) and slag, the filter, which contains graphite (in
combination with iodine or bromine and water), silica, and
slag, is unfolded, whilst it is still wet, on a watch-glass.
The contents are washed away from the filter (these should
ESTIMATION OF SILICON IN IEON AND STEEL. 411
only occupy the lower half of the filter whilst in the
funnel) with a very fine jet from a wash-bottle (so as not
to obtain too much water) into a platinum or silver crucible
of the capacity of 30 c.c. The loosening of the mass may
be facilitated by a fine camel-hair brush. The water in
the crucible is evaporated to about 6 c.c., then mixed with
3 c.c. of a saturated solution of sodium carbonate free
from silica, and the crucible put in a water-bath, It is
kept in the boiling water 1 hour, during which time the
liquid is stirred two or three times, and the insoluble mass
crushed with a platinum spatula. The supernatant liquid
is carefully poured from the insoluble mass on to a small
filter, and to the mass in the crucible are added 1 c.c. of
saturated solution of sodium carbonate, and 2 c.c. of water.
When this has been boiled 1 hour, the whole contents of
the crucible are thrown on the filter and washed. The
solution of silica in soda is acidified by hydrochloric acid,
and mixed with the iron solution, and the whole evaporated
to dryness on a water-bath. When the solution attains"
the thickness of ordinary syrup, it is stirred very often
with the glass rod until the mass becomes a dry powder,
and heated until the smell of hydrochloric acid has nearly
gone off; the beaker is then placed in boiling water for
6 hours, 15 c.c. of hydrochloric acid of 1-12 sp. gr. are
then added, and the beaker left on the water-bath 1 hour*
As soon as the red powder is entirely dissolved 50 c.c. of
water are added ; and when no crystals of iron chloride
are visible, the solution is thrown on a filter and washed
with cold water, warm water forming basic iron salts which
make the silica appear red. The filter containing the silica
is dried and ignited in a porcelain crucible, gradually in-
creasing the temperature to a full red heat, and weighed ;
if the silica is coloured red by ferric oxide, a little hydro-
chloric acid, 1*19 sp. gr., must be poured into the crucible,
One decigramme of ignited and pure silica obtained from
analysis will dissolve in the above manner in 6 c.c. of a
saturated soda solution and 12 c.c. of water. If any
residue is observed after the second boiling, this arises
from some impurity which has united in small quantities
412 ESTIMATION OF SILICON IN IKON AND STEEL.
with the silica, rendering it insoluble. When strong hydro-
chloric acid is boiled with this insoluble silica, it may
afterwards be dissolved. When the solution of silica is
diluted with water to the volume of 50 c.c. at the ordinary
temperature, it has no tendency to come into the form of
jelly. Quartz-powder is dissolved by the preceding
method, but very slightly, but ignited titanic acid and
finery slag are not acted upon, and the tersilicate slag from
blast-furnaces but very little.
When the silica is quickly exposed to a high tempera-
ture a considerable loss may arise from the spirting of the
water combined with the silica. Silica dried at 100 C.
has been proved to retain 1 equivalent of water to 3
equivalents of silica that is, about 6 per cent, of water,
which is lost by a strong ignition. The ferric oxide is
easily dissolved in the heat of a water-bath. The silica is
again thrown on a filter, w r ashed, dried, ignited, and
weighed.
To ascertain the purity of the silica, it may be mixed
in a platinum crucible with ten times its weight of pure
ammonium fluoride, diluted with water to the thickness of
syrup. The water must be evaporated on a water-bath,
and the crucible heated, with a cover on it, by a gradually
increasing heat over a spirit-lamp to a full red. If nothing
is left in the crucible, the silica was pure, and has passed
off as silicon fluoride ; but, if anything remains, the opera-
tion with ammonium fluoride must be repeated until a
constant weight is obtained. When iron contains tungsten,
some tungstic acid is formed, and this accompanies the
silica for the most part, being dissolved by the soda solu-
tion, but it is not volatilised by the use of ammonium
fluoride. Yanadic acid also accompanies the silica, behav-
ing as tungstic acid. Instead of using ammonium fluoride,
it is preferable to use hydrofluoric acid, with which the
silica is moistened, and the evaporation is conducted on a
water-bath. The mass left on the filter from the soda
solution may be composed of besides graphite slag, iron
oxide, titanium oxide, &c. (but not copper, at least when
the iron does not contain more than 1 per cent.); this is
ME. TURNERS PROCESS. 413
dried, ignited, and weighed. The method of separating
iron oxide and slag, when the iron or steel contains both
these, is not yet known. If the composition of slag were
always alike (which it is not) it would be easy to calculate
its amount from either the silica or iron oxide obtained in
the analysis. In a piece of very red-short Bessemer iron
which contained no sulphur, by several experiments 0'3
per cent, of iron oxide has been obtained, and only traces
of silicon. After ignition, the iron oxide may possibly be
found as sesquioxide. The amount of oxygen, in case the
redshortness is due to this, as it probably is, amounts to
less than Ol per cent.
When the iron or steel for analysis contains titanium, a
part of this substance follows the slag in the form of titanic
acid. If such is the case, this must be melted with ten
times its weight of acid potassium sulphate, by which it
is dissolved ; the mass is dissolved in cold water, and the
solution precipitated by boiling ; the weight is determined,
and subtracted from that of the slag.
Mr. Thomas Turner estimates silicon in iron and steel
in the following way :
For a sample of cast-iron, 2 grammes of metal in the
form of borings are placed in a clean beaker of about 8
ounces capacity, and covered with about 50 c.c. of water;
5 c.c. of sulphuric acid are then added, and the beaker
kept covered while solution proceeds. The liquid is then
evaporated till quite solid, after which the residue is ex-
tracted with 100 c.c. of boiling water. The insoluble portion
is washed with boiling hydrochloric or nitric acid, and
afterwards with water so long as any iron is extracted. It
is necessary in all cases to test the filtrate by potassium
ferrocyanide or thiocyanate. The insoluble portion is dried,
ignited, and weighed in the usual way.
In the analysis of steels containing only a small quan-
tity of silicon, results by this method agree fairly well
with those obtained by other trustworthy processes. In pig-
iron of specially good quality good results are also obtained.
The conclusions arrived at from Mr. Turner's experi-
ments are as follows :
414 BASIC CINDER IN IRON.
1. That with cast irons of specially good quality the
silicon can be correctly estimated by evaporation with
dilute sulphuric acid.
2. With phosphoric irons the residue obtained, though
white, is often impure, and should be further treated in
order to obtain accurate results.
3. "With phosphoric irons containing titanium, the
silica is contaminated not only with iron, but also with
titanic oxide and phosphoric acid. The residue may be
very nearly white and still contain 20 per cent, of sub-
stances other than silica.
4. On treatment with aqua regia the colour of the
residue is usually an indication of its purity.
The amount of silicon in grey charcoal pig-iron is about
2*7 per cent, and in spiegeleisen 0-8. per cent. The
amount of silicon in pig from coke blast-furnaces is rarely
more than 4 per cent. The least quantity of silicon in
grey cast iron is about 0*2 per cent., and in white (spiegel-
eisen) O'Ol per cent. The amount is usually from 1 to 2
per cent, in cast iron suitable for the Bessemer process,
and in pig-iron for puddling about 0'5 per cent. The
amount of silicon in iron of different degrees of hardness
from the same charge of the blast-furnace ought to be
pretty well judged by the fracture, after some estimations
have been made by analysis.
Estimation of Basic Cinder and Oxides in Manufac-
tured Iron. The principal value attached to estima-
tions of slag and oxides in defective iron is the almost
absolute certainty of discovering whether to the chemical
constituents of the metal or to its careless manufacture
may be attributed the faults observed. When we con-
trast the ' life ' of an iron rail of fair chemical purity with
' mild ' steel rails of nearly the same composition, it is
forcibly suggested to even the most superficial observer,
that the duration of a rail is proportionate to the cohesion
of its metallic particles, due principally to freedom from
mechanical impurities. Engineers have ever been alive
to the necessity of suitable machinery for expressing the
fluid cinder from the iron as it comes from the puddling
BASIC CINDER IN IRON. 415
and re-heating furnaces ; but how rarely is the estima-
tion of cinder asked for at the hands of the analyst! We
occasionally find extremely pure irons and steels pure
at least by the result of routine analysis which are con-
demned when subjected to mechanical tests. In some
cases a crystalline structure of the metal may weaken the
iron ; but it is usually the cinder which, by breaking the
lines of continuity in molecular structure of the metal,
determines its fracture.
We regularly estimate the ordinary chemical consti-
tuents of a manufactured iron : why not its contained
oxides of manganese and iron, and basic cinder, which,
when present in excess, weaken the iron and so detract
from its value ?
The cause of the omission seems to be that no process
with the exception of Eggertz's is suitable for technical
work. Fresenius's process is, without doubt, accurate ;
but what process can be admitted into an industrial labo-
ratory (where a complete analysis is required in two days
or less), which for each experiment requires a rod
of the metal and several days for the completion of the
analysis ?
An estimation of cinder, &c., by the following me-
thod, devised by Mr. Bettel, only occupies an hour or so,
and has the advantage of yielding constant results : Five
grammes of the borings are heated with a solution of 10
c.c. bromine and 35 grammes potassium bromide in 150
c.c. water. Heat is continued until the iron is dissolved,
the solution filtered through a 4-inch Swedish paper (pre-
viously washed with hydrochloric acid and boiling water),
drained, and washed with a solution of sulphurous acid
containing 5 per cent, hydrochloric acid. When the
filtrate is practically free from iron, wash with boiling
water containing \ per cent, hydrochloric acid ; then with
pure water rinse into small platinum dish ; evaporate to low
bulk, and dissolve out silica by means of hot solution of
carbonate of soda. Boil, dilute, filter, wash with hot
water ; then with a ^-per-cent. solution of hydrochloric
acid ; finally with water. Dry, ignite, and weigh. Esti-
416 BASIC CINDER IN IRON.
mate the silica in the residue in the ordinary manner.
The bromine in the first filtrate may be recovered as
bromide of potassium in an obvious way.
Some analysts object to the bromine process on account
of the vapour contaminating the air of the laboratory.
To those Mr. Bettel recommends the following process,
which has given good insults : Five grammes of the iron
rather finely divided are dissolved in 60 c.c. clear solution
of cupric chloride (1 in 2) mixed with 100 c.c. saturated
solution of potassium chloride. When no particles of
iron can be felt by the aid of a glass rod, add 50 c.c. of
dilute hydrochloric acid (1 in 20) ; boil and filter through
a 4-inch Swedish paper previously moistened with hot
hydrochloric acid (1 in 3), then with saturated solution of
potassium chloride. Wash the residue on the filter with
potassium chloride solution till all the copper is removed,
then with hot dilute hydrochloric acid (1 in 50), finally
with hot water. Separate the silica as before, ignite and
weigh. If the copper obstinately adheres to the paper, as
sometimes happens, slip over the tube of the funnel a piece
of india-rubber tubing with clip (or plugged with glass
rod) ; fill up funnel with strong liquid ammonia ; cover, and
allow to remain for half an hour ; then proceed with dilute
acid, &c., as before.
The results of both processes agree with Fresenius's
galvanic method.
Estimation of Phosphorus in Iron and Steel. The im-
portance of ascertaining the quantity of phosphorus in
iron is very great ; for although the presence of a very
small quantity of phosphorus in cast iron does not produce
any sensible modification in its proprieties, it nevertheless
loses its most essential qualities when the proportion of
phosphorus amounts to some thousandths. Almost all the
methods hitherto proposed consist in treating cast iron
with oxidising agents in such a manner as to cause the
phosphorus to pass into the condition of phosphoric acid,
which is estimated as a magnesian compound. Several
sources of error are inherent to this method, to avoid
which M. Tantin liberates the phosphorus as an hydrogen
ESTIMATION OF PHOSFHOKUS IX IRON AND STEEL. 417
compound. Experiment shows that there is not the least
trace of phosphorus in the residue after the complete at-
tack of the cast iron by hydrochloric acid, which fact is not
surprising if it be considered what strong affinities phos-
phorus has for hydrogen. The hydrogen phosphide pro-
duced by the action of hydrochloric acid upon cast iron
is almost always accompanied by sulphuretted, arseniu-
retted, and carburetted hydrogen. In order to effect the
separation of these gases, first pass them into a WoulfFs
flask containing a solution of potash, which absorbs the
sulphuretted hydrogen ; the other gases are then made to
traverse a solution of silver nitrate, which transforms the
arseniuretted hydrogen into silver arsenite, soluble in the
slightly acid liquid, whilst the phosphuretted hydrogen
precipitates the silver solution and forms an insoluble phos-
phide. The phosphorus being thus completely separated
from the sulphur and arsenic, its estimation is effected in
a simple manner. The phosphide of silver is treated with
aqua regia, and thus transformed into silver chloride and
phosphoric acid, which is precipitated in the state of am-
monia-magnesian phosphate.
Mr. J. B. Mackintosh has conducted several careful
experiments on M. Tantin's method, and finds that to
secure accuracy the steps to be followed are :
1. Solution in hydrochloric acid in a stream of oxygen
or air, absorbing the escaping gases in permanganate acidi-
fied with sulphuric acid.
2. Heating the solution to boiling, stopping the passage
of the oxygen current, and carefully adding an excess of
sulphurous acid solution, and continuing to boil till the
precipitated manganese binoxide in the absorption-flask
is re-dissolved. This boiling should last several minutes
to insure the completion of the reaction.
3. Disconnecting the junctions between the absorption-
flasks and solution-flask, and between the solution-flask
and the acid-flask, and allowing to cool.
4. Mixing solutions, filtering out residue which is
placed (with filter-paper) in a porcelain dish, oxidising
E E
418 ESTIMATION OF PHOSPHORUS IN IRON AND STEEL.
with nitric acid and potassium chlorate, and evaporating
to dryness.*
5. Boiling the solution till the excess of sulphurous
acid is expelled, adding a few c.c. of permanganate to per-,
oxidise a little of the iron, and precipitating basic acetates.
Boiling the nitrate for other precipitates, to insure getting
all the phosphoric acid present.
6. Dissolving these precipitates in hydrochloric acid
and adding to the solution of the residue, in which by this
time the paper will have been thoroughly destroyed.
7. Evaporating to dryness for silica, and proceeding as
usual with the molybdate precipitation.
The form of apparatus used in these estimations is
shown in fig. 104. It consists of a flask, A, to hold the acid ;
FIG. 104.
a flask, B, for the iron ; and flasks c and D, of which c is
generally left empty to catch condensed steam, and the
two flasks D contain the absorbent liquid.
The absorbent liquid in D is a solution of potassium
permanganate acidified with nitric acid. The iron having
been placed in B, and the necessary amount of dilute
hydrochloric acid to dissolve it in A, the whole apparatus
is filled with oxygen, a current of which is kept passing
through the whole time. When the air is all expelled
* It is better to filter out the residue instead of filtering it out with the
first basic acetate precipitate, because it is difficult to work accurately on a
muddy liquid, and because there is a probability of loss of phosphorus by the
continued action of the hot acid solution on the residue after the sulphurous
acid has been expelled. By proceeding as directed this chance of loss is
removed.
ESTIMATION OF PHOSPHORUS IN IRON AND STEEL. 419
the flask A is inverted above the level of B, so that its
contents flow into B and the solution of the iron com-
mences. The solution is heated to boiling for some
minutes ; and when all has dissolved with the exception
of the insoluble portion, the apparatus is disconnected,
the insoluble residue filtered from the solution, the solution
-oxidised with nitric acid and treated as described above.
The saving of time by this process is the greater as
the percentage to be estimated is less. In a steel of low
percentage, where it is necessary to use 10 grms. for the
-estimation, necessitating the use of 120 c.c. of nitric
acid for solution, by the old method, the time taken for
the evaporation of this amount of liquid, for the subse-
quent thorough drying of the residue, and for the reso-
lution and reduction of the iron to the ferrous state, is
evidently much greater, and the operations followed are
much more tedious, than by this method, where the phos-
phorus is concentrated at the start in but a few milli-
grammes of iron in very little bulk of solution, enabling the
subsequent evaporation to dryness and thorough drying
of the residue to be performed in a very short period.
In M. Tantin's experiments he used cast iron only. Mr.
Mackintosh's have been conducted on pig and wrought
iron. It is possible that the difference in the results may
be partially accounted for by different modes of occurrence
of phosphorus similar to the different forms of carbon ; but
this is only offered as a suggestion.
In calculating the percentage of phosphorus from the
weight of the yellow precipitate, Mr. Mackintosh uses in
all cases the figure of 1'63 per cent, phosphorus in the
precipitate.
Dr. J. Lawrence Smith was engaged off and on for two
or three years, examining the question of the estimation
of phosphorus in iron and steel, making several hundreds
of variously modified experiments, and repeating the
details of processes adopted by various chemists. The
following method was ultimately adopted as affording the
most speedy and accurate results.
Quantity of Iron employed. It is customary to employ
E E 2
420 ESTIMATION OF PHOSPHORUS IN IRON AND STEEL.
1 grm. for pig iron, and 2 to 3 grms. for malleable iron
and steel ; but Dr. J. L. Smith employs but 1 grm. for all
varieties of iron ; for even where the iron or steel contains
one- thousandth and less of phosphorus, as satisfactory
results are obtained as where 2 and 3 grms are employed.
Solution. The iron, say 1 grm., is placed in a por-
celain capsule of about 100 or 150 c.c., and 3 or 4
c.c. of water added ; the capsule is placed on a water-
bath, and 10 to 15 c.c. of aqua regia are added little by
little ; the aqua regia is prepared in advance in the usual
way with 2 parts hydrochloric acid and 1 part nitric acid.
The contents of the capsule are now evaporated to dry-
ness over the water-bath, or more speedily on an iron
plate ; the capsule with its contents is then placed in an
air-bath and heated from 140 to 150 C. for from 30
minutes to 1 hour, thus rendering all the silica insoluble ;
3 or 4 c.c. of hydrochloric acid with an equal quantity of
water are added to the dry residue, and then warmed
gently over a water-bath or lamp ; the iron is re-dissolved,
a little more water added, the solution filtered with the
filter-pump, the filtrate placed on a narrow graduated
measure of 100 c.c. capacity, and sufficient water added
to make the liquid contents 100 c.c. ; the whole is well
shaken to make the solution uniform. The next step is
to concentrate all the phosphorus into a limited amount of
the iron.
Concentration of the Phosphorus. From 90 to 92 c.c.
of the last solution is placed in a capsule of 300 or 400
c.c. capacity, either of porcelain or platinum the latter
by preference and 100 c.c. of water added ; the iron
oxide is now reduced to iron protoxide by ammonium
sulphite.* Two or three centimetres of the ammonium
sulphite are added to the iron solution and the contents of
the capsule are boiled until all the sulphurous acid is driven
off, this stage of the process being recognised by the sense
of smell. By putting a small drop of the solution on the
* Equal parts of ammonia and water are placed in a bottle and an excess
of sulphuric acid passed through ; the operation lasts for several hours, using
a mixture of charcoal and sulphuric acid. Once prepared, it keeps very well,
when kept from the light.
J. LAWRENCE SMITH'S PROCESS. 421
end of a glass stirrer into a weak ammonia solution we
readily recognise the complete conversion of the oxide, for
the precipitate is nearly white. Of course during the
whole of the above process the solution is acid, with the
excess of hydrochloric acid. Ammonia is now added
slowly to the warm solution until a little of the greenish
precipitate remains undissolved ; about 20 c.c. of acetic
acid is now added to the solution (which immediately re-
dissolves the precipitate), and then 1 or 2 c.c. of ammonia
acetate solution; finally, the 8 or 10 c.c. of original solu-
tion remaining in the graduated glass is added with 200
or 300 c.c. of water.
The whole contents of the large capsule is boiled
gently from one-half to one hour, and if necessary the
water renewed as it is evaporated. The result is the for-
mation of a basic per-salt of iron containing practically
all the phosphorus that was originally in the gramme of
iron used.
Separation of the Phosphorus from the above Precipi-
tate. With a filter-pump on a 3^-inch filter, the last pre-
cipitate is collected in 15 or 20 minutes. The precipitate
is not washed, but a mixture of 5 or 6 c.c. of hydrochloric
acid, with an equal quantity of water, is warmed in the
capsule in which the boiling has taken place, so as to
dissolve the adhering oxide of iron ; the hot acid solution
is thrown on the filter in the funnel, detached from the
pump ; the filtrate is readily dissolved, and passes into some
convenient vessel, and the filter washed once or twice ;
this solution is placed in a porcelain capsule and evapo-
rated to dryness over a water bath or on a hot plate. The
former is preferable, although it takes a longer time. To
the dry, but not over-heated residue is added 1 to 2 c.c.
of nitric acid, with an equal quantity of water. This will
furnish a clear solution if there be no titanium in the
iron ; if the latter be present, there will be formed a floccu-
lent precipitate that can be readily separated by a filter
prior to the last treatment.
The last Treatment. The solution now need not be
more than 10 or 20 c.c., to which ammonia is to be added
422 ESTIMATION OF PHOSPHORUS IN IRON AND STEEL.
until the precipitate first formed is no longer re-dissolved ;
then add a few drops of nitric acid to clear up the solution
completely, in which the phosphorus is supposed to have
been concentrated. 30 c.c. of molybdic acid solution is
now added to the last solution in a small beaker, which is
then warmed for 15 or 20 minutes to a temperature of
80 0., and agitated with a glass rod. The phosphorus is
precipitated as the double ammonia-salt, and settles as a
chrome-yellow powder in less than 30 minutes, and is
ready for collection on a double filter ; * although it is
better to allow two hours or more time to elapse before
filtering and washing with the filter-pump. As the filter
is very small, it is readily washed with a little distilled
water.
After washing, the double filter is placed in an air-bath
heated to about 120 C., and in about 30 minutes weighed
by separating the filters ; the complete dryness is verified
by a second heating in the air-bath.
Of the phospho-molybdate every ] 00 m.g. will contain
1-63 m.g. of phosphorus, or 3*74 m.g. of phosphoric acid.
The result of this method of analysis will indicate a very
minute quantity of phosphorus less than what is contained
in the iron, but so small as not to affect the practical result,
and will be more accurate, certain, and speedy than if
estimated as magnesian phosphate.
Estimation of Manganese in Iron. After the silicon
is estimated in the iron or steel by Eggertz's method
(p. 408), the manganese may be estimated in the same-
amount of material. The filtrate from the silicate is diluted
with water until it measures 400 c.c., and accurately
divided into two portions of 200 c.c. each, one of which is
set on one side, and in the other the manganese is estimated
in the following manner : (In the case of wrought iron
* When filtering a precipitate to be weighed on the filter, a double filter
is used, each of the same size ; they are weighed one against the other and
exactly balanced by the weights ; on the lighter one a -f mark is put with
pencil, and the number of m.g. that it is lighter than the other. As only a 2
or 3 inch filter is used, the difference in weight between the filters does not
usually exceed 10 or 20 m.g. A number of these double filters (with the differ-
ence marked on them) may conveniently be kept ready for this or any other
purpose.
ESTIMATION OF MANGANESE IN IRON. 423
and steel, where 3 grammes are taken, the solution is
diluted to 200 c.c., and the manganese estimated without
dividing the solution.) A saturated solution of sodium
carbonate is added to the solution until it becomes nearly
neutralised, appearing of a deep brown colour ; water
containing 5 per cent, of sodium carbonate is then added,
drop by drop, until a slight turbidity occurs in the solu-
tion ; and if, on standing in the cold, this turbidity rather
increases than diminishes, sufficient carbonate has been
added (if too much sodium carbonate has been added, and
a precipitate is thrown down, the excess must be neutralised
by hydrochloric acid) ; to the slightly turbid solution add
1-^ c.c. of hydrochloric acid, and heat on the water-bath
until the solution becomes clear ; dilute with about half
as much water as the volume of the solution, and add 30
c.c. of a saturated solution of sodium acetate ; boil for a
quarter of an hour ; allow the precipitated iron to settle,
and decant the clear liquid through a filter, washing the
iron by decantation with boiling water containing -J per
cent, of sodium acetate ; finally, throw the iron o$ the
filter, and continue the washing until bromine water gives
no reaction, showing that all the manganese has passed
through the filter ; evaporate the filtrate down to 400 or
500 c.c. ; and at the temperature of 50 C. add a few drops
of bromine to precipitate the manganese, and keep it near
to that temperature for twelve hours, stirring occasionally
with a glass rod. The solution after the addition of the
bromine becomes of a yellow or brownish colour, but
should be perfectly colourless before filtering. The man-
ganese is now thrown on a filter which has been dried at
100 C., and accurately weighed, washed with cold water
containing 1 per cent, of hydrochloric acid, dried at 100
C., and weighed. The precipitate is a hydrated manga-
nese oxide, containing 59'21 per cent, of manganese. The
.precipitate may also be ignited in a porcelain crucible at
a white heat, and is then an anhydrous manganese oxide,
containing 72-05 per cent, of manganese. If copper is
present it must be removed previous to the precipitation
of the manganese ; or the amount of copper found in the
424 ESTIMATION OF MANGANESE IN IRON.
ignited oxide, and then an equivalent amount of copper
oxide, subtracted from the total weight of the precipitate.
In using the first method, 20 grains of the finely
divided spiegeleisen are completely dissolved in hydro-
chloric acid, diluted, and a current of sulphuretted
hydrogen passed through the liquid. After standing for
twelve hours, the solution is filtered and washed with
water containing sulphuretted hydrogen ; the filtrate is
boiled, 10 grs. of potassium chlorate added, the iron sepa-
rated, and the manganese estimated in the usual manner.
If the method used be that of estimating the copper in
the precipitate, the estimation must be made with the
greatest care, on account of the small quantity of copper
present ; the solution must be decanted immediately the
zinc is completely dissolved, and excess of acid must be
carefully avoided ; otherwise the film of copper will par-
tially re-dissolve. It is evident that if the precipitation
be effected by ammonium sulphide or sodium carbonate
separation or estimation of the copper is likewise neces-
sary.
Mr. E. Eiley ('Chemical News,' April 27, 1877) gives
the following instructions on the estimation of manganese
in spiegeleisen, and in many auriferous iron ores :
There are two methods now in use (a) The Direct
Method. The pulverised spiegeleisen (about 1 grm.) is
dissolved in dilute nitric acid, sp. gr. 1*2, a little potassium
chlorate and hydrochloric acid added to destroy the soluble
organic matter from the combined carbon ; the solution,
diluted to about a litre, is neutralised with sodium or ammo-
nium carbonate, sodium or ammonium acetate added, the
solution boiled, the basic iron peracetate allowed to settle,
and filtered off. This precipitate is re-dissolved in hydro-
chloric acid, and the process repeated to insure complete
separation of the manganese. The filtrates are evaporated
to 1-^ litre, allowed to cool, 2 to 4 c.c. bromine added,
the solution well shaken, 0*88 ammonia added in excess,
the solution heated gradually for an hour, boiled for a
few minutes, the precipitate allowed to settle, filtered (the
filtrate should be evaporated and tested for manganese),
MR. EILEY'S PROCESS. 425
dried, and ignited in a muffle, or over a gas blowpipe for
half an hour.
(b) The Indirect Method. The finely powdered spiegel-
eisen (about 1 grin.) is dissolved in dilute sulphuric or in
hydrochloric acid, the liquid diluted with recently boiled
and cooled distilled water, and the iron estimated volu-
metrically ; to the percentage of iron thus obtained 5
per cent, is added for carbon and impurities ; the difference
is assumed to be manganese.
The results obtained by this method are usually too low,
from the formation of soluble organic matter during the
process of solution. This error can be obviated by using
nitric acid for a solvent, evaporating to dryness and heat-
ing ; the iron and manganese oxides are then dissolved in
hydrochloric acid, the solution largely diluted, and reduced
with sodium sulphite. The results thus obtained agree
very closely with the direct method. Thus, for all practical
purposes, the indirect method is sufficiently accurate, and
can be accomplished in one hour, the direct method re-
quiring five or six hours. The author strongly recom-
mends the use of ammonium acetate and carbonate, instead
of the corresponding soda salts in the direct method ; and
proves by check experiments with pure manganese sul-
phate, &c., the statements of Fresenius and others, that the
presence of ammoniacal salts prevents the complete pre-
cipitation of manganese by bromine and ammonia, to be
erroneous. On the other hand, if soda salts be used, the
ignited precipitate will contain soda. The author con-
siders that sulphur cannot exist in spiegeleisen. He es-
timates the carbon by dissolving the iron in neutral
copper chloride, and after complete solution of the iron
and precipitated copper, the carbon is filtered on asbestos,
and burnt with copper oxide in a current of oxygen.
The carbon estimations by the colour test are unsatis-
factory for high percentages of carbon. According to the
author, the percentage of carbon varies with the percentage
of manganese.
426 ASSAY OF MANGANESE IN SPIEGELEISEN.
Assay of Manganese in Spiegeleisen, Ferro-manganese, and
the most important ores.
W. Kalman and Alais Smolka have ascertained that
manganous oxide when opened up with a flux of borax
and potassium-sodium carbonate with access of air yields
an oxidation product containing 5 atoms available oxy-
gen and 6 atoms manganese. The flux is obtained by
melting in a platinum capsule 2 parts borax glass and
3 parts of the double carbonate, and pulverising the
very hygroscopic mass while still warm. There is be-
sides required a solution of ferrous sulphate and a per-
manganate solution. [The former is obtained by dissolving
about 100 grammes ferrous sulphate in 1,000 c.c. water,
acidulating with sulphuric acid, filtering, and mixing the
solution with 100 c.c. of pure undiluted sulphuric acid.
The permanganate solution is standardised for iron, 1
c.c. of the solution preferably representing 0-0025 grm.
iron. Its standard for manganese can be calculated from
the following proportion : iron standard : x 10-56 :
6-55.
For applying the method, from 0-15 to 0*30 grm.
of the sample, very finely ground, is ignited for fifteen
minutes in an open platinum crucible with a Bunsen
burner, and then more strongly with a blast. By the
ignition the manganese is chiefly converted into mangano-
manganic oxide. The crucible is let cool, covered, and
about 20 parts of the flux weighed in. Heat is slowly
applied till the mixture is melted, care being taken that
not much spirts up upon the lid, as such portions become
oxidised to manganate and make the result too high.
The contents of the crucible are kept in a state of fusion
for fifteen to twenty minutes, the lid is then removed, the
crucible placed slanting, and the fusion is continued for
five minutes longer, stirring with a platinum spoon.
Equal quantities (10 to 15 c.c.) of the iron solution
are poured into two beakers and diluted. The contents
of the crucible are dissolved in the solution in one of the
ASSAY OF MANOANESE IN SPIEGELEISEN. 427
beakers, adding a little strong sulphuric acid if requisite,
when the manganese compound formed oxidises a part
of the iron. The solution of ferrous sulphate in both
glasses is then titrated with permanganate. The difference
multiplied by the manganese standard gives the propor-
tion of manganese in the sample. In cases where the
manganese exists as a silicate the results are only approxi-
mate. It is suitable for all cases where the manganese in
the sample may be converted into mangano-manganic
oxide by simple ignition, where the proportion of man-
ganese is at least 1 to 2 per cent., and where no other
substance is present which can become capable of giving
up oxygen to a ferrous solution, such as chlorine.
The accurate estimation of manganese in spiegel-
eisen is of commercial importance ; as, being the most
important constituent, the value of the material is fre-
quently judged by the percentage of that element alone,
while the error introduced by the presence of copper is
aggravated by the fact that not only is copper worthless
but absolutely injurious.
The following method for the assay of manganese in
iron and steel by Mr. Samuel Peters, Bay State Iron Works,
South Boston, is not new in principle, but has given very
satisfactory results :
Dissolve O'l grm. pig iron or steel in 3 or 4 c.c.
nitric acid, about 1-2 sp. gr., and boil gently in a long
test-tube (about 8 inches long and f-inch diameter) for
five or ten minutes, or until solution is complete ; then
add an excess of plumbic oxide, say 0-2 or O3 grm.,
and boil again two or three minutes.* Cool the tube and
its contents in water. Filter through asbestos, washing
out the test-tube and the residue on the filter with dis-
tilled water until all the colour has been washed through.
Transfer to a graduated tube (f-inch in diameter), hold-
ing 50 or 60 c.c., graduated in 0*2 c.c., and compare with
a standard solution of permanganate held in a tube for
that purpose. The comparison is made in the same
* It is unnecessary to filter off graphite in pig iron before boiling with
plumbic oxide.
428 ASSAY OF MANGANESE IN SPIEGELEISEN.
manner as that in the Eggertz method when estimating
combined carbon in steel, &c. The solution under com-
parison is then diluted and well mixed with distilled
water (by pouring the contents of the graduated tube
into a small dish, and then transferring to the tube again),
until its colour is exactly of the same intensity as the
standard solution. Having attained to this point, the
number of c.c. is noted, and the result is obtained by
multiplying each c.c. by O'OOOl. Each c.c. is equivalent
to 01 per cent, manganese when Ol grm. of iron is
taken for analysis.
For irons containing O'lO to 0*35 per cent, manganese,
Ol grm. is the proper quantity ; but if there be, say,
0-8 to TOO per cent., it is best to take 0-1 grm. and
divide the solution (before adding the lead peroxide) in
four equal parts, and use O25 for the estimation, taking
another 0'25 for a second estimation. In case of a high
percentage, as 1 per cent., if O'l grm. is taken the re-
sults are too low on account of some of the manganese
escaping oxidation. This agrees with the observations
of others. With an unknown iron, one or two trials
with O'l grm., or half that quantity, will point out the
probable amount, and so be a guide for the next trial.
If the amount of iron taken does not yield more colour
than corresponds to 25 to 35 c.c. of standard hue, it may
be safely said that all the manganese is oxidised. It is
as well to take this volume as the guide to the quantity
of iron to be taken. The quantity of manganese in the
liquor to be tested should not exceed 0'4 of a millgramme,
and certainly not over half a milligramme. By taking 0*1
grm. of a spiegeleisen containing nearly 12 per cent,
manganese, and diluting to 50 c.c., and taking 2 c.c. or
0'04 for the estimation of the manganese, very nearly the
proper amount of manganese is obtained. This seems to
show that if the division of the solution can be accurately
made, and the bulk of the coloured liquid can be kept
down well, the amount of manganese in spiegeleisen can
be estimated very fairly.
Combined carbon in large quantity does not inter-
ASSAY OF MANGANESE IN SPIEGELEISEN. 429
fere with the accuracy of the method, for a steel con-
taining 2 per cent, combined carbon and only 0-8 per
cent, manganese was found to give good results by this
method.
The standard is made by diluting a permanganate of
potash solution of known strength until each c.c. = 0*00001
grm. manganese.
For example, a T n F solution will contain 3-16 grm.
permanganate in 1000 c.c. or 0*0011 grm. manganese
per c.c. ; if this be diluted 110 times it will give the
required strength. The standard is contained in a tube
of the same bore as the one used for the analysis ; or else
the standard is put in the latter one, and a solution of
permanganate put into a tube of nearly the same bore,
and diluted until it exactly corresponds with the standard
solution, when it will serve as a standard.
Permanganic acid of the proper hue keeps better than
permanganate of potash of the same hue, and is of course
easily made by adding nitric acid to the latter. The time
occupied in obtaining a result by this method is very
short (about half an hour), and it is a method that will
prove of advantage in analysing steel made by the Bes-
semer and Siemens-Martin processes.
Mr. Galbraith (< Chemical News,' February 4, 1876)
has given the following simple and accurate process for
the assay of manganese in spiegeleisen : 1 gramme of the
spiegeleisen is dissolved in nitric acid (sp. gr. 1-20) in a
small round-bottomed flask, and evaporated to dryness.
When dry, the flame, which may be either a spirit-lamp or
a Bunsen burner, is turned so that the bottom of the flask
is cherry-red for ten minutes. It is then allowed to cool
very gradually.
At this point a weighed quantity of ammonio-ferrous
sulphate, or ferrous sulphate of a known strength, is put
into the flask and then heated with rather dilute hydro-
chloric acid. The contents of the flask very soon dissolve ;
but it is well to keep shaking the solution while it is being
heated, to prevent loss of chlorine. It only remains now
to estimate the iron left unoxidised, in order to arrive
430 ESTIMATION OF TITANIUM IN IRON.
at the quantity of manganese, which can be done with
potassium bichromate solution. It is feared that the ferrous
solution may get oxidised by exposure to the air ; a small
piece of marble put into the flask, which can also be fitted
with a cork and tube, will readily prevent that.
In four successive experiments the following results
were obtained :
No. Fe Equal to
Oxidised Manganese p. c.
1 0-2018 19-82
2 0-2103 20-65
3 0-2396 23-53
4 0-2435 23-88
No. 2 gave by ammonium acetate method 20-55 per
cent., which was done with great care. No. 4 is a repeti-
tion of No. 3.
It is evident, of course, that there is nothing original or
new in the above method ; but it contrasts very favourably
with the usual methods of separating the iron with sodium
or ammonium acetate, and precipitating the manganese
from the filtrate with bromine. It is not at all trouble-
some, does not take long, and has the advantage that the
only chemicals and apparatus required are those which
are necessary for the assay of iron ores.
Estimation of Titanium in Iron. The detection of
titanium in iron is easy, although its estimation is difficult.
The best results have been obtained by following Kiley's
plan.* This is essentially as follows : A weighed portion
of the iron borings are treated with fuming nitric acid in a
flask, a few drops of hydrochloric acid added from time to
time, the whole being well boiled. The contents of the
flask are then transferred to a porcelain dish, evaporated to
dryness, and heated strongly. On cooling, it will be found
that the iron oxide readily detaches itself from the dish,
and can be easily transferred to a beaker, the portions
left on the dish being dissolved in hydrochloric acid, and
poured on the contents of the beaker ; the dish may be
rinsed out, if necessary, with strong hydrochloric acid.
The contents of the beaker are boiled for from two to three
* Chemical News, viii. 226, 233.
ESTIMATION OP TITANIUM IN IEON. 431
hours, until complete solution of the iron is effected ; and
as some quantity of hydrochloric acid is required for this,
the best plan is to allow a large portion of the excess of
acid to evaporate in the beaker, retaining only as much
as is requisite to keep the iron in solution. The silica
is filtered off in the usual way, after diluting with water
and adding a few drops of hydrochloric acid on the filter
to dissolve the basic salt formed by the water. By this
means the silica can be obtained nearly white after burning
off the graphite, and very little iron will be found with it
unless much phosphorus be present, as the silica invariably
contains more or less iron phosphate from the insoluble
iron phosphide, which cannot be completely dissolved out
by hydrochloric acid. Before estimating the titanium
the residue from the silica should be fused with potassium
bisulphate, dissolved in water, and added to the solution of
iron in which the titanium is to be estimated. The solu-
tion is reduced with sodium sulphite, and the excess of
sulphurous acid is driven off by boiling. The solution is
then near]y neutralised with ammonia, and ammonium or
sodium acetate added ; if there is only a small quantity of
phosphoric acid, there will always be sufficient peroxide
of iron to precipitate it, but if not, a few drops of nitric
acid must be added so that the precipitate produced is
distinctly red, and the solution boiled and filtered as
quickly as possible. The residue is fused with potassium
bisulphate, or, where nitric acid is used, this is driven off
with sulphuric acid. The result of the fusion with potas-
sium bisulphate is dissolved in cold water (when a little
iron phosphate, which remains insoluble, is separated),
boiled for some hours, and allowed to stand a night in a
warm place, when the titanic acid is filtered off and
washed with dilute sulphuric acid, dried, ignited, and
weighed.
The above process is not very satisfactory for the
quantitative estimation of titanic acid. The iron phos-
phate (insoluble in the potassium bisulphate) cannot be
washed without its passing through the filter ; and very
frequently, also, the small amount of iron keeps up the
432 ESTIMATION OF TITANIUM IN IRON.
titanic acid, as iron even in small quantities has a very
great effect in preventing the precipitation of titanic acid ;
so that it is always advisable to add a little sodium
sulphite, which reduces the iron oxide and facilitates the
precipitation of the titanic acid.
Titanium may, however, be found more satisfactorily
and more readily during the process usually adopted to
estimate the amount of graphite in pig iron, provided a
large quantity of the pig be operated on. About 200
grains of the pig are to be dissolved in dilute hydrochloric
acid ; when the pig is nearly all dissolved, and the action of
the acid has ceased, more hydrochloric acid is added, and
the solution well boiled, so as to thoroughly extract all the
iron. The solution is then thrown on dried counterpoised
filters encircling each other, and the filter well washed to
remove all the iron. It is then treated with dilute potash,
and washed once ; then re-treated with it so as to entirely
remove the silica. The potash is thoroughly washed out,
and the filter treated with hydrochloric acid, thoroughly
washed, and dried at 250 F. until the weight is constant.
This gives the graphite, on burning which a residue of a
dirty light brown colour is left, which, fused with potas-
sium bisulphate, and subsequent treatment as above ex-
plained, is seen to be nearly pure titanic acid.
Mr. W. Bettell (' Chemical News,' Aug. 22, 1873) pro-
poses the following modification of Mr. D. Forbes' process
(' Select Methods in Chemical Analysis,' second edition, pp.
194, 195) for the estimation of titanic acid, which may not
be unacceptable to those engaged in the analysis of titanic
ores :
Fuse about 0-5 grm. of the finely powdered ore with
6 grms. of pure potassium bisulphate (which has been
recently fused and powdered) in a platinum crucible at a
gentle heat, carefully increased to redness, and continued
till the mass is in a state of tranquil fusion.
Eemove from the source of heat, allow to cool, digest
for some hours in 5 or 6 oz. of cold distilled water (not
more than 10 oz. is to be used, as it generally causes a
precipitation of some titanic acid) ; filter off from a little
ESTIMATION OF HAKDNESS.
433
pure white silica, dilute to 45 or 50 oz., add sulphurous acid
till all the iron is reduced, then boil for six hours, replac-
ing the water as it evaporates.
The titanic acid is precipitated as a white powder,
which is now to be filtered off, washed by decantation,
a little sulphuric acid being added to the wash-water to
prevent it carrying titanic acid away in suspension. Dry,
ignite, allow to cool, moisten with solution of ammonium
carbonate, re-ignite, and weigh. The titanic acid is in-
variably obtained as a white powder, with a faint yellow
tinge, if the process has been properly carried out.
This method of fusing with potassium bisulphate
(' Select Methods in Chemical Analysis,' 2nd edition, p.
212) is preferable to all others for decomposing difficultly
soluble iron ores.
The Hardness of Iron and Steel.
Mr. T. Turner has given the following useful table
of the hardness of different varieties of iron and steel in
comparison with that of other bodies. The figures repre-
sent the weight in grammes necessary to produce a scratch
with a diamond on drawing its point over the smooth sur-
face of the metal.
Substances
Steatite
Lead (commercial)
Tin .
Eock salt
Zinc (pure annealled)
Copper (pure anne
Calcite .
Softest iron
Fluor-spar
Mild steel
Tyre steel
Good cast iron
Bar iron
Apatite .
Hard cast-iron scrap
Window -glass
Good razor-steel
Very hard white iron
Relative
hardness
'
1
)
1
2-5
4
ed)
6
aUed
t
8
12
15
*
L -
19
21
20^24
- -
2124
24
34
rap
36
60
\
60
ron
72
F F
434
CHAPTER X.
THE ASSAY OF COFFEE.
IN the assay of copper by the dry way, all minerals and
substances containing that metal may be divided into
three classes.
CLASS I. Comprises Sulphuretted Ores or Products, with
or without Selenium, Antimony, or Arsenic.
Copper glance, Cu 2 S, containing 79'7 p. c. of copper
Chalcopyrite, Cu 2 S, Fe 2 S 3 , 34-4
Erubesoite, 3Cu 2 S, Fe 2 S 3 , 55'7
Bournonite, 3Cu 2 S,SbS 3 + 2(3PbS,SbS 3 ) 12-7
Fahlerz, 4(Cu 2 S,FeS 5 ZnS,AgS,HgS).(SbS3,AsS 3 ,Bi 2 S 3 )
3048
Covelline CuS, 66'7
Wolfsbergite, Cu 2 S,SbS 3 , 24'9
Domeykite, Cu 6 As, 71'6
Copper regulus, Copper speiss, &c.
CLASS II. Oxidised Ores and Products.
Red copper, CojO, containing 88'7 per cent, of copper
Malachite, 2CuO, + H 2 0, 57'3
Azurite, 2CuO,C0 2 + CuO,H 2 0, 55-1
Cyanosite, CuO,S0 3 +5H 2 Q, 25'3
Phosphate of copper, 30 56
Arseniate of copper, , , 25 50
Chromate, Vanadate, and Silicate of Copper ; Slags, &c.
CLASS III. Copper and its Alloys.
The different methods of assaying copper are more
numerous than those for any other metal. They are in some
cases similar to each other, and in others based upon very
different principles.
CLASSIFICATION OF THE COPPER ASSAYS. 435
These methods may be divided into :
A. ASSAY IN THE DRY WAY.
a. For Rich Ores and Products of Class I.
I. English Copper Assay.
b. For Ores and Products of Class II.
1. Lake Superior Fire Assay,
B. ASSAYS IN THE WET WAY.
a. Colorimetric Copper Assay.
6. Volumetric Copper Assay.
c. Electrolytic Copper Assay.
A. ASSAYS IN THE DRY WAY.
a. For Rich Ores and Products of Class I.
I. ENGLISH COPPER ASSAY.
M. L. Moissenet has given, in the 'Annales des Mines,'*
a very complete description of the Cornish method of
assaying copper by the dry way. The following is from a
translation by Mr. W. W. Procter.
Each of the large Swansea copper-works keeps an
assayer at Cornwall, whose duty it is to estimate the
richness in copper of all the lots of minerals in the county
sold every Thursday at the Ticketing, and of all the
samples of foreign minerals and copper products which
may be useful to the smelter.
The copper being obtained in the state of a prill or
metallic button, the impurities (generally tin, antimony,
&c.) are thus made evident, and the hammer soon proves
the quality of the metal which we ought to expect to
obtain by metallurgic treatment. As for the accuracy of
the method, as far as regards the whole of the metal
obtained, we shall revert to this later on. We would,
however, observe that, within certain limits, the method
would not be less practical on account of being inexact ;
* Vol. xiii. p. 183.
p p 2
436 ENGLISH COPPER ASSAY.
for we must not forget that it has chiefly for its object
to teach the smelter the value of the mineral, even more
than its true richness.
For example, if we get too low an assay from a sample
of 2 or 3 per cent., we should only from this assent to the
opinion of the metallurgist, whose interest it is not to work
upon very poor minerals. The same remark will apply to
the case of minerals very antimonial, &c. Besides, in the
description of the method we shall discover the principal
phases of the Welsh process ; so that it is more just to con
sider the Cornish assay as a metallurgy on a small scale than
as a scientific laboratory method. From thence result also
the necessity of long practice and the almost uselessness of
theoretical knowledge-for those who purpose employing
this method alone.
Sir Henry de la Beche (' Eeport on the Geology of
Cornwall,' &c., p. 595), in giving a sketch of the method,
declares it to be rather rough and uncertain, and fails not
to add at the conclusion a translation of a passage relative
to the assay of copper pyrites from M. Berthier's treatise
on assays by the dry way.
These drawbacks upon the scientific value of the Cornish
method cannot injure the power of facts ; they constitute
but another reason which we may have for giving an ac-
count of the manner in which the first basis of the valua-
tion of the greater part of the copper minerals has been
fixed since so long a period.
Division Adopted. The rather complex operations
through which we have to pass will be better apprehended
by explaining in succession
1. The order of the operations, the nature and influence
of the fluxes employed, the kind of products obtained
(reactions).
2. The manipulations to which each operation gives
rise, the furnaces and apparatus used, the characters of the
principal products during the chief phases and at the end
of each (manipulations).
We shall add to these
3. Some information upon the influence of the princi-
ENGLISH COFFEE ASSAY. 437
pal foreign metals (tin, antimony, zinc, lead), and upon the
treatment of some special coppery matters.
4. Summary considerations on the result of the English
method compared with those of the analysis by the wet
way.
SECTION I. EEACTIONS.
At the very outset we distinguish two kinds of assays.
1. The roasted sample.
2. The raw sample.
The first only applies to cupreous pyrites or to samples
essentially formed of it that is to say, which contain
sulphur in excess ; the process begins by a roasting.
In the raw assay we dispense with the roasting ; we
have recourse to the addition of reagents, either oxidising
or sulphurising, according to the minerals ; we endeavour
to place them by these mixtures in the condition of a pro-
perly roasted pyritic mineral.
From this point, at least in general, the operations
become identical. They consist in
1. Fusion for regulus (regulus).
2. Calcining the regulus (calcining).
3. Fusing for coarse copper (coarse copper).
4. One or two fusions with fluxes (washings).
5. Trial by striking with a hammer, last refining (test-
ing, refining).
6. Treatment of slags for prill.
All the slags except those of the fusion for regulus have
been preserved. The fusion No. 6 gives a small sup-
plementary button of copper, which again undergoes, if
necessary, one or two washings.
As we have said, the roasting is used only for pyrites.
We shall return later on to the duration and the circum-
stances of this operation. Its evident aim is to drive off
the excess of sulphur, so as to cause the whole of the
copper, with a part only of the iron which abounds in the
pyrites, to pass into the state of sulphide at the time of
the fusion for regulus.
438 ENGLISH COFFEE ASSAY.
I. Eegulus.
1. Pyrites. The fusion for regulus of a properly
roasted pyrites is made by mixing with it equal volumes
of the three fluxes borax, fluor-spar in powder, lime
slaked in powder of each one ladle, and covering the
mixture with a layer of moist common salt. The matters
composing the gangue of the roasted mineral consist prin-
cipally of quartz, silica, and in general of more alumina
and magnesia than lime ; oxide of iron, resulting from the
roasting of the pyrites, is also present.
The borax only serves to give fusibility, the fluor-spar
contributes to the same end by forming a fluosilicate.
Otherwise it does not play an important part in the de-
composition that is to say, there is probably no produc-
tion of fluoride of silicon and calcium ; for this last base is
added here in considerable proportion, so as to form im-
mediately a silicate which may combine with the fluoride
of calcium.
The ferric oxide being so reduced as to pass into the
slag, and the different metallic oxides to pass into the re-
gulus, yield oxygen, which reacts on the remaining sulphur.
The disengagement of sulphurous acid which results from
this, joined to the water contained in the fluxes, justifies
to a certain extent the use of a layer of common salt,
designed to prevent the boiling over. Besides this, the
common salt, being without action on the metallic sul-
phides, does not here produce those important effects
which it exerts in the later fusions.
If the pyrites appear insufficiently roasted, we must
add a little nitre, the oxidising action of which again gives
off sulphur ; the opposite case, that of a roasting too
much prolonged, is rare ; we remedy it by the addition of
sulphur and tartar.
2. Very poor Pyrites. In a very poor pyrites that
of Bear Haven, in Ireland, for example the proportion of
sulphur does not require us to have recourse to the roast-
ing ; we employ the three fluxes and one ladle of nitre.
ENGLISH COPPER ASSAY. 439
3. Variegated Copper Ore. Peacock ore contains less
sulphur in proportion to the copper than pyrites ; we also
fuse with a little nitre.
4. Sulphide of Copper. The sulphur is here insuffi-
cient. We add together sulphur -J to 1 ladle, according
to the valuation ; tartar J to ^ ladle that is to say, half
the volume of the sulphur. The tartar is a powerful re-
ducing agent, and is supposed in small quantities to favour
the action of the sulphur by preventing its disengage-
ment as sulphurous acid by the oxidising matters in the
mineral ; but if used in excess, it acts as a desulphuriser,
as well by its carbon as by its alkali.
5. Carbonated Minerals. The addition of sulphur and
carbon is evidently still more necessary here.
6. Native Mixture : f sulphide of copper, ^ pyrites.
We add, in this case, nitre for the pyrites, and sulphur and
tartar for the sulphide of copper ; although these reagents
appear sure to neutralise each other, it is possible that their
simultaneous employment may be logical. The nitre pro-
bably decomposes the pyrites, which would without it fuse
and give a very ferrous regulus, whilst the free sulphur
would be of little use, on account of the sulphide of copper.
Be this as it may, this is the plan adopted.
During the progress of the fusion for regulus we have
still to introduce other matters, some incidentally, and
others in all cases.
If a blue flame persists in escaping from the crucible,
an index of the formation of sulphurous acid, we project
into it sulphur 1 ladle, tartar ^ a ladle. When the fusion
appears almost finished, in order to render the bath more
liquid, and to facilitate the collection of the button, we
throw in a little dried salt, and a flux composed before-
hand of lime, a little fluor-spar, and a very little borax
that is to say, of the elements in different proportions of
the mixture introduced originally.
The regulus obtained is composed principally of copper,
iron, and sulphur. We shall return to the aspect and the
richness which it ought to have according to the minerals
treated.
440 ENGLISH COPPER ASSAY.
II. Calcining.
The calcination of the regulus is one of the most im-
portant operations ; it ought to be quite complete.
III. Coarse Copper.
To the calcined regulus is added nitre J ladle, borax
1 ladle, charcoal -J ladle, dry salt 1 ladle (these quantities
remain the same, whatever mineral may be assayed) ; tartar
2 ladles for a regulus of medium richness. Covering of
moist salt, 2 ladles.
The nitre is designed to burn the sulphur which may
have escaped the calcining, and to insure the passage of
the easily oxidisable metals, especially of iron, into the
slag in the state of oxides. It is, besides, in too small pro-
portion to act upon the copper, especially in presence of
reducers whose effect is certainly later than the deflagra-
tion of the nitre.
The borax plays simply the part of a flux.
The dry salt has for its object to give fluidity to the
slag. Unfortunately, if the addition of the salt attains this
object, it also determines from this operation a sensible
loss of copper by carrying it away with the saline vapours.
We shall insist upon this point in describing the washing.
The charcoal and the tartar are especially the important
reagents in the fusion. The tartar, at the same time that
it is one of the most energetic reducers, is also a flux and
a desulphuriser. Its use is, then, perfectly justified here,
only the proportion of tartar added ought to be regulated
according to the quantity of copper, which the weight and
aspect of the regulus permit the experienced assayer to
estimate sufficiently closely ; an excess of tartar would re-
duce the foreign metals, and produce in consequence a
very impure coarse copper.
When the fusion appears complete, we throw in some
white flux,* which gives fluidity to the slag, and deter-
* This white flux is prepared in the laboratory by mixing in a mortar,
tartar 3 volumes, nitre 2 volumes, salt a little, then determining the combus-
tion by the introduction of a red-hot iron rod, which is turned round until the
matter ceases to deflagrate.
ENGLISH COPPER ASSAY. 441
mines by its partial decomposition, from which a disen-
gagement of carbonic oxide results, a stirring up of the
materials. These two effects facilitate the collection of
the metallic button. The potassium carbonate begins also,
without doubt, from this operation to refine the metal a
little by attacking the iron, zinc, and tin already reduced.
M. Berthier (' Essai par la Yoie Seche,' vol. i. p. 393)
points out this reaction : ' A part of the carbonic acid
which it contains being decomposed and changed into car-
bonic oxide, a compound is formed consisting of alkali,
carbonic acid, and metallic oxide, &c.
Lead, copper, and antimony are not attacked.
IV. Washings.
In the operation of washing we put into the crucible,
at the same time as the coarse copper, the following
fluxes : White flux, 1 ladle ; dry salt, 2 ladles.
It is evident that the white flux is here employed as
an oxidiser of the foreign metals, and with a view to the
application of the above-mentioned reaction.
As for the salt, it is both useful and injurious. If it were
only used with the view of augmenting the fluid mass so as
to preserve the metal from contact of air, &c., it would be
advantageously replaced by an excess of white flux ; but it
can form volatile chlorides with the arsenic and antimony
which the copper has retained in the form of arsenide
and antimonide. Common salt is, then, to be regarded as
one of the principal agents of purification put in operation
by the Cornish method. On the other hand, the loss of
copper which arises from the carrying off of this metal by
the vapours of common salt cannot be doubted. M. Ber-
thier has found that by heating equal weights of copper
and salt until the complete volatilisation of the latter, 3
per cent, of the metal is carried off.
In the event of the coarse copper appearing too im-
pure, we take care to add a little nitre. According to the
appearance of the button we recommend the washing or not.
442 ENGLISH COPPER ASSAY.
Y. Testing, Refining.
The button of metal is flattened on an anvil. We
thus recognise tin by the hardness, and antimony by the
brittleness, of the alloy. The button is then put alone in
the crucible. When it presents a proper appearance
that is, when the edges assume a bright colour, the centre,
which the assay er calls the eye, being dark ^we hasten to
put into the crucible the fluxes, which are the same as
for washing, only taken in rather smaller quantity.
In general, when we have operated well, the button
obtained is of a fine colour, and is regarded as pure ; if we
have passed the eye, it is covered with a layer of red
oxide ; if, on the contrary, we have put in the fluxes too
soon, the button is dull.
It is easy to give an account of the reactions which
take place during the refining, and which differ a little
from those of the washing.
In heating the button alone in the air in the crucible,
it is intended to submit it to an oxidation which ought to
act sufficiently on all the foreign metals more oxidisable
than copper without acting too much on the latter. The
proper point is indicated by the appearance of the eye ;
the projection of the fluxes puts an end to the atmos-
pheric oxidation, and causes the scorification of the oxides
which expel part of the carbonic acid of the potassium
carbonate, for which they substitute themselves, and give
rise to triple compounds of metallic oxides, alkali, and
carbonic acid.
The lead, tin, iron, and zinc oxides behave thus.
When we have passed the eye, there has been a con-
siderable formation of oxide, which leaves the button
reddened, as I have indicated. At the same time the slag
is strongly coloured red or green. If, on the contrary,,
the fluxes have been thrown in too soon, the oxidation
has been insufficient, and then the refiner just falls back
upon the preceding operation of washing an operation
less efficacious, and even without result, in the case of lead
and antimony.
ENGLISH COPPER ASSAY. 443
As for the physical phenomenon of the eye, perhaps it
corresponds to the very short instant when the oxides, less
dense than the copper, are concentrated at the top of the
button, .and there make a dark spot before attaining a
temperature sufficiently elevated to acquire the brightness
of the metal itself.
We may add that the minerals of Cornwall, generally
more impure than foreign minerals, require a notably
longer time for the appearance of the eye.
Extra Accidental Washing. More often the refining
gives a definite product, put aside to be weighed with the
prill extracted from the slag ; let the button be clear, burnt,
or dull. Even if the metal appeared too impure we would
not recommence the refining, but would have recourse to
an extra washing by putting at once into the usual crucible,
besides the button and the usual fluxes, the slag from the
refining.
VI. Slags for Prill.
All the slag from the fusion for coarse copper in-
clusively having been preserved, we fuse them altogether
with :
Tartar . . .1 ladle "1 Simple reducing
Charcoal . . . traces / mixture.
We obtain a small globule, variable with the circum-
stances of the different operations which have allowed
more or less copper to pass into the slag. If the prill is
not very small, and its appearance indicates a metal not
sufficiently pure, we submit it to one or two washings as
above.
SECTION II. MANIPULATIONS.
The sample, which has been taken with the utmost
care, arrives at the laboratory rather coarsely powdered,
still wet, and wrapped in strong packing-paper ; the paper
is opened and placed near a furnace on the cast-iron plate
which covers it ; the drying is rapidly done there.
The first requisite is to discover the kind or kinds of
minerals, so as to employ the warm or raw sample.
444 ENGLISH COPPEE ASSAY.
For this purpose we throw one or two large pinches of
the mineral into a flat-bottomed copper dish, and we wash
it very easily by putting in water several times and giving
a rotatory motion to the matters at the same time that
we incline the dish so as to cause the muddy parts to run
from the gangue. The small metallic fragments remain
distinctly visible, and we can often discern by simple in-
spection the presence of foreign metals.
We weigh 400 grs. of the dried mineral a quantity
upon which the assay is made.
The crucibles used in Cornwall are of three sizes :
1. Large.
2. Large second.
3. Small second.
The small seconds have externally the internal dimen-
sions of the large, into which they fit as into a nest ; the
first and third are sold the one in the other, and called
nested. They are the most used.
The large serve for the roasting and the fusion for
regulus: the small seconds for calcining the regulus and
all the fusions which follow.
The large seconds are only employed in place of the
former when we have to treat a very large regulus.
The crucibles are of a kind rather wrinkled, and as if
fused superficially ; they present the appearance of coarse
stoneware pottery. Their form, moderately wide, permits
us to make use of them successively for the roasting and
the fusion for regulus, and gives them sufficiently great
stability in the fire of a wind-furnace. They are, besides,
very resisting. They are made at Truro and Eedruth.
The wind-furnace has for its principal dimensions
Inches
Length from front to flue . . . 10
Breadth
Depth to the bars .
Opening of the flue
8
14
8
2
A sufficiently large space is reserved underneath the
fire, where the ashes accumulate without inconvenience,
but opening only by a contracted framework so as not to
allow too free an access of cold air.
ENGLISH COFFEE ASSAY. 445
The furnace serves either for roas tings or for fusions ;
in the latter case we cover it with two mounted bricks, very
easy to manage, and allowing us to only half-open it when
we wish to inspect the contents of the crucibles. We can
conduct ten roastings at once ; the crucibles are marked by
a brush with colcothar mixed with water. The furnace
having been recharged with coke, we put the crucibles on
the top, and after a few minutes, the substances beginning
to get warm, we stir them by means of iron rods. Each
crucible receives a rod, which we leave standing there
(leaning against the chimney) during the whole period of
the roasting, so as to avoid the loss which would take
place if we withdrew the rod. From time to time we
renew the surfaces by lightly taking hold of the rod with
the left hand by the upper end, whilst the right forefinger
and thumb make it turn at once upon itself and round the
crucible.
The duration of the roasting varies essentially with the
nature and the richness of the mineral ; it is never less
than six or seven minutes, and may reach half an hour.
When from the sandy appearance of the matters we con-
sider the operation finished, we withdraw the crucible,
raise the iron rod with care, and expose the crucible to the
air, allowing its contents to cool slowly. The roasting has
succeeded when the surface has the brown-red colour of
iron oxide and the bottom only is black. In this case we
proceed to the fusion for regulus by simply adding the
three fluxes (borax, fluor-spar, and Jime) ; if the bottom
of the crucible appear too black, we ought to complete
the oxidising action by the addition of a little nitre.
Fusion for Regulus.
The different substances above indicated are taken
from the box with a slightly concave ladle of If diameter,
then mixed in the crucible with a stirring-knife. We
ought to allow the heat of the wind-furnace to fall and
then to recharge, so as to have a gentle fire at the com-
mencement of the fusion for regulus. The crucibles
446 ENGLISH COPPER ASSAY.
placed upon the coke, and supported against the walls of
the furnace, which we then close with the two bricks.
After about a quarter of an hour, we open the front brick,
so as to observe the progress of the operation ; it is at this
stage that we throw the sulphur and tartar into those
crucibles from which a blue flame is disengaged. Some
minutes later that is to say, nearly seventeen minutes
from the commencement we add the salt and the flux
destined to collect the regulus ; then (twenty minutes from
the beginning) we run into a metal mould, not greased.
We make, in general, several fusions at once four,
for example ; we have in consequence two moulds into
which we pour the contents of the crucibles in an adopted
order, so as to avoid all confusion. The matters, very
rapidly solidified, are detached simply by a blow, and fall
in order on a metal plate fixed in front of the laboratory
window. We immediately seize them with the copper
tongs, put them into a basin of the same metal, and im-
merse them for a moment in cold water, where it is
important not to leave them too long. This immersion
allows us then to separate very easily the slag from the
button of regulus, itself very brittle. For this purpose the
mass is put on the metal plate, and by means of a hammer
we strike with care all round the slag, which breaks off
pretty cleanly. We hasten to detach from the surface of
the regulus the slag which may remain adherent, using a
small hand-chisel, without the hammer. The slags are
broken, and if we find any prills of regulus they are added
to the principal button. Sometimes in these breakings,
and especially in those analogous for the last fusions, we
surround the substances by an iron ring, placed on the
metal plate, so as to avoid loss of splinters. In a general
way, the slags of the fusion for regulus are rejected. We
shall see further on how it may become necessary to flux
them again when the mineral contains blende.
The aspect of the regulus is characteristic, and it is
easy to arrive at a pretty close estimation of its richness,
and consequently of the degree of success of the operation,
by simple inspection of the regulus.
ENGLISH COPPER ASSAY.
447
No. 1 . A very poor regulus (coarse) ; that is to say, too
much charged with iron, is bronzed and dull; the opera-
tion following would not be able to carry off the excess
of iron, at least without a corresponding loss of copper. A
like regulus evidently results from an imperfect warming,
or from an excess of sulphur, or from an insufficiency of
nitre, as the case may be.
It contains less than 40 per cent, of copper. There is
nothing for it but to reject it.
No. 2. A regulus of good appearance is in general
bronzed but rather shining ; it appears finer. Its richness
varies from 40 to 60 per cent.
No. 3. From oxides, carbonates, and from some mine-
rals charged with impurities (Sn, Sb) we desire to obtain
a fine bluish button of a greater richness 65 to 75 per
cent. We perceive, indeed, that from oxides and carbon-
ates, to which we have only to add sulphur, and which
also by their nature do not, like pyrites, contain combined
iron, it is easy to obtain a richer regulus without fearing
any loss of copper. As for the stanniferous and antimonial
minerals, we shall return to them further on.
No. 4. In every case a regulus, the richness of which
rises to 80 per cent., and of a very shining grey-blue
appearance, ought to be rejected, its richness indicat-
ing the loss of a certain quantity of copper left in the
slag.
Here is, in the preceding order, the result of the ana-
lyses of four buttons whose description agrees with that
which we have just given, excepting, perhaps, No. 2,
whose fracture is rather reddish :
No.
Copper
Iron
Balance ;
sulphur
and traces
of foreign
metals
1
2
3
4
Coarse, to be rejected .
Good in general (rather too fine) .
Good for a carbonate, &c. .
Too fine, to be rejected
36-00
60-00
65-60
80-16
32-90
14-70
10-50
2-10
31-10
25-30
13-90
17-74
If we compare these products with those obtained in
448
ENGLISH COPPER ASSAY.
tlie metallurgy of copper by the Welsh method, we find
(Le Play, ' Annales des Mines ') :
Matts of the operations
H. V. IV. VIII.
h
Si
a
a
d
2
(-H
Different
metals
1
a
"9
m
co
3
( Coarse matt (fusion of poor mine- j
II. \ rals, raw or calcined)
34-6
34-1
1-5
29-8
31-3
UCu 2 S + Fe 2 S 3 + 4(Fedif.met.)S j
I/ Blue matt (fusion of the cal- \
cined coarse matt with I
minerals of mean richness) [
57-2
18-5
1-0
23-3
24-3
1 0-8Cu + 3Cu 2 S + 2(Fe. d.m.)S J
1 Beddish variety, matte mince }
\ l-3Cu + 3Cu 2 S + 2(Fe. d.m.) S )
61-6
15-8
0-6
22-0
22-6
/White matt (fusion of the calcined
I course matt with rich minerals,
carbonates, and oxides)
IV J Metal very pure type
very blue variety .
77.4
64-8
0-7
9-0
0-9
3-6
21-0
22-6
21-9
26-2
mean .
73-2
6-3
20-5
8Cu 2 S + FeS
(Matt (roasting of extra white]
VIII A matt VII.)
81-1
0-2
18-5
( O2Cu + Cu 2 S J
These numbers show the evident analogy, the identity
almost, of the products of the laboratory and those of the
works ; we may sum up by saying that the regulus ought
to be richer than coarse metal, and in the case of ordinary
minerals to approach if not to attain (as in the case of
sample No. 2) to the composition of blue metal.
For carbonated and oxidised minerals we arrive
directly at the very bluish variety of white metal.
Finally, in no case must we have a button as rich as
regulus matt.
Calcining the Matt.
The matt is pounded fine in a bronze mortar ; we avoid
loss of fragments by means of a perforated cover and a
cloth which surrounds the pestle. To facilitate the pulver-
isation, and avoid the sulphide greasing, we add in the
mortar a small piece of coke. The pounded matt is care-
fully turned upon a sheet of paper, the mortar wiped out
with a hare's foot, and the substance put into a small
second or large-second crucible. The calcining is con-
ENGLISH COPPER ASSAY. 449
ducted as the roasting of a mineral ; it generally lasts
longer, for the expulsion of the sulphur is to be as com-
plete as possible. It is necessary to regulate the fire with
the greatest care, so as to avoid all agglomeration, and to
stir almost continually. When the matter adheres to the
rod, we withdraw the crucible for a moment ; this incon-
venience is chiefly produced if we have not detached
the slags sufficiently from the matte ; the calcining is then
much longer, the flames remain blue a long time, and the
fumes which are disengaged have an odour which is -not
purely that of sulphurous acid. When the fumes and the
odour cease, and the matter has taken a sandy appearance,
we raise the heat ; then withdraw the crucible, and allow
it to cool slowly in the air as when roasting.
The mean duration of calcining is half an hour.
Coarse Copper.
The fluxes above indicated are taken from a box No. 1,
except the dry salt called the refining flux, which forms
part of a second box. The ladle for this box No. 2 is a
little larger than for the first : it has a diameter of 1^
inch. At the beginning of the operation the furnace is
well filled and lighted ; the same fire ought to suffice for
all the following fusions, which it is very important to
conduct with great rapidity. After a moment, and if there
is any frothing, we throw in some dry salt, which calms
the ebullition. At the end of ten minutes, the fusion ap-
pearing complete, we throw in a pinch of white flux. A
little after we withdraw successively each of the crucibles,
pouring them carefully and by a single turn into each of
the principal cavities of the metal mould. These moulds
ought, this time only, to be greased with a cloth impreg-
nated with suet. The crucibles are immediately put back
again into the fire.
We detach the slag as previously, seize' each one
successively with the copper tongs, and plunge it for an
instant into a basin full of water. The rest is effected as
for the regulus, only the slags are preserved on the metal
G G
450 ENGLISH COPPER ASSAY.
plate, and in the order in which we have detached them.
The button of copper obtained appears more or less black ;
T have already indicated the influence of the tartar in
excess.
Washings.
We place the button and the fluxes in a large copper
shovel, lengthened and narrowed at the end, called a scoop,
and we pour them into the crucible, which is already at a
red heat. As the fusion is made in five or six minutes, it
would be inconvenient to prolong it on account of the loss
occasioned by the carrying off of copper with the vapours
of common salt. The pouring is made with care first into
one of the large cavities, then as soon as the metal has
fallen there we finish by casting the slag into one of the
small lateral cavities. This last slag, probably rich in
copper, is less fluid, and would adhere to the button,
which would be difficult to cleanse. The two buttons
being detached from the mould, we immerse the small one
first, then finish as in the preceding operation.
Testing and Refining.
The crucible has again been put back into the furnace,
after the pouring ; the button tried by the hammer is put
into the crucible by means of the tongs. At the end of
about three or four minutes it attains the colour of the
vessel, the eye manifests itself, and we rapidly throw in
the fluxes put into the scoop beforehand.
The pouring is done as for the washing, with the small
button of slag kept apart.
In general we get a button regarded as pure, clean
copper ; if not, as I have said, we proceed to an extra
washing by adding exceptionally in the scoop the last slag
obtained.
Prill.
The crucible this time has been left out of the furnace ;
put into it all the slags, collected for this purpose from
the metal plate into the scoop, and upon which we have
ENGLISH COPPER ASSAY. 451
put reducing reagents. The fusion lasts a quarter of an
hour ; pour all at once into the large cavity : before the
cooling, by means of a transverse blow, get rid of the
upper beds which are still liquid, and composed principally
of common salt. Then operate as above. Collect the
prill, which again undergoes, if necessary, a washing.
SECT. III. SOME MINERALS AND SUBSTANCES OF A SPECIAL
NATURE INFLUENCE OF FOREIGN METALS.
STANNIFEROUS MINERALS. Most often we only perceive
the presence of tin in a copper mineral when testing with
the hammer, which reveals the nature of the bronze ;
when we proceed to the refining of such a stanniferous
button, it is impossible to obtain the characteristic eye ;
that is to say, the surface of the metal becomes quite clear,
and we scarcely open the furnace when it again becomes
obscure. We free it from tin by two or three extra wash-
ings. If we suspect tin from the known produce of the
mineral, or the inspection of the sample in the basin, we
endeavour to obtain a fine regulus, which is accomplished
in the case of a warm sample by prolonging the calcining,
and for the raw sample by putting in more nitre or less
sulphur. It is clear that tin can only enter the regulus by
virtue of the excess of sulphur necessary to the formation
of the coppery matte, and that by restraining this excess of
sulphur we diminish the chance of tin entering the button.
The fine regulus ought to contain 70 to 75 per cent, of
copper, as for the carbonated copper minerals.
ANTIMONIAL MINERALS. Antimony is also detected in the
testing, the metal being rendered very brittle, We then
add one or two grammes (15 to 30 grains) of lead in the
refining operation. There forms an alloy of lead and an-
timony heavier than copper, which is poured into the small
cavity of the mould. When we suspect antimony we act
as for tin that is to say, we produce a fine regulus, a
most careful roasting expelling the antimony ; then we
have to make two washings, and in the second to add the
metallic lead.
G G 2
452 ENGLISH COPPER ASSAY.
We cause, then, three influences to act with a view of
expelling the antimony :
1. Slow oxidation at a low temperature, disengaging
antimony.
2. Eepeated chloridations, whence a formation of vola-
tile chlorides.
3. Affinity of the lead and mechanical separation of
the alloy.
ZINCIFEROUS MINERALS. One of the metals which is
most troublesome is zinc. We recognise it by the appear-
ance of the regulus and by its colour, which is that of
blende. Perhaps once out of ten the regulus collects suffi-
ciently to be able to detach it ; in this case we pound it,
add to it the slags, and borax 1 ladle, nitre ^ ladle. We
fuse anew, and obtain a good regulus, for the nitre has
caused the zinc to pass into the slag in the state of oxide.
Most often the zinciferous regulus does not collect, and
there is nothing for it but to begin anew by making a
very prolonged roasting of at least half an hour.
PLUMBIFEROUS MINERALS. Lead is not injurious, for it
does not alloy with copper. The warming is also pro-
longed. Lead passes into the regulus, which facilitates the
collection of the matter. In the last operation the lead
easily passes into the slag ; it also, in case of need, carries
off antimony. Thus the copper obtained from lead minerals
is most malleable.
Special Cupriferous Products.
EEGULUS OF CHILI. These are treated as those which
we obtain by the fusion for regulus. Their richness,
which rises to nearly 60 per cent., requires us to add
much tartar in the fusion for coarse copper.
SLAGS OF COPPER. To obtain regulus w^e add to the slag
sulphur, tartar, and nitre, this last maintaining metals
other than copper in the state of oxide in the slag.
OLD COPPER. For turnings, waste of workshops, &c.,
yielding 97 to 98 per cent, by the assay, and containing,
in fact, not much foreign matter except a little mixed dust
ENGLISH COPPER ASSAY.
453
or dirt, we take care first to glaze the crucible by fusing
in it a little borax and nitre ; then we treat the matters by
a simple washing, the slags of which we work for prill.
This last is often very considerable.
SECT. IV. SUMMARY CONSIDERATIONS COMPARISON OF THE
EESULTS WITH THE ANALYSIS BY THE WET WAY.
After this detailed account of the numerous operations
which the metal undergoes before attaining the state of
button and prill, it would be superfluous to insist upon the
practical difficulty of the Cornish method.
Nevertheless, in experienced hands, and in the case of
daily practice, it is still a rapid method, allowing us to
treat almost uniformly the different varieties of copper
mineral, and at the least to remedy during the operation
itself the impurities which show themselves.
As to the metallurgic accuracy, below is a small table
showing comparatively the produce by the dry way (es-
timated by a Cornish assayer) and that which we have
obtained by the most precise methods of the wet way. It
comprehends six samples, whose richness varies within
sufficiently great limits.
Nature of the sample
Dry way
Wet way
Difference
and produce
D.
W
W-D
Regulus of Chili .....
564 = 56-250
58-40
2-150
Green copper carbonate of Castile .
9f = 9-750
11-52
1-770
Variegated copper, Huel Damsel .
Pyrites, West Wheal Seton . .
10i = 10-500
8|= 8-375
11-30
8-40
0-800
0-025
United Mines ....
8 = 8-000
10-38
2-380
Devon Great Consols
4|= 4-625
5-60
0-975
S(W-D) .
.
8-100
6
1-350^
By adding the result given by the last five minerals
we find
= 9-44
b
By taking the ratio
454 ENGLISH COPPER ASSAY.
we see that we must add to the richness indicated by the
Cornish assay about f of that result, and by taking the
ratio
that the loss is -J of the copper, if we consider a mineral
of 9 or 10 per cent.
Without wishing to draw a conclusion altogether
general from so small a number of analyses, we neverthe-
less think they suffice to show that the Cornish method
occasions losses always sensible and sometimes considerable.
We think we may assert that upon the whole of the
Cornish minerals whose mean richness varies from 6 to 7
per cent., the loss by the assay is not less than 20 per
cent, of the contained copper, and that for certain pyrites
of 3 to 4 per cent, it attains 30 and 40 per cent, of the
metal.
The principal causes of these losses are (1st) the
quantity more or less great of copper left in the slag of
the regulus ; (2nd) and especially the carrying away of
copper by the vapours of common salt in the fusion for
coarse copper, the washing or washings, the refining and
the treatment of the slag for prill.
In consequence we think they ought to bear princi-
pally on the oxidised minerals for which we make a rich
regulus, and still more on the impure minerals, which
besides a rich regulus have undergone several washings.
Thus the minerals of Algeria, grey copper, assayed some
years ago at the School of Mines, have given a produce
much higher than that indicated by the Cornish assayers.
It may be said, therefore, that the results of the Cornish
assay do not fall short of the truth by a fixed quantity,,
but become more and more inaccurate the poorer the ore.
The smelter gets out of the ore more metal than the assay
indicates.
The most frequent alloys of copper, i.e. brass, German
silver, gun-metal, &c., cannot be assayed in a reliable
manner in the dry way ; German silver because the
ENGLISH COPPER ASSAY. 455
nickel could not be removed at all, or only with great
difficulty, and the rest because zinc and tin give such diffi-
cultly fusible oxides that they could not be properly
removed in the refining.
In the alloys of copper with silver -, gold, and platinum,
the copper may be estimated from, the loss arising from
cupellation with lead.
In all assays of copper in the dry way the silver or
auriferous silver contained in the assay sample cannot be
removed, and it is generally pretty completely collected
in the copper obtained. These copper assays give no-
where any indications whether gold or silver is present
or not ; and the amount of these metals which may be
present must therefore be both sought for and estimated
by a special assay for them. If they are found, and in
sufficiently large quantity, they are deducted from the
weight of the copper.
The dry assay is mostly found in practice in smelting
works, where, even in the hands of less scientifically edu-
cated than skilful assayers, with the character of the assay
substance once known, and suitable practice in following
out the separate manipulations, it gives results which
suffice for the business of working copper in the large
way.
b. For Ores and Products of Class II.
THE LAKE SUPERIOR FIRE ASSAY.
We are indebted to Dr. E. D. Peters's work on
' Modern American Methods of Copper Smelting ' for the
following excellent description of the dry method of assay
in use in America.* The ordinary English fire-assay is
little suited to American conditions. The Lake Superior
fire-assay, on the contrary, is not only quick and inexpen-
sive, but compares favourably in accuracy with the best
* In his description of this mode of assay Dr. Peters acknowledges his
indebtedness to Mr. Maurice B. Patch of Houghton, Michigan, for valuable
assistance in the preparation of this section on the Lake Superior assay. The
position held by Mr. Patch as chemist to the Detroit and Lake Superior
Copper Company is a sufficient guarantee of the accuracy of the following
description.
456 AMERICAN COPPER ASSAY.
wet methods. It is so peculiarly adapted to the conditions
that have given it birth, that no American work on the
metallurgy of copper would be complete without a detailed
account of it, especially as our docimastic literature up to
this time has made little mention of it. In the Swansea
assay, the substance under treatment consists usually of a
mixture of sulphides and gangue rock, which necessitates
a series of calcinations and fusions, culminating in a button
of impure copper, which has still to be refined at a con-
siderable loss. The Lake Superior assay er has the simpler
problem of dealing only with native or oxidised compounds
of copper that can be reduced to the metallic state at so
low a temperature as to preclude the adulteration of the
copper button with any other metallic substances, and
thus obviate the necessity of any refining process. In
spite of the apparent simplicity of this method, it demands
a good deal of skill and experience to obtain correct re-
sults ; but these once acquired, no assay can excel it in
accuracy and celerity.
A glance at the composition of the substances operated
on will render clear the objects to be accomplished. The
material assayed consists of the concentrates from the jigs,
tables, buddies, and other concentrating machines. This
material is technically termed ' mineral,' and varies greatly
in richness, composition, and size of particles, ranging in
copper from 10 to 97 percent., and in some instances con-
taining a gangue of nearly pure ferric oxide, while in
others it is highly silicious. Nearly all grades of mineral
contain a considerable proportion (from 3 to 10 per cent.)
of metallic iron from the stamp heads, while a sample
containing 50 per cent, of titanic iron-sand is no unusual
occurrence. It can readily be seen that no small skill is
required so to flux these various mixtures as to obtain a
clean and fusible slag, and a button of copper free from
iron or other metals that may be reduced with compara-
tive ease, and thus yield a far too high result.
Sampling. The mineral is received from the mines,
packed in strong barrels, weighing in the damp condition
in which it arrives from 500 to 2,000 pounds, its weight
AMERICAN COPPER ASSAY. 457
depending on its degree of concentration, the character of
its accompanying gangue, &c. As this material is to be
refined at once, the barrels are emptied on the iron plates
that form the floor in the neighbourhood of the charging-
door of the refining-furnace. After the contents of each
barrel have been thoroughly and separately mixed, a small
sample is taken from every package and put into a tightly
covered copper can. Only the samples from casks of the
same grade of mineral are placed in any one can, as each
quality is assayed by itself, although six or more different
grades of mineral may go to make up the sixteen barrels
that usually form a furnace charge. If two or more furnaces
are simultaneously in operation, the samples of the same
grade are mixed together, to avoid the unnecessary multi-
plication of assays ; the cans containing a little water in
the bottom, into which the tight copper cans are set, to
prevent any loss of moisture in the sample, which might
occur despite the close cover.
Fluxes. Sodium bicarbonate, borax, potassium bitar-
trate (cream of tartar), ferric oxides, sand, and slag from
the same operation are used to flux the gangue and other
worthless constituents, and effect the proper reduction of
the copper. The chief impurity to be dreaded is sulphur,
for which reason the best quality of sodium bicarbonate
must be purchased, and potassium bitartrate must be used
instead of argols. The borax and soda are prepared by
being melted in iron ladles, to drive off their water of
crystallisation, and then pulverised through a twenty-
mesh screen ; clean, well-fused slag from former operations
is reduced to the same degree of fineness, while. the oxide
of iron flux is prepared by pulverising selected fragments
of specular iron through a fifty-mesh sieve. Any clear
quartz sand answers for the silica needed.
Furnace. A common natural draught melting-furnace
is used, an inside measurement of 9-J by 18 inches being
large enough to accommodate six Hessian crucibles. These
are set in rows of three on two thin fire-bricks, the latter
resting on the longitudinal grate-bars, and serving to raise
the crucibles to the zone of greatest heat. Soft coal, broken
458 AMERICAN COPPER ASSAY.
to egg size, forms the customary fuel, and is carefully filled
in around the charged crucibles, which are not placed in
the furnace until the latter is in full heat. The crucibles
employed are four inches high and three inches in diameter,
and are provided with well-fitting covers made at the works
from a mixture of fireclay and sand ; these are the more
necessary because the assay often fills the crucible to
within half an inch of the top.
The skill of the assayer is nowhere more evident than
in the fluxing of the different grades of mineral, the com-
position of which was briefly noticed in the opening para-
graph of this chapter. It is, of course, familiar to all
chemists that sodium bicarbonate and ferric oxide act as
powerful bases, while the electro-negative elements are
represented by borax and sand : the potassium bitartrate
exercises a strong reducing action, as well as furnishing
an active base. The slag equalises the entire mixture,
being capable of neutralising a considerable amount of
either base or acid, and it covers the molten metal and
protects it from oxidation. It is not to his skill in fluxing
alone that the assayer trusts ; of almost equal importance
are the degree of temperature maintained and the length
of time that the assays are left in the furnace.
Good results can only be obtained by shortening the
period of fusion to the utmost. This demands a very hot
furnace at the outset, good fuel, and a lively draught.
Under these conditions an easily fusible assay will pro-
bably be entirely finished in 20 minutes, while from 25
to 30 minutes are required for different samples. It is
quite safe to assert that if the time necessary for a perfect
fusion is increased to 40 minutes, the resulting button
will contain sufficient impurities reduced from the slag to
give a result from 2-|- to 6 per cent, too high.
This assay is applicable to silicates as well as oxides and
native copper, and the results obtained from the assay of
both refining and blast furnace slags cannot be excelled
in accuracy by any other method.
A table of the different weights of fluxes used in assay-
ing the various grades of mineral from the Peninsular
AMERICAN COFFEE ASSAY.
459
Copper Company's works is annexed, as well as the mix-
ture adopted for reverberatory slags for very silicious
ores.
Minerals
Weight,
grains
Borax,
grains
Soda,
grains
Slag,
grains
Potassium
bitartrate,
grains
Sand,
grains
Iron ore,
grains
Xo.
Per cent.
copper
1
1
92
1,000
60
55 200
300
2
86
1,000
60
60
180
300
3
60
500
100
80
300
4
83
500
150
160
300
150
5
20
500
190
200
300
175
#
35
500
140
140
300
100
t
5 to 20
500
200
200
300
The percentage of slag-forming materials being so
small in Nos. 1 and 2, it requires but a slight amount of
borax and soda to flux them, while an addition of neutral
slag is necessary to protect the molten copper. A smaller
quantity of the ore is weighed out in the succeeding assays,
as they are so poor in copper that a large amount of flux
is required by the great quantity of gangue, so that the
capacity of the ordinary crucibles would be greatly ex-
ceeded if 1,000 grains were used. No. 3 mineral contains
just sufficient ferric oxide to form a good slag with the
mixture given ; while in Nos. 4 and 5 this substance, as
well as metallic iron, increases to such an extent as to
require the addition of a considerable proportion of sand
to flux this base and to prevent the adulteration of the
button with metallic iron.
The sample of Calumet and Hecla tail-house mineral
given is typical of the treatment of very silicious material.
There is nothing remarkable in the considerable proportion
of borax (an acid flux) used with even highly quartzose
ores ; for, in addition to the fluxing powers of the soda that
it contains, a boro-silicate is very much more fusible than a
simple silicate. No peculiarities exist in the execution
of this assay ; the ore and fluxes are thoroughly mixed on
glazed paper and covered with a thin layer of potassium
bitartrate after being poured into the crucible. In the
* Calumet and Hecla tail-house mineral. t Rich slag from refining.
460 AMERICAN COPPER ASSAY.
No. 1 mineral, which is nearly as coarse as split peas,
fragments of iron frequently exist, which come from the
stamp heads, and must be picked out of the sample after
weighing out for assay : not that cast iron will alloy with
copper, but that the fragments will be found imbedded in
the copper button after cooling.
The results obtained by this method are surprisingly
accurate. Duplicate estimations of the lower grade
samples seldom vary more than 0*1 or 0'2. A difference
of 0*4 per cent, is a rare occurrence even in the higher
classes of mineral, where the size of the metallic fragments
renders the sampling, and even the weighing out, of a
correct assay a matter of some uncertainty.
A few results from Mr. Patch's notes will confirm
these statements. An average series of tests on cupola
slags by the colorimetric method for the period of a month,
duplicated by the fire assay, gave a result 0-05 per cent,
lower for the latter test, the slag containing about 0*5 of
one per cent.
As an illustration of the results of this system when
applied to very rich ore, a comparative test was made for
eight days on No. 1 Calumet and Hecla mineral, with the
following results :
Battery assay .... 89-100 per cent.
Fire assay . . . 88-812
A similar test on No. 2 Calumet and Hecla mineral : -
Battery assay .... 77*590 per cent.
Fire assay . 77'657
A similar test with various samples :
No. Battery assay Fir
r89-501
>aa.ftn
mean = 89-544 ^ . >mean = 89*92
oy-/u
L 89-70 I
f 77-40]
> mean - 77-740 J ^'.^ >mean = 77'50
L77-40J
It is a somewhat curious fact that the slight loss of
about 0-25 per cent, of copper which results from the
passage of a minute portion of the metal into the slag
ASSAYS IX THE WET WAY. 461
is just about counterbalanced by the impurities in the
copper button from the reduction of ferric oxide, the
amount of which is indicated by the following analyses
of copper buttons the only weighable impurity being
iron :
Copper, per cent. Copper, per cent. Copper, per cent.
09-83 99-76 99-51
99-84 99-80 99-87
99-52 99-46 99-79
This account of a little-known process will doubtless
remove the impression sometimes held by chemists, that
the Lake Superior copper assay is a clumsy and imperfect
operation, and unworthy any advanced system of metal-
lurgy.
B. ASSAYS ix THE WET WAY.
a. Colorimetric Copper Assays.
These are based upon the fact that ammonia, added
in excess to the solutions of salts of copper, produces a
beautiful azure-blue colour, whose intensity depends upon
the quantity of copper dissolved. By comparing the
shades of blue colour in equally thick layers of the dis-
solved ammoniacal assay substance (assay fluid) with a
normal or standard ammoniacal fluid whose copper
contents are known, the quantity of copper in the former
can be calculated when its volume is measured.
To Heine, the superintendent of the smelting works in
Mansfield, belongs the merit of having first successfully
employed this reaction for the estimation of small per-
centages of copper ; and later it has been also extended
by Jacquelain, Von Hubert, and Muller, to the estimation
of larger quantities of copper.
1. HEIXE'S COLOEIMETRIC METHOD.
For the estimation of the quantity of copper in bodies
poor in this metal, e.g. in slags, lead matte, litharge,
crude lead, and 'other plumbiferous metallurgical pro-
ducts, tin, cupelled silver, &c. in short, in all substances
462 THE ASSAY OF COPPER.
which contain from a trace to about 1 per cent, or a little
more of copper, this method is the most advantageous to
be used.
After the assay sample has been reduced to as fine a
state of mechanical subdivision as possible, which with
slags is best attained by sifting or washing them, one
centner (3-4 grammes) of it is weighed out and dissolved,
or so completely decomposed by a suitable acid that in
the residue, which is to be filtered and well washed, no
more copper remains behind. For this purpose nitric acid
or aqua regia is employed, according to the character and
particular behaviour of the substance, and the nitric acid
is concentrated or somewhat diluted, as may be required.
The solution is either immediately, or after the copper
has been first precipitated by sulphuretted hydrogen gas
and again dissolved, strongly supersaturated with caustic
ammonia, and the precipitate, if any, thereby produced,
digested in caustic ammonia for a considerable time, with
frequent stirring at a very gentle heat (30-40 G.), then
filtered off and thoroughly washed. According to the
quantity of copper present, and according to the degree
of dilution, the solution obtained will appear more or
less strongly coloured blue. The volume of the solution
is measured in graduated vessels, and the intensity of the
colour compared with and estimated from fluids, which
have been previously prepared as standard fluids, and
which for a definite volume contain a definite, accurately
weighed quantity of copper, that has been dissolved in
nitric acid, precipitated by caustic ammonia, and redis-
solved in excess of the same. From the measured volume,
and the intensity found by comparison, the quantity of
copper is then found by calculation.
Heine proposes standard fluids with one, two, three,
and four assay loth of copper in one ounce (two loth,
commercial weight) of the ammoniacal fluid. These four
standard fluids are all-sufficient.
If the French weights and measures are used, standard
fluids are taken with -001, -002, -003, -004 gramme of
copper to every twenty -five cubic centimetres of the fluid.
COLORIMETRIC ASSAYS. 463
The graduated vessels (cylinders) required for the
preparation of the standard fluids, as well as for the
measuring of the assay fluid, can be easily prepared by
the assay er himself. One quarter of an ounce of water is
weighed out a number of times in succession and poured
into the cylinder, and each time the height of the fluid is
marked in a durable manner on the glass with a diamond,
or by etching it with hydrofluoric acid vapour, &c. Also
earthen or porcelain measures, that are prepared and
marked for the volumes that hold one, two, three, four,
&c., ounces of water, may be used.
It is not practicable to replace the volumetric measure-
ment by weighing, for the quality and quantity of those
substances which are soluble in acids and not precipitated
by ammonia, or are again dissolved by it, may vary greatly
in the assay.
In the formation of the normal fluids, two assay pounds
of chemically pure (galvanic) copper are weighed out on a
good balance, dissolved in nitric acid, the solution super-
saturated with caustic ammonia, and placed in a graduated
cylinder, which is divided to whole, half, and quarter
ounce volumes of water, and then water enough is added
to bring the fluid to the sixteen-ounce mark. The fluid
then contains -f-g-=4 loth of copper per ounce. Six ounces
of this four-loth solution are then taken, two ounces of
water added to it, and eight ounces of fluid obtained, with
^ = 3 loth of copper to one ounce of water. The two-loth
solution is formed in a similar way by diluting four ounces
of the four-loth solution to eight ounces ; the one-loth, by
diluting four ounces of the four-loth normal fluid to six-
teen ounces. In the measuring of the assay fluid it is
estimated within one-eighth of an ounce, which is suffi-
ciently close. If in the dilutions a mistake is actually
made of one-sixteenth of an ounce, the maximum of possi-
bility, the error amounts to about two cubic centimetres,
which in a whole mass of fluid of 200-500 cubic centi-
metres has no influence upon the solution that can be
detected with the eye.
The preservation of the standard fluids, as well as the
464 THE ASSAY OF COPPER.
comparison of the blue assay fluids with them, must take
place in glass vessels closed with ground-glass stoppers.
These vessels must have the same form and size, consist of
the same colourless glass, and have an equal thickness of
glass in the smooth side walls. The last condition is
obtained the surest by grinding. This grinding, however,
which notably increases the cost of the glasses, is not in-
dispensably necessary if the vessels are carefully formed
and blown in a good glass-house. An oblong form is most
advantageous for the vessels. They hold about an ounce
and a half of fluid, and are about two inches long, two and
a half inches high, and one inch wide, with walls about
one-eighth of an inch thick.
The glasses are very advantageously formed from
an unblemished sheet of plate glass of equal thickness
throughout, by cementing the sides together and the
insertion of a glass neck. The assayer has in the form of
vessel indicated a triple control in the comparison of the
assay fluid with the normal solution according as he looks
through the fluid in three different directions.
The digestion of the assay sample with acid may take
place in any suitable vessel whatever a glass flask, a
beaker covered with a watch-glass, &c., only no thumping
and spirting of the fluid should be possible in the process.
The nitric acid, &c., must be added little by little. The
time required for this may vary greatly. The solution of
cupelled silver* skimmings , &c., with nitric acid is finished
in a short time ; on the other hand, in the examination
of difficultly decomposable slags, with which concentrated
nitric acid or aqua regia will always be used, the digestion
often requires to be continued at a warm temperature for
two to three times in twenty-four hours. The mass must
be frequently stirred with a glass rod, because many slags
decompose rapidly with evolution of heat, form a thick
jelly, and deposit a crust on the bottom of the glass sub-,
mono-, and bi-silicate slags mostly decompose readily,
* With cupelled silver, after dissolving in nitric acid, the silver may be
precipitated with sodium chloride, the silver chloride filtered, washed, and the
solution then mixed with caustic ammonia.
COLORIMETRIC ASSAYS. 465
higher silicates resist complete decomposition by aqua
regia and then a preliminary solvent ignition or fusion
with potassium carbonate or calcined sodium carbonate,
or, better, a mixture of both, is, necessary, precisely in the
manner given in the wet assay of copper. Here also it
does no harm if some of the substance of the crucible
remains adhering to it.
The decomposition of the slags by acid is complete
when in the stirring with a glass rod no more grating can
be perceived.
After hot water has been added to the decomposed
assay, the residue is collected on a filter, well washed out,
without diluting the filtrate too largely, and the copper
precipitated from the solution, if necessary, with sulphu-
retted hydrogen gas, especially when a notable quantity of
alumina and iron is present, whose slimy precipitates from
the immediate precipitation with ammonia always retain
copper. This precipitation of the copper has also the
advantage that, as cobalt and nickel do not precipitate
with it, the colouring effects which they would produce,
if present, are removed. Since the copper sulphide re-
quires for its solution but a few drops of nitric acid, in
the succeeding treatment of the solution with ammonia,
but a small quantity of ammoniacal salt is formed, and the
specific gravity of the coloured fluid varies but very little
from that of water and the normal solution. With the
increase of the specific gravity of the assay solution, its
volume is considerably increased, and therefore it gives
too large a measure in the direct precipitation with am-
monia. If the precipitation with sulphuretted hydrogen
gas is completed in four to six hours, the copper sulphide
is filtered out, thoroughly washed with cold water contain-
ing sulphuretted hydrogen, the filter dried, ignited in a
porce 7 Crucible, the copper oxide formed warmed with
a few dro of nitric acid or aqua regia, supersaturated
with ainm< .a, filtered, and well washed, till the washings
are no longer tinged bluish.
A precipitation of the copper with iron wire, from a
solution evaporated with sulphuric acid, and a re-solution
H H
466 THE ASSAY OF COPPER.
of the copper in nitric acid, consumes less time. If the
copper is not precisely precipitated, errors of some 30
per cent, and more of the whole amount of copper may
occur. By repeated solution of the iron precipitate and
precipitation with ammonia, all the copper cannot, how-
ever, be extracted.
In the examination of litharge, the solution in nitric
acid may be dispensed with. The copper oxide can be at
once extracted from it with caustic ammonia ; however,
the litharge and ammonia must then be allowed to work
at least twenty-four hours on each other, with very diligent
stirring, and, moreover, the litharge must be ground very
fine.
The ammoniacal solution obtained from the assay is
now well stirred, so that it may mix with perfect unifor-
mity with the last washings ; then, either the whole, or a
part of it, is placed in a clean assay glass, and compared
with the standard fluids in similarly formed glass vessels
standing on a sheet of white paper. Should it correspond
with none of them in the intensity of its colour, the whole
of the fluid is diluted somewhat with water, until this is
the case. Its volume is thereupon measured in the glass
vessel graduated in ounces, &c., and noted. For a check,
the dilution may be carried still farther till the colour of
the assay corresponds to the next more faintly coloured
standard fluid, and then the increased volume be measured
anew. This might perhaps be still again repeated, but it
becomes more and more uncertain. The calculation of
the percentage of copper, from the intensity and the volume
found, then presents no further difficulty.
Suppose that the assay fluid agrees with the normal so-
lution of four loth of copper to the ounce of water, and
its quantity amounts to five ounces, then the quantity of
copper in the centner of the assay substance is 5 x 4=20
loth. This fluid further diluted till it equals the normal
solution with three loth of copper, must measure six and
two- third ounces if the obtained value of twenty loth is to
be confirmed.
According to Heine's experiments, the possible error
COLOEIMETEIC ASSAYS. 467
of observations in the comparisons and measurement de-
scribed amounts as a maximum with the stronger normal
solutions (with sixteen loth and over) to from three-quarters
to one loth, with the weaker ones to scarcely half a loth
of copper. In a centner of the assay substance, one loth
of copper '03 per cent, can still be estimated with cer-
tainty.
Le Play's Method. Le Play estimated in finely pul-
verised and carefully washed copper slags, the copper in
one gramme of the poorest slags to within half a milli-
gramme, and of the richest slags to within one milli-
gramme, by using twenty-six standard fluids with various
percentages of copper in cylindrical vessels. The com-
parison of colours in round vessels is more uncertain than
in oblong ones, since in the former the light is dissipated
and shadows are produced.
T. 0. Cloud's Method. According to Mr. T, 0. Cloud
the copper cannot be completely extracted from the finest
ground slags, even after three days' digestion with aqua
regia, and subsequent treatment with sulphuric acid. He
recommends fusing the slag with four parts of mixed
potassium and sodium carbonates and J part potassium
nitrate. The fused mass is extracted with dilute sul-
phuric acid, the liquid evaporated down, and the copper
estimated galvanically or colorimetrically.
Endemanris Method. As a delicate test for small quan-
tities of copper, Dr. Endemann (' Annalen der Chernie ')
adds to the dilute solution concentrated hydrobromic
acid, when a dark brownish-red or violet colour is at once
produced.
This reaction is so delicate that T -^Q- milligramme of
copper can be detected with certainty. One drop of a
solution containing this small quantity of copper is brought
on a watch-glass, then one drop of hydrobromic acid is
added, and the solution is then allowed to evaporate slowly
by standing the glass in a warm place. When the whole
has been concentrated to about one drop, this will distinctly
show a rose-red colour. The colour thus produced is
about three or four times as distinct as the one which is
H TT 2
408 THE ASSAY OF COPPEE.
obtained by the addition of potassium ferrocyanide. Of
other metals which are examined in this direction, we find
only iron to be apt to interfere with this reaction, and then
only when it is present in considerable quantity.
This reaction may also be utilised as a colorimetric
test for the quantitative estimation of small quantities of
copper.
If a substance contains so little copper that the fluid
does not equal the most faintly coloured standard fluid in
intensity of colour, the assayer must endeavour to remedy
the matter by evaporating till this is the case. An evapora-
tion is, however, to be avoided if possible, first because of
the loss of time, and also because other precipitates, carbo-
nate of lime, &c., are apt to be caused by it, and because,
when it has to be continued too long, so much ammonia is
apt to be volatilised, that a new addition of it becomes
necessary.
This method of assaying soon finds the limits of its
accuracy in an increasing percentage of copper in the
assay sample, since with fluids rich in copper, and there-
fore strongly coloured blue, the errors of observation soon
amount to several loth. And to seek then to better one-
self by diluting largely, yields no more accurate results^
since a small error of observation in estimating the
intensity of the colour, is so much the more multiplied in
the calculation of the value by the greater number of the
ounces.
If nickel is contained in the assay substance, the assay
cannot be conducted in the way prescribed, since the
nickel is extracted by the acids, and dissolves also in
caustic ammonia with a blue colour. The assay may also
become uncertain from the presence of much m,anganese,
cobalt, or chromium, since they render the hue of the blue
colour dingy. Chromium may be completely removed by
a slight boiling of the ammoniacal fluid ; not so cobalt.
The presence of vanadium or molybdenum does no harm.
If nickel, or much cobalt and manganese, is contained
in the assay substance, the solution obtained by acids and
filtered, though not further diluted, must first be decom-
COLOKIMETE1C ASSAYS.
posed by metallic iron. What is thrown down by the iron
is collected on a small filter, washed thoroughly, and then,
together with the filter, treated with dilute nitric acid.
When the copper is all dissolved, this solution is supersatu-
rated with caustic ammonia and then managed as above.
With higher percentages of copper the process of the
Swedish copper assay is used for estimating the value.
The precipitation of the copper may also be performed
with sulphuretted hydrogen gas.
Le Play removes the injurious influence of manganese,
nickel, and cobalt, by allowing the green or violet-coloured
ammoniacal solution to stand open to the air for some
time in a moderately warmed drying furnace, whereby a
few variously coloured gelatinous flocks are gradually
deposited, and the fluid, after the addition of a few drops
of ammonia, then becomes pure blue.
According to Jacquelain and Von Hubert, nickel and
cobalt are in a simple way rendered perfectly harmless by
gradually adding white pulverised marble to the solution
of the assay substance, until the effervescence ceases, and
then warming the whole on the sand-bath, whereby all the
copper is perfectly precipitated as carbonate, while nickel
and cobalt remain dissolved. It is now filtered, washed,
the residue dissolved in nitric acid, and the solution treated
as already explained, with ammonia. By the addition of
potassium carbonate to the ammoniacal fluid, and heating,
all the manganese precipitates, while the copper remains
dissolved in the excess of ammonia, and can be separated
from the manganese precipitate by filtration. The manga-
nese must have been present as oxide in the original solu-
tion in order that the precipitation by potassium carbonate
may be perfect.
The assayer may convince himself whether nickel
or cobalt is present, by slightly supersaturating a blue
ammoniacal solution^ obtained by the ordinary process of
assaying, with hydrochloric or sulphuric acid, then pre-
cipitating the copper completely with iron, filtering the
residual solution, concentrating somewhat, if necessary,
and then supersaturating with ammonia. If the fluid
470 THE ASSAY OF COPPEK.
remains colourless, neither of the two metals is present : a
blue colour indicates nickel, a red one cobalt.
Sometimes the normal solutions, which when freshly
prepared appear azure blue, assume a greenish hue, which
renders the comparison difficult, if not impossible. Copper
nitrate produces with ammonia a pure azure blue, copper
sulphate a lilac colour, and copper chloride greenish hues.
Sulphuric and hydrochloric acid are therefore to be avoided
as much as possible in the solution. But, nevertheless, an
assay fluid may sometimes e.g. by standing some time in
the air, or by slow filtration become green, in which case
the colour is destroyed by a few drops of nitric acid, and
ammonia added anew. But sometimes also the greenish
colour disappears, if the solution stands in a covered vessel
in the air, or by the addition of a few drops of red cobalt
ammonio-oxide.
According to Mliller, also, the colour stands in the
closest connection with the quantity of ammonia employed,
and it therefore leads to greater accuracy in the assay if- a
titrated solution of ammonia is used, and the volume of
the ammoniacal fluid noted, which, after neutralisation of
the residual free acid, is used for the solution of copper.
The solution appears more intense when viewed with a
grey background than with a white one. A greenish blue
colouring becomes the more noticeable the greater is the
excess of ammonia, or the more ammoniacal salts are in
the solution.
Camelly's Method. Dr. T. Camelry (Manchester Philo-
sophical Society) gives the following colorimetric method
for estimating small quantities of copper :
The method of analysis consists in the comparison of
the purple-brown colours produced by adding to a solu-
tion of potassium ferrocyanide first, a solution of copper
of known strength, and secondly, the solution in which
the copper is to be estimated.
The solution arid materials required are as follows :
(1) Standard Copper Solution. Prepared by dissolv-
ing 0-393 grammes of pure CuS0 4 .5H 2 in one litre of
water, 1 c.c. is then equivalent to 0-1 mgrm. Cu.
COLORIMETRIC ASSAYS. 471
(2) Solution of Ammonium Nitrate. Made by dissolv-
ing 100 grms. of the salt in one litre of water.
(3) Potassium Ferrocyanide Solution. Containing one
part of the salt in 25 parts of water.
(4) Two glass cylinders, holding rather more than 150
c.c. each, the point equivalent to that volume being marked
on the glass. They must, of course, both be of the same
tint and as nearly colourless as possible.
(5) A burette, marked T L c.c., for the copper solution,
a 5 c.c. pipette for the ammonium nitrate, and a small
tube to deliver the potassium ferrocyanide in drops.
The following is the method of analysis : Five drops
of the potassium ferrocyanide are placed in each cylinder,
and then a measured quantity of the neutral solution in
which the copper is to be estimated into one of them
(A), and both filled up to the mark with distilled water, 5
c.c. of the ammonium nitrate solution added to each, and
then the standard copper solution runs gradually into (B),
till the colours in both cylinders are of the same depth, the
liquid being well stirred after each addition. The number
of cubic centimetres used is then read off. Each cubic
centimetre corresponds to 0*1 mgrm. of copper, from
which the amount of copper in the solution in question
can be calculated.
The solution in which the copper is to be estimated
must be neutral, for if it contains free acid the latter lessens
the depth of colour, and changes it from a purple-brown
to an earthy brown. If it should be acid it is rendered
slightly alkaline with ammonia, and the excess got rid
of by boiling. The solution must not be alkaline, as the
brown coloration is soluble in ammonia and decomposed
by potash ; if it is alkaline from ammonia this is remedied
as before by boiling it off, while free potash, should it be
present, is neutralised by an acid and the latter by am-
monia.
Lead when present in not too large quantity has little
or no effect on the accuracy of the method. The precipi-
tate obtained on adding potassium ferrocyanide to a lead
salt is white, and this, except when present in compara-
472 THE ASSAY OF COPPER.
tively large quantity with respect to the copper, does riot
interfere with the comparison of the colours.
When copper is to be estimated in a solution contain-
ing iron the following is the method of procedure to be
adopted. To the solution a few drops of nitric acid are
added in order to oxidise the iron, the liquid evaporated to
a small bulk, and the iron precipitated by ammonia. Even
when very small quantities of iron are present this can be
done easily and completely if there is only a very small
quantity of fluid. The precipitate of ferric oxide is then
filtered off, washed once, dissolved in nitric acid, and repre-
cipitated by ammonia, filtered, and washed. The iron pre-
cipitate is now free from copper, and in it the iron can be
estimated by dissolving in nitric acid, making the solution
nearly neutral with ammonia, and estimating the iron by
the method given in the paper before referred to. The
filtrate from the iron precipitate is boiled till all the am-
monia is completely driven off, and the copper estimated
in the solution so obtained as already described.
2. JACQUELAIN'S AND VON HUBERT'S COLORIMETRIC ASSAYS.
Heine's method, for the reasons stated, is suitable only
for the estimation of small quantities of copper. Jacque-
lain has extended it to the examination of all cupriferous
substances, and this process has been further perfected
by Yon Hubert. According to the latter, a solution of any
cupriferous accurately weighed substance is prepared,
mixed with ammonia in excess, the ammoniacal solution
(assay solution) measured at a definite volume, and a small,
likewise measured portion of the measured solution diluted
with water, until its blue colour shows an equal intensity
with the blue colour of another solution (normal solution),
also cupriferous and ammoniacal, whose copper contents
are known once for all. Then, from the quantity of water
added, in order to make the two fluids equal to each other
in the intensity of their blue colours, the amount of copper
in the substance under examination can be estimated by
calculation.
The normal solution is prepared by dissolving -5 of a
COLORIMETRIC ASSAYS. 473
gramme of chemically pure copper in dilute nitric acid,
adding ammonia in excess, and diluting with distilled
water until the whole at 12 C. amounts to one litre = 1000
cubic centimetres. The solution is filtered, and pre-
served in a flask provided with a glass stopper ground in
to fit it.
For the preparation of the assay fluid, with substances
whose percentage of copper ranges from 1-5 to the highest
per cent., two grammes, and with the poorer substances
five grammes, are brought into ammoniacal solution with
the precautions specified in Heine's assay. This solution,
with over 5 per cent, of copper, is measured at two hun-
dred cubic centimetres, with 2 to 5 per cent, of copper at
one hundred and fifty cubic centimetres, and with 2 per
cent, and under, at one hundred ; and also, as may be
required, at 90, 80, 60, 50 c.c., according to the intensity
of the fluid. Only with an extremely small quantity of
copper is the assay fluid evaporated to a smaller volume,
.in order to be able to conduct the colorimetric test with
accuracy.
The comparison of the intensity of colour of the assay
fluid with the normal fluid is accomplished in two different
ways, according as the former, when measured at a defi-
nite volume, is darker or lighter than the latter. This
can be seen if a small arbitrary portion of each is poured
into a glass tube of nine millimetres interior diameter,
twelve centimetres in length, and uniform thickness,
and the two tubes are held in parallel positions over a
piece of white paper so that they rest firmly on it, and
are inclined to it at an angle of about 45, and direct
light falls upon them. Shadow should not fall upon the
tubes.
(a) The Assay Fluid is Darker than the Normal
Solution. By means of a pipette, five cubic centimetres of
the normal solution are placed in a glass tube closed at the
bottom and not graduated, and seven millimetres in interior
diameter and twelve centimetres long. Since 1000 c.c. of
the normal solution contain -5 of a gramme of copper, five
cubic centimetres contain exactly '0025, and the ratio
474 THE ASSAY OF COPPER.
5 : -0025 expresses once for all the known proportion of
copper in the normal solution.
Five cubic centimetres of the definitely measured assay
fluid are now also placed in a beaker and gradually
diluted with water till they show the same intensity of
colour as the normal solution. In the comparison the
assay fluid must be in a similar tube to that containing
the normal solution. With richer proportions a greater
accuracy is attained in this comparison if the assay fluid
is so far diluted that its intensity still appears as little as
possible darker than that of the normal solution, and then
water added carefully, drop by drop, till its intensity
is judged as little as possible lighter than that of the nor-
mal solution, whereupon the mean of the two volumes
noted is taken as the correct value. The measuring of the
diluted assay fluid is performed in glass tubes of nine
millimetres interior diameter and fifty centimetres in
length, which from their lower closed end to the circular
mark designated by 0, hold exactly five cubic centimetres,
and from upwards are divided into cubic centimetres
and tenths.
If, for example, two grammes of the assay substance
have been weighed out, the assay fluid measured at 200
cubic centimetres, and five cubic centimetres of it diluted
to 8-2 cubic centimetres, in order to obtain an equal inten-
sity of colour between the normal and assay fluid, then the
percentage of copper, #, follows from this according to the
following chain of ratios :
100 per cent.
200 c.c. assay fluid.
8-2 c.c. diluted assay = normal solution.
0025 gramme of copper in normal solution.
x = 8'2 per cent, of copper.
(b) The Assay Fluid is Lighter than the Normal Solution.
In this case five cubic centimetres of the normal solution
are diluted till their intensity is equal to that of the assay
solution that has been measured at a definite volume, and
for the comparison larger tubes of nine millimetres' interior
diameter are used.
COLORIMETRIC ASSAYS. 475-
If, for example, two grammes of the assay substance
have been weighed out, 150 cubic centimetres of assay
fluid obtained from it, and to get the same intensity of
colour, five cubic centimetres, of the normal solution
diluted to 8 -4 cubic centimetres, the quantity of copper x
amounts, according to the following chain of ratios, to
2-205 per cent. :
100 per cent.
x
2
8-4
5
150 c.c. assay solution.
5 c.c. normal solution.
0025 gramme of copper.
x = 16-8 | 37-5 = 2-205 per cent.
This assay is adapted for all cupriferous substances,,
since nickel, cobalt, and manganese, which would influence
the result unfavourably, can be removed without particu-
lar difficulty. It is also easy to be learned by those less
practised in analytical operations, can be completed in a
few hours, and is far less expensive than the dry assay.
From two to one-tenth per cent, of copper can also be
determined by it with accuracy.
Heine, however, prefers his method when a small per-
centage of copper is to be estimated, since by it even one
loth of copper in the centner = -03 per cent., can be esti-
mated, and there is less liability to error. While in slag
assays with nine to eighteen loth of copper in the centner,,
by Heine's method, errors of half a loth are not to be
avoided, variations of more than one loth occur by Von
Hubert's process. The latter works with a too deeply
coloured normal fluid, corresponding to a solution of over
fourteen loth of copper to one ounce of water, while Heine
does not exceed four loth. The process is surer if the fluids
are diluted and thicker layers of them compared, and the
hue thus made artificially deeper, than if small quantities of
stronger fluids are compared and the hue made artificially
lighter by comparing them in thinner layers, or especially
in tubes, where the light is dispersed and shade produced.
The comparison in oblong glasses is therefore to be pre-
ferred to that in tubes.
476 THE ASSAY OF COPPER.
By a comparison of Yon Hubert's assay with tnat of
the Oberhartz, it appears that, as Yon Hubert's experi-
ments themselves have shown, both give equally accurate
results for substances not too poor in copper (i.e. contain-
ing not less than 5 per cent.) The Oberhartz assay
allows a direct estimation of the copper, requires less ap-
paratus, is also very simple, and can be completed in a
shorter time. Since different individuals are differently
susceptible to colours, and the blue colour of the copper
ammoriio-oxide, in consequence of causes yet unknown,
sometimes inclines more or less to greenish, and thereby
renders observation difficult, therefore, for the sake of
greater certainty, though not of greater accuracy, those
assays by which an estimation of the copper is possible
by weight should in general be preferred to the colori-
metric methods, and this is the case with the Oberhartz
assay down to two per cent. With smaller percentages the
colorimetric assay must be called to our aid. It is not
yet settled that with higher percentages of copper the
principle of colorimetry is a correct one ; that is, that the
intensity of the colour is directly proportional to the quan-
tity of the colouring agent.
Since ammoniacal solutions poor in copper often show
a dash of green colour, Yon Hubert prepares a normal
solution for such by dissolving one decigramme of copper
and diluting to one litre of fluid.
b. Volumetric Copper Assays.
FLECK'S MODIFICATION OF MOHR'S METHOD.*
The proposal to take the action of solution of potas-
.sium cyanide on ammoniacal solution of copper, as the
foundation of a method for estimating copper, is due to
Carl Mohr.f
* This process is given by Fresenius, condensed from the ' Polytechn.
Oentralbl.' 1856, 1313.
t Annal. d. Chem. u. Pharm. 94, 198 ; Fr. Mohr's Lehrbuch der Titrier-
rnethode, 2, 91.
VOLUMETRIC ASSAYS. 477
In carrying out this estimation according to the direc-
tions of Mr. Parkes, a solution of potassium cyanide is slowly
added to a blue ammoniacal solution of copper, when the
latter gradually loses its colour, and finally becomes quite
colourless ; upon this chemical reaction the estimation of
copper by cyanide of potassium depends. By ascertaining
by direct experiment the amount of potassium cyanide
solution required to discharge the colour in an ammoniacal
solution containing a given weight of copper, it is easy
by a comparative experiment to estimate the amount of
copper in a given weight of ore.
For the preparation of the standard solution 2,000
grains of fused potassium cyanide are to be dissolved in
two quarts of water, to produce a solution of which 1,000"
grains measure will be equal to about ten grains of metallic
copper. The solution should be preserved in green glass
stoppered bottles, and kept as much as possible away from
the light : it is liable to a slow decomposition, which will
necessitate the standard being checked at intervals of one
or two weeks. In order to standardise the solution, a
burette, holding 1,000 grains measure, is filled to the zero
mark, and a piece of pure electrotype copper, previously
cleaned by means of dilute nitric acid, washed and dried,
is accurately weighed. About eight grains may be con-
veniently taken ; this is dissolved in a pint flask by dilute
nitric acid, and, after the energy of the first action has sub-
sided, the solution is warmed and ultimately boiled to expel
all the nitrous acid fumes. It is diluted with cold water
to the bulk of nearly half a pint, treated with ammonia in
excess, and to the deep blue solution the cyanide is added
from the burette until the colour is so nearly discharged
that a faint lilac tint only remains. This will generally
become quite bleached on standing at rest for a short time,
so that the cyanide must not be added too hastily towards
the end of the operation. It will be advisable to control
the standard by a second experiment upon another weighed
portion of copper, and to stop short of bleaching entirely
the faint lilac tint of the solution. A piece of white paper
folded and placed under and behind the flask during the
478 THE ASSAY OF COPPER.
decolorisation, will aid in recognising the proper tint of
the solution.
In applying this process to the examination of copper
ores, a known weight of the finely powdered sample is
introduced into a beaker provided with a glass cover, and
moistened with strong sulphuric acid ; strong nitric acid is
then added, and the whole digested on a sand-bath until
nitrous fumes are no longer given off. Should a small
quantity of sulphur be separated in the treatment of py-
ritic ores, the small globules may be taken out, burnt, and
the residual copper dissolved in a few drops of nitric acid
and mixed with the remainder. Water is now to be
added and left in contact for a short time to extract all
the metallic salt from the insoluble residue, which need not
be filtered off ; and so, likewise, when ammonia is added
in the next place, any ferric oxide which may thus be
precipitated is left in the solution, for it is apt to contain a
small proportion of copper when first thrown down ; but
this is entirely removed by the potassium cyanide later in
the experiment.
When the ore contains much iron it is considered
desirable to remove the hydrated peroxide by filtration, in
order to be enabled to observe with greater precision the
last effects of the potassium cyanide ; and in the event of
requiring to know the amount of iron present in the ore,
the precipitated ferric oxide on the filter is redissolved in
dilute sulphuric acid, reduced to the state of ferrous oxide
by metallic zinc, and then tested in the usual way with a
standard solution of potassium bichromate.
The metals which interfere with this mode of valuing
copper ores, are silver, nickel, cobalt, and zinc. The first
may readily be separated by adding a few drops of hydro-
chloric acid to the original solution : the other metals may
be excluded by following one of the methods pointed out
by the author for that purpose.
Fleck proposes the following modification in this pro-
cess :
Instead of caustic ammonia, use a solution of ammo-
nium sesquicarbonate (1 in 10), warm the mixture to
VOLUMETEIC ASSAYS. 479
about 60, and, in order to render the end reaction plainer,
add 2 drops of solution of potassium ferrocyanide (1 in
20) : the blue colour of the solution is not altered by this
addition, nor is its clearness affected. The value of the
potassium cyanide solution is first estimated by means of
copper solution of known strength, and it is then employed
on the copper solution to be examined. On dropping the
potassium cyanide into the blue solution warmed to 60,
the odour of cyanogen is plainly perceptible, and the
colour gradually disappears. As soon as the ammoniacal
double salt of copper is destroyed, the solution becomes
red from the formation of copper ferrocyanide, without
any precipitate appearing, and with the addition of a
final drop of potassium cyanide this red colour in its turn
vanishes, so that the fluid now appears quite colourless.
The method thus modified yields, it is true, better, but
still only approximate, results.* Where such are good
enough, the method is certainly convenient.
E. 0. BROWN'S METHOD BY SODIUM HYPOSULPHITE.
The process described by Mr. E. 0. Brown is particu-
larly applicable to the estimation of copper in gun-
metal, brass, and other alloys which contain no large
amounts of iron and lead. It is founded on the reactions
between salts of copper and the neutral iodides, and on
the conversion of the liberated iodine into hydriodic and
tetrathionic acids by a standard solution of sodium hypo-
sulphite.
The completion of the second reaction is manifested
by the bleaching effect produced upon the blue iodide of
starch by the addition of the hyposulphite. A convenient
strength of solution for this purpose may be made by dis-
solving 1,300 grains of the crystallised salt in two quarts
of water. The potassium iodide must be free from iodate ;
and a clear solution of starch employed.
* In six experiments, in which he had purposely added different quantities
of carbonate of ammonia, Fleck used for 100 c.c. copper solution, in the
minimum 15-2, in the maximum 15-75, in the mean 15'46 c.c potassium
cyanide solution.
480 THE ASSAY OF COPPER.
From eight to ten grains of the copper or alloy are
dissolved in dilute nitric acid, and the red nitrous fumes
expelled by boiling. The copper nitrate is converted into
acetate by adding sodium carbonate until a portion of
copper remains precipitated, and then re-dissolving in
acetic acid. The solution is diluted with water, and about
60 grains of potassium iodide in the form of crystals
dropped into the flask, and allowed to dissolve. The
standard solution of sodium hyposulphite is now poured
in from a burette, until the greater part of the dark-
coloured free iodine disappears. A little of the starch
solution is now added to make its presence more apparent,
and the addition of the hyposulphite continued until the
bleaching is completed, when the pale yellow colour of the
copper subiodide will alone be visible. The amount of
copper in the ore or alloy is calculated from the number
of divisions indicated upon the burette.
Copper ores containing much iron (which interferes by
reason of the dark red colour of the acetate) may be dis-
solved in nitric acid, and treated with sulphuretted hydro-
gen to precipitate the copper, the sulphide being collected
on a filter, washed, and re-dissolved in nitric acid to pro-
duce a solution suitable for testing by this process. Or
the hyposulphite may itself be employed to furnish a pre-
cipitate of copper disulphide.
c. Electrolytic Copper Assay.
ESTIMATION OF COPPER IN THE MANSFELD ORES BY
DR. STEINBECK'S PROCESS.
This method embraces three distinct operations, viz. :
1. The extraction of the copper from the ore; 2. The
separation ; 3. The quantitative estimation of that metal.
1. The Extraction of the Copper from the Ore. A proof
centner, equal to 5 grammes of pulverised ore, is put into
a flask, and there is poured over it a quantity of from 40
to 50 c.c. of crude hydrochloric acid, of a specific gravity
of 1-16, whereby all carbonates are converted into chlorides
STEINBECK S PROCESS. 481
while carbonic acid is expelled. After a while there is added
to the fluid in the flask 6 c.c. of a normal nitric acid, pre-
pared by mixing equal bulks of water and pure nitric acid
of 1*2 sp. gr. As regards certain ores, however, specially
met with in the district of Mansfeld, some, having a very
high percentage of sulphur and bitumen, have to be roasted
previously to being subjected to this process ; and others,
again, require only 1 c.c. of nitric acid instead of 6. The
flask containing the assay is digested on a well-arranged
sand-bath for half an hour, and the contents only boiled
for about fifteen minutes, after which the whole of the
copper and other metals occurring in the ore are in
solution as chlorides. The blackish residue, consisting of
sand and schist, has been proved by numerous experi-
ments to be either entirely free from copper, or at the
most only 0-01 to O03 per cent, has been left undissolved.
The extraction of the copper from the ore, according
to this method, is complete even in the case of the best
quality of ore, which contains about 14 per cent, of metal ;
while, at the same time, the very essential condition for
the proper and complete separation of the metal, viz. the
entire absence of nitric acid, or any of the lower oxides
of nitrogen, is fully complied with.
2. Separation of the Copper. The solution of metallic
and earthy chlorides, and some free hydrochloric acid,
obtained as just described, is separated by filtration from
the insoluble residue, and the fluid run into a covered
beaker of about 400 c.c. capacity ; in this beaker has
been previously placed a rod of metallic zinc, weighing
about 50 grammes, and fastened to a piece of stout pla-
tinum foil. The zinc to be used for this purpose should
be as much as possible free from lead, and at any rate
not contain more than from 0-1 to 0*3 per cent, of the
latter metal. The precipitation of the copper in the me-
tallic state sets in already during the filtration of the warm
and concentrated fluid, and is owing chiefly to the com-
plete absence of nitric acid completely finished in from
half to three-quarters of an hour after the beginning of
the filtration. If the fluid be tested with sulphuretted
I I
482 THE ASSAY OF COPPEE.
hydrogen, no trace even of copper will be detected ; the
spongy metal partly covers the platinum foil, partly floats
about in the liquid, and, in case either the ore itself or
the zinc applied in the experiment contained lead, small
quantities of that metal will accompany the precipitated
copper. After the excess of zinc (for an excess must be
always employed) has been removed, the spongy metal is
repeatedly and carefully washed by decantation with fresh
water, which need not be distilled, and care is taken to
collect together every particle of the spongy mass.
3. Quantitative Estimation of the Precipitated Copper.
To the spongy metallic mass in the beaker, wherein the
platinum foil is left, since some of the metal adheres to it,
8 c.c. of the normal nitric acid are added, and the copper
is dissolved, by the aid of moderate heat, as copper
nitrate : in the event of any small quantity of lead being
present, it will of course be contaminated with lead
nitrate.
When copper ores are dealt with which contain above
6 per cent, of copper, which may be sufficiently judged
from the larger bulk of the spongy mass of precipitated
metal, 16 c.c. of nitric acid, instead of 8, are employed
for dissolving the metal. The solution thus obtained
is left to cool, and next, immediately before titration
with potassium cyanide, mixed with 10 c.c. of normal
solution of ammonia, prepared by diluting 1 volume of
liquid ammonia, sp. gr. 0*93, with 2 volumes of distilled
water.
In the case of ores which yield over 6 per cent, of
copper, and when a double quantity of normal nitric acid
has consequently been used, the solution of copper in
nitric acid is diluted with water, and made to occupy a
bulk of 100 c.c. ; this bulk is then divided exactly into
two portions of 50 c.c. each, and each of these separately
mixed with 10 c.c. of the liquid ammonia solution just
alluded to, and the copper therein volumetrically esti-
mated. The deep blue-coloured solution of oxide of
copper in ammonia only contains, besides ammonium
nitrate, any lead which might have been dissolved, having
STEINBECK'S PROCESS. 483
been precipitated as hydrated lead oxide, which does not
interfere with the titration with potassium cyanide. The
solution of the last-named salt is so arranged that 1 c.c.
thereof exactly indicates 0-005 grm. of copper. Since,
for every assay, 5 grins, of ore have been taken, 1 c.c. of
the titration fluid is, according to the following, proportion
5 : O'OOo :: 100 : O'l equal to O'l per cent, of copper ;
it hence follows that, by multiplying the number of the
c.c. of potassium cyanide solution used to make the blue
colour of the copper solution disappear, by O'l, the per-
centage of copper contained in the ore is immediately
indicated.
As may be imagined, at the laboratory of the mine-
owners at Eisleben, such a large number of assays are
daily executed that, in this case, there can be no reason to
fear a deterioration of the cyanide solution, of which large
quantities are used and often fresh made ; but for security's
sake the solutions are purposely tested for control at least
once every week. According to the described plan, six
assays can be made within 4 hours ; and during a work-
ing day of from 7^ to 8 hours, twenty assays have been
often quite satisfactorily made by the umpires, as well as
by the workmen at Eisleben.
Special Observations on this Method. Dr. Steinbeck
considered it necessary to test this method specially, in
order to see what influence is exercised thereupon by (1)
ammonium nitrate, (2) caustic ammonia, (3) the presence of
lead oxide. The copper used to perform the experiments
for this purpose was pure metal, obtained by galvano-
plastic action, and was ignited to destroy any organic
matter which might accidentally adhere to it, and, next,
cleaned by placing it in dilute nitric acid. Five grammes
of this metal were placed in a litre flask, and dissolved in
266 '6 c.c. of normal nitric acid, the flask and contents
gently heated, and, after cooling, the contents diluted with
water, and thus brought to a bulk of 1,000 c.c. exactly.
Thirty c.c. of this solution were always applied to test and
titrate one and the same solution of potassium cyanide
under all circumstances. When 5 grammes of ore, con-
i i 2
484 THE ASSAY OF COPPER.
taining on an average 3 per cent, of copper, are taken for
assay, that quantity of copper is exactly equal to 0-15
gramme of the chemically pure copper. The quantity of
normal nitric acid taken to dissolve 5 grammes of pure
copper (266 -6 c.c.) was purposely taken, so as to corre-
spond with the quantity of 8 c.c. of normal nitric acid
which is applied in the assay of the copper obtained from
the ore, and this quantity of acid is exactly met with in
30 c.c. of the solution of pure copper.
As regards No. 1 and No. 2 (see above), the influence
of double quantities of ammonium nitrate and free caustic
ammonia (the quantity of copper remaining the same), and
the action of dilute solution of potassium cyanide there-
upon, will become elucidated by the following facts :
a. Thirty c.c. of the normal solution of copper, con-
taining exactly 0-15 gramme of copper, were rendered
alkaline with 10 c.c. of normal ammonia, and are found
to require, for entire decoloration, 29-8 c.c. of potassium
cyanide solution. A second experiment, again with 30 c.c,
of normal copper solution, and otherwise under identically
the same conditions, required 29-9 c.c. of cyanide solution.
The average of the two experiments is 29*85 c.c.
b. When to 30 c.c. of the normal copper solution 8 c.c.
of normal nitric acid are first added, and then 20 c.c. of
normal ammonia solution, instead of only 8, whereby the
quantity of free ammonia and of ammonium nitrate is
made double what it was in the case of the experiments
spoken of under a, there is required of the same cyanide
solution 30-3 c.c. to produce decoloration. A repetition
of the experiment, under exactly the same conditions, gave
30-4 c.c. of the cyanide solution employed ; the average
of both experiments is, therefore, 30-35 c.c.
The difference between 30-35 and 29-85 is equal to
0-5 c.c., and that figure is therefore the coefficient of the
influence of double quantities ; and supposing this to
happen with the ores in question, it would only be equiva-
lent to 0-05 per cent, of metallic copper. It is hence
clear that slight variations of from 0-1 to 0'5 c.c. in
the measuring out of 8 c.c. of normal nitric acid, used
STEINBECK S PEOCESS. 485
to dissolve the spongy copper, and of 10 c.c. of normal
ammonia, in order to render the nitric acid copper solu-
tion alkaline, are of no consequence whatever for the
technical results to be deduced from the assay. It should,
moreover, be borne in mind that the quantities of free
.ammonia and of ammonium nitrate in the actual assay of
ores, for which always a quantity of 5 grammes of ore is
taken, vary according to the richness or poverty of the
ores in copper ; and the quotation of the following results
of experiments proves that the influence of these sub-
stances is only very slightly felt in the accuracy of the
results :
Eight c.c. of the normal nitric acid have been found
to contain, by a series of experiments, 1*353 gramme
of anhydrous nitric acid ; and this quantity of acid is
-exactly neutralised by 7*7 c.c. of normal ammonia solu-
tion, which contains 0-6515 gramme of ammonium oxide ;
and 10 c.c. of the said normal solution contain 0-846
gramme of ammonium oxide.
One gramme of metallic copper requires, for complete
oxidation, 0-2523 gramme of oxygen, and this quantity of
-oxygen is given off by 0*5676 gramme of anhydrous nitric
acid ; while, at the same time, binoxide of nitrogen is
disengaged. From these data can be calculated (1) the
quantity of nitric acid which becomes decomposed when
variable quantities of metallic copper are dissolved
therein; (2) what quantity of nitric acid is left to form
neutral nitrate of ammonium ; and (3) what quantity of
free ammonia will be left after a portion of that alkali has
been combined with, and therefore neutralised by, copper
oxide ; and any remaining free nitric acid.
It is found that the quantitative variations between
ores containing 1 per cent, or 6 per cent, of metal vary
very little from the normal quantities exhibited by ores
containing 3 per cent, of metal. The relation is as 1 : 2 ;
-and, for technical purposes, this has been proved not to
be a disturbing quantity.
When, however, larger quantities of ammoniacal salts
are present in the fluid to be assayed for copper by means
486 THE ASSAY OF COPPER.
of a titrated solution of potassium cyanide, and especially
when ammonium carbonate, sulphate, and, worse still,
chloride are simultaneously present, these salts exert a
very disturbing influence.
The presence of lead oxide in the copper solution to
be assayed has the effect of producing, on the addition of
10 c.c. of normal ammonia, a milkiness along with the
blue tint ; but the presence of this oxide does not at all
interfere with the estimation of the copper by means of
the cyanide, provided the lead be not in great excess ; and
a slight milkiness of the solution even promotes the visi-
bility of the approaching end of the operation.
Dr. Steinbeck, however, purposely made some experi-
ments to test this point, and his results show that neither
50 nor 100 per cent, of addition of lead exerts any per-
ceptible influence upon the estimation of copper, from its
ores or otherwise, by means of potassium cyanide. A
small quantity of accidentally occurring lead will not,
therefore,- affect the results, and this the less so as, gener-
ally, no ores of both metals occur together wherein both
are met in sufficient quantity to make it worth while
working the ore for both metals at the same time.
Since it is well known that the presence of zinc very
perceptibly influences the action of a solution of potassium
cyanide, when applied to the volumetrical estimation of
copper, Dr. Steinbeck considered it necessary to institute
some experiments in order precisely to ascertain with
what quantity of zinc present along with copper this in-
fluence commences to become perceptible. The solution
of zinc applied was made by dissolving the metal in the
smallest possible quantity of nitric acid ; and 1 c.c. of that
solution contained 0*001 gramme of zinc. The results of
the experiments show that the presence of zinc does not
interfere with the visibility of the end of the reaction, viz.
the decoloration of the copper solution. They also prove
that a small quantity of zinc, less than 5 per cent, of the
quantity of copper present, or 0*0075 gramme by weight
of zinc, does not at all affect the action of the solution
of potassium cyanide ; but when the quantity of zinc
- LUCKOW'S PROCESS. 487
increases, a very perceptible effect is seen upon the
solution of cyanide : it is therefore necessary to bestow
due care while washing the spongy copper after it has
been precipitated by means of zinc from its solution.
Since it has been ascertained that the action of the
solution of potassium cyanide in researches of thi's kind
is also affected by an increased temperature of the solu-
tion of copper which is to be titrated, it is strictly neces-
sary never to operate with warm ammoniacal solutions of
copper, but to suffer the same to cool down to the ordinary
temperature of the air of the laboratory.
While 30 c.c. of copper solution, containing 0*15
gramme of copper, and 10 c.c. of normal ammonia solu-
tion, required at the ordinary temperature 30 c.c. of
cyanide solution, the same quantities required, at between
40 and 45 C., 28*8 c.c. of solution of cyanide ; and at. 45
C., 28'9 c.c. of the same solution, thus proving the injurious
effect of warm solutions.
ESTIMATION OF COFFEE IN THE MANSFELD ORES BY
M. C. LUCKOW'S PROCESS.
This gentleman has applied to the quantitative estima-
tion of copper a new method, based upon the. precipitation
of the metal in the metallic state from solutions containing
either free sulphuric or nitric acid, by means of a galvanic
current.
It is a great advantage of this method, that, while the
copper is precipitated, it is simultaneously separated from
metals witji which it is often found alloyed ; some of these,
such as tin and antimony, are separated by treatment with
nitric acid in an insoluble form ; while others, like silver,
can easily be removed in the form of chloride. It is,, at
the same time, another advantage that the state in which
the copper is obtained admits of its being accurately
weighed and estimated, while a great number of opera-
tions, which require much time and various apparatus, are
at the same time got rid of.
Although M. Luckow had previously discovered a
488 THE ASSAY OF COPPER.
method of electro-metallic analysis from fluids containing
free sulphuric acid, his researches on the same subject,
in the case of free nitric acid, belong to a recent period.
These researches brought very unexpectedly to light the
curious fact that even a weak galvanic current had the
power of completely precipitating copper in a pure metal-
lic state from nitric solutions, provided they did not con-
tain more than O'l gramme of anhydrous nitric acid to
the c.c. (nitric acid of 1'2 sp. gr. contains 0-32 gramme
of anhydrous nitric acid to the c.c.) ; while it was found
that the action was, at the same time, more regular, and
less dependent upon the power of the current than when
free sulphuric acid was present. The following more com-
monly occurring metals are not precipitated by galvanic
action from acid solutions : Zinc, iron, nickel, cobalt,
chromium, the metals of the earths and alkalies. The fol-
lowing are precipitated (a) in the shape of peroxides, at the
positive electrode : completely, lead and manganese ; in-
completely, silver. When the solution contains traces of
manganese, it becomes, in consequence of the formation of
a salt of manganese peroxide, or of permanganic acid, deeply
violet-coloured. This very sensitive reaction for man-
ganese also takes place when small quantities of chlorine
are present. The presence in the fluid of oxalic and
tartaric acids, and other readily oxidisable organic sub-
stances, and such protoxides as are readily peroxidised
for instance, iron protoxide retards the formation of
peroxides, as well as the occurrence of the reaction of
manganese.
(b) Mercury, silver, copper, and bismuth are precipi-
tated at the negative electrode in a metallic state. When
mercury is present in the solution simultaneously with
copper, the former metal is separated before the latter, in
the fluid metallic state. As soon, however, as the pre-
cipitation of copper commences, an amalgam of the two
metals is formed, when mercury is also present. Silver is
precipitated almost simultaneously with copper ; bismuth
only begins to be precipitated after the greater portion of
the copper has been separated. A complete separation of
LUCKOWS PROCESS. 481)
silver only ensues when some such substance as tartaric or
other similar acid is simultaneously present in the solution.
The separation of the three last-named metals by means
of galvanic action is, therefore, unsuccessful ; but, fortu-
nately, there are many other means to accomplish this end
completely.
(c) Metallic arsenic is only precipitated slowly, and
long after the complete separation of copper, if arsenic
acid happen to be present. The same remark applies to
antimony, since it is well known that small quantities of
antimonic acid are soluble in nitric acid.
The operations, according to Luckow's plan, are
1. Eoasting the ore; 2. Solution of the roasted product;
3. Precipitation of the copper ; 4. Weighing the copper.
1. Eoasting the Ore. Care should be taken to obtain
a finely ground average sample of the ore. Then weigh
off in small porcelain capsules, previously counterpoised,
quantities of from 1 to 3 grammes ; these quantities are then
placed on the inverted lid of an iron crucible, on the inner
surface of which the powdered ore is heated over the flame
of a Bunsen gas burner. The powder may be carefully
stirred up with a platinum wire, to promote the access of
air during the roasting ; the ignition of bituminous matter
and sulphur will be ended in about seven minutes. Ores
which do not contain bitumen at all need not be roasted.
It has been already stated that in the case of poor
copper ores (and those of the Mansfeld district are gener-
ally so), the quantities to be weighed off for assay should
not vary according to a presumed percentage of copper.
Two grammes are therefore taken, and, instead of roasting
the ore on the lid of an iron crucible, small porcelain
crucibles are used for that purpose.
2. Solution of the Roasted Product. The iron lid is
suffered to cool, the roasted powder placed on a piece of
glazed paper, and any powder adhering to the lid is re-
moved by means of a camel's- hair brush on to the paper.
The powder is next transferred to small beaker glasses,
and about 2 or 3 c.c. of nitric acid, of 1-2 specific gravity,
are added, along with about 10 to 15 drops of concentrated
490 THE ASSAY OF COPPER.
sulphuric acid. The beakers are then placed on a sand-
bath and moderately heated, at first ; but when the con-
tents have become nearly dry, the heat is increased, so as
to evaporate and expel all sulphuric acid. The beakers
should be covered with perforated watch-glasses. This
operation requires from about three-quarters of an hour
to one hour. The addition of sulphuric acid is made in
order to increase the oxidising action of the nitric acid,
and also to convert any lime which may happen to be
present in the ore into a difficultly soluble salt. It is very
useful, also, to add from 10 to 20 drops of hydrochtoric
acid to the mixture of the two acids just alluded to, since
the rapidity of the evaporation is thereby increased, and
the occasional spirting about of the fiuid is lessened.
The process just described may be modified, first, by
the use of porcelain capsules, the contents of which are
easily transferred to beakers with flat bottoms, and not
higher than about 2 inches altogether. It is better, also,
to use sulphuric acid, prepared with equal bulks of con-
FIG. 105. centrated acid and water, and to measure
off 4 c.c. for each assay ; while for each
assay, moreover, 6 c.c. of nitric acid and
about 25 drops of hydrochloric acid are
taken. Instead of covering the beaker
with a perforated watch-glass, the upper
part of a funnel is used, as represented
in fig. 105 ; with this arrangement the sulphuric acid
evaporates far more readily, and loss by spirting is pre-
vented. The beaker is heated on a well-arranged sand-
bath.
3. Precipitation of the Copper. As soon as the beaker
after removal from the sand-bath has become quite cool,
the funnel which has been used as a cover is washed on
both sides, inner and outer, with nitric acid of 1*2 sp. gr.,
diluted with six times its bulk of pure water ; the sides of
the beaker are next likewise washed, and it is then filled
to about half its height with the same acid. A few drops
of a concentrated solution of tartaric acid are added (this
acid is best kept in solution in open vessels, only slightly
LUCKOWS PROCESS.
491
covered with a piece of paper) ; this having been done,
the wire spiral, represented in fig. 106, is carefully placed
in the beaker. This spiral consists of a piece of platinum
wire, about l-12th of an inch thick, and 7-J- inches long,
two-thirds of its length being so wound that the straight
end of the wire projects as if it were the axis of the centre
of the spiral. The convolutions of the spiral are so large
that they touch the sides of the beaker, while the straight
portion just touches the centre of the bottom of the vessel.
When the heating has been carefully attended to,
the acid fluid added to the contents of the beaker, after
evaporation to dryness, will generally be quite clear ; if it
FIG. 106.
FIG. 107.
FIG. 108.
happens to be turbid, 1 or 2 c.c. of a concentrated solution
of barium nitrate may be added, and the thorough mixing
of this saline solution with the acid contents of the beaker
promoted by gently moving up and down the platinum
spiral just alluded to, and allowing the fluid to rest for -a
few minutes after. The copper present in the mass left
at the bottom of the beaker gradually dissolves, and it is
not actually requisite to wait before applying the galvanic
current until it is all dissolved.
The next point is to place in the beaker the platinum
492 THE ASSAY OF COPPER.
foil, represented at fig. 107, of which the dimensions are
length, 2^ inches ; width, 1J inch. The lower end of this
platinum foil should be kept about 1-1 6th of an inch apart
from the convolutions of the spiral. When the beaker is
only half filled with liquid, the platinum foil is immersed
in the same for more than three-fourths of its height. The
wire fastened to this foil is fixed, by means of a screw a,
to the arm a b of the stand, represented in fig. 108 ; the
other screw, b, serves to fasten a copper wire proceeding
from the zinc end of the galvanic battery. When the
small screw clamp c (fig. 108) has been fastened to the
platinum wire placed in the beaker, another wire is
fastened in the top opening of the clamp, and this wire
connected with the copper end of the battery, and the
galvanic circuit thus closed. In a few moments after this
has been done, the platinum foil, bent in the shape of the
cylinder and placed inside the beaker, as before described,
will be observed to become covered with a coating of
metallic copper, while from the platinum wire spiral
bubbles of gas escape, which facilitate, to some extent, the
solution of the copper oxide in the dilute acid.
In order to ascertain whether the whole quantity of
the copper has been precipitated, some more dilute nitric
acid is added to the fluid in the beaker glass. If, in ten
minutes after this, no more metallic copper is separated on
the clean portions of the platinum foil, the operation is
finished.
It must be here observed that continued practice has
proved that the addition of a concentrated solution of
barium nitrate acts injuriously on the process, as the
metallic copper, which becomes separated, gets mixed
with some insoluble barium sulphate, which increases the
weight of the substance to be weighed.
The time occupied by the complete precipitation of the
metal varies according to the force of the galvanic current.
It takes from three to even eight hours. In order to make
this point certain, all test assays are left, for eight hours
consecutively, to the action of the galvanic current, expe-
rience having proved that, after that lapse of time, even
LUCKOWS PROCESS. 493
with a weak current, the precipitation was so complete
that all chemical reagents for detecting the presence of
copper failed to discover the most minute trace of that
metal.
4. Weighing the Copper. The platinum cylinder to
which the copper adheres, and the platinum wire spiral,
are disconnected from the galvanic apparatus, the plati-
num cylinder carefully removed from the beaker and
immediately plunged into a beaker filled with fresh cold
water, and rinsed therein ; next washed with alcohol, by
means of a washing bottle, and then dried in a drying ap-
paratus, and weighed after cooling. Since the platinum
cylinder has been- very accurately weighed before the
experiment, its increase in weight will, of course, be that
of the copper obtained.
The process here described has been somewhat modi-
fied and greatly improved upon at Eisleben, where it is in
constant use, by the employment of a series of galvanic
elements. It is, in the first place, found better not to
disconnect the galvanic current while the copper is yet in
contact with acid, so that, instead thereof, the acid fluid in
the beaker is replaced by turning in a stream of water,,
and suffering the same to run over the sides of the beaker,
and to be received into a proper vessel to hold it. In
this manner all the acid is displaced, without risk of
any very small quantity of copper becoming acted upon
by the acid during the brief period elapsing between the
disconnecting of the galvanic current and the removal
from the beaker of the platinum cylinder and spiral wire.
These parts, on being removed, are carefully washed, first
with boiling water, next with alcohol, and then dried at a
temperature of about the boiling-point of water. The
cylinder is then weighed, the copper coating is removed
by means of nitric acid ; the platinum is next washed in
water, dried, and again weighed.
There are in use at Eisleben nine galvanic batteries
(lead and zinc elements) ; these yield eighteen assays ready
for weighing in twenty-four hours ; and it would not be
difficult for the person there employed to work with twelve-
494 THE ASSAY OP COPPER.
batteries each of three elements. In place of the Mei-
dinger elements, which do not remain constant for months,
a thermo-electric apparatus has been introduced, with the
best results. The results obtained are highly satisfactory.
The following observations may be made in reference to
this method :
(a) The quantity of ore taken for trial is 2 grammes ;
this is found sufficient, while it consumes less acid.
(b) The evaporation of the acid is carried on to com-
plete dryness on the sand-bath. Spirting of the fluid is
easily prevented.
When the copper has been precipitated properly it
will show its peculiar colour on the surface, and the good
success of the operation may also be judged from the fact
that no saline matter adheres to the platinum ; the com-
plete absence of this saline matter has been found to
be evidence of perfect removal of the copper from the
fluid.
The process just described is especially applicable for
rather poor ores, such as do not contain above 7 or 8 per
cent, of copper. Each assay, from beginning to end, takes
ten hours for complete analysis ; but it is evident that the
greater portion of this period does not give active employ-
ment to the assayer. The expense of working this process,
after the apparatus has been once purchased, is very small.
The process may also be applied to analyse richer ores,
and also alloys of copper, with some slight modifications
which will readily suggest themselves.
Assay of Copper Pyrites. The following method of
treating copper pyrites has been found more advantageous
than the ordinary process of oxidising the mineral with
aqua regia, and subsequently evaporating the solution re-
peatedly with hydrochloric acid, or with sulphuric acid, to
expel the last traces of nitric acid. It is thus described
by Mr. F. P. Pearson in the ' Chemical News ' :
Place a weighed quantity of the powdered mineral, to-
gether with some potassium chlorate, in a porcelain dish.
(Five grammes of a variety of a pyrites containing about
18 per cent, of copper was found to be enough for one
ASSAY OF COPPER PYRITES. 495
analysis ; and a quantity of potassium chlorate equal to a
small teaspoonful was added to the ore.) Invert a small
glass funnel with bent stem in the dish above the pyrites,
and pour upon the latter rather more ordinary strong
nitric acid than would be sufficient to completely cover
the powder. Place the dish upon a water-bath, and, from
time to time, throw into it small quantities of potassium
chlorate. The doses of the chlorate must be repeated at
frequent intervals, until free sulphur can no longer be seen
in the dish. If need be, add nitric acid, also, from time to
time, to replace that lost by evaporation.
As a general rule, it is safer and more convenient to
heat the mixture on a water-bath than upon sand, though
the oxidation of sulphur can be effected more easily and
quickly when the mixture of nitric acid and chlorate is
heated to actual boiling than at the temperature obtain-
able by means of a water-bath. When the last particles
of sulphur have been destroyed, remove the inverted funnel
from the dish, rinse it with water, and collect the rinsings
in a beaker by themselves. Allow the liquid in the evapo-
rating-dish to become cold, pour upon it a quantity of
ordinary strong hydrochloric acid rather larger than the
quantity of nitric acid taken at first, evaporate the mixed
solution to dryness, and heat the dry residue to render
silica insoluble, in case any silica be present.
Pour water upon the cold residue, and, without filter-
ing the liquor, wash the contents of the dish into the
beaker which contains the rinsings of the funnel. Heat
the liquid in the beaker nearly to boiling, add to it about
25 c.c. of a strong aqueous solution of ferrous sulphate
slightly acidulated with sulphuric acid, and keep the mix-
ture at a temperature near boiling during four or five
minutes, in order to destroy the small quantity of nitric
.acid which may have escaped decomposition in spite of
the dry evaporation with hydrochloric acid.
The ferrous salt seldom acts instantaneously, but the
reducing action proceeds rapidly and satisfactorily when
once begun. If need be, add more of the ferrous solution,
little by little, until the entire contents of the beaker
496 THE ASSAY OF COPPER.
become dark-coloured or black, and no more gas is dis-
engaged.
In order to be sure that all the nitric acid has been
reduced, it is as well, after the mixture of liquid and solu-
tion of ferrous sulphate has been duly heated, to place
a drop of the mixture upon porcelain, and test it with
potassium ferrocyanide. In general, however, the colora-
tion of the liquid in the beaker, due to the formation of
nitrous or hyponitric acid, will be a sufficient indication
that the sulphate of iron has done its work. The nitrous
fumes quickly disappear from the liquid at a subsequent
stage of operations when metallic iron is immersed in the
solution.
When enough of the ferrous sulphate has been added,
filter the mixed solution into a wide beaker, precipitate the
copper in the metallic state upon a sheet of iron in the
usual way, and ignite the copper in a porcelain crucible,
in a current of hydrogen, before weighing it.
By means of the ferrous salt, the last traces of nitric
acid may be got rid of far more quickly, conveniently, and
certainly than by the old system of evaporating the pyrites
solution with several successive portions of hydrochloric
acid. By treating the pyrites with potassium chlorate and
nitric acid it is easy to oxidise and dissolve every particle
of the sulphur in the mineral, so that no portion of the
latter can escape decomposition by becoming enveloped
in free sulphur. When aqua regia is used, on the other
hand, or a mixture of potassium chlorate and hydrochloric
acid, a certain proportion of sulphur almost invariably
remains undissolved, and might easily enclose portions of
the mineral, so as to prevent them from the solvent action
of the acids.
For the Estimation of Copper and Sulphur in Copper
Pyrites, E. Fresenius (' Zeitschrift fur Anal, Chemie,' 1877,
p. 355) proceeds as follows :
' After drying at 100 C., and carefully preparing the
sample, he takes for the estimation of the copper 5
grammes pyrites, heats with 6 to 7 c.c. hydrochloric acid
(sp. gr. 1'17), adding gradually nitric acid of sp. gr. 1-37, till
SULPHUR IX COPPER PYRITES. 497
no more action ensues, and then digests for some hours
at a gentle heat. The contents of the flask are poured
into a porcelain capsule, the flask is twice rinsed, out with
10 c.c. hydrochloric acid (sp. gr. 1*12) into the capsule,
and is then set aside. The contents, of the capsule are
then evaporated almost to dry ness on the water-bath,
20 c.c. hydrochloric acid of sp. gr. 1/12 are added, heated,
mixed with water, and filtered into a boiling-flask holding
500 c.c. The solution-flask is also rinsed upon the filter
with water. The filtrate is dried, incinerated in a porcelain
crucible, and the residue (in part lead sulphate) is treated
with 1 c.c. aqua regia, evaporated to dryness, the residue
treated with 5 c.c. hydrochloric acid (1*12 sp. gr.) slightly
diluted, and the solution, which may contain a little
copper, is filtered into the main solution. The solution is
then precipitated with sulphuretted hydrogen at 70 C.,
the precipitate is filtered, washed, dried, mixed with the
thoroughly burnt ash of the filter, heated with 5 c.c.
nitric acid (of sp. gr. 1-2), filtered, and washed. The filter
is incinerated, the small residue is again heated with 2 c.c.
of the same nitric acid, diluted, filtered to the main solu-
tion, and washed. The solution is mixed with 12 c.c.
dilute sulphuric acid (1 part acid to 5 water), evaporated
to expel nitric acid, filtered to remove lead sulphate,
washing with water acidulated with sulphuric acid. The
copper is then precipitated with sulphuretted hydrogen
at 70 C., the precipitate washed, dried, ignited along
with the ash of the filter in a current of hydrogen, and
weighed as copper sulphide.
6 For the estimation of sulphur the author fuses
^ grm. of the sample with 10 parts of a mixture of
2 parts sodium carbonate and 1 part potassium nitrate,
and estimates the sulphuric acid formed. For burnt ores
he uses 4 parts sodium carbonate to 1 part of potassium
nitrate.'
Professor Chapman, of Toronto, gives the following
directions for the detection of minute traces of copper in
iron pyrites and other bodies :
Although an exceedingly small percentage of copper
K K
498 THE ASSAY OF COPPER,
may be detected in blowpipe experiments by the reducing
process, as well as by the azure-blue coloration of the
flame when the test matter is moistened with hydrochloric
acid, these methods fail in certain extreme cases to give
satisfactory results. It often happens that veins of iron
pyrites lead at greater depths to copper pyrites. In this case,
according to the experience of the writer, the iron pyrites
wi]l almost invariably hold minute traces of copper. Hence
the desirability, on exploring expeditions more especially,
of some ready test by which, without the necessity of
employing acids or other bulky and difficultly portable
reagents, these traces of copper may be detected.* The
following simple method will be found to answer the pur-
pose : The test substance, in powder, must first be roasted
on charcoal, or, better, on a fragment of porcelain,f in
order to drive off the sulphur. A small portion of the
roasted ore is then to be fused on platinum wire with
phosphor-salt ; and some potassium bisulphate is to be
added to the glass (without this being removed from the
wire) in two or three successive portions, or until the glass
becomes more or less saturated. This effected, the bead
is to be shaken off the platinum loop into a small capsule,
and treated with boiling water, by which either the whole
or the greater part will be dissolved ; and the solution is
finally to be tested with a small fragment of potassium
ferrocyanide. If copper be present in more than traces,
this reagent, it is well known, will produce a deep red
precipitate. If the copper be present in smaller quantity
that is, in exceedingly minute traces the precipitate
* In blowpipe practice as far, at least, as this is possible the operator
should make it an essential aim to render himself independent of the use of
mineral acids and other liquids and inconvenient reagents of a similar
character. If these reagents cannot be dispensed with altogether, their use,
by improved processes, may be greatly limited.
t In the roasting of metallic sulphides, &c., the writer has employed, for
some years, small fragments of Berlin or Meissen porcelain, such as result
from the breakage of crucibles and other vessels of that material. The test
substance is crushed to powder, moistened slightly, and spread over the sur-
face of the porcelain ; and when the operation is finished, the powder is easily
scraped off by the point of a knife-blade or small steel spatula. In roasting
operations, rarely more than a dull red heat is required ; but these porcelain
fragments may be rendered white-hot, if such be necessary, without risk of
fracture.
ESTIMATION OF ARSENIC IN COPPER. 499
will be brown or brownish- black ; and if copper be en-
tirely absent, the precipitate will be blue or green assum-
ing, of course, that iron pyrites or some other ferruginous
substance is operated upon. In this experiment the pre-
liminary fusion with phosphor-salt greatly facilitates the
after solution of the substance in potassium bisulphate.
In some instances, indeed, no solution takes place if this
preliminary treatment with phosphor-salt be omitted.
Estimation of Arsenic in Copper.
The estimation of the small quantity of arsenic always
present in commercial copper, or the separation of a very
small quantity of arsenic from a large amount of copper,
is a matter of considerable difficulty, as the ordinary
methods of separation fail to give accurate results.
Having had a very extensive experience in the analysis
of copper, and knowing the extraordinary discrepancies
which occur in analyses of the same sample of copper by
different chemists, which can only arise from the use of
imperfect methods, Mr. A. Humboldt Sexton proposes
method which he has found to give quite accurate
results.
The copper is dissolved in nitric acid, a small quantity
of solution of ferric nitrate added, the solution nearly
neutralised with sodium hydrate (not ammonia), and excess
of sodium acetate added. The solution is then heated to
boiling, and filtered as rapidly as possible : the precipitate
after being well washed is dissolved in hydrochloric acid,
the solution made alkaline with ammonia and saturated
with sulphuretted hydrogen, and filtered from the precipi-
tated iron sulphide. The filtrate is acidified with hydro-
chloric acid and allowed to stand in a warm place for
some time. The arsenic and antimony sulphides are
filtered off, dried at 100 C., the precipitates removed
completely from the paper into a small beaker, treated
with red fuming nitric acid, a few drops of hydrochloric
acid being added as soon as the action has ceased. It is
then diluted, filtered, the arsenic precipitated as ammonia-
500 THE ASSAY OF COPPER.
magnesium arseniate, and weighed as usual. If the preci-
pitated sulphides cannot be perfectly removed from the
filter-paper, the paper must be treated with nitro-hydro-
chloric acid, filtered, and the filtrate added to the nitric
acid solution.
This method is very accurate, and each stage has been
carefully experimented upon. It requires, however, some
special precautions.
When the sodium acetate is added, the colour of the
solution should change from pale blue to dark green ; this
shows that the solution has been sufficiently neutralised.
The beaker must be removed from the heat immediately
the solution begins to boil ; if the solution be left boiling
(sometimes when it is not), a greenish white precipitate
of basic copper acetate- falls. This can generally be
removed by the addition of a few drops of hydrochloric
acid, but in cases where it has separated on the surface of
the beaker, or where it will not readily dissolve, it is best
to throw out the solution and commence again.
This is very troublesome to those using this method for
the first time, but after a little experience has been gained
it very rarely happens.
The precipitate should have the dark red colour of
ferric acetate ; if it is paler it is due either to there not
being sufficient iron, or to the co-precipitation of some
basic copper acetate. The filtrate should be blue or
pale green : sometimes it is dark green and turbid, from
the presence of iron acetate carried through the filter ; in
that case the first portions must be passed through the
filter again.
The precipitate must be washed till it is free from
copper, and when it is dissolved in hydrochloric acid the
solution must have the yellow colour of ferric chloride.
If it is at all green, the solution must be neutralised, a
little more sodium acetate added, and the iron and arsenic
re-precipitated.
Mr. Sexton has made a larger number of experiments
in order to ascertain the amount of iron necessary for
complete precipitation. With equal quantities of iron
ESTIMATION OF ARSENIC IN COPPER. 501
and arsenic, a small quantity of arsenic remained in solu-
tion, and the iron-arsenic precipitate was of a pale colour.
With 1-5 parts of iron to 1 of arsenic the precipitation
was complete. In order to make sure, it is well to add
about twice as much iron as it is expected there is arsenic
present. Then, even if a little iron remains unprecipitated,
all the arsenic will be thrown down.
Since copper sulphide retains so much arsenic, it might
be expected that iron sulphide would act in a similar
manner, but it does not ; if there be no copper present
the precipitate is quite free from arsenic, but if copper is
present a considerable quantity of arsenic may be retained.
Hence the importance of thoroughly washing the acetate
precipitate, and re -precipitating it if necessary.
502
CHAPTER XI.
THE ASSAY OF LEAD.
ALL minerals and substances containing lead may, for the
purposes of assay by the dry way, be divided into four
classes :
Class I. comprises sulphides, antimonial or otherwise
(galena, &c.)
Class II. includes all plumbiferous substances contain-
ing neither sulphur nor arsenic, or mere traces only of
these elements (litharge, minium, lead carbonate, native
and artificial, lead fume, cupel bottoms, furnace hearths,
lead slag, &c.)
Class III. comprises all substances into whose com-
position either sulphuric, arsenic, chromic, or phosphoric
acid, or a mixture of either, enters (pyromorphite, wolfram-
ite, &c.)
Class IV. Alloys of lead.
CLASS I.
Before describing the different modes of assaying sub-
stances of this class, it will be as well to pass in review the
action of various reagents on sulphides of lead, in order
that the rationale of the assay of those ores may be better
appreciated.
Action of Oxygen. If galena be roasted at a very
gentle temperature, care being taken to avoid fusion, it-
will be converted into a mixture of lead oxide and lead
sulphate, with evolution of sulphurous acid, thus:
2(PbS) + 70 =PbO + PbO,S0 3 + S0 2 .
Action of Metallic Iron. This metal completely and
ALKALIES AND ALKALINE CARBONATES. 503
readily decomposes lead sulphide, giving metallic lead in
a pure state, thus :
On the one side we have lead sulphide and metallic iron,
on the other metallic lead and iron sulphide.
The Alkalies and Alkaline Carbonates decompose lead
sulphide, but only partially ; pure lead is separated, and
at the same time a very fusible grey slag is formed, which
contains an alkaline sulphate and a compound of lead sul-
phide and an alkaline sulphide. A certain proportion of
the alkali is reduced by the sulphur, which is converted
into sulphuric acid, so that no lead oxide is produced.
This reaction may be thus expressed :
7(PbS) + 4(K 2 0)=4Pb + K 2 0,S0 3 + 3(PbS,K 2 S).
Potassium Nitrate completely decomposes lead sul-
phide, with the reduction of metallic lead and formation
of potassium sulphate and sulphurous acid, thus :
If the nitre be in excess, the lead will be oxidised in pro-
portion to the excess present ; and if there be a sufficiency
added, no metallic lead at all will be produced.
Argol. The presence of carbonaceous matter much
favours the decomposition of galena, by determining the
reduction of a larger quantity of potassium, and thereby
the formation of a larger quantity of alkaline sulphide.
With 4 parts of argol to 1 part of sulphide, 80 parts of
lead are reduced. If the reaction were complete, the
decomposition would be as follows :
For the reactions of lead oxide (litharge) and lead sul-
phate on sulphide of lead, see pages 187 and 188.
From the reactions above given, it will be seen that
there are many substances capable of completely reducing
the lead from its sulphide, and yet few can be used safely
504 THE ASSAY OF LEAD.
with any advantage, as so to use them would imply a
knowledge of how much sulphur and lead were in the ore
to be assayed, in order to tell the precise quantity of either
of the reagents required ; for it is evident that if either
more or less of some were added, a faulty result would be
the consequence : so that some systematic mode of assay,
which may be suitable for all classes of galena, whether
mixed with other sulphides or with gangue, must be
contrived. To facilitate this we now proceed to give an
outline of the methods generally adopted in the assay of
lead ores by various processes.
1. FUSION WITH POTASSIUM CARBONATE.
This plan is used at the Oberhartz, and described by
Kerl as follows :
One centner of the very finely pulverised assay sub-
stance is weighed out, mixed with three to four times its
weight of pure, dry, and finely pulverised potassium car-
bonate, and covered over, in a small clay crucible (fig. 61),
with a layer of decrepitated sodium chloride about one-
fourth of an inch thick. The assays thus prepared are
placed in the thoroughly heated muffle of a large assay
furnace (figs. 23, 24) having a strong draught. They re-
main in the highest temperature of the furnace, with the
mouth of the muffle closed with glowing coals, till they
have come into perfect fusion (about twenty to thirty
minutes). The draught opening is then closed, and at the
same time the muffle opened, until the temperature has
fallen so far that the crucibles appear brownish red, and
the vapours above them have greatly diminished, or have
disappeared. At this heat the crucibles whose contents
must, however, always remain in perfect fusion are main-
tained, according to the fusibility and composition of the
assay sample, and the draught of the furnace for a longer
or snorter time (ten to twenty-five, generally ten minutes).
This period, during which the heat is allowed to remain
low, is called the cooling of the assay.
The furnace is now again brought back to its first
FUSIOX WITH POTASSIUM CAEBONATE. 505
temperature by completely opening the draught and
closing the muffle. Ten to fifteen minutes of this last
heating are in most cases sufficient. Only poor ores, &c.,
which contain also a pretty large quantity of arsenic, or
of iron, zinc, and copper sulphides, are allowed to continue
hot five to ten minutes longer.
If many assays are to be made, it will be found advan-
tageous to mix those which contain larger quantities of
foreign sulphides, or, by reason of their earthy contents,
are difficultly fusible, with more or less borax ; or, instead
of this, to place them in the back and hotter part of the
muffle, while those that are very rich in lead and easily
fusible are placed in front, since the latter will be hot
enough here, and more easily reached by the air than
those deeper in the muffle.
The crucibles, when cold, are broken, the lead buttons
obtained are freed from all adhering slag or substance of
the crucible, and if the assay were otherwise successful
their weight found. The assays should not be too rapidly
cooled, because the slag is thus easily cracked, and the
still half-fluid button lying below is apt to be broken into
several pieces.
In a successful assay, the lead melted together to a
button deports itself under the hammer and knife like pure
lead, and possesses also its colour. If the slag shows, upon
its surface of separation from the metallic button, lead-
grey spots with metallic lustre, it will generally also be
found that a thin layer of not completely decomposed
glistening lead sulphide or subsulphide has at the same
time deposited itself upon the button. This layer, if the
above appearance presents itself in a high degree, can be
rubbed off or removed in fine scales. The lead button
itself then shows upon its surface a high metallic lustre,
which does not have the colour of pure lead, but a darker
and blackish hue. Assays of this kind are to be rejected ;
they have not been allowed to remain cool long enough,
or they have in the process become too cold: they give
the amount of lead too low, and often very considerably
so. In assays which have stood too long in the furnace
506 THE ASSAY OF LEAD.
in the last fusing heat, a very bright button of lead is also
found ; but here the layer of undecomposed lead sulphide
is wanting, as also the glistening spots on the surface of
the slag surrounding the button. If the influence of the
heat and air continues too long, then, besides a loss through
volatilisation of the lead, a slagging of the lead oxide may
take place. A button that is brittle, laminated, and bril-
liantly white in the fracture, indicates an insufficiency of
flux, or the presence of antimony and arsenic. In successful
assays the lead button generally has a bluish appearance,
which, although not dull, is at the same time not strongly
brilliant. The slag must be completely homogeneous, and
must have settled down uniformly towards the bottom of
the crucible, so that it does not stick in a thick layer to
the upper part of the sides of the crucible. It shows by
this that it has been in proper fusion. It must have
covered over the button in a thick layer (about one-fourth
of an inch thick). The sodium chloride covering, or a
more or less colourless slag that is formed containing
sodium chloride and potassium carbonate, overlies in a
still thicker layer the true dark-coloured slag containing
the foreign metallic oxides. A porous slag containing
metallic globules indicates a too small quantity of flux or
too low a temperature ; a brilliant vitreous slag, too high
a temperature and a slagging of lead. An assay and its
duplicate must, moreover, give equal results.
Lead matte and lead fume are smelted, with the addi-
tion of borax and coal-dust, with potassium carbonate, and
with the first the heat is allowed to last somewhat longer
(perhaps to three-quarters of an hour) than with ores.
The potassium carbonate assay gives for lead matte, with its
not inconsiderable lead contents (30 per cent, and over),
pretty satisfactory results.
The theory of this lead assay appears from the fol-
lowing.
If perfectly pure galena is intimately mixed with three
or four times its weight of good dry potassium carbonate,
placed in a clay retort, and this so arranged in the muffle
of the assay furnace that its neck projects from the mouth
FUSION WITH POTASSIUM CARBONATE. 507
of the muffle, while in the opening of the neck a glass tube
is closely fitted, which goes into a receiver, from which it
is further prolonged in a second tube, it will be observed
that at first only a little water collects in the receiver, pro-
ceeding from the small quantity of moisture always present
in the potassium carbonate. Later, with an incipient red
heat in the retort, a gas is disengaged, which upon closer
investigation proves to be pure carbonic acid gas, i.e. free
from sulphurous acid. The disengagement of gas becomes
more active with a stronger red heat, without yielding
gases of different composition, but ceases again after a
while. In order to obtain assurance of a complete decom-
position, the retort may be kept for an hour at a very
strong red heat. After the cooling and breaking of the
retort, some pure lead oxide and carbonate is found
deposited in the neck of it, then a pure lead button upon
the bottom, and over this a brown slag, free from little
globules of lead. It consists in by far the greatest part
of potassium sulphide and still undecomposed potassium
carbonate, but also in small part of potassium silicate de-
rived from the silica of the retort. If this slag is treated
with water till nothing further will dissolve, the substances
named can be easily shown to exist in the solution. The
solution is colourless, and when supersaturated with acids
disengages sulphuretted hydrogen, but throws down no
sulphur. In the treatment of the slag with water, lead
sulphide remains behind in black flocks, even the superfi-
cial character of which shows that it is not undecomposed
galena, but lead sulphide separated from a chemical com-
bination.
If the brown slag from the retort is placed in a small
uncovered crucible and brought back into the hot muffle
of the assay furnace and melted, then after some time,
whether the slag was covered with sodium chloride or not,
a button of lead again separates at the bottom of the cru-
cible, and the brown slag now shows itself decolourised.
If the crucible is removed from the furnace too soon,
only the upper layer of slag is decolourised, and that lying
below is still completely unchanged. The decolourised slag
508 THE ASSAY OF LEAD.
consists of potassium carbonate and sulphate, and no
longer contains any trace of potassium sulphide.
In the above-described lead assay, the process in the
strong preliminary heat proceeds as in the retort, i.e. the
potash of the potassium carbonate is reduced to potassium,
while it yields its oxygen to the sulphur of the galena and
with it forms sulphuric acid ; the liberated potassium takes
up sulphur from another portion of galena, forming potas-
sium sulphide. The galena would now in this double way
soon lose all its sulphur, if a combination a sulphur salt
of potassium sulphide with lead sulphide did not form,
which resists all further action of the potassium carbonate
[4 (K 2 0,C0 2 ) + 7PbS = 4Pb + 3 (K 2 S,PbS) +K 2 0,S0 3 + 4COJ.
The carbonic acid of the thus decomposed potassium car-
bonate escapes together with that set free by the sulphuric
acid formed, and causes a puffing up of the mass, by
which globules of lead already separated are raised up
with it, and may perhaps remain with some of the slag
sticking to the upper crucible walls. They would here
oxidise and produce yellow spots. The covering of sodium
chloride is designed to guard against loss of lead in this
and similar ways. It serves in a certain manner to rinse
down the sides of the crucible.
The atmospheric oxygen, in the open crucible, is not
entirely excluded by the covering of sodium chloride. In
the cooling of the assay, it oxidises the sulphur salt con-
tained in the upper part of the slag, forming potassium
sulphate and a portion of lead sulphate. The latter,
during the last high heat, decomposes the lead sulphide
still remaining in the slag, in such a way as to produce
metallic lead (PbS + PbO,S0 3 = 2Pb + 2S0 2 ). The reduced
particles of lead separate well from the slag thus rendered
thinly fluid. Mattes must be allowed to cool longer than
ores.
The potassium carbonate assay presupposes in general
great practice and close attention on the part of the
assayer ; and moreover, if one wishes to find the correct
value at once, without fruitless preliminary examinations,
and without the necessity of repeating the assay, a general
FUSION WITH POTASSIUM CARBONATE. 509
knowledge of the constituents of the assay sample, so far,
for example, as this can be obtained by the aid of minera-
logy, is necessary. The assay after this method, which
requires but little preparation, can only be conducted in
the muffle furnace, but then in pretty large number (as
many as fifty at once). For its success it is indispensably
necessary that the cooling of the assay be allowed and
stopped again at the right time and in the proper degree.
If it is allowed to cool too long, too much sulphate of lead
is formed in proportion to the lead sulphide still present
in the slag, and in the last heating up, by the action
of the two upon each other, easily scorifiable oxide of
lead is produced (PbS + 3PbO,S0 3 =4PbO + 4S0 2 ). If
the cooling is too soon interrupted, only a small part of
the lead sulphide in the sulphur salt is oxidised, and,
by the action of the oxidised portion upon the lead
sulphide, lead sub-sulphide is produced, which either
remains in the slag or settles upon the lead button
(2PbS + PbO,S0 3 = Pb 2 S + 2Pb + 2S0 2 ). Experience gives
the only means at hand to guide us here, but leaves
us easily in the lurch, so that the result of the assay
becomes more doubtful than in some of the methods-
hereafter described.
With substances containing antimony this assay deserves
the preference over the others, since most of the antimony
remains in the slag in the state of sulphide and oxide. An
addition of saltpetre works advantageously. Arsenic and
arsenic sulphide mostly go off in fumes during the smelt-
ing, but nevertheless always cause the formation of a
brittle metallic button. Copper sulphide remains in great
part in the slag, but a part of the copper is desulphurised
and goes into the lead. If the quantity of copper present
is very considerable, the button of metal may be considered
as black copper, and refined, and the loss thereby occur-
ring reckoned as lead.
Iron protosulphide, which occurs, for example, in lead
matte, is decomposed by potassium carbonate, forming
metallic iron, which desulphurises the galena. Iron pyrites ,
on the other hand, occasions the forming of a large quan-
510 THE ASSAY OF LEAD.
tity of potassium sulphide and, in consequence of this, of
a sulphur salt.
It follows, therefore, from the above, that ores which
contain much foreign sulphides are not suited to this
method of assaying, since they cause the production of a
large amount of potassium sulphide, which always retains
lead sulphide. By an addition of saltpetre to the potas-
sium carbonate these sulphides may, indeed, be partially
decomposed : only an oxidation of the lead is apt to be
produced, as well as a mechanical loss by the violent
action of the saltpetre.
From pure galena, by the potassium carbonate assay,
80 per cent, of lead at most can be obtained. Calcined
sodium carbonate is inferior to potassium carbonate as a
desulphurising agent, and always yields a few per cent,
less lead than the latter. According to .Phillips, 75 to 77
per cent, of lead is obtained from galena with sodium car-
bonate. With potassium cyanide, under certain circum-
stances, the same result can be obtained as with potassium
carbonate, and it does not require so high nor so long-
continued a temperature ; still it offers no real advantage
over potassium carbonate. An addition of 30 to 35 per
cent, of saltpetre to an assay, with which ten parts of
sodium carbonate are used, promotes, indeed, the desul-
phurising of the lead, but also increases its loss.
At the Oberhartz smelting-house the lead button is
weighed out to pounds, and a difference of five pounds is
allowed between different assayers. It is also a custom,
though not a correct one, to allow as many pounds differ-
ence as there are tens of pounds in the weight of the lead
button obtained. Thus, with a lead contents of thirty
and seventy pounds, the difference in the separate assays
might amount to three and seven pounds respectively.
2. FUSION WITH BLACK FLUX.
A modification of the preceding method of assaying,
which is sometimes employed, consists in using, instead of
the potassium carbonate, an equal quantity of black flux,
FUSION WITH METALLIC IRON. 511
or indeed of argol, or in mixing a few per cent, of pow-
dered charcoal or flour with the potassium carbonate, or
in replacing it in part by argol. Too great an addition of
carbon diminishes the fusibility of the mass, and hinders
the flowing together of the separated particles of lead. By
using argol the operation lasts longer, because the mass
remains pasty until most of the tartaric acid has been
decomposed ; but a greater product of lead is obtained.
The chemical reaction during the operation is thereby
modified so that the carbon of the black flux exerts an
influence upon the potash, and partially reduces it to
potassium ; the potassium, thus set free, works now, as
before, upon the galena. The latter is thus, without the
influence of the air, more completely decomposed than by
potassium carbonate alone, and the smelting is, therefore,
conducted in covered crucibles (fig. 61) in the wind fur-
nace. But since there is also potassium sulphide formed,
and this dissolves lead sulphide, it is more advisable, for
the completest possible separation of the lead, to perform
the smelting in open crucibles in the mulfle, in order to
allow the atmospheric oxygen to work at the same time
on the assay. The product of lead from pure galena does
not generally exceed 76 to 79 per cent.
At the Victor-Frederick smelting works in the Hartz,
one centner (= one hundred and fourteen assay pounds)
of galena is mixed with three or four times as much black
flux, and with pyritic ores ten pounds of borax-glass are
added. The mixture is covered with sodium chloride,
heated for about twenty-five minutes in the muffle furnace
with a charcoal fire, and then, after the mouth of the
muffle has been opened for about five minutes, taken out
of the furnace.
3. FUSION WITH METALLIC IRON.
Schlutter and many of the older assayers were aware
that iron would desulphurise galena, and ever after advised
the addition of a certain quantity of that metal to the
different fluxes which they used in lead assays ; but it was
512 THE ASSAY OF LEAD.
at the practical School of Mines, at Montiers, that iron was
first employed alone.
The process is extremely convenient and easy of exe-
cution ; it always succeeds, and requires no troublesome
attention. The fusion takes place quietly, without frothing
or bubbling ; and as the whole substance employed requires
but little space, very small pots may be employed, or a
very large quantity assayed. But this process can only be
employed for pure galenas, or those which contain at most
a few per cent, of gangue.
When galena is heated with iron, the metal is trans-
formed into protosulphide, whence it follows, that to de-
sulphurise galena 22 '6 per cent, is required ; but expe-
rience has shown that it is better to employ a little more,
and 30 per cent, can be used without inconvenience. The
iron employed ought to be in the state of filings, or wire
cut very small. The mixture is placed in a crucible,
which is three-fourths filled ; the whole is covered with a
layer of salt, sodium carbonate, or black flux, and exposed
to a full red heat. After the flux is perfectly fused, the
pot may be cooled and broken, and a button is obtained,
which at first sight has a homogeneous aspect, but on
being struck with the hammer separates into two distinct
parts. The lower part is ductile lead : the upper, a very
brittle matte, of a deep bronze colour, and slightly mag-
netic. Pure galena yields, by this process, 72 to 79 per
cent, of lead, so that there is a considerable loss, which
loss is entirely due to volatilisation. Berthier says that
it does not appear possible to avoid this loss, which
amounts from 6 to 13 per cent., giving as a reason that it
is probable galena begins to sublime before it arrives at
the proper heat for decomposition.
Antimonial galenas, or galenas mixed with iron pyrites,
may be assayed in the same manner ; but then a sufficiency
of iron must be added to reduce the antimony to the
metallic state, as well as to reduce the iron pyrites to the
minimum of sulphurisation. If the galena be mixed with
blende, the greater portion remains in the slag, because it
is only decomposed by iron at a very high temperature.
FUSION WITH METALLIC IRON. 513
Blende being infusible by itself, much diminishes the
fusibility of the mattes produced ; and if it exists in very
large quantity, it can even hinder their complete fusion ; in
which case some iron protosulphide and metallic iron must
be added to the assay, to make the slag more fusible.
All minerals are at a minimum of sulphurisation when
existing in mattes from metallurgical works ; therefore
much less iron may be used in their assay than if they
were pure ores. In very rich lead mattes, in which the
lead exists as subsulphide, from 10 to 12 per cent, is
sufficient. A small excess of iron may be employed with-
out inconvenience ; but if a larger proportion be added
than is necessary to execute the desulphurisation, the matte
contains some iron in the metallic state, and loses its
liquidity, and in consequence retains some globules of
lead.
The usual mode of assaying lead ores (galena) in the
lead mills is by a modification of this process : in lieu of
placing the ore in an earthen crucible, and adding nails or
filings, a given weight of the ore is projected into a red-
hot wrought- iron crucible, which is kept in the fire for
about a quarter of an hour, or until all the galena seems
decomposed. The lead thus reduced is poured into a
mould ; and if the scoriaceous matter be not well fused,
the iron crucible is returned to the fire and heated still
more strongly, and any lead that may be separated is
poured into the mould and weighed with the rest. This
is a very rude and imperfect process, and gives only
tolerable results with pure galenas, but is quite unsatisfac-
tory with those containing much earthy matter, as not
above half the lead is obtained, owing to volatilisation and
exposure to the air, and the loss of globules in the slag.
This process succeeds much better when a flux is added ;
this may be argol, or sodium carbonate, or a mixture of
both (see next process).
L L
514 THE ASSAY OF LEAD.
4. FUSION WITH SODIUM CARBONATE OR BLACK FLUX, AND
METALLIC IRON.
When galena is heated with an alkaline flux, out
of contact of air, the slag contains a double sulphide of
lead and the alkaline metal employed : if iron be thrown
into this fused mixture metallic lead separates, and the
iron combines with the sulphur formerly combined with the
lead, and the slag will contain a double alkaline sulphide,
containing iron sulphide instead of lead sulphide, thus :
PbS + K 2 S + Fe - Pb 2 + FeSK 2 S.
Any earthy substances the ore may contain will be
dissolved by the alkaline flux, without very much impair-
ing its fluidity. All these facts being considered, it may
be readily seen that the assay of all earthy bodies contain-
ing lead sulphide may be made in this manner, with as
much accuracy as this method of assay can be capable of.
Either sodium carbonate or black flux may be employed
as the alkaline reagent, and more of either of those
substances must be employed, in proportion to the in-
creased quantity of earthy matters the ore contains. Two
parts are nearly always more than sufficient for poor
ores, and are best for all cases, because an excess of
flux does not diminish the yield of lead ; nevertheless it
is sometimes convenient to employ, for the latter class,
but half a part. As to the iron, it is employed only to
separate that part of the lead which has been dissolved in
the state of sulphide by the alkali, but not decomposed ;
so that much less may be employed than is necessary for
the decomposition of the w r hole amount.
Experiment has shown that the maximum amount of
lead from pure galena may be obtained by the use of the
following mixtures :
2 parts of black flux, or sodium carbonate, and 10 to
12 of iron.
1 part of black flux, or sodium carbonate, and 20 of
iron
ROASTING AND REDUCING ASSAY. 515
\ a part of black flux, or sodium carbonate, and from
25 to 30 of iron.
When black flux is employed, and the iron is in the
state of filings, it would be inconvenient to employ too
much of the latter, especially if the assay were heated
very strongly, because the button of lead might be con-
taminated with iron ; but when sodium carbonate is used
with small iron nails instead of filings, the excess of iron
is not inconvenient, but rather useful, because the desul-
phurisation is certain to be complete.*
The following changes take place in both cases. That
portion of iron filings mixed with the sodium carbonate
which has not been sulphurised is brought to the state of
oxide by the carbonic acid of the alkaline carbonate, and
remains combined or neglected in the slag ; so that the
proportion of iron is never too great, and never becomes
mixed with the lead. When black flux is employed, the
same oxidation does not take place, on account of the
presence of carbonaceous matter, so that the portion of
filings not combined with sulphur, and which is merely
held in suspension in the flux, passes through it with the
globules of lead to the bottom of the crucible ; but if,
instead of filings, small nails are employed, they only suffer
corrosion at their surface, without change of form or soften-
ing, and after the assay are found fixed in the surface of
the button of lead, so that they can be detached very
readily, and, according to Berthier, without loss of lead.
This, however, we have found no easy task, and have
always sustained a notable loss.
5. ROASTING AND REDUCING ASSAY.
This mode is preferable for ores and substances which
contain a considerable quantity of foreign sulphides, or
arsenides and antimonides, and a greater or less amount of
* The French assayers use a piece of plate iron in the shape of a horse-
shoe, which is moved about in the melted mass until no more globules of lead
attach themselves to it.
In Germany a mass of iron wire is used. What iron is not consumed by
the assay is found still hanging together in a single mass.
L L 2
516 THE ASSAY OF LEAD.
earthy matter. It is used in many parts of Germany (at
the Unterhartz), and is described by Kerl thus :
Two assay centners of ore, matte, &c., are heated at first
at a low red heat in the muffle, on a roasting dish that has
been previously rubbed with chalk. After ten or fifteen
minutes they are taken out of the furnace, then again
roasted at a moderate temperature for ten or fifteen minutes
with frequent turning of the dish. The assay is then once
more taken from the furnace, allowed to cool, rubbed up
in an agate mortar, and again roasted for half an hour,
whereupon it is taken out of the furnace ; tallow is added
while it yet glows, and it is again brought to a strong red
heat. The rubbing up and calcining with tallow are
repeated several times more, and when afterwards the
assays have been exposed for two hours continuously to a
strong red heat, with the mouth of the muffle almost en-
tirely closed, if no more sulphurous acid vapours escape,
the roasting is considered as finished. This lasts from six
to twelve hours. The roasted sample is then portioned
out with the balance, each portion mixed with three or
four parts of black flux and an equal quantity of borax
and glass, placed in a small crucible covered with sodium
chloride, furnished with a little piece of coal as a cover,
and smelted in the wind furnace for about a quarter of an
hour after the fire is well ignited. Assays that have
worked well give nearly equal malleable buttons that do
not contain matte, and a black uniformly fused slag.
The purpose of the roasting is to convert the metallic
sulphides, arsenides, and antirnonides into oxides. But
since, in the process, sulphates, antimoniates, and arseniates
are produced, we seek to destroy these by repeated cal-
cining with tallow (see above), instead of an intermixture
of coal-dust or flour. By melting the roasted assay with
its charge at not too high a temperature, the lead oxide is
reduced, and the foreign oxides and earths contained in
the sample are, by the aid of the potash in the black flux,
as well as of the borax and glass, slagged off. If sulphates
or sulphides have remained behind in the roasted ore,
they will in the smelting be partially desulphurised by the
ASSAY WITH BLACK FLUX. 517
action of the oxides, especially of the iron oxide. An
addition of metallic iron would in this respect be advan-
tageous.
The roasting is a lengthy process, and one which causes
a not unimportant loss of lead. If it is not done tho-
roughly, then in the reduction smelting, sulphur salts
are formed, which always retain lead, as also a plumbi-
ferous matte which surrounds the lead button. By the
use of too high a temperature in the smelting a great part
of the foreign oxides is reduced, and the lead becomes
contaminated. The reduction, however, cannot be entirely
avoided, even with a rightly conducted temperature.
Galena melts less easily than metallic lead if the air
is excluded ; but is much more volatile than the latter,
and is decomposed by fusion into a higher sulphide
which is volatile, and a lower one (Pb 2 S) which remains
as a residue. Galena by roasting gives a mixture of lead
oxide and sulphate, from which last the sulphuric acid
cannot be separated, even at a fusing temperature. Lead
sulphate becomes soft by heat, fuses at a bright white
heat, and is converted by carbon, with a considerable loss
of lead through volatilisation, into lead oxide, metallic lead,
or lead subsulphide, acording to the quantity of carbon
used and the temperature employed. With lead oxide
the sulphate easily fuses together.
ADDITIONAL REMARKS ON THE LEAD ASSAY.
Comparison of the Different Methods for the Docimastic Esti-
mation of Lead in their Application to Various Products.
Markus has made the following comparative experi-
ments with the methods of assaying lead ores most in use
at the Austrian smelting works at Joachimsthal.
a. Assay with Slack Flux and Iron. One assay centner
(5*7 grammes) of the finely rubbed, sifted, and dried assay
substance was mixed with two assay centners of black flux,
made of sixteen saltpetre and forty argol, and sixty
of borax-glass in a mixing capsule, and put into a clay
crucible, on the bottom of which a piece of thick iron wire,
518 THE ASSAY OP LEAD.
one inch long and forty centners in weight, had been placed
in a vertical position. The crucible charge, covered over
with two centners of decrepitated sodium chloride, was
smelted in a mineral coal muffle furnace, with the mouth
of the muffle closed, and the draught half open, at a mode-
rate temperature, the temperature then lowered for six to
seven minutes by opening the mouth of the muffle, then
the muffle closed again for an equal period, and the final
heat then given. The cessation of the low crackling of
the assay was now carefully attended to, and this, ceasing
after seven to eight minutes, indicated the completion
of the assay, The duration of the assay was twenty
minutes.
b. Roasting and Reduction Assay with Iron. One assay
centner of galena was roasted, at first at a low tempera-
ture, for about thirty minutes on a roasting dish, and the
dish then pushed into the back part of the muffle for six
to eight minutes to destroy the sulphates formed. The
roasted ore was rubbed fine, intimately mixed with three
hundred centners of black flux, and fifty centners of borax-
glass, placed in a crucible with a piece of iron at the
bottom, weighing twenty centners covered with salt, and
smelted as above.
c. Roasting and Fusing with Black Flux. One centner
of the roasted ore was smelted as before with three hun-
dred centners of black flux and fifty centners of borax,
but without iron.
The results obtained proved
1 . That with all those products which contain tolerably
pure lead sulphide, especially with high percentages, the
iron assay, a, gives in a remarkably predominant degree
the most lead (as high as 96 per cent, of all the lead
present).
2. With impure lead ores, which contain more foreign
sulphides, the assay a gives likewise the highest per-
centage, though the assays b and c give only a few per
cent. less.
3. If foreign sulphides are present in predominant
ASSAY WITH SULPHURIC ACID. 519
quantity, the methods of b and c give a slightly higher
percentage than that of a.
LevoVs Fusion Assay with Potassium Ferrocyanide and
Cyanide.
According to Levol, the method of assaying galena for
its lead by smelting it with black flux and iron is defective
in two respects. First, it is difficult to choose precisely
the quantity of iron required for the reduction of the lead,
and a lack or excess of it either gives too little lead or a
button containing iron ; and second, in order that the
reaction may be complete and the lead unite to a button,
we are compelled to use a very high temperature, at which
lead volatilises. The first defect can indeed be removed
by the use of iron crucibles, but these are easily rendered
unserviceable, and require a pouring out of the fused
mass, and then globules of lead are apt to remain in the
slag.
By the use of a mixture of fifty parts of potassium
cyanide and one hundred of anhydrous potassium ferro-
cyanide to one hundred of galena, the loss of lead
diminishes to from 2 to 2^ per cent., probably in con-
sequence of the easy fusibility of the mixture, and the
extremely fine division of the iron in the potassium ferro-
cyanide. With antimonial galena this process is not
applicable, as the antimony is reduced and goes into the
lead. Potassium cyanide alone, gives, by reason of the
greater quantity of metallic sulphide which it retains, a
smaller product of lead.
6. ASSAY WITH SULPHURIC ACID.
The assay sample is rubbed as fine as possible. A
suitable quantity of it is then weighed out for an assay,
and boiled with four to eight times its weight of oil of vitriol
until all is decomposed. All excess of sulphuric acid is
then evaporated in a porcelain capsule, under a flue with a
good draught, and the mass carried to dryness. Boiling
520 THE ASSAY OF LEAD.
sulphuric acid decomposes the sulphides, changing iron,
copper, nickel, zinc, &c., into salts which dissolve readily
in water, and also at the same time changing the lead
sulphide into sulphate, which in water, especially when
cold, and containing free sulphuric acid, is practically in-
soluble. The composition of the ore is in general ascer-
tained by first heating it with nitric acid or aqua regia,
and then with the addition of sulphuric acid, evaporating
to dryness. The dry mass, when cold, is moistened with a
small quantity of sulphuric acid, then cold water ; it is
afterwards, by the aid of a small brush, brought without
loss upon a small filter, and washed with cold water until
the filtrate is colourless. Unnecessary prolonging of the
washing is to be avoided, for lead sulphate is not absolutely
insoluble. The filter, with its contents, is dried in the
funnel, until it can be easily taken out of it without tear-
ing. It is now put immediately into the clay crucible in
which the lead sulphate is afterwards to be reduced, and
this is placed in a very gentle stove warmth. (Some
potassium carbonate may be first poured into the bottom
of the crucible.) When completely dry, the crucible with
the cover laid over it is very gently heated, so that the
filter carbonises, which very soon happens, as the free
sulphuric acid is not completely soaked out. The filter is
now stirred up with a little rod, black flux or potassium
carbonate with coal-dust and iron are introduced into the
crucible, and intimately mixed with the lead sulphate and
the rest of the insoluble residue. About four or five times
the volume of the whole residue is taken of black flux,
and the assay is further treated as prescribed in the portion
which follows upon the assaying of lead sulphate.
In this way the lead is concentrated, and the foreign
sulphides, which were specified above as the cause of the
failure of the assay in such cases, completely removed.
The result obtained in this way is satisfactory, and de-
serves the same confidence as one obtained in favourable
circumstances by the ordinary lead assay from an ore with
a medium or high percentage of lead.
ASSAY OF GALENA. 52J
7. ASSAY OF GALENA IN THE WET WAY.
When in contact with metallic zinc, galena is readily
decomposed by acids. Even oxalic, acetic, and dilute
sulphuric acids are capable, when hot, of decomposing
galena metallic lead being deposited and sulphuretted
hydrogen gas set free, while with hydrochloric acid the
decomposition is peculiarly rapid and complete.
Galena is easily decomposed, also, even in the cold by
dilute nitric acid in presence of zinc ; but the reaction
differs in this case from that just described not metallic
lead but free sulphur is deposited, while lead nitrate goes
into solution.
The reaction with zinc and hydrochloric acid has been
employed with advantage by Mr. F. H. Storer, Professor of
Chemistry in the Massachusetts Institute of Technology,
for assaying galena, particularly the common American
variety, which contains no other heavy metal besides lead.
The details of the process are as follows : Weigh out 2 or
3 grammes or more of the finely powdered galena. Place
the powder in a tall beaker, together with a smooth lump
of pure metallic zinc. Pour upon the mixed mineral and
metal 100 or 150 c.c. of dilute hydrochloric acid which
has been previously warmed to 40 or 50 C. ; cover the
beaker with a watch-glass or broad funnel, and put it in a
moderately warm place.
Hydrochloric acid fit for the purpose may be prepared
by diluting 1 volume of the ordinary commercial acid with
4 volumes of water. For the quantity of galena above
indicated, the lumps of zinc should be about an inch in
diameter by a quarter of an inch thick ; they may be
readily obtained by dropping melted zinc upon a smooth
surface of wood or metal.
The zinc and acids should be allowed to act upon the
mineral during fifteen or twenty minutes in order to insure
complete decomposition. Any particles of galena which
may be thrown up against the cover or sides of the beaker
should of course be washed back into the liquid. It is
522 THE ASSAY OF LEAD.
well, moreover, to stir the mixture from time to time with
a glass rod.
When all the galena has been decomposed, as may be
determined by the facts that the liquid has become clear,
and that no more sulphuretted hydrogen is evolved, decant
the liquid from the beaker into a tolerably large filter of
smooth paper, in which a small piece of metallic zinc has
been placed. Wash the lead and zinc in the beaker as
quickly as possible with hot water, by decantation, until
the liquid from the filter ceases to give an acid reaction
with litmus paper ; then transfer the lead from the beaker
to a weighed porcelain crucible. In order to remove any
portions of lead which adhere to the lump of zinc, the
latter may be rubbed gently with a glass rod, and after-
wards with the finger or a piece of caoutchouc, if need be.
Wash out the filter into an evaporating dish, remove the
fragment of zinc, and add the particles of lead thus collected
to the contents of the crucible. Finally, dry the lead at a
moderate heat in a current of ordinary illuminating gas,
and weigh.
The lead may be conveniently dried by placing the
crucible which contains it in a small cylindrical air-bath
of Eammelsberg's pattern, provided with inlet and outlet
tubes of glass, reaching almost to the bottom of the bath.
When the process is conducted as above described,
the lead undergoes no oxidation ; hence there is no occa-
sion for igniting the precipitate in a reducing gas. The
precipitate needs only to be dried out of contact with
the air.
If desirable, the sulphur in the galena can be esti-
mated at the same time as the lead, by arresting the
sulphuretted hydrogen in the ordinary way.
If the mineral to be analysed is contaminated with a
silicious or other insoluble gangue, the metallic lead may
be dissolved in dilute nitric acid after weighing, and the
insoluble impurity collected and weighed by itself. In the
case of galenas which contain silver, antimony, copper, or
other metals, precipitable by zinc, the proportion of each
metal must be estimated by assay or analysis in the usual
ASSAY OF SUBSTANCES OF THE SECOND CLASS. 523
way, after the total weight of the precipitated metals has
been taken.
Besides galena, almost any of the ordinary lead com-
pounds may evidently be assayed by the method above
described. For example, metallic lead may be precipitated
quickly and completely from the sulphate, chromate,
nitrate, oxide, and carbonate and with peculiar ease from
the chloride by means of zinc and hydrochloric acid. The
method would also furnish an easy qualitative test for
the detection of baryta in white-lead. When applied
to the analysis of lead nitrate, it would probably be
best to decompose the nitrate by means of a solution of
sodium chloride before adding the zinc and hydrochloric
acid.
In all these cases the decomposition of the lead salt
by the zinc is so complete that no trace of coloration is
produced when sulphuretted hydrogen is added to the
liquid decanted from the metallic lead.
Attempts to estimate sulphur and lead in the same
portion of galena, by means of the reaction of zinc and
dilute nitric acid above described, have not given satisfac-
tory results. The free sulphur obtained by treating galena
with zinc and ordinary nitric acid, diluted with three,
four, and five volumes of water, always retains a small
quantity of lead, while a certain amount of sulphuric acid
is found in the clear liquid. It is, in short, well-nigh or
quite impossible to avoid the secondary reactions between
zinc and lead nitrate, and between sulphuric and nitric
acid, which set in as soon as, or just before, the last traces
of the galena have been decomposed.
CLASS II.
Assay of Substances of the Second Class. The assay of
these substances is very simple indeed. Litharge, minium,
lead carbonate, &c., may be assayed by simple fusion with
carbonaceous matter ; but when the operation is thus
conducted, loss of lead is sustained : it is therefore better
524 THE ASSAY OF LEAD.
to add some flux which will readily fuse, and allow the
globules of reduced lead to collect into one button. No
flux fulfils this condition better than a mixture of sodium
carbonate and argol, which is to be intimately mixed with
the assay. The following is the best mode of procedure :
To 200 grains of the finely pulverised substance add 100
grains of argol, and 300 of sodium carbonate, and inti-
mately mix : place the mixture in a crucible which it
about half fills, and cover with a layer of common salt
about J inch thick ; submit the crucible to a very gradu-
ally increasing temperature, keeping the heat at low
redness for about a quarter of an hour ; then urging it to
bright red until the contents of the crucible flow freely ;
take it from the fire and shake, tap it as directed in the
copper assay, and either pour the contents into the mould
or allow to cool in the crucible. If the operator be
pressed for time, the mould may be used, but it is recom-
mended to allow the assay to cool in the crucible, for
unless the operator be very careful, and have had some
considerable practice, he is very liable to lose a small
.quantity of metal in the pouring. After the contents of
the mould or crucible, as the case may be, are cold, the
lead may be separated from the slag by repeated gentle
blows from the hammer : if any of the slag or crucible
adhere to the button, the latter may be readily freed from
it by placing the button between the finger and thumb
with its edge on the anvil, and then gently hammering it.
The lead will be so altered in shape under the hammer
that the slag or crucible readily falls off; and by continu-
ing the process the whole may be removed. The cleaned
button may then be hammered into a cubical form, and
is ready for weighing.
In the assay of lead great care must be taken in the
management of the temperature, as lead is sensibly volatile
above a bright red heat, even when covered with flux,
and still more so if any portion be uncovered from want
of sufficient quantity of flux ; neither must the assay
remain in after the flux flows freely, for a loss may thereby
occur from oxidation, by decomposition of sodium carbo-
ASSAY OF SUBSTANCES OF THE SECOND CLASS. 525
nate, as explained in the reduction of copper ores and the
copper-refining process.
For the rationale of this mode of assay, refer to page
199, which explains the decomposition of lead oxide, with
the production of metallic ]ead, carbonic acid, and water,
by the agency of a substance, like argol, containing both
carbon and hydrogen.
Cupel bottoms, some lead fumes, and silicious slags
require a modified treatment in the assay, as the sub-
stances mixed with the oxide of lead (more particularly
bone-ash in the cupel bottoms) are very infusible ; and if
the flux already mentioned as applicable to the other
matters belonging to this class were employed, a very high
temperature would be necessary ; and as lead, as already
stated, is sensibly volatile above a bright red heat, an
evident loss of that metal would be the result.
Cupel bottoms may be thus assayed : 400 grains of the
finely pulverised bottoms to be mixed with 200 grains of
argol, 400 grains of sodium carbonate, and 400 grains of
pulverised fused borax ; the mixture is placed in a crucible
as already directed, covered with salt, and the fusion con-
ducted as just described.
Lead fumes and silicious slags require only half their
weight of fused borax, with 200 argol, 400 sodium car-
bonate, and 400 substance (fume or slag), covered with
salt.
The addition of the borax, which is a most powerful
flux, causes the fusion of the assay to take place almost as
readily with the last-named refractory substances as with
the former easily fusible and reducible matters. The
assay, however, is rather more subject to ebullition or
boiling over the sides of the crucible ; hence it must be
carefully watched ; and the instant it appears likely to do
so the crucible must be removed from the fire, gently
tapped on the furnace top, and when the effervescence
has subsided returned to the furnace, and this operation
repeated until the fusion proceeds tranquilly.
The lead obtained in these assays, if the ore or sub-
stance contained any foreign metal, is never pure ; if silver,
526 THE ASSAY OF LEAD.
copper, tin, or antimony be present, the whole of either
of these metals will be found alloyed with the lead pro-
duced ; but if the ore contains zinc, and it be heated
sufficiently, only a trace remains ; nevertheless the zinc
carries off with it a considerable quantity of lead.
The following experiments will show what an influence
the presence of zinc has upon the return of lead :
100 parts of litharge,
100 parts of zinc oxide,
300 parts of black flux,
were fused together, and 84 parts of lead were the result.
100 parts of litharge,
100 parts of zinc oxide,
600 parts of black flux,
were fused together, and but 70 parts of lead were pro-
duced instead of 90, which the pure litharge ought to
have given. Hence it will be seen that the more zinc is
reduced, the more lead is volatilised.
If iron oxide be present in the assay, it is reduced, but
it remains in suspension in the slag, and the lead does not
contain a trace when it has not been too strongly heated.
If the assay be made at a very high temperature, the iron
may be fused, and then the lead will be ferruginous ; this
may be ascertained by means of the magnet. A similar
result was obtained by many assayers, who thought for a
long time that lead and iron could thus combine together ;
but by careful examination it is easily ascertained that the
ferruginous buttons are but mechanical mixtures of lead
and iron in grains. Indeed, by careful hammering, nearly
all the iron may be removed from the lead, so that it loses
its magnetic properties.
The manganese oxides, when mixed with the ore, are
changed into protoxide, which remains in the flux, and is
not reduced.
Wet Assay of Ores of the Second Class. Pulverise the
substance very finely, and to 100 grains placed in a flask
add one ounce of nitric acid diluted with two ounces
of water (if minium be the substance to be analysed, it
ASSAY OF SUBSTANCES OF THE THIRD CLASS. 527
must be first heated to redness, so as to reduce the whole
of the lead it contains to the state of protoxide), and gently
heat, gradually raising the temperature to the boiling-
point : when all action seems to have ceased, pour the
contents of the flask into an evaporating basin, and evapo-
rate to dryness with the precautions directed in the analysis
of iron ore. Allow the dry mass to cool, add a little dilute
nitric acid, gently warm for an hour, then add water, boil
and filter. The whole of the lead now exists in the solu-
tion as nitrate : thus, say lead carbonate has been the
substance under analysis, then
PbO,C0 2 + N0 5 =PbO,N0 5 + C0 2 .
To the filtered solution containing the nitrate as above,
add solution of sodium sulphate, or dilute sulphuric acid,
until no further precipitation takes place ; insoluble sul-
phate of lead will now be thrown down : this must be
allowed to completely subside by standing in a warm place ;
and when the supernatant liquid is quite bright, the sul-
phate may be collected on a filter, washed, dried in the
water-bath, and weighed. It contains 68-28 per cent, of
metallic lead.
The decomposition of the lead nitrate by sodium sul-
phate may be thus expressed :
PbO,N0 5 + JSTa 2 0,S0 3 =PbO,S0 3 + Na 2 0,N0 5 .
Estimation of lead by standard solution will be
described at the end of this chapter.
CLASS III.
Assay of Substances of the Third Class. In the assay
of bodies belonging to this class a reducing agent must be
employed ; but if that alone be used, the sulphates and
arseniates produce sulphides and arsenides, and not pure
lead. The action of another reagent is therefore necessary,
in order to deprive the lead of the sulphur and arsenic
with which it is combined.
528 THE ASSAY OF LEAD.
There are two reagents employed for the sulphates
the alkaline carbonates and metallic iron ; but for the
arseniates and arsenites iron must be employed, because
the alkaline carbonates have no action on the arsenides.
In all cases black flux is employed ; this furnishes a
reducing agent for the oxides, and a flux for the earthy
matters. Iron is added when the arsenites or arseniates
are assayed ; but that metal may either be employed or
not when the sulphates are operated upon. It is, how-
ever, always better to use it.
When a mixture of black flux and iron is employed,
the assay is made in exactly the same manner as that of
the sulphides (large nails are preferable whenever the use
of iron is indicated in a lead assay). With the sulphate,
the iron sulphide formed combines in the slag with the
alkaline sulphide ; but it is not so with the arseniates and
arsenites. The arsenide produced mixes neither with the
lead nor the slag, but gives rise to the formation of a brittle
matte which adheres slightly to the button of lead.
When only black flux is employed, either of the two
following processes may be adopted : First, the ore can
be fused with four parts of common black flux ; then, as
in the case of sulphides, the excess of carbon determines
the formation of a large quantity of an alkaline sulphide,
and consequently produces a desulphurisation of the lead.
Secondly, it may be fused with such a proportion of black
flux, containing only the requisite proportion of carbon to
reduce the lead oxide, or with an equivalent mixture of
sodium carbonate and charcoal. Pure lead sulphate fused
with one part of sodium carbonate and 4 per cent, of char-
coal gives 66 of lead ; but in order to employ this method
the richness of the ore must be known, and the dry way
is then useless, excepting for the estimation of the silver
these substances always contain.
Wet Assay of Substances of the Third Class. These
are treated in precisely the same manner as those of the
preceding class.
In treating lead ores with nitric acid a loss generally
results from the formation of insoluble lead sulphate. The
ASSAY OF SUBSTANCES OF THE THIED CLASS. 529
solubility of these salts in sodium hyposulphite renders it
possible to avoid this inconvenience. After treatment with
nitric acid, J. Grceme (Bulletin de la Soc. Chimique de
Paris, November 5, 1873) proposes to exhaust the residue
with boiling water, until the soluble salts and the acid are
completely eliminated. It is then digested in the cold with
a concentrated solution of sodium hyposulphite. After this
treatment has been twice or thrice repeated the residue
is exhausted again with water ; the lead is then precipitated
from the filtrate by sulphuretted hydrogen or ammonium
sulphide ; to facilitate the agglomeration of the precipitate
and its washing, it is heated in the water-bath. The sul-
phide is then converted into sulphate, and its weight added
to that of the sulphate obtained directly.
F. Maxwell Lyte thinks the following process for the
management of the assay of lead in ores will be found
convenient, particularly where, as is often the case, the
lead to be estimated is mixed as sulphate with the matrix
insoluble in acid.
He dissolves the sulphate or chloride, as the case may
be, in ammonium acetate, makes the solution as neutral as
possible, and estimates the lead by a standard solution of
bichromate (a half-decinormal solution answers well) with
a silver nitrate indicator.
A. Mascazzini, previously to reducing the galena or
other lead ore to the metallic state, converts the lead
present in the ore into sulphate by igniting it in a porcelain
crucible with ammonium sulphate, after which the ore is
treated in the usual manner, The flux preferred by the
author is that recommended by Plattner, consisting of
13 parts of potassium carbonate, 10 of dry sodium car-
bonate, 5 of previously fused borax, and 5 of well-dried
starch.
To detect galena in mixtures, M. E. Jannesay throws
upon coarsely powdered galena a fragment of potassium
bisulphate, which gives a distinct evolution of sulphuretted
hydrogen. If the two bodies are ground , together the
odour becomes almost insupportable. Potassium bisul-
phate kept in fusion for half an hour produces the same
M M
530 THE ASSAY OF LEAD.
effect, perhaps with less intensity. Sulphuric acid, mixed
or even heated with galena, does not give rise to a sen-
sible disengagement of sulphuretted hydrogen. Blende
gives a sulphydric odour, but less intense. Antimony,
iron, mercury, and silver sulphides give off no sensible
odour. Boulangerite, zinbrenite, and in general the sul-
phides in which lead and sulphur do not form an isolated
combination, do not evolve sulphuretted hydrogen with
the potassium bisulphate.
CLASS IV.
ALLOYS OF LEAD.
ASSAY WITH SULPHURIC ACID.
No docimastic assay is known for exhibiting the lead
isolated from its alloys. In individval cases a serviceable
result may be attained, if the metal with which the lead is
combined be estimated, and its quantity then deducted.
This method is, however, in general the more unreliable
the smaller is the quantity of lead, or when the lead is
alloyed with several metals ; so that then the quantity of
lead can often only be estimated by the partial or com-
plete aid of the wet way.
For many products (e.g. crude lead, hard lead con-
taining antimony or arsenic plumbiferous copper, &c.)
the assay with sulphuric acid described on page 519 is
suitable. One assay centner of the substance is decom-
posed by nitric acid or aqua regia, then, with the addition
of sulphuric acid, evaporated to dryness, and the dry
mass treated as above directed. If the residue consists
only of lead sulphate, it can be brought upon a weighed
filter, and from the weight of the residue after drying,
the amount of lead may be calculated. 100 parts lead
sulphate contain 68-33 parts lead.
DOCIMASTIC ESTIMATION OF LEAD. 531
ESTIMATION OF LEAD BY MEANS OF STANDARD SOLUTIONS.
1. FLORES DUMONTE'S METHOD.
Estimation of Lead by means of Standard Solutions.
This process is due to M. Flores Dumonte, and may be
thus described : This mode of analysis is analogous to
that proposed by Pelouze for the estimation of copper ;
advantage is taken of the fact that oxide of lead is soluble
in caustic potash in the same manner that copper oxide is
soluble in ammonia ; and from either solution the respective
metal is precipitated by means of a standard solution of
sodium sulphide.
The solution of sodium sulphide may be conveniently
made by dissolving one ounce of sodium sulphide in one
quart of water, and estimating how much of it is necessary
to precipitate twenty grains of lead. To this end weigh
off twenty grains of lead, dissolve them in nitric acid, di-
lute with water, and add excess of caustic potash until the
oxide of lead first thrown down is completely dissolved.
The solution must now be heated to ebullition, and the
sodium sulphide gradually added from the burette ; at
each addition a black precipitate of lead sulphide falls.
The liquid is then boiled for a short time, by which means .
it brightens ; more sodium sulphide is then added, and the
whole again boiled, and these operations alternately con-
tinued until no further coloration or blackening is pro-
duced by the last drop of sulphide. The number of
divisions used is then read off, and the calculation made
as in Pelouze's copper assay, substituting lead for copper.
Having thus standardised the solution of sodium sul-
phide, the assay of a sample of ore may be thus made :
If the ore belong to the first class, dissolve it in dilute
nitric acid and evaporate to dryness ; to the dry mass add
excess of caustic potash solution, and boil ; after about
a quarter of an hour's ebullition, filter and throw down
the lead as directed, with the standard solution ; from the
amount used calculate the quantity of lead present ; if the
ore be of the second or third class, treat with strong nitric
M M 2
53-2 THE ASSAY OF LEAD.
acid and sodium carbonate, as already directed. The lead
carbonate so produced may be dissolved in either nitric or
acetic acid, and to the solution thus obtained add caustic
potash, &c.
2. SCHWARTZ'S METHOD.
Dissolve 14*730 grammes of pure potassium bichromate
in sufficient water to form one litre. One cubic centimetre
of this solution precipitates 0-0207 gramme of lead.
In the estimation of pure lead a certain quantity of
it should be dissolved in a minimum of nitric acid, the
solution diluted with water, carefully neutralised with
ammonia or sodium carbonate, excess of sodium acetate
added, and the solution precipitated by the potassium
bichromate solution. When the precipitation approaches
its end, or when the precipitate commences readily to
subside, some drops of a neutral solution of silver nitrate
are deposited on a porcelain plate, and the potassium
chromate solution only added by two or three drops at a
time to the liquid under examination ; after each addition
the whole is well stirred, allowed to subside, and a drop of
the clear supernatant liquor added to one of the drops of
the silver solution. As soon as the potassium bichromate
is in excess the two drops form a red colour, while the
precipitated lead chromate has no effect on the silver test,
but simply floats on the top as a yellow precipitate.
Should the solution assume a yellow colour before the
silver reaction has commenced, it would indicate that not
sufficient sodium acetate had been added in the first in-
stance, and it w r ould be necessary to add this now, and
also a cubic centimetre of a normal lead solution, contain-
ing 0-0207 of lead as nitrate. The slight turbidity which
first takes place soon goes off, and the operation may be
proceeded with as before. One cubic centimetre must
naturally, in such instance, be deducted from the amount
of chrome solution, on account of the extra addition of
lead.
Bismuth alone seems to interfere with the reaction,
BUISSON S VOLUMETRIC PROCESS. 533
.and behaves very like lead with chromic acid, and if present
it requires a different mode of proceeding.
The higher oxide of mercury is not precipitated by
potassium bichromate, not even in an acetic solution, while
the lower oxide is ; and, as it is difficult to peroxidise all
the mercury when united with lead, even by long-continued
boiling in nitric acid, it becomes necessary to evaporate and
calcine the residue till all the mercury is volatilised. To
obviate the formation of red-lead, the calcined residue has
to be moistened with a few drops of oxalic acid, and again
dried and carefully calcined and dissolved in acetic acid ;
after this the lead may be estimated as usual. To avoid
the above calcinations, the mercury may be precipitated
from the nitric acid solution by means of hydrochloric acid,
and the liquid boiled till the calomel is converted into the
higher chloride.
Copper, cadmium, zinc, iron, and cobalt do not in the
least interfere with the reaction, provided the iron is per-
oxidised. Of the different acids, hydrochloric acid some-
what disturbs the last silver reaction, but by using larger
drops, and allowing the reaction of silver chloride to go
off, we obtain the usual silver chromate reaction.
Lead sulphate has first to be converted into the state of
carbonate, by boiling with sodium carbonate, when it may
be dissolved in acetic acid. Lead phosphate and arsenite,
-or other lead salts insoluble in acetic acid, may be dis-
solved in nitric acid, and estimated according to the older
.method.
.<
o. BUISSON'S VOLUMETRIC PROCESS FOR ESTIMATING LEAD.
This process is based on the precipitation of lead by
potassium bichromate and the decomposition of the excess
of the reagent used by potassium iodide in a liquid
acidulated with sulphuric acid. 0'5 to 1 gramme of the
pulverised mineral to be tested is dissolved in dilute nitric
.acid. The solution is saturated with potash, and the preci-
pitate redissolved in acetic acid. The iron is removed by
boiling. To the solution separated from the iron 25 c.c.
534 THE ASSAY OF LEAD.
of potassium bichromate is added, and water to bring the
volume up to 250 c.c. After standing for some time the
solution is filtered through a dry filter-paper. To 100 c.c.
of .the clear liquid an excess of dilute sulphuric acid and
potassium iodide are added in such a way as to redissolve
the iodine set free, and then a few cubic centimetres of
solutions starch. By means of a graduated burette, sodium
hyposulphite is added until decolourisation of the starch
iodide takes place. The difference of standard obtained by
treating the bichromate alone, and after precipitation with
a known weight of lead, gives a basis for calculating the
amount of metal contained in the substance tested.
Silver, bismuth, copper, and baryta should be separated
from the lead before applying this process.
4. W. DIEHL'S PKOCESS.
W. Diehl proposes the following process for the
volumetrical estimation of lead ( 4 Zeitschrift fur Analy-
tische Chemie '). He employs a ^ normal solution of
potassium bichromate, containing 7 '38 grnis. per litre,
each c.c. representing 0*01035 grin, of lead, and a solution
of sodium hyposulphite, containing 4 to 5 grms. per litre.
To determine the relation between these two solutions,
20 to 30 c.c. water are mixed with 20 to 25 dilute sul-
phuric acid (1 vol. monohydrated acid, and 2 vols. of
water) ; a certain excess of sulphuric acid is indispensable,
hydrochloric acid being less convenient. The liquid is
brought to a boil, and the solution of hyposulphite is
added drop by drop. The solution becomes gradually
paler in colour. Towards the end, after the addition of a
few drops, it is let boil up again. The end of the reaction
may generally be distinguished by the liquid turning per-
fectly colourless, a result occasioned by a single drop. In
order to judge of the colour, the flask towards the end of
the operation may be set in a porcelain capsule. When
very large quantities of bichromate are used the liquid
does not become perfectly colourless, but slightly greenish.
In assaying ores in this manner they are dissolved
DIEHL'S PKOCESS. 535
in aqua regia and dilute sulphuric acid, the solution con-
centrated till the sulphuric acid begins to evaporate,
diluted with water, boiled to dissolve ferric sulphate, let
cool, and filtered through a smooth filter, washing with
water containing sulphuric acid. To the residue in the
flask as little as possible of which is thrown upon the
filter is added about 15 c.c. of a solution of neutral am-
monium acetate, and about 50 c.c. of water. The whole is
then boiled, and filtered through the same filter, into which
a drop of ammonia has been put, into a flask. The same
operation is then repeated with 5 c.c. ammonium acetate,
and the residue is finally well washed with boiling water,
to which a little of the same salt has been added. Thorough
washing is necessary, since filters have been found to
retain ammonium and lead acetate and tartrate with con-
siderable obstinacy. It is then advisable further to wash
the filter from its margin downwards with a little boiling
dilute hydrochloric acid (1 part hydrochloric acid at sp.
gr. 1-12 with 10 parts of water), and then to wash again
with hot water. In this manner every trace of lead is re-
moved from the filter. A thin filter-paper should be used,
and should be washed previously. The solution of lead
sulphate in ammonium tartrate thus obtained is then
titrated in the cold with potassium bichromate ; with the
aid of heat, ammonium acetate dissolves a little lead
chromate. The precipitate settles readily if the flask is
shaken, and the end of the reaction can be observed to
within 0-2 to 0*4 c.c.
An excess of at least 2 c.c. of the chromate solution
should be added, in order to obviate the solubility of the
lead salt. It is advisable in every experiment to take as
closely as possible an equal quantity. After thoroughly
shaking, it is allowed to stand for half an hour and filtered.
If the liquid passes through turbid, a few drops of a
solution of sodium acetate, acidulated with acetic acid, are
added. If, after all, a little lead chromate passes through
the filter, the filtration is repeated. The precipitate is
washed four times with cold water, and the solution is
acidulated with sulphuric acid and titrated as above.
636 THE ASSSAY OF LEAD.
Ammonium acetate is preferable to all other ammonium
salts as a solvent for lead sulphate. It is applied in a
neutral or faintly acid state. Free ammonia renders the
solution turbid. 1 grm. lead sulphate requires 15 c.c.
of the liquid acetate for solution. Ammonium tartrate
cannot be used.
537
GHAPTEE XII.
THE ASSAY OF TIN.
THIS metal is always found by the assayer in the state of
oxide.
Tin Oxide (Sn0 2 ). The appearance of this mineral
gives no indication, excepting to an experienced eye, that
metallic matter enters largely into its composition ; yet its
great density would lead one to suppose such to be the case.
Its colour varies from limpid yellowish white to brownish
black and opaque, passing from one to the other by all
intermediate shades. It usually possesses a peculiar kind of
lustre which cannot be readily described, but once seen can
scarcely be mistaken. It occurs crystallised in square prisms,
terminated by more or less complicated pyramids. These
crystals, derived from the octahedron, are often macled or
hemi tropic, so that they often possess re-entering angles,
which is to a certain extent characteristic. The principal
varieties are the following :
1. Crystallised Tin Oxide is found in more or less
voluminous crystals of the colour and form as above.
2. Disseminated Tin Oxide. This variety occurs in
grains of various sizes, sometimes so small as not to be
visible to the naked eye. It is found in the primitive
rocks.
3. Sandy Tin Oxide forms pulverulent masses often
of great extent ; in appearance it is merely a brown
sand.
4. Concretionary Tin Oxide, Wood Tin. This variety
occurs in small mamellated masses, the fibrous texture of
which resembles that of wood ; hence the name.
538 THE ASSAY OF TIN.
The following is an analysis of a sample of tin oxide
from Cornwall :
Tin 77-50
Oxygen 21-40
Iron -25
Silica . -75
The following remarks on tin ore and the minerals
which may be mistaken for it are from the pen of Dr. A.
Leibius, Senior Assayer of the Sydney branch of the Eoyal
Mint.
The colour of native tin ore varies from white to pink,
ruby-red, grey, greyish-black to black ; it therefore is cer-
tainly no very reliable criterion for distinguishing tin ore.
A safer characteristic is the weight, or specific gravity,
that of tin ore being 6-8 to 7*0. Unfortunately, however,
the specific gravity of iron tungstate is nearly the same
as that of tin oxide in fact, a little higher, being 7*19 to
7-55.
Titaniferou.s iron has a specific gravity of from 4 f 5 to
5-0, and magnetic iron 4-9 to 5*2.
Basaltic hornblende and iron silicate have also been
mistaken for tin ore ; but the specific gravity of the former
being only 3-1 to 3-4, and that of the latter 3-8 to 4*2,
ought to have prevented such mistakes. The colour of the
powdered ore forms a much better criterion than that of
the unpowdered ore. The powder of good tin ore varies
only from whitish-grey to dark drab, while iron tungstate
powders reddish-brown, and titaniferous iron black.
Most of the minerals appear to be mistaken for tin ore
on account of their dark granular pieces having been taken
for such ; but Mr. Leibius has seen a sample which con-
sists of blackish pieces with about 50 per cent, of small
indistinct crystals of a pink and dark ruby colour, with a
few small white crystals. The whole mixture being pretty
heavy has certainly at first sight all the appearance of
good tin ore. Even on closer inspection, when the darker
portion just referred to might have been recognised as an
iron compound, the ruby-coloured portion might readily
pass muster for tin ore unless chemically examined. On
further examination the whole sample was found to be
ASSAY OF PURE OXIDE OF TIN. 539>
free from tin. It was found to consist 1. Black portion,
about 50 per cent, of the sample, having a specific gravity
of 4-47, was found to be titaniferous iron. 2. Buby-coloured
and dark red pieces, about 50 per cent, of sample, with a
specific gravity of 4*57, were found to be zircons or hya-
cinths, showing the characteristic property, mentioned
in Professor Thompson's excellent ' Guide to Mineral Ex-
plorers ' (see ante, pp. 254,293), of becoming completely
and lastingly colourless when exposed to heat before the
blowpipe. 3. Besides these zircons were found a few small
topazes and garnets, and also a small sapphire. The
specific gravity of the mixed sample as received was
4-55.
As already mentioned, there was actually no tin in the
sample, and it forcibly illustrates the necessity of the
precaution, in dealing with tin ore, to have it carefully
assayed.
Assay of Pure Tin Oxide. Pure tin oxide may be
very readily assayed in the following manner : Weigh off
400 grains, place them in either a black-lead or charcoal-
lined crucible, cement on a cover by means of Stourbridge
clay, and place in the fire. The heat should for the first
quarter of an hour be a dull red, after which it may be
raised to a full bright red for ten minutes, and the crucible
removed with care so as not to agitate or disturb the con-
tents ; tapping in this case must not be resorted to. When
the crucible is cold, remove the cover, and a button of
pure tin will result ; this weighed and divided by four
gives the percentage. If the operation has not. been care-
fully conducted it sometimes happens the tin is not in one
button, but disseminated in globules either on the charcoal
lining or on the sides of the black-lead pot ; in this case
the charcoal on the one hand, or the black-lead crucible
on the other, must be pulverised in the mortar and passed
through a sieve ; the flattened particles of tin will be re-
tained by the sieve, and can be collected and weighed. If
any small particles escape the sieve, they may be separated
from the lining or crucible by vanning.
If a charcoal or black-lead crucible be not at hand, an
-540 THE ASSAY OF TIN.
ordinary clay pot may be used, but not so successfully,
excepting under certain circumstances to be hereafter
described. Indeed, in Cornwall the ordinary mode of con-
ducting this assay is in a naked crucible, thus : About 2
ounces of the ore are mixed with a small quantity ol
culm, and projected into a red-hot crucible. If the ore
seems to fuse or work sluggishly, a little fluor-spar is
added, and after about a quarter of an hour's fusing at a
good high temperature the reduced and fused tin is poured
into a small ingot mould, and the slag examined for metal
by pounding and vanning. This method never gives the
whole of the metal. To effect this, without fear of mis-
chance in the assay sometimes occurring, as already de-
scribed with both black-lead and charcoal-lined crucibles,
it may be thus conducted ; always supposing the oxide
to be pure, or nearly so, or at least containing little 'or no
silicious matter.
To 400 grains of ore add 100 grains of argol, 300 grains
of sodium carbonate, and 50 grains of lime ; mix well
together, place in a crucible which the mixture half fills,
cover with a small quantity of sodium carbonate and 200
grains of borax. Place the whole in the furnace with the
necessary precautions, raise the heat very gently, and keep
it at or below a dull red heat for at least twenty minutes ;
then gradually increase until the whole flows freely. Ee-
move the crucible, tap it as for copper assay, and allow to
cool. When cold, break it, and a button of pure metallic
tin will be found at the bottom, and a flux perfectly free
from globules and containing no tin.
There is. yet another process, which is more easy of
execution ; but the reagent employed is more expensive,
not so readily obtainable, and more difficult to keep without
decomposing than any of the substances above employed.
This reagent has been mentioned, in another part of this
volume, as a blowpipe flux, and, in the assay of copper
ores by standard solutions, as potassium cyanide. This is
the most effective reducing flux for tin ores yet known. It
acts by absorbing oxygen to form a compound known
as potassium cyanate.
ASSAY OF PURE OXIDE OF TIN. 541
The assay, by means of this substance, may be made
in ten minutes.
This method of estimating the value of tinstone has
been frequently practised by the writer, and has uniformly
furnished correct results with but little expenditure of
time and labour. The method of operating is as fol-
lows : The sample, having been carefully selected, is first
crushed by the hammer in a steel mortar, and then
further reduced to powder in an agate mortar. 100 grains
is a convenient quantity to be taken for analysis, and it
is always advisable to make two independent experiments
upon the same sample of ore, with the view of having a
control, and the highest result obtained is that upon which
to place reliance, since the error must always be on the
side of loss rather than excess. A couple of small Hessian
crucibles, of about 3 oz. capacity, are prepared in the first
instance by ramming into the bottom of them a small
charge of powdered potassium cyanide sufficient to form
a layer of about half an inch in depth ; the weighed quan-
tities of tin ore are then intimately mixed with from four
to five times their weight of the powdered cyanide, and
the mortar rinsed with a smull quantity of the pure flux,
which is laid upon the top of the mixture. The crucibles
are then heated in a moderate fire, or over a gas blowpipe,
and kept for the space of ten minutes at a steady fusion ;
they are then removed, gently tapped to facilitate the
formation of a single button, and allowed to cool. Upon
breaking the crucibles the reduced metal should present
an almost silvery lustre, with a clean upper layer of melted
flux. It is advisable to dissolve the latter in water, in order
to be certain as to the absence of any trace of reduced metal
or heavy particles of the original ore. There is always
contained in the commercial cyanide a sufficient quantity
of alkaline carbonate to secure the perfect fusion of the
silicious gangue, and other like impurities in the tin ore,
but the operator should assure himself of the absence of
copper and lead in the ore, either by preliminary treat-
ment with hydrochloric acid, in which tinstone is ab-
solutely insoluble, or by testing the button of reduced tin
542 THE ASSAY OF TIN.
after hammering or rolling for such metallic admixture.
We have usually found a minute trace of iron, and some-
times gold in the melted buttons, but not so much as to
add appreciably to their weight.
When worked with ordinary care, this process may be
relied upon as giving numbers true to within -J per cent.,
and we do not know any other method which exceeds this
in accuracy and rapidity of execution. The following are
a few analytical results taken at random from a number
of ores assayed in this manner :
Tin per cent.
L ~~n.
Sample No. 1 . . . . 45-6 45-8
No. 2 .... 57-2 57-6
No. 3 . . . . 68-4 68-7
Assay of Tin Oxide mixed with Silica. Although tin
oxide is completely reducible by charcoal or other carbo-
naceous matter, yet it has such an affinity for silica, that
whenever that substance is present the metal cannot be
wholly reduced, excepting at the highest temperature of a
wind furnace. The following experiments will show the
influence of silica on the return of tin in an assay of oxide
of that metal with black flux.
Ore . . 100 100 100 100 100
Quartz . 25 66 100 150 300
The first gave 52 per cent, of tin ; the second, 43 per
cent. ; the third, 28 per cent. ; the fourth, 10 per cent. ;
and the last, nothing.
The slags also produced in the treatment of tin ores in
the large way give no return with black flux. This mode
of assay, however, has been recommended by some, but
from the foregoing experiments is proved to be perfectly
fallacious : that is, unless the quantity of silica present be
very small in comparison to the amount of tin oxide ; and
even when the latter is present in four times the quantity
of the silica, as in experiment No. 1, a loss of 20 per cent,
of tin is sustained.
Assay of Tin Ores containing Silica and Tin Slags. It
having just been shown how injuriously the presence of
silica influences the produce of tin, both in ores and slags,
ASSAY OF TIN ORES CONTAINING SILICA. 543
other methods of assay than those just described must be
adopted for such substances. These will now be detailed.
Tin ores containing silica may be treated by two
methods : in the first the silica must be carefully separated
by vanning ; if the ore be well pulverised this is the best
and most expeditious method. In conducting this assay
take 400 or more grains of the pulverised ore, according
to its richness (if poor, as much as 2,000 grains may be
taken), van it carefully, dry the enriched product, which
will, if the operation has been properly conducted, be
nearly pure oxide of tin, and assay it as already described
for ores containing no silica. The other process of assay
may be thus conducted, and is dependent upon the fact that
iron displaces tin from its combination with silica : thus,
if a compound of tin oxide and silica be heated to white-
ness with metallic iron, a portion of the iron oxidises and
replaces the tin oxide, which was previously in combina-
tion with the silica as a tin silicate, and metallic tin and
iron silicate result, the tin so reduced combining with any
metallic iron that may be in excess, and the button thus
obtained is an alloy of tin and iron, whilst the slag is
entirely deprived of tin.
In this kind of assay mix 400 grains of the silicious
tin oxide with 200 grains of iron oxide (either pulverised
hematite or forge-scales will answer this purpose), 100
grains of pounded fluor-spar, and 100 grains of charcoal
powder : place the mixture in a crucible and cover with
a lid, gradually heat to dull redness, and keep at that
temperature for half an hour, then heat to whiteness for
another half-hour, and remove the crucible from the fur-
nace, allow to cool, and break. The button so obtained is
to be treated in the wet way, as hereafter described.
The assay of tin slags is conducted in the same manner,
or simply by mixing the pulverised slag with 20 per cent,
of iron filings, and fusing.
Assay of Tin Ores containing Arsenic, Sulphur -, and
Tungsten (Wolfram). In the assay of such ores it is neces-
sary to remove arsenic, sulphur, and tungsten before
attempting to obtain the tin in a pure state by the dry
544 THE ASSAY OF TIN.
assay. Ores of tin which contain either one or all of
these substances are most common ; hence this mode of
treatment will be generally required.
Most assayers usually submit the ore to the same mode
of treatment which it undergoes on the large scale by cal-
cination, or rather roasting, by which the greater part of
the arsenical and pyritic matter is removed ; this process
fails, however, to remove the whole of these substances,
and does not at all affect the tungsten. The following
process is therefore preferable, and is founded on the fact
that arsenical and other pyrites, as well as iron tungstate
(wolfram usually accompanying tin ores), are completely
decomposed by nitro-hydrochloric acid (aqua regia) at the
boiling temperature, the oxide of tin alone not being
affected : Take 400 grains or more of the impure tin
sample, place them in a flask, and add 1^ ounce of hydro-
chloric acid, and -J an ounce of nitric acid, heat gently for
about half an hour, and then boil until the greater part of
the mixed acids has evaporated ; the sulphur and arsenic
will by this time be converted into sulphuric and arsenic
acid, and the wolfram completely decomposed, its iron and
manganese having become soluble, and its tungstic acid
remaining in the insoluble state with the oxide of tin and
any silica that may be present. Allow the flasks and con-
tents to cool, add water, allow to settle, and decant, and so
on until the water passes off tasteless. The insoluble matter
in the flask is now tin oxide, silica, and tungstic acid ; to
remove the latter, digest for an hour at a very gentle heat
with one ounce of solution of caustic ammonia, with occa-
sional agitation ; add water, and van the remainder to
separate silica ; nothing remains now but tin oxide, with
perhaps a little silica : this is now to be dried and assayed
as directed for ores containing little or no silica.
If only an approximative assay be needed, it may be
accomplished after this treatment by taking the specific
gravity of the remaining oxide ; so that all ores of tin may
be thus roughly assayed, it being premised that the above
operation has been so carefully performed that nothing
but tin oxide and silica remain. The specific gravity of
ASSAY OF TIN OEES CONTAINING AESENIC, ETC.
545
the thus purified ore is to be taken. All now that is neces-
sary to be known is the specific gravity of tin oxide, its
percentage of pure tin, and the specific gravity of silica,
and a simple calculation gives the result. The following
is the formula :
Let a represent the specific gravity of tin oxide.
b silica.
c the mixture left after treatment
with acid, &c.
w weight of rough tin oxide or mixture left after
treatment with acid, &c.
x tin oxide.
y silica.
And y =
c (a b)
Or in arithmetical form thus :
1. From the specific gravity of the rough tin oxide (mixture of tin oxide
and silica) deduct the specific gravity of the silica.
2. Multiply the remainder by the specific gravity of the tin oxide.
3. Multiply the weight of the rough tin oxide by the last product, which
will make a second product, which may be called P.
4. From the specific gravity of tin oxide deduct the specific gravity of
silica.
5. Multiply the difference by the specific gravity of the rough tin oxide.
6. Take this product for a divisor to divide the above product P : the quo-
tient will be the weight of pure tin oxide in the rough oxide, and the quantity
of metal can now be readily calculated.
The following is an assay worked out in this manner :
400 grains of the ore are treated with nitro-hydrochloric acid and ammonia
as above described, washed, and dried. Suppose the dried matter weighs
250 grains. The 250 grains thus obtained are placed in the specific gravity
bottle, and the specific gravity is found to be 5*4.
Specific gravity of tin oxide (approximate) . * . 6*9
silica' . . 2*6
Sp. Gr.
Bough Oxide
5-4
2-8
Weight of
Bough Oxide
250
Sp. Gr.
Pure Oxide
6-9
4-3
Sp. Gr.
Silica
2-6
Sp. Gr.
Pure Oxide
6-9
19-32
Sp. Gr.
Silica
2-6
Sp. Gr.
Bough Oxide
5-4
4830 =208-4
2-8
19-32
4830
4-3
23-22
23-22
N N
546 THE ASSAY OF TIN.
208*4 grains is therefore the weight of pure oxide in the 400 grains of ore.
Now tin oxide contains 78-61 parts of pure tin, and
208-4 x 78-61
100
So that 400 grains of rough tin ore contain 163*72 grains of pure tin, and
=40-93.
The rough sample first operated on contains, therefore,
40*93 per cent, of metallic tin.
Mr. J. H. Talbott's Process for assaying Tin in the Pre-
sence of Tungsten. Another satisfactory method of assaying
tin in the presence of tungsten has been described by Mr. J.
H. Talbott.
The method is based on the fact already mentioned
that tin oxide is reduced by potassium cyanide with great
facility ; while tungstic acid undergoes no reduction, even
when heated with the cyanide at a high temperature. The
tin and tungsten oxides are to be heated in a porcelain
crucible with three or four times their weight of commer-
cial potassium cyanide previously fused, pulverised, and
thoroughly mixed with the two oxides. The mass is kept
fused for a short time, when the tin separates in the form
of metallic globules, while the tungstic acid unites with
the alkali of the potassium cyanate and carbonate present.
After cooling, the mass is to be treated with hot water,
which dissolves the alkaline tungstate and other salts, and
leaves the tin as metal ; this, is to be separated by filtra-
tion, washed, dried, and weighed as tin oxide, after oxida-
tion in the crucible with nitric acid. The tungstic acid
may be estimated by difference, or be precipitated by
mercury protonitrate, after boiling the solution with nitric
acid to decompose the excess of potassium cyanide present,
and then re-dissolving the precipitated tungstic acid by
means of an alkali.
Estimation of Tin by the Wet Method. There are
several methods of effecting this analysis, the chief diffi-
culty- being found in the intractable nature of the tin oxide,
it resisting the action of all acids. This, however, may be
overcome as first shown by Klaproth, who found that very
finely levigated tin oxide was soluble in hydrochloric acid
ESTIMATION OF TIN IN THE WET WAY. 547
after a prolonged fusion with caustic potash. The following
is his process :
Fifty grains of the tin ore, reduced to the most minute
state of division by levigation or otherwise, are mixed with
four times its weight of caustic potash. The best mode of
mixing is to place the caustic potash in a silver crucible,
add its own weight of water, and apply a gentle heat until
the potash is dissolved ; then stir in tin ore, and gradually
evaporate to dryness, stirring all the time to prevent loss by
spirting, as in the analysis of ironstone : when thoroughly
dry, enclose the silver crucible in one of clay, and submit
the whole to a dull red heat for at least half an hour ;
rather more than less renders the perfect solution of the
tin oxide more certain. When cold, act on the contents
of the crucible with dilute hydrochloric acid, transfer the
liquid and any undissolved matter to a flask, add some
strong hydrochloric acid, and boil for half an hour. If at
the end of this time any of the tin ore remains unacted on,
it must be separated by decantation or otherwise from the
solution, dried, again fused with potash, and then treated
with hydrochloric acid, in which it will now be found
totally soluble. This second operation will not be needed
if care has been taken to reduce the ore to the finest
possible state of division at first. The solution, however
obtained, is to be evaporated to dryness, and when cold
treated with a small quantity of hydrochloric acid, allowed
to stand for half an hour, then water added, boiled, and
filtered : the whole of the tin will pass through in solution
as tin chloride, and any silica or tungstic acid that may
be present will remain in the filter. If the or$ contains
copper, lead, and iron, these metals will also be in solution
at all events, the lead partially so ; but if the ore has,
previously to its fusion with caustic potash, been treated
with aqua regia, as already described, then it will contain
tin alone. It is always better thus to separate foreign
matters before attempting the solution of the tin, as the
after process is thereby simplified. Supposing, however,
that the rough ore has been submitted to fusion with
potash and then dissolved, the solution must be treated
Nur O
->
548 THE ASSAY OF TIN.
thus : A bar of zinc must be placed in the solution,
which will in course of time precipitate tin, copper, and
lead ; when all the metals are thus thrown down the zinc
is washed and removed, the precipitated metals well
washed and dried. To the dried metals strong nitric acid
is now to be added, the mass gently heated, and then
evaporated to dryness : when cold it is moistened with
dilute nitric acid, water added, and the whole filtered.
Lead and copper will pass through the filter as soluble
nitrates, and the tin will be found in the filter as insoluble
peroxide ; this is to be well washed, dried, ignited, and
weighed. It contains 78-61 parts of metallic tin. The
amount of tin thus obtained, when multiplied by two, will
represent the percentage of the ore.
If, before the action of caustic potash, the ore has
been submitted to the action of aqua regia, sulphuretted
hydrogen may be passed through the solution of tin chlo-
ride, when tin sulphide will be precipitated ; this is to
be washed, dried, gently calcined in a platinum crucible
until all smell of sulphurous acid has ceased, allowed to
cool, reheated with a fragment of ammonium carbonate,
as in the case of roasting copper sulphide, and when cold
weighed as pure oxide of tin. The calculation for metal
is made as above.
Mr. J. W. B. Hallet has found that tinstone is very
easily dissolved by fusion with three or four times its
weight of potassium fluoride. The mineral must be finely
pulverised. The fused mass is treated directly in the cru-
cible with sulphuric acid to expel fluorine, after which,
by adding water, filtering, and boiling the filtrate, the whole
of the tin is thrown down as stannic acid, which is to be
separated from traces of iron in the usual manner. This
method of dissolving the ore of tin is much more convenient
than fusion with caustic alkalies, or with sulphur and
sodium carbonate.
M. Moissenet precipitates the metal from a solution of
the chloride by means of zinc, and then melts the precipi-
tated metal in stearic acid. His process comprises five
operations :
ASSAY OF TIN IN GUN- AND BELL-METAL. 549
1. Purification of the ore by treatment with aqua regia.
2. Eeduction of the residue in the presence of charcoal.
3. Solution of the tin and iron in hydrochloric acid.
4. Precipitation of the tin by means of zinc.
5. Fusion of the precipitate into a button in stearic
acid.
The precipitation of tin by zinc is very rapid, and takes
place in strongly acid solutions ; but the amount of acid and
the dilution of the chloride influence the condition of the
precipitate. In some solutions it appears in brilliant
needles, but in very dilute solutions, and always towards
the end of an operation, it is only a muddy deposit. The
author recommends that a button of zinc be suspended in
the liquid by means of a copper wire. When the precipi-
tation is finished, the metal is collected and pressed into a
porcelain capsule. On applying heat the lump so formed
melts in a few minutes if a piece of stearine is added to it.
Assay of Tin in Gun- and Bell-metal. The follow-
ing process was employed for some years in H. Sainte-
Claire Deville's laboratory: Dissolve about 5 grms. of the
alloy in strong nitric acid contained in a flask provided
with a funnel in the neck to prevent loss by spirting.
When quite dissolved boil the strong solution for about
twenty minutes ; dilute with two or three times its bulk
of water, and boil again for the same time. Separate the
insoluble tin oxide by decantation or filtration, and weigh
after calcining it. (The tin oxide is sometimes rose-
coloured, owing to the presence of minute traces of gold ;
this may be disregarded.) The nitric acid solution freed
from the tin is evaporated on a small platinum or porce-
lain dish, and the residue is calcined at a dull red heat.
In this manner a mixture of oxides is obtained in a suf-
ficient quantity to serve for at least two analyses.
About 2 grms. of the finely pulverised oxides are
placed in a small platinum or porcelain boat, and thence
introduced into a small glass tube closed with a good cork
suitable for weighing. The boat, the tube, and the cork
having been previously weighed, the weight of the oxides
is obtained after they have been heated to dull redness in
550 THE ASSAY OF TIN.
the apparatus, through which a current of dry air circu-
lates. After having weighed the whole the current of air
is replaced by dry hydrogen, and the tube is heated over
a lamp until the contents cease to lose weight. It then
contains unreduced zinc oxide together with copper, lead,
and iron in the metallic state ; the colour of the product
shows the operator when the experiment is concluded. On
weighing again the loss of weight indicates with great
accuracy the amount of oxygen contained in the oxides of
these three metals.
If the iron and lead are present in inappreciable quan-
tities, multiplying this loss by 5 will give very nearly
the weight of copper present, and, in consequence, the
composition of the alloy itself.
For bronze, bell-metal, gun-metal, &c., E. Burse
('Zeitschrift fur Analytische Chemie,' 1878, p. 58) pro-
ceeds as follows : 1 grm. of the alloy, cut in pieces, is
placed in a beaker, covered with 6 c.c. nitric acid of sp.
gr. 1-5 : 3 c.c. water are then slowly poured in, and the
beaker quickly covered. When the whole is dissolved it
is heated to a boil and diluted with 500 c.c. boiling water.
The tin oxide, after it is completely settled, is washed with
boiling water, and weighed. The filtrate, for the expul-
sion of nitric acid, is evaporated with two grms. sulphuric
acid, and the copper, after the addition of sulphurous acid,
is precipitated with a solution of 2 grm. ammonium sulpho-
cyanide. The copper sulphocyanide, after settling, wash-
ing, and drying, is weighed as such, or is converted into
sulphide. The filtrate is concentrated with the addition
of nitric acid, and mixed with ammonia in excess. If
iron oxide is deposited, it is filtered off and weighed. The
ammoniacal solution of zinc is mixed with ammonium
sulphide, avoiding excess ; the sulphide when deposited is
filtered off, dried, and, after ignition with sulphur in a
current of hydrogen, weighed.
Estimation of Tin by means of a Standard Solution.
The first process to be described is due to M. Gaultier de
Clauby, and may be thus performed : The standard solu-
tion is made by dissolving 100 grains of iodine in 1 quart
ESTIMATION OF TIN BY A STANDARD SOLUTION. 551
of proof spirit (specific gravity "920), and is thus standard-
ised. Ten grains of pure tin are dissolved in excess of
hydrochloric acid, the solution boiled, and allowed to
cool : the burette is now filled with the solution of iodine
which is gradually added to that of the tin until the former
ceases to be decolourised ; as soon, therefore, as the tin
solution assumes a faint yellow tinge, sufficient iodine has
been added : the quantity thus found sufficient is then
noted, and the amount of tin each division of iodine
solution is equivalent to is calculated as for iron, copper,
and the other standard solutions.
In the actual assay of tin ore by means of this solution
it is necessary that the whole of the tin present be reduced
to the state of protochloride : this may be readily effected
by boiling the solution of tin for a quarter of an hour with
excess of metallic iron, and filtering. To the solution so
obtained the iodine is added as above. The tin ore is dis-
solved by any of the methods already described.
M. Lenssen* estimates tin by means of the iodine
.solution, but he operates in a liquid containing double
potassium and sodium tartrate, and sodium bicarbonate
in excess. The results M. Lenssen obtained by this
method are satisfactory, by using the atomic weight of tin
generally adopted (59). We shall see farther on why
M. Lenssen's results agree.
M. Stromeyer,f having recently occupied himself with
the same subject, has succeeded in solving the difficulty.
The solution of stannous chloride is carefully introduced
into an excess of ferric chloride. The salt of iron becomes
reduced to a minimum, according to the following equa-
tion :
2Sn + 2(Fe 2 Cl 6 ) = 2SnCl 2 + 4FeCl 2 .
It is then estimated by permanganate, as if it were a
salt of iron protoxide. The results M. Stromeyer obtains
in this way are very exact. The author adds that such a
method of estimating is applicable only in the absence of
copper or iron, as these two metals decompose potassium
* Annalen der Chemie und Pharmacie,' vol. cxiv. p. 114.
t Ibid. vol. cxvii. p. 261.
552 THE ASSAY OF TIN.
permanganate as well as the tin ; but it may be of great
use in the estimation of commercial salts of tin.
A method for the analysis of tin ore consists in reduc-
ing the finely powdered ore, in a hard porcelain vessel^
in a current of hydrogen gas. A bright red heat must
be applied and maintained for a considerable time. The
sample is allowed to cool in a current of hydrogen and
weighed, the operation being preferably repeated to make
sure that the reduction is complete. The loss of weight
gives the quantity of stannic acid, tin peroxide (1 part
by weight of oxygen representing 4- 673 parts tin per-
oxide), if no other oxide is present which may be reduced
at the same time. But as tinstone contains almost always
iron oxide, part of the loss is due to this compound, so
that the crucible, with its contents, should be digested in
a beaker with hydrochloric acid. When dissolved, it is
diluted, filtered into a large flask, the quantity of the
residue (silica, &c.) estimated, the liquid supersatu-
rated with ammonia, a sufficient quantity of ammonium
sulphide added with flowers of sulphur, so as to convert
the stannous sulphide into stannic sulphide, which then
dissolves in the ammonium sulphide. The whole is di-
gested in the flask, loosely stoppered, till the black iron
sulphide is separated from the yellow liquid, filtered under
cover, washed with the addition of a little ammonium
sulphide, dissolved in hydrochloric acid, oxidised with
potassium chlorate, and the ferric oxide (iron peroxide)
precipitated with ammonia. The tin may be directly
estimated by precipitating the solution of sulphide with
dilute sulphuric acid, using a slight excess. The vessel is
loosely covered with paper, and very gently heated, till it
no longer smells of sulphuretted hydrogen. The yellow
tin sulphide is washed upon the filter, dried enough to
enable it to be taken out of the funnel along with the
paper, and heated very gently, for a considerable time,
in a porcelain crucible, at first covered and then open.
When the odour of sulphurous acid is no longer percep-
tible the contents of the funnel are moistened with a few
drops of nitric acid, and are then gradually heated, with
ESTIMATION OF TIN BY A STANDARD SOLUTION. 553
excess of air, ultimately to full redness. The tin sulphide
is thus converted into stannic acid, which is heated for a
short time with a little ammonium carbonate to remove
every trace of sulphuric acid.
In another method for the assay of tinstone, the
finely ground sample is mixed in a porcelain crucible
with 3 parts sodium carbonate, and 3 parts of sulphur,
and the mixture is melted, covered, over the, lamp.
When cold, the crucible is laid in water, and heat is
applied till the mass dissolves, iron or other electro-
positive metals remaining as a black sulphide, which is
filtered off, washed, and when dry ignited in the air, in
order to convert it into oxide. From the dilute alkaline
solution the tin sulphide is precipitated by dilute sul-
phuric acid, as directed above.
In case of impure tin ores, the insoluble sulphide may
consist of iron, zinc, copper, and bismuth ; whilst the
soluble, along with the tin, may contain arsenic, tungsten,
and molybdenum. A process for the separation of tin
from tungsten has been given at p. 546.
Tin slags are decomposed by means of aqua regia, and
the tin, &c., precipitated by a current of sulphuretted
hydrogen.
Tin furnace products may be in part alloys of tin and
iron, lead, tungsten, cobalt, and arsenic. They are finely
powdered, digested with aqua regia, the tin thrown down
with sulphuretted hydrogen, and the tin sulphide treated
as already directed. If the proportion of tungstic acid is
considerable, after digestion in aqua regia and diluting, the
tungstic acid separates along with some stannic acid. This
deposit is filtered off, washed, dried, and repeatedly ignited
with sal-ammoniac, in a covered porcelain crucible, till
the weight of the residual tungstic acid is constant. The
stannic acid is found as difference. Sulphuretted hydro-
gen is then passed into the acid solution (from which the
tungstic acid has been deposited) ; the remaining tin is
thus precipitated as sulphide, and is filtered off. Iron
and manganese are estimated in the filtrate. (F. Rarn-
melsberg's Quant. Chem. Analysis.)
554 THE ASSAY OF TIN.
We have seen that M. Stromeyer, by a happy modifi-
cation, has reduced the estimation of tin to that of iron.
Applying the same principle, a salt of copper may be
substituted for a salt of iron. An equivalent quantity
of copper can thus be estimated in place of tin ; and
M. Mohr's as well as M. Terreil's * experiments show that
copper can be very exactly estimated by permanganate of
potash.
A double decomposition takes place on protochloride
of tin being added to nitrate or chloride of copper in
excess ; a salt of suboxide of copper forms, and the tin
passes to the maximum state of oxidation, according to
the following equation :
4CuO + 2SnCl 2 = 2Cu 2 + SnCl 4 + Sri0 2 -
To estimate tin it is, then, sufficient to transform it
into protochloride, to add to it a solution of nitrate of
copper slightly in excess, before diluting it with water,
and to titrate the liquid obtained by permanganate of
potash.
There are then three different processes for estimating
tin by potassium permanganate : 1. To operate with
water freed from air by boiling, protecting it from access
of air while cooling. 2. To oxidise protoxide of tin in
an alkaline medium. 3. To decompose stannous chloride
either by a salt of iron, as proposed by M. Stromeyer, or
by a salt of copper.
Alloys of tin and lead, such as solder, inferior tinfoil,
and the coating of terne plates, may be treated thus,
according to Eammelsberg :
Oxidise with nitric acid ; evaporate to dryness in the
water-bath ; heat the residue rather more strongly ;
moisten with nitric acid when cold ; dilute and separate
the stannic acid (tin peroxide) by filtration, washing till
the filtrate has no longer an acid reaction. When dry
the precipitate is detached from the filter, and placed in
a porcelain crucible; the paper is burnt on the lid, the
* Comptes-Rendus, vol. xlvi. p. 230.
ESTIMATION OF TIN BY A STANDARD SOLUTION. 555
ash added to the contents of the crucible, moistened with
a few drops of nitric acid, heated and finally ignited.
The tin is calculated from the stannic acid.
In the filtrate the lead is estimated as sulphate by
the addition of sulphuric acid. (See ' Separation of Lead
from Copper and Zinc.')
556
CHAPTEE XIII.
THE ASSAY OF ANTIMONY.
ANTIMONIAL substances susceptible of being assayed by the
dry way are divisible into two classes.
CLASS I. In this class are comprised native antimony
and all antimonial substances containing oxygen or chlorine,
and but little or no sulphur.
These substances are the following :
Native antimony, Sb,
Antimony oxide, Sb 2 O 3 ,
Antimonio.us acid, Sb 2 O 4 ,
Antimonic acid, Sb 2 5 .
CLASS II. includes antimony sulphide and all antimo-
nial ores containing much sulphur.
Antimony sulphide, Sb 2 S 3 ,
Antimony oxysulphide, Sb 3 + 2Sb 2 S 3 ,
Haidingerite, 2Sb 2 S 3 + 3FeS.
ASSAY OF ORES OF THE FIRST CLASS.
Antimony oxides are readily reduced by charcoal, so
that their assay presents no difficulty. The assay is con-
ducted in precisely the same manner as that of lead oxide ;
only, as antimony is much more volatile than lead, the
heat must be managed with care, and the assay taken
from the fire as soon as finished. When all suitable pre-
cautions are taken, the loss of antimony is not very con-
siderable ; but Berthier says it is never less than from 5
to 6 per cent. This, we think, is too high. The pure
protoxide gives 77 per cent, of metal, and antimonious
ASSAY OF ANTIMONY SULPHIDE. 557
acid 75. The reduction is readily made, without addition,
in a charcoal crucible ; but when the substance to be
assayed is mixed with impurities, some flux must be
added. It succeeds equally well with 3 parts of black
flux, with 1 part of tartar, with 1 part of sodium carbo-
nate, and 15 per cent, of charcoal, or any other equivalent
reducing flux.
When the substance under assay contains iron oxide,
the latter oxide is more or less reduced, and the metallic
iron alloys with the antimony.
Oxidised minerals which contain but a small quantity
of sulphur can also be assayed in this manner ; because
the sulphide gives up to black flux the small quantity of
antimony which it contains, so that but little remains in
the slag. The common glass of antimony produces by
this method of assay 70 per cent, of antimony, and occa-
sionally even more than that.
The ores of this class occur very seldom, and are only
in rare cases subject to assaying.
ASSAY OF ORES OF THE SECOND CLASS.
As pure antimony sulphide (antimonium crudum) as
well as metallic antimony (regulus of antimony) are mer-
cantile substances, the assays of the ores of this class have
for their object the estimation of both these bodies.
1. ASSAY OF PURE ANTIMONY SULPHIDE (ANTIMONIUM
CEUDUM).
Antimony sulphide is almost the only mineral from
which antimonium crudum is produced. This mineral
generally occurs intermixed with very refractory gangue
(gneiss, quartz, limestone, &c.) Antimony sulphide fuses
readily at a low red heat, and is not changed during
fusion if atmospheric air is precluded. At a white heat it
volatilises without change of composition.
The assay of antimony sulphide is now effected by a
liquation process, i.e. by heating the mineral sufficiently in
order to melt the antimony sulphide, and, by this means,
558 THE ASSAY OF ANTIMONY.
to separate it from the refractory gangue. The produc-
tion of antimony sulphide on a large scale is done in the
same way.
For the purpose of assaying, two pots or crucibles are
used, one standing in the other, and leaving sufficient
space between the two to receive the fused antimony
sulphide. The bottom of the inside crucible is furnished
with holes. The mineral to be assayed is put into the
inside crucible, the latter is then closed with a cover, and
hermetically luted ; the joints of the two crucibles are
also luted. The under crucible is then put on the hearth
of a furnace, enclosed with ashes or sand, in order to keep
it cool, and the upper crucible, as far as it is outside of
the under crucible, is covered with coal, and heated to a
moderate red heat. The antimony sulphide will then
melt and collect in the under crucible, from which it may
be taken out, after cooling, and weighed.
2. ASSAY OF EEGULUS OF ANTIMONY.
This assay may be made in two ways : first, by roast-
ing and fusing the oxidised r L matter with black flux ;
secondly, by fusing the crude ore with iron, or iron
scales, with or without the addition of black flux.
The roasting of antimony sulphide requires much
care, for it is very fusible and volatile, as is also the oxide
its decomposition gives rise to. The heat ought to be
very low during the operation, and the substance con-
tinually stirred. When no more sulphurous acid is given
off, we may feel [assured that it is perfectly roasted,
because no sulphate is ever formed in this operation.
The roasted sulphide is then fused with three parts of
black flux, or its equivalent.
Metallic iron very readily separates all the sulphur
from antimony sulphide ; but as iron sulphide has a
specific gravity near that of antimony, the separation is
very difficult to manage : a strong fire must be employed
when the desulphurisation is complete, to keep the whole
body in full fusion, for a considerable time. With these
precautions, two buttons are obtained, which separate very
ASSAY OF REGULUS OF ANTIMONY. 55
well : the one white, and in large plates, which is anti-
mony ; and the other a bronze-yellow, a little brighter
than the ordinary iron sulphide, because it is mixed with
a little metallic antimony. During the operation a con-
siderable portion of antimony is always volatilised, which,
by this process, is an inconvenience impossible to avoid.
It is, nevertheless, practised in the large way in some
factories ; but a good result is not generally obtained. It,,
however, appears that when all the necessary precautions
are taken, it can be employed with advantage.
The first precaution which is indispensable is, mixing
with the sulphide only the precise proportion of iron
necessary to effect its decomposition, which quantity
amounts to about 42 per cent, of its weight. If more be
used, the antimony, having a great tendency to play the
part of an electro -negative element, will combine with the
surplus, and an iron antimonide results, part of which will
remain in the antimony and part in the slag.
Further, the iron ought to be in the finest possible
state of division. If the masses be large, a portion of
antimony sulphide is volatilised before they can be fully
attacked. In general, 63 per cent, of antimony can be
extracted from its sulphide by the aid of iron in the small
way, but on the large scale it seems that 55 per cent, is
the maximum.
Cast iron cannot be employed instead of wrought,,
because sulphur has very little action on it. The desul-
phurisation is imperfect, and the slag adheres to the
reduced metal.
One of the greatest inconveniences in separating
sulphur from antimony by means of iron is the strong
heat necessary to separate the slag from the metal. This
may be remedied by making the slag more fusible and
less heavy, by the addition of some flux, as an alkaline
carbonate or sulphate.
If antimony sulphide be fused with an alkaline car-
bonate and charcoal, a regulus is obtained, and a slag com-
posed of an alkaline sulphide and antimony sulphide. If
metallic iron be thrown into this slag whilst in fusion, all
560 THE ASSAY OF ANTIMONY.
the antimony separates immediately, and a new slag is
formed as fluid as the former, containing iron sulphide
and a sulphide of the alkaline base employed. If, instead
of the above process, the iron be mixed intimately with
the antimony sulphide and carbonated alkali, the result is
the same 100 parts of sulphide, 42 of metallic iron, 50 of
sodium carbonate mixed with one-tenth of its weight of
charcoal, or 50 of black flux, give 65 to 66 of regulus ;
with the same proportion of iron, and only 10 of flux,
only 62 per cent, can be obtained. In these two cases the
fusion takes place very rapidly and without bubbling, and
the slag, which is very liquid, separates readily from
the metal. By employing 1 part of alkaline flux, the
proportion of iron can be reduced from 2 5 to 30 per cent.,
and the product of metal is always from 65 to 66 per
cent.
Hence, in making an assay of antimony sulphide, it is
always better to employ a smaller quantity of iron than is
necessary to complete the desulphurisation, and make
up for it by increasing the quantity of flux : then it may
be insured that no excess of iron will be present.
The alkaline sulphates are decomposed into alkaline
sulphides by the agency of charcoal at a slightly elevated
temperature. The sulphides of the alkaline metals, by
combining with the other metallic sulphides, augment
their fusibility very considerably. Thus when sodium
sulphate, mixed with about one-fifth of its weight of
charcoal, is added to a mixture of antimony sulphide
and metallic iron, the metallic antimony separates very
rapidly, and the slag almost instantly becomes perfectly"
fluid.
But it must be noted that the presence of an alkaline
sulphide diminishes the product of regulus, unless the
proportion of iron be augmented at the same time.
For instance, with
100 parts of antimony sulphide,
42 parts of iron,
100 parts of sodium sulphate,
20 parts of charcoal,
ASSAY OF EEGULUS OF ANTIMONY. 561
but 22 parts of regulus were furnished ; but with
100 parts of antimony sulphide,
42 parts of iron,
10 parts of sodium sulphate,
2 parts of charcoal,
62 parts of antimony were easily obtained.
Instead of metallic iron, pure iron oxide may be used,
or any ferruginous matter whatever, provided it is rich ;
but it is necessary to add, at the same time, an alkaline
flux and charcoal to reduce the iron oxide.
Not less than 40 parts of iron scales can be employed
for 100 of antimony sulphide ; and then, on the addition
of 50 to 100 parts of sodium carbonate and 8 to 10 of
charcoal, about 56 of regulus are obtained ; but if with
100 parts of sodium carbonate from 14 to 15 parts of
charcoal be employed, 65 parts of antimony are the result.
By augmenting the proportion of scales, that of soda may
be diminished. Thus, if from 56 to 60 parts of scales, 10
of soda, and 10 of charcoal be employed, 50 parts of
regulus are the result ; and if the proportion of soda be
50, and that of carbon 10, from 65 to 66, and even 67,
parts of regulus are obtainable.
The fusion always takes place quickly, and the slags
are very fluid.
When antimony sulphide is fused with forge slag
(ferrous silicate), sodium carbonate, and charcoal, a very
white crystalline regulus, in large plates, is obtained ; to-
gether with a bronze-yellow matte, and a black, opaque,
vitreous slag, shining like jet, in which the greatest
portion of the alkali employed appeared to be concen-
trated.
These three substances separate very readily from each
other.
100 parts of antimony sulphide,
80 parts of forge slag,
50 parts of sodium carbonate,
10 parts of charcoal,
produced very readily 60 parts of regulus.
o o
569! THE ASSAY OF ANTIMONY.
Antimony sulphide may also be analysed by boiling
with aqua regia. The residue consists of sulphur and
gangue. It is to be washed and dried, then weighed and
ignited. The loss will be sulphur, and the remainder is
pure gangue.
Water is then added to the filtered solution, which will
cause the precipitation of some of its contained antimony
as oxy chloride : this must be separated by filtration. The
solution is then to be saturated with potassium carbonate,
and a new precipitate will be formed. The solution is to
be filtered, and made slightly acid ; then barium nitrate
must be added to it to separate its sulphur as barium
sulphate, which is to be washed, dried, and weighed ; its
weight indicates the amount of sulphur : 116 parts are
equal to 16 parts of sulphur.
The precipitate by water of oxychloride which remains
on the filter is redissolved by hydrochloric acid, and its
antimony separated in the metallic state by means of zinc.
The precipitate formed by potassium carbonate may con-
tain lead, copper, iron, and antimony. It must be treated
with nitric acid ; this dissolves everything but the antimony,
which may then be estimated as antimonic acid.
It is always best, before conducting the analysis of
antimony sulphide, to moisten it with very dilute hydro-
chloric acid, in order to dissolve a portion of the calcium
carbonate which may form part of the gangue. As
the composition of the antimony sulphide is constant,
the following process is sufficient in the assay of an
antimonial ore :
Boil the ore, after treatment with dilute hydrochloric
acid, with concentrated hydrochloric acid, which dissolves
only antimony sulphide, and precipitate the metal as oxy-
chloride by means of water.
Or, after all gangue soluble in dilute hydrochloric acid
has been removed, the residue may be weighed, and then
acted on by boiling hydrochloric acid, until all action
ceases. The residue must be well washed with weak hydro-
chloric acid, dried, ignited, and weighed ; the loss of weight
corresponds to the percentage of pure antimony sulphide,
which contains 72-7 -per cent, of metal.
DETECTION OF ANTIMONY IN SUBLIMATES. 5(33
Franz Becker (' Zeitschrift flir Analyt. Chemie,' 1878,
p, 185) mixes 1 part of the ore with 3 parts sodium car-
bonate and 3 parts sulphur, melts in a porcelain crucible,
extracts with hot water, decomposes the filtrate with
hydrochloric acid, and converts the antimony sulphide
into oxide in the usual manner.
. The following method of estimating antimony is given
by Mr. Sutton :
The oxide of the metal, or any of its compounds, is
brought into solution as tartrate by tartaric acid and water ;
the excess of acid neutralised by neutral sodium carbonate,
then a cold saturated solution of sodium bicarbonate added
in the proportion of 20 c.c. to about CM grm. of Sb 2 3 ; to
the clear solution starch liquor and -^ iodine are added
until the blue colour appears ; the colour disappears
after a little time, therefore the first appearance of a per-
manent blue is accepted as the true measure of iodine
required.
1 c.c. yV iodine=0-0061 grm. Sb.
Detection of Antimony in Sublimates. In the examina-
tion of mineral bodies for antimony, the test substance is
often roasted in an open tube for the production of a
white sublimate. Dr. E. Chapman, Professor of Miner-
alogy at Toronto, recommends for the detection of anti-
mony in this substance the following process a method
more especially available when the operator has only a
portable blowpipe-case at his command : The portion of
the tube to which the chief part of the sublimate is
attached is to be cut off by a triangular file, and dropped
into a test-tube containing some tartaric acid dissolved
in water. This being warmed, or gently boiled, a part
at least of the sublimate will be dissolved. Some
potassium bisulphate either alone, or mixed with some
sodium carbonate and a little borax, the latter to prevent
absorption is then to be fused on charcoal in a reducing
flame ; and the alkaline sulphide thus produced is to be
removed by the point of the knife-blade, and placed in a
small porcelain capsule. The hepatic mass is most easily
o o 2
564 THE ASSAY OF ANTIMONY.
separated from the charcoal by removing it before it has
time to solidify. Some of the tartaric acid solution is
then to.be dropped upon it, when the well-known orange-
coloured precipitate of antimony sulphide will at once
result.
In performing this test, it is as well to employ a
somewhat large fragment .of the test substance, so as to
obtain a thick deposit in the tube. It is advisable also to
hold the tube in not too inclined a position, in order to
let but a moderate current of air pass through it ; and
care must be taken not to expose the sublimate to the
action of the flame, otherwise it might be converted
almost wholly into a compound of antimonious and anti-
monic acids, the greater part of which would remain
undissolved in the tartaric acid. A sublimate of arsenious
acid, treated in this manner, would of course yield a
yellow precipitate ; easily distinguishable by its colour y
however, from the deep orange antimonial sulphide. The
crystalline character, &c., of this sublimate would also
effectually prevent any chance of misconception.
To distinguish Arseniuretted Hydrogen from Antimo-
niuretted Hydrogen. On passing a mixture of these two
gases through a -tube containing solid pieces of caustic
potash, these become covered with a brilliant metallic
coating of antimony, whilst the arseniuretted compound
escapes undecomposed. A lye of potash, density 1-250,
only acts very slightly in a similar case. The fragments
of potash which have become metallised by the deposit of
antimony are altered in the air ; they soon become white
in water, the metallic coating falling to the bottom ; but
when it is attempted to collect them on a filter they
dissolve before the liquid has even passed through. In
the clear filtrate, antimony is found in solution.
Separation of Tin from Antimony and Arsenic. Dr.
Clemens Winckler (' Zeitschrift fur Analyt. Chemie,' 1875,
p. 163) proceeds as follows : If the substance is an alloy
it is dissolved in a mixture of 4 parts hydrochloric acid,
and 5 parts water, with the addition of a sufficient quan-
tity of tartaric acid. If the mixed metals exist as sul-
DETECTION OF ANTIMONY IN SUBLIMATES. 565
phicles they are collected on a filter, washed, and dis-
solved upon it in dilute potash lye. The nitrate is mixed
with tartaric acid, treated with a current of chlorine, or
with bromine in slight excess, and neutralised with hydro-
chloric acid. In either case the solution is introduced into
a beaker, diluted to 300 or 400 c.c., and so much solution
of calcium chloride of a known strength is added that the
subsequently precipitated calcium carbonate may exceed
the tin present by about 15 times in weight. The liquid
is then neutralised with potassium carbonate, potassium
cyanide is added, and afterwards a slight excess of potas-
sium carbonate, so that the lime may be totally precipi-
tated. The liquid is then heated till it begins to boil ; it is
allowed to settle, the liquid is poured upon a filter without
disturbing the sediment, which is then treated with fresh
water, boiled up, allowed to settle, and the clear liquid
poured upon the filter. In this manner the bulk of the
antimony is removed.
The precipitate in the beaker is dissolved in a little
concentrated hydrochloric acid, tartaric acid is added,
the liquid again neutralised with potassium carbonate,
and precipitated with potassium cyanide.
After boiling, the liquid is poured as above through
the same filter, three successive portions of water are
added, heating each time to a boil, and the precipitate
is finally brought upon the filter and completely washed.
All the arsenic and antimony are now found in the filtrate,
and all the tin, with an excess of calcium carbonate, in the
precipitate.
ASSAY OF ALLOYS OF LEAD AND ANTIMONY (Type Metal).
The comminuted alloy is digested in a flask with concen-
trated nitric acid till the antimony is oxidised. It is then
slightly diluted with water, supersaturated with ammonia,
and mixed with an excess of ammonium bi-hydrosulphide,
which must be concentrated and yellow. If recently
prepared a small quantity of flowers of sulphur is added.
The whole is digested in the flask for some time, heating
at last almost to a boil. The flask is then corked up and
let stand till the deposit, which must be of a pure black,
566 THE ASSAY OF ANTIMONY.
has subsided, while the supernatant liquid is yellow. If
this is not the case more ammonium bi-hydrosulphide is
needed. When cold the undissolved lead sulphide is
separated by nitration from the solution of the antimony
sulphide, washed with cold water, dried, placed in a tared
porcelain crucible, the filter burnt separately to ashes,
which are added to the precipitate along with some sul-
phur, and the whole is ignited in a current of hydrogen.
After the crucible with its contents has been weighed the
addition of sulphur and the ignition are repeated, in order
to find if the weight is constant. The lead is calculated
from the weight of the lead sulphide.
The solution of antimony sulphide is mixed, drop by
drop, with hydrochloric acid, stirring all the time till the
reaction becomes slightly acid. The beaker is covered
with a glass plate and allowed to stand till the antimony
sulphide is deposited. It is digested at a very gentle heat
till the odour of sulphuretted hydrogen has ceased ; when
cold it is filtered upon a filter dried at 120 C., and washed
with cold water to which a few drops of hydrochloric
acid have been added. When air-dry it is kept at the
heat of 120-130C. in the hot-air oven till its weight is
constant. A weighed portion of this substance is placed
in a bulb- tube and heated in a current of dry carbonic
acid, at first very gently, and afterwards at 200-300.
The black-grey tin sulphide remains, w r hich is allowed to
cool in a current of carbonic acid and weighed.
567
CHAPTER XIV.
THE ASSAY OF ZING.
ALL bodies containing zinc, usually found in the assay
office, may be divided into four classes :
Class I. Zinc ores, in which the metal exists as oxide
not combined with silica :
Earthy zinc oxide, ZnO.
Manganiferous zinc oxide, ZnO + MnO.
Zinc aluminate, Gahnite, ZnO,6AL 2 O 3 .
Franklinite, 3(FeO,ZnO) + (Fe 2 3 ,Mn 2 3 ).
Anhydrous zinc carbonate, ZnO,CO 2 .
Hydrated zinc carbonate, ZnO,3H 2 + 3ZnO,C0 2 .
Class II. Zinc ores, in which the metal exists, as in
the former class, as oxide, but partly or wholly combined
with silica :
Anhydrous zinc silicate.
Hydrated zinc silicate, electric calamine.
Class III. Zinc ores, in which the metal is partly or
wholly combined with sulphur.
Zinc sulphide (blende, Black Jack), ZnS.
Zinc oxysulphide. (This is rare.)
Zinc sulphate, ZnO, S0 3 ,7H 2 0.
Zinc selenide, ZnSe. (Very rare.)
Class IV. Alloys.
ASSAY OF ORES OF THE FIEST CLASS.
In order to reduce the zinc oxide contained in sub
stances of this class, it is sufficient to mix them with char-
coal and expose them to a white heat.
568 THE ASSAY OF ZINC.
At the moment of reduction the zinc is in a vaporised
state. Its vapours, however, are readily condensable, so
that the operation may be conducted in an ordinary retort,
and all the metal is deposited in the neck without loss.
It seems from this that nothing is so easy, at first sight,
as the assay of zinc oxide ; but it is not so. It is very
easy to reduce all the oxide, but it is not so easy to col-
lect all the zinc ; nor is it easy to condense it all in the
metallic state, and in consequence to estimate the pre-
cise proportion in the ore submitted to assay.
This difficulty consists, first, in the deposit being
extended over a large surface and often adhering very
strongly to the sides of the retort, so that it is nearly im-
possible to detach it ; and, secondly, as the neck of the
retort is open, the air, having access to it, brings to the
state of oxide all the vapour nearest the end of the neck.
The proportion of zinc oxidised is larger in proportion to
the smallness of the quantity submitted to assay, and is
always very considerable where no more than 200 to 400
grains are operated upon.
It is not, therefore, in the extraction of the zinc from
its oxide that the assay is rendered partially uncertain, but
in its collection.
The distillation of zinc requires a very high tempera-
ture, and cannot be performed in retorts of glass ; those of
earthenware must be employed. It is not necessary to
lute these retorts when they are of good quality ; and
they are better thin, because they heat more rapidly, and
are not so likely to crack.
After the mixture of oxide and charcoal has been
introduced into the retort it is placed in the fire. The
neck ought to have a long tube of glass, with a narrow
bore adapted to it, so as to collect all the zinc which may
escape from the wide part of the neck of the retort.
This disposition is also convenient, as it does not allow
such a free access of air.
It is heated gradually until it is white inside ; the zinc
is reduced and volatilised, and condensed in the neck : the
greater the heat, the nearer the orifice. The metal can be
ASSAY OF OEES OP THE FIRST CLASS. 569
detached readily from the neck, if it be well blackleaded
inside. It is necessary, from time to time, to observe the
state of the neck, because when very narrow it is often
obstructed, and, if not cleaned out with an iron rod, an
explosion might be caused.
When the operation is finished the apparatus is allowed
to cool, the retort taken out, placed carefully on its side
and broken, in order that any particles of zinc which have
condensed in its dome may be removed.
If the approximate proportion of metallic zinc alone be
required, all is collected and fused at a very gentle heat in
a crucible with some black flux ; but if the true quantity
of zinc is to be estimated it must be done in a more exact
manner. The deposit must be collected with all possible
care ; the neck must then be broken to pieces, and every
piece having adhering to it either zinc or oxide must be
placed on one side, and digested in hot nitric acid, which
takes up those substances. If any be in the glass tube, it
must be carefully cleansed by means of acid, and the solu-
tion added to that produced by the digestion of the broken
neck, and the deposit mechanically collected, in nitric acid.
The solution is then evaporated gradually to dryness and
heated to redness. The nitrate, by these means, is decom-
posed and transformed into oxide, four-fifths of the weight
of which is equal to the quantity of metallic zinc produced
in the assay.
The foregoing is the method of estimation by dis-
tillation ; the following is the method of estimation
by difference. Two plans of assay in this manner may
be adopted : first, at an ordinary assay temperature ;
secondly, at a very high temperature, as that of an iron
assay. In all cases it is necessary to commence with the
expulsion of all volatile bodies the ore may contain. If
water or carbonic acid alone be present, simple calcination
will do ; but if carbonaceous matter, roasting must be had
recourse to.
When the assay is made at an ordinary assay tempera-
ture, the sample is finely pulverised and mixed with from
15 to 20 per cent, of equally finely pulverised charcoal,
570 THE ASSAY OF ZINC.
and pressed into a crucible, on which a cover is placed,
but not luted, and rapidly heated to whiteness. When no
more zinc vapour is disengaged it is cooled, and the mixture
in the pot collected. The residue ought to be pulverulent ;
but as it is mixed with some charcoal, it is roasted, and
then weighed. It is evident that the loss represents the
zinc oxide. The charcoal added, it is true, leaves a small
quantity of ash, but it is too small to be worth accounting
for.
In making the assay in the manner described, it is to be
feared that a small quantity of the oxide might remain un-
decomposed, and that a part of the residue might adhere to
the crucible, and could not be detached ; and, lastly, there
is always a degree of uncertainty as to the state of oxida-
tion in which the iron will exist after roasting. No incon-
venience of this nature presents itself when the assay is
made at a very high temperature. This mode is the most
exact of all, and leaves nothing to be desired.
The assays of zinc at a high temperature are made
exactly as those of iron. They are made in a charcoal
crucible, with the addition of fixed fluxes, suitable to effect
the fusion of the gangues mixed with the zinc oxide, if
they be not fusible by themselves. The button is weighed ;
it is a compound of slag and grains of iron, which are col-
lected and their weight ascertained, and, by the difference,
that of the slag. The weight of oxygen which the iron
has lost during its reduction is then added to it, and by
deducting from the substance the weight of the button
and the oxygen so obtained we have the proportion of
zinc oxide reduced in the assay. On the other hand, by
deducting from the weight of the slag the weight of flux
added, the weight of earthy substances and irreducible
oxides which were mixed with the zinc oxide is ascer-
tained.
These results can be shown in a tabular form, in the
following manner :
Let m be the weight of the crude ore, n the weight of
the calcined ore, r the weight of the flux added, f the
Aveight of the cast iron, s the weight of the slag, o the
ASSAY OF OKES OF THE FIRST CLASS. 571
weight of oxygen combined with the iron, calculated from
the weight of metal produced, ' z the weight of the zinc
oxide, then:
in crude ore = calcined ore . n
r fixed fluxes ...... r
n + r
Gives metal . . /\ Tfttfll f , >>
Gives slag . . i/S^ }/+***
Zinc oxide n + r f s o
Flux added . . . r
Earthy matter . s - r
The following is an actual experiment by Berthier :
100 crude ore = calcined ore ..... 83'3
10 kaolin (china clay) acted on by acids . . . lO'O
7 marble = lime ....... 4-0
97-3
Gave metal . 45-3 \ Tfl fi1 .o^
Gave slag. . 16-3 j ^ W*j*<M . . 80-7
Zinc oxide . . 16'6
Fluxes added ...... . . 14-0
Earthy matters ....... 2-0
The above result was confirmed by wet analysis, show-
ing at once the exactitude of the process.
Estimation of Amount of Zinc by the Wet Process in
Ores of the First Class. Dissolve 50 grains of the finely
pulverised ore in nitric acid, evaporate to dryness, allow
to cool. Digest the cold mass with a little dilute nitric
acid, gently warming during the digestion ; add water, and
then filter. To the filtered solution add excess of caustic
ammonia, gently warm, and filter. The excess of caustic
ammonia dissolves the zinc oxide which it at first threw
down, as well as any manganese oxide that may be present.
This solution containing the zinc, and probably manganese,
must be separated from the precipitate produced by the
ammonia by filtration, the insoluble matter in the filter
washed with water containing a little ammonia, and the
washings so obtained added to the first strong filtrate. If
no manganese be present, ammonium sulphide may now be
added to the filtered liquid until it produces no further
white precipitate of zinc oxide. The liquid and precipitate
must now be allowed to stand in a warm place for about
572 THE ASSAY OF ZINC.
an hour, then filtered, and the zinc sulphide on the filter
washed with water containing a little ammonium sulphide.
After a few washings it is to be dissolved in dilute hydro-
chloric acid, and, if necessary, the solution filtered. To
the filtered solution is added excess of sodium carbonate ;
zinc carbonate is thrown down, which in its turn is col-
lected on a filter, washed, dried, separated from the filter,
ignited, and weighed. Four-fifths of its weight is metallic
zinc. If, by previous experiments by blowpipe or other-
wise, manganese were found to be present, the ammoniacal
solution containing the mixed oxides must be thus treated.
Excess of acetic acid is to be added to it, and a stream
of sulphuretted hydrogen gas passed through it until no
further precipitation takes place ; by this means the whole
of the zinc is deposited as sulphide, whilst the manganese
remains untouched in the liquid. The zinc sulphide is to
be collected on a filter and treated with hydrochloric acid,
&c., as just described.
ASSAY OF OKES OF THE SECOND CLASS.
Zinc silicates are not reducible by charcoal alone ; but
when in contact with substances which have the property
of combining with silica, they are reduced completely,
even at a moderate temperature. All the modes of assay
just described for ores of the first class apply to those of
the second, with the exception that the flux, instead of
being merely reducing, must have a true fluxing property
also : lime or magnesia -are good fluxes.
Wet Estimation of Zinc in Ores of the Second Class.
Ores of this class are best decomposed by strong hydro-
chloric acid with a small admixture of nitric acid. When
thoroughly decomposed, and the solution evaporated to
dryness, the residue is moistened with hydrochloric acid,
and treated exactly as described for Ores of the First Class.
ASSAY OF ORES OF THE THIRD CLASS. 573
ASSAY OF ORES OF THE THIRD CLASS.
In order to assay the substances containing sulphur
which belong to this class, they must be roasted, and then
treated as the ores of the first and second class. Zinc
sulphide may be roasted without difficulty ; and when
the operation is made with care, the roasted ore contains
neither sulphur nor sulphuric acid. The only precaution
necessary to observe is, that the heat must be carefully
regulated at first, in order to avoid fusion which might
take place, especially when a certain amount of iron sul-
phide is present. Towards the end the heat may be in-
creased to decompose any sulphate that may be formed.
Both a reducing and fusing substance must be added in
this case as in the last, in order to effect the fusion of
the gangue.
Wet Estimation of Zinc in Ores of the Third Class.
These ores are to be finely pulverised, treated with strong
nitric acid, at first with a gentle heat ; and, lastly, boiled
until the sulphur separates in bright yellow transparent
globules, as described under the Wet Assay of Copper
Ores of the Second Class. The solution so obtained
is to be evaporated to dryness, moistened with hydro-
chloric acid, and treated as described for ores of the first
class.
If ores of this class, or of either of the two former, con-
tain copper, they must be thus treated :
The ore is to be decomposed by an appropriate acid,
evaporated to dryness, moistened with hydrochloric acid r
water added, and the solution filtered. A current of sul-
phuretted hydrogen gas is now to be passed through the
solution until, even after violent agitation, it smells strongly
of it. It is now to be filtered, and the black precipitate
on the filter contains all the copper as copper sulphide,
that substance being insoluble in dilute acid, whilst in a
solution acidulated with either of the strong mineral acids
as nitric, hydrochloric, or sulphuric zinc is not at all
acted on by sulphuretted hydrogen. The solution, now
574 THE ASSAY OF ZINC.
freed from copper, is placed in an evaporating basin and
boiled for about a quarter of an hour ; nitric acid is then
added to per oxidise all the iron present, and the solution
allowed to cool. When cold, the zinc is separated by
means of ammonia, and the ammoniacal solution treated
as already described.
Assay of Cupriferous Blendes.
Mr. E. Monger has described the following rapid and
accurate process for the assay of cupriferous blendes :
Fifteen grains of the blende are taken and treated with
aqua regia and evaporated ; the residue is re-moistened
with hydrochloric acid, and re-evaporated, dissolved in
water with a drop or two of hydrochloric acid, 10 to
20 c.c. of ammonia added, and the iron filtered off and
washed.
The filtrate, which is now about 200 to 250 c.c. in
volume, is acidified with hydrochloric acid, heated, and
placed in a porcelain dish ready for buretting with ferro-
cyanide of O'Ol strength. To the solution of zinc and
copper add sodium sulphide solution enough to precipitate
the whole of the copper and leaving a little excess ; then
proceed with the buretting, with ferrocyanide and uranium
acetate as indicator.
In some experiments Mr. Monger took a non-cupri-
ferous blende and weighed out six samples, into three of
which he put some sulphate of copper solution, equal to
21 per cent, copper in the blende, and proceeded with
them as above explained, and obtained perfectly concordant
results.
FOURTH CLASS. ALLOYS.
The alloys of zinc with iron, copper, and tin may
be assayed by heating them to whiteness for about an
hour in a charcoal crucible with an earthy flux (calcium
silicate is the best), and weighing the resulting button :
the loss will be nearly equivalent to the quantity of zinc
present.
VOLUMETRIC DETERMINATION OF ZINC. 575
The Wet Estimation of Zinc in Substances of the Fourth
Class. These substances are treated precisely as described
under the heads Wet estimation of Zinc in First, Second,
and Third Classes.
VOLUMETRIC ESTIMATION OF ZINC.
Galetti 's Process. -A good method of volumetrically
estimating the amount of zinc in ores is given in the
' Zeitschrift fiir analytische Chemie ' for 1869, by Maurizio
Galetti. Chief Assayer at the Eoyal Assay Office, Genoa.
The following is a description of the process : Suppos-
ing zinc sulphide (blende) is to be assayed, about half a
gramme of the finely pulverised ore is to be treated with
concentrated nitric acid, and boiled to incipient dry ness,
until the sulphur left undissolved does not contain any
particles of undissolved ore. Then add strong hydrochloric
acid, and boil again until no nitric acid is left. Calamine
(zinc carbonate) should at once be acted upon with hydro-
chloric acid ; but, in order to make sure of the complete
oxidation of all the iron the ore may happen to contain,
it is best to add to the acid a few decigrammes of pure
potassium chlorate. After having boiled this solution for
a few minutes it is diluted with distilled water ; a large
excess of ammonia is added to the solution, which is then
boiled and slightly acidified with acetic acid. After brisk
agitation boil again for a few minutes, and then super-
saturate with ammonia. The liquid is then poured out of
the flask into a suitable glass vessel, and the flask is rinsed
out with a sufficient quantity of distilled water to bring
the bulk of the fluid up to half a litre. This having been
done, the fluid is very cautiously and gradually acidified
with dilute acetic acid, one part acid sp. gr. 1-07 to 10 of
distilled water. Any large excess of this should be avoided,
as the solution should only be very slightly acid. As soon
as the basic iron acetate has subsided, the precipitation of
the zinc by means of a standard solution of potassium
ferrocyanide may be proceeded with.
The ferrocyanide solution is made by dissolving 41-25
576 THE ASSAY OF ZINC.
grms. of the salt in as much distilled water as will make
the solution weigh exactly one kilogramme.
The presence of compounds of lead (as, for instance,
lead carbonate, sulphate, or sulphide) occurring along with
the ores of zinc does not interfere with the completeness
of the precipitation of zinc as zinc ferrocyanide. This even
holds good up to 10 per cent, of metallic lead. Since some
ores of zinc, especially calamine, often contain manganese,
it is best to add to the ammoniacal solution, before any
acetic acid is added, a few drops (from 2 to 4) of bromine,
in order to convert the manganese protoxide into proto-
sesquioxide, leaving the solution standing for twenty-four
hours after the addition of the bromine.
The ammoniacal solution of zinc chloride being colour-
less, there should be added to it, previous to cautious
acidification by means of dilute acetic acid, a few drops
of tincture of litmus, in order to more readily hit the
precise point of sufficient acidification, which is known by
the blue coloration changing to a rose-red.
The zinc ferrocyanide which is mixed with iron oxide
preserves its naturally white colour as long as the liquid
contains free zinc, but its colour changes to a greyish
white as soon as a very slight excess of the ferrocyanide
standard solution is present ; the liquid also then becomes
turbid, and the precipitate settles very slowly. By these
characteristic signs the end of the operation may be always
recognised. In order to make sure, the liquid should be
touched with a glass rod which has been just previously
moistened with a dilute solution of ammoniacal copper
nitrate ; this will have the effect of indicating any excess of
the ferrocyanide solution, by producing the more or less
intense colour characteristic of copper ferrocyanide. The
zinc solution should be at a temperature of from 40 to 50,
whereby the rapid subsidence of the zinc ferrocyanide is
promoted.
Filtration is not necessary, as the presence of the gela-
tinous silica (due to the decomposition of zinc silicates
occurring in the ores of that metal) does not interfere
with the correctness of this method of estimating zinc
quantitatively.
VOLUMETRIC ESTIMATION OF ZINC. 577
Fresenius * gives the following methods for the volu-
metric estimation of zinc :
1. SCHAFFNER'S METHOD^ MODIFIED BY C. KUNZEL^; AS EMPLOYED
IN THE BELGIAN ZINC- WORKS ; DESCRIBED BY C. GROLL.
a. Solution of the Ore and Preparation of the Ammoniacal
Solution.
Powder and dry the ore.
Take O5 grm. in the case of rich ores, 1 grm. in the
case of poor ores, transfer to a small flask, dissolve in
hydrochloric acid with addition of some nitric acid, by the
aid of heat, expel the excess of acid by evaporation, add
some water, and then excess of ammonia. Filter into a
beaker, and wash the residue with lukewarm water and
ammonia till ammonium sulphide ceases to produce a
white turbidity in the washings. The zinc oxide remain-
ing in the hydrated ferric oxide is disregarded. Its quan-
tity, according to Groll, does not exceed 0-3 0-5 per cent.
This statement probably has reference only to ores contain-
ing relatively little iron ; where much iron is present the
quantity of zinc left behind in the precipitate may be not
inconsiderable. The error thus arising may be greatly
diminished by dissolving the slightly washed iron precipi-
tate in hydrochloric acid, and adding excess of ammonia.
But the surer mode of proceeding is to add to the original
solution after evaporating off the greater part of the
free acid as above, and allowing to cool dilute sodium
carbonate nearly to neutralisation, then to precipitate the
ferric oxide with boiling sodium acetate, filter, and
wash. The washings, after being concentrated by evapo-
ration, are added to the nitrate, and the whole is then
mixed with ammonia till the first formed precipitate is
re-dissolved.
If the ore contains manganese provided approximate
results will suffice digest the solution of the ore in acids,
* 4th English edition, p. 653, published by Churchill and Sons.
t ' Journ. f. prakt. Chem.' 73, 410. t Ibid. 88, 486.
' Zeitschrift f. anal. Chem.' 1, 21.
P P
578 THE ASSAY OF ZINC.
after the addition of excess of ammonia and water, at a
gentle heat, for a long time, and then filter off, with the
iron precipitate, the hydrated manganese protosesquioxide
which has separated from the action of the air. The safer
course though undoubtedly less simple is, after sepa-
rating the iron with sodium acetate, to precipitate the
manganese by passing chlorine through, or by adding
bromine and heating.
If lead is present, it is separated by evaporating the
aqua regia solution with sulphuric acid, taking up the
residue with water and filtering ; then proceed as directed.*
bi Preparation and Standardising of the Sodium Sulphide
Solution.
The solution of sodium sulphide is prepared either by
dissolving crystallised sodium sulphide in water (about
100 grm. to 1,000-1,200 water), or by supersaturating a
solution of soda, free from carbonic acid, with sulphuretted
hydrogen, and subsequently heating the solution in a flask
to expel the excess of sulphuretted hydrogen. Whichever
way it is prepared, the solution is afterwards diluted, so
that 1 c.c. may precipitate about O'Ol grm. zinc. Prepare
a solution of zinc by dissolving 10 grm. chemically pure
zinc in hydrochloric acid, or 44-122 grm. dry crystallised
potassium and zinc sulphate in water, or 68-133 grm. dry
crystallised potassium and zinc sulphate in water, and
making the solution in either case up to 1 litre with water.
Each c.c. of this solution corresponds to 0-01 grm.
zinc. Now measure off 30-50 c.c. of this zinc solution
into a beaker, add ammonia till the precipitate is re-dis-
solved, and then 400-500 c.c. distilled water. Eun in
sodium sulphide as long as a distinct precipitate continues
to be formed, then stir briskly, remove a drop of the fluid
on the end of a rod to a porcelain plate, spread it out so
* Concerning the direct treatment of roasted zinc ores with a mixture of
carbonated and caustic ammonia, comp. E. Schmidt (Journ. f. prakt. Chew.
51, 257). By this treatment the zinc oxide, which was combined with car-
bonic acid, is dissolved, whilst that combined with silicic acid is, for the most
part, left undissolved.
VOLUMETRIC ESTIMATION OF ZINC. 579
that it may cover a somewhat large surface, and place in
the middle a drop of pure dilute solution of nickel chloride.
If the edge of the drop of nickel solution remains blue or
green proceed with the addition of sodium sulphide, testing
from time to time, till at last a blackish grey coloration
appears surrounding the nickel solution. The reaction is
now completed, the whole of the zinc is precipitated, and
a slight excess of sodium sulphide has been added. The
precise depth of colour of the nickel must be observed and
remembered, as it will have to serve as the stopping signal
in future experiments. To make sure that the zinc is
really quite precipitated, you may add a few tenths of a
c.c. more of the reagent, and test again ; of course the
colour of the nickel drop must be darker. Note the
number of c.c. used, and repeat the experiment, running
in at once the necessary quantity of the reagent less 1 c.c.,
and then adding 0'2 c.c. at a time till the end-reaction
is reached. The last experiment is considered the more
correct one. The sodium sulphide solution must be re-
standardised before each new series of analyses that is,
if it is kept in bottles containing air ; if, on the contrary,
oxygen is excluded by passing the air through an alkaline
solution of pyrogallic acid previously to its entering the
bottle, the solution would without doubt keep unaltered.
c. Estimation of Zinc in the Solution of the Ore.
Proceed in the same way with the ammoniacal solution
prepared in a as with the known zinc solution in b. Here
also repeat the experiment, the second time running in at
once the required number of c.c., less 1 of sodium sulphide,
and then adding 0*2 c.c. at a time till the end-reaction
makes its appearance. The second result is considered
the true one. There are three different ways in which
this repetition of the experiment may be made. You may
either weigh out at the first two portions of the zinc ore,
or you may -weigh out double the quantity required for
one experiment, and make the ammoniacal solution up to
1 litre, and employ -J litre for each experiment ; or, lastly,
r P 2
580 THE ASSAY OF ZINC.
having reached the end-reaction in the first experiment,
you may add 1 c.c. of the known zinc solution, which will
destroy the excess of sodium sulphide, and run in sodium
sulphide in portions of O2 c.c. till the end-reaction is
again attained. Of course, in this last process to obtain
the second result, you deduct from the whole quantity of
sodium sulphide used the amount of the same, correspond-
ing to 1 c.c. of the zinc solution.
If the ore contain copper, which frequently occurs in
the case of blendes, determine by a preliminary experi-
ment the number of c.c. of sodium sulphide which are
necessary to precipitate the copper, and, at the completion
of the zinc analysis, deduct them. In this case, let the drop
to be tested with nickel solution pass through a small
filter on its way to the porcelain plate, in order to avoid
the injurious influence of the copper sulphide on the nickel
reaction. If, however, the copper amounts to more than
2 per cent., remove it from the acid solution by sulphu-
retted hydrogen, evaporate the filtrate with nitric acid,
dilute, treat with ammonia, and estimate the zinc as
above.
In careful hands the error will, according to C. Kttnzel,
never exceed ^ per cent.
d. Further Modification of the Process.
To ascertain the point when the whole of the zinc is
precipitated and the sodium sulphide begins to predomi-
nate, Schafmer* employed flocks of hydrated ferric oxide,
which he produced by the addition of a few drops of ferric
chloride to the ammoniacal zinc solution, and which settled
at the bottom ; while Barreswil f used small pieces of white
porcelain, which were covered with ferric chloride, and
thrown into the ammoniacal zinc solution. Sodium sul-
phide is added till the flocks or the pieces of porcelain turn
black. In neither case is the end-reaction so exact as with
nickel solution.
* ' Journ. f. prakt. Chem.' 73, 410.
t ' Journ. de Pharm.' 1857, 431 ; ' Polyt. Centralbl.' 1858, 285.
VOLUMETRIC ESTIMATION OP ZINC. 581
With the help of lead-paper, however, the point may
be hit with great precision. Moisten a piece of white
filter-paper with solution of lead acetate, place it on a
layer of blotting-paper, drop some ammonium carbonate
upon it, so as to form a thin coating of lead carbonate, let
the blotting-paper absorb the excess of moisture, and then
spread the lead-paper on a porcelain plate. As soon as
you imagine the zinc to be nearly all precipitated, lay a
small piece of filter-paper on the lead-paper, and then dip
the end of a blunt glass rod in the fluid, and press it some-
what gently on the small piece of filter-paper. When the
sodium sulphide begins to be in excess, a brown spot forms
on the lead-paper. Fr.'Mohr* applies the lead reaction in
another manner. He makes an alkaline solution of lead
by warming together lead acetate, Eochelle salt, and solu-
tion of soda ; he first places a drop of this on filter-paper,
and then close by a drop of the precipitated zinc solution,
so that the circle formed by the spreading of the solution
to be tested may cut the circle of the lead solution. As
soon as the sodium sulphide begins to predominate, the
portion of the circumference of the lead circle which lies
in the other circle turns black.
2. H. SCHWARZ'S METHOD, f
Prepare an amnioniacal solution as in 1, a.
Heat gently, and mix with a moderate excess of am-
monium sulphide. Allow the precipitated zinc sulphide
to subside, then filter, using a tolerably large ribbed filter
of rapidly filtering paper, moistened with boiling water,
and warming the fluid to accelerate the operation, which
would otherwise require considerable time. Wash the
precipitate with warm water mixed with a little ammonia,
until the last drops no longer blacken a solution of lead
oxide in soda.
Transfer the filter with the precipitate to a beaker, add
* Mohr's * Lehrbuch der Titrirmethode,' 2 Aufl. 377.
t See his * Anleitung zu Maassanalysen,' Nachtrage, p. 29 (Brunswick).
Compare also v. Gellhorn ('Chem. Centralbl.' 1853, 291), who has made many
.analyses by Schwarz's method.
582 THE ASSAY OF ZINC.
a dilute solution of slightly acidified ferric chloride, cover
with a close-fitting glass plate, and let the mixture stand
for ten minutes ; then heat gently. Under these circum-
stances the zinc sulphide decomposes completely with the
ferric chloride to zinc chloride, ferrous chloride, and sul-
phur : Fe 2 Cl 6 + ZnS =ZnCl 2 + S + 2FeCl 2 .
Now add sulphuric acid, and heat gently until the
sulphur has agglutinated. Filter, and wash the filter, and
estimate the iron in the fluid as protochloride by per-
manganate,* 2 eq. iron correspond to 1 eq. zinc. If the
quantity of zinc sulphide is not very great, the filter may
be broken, and the zinc sulphide washed into a flask
which already contains the solution of ferric chloride.
The great objection to this method lies in the washing of
the zinc sulphide, which, as is well known, is a long and
troublesome operation. A possible loss of sulphuretted
hydrogen on mixing the zinc sulphide with ferric chloride
may be prevented by conducting the decomposition in a
flask, connected with a U-tube containing ferric chloride.
3. CARL MOHR'S METHOD, f
This method is based upon the following considera-
tions :
I. If a solution of zinc acetate, acidified with acetic
acid, is mixed with an excess of pota'ssium ferrocyanide,
the whole of the zinc is thrown down in the form of a
reddish-yellow precipitate of zinc ferrocyanide, Zn 3 (Cy 6 Fe 2 ).
II. If solution of potassium iodide is now added in ex-
cess, we have this decomposition : 2[Zn 3 (Cy 6 Fe 2 )] + 2KI +
2( A,HO) = 3[Zn 3 (Cy 3 Fe)] + 2(KO, A) + H 2 (Cy 3 Fe) + 21.
III. 1 eq. liberated iodine corresponds, accordingly, to-
3 eq. zinc.
IV. If potassium iodide is made to act upon zinc ferro-
cyanide in a neutral fluid, the liberated iodine acts upon
the potassium ferrocyanide present in that case, which
leads to the formation of a little potassium ferrocyanide ;
* Without doubt the ferric chloride might be replaced by ferric sul-
phate, by which means the presence of hydrochloric acid would be avoided.
t Dingler's ' Polyt. Journ.' 148, 115.
CARL MOHRS METHOD. 583
the remaining free iodine, therefore, will not indicate, with
accuracy, the quantity of zinc present. But whereas the
reaction actually takes place in acid solution of zinc ace-
tate, as above directed, it may be assumed that potassium
acetate and free hydroferrocyanic acid are formed ; and as
iodine exercises no appreciable action upon the latter sub-
stance, the iodine liberated in the process indicates, with
tolerable accuracy, the amount of zinc present.
The process is as follows :
Treat the ore with aqua regia, as in 1, a, and drive off
the greater part of the free acid ; nearly neutralise with
sodium carbonate, add sodium acetate in excess, boil,
filter, and wash with boiling water mixed with a little
sodium acetate. The solution is free from iron : it con-
tains the whole of the zinc, but, in presence of manganese,
also the whole of the latter metal. Hence the process is
not applicable in the presence of manganese.
Mix the solution of zinc, prepared as directed, with
potassium ferrocyanide in slight excess, i.e. until a
sample of the clear supernatant fluid gives a blue precipi-
tate with a ferrous salt. Then add a sufficient quantity
of potassium iodide. The fluid acquires a brown colour,
in consequence of the liberation of iodine ; the white pre-
cipitate of zinc ferrocyanide is suspended in the brown
fluid,
Now estimate the free iodine by means of sodium
hyposulphite, and calculate 3 eq. zinc for each eq. iodine.
The results obtained by C. Mohr are very satisfactory.
The method can be employed only if the acetic acid solu-
tion contains no other heavy metal besides zinc, and, more
particularly, no manganese.
For estimating the value of zinc powder, J. Drewson
proposes the following method :
He prepares two solutions, the one of pure fused
potassium dichromate say 40 grms. per 1,000 c.c. and
the other of crystalline ferrous sulphate, about 200 grms.
in 1,000 c.c. The iron solution must be strongly acidu-
lated with sulphuric acid to prevent oxidation. In order
to find the respective value of the two liquids, 10 c.c. of
584 THE ASSAY OF ZINC.
the iron solution are measured into a beaker, a little sul- .
phuric acid is added, and the other solution is dropped
in from a burette until a drop of the mixture is no longer
turned blue by potassium ferrocyanide. About 1 grm. of
the zinc powder is then weighed, placed in a beaker with
100 c.c. of the chromic solution ; 10 c.c. of dilute sul-
phuric acid are added, the whole is well stirred, 10 c.c.
more of the sulphuric acid are added, and allowed to act
for about a quarter of an hour, with diligent stirring.
When everything is dissolved except a small insoluble
residue, an excess of sulphuric acid is added, and 50 c.c.
of the iron solution, in order to reduce the greater part of
the excess of chromate ; more of the iron solution is then
added from a burette, till a drop displays a distinct blue
reaction with ferrocyanide, and the mixture is then titrated
back with chromate till this reaction disappears. From
the total number of c.c. of the iron solution consumed,
the quantity is deducted which corresponds to the ferrous
solution employed. The chromate contained in the re-
mainder, if multiplied by 0'66113, shows the metallic zinc
contained in the powder.
In order to separate copper from zinc by a single
precipitation with sulphuretted hydrogen, G. Larsen
(' Zeitschrift fur analyt. Chemie,' 1878, p. 312) passes
sulphuretted hydrogen into the solution, filters, washes
the precipitate first with hydrochloric acid of sp. gr.
1*05, through which sulphuretted hydrogen has been
passed, and then with pure sulphuretted hydrogen water.
Both the precipitation and the washing are effected by
heat.
Brass, pinchbeck, false gold-leaf, bronze not containing
tin, &c., are dissolved in nitric acid as directed for alloys
of silver and copper.
The acid liquid is diluted with water and a current of
sulphuretted hydrogen is introduced, the vessel being kept
covered with a glass plate. This is continued till the pre-
cipitate has subsided and the liquid becomes clear and
colourless. The precipitate copper sulphide is poured
upon a filter, and the filtrate collected in an evaporating
SEPARATION OF COPPER FROM ZINC. 585
basin. To prevent oxidation the filtration must be per-
formed rapidly, and the air excluded as far as possible by
keeping the funnel and the beaker covered with a glass
plate. The filter should be kept constantly full, till it
has received all the precipitate.
All particles adhering to the sides of the beaker and
to the gas delivery tube are washed into the filter by
means of a feather and the washing- bottle. The precipi-
tate is then washed on the filter moderately with cold
water, a few drops of sulphuretted hydrogen water being
added each time. The precipitate is dried, transferred to
a tared porcelain crucible, upon the lid of which the filter
is reduced to ashes, which are added to the precipitate.
The whole is mixed with a few centigrms. of flowers of
sulphur, and the crucible is covered with a peculiar lid, in
which a bent porcelain tube opens, through which dried
hydrogen is introduced into the crucible, which is then
placed over a lamp, and as soon as the common air is
expelled from the apparatus it is heated to redness. The
contents are then let cool in the current of hydrogen.
Thus is obtained pure cuprous sulphide, Cu 2 S, which is
weighed, and from its weight that of the copper is calcu-
lated.
The filtrate containing the zinc is concentrated by
evaporation to expel excess of sulphuretted hydrogen.
Crystallised sodium carbonate is then gradually added,
with stirring till the effervescence ceases, the precipitation
is effected, and the liquid has an alkaline reaction. In
order to lose nothing by spirting, the beaker is kept
covered with a glass plate, which is afterwards rinsed into
the beaker by means of the washing-bottle. The liquid is
brought to a boil, filtered, and the precipitate well washed
with boiling water. It is dried and strongly ignited, when
it becomes converted into zinc oxide. The filter is burnt
on the lid of the crucible. The zinc oxide is weighed, and
from it the zinc is calculated.
If lead is also present, the solution of the metals is
placed in an evaporating dish, and mixed with so much
moderately concentrated sulphuric acid as to convert the
586 THE ASSAY OF ZINC.
lead into sulphate and leave an excess, which is afterwards
chiefly expelled by evaporation. When cold, water and
a little alcohol are added ; the precipitate is poured upon
a filter, dried at 120, washed with cold alcoholic water,
dried, and weighed.
The filtrate is treated with sulphuretted hydrogen, as
for the estimation of copper, &c. (Rammelsberg).
587
CHAPTEE XV.
THE ASSAY OF MERCURY.
MERCURY is found in the native or metallic state, and as
sulphide or cinnabar :
Native mercury, Hg.
Mercury sulphide, cinnabar, Hg 2 S.
Bituminous mercury sulphide.
There are other minerals of mercury met with, but
hitherto not in sufficient quantity to be worked for the
metal. They are
Zinciferous mercury subsulphide.
Zinciferous mercury sulphide.
Mercury selenide.
Mercury subchloride.
Mercury iodide.
Silver amalgam (see Silver).
Assay of Mercurial Ores. The estimation of mercury
is generally made by distillation. When the mercury
is present in the form of native mercury, or mercury
oxide, it is distilled without any addition. The ore (say
from 500 to 1,000 grains) is placed in an iron or earthen-
ware retort, which is set over a suitable fire, and the
heat raised gradually, and kept up, until the whole of
the mercury has passed over. The mercury which passes
over is collected either in the neck of the retort or in a
receiver fitted for that purpose such as a glass flask kept
cool by affusion with water. When but a small quantity
is operated upon (say 150 to 200 grains), it is most con-
venient to use a glass retort, or bent tube retort, heating
it gradually over a charcoal fire, taking care to keep the
588 THE ASSAY OF MERCURY.
upper part so hot that no metallic mercury may adhere to
it. It must be heated nearly to the melting-point of the
glass, and until all the mercury has come over.
When the operation is finished, the neck is cut off,
weighed, the mercury detached, and weighed again : the
loss of weight is the amount of mercury. Or the metal
may be detached by means of a feather, and allowed to
fall into a basin of water, which, if heated for a few
seconds, will cause the mercury to collect into one glo-
bule : the water may be decanted, and the mercury dried
at the ordinary temperature, and weighed.
The mercury wholly condenses in the neck of the
retort, under the form of a metallic dew. Some may by
chance pass off; but in order to prevent such an occur-
rence, the beak of the retort is plunged into water, or a
loose plug of linen, moistened with water, is introduced
into the neck, the end of which is plunged into water, by
which means the neck of the retort is kept constantly cool,
and the mercury is found deposited on the linen, from
which it may be detached by shaking in water.
When large quantities of substances containing mer-
cury are operated upon, it is necessary to heat very
strongly towards the end, in order that the centre of the
mass may receive a sufficient amount of heat to effect its
decomposition. Naked glass retorts cannot be used, and
either coated glass or porcelain retorts must be employed.
In the large way, as in the distillation of amalgams, &c. 5
cast-iron retorts are used.
As before stated, all substances containing mercury,
either in its metallic state or as oxide, are distilled without
addition, but with the others it is necessary to employ
some reagent which will separate and retain the sulphur,
selenium, &c. ; this reagent may be a metal, as iron,
copper, or tin ; or black flux, or a mixture of quicklime
and charcoal : iron filings are most often used. For
cinnabar about 50 per cent, of iron filings is required, in
order to prevent any of it being sublimed ; the actual quan-
tity required is only about 24 per cent., but an excess is
necessary, in order, as before stated, to prevent loss : 50
ASSAY IN THE DEY WAY. 58$
per cent, of iron filings may be employed for the selenides,
&c. When black flux is used, from about 50 to 70 per
cent, is employed. Caustic lime may be employed in the
proportion of 30 per cent, mixed with 30 per cent, of its
weight of charcoal. After the ore to be assayed is care-
fully mixed with any of the above fluxes, it is always
advisable to cover it, w r hen in the retort, with a thin
layer of the flux employed, in order to avoid all chance
of any loss.
In estimating mercury by distillation it is necessary,,
especially if the metal is in the state of chloride or sul-
phide, to take certain precautions, without which a
portion of the sulphide or chloride would volatilise
without decomposition. H. Eose gives the following
directions for carrying out the operation : Introduce
into a glass tube capable of resisting fusion, closed at one
end, and measuring from 35 to 50 centimetres in length,
a column of sodium bicarbonate, then one of quicklime,
and then a w r ell-blended mixture of the mercurial com-
pound and quicklime, and, finally, a column of quick-
lime. The open end of the tube is drawn out and bent
round so as to enter a small flask containing water. The
tube is heated as if for an organic analysis, commencing
at the open end and finishing with the sodium bicar-
bonate. The operation ended, cut the bent end of the
tube. Collect all the mercury in the flask, dry with paper,
and afterwards over sulphuric acid, and then weigh it.
The quicklime should not be replaced by hydrate of
lime. That would occasion all the inconvenience of an
analysis of a sulphuretted combination of mercury. The
water acting on the calcium sulphide would form sul-
phuretted hydrogen, which, by dissolving in the water of
the receiver, would, in time, transform a portion of the re-
duced mercury into sulphide. It is advisable in some cases
to replace the sodium bicarbonate by magnesium carbonate-
Combinations containing mercury iodide are riot en-
tirely decomposed when treated as above. Biniodide and
protoiodide are condensed in the extremity of the tube
simultaneously with the metallic mercury.
590 THE ASSAY OF MERCURY.
To analyse these combinations recourse must be had
to metallic copper, the operation being similar to that
with quicklime.
Berthier, who experimented with an ore containing
arsenic, realgar, &c., and cinnabar from Huanca-Velica,
in Peru, found, after very many fruitless experiments, the
following method best adapted to its examination for
mercury :
The ore was heated in a retort with four to five times
its weight of litharge. From the litharge, the arsenic
sulphide, &c., a fusible slaggy mass was formed, while
the cinnabar was decomposed into sulphurous acid and
metallic mercury. The mercury volatilised completely at a
moderate heat, and collected in the fore part of the neck
of the retort and in the receiver. The single precaution
which must be observed for the success of the assay con-
sists in only gradually and moderately heating the clay or
glass retort, in order to prevent its being perforated by
the corroding effect of the litharge before the operation
is ended.
If the assay sample is extremely poor in mercury, the
ordinary assay method becomes somewhat inconvenient
and uncertain, on account of the large quantity which
must then be subjected to distillation in the assay. For
this case Berthier found it more appropriate to digest the
assay sample with aqua regia, wash it thoroughly, evapo-
rate the whole mass of fluid to dryness, and then treat the
dry mass, which contains all the mercury as chloride,
further in the dry way. He found that if mercury
chloride (corrosive sublimate) is heated with litharge, it
volatilises without undergoing ' any change. If, besides
the litharge, coal-dust is also added, or if instead of it
metallic lead is used in great excess, the chloride is
reduced to subchloride, which volatilises, but not the
smallest drop of mercury is thus produced. The best
reducing agent for the mercury chloride contained in the
dry mass is black flux, of which three parts by weight
are used. Since the mass to be subjected to distillation has
been greatly diminished by the treatment of aqua regia,
ASSAY IN THE DRY WAY. 591
and the subsequent evaporation, and no high heat is now
required for the decomposition, the distillation may be
performed in a glass retort. When the gangue in the
poor ore is calcium carbonate, all the lime must be dis-
solved out by moderately strong acetic acid before the
treatment with aqua regia.
By this method the smallest trace of mercury in an ore
or amalgamation product can be shown and estimated
by its weight.
EschkcCs process for assaying mercury ores is given
as follows in the 4 Chemical News ' for July 1872. The
method may be used for cinnabar, mercuriferous fahlerz,
&c. The ore should be weighed in a balance turning
with one milligramme. The quantity of ore for the assay
varies according to its richness, as follows :
Ore containing up to 1 per cent. .. .10 grammes
1 10 ,, . . . o
10 ,,30 . . .2
,, over 30 . . .1 gramme
The ore is introduced into a porcelain crucible, the edge
of which has been ground flat, and mixed with about
half its weight of clean iron filings by means of a glass
rod, and is then evenly covered with a layer of iron filings
about J to f inch thick. A concave cover, made of fine
gold, about two inches in diameter and 12 to 15 grammes
in weight, is now placed on the crucible after having been
carefully weighed ; the concavity is filled with distilled
water and the crucible placed on a triangle and heated for
ten minutes by a Bunsen burner or Argand spirit-lamp,
during which time the mercury is volatilised and deposits
itself on the gold. The gold cover is then removed, the
water poured off, and the mirror of mercury on the convex
side washed with alcohol from a wash -bottle. After being
dried in the water-bath, the cover is allowed to cool tho-
roughly, and is then weighed in a balance turning with -J
milligramme with 50 grammes in the pan. The increase
of weight gives the quantity of mercury in the ore.
During the weighing the cover is placed on an empty
592 THE ASSAY OF MERCURY.
porcelain crucible. The mercury is then driven off by
heating the cover gently in the flame of a Bunsen burner
or a spirit-lamp, in a place where there is a good draught,,
and the empty crucible and cover are subjected to a
second weighing as a check. In order that the assay
should succeed the following conditions must be fulfilled :
The cover must fit closely, so as to avoid loss of
mercury, and must be deep enough to hold a sufficient
quantity of water to keep it cool. The iron filings must
be free from grease, which would prevent the proper
formation of the mirror ; the washing with alcohol must
not be omitted, as it removes all the bituminous sub-
stances which spoil the mirror, and assists the drying ;
it must be dried in a water-bath for two or three minutes,
cooled in the desiccator, and weighed when fully cool.
When assaying rich ores the alcohol used in washing
the cover must be collected, as it may contain a little
amalgam ; it must be poured into the concavity of the
cover, which will take up any little globules of mercury.
The most exact results are obtained in the case of poor
ores containing less than 10 per cent.
Assay for the Amount of Cinnabar in an Ore. The ore
to be assayed is distilled, without addition, in a glass retort,
and the sublimed cinnabar collected and weighed. The
ores containing mercury combined with sulphur are often
mixed with bituminous matters and calcium carbonate :
then, when an assay is to be made for cinnabar, it often
happens that a portion of it is decomposed, either by the
carbon present or by the aid of the bituminous matter and
lime, and a little metallic mercury is driven off with the
cinnabar. In this case, having weighed the mixture of
cinnabar and mercury, the mixture is treated with nitric
acid, which dissolves only the latter, and pure cinnabar
remains, the weight of which is taken, and the quantity of
mercury dissolved ascertained by the difference, and from
that the quantity of cinnabar calculated which that quan-
tity of mercury would yield. Every 86 parts of mercury
furnish about 100 of cinnabar.
If the gangue of the ore be fixed in the fire, the assay
ASSAY IN THE WET WAY. 593
may be made by mere calcination, and the loss of weight
will correspond either to the metallic mercury, oxide, or
sulphide it may contain.
For the estimation of mercury in the wet way in
its usual ores, which are mixtures of cinnabar with lime-
stone, clay, iron oxide, and bituminous matters, the sample
is first treated with hydrochloric acid, which dissolves the
lime, &c< The liquid is then poured off and the residue
digested with aqua regia, when the mercury is dissolved
as chloride. The solution is filtered off from the insoluble
residue, the greater part of the free acid removed by
evaporation, diluted and heated with a solution of sul-
phurous acid, to reduce the ferric oxide to the ferrous con-
dition. Without previous filtration it is then treated with
sulphuretted hydrogen, the precipitate of mercury sulphide
is brought upon a weighed filter, and dried at 100-110 C.
Any free sulphur present may be removed by digestion
with a solution of sodium hyposulphite. The mercury
sulphide may be afterwards tested for the presence of
other volatile metals.
If it is intended to weigh mercury as mercurous
chloride (calomel) after reduction with stannous chloride,
the ore should be dissolved not in aqua regia, but in a
mixture of hydrochloric acid and potassium chlorate.
Free chlorine is expelled by heat, and the liquid is placed
in a flask and mixed with a clear solution of stannous
chloride, to which hydrochloric acid has been added, and
the whole is boiled for a few moments. When cold, the
liquid is decanted off, the mercury rinsed in a crucible,
washed with acid water, and dried in the desiccator.
In amalgams mercury is generally estimated by
placing the sample in a porcelain crucible and heating in
a current of hydrogen gas, when the mercury is volatilised.
(Rammelsberg . )
(This method, however, is not applicable to the amal-
gams of the alkaline metals.)
QQ
594 THE ASSAY OF MERCURY.
Electrolytic Estimation of Mercury.
Don Luis de la Escosura has described two processes
for the estimation of mercury by electrolysis. In the first
process he operates upon about 15 grains of ore, treated
with 20 c.c. of water mixed with 10 to 15 c.c. of hydro-
chloric acid. The whole is gently heated in a porcelain
capsule, and when the liquid is on the point of boiling
from 15 to 30 grains of potassium chlorate in powder are
added by small portions. When the action is over 50 c.c.
of water are added to the liquid, and it is boiled afresh
until the odour of chlorine is no longer perceptible. At
this moment 20 to 30 c.c. of a saturated solution of
ammonium sulphite are added, boiling again for a few
minutes ; the capsule is then withdrawn from the fire, and
let settle. Care must be taken to supply the water which
is lost by evaporation. The addition of the ammonium
sulphite is to precipitate selenium and tellurium before
electrolysing the liquid. After settling for half an hour
the liquid is filtered, and the insoluble residue is washed.
The volume of the filtrate should be about 200 c.c. This
liquid is put in a glass beaker, and the electrodes are
introduced. These are plates of metal ; the one may be
of platinum, but the other which communicates with the
zinc pole of the battery must be of pure gold. The plates
are suspended vertically in the liquid. The battery con-
sists of two Bun sen elements. In twenty-four to thirty
hours the operation is .completed ; the mercury is depo-
sited upon the gold plate. The increase of weight in this
corresponds to the quantity of mercury contained in the
weight of metal taken.
Second Process. Direct Electrolysis of the Mercury Ore
without Previous Solution. The ore, in fine powder, is
placed in a platinum capsule, in a mixture of water,
hydrochloric acid, and ammonium sulphite. For a 10 per
cent, ore the following proportions are employed : Ore,
6 grains ; acid, 10 c.c. ; water, 20 c.c. ; sulphite, 20 c.c. ;
so as to make up about 120 c.c. These proportions are
ELECTROLYTIC ESTIMATION OP MERCURY. 595
modified according to the richness of the ore. The
platinum capsule, 3^ inches in diameter, is placed on
a support, and a disc of gold is plunged into it ;
soldered to a rod of gold, communicating with the
zinc pole of the battery, the capsule is connected to
the other pole. In twenty-four hours the operation is
complete : the mercury is deposited upon the disc of gold,
the increase of weight indicating the quantity of the
metal.
In preference to a Bunsen battery, Don Luis de la
Escosura uses six elements formed of a cylindrical vessel
of glass, 0*1 metre in diameter and 0*16 metre in height,
in which a zinc disc is suspended by means of two copper
wires at two-thirds of the height of the glass above the
bottom. At the centre of the zinc disc is a copper wire
bent twice at right angles, so that the other extremity of
the wire plunges into the glass vessel of the next element.
To increase the surface of contact it is well to make the
copper wire end in a spiral. The copper wire immersed
in the liquid, except the spiral, is coated with an insu-
lating substance.
The glass is filled with plain water after the crystals of
copper sulphate have been introduced, and in a few hours
the battery is ready for use.
This second process, though much simpler than the
first, is strictly accurate, even in the case of very poor
ores containing less than 0*1 per cent, of mercury. The
accompanying impurities do not interfere.
VOLUMETRIC ESTIMATION OF MERCURY.
The process we have found most trustworthy is that of
M. J. Personne, described in the ' Comptes Eendus,' Ivi. 63,
as follows. The author says : ^
' The process at which I have arrived, after many
fruitless attempts, is founded on a well-known fact that
a combination of mercury iodide with potassium iodide,
forming the double iodide, gives a colourless solution.
Thus, two solutions in equal quantities, one containing one
Q Q 2
,596 THE ASSAY OF MERCURY.
equivalent of mercury bichloride, the other two equiva-
lents of potassium iodide, being mixed, by pouring the
mercurial solution into that of the potassium iodide, mer-
cury iodide will be produced by the contact of the two
solutions, which dissolves in proportion to its formation,
until the mercurial solution added is equal in volume to
that of the alkaline iodide used. The slightest excess of
bichloride causes the formation of a persistent red pre-
cipitate, giving the liquid a very perceptible red tint
even by artificial light. This coloration, which indicates
that the saturation is complete, gives to this mode of
estimation a precision and nicety quite as great as that
of litmus, used to ascertain the saturation of an acid by
a base. The mercurial solution must always be poured
into the alkaline iodide not the alkaline iodide into
the mercurial solution; otherwise, though the last reac-
tion may be the same, it is impossible to obtain exact
results, because the mercury iodide produced, not being
brought simultaneously with its formation (in a nascent
state) into contact with the alkaline iodide with which it
is to combine, becomes sufficiently cohesive to dissolve
but slowly in the potassium iodide. Thus, in operating
with the same liquids, the quantity of alkaline iodide
which must be added to dissolve the mercury iodide pre-
cipitated varies according to the time employed in effect-
ing the estimation, and that in considerable proportions.
I have no doubt that it is through operating in this way
that potassium iodide has hitherto been rejected as a
medium for the exact estimation of mercury.
' Two normal liquids are necessary to effect this esti-
mation.
6 1. Normal Standard Solution of Potassium Iodide.
Obtained by dissolving 3 3 '20 grammes of pure potassium
iodide in water enough to make 1 litre of solution. 10
cubic centimetres of this solution represent O'l gramme of
metallic mercury.
4 2. Normal Standard Solution of Mercury Bichloride.
Prepared by dissolving 13*55 grammes of mercury bi-
chloride in water, so as to make 1 litre of solution. The
VOLUMETRIC ESTIMATION OF MERCURY. 597
solution of mercurial salt is facilitated by the addition of
5 equivalents, or 30 grammes, of sodium chloride, which
has no influence on the reaction, like all neutral alkaline
salts ; 10 cubic centimetres of this solution also represent
O'l gramme of mercury. Of these 10 centimetres, divided
into 100 parts, each division represents O'OOl gramme of
mercury. This mercurial solution serves to test the purity
of the alkaline iodide solution or to give the standard of
an unknown solution.
' Liquids ten times more diluted may be prepared
without injuring the nicety of the reaction or the exact-
ness of the results ; fractions of a milligramme may thus
be estimated.
' The estimation is effected in the following manner :
10 cubic centimetres of a normal solution of iodide being
o
measured into a small saturating vessel, pour into it,
constantly shaking the vessel, the solution of bichloride,
measured in Guy-Lussac's burette. If the two liquids are
pure, it will require exactly 100 divisions of the burette
before the light red tint appears in the saturated liquid to
indicate the close of the operation. When the mercurial
solution is weak a proportionally larger quantity must be
added. If, on the other hand, it is too strong, less must
be added. As will be perceived, this is very similar to
the chlorometric process.
' This new method of estimating mercury being appli-
cable only to bichloride, it became desirable to extend its
application to a greater number of mercurial compounds,
if not to all. This side of the question presented diffi-
culties not easily resolved in a satisfactory manner. It
was, in fact, necessary to transform all the mercurial
compounds into a perfectly neutral solution of bichloride.
I was obliged to set aside successively the use of aqua
regia, and even of hypochlorous acid. The great volatility
of mercury bichloride, even in a boiling solution, caused
too great a loss. M. Eivot's process that is to say, the
action of chlorine in a solution of hydrate of potash or
soda is perfectly successful. Take, for instance, the
estimation of mercury in cinnabar. Eeduce one gramme
598 THE ASSAY OF MERCURY.
of cinnabar to a fine powder. Weigh it on paper, and
introduce it into a matrass. Pour into the matrass 20
cubic centimetres of a caustic soda solution, with which
mix the paper and its contents by quickly shaking ; then
send a current of chlorine, which need not be washed,
into the liquid. The action of the chlorine produces a
slight heat, which is gradually brought to boiling-point,
by which time all the matter will have disappeared. To
insure success, the temperature must be carefully managed
at the commencement. If it is raised too quickly, part of
the matter remains undissolved. The solution being com-
plete and saturated with chlorine, it is kept boiling long
enough to expel all the excess of chlorine. The boiling
may be prolonged without incurring any loss of bichloride,
which is not volatile in presence of alkaline chloride.
The solution when cooled is poured into a graduated
tube. The matrass as well as the tube for conducting the
chlorine is washed two or three times with water, and
the washing added to the original liquid, so as to form 100
cubic centimetres of solution. I effected the estimation
with the standard solution of iodide, of which 10 cubic centi-
metres represent 1 gramme of mercury. To saturate these
10 cubic centimetres it required 115 divisions of the chloro-
mercurial solution. These 115 divisions contain then 0*1
gramme of mercury. Now, as all the mercury contained in
the analysed cinnabar is spread through the 10,000 divi-
sions of solution, we have the quantity of mercury found
by the experiment by means of a simple proportion.'
Mr. G. Attwood gives the following process for the
quantitative blowpipe assay of mercury. The compounds
to be assayed may be divided into three classes. Class A,
containing metallic mercury, cinnabar, tiemanite, sub-
oxide, protoxide, and mixed sulphides. Class B, calomel,
corrosive sublimate, and iodide of mercury. Class C 5
amalgams of gold, silver, copper, lead, zinc, tin, &c.
Class A. 10 to 20 grains of the ore, finely powdered
and passed through a sieve, 2,000 holes to the square inch,
are mixed with 5 to 10 times their weight of powdered
litharge and distilled over a spirit-lamp in a small glass
BLOWPIPE ASSAY OF MERCURY. 599
retort, 1^ inch long and J inch in diameter. To this
retort is fitted, by means of a cork, a glass tube, slightly
curved, 2^ inches long and fths of an inch in diameter.
The end of this tube dips under water contained in a
small porcelain crucible. The operation lasts only a few
minutes. The mercury is carefully collected from the
glass tube and crucible. The retort is broken' up and its
contents carefully powdered and examined by a lens for
mercury. The globules are then united by gently warm-
ing under water, and the dry mercury weighed.
Class B. A quantity of the finely powdered ore, equal
to 10 grammes, is mixed with three times its volume of
potassium oxalate and one volume of potassium cyanide.
The apparatus closely resembles that used in Class A, but
the retort has a small bulb.
Class C. These amalgams are sometimes powdered
with difficulty, and it is often advantageous to add a
known weight of pure mercury, so as to render them
semi-fluid before distilling. 10 to 30 grammes of the amal-
gam are usually taken for an assay. A turned steel retort
is used for distillation, which is effected in a small charcoal
furnace heated by a blowpipe flame ; the head of the
retort is accurately ground to fit over the body. The
retort, including the cup and cap, is 1 inch high ; the
neck of the cap is 2 inches long. The paper contains
full-size illustrations of the different retorts, &c., which
are made by Casella. (See 'Journal of the Chemical
Society/ 1879.)
GOO
CHAPTEE XVI.
THE ASSAY OF SILVER.
ALL argentiferous substances may be divided into two
classes, as follows :
CLASS I. Minerals containing silver :
Silver glance (AgS), containing 87 per cent, of Ag.
Brittle silver ore (6AgS.SbS 3 ), containing 70-4 per cent, of Ag.
Light red silver ore (3AgS,AsS 3 ), containing 65-4 per cent, of Ag.
Dark red silver ore (3AgS,SbS 3 ), containing 59 per cent of Ag.
Light and dark fahlerz (argentiferous grey copper ore), containing from
5-7 to 18-31-8 per cent, of Ag.
Argentiferous copper sulphide (Cu 2 S,AgS), containing 53 per cent, of Ag.
Polybasite 9(Cu 2 S,AgS) + (SbS 3 ,AsS 3 ), containing 72-94 per cent of Ag.
Slags.
Cupel bottoms.
Dross.
Litharge, &c.
CLASS II. Metallic silver and alloys, either native or
otherwise.
General Observations on the Assay of Ores and Substances
of Class No. 1.
In order to separate silver from this class of sub-
stances, an alloy of the precious metal with lead must
be formed. The different methods by which this object
can be attained are the following : first, fusion with a
reducing flux ; secondly, fusion with oxidising reagents ;
thirdly, scorification.
All substances containing lead in the state of oxide,
such as carbonates, phosphates, &c., are fused directly
with a reducing flux, as also are slags, old cupels, litharge,
&c. All plumbiferous sulphides, &c., containing silver,
THE ASSAY- OF SILVER. 601
are assayed as for lead by the processes already pointed
out, taking care to follow the method which gives the
largest proportion of lead.
All argentiferous minerals containing copper may be
assayed as copper ores ; because an alloy of copper and
silver can be cupelled by means of lead.
In making assays of silver with lead or copper, it is
sometimes necessary to commence the operation by roast-
ing the ore ; under other circumstances, also, argentiferous
matters are roasted.
There is nothing very particular to be observed in
this roasting ; the temperature alone requires attention by
managing well at the commencement of the operation, in
order to avoid softening, and especially to avoid a very
rapid disengagement of arsenical vapours, because a very
considerable amount of silver may be lost by that means.
All substances which contain reducible oxides are
fused with a reducing flux, as also those from which char-
coal separates metals which alloy with lead, or metals
which do not hinder the process of cupellation ; but it is
necessary to add to the reducing flux a certain proportion
of litharge, in order to produce metallic lead, with which
the silver may alloy. A mixture of metallic lead and any
suitable flux may be substituted for that of litharge and a
reducing flux ; but the latter is preferable, because the
lead produced is uniformly diffused throughout the whole
mass of flux, &c., not allowing a particle of silver to escape
its action.
The reducing agent employed in nearly all assays is
charcoal, either in its ordinary state, or as it is found in
black flux. Starch and other analogous substances may
be, as before mentioned, substituted for it : crude argol is,
however, the best reducing agent. The portion employed
must be varied according to circumstances, so that the
silver -lead produced be not too rich, or that too great a
proportion of lead be reduced. If the silver-lead be too
rich, much of the precious metal may be lost in the slag,
and if too great a quantity of lead be produced, silver is
again lost, owing to the long exposure to the fire during
602 THE ASSAY OF SILVER.
cupellation ; and, indeed, this is the most fruitful cause of
loss, for more is lost in this manner than by having too
little lead produced. In order to know the right propor-
tions, the following data will serve as a guide : 1 part of
charcoal reduces about 30 parts of lead from litharge, and
1 part of black flux reduces about 1 part of lead.
The fluxes employed in this kind of assay are litharge,
black flux, potassium or sodium carbonate, and borax.
Litharge is an exceedingly convenient flux, because it
occupies very little room, and fuses without bubbling, pro-
ducing very liquid scoriae with nearly every substance.
Experiment has shown that nearly all argillaceous, stony,
and ferruginous substances fuse very well with from 8 to 12
or more parts of litharge. If from ^ to 1 part of black
flux, or 5^- to gV of charcoal, be added to 1 of ore, from
J a part to 1 part of silver-lead will be produced.
Black flux is employed in the fusion of all substances
containing a large proportion of alumina, or in which lime
is the predominant substance from 2 to 3 parts of this
flux generally suffice : 1 part of litharge is added to the
assay, which is wholly reduced, producing nothing but
lead.
Potassium or sodium carbonates produce exactly the
same effects as the alkali of the black flux. A certain
quantity of charcoal must, in this case, be added to the
assay.
Schlutter fuses the poor refuses of goldsmiths' work-
shops, mixtures of fragments of crucibles, glass, &c., with
2 parts of potassium carbonate, when they are very earthy,
and with 1 part only when they contain much glass, adding,
at the same time, to the mixture a little litharge and
granulated lead.
Borax has, like litharge, the advantage of being a
universal flux ; it is useful especially for the fusion of sub-
stances containing much lime ; but it is necessary to take
great care in the assay, in order to avoid the loss which its
boiling up might occasion. This only applies, however, to
its use in its ordinary state ; if previously fused that is,
used as glass of borax no particular care need be taken.
FUSION WITH OXIDISING REAGENTS. 60S
FUSION WITH OXIDISING EEAGENTS.
Litharge. The oxidising agents employed in the assay
of argentiferous substances are litharge and nitre. Litharge
attacks all the sulphides, arsenio-sulphides, &c., and
oxidises nearly all the elements, excepting silver, when
employed in sufficient quantity ; and a quantity of lead
equivalent to the oxidisable matters present is reduced, so
that there results from the assay a slag containing an
excess of lead oxide, and an alloy of lead and silver, very
little contaminated with foreign metals, if no copper be
present, and which can be submitted directly to cupella-
tion. This method of assay is exceedingly convenient and
quick.
The pulverised mineral is well mixed with litharge
and the mixture placed in a crucible, which may be very
nearly filled, as there is scarcely any boiling up when the
pot and its contents are submitted to the fire. A thin
layer of pure litharge is placed above the mixture, the
whole is then heated rapidly, and as soon as the litharge,
&c., is completely fused, the crucible is taken from the
fire. It is inconvenient to heat it for any length of time,
on account of the corrosive action litharge has on the
substance of the crucible, which it rapidly destroys.
The proportion of litharge which must be employed
depends upon the nature and quantity of oxidisable
matters present in the ore. It ought in general to be very
great, because it is absolutely necessary that no sulphurous
matters be present, so that the slag may not contain the
least trace of silver. But it is known how much litharge
is required to decompose the metallic sulphide. Pyrites
requires about 50 parts ; mispickel, blende, antimony sul-
phide, copper pyrites, grey cobalt, and grey copper require
from about twenty-five to about forty times their weight.
For sulphide of bismuth 10 are sufficient, and for galena
or silver sulphide but 4 or 5 parts need be employed.
The proportion of litharge will not be so great for a
mineral containing much stony gangue as for one entirely
004 THE ASSAY OF SILVER.
metallic. Experiment has proved that the assay of rough
schlichs, such as those treated in the large way by amalga-
mation, can be made very exactly with from 10 to 12 parts
of litharge.
Alloys of silver with the very oxidisable metals, such
as those of iron, antimony, tin, zinc, &c., can be assayed
by means of litharge ; but in order to have a successful
result the alloys should be reduced to a very fine state of
division, so that they must be at least granulated ; and it is
very often necessary to repeat the operation several, times
on the fresh alloy of lead produced.
The method of assay just pointed out is inconvenient,
on account of the large quantity of lead it produces ; pyrites
giving 8-J parts, copper pyrites and blende 7 parts, anti-
mony sulphide and grey copper about 6 parts, &c. In
order to avoid this inconvenience, part of the oxidation
can be performed by means of nitre. Nitre alone, em-
ployed in excess, oxidises all metallic and combustible
substances found with silver, and even, under certain cir-
cumstances, a portion of the silver itself ; but when the
proportion is insufficient to oxidise the whole, and when
the mixture contains at the same time litharge, after the
nitre has produced its action the litharge acts in its turn
on the remainder of the oxidisable substances, and the
resulting lead carries down the silver set free. So that, by
employing suitable proportions of nitre and litharge, all the
silver contained in oxidisable minerals may be extracted,
and any quantity of lead required may be thus alloyed
with it.
As to the requisite proportion of nitre, it can be come
at by practice, aided by the following data. It requires
about 2^ parts of nitre to completely oxidise iron pyrites,
1^ for sulphide of antimony, and f for galena.
This estimation can be ascertained at once as fol-
lows : Fuse 1 part of the mineral with 30 of litharge, and
weigh the resulting button of lead ; and having fixed upon
the quantity of lead necessary to carry on the cupellation
properly, deduct it from the whole weight of the button,
and the difference will be the amount of lead necessary to
SPECIAL DIRECTIONS FOR THE CRUCIBLE ASSAY. 605
leave in the slag in the state of oxide ; and as it has been
proved by experiment that 1 part of lead requires -25 to
30 of nitrethat is, from 25 to 30 per cent. it is easy
to calculate the quantity necessary to be added.
When the ore contains sulphur, the latter forms with
the nitre, potassium sulphate, which swims on the slag with-
out combining with it.
The assay of silver ores by means of nitre is advan-
tageous and useful in a variety of cases. If we wish to esti-
mate, for example, very exactly the percentage of silver
in a poor galena, a large quantity, say a quarter of a pound,
must be fused with about an ounce or an ounce and a half
of nitre, and a quarter of a pound of sodium carbonate,
or, better still, the same quantity of litharge, one of either
of which must be employed to flux the gangue and tem-
per the deflagration. After the fusion, all the contained
silver will be found alloyed with a very small quantity of
lead.
Sometimes the assay is made with a larger quantity of
nitre than is requisite for the oxidation, and when the
mixture is perfectly fused a certain quantity of metallic
lead is added, taking care to cover the whole surface of
the mixture, either by using granulated lead or a conve-
nient mixture of litharge and charcoal, or litharge and
galena. The shower of metallic lead passing through the
fluid mass alloys with all the silver it finds in its passage,
and so concentrates it.
This process, however, cannot always be confidently
employed. If an excess of nitre be employed with sub-
stances susceptible of forming peroxides capable of attack-
ing silver, such as some cupreous substances, the lead added
reduces the greater part, but not the whole of the silver in
the ore, so that the assay will not be perfect.
Special Directions for the Crucible Assay of Ores and
Substances of the First Class.
The ores and substances belonging to this class may,
for the convenience of assay, be further subdivided on the
606 THE ASSAY OP SILVER.
following principle. It has already been seen that sulphur,
and other substances having a great affinity for oxygen,
reduce metallic lead from litharge in proportion to the
amount of reducing matter present ; and as it is necessary
in this kind of assay that no more than a certain quantity
of lead alloy should be submitted to cupellation, some kind
of control must be exercised by thje assayer, to keep the
quantity of lead reduced in due and proper bounds. This
is readily accomplished by what is called a ' preliminary
assay,' by which all ores and substances of this class are
divided into three sections : 1st, ores which, on fusion
with excess of litharge, give no metallic lead, or less than
their own weight ; 2ndly, those which give their own
weight, or nearly their own weight, of metallic lead ; 3rdly,
those which give more than their own weight of metallic
lead. The preliminary or classification assay is thus con-
ducted :
Carefully mix 20 grains of the finely pulverised ore
(all silver ores must be passed through a sieve with 80
meshes to the linear inch) with 500 grains of litharge ;
place the mixture in a crucible which it only half fills ;
set the crucible, after careful warming, in a perfectly bright
fire, and get up the heat as rapidly as possible, so as to
finish the operation in a short time, to prevent the action
of the reducing gases of the furnace on the lead oxide ;
because if a great length of time were taken in the opera-
tion, a portion of the lead reduced might be traceable to
the furnace gases, and the result of the experiment vitiated.
After the contents of the crucible are fully fused, and the
surface perfectly smooth, the crucible may be removed
and allowed to cool, and when cold broken. One of
three circumstances may now present itself to the assayer :
1st, no lead, or less than 20 grains, has been reduced ;
2ndly, 20 or nearly 20 grains, more or less, may be
reduced ; and Srdly, more than 20 grains may have been
reduced.
Now, as it has been already stated, 200 grains of lead
alloy is a suitable amount to cupel, and as 200 grains is
the best quantity of ore to submit to assay, it will be
SPECIAL DIRECTIONS FOR THE CRUCIBLE ASSAY. 607
evident that ores and substances of the second section, or
those bodies which give their own weight, or nearly their
own weight, of lead alloy, simply require fusion with a suit-
able quantity of litharge and an appropriate flux. Ores of
the first section require the addition of a reducing agent, in
quantity equivalent to the standard amount of lead alloy
(200 grains) ; and ores of the third section require an equi-
valent quantity of an oxidising agent, or an amount of some
substance which will oxidise the lead in excess of 200
grains of alloy.
The reducing agent employed is argol ; the oxidising
agent potassium nitrate. It is necessary, before commenc-
ing an assay of a silver ore, to estimate how much lead a
given weight of the argol the assayer has in use will reduce ;
as also how much lead a given weight of potassium nitrate
will oxidise. These assays are thus made :
Assay of Reducing Power of ArgoL Carefully mix 20
grains of the argol to be tested with 500 grains of litharge
and 200 grains of sodium carbonate ; place the mixture in
a suitable crucible, and cover with 200 grains of common
salt. (It is best to mix two such quantities, and take the
mean of the results.) Fuse with the precautions pointed
out in assay of substances of the first class, containing
lead.
Weigh the resulting buttons, and take a note of the
mean weight, which will represent the amount of lead
reducible by 20 grains of argol.
Assay of Oxidising Power of Potassium Nitrate. Mix
20 grains of finely powdered potassium nitrate, 50 grains of
argol, 500 grains of litharge, and 200 grains of sodium
carbonate ; cover with 200 grains of common salt, and fuse
as above. Weigh the resulting button. Now calculate
the amount of lead which should have been reduced by 50
grains of argol, and the difference between that and the
amount of lead reduced in this experiment will represent
the amount of lead oxidised by 20 grains of potassium
nitrate.
Thirty to 32 grains of ordinary red argol reduce about
200 grains of lead ; and 23 grains of pure potassium nitrate
(508 THE ASSAY OF SILVEE.
oxidise about 100 grains of lead. The assayer must,
however, adopt the numbers found by himself by experi-
ment, as the samples of argol and nitre may be more or
less impure. He must also examine every fresh supply of
litharge for the amount of silver it contains, in the follow-
ing manner :
Assay of Litharge for Silver. Mix 1,000 grains of
litharge with 30 grains (or any other quantity that may,
by experiment, be found requisite) of argol, 200 grains of
sodium carbonate, and cover with salt, as already directed.
Fuse the mixture in a suitable crucible ; allow it to cool ;
break and cupel the button obtained, as hereafter to be
described ; take a note of the amount of silver obtained ;
and as 1,000 grains of litharge is the standard quantity for
a silver assay, the amount of silver, indicated as above, is
to be deducted from the amount of silver obtained in the
assay of any silver ore, until that quantity of litharge is
consumed.
Assay of Ores of the First Section. Make a preliminary
assay, as already described. Suppose 10 grains of lead
result ; then, as 20 have furnished 10 grains, so 200 grains
of ore would furnish 100 grains of lead, or 100 grains less
than the quantity best adapted for cupellation ; so that,
referring to the assay of argol, and finding that from 30 to
32 grains reduce 200 grains of lead, then it is clear that
the reducing power of from 15 to 16 grains of argol, in
addition to the reducing power of 200 grains of ore, is
necessary to furnish 200 grains of lead alloy. In this case
the ingredients required in the actual assay, or ' assay
proper,' would stand thus :
200 grains of ore.
200 grains of sodium carbonate.
1,000 grains of litharge.
15 to 16 grains of argol.
These materials are to be thoroughly well mixed,
placed in a crucible which they about half fill, and covered
first with 200 grains of common salt, and then 200 grains
of borax, and submitted to the fire with the usual precau-
ASSAY OF ORES OF THE THIRD SECTION. 609
tion ; when the flux flows smoothly the assay is complete ;
it may be removed and allowed to cool, the crucible broken,
and the button obtained must be hammered into a cubical
form, and should approximate to 200 grains, either more
or less, within 10 grains. Two crucibles must always
be prepared. It will also be here convenient to mention
that the argol and potassium nitrate are the only sub-
stances whose quantities vary in the assay of silver ores,
the amount of these variations being estimated by the
preliminary or classification assay.
Assay of Ores of the Second Section. If the preliminary
assays of the sample submitted to assay furnish from 18
to 22 grains of lead, then the assay proper may be thus
made :
200 grains of the ore,
200 grains of sodium carbonate,
1,000 grains of litharge,
well mixed and covered with salt and borax as above.
Fuse with due care, and reserve buttons of lead alloy for
cupellation.
Assay of Ores of the Third Section. If the sample on
preliminary assay furnished 40 grains of lead, then the 200
grains employed in assay proper would give 400 grains or
200 grains of lead in excess ; refer now to note-book for
quantity of lead oxidised by nitre ; suppose the nitre pure
as just stated, 23 grains will oxidise 100, therefore 46
grains are equivalent to 200, and the assay proper will
stand thus :
200 grains of the ore.
200 grains of sodium carbonate*
1,000 grains of litharge.
46 grains of potassium nitrate.
The potassium nitrate is to be weighed first, finely pul-
verised, and then well mixed with the remaining sub-
stances, and covered with salt and borax. The crucible
in this assay must be larger than in the two preceding
cases ; the mixture should not more than one-third fill it,
R R
010 THE ASSAY OF SILVEE.
as there is a considerable action set up between the oxygen,
of the nitre and the sulphur or arsenic, or any other sub-
stance that may be the reducing agent in the ore ; for in
fact the nitre does not directly oxidise the lead, which
sulphur, &c., might have reduced, but oxidises its equiva-
lent quantity of sulphur, or whatever other reducing
substance there may be in the ore, so as only to leave a
sufficient amount to reduce 200 grains of lead, in lieu of
the 400 as indicated by preliminary assay, or when the
reducing power of the ore was allowed to come into full
play. The buttons obtained in this case are also to be
reserved for cupellation.
Scarification. Scorification has, like fusion with lith-
arge, the effect of producing an alloy of lead capable of
cupellation, and a very fusible slag composed of lead oxide,
and all the matters foreign to silver, converted into the
state of oxide. In the crucible assay as just described the
oxidation of these substances takes place by the action
of the litharge, which furnishes at the same time by its
reduction the lead necessary to form the alloy, whilst in
scorification all the substances susceptible of oxidation
are oxidised in the roasting by means of the oxygen of
the air, and the litharge itself is produced by the oxidation
of part of the lead mixed with the ore to be assayed.
In this operation vessels termed scorifiers (see p. 144)
are employed. They are heated in the muffle of the
cupelling furnace, and as many assays may be made at
one time as the muffle holds scorifiers.
Before introducing the scorifiers into the muffle, a given
weight of the ore reduced to powder is mixed intimately
with a certain quantity of granulated lead, and placed in
each. They must then be heated gradually for about a
quarter of an hour, with the door of the muffle closed, in
order to fuse the lead ; then diminish the heat and allow
access of air by opening the door. The current thus esta-
blished in the muffle soon causes the commencement of the
roasting ; and this roasting goes on without its being neces-
sary to continually agitate the mass, as in the case of pul-
verulent substances.
SCOKIFICATION. 611
During the oxidation, a slag is formed on the fluid
metal, which is thrown towards the edges, and which, by
continually augmenting, at last entirely covers the bath.
This slag, which is often solid at the commencement, be-
comes softer and softer, and at last becomes perfectly fluid ;
because, in proportion to the advance of the operation, the
proportion of lead oxide continually increases. When it
is judged that the scorification has been carried far enough,
the melted matter is stirred with a rod of iron, in order to
mix with the mass the hard or pasty parts attached to the
bottom or sides of the scorifier. The fire is then urged so
as to completely liquefy the slags. It may be ascertained
when they are sufficiently fluid by plunging into them a
red-hot iron rod, which must only be covered with a slight
coating, capable of running off, and not solidifying into a
drop at the end.
This condition of liquidity is indispensable, in order to
enable the metallic globules to unite into a single button.
When this end is not attained, it is because the scorifica-
tion has not been carried sufficiently far, or because a
sufficient quantity of lead has not been added to form the
flux, in which case a fresh quantity must be added, or,
what is preferable, the assay recommenced with larger
proportions.
When the operation is finished, the scorifier must be
removed, and its contents immediately poured into a
circular or hemispherical ingot mould (see fig. 29, p. 69).
The metallic particles fall to the bottom, and as the cooling
proceeds they form ' a button covered by the slag, which
is readily detachable by a blow of a hammer ; it ought
to be very homogeneous and vitreous, and its colour vary-
ing from brown to greenish.
It is always advisable to examine the slag, and ascertain
if it contain metallic globules. The button ought to be as
ductile as ordinary lead ; if not, it cannot be cupelled, and
must be submitted to a fresh operation. It is in general
advantageous to push the scorification to its greatest ex-
tent, because experiment has proved that less silver is lost
than when a large button is cupelled. Nevertheless, there
R R 2
612 THE ASSAY OF SILVER.
is a limit, because if the silver-lead produced be too rich,
the least loss in the shape of globules would cause a not-
able one in the silver. Besides, as litharge exercises a very
corrosive action on earthy matters, if the scorification be
continued for a great length of time, it sometimes happens
the vessel is pierced, and the assay has to be recommenced.
The button of lead remaining ought to weigh about 200 to
300 grains, when the ores treated are of ordinary richness.
The length of time a scorification takes is from half an
hour to an hour. The scorifier can be rendered less per-
meable to the litharge by being rubbed inside with chalk,
or, better still, red ochre.
There may be distinguished three distinct periods in the
operation viz. the roasting, the fusion, and the scorifica-
tion. At first a strong fire is employed ; but the doors ol
the furnace are opened as soon as the mixture is fused.
The mineral, being specifically lighter than the lead, is then
seen floating on its surface, or forming masses in it ; the
roasting then commences, and from the appearance of the
vapours the nature of the combustible matter it contains
may be judged. Sulphur produces clear grey vapours ;
zinc, blackish vapours, and a brilliant white flame ; arsenic,
whitish-grey vapours ; antimony, fine red vapours, &c.
When no more fumes are seen, the mineral has disap-
peared, and the fused lead is perfectly uncovered, the
roasting has terminated ; this generally requires from
eighteen to twenty minutes. At this time the fire is
urged, so as to cause all the substances in the scorifier to
fuse. It can be ascertained that the fusion is complete by
the following signs : at the instant the muffle is opened the
button becomes whitish-red with a greyish- black band,
and there arise from the melted mass clear white fumes of
lead, and the slag appears like a ring encircling the metal.
The third period then commences : the furnace is cooled,
as in the roasting, and the lead is allowed to scorify until
it is entirely covered with fused oxide ; this last period
generally lasts about fifteen minutes. The fire is then
increased for about five minutes, and the contents of the
scorifier poured into the mould.
SCORIFICATIOtf* 613
The process of scorification is applicable to all argent-
iferous matters, and is at the same time the most exact
method of assay, as also the most convenient, when a large
number of assays are required at the same time, because
they are entirely executed in the muffle, which, with most
assay ers, is generally hot : it, however, requires a greater
number of vessels as cupels, &c.
When the silver ores are stony, the lead oxide formed
during the roasting combines with the gangue, forming a
fusible compound, whilst the remaining lead alloys with
the silver. When the ores are metallic, the oxidisable
bodies absorb oxygen from the atmosphere ; and the
oxides so formed combine with the litharge produced at
the same time, forming a compound which becomes very
fusible in proportion as the lead oxide increases ; and if
the scorification has not been pushed sufficiently far, the
button will contain, besides silver and lead, a little copper,
which will not, however, interfere with the cupellation.
There is this one peculiarity about scorification, that, how-
ever small the proportion of lead may be that is used, at
the end of the operation the slag does not contain any
oxysulphide. For instance, even when oxysulphides are
produced in the course of scorification, they are completely
decomposed in the roasting, and in consequence it is very
rarely that the slag retains any proportion of silver ; and
as to the proportion of lead employed, only just enough
to render the slag liquid, and to produce sufficient lead
for cupellation, is necessary.
It is different, however, when the sulphides and arsenic-
sulphides are assayed by means of litharge ; for from 30
to 50 parts of that substance must be employed to prevent
the scoriae retaining any silver, or, as already pointed out,
a certain proportion of nitre must be added.
All scorifications may be conducted by the simple
addition of lead ; but it has been proved that the opera-
tion proceeds more quickly, and with less danger to the
scorifier, when borax is employed. This salt dissolves the
oxides in proportion as they are produced, as also the
gangues, and forms a very liquid slag from the commence-
614 THE ASSAY OF SILVER.
ment of the operation, which does not happen when lead
alone is used, because litharge, which can alone cause the
fusion, is only present in the slag in sufficient proportion
at a very advanced stage of the operation.
When the slag is liquid at the beginning of the opera-
tion (as occurs in the use of borax), it is continually thrown
on the sides of the scorifier, and forms a ring on the sur-
face of the bath, leaving in the centre the metallic sub-
stance, having a considerable extent of surface which is
continually diminishing.
The current of air, being thus directly in contact with
the fused metals, rapidly causes their oxidation, which
does not take place when the semi-fluid substances float
here and there on the metallic bath. The proportion of
lead and borax necessary for a scorification varies exceed-
ingly, according to the nature of the substance under
assay, and ought to be greater in proportion as the sub-
stances, or resulting oxides, are difficult of fusion. In
ordinary cases 12 parts of lead and 1 of glass of borax,
are employed ; but sometimes 32 of lead, and 3 of borax,
are required. A large proportion of borax is useful,
especially when the substances contain much lime, zinc
oxide, or tin oxide.
Instead of borax, glass of lead may be employed. It
acts as a flux on silica ; but its action is much less effec-
tive than that of borax.
There are some substances which scorify with a small
proportion of lead. Thus, for galena and copper sulphide,
2 parts of lead suffice ; but 8 parts are required for ores
which contain much gangue.
Silver antimonide can be scorified with 8 parts of lead,
but according to experiments made in the Hartz it appears
that the slag retains about yy^th of silver ; with 16 parts
'of lead 2-^-0 th of fine metal is still lost ; but with 3 of borax
and 16 of lead not the slightest trace remains in the slag.
t It is very difficult to separate tin and silver by the dry
way. The best method is to roast the alloy in a scorifier,
adding to it 16 parts of lead and 3 of borax at least, and
operating as before described.
ASSAY OP SUBSTANCES OF THE FIRST CLASS. 615
Speiss almost always contains silver, and is one of the
most difficult substances to assay. If nickel be present,
the button cannot be cupelled. Generally, speiss may be
scorified with 16 parts of lead ; and the same operation is
gone through twice or thrice, adding each time a fresh
quantity of lead. The operation would probably succeed
by roasting the speiss in the scorifier before adding the . /
lead.
Special Instructions for the Scorification Assay of Ores
of the First Class. This mode of assay has an advantage
over the crucible assay just described, inasmuch as if
properly conducted no preliminary assay is required ; but
this is greatly counterbalanced by the fact that not more
than 50 grains of ore can be operated on in one scorifier,
and that good or trustworthy results cannot be obtained
by this method unless four scorifier s are employed for
each assay, so that in all 200 grains of ore may be em-
ployed. There are thus employed four scorifiers to three
crucibles, and four cupels to two cupels ; as in one case
four buttons are to be submitted to cupellation, and in
the other only two. When very rich copper ores, how-
ever, have to be assayed for silver, the plan by scorifica-
tion is very useful, as in the crucible operation much
copper is reduced with the lead, so as to require a very
large quantity of lead for its conveyance as oxide into the
cupel. This class of assay will, however, be particularly
noticed under the head Assay of the Alloys of Silver.
Assay in Scorifier. Weigh out 300 grains of granu-
lated lead, place them in a scorifier, then add 50 grains
of pulverised fused borax, and 50 grains of the ore to be
assayed, well mix them in the scorifier by aid of a spatula,
and cover the mixture with other 300 grains of granu-
lated lead : prepare in this way four scorifiers, place them
in the muffle with the tongs (b, fig. 28, page 68), and care-
fully watch them with all the precautions before pointed
out : when the surface of the metal is quite covered with
fused oxide, pour the contents of each scorifier into one
of the hollows of the mould depicted at fig. 29, page 69.
When the mass of slag and metal is cold, separate the
616 THE ASSAY OF SILVEE.
latter from the former by means of the hammer and anvil,
hammer the metal into the form of a cube, and reserve it
for cupellation.
Assay of Substances of the First Class admixed with
Native or Metallic Silver. The same kind of calculation
is necessary in the assay of ores as above, as in the case of
copper ores containing metallic copper. The sample must
be carefully weighed. Suppose it to weigh 2,500 grains.
It must be pulverised, and as much as possible passed
through the sieve with eighty meshes to the linear inch.
It will be thus divided into two parts: the one passing
through the sieve is mineralised silver that is, silver ore
of various kinds mixed with earthy matter, and a very
small quantity of metallic silver which has been sufficiently
divided to pass through a sieve of such a degree of fine-
ness ; the other, impure metallic silver, which has been
unable to pass through the sieve. The weights of both
portions are carefully taken, and thus noted
. Kough metallic silver . . . 5'07 grs.
Ore through sieve .... 2494*93
Total weight of sample . . . 2500-00
Assay the ore which passed through the sieve as
already directed, and the rough silver as directed under
the head Assay of Silver Alloys. Note the quantity of
silver obtained in each experiment. Thus : suppose 200
grains of ore yielded 2 grains of fine silver, and the 5 -07
grains of rough silver 4 grains of fine silver by cupellation,
the number of ounces of fine silver in the ton is thus
calculated.
On referring to Table III. in Appendix, it will be found
that if 200 grains of ore yield 2 grains of fine silver ,.
1 ton will yield 326 ozs. 13 dwts. 8 grs. of fine silver;
so that the average produce of the ore is the above
amount.
Then, if 5 -07 grains of rough silver yield 4 grains of
fine silver, 200 grains would yield, by calculation, 159;763
grains of fine silver.
Thus
ASSAY OP SJJBSTANCES OF THE FIRST CLASS. 617
Now, by referring to Table III. in the Appendix, it
will be found that 200 grains of ore give 159 grains of fine
silver=25,970 ounces per ton; and that 200 grains of
ore give '763 grain of fine silver =124 ozs. 12 dwts.
11 grains : therefore, the 5 -07 grains of rough silver con-
tain at the rate of 26,094 ozs. 12 dwts. 11 grs. per ton,
thus
25,970 ozs. + 124 ozs. 12 dwts. 11 grs. = 26,094 ozs. 12 dwts. 11 grs.
Thus we have
ozs. dwts. grs.
Average produce of ore .... 326 8
Average produce of rough silver . . 26,094 12 11
per ton of 20 cwts.
Then, as in the case of the copper, multiply the weight
and produce of each portion together, add the resulting
total products, and divide the sum by the weight of the
sample. For this purpose it is better to reduce the penny-
weights and grains to their decimal values. Thus 13 dwts.
8 grs. is nearly equal to *67 of an ounce, and 12 dwts. 11
grs. to -62 of an ounce ; therefore the quantities above will
stand thus, 326-67 ozs. and 26,094-62 ozs.
Then 326-67 x 2494-93 = 815018-7831
>and 26094-62 x 507 = 132296-7234
and 815018-7831 + 132299-7234
2500
= 378-9 oz.
or 378 ozs. 18 dwts. (nearly) per ton of the original sample,,
before pulverising and sifting.
In every case of assay yet described, it may be men-
tioned that if the sample contain gold, the whole of that
metal will be found with the silver, as obtained by cupel-
lation, and may be separated as stated in the chapter on
the Assay of Gold.
Cupellation. Cupellation is an operation that has been
known from time immemorial ; it has many characters in
common with scorification, and is effected in nearly the
same manner. Like that, it has for its end the separation
of silver and gold from different foreign substances, by
means of lead ; but it differs in this, that the scoriae pro-
duced are absorbed by the substance of the vessel named
618 THE ASSAY OF SILVER.
a cupel, in which the operation is made, instead of remain-
ing on the melted metal, the latter remaining uncovered
and in contact with the air, so that the extraneous metals
are not only oxidised, but also all the lead ; and there re-
mains nothing but the pure metals, silver and gold, or an
alloy of them, in the cupel.
Cupellation requires, as an indispensable condition,
that the slag should have the property of penetrating and
soaking into the body of the substance forming the cupel ;
it is, therefore, only applicable to a certain number of
substances, and not to all, like scorification. Lead and
bismuth oxides, in a state of purity, are the only oxides
which possess the property of soaking into the cupel ; but
by the aid of one or the other, various oxides which by
themselves form infusible scorise on the cupel, acquire the
property of passing through it : therefore, on making a
cupellation, it is necessary to fuse the substance with a
sufficient proportion of lead or bismuth, so that the oxides
they produce may combine with the oxides of all the foreign
metals produced in the operation, and carry them into the
body of the cupel.
This proportion varies with the nature of the sub-
stances cupelled, and other circumstances. The quantity
required in ordinary cases will be mentioned hereafter.
The cupels or porous vessels in which the operation is
made ought to have a sufficiently loose texture to allow
the fused oxides to penetrate them easily, and at the same
time to possess sufficient solidity to enable them to bear
handling without fracture ; and, moreover, they ought to
be of such a nature as not to enter into fusion with either
lead or bismuth oxide. For a description of their mode of
manufacture, see p. 141.
The following is the method in which an ordinary
-cupellation is conducted : The furnace being heated, the
bottom of the muffle is covered with cupels, placing the
largest towards the end ; and if they are required to be
heated as quickly as possible, they may be placed upside
down, and turned, at the instant of use, by means of the
tongs. When the interior of the muffle is reddish-white,
ASSAY OF SUBSTANCES OF THE FIRST CLASS. 610
the matters to be cupelled may be introduced. When the
cupels have been placed in their proper position, great
care must be taken from the commencement to blow out
of them all cinders, ashes, and other extraneous substances
which may have fallen into them.
The substance to be cupelled is sometimes an alloy,
which can pass without addition of lead, and sometimes
a compound, to which lead must be added. In the first
case, the alloy is taken hold of by a small pair of forceps,
and deposited gently in the cupel. In the second case, the
substance to be cupelled is enveloped in a sheet of lead of
suitable weight, and placed, as before, in the cupel ; or the
necessary quantity of lead may be first placed in the cupel,
and when the lead is fused, the substance to be cupelled
added, taking care not to agitate the melted mass and
cause loss by splashing. If the substance to be cupelled
is in very small pieces, as grains or powder, it must be
enveloped in a small piece of blotting-paper, or, still
better, in a piece of very thin sheet lead, giving it a slightly
spherical form, and dropping it gently into the mass ot
molten metal in the cupel. Sometimes the substance is
gradually added, by means of a small iron spoon ; but it
is preferable to use paper or thin lead, as just recom-
mended.
When the cupels are filled, the furnace is closed, either
by the door or by pieces of lighted fuel, so that the fused
metals may become of the same temperature as the muffle.
When this point has been gained, air is allowed to pass
into the furnace ; the metallic bath is then in the state
termed uncovered ; that is, it presents a convex surface,
very smooth and without slag. When the air comes in
contact with it, it becomes very lustrous, and is covered
with luminous and iridescent patches, which move on the
surface, and are thrown towards the sides. These spots
are occasioned by the fused oxide of lead which is continu-
ally forming, and which, covering the bath with a very
thin coating of variable thickness, presents the phenome-
non of coloured rings.
The fused litharge, possessing the power of moistening
020 . THE ASSAY OF SILVER
(so to speak) the cupel, is rapidly absorbed by it when
sufficiently porous, so that the metallic alloy is covered
and uncovered every instant, which establishes on its sur-
face a continual motion from the centre to the circumfer-
ence. At the same time a vapour rises from the cupels
which fills the muffle, and is produced by the vapour of
lead burning in the atmosphere. An annular spot is soon
observed on the cupel around the metal, and this spot
increases incessantly until it has reached its edges.
In proportion as the operation proceeds, the metallic
bath of silver-lead diminishes, becoming more and more
rounded ; the shining points with which it is covered
become larger and move more rapidly ; lastly, as the
whole of the lead separates, the button seems agitated by
a rapid movement, by which it is made to turn on its
axis ; it becomes very lustrous, and presents over its
whole surface all the tints of the rainbow : suddenly the
agitation ceases, the button becomes dull and immovable,
and after a few instants it takes the look of pure silver.
This last part in the operation of cupellation is termed the
brightening, fulguration, or coruscation.
If the button be taken from the muffle directly after
the brightening, it may throw off portions of its substance;
this must be avoided, especially when the button is large.
The button, when covered by mammillated and crystalline
asperities, is said to have ' vegetated.' The cause of this
effect seems to be, that when the fused buttons are suddenly
exposed to the cold air, the silver solidifies on the surface,
whilst that in the interior remains liquid. The solid crust,
contracted by cooling, strongly compresses the liquid
interior, which opens passages for itself, through which it
passes out, and around which it solidifies when in contact
with the cool air. But it sometimes happens that, when
the contraction is very strong, a small portion of the silver
is thrown off in the shape of grains, which are lost.
After brightening, the cupels must be left for a few
minutes in the furnace, and drawn gradually to the mouth,
before they are taken out, so that the cooling may be slow
and gradual. These precautions are nearly superfluous
CUPELLATION. 621
when the buttons are not larger than the head of an ordi-
nary pin.
As silver is sensibly volatile, it is essential, in order
that the smallest possible quantity be lost, to make the
cupellation at as low a temperature as may be. On the
other hand, the heat ought to be sufficiently great, so that
the litharge may be well fused and absorbed by the cupel ;
and, moreover, if the temperature be too low, the opera-
tion lasts a very long time, and the loss by volatilisation
will be more considerable than if 'the assay had been made
rapidly at a much higher temperature.
Experience has proved that the heat is too great when
the cupels are whitish and the metallic matter they contain
can scarcely be seen, and when the fume is scarcely visible
and rises rapidly to the arch of the muffle. On the con-
trary, the heat is not strong enough when the smoke is
thick and heavy, falling in the muffle, and when the
litharge can be seen not liquid enough to be absorbed,
forming lumps and scales about the assay. When the
degree of heat is suitable the cupel is red, and the fused
metal very luminous and clear.
In general, it is good to give a strong heat at the com-
mencement, so as to well uncover the bath, then to cool
down, and increase the heat at the end of the operation for
a few minutes, in order to aid the brightening. There can
be no inconvenience in urging the temperature at first, be-
cause the silver-lead is then poor, and much precious metal
cannot be lost by volatilisation. The increase of fire given
towards the end is for the purpose of separating the last
traces of lead, from which it is very difficult to free the
silver ; but this strong fire must not be continued long,
otherwise there might be a notable loss by volatilisation.
When the assay of very poor argentiferous matters is
made, the heat can be kept up nearly all through the
cupellation. It generally succeeds better when the tem-
perature is too high than too low.
The force of the current of air which passes through
the muffle is another very important thing in the success
of the operation. Too strong a current cools the cupel,
022 THE ASSAY OF SILVER.
oxidises too rapidly, and the assay would be spoilt. With
too feeble a current the operation proceeds slowly, the
assay remains a long time in the fire, and much silver is
lost by volatilisation.
When the litharge is produced more rapidly than it
can be absorbed by the cupel, or when it is not liquid
enough, which may happen from the furnace being too
cold, or when other oxides, produced at the same time,
diminish its fusibility, it accumulates gradually on the fluid
metal, forming at first a ring which envelopes its circum-
ference, and which, gradually extending, covers the whole
surface ; at this period the assay becomes dull, and all
movement ceases. When the operation is carefully at-
tended to, it is nearly always possible to avoid this acci-
dent. If at the first moment any signs are manifested of
this evil, the temperature of the muffle must be raised,
either by shutting the door, or placing in it burning fuel ;
the assay will, in a little time, resume its ordinary course.
But when the cause of the mishap is supposed to be the
abundance of foreign oxides in the assay, a fresh propor-
tion of lead must be added.
It can be ascertained whether an assay has passed well
by the aspect of the button. It ought to be well rounded,
white, and clear, to be crystalline below, and readily de-
tached from the cupel. When it retains lead it is brilliant
below and livid above, and does not adhere at all to the
cupel.
Ill order to detach the button, seize it with a strong
pair of pliers (see fig. 109), and examine with a microscope,
(see fig. HO), brushing it to detach small particles of
litharge which may adhere to it, and place it in the pan
of a balance (fig. 13, page 28), which will indicate the
T _i__tn of a grain. The weight of the silver furnished by
the lead or litharge employed in the operation ought to
be subtracted from the amount of silver obtained ; so that
it is necessary to ascertain the richness of these matters
beforehand, as they are never completely free from silver.
The poorest of them contain from TWoo tn to iw^o- tn -
Sometimes an equal quantity of lead is placed in another
CUPELLATION. 623
cupel, and the silver thus obtained placed in the balance-
pan containing the weights.
Cupellation does not give the exact proportion of silver
contained in an alloy. There is always a loss, and this loss-
is always greater than that which takes place in the large
way, as in the latter process a greater quantity is always
FIG. 109, FIG. 110.
obtained than that estimated by the assay. The loss of
silver is traceable to three causes .: 1st, to volatilisation ;
2ndly, to oxidation ; Srdly, and lastly, to the absorption of
minute globules of silver into the body of the cupel. It is
certain volatilisation takes place, because a notable quan-
tity of silver is always found deposited on the sides of the
furnace and chimney in the shape of dust ; and silver, which
is volatile by itself, becomes much more so when alloyed
with lead, and is carried away by the vapours of the latter,
and found in the pulverulent deposits, termed lead smoke
or fume, which proceed from the combustion of the latter
624 THE ASSAY OF SILVER.
metal in the air. Nevertheless, this cause of loss is not
very important, for it is rare that the fume contains more
than -roiro-oth f silver, and accurate experiments have
proved that in cupellation in the small way not more than
two to three per cent, of lead is volatilised. It is certain
that a portion of the silver found in cupels which have
been used for assays exist in the state of oxide, for no part
of their mass is free it is found even in the bottom ;
besides it is known that the lead carbonate precipitated
from lead acetate made from litharge contains silver, and a
notable quantity of that metal is found even in the lead
sulphate prepared by means of alum from the acetate
(unless the sulphate is repeatedly washed with water).
It has been remarked that the centres of cupels which
have been used for assays are richer in silver than the parts
near the circumference, and that under the button there is
a spot of bright yellow, which appears to be a compound
of silver. But the most important cause of loss in an assay
is the property which the alloys of silver and lead possess
of introducing themselves into the pores of the cupel. The
quantity thus lost is in proportion to the coarseness of the
cupel. For the same quantity of silver, the loss which
takes place in an assay varies according to the nature of
the alloy, and the circumstances under which the assay is
made ; so that it is not possible to form accurate tables of
correction. This loss is much augmented with the quantity
of lead employed, but without its being proportionate ; so
that when scorification is had recourse to, it is advantageous
to continue the operation for some length of time, in order
that the metallic button may be reduced to the smallest
suitable volume.
In the assay of rich alloys, the proportion to the total
amount of silver is very small, but notable ; and it has
been calculated for the alloys of copper employed in the
arts at ^ J^th ; but in the assay of poor ores, such as
galena and other minerals treated in the large way, the
loss is very great, for it is usually as high as -5^-0 th.
By extracting the lead from cupels used in this class
of assay, the metal furnished contains from about g- - oVo~o tn
CUPELLATION. 625
to -s-innnnrti 1 f silver. The following experiment will
give an idea of the influence of the proportion of lead on
the loss of silver : 100 grains of commercial litharge were
fused with 10 grains of black flux, and gave 27 grains of
lead, and a slag ; this was pulverised and reduced in the
same crucible with 15 grains of black flux, and a second
button was produced weighing 45 grains. These two
buttons, being cupelled separately, gave, the first -0035
and the second -001 only of silver. Three new quantities
of 100 grains of the same litharge were fused ; the first
with -J a part of starch, the second with 2^, and the third
with 10 of the same reducing agent. The resulting
buttons of lead weighed respectively 5, 28, and 79 grains.
These buttons were cupelled, and furnished "0035, '0035,
and -003 respectively. From these experiments it will be
seen that when the litharge is not reduced completely,
there remains a notable proportion of silver in the scoriae ;
but, nevertheless, in order to extract the largest possible
quantity, the whole must not be reduced. Indeed, but a
twentieth part only need be reduced, because more precious
metal is lost in the cupellation of a large quantity of lead
than remains in the portion not reduced. The loss of
silver in large cupellations is less than that which takes
place in an assay, because in the large way the litharge,
or the greater part of it, is run off; whilst in an assay the
cupel totally absorbs it, so that the latter presents, rela-
tively to the same mass of lead, a very much smaller
surface in the large than in the small way ; now it can be
readily seen that the quantity of silver lost by absorption
into the pores of the cupel must be proportioned to its
surface, all things being equal.
It has been ascertained by experiment that a cupel
absorbs about its own weight of litharge ; so that from
this fact a cupel of the proper size may be chosen, when
the weight of lead to be cupelled is ascertained. It is
always better to have the cupel about ^ or \ as heavy
again as the lead to be cupelled.
The various metals found in an alloy, which can be
submitted to cupellation, scorify in proportion to their
s s
626 THE ASSAY OF SILVER.
oxidisability. Those most oxidisable scorify with the
greatest rapidity, and vice versa ; so that those which
have the greatest affinity for oxygen accumulate in the
first portions of litharge formed, which, by that means
becoming less fusible, sometimes lose the property of
penetrating the cupel ; hence the reason why cupellations
always present more difficulties at the commencement of
the operation than towards the end, when the litharge
formed is nearly pure lead oxide, and can contain only
copper oxide.
The appearance of the cupel used in an assay will
give indications of the metals the alloy contained. Pure
lead colours the cupel straw-yellow, verging on lemon-
yellow. Bismuth, straw- yellow passing into orange-yellow.
Copper gives a grey, dirty red, or brown, according to its
proportion. Iron gives black scorise, which form at the
commencement of the operation, and are generally found
at the circumference of the -cupel. Tin gives a grey slag.
Zinc leaves a yellowish ring on the cupel, producing a
very luminous flame, and occasioning losses by carrying
silver in its vapour, and by projecting it from the cupel
in its ebullition. Antimony and lead sulphate in excess
give litharge-yellow scoriae, which crack the cupel ; but,
when not produced in too great a proportion, are gradually
absorbed by the litharge. If the lead alloy submitted to
cupellation is found to produce this effect, a fresh portion
must be mixed with its own weight of lead and scorified :
the button so obtained can now be cupelled.
Amalgamation. There are a certain number of argen-
tiferous matters which can be assayed by amalgamation,
as they are treated in the large way by that method.
Amongst these are native silver, chlorides, sulphides, and
arsenio-sulphides, which contain neither lead nor copper.
But this process is seldom had recourse to, because
it is long, troublesome, and less exact than those just
described.
ASSAY OF THE ALLOYS OF SILVER AND COPPER. 627
Substances of the Second Class.
Native silver.
Alloys of copper and silver.
Alloys of other metals and silver
(artificial).
Silver antimonide.
Silver telluride.
Auriferous silver telluride (see
gold).
Silver hydrargyride (amalgam),
Silver auride (see gold).
Silver arsenide.
The following method of separating silver from galena
is given in the ' Chemical News,' vol ii. p. 239.
' Galena consists, as is well known, of lead sulphide,
mixed with a variable proportion of silver sulphide, and
both these substances fuse together, or melt at a bright
red heat. Now, it so happens that, when silver sulphide
is fused with lead chloride, what is called a double de-
composition takes place ; that is to say, silver chloride
and lead sulphide are formed. Consequently, if we
fuse together a quantity of argentiferous galena and lead
chloride, we shall remove the whole of the silver from the
galena, and replace it by lead sulphide. This, then, is the
process : mix together the galena and lead chloride in the
proportion of 100 parts of galena, 1 part of lead chloride,
and 10 parts of sodium chloride or common salt ; or, if the
galena be very argentiferous, add a larger amount of lead
chloride. The whole is then fused together, when the
silver chloride and common salt rise to the surface, and
may be skimmed off, and the desilverised galena falls and
may be run out from the bottom. The mixture of silver
chloride and salt may then be decomposed by lime and
charcoal, or in any other manner, so as to reduce the silver
and a portion of the surplus lead chloride, by which a metal-
lic mass will result, suitable for the operation of the cupel.'
General Remarks on the Assay of the Alloys of Silver
and Copper. The assay of these alloys is nearly always
accomplished (at least in England) by cupellation. This
assay is most important, as it is by the results obtained in
the manner hereafter described that the price or value of
all kinds of silver bullion is decided.
This class of cupellation is effected without difficulty,
because the copper oxide forms so slowly, that the litharge
s s 2
628
THE ASSAY OF SILVER.
is always enabled to pass it into the body of the cupel.
After having weighed the lead and placed it in the cupel,
as soon as it is perfectly fused place in it the alloy to be
assayed, wrapped either in blotting-paper or thin leaf-lead.
It is essential, in this class of assay, to employ a sufficient
quantity of lead to carry away all the copper. We may
always be sure of succeeding, whatever the alloy may be,
by employing the maximum proportion of lead, that is to
say, the quantity necessary to pass pure copper ; but as
the loss which the silver undergoes increases with the
length of the operation and with the mass of the oxidised
matters, it is indispensable to reduce this loss as much as
possible by reducing the proportion of lead to that which
is strictly necessary. Long experience has proved that
silver opposes the oxidation of copper by its affinity, so
that it is necessary to add a larger amount of lead in pro-
portion to the quantity of silver present.
M. D'Arcet has obtained the following results by
accurate experiments :
Standard of
silver
Quantity of
copper alloyed
Quantity of lead
necessary
Relation of lead
to copper
. 1000
lo*hs
950
50
3
eotoi
900
100
7
701
800
200
10
50 1
700
300
12
40 1
600
400
14 35 1
500
500
16 to 17 321
400
600
16 17
27 1
300
700
16 17
23 1
200
800
16 17
20 1
100
900
16 17
18 1
pure copper
1000
16 17
16 1
It is remarkable that below the standard of 500, the
same proportion of lead must be employed, whatever that
of copper. This fact is repeatedly verified by experiment.
Whenever fine silver is fused in a cupel, it is always neces-
sary to add lead, in order to cause the button to unite
and form well. If less than fV ns f ^ eac ^ be employed,
the button will be badly formed ; the litharge cannot
separate but by the action of a very strong heat, and a
ASSAY OF THE ALLOYS OF SILVER AND COPPER.
629
considerable loss of silver ensues. If, on the contrary,
Y 3 ths of lead is exceeded, the cupellation goes on well, but
the loss is greater on account of the duration of the process.
These proportions also ought to vary with the temperature.
M. Chaudet has found that, to cupel an alloy containing
T 9 ^^ths of silver, 5 parts of lead are required in the middle
of the muffle, 10 in the front, and only 3 at the back.
The proportion of copper carried off by litharge varies
not only with the temperature, but even for the same
temperature in relation to the amount of copper and lead
the alloy contains. By cupelling 100 parts of copper with
different proportions of lead in the same furnace, M.
Karsten obtained the following results :
Lead added
Copper remaining after
cupellation
Quantity of lead consumed
in carrying off 1 of copper
100
78-75
3-
200
70-12
7-1
300
60-12
7-7
400
49-40
7-9
500
38-75
8-1
600
26-25
8-15
700
19-75
8-00
800
8-75
8-70
900
5-62
9-50
1000
1-25
10-10
1050
o-oo
10-50
From which we see that the lead carried away from -j^th
to yVth of its weight of copper. Much less lead can be
employed in a cupellation by making the alloy maintain
its richness of copper throughout the operation. This
can be accomplished by adding to the alloy in the cupel
small doses of lead, in proportion as that first added dis-
appears by oxidation. If, for example, an alloy composed
of 4 parts of copper and one of silver be fused with 10 of
lead, by adding successive small doses of the latter, as
already pointed out, but 7 parts will be consumed, although
in the regular way from 16 to 17 would be employed.
The proportion of copper oxide contained in the litharge
increases each instant, and goes on incessantly increasing
when an alloy of copper and lead is cupelled which con-
tains an excess of copper. According to M. Karsten, this
(330
THE ASSAY OF SILVER.
proportion is always about 13 per cent, at the commence-
ment, and 36, or more than a third, at the end of the
operation.
In the assay of the coined alloys of copper and silver,
the loss of silver may even amount to five thousandths ;
but the loss is variable, and is proportionately greater as
the standard of the alloy is lower.
The following Table contains the results of many
experiments made on this subject :
Loss, or the quantity of fine
Exact standard
Standard found by cupellation
metal to be added to the stan-
dard as obtained by cupellation
1000
998-97
1-03
975
973-24
1-76
950
947-50
2-50
925
921-75
3-25
900
896-00
4-00
875
870-93
4-07
850
845-85
4-13
825
820-78
4-22
800
795-70
4-30
775
770-59
4-41
750
745-38
4-52
725
720-36
4-64
700
695-25
4-75
675
670-27
4-73
650
645-29
4-71
625
620-30
4-70
600
595-32
4-68
575
570-32
4-68
550
545-32
4-68
. 525
520-32
4-68
500
495-32
4-68
475
470-50
4-50
450
445-69
4-31
425
420-87
4-13
400
396-05
3-95
875
371-39
3-61
350
346-73
3-27
325
322-06
2-94
300
297-40
2-60
275
272-42
2-58
250
247-44
2-56
225
222-45
2-55
200
197-47
2-55
175
173-88
2-12
150
148-30
1-70
125
123-71
1-29
100'
99-12
0-88
75
74-34
0-66
50 !
49-56
0-44
25
24-78
0-22
ASSAY PKOPER OF SILVEE BULLION. 631
These numbers, however, are not constant, and vary with
the circumstances under which the assays are made : two
assays made from the same ingot, by the same assayer,
may differ as much as four or five thousandths. Tillet
has remarked that the cupels can retain double as much
silver as is lost ; which proves, as has already been men-
tioned, that the silver obtained by cupellation is not per-
fectly pure, but may retain as much as 1 per cent, of lead.
Special Instructions for the Assay of the Alloys of Silver
and Copper.
As before stated, peculiar weights are employed in the
assay of silver bullion ; and the silver assay pound, with
its divisions, will be found described at pages 34-35.
In the ' General Eemarks on the Assay of the Alloys of
Silver and Copper,' it will be seen that the alloy must be
cupelled with a quantity of lead, varying with the amount
of copper present in the alloy. Standard silver cupels very
well with five times its weight of lead ; but when the
approximate composition of the alloy is not known, it
must be estimated by a preliminary assay.
Assay for Approximative Composition of Alloy. Weigh
off 50 grains of pure or test lead ; place them in a cupel
previously made red-hot ; when the ]ead is fused, and its
surface covered with oxide, place in it by means of the
light tongs (a, fig. 28, page 68) 2 grains of the alloy under
assay, wrapped in a small piece of thin paper. Allow the
cupellation to go on according to the instructions, and with
all the precautions already given, and when complete,
weigh the resulting button, and, according to its weight,
add lead in the actual assay in the quantity that is suffi-
cient, as exhibited in the Table at page 628.
Assay Proper of Silver Bullion. In this assay the ope-
rator requires silver known to be standard, and pure lead.
With the possession of the above substances the assay is
thus proceeded with : Place the 12 grains weight(=l Ib.)
in the scale pan, and exactly counterbalance it with stan-
dard silver. This is to serve as a check. Remove the
weight, and in its place add so much of the alloy to be
632 THE ASSAY OF SILVER.
assayed that the balance is again equal. In one cupel, that
destined to receive the check sample, place 60 grains of lead ;
and in another cupel place such a number of grains of lead
as may be found necessary by the preliminary assay. When
the lead in both cupels is fused, add the silver alloy, and
cupel with the necessary precautions. When the buttons in
the cupels are cold, seize them with the pliers, and if neces-
sary cleanse them with a hard brush, and place one in each
balance-pan. If they exactly balance each other, the alloy
operated on is standard silver ; if, however, it weighs less
than the button produced from the check sample by the
weight equivalent to 2 pennyweights, then it is 2 penny-
weights worse than standard : on the other hand, if it be
heavier by the same weight, it is 2 pennyweights better than
standard. Silver is also reported as so much fine : thus
standard silver may be reported as 11 ounces 2 penny-
weights fine, and so on. In case extreme accuracy be re-
quired, correction must be made according to the standard
as shown by the Table at page 630. The standard silver
in England is T V^- fine.
Assay of Alloys of Copper and Silver. In the treat-
ment on the large scale of copper ores containing silver,
the contained silver is found alloyed with the copper, and
it often falls under the assayer's province to estimate the
quantity of precious metal. An assay of this kind is most
conveniently accomplished by scorification before cupel-
lation, thus : Prepare four scorifiers ; weigh into each of
them 50 grains of the alloy, 50 grains of fused borax, and
600 grains of lead, and proceed as already described under
the head ' Assay of Ores of the First Class by Scorifica-
tion.' When the four buttons of lead are obtained, place
them together in another scorifier, and submit to the
furnace until the contents of the scorifier are completely
covered with oxide ; pour as usual, and cupel the re-
sulting mass of lead.
Alloys of Platinum and Silver. -If any substance con-
taining platinum as well as silver were assayed as already
described, the button resulting from the cupellation would,
in addition to the silver, contain the whole of the platinum.
SEPARATING SILVER FROM THE BASER METALS. 633
In such a case the button so obtained must be thus
treated :
If the alloy contain much platinum, it must be fused
with twice its weight of silver ; then treated with hot nitric
acid ; evaporate the solution nearly to dryness ; add water
and hydrochloric acid, until no further precipitation of
silver as a white curdy precipitate (silver chloride) takes
place. The silver chloride may be collected either on a
filter or by decantation. The solution containing the
platinum is treated with excess of sal-ammoniac solution
until no further precipitation takes place ; the solution
evaporated to dryness. When cold, dilute alcohol is added ;
and the insoluble yellow matter (platinum ammonio-
chloride) collected on a filter, washed with alcohol, dried,
and ignited. The ignited residue is metallic platinum,
which is weighed. The loss of weight which the alloy
from cupel has sustained represents the amount of silver
previously alloyed with it.
Alloy of Platinum, Silver, and Copper. Treat such an
alloy as above ; and the liquid, filtered from the platinum
ammonio-chloride, will contain the copper. Acidulate it
with hydrochloric acid, add metallic zinc, and proceed as
directed under the head ' Wet Copper Assay.'
Native Silver, Rough Silver left on Sieve during Pulveri-
sation of Silver Ores of First Class, and Native Alloys of
Silver as Antimonides, c. are treated by scorification
and cupellation in precisely the same manner as just de-
scribed for alloys of copper and silver.
Dr. W. Dyce proposed, in ' Tilloch's Philosophical
Magazine ' for 1805, the following process for separating
gold and silver from the baser metals :
' Hitherto the process has always been, as far as I have
understood it, attended with considerable difficulty in the
execution ; but, by that which I am about to describe, it
is done with exact certainty. It was discovered and com-
municated to me by a gentleman in the neighbourhood.
The process consists in mixing not less than two parts of
powdered manganese with the impure or compound metal,
which should be previously flattened or spread out so as
634 THE ASSAY OF SILVEE.
to expose as large a surface as possible, and broken or cut
into small pieces for the convenience of putting the whole
into a crucible, which is then to be kept in a sufficient
heat for a short time. On removing the whole from the
fire, and allowing it to cool, the mixture is found to be
converted into a brownish powder, which powder or oxide
is then to be mixed with an equal proportion of powdered
glass, and then submitted in a crucible to a sufficient heat,
so as to fuse the whole, when the perfect metals are
found at the bottom in a state of extreme purity a cir-
cumstance of no small importance to the artist and the
chemist, the latter of whom will find no difficulty in sepa-
rating the one from the other, with so little trouble com-
pared with the usual processes that I have no doubt it
will always be practised in preference to the cupel.'
Assay of Silver Bullion by the Wet Method. From
what has been stated under the head of ' Cupellatiori,' it
will be observed that there are many sources of error ;
such as volatilisation of the precious metal, its oxidation
in the presence of excess of lead oxide and atmospheric
oxygen, and, lastly, its absorption into the body of the
cupel either as oxide or metal, or in both states. These
losses, as before stated, vary with the temperature, the
amount of lead employed, and the texture of the cupel ;
and, as may be seen from the table of corrections as
drawn up by D'Arcet, give a very erroneous assay, unless
the addition necessary for each standard be made.
Considerable attention was called to this matter in
France some years since, and a Special Commission was
appointed to examine the subject thoroughly, and, if
possible, to devise some means of assay which might be
both easy and accurate. The result of this examination
was the invention of a process of assay at once elegant
and trustworthy : and as a full account of this method has
not, to the author's knowledge, been translated and pub-
lished in this country,* he has prepared the present from
M. Gay-Lussac's Eeport, which formed a part of a com-
* Some portion of this report has been published in Dr. Ure's ' Dictionary
of Arts, Mines, and Manufactures.'
ASSAY OF SILVER BULLION IN THE WET WAY. 635
munication from M. Thiers to Earl Granville, and which
appeared in the original language in the year 1837, in a
Eeport on the Eoyal Mint.
The process of assay about to be described consists in
estimating the fineness of silver bullion by the quantity
of a standard solution of common salt necessary to fully
and exactly precipitate the silver contained in a known
weight of alloy. This process is based on the following
principles :
The alloy, previously dissolved in nitric acid, is mixed
with a standard solution of common salt, which precipi-
tates the silver as chloride, a compound perfectly insoluble
in w r ater, and even in acids.
The quantity of silver chloride precipitated is esti-
mated not by its weight, which would be less exact and
occupy too much time, but by the weight or volume of
the standard solution of common salt necessary to exactly
precipitate the silver previously dissolved in nitric acid.
The term of complete precipitation of the silver can
be readily recognised by the cessation of all cloudiness
when the salt solution is gradually poured into that
of the nitrate of silver. One milligramme of that metal
is readily detected in 150 grammes of liquid ; and even
a half or a quarter of a milligramme may be detected,
if the liquid be perfectly bright before the addition of
the salt solution.
By violent agitation during a minute or two, the liquid,
rendered milky by the precipitation of silver chloride,
becomes sufficiently bright after a few moments' repose to
allow of the effect of the addition of half a milligramme of
silver to be perceptible. Filtration of the liquid is more
efficacious than agitation ; but the latter, which is much
more rapid, generally suffices. The presence of copper ,.
lead, or any other metal, with the exception of mercury
(the presence of the latter metal requires a slight modifi-
cation of the process, which will be hereafter pointed out) r
in the silver solution, has no sensible influence on the
quantity of salt required for precipitation : in other words,
the same quantity of silver, pure or alloyed, requires for
636 THE ASSAY OF SILVER.
its precipitation a constant quantity of the standard salt
solution.
Supposing that 1 gramme of pure silver be the quan-
tity operated on, the solution of salt required to exactly
precipitate the whole of the silver ought to be of such
strength that, if it be measured by weight, it shall weigh
exactly 100 grammes, or if by volume 100 cubic centi-
metres. This quantity of salt solution is divided into
1000 parts, called thousandths.
The standard of an alloy of silver is generally the
number of thousandths of solution of salt necessary to
precipitate the silver contained in a gramme of the alloy.
Measurement of the Solution of Common Salt. The
solution of common salt will hereafter be termed the
normal solution of common salt. It can be measured by
weight or volume. The measure by weight gives greater
FIG ill precision, and it has the special advan-
tage of being independent of temperature ;
but it requires too much time in nume-
rous assays. The measure by volume
gives a sufficient exactitude, and requires
much less time than the measure by
weight ; it is, indeed, liable to the in-
fluence of temperature, but tables for
correction will be appended.
Measure of the Normal Solution of
Salt by Weight This solution should be
so made that 100 grammes will exactly
precipitate 1 gramme of pure silver
dissolved in nitric acid. In order to
point out the method of taking the
weight it must be supposed to have
been previously prepared. After the
process of taking the weight is described,
the mode of preparing the solution will
be given.
The solution is weighed in a burette
(fig. Ill) whose capacity is from 115 to 120 grammes of
the solution, and divided into grammes. These divisions are
MEASUKEMENT OF THE SOLUTION OF COMMON SALT. 637
for the purpose of approximate! vely estimating the weight
of solution, so as to shorten the operation of weighing. The
burette is represented as closed by a cork, B, in order to
prevent evaporation of the solution when the instrument is
not in use. It is also easy to remedy the inconvenience of
evaporation, by rinsing the burette with a small quantity
of the fresh solution. On pouring the solution from the
orifice, (9, of the burette, each division will furnish from
8 to 10 drops ; and consequently the weight of a drop is
about a decigramme. The burette is filled with solution
to the division o ; it is then tared in a balance capable
of turning with a centigramme. The burette is then re-
moved, and its place supplied with a weight equivalent to
the amount of solution required 100 grammes, for in-
stance. The solution is then gradually poured from the
burette into a bottle appointed for its reception, until the
equilibrium is nearly established. It is not easy to attain
the point exactly, as no smaller quantity than a drop can
be poured from the burette. This, however, is a matter
of indifference ; it suffices to know the exact weight of
the solution poured out : suppose it to be 99 gr. 85 c.
The mode of more nearly approximating the
required weight of 100 grammes will now be
pointed out.
It must be remarked that it is not the
amount of water contained in the 100 grammes
that is of consequence, but only the quantity
of salt found in solution ; this should exactly
represent 1000 thousandths of pure silver. If
now 100 grammes of the normal solution be
mixed with 900 grammes of water, it is evi-
dent that 1 gramme of this new solution is
equivalent to a decigramme of the first, and
consequently it will be easy to obtain 100
grammes of the normal solution, or rather the
1000 thousandths of salt it ought to contain ;
it will now be sufficient to add to the 99
grammes already poured from the burette, 1 gramme
of the new solution. It can be weighed, like the normal
FIG. 112.
038 . THE ASSAY OF SILVER.
solution, to a drop nearly, in the burette (fig. 112), which
is of such a diameter that each small division represents
a decigramme of liquid, and consequently a centigramme
of the normal solution ; but it is more readily measured
by volume, preparing it in the manner to be hereafter
pointed out. To avoid all confusion, a solution to be
termed a decime solution of common salt is one containing
the same quantity of salt as the normal solution, in a
weight or volume ten times greater.
A decime solution of silver is a solution of silver equi-
valent to the latter, both mutually suffering complete
decomposition.
Preparation of the Decime Solution of Common Salt.
One hundred grammes of the normal solution of common
FIG. 113. FIG. 114.
salt are weighed in a flask (fig. 113) containing a kilo-
gramme of pure water, when filled up to the mark a b,
or 1000 cubic centimetres ; this quantity is made up with
pure water, taking care to agitate the whole well, to
render the mixtuise, homogeneous. A cubic centimetre of
this solution represents 1 thousandth of silver. This
quantity is readily obtained by means of a pipette (fig.
114), gauged so that when filled up with water to the
mark c d, it shall allow 1 gramme, or 1 cubic centimetre,
to run freely, the small quantity of liquid remaining in
the pipette not forming part of the gramme. In pouring
the liquid by drops, a little more or a little less than
PREPARATION OP THE DECIME SOLUTION OF SALT. 639
twenty may be counted, according to the size of the
orifice, o. This number will not vary more than one
drop. Half a cubic centimetre will consequently be re-
presented by 10 drops, and a quarter by 5. The precision
arrived at by this method of measurement suffices, since
the possible error on the cubic centimetre will be but one-
twentieth of that quantity, or one-twentieth of a thou-
sandth ; if, however, many measures be required, then
compensation must be made.
The decime solution of common salt requisite for assays
must be kept in a bottle (fig, 114) closed by a cork, tra-
versed by the pipette, firmly fixed in a FlG< 115<
hole bored for that purpose. To mea-
sure a thousandth with the pipette,
the bottle is held with one hand, and
the pipette with the other (fig. 115).
The pipette is taken from the solution
after its upper orifice has been closed
by the forefinger ; the lower orifice
is then inclined against the edge of
the flask to remove the liquid, which
without this precaution would remain
there : the mark c d is then raised to
the level of the eye, and by a suitable
pressure of the forefinger on the upper
orifice, which may be obtained by giving the pipette a
slight alternating circular movement between the fingers,
the solution is allowed to run out gradually. The instant
the concave surface of the liquid is at the level c d, the
pipette is firmly closed by pressure of the forefinger on its
orifice, which is held above the bottle into which the solu-
tion is to be poured, and the forefinger removed so that it
can be emptied. It is here necessary to remark that in
order to regulate the slow and regular runnings of the
liquid from the pipette, by the pressure of the forefinger,
the latter ought to be neither too moist nor too dry : if too
dry, it will not perfectly close the orifice, even by strong
pressure ; if too moist, it prevents the entrance of air, and
the liquid will not run, or if it do, it will be irregularly.
640 THE ASSAY OF SILVER.
This observation should not be lost sight of in the use of
the large burettes mentioned hereafter.
Preparation of the Decime Solution of Silver.*- The
decline solution of silver is prepared by dissolving 1
gramme of pure silver in nitric acid, in a flask holding 1
litre (see fig. 113), and then diluting the solution with
distilled water, so that, cooled at the ordinary temperature
FIG. 116.
of the air, it shall occupy exactly the volume of one litre.
It is measured in precisely the same manner as the decime
salt solution.
Weighing the Normal Solution of Common Salt. To
execute this operation with rapidity, a balance similar
to that represented at fig. 116 is employed. The arms
are divided as in the assay balance described at p. 28 ;
each of the arms, c B, is furnished with a rider, c, of
such a weight (about 5 decigrammes) that, moved from the
right or the left of the centre, c, of each arm, it indicates
PKEPAKATIOX OF THE NORMAL SALT SOLUTION. 641
two decigrammes. The space traversed by the rider is
divided into twenty equal parts, representing an equal
number of centigrammes.
We will take for example the weighing of 100 grammes
of normal solution of common salt, which is that most fre-
quently made in the estimation of the standard of all
varieties of argentiferous matter.
There are two weights, one, P, equal to the tare of the
burette when full of solution to the mark o, the other, P',
equals 100 grammes. The burette is filled with solution,
and placed on the right-hand pan of the balance, on which
it is kept in position by the collar d 0, and through which
it is passed before placing it on the pan. The tare, P, of
the burette is- supposed to be on the opposite side. If the
equilibrium be not perfect, it is effected by the rider on the
left ; the burette is then removed, and 100 grammes of the
solution (either more or less to one or two decigrammes)
poured out. The burette is then again placed in the
balance, with the 100-gramme weight P', the upper part
of which is slightly concave to receive the bottom of the
burette, in order to prevent it sliding off. The equilibrium
is again established by the aid of the rider on the right,
If, for instance, it is found necessary to remove the rider,
which represents 15 centigrammes, 15 divisions towards B,
the weight of the solution poured out of the burette will
be equal to 100 gr. - 0-15 gr. = 99'85 gr. If, on the other
hand, it is necessary to move the rider six divisions towards
c, the weight of the solution will be 100 gr. + 0-06
gr.=100-06 gr.
The above method of weighing the salt solution ap-
pears to be the most convenient that can be employed,
although it is not very expeditious. Other methods of
weighing and measuring will be given in an appendix to
this article.
Preparation of the Normal Solution of Common Salt
when measured by weight. After having pointed out the
method of weighing the normal solution of salt, and of
taking very small quantities, its preparation will be de-
scribed.
T T
042 THE ASSAY OP SILVER.
Supposing the salt as well as the water to be employed
are pure, the two substances have only to be taken in the
following proportions : 0-5427 kilogramme of salt and
99*4573 kilogrammes of water, to form 100 kilogrammes
of solution, of which 100 grammes will exactly precipitate
1 gramme of silver. But instead of pure salt, which is
difficult to procure, and which besides rapidly alters by
the absorption of atmospheric moisture, it is preferable to
employ a concentrated solution of commercial salt, which
can be prepared in large quantities, and kept for use as
needed. The quantity of salt it contains can be ascer-
tained " s by evaporating a portion to dryness, and by a few
experiments it is easy to estimate in what proportion
it shall be mixed with water to produce a solution, 100
FIG 117 grammes of which shall exactly precipitate
1 gramme of silver.
Suppose, for example, that the salt
solution contains 250 grammes of salt per
kilogramme, and that it is necessary to
prepare 100 kilogrammes of the normal
solution. Now, since for the preparation
of this quantity 0-5427 kilogramme of pure
salt is required, we have the following pro-
portion :
0-25 : 1 :: 0-5427 : ^=2-1708 kilogs.
To this last weight enough water is added
to make up 100 kilogrammes, that is to say,
97-8292 kilogrammes, which quantity can
be readily measured by means of a flask
containing 5 or 6 kilogrammes previously
gauged.
The mixture must be well agitated by
means of the agitator (fig. 117), which is
made of an osier twig, split into four
branches, to the extremities of which is attached a small
square piece of silk. This'substance is employed to avoid
the separation of filaments which would ensue from the
use of any other material. This agitator can be intro-
PREPARATION OP THE NORMAL SALT SOLUTION. G43
duced into very small openings, and is exceedingly ser-
viceable in agitating large masses of liquid.
When well mixed, the solution must be assayed. To
effect this, dissolve 1 gramme of silver in nitric acid, sp.
gr. 1-290, in a stoppered bottle (fig. 118) Fm. us.
holding about 200 grammes of water, tare
the burette (fig. 1] 1) filled with the solution,
and pour rather more than less into the
bottle ; in proportion as the salt employed is
impure, more than 100 grammes will be re-
quired to precipitate 1 gramme of silver. The
mixture is at first milky, but, by vigorously
shaking the bottle, having its stopper firmly
fixed, for about a minute, and then allowing it
to remain at rest for a short time, the liquid will become
perfectly bright : two drops of the solution must then be
poured into it from the burette : if a cloudiness is pro-
duced, it is agitated again to brighten it, and two drops
more added. This must be continued until the last two
drops added give no precipitate. The operation is then
terminated, and nothing remains to state but the result.
Supposing the total weight of solution poured from the
burette is 101-88 grammes, the last two drops must not
be reckoned, because they produce no effect ; the two pre-
ceding drops were necessary, but in part only ; that is
to say, the number of drops to be deducted is less than
four, and more than two, or rather it is the mean term,
three. Or the weight of a drop can be known exactly
by taking that of a dozen: suppose it is equal to 0*082
gramme, three times that number must be deducted,
or 0*255 grammes from 101-880 grammes: there will
remain 101-625 grammes, representing the quantity of
normal solution necessary to precipitate 1 gramme of
silver.
The solution is thus found to be too weak ; to bring it
to its proper standard it is necessary to remove 1-625
gramme of water from the 101-625 grammes of solution,
or, what is the same thing, to add to the normal solution
a certain quantity of the concentrated solution of common
T T 2
644 THE ASSAY OF SILVER.
salt, which quantity may be found by the following pro-
portion :
100 : -1-625 :: 2-1708 kilogrs. of silver solution : ^ = 0*0353.
After the addition of this quantity of salt to the normal
solution, a fresh assay is made, proceeding in precisely the
same manner as before ; taking care, however, to pour
from the burette a weight of solution slightly under 100
grammes, or 1000 decigrammes ; for instance, 998*4
decigrammes, because it is not possible, in pouring the
solution by drops, to arrive at the exact weight, 1000
decigrammes. To ascertain the true standard in the most
exact manner possible, a decime solution must be prepared
by weighing 100 grammes of the normal solution, and
diluting it with pure water, so that it shall occupy one
litre : a cubic centimetre of this solution will represent a
decigramme of the normal solution. This decime solution
will not be rigorously exact, since the normal solution has
not been truly standardised ; but it is easily perceived
that the error thus committed is very small, and that it
may be neglected. Nevertheless, as soon as the normal
solution is perfectly standardised, it is better to prepare
another decime solution.
A decime solution may be immediately obtained by
dissolving 0*5427 gramme of pure sea-salt in such a quan-
tity of water that the whole will occupy one litre ; yet the
first process is preferable.
With the decime solution the assay may be thus con-
tinued, remembering that the pipette described at fig. 114
is a cubic centimetre containing 20 drops ; that the half,
therefore, is represented by 10 drops, and the fourth by 5.
To the 998-4 decigrammes of normal solution already
added, pour one pipette and 12 drops of the decime
solution, which will exactly complete the weight of 1000
decigrammes of normal solution. The mixture is agitated
to brighten it, and one-thousandth of common salt or one
pipette of the decime solution added. If this causes a
cloudiness, it is agitated and a second thousandth added.
This last should produce no opalescence. The weight
PREPARATION OF THE NORMAL SALT SOLUTION. 645
of normal solution necessary to exactly precipitate one
gramme of silver will be between 1000 and 1001 deci-
grammes ; that is to say, the mean will be equal to 1000-5.
The standard of the normal solution is then too weak by
half a thousandth ; to correct this a quantity of concen-
trated salt solution must be added equal to half a thou-
sandth of that already added (2-1708 + 0-0353 = 2-2061
kilogrammes) ; that is to say, 1-1 gramme.
A new assay is then made for verification.
When the standard of a solution is very nearly arrived
at, it is well to employ filtration to detect the slightest
opalescence, at least when sufficient time is not .allowed for
the liquid to become perfectly bright. The surest method,
when the standard is nearly attained, is to place some of
the liquid in two test-glasses, and pour into one a few
drops of the decime solution of common salt, and into the
other a corresponding number of drops of the decime
solution of silver nitrate. It may then be determined on
which side the opalescence is manifested, and the assay of
the normal solution may be continued after the mixture of
the liquid in the two glasses, since the two quantities of
the decime solutions of common salt and silver nitrate
mutually decompose each other, and do not interfere with
the assay. Once the standard of the normal solution is
definitely fixed, the sum of the quantities of the concen-
trated solution of common salt which have been employed,
as well as those of the water, must be noted, and in the
preparation of a new normal solution the proportions
found as above would only have to be mixed to obtain at
once a solution having very nearly its true standard.
In determining the standard of the normal solution,
supposing that it were always too weak, it would be
necessary to add to the solution a certain quantity of
common salt ; but if the true amount had been exceeded,
and it had been found too strong, the solution would
have to be precipitated with the decime solution of silver ;
and knowing the number of cubic centimetres or thou-
sandths of silver which had been necessary to ; precipitate
the excess of common salt, it could be estimated what
646 THE ASSAY OF SILVER.
amount of water must be added to reduce the normal
solution to standard. For instance, if 2 thousandths of
the decime solution of silver had been consumed, 2 thou-
sandths of its weight of water would have to be added
to the total amount of solution ; that is to say, 0-2 kilo-
gramme or 200 grammes.
Preservation of the Normal Solution of Common Salt.
The most suitable vessel for containing the normal solution
of common salt is one of glass, because that cannot affect
the standard. Large glass bottles, termed carboys, are
found in commerce. These bottles contain from 50 to 60
litres, and are very applicable for this purpose. Fig. 119
represents one of these bottles fixed in a stand formed of a
sieve hoop. It is graduated into litres or kilogrammes of
water, and a paper scale fixed on its side shows at any
time the quantity of contained liquid. It is closed by an
hydraulic valve, made of sheet iron, but the bell or cover
is of glass. The detail of this valve is shown at fig. 120.
FIG. 119. The a ^ r can on ty en ter the bottle by
the narrow tube T 7 , and cannot pass
out by it ; consequently evaporation
is not to be feared. The neck of
the valve should be about a deci-
metre deep, into which mercury
should be poured, but only to about
one-third of its height.
The solution is drawn from the
bottle by the syphon S. This is
furnished with a stopcock ; but this
syphon being brittle, at least when
not of metal, is not convenient in
use, since it is incorporated with the-
bell of the valve : it is, therefore,
preferable to pierce the bottom of
the bottle (fig. 121), and fix a metal tube (T) by means of
a plate moulded on the bottom and cemented to it. This
tube is raised a little above the bottom of the bottle, and
covered by a small cup, the object of which is to protect
it from any of the mercury which might fall into it. It is
APPLICATION OF GAY-LUSSAC's PKOCESS.
647
terminated at its other extremity by a very narrow tube,
so that the flow of the solution may not be too rapid-
FIG. 120.
FIG. 121.
Fifl. 122.
Hereafter a metal reservoir will be described which has
all the advantages of a glass vessel without its inconve-
niences.
Application of the Process described in the Estimation
of the Standard of a Silver Alloy. The alloy is supposed
to be that made into coin, the mean standard
of which is fixed at 900 thousandths, but
which may vary from 897 to 903 thousandths
without ceasing to be legal (French standard
for coin). One gramme is dissolved in the
bottle (fig. 118) by about 10 grammes of
nitric acid, sp. gr. 1-290. This quantity of
nitric acid can be readily taken by means of
the pipette P (fig. 122), which contains 7'7
grammes of water to the mark a b. The
solution may be accelerated by placing the
bottle in a small pan of hot water, the bottom
of which must be covered with a piece of
cloth, so as to prevent contact of the glass and metal.
The solution finished, and the flask slightly cooled, the
648 THE ASSAY OF SILVER.
nitrous vapour must be removed 4 by a blower (see fig. 123),
the nozzle of which is formed of a piece of bent glass tube,
connected by a cork with a copper socket
D, having a screw inside. This operation
ought to be effected, as well as the solution
of the alloy in nitric acid, under a chimney
with a strong current of air, to carry off the
nitrous vapour.
The burette (fig. 99), being filled with
the normal solution of common salt, and
tared, about 90 grammes are poured into the
solution of the alloy ; say 89 -85 grammes.
After agitating the liquor, a cubic centimetre
of the decime solution of common salt is added
representing one-thousandth of silver. If a cloudiness
be observed, agitate again, and add a second thousandth
of common salt, and so on, until the last thousandth
gives no precipitate. Suppose it to be the fourth :
that must not be counted, because it has produced no
effect ; and only a half of the third must be taken,
because only a portion of that was necessary. The
standard of the alloy would be consequently equal to
If it be desirable to approach still nearer to the true
standard of the alloy, half-thousandths must be added until
the last half-thousandth gives no precipitate ; and in order
to avoid all confusion, it is better to write with chalk on
a blackboard the thousandths of common salt, preceding
them by the plus sign +, and on the other side the thou-
sandths of the silver nitrate, preceding them by the sign
minus.
In the above example, after the addition of the 4
thousandths of common salt, the last of which has pro-
duced no cloudiness, 1^ thousandth of nitrate of silver is
added, which destroys 1^ thousandth of common salt, and
brightens the liquid. If another half-thousandth of nitrate
of silver praduce no precipitate, it is not taken into account,
and is struck off from the table. From whence it is con-
cluded that the quantity of nitrate of silver necessary to
APPLICATION OF GAY-LUSSAC'S PROCESS. 649
destroy the excess of common salt is more than 1 and less
than li ; that is to say, nearly the J of a thousandth, and
is equal to 1J. Thus the number of thousandths of salt
really used is 41-25 = 275. The standard of the alloy,
therefore, is 898-50 + 275 = 901-25.
Another example, everything else remaining as above :
Suppose the first thousandth of salt did not precipitate.
This is a proof that too much normal solution of common
salt has been employed, and that there is an excess of salt
in the liquid. Add one- thousandth of silver, and agitate :
things are now as at first, but it is nevertheless known
that it is with nitrate of silver the process must be con-
tinued. One-thousandth has been added, which produced
a precipitate ; the second does not. The standard of the
.alloy is consequently 898-5 0-5 = 898. To approach still
nearer to the real standard, destroy the last 2 thousandths
of silver by 2 thousandths of common salt, and add half a
thousandth of silver a cloudiness is produced, as already
known ; but another half-thousandth does not precipitate.
The standard of the alloy is therefore 898-50-0-25 =
898-25.
This process, on which it would be useless to enlarge
further at present, because many other parts of the pro-
cess to be presently described apply to it, is general, and
gives exactly the standard of an alloy when it is known
approximately, which can always be ascertained by a
previous rough assay.
Correction of the Standard of the Normal Solution of
Salt when the Temperature Varies. It has been assumed
that, in the estimation of the standard of the normal
solution of salt, the temperature has remained constant.
Assays made under these circumstances need no correc-
tion ; but if the temperature changes, the same measure
of solution will not contain the same amount of salt. Sup-
posing the solution of salt has been standardised at 15.
If, at the time an experiment is made, the temperature
is 18, for instance, the solution will be found too weak,
since it has expanded, and the pipette holds less than its
proper weight. If, on the other hand, the temperature
650 THE ASSAY OF SILVEE.
falls to 12, .the solution becomes concentrated, and is
found too strong. It is therefore necessary to estimate
the correction to be made for any variation of temperature
that may occur.
To this end the temperature of a solution of common
salt has been gradually raised from 0... .5. ...10... .15. ...20....
25.. ..30 degrees, and three pipettefuls of the solution
exactly weighed at each of the above temperatures. One-
third of the total weight gives the mean weight of the
contents of a pipette. The corresponding weights of a
pipetteful of solution are then entered, and form the second
column of the following table, called ' Table of Correction
for the Variations of Temperature in the Normal Solution
of Salt/ By this table correction may be made for any
temperature between and 30, when the solution of
salt has been standardised within the same limits. Sup-
pose, for example, the solution had been standardised at
15, and that at the time it was used its temperature was
18. On referring to the second column of the table, it
will be seen that the weight o a measure of solution at
15 is 100-099 gr. ; and at 18, 100-065 gr. ; the difference
0-034 gr. is the quantity of solution taken too little, and
consequently it must be added to the normal measure, so
that it may be equal to one thousand thousandths. If the
temperature of the solution had fallen to 10, the differ-
ence of weight between a measure at 10 and a measure at
15 will be 0-019 gr., which must, on the contrary, be de-
ducted from the measure, as it has been taken in excess.
These differences of weight of a measure of solution at 15,
and that of a measure for any other temperature, form the
column 15 in the table, where they are expressed in thou-
sandths. They are written on the same horizontal line as the
temperatures to which each corresponds, with the sign +
when they are to be added, and the sign when to be
subtracted. The columns 5, 10, 20, 25, 30 have been
calculated in the same manner, to meet cases in which the
normal solution had been graduated at each of the above-
named temperatures. Thus, to calculate the column 10,
take the number 100-118 from the column of weights as
CORRECTION OF THE STANDARD OF SOLUTION OF SALT. 651
a point of departure, and find the difference for all the
other numbers in the same column.
An application of this Table will be given hereafter.
TABLE OF CORRECTIONS FOR VARIATIONS IN TEMPERATURE OF THE NORMAL
SALT SOLUTION.
Temperature
Weight
5
10
15
20
25
30
Degrees
Grammes
Mill.
Mill.
Mill.
Mill.
Mill.
Mill.
4 100-109
o-o
-o-i
+ 0-1
+ 07
+ 1-7
+ 2-7
5 100-113
o-o
-0-1
+ 0-1
+ 0-7
+ 1-7
+ 2-8
6 100-115
o-o
o-o
+ 0-2
40-8
hl-7
+ 2-8
7 100-118
+ 0-1
o-o
+ 0-2
+ 0-8
+ 1-7 +2-8
8 100-120
+ 0-1
o-o
+ 0-2
+ 0-8
+ 1-8
+ 2-8
9 100-120
+ 0-1
o-o
+ 0-2
+ 0-8
+ 1-8
+ 2-8
10 100-118
+ 0-1
o-o
+ 0-2
+ 0-8
4-1-7
+ 2-8
11 100-116
o-o
o-o
+ 0-2
+ 0-8
+ 1-7
+ 2-8
12
100-114
o-o
o-o
+ 0-2
+ 0-8
+ 1-7
+ 2-8
13
100-110
o-o
-o-i
+ 0-1
+ 0-7
+ 1-7
+ 2-7
14
100-106
-0-1
-0-1 +0-1
+ 0-7 +1-6
+ 2-7
15
100-099
-0-1
-0-2 0-0
+ 0-6 +1-6
+ 2-6
16
100-090
-0-2
-0-3 -0-1
+ 0-5 + 1-5
+ 2-5
17
100-078
-0-4
-0-4
-0-2
+ 0-4
+ 1-3
+ 2-4
18
100-065
-0-5
-0-5 -0-3
+ 0-3
+ 1-2
+ 2-3
19
100-053
-o-o
-0-7
-0-5
+ 0-1
+ 1-1
+ 2-2
20
100-039
-0-7
-0-8
-0-6
0-0
+ 1-0
4-2-0
21
100-021
-0-9
-1-0
-0-8
-0-2
+ 0-8
+ 1-9
22
100-001
-M
-1-2
-1-0
-0-4
+ 0-6
4-1-7
23
99-983
-1-3
-1-4
-1-2 -0-6
+ 0-4
+ 1-5
24 99-964
-1-5
-1-5
-1-4
-0-8
+ 0-2
+ 1-3
25 99-944
-1-7
-1-7
-1-6 1 -1-0
o-o
+ 1-1
26
99-924
-1-9
-1-9
-1-8 1 -1-2 -0-2
+ 0-9
27
99-902
-2-1
-2-2
-2-0 -1-4
-0-4
+ 0-7
28
99-879
-2-3
-2-4
-2-2
-1-6
-0-7
+ 0-4
29
99-858
-2-6
-2-6
-2-4
-1-8
-0-9
-t-0-2
30
99-836
-2-8
-2-8
-2-6
-2-0
-1-1
0-0
TABLE FOR THE ASSAY, BY THE WET METHOD, OF AN ALLOY
CONTAINING ANY PROPORTIONS WHATEVER OF SILVER, BY
THE EMPLOYMENT OF A CONSTANT MEASURE OF THE NORMAL
SOLUTION OF COMMON SALT.
The process by which the normal solution of salt is
measured by weight is applicable to the assay of every
kind of alloy, since it suffices to take a weight of the solu-
tion corresponding to the presumed standard of the silver,
and complete the assay by means of the decime solu-
tion ; the process by volume, however, has not the same
652 THE ASSAY OF SILVER.
advantage, because the volume of normal solution cannot
be varied in the same manner as the weight. This incon-
venience, however, is of no very great consequence, for,
by keeping the volume of normal solution constant, it
suffices to vary the weight of the alloy, taking in each
particular case a weight which contains approximative] y
one gramme of pure silver. Suppose the alloy has a
standard of about 900 thousandths, we have the following
proportion :
900 thousandths : 1000 of alloy :: 1000 thousandths:
x = 1111-1.
If that weight be now taken to ascertain the standard
of the alloy, it may be found, for instance, that to the mea-
sure of 1000 thousandths of salt it is yet necessary to add
4 thousandths of salt to precipitate the whole of the silver ;
that is to say, that 11.11-1 of alloy really contain 1004 of
silver. From this result the real standard of the alloy may
be found to be 903-6, by the following equation :
1111-1 : 1004 : : 1000 : x = 903-6.
But such calculations, however simple, should be
.avoided where numerous daily assays are made, not only
on account of the time consumed, but still more from the
errors to which such operations are necessarily exposed.
Fortunately, all these inconveniences may be avoided by
the use of tables, which entirely free the assay er from
calculation.
Wishing in weighing the alloy to avoid fractions of
thousandths, and only making use of tenths and half-tenths
of thousandths, the weight of alloy increases, starting from
a gramme, from 5 to 5 thousandths, and the correspond-
ing standard for each of these weights has been sought, all
containing one gramme of pure silver. Thus the weight
1020 of alloy, in which there are 1000 of silver and 20 of
copper, corresponds to the standard 980-39, obtained by
the proportion
1020 : 1 000 : : 1000 : x = 980-39.
EXPLANATION OF THE FOLLOWING TABLES. 653
On this principle are formed the first and second
columns of the table marked Salt. The first contains the
weight of each alloy, and the second its corresponding
standard. The following columns, 1, 2, 3, to 10, give the
standard of the alloy, when, instead of the 1000 milli-
grammes of silver it was supposed to contain, it really
contained 1, 2, 3, &c. more, and consequently 1, 2, 3, (fee-
milligrammes of copper less.
Another table, constructed in the same manner as the
preceding, and marked Nitrate of Silver, gives the standard
of the alloy when, under the weight given in the first
column, it contains 1, 2, 3, <fec. milligrammes less silver,
and as much more copper. Thus, for example, an alloy of
the weight of 1020 (1000 silver and 20 copper) has for its
standard 980-4 in both tables. If it always contains in the
same weight 4 more silver and consequently 4 less copper,
its standard would be 984*3, and would be found in the
4 Salt ' table at the intersection of the column 4, and the
horizontal line 1020. If, on the contrary, it contains 4
less of silver and 4 more of copper, its standard will be
976-5, and will be found in the 'Nitrate of Silver ' table, at
the intersection of the column 4, and the horizontal line
1020.
654
THE ASSAY OF SILVER.
Tables for Estimating the Standard of any Silver
approximativcly containing
NITRATE OF
Weight of
Assay in
0.
i.
2.
3.
4.
Milligrs.
1000
1000-0
999-0
998-0
997-0
996-0
1005
995-0
994-0
993-0
992-0
991-0
1010
990-1
989-1
988-1
987-1
986-1
1015
985-2
984-2
983-2
982-3
981-3
1020
980-4
979-4
978-4
977-4
976-5
1025
975-6
974-6
973-7
972-7
971-7
1030
970-9
969-9
968-9
968-0
967-0
1035
966-2
965-2
964-2
963-3
962-3
1040
961-5
960-6
959-6
958-6
957-7
1045
956-9
956-0
955-0
954-1
953-1
1050
952-4
951-4
950-5
949-5
948-6
1055
947-9
946-9
946-0
945-0
944-1
1060
943-4
942-4
941-5
940-6
939-6
1065
939-0
938-0
937-1
936-1
935-2
1070
934-6
933-6
932-7
931-8
930-8
1075
930-2
929-3
928-4
927-4
926-5
1080
925-9
925-0
924-1
923-1
922-2
1085
921-7
920-7
919-8
918-9
918-0
1090
917-4
916-5
915-6
914-7
913-8
1095
913-2
912-3
911-4
910-5
909-6
1100
909-1
908-2
907-3
906-4
905-4
1105
905-0
904-1
903-2
902-3
901-4
1110
900-9
900-0
899-1
898-2
897-3
1115
896-9
896-0
895-1
894-2
893-3
1120
892-9
892-0
891-1
890-2
889-3
1125
888-9
888-0
887-1
886-2
885-3
1130
885-0
884-1
883-2
882-3
881-4
1135
881-1
880-2
879-3
878-4
877-5
1140
877-2
876-3
875-4
874-6
873-7
1145
873-4
872-5
871-6
870-7
869-9
1150
869-6
868-7
867-8
867-0
866-1
1155
865-8
864-9
864-1
863-2
862-3
1160
862-1
861-2
860-3
859-5
858-6
1165
858-4
857-5
856-6
855-8
854-9
1170
854-7
853-8
853-0
852-1
851-3
1175
851-1
850-2
849-4
848-5
847-7
1180
847-5
846-6
845-8
844-9
844-1
1185
843-9
843-0
842-2
841-3
840-5
TABLE FOR THE WET ASSAY OF SILVER.
655
Alloy by employing an Amount of Alloy always
the same Amount of Silver.
SILVER.
5.
6.
7.
8.
9.
10.
995-0
994-0
993-0
992-0
991-0
990-0
990-0
989-0
988-1
987-1
986-1
985-1
985-1
984-2
983-2
982-2
981-2
980-2
980-3
979-3
978-3
977-3
976-4
975-4
975-5
974-5
973-5
972-5
971-6
970-6
970-7
969-8
968-8
967-8
966-8
965-8
966-0
965-0
964-1
963-1
962-1
961-2
961-3
960-4
959-4 | 958-4
957-5
956-5
956-7
955-8
954-8
953-8
952-9
951-9
952-1
951-2
950-2
949-3
948-3
947-4
947-6
946-7
945-7
944-8
943-8
942-9
943-1
942-2
941-2
940-3
939-3
938-4
938-7
937-7
936-8
935-8
934-9
934-0
934-3
933-3
932-4
931-4
930-5
929-6
929-9
929-0
928-0
927-1
926-2
925-2
925-6
924-7
923-7
922-8
921-9
920-9
921-3
920-4
919-4
918-5
917-6
916-7
917-0
916-1
915-2
914-3
913-4
912-4
912-8
911-9
911-0
910-1
909-2
908-3
908-7
907-8
906-8
905-9
905-0
904-1
904-5
903-6
902-7
901-8
900-9
900-0
900-4
899-5
898-6
897-7
896-8
895-9
896-4
895-5
894-6
893-7
892-8
891-9
892-4
891-5
890-6
889-7
888-8
887-9
888-4
887-5
886-6
885-7
884-8
883-9
884-4
883-6
882-7
881-8
880-9
880-0
880-5
879-6
878-8
877-9
877-0
876-1
876-7
875-8
874-9
874-0
873-1
872-3
872-8
871-9
871-0
870-2
869-3
868-4
869-0
868-1
867-2
866-4
865-5
864-6
865-2
864-3
863-5
862-6
861-7
860-9
861-5
860-6
859-7
858-9
858-0
857-1
857-8
856-9
856-0
855-2
854-3
853-4
854-1
853-2
852-4
851-5
850-6
849-8
850-4
849-6
848-7
847-9
847-0
846-1
846-8
846-0
845-1
844-3
843-4
842-5
843-2
842-4
841-5
840-7
839-8
839-0
839-7
838-8
838-0
837-1
836-3
835-4
(156
THE ASSAY OF SILVER.
NITKATE OF
Weight of
Assay in
Milligrs.
0.
i.
2.
3.
4.
1190
840-3
849-5
838-7
837-8
837-0
1195
836-8
836-0
835-1
834-3
833-5
1200
833-3
832-5
831-7
830-8
830-0
1205
829-9
829-0
828-2
827-4
826-6
1210
826-4
825-6
824-8
824-0
823-1
1215
823-0
822-2
821-4
820-6
819-7
1220
819-7
818-8
818-0
817-2
816-4
1225
816-3
815-5
814-7
813-9
813-1
1230
813-0
812-2
811-4
810-6
809-8
1235
809-7
808-9
808-1
807-3
806-5
1240
806-5
805-6
804-8
804-0
803-2
1245
803-2
802-4
801-6
800-8
800-0
1250
800-0
799-2
798-4
797-6
796-8
1255
796-8
796-0
795-2
794-4
793-6
1260
793-6
792-9
792-1
791-3
790-5
1265
790-5
789-7
788-9
788-1
787-3
1270
787-4
786-6
785-8
785-0
784-2
1275
784-3
783-5
782-7
782-0
781-2
1280
781-2
780-5
779-7
778-9
778-1
1285
778-2
777-4
776-6
775-9
775-1
1290
775-2
774-4
773-6
772-9
772-1
1295
772-2
771-4
770-7
769-9
769-1
1300
769-2
768-5
767-7
766-9
766-1
1305
766-3
765-5
764-7
764-0
763-2
1310
763-4
762-6
761-8
761-1
760-3
1315
760-5
759-7
758-9
758-2
757-4
1320
757-6
756-8
756-1
755-3
754-5
1325
754-7
754-0
753-2
752-4
751-7
1330
751-9
751-1
750-4
749-6
748-9
1335
749-1
748-3
747-6
746-8
746-1
1340
746-3
745-5
744-8
744-0
743-3
1345
743-5
742-7
742-0
741-3
740-5
1350
740-7
740-0
739-3
738-5
737-8
1355
738-0
737-3
736-5
735-8
735-1
1360
735-3
734-6
733-8
733-1
732-4
1365
732-6
731-9
731-1
730-4
729-7
1370
729-9
729-2
728-5
727-7
727-0
1375
727-3
726-5
725-8
725-1
724-4
1380
724-6
723-9
723-2
722-5
721-7
1385
722-0
721-3
720-6
719-9
719-1
1390
719-4
718-7
718-0
717-3
716-5
1395
716-8
716-1
715-4 714-7
714-0
1400
714-3
713-6
712-9 712-1
711-4
TABLE FOR THE WET ASSAY OF SILVER.
657
SILVER continued.
6.
6.
7.
8.
9.
10.
836-1
835-3
834-5
833-6
832-8
831-9
832-6
831-8
831-0
830-1
829-3
828-4
829-2
828-3
827-5
826-7
825-8
825-0
825-7
824-9
824-1
823-2
822-4
821-6
822-3
821-5
820-7
819-8
819-0
818-2
818-9
818-1
817-3
816-5
815-6
814-8
815-6
814-7
813-9
813-1
812-3
811-5
812-2
811-4
810-6
809-8
809-0
808-2
808-9
808-1
807-3
806-5
805-7
804-9
805-7
804-9
804-0
803-2
802-4
801-6
802-4
801-6
800-8
800-0
799-2
798-4
799-2
798-4
797-6
796-8
796-0
795-2
796-0
795-2
794-4
793-6
792-8
792-0
792-8
792-0
791-2
790-4
789-6
788-8
789-7
788-9
788-1
787-3
786-5
785-7
786-6
785-8
785-0
784-2
783-4
782-6
783-5
782-7
781-9
781-1
780-3
779-5
780-4
779-6
778-8
778-0
777-3
776-5
777-3
776-6
775-8
775-0
774-2
773-4
774-3
773-5
772-8
772-0
. 771-2
770-4
771-3
770-5
769-8
769-0
768-2
767-4
768-3
767-6
766-8
766-0
765-2
764-5
765-4
764-6
763-8
763-1
762-3
761-5
762-4
761-7
760-9
760-1
759-4
758-6
759-5
758-8
758-0
757-2
756-5
755-7
756-6
755-9
755-1
754-4
753-6
752-8
753-8
753-0
752-3
751-5
750-8
750-0
750-9
750-2
749-4
748-7
747-9
747-2
748-1
747-4
746-6
745-9
745-1
744-4
745-3
744-6
743-8
743-1
742-3
741-6
742-5
741-8
741-0
740-3
739-5
738-8
739-8
739-0
738-3
737-5
736-8
736-1
737-0
736-3
735-6
734-8
734-1
733-3
734-3
733-6
732-8
732-1
731-4
730-6
731-6
730-9
730-1
729-4
728-7
727-9
728-9
728-2
727-5
726-7
726-0
725-3
726-3
725-5
724-8
724-1
723-4
722-6
723-6
722-9
722-2
721-4
720-7
720-0
721-0
720-3
719-6
718-8
718-1
717-4
718-4
717-7
717-0
716-2
715-5
714-8
715-8
715-1
714-4
713-7
712-9
712-2
713-3
712-5
711-8
711-1
710-4
709-7
710-7
710-0
709-3
708-6
707-9
707-1
u u
658
THE ASSAY OF SILVER.
NITRATE 01
Weight of
Assay in
Milligrs.
0.
i.
2.
3.
4.
1405
711-7
711-0
710-3
709-6
708-9
1410
709-2
708-5
707-8
707-1
706-4
1415
706-7
706-0
705-3
704-6
703-9
1420
704-2
703-5
702-8
702-1
701-4
1425
701-8
701-0
700-3
699-6
698-9
1430
699-3
698-6
697-9
697-2
696-5
1435
696-9
696-2 i 695-5
694-8
694-1
1440
694-4
693-7
693-1
692-4
691-7
1445
692-0
691-3
690-7
690-0
689-3
1450
689-7
689-0
688-3
687-6
686-9
1455
687-3
686-6
685-9
685-2
684-5
1460
684-9
684-2
683-6
682-9
682-2
1465
682-6
681-9
681-2
680-6
679-9
1470
680-3
679-6
678-9
678-2
677-5
1475
678-0
677-3
676-6
675-9
675-2
1480
675-7
675-0
674-3
673-6
673-0
1485
673-4
672-7
672-0
671-4
670-7
1490
671-1
670-5
669-8
669-1
668-5
1495
668-9
668-2
667-6
666-9
666-2
1500
666-7
666-0
665-3
664-7
664-0
1505
664-5
663-8
663-1
662-5
661-8
1510
662-3
661-6
660-9
660-3
659-6
1515
660-1
659-4
658-7
658-1
657-4
1520
657-9
657-2
656-6
655-9
655-3
1525
655-7
655-1
654-4
653-8
653-1
1530
653-6
652-9
652-3
651-6
651-0
1535
651-5
650-8
650-2
649-5
648-9
1540
649-4
648-7
648-0
647-4
646-7
1545
647-2
646-6
645-9
645-3
644-7
1550
645-2
644-5
643-9
643-2
642-6
1555
643-1
642-4
641-8
641-2
640-5
1560
641-0
640-4
639-7
639-1
638-5
1565
639-0
638-3
637-7
637-1
636-4
1570
636-9
636-3
635-7
635-0
634-4
1575
634-9
634-3
633-6
633-0
632-4
1580
632-9
632-3
631-6
631-0
630-4
1585
630-9
630-3
629-6
629-0
628-4
1590
628-9
623-3
627-7
627-0
626-4
1595
627-0
626-3
625-7
625-1
624-4
1600
625-0
624-4
623-7
623-1
622-5
1605
623-1
622-4
621-8
621-2
620-6
1610
621-1
620-5
619-9
619-2
618-6
1615
619-2
618-6
618-0
617-3
616-7
TABLE FOE THE WET ASSAY OF SILVER.
SILVEK continued.
10.
5.
6.
7.
8.
9.
708-2
707-5
706-8
706-0
705-3
704-6
705-7
705-0
704-3
703-5
702-8
702-1
703-2
702-5
701-8
701-1
700-3
699-6
700-7
700-0
699-3
698-6
697-9
697-2
698-2
697-5
696-8
696-1
695-4
694-7.
695-8
695-1
694-4
693-7
693-0
692-3
693-4
692-7
692-0
691-3
690-6
689-9
691-0
690-3
689-6
688-9
688-2
687-5
688-6
687-9
687-2
686-5
685-8
685-1
686-2
685-5
684-8
684-1
683-4
682-8
683-8
683-2
682-5
681-8
681-1
680-4
681-5
680-8
680-1
679-4
678-8
678-1
679-2
678-5
677-8
677-1
676-4
675-8
676-9
676-2
675-5
674-8
674-1
673-5
674-6
673-9
673-2
672-5
671-9
671-2
672-3
671-6
670-9
670-3
669-6
668-9
670-0
669-4
668-7
668-0
667-3
666-7
667-8
667-1
666-4
665-8
665-1
664-4
665-5
664-9
064-2
663-5
662-9
662-2
663-3
662-7
662-0
661-3
660^7
660-0
661-1
660-5
659-8
659-1
658-5
657-8
658-9
658-3
657-6
656-9
656-3
655-6
656-8
656-1
655-4
654-8
654-1
653-5
654-6
653-9
653-3
652-6
652-0
651-3
652-5
651-8
651-1
650-5
649-8
649-2
650-3
649-7
649-0
648-4
647-7
647-1
648-2
647-6
646-9
646-2
645-6
644-9
646-1
645-4
644-8
644-2
643-5
642-9
644-0
643-4
642-7
642-1
641-4
640-8
641-9
641-3
640-6
640-0
639-3
638-7
639-9
639-2
638-6
637-9
637-3
636-7
637-8
637-2
636-5
635-9
635-3
634-6
635-8
635-1
634-5
633-9
633-2
632-6
633-8
633-1
632-5
631-8
631-2
630-6
631-7
631-1
630-5
629-8
629-2
628-6
629-7
629-1
628-5
627-8
627-2
626-6
627-8
627-1
626-5
625-9
625-2
624-6
625-8
625-2
624-5
623-9
623-3
622-6
623-8
623-2
622-6
621-9
621-3
620-7
621-9
621-2
620-6
620-0
619-4
618-7
619-9
619-3
618-7
618-1
617-4
616-1
618-0
617-4
616-8
616-1
615-5
614-9
616-1
615-5
614-9
614-2
613-6
613-0
660
THE ASSAY OP SILVEK.
NITKATE OF
Weight of
Assay in
Milligrs.
0.
l.
2.
3.
4.
1620
617-3
616-7
616-0
615-4
614-8
1625
615-4
614-8
614-1
613-5
612-9
1630
613-5
612-9
612-3
611-7
611-0
1635
611-6
611-0
610-4
609-8
609-2
1640
609-8
609-1
608-5
607-9
607-3
1645
607-9
607-3
606-7
606-1
605-5
1650
606-1
605-4
604-8
604-2
603-6
1655
604-2
603-6
603-0
602-4
601-8
1660
602-4
601-8
601-2
600-6
600-0
1665
600-6
600-0
599-4
598-8
598-2
1670
598-8
598-2
597-6
597-0
596-4
1675
597-0
596-4
595-8
595-2
594-6
1680
595-2
594-6
594-0
593-4
592-9
1685
593-5
592-9
592-3
591-7
591-1
1690
591-7
591-1
590-5
589-9
589-3
1695
590-0
589-4
588-8
588-2
587-6
1700
588-2
587-6
587-1
586-5
585-9
1705
586-5
585-9
585-3
584-7
584-2
1710
584-8
584-2
583-6
583-0
582-5
1715
583-1
582-5
581-9
581-3
580-8
1720
581-4
580-8
580-2
579-6
579-1
1725
579-7
579-1
578-5
578-0
577-4
1730
578-0
577-5
576-9
576-3
575-7
1735
576-4
575-8
575-2
574-6
574-1
1740
574-7
574-1
573-6
573-0
572-4
1745
573-1
572-5
571-9
571-3
570-8
1750
571-4
570-9
570-3
569-7
569-1
1755
569-8
569-2
568-7
568-1
567-5
1760
568-2
567-6
567-0
566-5
565-9
1765
566-6
566-0
565-4
564-9
564-3
1770
565-0
564-4
563-8
563-3
562-7
1775
563-4
562-8
562-2
561-7
561-1
1780
561-8
561-2
560-7
560-1
559-5
1785
560-2
559-7
559-1
558-5
558-0
1790
558-7
558-1
557-5
557-0
556-4
1795
557-1
556-5
556-0
555-4
554-9
1800
555-6
555-0
554-4
553-9
553-3
1805
554-0
553-5
552-9
552-3
551-8
1810
552-5
551-9
551-4
550-8
550-3
1815
551-0
550-4
549-9
549-3
548-8
1820
549-4
548-9
548-3
547-8
547-2
1825
547-9
547-4
546-8
546-3
545-7
1830
546-4
545-9
545-4
544-8
544-3
TABLE FOR THE WET ASSAY OF SILVER.
661
SILVER continued.
5.
6.
7.
8.
9.
10.
614-2
613-6
613-0
612-3
611-7
611-1
612-3
611-7
611-1
610-5
609-8
609-2
610-4
609-8
609-2
608-6
608-0
607-4
608-6
607-9
607-3
606-7
606-1
605-5
606-7
606-1
605-5
604-9
604-3
603-7
604-9
604-3
603-6
603-0
602-4
601-8
603-0
602-4
601-8
601-2
600-6
600-0
601-2
600-6
600-0
599-4
598-8
598-2
599-4
598-8
598-2
597-6
597-0
596-4
597-6
597-0
596-4
595-8
595-2
594-6
595-8
595-2
594-6
594-0
593-4
592-8
594-0
593-4
592-8
592-2
591-6
591-0
592-3
591-7
591-1
590-5
589-9
589-3
590-5
589-9
589-3
588-7
588-1
587-5
588-8
588-2
587-6
587-0
586-4
585-8
587-0
586-4
585-8
585-2
584-7
584-1
585-3
584-7
584-1
583-5
582-9
582-3
583-6
583-0
582-4
581-8
581-2
580-6
581-9
581-3
580-7
580-1
579-5
578-9
580-2
579-6
579-0
578-4
577-8
577-3
578-5
577-9
577-3
576-7
576-2
575-6
576-8
576-2
575-6
575-1
574-5
573-9
575-1
574-6
574-0
573-4
572-8
572-2
573-5
572-9
572-3
571-8
571-2
570-6
571-8
571-3
570-7
570-1
569-5
569-0
570-2
569-6
569-0
568-5
567-9
567-3
568-6
568-0
567-4
566-9
566-3
565-7
566-9
566-4
565-8
565-2
564-7
564-1
565-3
564-8
564-2
563-6
563-1
562-5
563-7
563-2
562-6
562-0
561-5
560-9
562-1
561-6
561-0
560-4
559-9
559-3
560-6
560-0
559-4
558-9
558-3
557-7
559-0
558-4
557-9 *
557-3
556-7
556-2
557-4
556-9
556-3
555-7
555-2
554-6
555-9
555-3
554-7
554-2
553-6
553-1
554-3
553-8
553-2
552-6
552-1
551-5
552-8
552-2
551-7
551-1
550-6
550-0
551-2
550-7
550-1
549-S
549-0
548-5
549-7
549-2
548-6
548-1
547-5
547-0
548-2
547-7
547-1
546-6
546-0
545-5
546-7
546-2
545-6
545-1
544-5
544-0
545-2
544-7
544-1
543-6
543-0
542-5
543-7
543-2
542-6
542-1
541-5
541-0
(56-2
THE ASSAY OF SILVER.
NITKATE OF
Weight of
Assay in
Milligrs.
0.
i.
2.
3.
4.
1835
545-0
544-4
543-9
543-3
542-8
1840
543-5
542-9
542-4
541-8
541-3
1845
542-0
541-5
540-9
540-4
539-8
1850
540-5
540-0
539-5
538-9
538-4
1855
539-1
538-5
538-0
537-5
536-9
1860
537-6
537-1
536-6
536-0
535-5
1865
536-2
535-7
535-1
534-6
534-0
1870
534-8
534-2
533-7
533-2
532-6
1875
533-3
532-8
532-3
531-7
531-2
1880
531-9
531-4
530-8
530-3
529-8
1885
530-5
530-0
529-4
528-9
528-4
1890
529-1
528-6
528-0
527-5
527-0
1895
527-7
527-2
526-6
526-1
525-6
1900
526-3
525-8
525-3
524-7
524-2
1905
524-9
524-4
523-9
523-4
522-8
1910
523-6
523-0
522-5
522-0
521-5
1915
522-2
521-7
521-1
520-6
520-1
1920
520-8
520-3
519-8
519-3
518-7
1925
519-5
519-0
518-4
517-9
517-4
1930
518-1
517-6
517-1
516-6
516-1
1935
516-8
516-3
515-8
515-2
514-7
1940
515-5
514-9
514-4
513-9
513-4
1945
514-1
513-6
513-1
512-6
512-1
1950
512-8
512-3
511-8
511-3
510-8
1955
511-5
511-0
510-5
510-0
509-5
1960
510-2
509-7
509-2
508-7
508-2
1965
508-9 .
508-4
507-9
507-4
506-9
1970
507-6
507-1
506-6
506-1
505-6
1975
506-3
505-8
505-3
504-8
504-3
1980
505-0
504-5
504-0
503-5
503-0
1985
503-8
503-3
502-8
502-3
501-8
1990
502-5
502-0
501-5
501-0
500-^
1995
501-3
500-7
500-2
499-7
499-2
2000
500-0
499-5
499-0
498-5
498-0
TABLE FOR THE WET ASSAY OF SILVER.
663
SILVER continued.
5.
6.
7.
8.
9.
10.
542-2
541-7
541-1
540-6
540-0
539-5
540-8
540-2
539-7
539-1
538-6
538-0
539-3
538-7
538-2
537-7
537-1
536-6
537*8
537-3
536-8
536-2
535-7
535-1
536-4
535-8
535-3
534-8
534-2
533-7
534-9
534-4
533-9
533-3
532-8
532-3
533-5
533-0
532-4
531-9
531-4
530-8
532-1
531-5
531-0
530-5
529-9
529-4
530-7
530-1
529-6
529-1
528-5
528-0
529-3
528-7
528-2
527-7
527-1
526-6
527-8
527-3
526-8
526-3
525-7
525-2
526-5
525-9
525-4
524-9
524-3
523-8
525-1
524-5
524-0
523-5
523-0
522-4
523-7
523-2
522-6
522-1
521-6
521-0
522-3
521-8
521-3
520-7
520-2
519-7
520-9
520-4
519-9
519-4
518-8
518-3
519-6
519-1
518-5
518-0
517-5
517-0
518-2
517-7
517-2
516-7
516-1
515-6
516-9
516-4
515-8
515-3
514-8
514-3
515-5
515-0
514-5
514-0
513-5
512-9
514-2
513-7
513-2
512-7
512-1
511-6
512-9
512-4
511-9
511-3
510-8
510-3
511-6
511-0
510-5
510-0
509-5
509-0
510-3
509-7
509-2
508-7
508-2
507-7
508-9
508-4
507-9
507-4
506-9
506-4
507-6
507-1
506-6
506-1
505-6
505-1
506-4
505-8
505-3
504-8
504-3
503-8
505-1
504-6
504-1
503-5
503-0
502-5
503-8
503-3
502-8
502-3
501-8
501-3
502-5
509-0
501-5
501-0
500-5
500-0
501-3
500-8
500-2
499-7
499-2
498-7
500-0
499-5
499-0
498-5
498-0
497-5
498-7
498-2
497-7
497-2
496-7
496-2
497-5
497-0
496-5
496-0
495-5
495-0
C64
THE ASSAY OF SILVER.
Tables for Determining the Standard of any Silver
approximatively containing
COMMON
Weight of
Assay in
Milligrs.
0.
i.
2.
3.
4.
1000
1000-0
1005
995-0
996-0
997-0
998-0
999-0
1010
990-1
991-1
992-1
993-1
994-1
1015
985-2
986-2
987-2
988-2
989-2
1020
980-4
981-4
982-4
983-3
984-3
1025
975-6
976-6
977-6
978-5
979-5
1030
970-9
971-8
972-8
973-8
974-8
1035
966-2
967-1
968-1
969-1
970-0
1040
961-5
962-5
963-5
964-4
965-4
1045
956-9
957-9
958-8
959-8
960-8
1050
952-4
953-3
954-3
955-2
956-2
1055
947-9
948-8
949-8
950-7
951-7
1060
943-4
944-3
945-3
946-2
947-2
1065
939-0
939-9
940-8
941-8
942-7
1070
934-6
935-5
936-4
937-4
938-3
1075
930-2
931-2
932-1
933-0
933-9
1080
925-9
926-8
927-8
928-7
929-6
1085
921-7
922-6
923-5
924-4
925-3
1090
917-4
918-3
919-3
920-2
921-1
1095
913-2
914-2
915-1
916-0
917-0
1100
909-1
910-0
910-9
911-8
912-7
1105
905-0
905-9
906-8
907-7
908-6
1110
900-9
901-8
902-7
903-6
904-5
1115
896-9
897-8
898-6
899-5
900-4
1120
892-9
893-7
894-6
895-5
896-4
1125
888-9
889-8
890-7
891-6
892-4
1130
885-0
885-8
886-7
887-6
888-5
1135
881-1
881-9
882-8
883-7
884-6
1140
877-2
878-1
878-9
879-8
880-7
1145
873-4
874-2
875-1
876-0
876-9
1150
869-6
870-4
871-3
872-2
873-0
1155
865-8
866-7
867-5
868-4
869-3
1160
862-1
862-9
863-8
864-7
865-5
1165
858-4
859-2
860-1
860-9
861-8
1170
854-7
855-6
856-4
857-3
858-1
1175
851-1
851-9
852-8
853-6
854-5
1180
847-5
848-3
849-2
850-0
850-8
1185
843-9
844-7
845-6
846-4
847-3
TABLE FOB THE WET ASSAY OF SILVER.
665
Alloy by employing an Amount of Alloy always
the same Amount of Silver.
SALT.
5.
6.
7.
8.
9.
10.
1000-0
995-0
996-0
997-0
998-0
999-0
1000-0
990-1
991-1
992-1
993-1
994-1
995-1
985-3
986-3
987-2
988-2
989-2
990-2
980-5
981-5
982-4
983-4
984-4
985-4
975-7
976-7
977-7
978-6
979-6
980-6
971-0
972-0
972-9
973-9
974-9
975-8
966-3
967-3
968-3
969-2
970-2
971-1
961-7
962-7
963-6
964-6
965-5
966-5
957-1
958-1
959-0
960-0
960-9
961-9
952-6
953-5
954-5
955-4
956-4
957-3
948-1
949-1
950-0
950-9
951-9
952-8
943-7
944-6
945-5
946-5
947-4
948-4
939-3
940-2
941-1
942-1
943-0
943-9
934-9
935-8
936-7
937-7
938-6
939-5
930-6
931-5
932-4
933-3
934-3
935-2
926-3
927-2
928-1
929-0
930-0
930-9
922-0
922-9
923-8
924-8
925-7
926-6
917-8
918-7
919-6
920-5
921-5
922-4
913-6
914-5
915-4
916-4
917-3
918-2
909-5
910-4
911-3
912-2
913-1
914-0
905-4
906-3
907-2
908-1
909-0
909-9
901-3
902-2
903-1
904-0
904-9
905-8
897-3
898-2
899-1
900-0
900-9
901-8
893-3
894-2
895-1
896-0
896-9
897-8
889-4
890-3
891-1
892-0
892-9
893-8
885-5
886-3
887-2
888-1
889-0
889-9
881-6
882-5
883-3
884-2
885-1
886-0
877-7
878-6
879-5
880-3
881-2
882-1
873-9
874-8
875-7
876-5
877-4
878-3
870-1
871-0
871-9
872-7
873-6
874-5
866-4
867-2
868-1
869-0
869-8
870-7
862-7
863-5
864-4
865-2
866-1
866-9
859-0
859-8
860-7
861-5
862-4
863-2
855-3
856-2
857-0
857-9
858-7
859-6
851-7
852-5
853-4
854-2
855-1
855-9
848-1
848-9
849-8
850-6
851-5
852-3
6G6
THE ASSAY OF SILVER.
COMMON
Weight of
Assay in
Milligrs.
0.
i.
2.
3.
4.
1190
840-3
841-2
842-0
842-9
843-7
1195
836-8
837-7
838-5
839-3
840-2
1200
833-3
834-2
835-0
835-8
836-7
1205
829-9
830-7
831-5
832-4
833-2
1210
826-4
827-3
828-1
828-9
829-7
1215
823-0
823-9
824-7
825-5
826-3
1220
819-7
820-5
821-3
822-1
822-9
1225
816-3
817-1
818-0
818-8
819-6
1230
813-0
813-8
814-6
815-4
816-3
1235
809-7
810-5
811-3
812-1
813-0
1240
806-5
807-3
808-1
808-9
809-7
1245
803-2
804-0
804-8
805-6
806-4
1250
800-0
800-8
801-6
802-4
803-2
1255
796-8
797-6
798-4
799-2
800-0
1260
793-6
794-4
795-2
796-0
796-8
1265
790-5
791-3
792-1
792-9
793-7
1270
787-4
788-2
789-0
789-8
790-5
1275
784-3
785-1
785-9
786-7
787-4
1280
781-2
782-0
782-8
783-6
784-4
1285
778-2
779-0
779-8
780-5
781-3
1290
. 775-2
776-0
776-7
777-5
778-3
1295
772-2
773-0
773-7
774-5
775-3
1300
769-2
770-0
770-8
771-5
772-3
1305
766-3
767-0
767-8
768-6
769-3
1310
763-4
764-1
764-9
765-6
766-4
1315
760-5
761-2
762-0
762-7
763-5
1320
757-6
758-3
759-1
759-8
760-6
1325
754-7
755-5
756-2
757-0
757-7
1330
751-9
752-6
753-4
754-1
754-9
1335
749-1
749-8
750-6
751-3
752-1
1340
746-3
747-0
747-8
748-5
749-2
1345
743-5
744-2
745-0
745-7
746-5
1350
740-7
741-5
742-2
743-0
743-7
1355
738-0
738-7
739-5
740-2
741-0
1360
735-3
736-0
736-8
737-5
738-2
1365
732-6
733-3
734-1
734-8
735-5
1370
729-9
730-7
731-4
732-1
732-8
1375
727-3
728-0
728-7
729-4
730-2
1380
724-6
725-4
726-1
726-8
727-5
1385
722-0
722-7
723-5
724-2
724-9
1390
719-4
720-1
720-9
721-6
722-3
1395
716-8
717-6
718-3
719-0
719-7
1400
714-3
715-0
715-7
716-4
717-1
TABLE FOE THE WET ASSAY OF SILVEE.
SALT. continued.
5.
6.
7.
8.
9.
10.
844-5
845-4
846-2
847-1
847-9
848-7
841-0
841-8
842-7
843-5
844-3
845-2
837-5
838-3
839-2
840-0
840-8
841-7
834-0
834-8
835-7
836-5
837-3
838-2
830-6
831-4
832-2
833-1
833-9
834-7
827-2
828-0
828-8
829-6
830-4
831-3
823-8
824-6
825-4
826-2
827-0
827-9
820-4
821-2
822-0
822-9
823-7
824-5
817-1
817-9
818-7
819-5
820-3
821-1
813-8
814-6
815-4
816-2
817-0
817-8
810-5
811-3
812-1
812-9
813-7
814-5
807-2
808-0
808-8
809-6
810-4
811-2
804-0
804-8
805-6
806-4
807-2
808-0
800-8
801-6
802-4
803-2
804-0
804-8
797-6
798-4
799-2
800-0
800-8
801-6
794-5
795-3
796-0
796-8
797-6
798-4
791-3
792-1
792-9
793-7
794-5
795-3
788-2
789-0
789-8
790-6
791-4
792-2
785-2
786-0
786-7
787-5
788-3
789-1
782-1
782-9
783-7
784-4
785-2
786-0
779-1
779-8
780-6
781-4
782-2
782-9
776-1
776-8
777-6-
778-4
779-1
779-9
773-1
773-8
774-6
775-4
776-1
776-9
770-1
770-9
771-6
772-4
773-2
773-9
767-2
767-9
768-7
769-5
770-2
771-0
764-3
765-0
765-8
766-5
767-3
768-1
761-4
762-1
762-9
763-6
764-4
765-2
758-5
759-2
760-0
760-7
761-5
762-3
755-6
756-4
757-1
757-9
758-6
759-4
752-8
753-6
754-3
755-1
755-8
756-6
750-0
750-7
751-5
752-2
753-0
753-7
747-2
748-0
748-7
749-4
750-2
750-9
744-4
745-2
745-9
746-7
747-4
748-1
741-7
742-4
743-2
743-9
744-6
745-4
739-0
739-7
740-4
741-2
741-9
742-6
736-3
737-0
737-7
738-5
739-2
739-9
733-6
734-3
735-0
735-8
736-5
737-2
730-9
731-6
732-4
733-2
733-8
734-5
728-3
729-0
729-7
730-4
731-2
731-9
725-6
726-3
727-1
727-8
728-5
729-2
723-0
723-7
724-5
725-2
725-9
726-6
720-4
721-1
721-9
722-6
723-3
724-0
717-9
718-6
719-3
720-0
720-7
721-4
668
THE ASSAY OF SILVER.
COMMON
Weight of
Assay in
Milligrs.
0.
1.
2.
3.
4.
1405
711*7
712-5
713-2
713-9
714-6
1410
709-2
709-9
710-6
711-3
712-1
1415
706-7
707-4
708-1
708-8
709-5
1420
704-2
704-9
705-6
706-3
707-0
1425
701-8
702-5
703-2
703-9
704-6
1430
699-3
700-0
700-7
701-4
702-1
1435
696-9
697-6
698-3
698-9
699-6
1440
694-4
695-1
695-8
696-5
697-2
1445
692-0
692-7
693-4
694-1
694-8
1450
689-7
690-3
691-0
691-7
692-4
1455
687-3
688-0
688-7
689-3
690-0
1460
684-9
685-6
686-3
687-0
687-7
1465
682-6
683-3
684-0
684-6
685-3
1470
680-3
680-9
681-6
682-3
683-0
1475
678-0
678-6
679-3
680-0
680-7
1480
675-7
676-3
677-0
677-7
678-4
1485
673-4
674-1
674-7
675-4
676-1
1490
671-1
671-8
672-5
673-1
673-8
1495
668-9
669-6
670-2
670-9
671-6
1500
666-7
667-3
668-0
668-7
669-3
1505
664-5
665-1
665-8
666-4
667-1
1510
662-3
662-9
663-6
664-2
664-9
1515
660-1
660-7
661-4
662-0
662-7
1520
657-9
658-5
659-2
659-9
660-5
1525
655-7
656-4
657-0
657-7
658-4
1530
653-6
654-2
654-9
655-6
656-2
1535
651-5
652-1
652-8
653-4
654-1
1540
649-4
650-0
650-6
651-3
651-9
1545
647-2
647-9
648-5
649-2
649-8
1550
645-2
645-8
646-4
647-1
647-7
1555
643-1
643-7
644-4
645-0
645-7
1560
641-0
641-7
642-3
642-9
643-6
1565
639-0
639-6
640-3
640-9
641-5
1570
636-9
637-6
638-2
638-8
639-5
1575
634-9
635-6
636-2
636-8
637-5
1580
632-9
633-5
634-2
634-8
635-4
1585
630-9
631-5
632-2
632-8
633-4
1590
628-9
629-6
630-2
630-8
631-4
1595
627-0
627-6
628-2
628-8
629-5
1600
625-0
625-6
626-2
626-9
627-5
1605
623-1
623-7
624-3
624-9
625-5
1610
621-1
621-7
622-4
623-0
623-6
1615
619-2
619-8
620-4
621-0
621-7
TABLE FOR THE WET ASSAY OF SILVER.
SALT continued.
5.
6.
7.
8.
9.
10.
715-3
716-0 716-7
717-4
718-1
718-9
712-8
713-5 714-2
714-9
715-6
716-3
710-2
710-9
711-7
712-4
713-1
713-8
707-7
708-4
709-2
709-9
710-6
711-3
705-3
706-0
706-7
707-4
708-1
708-8
702-8
703-5
704-2
704-9
705-6
706-3
700-3
701-0
701-7
702-4
703-1
703-8
697-9
698-6
699-3
700-0
700-7
701-4
695-5
696-2
696-9
697-6
698-3
699-0
693-1
693-8
694-5
695-2
695-9
696-6
690-7
691-4
692-1
692-8
693-5
694-2
688-4
689-0
689-7
690-4
691-1
691-8
686-0
686-7
687-4
688-0
688-7
689-4
683-7
684-3
685-0
685-7
686-4
687-1
681-4
682-0
682-7
683-4
684-1
684-7
679-1
679-7
680-4
681-1
681-8
682-4
676-8
677-4
678-1
678-8
679-5
680-1
674-5
675-2
675-8
676-5
677-2
677-8
672-2
672-9
673-6
674-2
674-9
675-6
670-0
670-7
671-3
672-0
672-7
673-3
667-8
668-4
669-1
669-8
670-4
671-1
665-6
666-2
666-9
667-5
668-2
668-9
663-4
664-0
664-7
665-3
666-0
666-7
661-2
661-8
662-5
663-2
663-8
664-5
659-0
659-7
660-3
661-0
661-6
662-3
656-9
657-5
658-2
658-8
65?-5
660-1
654-7
655-4
656-0
656-7
657-3
658-0
652-6
653-2
653-9
654-5
655-2
655-8
650-5
651-1
651-8
652-4
653-1
653-7
648-4
649-0
649-7
650-3
651-0
651-6
646-3
646-9
647-6
648-2
648-9
649-5
644-2
644-9
645-5
646-1
646-8
647-4
642-2
642-8
643-4
644-1
644-7
645-4
640-1
640-8
641-4
642-0
642-7
643-3
638-1
638-7
639-4
640-0
640-6
641-3
636-1
636-7
637-3
638-0
638-6
639-2
634-1
634-7
635-3
636-0
636-6
637-2
632-1
632-7
633-3
634-0
634-6
635-2
630-1
630-7
631-3
632-0
632-6
633-2
628-1
628-7
629-4
630-0
630-6
631-2
626-2
626-8
627-4
628-0
628-7
629-3
624-2
624-8
625-5
626-1
626-7
627-3
622-3
622-9
623-5
624-1
624-8
625-4
670
THE ASSAY OF SILVER.
COMMON
Weight of
Assay in
0.
i.
2.
3.
4.
Milligrs.
1620
617-3
617-9
618-5
619-1
619-7
1625
615-4
616-0
616-6
617-2
617-8
1630
613-5
614-1
614-7
615-3
615-9
1635
611-6
612-2
612-8
613-5
614-1
1640
609-8
610-4
611-0
611-6
612-2
1645
607-9
608-5
609-1
609-7
610-3
1650
606-1
606-7
607-3
607-9
608-5
1655
604-2
604-8
605-4
606-0
606-6
1660
602-4
603-0
603-6
604-2
604-8
1665
600-6
601-2
601-8
602-4
603-0
1670
598-8
699-4
600-0
600-6
601-2
1675
597-0
597-6
598-2
598-8
599-4
1680
595-2
595-8
596-4
597-0
597-6
1685
593-5
594-1
594-7
595-2
595-8
1690
591-7
592-3
o92-9
593-5
594-1
1695
590-0
590-6
591-1
591-7
592-3
1700
588-2
588-8
589-4
590-0
590-6
1705
586-5
587-1
587-7
588-3
588-9
1710
584-8
585-4
586-0
586-5
587-1
1715
583-1
583-7
584-3
584-8
585-4
1720
581-4
582-0
582-6
583-1
583-7
1725
579-7
580-3
580-9
581-4
582-0
1730
578-0
578-6
579-2
579-8
580-3
1735
576-4
576-9
577-5
578-1
578-7
1740
574-7
575-3
575-9
576-4
577-0
1745
573-1
573-6
574-2
574-8
575-4
1750
571-4
572-0
572-6
573-1
573-7
1755
569-8
570-4
570-9
571-5
572-1
1760
568-2
568-7
569-3
569-9
570-4
1765
566-6
567-1
567-7
568-3
568-8
1770
565-0
565-5
566-1
566-7
567-2
1775
563-4
563-9
564-5
565-1
565-6
1780
561-8
562-4
562-9
563-5
564-0
1785
560-2
560-8
561-3
561-9
562-5
1790
558-7
559-2
559-8
560-3
560-9
1795
557-1
557-7
558-2
558-8
559-3
1800
555-6
556-1
556-7
557-2
557-8
1805
554-0
554-6
555-1
555-7
556-2
1810
552-5
553-0
553-6
554-1
554-7
1815
551-0
551-5
552-1
552-6
553-2
1820
549-4
550-0
550-5
551-1
551-6
1825
547-9
548-5
549-0
549-6
550-1
1830
546-4
547-0
547-5
548-1
548-6
TABLE FOR THE WET ASSAY OF SILVER.
071
I
SALT continued.
5.
6.
7.
8.
9.
10.
620-4
621-0
621-6
622-2
622-8
623-5
618-5
619-1
619-7
620-3
620-9
621-5
616-6
617-2
617-8
618-4
619-0
619-6
614-7
615-3
615-9
616-5
617-1
617-7
612-8
613-4
614-0
614-6
615-2
615-8
610-9
611-5
612-2
612-8
613-4
614-0
609-1
609-7
610-3
610-9
611-5
612-1
607-2
607-8
608-5
609-1
609-7
610-3
605-4
606-0
606-6
607-2
607-8
608-4
603-6
604-2
604-8
605-4
606-0
606-6
601-8
602-4
603-0
603-6
604-2
604-8
600-0
600-6
601-2
601-8
602-4
603-0
598-2
598-8
599-4
600-0
600-6
601-2
596-4
597-0
597-6
598-2
598-8
599-4
594-7
595-3
595-9
596-4
597-0
597-6
592-9
593-5
594-1
594-7
595-3
595-9
591-2
591-8
592-3
592-9
593-5
594-1
589-4
590-0
590-6
591-2
591-8
592-4
587-7
588-3
588-9
589-5
590-1
590-6
586-0
586-6
587-2
587-8
588-3
588-9
584-3
584-9
585-5
586-0
586-6
587-2
582-6
583-2
583-8
584-3
584-9
585-5
580-9
581-5
582-1
582-7
583-2
583-8
579-2
579-8
580-4
581-0
581-6
582-1
577-6
578-2
578-7
579-3
579-9
580-5
575-9
576-5
577-1
577-6
578-2
578-8
574-3
574-9
575-4
576-0
576-6
577-1
572-6
573-2
573-8
574-4
574-9
575-5
571-0
571-6
572-2
572-7
573-3
573-9
569-4
570-0
570-5
571-1
571-7
572-2
567-8
568-4
568-9
569-5
570-1
570-6
566-2
566-8
567-3
567-9
568-4
569-0
564-6
565-2
565-7
566-3
566-8
567-4
563-0
563-6
564-1
564-7
565-3
565-8
561-4
562-0
562-6
563-1
563-7
5t>4-^
559-9
560-4
561-0
561-6
562-1
562-7
558-3
558-9
559-4
560-0
560-6
561-1
556-8
557-3
557-9
558-4
559-0
559-6
555-2
555-8
556-3
556-9
557-5
558-0
553-7
554-3
554-8
555-4
555-9
556-5
552-2
552-7
553-3
553-8
554-4
554-9
550-7
551-2
551-8
552-3
552-9
553-4
549-2
549-7
550-3
550-8
551-4
551-9
672
THE ASSAY OF SILVER.
COMMON
Weight of
Assay in
Milligrs.
0.
i.
O "
3.
4.
1835
545-0
545-5
546-0
546-6
547-1
1840
543-5
544-0
544-6
545-1
545-6
1845
542-0
542-5
543-1
543-6
544-2
1850
540-5
541-1
541-6
542-2
542-7
1855
539-1
539-6
540-2
540-7
541-2
1860
537-6
538-2
538-7
539-2
539-8
1865
536-2
536-7
537-3
537-8
538-3
1870
534-8
535-3
535-8
536-4
536-9
1875
533-3
533-9
534-4
534-9
535-5
1880
531-9
532-4
533-0
533-5
534-0
1885
530-5
531-0
531-6
532-1
532-6
1890
529-1
529-6
530-2
530-7
531-2
1895
527-7
528-2
528-8
529-3
529-8
1900
526-3
526-8
527-4
527-9
528-4
1905
524-9
525-4
526-0
526-5
527-0
1910
523-6
524-1
524-6
525-1
525-6
1915
522-2
522-7
523-2
523-8
524-3
1920
520-8
521-3
521-9
522-4
522-9
1925
519-5
520-0
520-5
521-0
521-6
1930
518-1
518-6
519-2
519-7
520-2
1935
516-8
517-3
517-8
518-3
518-9
1940
515-5
516-0
516-5
517-0
517-5
1945
514-1
514-6
515-2
515-7
516-2
1950
512-8
513-3
513-8
514-4
514-9
1955
511-5
512-0
512-5
513-0
513-5
1960
510-2
510-7
511-2
511-7
512-2
1965
508-9
509-4
509-9
510-4
510-9
1970
507-6
508-1
508-6
509-1
509-6
1975
506-3
506-8
507-3
507-8
508-3
1980
505-0
505-6
506-1
506-6
507-1
1985
503-8
504-3
504-8
505-3
505-8
1990
502-5
503-0
503-5
504-0
504-5
1995
501-3
501-8
502-3
502-8
503-3
2000
500-0
500-5
501-0
501-5
502-0
TABLE FOR THE WET ASSAY OF SILVER.
673
SALT continued.
5.
G.
I .
8.
9.
10.
547-7
548-2
548-8
549-3
549-9
550-4
546-2
546-7
547-3
547-8
548-4
548-9
544-7
545-3
545-8
546-3
546-9
547-4
543-2
543-8
544-3
544-9
545-4
545-9
541-8
542-3
542-9
543-4
543-9
544-5
540-3
540-9
541-4
541-9
542-5
543-0
538-9
539-4
539-9
540-5
541-0
541-5
537-4
538-0
538-5
539-0
539-6
540-1
536-0
536-5
537-1
537-6
538-1
538-7
534-6
535-1
535-6
536-2
536-7
537-2
533-2
533-7
534-2
534-7
535-3
535-8
531-7
532-3
532-8
533-3
533-9
534-4
530-3
530-9
531-4
531-9
532-4
533-0
528-9
529-5
530-0
530-5
531-0
531-6
527-6
528-1
528-6
529-1
529-7
530-2
526-2
526-7
527-2
527-7
528-3
528-8
524-8
525-3
525-8
526-4
526-9
527-4
523-4
524-0
524-5
525-0
525-5
526-0
522-1
522-6
523-1
523-6
524-2
524-7
520-7
521-2
521-8
522-3
522-8
523-3
519-4
519-9
520-4
520-9
521-4
522-0
518-0
518-6
519-1
519-6
520-1
520-6
516-7
517-2
517-7
518-2
518-8
519-3
515-4
515-9
516-4
516-9
517-4
517-9
514-1
514-6
515-1
515-6
516-1
516-6
512-8
513-3
513-8
514-3
514-8
515-3
511-4
512-0
512-5
513-0
513-5
514-0
510-1
510-7
511-2
511-7
512-2
512-7
508-9
509-4
509-9
510-4
510-9
511-4
507-6
508-1
508-6
509-1
509-6
510-1
506-3
506-8
507-3
507-8
508-3
508-8
505-0
505-5
506-0
506-5
507-0
507-5
503-8
504-3
504-8
505-3
505-8
506-3
502-5
503-0
503-5
504-0
504-5
505-0
XX
674 THE ASSAY OF SILVER.
APPLICATION.
Assay of Pure, or nearly Pure, Silver, the Temperature of
the Normal Solution of Salt being that at which it was
standardised.
First example. Let the ingot of silver have an ap-
proximative standard of from 995 to 1000 thousandths.
Take one gramme ; dissolve it in ten grammes of nitric
acid, in the bottle, fig. 118. Then pour into the bottle an
exact measure of the normal solution of salt, and brighten
by agitation. The silver not being supposed to be quite
pure, the standard is not further sought for by the decline
solution of salt, but that of silver nitrate is employed.
One thousandth of this latter solution is poured into
the bottle ; it becomes cloudy, and is well agitated. A
second and a third thousandth also give a precipitate, but
not so a fourth. From these data the following is the
method of ascertaining the standard of the alloy :
The last thousandth of the decime solution of silver,
having produced no cloudiness, is not to be counted. The
third was necessary, but only partially so ; consequently
the number of thousandths of silver necessary to decom-
pose the excess of salt is more than 2 and less than 3 ; in
other words, it is equal to the mean, 21 ; but since 2-|
thousandths of silver have been required to complete the
precipitation of salt representing 1000 thousandths of
silver, it is evident that the silver submitted to assay con-
tained 21 thousandths of alloy, and that its standard, to
within nearly half a thousandth, is but 997-J.
If it be considered necessary to arrive nearer the true
standard, the following proofs must be employed : Pour
into the solution 11 thousandths of salt, which will decom-
pose a like number of thousandths of silver.* After due
It has already been stated how a thousandth of the decime solution may
be subdivided by the number of drops furnished by the pipette. If, for in-
stance, it contains 20 drops, 10 will give the half, 5 the quarter, &c. Half a
thousandth may also be obtained by diluting the solution with its volume of
water, and using a whole pipetteful. This latter plan has been found the best
in practice.
ASSAY OF PURE, OR NEARLY PURE, SILVER. 675
.agitation, add half a thousandth of silver nitrate. Sup-
posing a cloudiness is produced, no further addition must
be made ; for it is already known that above the third
thousandth no precipitate is formed in the liquid by silver
nitrate, and consequently only half of the last half-thou-
sandth must be calculated, as only a portion of it was
necessary. From which, the entire number of thousandths
of silver nitrate being 4J, and those of salt 1-J, there
remains 2f for the number of thousandths of nitrate of
silver addded to the normalsolution ; and consequently the
standard of the aUoy is 1000 -2f =997J. If, on the other
hand, the last half-thousandth of the silver nitrate had
produced no cloudiness, it would not have to be reckoned,
and only half of the preceding half-thousandth would have
been taken. Thus from the 4 thousandths of silver
nitrate employed a quarter of a thousandth is deducted ;
and from the difference, of, is yet deducted 1-J of salt, the
final remainder being 2^ thousandths of silver nitrate which
have been added to the normal solution : the standard of
the alloy would be 1000 -2J= 997 j.
Although the above-described operation is very simple,
yet it is desirable, in order to avoid al] confusion, to note
in writing such thousandths of salt or silver nitrate as are
added. The thousandths of salt indicating an increase of
standard should be preceded by the sign + ; and the thou-
sandths of silver nitrate announcing a diminution of stan-
dard, by the sign .
Second Example. Suppose the ingot has a presumed
standard of 895 thousandths, and the temperature of the
normal solution supposed invariable.
Find in the table of standards (Salt Table), first column,
that which approaches the nearest to 895 ; it will be
found to be 896- 9, corresponding to the weight of 1115
milligrammes. This weight of the alloy is taken and
dissolved in nitric acid, a measure of normal solution of
salt added, and the whole well agitated. The operator is,
however, doubtful whether the assay must be proceeded
with by the decime salt solution, or the silver nitrate
decime solution. If the former produces a precipitate^
x x2
676 THE ASSAY OF SILVER.
it is gone on with ; but if it does not precipitate, that
already added is decomposed by a similar addition of the
second, and the solution rendered bright by agitation. A
starting-point has now been arrived at for the continuance
of the assay, for it is known that the silver nitrate solution
must be employed.
Suppose, then, that the alloy, after the addition of the
measure of normal solution, yet gives a precipitate with
the decime solution of salt. The first 5 thousandths
produce a precipitate, but not the sixth, which conse-
quently is not counted. The fifth has only been partially
required, so that it is more than 4 thousandths, and less
than 5, or the mean, 4^, is the quantity required to
entirely precipitate the excess of silver in the alloy sub-
mitted to assay. But by neglecting at first the fraction
0*5, seek in the Salt Table of Standards the number found
on the longitudinal line of the weight 1115, under column
4 ; it is 900-4, and on adding 0-5 to this number we have
900-9, or 901, for the required standard.
Supposing, however, that the same alloy, after the
addition of the normal measure of salt, gives a precipitate
with silver nitrate, and that the 3 first thousandths pro-
duce a cloudiness, but not the fourth. The number of
thousandths of silver nitrate really necessary for complete
precipitation will be very nearly 2-J. To ascertain the
real standard of the alloy of which 1115 thousandths were
supposed to contain about 1000 thousandths of silver, take
the number found in the horizontal line 1115, and in the
column 2 of the Silver Nitrate Table. This number, which
is 895-1, diminished by the fraction 0-5, gives 894*6 for
the standard of the alloy to within half a thousandth.
Third Example. The actual temperature of the normal
solution of salt being 18, when it was standardised at 15.
The ingot of silver submitted to assay has an approxi-
mative standard of 795 thousandths. Find in the Salt
Table of Standards, first column, that which is nearest
to it ; it is 793*7, corresponding to the weight 1260.
This weight of the alloy is taken, and the operation pro-
ceeded with as already described. Supposing it had
GRADUATION OF THE NORMAL SOLUTION OF SALT. 077
required 6'5 thousandths of salt to precipitate the whole
of the silver contained in the alloy to within half a
thousandth, the required standard, without correction
for temperature, will be 798-4 + 0'4 = 798'8. But, making
this correction, recourse must be had to the table, page
651, column 15 : the number 0'3, which will be found in
the horizontal line 18 and the column 15, possesses the
sign ; consequently it must be deducted from 798*8, and
the remainder, 798'5, will be the standard weight. If the
temperature of the solution, instead of being 3 higher than
at the time it was standardised, was 3 lower, that is 12,
the correction must be added, and would be equal to +0-2.
The standard of the alloy would conse- FrG 12 4.
quently be 798-8 + 0-2 = 799.
Graduation of the Normal Solution of Salt,
the Temperature being different to that at
which it was wished to be graduated.
Two equally ready processes can be
employed. The one consists in reducing
the temperature of the solution to the
desired degree before standardising ; the
other, in estimating its standard without
regard to the temperature of the solution,
and then correcting its influence by the
aid of the tables of correction already
given.
First Process. Place the liquid to be
graduated in a bottle, F, fig. 124. Intro-
duce a thermometer, and heat to a deter-
minate degree, say 20 for instance. This
done, place the jet of the pipette in the
bottle ; raise the liquid by aspiration by
means of the conical tube, T, fig. 124,
which is adapted to the opening of the
air-cock, R. As soon as the liquid is
raised a little, above the mark a 6, which determines the
capacity of the pipette, close the stopcock, and complete
678 THE ASSAY OF SILVEK.
the measurement as usual. This same means of filling
the pipette by aspiration may be employed to fill it either
with caustic alkali or nitric acid, as the case may be, to
cleanse it instead of taking it to pieces.
Second Method. The solution of salt being supposed
at a temperature of 16, and it is desired to graduate it
at that of 20. Proceed with the standardising without
regard to temperature ; but when it is obtained in each
trial assay, it is necessary to make the correction required
by the temperature.
If, for example, in an approximative assay the standard
of the solution was expressed by 1001*5 this standard
would not only be too weak by 1*5 thousandth, but,
according to the table of temperatures, by yet another
0-5, for the solution is weakened by this quantity, by
passing from 16 to 20. The standard, if taken at this
last temperature, would be too low by 2 thousandths, and
must consequently be corrected.
If, on the other hand, the standard of the solution were
too high instead of too low, and expressed by 998*5 at
the temperature of 16 ; at that of 20, the solution being
weakened by 0'5, the standard would only be but one
thousandth too high, and it must be corrected by that
quantity.
Approximative Estimation of the Standard of an
Unknown Alloy.
It has always been supposed, in the experiments already
detailed, that the approximative standard of the alloy sub-
mitted to assay was known : and this, indeed, is nearly
always the case. If, however, this be unknown, two
means are available for obtaining the necessary know-
ledge. A decigramme of the alloy is cupelled with one
gramme of lead ; or if it be desirable not to use the cupel,
it may be ascertained by the wet method, in the follow-
ing manner :
The assayer supposes the standard of the alloy known
to be about a twentieth, and it can always be found nearer
than that by touch, density, &c. A weight relative to its
MODES OF ABRIDGING MANIPULATION. 679
supposed standard is taken, and its standard sought by
adding the decime liquid by 10 thousandths at a time,
by means of pipettes of this capacity (see fig. Fm 125
125). The term of complete precipitation is
soon passed, and the standard of the alloy to
about 5 thousandths is thus ascertained. The
approximate standard to 2-J- thousandths may
be obtained by adding only 5 thousandths oi
solution at a time.
Suppose the alloy 840 thousandths. . Take
the weight 1190, corresponding to this standard,
and proceed "as in an ordinary assay, adding
each time, for example, a pipette of 10 thou-
sandths of salt solution. It is found the fifth
pipette gives no precipitate, and consequently
the number of thousandths of salt for the pre-
cipitate of the silver to within 5 thousandths is
35. The 1199 of alloy will therefore contain
1000 + 35 = 1035 of silver ; and the approxi-
mative standard will be obtained by the pro-
portion
1120: 1035:: 1000:0=869-7.
Modes of Abridging Manipulation.
In the statement already given of the mode of con-
ducting the assay by the wet method, only such instructions
have been given as were necessary for its full comprehen-
sion, and everything that might call away or FIG. 126.
fatigue the attention has been omitted. Never-
theless, here it will be convenient to describe
some methods of abridging the necessary mani-
pulations, supposing that ten, or at least five,
assays are made at once.
Bottles. It is necessary to have these all,
as nearly as possible, of the same height and
diameter. They are marked progressively on the shoulder,
as are also their stoppers (fig. 126), thus 1, 2, 3, 4, &c.
They are taken successively by tens, in the natural order.
680
THE ASSAY OF SILVER.
The stoppers are placed on a support, numbered in the same
manner (fig. 127). The support is pierced with ten holes,
127. distinguished in precedence by
a mark between the fifth and
sixth.
Stand. Each ten flasks
are in turn placed in a case or
stand of japanned tin-plate (fig. 128), having ten compart-
ments numbered from 1 to 10. Each of these compart-
ments is cut anteriorly to about half its length, so as to
FIG. 128.
FIG. 129.
FIG. 130.
allow the numbers, of the bottles to be seen. The same
stand serves for all the series, by making the same units
correspond : thus, No. 23 of the third series is placed in
stand No. 3, &c. When each
flask is charged with the
alloy, about 10 grammes of
nitric acid, 40 C., are mea-
sured by a pipette (fig. 125)
introduced into the bottles by
means of a funnel with a large
neck (fig. 129). The whole
are then exposed to the heat
of a water-bath, to facilitate
the solution of the alloy.
Water-bath. This is an
oblong tin-plate vessel, cal-
culated to receive 10 bottles (fig. 130). It has a move-
able double bottom, pierced with small holes,, the prin-
MODES OF ABE1DGING MANIPULATION. 681
cipal object of which is to prevent the fracture of the
bottles by isolating them from the bottom of the vessel,
which is immediately exposed to the heat. On the
moveable bottom are soldered the cylinders c c, three or
four centimetres in height, and above which, at the distance
of eight centimetres, is a sheet of tin plate, p p, pierced with
ten holes, corresponding to the cylinders, and connected
with the movable bottom by the supports, s s. These
cylinders, and the sheet of tin plate, are destined to isolate
the bottles, F F, one from the other in the bath, and to
keep them some time suspended over it, when the water is
boiling, before complete immersion. The water-bath may
be replaced by a steam-bath ; the bottles will then be sup-
ported by a grating above the surface of the water. The
solution of the alloy in the nitric acid takes place rapidly,
and as it gives rise to an abundant evolution of nitrous
vapour, it must be made under a flue having a good
draught.
Flue. This is represented at fig. 131. C C is a flue
resting on a table or support, T T, about 90 centimetres
high. The anterior side in the FlG 1B1
figure is removed to show^ the
water-bath 5, and the furnace
F. The opening, 0, of the flue
is closed by the wooden door, p,
movable on two eccentric pivots,
which keep it up during the
solution, and allow it to fall so
that the flask may be placed
upon it. The nitrous vapour is
removed from the bottles with the blower (fig. 123). The
hood, ff, prevents the diffusion of the nitrous vapour in
the laboratory..
Agitator. Figure 132 gives a sufficiently exact idea of
this apparatus, and dispenses with a long description. It
has ten cylindrical compartments, numbering from 1 to 10.
The bottles, after solution of the alloy, are placed in it in
the order of their numbers. The agitator is then placed
by the side of the pipette, by which is measured the normal
682
THE ASSAY OF SILVER.
solution ol salt, and into each flask is poured a pipetteful
of the solution. The bottles are fitted with their stoppers,
previously moistened with distilled water (fig. 133); they
are then fixed in order with wooden wedges (fig. 134).
The agitator is suspended to a spring, 7?, and a rapid
alternating movement given to it with both hands, by which
the solution is agitated, and in less than a minute rendered
FIG. 132.
FIG. 133.
FIG. 134.
as clear as water. The movement is assisted by a spiral
spring, B, fixed to the agitator and its stand. The agitation
finished, the wedges are removed, and placed in the vacant
spaces between the compartments. The agitator is taken
from the spring, and the bottles^placed in order on a table
prepared to receive them.
Table. This table (fig. 135) has a double bottom ; the
upper is pierced with ten holes, a little larger than the
MODES OF ABKIDGING MANIPULATION.
083-
diameter of the bottles, and of such a distance from the
lower portion, or false bottom, that the flasks do not rise
above its edge, or at least but little. This disposition is to
FIG. 135.
FIG. 136.
protect the silver chloride from the light, for it decomposes
in contact with water, and a little hydrochloric acid is
produced, which requires for its precipitation a certain
quantity of silver nitrate, and so lowers the standard of
the alloy. This cause of error is, however, not very great,
at least when the light does not fall directly on the
chloride ; but it is easy to avoid, and should not be neg-
lected. The disposition already pointed out does not at all
complicate the process, and is moreover useful, as it pre-
vents the fracture or upsetting of the bottles. When but
one bottle is operated on, it is placed for agitation in a
japanned tin-plate cylinder, which is held as shown at fig.
136. On placing the bottles in their respective places on
the table, a brisk circular movement is given to them, so as
to remove any silver chloride
adhering to the sides ; their
stoppers are removed and sus-
pended by spring pincers, a a.
These are formed of sheet-iron
wire (see fig. 137). A thou-
sandth of the decime solution is then poured into each
bottle, and before this has been completed there will have
FIG. 137.
<>84 THE ASSAY OF SILVER.
formed in the first bottles where there is any precipitate, a
well-marked nebular layer about a centimetre in thickness.
At the back of the table is a black board, PP, divided
into compartments numbered from 1 to 10, on each of
which is marked with chalk the number of thousandths of
decime liquid added to the contents of the corresponding
bottle. The thousandths of salt announcing augmentation
of standard are preceded by the sign + , those of silver
nitrate by the sign .
Lastly, the black board carries a small shelf pierced
with holes, 1 1, and these receive the funnels or drain the
bottles ; on this shelf also are fastened the pincers for sus-
taining the stoppers.
Cleaning the Bottles. The assays terminated, the liquid
from each flask is poured into a large vessel in which there
is always a slight excess of common salt ; and when it is
full, the clear supernatant fluid is removed by means of a
syphon. (Immediately will be given the means of reducing
the silver chloride so collected to the metallic state.) The
bottles, to the number of ten, are first rinsed with the same
water passed from one to the other, then a second, and then
a third time with fresh water. They are then placed to
FIG 138 drain on the board just mentioned,
and the stoppers are placed in a
stand by series of tens (see figs. 138
and 127). It is important to remark,
that when a glass has been rinsed
with distilled water, care must be
taken not to rub it with the fingers, for water poured in
such a vessel would always be clouded on the addition of
silver nitrate. This effect is due to the chlorides contained
in the perspiration, and is of course more to be guarded
against in summer.
Reduction of Silver Chloride , obtained in the Assay of
Alloys by the Wet Method.
Silver chloride can be reduced without sensible loss,
after haying been well washed, by plunging it into scraps
PREPARATION OF PUKE SILVER. 685
of iron or zinc, and adding dilute sulphuric acid in suffi-
cient quantity to set up a slight disengagement .of hydro-
gen gas. The whole can be left to itself, and in the course
of a few days the silver is completely reduced. This point
can be easily estimated by the colour and nature of the
product, but better still by treating a small quantity with
ammonia, which, if the chloride is perfectly reduced, will
give no precipitate or cloudiness on treatment with an acid.
The chlorine remains in solution in the water combined
with zinc or iron. The residue must now be washed ; the
first washings are made with acidulated water, to dissolve
iron oxide which might have formed, and the following
with ordinary water : after having completed the washing
as much water as may be left is decanted, the mass dried,
and a little powdered borax added. Nothing now remains
but to fuse it. The powdered silver being voluminous, it
is placed by separate portions in the crucible, in propor-
tion as it sinks. The heat should at first be moderate, but
towards the end of the operation should be sufficiently
high to reduce the silver and slag to a state of complete
liquidity. If it be found that not quite all the chloride
was reduced by the iron or zinc, a little potassium or
sodium carbonate may be added to the powdered silver.
The standard of silver thus obtained is from 999 to 1000
thousandths.
Preparation of Pure Silver.
Take the silver prepared as above, dissolve it in nitric
acid, and leave the solution some time in perfect rest in
water, to deposit any gold it might contain. Decant the
solution, and precipitate with common salt, well wash the
precipitate, and reduce it, when the resulting silver will be
pure.
M. Gay-Lussac here gives a description of a process for
the precipitation of chlorine from nitric acid for use in the
mode of assay already described ; but as that acid in a state
of ordinary purity forms an article of commerce, and can be
obtained at most operative chemists, the process will riot be
here reproduced.
C86 THE ASSAY OF SILVER.
Modifications required in the Assay of Silver Alloys
containing Mercury.
Whenever mercury is present in solution with silver, it
is thrown down as insoluble chloride, and the assay ren-
dered inaccurate. The presence of mercury in silver can
be readily detected by the remarkable change which occurs
in silver chloride on exposure to light (viz. blackening)
when free from mercury ; but if the smallest quantity of
the latter metal be present, no blackening will ensue. This
source of error was removed by M. Levol in the following
manner : The sample being dissolved, as usual, in nitric
acid, it was supersaturated with 25 cubic centimetres of
caustic ammonia ; the pipetteful of normal solution was
then added, and the excess of ammonia supersaturated
with 20 cubic centimetres of acetic acid, and the operation
continued in the usual way.
Some little time after the publication of this, M. Gray-
Lussac examined the above process himself, and very con-
siderably simplified it. He says : ' After having confirmed
by several experiments the accuracy of M. Level's pro-
cess, I thought it might be simplified by adding to the
nitric solution of silver the ammonia and acetic acid at one
and the same time, but in sufficient quantity to saturate
the whole of the nitric acid, both that in combination with
the silver and that in the free state. Ten grammes of ammo-
nium acetate, were added, with a little water, to the silver
dissolved in nitric acid, and the assay finished in the
ordinary manner. The quantity indicated by synthesis
was found very accurately, although 100 thousandths of
mercury had been added/ Finally, M. Gay-Lussac found
that 10 grammes of sodium acetate, in crystals, also fully
answered the purpose ; and as that is a very cheap com-
mercial salt, it is the best adapted for overcoming the diffi-
culty in this class of assay, as regards the presence of
mercury.
METHOD OF ASSAYING SILVER BARS.
087
FIG. 139
Method of Taking the Assay from the Ingot.
The ingots are so rarely perfectly homogeneous, even
taking as a starting-point the standard 950 thousandths,
that the differences remarked between the assays of samples
made in different places should
rather be attributed to the
above cause than to the assay
itself. It is important, there-
fore, to take a sample in a
uniform manner, and from
the same depth, on the upper
surface of the ingot as on the
lower. This condition is per-
fectly fulfilled by boring the
ingots with a kind of drill,
similar to that employed by
the smith, and which is re-
presented at fig. 139. The
ingot, Z, is placed in a copper
tray, C\ and in order to
retain the borings, which
might otherwise be thrown
out, the drill,/, is surrounded
by a casing, m, which does
not impede its motion, and
stands freely on the ingot.
After a few turns of the drill,
the first borings, which are not pure, are removed by
means of a feather, and only those following are collected
and reserved for assay. If it be desirable to try the lateral
faces, it is necessary to employ a pressure screw, to keep
the ingot in the position that may be deemed necessary.
Method of Assaying Silver Bars adopted in the Assay
Offices of II.M. Indian Mints.
According to the report of H. E. Busteed, M.D., H.M.
Madras Army, refining assay master, Calcutta Mint, the
038 THE ASSAY OF SILVER.
silver bars are assayed by estimating the silver as chlo-
ride. The cupellation assay is not correct enough to
satisfy the sellers and purchasers : the Gay-Lussac assay
is not used, because most of the bullion contains mercury,
lead, and other base metals ; because a previous cupel-
lation assay is required ; because the high temperature of
the climate Causes an evaporation of the salt solution ; and
because a large number of persons would be necessary on
account of the large daily number of assays.
The credit is due to Mr. J. Dodd, a former assay
master of the Calcutta Mint (and a surgeon in the
Madras army), of having overcome those difficulties of
manipulation, inasmuch as he modified and simplified
them, and in short so systematised the whole practical
working of the process as to render its application to
the assaying of silver, to any amount, easy, accurate,
and economical.
The process is technically known as the ' chloride
process.' It is thus described by Dr. Busteed in the
' Chemical News ' for November 1871 :
The samples (or ' musters ') for assay are, to save time,
first approximately weighed by an assistant ; they are then
placed (each sample in duplicate) in small shallow saucers
of polished copper, and so brought in batches of 40 on a
board, containing in numerical order receptacles for the
little saucer, to the assay master, who, in the delicate assay
balance, exactly brings each sample to the one required
weight.*
As each sample is weighed, it is transferred from the
platinum skiff of the balance to a bottle on the left hand
of the assayer, by means of a small copper funnel. The
bottles f for this purpose are held in readiness for the
musters by an assistant, and, on receiving them, are
removed into the laboratory in batches of six.
On being taken to the laboratory, they are ranged on
a circular platform or turn-table, and there one of the
* The amount of this weight will be more particularly referred to farther on.
t The chief appliances will be described more fully at the conclusion of
the description of the process.
THE CHLOKIDE PROCESS. 689
(European) assistants adds by means of a pipette 1^ drm.
of nitric acid to each bottle, which are then (without their
stoppers) transferred to a sand-bath and exposed to a
considerable degree of heat, till solution of the contents is
effected.
The specific gravity of the nitric acid used is generally
1-200, i.e. in the case of known alloys of only copper and
silver, such as the standard meltings, coins, &c. ; but
when the nature of the alloy is uncertain, such as bazaar
silver, or some sycee (where the presence of mercury may
be suspected), a stronger acid of sp. gr. 1*320 is used. It
has been found, too, by experience, that the chlorides
from fine bar silver agglomerate better when the solution
has been effected in the stronger acid.
When the samples have been completely dissolved,*
the bottles are brought back to the platform, and there
each receives through a glass funnel f about six ounces of
cold distilled water.
There is then added to each bottle, through a glass
pipette as before, 1^ drm. of hydrochloric acid, sp. gr.
TOGO, which immediately converts the silver present into
the characteristic white precipitate of silver chloride, which
forms in slow-falling curdy volumes.
The stoppers (previously dipped in distilled water) are
then carefully replaced, and the bottles are allowed to
stand for five minutes.
The bottles are next well shaken two and two by the
laboratory workman for three or four minutes, till the
chloride aggregates and rapidly falls down ; any particles
which may remain attached to the neck or upper part of
the bottles are washed down by a quick circular motion,
and, more distilled water being added to within about
two inches of the neck (great caution being observed in
removing and returning the stoppers), the bottles are then
allowed to rest each in its assigned place on the platform
for four hours.
* A slight residuum of gold, as a black powder, is very generally seen.
t The portion of this which enters the neck of the bottle is protected or
sheathed, with an inch of india-rubber tubing, to prevent chipping, if struck
against the neck of the bottle.
Y Y
690 THE ASSAY OF SILVER.
At the expiration of that period, the clear supernatant
liquid (blue-coloured when copper is present) is removed
by a glass syphon, which is lowered to within an inch of
the deposited chloride, the greatest care being taken that
none of it is drawn up into the syphon. As each platform
is made to revolve on its centre, according as each bottle
is syphoned, the operator sitting in one place brings the
platform round till the next bottle in order gets under the
syphon, which is thus in turn lowered into each. The
fluid escapes from the long leg of the syphon through a
funnel fitted in the table to ajar placed underneath.
After the first syphoning, the bottles are immediately
filled again with distilled water, and each gets a quiet cir-
cular motion for a few moments, and the precipitate is again
allowed to settle as evenly as possible ; this time it will be
sufficient to allow them to rest for two hours, when they
are again syphoned as before and the stoppers returned.
Under ordinary circumstances these two washings are
sufficient ; but if the silver is evidently ' coarse,' a third or
fourth washing is similarly given.
When it is considered that the chlorides have been
sufficiently washed, the bottles are placed for half an hour
in a reclining position on their platforms ; this causes the
chloride to fall and settle to one spot, and renders its
removal from the bottles more easy.
Meantime a pneumatic trough has been got ready,
capable of containing a batch of twenty inverted bottles ;
the trough is filled with distilled water ; for each bottle
there is placed on the floor of the trough a small porcelain
saucer holding a little Wedgwood crucible or cup, each
numbered to correspond to the bottles. A laboratory
workman then removes the stoppers from the bottles, and
hands them one by one to an assistant at the trough, who,
placing his forefinger over the mouth of each bottle,
inverts it over its corresponding cup, and does not remove
his finger till the neck of the bottle has passed down
through the water and well into the cup ; then the finger
being taken away, the bulk of the chloride falls by its
own weight to the bottom of the cup.
THE CHLORIDE PROCESS. 691
The bottle is held in the position by two rings, one
(the larger) above the other, which are fixed to the sides
of the trough ; this arrangement retains each bottle in
situ, at the proper slant, and admits of the operator gently
revolving or slightly raising the bottle with his left hand,
while with the right he patiently taps the bottom and
sides till the whole of the chloride has been dexterously
got out ; the finger is then again placed over the mouth,
and the bottle raised up through the rings and handed
(mouth upwards) to the assayer, or to the supervising
assistant standing by, who carefully examines it to see
that every particle of chloride has been dropped into the
cup. When this part of the manipulation has been neatly
done, none of the chloride falls over into the saucer which
is placed as a precautionary measure under each cup.
When the chloride falls into the cup, it is in an uneven
lumpy state, and not in a favourable condition for being
uniformly dried ; it has, therefore, next to be broken up.
For this purpose the cups (containing the chlorides, and
water to the brim) on removal from the trough are taken
in batches on a tray to an assistant seated at a steady
table, who first carefully decants off about half the water,
and then with a finely polished glass rod (four inches long
and one-third of an inch thick) gently stirs and beats the
lumpy precipitate, while revolving the cup on the table ;
this causes it to lie evenly and loosely at the bottom of
the cup as a purplish grey powder, not too fine.
He next washes the rod over the cup with distilled
water from a drop bottle, lest any of the chloride may be
adhering to it, and sprinkles a drop or two from it on
to the surface of the water in each cup, so as to cause
to sink any minute particles that may happen to remain
floating. He then, after an interval of ten minutes, drains
off about three-fourths of the supernatant water, which he
lets run down the rod into a vessel near him, arid with a
tap or two of the rod on the outside of the cup to still
further loosen the deposit, this part of the manipulation is
concluded.
The crucibles are next taken to the drying chamber,
TT 2
092 THE ASSAY OF SILVEE.
where a steam bath is ready to receive them ; on the per-
forated upper plate of this they are ranged, and allowed
to remain for about an hour. This gradually, and without
spurting, frees the chlorides from moisture, which may be
known by their caking, i.e. leaving the sides of the cups
round the edges and forming at the bottom of each a
loose cake, resembling somewhat a gun-wad. The crucibles
are then arranged on a hot-air plate and there exposed to
a temperature of between 300 and 350 (F.) for about
two hours, till thoroughly dried, when they are ready for
weighing.* When the above manipulations have been
carefully and satisfactorily gone through, each little cup
contains an unbroken, tolerably firm cake of chloride of
silver, lying unattached, which admits of being easily
grasped with a pair of forceps, and cleanly lifted out of
the cup and conveyed to the skiff of the assay balance in
which it is weighed. The cups are generally brought
from the laboratory to the assayer at the balance in
batches of eight or ten. A 6 standard,' synthetically pre-
pared of pure silver and copper, and an assay pound of
pure silver, are introduced with each day's set of assays,
and their chlorides dried with the others, and the analysis
of them verified before weighing the rest. Occasionally
these ' checks ' are also fused and weighed in a porcelain
capsule, but the weight found never differs from that of
the chloride merely dried as above.
Once or twice a month, the silver is recovered from
the accumulated chlorides, which are well pounded in a
mortar and brought to a powder and then mixed with a
proper proportion of chalk and charcoal, and put into
a wrought-iron crucible and reduced with heat. The
metallic silver so recovered is transferred to the mint.
Under the circumstances of the solution and of the
precipitation as detailed above, should any gold happen
to be present in the sample operated on, it is not dissolved,
and therefore becomes entangled with the precipitated
* The chlorides are weighed warm, to obviate the risk of their absorbing
moisture ; a precaution especially necessary in the heavy monsoon weather
in this country
THE CHLORIDE PROCESS. 693
silver chloride and dried and weighed with it, and ac-
cordingly comes to be regarded and valued as silver. In
this the chloride process resembles that by cupellation,
which likewise takes no distinguishing cognisance of gold;
and both these processes contrast in this respect with the
volumetric one, which is a rigid analysis for silver alone ;
so that, strictly speaking, an assay conducted by either of
the first-named methods ascertains the proportion present
of ' the precious metals,' i.e. silver and gold.*
Should mercury be present it does not interfere with
the result, when the solution has been effected in excess
of nitric acid with strong heat. Thus the mercury be-
comes per oxidised, and hydrochloric acid forms no pre-
cipitate in solutions of mercuric salts ; any mercuric
chloride resulting from the combination would remain
in solution, and be washed away in the course of the
process.
Should lead happen to be present, hydrochloric acid
gives no precipitate in a dilute solution, the lead chloride
being soluble in a certain proportion of distilled water ;
but even were the proportion of lead to silver tolerably
large, and the lead chloride happened to be thrown down,
the repeated washings would dissolve and get rid of it.
With regard to the weight of the small portion taken
to represent the mass, the system prevails in the Indian
mints of taking samples for assay by granulating a small
portion of the contents of each melting-pot : when the
metal is in a thorough state of fusion and has just been
well stirred, a small ladleful of the molten metal is quickly
poured from a tolerable height into a vessel of water, and
.the granules so formed received on a strainer, lifted out,
and perfectly dried. f The weight of this specimen repre-
* Much of the silver which finds its way to the Indian mints is rich in
gold; for instance, sycee contains on an average somewhat about 12 grs. in
the troy pound. This in minting operations is considered as silver, and as
such it enters into the coinage. There being as yet no refineries established
here, through which such silver could pass to the mechanical departments of
the mints, the silver coins made during a period when a heavy importation of
sycee had been worked up, contain as much as 4 or 6 grains of gold in every
32 tolas or 1 pound troy.
t The introduction into the Calcutta Mint of this system of taking
musters is attributable to Dr. Boycott, late Assay Master, and to Dr. Shekelton,
694 THE ASSAY OF SILVER.
senting each pot was first fixed at 24 grs., technically
called the ' assay pound : ' this in the case of pure silver
yielded 31-87 grs. of silver chloride, while the same quan-
tity of Indian standard silver (which is -j^-ths silver plus
T Vth copper= 916-66 in 1000) yielded ^th less, or 29-21
grs. : on the weight of chloride being ascertained in
each case, a table which was calculated and prepared for
the purpose was referred to, and the equivalent fineness
assigned to the -Jdwt., plus the odd grs., when any. But
when it became desirable to prepare for the decimal form
of notation, a number more convenient than 31-97 was
looked for to represent purity or 1000, and 25 was fixed
on as a desirable starting-point, particularly as the quan-
tity of pure silver yielding that amount of chloride, viz.
18*825 grs., was quite large enough to represent each pot.*
The weight, therefore, of the ' assay pound ' in use at
present is 18*825 grs. This produces (with chlorine) in
the case of pure silver 25 grs. of silver chloride.f
But to obviate the necessity of constant reference to a
calculated table to find the equivalent in pure silver of the
amount of silver chloride found in each case, it was inge-
niously arranged to stamp each of the assay weights not
with its actual weight, but with the figures representing
the proportion per mille of pure metal which such a
weight of chloride so found corresponds to : thus, sup-
who, by a number of interesting experiments, satisfied themselves that
samples so taken represent the mass of mixed metal to be valued much more
fairly than samples of the same mass cut or gouged from it after it has been
poured and allowed to cool in the ingot moulds, where a partial separation of
the copper from the silver seems to take place ; the result being, according to
the above experiments, that in the case of ingots cast in upright moulds all
the outside is much below the average fineness of the mass on assay, and the
centre much above it. This refers to alloys of silver and copper mixed in or
about the proportion of ' standard.' According to M. Levol, however, it would
appear that when an alloy of silver and copper in which the proportion of the
latter is very high (viz. over 28 per cent.) has been melted, poured, and
allowed to cool, an opposite result to the above is found, viz. the outside
of the ingots is above the average fineness. An assay, therefore, from a
granulated sample must give a much nearer approximation to truth than one
from a cut sample.
* The average weight of the contents of each melting-pot is 12.500 tolas,
or about 390 pounds troy, so that the specimen taken to represent this is but
about the 119,000th part ; each sample is assayed in duplicate.
t The basis for these numbers was founded on the proportion in which,
according to Turner, silver combines with chlorine, viz. 100 parts with 32'80.
THE CHLORIDE PEOCESS. 695
posing a melting of five-franc pieces was being assayed,
and the chloride resulting from the assay pound operated
on weighed 22-5 grs. (showing the actual pure contents in
the sample to be 16-94 grs.), instead of referring to a table
to see the equivalent per mille-age of silver, that weight
which is actually 22*5 grs. has 900 marked on it, and the
assayer simply reads the c touch ' from it.
Accordingly, the assay weights are as follows :
Actual weight in grs. Figures marked on the weights
25-000 ....... 1000-00
22-910 (Std.) 916-66
22-500 900-00
20-000 . . ... . . 800-00
17-500 . . ... . . . 700-00
15-000 600-00
12-500 . 500-00
10-000 400-00
7-500 . 300-00
5-000 200-00
2-500 100-00
1-250 . ' . .... 50-00
1-000 40-00
0-750 30-00
0-500 20-00
0-250 10-00
0-125 5-00
0-100 4-00
0-075 . 3-00
0-050 2-00
0-025 1-00
Assay lb., weight=18*825 grs.
An ordinary day's work consists of eighty assays,*
estimating imported bullion to the value of four lacs of
rupees, and standard meltings and coins to the value of
five lacs. But on emergencies, in time of heavy pressure,
by working extra hours, as many as 164 assays have been
daily conducted, estimating to the value of eight lacs of
rupees, and standard coins and meltings to the value of
fourteen lacs.
Such is an outline of the method of assay worked on a
large scale ; of course successful results from it cannot be
expected unless each step in the manipulation be con-
ducted with great care and accuracy, and only then after
much practice and experience.
* Exclusive of any gold assays which may be going on.
698 THE ASSAY OF SILVER.
The natives of this country possess great aptitude in
acquiring the skill and consequent lightness of touch so
essential for delicate manipulation ; this, added to their
characteristic patience, makes them admirable subordi-
nates in an assay laboratory, under judicious supervision ;
moreover, their labour is cheap ; so that, on the whole,
the process seems to be especially suitable for an Indian
mint.
When bar silver is imported from the Continent, the
assays of it, made" here, almost invariably correspond most
closely with those previously made of it in Paris by the
volumetric method. But were further proof needed of
the practical accuracy of this system, it is to be found in
the very close proximity to the legal standard at which
the large Indian coinage has been maintained for many
years, as annually reported by the assayers to the Eoyal
Mint of Great Britain, who test the fineness of the Indian
pyx coins by the French process.
Without this me tBod (improved and made more per-
fect, as it has been, in the hands of successive assay
officers), it would have been very difficult for the assay
establishments of the Indian mints to have dealt with, in
the same time and with the same accuracy, the immense
importation of silver to India during "the last fifteen years.
In the single year 1865-66 there was poured into the
Indian mints, and manufactured into coin, silver alone
reaching in value to the prodigious amount of over four-
teen millions sterling.
The system which enabled the assay officers to value
such a rapid and heavy influx with accuracy, and with
satisfaction to the importer on the one hand, and to the
mint (the buyer) on the other, and to faithfully maintain
the immense resulting coinage close to legal standard, has
been put to a severe test. If success be the criterion of
merit, the twenty years' large experience of this method
gained in the Indian mints goes to show that it is worthy
of a yet wider field of utility.
THE CHLORIDE PROCESS. 697
Apparatus and Appliances required.
(1.) The bottles used in this process are of thin (but
strong) white glass, and contain about 12 fluid oz. ; about
6 inches in height and 2^ inches in diameter at the bottom,
which should present a perfectly even, level floor ; they
are without any abrupt shoulder, but become gradually
pyramidal from about half-way up to the neck ; this shape
favours the easy dropping out of the chloride. The neck
is about one inch in length, polished on its inner surface ;
the stoppers are of ground glass, polished, with globular
heads, and are made to fit with the utmost accuracy and
smoothness. The bottles and stoppers are numbered, to
correspond with the number on the muster board and also
on the cups.
(2.) The ' cups' are Wedgwood crucibles, smooth and
thin, about 1^ inches in height, 1^ inches in diameter
above, and a little less than 1 inch in outside diameter at
the bottom. The floor should be perfectly level, and
neither it nor the sides should present any roughness likely
to retain the chloride. The cups are all numbered.
(3.) The porcelain saucers are shallow, f of an inch in
depth ; the upper diameter is about 4 inches, the lower 2-J
inches.
(4.) The turn-table is a circular board of about 3 feet
in diameter, fenced by a brass railing (or by a simple ledge) ;
its centre is occupied by a raised platform about 2 feet in
diameter, between which and the rail the bottles (20 on
each) stand, the round outer edge of the platform having
semi-lunar niches cut in it, into which the bottles fit ;
opposite to each niche on the platform is a little concavity
in which the stoppers rest when not in the bottles. Each
turn-table is made to revolve on its centre in either direc-
tion, and is raised about 6 inches above the long general
table on which all are supported ; close to each a funnel
is fitted into the lower (supporting) table for conducting
away the fluid syphoned from each set of bottles.
(5.) The trough is a basin of cast iron (painted) ; it
698 THE ASSAY OF SILVER.
may be oblong or round, raised to about the height of
3 feet from the ground ; when round and large enough
for twenty bottles, space and distilled water may be
economised by having a platform in 'the centre. This is
convenient for resting the bottles oh after the chlorides
have been got out. A trough of this kind may
be about 2^ feet in diameter, having a space 7 inches
broad and 4 deep all round between the circumference of
the basin itself and the outer edge of the island platform.
Into this space is poured distilled water to the depth of
3 inches. From the rim of the trough hang as many
brass supports as there are bottles to be inverted ; there
are two circular clasps connected at the back to a bar
common to both ; one, the larger, is 1-| inch above the
smaller and lower one, which is under water ; they are
open in front (or towards the centre of the basin) to about
J of an inch in width. The openings of both are in the
same line, owing to the lower (smaller) segment being
projected towards the centre by an abrupt curve in the
connecting bar, by which they hang from the brim. This
arrangement receives and fixes the inverted bottles in the
required position. The distilled water is removed from
the trough by the withdrawal of a plug. These troughs
are sometimes made to revolve on the centre.
(6.) The drop bottle used for washing down the glass
rod when breaking up the chlorides, and for sprinkling the
surface of the cups, is small-sized, round, so as to be easily
grasped ; it holds about 6 oz. The stopper is hollow,
with two small tubes leading from its head, one opposite to
the other. Glass is so liable to break or chip, that a hollow
silver stopper is now generally substituted.
(7.) The steam-bath is simply a square vessel made
of sheet copper, between 3 and 4 inches deep, the top or
upper plate of which has a number of circular openings
about two-thirds of the diameter of a Wedgwood crucible.
There is also a steam escape pipe leading from the centre
below to about a foot in height. They are of various sizes,
to contain from 10 to 150 pots : they are raised or moved
by two lateral handles.
EFFECT OF BISMUTH ON SILVER. G99
(8.) Hot-air plate of thin sheet iron bored with holes
for the reception of the crucibles, raised by iron feet about
1-| inch above the furnace plate. It is furnished with a
square tin cover, which fits over it. This is provided with
lateral apertures for the escape of heated air, and with a
tube from its roof for the reception of a thermometer.
The drying furnace on which the above rest is sur-
mounted by a hood, the glazed door of which slides up
and down by weights and pulleys ; the plate is heated by
means of gas jets ; it has a good draught to carry off the
nitrous fumes, as on it the musters are dissolved in the first
instance on a sand bath.
(9.) The forceps for removing the cake of chloride
from each cup to the skiff of the balance should not be
too sharp in its grasp ; it is much improved by having
the blades tipped for about an inch from the points with
platinum about ^ inch in width.
(10.) It is a convenience to have the assay weights
arranged in a set of ivory compartments in the weight
box ; on the floor of each compartment are engraved the
figures corresponding to those engraved on the weight
which occupies it ; by this means the assayer has merely
to glance at his weight box to see what weights are in the
pan of the balance, and to read off the ' touch ' when each
chloride is counterpoised.
Effect of Bismuth on the Ductility of Silver.
The effects produced by small quantities of bismuth on
the ductility of silver have been carefully investigated by
Surgeon-Major J. Scully, Assay Master, Calcutta.
The following are extracts from an elaborate paper
which he published in the Chemical News for 1887 :
It is well known that alloys of silver and bismuth, in
certain proportions, are brittle. In Dr. Percy's valuable
work on Metallurgy (Silver and Gold Part I.) it is stated
that alloys of silver w^ith bismuth, in the proportion of
50 per cent, and 33 per cent, of the latter metal, are
brittle ; while an ore of silver and bismuth, called
700 THE ASSAY OF SILVER.
Chilenite, in which bismuth occurs only to the extent of
14*4 to 15*3 per cent., is said to be malleable. The least
amount of bismuth, however, which will injuriously affect
the ductility of silver, for example, in such an operation
as the lamination of silver bars for coinage, does not, so
far as I am aware, appear to have been experimentally
investigated. It may here be mentioned at the outset
that an alloy of silver and bismuth may, by careful
hammering, be extended considerably, so as to pass muster
as malleable ; although, if subjected to lamination by
means of steel rolls, the same alloy will crack at the
edges and thus show a deficient ductility, as compared
with pure silver or some silver-copper alloys. It is to the
deficiency in ductility, as tested by rolling, of silver con-
taining only very small proportions of bismuth that I here
wish to call attention.
My attention was first prominently directed, about two
years ago, to the injurious effects caused by small quan-
tities of bismuth in silver by the circumstance that some
silver bullion, in the shape of English refined bars of as
high a fineness as 990 per mille, proved so brittle as to be
unfit for mintage. Attention was first attracted to this
matter by the peculiar behaviour under assay of the
granulated samples taken from this silver after melting.
The appearances noticed under assay will be referred to
presently, but they led to the bullion being at once tested
for brittleness. A bar, about 21 inches long, 2-25 broad,
and 1 inch thick, was hammered out at one end without
cracking, but on being passed through the rolls it cracked
badly at the edges and was pronounced to be ' brittle.'
in the Mint sense of the term. The bullion was then re-
melted in five plumbago pots, and a partial refinement of
it attempted in the ordinary way with nitre, about 8 to
10 Ibs. of this salt being used for each pot. The resulting
silver bars were not appreciably improved by this treat-
ment ; hammering again proved an inconclusive test, but
a bar of the size I have mentioned broke in two by merely
dropping on the floor of the melting-room.
In the meantime the assay had shown that the brittle
EFFECT OF BISMUTH OS SILVER. 701
bullion contained bismuth, and that this was the only
substance present likely to be the cause of brittleness.
The Indian process of assaying silver has already been
described (p. 688) by Dr. Busteed. The main features of
the process may here be briefly recapitulated for the
purpose I have in view.
A fixed weight of the silver bullion to be assayed is dis-
solved in an assay bottle, by means of nitric acid aided by
heat ; the solution is diluted with water and an excess of
hydrochloric acid is added, to precipitate all the silver pre-
sent as chloride. The silver chloride having been caused
to aggregate and settle by vigorous shaking, the bottle is
filled up with water, and the supernatant fluid is subse-
quently syphoned off, to remove all the now dissolved
matter which may have been contained in the bullion.
Under these conditions of solution, precipitation, and
dilution with water, chemists will readily understand that
even a small trace of bismuth, if it be in the silver, will
reveal its presence by the partial formation of insoluble
oxy chloride of bismuth. Now, in the assay of the brittle
bullion under consideration, solution in nitric acid had
been readily and completely effected by the aid of heat ;
antimony and tin were consequently absent. After the
addition of water and hydrochloric acid, however, the
solution in the assay bottles could not be cleared by
shaking ; the bulk of the silver chloride collected at the
bottom of the bottles, but the supernatant fluid remained
turbid.
Tin and antimony being excluded, only two metals
could produce this result in the wet assay of silver
namely, mercury and bismuth. To determine which of
these is the interfering metal it is only necessary to note
the effect of solar light on the silver chloride formed ;
when mercury is present the silver chloride maintains its
pure white colour unaltered, while in the presence of
bismuth the chloride immediately acquires the well-known
purple colour under the influence of daylight. Our
assays, then, being turbid after precipitation and yet the
silver chloride blackening readily under the influence of
702 THE ASSAY OF SILVER.
daylight, it was evident that bismuth was present. The
turbidity produced was due to the partial formation of
bismuth oxychloride ; and this compound diffusing itself
in its characteristic manner through the solution had
broken up part of the silver salt into very fine powder, so
that some hours had to elapse before the supernatant fluid
cleared by the gradual subsidence of both bismuth oxy-
chloride and the finely divided silver chloride. The assay
was of course thus rendered unreliable, since the silver
chloride to be weighed, and on which the calculation of
the fineness rested, was contaminated with bismuth oxy-
chloride. A cupellation assay of this bullion was at once
had recourse to for ascertaining its fineness.
So far, then, this tender of silver bullion seemed to
establish the following points:
1. Silver bullion of as high fineness as 990 per mille
is rendered unfit for coinage purposes by an amount of
bismuth which, in this particular case, could not possibly
have exceeded 1 per cent., and was probably less than that
proportion.
2. Hammering a bar of silver bullion is not a good
test for detecting brittleness, as far as Mint purposes are
concerned.
3. The toughening of silver bullion 990 fine, and con-
taining only a small amount of bismuth, by the aid of
nitre in plumbago melting-pots, is not readily affected.
4. The presence of a trace of bismuth in silver of
high fineness is immediately detected in the ordinary course
of assay by the Indian method, but this bismuth interferes
with the perfect accuracy of the results obtained by that
process.
A comprehensive research seemed, therefore, called for
to elucidate the whole subject, and the necessity for this
investigation has since been emphasised by the fact that
silver bullion contaminated with bismuth has frequently
found its way to the Mint since its first discovery here.
The points to be investigated seemed naturally to group
themselves under the following heads :
1. Is our ordinary wet assay of silver susceptible of
EFFECT OF BISMUTH ON STLVEE. 703
such easy modification as will enable us to obtain perfectly
accurate results by it, in presence of bismuth, without
having recourse to the confessedly less accurate assay by
cupellation ? And, how may small quantities of bismuth
in silver be readily estimated with the despatch indispens-
able for mint operations ?
H. What is the smallest amount of bismuth in silver
that will render it unfit for coinage, when present in bars
of the Indian standard fineness of 916*6 ? And, what is
the amount of bismuth that may be tolerated in such bars
without materially injuring the ductility ?
III. How is silver bullion containing bismuth, which
may be tendered to the Mint, to be dealt with, supposing
that establishment accepts any metal that is brittle ; and
how is the presence of bismuth in refined bars to be
accounted for ?
1. As the purity of the bismuth to be used in the
experiments now to be detailed was a matter of first im-
portance, I may briefly mention the steps taken to insure
the purity of the metal. Eefined bismuth was dissolved
in nitric acid, precipitated as basic nitrate by diluting
largely with distilled water, the nitrate digested in solu-
tion of caustic potash, and then well washed, dried, and
reduced by heating with charcoal in a clay crucible. A
series of synthetical assays, made by dissolving together
pure silver and pure bismuth, the latter in the proportion
of from 1 to 5 thousandths, showed that our ordinary
process of assay, under such conditions, gave unreliable
results, there being a surcharge, or higher report than
should have been obtained, which varied from 0'7 to 2- 7
mils., when the proportion of bismuth was from 3 to 5
thousandths. A modification in our process of assay was
evidently required if it were to be used for estimating the
fineness of silver bullion containing bismuth ; and the neces-
sary steps to this end were, after repeated experiment,
found to consist in adding the smallest possible amount of
hydrochloric acid for the precipitation of the silver, and
increasing the amount of nitric acid in which it was first
dissolved. We use ordinarily for the precipitation of an
704 THE ASSAY OF SILVER.
assay pound of silver 5-4 c.c. of hydrochloric acid, of
sp. gr. 1075 ; but 2*5 c.c. of acid of this strength suffices
for the complete precipitation of an assay pound of even
fine silver ; so that we have here at once a means of
diminishing the tendency of any bismuth in the silver to
form insoluble oxy chloride. If, in addition to diminishing
the amount of hydrochloric acid, we added a considerable
excess of nitric acid to the solution (which acid would
not in any way interfere with the silver chloride formed),
all risk of the partial formation of insoluble bismuth salts
seemed removed. This in fact has proved to be the case,
and the successful modified process for the assay of silver
containing bismuth is as follows :
The assay pound of silver bullion containing 'bismuth
is dissolved in 5-5 c.c. of nitric acid, sp. gr. 1200, with
the aid of heat, about 5 ounces of water are added,
and then 10 c.c. of nitric acid, sp. gr. 1320. The silver
is now precipitated by the addition of 2-5 c.c. of hydro-
chloric acid, and after vigorous shaking the supernatant
fluid will be found perfectly clear, and it will remain so
when the bottle is filled up with water, all the bismuth
present being in solution. Whenever samples of silver
now show the presence of bismuth during the assay, a
fresh set is taken up and worked by the modified process,
the delay thus caused not amounting to more than a
few minutes. It may be mentioned here that all our
assays are reported to one-tenth of a millieme (0*1 per
mille).
Having thus ascertained the presence of bismuth in
silver bullion, and put in practice a modification of the
assay process which renders us indifferent to its presence,
it is still of importance to ascertain the exact proportion
of bismuth which is present in the bullion, and, to be of
practical use for mint work, this estimation must be
effected rapidly and as simply as possible. The ordinary
directions given for the separation of bismuth in the pre-
sence of silver, by first removing the latter as chloride
and then precipitating the bismuth as carbonate, do not,
I find, give accurate results when silver is present in such
EFFECT OF BISMUTH ON SILVEK. 705
overwhelming proportions as obtain in the cases under
consideration.
I have therefore adopted the following plan, which a
number of synthetically prepared solutions have proved
to give quick and good results, though sometimes the
amount of bismuth present is very slightly underestimated.
The ordinary silver assay having given a rough visual
estimate of the amount of bismuth likely to be present,
enough of the bullion is taken to yield a fairly weighable
amount of bismuth oxide in the final result. The bullion
is dissolved in a small amount of nitric acid, the solution
carefully diluted, and an excess of ammonium carbonate
at once added, the precipitation being aided by heating.
The carbonates of silver and copper at first formed are
re-dissolved, and the carbonate of bismuth after a time
settles completely at the bottom of the beaker. The con-
tents of the beaker are then passed through a filter, of
which the weight of ash yielded by incineration is known,
and the carbonate of bismuth on the filter washed quite
free of all traces of silver. The filter is then dried, its
contents transferred to a porcelain crucible for ignition,
the filter-paper being ignited separately, treated with a
drop or two of nitric acid to re-oxidise any bismuth oxide
reduced by contact with the carbon of the filter, and the
ash added to the crucible. From the weight of bismuth
oxide thus found, after deducting the weight of the filter
ash, the amount of metallic bismuth present in the sample
of bullion taken for analysis can be at once found.
There are only two metals likely to interfere with the
accuracy of the process here described namely, cadmium
and lead ; the carbonates of both these metals being as
insoluble in excess of the precipitant employed as bismuth
carbonate. Cadmium is very unlikely to be found in
silver bullion and its consideration may be neglected, but
if the presence of lead is suspected the carbonate filtered
from the silver solution is dissolved in nitric acid, evapo-
rated down with the addition of sulphuric acid, and the
lead sulphate formed (if any) collected and weighed in
the usual way. The bismuth is again precipitated as
z z
706 THE ASSAY OF SILVER.
carbonate and treated as before directed. Many experi-
ments have been made with synthetically prepared mix-
tures of silver, copper, lead, and bismuth, the latter two
metals being in very small proportion to the silver, so as
to imitate the composition of some refined bars. Ullgreen's
plan for the separation of the carbonates of lead and
bismuth, by dissolving them in acetic acid and then pre-
cipitating the bismuth by means of a lead rod, does not-
work satisfactorily and requires too long a time for the
precipitation.
II. As it seemed likely that a large number of ex-
periments would be required to estimate accurately the
smallest amount of bismuth that would injure the ductility
of our coinage alloy, and the still smaller proportion that
would not sensibly affect this ductility, it was determined
to begin the inquiry by a number of laboratory experi-
ments on small bars of silver, before trying the effects of
bismuth on ordinary coinage bars and with the procedure
for lamination carried out in the Mint. These laboratory
experiments were made in the following way : Pure
silver prepared for assay check purposes, or an alloy of
silver and copper of which the exact composition had
been estimated by assay, was melted in a clean plum-
bago crucible under charcoal. When the metal was in
fusion the necessary amount of bismuth was rolled in a
piece of paper, carried down at once to the bottom of the
silver bath, and then thoroughly mixed with the silver by
stirring. The calculated composition was confirmed by
assay of the silver. When this mixture had been accom-
plished, the contents of the crucible were poured into an
open iron ingot mould, and after cooling either quickly
by plunging the casting into water or slowly in contact
with the mould the bar so cast was tested for brittleness
by hammering and rolling. The bars cast were of two
sizes, one set being 3'75 inches long, 1-125 broad, 0-375
thick, and weighing about 6*2 troy ounces ; and another
set 2-69 inches long, 1-125 broad, 0-25 thick, and weigh-
ing about 4-1 troy ounces. When reduced to the fullest
extent by rolling, these bars were converted into straps
EFFECT OF BISMUTH ON SILVER. 707
about 0/015 inch in thickness. In laminating them they
were twice annealed, first after having undergone four
pinches in the rollers, and again after the tenth pinch
from the beginning. Similarly shaped bars of silver,
without bismuth, were occasionally laminated in the same
way to obtain a sure means of comparison. Before any
result was accepted as to brittleness or its absence, the
bar under experiment was always re-melted and tried at
least a second time. The number of experiments in this
series amounted to fifty -three, and the following is a
summary of the results obtained :
Fine silver when alloyed with only 1 per mille (one-
thousandth part of its weight) of bismuth, and the casting
rapidly cooled by plunging it into water as soon as it has
set, has its ductility, as tested by lamination, sensibly but
slightly impaired, the straps resulting from rolling having
slightly jagged edges. When the proportion of bismuth
is increased to 2, 3, 4, and 5 per mille, the plan of cooling
remaining the same, the raggedness of the edges of the
straps was somewhat increased, but not very markedly.
If, however, the casting was allowed to cool down com-
pletely, but very slowly, in contact either with the mould
or a stone floor, the results were very different. Under
this condition of cooling, a bar composed of fine silver
with 4 per mille of bismuth was completely brittle ; it was
readily broken, and its fracture was strongly crystalline.
On laminating it small cracks appeared all over the sur-
face on the second pinch, the bar emitting a crackling
sound under the rolls, much like the ' cry ' of tin, and on
the fourth pinch the bar cracked deeply at the edges.
This remarkable effect on the molecular structure of this
alloy of silver and bismuth, as due solely to the mode of
cooling the casting, was repeatedly verified on the same
metal by re -melting and cooling rapidly and slowly alter-
nately. The case seems analogous to that of bronze,
where slow cooling of the alloy after casting is said to
make it hard and brittle.
Fine silver with 6 per mille of bismuth, rapidly cooled,
was distinctly cold-short and crystalline on fracture ; the
z 2
708 THE ASSAY OF SILVEK.
bar cracked on the surface at the fourth pinch. With
7 per mille of bismuth these evidences of diminished
ductility were slightly more pronounced. With 8 per
mille of bismuth the silver was still more brittle, the bar
broke readily when hammered, and cracked all over the
surface on the fourth pinch from the rolls. With 9, 10 r
and 11 per mille of bismuth, the bar of silver could be
readily broken in two by merely striking it against the
edge of an anvil, the fracture was coarsely crystalline, and
the bar, in one case, proved to be very red-short, a mere
tap from the tongs sufficing to break it in two when
heated for the purpose of annealing. Although these bars
were so very brittle, it was still possible to roll them into
thin straps after careful annealing ; but the edges of the
straps so produced were deeply jagged and indented by
cracks. These bars also all emitted the peculiar crack-
ling noise under the rolls which has before been men-
tioned.
An alloy containing 990 parts of silver and 10 of
copper then had added to it successively 1, 2, 3, 4, and 5
per mille of bismuth, the castings being rapidly cooled.
The remarks already made with reference to fine silver
alloyed with the same proportions of bismuth would
apply here almost exactly that is to say, the bars were
rolled out to a thickness of 0-015 with somewhat ragged
edges, so that although ductility, as thus tested, was
impaired, it was only slightly so. With 6 per mille of
bismuth (fineness of metal on assay 983-9) the edges
cracked a little, and, after annealing and rolling out, the
strap had decidedly jagged edges and was split for some
distance at one end. The bars containing 4, 5, and 6 per
mille of bismuth were now re-melted and allowed to cool
slowly and completely in the mould. They were all found
to be highly brittle, broke easily under the hammer
the fracture being granular and not crystalline and on
being rolled they cracked badly, all over the surface and
at the edges, on the first or second pinch ; in one case the
bar broke in two on the second pinch. That these very
different results were again solely due to the manner of
EFFECT OF BISMUTH ON SILVER. 709
cooling was proved by re-melting and rapidly cooling the
castings, when the same metal proved comparatively
ductile, as first stated.
Silver of the Indian standard of 916*6 per mille (the
rest being copper) to which 2 per mille of bismuth was
added, gave on lamination straps with slightly jagged
edges and proved to be red-short. With 4 per mille of
bismuth the bars showed a few surface cracks on being
rolled, and the resulting straps had decidedly jagged
edges. Slow cooling of these castings did not affect their
ductility, thus showing a marked contrast to what had
been observed in the case of fine silver and the alloy con-
taining only 10 per mille of copper. When the amount
of bismuth was increased to 5 per mille, the copper
present remaining at 834 per mille, the bars were de-
cidedly brittle and cracked readily on hammering the
fracture being again granular, and not crystalline as in
the case of fine silver. On lamination both surface and
edge cracks developed after four pinches from the rolls,
and in annealing one of these bars the whole surface
blistered considerably, no doubt owing to the temperature
having been carried a little too high. Standard silver
with 10 per mille of bismuth, reducing the fineness as
ascertained by assay to 906*6, was very brittle, the bars
breaking easily under the hammer, and on the fourth
pinch from the rolls splitting and cracking all over the
surface. In the course of these latter experiments it was
ascertained that, with from 83-5 to 70 per mille of copper
present, slow or rapid cooling of silver alloys containing
bismuth made no appreciable difference in their ductility.
The foregoing experiments having furnished some in-
formation as to the amount of bismuth that might be
expected to injure our coinage alloy, it was now decided
to test that point practically, by operating on coinage bars
subjected to the regular procedure for the manufacture of
rupees in the Calcutta Mint. The experiments made in
this connection were fourteen in number. The bars used
here for coinage weigh about 253 troy ounces, and are
about 20 inches long, 2-25 broad, and 1 inch thick ; they
710 THE ASSAY OF SILVER.
are cast in vertical iron moulds. In lamination they are
first reduced by eleven pinches to a thickness of 0'23 in. ;
they are then annealed, and finally reduced by twelve
additional pinches to a thickness of 0*06 inch. A number
of bars, poured from a pot of which the contents had
proved on assay of a granulated sample to be 916-6 fine,
were selected for the experiments, and as a preliminary
step one of the bars was laminated to test its ductility,
It rolled out with smooth ' wire ' edges, and indeed its
ductility was beyond suspicion, as it resulted from a
melting of good coins. Another bar of the same batch
was now melted and 1 per mille of bismuth added to it,
the result of the addition being checked, in this and all
following cases, by the assay of a granulated sample of
the metal, taken after thorough stirring. At the eighth
pinch both edges of the lower half of this bar began to
crack, and at the eleventh pinch these cracks extended
towards the middle line of the strap for about a quarter
of an inch, and occurred at about every half inch of the
edge. After annealing, and in the subsequent lamination
to a thickness of 0-065 inch, these cracks increased con-
siderably in number, but did not become sensibly deeper.
The strap as finished was pronounced unfit for coinage
purposes ; for although two blanks could have been cut
from its width, the edges were too jagged to admit of the
blanks being obtained exactly along the line from which
it was desired to cut them this position being attained
by means of a fixed lateral guide against which the edge
of the strap had to be maintained in cutting. With 2 per
mille of bismuth the results obtained on rolling were not
much worse than with 1 per mille. But the side cracks
opened out more, and here again it was noticed that the
lower portion of the bar (upper and lower here having
reference to the casting in upright moulds) was somewhat
less ductile than the upper part.
With 3 per mille of bismuth (fineness on assay 913'8)
the bar began to crack on both edges at the ninth pinch ;
at the eleventh pinch there were many cracks quite a
quarter of an inch deep, and after annealing the bar these
TITRATION OF SILVER IN PRESENCE OF COPPER, ETC. 711
cracks increased at every pinch, so that at the twenty-first
pinch the strap was cracked all along both edges very
badly. It would only have been possible to obtain one
blank from the width of this strap.
As it was perfectly clear that no further experiments
were required with larger proportions of bismuth, the
subsequent trials were made on coinage bars containing
0-5 per mille, 0-25 per mille, and, by dilution of the latter
bars with standard silver, to even half and a quarter of
the lesser proportion just stated. Here the results were
rather discordant ; they appear to have been somewhat
influenced by the state of different rolls, and by quick or
slow annealing. The general outcome of the tests, how-
ever, was that although some of the straps containing the
proportions given of bismuth were jagged at the edges,
and so would have yielded a diminished percentage in
outturn of good blanks, others were not materially worse
than the average of straps without any bismuth at all.
As a result of this part of the inquiry, it may, I think, be
fairly concluded that if our coinage bars contain less than
0-5 per mille of bismuth their ductility will not be mate-
rially affected. It must be borne in mind that these results
only apply to bars of the size and shape of those experi-
mented on, and with the particular treatment in lamina-
tion above detailed. With thinner bars and a different
method of rolling, different results may be expected. The
system of cutting out blanks has also to be considered,
for in some mints straps with saw edges are not so pre-
judicial as in others.
Titration of Silver in Presence of other Metals by means of
Ammonium Sulpho cyanide.
Mr. J. Volhard gives the following process for the
volumetric assay of silver.
In a nitric solution of silver to which a soluble ferric,
salt has been added, a permanent redness does not appear
on the gradual addition of a dilute solution of ammonium
or potassium sulphocyanide, until all the silver has been,
712 THE ASSAY OF SILVER.
thrown down as a sulphocyanide. If we know how much
of the sulphocyanide solution is required to precipitate a
known weight of silver we can estimate volumetrically
the quantity of silver present in any acid argentic solution,
the ferric salt serving as an indicator. For the prepara-
tion of the standard solution the author uses ammonium
sulphocyanide, though Lindermann prefers the potassium
salt ; both in dilute solution are permanent. But the am-
monium salt is less frequently contaminated with chlorides,
which interfere greatly if present in more than mere traces.
The solution may be conveniently made of such a strength
that 1 c.c. indicates 1 centigramme of silver. The weighed
quantity of the salt is dissolved in water, and diluted in a
graduated flask so far that 7*5 grins, (or 8 grins, if the
salt appears very damp) may be contained in each litre.
For the precipitation of 1 grm. of silver 0-704 grm. of
pure ammonium sulphocyanide is requisite ; 0'5 grm. of
chemically pure silver is then weighed out, dissolved in 8
or 10 c.c. of pure nitric acid of sp. gr. 1*20, and after the
complete solution of the metal the nitrous acid is expelled
by boiling or prolonged heating on the sand-bath, and the
solution is allowed to cool. It is then diluted with 200
c.c. of water, and 5 c.c. of a cold saturated solution of
ammonia-iron alum are added. If the colour of the ferric
salt is perceptible, a little pure colourless nitric acid is
added till it disappears. The sulphocyanide solution is
then added from a burette. At first a white precipitate is
produced, which remains suspended in the liquid like silver
chloride, rendering it milky. On the further addition of
sulphocyanide each drop produces a blood-red cloud, which
quickly disappears on agitation. As the point of satura-
tion is reached, the silver sulphocyanide collects in flocks,
and the liquid grows clearer, without becoming perfectly
limpid, as long as a trace of silver remains in solution. As
soon as all the silver is precipitated the flocculent precipi-
tate quickly deposits, and the supernatant liquid becomes
quite clear. The sulphocyanide solution is added by drops
till this point is attained, and till a very faint light-brown
colour appears in the liquid, which does not vanish on
TITRATION OF SILVER IN PBBSBNCB OF COPPER, ETC 713
repeated agitation. The colour is most easily perceived
if the liquid is held, not up to the light, but against a white
wall turned away from the window.
For a repetition of the experiment it is convenient to
use a silver solution of a known strength. For this pur-
pose 10 grms. of pure silver are dissolved in nitric acid
in a litre flask, the nitrous acid is expelled, the liquid is
allowed to cool, and diluted up to the volume. For use
50 c.c. are taken with a pipette. If the above-mentioned
proportions are preserved for 0'5 grin, silver, or 50 c.c. of
the silver solution, somewhat less than 50 c.c. of the sul-
phocyanide solution will be required, whence the needful
dilution can be calculated. If 48*5 c.c. sulphocyanide
have been used, then to every 48-5 c.c. of the solution 15
c.c. of water must be added.
The solution is perfectly permanent, even on being
kept for two years. The silver solution must have a
decidedly acid reaction from free nitric acid, which it is
unnecessary to neutralise, although it is important that
the proportion of acid in different experiments should be
approximately equal. It is, however, essential for the con-
stancy of the results that the ferric salt should be in large
excess, and should be used approximately in one and the
same proportion to the total volume of the fluid. It must
also be remembered that the colour of ferric sulphocyanide
is destroyed by nitrous acid at common temperatures, and
by nitric acid on the application of heat. Hence follows
the necessity of completely expelling all nitrous acid, and
of allowing the liquid to become cold before the operation
is begun. The nitric acid used in this process should be
kept in the dark.
Copper. The proportion of copper in an alloy may
reach 70 per cent, without in the least affecting the accu-
racy of the process. Beyond this limit the recognition of
the final reaction is somewhat doubtful, since after the
precipitation of the silver the liquid is rendered opaque
and discoloured by the formation of black copper sulpho-
cyanide. The only remedy in such cases is to add a
known weight of pure silver to the alloy, so as to reduce
711 THE ASSAY OF SILVER.
the proportion of copper below 70 per cent., as is done in
Gay-Lussac's process.
Mercury. In presence of this metal silver cannot be
titrated with sulphocyanide solution. This defect is of
little consequence, as mercury is readily expelled.
Palladium renders the titration of silver with sulpho-
cyanide inaccurate, this metal appearing in the result as
silver. This is an important circumstance from a technical
point of view, since palladium, though frequently present
in silver, occurs in very small quantities.
As regards other metals soluble in nitric acid and
found along with silver in alloys and ores, lead, cadmium,
thallium, bismuth, zinc, iron, and manganese are without
influence. The recognition of the final reaction in solu-
tions strongly coloured by salts of cobalt or nickel,
requires some practice. At first a few drops of sulpho-
cyanide will always be added in excess. It is then recom-
mended to titrate back with a solution of silver, when the
pure colour of the cobalt or nickel will appear so suddenly
and distinctly, that it will conversely be easy to seize the
exact point, when the colour of the solution is modified
to a yellowish brown by the addition of the colour of fer-
ric sulphocyanide. When the change of shade has been
observed four or five times by titrating backwards and
forwards with exactly corresponding solutions of silver
and of sulphocyanide, such a certainty in the recognition
of the final reaction will be attained that there can
be no doubt as to the addition of a half drop more or
less.
The presence of tin, antimony, and arsenic does not
interfere with the accuracy of the process.
Mr. C. A. M. Balling gives the following instructions
for the direct estimation of silver in galena by Vol-
hard's process : From 2 to 5 grins, of the galena, accord-
ing to its supposed richness in silver, are very finely ground
and intimately mixed in a porcelain mortar with from
three to four times its weight of a flux composed of equal
parts of soda and saltpetre, placed in a porcelain crucible,
covered, and heated over a burner to thorough fusion,
BLOWPIPE ASSAY OF SILVER. 715
when the mixture is well stirred with a glass rod. It is
then let cool and placed in an evaporating dish partly
filled with water, in which the melted matter is softened,
dissolved out of the crucible into the dish, which is then
heated, and the watery solution is filtered into a flask.
The residue on the filter, after being well washed, is rinsed
back into the dish, very dilute nitric acid is added, and
the whole evaporated to dryness. The dry residue is
taken up in water acidulated with nitric acid, heated, and
filtered into the same flask in which is the aqueous solu-
tion. The residue is washed with hot water, the filtrate
is allowed to cool in the flask, ferric sulphate or iron
alum is added, and the liquid is titrated.
BLOWPIPE ASSAY OF SILVER.
The following very complete method for the blowpipe
assay of silver and its ores is given by David Forbes,
F.K.S., in the 'Chemical News,' JSTos. 380, 384, 392, 396,
398, and 412.
The blowpipe assay of silver ores was first described
in 1827 by Harkort,* and subsequently considerably im-
proved by Plattner. This assay process is in all cases
based upon the reduction to a metallic state of all the
silver contained in the compound in question along with
more or less metallic lead, which latter metal, when not
already present in sufficient quantity in the substance
itself under examination, is added in the state of granu-
lated lead to the assay previous to its reduction. The
globule of silver-lead thus obtained, if soft and free from
such elements as would interfere with its treatment upon
the cupel, may then be at once cupelled before the blow-
pipe until the pure silver alone remains upon the bone-ash
surface of the cupel ; but if not, it is previously submitted
to a scorifying or oxidising treatment upon charcoal until
all such substances are either slagged off or volatilised,
and the resulting silver-lead globule cupelled as before.
* 'Die Probirkunst mit dem Lothrohre.' Freiberg, 1827, I. Heft (all
published).
716
THE ASSAY OF SILVER.
As, therefore, the final operation in all silver assays is
invariably that of cupelling the silver-lead alloy obtained
from the previous reduction of the substance, effected by
methods differing according to the nature of the argen-
tiferous ore or compound under examination, it is here con-
sidered advisable to introduce the description of the silver
assay by an explanation of this process.
In the ordinary process of cupellation in the muffle,
bone-ash or other cupels are employed of a size large
enough to absorb the whole of the litharge produced from
the oxidation of the lead in the assay.
This, however, should not be the case when using the
blowpipe ; for as .the heating powers of that instrument
are limited, it is found in practice much better to accom-
plish this result by two distinct operations the first being
a concentration of the silver-lead in which the greater
part of the lead is converted by oxidation into litharge re-
maining upon, but not, or only very slightly, absorbed by,
the bone -ash cupel : and the second in cupelling the small
concentrated metallic bead so obtained upon a fresh cupel
until the remaining lead is totally absorbed by the cupel
and the silver left behind in a pure state. By this means a
much larger weight of the silver-lead alloy can be submitted
to assay, and, for reasons hereafter to be explained, much
more exact results are obtained than
would be the case when the cupella-
tion is conducted at one operation in
the ordinary manner.
The apparatus used by the author
for these operations" are shown to a
scale of one-half their real size in the
woodcut fig. 140 (a to d).
In fig. 140, a represents in section a
small cylindrical mould of iron, seven-
tenths of an inch in diameter, and
about four-tenths high, in which is
turned a cup-shaped nearly hemisphe-
rical depression two-tenths of an inch
deep in centre, the inner surface of which is left rough, or
FIG. 140.
FOKBES'S BLOWPIPE ASSAY. 717
marked with minute ridges and furrows for the purpose of
enabling it to retain more firmly the bone-ash lining which
is stamped into it by means of the polished bolt, also shown
in the figure. This mould rests upon the stand d, having
for this purpose a small central socket in its base, into
which the central pivot of the stand enters. This socket
is seen in the ground plan, 5, of the base of the mould,
which shows likewise three small grooves or slots made to
enable a steady hold to be taken of it, when hot, by the
forceps. The stand itself is composed of a small turned
ivory or wood base, fixed into a short piece of strong glass
tubing, which, from its non-conducting powers, serves as an
excellent handle. In the centre of the base a slight iron
rod rising above the level of the glass outer tube serves as
a support for the cupel mould, into the socket in the base
of which it enters.
Bone-ash is best prepared by burning bones which have
previously been boiled several .times, so as to extract all
animal matter. The best bone-ash is made from the core-
bone of the horns of cattle well boiled out and burned.
The ash from this is more uniform than from the .other
bones, which have in general a very compact enamel-like
exterior surface, whilst the interior is of a much softer
nature.
Concentration of the Silver-lead. A cupel is prepared
by filling the above-described cupel mould with bone-ash
powder not finer than will pass through a sieve containing
from forty to fifty holes in the linear inch, and should be
well dried and kept in an air-tight bottle, and the whole
pressed down with, the bolt, using a few taps of the hammer,
It is then heated strongly in the oxidising blowpipe flame,.
in order to drive off any hygroscopic moisture. The bone-
ash surface of the cupel, after heating, should be smooth,
and present no cracks ; if the reverse, these may be re-
moved by using the bolt again and reheating.* The silver-
lead, beaten on the anvil into the form of a cube, is placed
* These precautions are very important, as the slightest trace of moisture
in the substance of the bone-ash would inevitably cause a spurting of the
metal during the operation.
718 THE ASSAY OF SILVER.
gently upon the surface of the bone-ash, and, directing a
pretty strong oxidising flame on to its surface, it is fused,
and quickly attains a bright metallic appearance, and
commences to oxidise with a rapid rotary movement.
(Occasionally, when the assay is large, and much copper or
nickel present, the globule may, under this operation, cover
itself with a crust of lead oxide or solidify ; in such cases
direct the blue point of a strong flame steadily on to one
spot of the surface of the lead globule, until it commences
oxidising and rotating. In some cases where much nickel
is present, an infusible scale, impeding or even preventing
this action, may form, but will disappear on adding more
lead say from three to six grains, according to the thick-
ness of this scale or crust.) When this occurs the cupel
is slightly inclined from the lamp, and a fine blue point
obtained by placing the blowpipe nozzle deeper into the
flame, and the lamp is directed at about an angle of 30
on to the globule not, however, so near as to touch it
with the blue point, but only with the outer flame, so
moderating it as to keep the assay at a gentle red heat,
and not allowing the rotation to become too violent.
This oxidising fusion should be carried on at the lowest
temperature sufficient to keep up the rotatory movement,
and to prevent a crust of litharge accumulating upon the
surface of the globule, but still sufficiently high to hinder
the metallic globule from solidifying. Should this, how-
ever, happen, a stronger flame must be employed for a
moment until the metal is again in rotation ; such inter-
ruptions should, however, be avoided. The proper tem-
perature can only be learned by practice. A too high
temperature is still more injurious, causing the lead to
volatilise, and, if rich in silver, carry some of that metal
mechanically along with it. The litharge, also, instead of
remaining on the cupel, would be absorbed by the bone-
ash, and as the surface of the metallic globule is covered
by a too thin coating of fused litharge, some silver may be
absorbed along with the litharge. In this operation, in
order to avoid loss of silver, the fused globule should be
always kept in contact with the melted litharge.
FORBES'S BLOWPIPE ASSAY. 719
By the above treatment, the air has free access to the
assay, and the oxidation of the lead and associated foreign
metals goes on rapidly. The surface of the melted globule,
when poor in silver, shows a brilliant play of iridescent
colours, which does not take place when very rich in silver.
The litharge is driven to the edge of the globule, heating
itself up and solidifying behind and around it. When the
globule becomes so hemmed in by the litharge as to present
too small a surface for oxidation, the cupel is moved so as
to be more horizontal (having been previously kept in an
inclined position), thus causing the lead globule to slide
by its own weight on one side, and expose a fresh surface
to the oxidising action. When the lead is pure, the litharge
formed has a reddish-yellow colour, but if copper is
present it is nearly black.
In concentrating silver-lead, it must be remembered
that an alloy of lead and silver, if in the proportion of
about 86 per cent, silver along with 14 per cent, lead,
when cooled slowly in the litharge behaves in a manner
analogous to the spitting of pure silver, throwing out a
whitish-grey pulverulent excrescence rich in silver. For
this reason, therefore, the concentration above described
should be stopped when the globule is supposed to contain
about six parts silver along with one part in weight of
lead. In case, however, this limit should have been ex-
ceeded, it is advisable at once to push the concentration
still further until the silver globule contains but very little
lead. In practice with poor ores it is usual to concentrate
the lead until the globule is reduced to the size of a small
mustard -seed, or in rich ores to some two or three times
that size. Upon arriving at this point, the cupel is with-
drawn very gradually from the flame, so that the cooling
shall take place as slowly as possible until the globule has
solidified in its envelope of litharge. If cooled too quickly,
the litharge, contracting suddenly, would throw out the
globule, or even cause it to spirt; in such case it should
be touched by the point of the blue flame, so as to fuse it
to a round globule, which is cooled slowly, as before de-
scribed. The globule is now reserved for the next opera-
720 THE ASSAY OF SILVER.
tion, for which purpose it is, when quite cold, extracted
from the litharge surrounding it.
Cupellation.- The bone- ash required for this process
should be of the best quality and in the most impalpable
powder, prepared by elutriating finely ground bone ash,
and drying the product before use.
The cupel, still hot from the last operation, is placed
upon the anvil, and the crust of litharge, with its enclosed
metallic bead, gently removed, leaving the hot coarse bone-
ash beneath it in the mould ; upon this a small quantity
of the elutriated bone-ash is placed, so as to fill up the
cavity, and the whole, whilst hot, stamped down by the bolt,
previously slightly warmed, with a few taps of the ham-
mer. The cupel thus formed is heated strongly in the
oxidising flame, which should leave the surface perfectly
smooth, and free from any fissures or scales ; if such
appear, the bolt must again be used, and the cupel re-
heated. In this process it is very important that the cupel
should possess as smooth a surface as possible, whilst at
the same time the substance of the cupel beneath should
not be too compact, so as thereby to permit the litharge
to filter through and be readily absorbed, leaving the silver
bead upon the smooth upper surface.
The bead of silver-lead obtained from the last operation
is taken out of the litharge in which it is embedded, and,
after removing any trace of adherent bone-ash or litharge,
is slightly flattened to prevent its rolling about upon the
surface of the cupel.
It is now put into the cupel prepared as before de-
scribed, placing it on the side farthest from the lamp and
a little above the centre of the cupel, which is now in-
clined slightly towards the lamp, and is heated by the
oxidising flame directed downwards upon it, this causing
the globule, when fused and oxidising, to move of itself
into the centre of the cupel. The cupel is now brought
into a horizontal position, and the flame, directed on to it at
an angle of about forty-five degrees, is made to play upon
the bone-ash surface immediately surrounding the globule,
without, however, touching it, so as to keep this part of
FORBESS BLOWPIPE ASSAY. 721
the cupel at a red heat sufficiently strong to insure the
globule being in constant oxidising fusion, and at the same
time to cause the perfect absorption of the litharge, so as
to prevent any scales of litharge forming upon the surface
of the cupel under the globule, which would impede the
oxidation, as well as prevent the silver bead being easily
detached at the conclusion of the operation. Should the
heat at any time be too low and the globule solidify, it
must be touched for an instant with the point of the flame
and proceeded with as before. Should (in consequence of
the bone-ash not having been sufficiently heated to absorb
the litharge perfectly) a little litharge adhere pertinaciously
to the globule, or a particle of the bone-ash cupel attach
itself, the cupel should be slightly inclined, so as to allow
the globule to move by its own weight on to another and
clean part of the cupel, leaving the litharge or bone-ash
behind it ; but, if not sufficiently heavy to do so, a small
piece of pure lead may be fused to it in order to increase its
weight, and so allow of the same proceeding being adopted.
By slightly inclining the cupel-stand, and moving it so
as to present in turn all parts of the surface surrounding
the globule to the action of the flame, the cupellation pro-
ceeds rapidly. If, however, the assay contains very little
silver, it will be found necessary to move the globule from
one spot to another on the cupel, in order to present a fresh
surface for absorbing the litharge formed ; this is done by
simply inclining the cupel-stand, remembering that the bone-
ash surrounding the globule must always be kept at a red
heat, without ever touching the globule itself by the flame.
In assays rich in silver a play of iridescent colours
appears some seconds before the ' brightening,' which dis-
appears the moment the silver becomes pure ; as soon as
this is observed the cupel should be moved in a circular
manner, so that the globule is nearly touched all round by
the point of the blue flame, and this is continued until the
surface of the melted silver is seen to be quite free from
any litharge, upon which it is very gradually withdrawn
from the flame so as to cool the assay by degrees very
slowly, in order to prevent ' spitting.'
3 A
722 THE ASSAY OF SILVER.
When the silver-lead is very poor, this play of colours
is not apparent, and as soon as the rotatory movement of
the globule ceases, the heat must be increased for an
instant, in order to remove the last thin but pertinacious
film or scale of litharge, and subsequently the assay is
cooled gradually ; when cold it should, whilst still upon
the cupel, be examined by a lens, to see whether the bead
possesses a pure silver colour, as, if not, it must be reheated.
Frequently, when the ' brightening ' takes place, the
silver globule is found to spread out, and, after cooling,
although of a white colour, is found to appear somewhat
less spherical or more flattened in shape than a correspond-
ing globule of pure silver would be. This arises from the
presence of copper still remaining in the silver, and in such
cases a small piece of pure lead (about from one- half to
one and a half grain in weight, according to size of assay)
should be fused on the cupel along with the silver, and the
cupellation of the whole conducted as before on another
part of the cupel, when the silver globule will be obtained
pure, and nearly spherical in shape. Sometimes the silver
globule in ' brightening ' may still remain covered with a
thin film of litharge, although otherwise pure ; this arises
from too little heat having been employed in the last stage
of the operation, and consequently the bead should be
re-heated in a strong oxidising flame until this litharge is
absorbed, and the globule, after slow cooling, appears pure.
If the instructions here given be strictly attended to, it
will be found after some practice that very accurate results
may be obtained in the blowpipe assay for silver, and that
no difficulty will be found in detecting the presence and
estimating the amount of silver present, even when in as
small a quantity as half an ounce to the ton. When sub-
stances containing very little silver or less than that amount
are examined, several assays should be made, and the
silver-lead obtained concentrated separately, after which
the various globules should be united and cupelled together
in one operation.
It is hardly necessary to remark that the lead employed
in assaying should be free from silver, or, if not, its actual
FORBES'S BLOWPIPE ASSAY. 723
contents in silver should be estimated, and subtracted from
the amount found in the assay.
Assay lead containing less than one quarter of an ounce
to the ton of lead can readily be obtained, or can be made
by precipitating a solution of acetate of lead by metallic
zinc, rejecting the first portion of lead thrown down. In
all cases the lead should be fused and granulated finely
the granulated lead for use in these assays being previously
passed through a sieve containing forty holes to the linear
inch. It is also useful to have some lead in the form of
wire, as being very convenient for adding in small portions
to assays when on the cupel.
Estimation of the Weight of the Silver Globule obtained
on Cupellation. As the amount of lead which can, by
the method before described, be conveniently cupelled
before the blowpipe is necessarily limited, the silver glo-
bule which remains upon the bone-ash surface of the cupel
at the end of the operation is, when substances poor in
silver have been examined, frequently so very minute that
its weight could not be estimated with correctness by
the most delicate balances in general use.
The blowpipe balance employed by the author turns
readily with one-thousandth of a grain, but could not be
used for estimating weights below that amount.
Globules of silver of far less weight than one-thousandth
are distinctly visible to the naked eye a circumstance
which induced Harkort to invent a volumeti|ical scale
based upon the measurement of the diameters of the glo-
bules, which scale in practice has been found of very great
utility in the blowpipe assay of silver.
The scale for this purpose which is employed by the
author is shown in full size in the woodcut on p;. 724.
This figure represents a small strip of highly polished
ivory about 6J inches long, f inch broad, and inch in
thickness, on which are drawn, by an extremely fine point,
two very fine and distinct lines emanating from the lower
or zero point, and diverging upwards until, at the distance
of exactly six English standard inches, they are precisely
four hundredth parts of an inch apart. This distance
3 A 2
724
THE ASSAY OF SILVER.
(six inches) is, as shown in woodcut, divided into 100 equal
parts by cross lines numbered in accordance from zero
FIG. i4i. upwards. It is now evident if a small globule
of silver be placed in the space between these
two lines, using a magnifying glass to assist
the eye in moving it up or down until the
diameter of the globule is exactly contained
within the lines themselves, that we have at
once a means of estimating the diameter of
the globule itself, and therefrom are enabled
to calculate its weight.
As the silver globules which cool upon
the surface of the bone-ash cupel are not true
spheres, but are considerably flattened on the
lower surface, where they touch and rest upon
the cupel, it follows that the weight of glo-
bules corresponding in diameter to the extent
of divergence at the different degrees of the
scale cannot be calculated directly from their
diameters as spheres, but require to have
their actual weight experimentally determined
in the same manner as employed byPlattner.
.. -The table here appended has been cal-
culated by the author, and in one column
shows the diameter in English inches corre-
sponding to each number or degree of the
scale itself, and in the two next columns the
.respective weights of the flattened spheres
which correspond to each degree or diameter ;
for convenience these weights are given in
the different columns in decimals, both of
English grains and of French grammes.
These weights are calculated from the
following data, found as the average result of
several very careful and closely approximating
assays, which showed that globules of silver
exactly corresponding to No. 95 on this scale,
or 0*038 inch in diameter, possessed a weight
of 0-0475573 grains or 0-003079 grammes. Erom this the
100
98
96
94
92 _:
90 :
88 "
1- :
86
84
82
80 ~
78
76 ~
74
72
70
68 -
66 -f
64
62 -
60
68 -
56 7
:~-
50 :
48 -
46
44
v
42 -
40 -
38
36 -
'34
; 32 -
5
28-^
W
24
22
SO
18 -^
16 -3
14
12
|
i 10
6 ~
o *
FORBES 8 BLOWPIPE ASSAY.
725
No. on
scale
Greatest diameter
in inches
Weight of globule in
grains
Weight of globule in
grajnmes
1
0-0004
0-00000005
0-000000003
2
0-0008
0-00000044
0-000000028
3
0-0012
0-00000149
0-000000096
4
0-0016
0-00000355
0-000000229
5
0-0020
0-0000069
0-00000044
6
0-0024
0-0000119
0-00000077
7
0-0028
0-0000190
0-00000120
8
0-0032
0-0000284
0-00000184
9
0-0036
0-0000403
0-00000262
10
0-0040
0-0000554
0-00000359
11
0-0044
0-0000736
0.00000478
12
0-0048
0-0000958
0-00000620
13
0-0052
0-0001218
0-00000789
14
0-0056
0-0001522
0-00000985
15
0-0060
0-0001872
0-00001203
16
0-0064
0-0002272
0-00001471
17
0-0068
0-0002725
0-00001764
18
0-0072
0.0003234
0-00002094
19
0-0076
0-0003804
0-00002463
20
0-0080
0-0004437
0-00002872
21
0-0084
0-0005137
0-00003327
22
0-0088
0-0005906
0-00003823
23
0-0092
0-0006748
0-00004367
24
0-0096
0-0007668
0-00004964
25
0-0100
0-0008667
0-00005611
26
0-0104
0-0009749
0-00006311
27
0-0108
0-0010918
0-00007068
28
0-0112
0-0012176
0-00007883
29
0-0116
0-0013528
0-00008758
30
0-0120
0-0014976
0-00009696
31
0-0124
0-0016524
0-00010698
32
0-0128
0-0018176
0-00011677
33
0-0132
0-0019934
0-00012817
34
0-0136
0-0021801
0-00014114
35
0-0140
0-0023786
0-00015397
36
0-0144
0-0025879
0-00016755
37
0-0148
0-0028097
0-00018190
38
0-0152
0-0030437
0-00019705
39
0-0156
0-0032903
0-00021302
40
0-0160
0-0035550
0-00022983
41
0-0164
0-0038230
0-00024751
42
0-0168
0-0041096
0-00026606
43
0-0172
0-0044111
0-00028553
44
0-0176
0-0047250
0-00030589
45
G'0180
0-0050546
0.00032725
46
0-0184
0-0053991
0-00034955
47
0-0188
0-0057590
0-00037285
48
0-0192
0-0061344
0-00039716
49
0-0196
0-0065258
0-00042250
50
0-0200
0-0069335
0-00044890
51
0-0204
0-0073581
0-00047638
52
0-0208
0-0077799
0-00050495
53
0-0212
0-0082580
0-00053464
54
0-0216
0-00873438
0-00056549
55
0-0220
0-00922854
0-00059748
56
0-0224
0-0097412
0-00063067
726
THE ASSAY OF SILVEK.
Jfo. on
scale
Greatest diameter
in inches
Weight of globule in
grains
Weight of globule in
grammes
57
0-0228
0-0102725
0-00066506
58 0-0232
0-0108228
0-00070021
59 0-0236
0-0113922
0-00073753
60 0-0240
0-0119815
0-00077570
61 0-0244
0-0125901
0-00081513
62 0-0248
0-0132119
0-00085588
63 0-0252
0-0138901
0-00089797
64 0-0256
0-0145440
0-00094141
65 0-0260
0-0152311
0-00098623
66
0-0264
0-0159472
0-00103245
67
0-0268
0-0166828
0-00108010
68 0-0272
0-0174414
0-00112918
69
0-0276
0-0182220
0-00117974
70
0-0280
0-0190256
0-00123177
71
0-0284
0-0198529
0-00128535
72
0-0288
0-0207035
0-00134041
73
0-0292
0-0215782
0-00139704
74
0-0296
0-0224469
0-00145525
75 0-0300
0-0234010
0-00151504
76 0-0304
0-0243496
0-00157645
77
0-0308
0-0253224
0-00163950
78
0-0312
0-0263228
0-00170422
79
0-0316
0-0273484
0-00177060
80
0-0320
0-0284000
0-00183869
81
0-0324
0-0294789
0-00190852
82
0-0328
0-0305838
0-00198008
83
0-0332
0-0317162
0-00205340
84
0-0336
0-0328768
0-00212851
85
0-0340
0-0340649
0-00220549
86
0-0344
0-0349739
0-00228400
87
0-0348
0-0364422
0-00235938
88
0-0352 0-0378008
0-00244730
89
0-0356 0-0390138
0-00253168
90
0-0360 0-0404368
0-00261797
91
0-0364 0-0417943
0-00270790
92
0-0368 0-0431930
0-00279642
93
0-0372 0-0446162
0-00288860
94
0-0376 0-0460718
0-00298279
95
0-0380 0-0475573
0-00307900
96
0-0384 0-0465239
0-00317728
97
0-0388 0-0506249
0-00327759
98
0-0392 0-0522069
0-00338020
99
0-0396 0-0538215
0-00348452
100
0-0400 0-0554688
0-00359138
respective weights of all the other numbers or degrees on
this scale were calculated, on the principle that solids were
to one another in the ratio of the cubes of their diameters.
This mode of calculation is not, however, absolutely correct
in principle, for the amount of flattening of the under sur-
face of the globule diminishes in reality with the decreasing
FORBES'S BLOWPIPE ASSAY. 727
volume of the globule. In actual practice, however, this
difference may be assumed to be so small that it may be
neglected without injury to the correctness of the results.
The smaller the diameter of the globule, the less will
be the difference or variation in weight in descending the
degrees of this scale, since the globules themselves vary in
weight with the cubes of their diameters : for this reason,
also, all such globules as come within the scope of the
balance employed should be weighed in preference to being
measured, and this scale should be regarded as more spe-
cially applicable to the smaller globules beyond the reach
of the balance.
Cupellation Loss. This term is applied to indicate a
minute loss of silver, unavoidably sustained in the process
of cupellation, which arises from a small portion of that
metal being mechanically carried along with the litharge
into the body of the cupel. The amount of this loss in-
creases with the quantity of lead present in the assay
(whether contained originally in the assay or added subse-
quently for the purpose of slagging off the copper, &c.) ;
it is relatively greater as the silver globule is larger, but
represents a larger percentage of the silver actually con-
tained in the assay, in proportion as the silver globule
obtained diminishes in size. It has, however, been experi-
mentally proved that, in assays of like richness in silver,
this loss remains constant when the same temperature has
been employed, and similar weights of lead have been
oxidised in the operation.
In the blowpipe assay this loss is not confined to the
ultimate operation of cupellation, but occurs, though in a
less degree, in the concentration of the silver-lead, and in
the previous scorification of the assay, had such operation
preceded the concentration. The total loss in the blow-
pipe assay is found, however, to be less than in the ordinary
muffle assay, since in the latter case the whole of the oxi-
dised lead is directly absorbed by the cupel.
In mercantile assays of ore it is not customary to pay
attention to the cupellation loss, and the results are usually
stated in the weight of silver actually obtained. Where,
728
THE ASSAY OP SILVER.
however, great accuracy is required, especially when the
substances are very rich in silver, the cupellation loss is
added to the weight of the silver globule obtained, in order
to arrive at the true percentage.
The amount to be added for this purpose is shown
in the annexed table, which is slightly modified from
Plattner's :
Actual per-
centage of
silver found by
assay
Cupellation loss, or percentage of silver to be added to the actual percentage
found by assay in order to show the true percentage of silver contained in same.
The entire amount of lead in or added to the assay being the folio wing multiples
of the original weight of assay :
1
2
3
4
5
6 I 8
1
11
13
16
1
99-5 /
0-25
0-32
0-39
0-45
0-50
90
0-22
0-29
0-36
0-42
0-47
0-69
0-83
80
0-20
0-26
0-33
0-39
0-44
0-64
0-75
70
0-18
0-23
0-29
0-35
0-40
0-58
0-68
0-82
60
0-16
0-20
0-26
0-30
0-36
0-52
0-61
0-74
50
0-14
0-17
0-23
0-26
0-32
0-46
0-54
0-65
40
0-12
0-15
0-20
0-22
0-27
0-39
0-46
0-55
0-62
35
0-11
0-13
0-18
0-18
0-25
0-36
0-42
0-50
0-57
30
0-10
0-12
0-16
0-16
0-22
0-32
0-38
0-45
0-51
25
0-09
0-10
0-14
0-14
0-20
0-29
0-34
0-40
0-45
20
0-08
0-09
0-12
0-12
0-17
0-25
0-29
0-35
0-39
0-45
15
0-07
0-08
0-10
0-11
0-15
0-20
0-23
0-28
0-32
0-37
12
0-06
0-07
0-09
0-10
0-13
0-17
0-19
0-23
0-26
0-22
10
0-05
0-06
0-08
0-09
0-11
0-15 i 0-17
0-20
0-23
0-27
9
0-04
0-05
0-07
0-08
0-10
0-14
0-16
0-18
0-21
0-25
8
0-03
0-04
0-06
0-07
0-09
0-13
0-15
0-16
0-18
0-22
7
0-02
0-03
0-05
0-06
0-08
0-12 i 0-13
0-14
0-16
0-20
6
0-01
0-02
0-04
0-05
0-07
0-10
0-11
0-12
0-14
0-17
5
0-91
0-03
0-04
0-06
0-09
0-10
0-11
0-12
0-14
4
0-02
0-03
0-05 i 0-07
0-08
0-09
0-10
0-11
3
0-01
0-02
0-04
0-05
0-06
0-07
0-08
0-09
2
0-01
0-03
0-04
0-04
0-05
0-06
0-07
1
,
0-01
0-03
0-03
0-04
0-04
0-05
The use of this table is best explained by an example,
as the following : An assay to which there had been
added, in all, five times its weight of assay lead, gave a
globule of silver equivalent to six per cent. Upon referring
to the table it will be seen that the cupellation loss for this
would be 0*07 ; consequently the true percentage of silver
contained in the assay would be 6'07. This table is only
extended to whole numbers, but fractional parts can easily
be calculated from the same.
FOEBES'S BLOWPIPE ASSAY. 729
When the globules of silver are so minute that they
cannot be weighed, but must be measured upon the scale,
the cupellation loss should not be added, since, as a rule,
it would be less than the difference which might arise from
errors of observation likely to occur when measuring their
diameters upon the scale.
In the case of beginners, it will be found that the cupel-
lation is usually carried on at too high a temperature, and
that thereby a greater loss is occasioned than would be
accounted for by the above table. After some trials the
necessary experience will be acquired in keeping up the
proper temperature at which this operation should be
effected.
It now becomes necessary to consider in detail the
processes requisite for extracting the silver contents (in
combination with lead) from the various metallic alloys of
silver which are met with in nature or produced in the
arts.
In considering these, the following classification of the
substances will be found convenient :
METALLIC ALLOYS.
A. Capable of direct Cupellation.
a. Consisting chiefly of lead or bismuth ; silver-lead and
argentiferous bismuth, native bismuthic silver.
6. Consisting chiefly of silver : native silver, bar silver, test
silver, precipitated silver, retorted silver amalgam, stan-
dard silver, alloys of silver with gold and copper.
c. Consisting chiefly of copper : native copper, copper ingot,
sheet or wire, cement copper, copper corns, copper-
nickel alloys.
B. Incapable of direct Cupellation.
a. Containing much copper or nickel, with more or less
sulphur, arsenic, zinc, &c. ; unrefined or black copper,
brass, German silver.
6. Containing tin ; argentiferous tin, bronze, bell-metal, gun-
metal, bronze coinage.
c. Containing antimony, tellurium, or zinc.
d. Containing mercury : amalgams.
e. Containing much iron : argentiferous steel, bears from
smelting furnaces.
730 THE ASSAY OF SILVER.
A. METALLIC ALLOYS CAPABLE OF DIBECT CUPELLATION.
a. Consisting chiefly of Lead or Bismuth. In estimat-
ing the silver contained in these alloys, it is only requisite
to place a clean piece of the same, weighing about from
one to ten grains according to its probable richness in
silver, upon a cupel of coarse bone-ash, and proceed by
concentration and cupellation exactly as has been already
described under these heads.
Should the substance be not altogether metallic, or
not free from adherent slag, earthy or other extraneous
matter, it should previously be fused on charcoal with a
little borax in the reducing blowpipe flame, and the clean
metallic globule then removed from the charcoal, and
treated as before. In order to remove the globule from
the adherent borax-glass, it may be allowed to cool, and
then detached ; or, after a little practice, it will be found
easy, by a quick movement of the charcoal, to cause the
globule, still melted, to detach itself completely, and drop
on the anvil in the form of a single somewhat flattened
globule, without suffering any loss of lead adhering to the
charcoal.
In the case of argentiferous bismuth alloys the process
is carried on in all respects the same as if silver-lead were
being treated. As, however, the bismuth globule is very
brittle, care must be taken when separating the concen-
trated globule from the litharge, as, if not carefully done,
a loss may easily be sustained from a portion of the glo-
bule remaining behind adherent to the litharge. It is
better, therefore, to remove the litharge by degrees from
the globule with the aid of the forceps.
Argentiferous bismuth, free from lead, when cupelled
alone, invariably leaves a globule of silver, having a dull
frosted surface. If, however, at the end of the operation
a small quantity of lead 4 (i to ^ a grain) be added, and
fused along with it, the silver globule then obtained will
be perfectly bright and free from all bismuth.
In the case of native bismuthic silver it is advisable to
FOKBES'S BLOWPIPE ASSAY. 731
fuse the previously weighed mineral with a little lead and
borax-glass on charcoal in the reducing flame, so as to
free it from any adherent earthy matter, and then proceed
by concentration and cupellation, as before described.
b. Consisting chiefly of Silver: native silver, bar, test,
and precipitated silver, retorted silver amalgam, standard
silver, silver coin and other alloys of silver with gold and
copper. These alloys may be at once fused with lead
on the cupel itself, and the operation finished as before
described. In general, however, it is better to fuse the
weighed assay previously with the requisite amount of
pure lead and a little borax-glass, say from a quarter to
half the weight of assay, in the reducing flame at a low
heat on charcoal until the globule commences to rotate.
This insures having a perfectly clean button of silver-lead,
which is then cupelled in the ordinary manner.
In most cases the quantity of lead to be added need
not exceed that of the weight of the alloy, but when
several percentages of copper are present in the assay, as
in the case of many coins, &c., the lead should be increased
to some three, or even five, times the weight of the assay
in proportion to the amount of copper actually contained
in the substance under examination, and which will be
treated of more at length under the head of copper-silver
alloys.
When no more lead has been added to the assay than
its own weight, the cupellation may be concluded in one
operation by inclining the stand, and so moving the glo-
bule to a clean part of the cupel ; but when more copper
is present, it is preferable to concentrate first and cupel
subsequently, in order thereby to reduce the cupellation
loss to its minimum.
In the concentration as much copper as possible should
be slagged off with the lead, which is effected by inclining
the cupel somewhat more than usual, so that its surface
may be less covered up with the litharge and exposed as
much as possible to oxidation, by which means the litharge,
as it forms, is enabled to carry off more of the copper
contained in the silver-lead.
732 THE ASSAY OF SILVER.
Should the silver globule after cupellation show indi-
cations of still containing copper, as before noticed, when
treating of cupellation, a small quantity of lead must be
fused along with it, and the cupellation finished as usual.
As at the present time no means are known by which
silver can be separated from gold by the use of the blow-
pipe in all cases of alloys containing gold, this metal
remains to the last along with the silver, and the result in
such cases always indicates the combined weight of both
these metals contained in the alloy under examination.
The wet assay must be resorted to for effecting their
separation.
c. Containing chiefly Copper : native copper, ingot, wire
or sheet copper, cement copper, copper coins, copper-nickel
alloys. Under the most favourable conditions in cupella-
tion, the amount of lead requisite, when converted into
litharge, to slag off one part of copper along with it as
oxide, amounts to between seventeen and eighteen times
its weight. In the blowpipe assay it is usual to add to
any cupriferous alloy an amount of pure lead equal to
twenty times the amount of copper contained in the alloy,
in order to insure the whole of the copper being separated
in the litharge. In the case of nickel the amount of lead
required is somewhat less than with copper, but in prac-
tice the same amount of lead may be employed.
When the copper is quite clean the requisite amount
of lead may be added to it in a single piece on the cupel,
fused and cupelled as usual, after previous concentration
of the silver-lead to a small-sized globule.
It is generally found, however, that traces of iron, slag,
gangue, or other foreign matter, are present ; and, conse-
quently, it is usually advisable to fuse the assay along
with the requisite amount of lead, and about one-half its
own weight of borax-glass in the reducing flame, until the
whole of the substance is seen to have perfectly combined
or alloyed with the lead, and the globule has entered into
brisk rotation, whilst at the same time no detached
metallic globules are seen in the borax-glass.
The concentration of the silver-lead and cupellation
FORBES'S BLOWPIPE ASSAY. 733
are then conducted as usual, taking care when concentrat-
ing to incline the cupel-stand so as to expose as much as
possible of the metallic surface of the melted globule to
the oxidising action of the air, with a view of enabling the
litharge whilst forming to carry off as much copper along
with it as possible.
Should the silver globule obtained after cupellation
spread out, or appear to the eye more flattened than
usual with globules of pure silver, it indicates that some
copper still remains, and a small piece of assay lead (-J to
1 grain weight) should be placed alongside it whilst still
on the cupel, fused together, and the cupellation finished
on a clean part of the same cupel as usual.
Precipitated or cement copper, especially that which
is in the crude state, and has not been melted and run
into ingots, is often very impure, containing so much iron,
lead, arsenic, earthy matter, &c., as not to admit of direct
cupellation, and in such case should be treated as pertain-
ing to class B. a.
B. METALLIC ALLOYS INCAPABLE OF DIRECT CUPELLATION.
a. Containing much Copper or Nickel, with frequently
some little sulphur, arsenic, zinc, iron, cobalt, fyc. ; as unre-
fined or black copper, brass, German silver, fyc. As the
presence of these extraneous matters would interfere with
the cupellation, either by causing a loss of silver-lead
projected from the cupel upon the evolution of the
volatile substances present, or by forming oxides which
could not be absorbed by the cupel, it is necessary to
eliminate such substances by a scorification with borax
on charcoal, previous to concentration or cupellation.
In the case of unrefined and black copper, the portion
used in the examination is placed in the scoop with twenty
times its weight of assay lead, and its own weight of pow-
dered borax-glass, mixed with the spatula, and transferred
to a soda-paper cornet. It is then fused on charcoal in
the reducing flame, which should be constant and uninter-
rupted, until all particles have completely united, and a
734 THE ASSAY OF SILVER.
brisk rotation sets in, which is kept up for a short time,
when the silver-lead globule, which should appear bright
on the surface after cooling, is concentrated and cupelled
precisely as is directed under A. c. By this preliminary
scorification the sulphur, arsenic, and zinc are volatilised,
and any lead, cobalt, or iron slagged off into the borax-
glass.
In the assay of brass and German silver, the quantity
employed is fluxed with its own weight of borax-glass,
but only requires ten times its weight of assay-lead. The
operation is commenced as before, but the globule is kept
somewhat longer in rotation (always keeping the flame
directed only on to the borax glass), so as to allow the zinc
present to be completely volatilised, which is evident when
the surface of the silver-lead becomes bright, on which the
heat is increased for a few moments to expel the last traces
of that metal, and the silver-lead thus obtained is concen-
trated and cupelled as before.
The silver globule obtained from the cupellation of
substances rich in copper generally requires the addition
of a small quantity of lead and re-cupellation (as before
described), in order to insure its freedom from copper.
b. Containing Tin: argentiferous tin, bronze, bell and
gun metal, bronze coinage, fyc. Alloys of silver with other
metals containing tin do not admit of being cupelled, since
the oxide of tin formed by the oxidation of that metal is
not absorbed by the bone-ash of the cupel along with the
litharge ; it consequently remains upon the surface of the
cupel, and if present in any quantity interferes with the
operation. As tin is not volatile when heated on charcoal
either in the oxidising or reducing blowpipe flame, it
cannot be so dissipated, and, in consequence, the entire
amount of tin contained in any alloy under examination
must be removed by oxidation or scorification from the
silver-lead, previous to its being submitted to cupellation.
For this purpose, 1 part of the stanniferous alloy is
fluxed with from 5 to 15 parts of granulated assay lead (ac-
cording to the amount of copper suspected to be present in
the alloy), O5 part anhydrous sodium carbonate, and 0'5
FORBES'S BLOWPIPE ASSAY. 735
part pulverised borax-glass, made up as usual in a soda-
paper cornet, and the whole at first gently heated in the
reducing flame until the soda-paper is charred and the
alloy has afterwards united with the lead to form a single
globule, whilst the borax and soda have combined as a
glass or slag in which the soda prevents the easily oxidis-
able tin becoming oxidised to any extent before a perfect
alloy has been formed with the lead, which then contains
the whole of the silver.
As soon as this is effected, the blowpipe flame is altered
to an oxidising one, and the metallic globule is kept at
the point of the blue flame, which should touch it so as to
cause the tin to become oxidised and be at once taken up
by the glass surrounding it.
Should, however, it be seen that minute globules of
metallic tin made their appearance on the outer edge of
the slag or glass,* the operation must be at once discon-
tinued, and the assay allowed to cool ; after cooling the
metallic globule is detached from the slag surrounding it,
and, being placed in a cavity on charcoal, is fused in the
reducing flame along with a small piece of borax-glass
and afterwards treated with the oxidising flame exactly as
before (and if necessary, which is seldom the case unless
when treating argentiferous block-tin, this operation may
again require to be repeated), until it is seen that the
surface of the metallic silver-lead globule does not any
longer become covered with a crust or scales of tin oxide,
but presents a pure and bright metallic surface.
The silver-lead globule is now quite free from tin, and
can be cupelled and the amount of silver estimated as
usual.
c. Metallic alloys containing much antimony, tellurium,
or zinc : antimonial silver and argentiferous antimony, tel-
luric silver, and argentiferous zinc. Alloys of antimony
with silver when treated on charcoal in the oxidising
flame give off all their antimony, leaving the silver behind
as a metallic globule having a frosted external appearance ;
* This occurs when the flux has become so saturated with tin oxide that it
cannot take up any more.
736 THE ASSAY OF SILVEft.
telluric silver, on the contrary, however, when treated
in a similar manner, only evolves a part of its tellurium,
and even after cupellation with lead a small amount
of tellurium generally remains behind alloyed with the
silver.
All these compounds may be assayed as follows :
One part of the alloy is placed in a soda-paper cornet
along with 5 parts of granulated assay lead, and 0-5 part
of pulverised borax-glass, and fused in the reducing flame
until the globule and slag are well developed ; the oxidising
flame is now directed on to the globule, causing the
whole of the zinc, along with most of the antimony and
part of the tellurium, to volatilise before the lead com-
mences oxidising. The last traces of antimony are re-
moved with some difficulty, during which operation some
portion of the lead becomes oxidised. On cooling, the
globule is separated from the slag and concentrated upon
a coarse bone-ash cupel as usual, and if 110 tellurium were
present in the concentrated silver-lead, this may now be
cupelled as usual.
If tellurium be present as is seen by the concentrated
globule of silver-lead possessing a dark-coloured exterior
it must be re-melted with 5 parts of assay lead and again
concentrated ; and these operations, if necessary, must be
repeated until the surface of the concentrated globule is
found to be clean and bright, as is usual with pure silver-
lead, when it may be cupelled fine and the silver globule
weighed.
It sometimes happens, even after all these precautions
have been taken, that the silver globule after cupellation
shows a crystalline, greyish-white, frosted appearance, from
its still containing tellurium ; in such cases its own weight
of assay lead (in one piece) should be placed beside it on
the cupel, melted together, and the globule again cupelled
fine on another part of the surface of the same cupel. In
assaying substances very rich in tellurium the results ob-
tained are, however, not very satisfactory, and may be as
much as one or two per cent, too low, even after employ-
ing all precautions.
FOKBES'S BLOWPIPE ASSAY. 737
d. Compounds of Silver with Mercury : arquerite, native
and artificial amalgams and argentiferous mercury. The
assay of these compounds is very simple. A weighed
quantity of the liquid or solid amalgam is placed in a
small bulb-tube, and heated over the lamp very gradu-
ally in order to avoid spirting and to allow the mercury
to volatilise quietly ; * the heat is increased by degrees
as long as any mercury is driven off, and the residue is
heated for some time at a red heat in order to drive off
as much mercury as possible without fusing the glass or
causing the residual silver to adhere to it. The mercury
expelled condenses above the bulb on to the upper
part of the tube, and by gently tapping will collect in
globules, which, by carefully turning the tube, unite and
can be poured out of the tube ; after which the silver,
left behind as a porous mass, may be removed from the
tube, and after being fluxed with an equal weight of
granulated assay lead and half its weight of borax-glass,
must be fused on charcoal in the reducing flame, and the
button, on cooling, cupelled as usual. Should, however,
much copper have been present in the amalgam, a pro-
portionately larger amount of assay lead is required to be
added.
When the argentiferous residue is extremely small, as
is often the case when assaying argentiferous mercury,
this may adhere firmly to the glass of the tube. On such
occasions this part of the tube must be cut off with the
adherent residue, and the whole fused in a strong reducing
flame along with its own weight of granulated assay lead,
and with half its weight of anhydrous sodium carbonate.
Upon cooling, the globule of silver-lead thus obtained is
cupelled as usual.
e. Compounds chiefly consisting of Iron: argentiferous
steel ; cast iron ; bears from smelting furnaces. Compounds
consisting principally of iron with a small percentage of
silver, although occasionally produced in the arts inten-
In the case of solid amalgams, which often spirt very violently, this
may be obviated by wrapping the assay in a small piece of tissue paper, and
heating it in a blowpipe crucible, when all the mercury is given off quietly,
leaving the silver behind.
SB
738 THE ASSAY OF SILVER.
tionally, as, for example, the so-called silver-steel, are
commonly found on the blowing-out of furnaces used in
the smelting of silver and copper ores, and are frequently
rich in silver, as is the case with the bears from the silver
furnaces at Kongsberg in Norway. An alloy of iron with
silver is occasionally also found appearing in small quanti-
ties on the surface of melted silver in the process of cast-
ing, and in some cases at least this may be due to the action
of the melted silver on the iron rods used for stirring up
the molten metal.
As iron cannot be made to alloy itself with lead before
the blowpipe, it becomes necessary to extract the silver by
a more indirect process than is used in the case of other
alloys containing that metal. In order to remove the iron
the alloy must first be converted into iron and silver sul-
phide, and to effect this the iron or steel must be reduced
to powder, or fragments none greater than about a quarter
of a grain in weight ; for which purpose steel when har-
dened may require to be softened previously.
One part of the finely divided iron or steel is now mixed
with O75 part sulphur, eight parts granulated assay lead,
and one part pulverised borax-glass : the mixture after
being placed in a soda-paper cornet is carefully fused in a
cavity on charcoal in the reducing flame, until the whole
appears as a fluid globule containing both the lead and
iron in combination with the sulphur. Without removing
either this globule or the glass surrounding it from the
charcoal, an amount of borax-glass in one or more frag-
ments (in all about equal in weight to the original amount
of iron employed), is now added (in order to combine with
and slag off the whole of the iron), and fused along with the
former globule, after which the whole is submitted to a
strong oxidising flame until the impure lead globule shows
itself protruding from the slag.
The charcoal is then inclined, so that the lead is alone
subjected to the action of the outer flame, in order to vola-
tilise the sulphur, and at the same time oxidise the iron
which goes into the slag ; this operation is continued until
the globule of lead appears with a bright metallic surface ;
ASSAY OF ALLOYS OF SILVER AND COPPER. 739
should it on cooling, however, be found to possess a black
colour, and to be brittle, it must be still further oxidised
as before described.
The silver-lead thus obtained will now be found to
contain all the silver, and at the same time to be free from
both iron and sulphur, and can be cupelled as usual.
No notice is here taken of alloys of silver and gold,
since these metals cannot be separated before the blowpipe
by any process yet known ; and in all cases where gold
may be present in an alloy treated as here directed for
obtaining its contents in silver, the gold also will be found
to follow along with the silver, and must be parted from
that metal by the wet method, in order to enable the
true amount of silver present in the substance to be ascer-
tained.
Alloys of Silver and Copper. A quantity of from 1
to 2 grms. is placed in a small flask or beaker, treated
with a sufficient quantity of pure nitric acid of a moderate
strength, covered with a watch-glass and digested at a
gentle heat till totally dissolved. The solution is then
diluted with water, and the silver precipitated with hydro-
chloric acid, which is added drop by drop as long as any-
thing is thrown down. It is then allowed to digest at a
very gentle heat till the supernatant liquid is clear. The
liquid is then poured upon a dried and counterpoised filter ;
the silver chloride is stirred up Avith a little water, and
brought upon the filter, the glass being perfectly rinsed out
by aid of the washing-bottle. A few drops of nitric acid
may be usefully added to the rinsing-water. The washing
is afterwards continued with pure water till the droppings
have no longer an acid reaction. The filter with its con-
tents is dried, and from its weight that of the silver is
calculated.
740
CHAPTEE XVII.
THE ASSAY OF GOLD.
FOR the purposes of assay, all substances containing gold
may be divided into two classes, as in the case of silver.
The First Class comprises ores containing gold in a
mineralised form.
The Second Class, comprises all alloys of gold, native or
artificial.
CLASS I. Minerals.
Graphic Tellurium, (AuAg) Te 3 , containing 30 per cent.
of gold, and 10 per cent, of silver.
Foliated Tellurium, (Pb,Au,Ag) 2 (Te,Sb,S) 3 containing
about 9 per cent, of gold.
CLASS n. Alloys of Gold.
Native Gold, Au Ag, containing 65-99 per cent, of
gold.
Palladium Gold, AuPd, containing about 86 per cent,
of gold.
Rhodium Gold, AuEh, containing 59-66 per cent, of
gold,
Gold Amalgam, (Au,Ag) 2 HgJ, containing 38 per cent,
of gold.
Artificial Alloys, Gold coin, jewellery, &c.
Of the foregoing list, native gold alone occurs in nature
in sufficient abundance to acquire any great commercial
value. It is commonly found in a quartzose gangue, and
nearly always associated with one, or more, of the following
. NATIVE GOLD. 741
minerals : iron and copper pyrites, mispickel or arsenical
pyrites, blende, galena, many antimonial minerals, and
nearly all the primitive rocks. All auriferous slags, amal-
gamation residues and tailings, belong to this class.
If silver or platinum is associated with the gold in the
mineral subjected to assay, it will be found combined with
the gold obtained by cupellation ; and all gold so obtained
must be submitted to the ' parting process,' which see
further on.
When gold is associated in quantity with quartz, its
percentage can be approximately ascertained in the same
manner as that of pure tin-stone when mixed with quartz
(see page 545). If possible, a fragment of the gold must
be detached from the quartz, and its specific gravity
taken : if this be not possible, and the gold is nearly
fine, the number 19 may be adopted. It is better, how-
ever, to estimate experimentally the specific gravity of
both quartz and gold.
Native Gold and Aurides of Silver (Native) are found
in variously contorted and branched filaments, in scales,
in plates, in small irregular masses, in the crevices or on
the surface of common ferruginous and other quartz.
The author has received two specimens of gold one
from Wales, and the other from the Britannia Mine, Devon
and found both to be absolutely fine gold.
Artificial Alloys of Gold. The only one of these alloys
which will be specially noticed here is the standard gold
of this realm. It is composed of 22 parts of fine gold and
2 parts of alloy, constituting 22-carat or standard gold.
ASSAY OF GOLD ORES.*
The assay of gold ores embraces the following steps :
1. Preparation of the sample. 2. Collection of the
gold in a button of metallic lead. 3. Cupellation of the
lead button, by which the lead is oxidised and absorbed by
* For some portions of the following method of assaying gold ores, the
Editor is indebted to Mr. T. M. Blossom, who has had great experience in the
laboratory of the School of Mines, Columbia College, U.S.A. (See ' Chemical
News,' Nos. 607, 609, 628, 635, 636.)
742 THE ASSAY OF GOLD.
the cupel, leaving behind a bead of the gold and silver.
4. Weighing the bead. 5. Inquartation, parting, and
cupellation of the gold residue. 6. Weighing the gold
bead.
1. Preparation of the Sample. It is essential, in the
first place, to obtain a fair average sample of the ore,
otherwise the results of the assay may be commercially
worthless. Selection must be left to the judgment of the
assayer. The sample must be dried, if necessary, care
being taken not to roast it. It must then be pounded in
an iron mortar and passed through a sieve of eighty meshes
to, the linear inch. If any native metal, in the form of
scales or filaments, remain upon the sieve, take the weights,
separately, of what has passed through and of what is left
upon the sieve. The latter must be assayed according to
* Assay of Alloys,' and the result referred to the whole
amount of ore. It is essential that the whole of the sample,
except the malleable portion, be passed through the sieve.
Mix thoroughly the sifted ore.
2. The Collection of the Gold and Silver in a button of"
metallic lead is effected in a crucible, or in a scorifier,
whence two methods of assay :
~\a) Crucible assay. (b) Scorification assay.
The former is applicable to all ores ; the latter is
lima ted /practically, by the small size of scorifiers, to the
richer ores.
-(a) .Crucible Assay. An ore of gold and silver is com-
posed of precious metal, gangue, and oxides, sulphides,
&c., of foreign metals. To collect the precious metals in
a button of lead, the ore is mixed with litharge, suitable
fluxes, an oxidising or a reducing agent, and fused in the
crucible. Litharge is reduced to metallic lead ; the latter
seizes upon the precious metals and collects in a button at
the bottom of the crucible, while the foreign materials
form, with the fluxes, a fusible slag above the lead
button. The crucible is broken when cold, and the malle-
able button detached from the slag by hammering on an
anvil.
GOLD OKES. 743
The Charge. The weight of ore taken for an assay
depends upon its supposed richness or poverty, since it is
required to obtain finally a bead of precious metal of con-
venient size for weighing, and, at the same time, neither
too large for cupellation (vide cupellation) nor so small
as seriously to affect the calculated results, through losses
sustained 'in the assay. As a rule, it is usual to take
500, 1000, or 2,000 grains for gold ores.
The ores require the following reagents : Litharge,
sodium carbonate, and one of the reducing agents, argol
and charcoal, or an oxidising agent, as nitre, with invariably .
a cover of salt one-quarter to half an inch in depth.
Borax, silica, and other reagents are very useful at times,
but no general rule can be given for their employment.
This matter must be left to the judgment of the assay er,
guided by the known properties of the reagents and by
the composition of the ore. It is well, in this connection,
to bear in mind the principle that for basic impurities
an acid flux is needed, and for an acid gangue a basic
flux. As a rule, employ a weight of litharge twice that
of the ore, and of sodium carbonate, the same as of ore.
These proportions also may be modified to advantage,
according to the composition of the ore. The propor-
tion of nitre, or of reducing agent, depends upon the re-
ducing power of the ore, hence it is variable in every case.
These reagents are added to control the size of the lead
button.
Size of the Lead Button. There are two limits to the
size of the button : (1) it must be large enough, or, in
other words, enough litharge must be reduced throughout
the mass to collect all the precious metal, and, at the same
time, (2) there should not be a useless excess of lead,
which would occasion loss of silver in the subsequent
cupellation.
These two conditions cannot always be fulfilled, but
in this case the latter must be sacrificed. It has been
found that a button of 225 grains is the best size
for a weight of ore from 500 to 2,000 grains. This is
also a convenient size for a cupellation. A button
744 THE ASSAY OF GOLD.
that is too large for cupellation can, however, always be
reduced in size by scorifying.
These requirements necessitate, in many instances, a
preliminary assay of the ore to estimate its reducing
power. The reducing power of an ore is due to the
presence of sulphur, arsenic, antimony, zinc, &c., but
generally to sulphur contained in pyrites, &c.
Preliminary Assay of Ore.
Charge. Ore . . . . .20 grains.
Litharge . . ; .' ' .250
Sodium Carbonate . .100
Salt . . . . . Cover.
A duplicate assay need not be made.
Warm the crucible before placing it in the fire, which
should be perfectly bright, and should be urged to effect
complete fusion in the shortest possible time. When the
contents of the crucible are in quiet fusion, it must be
withdrawn, to prevent further reducing action of the
furnace gases. Tap the crucible gently, and when cold
break it open.
Three cases may arise here. Twenty grains of ore may
yield
1. No lead, or less than thirty grains.
2. Thirty grains lead.
3. More than thirty grains.
The reducing power is stated thus : 20 grains ore
=# grains lead. Let us suppose that we shall take for the
regular assay 150 grains of ore, and that the reducing
power is found to be 20 grains ore=15 grains lead:
150 grains (about) ore will in this case reduce 112-5
grains lead, and as the required button is 225 grains, we
must add enough argol, or charcoal, to reduce 112-5 grains
in addition ; taking argol as 5*6, we shall require 112'5-^5'6
grains, or 20 grains, or charcoal 112-5-*-28 = 4 grains.
If the reducing power correspond to the third case, a
similar calculation will indicate how much nitre is needed
GOLD ORES. 745
to oxidise part of the sulphur, arsenic, or other reducing
agent, and thus prevent the reduction of a button larger
than 225 grains. One part of lead requires 0*25 part of
nitre to oxidise it. In the second case, 150 grains of
ore would reduce a button of 225 grains, and neither argol
nor nitre would be required.
The character of slag obtained in the preliminary assay
may also suggest some modification of the regular charge.
If it be earthy, for instance, we would add borax-glass or
silica. Experience will often enable the assayer to judge
of the reducing power with sufficient exactness, without an
extra assay. Cases will arise, however, in which he must
make a preliminary assay, or roast the ore, and it is
always best to roast when there is a large amount of
sulphur in the ore. He will then have an ore of no
reducing power. (Case I.)
Ores to be Roasted. Ores containing a large amount of
sulphur, or arsenic, antimony, or zinc, should always be
roasted. In the former case, if the ore be not roasted,
there will be danger of the formation of oxysulphides,
which are very fusible, but are not decomposed even at a
white heat, and enter the slag carrying silver with them.
A large quantity of nitre also subjects the contents of the
crucible to the liability of boiling over : even should this
mishap not occur, the great evolution of nitrous and sul-
phurous vapours puffs up the mass throughout the cru-
cible, and the globules of lead afterwards reduced may be
left adhering to the sides, not being washed down by the
retreating charge. Arsenic and antimony produce arseni-
ates and antimoniates, which carry silver into the slag.
Zinc increases the loss of silver by volatilisation and also
in the slag.
Roasting the Ore. The ore may be roasted conveniently
in a cast-iron pan over the crucible furnace. There ought
to be a hood over the furnace, to carry off the fumes of
sulphurous, arsenious, and other acids. The pan should
be covered with a coating of red ochre, or, still better,
of chalk. The former is put on wet with a brush. An
excellent even coating of the latter may v be obtained by
746 THE ASSAY OF GOLD.
making a chalk paste, of proper consistence, in the pan,
and turning with the hand so as to spread the paste while
the pan is being held over the fire to dry. The coating
prevents a loss of ore and injury to the pan through the
sulphides attacking the iron.
The weighed sample must be spread over the pan and
stirred with a bent wire until all danger of fusion is past.
The pan must be heated gradually at first, not above a
dull red heat for some time, and may be brought, finally,
to a full red, or higher heat. Too high a heat at the
outset might cause the fusion of sulphides and the forma-
tion of matter troublesome to roast. A rapid disengage-
ment of arsenic, antimony, or zinc would cause, also, a
mechanical loss of silver, by carrying it off in their vapours.
Should fusion take place, it is better to weigh out a fresh
portion of ore and roast again with more care. If this be
not done the fused portions must be pulverised in a mortar
and re-roasted. Generally, in roasting sulphides, the
operation may be considered finished when, after keeping
the pan at a full red heat for some time, no more fumes of
sulphurous acid can be perceived. As there is danger,
however, of the formation of sulphates, especially if copper
pyrites be present, it is best always to mix some ammo-
nium carbonate with the ore, after the fumes have ceased,
and to cover the pan. The ore is thus confined in an
atmosphere of ammonium carbonate, which decomposes
the sulphates with formation of volatile ammonium sul-
phate.
Arsenic and antimony require the addition of fine
charcoal, to reduce any arseniates and antimoniates that
may have been formed during the roasting. Care must
be taken to burn out all the charcoal.
If the ore contain a very fusible sulphide, as antimony-
glance or galena, it may be mixed with some fine sand
previous to roasting
The roasting of ores may also be done in the muffle in
an earthen saucer.
GOLD ORES.
747
Fusion.
The charge prepared according to the foregoing direc-
tions is thoroughly mixed and placed in a crucible, which
it should not more than two-thirds fill. A hot fire should be
employed, and the crucible removed when complete fusion
has taken place. This ought to require from twenty to
twenty-five minutes. The crucible is tapped, as usual, and
broken when cold.
(b) Scarification Assay. The reagents necessary for a
scorification assay are test-lead and borax-glass. The ore
is mixed with these in suitable proportions, the mixture
put in a scorifier and fused in a muffle. The operation
affords an alloy of lead with the precious metals, and a
slag composed of litharge with the impurities and gangue
of the ore. The assay might be made with lead only,
but it is advantageous to add some borax. There will be
required of lead only enough to render the slag liquid and
to furnish lead for the button. The proportions of both
lead and borax will vary, and should be greater in pro-
portion as the gangue and metallic oxides are more diffi-
cult of fusion. The following table exhibits the propor-
tions found by experience to be best adapted to the
different gangues. The proportions are referred to one
part of ore.
Parts
Test-Lead
Parts
Borax
8
8
5-6
16
16
12-16
10-15
10-15
0-25-1-00
0-15
0-10-0-50
0-10-1-00
0-10-0-15
0-10-0-20
0-10-0-20
Character of Gangue
Quartz ose
Basic (Fe 2 O 3 ,AL0 3 CaO), &c
Galena .
Arsenical
Antimonial
Fahlerz
Iron pyrites
Blende .
No preliminary roa.sting is required. As in the cru-
cible assay, the weight of ore taken depends very much
upon its richness, but is generally a third, sixth, or tenth
of an assay ton.* If one scorifier will not contain the
charge, it is best to weigh equal fractional parts for the
* An assay ton is about 450 grains.
748 THE ASSAY OF GOLD.
number required, rather than to weigh the whole charge
and roughly divide it between the scorifiers. In case of
an accident to one of the scorifiers, the known loss can
easily be restored, if the former course be followed. The
exact parts that lead and borax perform in scorification
can be understood best from a description of the operation
in detail.
Three distinct periods maybe noted in the working :
1. Eoasting ; 2. Fusion ; 3. Scorification. A strong-
heat is maintained, at first, in order to melt the lead.
'This is effected by closing the muffle and regulating the
draught. As soon as the lead is fused the muffle is
opened, and the ore is seen floating upon the surface of
the lead.
1. The roasting now commences, and is continued at
a moderate heat until no more fumes are seen, and the
ore has disappeared.
2. The heat should now be raised in order to fuse
thoroughly all the material in the scorifier. When the
fusion is complete, clear white fumes of lead may be seen
arising from the scorifier, there is an intermittent play of
colours across the bright surface of the lead, and the slag
produced encircles the metallic bath like a ring. The
borax plays an important part just here, by giving liquidity
to the slag, thus permitting it to be thrown to the side as
fast as formed, exposing a clear surface of lead -for oxida-
tion. If borax be not added and the ore contain a diffi-
cultly fusible gangue, the scoriae will float in detached
masses over the lead, impeding the oxidation, until suffi-
cient litharge has been formed to give it liquidity.
3. When the fusion is complete, the heat may be
lowered and the third period of scorification continued
until the ring of slag, ' which is continually growing
smaller, closes over the residue of metallic lead. The heat
.should again be raised, to liquefy the slag and allow the
metallic lead to settle, after which the scorifier is removed
from the furnace. If the button is wanted immediately
for cupellation, the contents of the scorifier may be poured
into an iron or copper mould coated with reel ochre. It
CUPELLATIOiV. 749
is thus soon cooled. Otherwise allow the contents to
solidify in the scorifier, and break this for the button
when cold. Hammer the button as usual.
The whole assay occupies about thirty-five minutes.
In making the charge, it is customary to mix the ore
with a part of the test-lead and with borax, and to cover
this in the scorifier with the rest of the lead. This will
prevent loss of ore prior to fusion. Too much borax
should not be added at first. If a large amount is needed,
it is better to mix only a portion with the ore ten grains,
for instance and to introduce the rest wrapped up in
paper, as needed during the operation. Some of it may
be reserved for the final heating after the lead is slagged
over. If too much were added at first, the lead would be
slagged over before the necessary reactions had taken place.
The Lead Button. The lead button submitted to cupel-
lation must be malleable and of the proper size for the
cupel. A good cupel will absorb its own weight of
litharge, but it is better to use a cupel one-quarter to a
third as heavy again as the lead-bottom. The cupels in
ordinary use weigh about 250 grains, hence a button of
180 to 230 grains is the proper size for them. If the
button be too large it may be reduced in size by scorifica-
tion. In case of doubt it is better to scorify, since there is
less loss in this operation than in cupellation. A brittle
button may be due to the presence of sulphur, arsenic,
antimony, zinc, or litharge. In either case it must be
scorified before cupellation, and with test-lead if necessary.
If the button contain copper it must be scorified until no
more copper can be seen on hammering out the button,
If nickel be present the button cannot be cupelled. This,
however, will occur but rarely.
3. Cupellation. The lead button obtained by the fore-
going operations is next cupelled. This operation is
similar in many respects to scorification, but differs from
it in that the scoriae formed are absorbed entirely by the
cupel, leaving a pure bead of the precious metals. This
property of the absorption of scorise is an indispensable
condition in cupellation, so that, unlike scorification, it is
750 THE ASSAY OF GOLD.
limited to a few substances capable of being absorbed by
the cupel. The oxides of lead and bismuth alone can be
absorbed in a state of purity, but they can carry along
with them certain proportions of other substances which,
by themselves, would form infusible scoria. It is thus
that we are enabled to get rid of small amounts of copper,
iron, arsenic, &c., remaining in the lead button, Bismuth is
seldom, if ever, used for this purpose. The proportion of
lead required to carry off the different impurities varies
according to circumstances. In the case of lead buttons
from an assay, there is always an excess of lead, if they
have been properly purified. Other cases will be treated
under the head of Assay of Alloys, where the necessary
tables will be given.
The operation of cupelling a lead button is conducted
t as follows :
The cupels must be carefully dried before use, and
must be free from cracks, which would cause a loss of
precious metal. The bottom of the muffle should be
covered with sand to prevent injury to it in case of the up-
setting of a cupel. A cupel having been selected, it should
be wiped carefully with the finger, and all extraneous
matter blown out. It may then be placed in the muffle.
Before the introduction of the lead button, the muffle
should have attained a reddish-white heat, and the cupel
should be of the same temperature. This being attained,
the button is placed in the cupel with a pair of forceps,
gently, so as not to injure it. The muffle should now be
closed, either by a door or by a piece of lighted charcoal,
to bring the fused button to the same temperature. This
done, the muffle is opened and air allowed to enter ; the
button, which at first appears bright and uncovered, is
soon covered with a film of oxide moving in luminous
patches over the surface, and being continually thrown
toward the edge, where it is absorbed by the cupel. White
fumes of lead arise, and the button is surrounded by a
ring of the absorbed litharge, which continually widens
until it reaches the edge of the cupel. The button thus
gradually diminishes in size by oxidation and absorption,
CUPELLATIOtf. 751
and becomes more convex ; the luminous patches become
larger and move more quickly ; and when, finally, the last
of the lead is absorbed, the button appears to revolve
rapidly on its axis, becomes very brilliant, and is suffused
with all the tints of the rainbow ; the movement is sud-
denly checked, the button becomes dull for a few instants,
and then presents the appearance of the pure precious
metals. The latter part of the operation is called the
4 brightening ' of the button. Should the bead be large,
and composed, in great part at least, of silver, it must be
removed slowly and gradually from the furnace, to pre-
vent loss by ' spitting,' which might happen if the cupel
were removed at once from the muffle. When a cupel is
withdrawn directly after brightening, the button is liable
to be covered by mammillated and crystalline protuber-
ances, and is then said to have ' vegetated ' or ' spit.'
Portions of the metal are sometimes thrown off and lost.
Whatever the cause may be, this does not occur if the
button be withdrawn gradually, so as to permit a slow and
gradual cooling. In case the bead be very large, say
2 to 5 grains, it is well to cover it with a hot cupel,
which will retard the cooling and prevent the loss of small
particles that may be thrown off despite all care. If the
bead is not larger than the head of an ordinary pin, the
danger of vegetation is slight, and no great precautions
need be taken in its removal.
It is well to raise the heat of the furnace just before
the brightening, or to push the cupel into a hotter part
of the muffle, in order to aid the brightening and get rid
of the last traces of lead, which are somewhat difficult to
remove. The button should also, for the latter purpose,
be heated strongly a few seconds after brightening.
Silver is sensibly volatile at a high heat ; the loss of
gold is slight, and may be disregarded. In the assay of
ores this loss need not be considered, but in bullion assay
a correction is necessary, for which a table will be given
under Assays of Alloys. The loss of silver increases with
the temperature, but we must avoid, in cupellation, the
two extremes of a high heat and quick work and a low
752 CUPELLATION.
heat with prolonged work. Of the two extremes the latter
is worse. The following are indices of favourable work-
ing : The muffle is reddish-white, the cupel is red, the
fused metal very luminous and clear, the lead fumes rise
slowly to the top of the muffle, and the litharge is com-
pletely absorbed by the cupel.
The heat is too great when the cupels are whitish, when
the fused metal is seen with difficulty, and the scarcely
visible fumes rise rapidly in the muffle.
The heat is too low when the fumes are thick and fall
in the muffle, and when the un absorbed litharge is seen
forming lumps and scales about the button.
The degree of heat should bear some relation to the
richness of the alloy, and may be greater according as
the lead is poorer in silver. By bearing this in mind the
assayer may often hasten the operation without detriment
to the assay. The draught of air through the muffle must
also be regulated. Too strong a current cools the cupel
and oxidises the lead faster than it can be absorbed, thus
endangering the assay. Too slow a current prolongs the
operation and increases the loss by volatilisation, In
ordinary work this matter, however, occasions very little
trouble.
It happens sometimes that the material in the cupel
becomes solidified in the midst of an operation, stopping
all further action. This disaster is called the ' freezing ' of
the button, and is occasioned by the following conditions :
a production of litharge more rapidly than it can be ab-
sorbed by the cupel, and infusible scoriae, due to a cold
furnace, or to an excess of foreign oxides. In either case
the scoriae gradually extend over the surface of the button
until it is entirely covered, when further movement ceases.
This disaster may sometimes be prevented or remedied by
raising the heat of the muffle ; if this fails, or if the acci-
dent be due to foreign oxides, an addition of pure lead
must be made to the assay ; in either case, the results are
unreliable.
An assay that has passed well, furnishes a bead well-
rounded and clear, crystalline below, and readily to be
WEIGHT OP MINUTE SPHERES OF GOLD. 753
detached from the cupel. If the bead contain lead, it is
brilliant below, and does not adhere at all to the cupel.
If the bead exhibit rootlets passing down into the sub-
stance of the cupel, the results are inaccurate.
ESTIMATING THE WEIGHT OF MINUTE SPHERES OF GOLD.
M. G. A. Gozdorf has given, in the ' Chemical News '
for 1886, an ingenious plan for the production and measure-
ment of gold and other minute metallic spheres to esti-
mate their weight.
In making assays for gold where the amount of gold
is very small a little silver is required in which the gold
may be collected. As nearly all commercial litharge con-
tains silver it is rarely necessary to add any separately for
this purpose.
Having obtained a prill in which the amount of gold is
a third or less than the silver, the prill is boiled in dilute
nitric acid in a porcelain capsule to dissolve the silver, and
where the amount of gold is more than 1 dwt. to the ton
a second boiling in strong nitric acid should be given. If
care is taken in using dilute acid at first and boiling gently
the gold will be left in one piece of a nearly black colour.
The acid is now decanted off and the gold washed two or
three times in distilled water. The gold may be now placed
on an aluminium or other polished metal plate by invert-
ing the capsule and leading the last drop of water and the
gold with a glass rod on to the plate, the water is drawn
off by a piece of filter-paper, and the plate gently heated
till dry.
Having thus obtained the gold in a pure state a bead is
made of boracic acid on a platinum wire loop and pressed
on the gold while still red-hot ; the gold adheres without
difficulty, and by heating the bead before the blowpipe the
gold is obtained as an almost perfect sphere.
Should the resulting sphere of gold be very minute it
is better to measure it under the microscope while in the
bead, but if large enough to be seen with the naked eye
it can be measured more accurately after dissolving the
3c
754 THE ASSAY OF GOLD.
boracic acid bead in a watch-glass with hot water and
placing the sphere of gold on a glass slide.
The plan of measuring minute prills of silver and gold
to estimate their weight was first introduced by Harkort,
who used an ivory scale engraved with two fine lines meet-
ing at an acute angle and divided into fifty equal parts
(see fig. 141, p. 724).
According to the fifth edition of Plattner's 'Probir-
kunst,' p. 520, Goldschmidt estimates the weight of
silver and gold prills by measurement with the micro-
scope.
Harkort and Plattner, in making scales for the esti-
mation of the weight of gold and silver prills, weighed
prills corresponding to the larger divisions of the scale,
and from their weight calculated the weight for the smaller
divisions. These prills were taken direct from the cupel,
and at the point of contact are flattened ; but as the
amount of flattening is not always the same, and hardly
varies in extent with the size of the prill, and as the con-
verging lines on the scale cannot be very sharply defined,
this method is not capable of the same accuracy as where
the almost perfect spheres are measured with a microscope.
No other flux seems to possess advantages equal to
those of boracic acid for obtaining a sphere of gold. Borax
and other fluxes are so fluid when hot that the gold is very
liable to alloy with the platinum wire : this rarely occurs
with boracic acid, on account of its great viscosity, even
when white-hot. Boracic acid is also easily soluble in
water, so that the gold spheres can be separated without
loss of time.
The following rules and figures may be useful to any
one wishing to adopt the system here described.
1. The weight of a sphere increases as the cube of the
diameter.
2. The weight of a sphere of any substance of which
the specific gravity is known is obtained by mul-
tiplying the weight of a unit sphere of water by
the specific gravity of the substance and the cube
of the diameter.
WEIGHT OP MINUTE SPHERES OF GOLD. 755
Constants for Use with Gramme Weights.
1. Weight of a sphere of water 0-01 m.m. in diameter
-0-000,000,000,523,6 of a grm.
2. Weight of a sphere of gold 0-01 m.m. in diameter=
0-000,000,010,210,2 of a grm.
3. Weight of a sphere of gold 0'0# m.m. in diameter,
a* x 0-000,000,010,210,2 of a grm.
4. If 20 grms. of ore are taken for assay the number
of grains of gold per ton is found by x* x 0*008004,
in which ^?=the diameter of the sphere of gold in
hundredth s of a millimetre.
Constants for Use with Grain Weights.
1. Weight of a sphere of water 0*001 inch in diameter,
0-000,000,132,4 of a grain.
2. Weight of a sphere of gold 0*001 inch in diameter,
0-000,002,582 of a grain.
3. Weight of a sphere of gold 0*00# inch in diameter,
x* x 0-000,002,582 of a grain.
4. 200 grains of ore being taken for assay, the num-
ber of grains of gold per ton is found by
x 3 x 0-2045288, in which #==the diameter of the
sphere of gold in thousandths of an inch. By
taking 978 grains for assay # 3 = grains of gold
per ton.
To test the accuracy of the above figures a compara-
tively speaking large sphere of gold from an assay was
measured and found to be 0*593 m.m. or -^ of a m.m.
in diameter; 59*3 3 x 0*000,000,010,210,2=0-002,129 of a
grm. When weighed on a very delicate balance it was
found to weigh 0*0021 grm., and as this balance does not
indicate beyond the fourth decimal the results may be
considered identical. This sphere indicated gold in the
sample tried at the rate of 3 oz. 9 dwts. 13 grs. per
ton.
The smallest sphere of gold yet measured was 0*024 m.m.
3 o 2
756 THE ASSAY OF GOLD.
in diameter, and by applying the above rule the weight
would be 2-4 3 x 0-000,000,010,210,2 x 15-43235 (to convert
grammes to grains) = 0- 000, 002,178, or a trifle over two
millionths of a grain.
Spheres of silver may be obtained and measured in a
similar manner. The boracic acid acts slightly on the
silver, but the quantity dissolved is inappreciable, as the
action is not prolonged. The specific gravity of silver
being 10-53, the weight of l-100th m.m. would be
0-000,000,000,523,6x10-53 = 0-000,000,005,513,508 of a
grm. In a test assay made with silver the sphere measured
0-57=f$j- of a m.m., from which the weight deduced
would be 0-001,020,96 of a grm., the balance showing the
weight as 0*0010 of a grm.
Copper, lead, and other metals cannot be melted in
boracic acid on platinum wire without dissolving to a
perceptible amount, but may with care be melted in sodic
carbonate, and, by dissolving the latter in hot water, the
sphere of copper, &c., obtained and measured.
GENERAL OBSERVATIONS ON THE ASSAY OF GOLD ORES.
Gold and Copper, Proportion of Lead. The alloys of
gold and copper are cupelled like the alloys of gold and
silver ; but as copper has a very great affinity for gold, it
is necessary to use a larger proportion of lead to insure
its oxidation when combined with gold than when united
with silver. This proportion varies according to the
standard and the temperature. It is admitted that for the
same standard there must, under similar circumstances, be
twice as much lead used in the cupellation of gold as for
that of silver. Thus, 14 parts, at least, ought to be em-
ployed in common furnaces for an assay of gold coin
which contains 0*1 of copper. There is no inconvenience
in employing a little more, as it does not increase the loss
of gold. However great the proportion of lead may be
that is added to the cupreous gold for the purpose of
cupellation, the button retains always a very small quantity
of copper, which a fresh cupellation does not free it from,
EXAMINATION ON THE TOUCHSTONE. 767
.and which occasions what is termed the surcharge. This
.surcharge, being very slight, can be neglected in assays of
minerals ; but it is necessary to take notice of it in the
assay of alloys. But it is known that the presence of
silver much facilitates the separation of copper from gold,
and it is rare that an alloy of cupreous gold does not con-
tain a little silver, which must be separated : and when that
is not the case, a small quantity of that metal can be
introduced into the alloy, so as to be in about the propor-
tion of 3 parts to 1 of gold. When an assay is to be made
of an alloy of gold and copper, a sufficient quantity of
silver is to be added to fulfil this condition according to
the presumed standard, which is estimated approxi-
matively by a preliminary assay, and then cupelled with
lead.
Examination on the Touchstone. This method is based
upon the fact that the richer an alloy is in gold the more
clearly does a streak drawn with it on a black ground
present a pure gold-yellow colour, and the less is it
attacked by pure nitric acid or by a test acid. This test
-acid consists of ninety-eight parts pure nitric acid of
1*34 sp. gr. (37 Beaume), two parts pure hydrochloric
acid of 1-173 sp. gr. (21 B.), and twenty-five parts
distilled water. To judge of the richness of the alloy to
be examined, its streak is compared with marks drawn
with alloys (the touch-needles) whose richness is accurately
known. In order to get correctly the streak of the alloy
to be tested, the surface of the metal must first be some-
what filed away, since this may be impure, or, as with
coins and jewellery, it may have been made somewhat
richer by boiling with acid, and the so-called colouring of
the goldsmith, and a clean fracture is rarely to be obtained.
.Five series of prepared touch-needles are required. The
first series consists of copper and gold, and is called the
.red series, and the proportion of gold increases by half-
carats in the successive needles. The second series, the
white series, contains needles of gold and silver, in which
the proportion of gold likewise increases by half-carats.
The third series, a mixed one, contains needles in which the
758 THE ASSAY OF GOLD.
quantities of silver and copper are equal, and the propor-
tion of gold also increases by half-carats. The fourth con-
sists also of needles for a mixed series, in which the silver
is to the copper as 2 : 1, and the gold increases by half-
carats ; and the fifth is also formed of needles for a mixed
series, in which the quantity of silver is to that of the
copper as 1 : 2. Moreover, in mints and stamping bureaux,
alloys are used which correspond precisely to the legal
standards. The testing upon the touchstone begins by
determining to which series the alloy to be examined be-
longs. Then those touch-needles are rubbed against the
stone whose marks most nearly approximate in colour to
that of the alloy. The marks must form a thin continuous
layer. A drop of pure nitric acid is now placed upon
them with a glass rod, and its comparative effect observed.
The acid is allowed to work a short time, and then wiped
off, in order to see whether the streak appears unchanged,
or whether it has more or less disappeared. The test acid
above is also used. This is so composed that it does not
work at all upon an alloy containing eighteen carats and
more of gold, and with such an alloy the streak, after
using the acid, will not be wiped off with a fine linen rag,
provided that stone and acid had a temperature of 10 to
12 C. Pure nitric acid produces almost no effect upon an
alloy of fifteen or sixteen carats fine, and over. The test-
ing on the touchstone can indeed make no pretension to
accuracy, especially where the amount of gold is small,
but it yields sufficiently useful results for a preliminary
test. It requires, however, a sharp and very practised eye.
Moreover, the preparation of the touch-needles is weari-
some, as the required proportion is not always quickly
reached, nor are good malleable alloys always obtained.
The touchstone, therefore, is in general only used where
frequent gold assays are to be made of alloys varying
in richness, or where (as frequently with gold plate) aiL
examination on the touchstone will suffice.
CUPELLATION OF GOLD AND COPPER.
769
TABLE FOR PROPORTION OF LEAD TO BE EMPLOYED IN THE
CUPELLATION OF GOLD AND COPPER.
Gold in alloy
1000 thousandths
Lead required
Eatio of lead in
the assay to the
copper, <fcc.
900
100 000 1
800
16
80 000 * 1
700
22
70 qqq . -i
600
24
fift 000 1
500
26
no 00 1
400 v
300
200 I
34
/ 56,666 : 1
48,571 : 1
{ 4.0 K()f) . 1
100
50]
(37,377 : 1
Kandelhardt gives the ratio in the following table :
Gold in 1000 parts
1000 fine gold
980 920
920 875
875750
750 600
' 600 350
350
Quantity of lead required
8 times the weight of the alloy
12
16
20
24
28
32
Gold, Silver, Platinum, and Copper. The presence of
platinum in an alloy renders- the separation of the oxi-
disable metals, more especially copper, very difficult by
cupellation. It appears, indeed, that it would be almost
impossible to arrive at it, if the alloy of copper contained
nothing but gold and platinum. It is necessary that silver
be present at the same time. When this metal is absent,
it is requisite to add a quantity of it, which ought to be
equivalent to double the weight of the gold and platinum
united, and cupel at the strongest heat which can be ob-
tained in a good muffle with a suitable proportion of lead.
This proportion varies much according to the composition
of the alloy, and the temperature at which the operation
is carried on.
Experience has shown that the copper can be more
completely separated and less silver lost by cupelling at a
high temperature, with the least possible quantity of lead,
than by employing more lead, and working at a lower tem-
perature. M. Chaudet has made several assays, in order
760 THE ASSAY OF GOLD.
to estimate the proportion of lead required for the cupel-
lation of the three following alloys :
1. 2. 3.
Gold .... .0-100 0-020 0-005
Platinum 0-100 0-200 0-300
Silver 0-250 0-580 0-595
Copper 0-550 0-200 0-100
and has found, for the first, that by employing 20 parts of
lead the separation is very nearly complete ; but that at a
high temperature there is a loss of silver, and in order to
render the assay correct it must be cupelled at the latter
temperature, with only 14 of lead ; for the second, 8 of
lead, at a high temperature ; and for the third, 30 parts
of lead are necessary, at the same high temperature of the
muffle ; but it is almost impossible to separate all the
copper, and no advantage can be obtained by increasing
the quantity of lead. When almost the last traces of the
copper are separated, the button must be cupelled afresh,
with a small quantity of lead ; but a small quantity of
silver is nearly always lost. In all cases, in order that no
lead shall remain, it is necessary to leave the assay button
some few minutes in the muffle after cupellation is finished.
The alloys of gold and silver which contain platinum
show, either by cupellation or parting, certain characters
which prove the presence of the metal. If the assay be
not heated very strongly, it does not pass, and the button
becomes flat : this effect becomes very sensible when the
platinum is to the gold in the proportion of 2 to 100.
Under the same circumstances, the nitric acid solution
proceeding from the parting is coloured straw-yellow.
At the moment an assay of an alloy containing platinum
terminates, the motion is slower, and the coloured bands
are less numerous, more obscure, and remain a much
longer time than when there is no platinum ; the button
does not uncover, and the surface does not become as
brilliant as that of an alloy of gold or silver, but it remains
dull and tarnished. When the assay is well made, it is to
be remarked that the edges of the button are thicker and
more rounded than in ordinary assays, and it is of a dull
white, approaching a little to yellow; and lastly, its sur-
GOLD AND SILVER PARTING. 761
face is wholly or in part crystalline. These effects are
sensible even when the gold does not contain more than
01 of platinum. When the alloy contains more than 10
parts of platinum to 90 of gold, the annealed cornet pro-
duced in the parting process is of a pale yellow, or tar-
nished silver colour.
Gold alloyed with Silver. The separation of gold from
silver is termed parting. Parting is not only used to sepa-
rate silver from gold, but for the separation of other metals,
such as copper, when cupellation does not separate it
entirely. Parting by the wet process is carried on by the
means of nitric acid, aqua regia, or sulphuric acid.
When an alloy of gold and silver has been reduced by
a flatting mill to very thin plates, it is sufficient that it
contains 2^ of silver to 1 of gold in order that the parting
may be effected completely by nitric acid, and it takes place
much less easily when the silver in the alloy is in larger
proportion ; but when this proportion exceeds 3 parts of
silver for 1 of gold, then the latter is obtained in leaves so
fine that there is risk incurred of losing some in the sub-
sequent manipulation, and even by the act of boiling the
acid liquid.
We must always, therefore, when a very exact assay is
required, contrive that the alloy shall contain a little less
than 3 parts of silver to 1 of gold, a proportion which
long experience has demonstrated to be the best. If the
alloy contain less than 2-| of silver to 1 of gold, the silver
does not wholly dissolve, because there is a part of it so
enveloped in the gold that the strongest acid does not act
on it.*
Inquartation. The operation by which the alloy is
brought to this standard is termed quartation, or inquar-
tation. It consists in fusing the alloy in a cupel, with 2
parts of lead and the quantity of fine silver, or fine gold,
necessary to bring it to the desired composition. This
quantity is estimated according to the approximative esti-
mation of the standard of the alloy, which ought to be
* Pettenkoffer and others have shown that less than two parts of silver
will suffice, and be even advantageous.
762 THE ASSAY OF GOLD.
made either by means of a preliminary assay, as hereafter
described, or by means of the touchstone. If we do not
employ the whole of the alloy the assay will not be exact,
because the gold and silver are not always found dis-
tributed in an uniform manner ; at least every time it is
not poured into a cold ingot mould.
Operation. The cupelled and quartated button is
flattened on an anvil and annealed, in order to soften it.
It is laminated to give it a certain thickness, and is then
annealed afresh, and rolled into a cornet or spiral around
the quill of a pen. It is necessary that the alloy should
be reduced to a suitable thickness, on the one hand, in
order that the silver may be dissolved completely ; and, on
the other, that the plate of gold may remain whole after
the operation. The following is that which experience has
proved best. The quantity of matter operated upon, or
taken for the assay, should be about 12 grains, and the
alloys resulting from these 12 grains, and the silver em-
ployed in the inquartation, should be made into a plate of
from 18 to 20 lines * in length and 4 or 5 in breadth.
The cornet for assay is placed in a glass matrass,,
capable of containing about three ounces of water ; pure
nitric acid is added at different times, and heat applied.
When all the silver is dissolved, it is washed by decanta-
tion with water ; the matrass is reversed into a small
crucible, the cornet falls out and is dried. In this state
the cornet is very fragile, and of a dull red colour ; it is
annealed in a muffle, and heated gradually without fusion.
It becomes thereby much contracted, and acquires a
metallic lustre, and so much solidity that it can be weighed
without fear of breaking it. Its weight can be ascertained
in the assay balance.
There are many ways of employing nitric acid. For-
merly 2-|- ounces (thirty-five times the weight of the alloy)
of nitric acid (1*15 sp. gr.) was poured upon the inquar-
tated cornet, and boiled gently for fifteen or twenty
minutes, the liquid decanted and replaced by 1-J- of acid
* A line is the f an inch. ,
INQUARTATION. 763
(sp. gr. 1-24 or T26), twenty-four times the weight of the
alloy, boiling for twelve minutes, then decanting and wash-
ing, &c. Vauquelain advised, in his ' Manuel de 1'Essayeur,'
to pour on the quartated cornet the weight of the assay
being 7*7 grains 554 to 770 grains of nitric acid (1*16
sp. gr.), which ought to fill the matrass half or two-thirds,
and boil gently for twenty, or twenty-two minutes at most,
to decant and replace the liquid by 500 to 800 grains of
acid (1-26 sp. gr.), and to boil for eight or ten minutes. The
assay is to be acted on always twice, because if we employ
at once very strong acid, the action will be too brisk, and
the cornet might be broken or carried out of the matrass,
and, on the other side, the acid of 1*16 sp. gr. cannot dis-
solve the last portions of silver, which are very difficult to
separate from the gold.
Surcharge. It is remarked that by following this
method the cornet always retains a small quantity of silver,
so that fine gold submitted to quartation and parting
always weighs more after than before the operation. The
augmentation of weight which it undergoes is termed the
surcharge: this surcharge is commonly from 0*001 to
0-002. M. Chaudet has found means to avoid it. In order
to do so, pour on to the quartated cornet nitric acid of
1*16 sp. gr., and heat for three or four minutes only ;
replace this acid by acid of 1'26 sp. gr., and boil during
ten minutes ; decant and make a second boiling with acid
of 1'26 sp. gr., which boil for eight or ten minutes. The
assay requires but from twenty to twenty-three minutes,
and, according to M. Chaudet, gives perfectly pure gold.
Mr. W. F. Lowe describes in the 'Chemica] News ' a use-
ful piece of apparatus for boiling gold assays. This appara-
tus consists of (1) a porcelain basin (a) 8^- ins. in diameter,
having a capacity of fifty oz. ; (2) a porcelain cover (b)
perforated with thirty holes, each hole being numbered in
black enamel; and (3) of a number of glass tubes (c).
These tubes are made to slip loosely through the holes in
the cover, and in order that they may not come against
the bottom of the basin the glass is bulged out into a ring
near the centre of the tube, which rests upon the cover ;
764
THE ASSAY OF GOLD.
they have two small holes, one on each side of the bottom,
one in the centre of the bottom, and also one in the side
an inch above.
The method of employing the apparatus is the follow-
ing : Two basins containing a sufficient quantity respec-
tively of strong and weak nitric acid are heated over the
lamp, the weak acid basin being covered with the perfor-
ated cover carrying the boiling tubes, and the strong acid
FIG. 142.
(a) Section of basin, tubes, and cover, one-fourth size, (b) Cover,
one-fourth size, (c) Tube, half size.
'being covered with an ordinary dinner-plate. The assays,
which can be flattened and rolled while the acid is being
heated, are dipped into the tubes ; and in about five
minutes, if the acid is boiling, the whole thirty may be
lifted off by the cover, washed by being dipped into a
.basin of water, and then transferred to the basin of strong
.acid, where they are boiled for half an hour or more, after
INQUARTATION. 765
which they are all lifted off by the cover, and transferred
to a similar basin full of water. Each tube is taken out
and plunged over head in water, so as to fill the tube, and
the assay is transferred to the crucible in the same way as
from a flask.
Mr. Lowe says that he has had this apparatus at work
for more than twelve months, and has found it a great
saving of labour, besides requiring so little attention ; in
fact, it can be left for an. hour without the assays taking
any harm, and he considers it preferable to boiling in?
flasks, for all the assays are under exactly the same con-
ditions, and can be boiled in the acid for a much longer
time. Another recommendation is that it is of very
moderate cost. For boiling the basins he uses two of
Fletcher's radial burners, which are very suitable, as they
are little affected by the fumes.
A source of loss occurs in parting operations and refin-
ing on the large scale, from the solution of gold in nitric
acid, even when it is quite free from hydrochloric acid, in
consequence of the formation of nitrous acid. To ascer-
tain the amount of loss from this source in ordinary assay
operations, Mr. Makin took four specimens of pure gold
accurately weighed, added the usual proportions of fine-
silver and lead, and then cupelled them. The resulting
buttons were rolled, coiled, and parted with nitric acid,,
the cornets being boiled in two acids of different strengths
a different number of times. Calling the weighings before^
the operation 1000, the results were as follows :
1. Boiled in acid twice 999'6
2. three times 999-2
3. four 998-7
4. five ....... 997-9
The loss is thus seen to increase as the boilings are multi-
plied.
When silver is present in large quantity, Mr. Makin
believes that the solvent action of nitrous acid is restrained
by electrical action, the gold becoming the negative and
the silver the positive pole of a circuit ; but as the silver is
removed, the solution of the gold goes on more rapidly.
766 THE ASSAY OF GOLD.
The cause of the evolution of nitrous acid is evident as long
as there is any silver present, and it often results from the
use of charcoal to prevent ' bumping.' When charcoal
is thoroughly carbonised, it does not materially affect the
acid ; but if it contain woody matter, nitrous acid is sure to
be set free. Mr. Makin has given up the use of charcoal
on this account.
The commercial importance of this subject will be
admitted, when we remember the enormous value of the
metals dealt with in this country, and that the question of
.profit and loss in commercial transactions with them are
-almost entirely in the hands of the assayer. A knowledge
of these facts may also serve to account for some of the
discrepancies between assayers.
In the assay of auriferous ores, the button produced
by cupellation commonly contains silver. When the pro-
portion of this metal surpasses that of inquartation, the
button is flattened between two pieces of paper, and treated
by pure nitric acid. The gold remains under the form of
a yellowish-brown powder, which is weighed immediately,
or fused in the cupel enveloped in a sheet of lead.
When the quantity is extremely small and imponderable,
we can assure ourselves at least of its presence by treat-
ing the residue left by nitric acid with aqua regia : if
it contain gold, it dissolves and gives a yellowish liquid,
in which a drop of solution of chloride of tin or the crys-
tallised chloride forms a deposit of purple of Cassius of
a violet colour : this character proves the presence of
the smallest traces of gold. When the gold predomin-
ates in the! bufton, o it is necessary to re-fuse it with three
times or less its weight of silver, and recommence the
assay.
Aqua Regia. When gold is the largest portion of the
alloy, and when there are reasons for not adding silver,
the parting can be made by aqua regia. In this case, all
the gold is dissolved, and the silver converted into chloride ;
the chloride is washed, dried perfectly, and weighed. When
the gold is precipitated by ferrous sulphate, it is washed
with a little hydrochloric acid, and annealed strongly
SEPARATION OF GOLD AND SILVER. 767
before weighing, or even annealed so far as to fuse it, and
then cupelled with lead.
If an alloy, containing much silver, be treated by this
process, it sometimes happens that the excess of silver
chloride prevents the complete solution of the gold. In
this case it is necessary to reduce the alloy to an exces-
sively thin plate, to dissolve the chloride in ammonia, and
to treat afresh with aqua regia. This process can rarely
be made use of on the large scale, because the precipitation
of gold by ferrous sulphate is long and troublesome.
M. G. Eose fuses the alloy with lead, over a spirit-
lamp, in a porcelain crucible, acts on it with nitric acid,
which dissolves the silver and lead, precipitates the silver
by a solution of lead chloride ; lastly, the auriferous
residue is dissolved by aqua regia, and the gold precipitated
by ferrous chloride.
For the separation of gold and silver and their esti-
mation in alloys, H. von Jtiptner ('Zeitschrift ftir Analy-
tische Chemie,' 1879, 105) alloys the metal with 5 to 8
parts of zinc, and dissolves in nitric acid ; zinc, silver,
copper, &c., dissolve, whilst gold and the platinum metals
remain, and also tin as oxide, if present. The zinc alloy
is easily refined, and the metals in the crucible are best
covered with resin to prevent oxidation.
If it is known that neither tin nor platinum is present,
it is sufficient to decant, dry, and weigh in order to find the
weight of the gold.
If the platinum metals or tin are suspected the residue
is dissolved in aqua regia, the free chlorine is expelled by
boiling, the gold is reduced with ferrous ammonium sul-
phate of known strength, and the excess of ferrous oxide
titrated with potassium permanganate. From the quantity
of ferrous oxide consumed in the reduction of the gold
its proportion may be calculated.
All the silver is contained in the nitric acid solution of
the zinc alloy.
By another method for the separation of gold and silver
the metal is alloyed with 5 to 8 parts of zinc, for which
the heat of a Bunsen burner is sufficient. The alloy is
768
THE ASSAY OF GOLD.
dissolved in nitric acid, when gold (with platinum and tin
as stannic oxide, if present) remains imdissolved. To-
separate gold from platinum and tin they are dissolved in
aqua regia, the platinum metals are precipitated with am-
monia, the free chlorine is expelled, and the liquid is mixed
with excess of ammonio-ferrous sulphate, the excess of
which is estimated by titrating back with permanganate
(< Zeitschrift fur Anal. Chemie,' 18, 104).
Standard of the Alloys of Gold. The real standard
of the alloys of gold is expressed in fractions of unity, as in
the case of alloys of silver. We suppose 24 carats in unity,
and 32-32nds in the carat ; the unity contains then
768-32nds. After these data the following table has been
formed, which expresses the relation of 32nds and carats to-
decimal fractions of the unity :
32nds
i
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
Decimals
Carats
0-001302
1
0-002604
2
0-003906
3
0-005208
4
0-006510
5
0-007912
6
0-009115
7
0-010415
8
0-011718
9
0-013021
10
0-014323
11
0-015625
12
0-016927
13
0-018230
14
0-019531
15
0-020833
16
0-022135
17
0-023436
18
0-024740
19
0-026042
20
0-027343
21
0-028644
22
0-029948
23
0-031250
24
0-032552
0-033854
0-035156
0-036460
037760
039062
0-040364
0-041667
Decimals
0-041667
0-083334
0-125001
0-166667
0-208333
0-250000
0-291666
0-333333
0-374999
0-416667
0-458630
0-500000
0-541667
0-583333
0-624555
0-666667
0-707333
0-750000
0-791666
0-833333
0-874999
0-916666
0-958333
1-000000
ASSAY OP ALLOYS (ASSAY PKOPER).
Assays of Gold Coin and Bullion (Alloys of Gold and
Copper, or Gold, Silver, and Copper).
Preliminary Assay. As in the case of silver assaying
the quantity of lead to be employed is of importance, a
preliminary assay must be made when the standard of the
alloy to be examined is not approximatively known. It is
thus effected : To 2 grains of the alloy add 6 grains of
fine silver and 50 grains of pure lead. The lead must be
introduced into a hot cupel, and when fused, and its sur-
face fully uncovered, the alloy and silver may be added,
wrapped either in thin paper or a small quantity of lead
foil. The cupellation finished, and the cupel cold, the
button of gold and silver must be removed from the cupel
by aid of the pliers, and if necessary cleansed. Hammer it
to a thin plate on the anvil, place it in a small evaporating
basin, and treat it with half an ounce of nitric acid. (It
may be here mentioned, that the nitric acid employed in the
assay of gold must be chemically pure, and special care
must be taken that it contains no trace of chlorine.) The
evaporating basin is gently heated until all action ceases.
The brownish residue is repeatedly washed with hot water,
dried, ignited, and weighed ; and from its weight the
amount of lead and silver to be added in the actual assay
may be estimated. The presence of copper in the 'alloy
is indicated by the blackness of the cupel where it. is satu-
rated with oxide.
Assay Proper. In this case it will be supposed that
standard gold is the alloy operated on, and that prelimi-
nary assay has given about 91^ per cent, of gold. On
referring to the table (page 759), it will be found that
between 27 and 30 parts of lead are required for such per-
centage of gold, and that, according to the general observa-
tions on this class of assay, three times its weight (that is,
the weight of fine silver) will be required to so dilute the
gold that nitric acid can attack and dissolve out the whole
of the silver combined with it.
Place the weight representing 24 carats in the pan of
3D
770 THE ASSAY OF GOLD.
the balance, and exactly counterpoise it with the gold to be
assayed ; two portions should be thus weighed. Two por-
tions of fine silver must now be weighed ; 33 grains will be
required for each 24 carats of gold, as 22 carats, or 11
grains, of fine gold exist in the 24 carats, and three times
the quantity of silver is necessary. 300 grains of lead must
be placed in a hot cupel (two being thus prepared), and, as
in the preliminary assay, when the surface is fully uncovered,
the gold and silver are added, and the cupellation proceeded
with, taking all the precautions already fully pointed out
elsewhere.
The button so obtained is cleaned, hammered on the
anvil, then annealed and passed between the rollers of a
small flatting-mill ; being occasionally annealed, in order to
prevent the laminated button cracking at the edges. When
reduced to the desired degree of thinness it is again an-
nealed and rolled round a quill or glass rod into a spiral,
termed a cornet. This cornet is placed in a parting flask
with 1^ oz. of nitric acid, sp. gr. 1'16, very gently heated
to the boiling point, and at that maintained for ten minutes.
The acid is then to be poured off, and 2 oz. of nitric acid,
sp. gr. 1*26, added, and again boiled for ten minutes. This
second acid is also poured off,' and a third quantity of like
specific gravity added and boiled. The cornet is then well
washed with distilled water, and the flask, filled with dis-
tilled water, is inverted, having its mouth closed with the
thumb. . The cornet will fall through the water without
breaking, and can be introduced, together with some of the
water, into a small crucible (cornet crucible), the water
poured off, the crucible and gold gradually dried, and then
heated to redness. When cold, the final operation of
weighing may be performed, thus : The weight represent-
ing 22 carats is placed in one pan of the balance, and the
cornet in the other : as the gold employed was supposed to
be standard, it ought to weigh exactly 22 carats. If, how-
ever, gold of greater or less fineness had been submitted to
assay say of 23 and 21 carats respectively 1 carat weight
would have been required in the pan containing the 22-
carat weight, to counterbalance the gold carat ; in this case
ASSAY OF ALLOYS (ASSAY PKOPEE,). 771
the gold would be 23 carats fine, or, in the usual mode of
reporting, ' one carat better.' If, on the other hand, the 1
carat weight had been found necessary in the pan containing
the cornet, the gold would be 21 carats fine, or ' one carat
worse.'
In cases where it is known that the gold under exami-
nation contains no silver, the only alloy being copper, its
fineness can be estimated by cupelling 24 carats with its
proper portion of lead, and weighing the resulting button,
which should represent the amount of fine gold in the alloy
assayed.
Parting Assays. Parting assays are those assays by
which the amount of fine gold and fine silver in any alloy is
estimated. When the amount of gold exceeds that of the
silver, it is called ' gold parting ' ; when the amount of
silver exceeds that of the gold, ' silver parting.'
In this assay the weights employed in the silver assay
are employed, as the report is made in ounces of fine metal
per pound Troy.
12 grains (representing 1 Ib. Troy) of the alloy are
weighed off, cupelled with 300 grains of lead, and the
resulting button, containing only gold and silver, is weighed.
Suppose it weigh 10 grains, then 2 grains =2 ounces in
the pound of alloy, is copper or some other metal, which
has been oxidised and carried into the cupel with the
litharge. A preliminary assay must be made of the alloy,
to ascertain the approximative quantity of silver and gold,
so as to apportion the amount of silver in the assay proper :
this amount being found, it is to be weighed off, added to
the button of fine gold and silver obtained as above, and
the whole cupelled with 200 grains of lead ; the cupelled
mass of gold and silver laminated and treated with nitric
acid, as already described, and the resulting gold weighed.
Suppose the weight to be 8 grains =8 ounces, the result
would stand thus :
Copper or other base metal . . . 2 oz.
Gold 8
Silver 2
12 oz.
3 D 2
77'2
THE ASSAY OF GOLD.
The above arrangement is very convenient for accom-
plishing gold assays, and is the one employed in the assay
office of the French Mint. The annexed cut (fig. 143)
represents this apparatus.
The assay flask, M 9 being charged with the cornet, a
constant amount of acid is added with a pipette. On the
addition of the second acid a small piece of charcoal is
FIG. 143.
placed in the flask ; this serves to prevent bumping during
ebullition. The flasks are supported on a plate of sheet
iron, P, pierced with holes or by a grating, and the acid
vapours, before escaping by the flue, pass into glass tubes,
JT, about half an inch in diameter, and four feet long : at
each end a narrower tube, , is fused. The lower tube
freely enters the neck of the flask ; and as the space
ASSAY OF PYRITES FOR GOLD. 77a
between is so small that a layer of acid remains suspended
and obstructs the passage of the acid vapours, they are
thus forced to pass into the large tube, where, for the
greater part, they condense and fall into the flasks. By
this means the quantity of acid employed in the assay can
be diminished, as there is no loss by evaporation, and the
results are found to be more constant. In order that the
passage to the large tube for the acid vapours may always
remain free, the end of the narrow tube passing into the
flask must be cut at an angle (see P). The drops of acid
collect at this part, and never close the tube.
Assay of Pyrites for Gold.
Mr. J. M. Merrick gives the following method for
assaying pyrites for gold :
' One pound, or even 18 ounces (avoirdupois), of fine
marble-dust is mixed with 8 ounces of finely pulverised
and sifted pyrites ; the whole then re-sifted and put into
a Hessian crucible, which should be about one-third filled
by the mixture. The crucible is set as usual on a fire-
brick, and a fire of hard coal is made around it, the coals
being heaped up to within an inch of the top. The cm*
cible is covered with a piece of brick or a piece of sheet
iron. During the first half-hour the contents should be
stirred once or twice. As the fire grows brisker the car-
bonic acid evolved keeps the contents of the crucible in
brisk ebullition, and the mixture should be stirred well
every five or ten minutes. On stirring during this time, the
iron rod used seems to meet with but little resistance
from the light mass, but at the end of about 1^ hour the
evolution of gas suddenly ceases, the red-hot mass becomes
heavy, sinks, and requires considerable force to keep it
stirred. It must be stirred well and vigorously, however,
for about half an hour, not leaving it unstirred for more
than a minute, otherwise the mass will fuse or cake, and
the assay will be almost inevitably ruined.
' When a sample taken out in an iron spoon gives off
no smell of sulphur, the entire contents of the crucible
774 THE ASSAY OF GOLD.
must be turned into a stoneware pot or a wooden bucket
half filled with water, and well stirred. When the powder
which should be uniform and free from lumps or fused
pieces has settled, the water must be poured off, the
wet mass allowed to drain, and then transferred to a large
earthen bowl or porcelain mortar. Here it is to be amal-
gamated with about 2 ounces of mercury, to which a little
bit of sodium amalgam has been added. The amalgama-
tion, as well as the stirring in the fire, is a tedious process,
and one which it is as well to do by proxy. It does not
consist in merely grinding with a pestle the mercury in
among the particles of the roasted ore, but this ore itself
must be ground in contact with the mercury, until the
particles are so fine that they will float suspended in water
for several seconds. At the end of say 10 minutes'
thorough grinding, the contents of the bowl are to be
brought into one mass in the bottom of the vessel, the
bowl then sunk in a tub of water, and the contents
* washed down ' an operation not easily described, but
familiar enough to every old miner. It consists essen-
tially in shaking the bowl half full of ore and water
in such a way that the mercury, gold particles, and un-
ground ore sink to the bottom, while the light and
finely ground ore is floated off into the tub. The ore re-
maining is re-ground and re-washed, and these processes
are repeated till nothing but the mercury remains in the
bottom of the bowl or mortar. This mercury is then
dried with filter-paper, and heated in a porcelain capsule
over a Bunsen flame, very gently, until it is sublimed and
the gold remains behind. The film of gold may then be
scraped up and melted, with a little borax and nitre, in
the very smallest-sized Hessian crucible, either with the
foot blowpipe or in a charcoal furnace, by which means a
round, clean button of gold, suitable for weighing, will be
obtained.
' This method has its disadvantages and its counter-
balancing merits. On the one hand, it must be admitted
to be tedious, laborious, and to a considerable degree
uncertain. Some analysts fail with it altogether, while
ASSAY OF PYRITES FOR GOLD. 775
none who have tried it, so far as I know, get closely
agreeing results.
' But, on the other side, it is as certain that this
method will indicate the presence of gold, and will bring
out the gold in a weighable form from pyritic ores, where
the assay by smelting will not show a remote trace of the
precious metal ; and that where the fire assay shows a
certain percentage this will invariably bring out a larger
-amount. Large returns have been obtained by this
amalgamation method from iron pyritic ores, which have
been repeatedly assayed in the ordinary way, by chemists
of great eminence, with uniformly negative results.'
Treatment of Gold- and Silver-Bearing Copper Ores.
A very few words may serve to indicate the present prac-
tice in the separation of the precious metals from copper.
The older processes employed for this purpose were by
far the most complicated and wasteful operations known
to metallurgy, and it is only since the discovery and intro-
duction of the various ' wet ' processes that any but the
richest coppers could be advantageously treated for the
precious metals.
The Ziervogel process has only been successful in a
few isolated cases, and demands such pure material, and
such skill in manipulation, as to debar its use in ordinary
instances ; nor does it provide for the extraction of gold.
It is indisputable that the electrolytic methods are
rapidly advancing to the front in the treatment of gold-
and silver-bearing metallic copper, and have the great
advantages of producing a copper of the best quality,
but are yet largely in the experimental stage, and require
a bulky and expensive plant.
The' new Hunt & Douglas method, as applied to copper
ores or mattes, seems to fill the gap more completely
than any previous invention. By this method the copper
is extracted from the ore or matte after a very imperfect
roasting, and, being precipitated as a dioxide by sulphurous
acid generated from pyrites, it is decomposed by about
one-half its weight of metallic iron, the resulting cement
being fit for immediate refining. The copper is obtained
776
THE ASSAY OP GOLD.
in a state of absolute purity even in the presence of ar-
senic and antimony ; while the residues, containing every
trace of the gold, silver, and lead originally present, may
be smelted with lead ores in a blast furnace. The pro-
cess has long passed the experimental stage, and offers
advantages peculiar to itself and unshared by any other.
The ease with which the small amount of gold some-
times present in cupriferous pyrites may be won is not
realised by all copper smelters, although the method is
extensively practised in this country, as well as at Swansea
and in Chili.
Owing to its great affinity for metallic copper, the gold
contained in white metal may be concentrated into a very
small bulk of the former by exposing the pigs of matte
to a slow oxidising fusion, exactly as in the process
for making blister copper. The operation, however, is in-
terrupted as soon as a certain quantity of metallic copper
is formed, when the furnace is tapped, and the product
now advanced to pimple metal, or even regulus, from 82 to-
88 per cent. being examined, bottoms of metallic copper
will be found under the first few pigs. This is the method
pursued in making best selected copper, for not only does
the small quantity of metallic copper extract the gold,,
but also the greater part of other foreign and injurious
substances, such as arsenic, antimony, tellurium, tin, &c..
The proportion of bottoms formed must vary with the
quantity of gold present ; in some instances, even a repe-
tition of the processes being required to fully extract the
more valuable metal. Silver is but slightly concentrated
by this operation, as will be observed from the following
assays made under the author's direction :
Assay of Original
White Metal
Propor-
tion of
Bottoms
Assay of Bottoms
Proportion
thus
Extracted
Assay
of Residual
Pimple
Metal
Gold
Silver
Gold
Silver
Gold
Silver
Gold
Silver
Ounces
Ounces
Per cent.
Ounces
Ounces
Per cent.
Per cent.
Ounces
Ounces
0-64
93-3
6-4
9-60
213-4
93-7
14-8
0-030
78-7
2-37
16-6
9-0
19-10
36-2
90-2
18-5
0-110
14-2
0-11
5-4
1-73
88-4
-
0-012
DETECTION OF MINUTE TRACES OF GOLD IN MINERALS. 777
In examining this table it must be remembered that a
considerable concentration has taken place in the matte
itself, as well as in the copper bottoms, so that the results
do not seem to agree ; but the figures given are sufficient
to indicate the general results of the process. Unless the
furnace bottom is already well saturated with auriferous
metal, a heavy loss in gold must be expected.
Detection of Minute Traces of Gold in Minerals. Mr.
Skey, analyst to the Geological Survey of New Zealand,
has devised a plan which gives very good results, even
when small quantities of mineral are operated on. He
employs iodine or bromine for the purpose of dissolving
out the gold. Both of these substances differ from chlo-
rine, especially in their relatively feeble affinities for hydro-
gen, so that there is less fear that from the generation
of hydrogen acids any great preponderance of other
matters would be dissolved along with the gold. Either
of these substances can be safely and advantageously
employed for the separation of gold from its matrix.
The following particulars of experiments made by this
method will be useful in showing what is approximately
the smallest quantity of gold that can be positively sepa-
rated and identified when operating upon a limited
quantity.
1st. 2 grms. of roasted ' buddle headings ' from a
quartz mine at the Thames, N.Z., known to contain gold
at the rate of 1 oz. or so to the ton, was well shaken for a
little while with its volume of alcoholic solution of iodine,
then allowed to subside.' A piece of Swedish filter-paper
was then saturated with the clear supernatant liquid, and
afterwards burned to an ash ; the ash, in the place of
being white, as it would be if pure, was coloured purple ;
the colouring matter was quickly removed by bromine
a clear indication of the presence of gold. The time
occupied by the whole process was twenty minutes.
2nd. 1 grm. of the same { buddle headings,' mixed
with such a quantity of earth as to reduce the proportion
of gold present to 2 dwts. per ton, was kept in contact
with its own volume of the tincture of iodine for two
778 THE ASSAY OF GOLD.
hours, with occasional stirring ; a piece of filter-paper was
then saturated with the liquid, and dried, five times con-
secutively, and finally burnt off as before : in this case,
also, the colour of the residual ash was purple, and it
gave the reaction of gold.
3rd. 32 grms. of siliceous hematite, finely pounded,
were thoroughly mixed with precipitated gold to the
amount of % dwts. per ton ; then ignited and treated with
bromine water. After two hours the solution was filtered,
and evaporated to a bulk of 20 minims ; this gave a good
reaction of gold to the ' tin chloride ' test.
4th. 100 grms. of the hematite, with precipitated gold
at the rate of -J dwt. per ton, treated as before, but this
time well washed at the expiration of two hours ; the
washings evaporated along with the first filtrate gave
a fainter, but still decided, reaction of gold to the same
test.
5th. Iodine, as tincture, substituted for bromine in
Experiments 3 and 4, gave similar results ; the only varia-
tion made was, that, as a precautionary measure allowing
for its slower action, they were kept in contact for twelve
hours.
Careful experiments have been made to compare the
results of the common amalgamating process with the
foregoing, and it has been found that it is not certain,
with the same expenditure of labour, to get reliable indi-
cations of gold, when present in less quantity than 2 dwts
per ton, operating upon about 100 grms. of material.
In summing up the results of these experiments, it
appears, then, that for qualitative examinations for gold,
or for quantitative estimations in certain cases, iodine
and bromine are each superior to mercury. It also appears
that a proportion of gold equal to \ dwt. per ton, upon a
bulk of about 4 oz. of ferruginous matters, can be easily
and rapidly detected. Of course, by operating upon
larger bulks, gold could be discovered by this process,
were it present in far less quantities, but this is sufficiently
near for the majority of cases.
These processes are especially adapted for the sepa-
DETECTION OF GOLD IN MINERALS. . 779
ration of gold from sulphides, as the preliminary roasting
is extremely favourable to them, the loss in the substitu-
tion of oxygen for sulphur amounting to 25 per cent, by
weight, while the volume remains constant (or nearly so) ;
hence there is a corresponding porosity in the product,
by which every particle of it is thrown open to contact
with the solution. This mechanical accessibility obviously
cannot be taken advantage of by mercury.
With sulphides these processes are practically ex-
haustive, while at the same time the simultaneous extrac-
tion of other matters is so trifling, that the proper tests
for gold can be safely applied directly to the concentrated
solution. In the roasting of pyrites it is necessary to
raise the temperature towards the end to a full red heat,
in order to decompose the ferruginous sulphates, since if
these remained iron would get into the solution. In the
case of an excess of calcium carbonate being present, it is
proper to gently reignite the roasted mineral, &c., with
ammonium carbonate, or much lime might get into the
iodine or bromine solution. On the other hand, a very
high temperature is to be avoided, for a considerable
quantity of fine gold can escape detection in this way
by the partial vitrification of the more fusible of the
silicates.
The identification of gold by the combustion of its
salts with filter- paper seems to promise a rapid method of
estimating it, comparatively, by the aid of a series of pre-
pared test-papers, representing gold in different degrees of
dilution.
Assay by the Spectroscope. It seemed at one time
possible that the assay of gold and silver alloys might be
simply and rapidly effected by the aid of the spectroscope.
The researches of Professor Chandler Eoberts, F.K.S.,
Chemist to the Mint, and of Mr. A. E. Outerbridge, As-
sistant in the Assay Department of the United States Mint,
show that for the present at least these expectations are
groundless.
It has been shown by Mr. Capel that the ^Vo f a
milligramme of gold will show a spectrum, if the spark be
780 THE ASSAY OF GOLD.
passed through a weak solution of the pure metal. But
when operating on a slip of alloy formed of
Silver 708
Copper 254
Gold 38
1,000
the spectra of copper and silver alone were visible. In
an alloy of gold and copper containing from 200 to 250
parts in the thousand of the precious metal, the gold
spectrum is barely visible. On the other hand, in an
alloy of gold and copper containing 1 per cent, of the
latter, the copper spectrum was distinctly shown. In
copper alloyed with 20 per cent, of nickel, the spectrum
of the latter is not visible. Hence we arrive at the in-
teresting fact that when two or more metals are present,
the spark will to some extent elect for its vehicle the one
which is most rapidly volatilised.
It is also not possible to obtain alloys of gold so per-
fectly homogeneous that the quantity of metal volatilised
and giving the spectrum may safely represent the whole
melt.
781
CHAPTER XVIII.
THE ASSAY OF PLATINUM.
PLATINUM is found in a native or metallic state. It occurs
very rarely, yet it is exceedingly probable that wherever
gold is found this metal will more or less accompany it.
It is found disseminated in sand, in the form of grains
varying in size from gunpowder to hempseed : this last
size they rarely exceed, yet, as in the case of gold, nuggets
have been found of large size and weight. Its colour is
steel-grey, or, rather, a tinge between silver- white and steel-
grey.
The sands from which platinum is derived are remark-
able, from the number and importance of their principal
constituents. With the platinum may be found Au, Ag,
Hg, Fe, Cu, Cr, Ti, Ir, Os, Eh, Eu, and Pd. Besides all
these metals, precious stones have also been found as-
sociated with it.
The following plan will serve to detect platinum in
admixture with gold and other heavy matters obtained by
washing or vanning sands, earths, &c.
Act on a small quantity by mercury, and separate the
amalgam ; by this means the gold is removed. To the
residue add aqua regia and boil, evaporate the solution to
dryness, add a little hydrochloric acid and water, boil and
filter. To the filtered solution add a strong solution of sal
ammoniac (ammonium chloride). If a bright yellow, or
reddish-yellow, granular precipitate falls, platinum is pre-
sent in the sand.
A still more ready method is the following : Separate
as much earthy matter as possible by careful washing. If
gold is present, separate that by amalgamation. Dry the
782
THE ASSAY OF PLATINUM.
residue and take its specific gravity ; if it be above 10,
platinum is most likely present. The specific gravity of
native platinum, free from earthy matter, is from 16 to 19.
Analysis of Platinum Ores. The platinum sands often
contain metallic compounds of iron and platinum, not only
capable of being attracted by the magnet, but possessed
even of polarity. These grains have a different composition
from those not magnetic, as shown in the two following
analyses by Berzelius :
Analysis of the non-magnetic grains :
Platinum 78-94
Iridium
4-97
Rhodium
86
Palladium
28
Iron .
11-04
Copper
70
C ]Q 2fI*tlIlS
1-00
LUm \in scales
96
Analysis of the magnetic grains :
Platinum
Iridium
Rhodium
Palladium
Iron '.'
Copper
Insoluble matters
98-75
73-58
2-35
1-15
30
12-98
5-20
2-30
97-86
These grains being separated, their relative proportion
is estimated.
Bunsen's method of analysing platinum ores is as fol-
lows : The ores employed contain no osmium, and were
relatively rich in rhodium.
Platinum, and Palladium. It is easy to effect the
almost complete separation of platinum and palladium
from rhodium, iridium, and ruthenium. The original
material is mixed in a Hessian crucible, with from ^ to J-
its weight of ammonium chloride, heated until the latter
is completely volatilised, allowed to glow gently until only
the vapours of ferric chloride show themselves, and then
placed in a porcelain dish, with from two to three times
ASSAY OF PLATINUM ORES. 783
its weight of raw commercial nitric acid, and evaporated
to a syrupy consistency. By this treatment with am-
monium chloride the metals present not belonging to the
platinum group will have been partially converted to lower
chlorides, the rhodium, iridium, and ruthenium will have
been rendered insoluble, and the silica present as gangue
converted from a gelatinous mass to a finely pulverulent
condition, in which state it will admit of speedy filtering.
The chlorine compounds, produced by the ammonium
chloride, give, upon digestion with nitric acid, just enough
hydrochloric acid to dissolve the platinum to bichloride,
while the metallic copper and iron present act so far re-
ducingly upon the palladium (in solution in nitric acid)
that it remains in solution, not as bichloride, but as the
protochloride, which latter is not precipitated with potas-
sium chloride. The mass is diluted with water, filtered,
and the solution saturated with potassium chloride, and
the greater part of the platinum separated pure as potas-
sium platinochloride, which is washed out first with potas-
sium chloride, and later with absolute alcohol (the last
washings must not be added to the solution).
The filtrate is poured into a large flask (which can be
made airtight), which will not be more than half-filled with
it. Chlorine gas is passed into this flask, and it is from
time to time shaken vigorously, until no further absorption
of gas takes place, when all the palladium will have sepa-
rated as a cinnabar-red precipitate of potassium palladio-
chloride (somewhat impure, however, from traces of
platinum, iridium, and rhodium). The fluid from which
these precipitates were obtained is now evaporated, not
quite to dryness, with hydrochloric acid ; and, upon ad-
dition of just so much water as is necessary to dissolve out
the potassium chloride and other soluble salts (aiding the
operation by rubbing with a pestle), there remains behind
a dirty, yellow-coloured precipitate. This is separated by
filtration, boiled with caustic soda and a few drops of
alsolute alcohol. Hydrochloric acid is added to dissolve
the precipitate formed, and the liquid then saturated with
potassium chloride ; the result is a precipitate of chemically
784 THE ASSAY OP PLATINUM.
pure potassium platinochloride. The mother-liquid con-
tains only copper and no platinum metals.
The purification of the cinnabar-red precipitate of pal-
ladium is accomplished as follows : Dissolve in boiling
water, whereby a portion of the chloride dissolves, with
evolution of chlorine, to palladium protochloride. Then
evaporate with 2^ times its weight of oxalic acid, and dis-
solve again in a solution of potassium chloride ; where-
upon potassium platinochloride remains behind, chemically
pure. Wash out as before.
The brown liquid is then somewhat concentrated upon
the water-bath : and upon cooling, there separate bright
green, well-formed crystals of potassium palladio-proto-
chloride (with some potassium chloride), which upon
testing proves free from the other platinum metals.
The fluid poured off from these crystals is then
neutralised carefully with caustic soda, and gives a slight
precipitate of copper and iron, which is filtered off.
Upon adding potassium iodide to the filtrate, all the
palladium separates as palladium iodide. To avoid adding
an excess of the reagent, it is best to take from time to
time a drop from the fluid with a capillary tube, and put
the same upon a watch-glass. As long as the precipitation
is incomplete, the drop appears, upon a white background
brown ; when complete, it is colourless ; when the reagent
is present in excess it is red. This is tested for its purity
by reducing it to metallic palladium, and then heating and
dissolving in nitric acid ; when pure, it must dissolve
completely. The whole mass is now reduced in a slow
stream of hydrogen gas (whereby the iodine can be
obtained again, as hydriodic acid, by absorbing with
water). At last the mass must be strongly heated, to
decompose slight traces of the palladium subiodide which
are formed.
The mother-liquid from which all this platinum and
palladium have been obtained may contain some iridium
and rhodium ; it is, therefore, evaporated to dryness with
a little potassium iodide, whereby a mixture of rhodium
and iridium iodides separates. This can either be dissolved
ASSAY OF PLATINUM OKES. 75
in aqua regia, and the two metals separated (as will here-
after be described) by sodium bisulphite, or it can be
united with the next portion from which these metals will
be obtained.
Ruthenium, Rhodium, and Tridium. The residue from
the original material which remains, after treatment with
ammonium chloride and nitric acid, is treated as follows,
to get the metals in a form adapted to further chemical
treatment.
The method depends upon the behaviour of zinc
chloride to zinc. If a piece of zinc be melted, it rapidly
covers itself with a stratum of oxide. If, to the melted
metal, a metal like iridium be added, the oxide stratum
hinders the latter from coming into contact with the zinc,
even though it be pushed beneath the surface. If, how-
ever, a few grains of ammonium chloride be added to it,
ammonia, hydrogen, and zinc chloride will be formed,
which last dissolves the oxide stratum to basic zinc chlo-
ride. The zinc below resembles mercury in lustre and
mobility. As soon as the chloride has dissolved as much
of the oxide as is possible for it, the oxide stratum again
forms, and is instantly removed again by the addition of
more ammonium chloride. The melted zinc, strewn with
ammonium chloride, also possesses, like mercury, the
property of attacking other metals, if the affinity exists
of forming alloys with them. By strewing ammonium
chloride upon the melted zinc, a quiet surging is kept
up, as the ammonia and hydrogen are given off. Many
oxides and chlorides (among which are those of the
platinum metals), when they come into contact with this
atmosphere of reducing gases, and with the basic zinc
chloride, are instantly reduced and dissolved to alloys by
the zinc. In making the solution, the zinc, in a porcelain
dish, should be constantly rotated : the gangue remains
in the basic chloride. The regulus, immediately upon
solidifying, should be taken from the capsule, out of the
yet fluid basic chloride, and washed off with acetic acid
until all the basic chloride is dissolved away. The gangue
can be quantitatively estimated by filtration and weighing.
3E
786 THE ASSAY OF PLATINUM.
If the regulus is not immediately removed, the containing
vessel will be broken, owing to the unequal expansion of
the porcelain and the metal.
The best proportions for a quantitative separation
are, to 1 part of the platinum metals, from 20 to 30 parts
of zinc. For an ordinary separation 7 parts of zinc are
sufficient.
For the extraction of the residues remaining after the
treatment with nitric acid, this method is admirably
adapted. By fusing only once with zinc for two or three
hours, all the platinum metals are extracted. The opera-
tion is as follows :
From 3 to 3*5 kilos, of commercial zinc are fused in a
2-litre Hessian crucible, ammonium chloride from time
to time strewn upon it ; 400 grms. of residue, previously
heated to faint glowing with ammonium chloride, are
added, and the temperature kept, for two or three hours,
just above the fusing-point of the alloy, by adding, when-
ever the mass threatens to solidify, some ammonium
chloride. The mass is divided into three strata after
solidification has taken place.
The outer stratum, easily broken away by a blow from
a hammer, contains no platinum metals. The next con-
tains some particles of the zinc and platinum alloy, im-
bedded in the basic zinc chloride ; it is porous, and not
very thick. The inner stratum consists of a beautiful
crystalline regulus.
To obtain the alloy from the middle stratum, it is
only necessary to wash repeatedly with water ; and the
alloy gained is, of course, to be added to the regulus. To
obtain this regulus as pure as possible, it is again fused
with 500 grms. of zinc and some ammonium chloride, then
granulated in water, and the granules dissolved in fuming
hydrochloric acid. The acid attacks the regulus with
greatest energy, and the solution is complete in less than
an hour. The zinc chloride can be used for the next
operation.
The platinum metals are found at the bottom of the
vessel, in the form of a finely divided black powder, which
ASSAY OF PLATINUM ORES. 787
is contaminated with zinc, and with traces of iron, copper,
&c., from the latter. It cannot be purified with nitric
acid, nor with aqua regia, for part of the platinum metals
will thereby be dissolved, or, at best, so suspended in the
fluid that filtration is impossible. If, however, the powder
is treated with hydrochloric acid, singularly enough, all
the impurities are dissolved ; not only zinc and iron, but
also lead and copper, dissolve readily with the generation
of hydrogen. The explanation is readily found in elec-
trical currents produced by the contact of the metals, the
stream passing from the positive zinc, iron, &c., to the
negative platinum metals, hydrogen being given from the
latter, and chlorine from the former, and uniting with
them. The metallic powder, after thorough washing,
possesses the property, upon being gently heated, of ex-
ploding weakly, and, when highly heated, with violence,
the explosion being accompanied with the evolution of
light, although neither hydrogen, chlorine, nitrogen, nor
aqueous vapour is given off; and as these are the only
elements which it is possible that the metallic powder
could have taken up, it must be assumed that these metals
are, by this treatment, converted into an allotropic con-
dition, and that, upon heating, they return, with more
or less energy, to their original condition. The powder
contains, mainly, rhodium and iridium ; but there are
traces present of platinum, palladium, lead, copper, iron,
and zinc.
It is intimately mixed with about 3 or 4 times its
weight of completely anhydrous barium chloride, and a
stream of chlorine gas led over it at a tolerably high
temperature. The operation is concluded when particles
of ferric chloride show themselves on the neck of the flasks
containing the powder. These are carefully brushed away
with filter-paper. Some water is now added, and the
mass of the platinum metals dissolves with the evolution
of heat. There remains behind insoluble matter, which,
upon reduction with hydrogen, alloying with zinc, and
treatment with hydrochloric acid, furnishes ruthenium.
From the solution all the barium chloride is removed by
3 E 2
788 THE ASSAY OF PLATINUM.
careful addition of sulphuric acid. The platinum metals
are now completely freed from all other metals by reduc-
tion with hydrogen, the temperature being, throughout
the operation, maintained at nearly 100 C., by means of
a constant water-bath. Platinum and palladium chiefly
separate first ; then mainly rhodium ; and the last por-
tions consist almost entirely of iridium. It is best to
break off the operation when the fluid has assumed a
greenish-yellow colour. The last portions of iridium
(obtained by evaporating the solution to dryness, fusing
with sodium carbonate, and treatment with aqua regia)
are added to the portion, afterwards to be again rendered
workable by renewed treatment w r ith barium chloride.
The operation of reduction is hastened by concentrating
the fluid ; in doing which care must be taken to guard
against explosion, on account of the hydrogen. The
separated metals are treated with aqua regia, and the
platinum and palladium thus dissolved separated from
each other as already described. The traces of rhodium
and iridium in the mother-liquid can be removed entirely
by continued boiling with potassium iodide (whereby they
precipitate as iodides) ; they are then dissolved in aqua
regia and added to the insoluble portion.
This insoluble and partly oxidised portion is now again
reduced by hydrogen gas, treated as before described, with
barium chloride, and, after the removal of the barium,
the last traces of platinum and palladium are removed by
boiling with caustic soda. Ehodium and iridium now alone
remain to be separated.
The brown-red fluid is, for this purpose, evaporated
with hydrochloric acid, and, after filtration, treated with
sodium bisulphite in great excess, and the whole allowed
to remain quietly in the cold for several days. The double
rhodium and sodium sulphite separates slowly, giving a
lemon-yellow precipitate. The solution becomes lighter
and lighter, and finally almost colourless. The colour of
the precipitate changes with that of the fluid, becoming,,
with it, lighter. This precipitate, upon washing, contains
the rhodium almost pure.
ASSAY OF PLATINUM ORES. 780
Upon heating the fluid gently, a yellow-white preci-
pitate separates, which consists mainly of rhodium, but
contains also some iridium. After filtering off this pre-
cipitate, the solution, upon being concentrated to a small
volume, gives yet two precipitates
1. A curdy, slowly separating, yellowish white pre-
cipitate containing nearly chemically pure iridium, with
but the faintest traces of rhodium.
2. A heavy crystalline powder, quickly separating,
which is readily freed from the first by decantation. Upon
testing, it gives all the reactions for iridium, but likewise
some peculiar reactions not shown by the latter.
The complete separation of rhodium from iridium is
accomplished by treating the yellow precipitates with
concentrated sulphuric acid. They are added in small
portions to the acid, heated in a porcelain capsule until
all the sulphurous acid has escaped, and then left upon the
sand-bath until the free sulphuric acid has been driven off
and sodium sulphate formed. Upon boiling the mass in
water, all the iridium dissolves as sulphate, with a chrome-
green colour, while the rhodium remains behind as a flesh-
coloured double salt of sodium and rhodium. The latter
is boiled in aqua regia, and washed by decantation. It
is insoluble in water, hydrochloric or nitric acids, and in
aqua regia. The rhodium and iridium are now completely
separated.
The first yellow precipitate obtained in the cold by the
sodium bisulphite gives, by this treatment, the rhodium
quite pure. The second and third precipitates, containing
much iridium, give very fine rhodium, but still slightly
contaminated with iridium. The products, therefore,
obtained by this treatment with sulphuric acid (which
betray their contamination with iridium by their somewhat
brownish colour) are collected for themselves, the rhodium
separated therefrom by glowing, treated again with barium
chloride, and the operation of separation repeated. The
green solution, containing only iridium, is gradually heated
over an ordinary burner, in a porcelain capsule, and after-
wards upon the sand-bath, to remove the excess of sul-
790 THE ASSAY OF PLATINUM.
phuric acid ; and, finally, the capsule and its contents are
highly heated in a Hessian crucible. There is formed
thereby sodium sulphate and iridium sesquioxide. Upon
boiling the mass with water, the last remains behind as
a black, insoluble powder, which is readily washed by
decantation.
C. Lea's Process for Analysing Platinum Ores. The
ores on which these analyses were performed contained
chiefly iridium, together with ruthenium, osmium, rhodium,,
and platinum. It was a Californian osni-iridium which had
already undergone a preliminary fusion with nitre and
caustic potash.
This material is boiled with aqua regia to extract all
the soluble portions, the residue then ignited with nitre
and caustic soda,* and the fused mass heated with water.
From the resulting solution small portions of potassium
osmite crystallise out. The metallic oxides are next pre-
cipitated, and this precipitate, together with the portions
insoluble in water, is boiled again with aqua regia, ignited
again, &c. These ignitions still leave a small portion of
un at tacked residue.
The boiling with aqua regia is continued for a long
time, in order to get rid as thoroughly as possible of the
osmic acid. Even 200 hours' boiling, however, still leave
osmium in the solution in easily recognisable, but in com-
paratively small, quantity. The greatest advantage is
found throughout the whole of this part of the operation
from the use of a blowing-apparatus, by the aid of which
all inconvenience from the fumes of osmic acid is avoided.
The apparatus is constantly swept clear by a powerful air-
current, and the osmic acid is removed as fast as it is
volatilised. As the ignition of the ore with alkaline nitrate
and caustic alkali scarcely drives off any osmium, and
as almost all inconvenience in manipulating the resulting
* Attention is necessary to the order in which these substances are em-
ployed. If the caustic soda is melted first, it attacks the iron vessel strongly,
and may even go through. If added last, it causes sudden and violent effer-
vescence, with danger of boiling over. Therefore, place the nitre first in the
vessel, and when it is fused add the caustic soda. When a red heat is obtained
add the osm-iridium by degrees.
c. LEA'S PROCESS. 791
solutions can be avoided by throwing down. the metals with
alcohol from the hot alkaline solution, in place of using
acid, it is clear that the difficulties arising from the noxious
effects of osmic acid can be almost wholly removed from
each of the various stages of the process.
A very prolonged treatment with aqua regia is found to
have the great advantage of converting nearly the whole
of the ruthenium into bichloride. The separation of
ruthenium in this form from the other metals is so easy in
comparison with the difficulties presented by the separation
of the sesquichloride, that this advantage cannot be looked
upon as other than a very material one.
Sal-ammoniac is next added to the mixed solution in
quantity sufficient to saturate it. The sandy crystalline pre-
cipitate (A) is thoroughly washed out, first with saturated,
and then with dilute sal-ammoniac solution. The saturated
solution of ammonium salt carries through with it nearly
the whole of the ruthenium as bichloride (B) ; the dilute
solution is found to contain small quantities of iridium,
rhodium, and ruthenium (C).
Over (A), water acidulated with hydrochloric acid is
placed, and allowed to stand for some days. This is treated
with ammonia and boiled. The precipitate, when treated
with hydrochloric acid, furnishes green osmium chloride,
with traces of ruthenium.
In these preliminary steps, Claus's process has been
followed, which undoubtedly offers advantages over any
other, and best brings the metals into a convenient state
for separation, varying it only by prolonging the treatment
with aqua regia^ and converting the ruthenium principally
into bichloride instead of sesquichloride.
We have now three portions of material: (A), con-
sisting of ammonium iridiochloride, containing also ruthe-
nium, osmium, rhodium, and platinum in small quantities.
(The ore under examination contained no palladium, which
metal, if present, has always its own peculiar mode of
separation, and does not increase the difficulties of opera-
tion.) (B), containing ruthenium bichloride, together
with iron in quantity, copper, and other base metals which
792 THE ASSAY OF PLATINUM.
may be present. Finally (C), containing chiefly ruthenium
bichloride, mixed with small quantities of iridium and
rhodium.
The next step in the process is to introduce the am-
monium iridio-chloride (A) into a large flask with twenty
to twenty-five times its weight of water, and apply heat
until the solution is brought to the boiling-point ; the
whole of the ammonium iridio-chloride should be brought
into solution in order that the reduction to be effected
may not occupy too long a time, as otherwise the platinum
and ruthenium salt, if any be present, might likewise be
attacked. Crystals of oxalic acid are thrown in as soon
as the solution actually boils, whereupon a lively efferves-
cence takes place, and the iridium salt is rapidly reduced.
As fast as the effervescence subsides, more oxalic acid is
added until further additions cease to produce any effect.
When this is the case, the liquid is allowed to boil for two
or three minutes longer, not more ; the heat is to be re-
moved, and the flask plunged into cold water.
By this treatment any platinum present is unaffected.
Sal-ammoniac in crystals is added, about half enough to
saturate the quantity of water present. The sal-ammoniac
may be added immediately before the flask is removed
from the fire. After cooling, the solution should be left for
a few days in a shallow basin, whereby the ammonium
platino-chloride will separate out as a yellow, reddish, or
even (especially if the quantity of water used was insuf-
ficient) as a black crystalline powder, according to the
quantity of iridium which it may contain.
The mother-liquor is to be again placed in a flask and
boiled with aqua regia. On cooling, the ammonium iridio-
chloride crystallises out, and any traces of rhodium and
ruthenium which maybe present remain in solution. The
iridium salt is to be washed with a mixture of two parts
of a saturated solution of sal-ammoniac and three parts of
water, and may then be regarded as pure.
The treatment by oxalic acid affords iridium free from
all traces of ruthenium.
The treatment of solutions (B) and (C) presents no
c. LEA'S PKOCESS. 793
difficulty. With (B) the best plan is to place the solution
aside in a beaker covered with filter-paper for some time.
Treated in this way, the bichloride gradually crystallises
out, and by re-crystallisations may be obtained in a state
of perfect purity.
Solution (C) is to be evaporated to dryness, and re-
duced to an impalpable powder. It is then to be thrown
upon a filter, and thoroughly washed with a perfectly
saturated solution of sal-ammoniac. The ruthenium bi-
chloride is thus carried through, with perhaps a trace of
rhodium sesquichloride, from which, however, it is easily
freed by crystallisation. From the residue, the rhodium
and ammonium sesquichloride is removed by a dilute solu-
tion of sal-ammoniac, perfectly free from the iridium, w r hich
is left behind.
In connection with this separation, Mr. Lea makes a
remark which, though of special reference to this par-
ticular case, is also applicable to all those cases in which
the double chlorides of the platinum metals are to be
separated by their various solubilities in solution of sal-
ammoniac. This most valuable process, for which we are
indebted, as for so much else, to Glaus, whose untiring
labours have made him the father of this department of
chemistry, requires to be applied with some attention to
minutiae.
The crystalline matter must be reduced to the finest
powder, and after being thrown upon the filter it must be
washed continuously until the separation is effected. Any
interruption of the washing is followed by more or less
crystallisation of sal-ammoniac through the material, which
precludes an effectual separation. The same material, which
in a state of coarse powder will hardly yield up enough
ruthenium bichloride to colour the sal-ammoniac solution,
will, when thoroughly pulverised, give an almost opaque
blood-red filtrate.
Solution (C) may be subjected to a different treatment
from the foregoing, and oxalic acid may be used to effect
the separation. The solution is to be brought to the boil-
ing-point, and oxalic acid added as long as effervescence
794 THE ASSAY OF PLATIXUM.
is produced. The iridium bichloride is thereby reduced ;
the ruthenium bichloride and the rhodium sesquichloride
are not affected. Sal-ammoniac is then to be dissolved
in the solution to thorough saturation. By standing and
repose the double rhodium and ammonium chloride sepa-
rate out. The solution is then re-oxidised by boiling with
aqua regia ; by standing for some days in a cool place, the
ammonium iridio-chloride crystallises out, and the super-
natant solution contains the double ammonium chloride
and ruthenium bichloride, which may be rendered pure
by several re-crystallisations.
For purifying the double iridium and ammonium chlo-
ride the oxalic process is decidedly the best. It is simple
and less troublesome, and there is the further advantage
that the platinum is left in the condition of double chloride ;
whereas when the usual method of treating with aqueous
sulphuretted hydrogen is used, the platinum is apt to be
converted partly into sulphide, together with any traces
of rhodium and ruthenium which may be present. When
oxalic acid is used, the platinum remains behind as a red-
dish powder, containing some iridium, from which it may
be freed in the ordinary manner, if it is present in quantity
sufficient to be worth working.
For treating a mixture such as that which is here
designated as (C), containing no platinum, and only ruthe-
nium present in the form of ammonium ruthenio-chloride,
it is unnecessary to apply reducing-agents, and the first
method described is the best. But if it be proposed to
effect the separation by the reduction of the iridium com-
pound, the method here described is preferable to that
based on the use of sulphuretted hydrogen even in this
case.
The action of oxalic acid on the platinum metals is
interesting ; its reducing effect upon iridium bichloride at
the boiling-point is immediate. On ruthenium bichloride
it seems to have no effect whatever, and they may be boiled
together for a length of time without sensible result. In
a trial made with ruthenium and ammonium sesquichlo-
ride, the oxalic acid was boiled with the metallic salt for a.
DEVILLE AND DEBRAY S PROCESS. 795
considerable time without any effect becoming visible, but
by long-continued boiling a gradual precipitation took
place. When ammonium platino-chloride was boiled with
oxalic acid, no effect was produced for a considerable
time, but gradually the platinum salt diminished in quan-
tity, and the liquid acquired a stronger yellow colour,
perhaps owing to formation of soluble platinic oxalate.
This process will not, however, furnish an easy and conve-
nient method of purifying commercial platinum from the
iridium always found in it, as the reduction of very small
quantities of double iridium and ammonium chloride in
the presence of a large proportion of the corresponding
platinum salt is difficult and slow, and the platinum salt
itself is evidently attacked.
The following method of Analysis of Platinum Ores, by
MM. Deville and H. Debray, may be useful. The ores of
platinum contain the following substances :
1. Sand. The whole of the sand is never removed
by washing the ore ; and the sand contains quartz,
zircon, chromate of iron, and, in the Russian ores, titanate
of iron.
2. Osm-iridium.
3. Platinum, iridium, rhodium, and palladium, com-
bined, no doubt, in the form of an alloy.
4. Copper and iron, which exist in the ores as an
alloy, for the iron found in the sand is not soluble in
acids.
5. Gold, and, oftener than is supposed, a little silver.
The latter metal is generally found with the palladium,
and it is very rarely that palladium is obtained quite free
from silver when it is prepared by the old processes.
1. Sand. To estimate the sand, we take a small assay-
crucible, or an ordinary crucible with smooth sides, and
melt in a little borax, so as to glaze the inside. We now
introduce from 7 to 10 grammes of pure granulated silver,
and 2 grammes of the ore fairly taken, and weighed very
accurately. Over the platinum we put 10 grammes of
fused borax, and one or two small pieces of wood charcoal.
The silver is now melted, and care must be taken to keep
796 THE ASSAY OF PLATINUM.
it for some time a little hotter than the melting-point, so
that the borax may be very liquid, and may dissolve the
vitreous matters which accompany the platinum and con-
stitute the sand. The crucible is now allowed to cool, and
when it is cold the button, which will contain the silver,
osmium, platinum, and all the other metals, is detached,
and if necessary digested for a time with weak fluoric acid
to remove the last portions of borax. It is now heated to
a faint redness, and then weighed. The weight of the
button, subtracted from the sum of the weight of the ore
and silver employed, will give the amount of sand con-
tained in the ore. For example :
Milligr.
Californian ore ...... 2000
Silver ......... 7221
9221
After fusion, the button weighed . . 9162
Consequently, the ore contained, sand . 59
It is very important to know this number, for it represents
the only matter absolutely destitute of value which the
ore contains ; and this simple operation may be considered
the most important performed in estimating the value of
an ore. It is, besides, performed so quickly, that it is as
well to do at the same time two or three specimens, taken
from different parts of a lot of platinum powder.
2. Osm-iridium. Another 2 grammes of the ore weighed
very accurately are treated with aqua regia at 70 (Cent.)
until the platinum is entirely dissolved. The aqua regia
must be renewed occasionally for 12 or 15 hours, or until
it is no longer coloured. It is best to perform this opera-
tion in a large beaker, and to place a cover over it to pre-
vent loss. The solution must be decanted with the greatest
care from the metallic spangles of the osm-iridium and the
sand which remain at the bottom of the beaker. If neces-
sary it may be filtered, but as little as possible of the
osmide must be allowed to go on the paper. The insoluble
residue must be washed by decantation, then dried and
weighed, after having added what remained on the filter-
By subtracting the weight of the residue from the weight
DEVILLE AND DEBRAY's PROCESS. 797
of the sand obtained in the former operation, we obtain the
weight of the osm-iridium. For instance, in the California^
ore we had :
Milligr.
Osm-iridium and sand 81
Sand .59
Osrn-iridium ,22
The button obtained in estimating the sand might be
employed in this operation. In that case it is necessary to
dissolve out the silver with nitric acid, and then proceed
with the residue, as we have just directed.
3. Platinum and Iridium. The solution in aqua regia
obtained in the last operation is evaporated to dryness at a
low temperature, and the residue is re-dissolved in a small
quantity of water (if it should not entirely dissolve in the
water, some more aqua regia must be added, and the evapo-
ration repeated), to which is added about twice as much
pure alcohol ; lastly, we add a great excess of sal-ammoniac
in crystals. The whole is now slightly warmed to com-
plete the solution of the sal-ammoniac ; it is then stirred,
and afterwards set aside for 24 hours. The orange-yellow,
or even reddish -brown, precipitate which is formed con-
tains the platinum and the iridium, but some remains in
the solution. The precipitate must be thrown on a filter
and washed with alcohol. Afterwards the filter is dried
in a platinum crucible, placed, for greater safety, within a
larger one, and then heated by degrees to low redness.
The crucibles are now uncovered, and the filter is burnt at
the lowest possible temperature. Once or twice after the
incineration of the filter a piece of paper saturated with
turpentine should be introduced into the crucible, by
which means the iridium oxide will be reduced, and the
expulsion of the last traces of osmium will be effected.
The crucible is now heated to whiteness until it no longer
loses weight, or the reduction is finished in a current of
hydrogen.
The liquid separated from the platinum -yellow by fil-
tration is evaporated until the ammonium chloride crys-
tallises in great quantity. It is allowed to cool, is then
798 THE ASSAY OF PLATINUM.
decanted, and on a filter is collected a small quantity of
a deep violet-coloured salt, which is the iridium ammonio-
chloride mixed with a little of the platinum salt. This is
first washed with a solution of sal-ammoniac, and then with
alcohol. The salt is then ignited, and if necessary reduced
by hydrogen like the platinum salt. The mixture of pla-
tinum and iridium obtained by the two reductions is then
weighed. The two metals are now digested at about 40
or 50 (Cent.) in aqua regia, diluted with about 4 or 5
times its weight of water the aqua regia being renewed
until it is no longer coloured. The residue is pure iridium.
To obtain the weight of the platinum the weight of the
iridium is subtracted from that of the mixture of the two.
This method of separating the two metals is very accurate
If the aqua regia used be weak, and the contact with it
prolonged.
4. Palladium , Iron, and Copper. The liquid charged
with sal-ammoniac and alcohol, from which the platinum
.and iridium have been separated, is evaporated to get rid
of the alcohol, and then treated with an excess of nitric
acid, which transforms the ammonium chloride into nitro-
gen and hydrochloric acid. It is now evaporated almost
to dryness. The residue is removed to a covered porcelain
crucible, which is weighed with great care. When the
matter is dry it is moistened with concentrated ammonium
sulphide, and afterwards dusted over with 2 or 3 grammes
of pure sulphur. When dry this crucible is placed within
a larger one of clay, and surrounded with pieces of wood
charcoal. The two, covered, are now set in a cold furnace
which is filled up with charcoal, and the fire is lighted
at the top to avoid the projection of any matter from the
crucible if it were too quickly heated. After reaching a
bright red heat the crucibles are allowed to cool. The
porcelain crucible now contains palladium in a metallic
state, with the sulphides of iron and copper, and also the
gold and rhodium. This mixture is moistened with con-
centrated nitric acid, which, after prolonged digestion at
70, dissolves the palladium, iron, and copper, forming at
the same time a little sulphuric acid. The solution of the
GUYABDS PEOCESS. 799
nitrate is poured off from the residue, which is washed by
decantation, and the solutions and washings are evaporated
to dryness, and then calcined at a strong red heat. In this
way the palladium is reduced, and the iron and copper
pass to the state of oxides, which are easily separated from
the palladium by means of strong hydrochloric acid. The
palladium remains in the crucible, in which it is again
strongly ignited and then weighed.
The iron and copper chlorides are now evaporated to
dryness at a temperature but little above 100 (Cent.), and
are then treated with ammonia. The ferric chloride, having
lost nearly all its acid, has become insoluble ; but the copper
chloride is readily dissolved, and may be filtered from the
iron, which is washed, ignited, and weighed. The copper
solution is now evaporated almost to dryness, and then mixed
with excess of nitric acid, and heated to drive off the am-
monium chloride. Afterwards the copper nitrate is ignited
and weighed. The weight of the copper is always so small
that the hygroscopic water the copper oxide may absorb
may be neglected.
5. Gold and Platinum. The residue insoluble in nitric
acid is weighed and treated with very dilute aqua regia
which takes up the gold, and sometimes, but very rarely,
traces of platinum. To ascertain if platinum be present,
evaporate to dryness, and re-dissolve by alcohol and am-
monium chloride. If any platinum-yellow remain, it must
be ignited and weighed. The difference in the weight of
the porcelain crucible before and after the treatment by
aqua regia gives the weight of the gold, from which, if
any be found, the weight of the platinum must be deducted.
6. Jthodium. The residue left in the crucible is rho-
dium, which must be reduced in a current of hydrogen.
We append the results of some analyses of platinum
ores, by MM. Deville and Debray. (See Table on next page.)
M. A. Guyard gives the following process for the ex-
traction of metals from platiniferous residues :
' This process comprises three different operations, which
I will succinctly describe.
' 1 . Solution of the Residues. The mother- liquors which
800
THE ASSAY OP PLATINUM.
ANALYSES OF PLATINUM ORES FROM VARIOUS SOURCES.
Columbia
California
Oregon
Spain
Australia
Russia
Platinum
80-00
79-85
51-45
45-70
59-80
77-50
Iridium
1-55
4-20
0-40
0-95
2-20
1-45
Rhodium
2-50
0-65
0-65
2-65
1-50
2-80
Palladium
1-00
1-95
0-15
0-85
1-50
0-85
Gold .
1-50
0-55
0-85
3-15
2-40
(*)
Copper
Iron .
0-65
7-20
0-75
4-95
2-15
4-30
1-05
6-80
1-10
4-30
2-15
9-60
Osm-iridium
1-40
4-95
37-30
2-85
25-00
2-35
Sand .
4-35
2-60
3-00
35-95
1-20
1-00
Osmium and loss
0-05
0-05
0-80
2-30
100-15
100-00
100-25
100-00
100-00
100-00
remain after the precipitation of platinum by sal-ammoniac
come from solutions of crude or commercial platinum.
They always contain iron, mostly produced from the iron
sulphate used for the precipitation of gold, lead, copper,
palladium, platinum, and especially rhodium. These mother-
liquors are acidulated by hydrochloric acid, and are then
ready to be investigated. To recall their composition, I
shall distinguish them here only as residues in solution. It
need only be mentioned that iron, which is generally used
for the precipitation, must be avoided.
' Solid residues are melted at once with three times
their weight of a mixture of equal parts of soda and sodium
nitrate. The fusion is effected at a bright red heat in a
thick iron vessel. It is accomplished without bubbling or
projection, and requires about an hour. During the last
twenty minutes the mass must be constantly stirred with
an iron spoon. The operation is extremely simple.
4 These residues contain osm-iridium, unattackable by
all chemical agents, attackable osmide, some grains of triple
alloy of platinum, iridiurn, and rhodium, which aqua regia
will not dissolve, but which nitre completely oxidises and
completely breaks up. They also contain the gangue
characteristic of platinum ores quartz, silicates of all
bases, titanates, hyacinth, &c., &c.
' The mixture I make use of oxidises all that is oxidis-
* Gold, if any, counted in the loss.
GUYAED S PEOCESS. 801
able, and breaks up the gangue, which it partly dissolves.
The melted mass contains all the bodies above mentioned,
besides a large quantity of iron oxide taken from the sides
of the vessel in which the operation is performed. The
fused mass is poured into cast-iron moulds. When solid
it is broken into fragments and boiled with sufficient water
to obtain a strong solution of soda, capable of holding all
the gelatinous acids in solution. It also contains osmium
in the state of osmiate.* It is filtered from insoluble
matter, and then supersaturated with hydrochloric acid.
The insoluble oxides are freed by washing from the excess
of alkali, and are then dissolved in aqua regia.
' This solution contains iron, copper, lead, iridium,
rhodium, platinum, and ruthenium. It is separated from
the undissolved osmide, evaporated to expel the excess of
aqua regia and dissolved in water and hydrochloric acid.
4 2. Precipitation of Liquids by Sulphuretted Hydrogen.
Liquids obtained as above are ready for precipitation by
.hydrosulphuric acid.
4 The apparatus in which all the liquids are precipitated
is composed of a sulphuretted hydrogen gas generator by
the action of sulphuric acid on iron sulphide. This gene-
rator communicates with four or five large earthenware
jars holding about seventy litres, arranged precisely as in
Wolffs apparatus. A special tube conducts to each of
them the vapour destined to heat the liquid which they
contain.
c The whole apparatus is enclosed in a well-fitted wood
stove placed near a chimney, with which it communicates.
As to the small quantities of unabsorbed gas, they are
conducted into the chimney, where the fire creates a strong
draught. By this means, also, all smell is avoided during the
precipitation ; but after the operation air is forced through
the apparatus from large gasometers. It expels the hydro-
sulphuric acid which saturates the mother-liquors, and
these can then be manipulated free from smell, f
* This solution is separately precipitated by hydrosulphuric acid. Osmium
sulphide is thus isolated.
t A carbonic acid generator may be substituted for the gasometers and the
air with no difference in the result.
3F
802 THE ASSAY OF PLATINUM.
' The experiment is carried on during the precipitation
in the following manner : when the generator begins to
disengage gas, the temperature of the liquids is raised to
about 70. This temperature is maintained for nearly fif-
teen hours, that being the time required for the complete
precipitation of the sulphides, which collect better under
the influence of heat. The operation is concluded when
there remains but a very slight yellow tint in the mother-
liquor, arising from the presence of a little soluble iridium
sulphide. This mother-liquor is poured from the precipi-
tated sulphides into a vessel with pieces of iron, which
takes off a little of the iridium. The sulphides are filtered
through linen filters.
' 3. Purification and Treatment of the Sulphides. The
mass of sulphides thus separated from the iron and from
all other bodies not precipitated by the sulphuretted gas,
contains, in addition to the sulphides of the platinum
metals, a large proportion of sulphur and the sulphides
of copper and lead. To get rid of these bodies, I have
thought of concentrated sulphuric acid, which changes
them to sulphurous acid and sulphates, while it does not
act on the sulphides of the precious metals. This refining
can be effected in an iron vessel ; but Mr. Matthey, who
neglects nothing to insure the certainty and exactness of
the results, makes use of platinum.
' When, after prolonged boiling, no more sulphurous
acid is given off, the refining is complete.
' The mass of sulphides, diluted with a quantity of
water, is thrown on filters, and thoroughly w r ashed, until
ammonia no longer shows any trace of copper or iron in
the filtered liquid.
' At this point the precious metals are entirely freed
from iron, which is so detrimental to them, and from
copper, and contain only a little lead sulphate, which
separates by itself during an ulterior reaction. They are
then, moreover, in a condition to be dissolved by simple
nitric acid or by aqua regia, and this is not their least
valuable condition.
' Treatment of the Sulphides. The sulphides are next
ANALYSIS OF OSM-IRIDIUM. 803
dissolved in aqua regia, which should not be previously
prepared, because its action on sulphates is sudden and
energetic ; it heats so rapidly, and the disengagement of
gas is so great, that, were it previously prepared, it would
certainly be thrown from the vessels.
' I add then moderately strong cold nitric acid, and
add it gradually, because its action is strong. A quantity
of rutilant vapours are disengaged. Hydrochloric acid is
added when the effervescence ceases. It is then gradually
heated to boiling, which is necessary to obtain a complete
solution.
' The solution is poured from the deposited lead
chloride, and the ordinary method with sal-ammoniac is
used to separate the different metals it contains. Experi-
ments on large quantities of material have fully proved
the advantages of this process.'
Analysis of Osm-iridium. Wohler's method of resolv-
ing osm-iridium consists in passing moist chlorine over
the ore mixed with common salt and heated to low red-
ness in a glass or porcelain tube. This method is invalu-
able in analysis, and gives excellent results in working the
ore upon a small scale. In all cases, however, several
repetitions of the process are necessary for complete
resolution or reduction to a soluble form. On the other
hand, it can scarcely be doubted that this method could be
advantageously employed upon the large scale, if vessels
of porcelain of large size and of a proper shape could
be obtained. Such vessels might be constructed in the
form of long and flattened ellipsoids, furnished at each
extremity with wide tubes several inches in length, and
would be of great utility in various chemical processes.
No process of fusion with oxidising agents compares with
Wohler's method in point of elegance, as no iron or other
impurities afterwards to be removed are introduced by
the process itself.
Claus's method of resolving the ore consists in fusing
for an hour, at a red heat, a mixture of one part of ore
with one part of caustic potash and two of saltpetre. The
fused mass is to be poured out upon a stone, allowed to
3 F 2
804 THE ASSAY OF PLATINUM.
cool, broken into small pieces or powdered, and then
introduced into a flask, which is to be filled with cold
water and allowed to stand for twenty-four hours. The
clear deep orange-red solution of potassium osmiate and
rutheniate is then to be drawn off by means of a syphon,
and the black mass remaining again washed in the same
manner. The finely divided oxidised portion of the in-
soluble matter may now be separated from, the unattacked
ore by diffusion in water and pouring off, after the sub-
sidence of the heavier ore. The unattacked ore is then
to be fused a second time with potash and saltpetre and
treated as before. Clans asserts that he has been able in
this manner to resolve the Siberian osm-iridium completely
in two operations.
Dr. Wolcott Gibbs, to whom the chemistry of the
platinum metals is so greatly indebted, recommends the
following process for the analysis of osm-iridium : The
ore, which is usually very impure, is in the first place-
to be fused with three times its weight of dry sodium
carbonate. The fused mass after cooling is to be treated
with hot water, to remove all the soluble portions, and
then the lighter portions are to be separated by washing
from the heavy unattacked ore. In this manner the
greater part of the silica and other impurities present rnay
be removed. A previous purification of this kind is not
indispensable, and may be omitted altogether when the
ore is in plates or large grains ; but it is very desirable
when the ore is in fine powder, and greatly facilitates the
subsequent action of the oxidising mixture. By cutting
off- the top of a mercury bottle a wrought-iron crucible is
obtained, in which 600 grms. of osm-iridium may be fused
at one operation with potash and saltpetre as above.
There is usually little or no foaming, and if any occur it
may easily be checked by stirring with an iron rod. No
sensible quantity of osmic acid is given off during the
process, which with a little care is entirely free from
danger. In this manner 1,500 grms. of ore have been
worked up in a few hours in three successive operations.
The fused mass is to be broken into pieces with a hammer,
WOLCOTT GIBBS'S PROCESS. 805
and placed in a clean iron pot. Boiling water, contain-
ing about one-tenth of its volume of strong alcohol, is
then to be added, and the whole is to be boiled over an
open fire until the fused mass is completely disintegrated.
The potassium osmiate is, in this manner, reduced to
osmite, while the potassium rutheniate is completely
decomposed, the ruthenium being precipitated as a black
powder. It is advantageous, after boiling for some time,
to pour off the supernatant liquid with the lighter por-
tions of the oxides, and boil a second time with a fresh
mixture of alcohol and water. In this manner we obtain
a solution of potassium osmite, a large quantity of black
oxides, and a heavy black and coarse powder. This last
consists chiefly of undecomposed ore, mixed with a small
quantity of the iridium oxides, &c., with scales of iron
oxide from the crucible, and, if the ore has not been
previously purified, with the impurities of the ore itself.
The greater specific gravity of this residual mass renders
it very easy to pour off from it the mixture of black
oxides with the solution of osmite of potash and alkaline
salts. . This solution with the suspended powder is to be
poured into a beaker and allowed to settle. The heavy
black powder remaining in the iron pot is then to be
perfectly dried over the fire, and fused a second time with
potash and saltpetre as before. The fused mass is to be
treated exactly as after the first fusion. The heavy por-
tions remaining after this operation may be fused a third
time with the oxidising mixture. When, however, the ore
has been previously purified by fusion with sodium carbon-
ate, or when it was originally in the form of clean scales,
the heavy portion remaining after two successive oxidations
will be found to consist chiefly of scales of iron oxide.
The solutions containing potassium osmite and alka-
line salts are to be carefully drawn off by a syphon from
the black oxides which have settled to the bottom of the
Containing vessels. The oxides may then be washed with
hot water containing a little alcohol, and introduced into
a capacious retort. By this process, when carefully
executed, no trace of osmic acid escapes an advantage
800 THE ASSAY OF PLATINUM.
not to be despised, as the deleterious effects of this body
upon the lungs have not been exaggerated, and too much
care cannot be taken to avoid inhaling it.
The solution of alkaline salts contains only a portion
of the osmium in the ore. The other portion exists in the
mixture of oxide, and must be separated by distillation.
For this purpose the retort should be provided with a
safety-tube, passing through the tubulure, and with a
receiver kept cold, and connected by a wide bent tube
with a series of two or three two-necked bottles contain-
ing a strong solution of caustic potash with a little alcohol,
and also kept cold. All the tubulures and connections
must be made perfectly tight. Strong hydrochloric acid
is then to be cautiously poured into the retort, through
the safety-tube, in small portions at a time. The reaction
which ensues is often violent ; great heat is evolved, and
a portion of the osmic acid distils over immediately, and
condenses in the receiver in the form of colourless needles.
When a large excess of acid has been added, the action
entirely ceased, and the retort become cold, heat may
be applied by means of a sand-bath. The osmic acid
gradually distils over, and condenses in the receiver and
in the two-necked bottles. Especial care must be taken
that the beak of the retort is not too small at the ex-
tremity, as it may otherwise become completely stopped
up with the condensed osmic acid. The same applies to
the tubes which connect the receivers and two-necked
bottles. The distillation should be continued for some
time after osmic acid ceases to appear in the neck of the
retort ; when this has once become hot, the acid condenses,
and passes into the receiver in the form of oily drops.
When the distillation is finished, the retort is to be
allowed to cool, and then separated from the receiver,
which is to be immediately closed with a cork. By gently
heating the receiver in a water-bath, the contained osmic
acid may be driven over into the two-necked bottles,
where it condenses in the alkaline solution, and is reduced
by the alcohol to potassium osmite. The solution thus
obtained may be added to that obtained directly from the
NELSON w. PERRY'S PROCESS. 807
fused mass of ore, and, on evaporation in a water-bath and
cooling, will yield crystals of potassium osmite, the salt
being but slightly soluble in strong saline solutions. The
mother- liquor from the crystals contains only traces of
osmium, and may be thrown away as worthless.
The dissolved portions drawn off from the retort have
a very dark brown-red colour. The solution is to be
evaporated to dryness, re-dissolved in hot water and again
evaporated, after adding a little hydrochloric acid, and
this process repeated till no smell of osmic acid can be
perceived. A cold and saturated solution of potassium
chloride is then to be added in large excess. This dissolves
the iron and palladium chlorides which may be present,
leaving platinum, iridium, rhodium, and ruthenium as
double chlorides, insoluble in a strong solution of the
alkaline chloride.
The undissolved mass is to be well washed with a
saturated solution of potassium chloride, which is prefer-
able to sal-ammoniac. In this manner nearly the whole
of the iron and palladium may be removed, while any
insoluble impurities contained in the ore remain with the
mixed double chlorides.
For the separation of osmium from the other metals of
the group, the best plan seems to be the one which is
universally employed namely, the volatilisation of the
osmium, in the form of osmic acid.
Mr. Nelson W. Perry (' Engineering and Mining Jour-
nal ') proposes the following method for the assay of
platinum alloys containing base metal, silver, platinum,
gold, and osm-iridium.
Charge, platinum alloy 200 milligrammes, pure silver
150 milligrammes, or sufficient to produce perfect cu-
pellation.
Wrap charge in sheet lead and cupel. Weigh button.
Loss = base metal.
Flatten button, anneal, roll out thin, anneal again, and
make into cornet as in gold bullion assay. Introduce
cornet into parting flask and part with concentrated
sulphuric acid. Wash, anneal, and weigh. Loss from
808 THE ASSAY OF PLATINUM.
previous weight = silver in original alloy + silver added
for cupellation.
Alloy cornet with at least twelve times the amount of
silver that there is platinum present, and as before, form
cornet, and part first with nitric acid, sp. gr. 1*16, and
then nitric acid, sp. gr. 1-26. Wash thoroughly, anneal in
annealing cup, and weigh.
Treat residue with aqua regia, obtain gold by loss, the
residue is osm-iridium.
Time to complete assay in duplicate, 2 hours 45 min.
The quality of the silver added should at least be
sufficient, so that after the addition of the silver the alloy
will be to the gold as 3 : 1.
As platinum and osm-iridium add greatly to the in
fusibility of the compound, silver in sufficient quantity
must be added to prevent ' freezing ' and give a perfect
cupellation. Any large excess over these requirements is
to be avoided, first, because the residue, after parting, will
in that case be non-adherent, and in a more or less fine
state of subdivision, which .may occasion loss in washing
by decantation ; second, the larger the button cupelled,
the more difficult it is to obtain a good cupellation, and
the greater loss of silver during the process. It may,
for this reason, sometimes be necessary to use only 100
milligrammes of the alloy for assay instead of 200 milli-
grammes as above.
The cupellation should take place at a moderate
temperature, until near the ' blick,' when the assay should
be thrust back into the hottest part of the furnace to
prevent ' freezing.' The button must remain in the muffle
until all the lead is gone.
In parting with sulphuric acid boil for several minutes.
In other respects this operation is identical with -the gold
bullion assay. Any large excess of silver over twelve
times the amount of platinum in alloy is to be avoided, as
it causes the residue, after parting, to be too fine and float,
thereby occasioning loss in washing. Insufficient silver is
even worse, as the platinum will then be only incompletely
dissolved.
CHAPTEK XIX.
THE ASSAY OF BISMUTH.
THE following varieties of bismuth ores are met with, but
are somewhat rare :
Native Bismuth.
Bismuth Oxide.
Bismuth Sulphide.
Bismuth Persulphide.
Cupriferous Bismuth Sulphide.
Plumbo-cupriferous Bismuth Sulphide.
Plumbo-argentiferous Bismuth Sulphide.
Native Bismuth possesses a tolerably bright metallic
lustre ; its colour yellowish-white, often iridescent. It
fuses in the candle flame. It is generally found in small
amorphous lamellar masses, yet it occasionally occurs in
acute rhomboidal as well as cubical and octahedral
crystals.
This substance does not seem to form veins by itself,
but generally accompanies other minerals, particularly
those of cobalt, nickel, arsenic, and lead.
To within a recent period the chief source of the com-
mercial product has been native bismuth. But this limited
source is becoming well-nigh exhausted, whilst the de-
mand for this metal, especially in a great state of purity,
is increasing every day. It has thus become a matter of
necessity to look for fresh fields of exploration, for new
deposits : and as bismuth ores of every description mixed
up with other ores of various kinds are now used for the
extraction of bismuth, the assay of this metal has lost some
of its former simplicity. Mr. Hugo Tamm has done more
than any metallurgist towards perfecting the assay of
810 THE ASSAY OF BISMUTH.
bismuth, and from his papers on this subject in the
4 Chemical News,' Nos. 639 and 640, the following method
of assay is condensed :
Assaying of Bismuth Ores. Whenever the ore to be
tried is of a simple nature, is free from admixture with
other ores, and contains bismuth in the metallic state, or
in the state of sulphide, of oxide, or of carbonate, or, as
sometimes occurs, consists of a mixture of oxide, carbon-
ate, sulphate, and oxychloride, the assaying of bismuth is
reduced to the very simple operation of mixing the ore with
as fusible a flux as can easily be obtained, to which a
reducing substance, generally charcoal powder, is added
in proper quantity.
It is of course useless to lay down particular rules
concerning the nature or the quantity of the flux, and of
the reducing substance to be employed in this operation ;
indeed, it is not advisable to do so, and it is by far the
best to be guided by the nature of the materials at hand,
and by the results of a few trials with varied proportions
of flux and of the reducing agent ; the aim of the assayer
being the highest amount of metal that can be obtained in
a given instance. Still, one of the best fluxes, as well as
one of the most simple, consists of a mixture of two parts
of potassium or sodium carbonate, and one part of sodium
chloride, to which a proper amount of red argol or of
potassium cyanide on the small scale, and powdered
charcoal on the large scale, are added.
Assaying of Bismuth in Ores containing a large amount
of Copper. The problem of the direct separation of bis-
muth from ores containing large proportions of copper has
hitherto been one of difficulty, and its solution, which was
of considerable importance, offered great interest. The
difficulty consisted chiefly in the fact that both copper
and bismuth behave, in nearly every instance, in an
identical manner with docimastic reagents ; but Mr.
Tamm has happily hit upon a most simple and practical
means of effecting the direct separation of those two
metals.
The chief kinds of ores containing both bismuth and
ASSAYING BISMUTH ORES WITH MUCH COFFEE. 811
copper are the bismuth-copper pyrites or sulphuretted
ores, and the double bismuth and copper oxides or car-
bonates, or oxidised ores.
Both kinds of ores may be, and generally are, con-
taminated with other metals, but these foreign metals con-
stitute only, as a rule, a small fraction of the w r hole, and
the problem of their elimination will be given further on..
The reaction upon which the separation of bismuth
from copper is founded consists in the fact that, in pre-
sence of alkaline fluxes, carbonaceous reagents, and, of
course, among them carbon itself, reduce bismuth sul-
phide to the metallic state, while copper sulphide is not
reduced.
In the treatment of sulphuretted ores, both metals
being already in the state of sulphides, all that is required
is to run them down with a mixture of potassium carbon-
ate or soda and salt, to which a little flowers of sulphur
or ground sulphur, and charcoal or any other carbonaceous
substance, are added.
In this operation metallic bismuth is extracted quite
easily, and the metal thus obtained is tolerably free from
copper. It is recommended to add a little sulphur in
order to insure a complete sulphurisation of copper during
the whole of the operation, and especially to prevent any
desulphurisation of copper by the alkali, and, consequently,,
to prevent, as much as possible, this metal from being
reduced.
With oxidised ores the operation is very similar in
every respect to the one just described, and it differs from
it only in the amount of sulphur used, which is greater
in this instance, since the whole of the metals have to be
sulphurised.
Three parts of the ore are mixed with from two to
three parts of a flux composed of
Sodium carbonate .... 5 parts
Salt 2
Sulphur 2
Charcoal powder . . . . .1 part
Both the composition of the flux and the amount to be
used may be altered with advantage to suit each particular
812 THE ASSAY OF BISMUTH.
case. A few synthetical trials, in the hands of a person
accustomed to metallurgical operations, are all that are
required to make the best use of this reaction.
In general, it is to be observed that the amount of flux
and of reagents required for the assaying may be con-
siderably reduced when the operation is carried on on a
larger scale. On the other hand, it is scarcely worth
while mentioning that, in the operation of assaying, potas-
sium cyanide forms an admirable substitute for carbon.
During the process of extracting bismuth by means of
sulphur and carbon there is a loss of about 8 per cent, of
the bismuth contained in the ore. This loss is unavoidable,
but there is a more than proportionate loss of the metals
arsenic, antimony, and lead, which in this operation are
reduced with bismuth, and the crude metal obtained by
this process is not so impure as the corresponding metal
obtained by the direct reduction of the oxidised ores ;
besides the whole of the copper remains in the slag.
Whenever the sulphur-carbon process is employed, the
use of iron stirrers must be carefully avoided, for the reason
that copper sulphide is rapidly reduced to the metallic state
by this metal, especially in the presence of alkalies.
This process for the separation of bismuth from copper
will be found chiefly useful and important for the separa-
tion of bismuth in minerals containing large quantities of
copper. When, on the contrary, this metal exists only in
smaller proportions, it is more advantageous to run down
the whole of the metals, and to separate them afterwards
in the special operations of refining. But it is recom-
mended that the sulphur-carbon process be used for the
treatment of the somewhat abundant ores of bismuth
formed of bismuth and lead oxides, and small proportions
ofarsenious acid and antimonious acid, with a lit tie copper
oxide ; for there is as yet no direct means of smelting pure
bismuth from ores containing large proportions "of lead,
but it has been observed that bismuth' extracted by the
sulphur process contains less lead than the corresponding
metal obtained direct from the oxidised ore. The same
remark applies to arsenic and antimony, and this is in
PURIFICATION OF BISMUTH. 813
accordance with the behaviour of the sulphides of these
metals with alkaline sulphides.
Refining Crude Bismuth. The various ores of bismuth
above described, whether sulphuretted or oxidised, are
seldom formed of bismuth and iron only, or of only
bismuth, copper, and iron. They nearly always are con-
taminated by various proportions of lead, arsenic, or anti-
mony, metals which are reduced with bismuth, partially at
least, whatever process has been used for the extraction of
bismuth, and besides, the metal obtained by the sulphur
process from copper bismuth ores still contains a small
quantity of copper, which it is important to remove.
Bismuth extracted by any process is so generally free
from iron that no notice need be taken of this metal, which
remains wholly in the slags.
The fracture of good bismuth and that of its various
alloys is so characteristic that it is not often necessary to
have recourse to tests in order to determine what particular
processes will have to be used for the refining of the crude
metal.
Pure bismuth is tougher than most of its alloys. Its
fracture is bright, and it possesses a fine reddish colour.
Bismuth containing arsenic gives a beautiful fracture con-
sisting of large laminae of a whiter colour than that of pure
bismuth. Copper mixes with bismuth without alloying
with it, and is almost always discernible. The fracture of
bismuth containing antimony is dull and is mostly composed
of very small crystals. Lead does not prevent bismuth from
crystallising in large crystals, but these crystals are studded
all over with fine crystals. Sulphur imparts a black tinge
to metallic bismuth.
To these appearances, which almost suffice to an ex-
perienced eye, may be added a few simple tests.
It is difficult to detect arsenic in the presence of a large
quantity of bismuth by means of reagents, and the most
simple way of detecting this substance is to heat the bis-
muth on charcoal, with the oxidising flame of the blow-
pipe. Very small quantities of arsenic may be detected
by the garlic odour evolved.
814 THE ASSAY OF BISMUTH.
To detect copper the metal is dissolved in nitric acid,
the solution is supersaturated by ammonia, and filtered.
The blue colour of the filtrate indicates the presence of
copper.
When bismuth dissolves in strong nitric acid, with the
formation of a cloudy white precipitate which does not
disappear on the addition of w r ater, it is because antimony
is present.
When bismuth dissolves in strong nitric acid, with the
formation of a very white granular or crystalline precipi-
tate which dissolves freely on the addition of water, the
presence of lead is indicated.
But to detect with absolute certainty the presence of
even very small proportions of lead, the metal is dissolved
in nitric acid. The solution is supersaturated with am-
monia, and re-acidulated with the smallest amount of hydro-
chloric acid which will give a clear liquor. This liquor is
then precipitated by a large excess of boiling water. Water
must be added until no further precipitation takes place.
The whole is then filtered, and the filtrate is saturated with
a mixture of ammonia and ammonium carbonate ; when
a yellowish-white precipitate is formed it is because lead
exists in the bismuth.
It may be useful to submit the metal to be refined to
these various tests in order to ascertain beforehand which
refining process should be used. But it is essential to apply
each test to the refined metal, so as to verify its degree of
purity.
PURIFICATION OF THE KEDUCED BISMUTH.
Purification of Bismuth from Arsenic. The separation
of bismuth from arsenic is founded on the almost absolute
want of affinity of bismuth for iron, on the readiness with
which arsenic combines with iron, and on the fact that the
iron arsenide thus formed does not alloy with bismuth.
This operation is conducted in the following manner :
Bismuth is melted at a relatively high temperature, at
a bright red heat, under cover of borax or fiux, to avoid
loss of bismuth by volatilisation, and strips of iron are
PURIFICATION OF BISMUTH. 815
plunged into the molten metal. Iron is, according to the
technical expression, rapidly ' eaten away,' forming iron
arsenide, which rises to the surface of the metal.
When it is ascertained that fresh pieces of iron are no
longer attacked, the whole is allowed to cool. The iron
arsenide sets rapidly, and the bismuth, which is still fluid,
is poured out of the crucible into moulds. Singularly
enough, this process, which succeeds in perfection for the
separation of arsenic, is valueless when applied to the
separation of bismuth from antimony ; although, be it
noticed, the affinity of this metal for iron is very great.
Some antimony is removed by this process, but part of it
only, and it must be admitted that bismuth has as much,
or more, affinity for antimony than iron.
Purification of Bismuth from Antimony. The best way
of separating the two metals is to melt the alloy with a
quantity of bismuth oxide, equal to two and a half or three
times the weight of the antimony contained in the alloy.
The bismuth oxide is instantaneously reduced to the
metallic state, and antimony is liberated under the form of
antimony oxide, which combines with a little bismuth oxide,
and floats on the surface of the pure metal, whence it can
easily be removed.
This operation must be performed in clay crucibles, and
both carbon and iron must be carefully excluded, to avoid
any reduction of antimony oxide. The least traces of
antimony may be removed by this process without any
difficulty whatever.
Purification of Bismuth from Copper. When bismuth
ores contain only a small percentage of copper, and when
the ores are oxidised ores, it is advantageous to reduce
them at once by carbon and fluxes, without going through
the sulphurising process ; and, as a matter of course, all
the copper is alloyed with the bismuth.
On the other hand, bismuth extracted from copper
ores by the sulphur process contains, even in the best con-
ducted operation, a certain proportion of copper which
must be removed. This elimination has hitherto presented
very great difficulties, and could not be effected without
816 THE ASSAY OF BISMUTH.
losing a large amount of bismuth. Mr. Tamm has, how-
ever, devised the following method, by melting the alloy
with potassium sulphocyanide.
The sulphocyanide is prepared during the process of
refining, by mixing together eight parts of potassium cya-
nide and three parts of flowers of sulphur. One part of
this mixture is thrown on to sixteen parts of the metal
melted at a low temperature.
A reaction soon takes place, by which the mass of the
metal is brought to a bright red heat, and, at the same
time, the sulphocyanide begins to burn vividly, throwing,
in every direction, showers of scintillating sparks emitting
a blue light.
The crucible is covered over, and great care must be
taken to prevent the heat from rising above the burning
point of the sulphocyanide, a temperature at which bismuth
sulphide begins to volatilise.
The reaction is allowed to exhaust itself, and, when
all is quiet, and after the metal has been well stirred with
a clay stirrer (iron must be avoided), the flux is allowed to
set, and the metal, which is still fluid, is poured out into
moulds.
Purification of Bismuth from Sulphur. The metal ob-
tained in the above operation contains some sulphur. To
remove this substance, the metal is melted with iron or
carbon ; the separation is thus effected easily.
The several processes here proposed are chiefly useful
for the refining of bismuth alloyed with one metal.
There is no dry method of refining by one process
bismuth alloyed with several metals ; but the succes-
sive use of these different methods can safely be recom-
mended.
Copper should be removed first, for the reason that
some lead, antimony, and arsenic are eliminated at the same
time.
Bismuth should next be freed from antimony, and,
lastly, from arsenic and sulphur.
Herr Thtirach (' Journal fur Prakt. Chemie,' P. S. 14,
315) precipitates with sulphuretted hydrogen from a hot
VOLUMETRIC ASSAY OF BISMUTH. 817
solution, washes with hot water, heats for a considerable
time to 200-300 in a covered crucible, roasts in the open
crucible, and finally ignites strongly arid weighs as bismuth
oxide.
VOLUMETEIC ASSAY OF BISMUTH.
Mr. E. W. Pearson has given the following process for
the Yolumetric Assay of Bismuth.
Preparation of Standard Solution. '7135 grain of
pure crystallised potassium bichromate is dissolved in
100 grains of water. Call this solution the bichrome test
A. In a similar way, prepare a second solution, one-tenth
the strength of bichrome test A ; -07135 grain of potassium
bichromate, dissolved in 100 grains of water, will furnish
such a solution ; call it the bichrome test B. Bichrome
test C, one-tenth the strength of solution B, is also prepared
by dissolving -007135 grain of the potassium bichromate in
100 grains of water.
These figures can be multiplied to any convenient
number. These solutions will contain potassium bichro-
mate, in 100 grains of bichrome test A, equal to 1 grain of
bismuth ; in 100 grains of bichrome test B, equal to O'JL
grain of bismuth; and in 100 grains of bichrome test C,
equal to -01 grain of bismuth.
The bismuth should be in the form of nitrate, and the
solution kept hot during the experiment, as the precipi-
tated chromate collects more readily then; after com-
plete precipitation of the bismuth the solution will exhibit
a characteristic colour, produced by excess of potassium
bichromate.
By employing a standard solution of bismuth it has
been ascertained that 71*35 parts of potassium bichromate
are required to combine with 100 parts of bismuth.
For alloys of tin, lead, and bismuth (fusible metal) the
alloy, finely divided, is oxidised in a roomy flask with
nitric acid, the liquid is somewhat diluted, supersaturated
with ammonia, and digested for a long time and with -fre-
quent agitation with ammonium hydrosulphide, to which
3 G
818 . THE ASSAY OF BISMUTH.
a little sulphur has been added. The tin is thus dissolved
as sulphide. The lead and bismuth sulphides are filtered
off and washed with cold water. The liquid containing
the tin sulphide is slightly supersaturated with dilute sul-
phuric acid and very gently warmed, the vessel being
loosely covered with paper till the odour of sulphuretted
hydrogen has disappeared. The yellow tin sulphide is
washed on a filter with cold water, and treated as directed
for the analysis of tin ore.
The lead and bismuth sulphides are dried, detached from
the filter, which is incinerated with the usual precautions,
the whole digested with nitric acid, and the solution con-
centrated at last with the addition of hydrochloric acid till
it is reduced to a small bulk, and the greatest part of the
lead is separated as chloride. The proportion of acid must
be so large that the liquid is not rendered turbid by a
slight addition of water. Sulphuric acid is added, and the
whole let stand, but with frequent stirring. After the lead
chloride has been converted into sulphate a little alcohol
of O80 sp. gr. is added, and the whole brought upon
a weighed filter, where it is washed, first with alcohol
acidulated with hydrochloric acid, and afterwards with
water.
In the filtrate all the bismuth is precipitated as basic
chloride by the addition of water in large excess, and fil-
tered off (the filtrate being tested with sulphuretted hydro-
gen), washed with cold water, dried, and melted at a
moderate heat with four parts potassium cyanide in a
covered porcelain crucible (Eammelsberg).
Mr. M. Patteson Muir (< Chem. News,' April 27, 1877)
finds the following solution a most sensitive test for bis-
muth : 12 grms. crystalline tartaric acid, and 4 grms.
stannous chloride are dissolved in caustic potash, producing
a clear liquid of a decidedly alkaline reaction, which
should remain clear at 60 to 70 C. To the liquid to be
tested is added a considerable quantity of tartaric acid.
It is warmed and made alkaline with caustic potash. A
few c.c. of the stannous chloride solution (called from its
VOLUMETRIC ASSAY OF BISMUTH. 819
first discoverer Schneider's reagent) are added, and the
liquid warmed to 60 to 70 C. for a few minutes. If bis-
muth is present to the extent of one part in 210,000 parts,
a brownish-black colour is produced. Mercury must be
absent ; copper and manganese interfere slightly.
3 G 2
820
CHAPTER XX.
THE ASSAY OF CHROMIUM.
THE principal ore of this metal which occurs in commerce
is known as chrome iron, or chrome iron ore. It is found
in amorphous masses of a brownish-black colour, approach-
ing an iron grey. Its fracture is uneven, sometimes la-
mellar ; and its powder is greyish.
The two following analyses will give a general idea of
its composition :
Chromium oxide .... 36'0 43'7
Ferric oxide 37'0 34-7
Alumina . . . . . 21-5 20-3
Silica . . . .* ... 5-0 2'0
99-5 100-7
Assay of Chrome Iron Ore.
Chrome iron ore, like native tin oxide, is very difficultly
decomposable by ordinary reagents.
Dr. Genth, of Philadelphia, who has had much experi-
ence in the analysis of chrome-iron ore, gives the following
process ; it is very trustworthy, although long and some-
what tedious.
Of the chrome ore, reduced to an impalpable powder,
put 0-5 gramme in a platinum crucible about 2 inches
high, nearly If inch wide, and holding 52 grammes of
water, and place upon it 6 grammes of pure fused potas-
sium bisulphate, and heat with the greatest care for about
15 minutes, at a temperature scarcely above the fusing
point of the bisulphate ; then the heat is gradually raised,
but not higher than to make the bottom of the crucible
red hot, and kept at this temperature from 15 to 20
minutes. Never permit the mass to rise to half the height
ASSAY OF CHROME IRON ORE. 521
of the crucible. (If the fusion with potassium bisulphate
is done too rapidly a portion of the analysis is very apt to
be lost by spluttering, from the escape of sulphurous acid,
resulting from the oxidation of the ferrous oxide by the
sulphuric acid.) The mass begins now to fuse quietly,
and vapours of sulphuric acid go off more freely ; it should
then be kept at a red heat for about twenty minutes, and
the heat next raised as high as necessary to drive off the
second equivalent of sulphuric acid, and even to decom-
pose a portion of the iron and chromium sulphates. To
the fused mass add about three grammes of pure sodium
carbonate, and fuse the mixture, and then, by degrees,
keeping the temperature for about one hour at a dull red
heat, about the same quantity of saltpetre ; next heat for
fifteen minutes at a bright red heat. The fused mass is
dissolved in boiling water, filtered whilst boiling, and
washed with boiling water.
The insoluble residue, containing the greater portion of
the silicic acid, titanic acid, and alumina, the ferric oxide,
zirconia, and if the fusion has been conducted at a tem-
perature sufficiently high to convert the saltpetre into
caustic potash, and the above precautions have been used
! all the magnesia, is re-dissolved in dilute warm hydro-
chloric acid, which generally dissolves it readily and com-
pletely, and rarely leaves un decomposed ore behind ; but
if so, this residue must invariably be fused in a small
Crucible as before, adding, after the separation of the in-
soluble portion, the solution containing the small quantity
of chromic acid to the first filtrate.* (The certainly less
troublesome method, to deduct the insoluble portion from
the original weight, is bad ; such residues have never the
.composition of the original ore.) The filtrate contains the
whole quantity of the chromium as chromic acid, some-
times a trace of manganic acid, small quantities of silicic
acid, alumina, and rarely titanic acid. To this solution
add an excess of ammonium nitrate, and evaporate over a
water-bath nearly to dryness, and until all the liberated
ammonia has been expelled. The precipitate, remaining
ron addition of water, contains the silicic acid, titanic acid,
822 THE ASSAY OF CHEOMIUM.
alumina, and manganic oxide, which had gone into solution
with the chromic acid ; it is filtered off, and the filtrate
made strongly acid with sulphurous acid, carefully heated
to boiling, precipitated with a slight excess of ammonia,
boiled for a few minutes, and filtered. Dr. Genth says he
formerly acidulated the chromic acid solution by hydro-
chloric acid and then added sulphurous acid, but he several
times observed that, although an excess of sulphurous
acid had been used, a small portion of the chromic acid
escaped reduction, the filtrate from the ammonia pre-
cipitate being yellow. He has in vain tried to find the
reason for this singular behaviour. Since using sul-
phurous acid only, he has never been troubled with any-
thing similar.
It is exceedingly difficult to wash out the chromic
oxide ; it succeeds best in the following way : After the
precipitate has settled, the clear liquid is passed through
the filter, then boiling water is added to the precipitate,
and, after settling, the supernatant liquid is filtered ; the
precipitate then is put on the filter, and washed twice or
three times with boiling water ; it is then washed back
again into the dish and boiled with water until the little
lumps which clog together are completely broken up, and
it is then filtered again, and this operation repeated until
the wash- waters do not show the presence of any sul-
phates when tested with barium chloride. The precipi-
tate is then dried and burned. No matter how well it
may have been washed, it almost invariably contains
minute quantities of alkalies, in the presence of which a
little chromic oxide is converted into chromic acid. The
ignited precipitate is therefore put into a dish, boiled
with water, a few drops of sulphurous acid added, precipi-
tated by ammonia, filtered, washed, dried, ignited, and
weighed.
In this manner the chromic oxide is obtained quite
pure, and repeated analyses of the same sample of ore
never vary 0-25 per cent, of chromic acid.
Mr. O'Neill uses a volumetric method to estimate the
chromic acid, depending upon the capability of sulphurous
WOLCOTT GIBBS'S PEOCESS. 823
acid to deoxidise chromic acid at the ordinary temperature
in the presence of free sulphuric acid. Prepare a strong
solution of sodium bisulphite, by passing sulphurous acid
through caustic soda to saturation, and then make it
alkaline with caustic soda, so as to have a neutral sulphite,
which is less readily oxidised by keeping than the bisulphite.
Use a dilute solution made from this concentrated sulphite^
of such a strength that one grain of pure potassium bichro^
mate requires about, and not less than, 200 grains measure
of the sulphite to deoxidise it. The value of the sulphite
must be estimated for every operation, since it is con-
tinually absorbing oxygen. This is done twice by weighing
out three grains, and four grains of pure bichromate, dis-
solving each of them in ten ounces of water and acidulating
freely with sulphuric acid, then adding the sulphite from
the burette, with continual stirring, until the chromic acid
is destroyed. The stopping-point may be ascertained by
the colour when one is accustomed to the reaction, but
even an experienced eye will often be glad of additional
evidence. A mixture of potassium iodide and boiled
starch, slightly acidulated, forms a delicate test ; it has
usually a faint colour, which is even preferable to a colour-
less mixture. An exceedingly small quantity of chromic
acid develops the blue colour in spots of this mixture,
and a very slight excess of sulphite makes it colourless.
One division of the sulphite test-liquor, or 0-05 grain of
potassium bichromate in twelve ounces of water, easily
and quickly influences the test mixture. The chr ornate
from the chrome ore is tested in the same manner, and the
quantity of chromium oxide or chromic acid calculated
from the equivalents of potassium bichromate. A tenth
of a grain more than five is taken to allow for all losses,
and the results are multiplied by 20 for the percentage.
151 of potassium bichromate is reckoned equivalent to 80
of green oxide of chromium, and 104 of chromic acid.
An estimation can be made by this process in three
or four hours, and a double estimation in a little longer
time.
Dr. Wolcott Gibbs has shown that chrome-iron ore
824 THE ASSAY OF CHROMIUM.
may be completely resolved by fusion with fluohydrate of
potassium fluoride. In this and in all similar applications
of the fluohydrate it is best to evaporate the finely pul-
verised mineral to dryness with a concentrated solution of
the salt. On subsequently heating to low redness, the
resolution of the mineral is effected with the utmost ease,
a portion of the chromium being usually oxidised to
chromic acid by the oxygen of the air. After expelling
the fluorine by heating the fused mass with sulphuric acid,
the remaining mass is dissolved in water, rendered nearly
neutral by a solution of sodium carbonate, and sodium
acetate is added in excess. A current of chlorine gas, or
a solution of chlorine water, then readily converts the
whole of the chromium present into chromic acid, espe-
cially when the solution is hot, and when it is kept nearly
neutral by the occasional addition of sodium carbonate.
The excess of chlorine is easily got rid of by boiling.
The iron and alumina may then be precipitated to-
gether by boiling the solution in the presence of excess of
sodium acetate. It is more convenient and equally accu-
rate to neutralise the solution with ammonia, separate the
ferric oxide by filtration, and estimate the chromium in
the filtrate by reduction and precipitation with ammonia.
For the technical estimation of chromium in chro-
mite, Mr. Clarke fuses with cryolite and potassium bisul-
phate. The mass is then treated with a little strong
hydrochloric acid, and allowed to digest for about ten
minutes. Then upon boiling with water the whole dis-
solves. The solution should then be neutralised, sodium
acetate added, and the chromium oxidised to chromic acid
by a current of chlorine gas, or by boiling with sodium
hypochlorite solution. The chromium may then be sepa-
rated from other substances, as directed by Dr. W. Gibbs.
When chromite is fused with potassium bisulphate and
cryolite, and saltpetre is added to the mass, as soon as clear
fusion is obtained, the chromium is nearly all oxidised to
chromic acid. If the mass be boiled with a solution of
sodium carbonate, and the liquid filtered, a filtrate is
obtained which contains nearly all, but not quite all, the
chromium as alkaline chromates, free from iron or alumina ;
ESTIMATION OF CHROMIUM IN CHROME IRON ORE. 825
but invariably the residue upon the filter contains traces
of chromium. When chromite is fused with the acid
potassium fluoride, a part of the chromium is usually
oxidised to chromic acid by the oxygen of the air ; and in
one case that came under Mr. Clarke's observation, when
he heated the resulting mass with sulphuric acid, red fumes
were given off, which were probably the so-called chro-
mium terfluoride.
When potassium bisulphate alone is used for the de-
composition of chromite, &c., it is necessary that the
mineral should be reduced to extremely fine powder ; but
when the mixture of bisulphate and fluoride is employed,
although the mineral should be in fine powder, such an
extreme state of subdivision is by no means required, and
thus much labour is saved.
Estimation of Chromium in Chrome Iron Ore. When
an estimation of the chromium only is sought, the decom-
position of chrome iron ore, according to H. N. Morse
and W. C. Day, can be best accomplished by fusing the
material with potassium hydroxide in a wrought-iron
crucible.
The method here described has, without exception,
given satisfactory results. From 6 to 10 grms. of potas-
sium hydroxide are placed in a wrought-iron crucible
(having the form of the ordinary porcelain crucible and a
capacity of about 100 c.c.) and gently heated until the
evolution of steam ceases and the fused mass becomes
tranquil. After cooling, the finely pulverised material,
weighing not more than 0-5 grm., is placed upon the
potassium hydroxide and evenly distributed over the
surface. A flame just sufficient to thoroughly fuse the
alkali is applied to the uncovered crucible, and the con-
tents, as long as they remain in a fluid condition, fre-
quently stirred with a piece of iron wire, which is allowed
to remain in the crucible. The decomposition progresses
rapidly, and the potassium hydroxide together with the
soluble products of the decomposition soon begins to rise
upon the sides of the crucible, where it deposits itself in
forms somewhat resembling the cauliflower. Within two
or three hours the decomposition is complete, and the
826 THE ASSAY OF CHEOMIUM.
bottom of the crucible becomes dry. The crucible is
then turned upon its side and the temperature of its under
surface raised to a dull red heat. The incrustation on the
interior of the crucible does not fuse at this temperature,
but becomes rapidly yellow, owing to the oxidation of
the chromium to chr ornate. At the end of two or three
hours the oxidation is perfect. Portions of the incrusta-
tion retain a greenish colour, however long the heating is
continued ; but this is due to the presence of iron or
manganese, and not to unoxidised chromium.
After cooling, the crucible is placed in a porcelain
evaporating dish and the contents removed by means of
hot water. The solution, which at first has a greenish
appearance owing to the presence of iron dissolved in the
caustic potash, is heated for some time in order to effect
complete precipitation of the iron. The filtrate, which
has a clear yellow colour, is rendered slightly acid with
pure dilute nitric acid, the aluminium precipitated with
ammonia and washed by decantation. The potassium
chromate is then reduced and the silica rendered insoluble
by evaporating to perfect dryness with an excess of hydro-
chloric acid. The residue is moistened with hydrochloric
acid and treated with water.
It only remains to separate the chromium in the
filtrate from magnesium and to estimate it as chromic
oxide. To do this the authors prefer in each instance to
first precipitate with barium carbonate.
We give below the data of twelve estimations upon
material whose origin is unknown to us. The analyses
reported were in each case made consecutively.
Wt. of Ore Taken. Wt. of Cr 2 3 Found. Percentage.
1. 0-3250 0-1343 41-32
2. 0-3465 0-1433 41-35
3. 0-3248 0-1343 41-36
4. 0-3383 0-1392 41-14
5. 0-2723 0-1120 41-12
6. 0-3955 0-1629 41-19
7. 0-3311 0-1358 41-02
8. 0-31285 0-1290 41-23
9. 0-3387 0-1392 41-10
10. 0-2995 0-1235 41-25
11. 0-3606 0-1478 . 41-00
12. 0-3007 0-1236 41-11
ESTIMATION OF CHROMIUM IN IRON AND STEEL. 827
Estimation of Chromium by means of Standard Solu-
tion. This process is the converse of the estimation of iron
by means of solution of potassium chromate.
The chrome ore is treated with potassium nitrate and
sodium carbonate, as above described ; and the solution of
potassium chromate so obtained has an excess of hydro-
chloric acid added to it.
It is stated at page 324, under the head of Assay of
Iron in the Wet Way, that 100 parts of metallic iron cor-
respond to and are represented by 88-6 grains of potassium
bichromate. Now 88-6 grains of potassium bichromate
contain 32-96 grains of chromium ; therefore 100 grains of
iron are equal to 32-96 of chromium. From these data a
standard solution can be readily made : thus Dissolve 50
grains of pianoforte wire in excess of hydrochloric acid ;
place the solution in a burette, and fill up to 100 on the
instrument with water, and well mix : it is now evident
that every division of the burette will equal or represent
0-1648 grain of chromium. The assay is now thus proceeded
with : Gradually add the standard solution of iron to the
solution of potassium bichromate acidulated with hydro-
chloric acid, until a drop of the solution mixed with a drop
of solution of potassium ferrocyanide gives a pale blue
colour : a slight excess of ferrous oxide is then present,
showing that all the chromic acid has been reduced to the
state of chromium oxide. Now observe how many divi-
sions of the iron solution have been required, and multiply
them by -1648 : the resulting number will represent the
amount of metallic chromium in the sample submitted to
assay.
For the estimation of chromium in iron and steel
Mr. J. 0. Arnold (' Chem. News,' Dec. 10, 1880, p. 285)
weighs out from 1 to 5 grms. of steel (in drillings) accord-
ing to the amount of chromium present (this may be
ascertained by a rough colorimetric test). Place the
metal in a wide, covered beaker, and add 20 c.c. of strong
hydrochloric acid ; heat till all action is at an end ; rinse
the cover and sides of the beaker from splashings, and eva-
porate the solution gently to complete dryness. If the
g28 THE ASSAY OF CHEOMIUM.
evaporation has not been too rapid, the chlorides may be
almost entirely detached from the bottom of the beaker in
a brittle cake. This is broken up into small pieces by means
of a platinum spatula, and as much as possible is brushed
out into a clean dry porcelain dish. A small quantity of
the chlorides will, however, remain adhering to the beaker :
this may be removed with 2 or 3 c.c. of dilute hydro-
chloric acid. The solution is poured into a deep platinum
crucible, the beaker rinsed with 1 or 2 c.c. of water, the
washings added to the crucible, the contents of which are
now evaporated to dryness on a sand-bath. When dry
the main quantity of the chlorides is carefully brushed out
of the porcelain into the platinum dish, and is reduced to
a fine powder by means of a little pestle made from a glass
rod. The spatula and pestle are cleaned into the crucible.
The finely divided chlorides are now thoroughly mixed
with an excess of fusion mixture (1 part dry sodium car-
bonate to 1 part powdered nitre), a cover is placed over
the crucible, and its contents are fused over a gas blow-
pipe till quite liquid. By this treatment the iron is con-
verted into insoluble oxide, the manganese, silica, and
chromium respectively into alkaline manganate, silicate,
and chromate. When cool the crucible is placed in a
beaker containing 80 c.c. of boiling water, and is gently
boiled till the fused mass is detached and dissolved out.
This may be assisted by occasional stirring with a glass
rod. When clear from oxide, remove the crucible and
cover, washing them well with hot water. Add 3 or 4
drops of alcohol to decompose the manganate, and allow
the iron and manganese oxides to settle thoroughly.
When the supernatant liquid is quite clear it is decanted
on a close double filter, the filtrate being received into a
clean beaker. The precipitates are disturbed as little as
possible. When all the clear liquid has passed through, the
filter is well washed with hot water. The precipitates are
now washed twice by decantation with 30 c.c. of hot water ;
at the second washing the contents of the beaker are allowed
to drain gently into the filter, which is now allowed to
drain thoroughly, and is removed without further washing.
VOLUMETRIC ESTIMATION OF CHROMIUM. 829
These precautions in washing must be strictly carried out,
as the ferric oxide is in such an exceedingly fine state of
division that any attempt to wash or disturb it on the
filter will inevitably cause some of it to pass through into
the chromium solution. The clear yellow filtrate contains
chromium and silica, to it is added 20 c.c. of hydrochloric
acid, the cover being kept on the beaker, to prevent the
projection of the solution by the evolved carbonic acid.
The solution is now well boiled until all the carbonic acid
and nitrous fumes are driven off. Its colour will now be
green, owing to reduction to chloride. Dilute ammonia
is added until alkaline, and the solution heated nearly to
boiling. The resulting precipitates consist of chromium
silicate mixed with alkaline salts. It is collected on a
filter (previously well washed with hot dilute hydrochloric
acid to free it from iron), and is slightly washed. When
the washings have drained through, the precipitate is dis-
solved off the filter with hot dilute hydrochloric acid, the
filtrate being received into the beaker in which the precipi-
tation took place. The solution is now evaporated to dry-
ness to relider the silica insoluble. The soluble portion is
taken up with 10 c.c. hydrochloric acid and 90 c.c. of water,
and is filtered through a washed filter into a clean beaker,
the filter being well washed. The precipitation is now
repeated as above, and the chromic hydrate comes down
free from silica and alkaline salts. It is collected, washed,
dried, ignited, and weighed as chromic oxide.
Precautions. Only three or four drops of alcohol
should be added, as this quantity is sufficient, not only to
precipitate the manganese, but also to effect the reduction
to chloride. If too much is added organic compounds are
formed, which tend to prevent the complete precipitation
of the hydrate. The ammonia in the last precipitation
should be added in the least possible excess, and the solu-
tion should be heated gently nearly to boiling. If any
great excess of ammonia be present, and if the solution is
boiled, the glass of the beaker is attacked and the result is
high. The method, if properly carried out, is accurate.
Mr. W. J. Sell has devised the following method for
830 THE ASSAY OF CHROMIUM.
the volumetric estimation of chromium. The solution,
containing chromium acidified with sulphuric acid, is
boiled, and a dilute solution of permanganate added to the
boiling liquid until a purplish tint remains after boiling for
three minutes. The solution is then rendered slightly
alkaline with sodium carbonate, alcohol is added, and the
manganese filtered off. The chromic acid in the filtrate is
estimated, by titration with iodine and sodium hyposul-
phate. The author has successfully applied the method
to the estimation of chromium in chromic iron ore. He re-
commends the following plan for effecting its decomposition.
The chromic iron ore is placed on the top of about ten times
its weight of a mixture, made in the proportion of one
molecule of well-fused and powdered sodium bisulphate
to two molecules of sodium fluoride, and the whole is
ignited for fifteen minutes. An amount of sodium bisul-
phate is now added equal to that of the mixture taken,
and when thoroughly fused a further addition of an equal
quantity of bisulphate is made, the mass fused, and then
rapidly cooled. The fused mass so obtained dissolves
completely in boiling water acidified with sulphuric acid.
In this way an estimation can be made in an hour and a
quarter.
831
CHAPTER XXI.
THE ASSAY OF AKSENIC.
THE minerals from which arsenic is produced are the
following :
Native arsenic.
Arsenical pyrites, FeS 2 + FeAs, containing 46-6 As and 19-6 S.
Arsenical pyrites, Fe 4 As 3 , containing 66'8 As.
Speiskobalt (Co,NiFe), As.
Glanzkobalt, CoS 2 + CoAs,
Coppernickel, NijAs.
Nickel and cobalt arsenical pyrites (Co,Ni,Fe)S 2 + (Co,Ni,Fe), As,
White arsenical nickel, NiAs; Tennantite (Cu2S,SeS) 4 ,AsS 3 .
Eealgar, AsS 2 and yellow arsenic AsS 3 .
Assay of Arsenic. 50 grains of the finely pulverised
mineral are deflagrated with 200 of potassium nitrate and
200 of sodium carbonate in a porcelain crucible. When
the crucible is cold, it and its contents are to be treated
with water, as in the case of chromium. The solution will
contain potassium arseniate, and (if the ore had in its con-
stitution sulphur, which is most likely) potassium sulphate.
Lead nitrate must be added to the solution (made neutral
with nitric acid, if requisite) : a mixture of lead arseniate
and sulphate is precipitated ; this precipitate is well washed
on a filter, and digested with dilute nitric, acid ; this agent
dissolves out the lead arseniate, and leaves the sulphate.
Filter, and saturate the filtered solution with soda, which
will throw down the arseniate ; this must be collected on
a filter, washed, dried, and weighed. Every 100 parts cor-
respond to 22-2 of metallic arsenic, or 29 parts of arsenious
acid (the white arsenic of commerce).
This method is only approximative : the following is
the better plan :
Digest the ore in strong nitric acid until nothing more is
832 THE ASSAY OF AESENIC.
taken up (the action may be facilitated by the occasional
addition of a crystal or two of potassium chlorate), and all
action on the addition of fresh acid is at an end : dilute
with water, and filter : to the filtered solution add lead
nitrate, and proceed as above.
For estimating the amount of arsenic in ores, Mr.
Parnell says that the neatest, simplest, and most accurate
mode of procedure is to heat the finely divided sample in
a gentle stream of chlorine gas to a temperature of about
200 C., and to collect the escaping arsenic chloride in
chlorine-water. If free from antimony, the liquid may be
well boiled, to expel free chlorine, and the arsenic preci-
pitated with sulphuretted hydrogen, and weighed as penta-
sulphide.
In cases where the arsenic is obtained in the form of
arsenio-magnesian phosphate (as in the separation of the
metal from antimony or copper), the most accurate plan
would be to dissolve the precipitate in hydrochloric acid,
and precipitate the arsenic as pentasulphide. When the
amount of arsenic is small, it may be weighed as the
double arseniate. The sample should not, however, be
dried at a higher temperature than that of an ordinary
water-bath namely, about 95 C. Perfectly accurate
results could, no doubt, be obtained by drying the pre-
cipitate over sulphuric acid, when it retains its six equi-
valents of water. The only objection is that it would
take many days for a filter containing a precipitate to be
properly dried by this means.
833
''*))
CHAPTER XXII.
THE ASSAY OP MANGANESE.
THE following are the commercially valuable minerals
containing manganese.
Pyrolusite, MnO 2 , containing 18-0 p.c. of available oxygen
Braunite, Mn 2 3 , 10-0
Manganite, Mn 2 O 3 , 9'0
Varvicite, Mn" 2 O 2 + Mn,0 3 ,HO, 13-8
Hausmannite, MnO + Mn 2 6 3 , 6*8
Psilomelane, Mn 2 O 3 .
The assay of this metal is confined to the amount of
peroxide any one of its ores may contain. There are
several methods of effecting this, and the best of these will
be described below.
Valuation of Manganese Ores. The best methods used
for the valuation of manganese ores are not necessarily
those which give in the most rapid and accurate manner
the absolute amount of manganese peroxide present in the
ore. The analyst must bear in mind that the commercial
value of manganese ore depends on its power of liberating
chlorine from hydrochloric acid ; and it not unfrequently
happens that an ore, which on accurate analysis would be
reported to contain a high percentage of manganese per-
oxide, likewise contains some other mineral (iron protoxide
or magnetic oxide), which will materially reduce the value
of the manganese as a chlorine-yielding ore. It is on this
account that some processes excellent though they be
from a purely analytical point of view have fallen into
discredit amongst manufacturers, whilst other processes
which do not profess to give the amount of manganese per-
oxide actually present, but only that available for liberating
chlorine, are now generally adopted. In the following
3H
834 THE ASSAY OF MANGANESE.
pages are given the methods of testing manganese ore for
the available peroxide which have best stood the test of
practical experience.
Messrs. Scherer and Kumpf, after examining all the most
approved methods in Dr. Fresenius's laboratory at Weis-
baden, have come to the conclusion that Bunsen's method
is the best for rapidly giving the amount of available man-
ganese in an ore. This process is carried out by dissolving a
weighed quantity of the sample in strong hydrochloric acid
in a small flask, until complete decomposition has taken
place. The escaping chlorine is received in a strong solution
of potassium iodide, and the liberated iodine subsequently
estimated by means of a standard solution of sodium hypo-
sulphite and a solution of starch. To prevent the solution
of potassium iodide from being sucked back into the gene-
rating-flask, a few small pieces of magnesite are introduced
with the manganese, so that a continual slight escape of
carbonic acid takes place through the solution. The solu-
tion of sodium hyposulphite is tested by means of carefully
prepared pure iodine, dissolved in potassium iodide. The
solution should be of such a strength that 1000 c.c. of
sodium hyposulphite solution corresponds to from 2 to
3 grms. of manganese peroxide. In this estimation the
iodine liberated by the chlorine should be tested as soon
as possible after the decomposition ; it gives higher results
after standing 24 hours than before. These higher results
are caused by the liberation of iodine by spontaneous
decomposition of hydriodic acid, set free by the hydro-
chloric acid, distilled over during the process. The follow-
ing experiment proves this : A few drops of hydrochloric
acid were added to a solution of potassium iodide. The solu-
tion remained for some hours colourless, but, after standing
twenty-four hours, had become quite yellow, and was found
to contain free iodine sufficient to indicate 8 per cent, of
manganese peroxide when titrated with hyposulphite.
Messrs. Scherer and Eumpf have made the suggestion
that the value of manganese ores should be measured by
chlorometrical degrees rather than by the actual percentage
of binoxide ; thus tending in the same direction as the
VALUATION OF MANGANESE ORES. 835
resolution* passed by the Association of Alkali Manu-
facturers in 1869, in reference to this subject a decision
which would seem also "to indicate a desire on the part of
manufacturers that tests of manganese ore should express
the amount of peroxide available for liberating chlorine,
and not the amount actually present in the ores.
For the above reasons, Dr. Paul adopts Mohr's method
of using a known quantity of a standard solution of oxalic
-acid, together with excess of sulphuric acid, for dissolving
the ore ; if necessary, boiling until the ore is completely
dissolved and then, by means of a standard solution of
permanganate, estimating the quantity of oxalic acid
remaining undecomposed. This method is very conve-
nient for testing manganese ores, and involves only one
weighing for each test. The results obtained are also
very uniform.
This method has also the advantage of giving results
which fairly represent the amount of available peroxide
in manganese ores ; for any iron that may be present as
metal or protoxide would consume an equivalent quantity
of permanganate solution, and thus apparently reduce the
quantity of oxalic acid decomposed by the peroxide to
an extent proportionate to the amount of iron existing in
the ore. Thus, for instance, if the quantity of oxalic acid
decomposed by 100 grains of manganese ore free from
iron or ferrous oxide were 109;53 grains, the ore would
contain 76*5 per cent, of peroxide, and the whole of that
would be available. But if the 100 grains of ore also
contained 5*6 grains of metallic iron, or an equivalent of
protoxide, the permanganate solution required for per-
oxidising that iron would represent 6-3 grains of oxalic
acid, and the quantity of oxalic acid decomposed by the
peroxide would appear so much less than it really was, or
103-23 grains instead of 109-53 grains. Accordingly, the
amount of peroxide would be represented as 72-1 per
cent, instead of 76-5 per cent. : and that would, in fact,
* ' That as the testing of manganese according to the method of Will and
Fresensius is, in the opinion of the meeting, incorrect, and yields uncertain
results, it is recommended to members of this association not to buy by that test.'
3 H 2
836 THE ASSAY OF MANGANESE.
be the amount of peroxide available for generating
chlorine.
This method of testing recommends itself by its sim-
plicity, and by the fact that the standard solutions of
oxalic acid and permanganate will keep for a long time
without alteration of value. The oxalic acid solution
contains 63 grms. in the litre, and 1 c.c. is equivalent to 5
c.c. of the permanganate solution.
Mr. John Pattinson prefers a modification of Otto's
process for the valuation of manganese, which consists in
boiling the ore with a known weight of a proto-salt of iron,
and then estimating the excess of iron by a standard
solution of potassium bichromate. This modified process,
in his opinion, requires less skill and care at the hands
of the operator than Bunsen's method, as described by
Messrs. Scherer and Eumpf. 30 grs. of clean iron wire
are placed in a 20-oz. flask, along with 3 oz. of dilute
sulphuric acid, made by adding 3 parts of water to 1 of
oil of vitriol. A cork through which passes a tube bent
twice at right angles is inserted in the neck of the flask,
and the flask is heated over a gas-flame until the iron is
dissolved. The bent tube is placed so as to dip into a
small flask or beaker containing a little water. When the
iron is quite dissolved, 30 grs. of the finely pounded and
dried sample of manganese ore to be tested are put into
the flask, the cork replaced, and the contents again made
to boil gently over a gas-flame until it is seen that the
whole of the black part of the manganese is dissolved.
The water in the small flask or beaker is then allowed to
recede through the bent tube into the larger flask, more
distilled water is added to rinse out the small flask or
beaker and bent tube, the cork well rinsed, and the con-
tents of the flask made up to about 8 or 10 oz. with dis-
tilled water. The amount of iron remaining unoxidised
in the solution is then ascertained by means of a standard
solution of potassium bichromate. The amount the bi-
chromate indicates, deducted from the total amount of
iron used, gives the amount of iron which has been
peroxidised by the manganese ore, and from this can be
VALUATION OF MANGANESE ORES, 837
calculated the percentage of manganese peroxide contained
in the ore. Thus, supposing it were found that 4 grs. of
iron remained unoxidised, then 30 4=26 grs. of iron,
which have been oxidised by the 30 grs. of ore. By a
simple calculation it is found that this 26 grs. of iron
are equivalent to 20-43 grs. of manganese peroxide, the
amount of peroxide in the 30 grs. of ore. The percent-
age is, therefore, 68*10.
It must be understood that an ore may contain a mix-
ture of one atom of manganic oxide, and two atoms of
magnetic iron oxide, or 27'3 per cent, of the former, with
72 -7 per cent, of the latter ; in such a mixture the method
of Fresenius and Will will indicate with precision the
amount of manganese peroxide, but on adding hydrochloric
acid to this mixture not a trace of chlorine will be given
off, since the free atom of oxygen of the manganese
peroxide is just sufficient for the oxidation of the 2 atoms
of iron protoxide of the magnetic iron ore ; in the same
way a mixture of manganese binoxide with iron protosul-
phate or protocarbonate of that metal will be perfectly
worthless as an article for chlorine-making use.
Dr. Mohr accordingly recommends that manganese
ores and samples of manganese peroxide should be
always tested, previous to analysis, with an astatic mag-
netic needle, and he further recommends Dr. Bunsen's
process (given on p. 834) as the best and surest method
of analysis. This process is really the same as that which
the manufacturer employs for making chlorine ; any
magnetic iron ore present will become oxidised in both
processes, and a special examination for magnetic iron
oxide is rendered unnecessary, while the available manga-
nese for the production of chlorine only is estimated.
Mr. J. Pattinson proposes the following method of
precipitating manganese entirely as peroxide, and applies
it to the volumetric estimation of manganese. He
finds that the whole of the manganese in a solution of
manganous chloride can be precipitated as peroxide, if a
certain amount of ferric chloride be present, by a suf-
ficient excess of a solution of chloride of lime or bromine
838 THE ASSAY OF MANGANESE.
water, adding, after heating the solution to from 140 to-
160 F., an excess of calcium carbonate, and then well
stirring the mixture. Zinc chloride may be substituted
for ferric chloride. The author recommends the follow-
ing solutions, etc. The clear liquid obtained by decanta-
tion from a 1*5 per cent, solution of bleaching-powder ;
light granular calcium carbonate obtained by precipitating
an excess of calcium chloride by sodium carbonate at
180 F. ; a 1 per cent, solution of ferrous sulphate in
dilute (1 in 4) sulphuric acid ; standard solution of potas-
sium dichromate equivalent to 1 part of iron in 100 of
solution. The application of the process to manganiferous
iron ores is as follows : 10 grains of the ore, dried at
212, are dissolved in a 20-oz. beaker in about 100 fluid
grains of hydrochloric acid (sp. gr. 1*18). Calcium
carbonate is then added, until the free acid is neutralised
and the liquid turns slightly reddish. 6 or 7 drops of
hydrochloric acid are now added, and 1000 grains of
the bleaching-powder solution, or 500 grains of saturated
bromine water, and boiling water run in until the tempe-
rature is raised from 140 to 160 F. ; 25 grains of calcium
carbonate are added, and the whole well stirred. If the
supernatant solution has a pink colour, the permanganate
is reduced by a few drops of alcohol. The precipitated
oxides of iron and manganese are filtered off and washed.
1000 grains of the acidified ferrous sulphate solution are
carefully measured into the 20-oz. beaker already used ;
the filter, with its washed contents, added. A certain quan-
tity of the ferrous sulphate is oxidised by the manganese*
binoxide ; this quantity is estimated with the standard
dichromate solution, when the quantity of manganese bin-
oxide can easily be calculated. The iron present must be
at least equal in weight to the manganese during the pre-
cipitation, in order to insure the absence of lower oxides.
839
CHAPTEE XXIII.
THE ASSAY OF NICKEL AND COBALT ORES.
ORES OF NICKEL.
Nickel oxide.
Nickel sulphide.
Nickel arsenide ; kupfernickel.
Nickel arsenio-sulphide ; grey nickel.
Nickel antimonio-sulphide.
Nickel arseniate.
Nickel arsenite.
Nickel silicate.
ORES OF COBALT.
Cobalt oxide.
Cobalt sulphide ; cobaltine.
Cobalt sulphate.
The cobalt arsenides.
Arsenio-sulphide, or grey cobalt.
Cobalt arsenite.
THE analysis of cobalt ores is the most tedious, with the
exception of those of platinum, of any that fall under the
assayer's notice, the greatest difficulty being in the sepa-
ration of cobalt and nickel. The following is Mr. Hadow's
process :
The only examination which the ore need undergo
previously to the solution of a weighed quantity is with the
view of obtaining a rough idea as to the amount of arsenic
and cobalt or nickel present in the sample ; for this purpose
a little may be roasted on charcoal, or ignited in a tube?
to see whether arsenic readily sublimes ; another portion,
of a few grains' weight, may be dissolved in aqua regia in a
test-tube, when the depth of the blue or green colour will
serve as an indication of the degree of richness of the ore
in cobalt and nickel.
If the ore is rich, from 20 to 30 grains ; if poor, from
840 THE ASSAY OF NICKEL AKD COBALT.
50 to 100 grains, in a state of fine division, are weighed
out for the analysis. If much arsenic has been found, the
portion, after weighing, had better be ignited in a small
porcelain capsule or crucible over a gauze burner, when it
generally ignites and smoulders away, evolving abundance
of arsenious acid. The powder ready for solution is trans-
ferred to a small 4-oz. flask by means of glazed letter-paper
and a camel's-hair paint-brush to sweep in the last par-
ticles ; the mouth of the flask is then partially closed by a
small funnel placed to catch the drops projected during
solution. The ore is then drenched with hydrochloric
acid, nitric acid being added from time to time, until all
heavy metallic- looking particles are found to have disap-
peared from the bottom of the flask. The solution may
then be decanted from the insoluble matters into a half-
pint beaker, together with the washings of the flask ; and
as sulphur frequently remains, entangling portions of undis-
solved ore, it is advisable to transfer the undissolved residue
fr6m the flask into a capsule, drying and igniting the con-
tents of the latter, and then digesting again the ignited
matters in a little more aqua regia ; the whole of the latter,
both dissolved and undissolved, may now be added to the
first portion in the half-pint beaker.
To separate out iron, arsenic, phosphoric acid, and
aluminium from the solution, sodium acetate may be added
at once, and the liquid boiled ; a far better mode, however,
is to effect a partial separation of these ingredients by the
addition of calcium carbonate in excess to the solution of
the ore, and after filtering out the solution containing the
greater portion of the cobalt and nickel, and partly wash-
ing the precipitate, to extract the last traces of cobalt and
nickel from the latter by dissolving it in hydrochloric acid,
adding excess of sodium acetate and boiling. The first
filtrate from the precipitate by calcium carbonate had
better be collected apart from the second filtrate from the
precipitate produced by sodium acetate, and received in a
beaker capable of holding at least a quart. The solution
of the precipitate by calcium carbonate is best effected in
a beaker, after the removal of the precipitate from the
COBALT AND NICKEL ORES. 841
filter. This is easily effected by inclining the funnel over
the beaker and sending a stream of water from the wash-
bottle between the filter and the upper edge of the mass
of precipitate, when the latter will soon become detached
and slide off into the -beaker below : it is here treated with
dilute hydrochloric acid, to dissolve all but the insoluble
residues of the ore which had not been previously filtered
off, and then a solution of sodium acetate is added in
excess (indicated by the deep red colour of liquid) and the
whole, heated to boiling, may be filtered at once. Iron
thus separated out, in presence of free acetic acid, has less
tendency to retain cobalt than when precipitated by means
of calcium carbonate ; besides which the cobalt and nickel
in the filtrate are left in the condition of acetates, a
necessary step preparatory to their separation from man-
ganese, &c.
This method of separating out iron, &c., though very
effectual, was often at first found to be attended with diffi-
culties ; for if much arsenic were not present the basic iron
acetate frequently became slimy towards the end of the
filtration, only allowing the boiling washing-water to pass
with such extreme slowness as to render the method almost
useless, until it was found that the addition of a little
sodium sulphate during the washing at once and perma-
nently effected a cure, causing filtration to proceed rapidly,
and diminishing the tendency of the iron to pass the filter.
Another difficulty was, that when sodium acetate was added
at once to the original solution of the ore, the solution,
often containing much cobalt and nickel as acetates, and
filtered in a concentrated state, yielded to the filter-paper
sufficient cobalt and nickel to occasion distinct loss. This
was avoided by separating out the great bulk of the cobalt
and nickel in solution as chlorides by means of calcium
carbonate, as above recommended, and then the weaker
solution, being comparatively strongly acid, could be
filtered without loss. This second filtrate may still retain
traces of iron ; a little sodium acetate may be added to
make sure that none remains in the condition of chloride,
which would be indicated at once by a reddening of the
842 THE ASSAY OF NICKEL AND COBALT.
liquid, and the whole is then boiled thoroughly once more ;
if rendered at all turbid passed, through a filter again, then
nearly neutralised with ammonia, and finally added to the
bulk of the cobalt and nickel solution in the quart beaker.
There will in all probability be enough of the sodium and
ammonium acetates present to convert the entire quantity
of cobalt and nickel into acetates without further addition,
and rendering it thus ready for the next operation.
If sulphuretted hydrogen be now transmitted through
the solution containing cobalt and nickel, these metals are
perfectly and completely separated without a trace of
manganese, magnesium, calcium, aluminium, or soluble
silica, which, when present, invariably accompany the sul-
phides precipitated by ammonium sulphide ; the sulphides,
moreover, thus precipitated from an acetic solution have
much less tendency to oxidise while on the filter, so that
their washing may be more perfectly accomplished than
in the former case. The passage of sulphuretted hydrogen
may be conveniently effected at the end of the day, and
the next morning the sulphides will be found perfectly
settled at the bottom of the beaker, permitting the great
bulk of the liquid (tested first to make sure of the removal
of cobalt and nickel) to be drawn off and thrown away, or
at least rapidly run through a filter ; the sulphides collected
at the bottom, together with that which always adheres to
the sides of the beaker, and which may be detached with-
out loss by a caoutchouc- covered glass rod, are then well
washed on the filter with boiling water until all soluble
matters are perfectly removed. The sulphides, perfectly
washed, are now to be dried by placing the funnel with
the filter in a broken beaker on wire gauze, at a safe
distance over a lamp, and when dry they may be detached
from the filter into a small beaker of from 1 to 2 oz. capa-
city, capable of being covered with a watch-glass ; the
filter itself is ignited, and the well-burnt ashes added to
the sulphides, which are then to be cautiously treated with
nitric acid, the action being rather violent, and, if care be
not taken, liable to occasion loss. With the aid of a little
heat, the whole should pass into solution.
COBALT SPEISS. 84$
In addition to cobalt and nickel the solution may still
contain zinc, together with copper, and other metals pre-
cipitable by sulphuretted hydrogen from hydrochloric
solutions. By passing sulphuretted hydrogen now through
the nitric solution, somewhat diluted, these latter are
readily precipitated and removed by filtration. Zinc, how-
ever, may still remain, to detect and remove which it is
necessary to expel the sulphuretted hydrogen still remain-
ing in the solution by boiling, to add solution of ammonia
until a precipitate occurs, and then to acidify pretty
strongly with acetic acid. If sulphuretted hydrogen slowly
transmitted, or fresh sulphuretted hydrogen water, occa-
sions a milkiness, zinc is present, and the slow passage of the
gas is to be continued until the precipitate begins to show
signs of darkening. The liquid is then filtered. The zinc
may be identified as such by collecting and igniting the
precipitate, when a trace of cobalt carried down with it
(and which may be separated out, if desired, by a repetition
of the process on the precipitate) will produce the beautiful
and well-known Einmann's green.
The filtrate, containing only nickel, cobalt, and salts
of ammonia, is treated with some pure sulphuric acid and
evaporated to dryness in a weighed capsule, and heated
sufficiently to expel the excess of sulphuric acid and all
the ammoniacal salts. The residual cobalt and nickel sul-
phates may now be weighed in a covered crucible. This
form of weighing these metals is easy, exact, and may be
rapidly executed. The weight of the ash of a filter of the
size used for collecting the sulphides must be ascertained
after treatment with sulphuric acid, and subsequent ex-
pulsion of the excess, and this weight deducted from the
total sulphates, in order to obtain perfectly correct results.
Decomposition of Cobalt Speiss or Arsenical Alloys of the
Same. The following is an outline of a new method de-
vised by Mr. H. Warren for decomposing cobalt speiss, or
arsenical alloys of the same : The regulus to be decom-
posed is first broken into pieces, about 1 Ib. in weight, and
suspended, by means of oiled string or other suitable sup-
port, in a vessel containing crude hydrochloric acid, to
844 THE ASSAY OF NICKEL AND COBALT.
which has been added about 1 oz. of copper nitrate, and
the whole allowed to remain for the course of a day or so
to undergo dissolution, the copper nitrate reacting with
the hydrochloric acid present, forming cupric chloride,
while the nitric acid evolved reacts on the metals compos-
ing the regulus to form nitrates of the same, which in
their turn are again decomposed, generating by so doing
sufficient nitric acid to react on a further portion of un-
deconiposed regulus.
The alloy by this treatment is deprived largely of its
nickel and cobalt, besides other metals present, such as
arsenic and bismuth, which have passed into solution ;
while the remaining portion, having become sufficiently
brittle to be readily reduced to a powder, but still con-
taining notable quantities of both nickel and cobalt, is
calcined at a low red heat in a plentiful supply of air, by
which treatment the residue consisting of arsenides and
sulphides are wholly converted into arseniates and sul-
phates, which are readily brought into solution by means
of crude hydrochloric acid, and added to the original
solution. Metallic iron, in the form of bars, is now
brought into contact with the solution, by which means
the whole of the copper present is removed as metallic
copper, together with a considerable quantity of the arsenic
and bismuth present, the iron passing into solution as
chloride, which is again separated by the addition of milk
of lime, carrying with it the remainder of the arsenic as
basic arseniate of iron. The remaining salts of nickel and
cobalt still existing in solution are precipitated together
by means of sodium carbonate, as carbonate of nickel and
cobalt ; they are next disseminated through water, and
chlorine gas passed through until saturation, the whole of
the nickel by this means passing into solution, the cobalt
remaining precipitated. The solution containing the nickel
is brought to a state of ebullition, in order to free it from
the excess of chlorine present ; the nickel is precipitated as
hydrate, by means of caustic soda, ignited to expel the
water present, and reduced to the metallic state -by the
usual method. It has been found preferable, for various
ASSAY OF METALLIC NICKEL. 845
reasons, to separate the calcium salts present, by means of
dilute sulphuric acid before proceeding to separate the
nickel and cobalt.
Assay of Nickel Ores. The ore of nickel usually met
with is the arsenide, containing variable quantities of
cobalt and iron, and frequently also copper and bismuth.
It is finely powdered, mixed with two parts of solid caustic
soda and 1^ part of sulphur, and fused in an earthen
crucible, gradually increasing the heat to dull redness, at
which temperature it is to be kept for some time. The
mass is then digested in water, which dissolves the soluble
sodium sulpho-arseniate, and leaves, when washed by de-
cantation, crystallised nickel sulphide. Attack this with
warm hydrochloric acid containing a little nitric acid.
The solution heated to about 70 C. is then submitted
to a current of sulphuretted hydrogen, which must be
continually passed until the liquid is cold. It is then to
be covered over and left at rest for 24 hours ; the arsenic
copper and bismuth come down as sulphides. These are
filtered off, and the filtrate is heated with potassium chlo-
rate or sodium hypochlorite to bring the iron to the state
of sesquioxide, which is then precipitated by ebullition
with excess of sodium acetate. The filtrate from the basic
iron acetate is concentrated by evaporation and mixed
with a saturated solution of potassium nitrite,* which will
precipitate all the cobalt.
The yellow precipitate washed with a saturated solu-
tion of potassium chloride may be treated as described
below. The nickel is precipitated from the liquid either
in the state of oxide by means of caustic potash, or after
concentration by a saturated and warm solution of potas-
sium binoxalate. The precipitated oxalate upon calcination
leaves pure nickel.
Assay of Commercial Metallic Nickel. Dissolve the
* Potassium nitrite is prepared by fusing, in an iron crucible, 1 part of
nitre with 2 parts of granulated lead, stirring well with an iron spatula, and
then heating to redness until the lead is completely oxidised. The fused mass,
after cooling, is extracted with water, and the small amount of lead which is
dissolved is precipitated by carefully adding a mixture of caustic ammonia
and ammonium carbonate or sulphide.
846 THE ASSAY OF NICKEL AND COBALT.
metal in hydrochloric acid containing a little nitric acid.
Pass sulphuretted hydrogen through the solution until the
metallic impurities are thrown down, and then precipitate
the nickel and cobalt by a warm saturated solution of
potassium binoxalate. The precipitate after being washed
,and calcined leaves the nickel (containing a little cobalt)
in the metallic state.
Assay of Cobalt Ores. Metallic cobalt may be prepared
from its ores (arsenide or sulphide) by a similar process
to that adopted in the case of nickel. When the mineral
contains more than 70 per cent, of arsenic, a preliminary
fusion should be performed with chloride of sodium, to
remove most of the arsenic. This may be continued by
roasting or by fusion with a mixture of sodium carbonate
and sulphur.
As nickel is almost invariably present in cobalt ores,
this metal will require to be separated. Wohler recom-
mends for this purpose the potassium nitrite process. The
yellow cobalt precipitate is dissolved in as small a quantity
of hydrochloric acid as possible, and sodium acetate is then
added ; the addition of a warm saturated solution of oxalic
acid now precipitates the cobalt as oxalate. This oxalate
after being washed and dried may be packed closely in a
crucible of biscuit porcelain, protected by enclosure in a
Hessian crucible. The covers being well luted on, the whole
is heated in a wind-furnace or a forge. If a sufficient tem-
perature has been obtained the cobalt will be in the form
.of a fused button.
Separation of Nickel and Cobalt. A method of sepa-
rating these metals, given some years since by Liebig,
consists in boiling the mixed double nickel and potassium
cyanides, cobalt and potassium cyanides, with oxide of
mercury. Nickel oxide is precipitated, while an equiva-
lent quantity of mercury is dissolved as cyanide. The
method certainly gives good results, but is not free from
objection. Long boiling is necessary before the precipi-
tation is complete, and it is difficult to prevent bumping
-during ebullition. The excess of mercury oxide must be
separated from the nickel oxide by a special operation,
SEPARATION OF NICKEL AND COBALT. 847
and the nickel afterwards again precipitated by caustic
alkali.
According to Wolcott Gibbs,* these inconveniences
may be completely avoided by employing, instead of the
oxide alone, a solution of the oxide in the mercury
-cyanide. When this solution is added to a hot solution
of the double cyanide of nickel and potassium, the whole
of the nickel is immediately thrown down as a pale green
hydrate of the protoxide. Under the same circumstances
cobalt is not precipitated from the double cobalt and potas-
sium cyanide. Mr. W. N. Hill, who has repeatedly em-
ployed this method and carefully tested it, has found that
the separation effected is complete. JSFo cobalt can be
detected in the precipitated nickel oxide by the blow-
pipe, nor can the nickel be detected in the cobalt (finally
separated as oxide) by Plattner's process with the gold
bead. The solution of mercury oxide is easily obtained
by boiling the oxide with a strong solution of the cyanide
and filtering. According to Kuhn, the cyanide formed in
this manner has the formula HgCy-f 3HgO. The hydrated
nickel oxide precipitated may be filtered off, washed, dried,
ignited, and weighed. The cobalt is more readily and
conveniently estimated by difference, when, as it is always
possible, the two metals have been weighed together as
sulphates. We are not prepared to say that this modifi-
cation of Liebig's method of separating nickel and cobalt
gives better results than Stromeyer's process by means of
potassium nitrite, but it is at least very much more con-
venient, and requires much less time. The complete pre-
cipitation of cobalt in the form of Co 2 3 ,2N0 3 + 3KO,N03
usually requires at least forty-eight hours, and rarely
succeeds perfectly except in experienced hands.
M. Terreil has proposed a very excellent method for
separating these two metals. The author's method is
founded (1) on the insolubility of roseocobaltic hydro-
chlorate in acid liquids and ammoniacal salts, discovered
by M. Fremy ; (2) on the rapid transformation of ordinary
salts of cobalt into roseocobaltic salts, under the double
* * Chemical News,' March 17, 1865.
848 THE ASSAY OF NICKEL AND COBALT.
influence of ammonia and oxidising bodies such as potas-
sium permanganate and alkaline hypoehlorites ; (3) on
the complete precipitation of manganese in ammoniacal
liquids by alkaline hypochlorites, and potassium perman-
ganate.
To separate cobalt from nickel, operate in the follow-
ing manner :
To the solution of the two metals add an excess of
ammonia, which re-dissolves the two oxides ; add to the
hot ammoniacal liquid a solution of potassium permanga-
nate, sufficient to cause the liquid to remain coloured
violet for a few instants by the excess of permanganate.
Boil the liquid for a few minutes, then add a slight excess
of hydrochloric acid, to re-dissolve the manganese oxide
which will have formed. Heat the liquid gently for twenty
or twenty-five minutes ; then let it stand for about twenty-
four hours. All the cobalt will then be deposited in the
form of a beautiful red-violet crystalline powder ; the pre-
cipitate is roseocobaltic hydrochlorate, which collect on a
weighed filter ; wash it on the filter with cold water, then
with diluted hydrochloric acid, or with a solution of am-
moniacal salt, and then with ordinary alcohol, which frees
it from ammoniacal salt. Dry it at 110, and weigh. 100
parts of roseocobaltic hydrochlorate correspond to 22*761
of metallic cobalt, or to 28*959 of protoxide of cobalt.
It is, however, better to take a given quantity of the
roseocobaltic salt, and reduce it by dry hydrogen ; this
leaves perfectly pure cobalt to be weighed.
Next boil the solution containing nickel to expel the
alcohol which has been introduced in washing the cobaltic
salt ; saturate it with ammonia, add another small excess
of permanganate of potash, and boil. All the manganese
will be precipitated ; filter the liquid, and all the nickel
will be found in the filtrate, from which it may easily be
separated in the state of sulphide, and then transformed
into oxide.
By this process the presence of a ten-thousandth part
of cobalt in a salt of nickel may be ascertained.
In this operation an alkaline hypochlorite may take
DETECTION OF NICKEL BEFORE THE BLOWPIPE. 849
the place of the potassium permanganate, but then the
deposit of roseocobaltic salt takes place with extreme
slowness, and several days are required to complete it.
This reagent is preferable to permanganate when manga-
nese is to be separated from nickel and cobalt.
Quantitative Assay of Small Proportions of Cobalt in
Nickel. The following method is proposed by Dr. Fleit-
mann. It is well known that from solutions containing
both these metals the cobalt is first precipitated by hypo-
chlorites, as a brown hydrate, and the nickel hydrate does
not fall until after a further addition. The partial pre-
cipitation with sodium hypochlorite is so conducted that
at least two parts of nickel may be thrown down to one of
cobalt. The proportion may be judged by the colour of
the solution of the precipitate. If this solution is decidedly
red, more of the precipitant must be added. After a
slight washing the precipitate is dissolved on the filter
with warm hydrochloric acid,. the excess of chlorine re-
moved by boiling, and the mixture of cobaltous and
nickelous oxides is precipitated from the warm solution
with potash lye. The precipitate is filtered, slightly washed
with water, dissolved in acetic or nitric acid, and the
cobalt is then precipitated in the ordinary manner with
potassium nitrite.
Detection of Nickel before the Blowpipe.- The following
is Plattner's method for detecting nickel when contained
in large quantities of cobalt :
Fuse in the oxidising flame a moderate quantity of
borax to a bead in the loop of platinum wire, with suffi-
cient oxide of cobalt to give an opaque glass ; remove the
assay, and prepare one or two similar beads, and place the
whole in a charcoal cavity, with a button of pure gold
weighing from fifty to eighty milligrammes. The operator
must now heat in the reducing flame until he is satisfied
that the whole of the nickel is in a metallic state ; the
charcoal during the action must be inclined alternately
backwards and forwards, so that the gold button may
flow through the melted glass, and form an alloy with the
reduced particles of nickel. When the golden globule
3 i
850 THE ASSAY OF NICKEL AXD COBALT.
solidifies, it must be extracted with a forceps, placed be-
tween paper, and struck with a hammer, so as to detach
all the adhering vitreous parts. The auriferous button,
which has become more or less grey from the presence of
nickel, and also more brittle than pure gold, is now to be
mixed with microcosmic salt, and heated for some time in
the oxidising flame. If the borax-glass has not been in
the first instance oversaturated with oxide of cobalt, a
bead will now be obtained, which is coloured only by
oxide of nickel, and will therefore appear brownish-red
while hot, and when cold reddish yellow. Should por-
tions of oxide of cobalt be also reduced, as the cobalt is
oxidised before the nickel, either a blue glass, coloured
by oxide of cobalt, or a green one if some nickel was
also oxidised will be obtained. In either case the glass
must be separated from the button, mixed with more
microcosmic salt, and heated in the oxidising flame
until it acquires a tinge. If the borax-glass had not
been over-saturated at the commencement, the colour
now obtained will proceed from nickel, although the
cobalt oxide contains a trace only ; but if nickel oxide
be not present, the microcosmic bead remains perfectly
colourless.
In the assay of substances containing cobalt, nickel,
and zinc, Alex. Olassen (' Zeitschrift fur Anal. Chemie,'
1879, p. 189) proceeds as follows:
To the solution, rendered as neutral as possible, he adds
so much neutral potassium oxalate (1 part of the salt and
3 parts water) that the precipitate is redissolved. He then
adds, with stirring, acetic acid about equal to the volume
of the liquid. The precipitate becomes crystalline at a
temperature of 50 to 60 C., and the liquid remains clear.
The precipitate is washed with a mixture of equal volumes
of concentrated acetic acid, alcohol, and water. The dry
precipitate, after the filter has been burnt on a platinum
wire, is first ignited very slightly in a covered platinum
crucible, so that no particles may be expelled by the
escaping carbonic oxide, and finally are ignited in an
open crucible.
NICKEL AND COBALT GLANCE. 851
Ores containing Sulphur, Arsenic, Nickel Cobalt, and
Iron. (Arsenical nickel glance, cobalt-glance, red and
white nickel pyrites, cobalt speiss, commercial nickel.)
In nickel and cobalt glance sulphur is present in great
quantity, and is rarely absent in the remaining ores. Its
estimation is best effected in a separate portion. For
the separation and estimation of the metals the finely
powdered sample is oxidised either with aqua regia or
with hydrochloric acid and potassium chlorate, and the
arsenic is estimated as directed for the arsenides and sulph-
arsenides of iron.
The liquid filtered from the arsenic sulphide is freed
from sulphuretted hydrogen by heat, the iron is oxidised
with potassium chlorate, the free chlorine expelled by
heat and the addition of a little alcohol, the liquid is
largely diluted, placed in a basin, and mixed with sodium
carbonate till the acid reaction becomes very faint.
Barium carbonate levigated in water to a fine paste is
then added till it lies at the bottom of the vessel. After
repeated stirring the iron is all precipitated. It is filtered
off, dissolved in dilute hydrochloric acid, sulphuric acid
is added, the precipitate of barium sulphate removed,
and the ferric oxide precipitated with ammonia. The
liquid containing the nickel and cobalt is freed from
soluble barium compounds by means of sulphuric acid,
and the barium sulphate is removed by filtration.
The filtrate, containing the nickel and cobalt, is placed
in a basin, supersaturated with potash, heated to a boil,
filtered, washed with hot water, and the mixture of both
oxides dissolved in acetic acid. To this liquid is added a
concentrated solution of potassium nitrite (if it contains
free potash an excess of acetic acid must be present), and
the whole is let stand for 24 hours.
A yellow precipitate of potassium cobaltic nitrite is
produced, which is filtered off and washed in the cold with
a saturated solution of potassium chloride. More potas-
sium nitrite is added to the filtrate to ascertain whether
anything further is deposited. The yellow precipitate is
digested with hydrochloric acid, in which it dissolves,
3 i 2
852 THE ASSAY OF NICKEL AND COBALT.
diluted, filtered into a basin, and the cobalt oxide is pre-
cipitated by boiling with an excess of potash. The pre-
cipitate is washed hot, and when dry it is placed in a
porcelain crucible and strongly ignited in a current of
hydrogen. After being allowed to cool in the same current
it is weighed in the covered crucible as metallic cobalt. It
is then well washed with water, ignited again in the stream
of hydrogen, and finally weighed.
In the liquid filtered from the cobaltic precipitate the
nickel is thrown down by boiling with potash. When care-
fully washed, dried, and ignited, it yields pure nickel oxide,
from the weight of which that of the metal is calculated.
Many substances of this class contain other constituents
in smaller proportion.
Commercial nickel contains silicon, which on dissolving
the metal in nitric acid, &c., is separated out as silica. For
its estimation the whole is evaporated to dryness in the
water-bath, the residue when cold is moistened with acid
and treated with water. The silica is then filtered off, and
the other constituents are estimated in the filtrate in the
usual manner.
The ores of nickel and cobalt and speiss-nickel contain
antimony. It is precipitated by sulphuretted hydrogen
along with arsenic. The precipitate is dissolved in aqua
regia, and the two metals are separated in the usual
manner.
The above ores and furnace products may also contain
copper, bismuth, and lead. In presence of small quantities
of these metals sulphuretted hydrogen should be passed
through the solution of the mixture for a rather shorter
time at first. These are precipitated before arsenic, but are
accompanied by a part of it. The precipitate is filtered off,
and the passage of the sulphuretted hydrogen through the
filtrate is continued till the arsenic is completely deposited.
The precipitated sulphides whilst still moist are placed
in a flask together with the filter, and are digested for a
considerable time with concentrated yellow ammonium
hydrosulphide. When cold the solution is diluted and
filtered with exclusion of air, washed with water, to which
ALLOYS OF COPPER, ZINC, AND NICKEL. 853
a few drops of ammonium hydrosulphide have been added ;
the filtrate is slightly supersaturated with hydrochloric
acid, and the precipitate of arsenic sulphide is filtered off
and added to the main arsenical precipitate. The sulphides
(copper, bismuth, and lead) if their quantity permits are
separated by methods already given. If the quantities of
these metals are considerable, as is the case in many furnace
products, their assay is conducted as pointed out for lead-
and copper-speiss. Rammelsberg .
Alloys of Copper, Zinc, and Nickel The assay is con-
ducted as has been directed for alloys of copper and zinc.
The solution filtered from the copper sulphide, and con-
taining the zinc and nickel, is concentrated by evaporation
in order to remove the excess of sulphuretted hydrogen.
It is then poured into a flask, supersaturated with pure
potash and hydrocyanic acid enough to dissolve the whole
to a yellow liquid. From this solution the zinc is precipi-
tated as zinc sulphide by means of potassium monosulphide
(prepared by reducing potassium sulphate with charcoal),
and caused to settle by digestion. It is then passed
through a covered filter, the filtrate being collected in a
flask, washed with cold water, to which a little potassium
sulphide has been added, redissolved by digestion with
hydrochloric acid in a covered beaker till the odour
of sulphuretted hydrogen has entirely disappeared ; the
diluted solution is filtered into a capsule, and the zinc
oxide is thrown down by means of sodium carbonate.
The filtrate from the zinc sulphide is boiled in the flask
with aqua regia till the odours both of sulphuretted
hydrogen and hydrochloric acid are expelled, and do not
return on the addition of a little acid. The liquid is then
placed in a basin, supersaturated with potash, and kept at
a boil for a few minutes. The hydrated nickel-oxide is
washed with hot water, dried, ignited, and weighed, the
proportion of the metal being calculated from the weight
of the nickel-oxide.
In the electrolytic assay of copper and nickel Herpin
proceeds as follows in the assay of alloys of these metals.
He dissolves 1 grm. of the sample in nitric acid in a flask
854 THE ASSAY OF NICKEL AND COBALT.
capable of holding 250 c.c., evaporates almost to dry ness,
and adds 4-5 c.c. sulphuric acid, and water enough to
make up a volume of 60-70 c.c. The liquid is rinsed into
a platinum capsule and submitted to electrolysis. The
copper only is deposited from an acid solution.
The liquid, still containing the nickel, is poured into a
flask like the one used for dissolving the alloy ; the plati-
num capsule is rinsed first with water, then with alcohol,
dried and weighed to estimate the copper.
The nickeliferous liquidate the washings, is heated to
a boil, partially neutralised with sodium carbonate, and
supersaturated with ammonia till it takes a blue colour.
It is then placed in the platinum capsule, and submitted to
electrolysis. Traces of lead and iron do not interfere.
W. Ohl (' Zeitschrift fur Anal. Chemie,' 1879, 523) gives
the following process for the assay of a nickel speiss :
1 grm. finely ground is placed in a beaker holding 300 c.c.,
and covered with nitric acid or aqua regia. The beaker
is covered with a watch-glass and set on the sand-bath.
When it is completely dissolved the watch-glass is taken
off and the liquid evaporated to dryness. About 5 c.c. of
pure concentrated hydrochloric acid are added, and after
the mass is dissolved the beaker is half-filled with water.
When the solution is hot, sulphuretted hydrogen is passed
through it till cold. It is again set to warm, and again
treated with the same gas till cold. The precipitate of
copper and arsenic is quickly deposited, and the super-
natant liquid becomes clear. As arsenic sulphide is slightly
soluble in water containing sulphuretted hydrogen, the
beaker is set in a warm place till the smell becomes very
faint. If the precipitate is a fine uniform yellow it is
washed on filtering with cold pure water ; if darker, and
therefore containing more water, it is washed with sulphu-
retted hydrogen water.
The filtrate containing cobalt and nickel is evaporated
to dryness in a capsule holding f litre, adding a little
potassium chlorate to oxidise iron. The residue is taken
up with a little hot water and hydrochloric acid, preci-
pitated with pure solution of soda till the reaction is
LEAD- AND COPPEK-SPE1SS. 355
alkaline, redissolved in pure acetic acid, largely diluted
and heated to a boil. The iron is all deposited as basic
ferric acetate, which is filtered off and washed with hot
water till a drop of ammonium sulphide produces no
turbidity in a drop of the washings. The solution, freed
from the iron, is evaporated to dryness, dissolved in water
and a few c.c. of dilute sulphuric acid, placed in a beaker
holding 600 c.c., supersaturated with ammonia, and sub-
mitted to the electric current.
When the electrolysis is complete, a drop of the liquid
is withdrawn with a pipette, filtered, and mixed with a
drop of ammonium sulphide. If no turbidity is formed the
platinum cone is withdrawn, washed first in water and then
in absolute alcohol, and dried. The increase of weight
gives the sum of the cobalt and nickel.
LEAD- AND COPPEE-SPEISS.
These substances are very complicated products formed
during the metallurgical treatment of arseniferous and
antimoniferous ores of lead and copper. They may con-
tain copper, lead, iron, nickel, cobalt, zinc, bismuth, silver,
arsenic, antimony, and sulphur.
A. Speiss containing little or no Lead or Antimony
One portion is taken for the determination of sulphur,
and another is dissolved in aqua regia, or in a mixture of
hydrochloric acid and potassium chlorate. The proportion
of silver is generally so small that no silver chloride
remains. If any lead chloride is separated it is dissolved
by heating in water. The solution is treated with sulphu-
retted hydrogen, when lead, copper, bismuth, antimony,
and arsenic are deposited. To ascertain that the arsenic
is completely precipitated the liquid is heated and again
treated with sulphuretted hydrogen. The metallic sul-
phides are digested with concentrated yellow ammonium
hydrosulphide, in order to dissolve the arsenic and anti-
mony sulphides.
After filtration both are precipitated by hydrochloric
acid, and separated as directed under arsenical iron.
856 THE ASSAY OF NICKEL AND COBALT.
The undissolved sulphides of copper, lead, and bismuth
are allowed to become air-dry, and are then detached from
the filter, which is incinerated in a small porcelain crucible.
First, these ashes, and then the total sulphides, are dis-
solved in nitric acid. The sulphur, which is not quite
pure, is filtered off and gently heated in a porcelain
crucible, the slight residue being digested again with nitric
acid and added to the main solution (if lead sulphate
remains unattacked it must be collected on a weighed
filter). The nitric solution of the three metals is concen-
trated by evaporation, and the lead separated as already
directed. The filtrate is saturated with sodium carbonate,
potash is added, the whole heated to a boil, the oxides
are filtered off, washed slightly, dissolved in the smallest
quantity of hydrochloric acid, and the bismuth is precipi-
tated by the addition of much water as basic oxychloride,
which is reduced by potassium-cyanide, as already de-
scribed. Copper is precipitated from the filtrate by sul-
phuretted hydrogen.
In the filtrate from the first sulphuretted hydrogen
precipitate there are still found iron (as ferrous oxide),
nickel, cobalt, and zinc. The liquid is concentrated to
expel sulphuretted hydrogen and most of the free acid ;
the iron is peroxidised by the addition of a little potassium
chlorate, diluted, the free chlorine expelled by heating
with a few drops of alcohol, and the iron oxide is separated
from zinc, nickel, and cobalt as directed for arsenical
nickel glance, &c.
B. Speiss containing much Lead or Antimony. The
process to be followed is tedious, and requires much care.
The substance, finely powdered, is decomposed by heating
in a current of chlorine gas.
The apparatus requisite consists of a capacious flask, in
which the necessary chlorine gas is evolved from a mixture
of 2 parts black oxide of manganese, 3 parts of common
salt and dilute sulphuric acid (1J to 2 parts water to 1
part of the monohydrated acid). The flask is closed with
a cork, through which passes a tube bent twice at right
angles. Its longer leg dips into concentrated sulphuric
LEAD- AND COPPEK-SPEISS. 857
acid contained in a second flask, where the gas is partially
dried, and whence it issues to pass by another tube into a
chloride of calcium apparatus, and thence into the bulb-
tube in which the reaction is to take place. The tube is
tared, from 1 to 2 grms. of the finely ground sample is
inserted into the bulb through its wider tube, both ends
are cleaned with a feather, so that none remains except in
the bulb, and the whole is weighed to find the exact quantity
which has been taken for analysis. The shorter tube is
then connected to the chloride of calcium apparatus, and
the longer, or bent tube, is conducted through a cork into
a receiver, which resembles a Liebig's bulb-tube on a larger
scale, and with parallel limbs. It is filled with dilute
hydrochloric acid, to which a little tartaric acid is added,
if the body under analysis contains antimony. From the
second limb of this receiver a bent tube passes into a
Woolf's bottle, and dips into the same solution. From
the other aperture passes a tube which carries away the
escaping chlorine. When all parts of the apparatus are
full of chlorine heat is carefully and gradually applied to
the bulb, which must not reach visible redness. After the
completion of the operation the bulb contains the non-
volatile lead, silver, bismuth, copper, nickel, cobalt chlo-
rides, and in part those of iron and zinc. They are dis-
solved out by means of water and hydrochloric acid ;
undecomposed portions of the sample and silver chloride
may remain, and are collected upon a weighed filter, and
after weighing treated with ammonia, which dissolves the
silver chloride, leaving the undecomposed matter behind.
The hydrochloric solution is filtered into dilute sulphuric
acid, evaporated, and the lead separated as lead sulphate.
The filtrate is precipitated with sulphuretted hydrogen,
and copper and bismuth are separated by dissolving the
sulphides in hydrochloric acid 5 to which a little nitric acid
is added, the solution concentrated to a small bulk, the
bismuth precipitated by water, the basic bismuth chloride
reduced by potassium cyanide, and in the filtrate the
copper precipitated by sulphuretted hydrogen in the ordi-
nary manner. The filtrate from which the copper and
858 THE ASSAY OF NICKEL AND COBALT.
bismuth sulphides were removed contains iron, zinc, nickel,
and cobalt, and is treated as already pointed out.
From the volatile chlorides contained in the liquid
from the receivers the sulphuric acid is removed by barium
chloride, and any excess of the latter by the cautious addi-
tion of sulphuric acid. Antimony and arsenic are precipi-
tated by sulphuretted hydrogen, and separated in the usual
manner. The filtrate contains the volatilised portions of
iron and zinc, which are also dealt with by ordinary
methods.
CHAPTEE XXIV.
THE ASSAY OF SULPHUE.
THE only commercially valuable Sulphur-minerals are :
I. Sulphurous Earth (native sulphur).
In Sicily these minerals are divided into five classes :
1. Very rich ores, containing 32 34 per cent, sulphur
2. Kich 2426
3. Good 16-18
4. Middling ,, 8 9
5. Poor 3 5
II. Iron and Copper Pyrites (FeS 2 ), and (Cu 2 S, Fe 2 S 3 ).
These ores are used to a very large extent for the
manufacture of sulphuric acid.
In order to approximately estimate the value of ores
of the first class for such manufacture, the following method
of assay may be used.
Assay by Distillation. A certain quantity of the pul-
verised sulphurous earth is heated in a glass retort, which
is furnished with a receiver. The retort is then heated,
gradually raising the temperature, till no more sulphur is
evolved. The latter will collect in all cases in the neck of
the retort and receiver, which may be of glass and must
be kept cool.
The sulphur derived from sulphurous earth is generally
pure, whilst that from pyrites frequently contains arsenic
and selenium, and sometimes traces of thallium.
ASSAY OF IRON AND COPPER PYRITES.
A. Assay of Sulphur in the Dry Way. Fuse the
weighed ore with a weighed quantity of anhydrous sodium
carbonate, twice as much potassium chlorate as ore, and
860 THE ASSAY OF SULPHUK.
from 12 to 20 times as much sodium chloride (added to
moderate the action) ; carbonic acid is expelled, potassium
chloride formed, and all the sulphur converted into sodium
sulphate. By dissolving the residue in water and esti-
mating alkalimetrically the unaltered sodium carbonate by
a standard acid solution, the portion converted into sul-
phate, and hence the sulphur in the ore, is known. Besides
the difficulty of preventing loss by deflagration, this method
is open to the small errors caused by the reckoning all
arsenic present to be sulphur : this, however, is usually of
no moment for commercial purposes ; and calcium carbo-
nate in the ore may, if required, be previously dissolved
out by dilute hydrochloric acid.
In performing fusions of sulphur compounds with
nitre or potassium chlorate the operator must bear in mind
a source of error, first pointed out by Dr. David S. Price,
in consequence of sulphur compounds being contained in
the coal-gas which frequently serves as fuel in these experi-
ments. By exposing a small quantity of fused nitre, on
the outside of a platinum capsule, to the flame of a Bunsen
gas-burner for three-quarters of an hour, Dr. Price suc-
ceeded in detecting the presence of sulphuric acid to an
amount equivalent to 12 milligrammes of sulphur. This
sulphuric acid had been formed by the oxidation of the
sulphur in the coal-gas, and when dissolved in water it
gave an immediate precipitate with chloride of barium.
By making a similar experiment with the use of a spirit-
lamp as the source of heat, no trace of potassium sulphate
was formed ; nor was any appreciable amount of sulphuric
acid generated in another trial made by fusing a small
quantity of nitre inside a platinum capsule heated over
gas ; but whenever the fused salt crept over the edges of
the capsule, some potassium sulphate was sure to be formed.
This observation may become a matter of importance when
the amount of sulphur in pig-iron is estimated by fusion
with pure nitre, for the author has remarked that samples
containing much manganese are especially liable to impart
to the fused salt a tendency to creep up and escape over the
sides of the crucible.
ASSAY OF SULPHUR IN PYRITES. 861
B. Assay of Sulphur in the Wet Way. Mr. C. E. A.
Wright recommends the following process as being the one
best adapted for commercial purposes : A known weight
of the ore reduced to fine powder is oxidised (best in a
small flask with a funnel in the mouth to avoid loss by
spirting, and heated on a sand-bath), either by strong
nitric acid, or aqua regia, perfectly free from sulphuric
acid ; after the oxidation is complete, the liquid is evapo-
rated down as far as possible to expel the majority of the
remaining nitric or hydrochloric acid ; the residue is boiled
with a little water, and almost but not quite neutralised by
ammonia ; a solution of barium chloride of known strength
is then added until no further precipitate is produced, the
exact point being found by filtering off a little of the liquid
after each addition of barium chloride, and adding to it a
few more drops of the standard solution, care being always
taken, in case of a further precipitate being thus produced,
to add this filtrate to the original solution, and mix well
before filtering a second time. In case of overstepping the
mark, it is convenient to have at hand a solution of sodium
sulphate of strength precisely equal to that of the barium
chloride ; this solution may then be cautiously added
with repeated filtration and examination of the filtrate
with the sulphate solution, until the point is just reached
when addition of sulphate solution produces no further
precipitate ; by subtracting the volume of sulphate solu-
tion thus used from the total volume of barium solution
added, the exact quantity of this latter consumed is
known. If 1 grm. of sulphur ore be taken, and 32 -5
grms. of pure anhydrous barium chloride be dissolved in
a litre of fluid, each cubic centimetre of barium solution
used will represent 0*5 per cent, of sulphur in the ore
examined : 22-19 grms. of anhydrous sodium sulphate
being dissolved to a litre for the second solution. In case
of lead being contained in the ore, an error is introduced
from the formation of insoluble lead sulphate. As lead,
however, rarely occurs in any perceptible quantity, this
error is negligible, the process only giving approximate
results.
86*2 THE ASSAY OF SULPHUR.
Where greater accuracy is required, it is advisable to
precipitate the sulphuric acid formed from the original
liquid (filtered from insoluble residue) by barium nitrate
or chloride, and to weigh the barium sulphate produced.
Instead of oxidising by acids, the powdered ore may be
suspended in caustic potash (free from sulphate), and
oxidised by passing washed chlorine into the liquid ; lead,
being converted into dioxide, is thus rendered non-injurious ;
the alkaline liquid obtained is acidified, and precipitated
by chloride of barium as before. In the volumetric es-
timation usually pursued, a curious circumstance is occa-
sionally observable when much free acid exists in the
solution viz. that a point may be reached when the fil-
tered liquid is clear, and remains so even on standing for
a short time, but yields a cloud, or even a precipitate, on
the addition either of barium solution or sulphate solution.
This source of error is mostly avoidable by nearly neutral-
ising the free acid with ammonia.
Instead of chlorine, hypochlorous acid may be used to
transform the sulphur of pyrites into sulphuric acid, which
is then estimated by barium chloride. Finely pulverise
the mineral and suspend it in water, through which a
current of gaseous hypochlorous acid, or, better still,
hypochloric acid, is passed ; this entirely dissolves the
pyrites. Hypochlorous acid is prepared by heating a milk
of calcium carbonate through which a current of chlorine
is passed to saturation. Hypochloric acid is obtained by
heating in a water-bath a tube, supplied with a cork and
delivery tube, and containing a mixture of nine equivalents
of oxalic acid and one equivalent of chlorate of potash.
Mr. A. H. Pearson has given the following very accu-
rate method of estimating sulphur in pyrites : Weigh out
1 grm. or less of the powdered ore ; place the powder in a
porcelain dish, together with a small quantity of potassium
chlorate ; pour upon it some 50 c.c. of pure nitric acid of
39 B., and cover the mixture with an inverted glass funnel
with bent stem. Set the dish upon a water-bath and
heat the water to boiling. From time to time throw
crystals of potassium chlorate into the hot acid. By adding
ASSAY OF SULPHUR IN PYRITES. 863
rather large crystals of the chlorate at frequent intervals,
it is easy to oxidise the whole of the sulphide in half an
hour ; but since the solution obtained in that case is
highly charged with saline matter, it will usually be found
more advantageous to use less of the potassium chlorate,
and to allow a somewhat longer time for the process of
oxidation.
When all the sulphur has been oxidised, rinse the
funnel with water and remove it from the dish. Evapo-
rate the liquid to a small bulk, then add to it a little
concentrated hydrochloric acid, and again evaporate to
absolute dryness, in order to render silicic acid insoluble.
Moisten the residue with concentrated hydrochloric acid,
mix it with water, and filter to separate silicic acid and
gangue.
To the filtrate from the silicic acid add a quantity of
solid tartaric acid, about as much as that of the pyrites
originally taken ; heat the liquid almost to boiling, and
add to it an excess of barium chloride, to precipitate the
sulphuric acid. After the barium sulphate has been
allowed to subside, wash it thoroughly by decantation,
first with hot water, and afterwards with a dilute solution
of ammonium acetate (the latter may be prepared at the
moment of use by mixing ammonia-water and acetic acid).
The purpose of the ammonium acetate is to dissolve any
barium nitrate which may adhere to the sulphate : that of
the tartaric acid is to prevent the precipitation of iron
compounds together with the barium sulphate. In an
experiment where 0'7 grm. of pyrites was oxidised with
potassium chlorate and nitric acid, and the filtrate from
silica was acidulated with hydrochloric acid without the
addition of tartaric acid, there was thrown down, on the
addition of barium chloride, a bright yellow precipitate,
which became darker-coloured when the solution was
boiled. It was not only found to be impossible to wash
out the iron with which this precipitate was contaminated,
but the consistency of the precipitate was such that it was
a difficult matter even to wash away the saline liquor in
which it was formed.
864 THE ASSAY OF SULPHUK.
M. A. Houzeau, in his so-called gravi-volumetric
process, attacks 1 gramme of the powdered ore with a
mixture of 4 parts pure potassium nitrate, and 3 parts
sodium carbonate, likewise pure. The saline mass is
dissolved in hot water, and the ferric oxide filtered from
the alkaline sulphate. The washing-waters are added to
the filtrate, and, after cooling, it is made up with distilled
water to half a litre. A portion is then taken (10 c.c.)
and acidulated with a few drops of pure acetic acid, and
the sulphuric acid is rapidly estimated, as the author
has elsewhere indicated for selenitic waters, by making use
of a standard solution of barium chloride, applied by the
aid of the gravi-volumeter in place of the ordinary burette,,
the use of which, in such estimations, yields merely erro-
neous results. In the gravi-volumeter the weight of the
standard solution shows the quantity of the reagent which
has been used. But each drop of the barytic liquid
deli vered by the gravi-volumeter weighs at the tempera-
ture of 13 exactly 0-050 grm.
In the Frieberg works, 1 grm. of finely ground ore is
mixed with 3 grms. anhydrous sodium carbonate, and an
equal weight of saltpetre. This mixture is placed in an
iron crucible, melted in a muffle. At a red heat, the mass
is dissolved in hot water, and the liquid is filtered into
a beaker, in which there is a little hydrochloric acid to
saturate the excess of soda. The liquid, which should
have an acid reaction, is then boiled for a short time,,
and the sulphuric acid is estimated volumetrically with a
solution of barium chloride, standardised so that 1 c.c.
indicates 2 per cent, of sulphur.
B. Deutecom adopts the following process : 1 grm.
pyrites is mixed in a large covered crucible with 8 grms.
of a mixture of equal parts potassium chlorate, sodium
carbonate, and sodium chloride. The crucible is heated at
first gently, so as to dry the contents, which are after-
wards melted at a high temperature. The mass, when
cold, is treated with boiling water, and the solution, to-
gether with the deposit, is introduced into a measuring-
flask of 200 c.c., filled up, filtered, and the sulphuric acid
ASSAY OF SULPHUR IN PYRITES. 865
is estimated in aliquot parts, say 50 c.c. The insoluble
residue does not retain any sulphuric acid. In this manner
the use of nitric acid is avoided. The decomposition of
the potassium chlorate is complete.
In assaying pyrites for sulphur only by fusion, Mr.
P. Holland has obtained good results by the following
process, which may be useful in such laboratories as do
not possess large platinum crucibles. A test-tube, or piece
of sealed combustion tube, about 6 inches long and
half an inch in internal width, is fitted with a cork and
delivery tube, the latter bent at a right angle, and long
enough to reach to the bottom of the flask in which it is
intended to make the titration. The fusion mixture con-
sists of equal parts of nitre and ignited sodium bicar-
bonate, both free from sulphur, dry, and in fine powder.
Nine to ten grammes are taken in an operation, together
with one of pyrites ; the latter must be in exceedingly fine
powder. The two are mixed in a warm porcelain dish or
agate mortar, and transferred to the tube without loss.
The delivery tube is then inserted, with its extremity
dipping into the flask. A channel is made on the surface
of the mixture, and the tube, suitably supported, is heated
by small portions at a time with a Bunsen gas flame, com-
mencing as usual with the anterior portion. When the
operation is progressing favourably, the deflagration
proceeds for a few seconds after removing the flame.
There is no danger to be apprehended, and the tube
does not crack or blow out with proper care. When the
tube has been heated throughout, and the deflagration
has ceased, it is then more strongly heated with a Hera-
path or powerful gas flame. It is a good plan at this
stage to slip a coil of wire gauze over the tube, which
helps to accumulate the heat. It is not, however, neces-
sary that the contents should be fused a second time ; at
least this has not been done in the experiments appended.
The sulphur ores examined have yielded their sulphur
readily.
The gaseous products of the combustion, which
mechanically carry over with them small quantities of
866 THE ASSAY OF SULPHUR.
sulphates or sulphuric acid, \being heavier than air, collect
in the flask, and are washed by shaking with a little
water, closing the flask with the palm of the hand. The
delivery tube is also washed. That containing the fused
mass is carefully broken and put in the flask, together
with sufficient hydrochloric acid to dissolve nearly the
whole of the iron oxide ; then ammonia is added, until a
precipitate of oxide reappears ; and lastly, as much hydro-
chloric acid and water as is necessary to bring the fluid
to the conditions which were obtained when the barium
solution was standardised. The author has used 2 c.c. of
free acid, and the total volume of solution was 200 c.c.
F. Bceckmann recommends the following modification
of the potassium chlorate method : Half a grm. of finely
ground pyrites (sifting is not absolutely necessary) is
mixed in a large platinum capsule with the well-known
mixture of 6 parts sodium carbonate and 1 part potassium
chlorate. The mixing is effected with a platinum spatula,
and is then made more complete by gentle rubbing with
an agate pestle fixed to a wooden handle. The whole is
then fused over the blast-lamp. The aqueous solution of
the melt is first poured into a beaker to avoid spirting,
and thence into another tall beaker containing an excess of
hydrochloric acid. The filtered solution is heated and
precipitated with hot barium chloride, heated gently upon
the sand-bath for a time until the liquid standing above
the precipitate has become clear, and is filtered at once.
The burnt ores in sulphuric acid works have been for a
long time assayed for sulphur by this process. Take
about 2 grms. of burnt ore to from 20 to 25 grms. of
chlorate mixture.
867
CHAPTER XXV.
DISCRIMINATION OF GEMS AND PRECIOUS STONES.
SIMPLE characteristics of and means of recognising many
gems and precious stones have been given at page 291, in
the section on the discrimination of minerals.
The present chapter contains much information which
could not appropriately be introduced in the previous
chapter, which was intended chiefly for the use of tra-
vellers and explorers. Some trifling repetitions occur
purposely to save the inconvenience of referring back.
The principle sources of recognition are colour, crys-
talline form, specific gravity, and hardness. In the present
chapter will be introduced all the most constantly oc-
curring natural forms of the gems and precious stones
mentioned.
The specific gravity or density of a substance is the
proportion of its weight to its volume, and it forms a
characteristic property of substances. A full description
of the method of taking specific gravities has been given
at pages 241, 242, 243, 244.
COLOURLESS STONES.
THE DIAMOND.
(See also p. 257.)
Specific gravity, 3'48 to 3-52 ; hardness, 10. The
diamond is the hardest of all known substances. It is
the only substance which is capable of cutting glass,
although most gems will scratch glass ; hence it is the:
3 K 2
868 GEMS AND PRECIOUS STONES.
utmost term of hardness. When cut and polished, it
is the most brilliant gem. It occasionally becomes phos-
phorescent on exposure to light. The greater number of
diamonds are limpid and colourless, but many coloured
specimens are found ; as rose, yellow, orange, blue, green,
brown, or even black. It occurs in regular crystals,
octahedrons, dodecahedrons, and more complex forms ; see
figs. 131, 132, 133, 134.
The crystalline faces are often curved. The cleavage
is octahedral and highly perfect ; hence, although dia-
monds are so exceedingly hard, they are somewhat brittle,
owing to their tendency to facile cleavage. Like most gems,
FIG. 144. FIG. 145.
FIG. 147.
FIG. 146.
they become electrical by friction ; but it has been re-
marked that other gems do not, unless they have been
previously polished.
Composition (C) : Pure carbon.
The Matrix of the Diamond.
Professor H. Carvill Lewis gives the following interest-
ing notes on the matrix of the diamond, arrived at from a
microscopical study of the remarkable porphyritic perido-
tite which contains the diamonds in South Africa.
MATRIX OF THE DIAMOND. 869
The olivine, forming much the most abundant con-
stituent, is in porphyritic crystals, sometimes well bounded
by crystal faces, at other times rounded and with corrosive
cavities, such as occur in it in basaltic rocks. It rarely
encloses rounded grains of glassy bronzite, as has been
observed in meteorites. The olivine alters either into
serpentine in the ordinary way, or into an aggregate of
acicular tremolite crystals, the so-called 'pilit,' or becomes
surrounded by a zone of indigo blue bastite a new variety
of that substance. The olivine is distinguished by an un-
usually good cleavage in two directions.
Bronzite, Chrome diallage, and Smaragdite occur in
fine green plates, closely resembling one another. The
bronzite is often surrounded by a remarkable zone, with a
centric, pegmatitic, or chondritic structure, such as occurs
in certain meteorites. This zone is mainly composed of
wormlike olivine grains, but a mineral having the optical
characters of cyanite also occurs in this zone.
Biotite, a characteristic constituent, occurs in conspicu-
ous plates, often twinned, generally rounded, and distin-
guished by its weak pleochroism, a character peculiar to
the biotite of ultra-basic eruptive rocks. It alters by
decomposition into the so-called Vaalite.
Perofskite occurs in very numerous but small crystals,
which optically appear to be compound rhombic twins.
Pyrope is abundant in rounded red grains. Titanic
iron, chromic iron, and some fifteen other minerals are
also found. Eutile is formed as a secondary mineral
through the alteration of olivine into serpentine, being a
genesis of rutile not heretofore observed.
The chemical composition shows this to be one of the
most basic rocks known, and is a composition which by
calculation would belong to a rock composed of equal
parts of olivine and serpentine, impregnated by calcite.
The structure is at the same time porphyritic and brec-
ciated, being one characteristic of a volcanic rock which,
after becoming hard, had been subjected to mechanical
movements. It is a volcanic breccia, but not an ash or
tuff, the peculiar structure being apparently due to succes-
870 GEMS AND PRECIOUS STONES.
sive paroxysmal eruptions. A similar structure is known
in meteorites, with which bodies this rock has several
analogies. A large amount of the adjoining bituminous
shale is enclosed, and has been more or less baked and
altered. The occurrence of minute tourmalines is evidence
of fumarole action.
The microscopical examination supports the geological
data in testifying to the igneous and eruptive character of
the peridotite, which lies in the neck or vent of an old
volcano.
While belonging to the family of peridotites, this rock
is quite distinct in structure and composition from any
member of that group heretofore named. It is more basic
than the picrite porphyrites, and is not holocrystalline
like dunite or saxonite. It is clearly a new rock type,
worthy of a distinctive name. The name Kimberlite, from
the famous locality where it was first observed, is there-
fore proposed.
Kimberlite probably occurs in several places in Europe,
certain garnetiferous serpentines belonging here. It is
already known at two places in the United States : at
Elliott County, Kentucky, and at Syracuse, New York ;at
both of which places it is eruptive and post-carboniferous,
similar in structure and composition to the Kimberley
rock.
At the diamond localities in other parts of the world
diamonds are found either in diluvial gravels or in con-
glomerates of secondary origin, and the original matrix
is difficult to discover. Thus, in India and Brazil the
diamonds lie in a conglomerate with other pebbles, and
their matrix has not been discovered. Recent observations
in Brazil have proved that it is a mistake to suppose that
diamonds occur in itacolumite, specimens supposed to show
this association being artificially manufactured. But at
other diamond localities, where the geology of the region
is better known than in India or Brazil, the matrix of the
diamond may be inferred with some degree of certainty.
Thus, in Borneo diamonds and platinum occur only in
those rivers which drain a serpentine district, and on Tanah
MATRIX OF THE DIAMOND. 871
Laut they also lie on serpentine. In New South Wales,
near each locality where diamonds occur, serpentine also
occurs, and is sometimes in contact with carboniferous
shales. Platinum, also derived from eruptive serpentine,
occurs here with the diamonds. In the Urals, diamonds
have been reported from four widely separated localities,
and at each of these, as shown on Murchison's map,
serpentine occurs. At one of the localities the serpentine
has been shown to be an altered peridotite. A diamond
has been found in Bohemia in a sand containing pyropes,
-and these pyropes are now known to have been derived
from a serpentine altered from a peridotite. In North
Carolina a number of diamonds and some platinum have
been found in river sands, and that state is distinguished
from all others in eastern America by its great beds
of peridotite and its abundant serpentine. Finally, in
northern California, where diamonds occur plentifully
.and are associated with platinum, there are great out-
bursts of post-carboniferous eruptive serpentine, the ser-
pentine being more abundant than elsewhere in North
America. At all the localities mentioned chromic and
titanic iron ore occur in the diamond-bearing sand, and
both of these minerals are characteristic constituents of
serpentine.
All the facts thus far collected indicate serpentine,
in the form of a decomposed eruptive peridotite, as the
original matrix of the diamond.
Speaking of diamonds, Professor Orton says that few
things are so unpromising and unattractive as these gems
in their native state. Hence their slow discovery. There
is little doubt that diamonds exist in many places as yet
unknown, or where their presence is unsuspected. It is
very difficult for the unpractised eye to distinguish them
from crystals of quartz or topaz. The colour constitutes
the main difficulty in detecting their presence. They are
of various shades of yellowish -brown, green, blue, and rose-
Ted, and thus closely resemble the common gravel by which
they are surrounded. Often they are not unlike a lump
of gum arabic, neither brilliant nor transparent. The
872 GEMS AND PRECIOUS STONES.
finest, however, are colourless, and appear like rock
crystals.
In Brazil, where great numbers of diamonds, chiefly of
small size, have been discovered, the method of searching
for them is to wash the sand of certain rivers in a manner
precisely similar to that employed in the gold fields
namely, by prospecting pans. A shovelful of earth is
thrown into the pan, which is then immersed in water, and
gently moved about. As the washing goes on, the pebbles r
dirt, and sand are removed, and the pan then contains
about a pint of thin mud. Great caution is now observed,
and ultimately there remains only a small quantity of sand.
The diamonds and particles of gold, if present, sink to the
bottom, being heavier, and are selected and removed by
the practised fingers of the operator. But how shall the
gems be detected by one who in a jeweller's shop could
not separate them from quartz or French paste ? The
difficulty can only be overcome by testing such stones as
may be suspected to be precious. Let these be tried by
the very sure operation of attempting to cut with their
sharp corners glass, crystal, or quartz. When too minute
to be held between the finger and thumb, the specimens
may be pressed into the end of a stick of hard wood and
run along the surface of window glass. A diamond will
make its mark, and cause, too, a ready fracture in the line
over which it has travelled. It will also easily scratch
rock crystal, as few crystals will.
But a more certain and peculiar characteristic of the
diamond lies in the form of its crystals. The ruby and
topaz will scratch quartz, but no mineral which will scratch
quartz has the curved edges of the diamond. In small crys-
tals this peculiarity can be seen only by means of a magnify-
ing glass ; but it is invariably present. Interrupted, convex,,
or rounded angles are sure indications of genuineness.
Quartz crystal is surrounded by six faces, the diamond
by four. The diamond breaks with difficulty ; and hence
a test sometimes used is to place the specimen between
two hard bodies, as a couple of coins, and force them
together with the hands. Such a pressure will crush a
THE DIAMOND. 87$
particle of quartz, but the diamond will only indent the
metal.
The imperfections of the diamond, and, in fact, of all
cut gems, are made visible by putting them into oil of
cassia, when the slighest flaw will be seen.
If a rough diamond resemble a drop of clear spring
water, in the middle of which you perceive a strong light ;
or if it has a rough coat, so that you can hardly see through
it, but white, and as if made rough by art, yet clear of
flaws or veins ; or if the coat be smooth and bright, with
a tincture of green in it, it is a good stone. If it has a
milky cast or a yellowish -green coat, beware of it. Eough
diamonds with a greenish crust are the most limpid when
cut.
Diamonds are found in loose pebbly earth, along with
gold, a little way below the surface, towards the lower
outlet of broad valleys, rather than upon the ridges of the
adjoining hills.
Prof. Silliman, on examining with the microscope a
small parcel of the sand resulting from the hydraulic treat-
ment of ores, found that they abounded in fine colourless
zircons, along with crystals of topaz, fragments of quartz,,
grains of chrome iron, and titanic acid, and globular bodies
of a very high refractive power, which he believes to be
diamonds.
Mr. J. Torry, in a single sample of the sands washed
from the gold ores of Nicaragua, found twenty mineral
species, some of them very rare.
For estimating the specific gravity of certain minerals,,
and separating diamond dust or small diamonds and other
gems from quartz, sand, &c., Mr. E. Sonstadt * uses a
solution in water of pure potassium iodide and pure mer-
curic iodide in cold water. It should be diluted to such
a strength that quartz will just float in it. (See ante>
pages 243, 244.)
* ' Chemical News,' March 20, 1874.
874
GEMS AND PRECIOUS STONES.
QUARTZ.
Specific gravity, 2 -55 to 2*7 ; hardness, 7. Quartz
occurs in many forms, and has often by inexperienced
persons been mistaken for the diamond, owing to the lustre
of its crystals and its considerable hardness. It, however,
can always be distinguished from the diamond by its crys-
FIG. 150.
FIG. 148.
FIG. 149.
FIG. 151.
FIG. 152.
talline faces, hardness, and specific gravity (see example in
Table I.)
It usually occurs in six-sided prisms, more or less
modified, terminated with six-sided pyramids. Traces of
cleavage are seldom or ever apparent. Some of its salient
forms are shown in figs. 148, 149, 150, 151, 152, 153, 154.
COLOURLESS STONES.
875
FIG. 153.
FIG. 154.
Some crystals are as pellucid as glass ; others, however,
assume all the shades of colour mentioned in the case of
the diamond.
Composition (Si0 2 ) : Pure silica or silicic acid.
WHITE ZIRCON.
Specific gravity, 4'44 to 4-8 ; hardness, 7'5. This stone
is often found crystallised in nature in four-sided prisms,
terminated by four-sided or rhomboidal or triangular
FIG. 156.
FIG. 157.
FIG. 155.
pyramids, and other forms. See figs. 155, 156, 157, 158,
159.
These stones are often employed in jewellery under the
name of ' rough diamonds.' They often occur brownish-
red and brown, red, yellow, and grey ; these varieties will
876 GEMS AND PEECIOUS STONES.
be treated under their appropriate heads. It can be readily
distinguished from the diamond and quartz by hardness
FIG. 158. FIG. 160.
arid specific gravity ; also by the action of strong hydro-
chloric acid, which, if dropped on the diamond or quartz,
and allowed to remain for a little time, produces no change,
but if a zircon be so treated, the spot on which the acid
was placed remains dull.
Composition (Zr 2 3 ,Si0 2 ) :
Zirconia . 67'2
Silicic acid . . . . . . 33'5
100-7
WHITE SAPPHIRE.
Specific gravity, 3-97 to 4-27 ; hardness, 9. This stone,
in hardness, is next to the diamond. The sapphire occurs
variously coloured ; other colours will be discussed under
their appropriate heads. It crystallises in the rhombohe-
dric system, usually in six-sided prisms, but often so very
rough as not to be readily distinguishable. May be dis-
tinguished by gravity and hardness from all the preceding.
Composition (A1 2 3 ) : Pure alumina.
WHITE TOPAZ.
Specific gravity, 3'54 ; hardness, 9. This variety of
topaz, known for its limpidity by the term ' gouttes d'eau,"
when polished has nearly the same lustre as the diamond ;
COLOURLESS STONES.
877
the topaz, however, occurs of many colours see hereafter.
It crystallises in the right rectangular prismatic system.
Some of its natural forms are shown in figs. 161, 162, 163,
164, 165, 166.
It is readily rendered electric, and retains its electricity
for a very considerable time ; it is also pyro-electric, or
becomes electric when heated, a property by which it is
FIG. 161. FIG. 162. FIG. 163.
FIG. 164.
FIG. 165.
FIG. 166.
distinguished from the diamond, its specific gravity being
so similar that it cannot be made available as a means of
discriminating between the two stones. From the other
stones in this group, with the exception of the sapphire,
it is readily distinguished by its hardness and gravity, and
from the latter by its gravity and pyro-electricity.
Composition :
Silica
Alumina
Fluorine
34-2
57-5
7-8
99-5
878
GEMS AND PRECIOUS STONES.
TABLE I.
COMPARATIVE TABLE OF THE WEIGHTS OF COLOURLESS STONES
WEIGHED IN AlR AND WATER.
Weight in
Weight in Water
Air
White
White
White
White
White
Grains
Zircon
Sapphire
Topaz
Diamond
Quartz
1
0-775
0-766
0-716
0-715
0-611
4
3-10
3-06 2-86
2-86
2-42
8
6-20
6-12 5-72
5-72
4-86
12
9-30
9-18 8-58 8-58
7-31
16
12-40
12-25 11-55 11-45
9-75
20
15-50
15-31
14-42 14-31
12-19
24
18-60
18-37
17-28 17-17
14-64
28
21-70
21-44 | 20-16
20-13
17-08
32
24-80
24-51
23-01
22-90
19-53
36
27-90
27-57
25-88
25-76
11-98
40
31-00
30-64
28-75
28-63
24-43
44
34-10
33-71
31-61
31-49
26-88
48
37-20 36-76
34-47
34-35
29-32
52
40-30
39-82
37-34
37-21
31-77
56
43-40
42-89
40-20
40-17
34-21
60
46-50
45-95
43-06
42-94
36-66
64
49-60
49-01
45-93
45-80
39-11
68
52-70
52-07
48-90
48-66
41-56
72
55-80
55-14
51-77
51-52
44-00
76
58-90
58-21
54-63
54-38
46-44
80
62-00
61-28
57-49
57-24
48-88
84
65-10
64-34
60-35
60-12
51-32
88
68-20
67-41
63-22
62-97
53-76
92
71-30
70-47
66-08
65-30
56-21
96
74-40
73-54
68-94
68-69
58-65
100
77-50
76-60
71-80
71-55
61-09
Specific
Gravity
\ 4-44
4-27
3-54
3-52
2-55
Example of the use of Table /.* A colourless stone,
weighing 40 grains in air, is reduced to 24-43 in water.
Look in the first column to 40, and then trace along its
horizontal line until a number very nearly approaching
24-43 is found ; refer then to the heading of the table,
above the number found, and the name there expressed
will be that of the stone examined. Supposing, however,
the weight of the stone be 41 grains, still the number
24-43 will be the nearest in the table, and -611 must be
added to it, as that sum would be the weight of 41 grains
* The Tables of Comparative Weights were calculated by Brard.
YELLOW STONES. 879
of quartz or water. From the numbers obtained by cal-
culation also can the specific gravity be estimated. If
this course be pursued, refer to the bottom line of the
table for a corresponding number, and to the heading of the
table for the name of the stone. When the weight is any
even number of grains (that is, without fractions), the
readiest way is to refer to the table (first column) for the
number of grains, and then to the horizontal line to corre-
sponding number obtained, which is the weight in water.
Diamond and topaz, however, have very nearly equal
densities, and a second characteristic must be had recourse
to in order to determine the nature of two stones which
have an equal weight in water. This auxiliary character
is the development of electricity by heat, a phenomenon
exhibited by the topaz but not by the diamond. The test
of hardness may be also resorted to.
YELLOW STONES.
YELLOW ZIRCON ( JARGON).
The crystalline form, characteristics, and composition
of this stone have been described under the head ' White
Zircon.'
YELLOW SAPPHIRE.
Characteristics, &c., described under ' White Sapphire.'
CYMOPHANE (CHRYSO BERYL).
Specific gravity, 3'65 to 3-89 ; hardness, 8-5. The
cymophane is nearly as hard as the sapphire, harder than
the topaz and the emerald ; it readily scratches quartz.
Its colour is greenish-yellow, and has been placed in the
list of yellow stones rather than green, because usually the
yellowish tint is the most decided. This tint, which is very
agreeable in itself, is often relieved by a small spot of light
of a bluish- white tinge, which moves from point to point
of the stone as the position of the latter is varied. It is
S80
GEMS AND PRECIOUS STONES.
rarely found in regular crystals, but more generally occurs
in rolled and rounded masses. For some of its iorms,
however, see figs. 167, 168, 169, 170.
FIG. 167.
FIG. 168.
FIG. 169.
FIG. 170.
Composition : No. 1 is a sample from the Brazils ;
No. 2, from Siberia.
i.
Alumina 78-10
Glucina 17-94
Iron oxide . . . . . 4*46
Chromium oxide ....
Copper and lead oxides ...
100-50
2.
78-92
18-02
3-12
0-36
0-29
100-71
YELLOW TOPAZ.
The general characteristics of this stone are described
under ' White Topaz.'
YELLOW TOURMALINE.
Specific gravity, 3-00 to 3-22 ; hardness, 7 to 7'5. The
tourmaline becomes electrical by heat ; one portion of a
crystal attracts light bodies, the other repels them. Its
YELLOW STONES.
881
colour is very varied. The tourmaline has a vitreous
fracture. It occurs in semicrystalline prisms of irregular
FIG. 171.
FIG. 172.
FIG. 173.
FIG. 174.
FIG. 175.
FIG. 176.
form, generally deeply striated, and in prisms of six or
more sides, variously terminated, one end usually differing
from the other.
Figs. 171, 172, 173, 174, 175, and 176 represent some
of the forms of this mineral.
YELLOW EMEKALD.
Specific gravity, 273 to 2-76 ; hardness, 7-5 to 8. The
emerald occurs of many colours ; its tint par excellence is-
3 L
882
GEMS AND PKECIOUS STONES.
green ; but there are many varieties tinged more or less
yellow or blue, and they even occur white. Its fracture
FIG. 178.
FIG. 177.
FIG. 179.
FIG. 180.
FIG. 181.
FIG. 183.
FIG. 182.
is vitreous, brilliant, and undulating. Its common form
is the hexahedral prism, sometimes deeply striated longi-
YELLOW STONES.
883
tudinally. It readily cleaves parallel to all the planes of
its primary form the hexahedral prism.
The above are some of the forms it assumes :
178, 179, 180, 181, 182, and 183.
Composition :
figs. 17
Glucina
Silica
Alumina
Iron oxide
15-50
66-45
16-75
60
99-30
The
mium
ie green varieties contain a small quantity of chro-
oxide.
TABLE II.
COMPARATIVE TABLE OF THE WEIGHTS OF YELLOW STONES
WEIGHED IN AlE AND WATER.
Weight
in air
Weight in Water
Grrftins
Yellow
Yellow
Yellow
Yellow
Yellow
Yellow
Yellow
Zircon
Sapphire
Cymophane
Topaz
Tourmaline
Emerald
Quartz
i
0-775
0-766
9-738
0-716
0-690
0-633
0-611
4
3-10
3-06
2-95
2-86
2-76
2-53
2-42
8
6-20
6-12
5-90
5-72
5-52
5-06
4-86
12
9-30
9-18
8-85
8-58
8-28
7-59
7-31
16
12-40
12-25
11-80
11-55
11-04
10-12
9-75
20
15-50
15-31
14-75
14-42
13-80
12-65
12-19
24
18-60
18-07
17-70
17-28
16-56
15-19
14-04
28
21-70
21-44
20-65
20-15
19-32
17-72
17-08
32
24-80
24-51
23-60
23-01
20-08
20-25
19-53
36
27-90
27-57
26-55
25-88
24-84
22-77
21-98
40
31-00
30-64
29-50
29-75
27-60
25-30
24-43
44
34-10
33-71
32-45
31-61
30-36
27-83
26-88
48
37-20
36-76
35-40
34-47
33-12
30-36
29-32
52
40-30
39-82
38-35
37-34
35-88
32-89
31-77
56
43-40
42-89
41-30
40-20
38-64
35-43
34-21
60
46-50
45-95
44-25
43-06
41-40
37-94
36-66
64
49-60
49-01
47-20
45-93
44-16
40-47
39-11
68
52-70
52-08
50-15
48-90
46-92
43-00
41-56
72
55-80
55-14
53-10
51-77
49-68
45-53
44-00
76
58-90
58-21
56-05
54-63
52-44
48-07
46-44
80
62-00
61-28
59-00
57-49
55-20
50-60
48-88
84
65-10
64-34
61-95
60-35
57-96
53-13
51-32
88
68-20
67-41
64-90
63-22
60-72
55-66
53-76
92
71-30
70-47
67-85
66-08
63-48
58-19
56-21
96
74-40
73-54
70-80
68-94
66-24
60-72
58-65
100
77-50
76-60
73-75
71-80
69-00
63-25
61-09
Specific
Gravity
j- 4-44
4-27
3-89
3-53
3-22
2-72
2-55
3 L 2
884
GEMS AND PRECIOUS STONES.
YELLOW QUARTZ.
For the characteristics, hardness, &c., of this mineral,
see ' White Quartz/
BKOWN AND FLAME-COLOURED STONES.
ZIRCON (HYACINTH).
For characteristics, &c., see ' White Zircon/
VERMEIL GARNET, NOBLE GARNET, ALMANDINE.
Specific gravity, 4 to 4*2 ; hardness, 6-5 to 7*5. There
are very many varieties of garnet, variously coloured ; but
FIG. 184.
FIG. 185.
FIG. 186
their crystalline form a rhombic dodecahedron, more
or less modified is a distinguishing characteristic. The
colouring matter of the garnet is iron. Figs. 184, 185, 186,
187, and 188 represent some of its crystalline forms.
BROWN AND FLAME-COLOURED STONES.
FIG. 187. FIG. 188.
885
Composition :
Silica
Alumina .
Iron oxide
Manganese oxide
33-75
27-25
36-00
25
97-25
TABLE III.
COMPARATIVE TABLE OF THE WEIGHTS OF BROWNISH AND FLAME-
COLOURED STONES WEIGHED IN AIR AND WATER.
Weight in
Air
Weight in water
Grains
Hy acinthine Zircon
Vermeil Garnet
Essonite
Tourmaline
1
0-775
0-750
0-710
0-690
4
3-10
3-00
2-87
2-76
8
6-20
6-00
5-74
5-52
12
9-30
9-00
8-61
8-28
16
12-40
12-00
11-48
11-04
20
15-50
15-00
14-35
13-80
24
18-60
18-00
17-22
16-56
28
21-70
21-00
20-09
19-32
32
24-80
24-00
22-96
22-08
36
27-90
27-00
25-83
24-84
40
31-30
30-00
28-70
27-60
44
34-10
33-00
31-57
30-36
48
37-20
36-00
34-44
33-12
52
40-30
39-00
37-31
35-88
56
43-40
42-00
40-18
38-64
60
46-50
45-00
43-05
41-40
64
49-60
48-00
45-92
44-16
68
52-70
51-00
48-79
46-92
72
55-80
54-00
51-66
49-68
76
58-90
57-00
54-53
52-44
80
61-00
60-00
57-40
55-20
84
65-10
63-00
60-27
57-96
88
68-20
66-00
63-14
60-72
92
71-30
69-00
66-01
63-48
96
74-40
72-00
68-88
66-24
100
77-50
75-00
71-75
69-00
Specific
Gravity
} 4-44
4-00
3-54
3-22
886 GEMS AND PKECIOUS STONES.
ESSONITE, CINNAMON STONE.
Specific gravity, 3*5 to 3 -6. This stone has an agree-
able orange-yellow tinge, which becomes a warm and
brilliant tint when 'the mass is large. This stone is not
usually found crystalline, but in irregular forms and
masses, which are characterised by fissures in all direc-
tions.
Composition :
Silica 38-80
Alumina 21-20
Lime 31-25
Iron oxide with small quantities of Potash "1 Q-Q
and Magnesia /
97-75
TOURMALINE.
For the characteristics of this mineral see 'Yellow
Tourmaline.'
KED AND KOSE-COLOURED STONES.
EED SAPPHIEE (ORIENTAL RUBY).
For characteristics, crystalline form, &c., see ' White
Sapphire.'
DEEP RED GARNET, NOBLE GARNET.
For characteristics, &c., see ' Vermeil Garnet.'
SPINEL RUBY.
Specific gravity, 3*5 to 3-6 ; hardness, 8. The spinel
readily scratches quartz, but is scratched by the sapphire.
Its special colour is red, approaching a rose tint : this
tinge, however, undergoes various modifications, such as
scarlet, red, rose, yellowish-red, and reddish-purple : it
is also found blue and black. Its fracture is flattish-
conchoidal, with a splendent vitreous lustre. It occurs
crystallised in regular octahedrons, sometimes having
their edges replaced as in macles : sometimes it assumes
the globular form. The spinel may be distinguished from
the true ruby and the garnet by hardness and specific
EED AND KOSE-COLOUKED STONES.
887
gravity ; and from reddish topaz, which possesses nearly
the same specific gravity, by its electric properties.
Composition of spinel ruby :
Silica -02
Alumina . 69-01
Magnesia 26'21
Iron protoxide 0'71
Chromium oxide . . . . . 1-11
99-06
EEDDISH TOPAZ.
For characteristics, &c., see ' White Topaz.'
KED TOURMALINE.
For characteristics, &c., see 4 Yellow Tourmaline.'
TABLE IV.
COMPARATIVE TABLE OF THE WEIGHTS OF EED OR EOSE-COLOURED
STONES WEIGHED IN AIR AND WATER.
Weight
Weight in Water
in Air
Grains
Red Sapphire
(True Ruby)
Deep Garnets
Spinel
Smoke or Red
Topaz
Red Tourmaline
1
0-766
0-750
0-722
0-716
0-690
4
3-060
3-700
2-880
2-860
2-760
8
6-120
6-000
5-770
5-720
5-520
12
9-180
9-000
8-660
8-585
8-280
16
12-250
12-000
11-550
11-550
11-040
20
15-310
15-000
14-440
14-420
13-800
24
18-370
18-000
17-330
17-280
16-560
28
21-440
21-000
20-220
20-150
19-320
32
24-510
24-000
23-110
23-610
22-080
36
27-570
27-000
26-000
25-880
24-840
40
30-640
30-000
28-880
28-750
27-600
44
33-710
33-000
31-770
31-610
30-360
48
36-760
36-000
34-660
34-470
33-120
52
39-820
39-000
37-550
37-340
35-880
56
42-890
42-000
40-440
40-200
38-640
60
44-950
45-000
43-300
43-060
41-400
64
49-010
48-000
46-220
45-930
44-160
68
52-080
51-000
49-110
48-900
46-920
72
55-140
54-000
51-990
51-770
49-680
76
58-210
57-000
54-880
54-630
52-440
80
61-280
60-000
57-770
57-490
52-200
84
64-340
63-000
60-660
60-350
57-960
88
67-410
66-000
63-550
63-220
60-720
92
70-470
69-000
66-440
66-080
63-480
96
73-540
72-000
69-330
68-940
66-240
100
76-600
75-000
72-220
71-800
69-000
Specific
Gravity
} 4-270
4-000
3-600
3-530
3-220
888 GEMS AND PRECIOUS STONES.
BLUE STONES.
BLUE SAPPHIEE.
For characteristics, &c., see ' White Sapphire.'
DISTHENE, CYANITE.
Specific gravity, 3'5 to 3*7 ; hardness, 5 to 7. Fine
specimens of disthene possess a bright blue colour, which
passes insensibly into a deep sky blue. Its transparency
is nearly perfect, and it presents small pearly reflections,
which add to the beauty of its colour. The primary form
of its crystals is a doubly oblique prism, and they cleave
very readily in the direction of their length. It can be
FIG. 189.
FIG. 190, FIG. 191.
readily distinguished from the sapphire by its being less
hard, as also by its specific gravity. Figs. 189, 190, and
191 represent some of its crystalline forms.
Composition of a specimen from St. Gothard :
Silica 43-0
Alumina 55-0
Iron oxide -5
98-5
BLUE TOPAZ.
For characteristics, &c., see fc White Topaz.' Blue
topaz and disthene, having the same specific gravity, may
by that test alone be confounded with each other ; but
the appearance of each is so different that they can
rarely be confounded. If, however, the electrical test be
applied, no fear of mistaking one for the other need be
entertained, as only the topaz becomes electrical.
BLUE STONES.
BLUE TOURMALINE.
For characteristics, &c., see ' Yellow Tourmaline.'
BLUE BERYL.
For characteristics, &c., see 'Emerald.' The tint and
appearance of this stone and that of the blue topaz are
so similar that they cannot be distinguished by that test ;
their specific gravities, however, are so different that they
may, by this simple means, be readily discriminated.
DICHROITE, WATER SAPPHIRE.
Specific gravity, 2*56 to 2-65 ; hardness, 7 to 7'5. The
chief characteristic of this stone is that it possesses a
double colour ; that is, it is a fine blue or a normal yellow,
as it is viewed in the direction of its base, or the planes
of a hexahedral prism, which is its crystalline form. It
can be thus readily distinguished, as also by its having
nearly the same specific gravity as quartz, and thus being
the lightest of the blue stones. Composition :
Silica. . .
Alumina
Magnesia .
Iron protoxide .
Manganese protoxide
Loss in fire (water ? )
48-35
31-71
10-16
8-12
33
60
99-27
TURQUOISE.
Specific gravity, 2-8 to 3 ; hardness, 5 to 6. This stone
has not been placed in the list of specific gravities, as it
can be so readily detected by its appearance. It is bright
or greenish-blue in colour ; its aspect is earthy or compact.
It scratches apatite, and even glass ; but is scratched by
quartz. It occurs filling fissures, or forming concretions
in siliceous and argillo-ferruginous rocks. Composition :
Phosphoric acid 17-86
Alumina .
Silica
Iron peroxide .
Lime .
Water and fluoric acid
10-01
8-90
36-82
0-15
25-95
99-69
890
GEMS AND PRECIOUS STONES.
TABLE V.
COMPARATIVE TABLE OF THE WEIGHTS OF BLUE STONES WEIGHED
IN AIR AND WATER.
Weight iii Water
Weight
in Air
|
Grains
Blue
Sapphire
Disthene
Cyanite
Blue
Topaz
Tourmaline
Blue
Beryl
Dichroite,
Water
Sapphire
1
0-766
0-717
0-716
0-690
0-633
0-622
4
3-06
2-87
2-86
2-16
2-53
2-49
8
6-12
5-74
5-72
5-52
5-06
4-98
12
9-18
8-61
8-58
8-28
7-59
7-47
16
12-25
11-48
11-45 11-04
10-12
9-96
20
15-31
' 14-35
14-42
13-80
12-65
12-45
24
18-37
17-22
17-18
16-56
15-19
14-94
28
21-44
20-09
20-05
19-32
17-72
17-43
82
24-51
22-96
22-91
20-08
20-25
19-92
36
27-57
25-83
25-78
24-84
22-77
22-41
40
30-64
28-70
28-65
27-60
25-30
24-90
44
33-71
31-57 31-51
30-66
27-83
27-39
48
36-76
34-44 34-37
33-12
30-36
29-88
52
39-82
37-31 37-24
35-88
32-89
32-37
56
42-89
40-18 40-10
38-64
35-43
34-88
60
45-95
43-05 42-96
41-40
37-94
37-35
64
49-01
45-92 45-83
44-16
40-47
39-84
68
52-08
48-79 48-80
46-92
43-00
42-33
72
55-14
51-66 51-67
49-68
45-53
44-82
76
58-21
54-53 54-53
52-44
48-07
47-31
80
61-28
57-40 57-49
55-20
50-60
49-80
84
64-34
60-27 60-25
57-96
53-13
52-29
88
67-41
63-14 63-12
60-72
55-66
54-78
92
70-47
66-01
65-98
63-48
58-19
57-27
96
73-54
68-88
63-84
66-24
60-72
59-76
100
76-60
71-75
71-70
69-00
63-25
62-25
Specific
Gravity
| 4-27
3-54
3-53
3-22
2-72
2-65
VIOLET STONES.
VIOLET SAPPHIRE.
For characteristics, &c., see 4 White Sapphire.'
VIOLET TOURMALINE.
For characteristics, &c., see ' Yellow Tourmaline.'
VIOLET QUARTZ, AMETHYST.
For characteristics, &c., see ' White Quartz.'
GREEN STONES.
891
TABLE VI.
COMPARATIVE TABLE OF THE WEIGHTS OF VIOLET STONES WEIGHED
IN AIR AND WATER.
Weight in Air
Weight in Air
Grains
Violet Sapphire
Violet Tourmaline
Amethystine Quartz
(Amethyst;
1
0-766
0-690
0-611
4
3-06
2-76
2-42
8
6-12
5-52
4-86
12
9-18
8-28
7-31
16
12-25
11-04
9-75
20
15-31
13-80
12-19
24
18-37
16-56
14-64
28
21-44
19-32
17-08
32
24-51
20-08
19-53
36
27-57
24-84
21-98
40
30-64
27-60
24-43
44
33-71
30-36
26-88
48
36-76
33-12
29-32
52
39-82
35-88
31-77
56
42-89
38-64
34-21
60
45-95
41-40
36-66
64
49-01
44-16
39-11
68
52-02
46-92
41-56
72
55-14
49-68
44-00
76
58-21
52-44
46-44
80
61-28
55-20
48-88
84
64-34
57-96
51-32
88
67-41
60-72
53-76
92
70-47
63-48
56-21
96
73-54
66-24
58-65
100
76-60
69-00
61-09
Specific Gravity
4-27
3-22
2-55
GBEEN STONES.
GREEN SAPPHIRE.
For characteristics, &c., see 'Yellow Emerald.'
PERIDOT, CRYSOLITE.
Specific gravity, 3*3 to 3-5 ; hardness, 6*5 to 7.. This
stone has a more or less deep olive or yellowish -green
colour. It is more generally found in rolled grains than
in regular prismatic crystals. It is possessed in a very
high degree of double refraction. Figs. 192, 193, 194,
and 195 represent some of its crystalline forms.
392
GEMS AND PEECIOUS STONES.
GEEEN TOUEMALINE.
For characteristics, see ' Yellow Tourmaline.
EMEEALD.
For characteristics, see ' Yellow Emerald.'
FIG. 192.
FIG. 193.
FIG. 194.
FIG. 195.
AQUA-MAEINE.
This stone possesses a very pale green tinge. For
other characteristics, see ' Yellow Emerald.'
CHEYSOPEASE.
This mineral is a green-coloured quartz, and can be
readily recognised by referring to the characteristics of
quartz.
GREEN STONES.
TABLE VII.
COMPARATIVE TABLE OF THE WEIGHTS OF GREEN STONES WEIGHED
IN AIR AND WATER.
Weight
1T1 A !
Weight in Water
in Air
Grains
Green
Sapphire
Peridot
Green
Tourmaline
Emerald
Aqua-marine
Chrysoprase
i
0-766
0-708
0-690
0-633
0-633
0-611
4
3-06
2-83
2-76
2-53
2-53
2-42
8
6-12
5-66
5-52
5-06
5-06
4-86
12
9-18
8-49
8-28
7-59
7-59
7-31
16
12-25
11-32
11-04
10-12
10-12
9-75
20
15-31
14-16
13-18
12-65
12-65
12-19
24
18-37
16-99
16-56
15-19
15-19
14-64
28
21-44
19-82
19-32
17-72
17-72
17-08
32
24-51
22-65
22-08 20-25
20-25
19-53
36
27-57
25-48
24-84
22-77
27-77
21-98
40
30-64
28-32
27-60
25-30
25-30
24-43
44
33-71
31-15 30-36
27-83
27-83
36-88
48
36-76
33-98
33-12
30-36
30-36
29-32
52
39-82
36-81
35-88 32-89
32-89
31-77
56
42-89
39-64
38-64
35-43
35-43
34-21
60
45-95
42-48
41-40
37-94
37-94
36-66
64
49-01
45-31
44-16
40-47
40-47
39-11
68
52-08
48-14
46-92
43-00
43-00
41-56
72
55-14
50-97
49-68
45-53
45-53
44-00
76
58-21
53-80
52-44
48-07
48-07
46-44
80
61-28
56-64
55-20
50-60
50-60
48-88
84
64-34
59-47
57-96
53-13
53-13
51-32
88
67-41
62-30
60-72
55-66
55-66
53-76
92
70-47
65-13
63-48
58-19
58-19
56-21
96
73-54
67-96
66-24
60-72
60-72
58-65
100
76-60
70-80
69-00
63-25
63-25
61-09
Specific
Gravity
| 4-27
3-42
3-22
2-72
2-72
2-56
STONES POSSESSING A PLAY OF COLOUES
(CHATOYANT).
In the following list of stones no regard has been paid
to absolute colours, but only to the play of colours the
stones exhibit. This play or reflection is of two kinds :
in some, as the sapphires, it appears as a white star with
six rays, on a blue, red, or yellow ground ; or on a purple
ground in the garnet. In others it is but a point or mass
of pearly light, which sometimes appears to occupy the
whole of the stone, and varies according to the inclination
given to the stone. The cymophane, cry soli te quartz,
Egyptian emerald, felspar, and cat's eye belong to this
class.
GEMS AND PRECIOUS STONES.
The specific gravities of such stones as the opal, &c.,
have not been given, as their appearance sufficiently charac-
terises them.
SAPPHIRE.
For characteristics, &c., see 'White Sapphire.'
GARNET.
For characteristics, &c., see ' Vermeil Garnet.'
See ' Cymophane.'
CYMOPHANE.
ANTIQUE EMERALD.
For characteristics, &c., see ' Yellow Emerald.'
QUARTZ.
See ' White Quartz.'
FIG. 196. FIG. 197. FIG. 198.
FIG. 199.
FIG. 201.
FIG. 200.
STONES POSSESSING A PLAY OF COLOURS.
895
FELSPAR, NACREOUS FELSPAR, FISH-EYE, ETC.
Specific gravity, 2'3 to 2-5 ; hardness 4'5 to 5. This
species of felspar has a lamellar texture. It will be seen
by the lowness of its specific gravity that it cannot be
readily confounded with other stones. In appearance its
transparency is nebulous, and it presents pearly white
reflections, which float about and vacillate in proportion
as its position changes. The foregoing are some of the
forms of felspar ; see figs. 183, 184, 185, 186, 187, and 188.
Composition :
Potash 5-26
Silica 52-90
Lime 25-20
Water ....... 16-00
Fluoric acid 0*82
100-18
TABLE VIII.
COMPARATIVE TABLE OF THE WEIGHTS OF STONES POSSESSING A
PLAY OF COLOURS (CHATOYANT).
Weight in Water
Weight
in Air
Grains
Sapphires
Garnets
Gymophane
Antique
Emerald
Quartz
Felspar
1
0-766
0-750
0-738
0-633
0-611
0-592
4
3-06
3-00
2-95
2-53
2-42
2-37
8
6-12
6-00
5-90
5-06
4-86
4-74
12
9-18
9-00
8-85
7-59
7-31
7-11
16
12-25
12-00
11-80
10-12
9-75
9-47
20
15-31
15-00
14-75
12-65
12-19
11-84
24
18-37
18-00
17-70
15-19
14-64
14-20
28
21-44
21-00
20-65
17-72
17-08
16-57
32
24-51
24-00
23-60
20-25
19-53
18-94
36
27-57
27-00
26-55
22-77
21-98
21-31
40
30-64
30-00
29-50
25-30
24-43
23-68
44
33-71
33-00
32-46
27-83
26-88
26-05
48
36-76
36-00
35-40
30-36
29-32
28-42
52
39-84
39-00
38-35
32-89
31-77
30-79
56
42-89
42-00
41-30
35-43
34-21
33-15
60
45-95
45-00
44-25
37-94
36-66
35-52
64
49-01
48-00
47-20
40-47
39-11
37-88
68
52-07
51-00
50-15
43-00
41-56
40-25
72
55-14
54-00
53-10
45-53
44-00
42-62
76
58-21
57-00
56-05
48-07
46-44
44-99
80
61-28
60-00
59-00
50-60
48-88
47-36
84
64-34
63-00
61-95
53-13
51-32
49-73
88
67-47
66-00
64-90
55-66
53-76
52-10
92
70-47
69-00
67-85
58-19
56-21
54-47
96
73-54
72-00
70-80
60-72
58-65
56-84
100
76-60
75-00
73-75
63-25
61-09
59-21
Specific
Gravity
| 4-27
4-00
3-89
2-72
2-55
2-45
896 GEMS AND PRECIOUS STONES.
GLASS AND AETIFICIAL GEMS.
Glass is often used to imitate gems, but can be easily
distinguished by the following characters :
1. Inferior brilliancy. Though in many cases artificial
gems have a fine lustre, they are invariably soft. The
materials which communicate brilliancy to glass impair
its hardness ; and the result is that glass gems, when
examined by a lens, are generally found to have blunt or
jagged corners and edges, and surfaces covered with
minute, irregular scratches. This is invariably the case
after the glass gem has been a little in use, and thus the
brilliancy soon becomes impaired.
2. Inferior hardness. Artificial gems can be scratched
with a knife, using a slight pressure. Faint scratches are
made visible by breathing upon them, whereas true gems
retain the original brilliancy which they possessed on
leaving the lapidary's wheel. The polished faces of true
gems cannot be scratched with a knife ; moreover, after
long wear, they show no signs of scratches.
3. Polished gems become readily electric by friction,
particularly the topaz and diamond ; but glass imitations
require much longer friction to produce the same effect,
and also retain the electric power for a shorter time.
4. Glass is fusible in the blowpipe flame.
APPENDIX.
TABLE L
Showing the Quantity of FINE GOLD in 1 oz. of any ALLOY to
of a Carat Grain and the MINT VALUE of 1 oz. of each
Alloy.
FINE GOLD,
Per Ounce
CARAT GOLD,
Per Ounce
STERLING VALUE,
Per Ounce
Oz.
Drvts. Grs.
Carats
GTS. E
'sjhtlis
s.
d.
1
000
24
4
4
11-4545
19
23
375
23
3
7
4
4
10-1271
19
22
750
23
3
6
4
4
8-7997
19
22
125
23
3
5
4
4
7-4723
19
21
500
23
3
4
4
4
6-1448
19
20
875
23
3
3
4
4
4-8174
19
20
250
23
3
2
4
4
3-4900
19
19
625
23
3
1
4
4
2-1626
19
19
000
23
3
4
4
0-8352
19
18
375
23
2
7
4
3
11-5078
19
17
750
23
2
6
4
3
10-1804
19
17
125
23
2
5
4
3
8-8529
19
16
500
23
2
4
4
3
7-5255
19
15
875
23
2
3
4
3
6-1981
19
15
250
23
2
2
4
3
4-8707
19
14
625
23
2
1
4
3
3-5433
19
14
000
23
2
4
3
2-2159
.0
19
13
375
23
1
7
4
3
0-8885
19
12
750
23
1
6
4
2
11-5610
19
12
125
23
1
5
4
2
10-2336
19
11
500
23
1
4
4
2
8-9062
19
10
875
23
1
3
4
2
7-5788
19
10
250
23
1
2
4
2
6-2514
19
9
625
23
1
1
4
2
4-9240
19
9
000
23
1
4
2
3-5965
19
8
375
23
7
4
2
2-2691
19
7
750
23
6
4
2
0-9417
19
7
125
23
5
4
1
11-6143
19
6
500
23
4
4
1
10-2869
19
5
875
23
3
4
1
8-9595
19
5
250
23
2
4
1
7-6321
19
4
625
23
1
4
1
6-3047
19
4
000
23
4
1
4-9772
19
3
375
22
3
7
4
1
3-6498
19
2
750
22
3
6
4
1
2-3224
GOLD-VALUING TABLE.
Ill
FINE GOLD,
Per Ounce
CARAT GOLD,
Per Ounce
STERLING VALUE,
Per Ounce
O*.
DivU
Grs.
Carats Grs.
Eighths
s.
d.
19
2-125
22
3
5
4
1
0-9950
19
1-500
22
3
4
4
11-6676
19
0-875
22
3
3
4
10-3402
19
0-250
22
3
2
4
8-0127
18
23-625
22
3
1
4
7-6854
18
23-000
22
3
4
6-3579
18
22-375
22
2
7
4
4-0305
18
21-750
22
2
6
4
3-7031
18
21-125
22
2
5
4
2-3757
18
20-500
22
2
4
4
0-0482
18
19-875
22
2
3
3
19
11-7208
18
19-250
22
2
2
3
19
10-3934
18
18-625
22
2
1
3
19
8-0660
18
18-000
22
2
3
19
7-7386
18
17-375
22
1
7
3
19
6-4112
18
16-750
22
1
6
3
19
4-0838
18
16-125
22
1
5
3
19
3-7563
18
15-500
22
1
4
3
19
2-4289
18
14-875
22
1
3
3
19
0-1015
18
14-250
22
1
2
3
18
11-7741
18
13-625
22
1
1
3
18
10-4467
18
13-000
22
1
3
18
8-1193
18
12-375
22
7
3
18
7-7919
18
11-750
22
6
3
18
6-4644
18
11-125
22
5
3
18
4-1370
18
10-500
22
4
3
18
3-8096
18
9-875
22
3
3
18
2-4822
18
9-250
22
2
3
18
0-1548
18
8-625
22
1
3
17
11-8274
18
8-000
22
3
17
10-5000
18
7-375
21
3
7
3
17
.8-1725
18
6-750
21
3
6
3
17
7-8451
18
6-125
21
3
5
3
17
6-5177
18
5-500
21
3
4
3
17
4-1903
18
4-875
21
3
3
3
17
3-8629
18
4-250
21
3
2
3
17
2-5355
18
3-625
21
3
1
3
17
0-2081
18
3-000
21
3
3
16
11-8806
18
2-375
21
2
7
3
16
10-5532
18
1-750
21
2
6
3
16
8-2258
18
1-125
21
2
5
3
16
7-8984
18
0-500
21
2
4
3
16
6-5710
17
23-875
21
2
3
3
16
4-2436
17
23-250
21
2
2
3
16
3-9162
?> M 2
IV
GOLD-VALUING TABLE.
FINE
Per
GrOLD,
Ounce
CARAT GOLD,
Per Ounce
STERLING VALUE,
Per Ounce
Oz.
Dn-ts. Grs.
Carats
Grs.
Eighths
s.
d.
17
22-625
21
2
1
3 16
2-5887
17
22-000
21
2
3 16
1-2613
17
21-375
21
1
7
3 15
11-9339
17
20-750
21
1
6
3 15
10-6065
17
20-125
21
1
5
3 15
9-2791
17
'19-500
21
1
4
3 15
7-9517
17
18-875
21
1
3
3 15
6-6243
17
18-250
21
1
2
3 15
5-2968
17
17-625
21
1
1
3 15
3-9694
17
17-000
21
1
3 15
2-6420
17
16-375
21
7
3 15
1-3146
17
15-750
21
6
3 14
11-9872
17
15-125
21
5
3 14
10-6598
17
14-500
21
4
3 14
9-3324
17
13-875
21
3
3 14
8-0049
17
13-250
21
2
3 14
6-6775
17
12-625
21
1
3 14
5-3501
17
12-000
21
3 14
4-0227
17
11-375
20
3
7
3 14
2-6953
17
10-750
20
3
6
3 14
1-3678
17
10-125
20
3
5
3 14
0-0404
17
9-500
20
3
4
3 13
10-7130
17
8-875
20
3
3
3 13
9-3856
17
8-250
20
3
2
3 13
8-0582
17
7-6^5
20
3
1
3 13
6-7308
17
7-000
20
3
3 13
5-4034
17
6-375
20
2
7
3 13
4-0759
17
5-750
20
2
6
3 13
2-7485
17
5-125
20
2
5
3 13
1-4211
17
4-500
20
2
4
3 13
0-0937
17
3-875
20
2
3
3 12
10-7663
17
3-250
20
2
2
3 12
9-4389
17
2-625
20
2
1
3 12
8-1115
17
2-000
20
2
3 12
6-7840
17
1-375
20
1
hr
t
3 12
5-4566
17
0-750
20
1
6
3 12
4-1292
17
0-125
20
1
5
3 12
2-8018
16
23-500
20
1
4
3 12
1-4744
16
22-875
20
1
3
3 12
0-1470
16
22-250
20
1
2
3 11
10-8196
16
21-625
20
1
1
3 11
9-4921
16
21-000
20
1
3 11
8-1647
16
20-375
20
7
3 11
6-8373
16
19-750
20
6
3 11
5-5099
GOLD-VALUING TABLE.
FINE
Per
GrOLD,
Ounce
CARAT G-OLD,
Per Ounce
STERLING VALUE,
Per Ounce
Oz.
Dmts
Grs.
Carats Grs.
Eighths
s.
d.
16
19-125
20
5
3
11
4-1825
16
18-500
20
4
3
11
2-8551
16
17-875
20
3
3
11
1-5277
16
17-250
20
2
3
11
0-2002
16
16-625
20
1
3
10
10-8728
16
16-000
20
3
10
9-5454
16
15-375
19
3
7
3
10
8-2180
16
14-750
19
3
6
3
10
6-8906
16
14-125
19
3
5
3
10
5-5632
16
13-500
19
3
4
3
10
4-2357
16
12-875
19
3
3
3
10
2-9083
16
12-250
19
3
2
3
10
1-5809
16
11-625
19
3
1
3
10
0-2534
16
11-000
19
3
3
9
10-9260
16
10-375
19
2
7
3
9
9-5986
16
9-750
19
2
6
3
9
8-2712
16
9-125
19
2
5
3
9
6-9437
16
8-500
19
2
4
3
9
5-6163
o
16
7-875
19
2
3
3
9
4-2889
16
7-250
19
2
2
3
9
2-9615
16
6-625
19
2
1
3
9
1-6341
16
6-000
19
2
3
9
0-3067
16
5-375
19
1
7
3
8
10-9793
16
4-750
19
1
6
3
8
9-6518
16
4-125
19
1
5
3
8
8-3244
16
3-500
19
1
4
3
8
6-9970
16
2-875
19
1
3
3
8
5-6696
16
2-250
19
1
2
3
8
4-3422
16
1-625
19
1
1
3
8
3-0148
16
1-000
19
1
3
8
1-6874
16
0-375
19
7
3
8
0-3599
15
23-750
19
6
3
7
11-0325
15
23-125
19
5
3
7
9-7051
15
22-500
19
4
3
7
8-3777
15
21-875
19
3
3
7
7-0503
15
21-250
19
2
3
7
5-7229
15
20-625
19
1
3
7
4-3955
15
20-000
19
3
7
3-0681
15
19-375
18
3
7
3
7
1-7407
15
18-750
18
3
6
3
7
0-4133
15
18-125
18
3
5
3
6
11-0859
15
17-500
18
3
4
3
6
9-7585
15
16-875
18
3
3
3
6
8-4311
15
16-250
18
3
2
3
6
7-1036
VI
GOLD-VALUING TABLE.
FINE
GOLD,
CARAT GOLD,
STERLING VALUE,
Per Ounce
Per
Ounce
Per
Ounce
6b.
JJivts.
Grs.
Car a 's
Grs.
Eighths
s.
d.
15
15-625
18
3
1
3
6
5-7762
15
15-000
18
3
3
6
4-4488
15
14-375
18
2
7
3
6
3-1214
15
13-750
18
2
6
3
6
1-7940
15
13-125
18
2
5
3
6
0-4666
15
12-500
18
2
4
3
5
11-1392
15
11-875
18
2
3
3
5
9-8117
15
11-250
18
2
2
3
5
8-4843
15
10-625
18
2
1
3
5
7-1569
15
10-000
18
2
3
5
5-8295
15
9-375
18
1
7
3
5
4-5021
15
8-750
18
1
6
3
5
3-1747
15
8-125
18
1
5
3
5
1-8473
15
7-500
18
1
4
3
5
0-5198
15
6-875
18
1
3
3
4
11-1924
15
6-250
18
1
2
3
4
9-8650
15
5-625
18
1
1
3
4
8-5376
15
5-000
18
1
3
4
7-2102
15
4-375
18
7
3
4
5-8828
15
3-750
18
6
3
4
4-5554
15
3-125
18
5
3
4
3-2279
15
2-500
18
4
3
4
1-9005
15
1-875
18
3
3
4
0-5731
15
1-250
18
2
3
3
11-2457
15
0-625
18
1
3
3
9-9183
15
o-ooo
18
3
3
8-5909
14
23-375
17
3
7
3
3
7-2634
14
22-750
17
3
6
3
3
5-9360
14
22-125
17
3
5
3
3
4-6086
14
21-500
17
3
4
3
3
3-2812
14
20-875
17
3
3
3
3
1-9538
14
20-250
17
3
2
3
3
0-6264
14
19-625
17
3
1
3
2
11-2990
14
19-000
17
3
3
2
9-9715
14
18-375
17
2
7
3
2
8-6441
14
17-750
17
2
6
3
2
7-3167
14
17-125
17
2
5
3
2
5-9893
14
16-500
17
2
4
3
2
4-6619
14
15-875
17
2
3
3
2
3-3345
14
15-250
17
2
2
3
2
2-0071
14
14-625
17
2
1
3
2
0-6796
14
14-000
17
2
3
1
11-3522
14
13-375
17
1
7
3
1
10-0248
14
12-750
17
1
6
3
1
8-6974
GOLD-VALUING TABLE.
Vll
FINE G-OLD,
Per Ounce
CARAT GOLD,
Per Ounce
STERLING VALUE,
Per Ounce
Oz.
Divts.
Grs.
Carats Grs,
Eighths
s.
d.
14
12-125
17
1
5
3
I
7-3700
14
11-500
17
1
4
3
1
6-0426
14
10-875
17
1
3
3
1
4-7152
14
10-250
17
1
2
3
1
3-3877
14
9-625
17
1
1
3
1
2-0603
14
9-000
17
1
3
1
0-7329
14
8-375
17
7
3
11-4055
14
7-750
17
6
3
10-0781
14
7-125
17
5
3
8-7507
14
6-500
17
4
3
7-4233
14
5-875
17
3
3
6-0958
14
5-250
17
2
3
4-7684
14
4-625
17
1
3
3-4410
14
4-000
17
3
2-1136
14
3-375
16
3
7
3
0-7862
14
2-750
16
3
6
2
19
11-4588
14
2-125
16
3
5
2
19
10-1313
14
1-500
16
3
4
2
19
8-8039
14
0-875
16
3
3
2
19
7-4765
14
0-250
16
3
2
2
19
6-1491
13
23-625
16
3
1
2
19
4-8217
13
23-000
16
3
2
19
3-4943
13
22-375
16
2
7
2
19
2-1669
13
21-750
16
2
6
2
19
0-8394
13
21-125
16
2
5
2
18
11-5120
13
20-500
16
2
4
2
18
10-1846
13
19-875
16
2
3
2
18
8-8572
13
19-250
16
2
2
2
18
7-5298
13
18-625
16
2
1
2
18
6-2024
13
18-000
16
2
2
18
4-8750
13
17-375
16
1
7
2
18
3-5475
13
16-750
16
1
6
2
18
2-2201
13
16-125
16
1
5
2
18
08927
13
15-500
16
1
4
2
17
11-5653
13
14-875
16
1
3
2
17
10-2377
13
14-250
16
1
2
2
17
8-9103
13
13-625
16
1
1
2
17
7-5829
13
13-000
16
1
2
17
6-2554
13
12-375
16
7
2
17
4-9280
13
11-750
16
6
2
17
3-6006
13
11-125
16
5
2
17
2-2732
13
10-500
16
4
2
17
0-9458
13
9-875
16
3
2
16
11-6184
13
9-250 16
2
2
16
10-2909
Vlll
GOLD-VALUING TABLE.
FINE
Per
GOLD,
Ounce
CARAT GOLD,
Per Ounce
STERLING VALUE,
Per Ounce
Oz.
Divts. 6frs.
Carats Grs. M
\ghtJis
s.
d.
13
8-625
16
1
2
16
8-9635
13
8-000
16
2
16
7-6363
13
7-375
15
3
7
2
16
6-3089
13
6-750
15
3
6
2
16
4-9815
13
6-125
15
3
5
2
16
3-6541
13
5-500
15
3
4
2
16
2-3267
13
4-875
15
3
3
2
16
0-9992
13
4-250
15
3
2
2
15
11-6718
13
3-625
15
3
1
2
15
10-3444
13
3-000
15
3
2
15
9-0170
13
2-373
15
2
7
2
15
7-6896
13
1-750
15
2
6
2
15
6-3622
13
1-125
15
2
5
2
15
5-0348
13
0-500
15
2
4
2
15
3-7073
12
23-875
15
2
3
2
15
2-3799
12
23-250
15
2
2
2
15
1-0525
12
22-625
15
2
1
2
14
11-7251
12
22-000
15
2
2
14
10-3976
12
21-375
15
1
7
2
14
9-0702
12
20-750
15
1
6
2
14
7-7428
12
20-125
15
1
5
2
14
6-4154
12
19-500
15
1
4
2
14
5-0880
12
18-875
15
1
3
2
14
3-7606
12
18-250
15
1
2
2
14
2-4332
12
17-625
15
1
1
2
14
1-1057
12
17-000
15
1
2
13
11-7783
12
16-375
15
if
i
2
13
10-4509
12
15-750
15
6
2
13
9-1235
12
15-125
15
5
2
13
7-7961
12
14-500
15
4
2
13
6-4687
12
13-875
15
3
2
13
5-1413
12
13-250
15
2
2
13
3-8138
12
12-625
15
1
2
13
2-4864
12
12-000
15
2
13
1-1591
12
11-375
14
3
7
2
12
11-8316
12
10-750
14
3
6
2
12
10-5042
12
10-125
14
3
5
2
12
9-1768
12
9-500
14
3
4
2
12
7-8494
12
8-875
14
3
3
2
12
6-5220
12
8-250
14
3
2
2
12
5-1946
12
7-625
14
3
1
2
12
3-8671
12
7-000
14
3
2
12
2-5397
12
6-375
14
2
7
2
12
1-2123
12
5-750
14
2
6
2
11
11-8849
GOLD-VALUING TABLE.
FINE GOLD,
Per Ounce
CARAT GOLD,
Per Ounce
STERLING VALUE,
Per Ounce
Oz.
Dwts.
Grs.
Carats Grs.
Eighths
s.
d.
12
5-125
14
2
5
2
11
10-5575
12
4-500
14
2
4
2
11
9-2301
12
3-875
14
2
3
2
11
7-9027
12
3-250
14
2
2
2
11
6-5752
12
2-625
14
2
1
2
11
5-2478
12
2-000
14
2
2
11
3-9204
12
1-375
14
1
7
2
11
2-5930
12
0-750
14
1
6
2
11
1-2656
12
0-125
14
1
5
2
10
11-9382
11
23-500
14
1
4
2
10
10-6107
11
22-875
14
1
3
2
10
9-2833
11
22-250
14
1
2
'2
10
7-9559
11
21-625
14
1
1
2
10
6-6285
11
21-000
14
1
2
10
5-3011
11
20-375
14
7
2
10
3-9737
11
19-750
14
6
2
10
2-6463
11
19-125
14
5
2
10
1-3188
11
18-500
14
4
2
9
11-9914
11
17-875
14
3
2
9
10-6640
11
17-250
14
2
2
9
9-3366
11
16-625
14
1
2
9
8-0092
11
16-000
14
2
9
6-6818
11
15-375
13
3
7
2
9
5-3544
11
14-750
13
3
6
2
9
4-0269
11
14-150
13
3
5
2
9
2-6995
11
13-500
13
3
4
2
9
1-3721
11
12-875
13
3
3
2
9
0-0447
11
12-250
13
3
2
2
8
10-7173
11
11-625
13
3
1
2
8
9-3899
11
11-000
13
3
2
8
8-0625
11
10-375
13
2
7
2
8
6-7350
11
9-750
13
2
6
2
8
5-4076
11
9-125
13
2
5
2
8
4-0802
11
8-500
13
2
4
2
8
2-7528
11
7-875
13
2
3
2
8
1-4254
11
7-250
13
2
2
2
8
0-0980
11
6-625
13
2
1
2
7
10-7705
11
6-000
13
2
2
7
9-4431
11
5-375
13
1
7
2
7
8-1157
11
4-750
13
1
6
2
7
6-7883
11
4-125
13
1
5
2
7
5-4609
11
3-500
13
1
4
2
7
4-1335
11
2-875
13
1
3
2
7
2-8061
11
2-250
13
1
2
2
7
1-4786
GOLD-VALUING TABLE.
FINE GOLD,
Per Ounce
CAKAT GOLD,
Per Ounce
STERLING VAI/~E,
Per Our-*
Oz.
Drvts.
Grs.
Carats Grs.
Eighths
s.
d.
11
1-625
13
1
1
2
7
0-1512
11
1-000
13
1
2
6
10-8238
11
0-375
13
7
2
6
9-4964
10
23-750
13
6
2
6
8-1698
10
23-125
13
5
2
6
6-8416
10
22-500
13
4
2
6
5-5142
10
21-875
13
3
2
6
4-1867
10
21-250
13
2
2
6
2-8593
10
20-625
13
1
2
6
1-5319
10
20-000
13
2
6
0-2045
10
19-375
12
3
7
2
5
10-8771
10
18-750
12
3
6
2
5
9-5497
10
18-125
12
3
5
2
5
8-2223
10
17-500
12
3
4
2
5
6-8948
10
16-875
12
3
3
2
5
5-5674
10
16-250
12
3
2
2
5
4-2400
10
15-625
12
3
1
2
5
2-9126
10
15-000
12
3
2
5
1-5852
10
14-375
12
2
7
2
5
0-2578
10
13-750
12
2
6
2
4
10-9303
10
13-125
12
2
5
2
4
9-6029
10
12-500
12
2
4
2
4
8-2755
10
11-875
12
2
3
2
4
6-9481
10
11-250
12
2
2
2
4
5-6207
10
10-625
12
2
1
2
4
4-2933
10
10-000
12
2
2
4
2-9659
10
9-375
12
1
7
2
4
1-6384
10
8-750
12
1
6
2
4
0-3110
10
8-125
12
1
5
2
3
10-8366
10
7-500
12
1
4
2
3
9-6562
10
6-875
12
1
3
2
3
8-3288
10
6-250
12
1
2
2
3
7-0014
10
5-625
12
1
1
2
3
5-6740
10
5-000
12
1
2
3
4-3465
10
4-375
12
7
2
3
3-0191
10
3-750
12
6
2
3
1-6917
10
3-125
12
5
2
3
0-3643
10
2-500
12
4
2
2
11-0369
10
1-875
12
3
2
2
9-7095
10
1-250
12
2
2
2
8-3821
10
0-625
12
1
2
2
7-0546
10
0-000
12
2
2
5-7272
9
23-375
11
3
7
2
2
4-3998
9
22-750
11
3
6
2
2
3-0724
GOLD-VALUING TABLE.
FINE GOLD,
Per Ounce
CARAT GOLD,
Per Ounce
STERLING VALUE,
Per Ounce
Oz.
Drvts
Grs.
Carats
Grs.
Eighths
s.
d.
9
22-125
11
3
5
2
2
1-7450
9
21-500
11
3
4
2
2
0-4176
9
21-875
11
3
3
2
1
11-0901
9
20-250
11
3
2
2
1
9-7627
9
19-625
11
3
1
2
1
8-4353
9
19-000
11
3
2
1
7-1079
9
18-375
11
2
7
2
1
5-7805
9
17-750
11
2
6
2
1
4-4531
9
17-125
11
2
5
2
1
3-1257
9
16-500
11
2
4
2
1
1-7982
9
15-875
11
2
3
2
1
0-4708
9
15-250
11
2
2
2
11-1434
9
14-625
11
2
1
2
9-8160
9
14-000
11
2
2
8-4886
9
13-375
11
1
7
2
7-1612
9
12-750
11
1
6
2
5-8338
9
12-125
11
1
5
2
4-5063
9
11-500
11
1
4
2
3-1789
9
10-875
11
1
3
2
1-8515
9
10-250
11
1
2
2
0-5241
9
9-625
11
1
1
1
19
11-1967
9
9-000
11
1
1
19
9-8693
9
8-375
11
. 7
1
19
8-5419
9-
7-750
11
6
1
19
7-2144
9
7-125
11
5
1
19
5-8870
9
6-500
11
4
1
19
4-5596
9
5-875
11
3
1
19
3-2322
9
5-250
11
2
1
19
1-9048
9
4-625
11
1
1
19
0-5774
9
4-000
11
1
18
11-2500
9
3-375
10
3
7
1
18
9-9225
9
2-750
10
3
6
1
18
8-5951
9
2-125
10
3
5
1
18
7-2677
9
1-500
10
3
4
1
18
5-9403
9
0-875
10
3
3
1
18
4-6129
9
0-250
10
3
2
1
18
3-2855
8
23-625
10
3
1
1
18
1-9580
8
23-000
10
3
1
18
0-6306
8
22-375
10
2
7
1
17
11-3032
8
21-750
10
2
6
1
17
9-9758
8
21-125
10
2
5
1
17
8-6484
8
20-500
10
2
4
1
17
7-3210
8
19-875
10
2
3
1
17
5-9936
8
19-250
10
2
2
1
17
4-6661
Xll
GOLD-VALUING TABLE.
FINE GOLD,
Per Ounce
CARAT GOLD,
Per Ounce
STERLING VALUE,
Per Ounce
Oz.
Diets
. Grs.
Carats
Grs.
Eighths
s.
d.
8
18-625
10
2
1
1
17
3-3387
8
18-000
10
2
1
17
2-0113
8
17-375
10
1
7
1
17
0-6839
8
16-750
10
1
6
1
16
11-3565
8
16-125
10
1
5
1
16
10-0291
8
15-500
10
1
4
1
16
8-7017
8
14-875
10
1
3
1
16
7-3742
8
14-250
10
1
2
1
16
6-0468
8
13-625
10
1
1
1
16
4-7194
8
13-000
10
1
1
16
3-3920
8
12-375
10
7
1
16
2-0646
8
11-750
10
6
1
16
0-7372
8
11-125
10
5
1
15
11-4098
8
10-500
10
4
1
15
10-0823
8
9-875
10
3
1
15
8-7549
8
9-250
10
2
1
15
7-4275
8
8-625
10
1
1
15
6-1001
8
8-000
10
1
15
4-7728
8
7-375
9
3
7
1
15
3-4454
8
6-750
9
3
6
1
15
2-1179
8
6-125
9
3
5
1
15
0-7905
8
5-500
9
3
4
1
14
11-4631
8
4-875
9
3
3
1
14
10-1357
8
4-250
9
3
2
1
14
8-8083
8
3-625
9
3
1
1
14
7-4809
8
3-000
9
3
1
14
6-1535
8
2-375
9
2
7
1
14
4-8260 .
8
1-750
9
2
6
1
14
3-4986
8
1-125
9
2
5
1
14
2-1712
8
0-500
9
2
4
1
14
0-8438
7
23-875
9
2
3
1
13
11-5164
7
23-250
9
2
2
1
13
10-1890
7
22-625
9
2
1
1
13
8-8616
7
22-000
9
2
1
13
7-5341
7
21-375
9
1
7
1
13
6-2067
7
20-750
9
1
6
1
13
4-8793
7
20-125
9
1
5
1
13
3-5519
7
19-500
9
1
4
1
13
2-2245
7
19-875
9
1
3
1
13
0-8971
7
18-250
9
1
2
1
12
11-5697
7
17-625
9
1
1
1
12
10-2422
7
17-000
9
1
1
12
8-9168
7
16-375
9
7
1
12
7-5874
7
15-750
9
6
1
12
6-2600
GOLD-VALUING TABLE.
Xlll
FINE
Per
GrOLD,
Ounce
CARAT G-OLD,
Per Ounce
STERLING VALUE,
Per Ounce
Oz.
Ihuts. Grs.
Carats Grs.
Eighths
s.
d.
7
15-125
9
5
1
12
4-9326
7
14-500
9
4
1
12
3-6052
7
13-875
9
3
1
12
2-2778
7
13-250
9
2
1
12
0-9503
7
12-625
9
1
1
11
11-6229
7
12-000
9
1
11
10-2954
7
11-375
8
3
7
1
11
8-9680
7
10-750
8
3
6
1
11
7-6406
7
10-125
8
3
5
1
11
6-3132
7
9-500
8
3
4
1
11
4-9857
7
8-875
8
3
3
1
11
3-6583
7
8-250
8
3
2
1
11
2-3309
7
7-625
8
3
1
1
11
1-0035
7
7-000
8
3
1
10
11-6761
7
6-375
8
2
7
1
10
10-3487
7
5-750
8
2
6
1
10
9-0213
7
5-125
8
2
5
1
10
7-6938
7
4-500
8
2
4
1
10
6-3664
7
^3-875
8
2
3
1
10
5-0390
7
3-250
8
2
2
1
10
3-7116
7
2-625
8
2
1
1
10
2-3843
7
2-000
8
2
1
10
1-0568
7
1-375
8
1
7
1
9
11-7294
7
0-750
8
1
6
1
9
10-4019
7
0-125
8
1
5
1
9
9-0745
6
23-500
8
1
4
1
9
7-7471
6
22-875
8
1
3
1
9
6-4197
6
22-250
8
1
2
1
9
5-0923
6
21-625
8
1
1
1
9
3-7649
6
21-000
8
1
1
9
2-4375
6
20-375
8
7
1
9
1-1100
6
19-750
8
6
1
8
11-7826
6
19-125
8
5
1
8
10-4552
6
18-500
8
4
1
8
9-1278
6
17-875
8
3
1
8
7-8004
6
17-250
8
2
1
8
6-4730
6
16-625
8
1
1
8
5-1455
6
16-000
8
1
8
3-8181
6
15-375
7
3
7
1
8
2-4907
6
14-750
7
3
6
1
8
1-1633
6
14-125
7
3
5
1
7
11-8359
o
6
13-500
7
3
4
1
7
10-5085
6
12-875
7
3
3
1
7
9-1811
6
12-250
7
3
2
1
7
7-8536
XIV
GOLD-VALUING TABLE.
FINE
Per
GOLD,
Ounce
CARAT GOLD,
Per Ounce
STERLING VALUE,
Per Ounce
Oz.
Dwts. GTS.
Carats
Grs.
Eighths
s.
A
6
11-625
7
3
1
I
7
6-5262
6
11-000
7
3
1
7
5-1988
6
10-375
7
2
7
1
7
3-8714
6
9-750
7
2
6
1
7
2-5440
6
9-125
7
2
5
1
7
1-2166
6
8-500
7
2
4
1
6
11-8892
6
7-875
7
2
3
1
6
10-5617
6
7-250
7
2
2
1
6
9-2343
6
6-625
7
2
1
1
6
7-8069
6
6-000
7
2
1
6
6-5795
6
5-375
7
1
7
1
6
5-2521
6
4-750
7
1
6
1
6
3-9247
6
4-125
7
1
5
1
6
2-5973
6
3-500
7
1
4
1
6
1-2698
6
2-875
7
1
3
1
5
11-9424
6
2-250
7
1
2
1
5
10-6150
6
1-625
7
1
1
1
5
9-2876
6
1-000
7
1
1
5
7-9602
6
0-375
7
7
1
5
6-6328
5
23-750
7
6
1
5
5-3054
5
23-125
7
5
1
5
3-9779
5
22-500
7
4
1
5
2-6505
5
21-875
7
3
1
5
1-3231
5
21-250
7
2
1
4
11-9957
5
20-625
7
1
1
4
10-6683
5
20-000
7
1
4
9-3409
5
19-375
6
3
7
1
4
8-0134
5
18-750
6
3
6
1
4
6-6860
5
18-125
6
3
5
1
4
5-3586
5
17-500
6
3
4
1
4
4-0312
5
16-875
6
3
3
1
4
2-7038
5
16-250
6
3
2
1
4
1-3764
5
15-625
6
3
1
1
4
0-0490
5
15-000
6
3
1
3
10-7216
5
14-375
6
2
7
1
3
9-3941
5
13-750
6
2
6
1
3
8-0667
5
13-125
6
2
5
1
3
6-7393
5
12-500
6
2
4
1
3
5-4119
5
11-875
6
2
3
i
3
4-0845
5
11-250
6
2
2
1
3
2-7571
5
10-625
6
2
1 > l 1
3
1-4297
5
10-000
6
2
1
3
0-1022
5
9-375
6
1
7
1
2
10-7748
i o
5
8-750
6
1
6
1
2
9-4474
GOLD-VALUING TABLE.
XV
FINE
Per
GrOLD,
Ounce
CARAT GOLD,
Per Ounce
STERLING VALUE,
Per Ounce
to.
Dmts. Grs.
Carats Grs.
Eighths
s.
d.
5
8-125
6
1
5
I 2
8-1200
5
7-500
6
1
4
1 2
6-7926
5
6-875
6
1
3
1 2
5-4652
5
6-250
6
1
2
1 2
4-1377
5
5-625
6
1
1
1 2
2-8103
5
5-000
6
1
1 2
1-4829
5
4-375
6
7
1 2
0-1555
5
3-750
6
6
1 1
10-8281
5
3-125
6
5
1 1
9-5007
5
2-500
6
4
1 1
8-1733
5
1-875
6
3
1 1
6-8458
5
1-250
6
2
1 1
5-5184
5
0-625
6
1
1 1
4-1910
5
o-ooo
6
1 1
2-8636
4
23-375
5
3
7
1 1
1-5362
4
22-750
5
3
6
1 1
0-2088
4
22-125
5
3
5
1
10-8813
4
21-500
5
3
4
1
9-5539
4
20-875
5
3
3
1
8-2265
4
20*250
5
3
2
1
6-8991
4
19-625
5
3
1
1
5-5717
4
19-000
5
3
1
4-2443
4
18-375
5
2
7
1
2-9169
4
17-750
5
2
6
1
1-5894
4
17-125
5
2
5
1
0-2620
4
16-500
5
2
4
19
10-9346
4
15-875
5
2
3
19
9-6072
4
15-250
5
2
2
19
8-2798
4
14-625
5
2
1
19
6-9524
4
14-000
5
2
19
5-6250
4
13-375
5
1
7
19
4-2975
4
12-750
5
1
6
19
2-9701
4
12-125
5
1
5
19
1-6427
4
11-500
5
1
4
19
0-3153
4
10-875
5
1
3
18
10-9879
4
10-250
5
1
2
18
9-6605
4
9-625
5
1
1
18
8-3331
4
9-000
5
1
18
7-0056
4
8-375
5
7
18
. 5-6782
4
7-750
5
6
18
4-3508
- 4
7-125
5
5
18
3-0234
4
6-500
5
4
18
1-6960
4
5-875
5
3
18
0-3686
4
5-250
5
2
17
11-0411
XVI
GOLD-VALUING TABLE.
FINE GrOLD,
Per Ounce
CARAT G-OLD,
Per Ounce
STERLING VALUE,
Per Ounce
Oz.
Drvts
Grs.
Carats
Grs.
Eighths
s.
d.
4
4-625
5
1
17
9-7137
4
4-000
5
17
8-3863
4
3-375
4
3
7
17
7-0589
4
2-750
4
3
6
17
5-7315
4
2-125
4
3
5
17
4-4041
4
1-500
4
3
4
17
3-0767
4
0-875
4
3
3
17
1-7492
4
0-250
4
3
2
17
0-4218
3
23-625
4
3
1
16
11-0944
3
23-000
4
3
16
9-7670
3
22-375
4
2
7
16
8-4396
3
21-750
4
2
6
16
7-1122
3
21-125
4
2
5
16
5-7848
3
20-500
4
2
4
16
4-4573
3
19-875
4
2
3
16
3-1299
3
19-250
4
2
2
16
1-8025
3
18-625
4
2
1
16
0-4751
3
18-000
4
2
15
11-1477
3
17-375
4
1
7
15
9-8203
3
16-750
4
1
6
15
8-4929
3
16-125
4
1
5
15
7-1655'
3
15-500
4
1
4
15
5-8380
3
14-875
4
1
3
15
4-5106
3
14-250
4
1
2
15
3-1832
3
13-625
4
1
1
15
1-8558
3
13-000
4
1
15
0-5284
3
12-375
4
7
14
11-2009
3
11-750
4
6
14
9-8735
3
11-125
4
5
14
8-5461
3
10-500
4
4
14
7-2187
3
9-875
4
3
14
5-8913
3
9-250
4
2
14
4-5639
3
8-625
4
1
14
3-2365
3
8-000
4
14
1-9090
3
7-375
3
3
7
14
0-5816
3
6-750
3
3
6
13
11-2542
3
6-125
3
3
5
13
9-9268
3
5-500
3
3
4
13
8-5994
3
4-875
3
3
3
13
7-2720
3 '
4-250
3
3
2
13
5-9446
3
3-625
3
3
1
13
4-6171
3
3-000
3
3
13
3-2897
3
2-375
3
2
7
13
1-9623
3
1-750
3
2
6
13
0-6349
GOLD -VALUING TABLE.
XV11
FINE GOLD,
Per Ounce
CARAT GOLD,
Per Ounce
STERLING VALUE,
Per Ounce
Oz.
Dwts.
r. ? .
Carats
r*.
Eighths
I
s.
d.
3
1-125
3
2
5
12
11-3075
3
0-500
3
2
4
12
9-9801
2
23-875
3
2
3
12
8-6527
2
23-250
3
2
2
12
7-3250
2
22-625
3
2
1
12
5-9978
2
22-000
3
2
12
4-6704
2
21-375
3
1
7
12
3-3430
2
20-750
3
1
6
12
2-0156
2
20-125
3
1
5
12
0-6882
2
19-500
3
1
4
11
11-3607
2
18-875
3
1
3
11
10-0333
2
18-250
3
1
2
11
8-7059
2
17-625
3
1
1
11
7-3785
2
17-000
3
1
11
6-0511
2
16-375
3
7
11
4-7237
2
15-750
3
6
11
3-3963
2
15-125
3
5
11
2-0688
2
14-500
3
4
11
0-7414
2
13-875
3
3
10
11-4140
2
13-250
3
2
10
10-0866
2
12-625
3
1
10
8-7592
2
12-000
3
10
7-4318
2
11-375
2
3
7
10
6-1044
2
10-750
2
3
6
10
4-7769
2
10-125
2
3
5
10
3-4495
2
9-500
2
3
4
10
2-1221
2
8-875
2
3
3
10
0-7947
2
8-250
2
3
2
9
11-4673
2
7-625
2
3
1
9
10-1399
2
7-000
2
3
9
8-8125
2
6-375
2
2
7
9
7-4850
2
5-750
2
2
6
9
6-1576
2
5-125
2
2
5
9
4-8302
2
4-500
2
2
4
9
3-5028
2
3-875
2
2
3
9
2-1754
2
3-250
2
2
2
9
0-8480
2
2-625
2
2
1
8
11-5205
2'
2-000
2
2
8
10-1931
2
1-375
2
1
7
8
8-8657
2
0-750
2
1
6
8
7-5383
2
0-125
2
1
5
8
6-2109
1
23-500
2
1
4
8
4-8835
1
22-875
2
1
3
8
3-5561
]
22-250
2
1
2
8
2-2286
3 N
XV111
GOLD -VALUING TABLE.
FINE
Per
GrOLD,
Ounce
CAKAT (TOLD,
Per Ounce
STERLING VALUE,
Per Ounce
Oz.
Dwts. Grs.
Carats
Grs.
Eighths
5.
d.
1
21-625
2
1
1
8
0-9012
1
21-000
2
1
7
11-5738
1
20-375
2
7
7
10-2464
1
19-750
2
6
7
8-9190
1
19-125
2
5
7
7-5916
1
18-500
2
4
7
6-2642
1
17-875
2
3
7
4-9367
1
17-250
2
2
7
3-6093
1
16-625
2
1
7
2-2819
1
16-000
2
7
0-9545
1
15-375
1
3
7
6
11-6271
1
14-750
1
3
6
6
10-2997
1
14-125
1
3
5
6
8-9723
1
13-500
1
3
4
6
7-6448
1
12-875
1
3
3
6
6-3174
1
12-250
1
3
2
6
4-9900
1
11-625
1
3
1
6
3-6626
1
11-000
1
3
6
2-3352
1
10-375
1
2
7
6
1-0078
1
9-75C
1
2
6
5
11-6803
1
9-125
1
2
5
5
10-3529
1
8-500
1
2
4
5
9-0255
1
7-875
1
2
3
5
7-6981
1
7-250
1
2
2
5
6-3707
1
6-625
1
2
1
5
5-0433
1
6-000
1
2
5
3-7159
]
5-375
1
1
7
5
2-3884
1
4-750
1
1
6
5
1-0610
i
4-125
1
1
5
4
11-7336
1
3-500
1
1
4
4
10-4062
1
2-875
1
1
3
4
9-0788
1
2-250
1
1
2
4
7-7514
1
1-625
1
1
1
4
6-4240
1
1-000
1
1
4
5-0965
1
0-375
1
7
4
3-7691
23-750
1
6
4
2-4417
23-125
1
5
4
1-1143
22-500
1
4
3
11-7869
21-875
1
3
3
10-4595
21-250
1
2
3
9-1321
20-625
1
1
3
7-8046
20-000
1
3
6-4772
o
19-375
3
7
3
5-1498
18-750
3
6
3
3-8224
(iOLD-VALUING TABLE.
XIX
FINE
Per
GrOLD,
Ounce
CARAT GOLD,
Per Ounce
STERLING VALUE
Per Ounce
Os.
Dwts. Grs.
Carats
Grs.
Eighths
s.
d.
C
18-125
3
5
3
2-4950
17-500
3
4
3
1-1676
16-875
3
3
2
11-8401
16-250
3
2
2
10-5127
15-625
3
1
2
9-1853
15-000
3
2
7-8579
14-375
2
7
2
6-5305
13-750
2
6
2
5-2031
13-125
2
5
2
3-8757
12-500
2
4
2
2-5482
11-875
2
3
2
1-2208
11-250
2
2
1
11-8934
10-625
2
1
1
10-5660
10-000
2
1
9-2386
9-375
1
7
1
7-9112
8-750
1
6
1
6-5838
8-125
1
5
1
5-2563
7-500
1
4
1
3-9289
6-875
1
3
1
2-6015
6-250
1
2
1
1-2741
5-625
1
L
11-9467
5-000
1
10-6193
4-375
7
9-2919
3-750
6
7-9644
3-125
5
6-6370
2-500
4
5-3096
1-875
3
3-9822
1-250
2
2-6548
0-625
1
1-3274
XX
GOLD-VALUING TABLE.
To convert MINT VALUE into BANK VALUE when the Standard
is expressed in Carats, Grains, and Eighths. This can be
readily accomplished for every report by the following
Tables :
TABLE A.
CARATS
VALUE IN PENCE
CARATS
VALUE IK PENCE
1
0681
13
8863
2
1363
14
9545
3
2045
15
1-0227
4
2727
16
1-0909
5
3409
17
1-1590
6
4090
18
1-2272
7
4772
19
1-2954
8
5454
20
1-3636
9
6136
21
1-4318
10
6818
22
1-5000
11
7500
23
1-5681
12
8181
24
1-6363
TABLE B.
CAEAT GRAINS
VALUE IN PENCE
CARAT GRAINS
VALUE IN PESCE
1
2
0170
0340
3
4
0511
0681
TABLE C.
EIGHTH CARAT
GRAINS
VALUE IN PENCE
EIGHTH CARAT
GRAINS
VALUE IN PENCE
1
0021
5
0106
2
0042
6
0127
3
0063
7
0149
4
0085
8
0170
GOLD-VALUING TABLE.
XXI
Table A gives the difference in price between Mint and Bank
value for each carat up to fine gold ; Table B the same for carat
grains ; and Table C the same for eighths of carat grains.
Now as the Bank value of gold is 3 17s. 9d. per oz. standard
against Mint value of 3 17s. lO^d., it follows by calculation
that fine gold would fetch, Bank price, only 4 4s. 9-8182d,
instead of 4 4s. ll-4545<i., as shown by Table I. of Mint
Values ; and the Bank value of 1 oz. of gold, of any standard
whatever, may be readily ascertained by the above Tables A,
B, and C, and Table I. the Tables A, B, and C, giving the
quantities in pence to be deducted from the corresponding stan-
dard in Table I. Thus, suppose it is necessary to ascertain the
Bank value of 1 oz. of gold of 14 carats 2 grains 5 eighths fine :
refer to Table A, at 14 carats is found *9545c. ; at 2 grains in
Table B is found -0340d. ; and at 5 eighths in Table C -0106d
Now -9545 + -0340 -f- -0106 = -9991, which has to be deducted
from 2 lls. 10-5575d. (see Table L), leaving 2 11s. 9'5564d. as
the Bank value of 1 oz. of gold of the above fineness.
TABLE II.
TABLE of relative proportions of FINE GOLD and ALLOY, with
the respective Mint Values of 1 oz. of each Alloy when the
Standard is expressed in Thousandths.
FINE
GOLD
ALLOY
VALUE
FINE
GOLD
ALLOY
VALUE
s. d.
s. d.
1000
ooo
4 4 11-4545
986
014
4 3 9-1821
999
001
4 4 10-4350
985
015
4 3 8-1627
998
002
4 4 9-4156
984
016
4 3 7-1432
997
003
4 4 8-3961
983
017
4 3 6-1238
996
004
4' 4 7-3767
982
018
4 3 5-1043
995
005
4 4 6-3572
981
019
4 3 4-0849
994
006
4 4 5-3378
980
020
4 3 3-0654
993
007
4 4 4-3183
979
021
4 3 2-0459
992
008
4 4 3-2989
978
022
4 3 1-0265
991
009
4 4 2-2793
977
023
4 3 0-0070
990
010
4 4 1-2600
976
024
4 2 10-9876
989
Oil
4 4 0-2405
975
025
4 2 9-9681
988
012
4 3 11-2210
974
026
4 2 8-9487
987
013
4 3 10-2016
973
027
4 2 7-9292
XX11
GOLD-VALUING TABLE.
GOLD
ALLOY
VALUE
PINE
GOLD
ALLOY
VALUE
s. d.
s. d.
972
028
4 2 6-9098
929
071
3 18 11-0732
971
029
4 2 5-8903
928
072
3 18 10-0538
970
030
4 2 4-8709
927
073
3 18 9-0343
969
031
4 2 3-8504
9^6
074
3 18 8-0149
968
032
4 2 2-8319
925
075
3 18 6-9954
967
033
4 2 1-8125
924
076
3 18 5-9759
966
034
4 2 0-7930
923
077
3 18 4-9565
965
035
4 1 11-7736
922
078
3 18 3-9370
964
036
4 1 10-7541
921
079
3 18 2-9176
963
037
4 1 9-7347
920
080
3 18 1-8981
962
038
4 1 8-7152
919
081
3 18 0-8787
961
039
4 1 7-6958
918
082
3 17 11-8592
960
040
4 1 6-6763
917
083
3 17 10-8398
959
041
4 5-6569
916*
084
3 17 9-8203
958
042
4 1 4-6374
915
085
3 17 8-8009
957
043
4 3-6179
914
086
3 17 7-7814
956
044
4 2-5985
913
087
3 17 6-7619
955
045
4 1-5790
912
088
3 17 5-7425
954
046
4 1 0-5596
911
089
3 17 4-7230
953
047
4 11-5401
910
090
3 17 3-7036
952
048
4 10-5207
909
091
3 17 2-6841
951
049
4 9-5012
908
092
3 17 1-6647
950
050
4 8-4818
907
093
3 17 0-6452
949
051
4 7-4623
906
094
3 16 11-6258
948
052
4 6-4429
905
095
3 16 10-6063
947
053
4 5-4234
904
096
3 16 9-5869
946
054
4 4-4039
903
097
3 16 8-5674
945
055
4 3-3835
902
098
3 16 7-5479
944
056
4 2-3650
901
099
3 16 6-5285
943
057
4 1-3456
900
100
3 16 5-5090
942
058
4 0-3261
899
101
3 16 4-4896
941
059
3 19 11-3067
898
102
3 16 3-4701
940
060
3 19 10-2872
897
103
3 16 2-4507
939
061
3 19 9-2678
896
104
3 16 1-4312
938
062
3 19 8-2483
895
105
3 16 0-4118
937
063
3 19 7-2289
894
106
3 15 11-3923
936
064
3 19 6-2094
893
107
3 15 10-3729
935
065
3 19 5-1899
892
108
3 15 9-3534
934
066
3 19 4-1705
891
109
3 15 8-3339
933
067
3 19 3-1510
890
110
3 15 7-3145
932
068
3 19 2-1316
889
111
3 15 6-2950
931
069
3 19 1-1121
888
112
3 15 5-2756
930
070
3 19 0-0927
887
113
3 15 4-2561
* 916-666 Standard -083-333 3 17s. 10'5000<f.
GOLD-VALUING TABLE.
XX111
FINE
GOLD
ALLOY
VALUE
FINE
GOLD
ALLOY
VALUE
8. d.
s. d.
886
114
3 15 3-2367
841
159
3 11 5-3612
885
115
3 15 2-2172
840
160
3 11 4-3418
884
116
3 15 1-1978
839
161
3 11 3-3223
883
117
3 15 0-1783
838
162
3 11 2-3029
882
118
3 14 11-1589
837
163
3 11 1-3834
881
119
3 14 10-1394
836
164
3 11 0-2639
880.
120
3 14 9-1199
835
165
3 10 11-2445
879
121
3 14 8-1005
834
166
3 10 10-2250
878
122
3 14 7-0810
833
167
3 10 9-2056
877
123
3 14 6-0616
832
168
3 10 8-1861
876
124
3 14 5-0421
831
169
3 10 7-1667
875
125
3 14 4-0227
830
170
3 10 6-1472
874
126
3 14 3-0032
829
171
3 10 5-1278
873
127
3 14 1-9838
828
172
3 10 4-1083
872
128
3 14 0-9643
827
173
3 10 3-0889
871
129
3 13 11-9449
826
174
3 10 2-0694
870
130
3 13 10-9254
825
175
3 10 1-0499
869
131
3 13 9-9059
824
176
3 10 0-0305
868
132
3 13 8-8865
823
177
3 9 11-0110
867
133
3 13 7-8670
822
178
3 9 9-9916
866
134
3 13 6-8476
821
179
3 9 8-9721
865
135
3 13 5-8281
820
180
3 9 7-9527
864
136
3 13 4-8087
819
181
3 9 6-9332
863
137
3 13 3-7892
818
182
3 9 5-9138
862
138
3 13 2-7698
817
183
3 9 4-8943
861
139
3 13 1-7503
816
184
3 9 3-8749
860
140
3 13 0-7309
815
185
3 9 2-8554
859
141
3 12 11-7114
814
186
3 9 1-8359
858
142
3 12 10-6919
813
187
3 9 0-8165
857
143
3 12 9-6725
812
188
3 8 11-7970
856
144
3 12 8-6530
811
189
3 8 10-7776
855
145
3 12 7-6336
810
190
3 8 9-7581
854
146
3 12 6-6141
809
191
3 8 8-7387
853
147
3 12 5-5947
808
192
3 8 7-7192
852
148
3 12'' 4-5752
807
193
3 8 6-6998
851
149
3 12 3-5558
806
194
3 8 5-6803
850
150
3 12 2-5363
805
195
3 8 4-6609
849
151
3 12 1-5169
804
196
3 8 3-6414
848
152
3 12 0-4974
803
197
3 8 2-6219
847
153
3 11 11-4779
802
198
3 8 1-6025
846
154
3 11 10-4585
801
199
3 8 0-5830
845
155
3 11 9-4390
800
200
3 7 11-5636
844
156
3 11 8-4196
799
201
3 7 10-5441
843
157
3 11 7-4001
798
202
3 7 9-5247
842
158
3 11 6-3807
797
203
3 7 8-5052
XXIV
GOLD-VALUING TABLE.
FINE
GOLD
ALLOY
VALUE
FINE
GOLD
ALLOY
VALUE
I s. d.
s. d.
796
204
3 7 7-4858
751
249
3 3 9-6103
795
205
3 7 6-4663
750
250
3 3 8-5909
794
206
3 7 5-4469
749
251
3 3 7-5714
793
207
3 7 4-4274
748
252
3 3 6-5519
792
208
3 7 3-4979
747
253
3 3 5-5325
791
209
3 7 2-3885
746
254
3 3 4-5130
790
210
3 7 1-3690
745
255
3 3 3-4936
789
211
3 7 0-3496
744
256
3 3 2-4741
788
212
3 6 11-3301
743
257
3 3 1-4547
787
213
3 6 10-3107
742
258
3 3 0-4352
786
214
3 6 9-2912
741
-259
3 2 11-4158
785
215
3 6 8-2718
740
260
3 2 10-3963
784
216
3 6 7-2523
739
261
3 2 9-3769
783
217
3 6 6-2329
738
262
3 2 8-3574
782
218
3 6 5-2134
737
263
3 2 7-3379
781
219
3 6 4-1939
736
264
3 2 6-3185
780
220
3 6 3-1745
735
265
3 2 5-2990
779
221
3 6 2-1550
734
266
3 2 4-2796
778
222
3 6 1-1356
733
267
3 2 3-2601
777
223
3 6 0-1161
732
268
3 2 2-2407
776
224
3 5 11-0967
731
269
3 2 1-2212
775
225
3 5 10-0772
730
270
3 2 0-2018
774
226
3 5 9-0578
729
271
3 1 11-1823
773
227
3 5 8-0383
728
272
3 1 10-1629
772
228
3 5 7-0189
727
273
3 1 9-1434
771
229
3 5 5-9994
726
274
3 1 8-1239
770
230
3 5 4-9799
725
275
3 1 7-1045
769
231
3 5 3-9605
724
276
3 1 6-0850
768
232
3 5 2-9410
723
277
3 1 5-0656
767
233
3 5 1-9216
722
278
3 1 4-0461
766
234
3 5 0-9021
721
279
3 1 3-0267
765
235
3 4 11-8827
720
280
3 1 2-0072
764
236
3 4 10-8632
719
281
3 1 0-9878
763
237
3 4 9-8438
718
282
3 11-9683
762
238
3 4 8-8243
717
283
3 10-9489
761
239
3 4 7-8049
716
284
3 9-9294
760
240
3 4 6-7854
715
285
3 8-9099
759
241
3 4 5-7659
714
286
3 7-8905
758
242
3 4 4-7465
713
287
3 6-8710
757
243
3 4 3-7270
712
288
3 5-8516
756
244
3 4 2-7076
711
289
3 4-8321
755
245
3 4 1-6881
710
290
3 3-8127
754
246
3 4 0-6687
709
291
3 2-7932
753
247
3 3 11-6492
708
292
3 1-7738
752
248
3 3 10-6298
707
293
3 0-7543
GOLD-VALUING TABLE.
XXV
FINE
GOLD
ALLOY
VALUE
FINE
GOLD
ALLOY
VALUE
s. d.
s. d.
706
294
2 19 11-7349
661
339
2 16 1-8594
705
295
2 19 10-7154
660
340
2 16 0-8399
704
296
2 19 9-6959
659
341
2 15 11-8205
703
297
2 19 8-6765
658
342
2 15 10-8010
702
298
2 19 7-6570
657
343
2 15 9-7816
701
299
2 19 6-6376
656
344
2 15 8-7621
700
300
2 ]9 5-6181
655
345
2 15 7-7427
699
301
2 19 4-5987
654
346
2 15 6-7232
698
302
2 19 3-5792
653
347
2 15 5-7038
697
303
2 19 2-5598
652
348
2 15 4-6843
696
304
2 19 1-5403
651
349
2 15 3-6649
695
305
2 19 0-5209
650
350
2 15 2-6454
694
306
2 18 11-5014
649
351
2 15 1-6259
693
307
2 18 10-4820
648
352
2 15 0-6065
692
308
2 18 9-4625
647
353
2 14 11-5870
691
309
2 18 8-4430
646
354
2 14 10-5676
690
310
2 18 7-4236
645
355
2 14 9-5481
689
311
2 18 6-4041
644
356
2 14 8-5287
688
312
2 18 5-3847
643
357
2 14 7-5092
687
313
2 18 4-3652
642
358
2 14 6-4898
686
314
2 18 3-3458
641
359
2 14 5-4703
685
315
2 18 2-3263
640
360
2 14 4-4509
684
316
2 18 1-3069
639
361
2 14 3-4314
683
317
2 18 0-2874
638
362
2 14 2-4120
682
318
2 17 11-2680
637
363
2 14 1-3925
681
319
2 17 10-2485
636
364
2 14 0-3730
680
320
2 17 9-2290
635
365
2 13 11-3536
679
321
2 17 8-2096
634
366
2 13 10-3341
678
322
2 17 7-1901
633
367
2 13 9-3147
677
323
2 17 6-1707
632
368
2 13 8-2952
676
324
2 17 5-1512
631
369
2 13 7-2758
675
325
2 17 4-1318
630
370
2 13 6-2563
674
326
2 17 3-1123
629
371
2 13 5-2369
673
327
2 17 2-0929
628
372
2 13 4-2174
672
328
2 17 1-0734
627
373
2 13 3-1979
671
329
2 17 0-0540
626
374
2 13 2-1785
670
330
2 16 11-0345
625
375
2 13 1-1590
669
331
2 16 10-0151
624
376
2 13 0-1396
668
332
2 16 8-9956
623
377
2 12 11-1201
667
333
2 16 7-9761
622
378
2 12 10-1007
666
334
2 16 6-9567
621
379
2 12 9-0812
665
335
2 16 5-9372
620
380
2 12 8-0618
664
336
2 16 4-9178
619
381
2 12 7-0423
663
337
2 16 3-8983
618
382
2 12 6-0229
662
338
2 16 2-8789
617
383
2 12 5-0034
XXVI
GOLD- VALUING TABLE.
FINE
GOLD
ALLOY
VALUE
FINE
GOLD
ALLOY
VALUE
s. d.
s. d.
616
384
2 12 3-9839
571
429
2 8 6-1085
615
385
2 12 2-9645
570
430
2 8 5-0890
614
386
2 12 1-9451
569
431
2 8 4-0696
613
387
2 12 0-9256
568
432
2 8 3-0501
612
388
2 11 11-9061
567
433
2 8 2-0307
611
389
2 11 10-8867
566
434
2 8 1-0112
610
390
2 11 9-8672
565
435
2 7 11-9918
609
391
2 11 8-8478
564
436
2 7 10-9723
608
392
2 11 7-8283
563
437
2 7 9-9529
607
393
2 11 6-8089
562
438
2 7 8-9334
606
394
2 11 5-7894
561
439
2 7 7-9140
605
395
2 11 4-7699
560
440
2 7 6-8945
604
396
2 11 3-7505
559
441
2 7 5-8751
603
397
2 11 2-7311
558
442
2 7 4-8556
602
398
2 11 1-7116
557
443
2 7 3-8361
601
399
2 11 0-6921
556
444
2 7 2-8167
600
400
2 10 11-6727
555
445
2 7 1-7972
599
401
2 10 10-6532
554
446
2 7 0-7778
598
402
2 10 9-6338
553
447
2 6 11-7583
597
403
2 10 8-6143
552
448
2 6 10-7389
596
404
2 10 7-5949
551
449
2 6 9-7194
595
405
2 10 6-5754
550
450
2 6 8-6999
594
406
2 10 5-5559
549
451
2 6 7-6805
593
407
2 10 4-5365
548
452
2 6 6-6611
592
408
2 10 3-5170
547
453
2 6 5-6416
591
409
2 10 2-4976
546
454
2 6 4-6221
590
410
2 10 1-4781
545
455
2 6 3-6027
589
411
2 10 0-4587
544
456
2 6 2-5832
588
412
2 9 11-4392
543
457
2 6 1-5638
587
413
2 9 10-4198
542
458
2 6 0-5443
586
414
2 9 9-4003
541
459
2 5 11-5249
585
415
2 9 8-3809
540
460
2 5 10-5054
584
416
2 9 7-3614
539
461
2 5 9-4859
583
417
2 9 6-3419
538
462
2 5 8-4665
582
418
2 9 5-3225
537
463
2 5 7-4470
581
419
2 9 4-3030
536
464
2 5 6-4276
580
420
2 9 3-2836
535
465
2 5 5-4081
579
421
2 9 2-2641
534
466
2 5 4-3887
578
422
2 9 1-2447
533
467
2 5 3-3692
577
423
2 9 0-2252
532
468
2 5 2-3498
576
424
2 8 11-2058
531
469
2 5 1-3303
575
425
2 8 10-1863
530
470
2 5 0-3109
574
426
2 8 9-1669
529
471
2 4 11-2914
573
427
2 8 8-1474
528
472
2 4 10-2719
1 57?
428
2 8 7-1279
527
473
9 4 9-2525
GOLD -VALUING TABLE.
XX VU
FINE
GOLD
ALLOT
VALUE
FINE
GOLD
ALLOY
VALUE
s. d.
s. d.
526
474
2 4 8-2330
481
519
2 10-3576
525
475
2 4 7-2136
480
520
2 9-3381
524
476
2 4 6-1941
479
521
2 8-3187
523
477
2 4 5-1747
478
522
2 7-2992
522
478
2 4 4-1552
477
523
2 6-2798
521
479
2 4 3-1358
476
524
2 5-2603
520
480
2 4 2-1163
475
525
2 4-2409
519
481
2 4 1-0969
474
526
2 3-2214
518
482
2 4 0-0774
473
527
2 2-2020
517
483
2 3 11-0579
472
528
2 1-1825
516
484
2 3 10-0385
471
529
2 0-1630
515
485
2 3 9-0190
470
530
1 19 11-1436
514
486
2 3 7-9996
469
531
1 19 10-1241
513
487
2 3 6-9801
468
532
1 19 9-1047
512
488
2 3 5-9607
467
533
1 19 8-0852
511
489
2 3 4-9412
466
534
1 19 7-0658
510
490
2 3 3-9218
465
535
1 19 6-0463
509
491
2 3 2-9023
464
536
1 19 5-0269
508
492
2 3 1-8829
463
537
1 19 4-0074
507
493
2 3 0-8634
462
538
1 19 2-9879
506
494
2 2 11-8439
461
539
1 19 1-9685
505
495
2 2 10-8245
460
540
1 19 0-9490
504
496
2 2 9-8051
459
541
1 18 11-9296
503
497
2 2- 8-7856
458
542
1 18 10-9101
502
498
2 2 7-7661
457
543
1 18 9-8907
501
499
2 2 6-7467
456
544
1 18 8-8712
500
500
2 2 5-7272
455
545
1 18 7-8518
499
501
2 2 4-7078
454
546
1 18 6-8323
498
502
2 2 3-6883
453
547
1 18 5-8129
497
503
2 2 2-6689
452
548
1 18 4-7934
496
504
2 2 1-6494
451
549
1 18 3-7739
495
505
2 2 '' 0-6300
450
550
1 18 2-7545
494
506
2 1 11-6105
449
551
1 18 1-7351
493
507
2 1 10-5911
448
552
1 18 0-7156
492
508
2 1 9-5716
447
553
1 17 11-6961
491
509
2 1 8-5521
446
554
1 17 10-6767
490
510
2 1 7-5327
445
555
1 17 9-6572
489
511
2 1 1-5132
444
556
1 17 8-6378
488
512
2 1 5-4938
443
557
1 17 7-6183
487
513
2 1 4-4743
442
558
1 17 6-5989
486
514
2 1 3-4549
441
559
1 17 5-5794
485
515
2 1 2-4354
440
560
1 17 4-5599
484
516
2 1 1-4159
439
561
1 17 3-5405
483
517
2 1 0-3965
438
562
1 17 2-5211
482
518
2 11-3770
437
563
1 17 1-5016
XXV111
GOLD-VALUING TABLE.
GOLD
ALLOY
VALUE
FINE
GOLD
ALLOY
VALUE
I
s. d.
t. d.
436
564
1 17 0-4821
391
609
1 13 2-6067
435
565
1 16 11-4627
390
610
1 13 1-5872
434
566
1 16 10-4432
389
611
1 13 0-5678
433
567
1 16 9-4238
388
612
1 12 11-5483
432
568
1 16 8-4043
387
613
1 12 10-5289
431
569
1 16 7-3849
386
614
1 12 9-5094
430
570
1 16 6-3654
385
615
1 12 8-4899
429
571
1 16 5-3459
384
616
1 12 7-4705
428
572
1 16 4-3265
383
617
1 12 6-4511
427
573
1 16 3-3070
382
618
1 12 5-4316
426
574
1 16 2-2876
381
619
1 12 4-4121
425
575
1 16 1-2681
380
620
1 12 3-3927
424
576
1 16 0-2487
379
621
1 12 2-3732
423
577
1 15 11-2292
378
622
1 12 1-3538
422
578
1 15 10-2098
377
623
1 12 0-3343
421
579
1 15 9-1903
376
624
1 11 11-3142
420
580
1 15 8-1709
375
625
1 11 10-2954
419
581
1 15 7-1514
374
626
1 11 9-2759
418
582
1 15 6-1319
373
627
1 11 8-2565
417
583
1 15 5-1125
372
628
1 11 7-2370
416
584
1 15 4-0930
371
629
1 11 6-2176
415
585
1 15 3-0736
370
630
1 11 5-1981
414
586
1 15 2-0541
369
631
1 11 4-1787
413
587
1 15 1-0347
368
632
1 11 3-1592
412
588
1 15 0-0152
367
-633
1 11 2-1398
411
589
1 14 10-9958
366
634
1 11 1-1203
410
590
1 14 9-9763
365
635
1 11 0-1009
409
591
1 14 8-9569
364
636
1 10 11-0814
408
592
1 14 7-9374
363
637
1 10 10-0620
407
593
1 14 6-9179
362
638
1 10 9-0425
406
594
1 14 5-8985
361
639
1 10 8-0230
405
595
1 14 4-8790
360
640
1 10 7-0036
404
596
1 14 3-8596
359
641
1 10 5-9841
403
597
1 14 2-8401
358
642
1 10 4-9647
402
598
1 14 1-8207
357
643
1 10 3-9452
401
599
1 14 0-8012
356
644
1 10 2-9258
400
.600
1 13 11-7818
355
645
1 10 1-9063
399
601
1 13 10-7623
354
646
1 10 0-8869
398
602
1 13 9-7429
353
647
1 9 11-8674
397
603
1 13 8-7234
352
648
1 9 10-8479
396
604
1 13 7-7039
351
649
1 9 9-8285
395
605
1 13 6-6845
350
650
1 9 8-8090
394
606
1 13 5-6651
349
651
1 9 7-7896
393
607
1 13 4-6456
348
652
1 9 6-7701
392
608
1 13 3-6261
347
653
1 9 5-7507
GOLD- VALUING TABLE.
XXIX
FINE
GOLD
ALLOY
VALUE
FINE
GOLD
ALLOY
VALUE
s. a.
8. d.
346
654
1 9 4-7312
301
699
1 5 6-8558
345
655
1 9 3-7118
300
700
1 5 5-8363
344
656
1 9 2-6923
299
701
1 5 4-8169
343
657
1 9 1-6729
298
702
1 5 3-7974
342
658
1 9 0-6534
297
703
1 5 2-7779
341
659
1 8 11-6339
296
704
1 5 1-7585
340
660
1 8 10-6145
295
705
1 5 0-7390
339
661
1 8 9-5951
294
706
1 4 11-7196
338
662
1 8 8-5756
293
707
1 4 10-7011
337
663
1 8 7-5561
292
708
1 4 9-6807
336
664
1 8 6-5367
291
709
1 4 8-6612
335
665
1 8 5-5172
290
710
1 4 7-6418
334
666
1 8 4-4978
289
711
1 4 6-6223
333
667
1 8 3-4783
288
712
1 4 5-6029
332
668
1 8 2-4589
287
713
1 4 4-5834
331
669
1 8 1-4394
286
714
1 4 3-5639
330
670
1 8 0-4199
285
715
1 4 2-5445
329
671
1 7 11-4005
284
716
1 4 1-5251
328
672
1 7 10-3811
283
717
1 4 0-5056
327
673
1 7 9-3616
282
718
1 3 11-4861
326
674
1 7 8-3421
281
719
1 3 10-4667
325
675
1 7 7-3227
280
720
1 3 9-4472
324
676
1 7 6-3032
279
721
1 3 8-4278
323
677
1 7 5-2838
278
722
1 3 7-4083
322
678
1 7 4-2643
277
723
1 3 6-3889
321
679
1 7 3-2449
276
724
1 3 5-3694
320
680
1 7 2-2254
275
725
1 3 4-3499
319
681
1 7 1-2059
274
726
1 3 3-3305
318
682
1 7 0-1865
273
727
1 3 2-3110
317
683
1 6 11-1670
272
728
1 3 1-2916
316
684
1 6 1Q.-1476
271
729
1 3 0-2721
315
685
1 6 9-1281
270
730
1 2 11-2527
314
686
1 6 8-1087
269
731
1 2 10-2332
313
687
1 6 7-0892
268
732
1 2 9-2138
312
688
1 6 6-0698
267
733
1 2 8-1943
311
689
1 6 5-0503
266
234
1 2 7-1749
310
690
1 6 4-0309
265
735
1 2 6-1554
309
691
1 6 3-0114
264
736
1 2 5-1351
308
692
1 6 1-9919
263
737
1 2 4-1165
307
693
1 6 0-9725
262
738
1 2 3-0970
306
694
1 5 11-9530
261
739
1 2 2-0776
305
695
1 5 10-9336
260
740
1 2 1-0581
304
696
1 5 9-9141
259
741
1 2 0-0387
303
697
1 5 8-8947
258
742
1 1 11-0192
302
698
1 5 7-8752
257
743
1 1 9-9998
XXX
GOLD-VALUING TABLE.
FINE
GOLD
ALLOY
VALUE
FINE
GOLD
ALLOY
VALUE
s. d.
s. d.
256
744
1 8-9803
211
789
17 11-1049
255
745
1 7-9609
210
790
17 10-0854
254
746
1 6-9414
209
791
17 9-0659
253
747
1 5-9219
208
792
17 8-0465
252
748
1 4-9025
207
793
17 7-0270
251
749
1 3-8830
206
794
17 6-0076
250
750
1 2-8636
205
795
17 4-9881
249
751
1 1 1-8441
204
796
17 3-9687
248
752
1 1 0-8247
203
797
17 2-9492
247
753
1 11-8052
202
798
17 1-9298
246
754
1 10-7858
201
799
17 0-9103
245
755
1 9-7663
200
800
16 11-8909
244
756
1 8-7469
199
801
16 10-8714
243
757
1 7-7274
198
802
16 9-8519
242
758
1 6-7079
197
803
16 8-8325
241
759
1 5-6885
196
804
16 7-8130
240
760
1 4-6690
195
805
16 6-7936
239
761
1 3-6496
194
806
16 5-7741
238
762
1 2-6301
193
807
16 4-7547
237
763
1 1-6107
192
808
16 3-7352
236
764
1 0-5912
191
809
16 2-7158
235
765
19 11-5718
190
810
16 1-6963
234
766
19 10-5523
189
811
16 0-6769
233
767
19 9-5329
188
812
15 11-6574
232
768
19 8-5134
187
813
15 10-6379
231
769
19 7-4939
186
814
15 9-6185
230
770
19 6-4745
185
815
15 8-5990
229
771
19 5-4551
184
816
15 7-5796
228
772
19 4-4356
183
817
15 6-5601
227
773
19 3-4161
182
818
15 5-5407
226
774
19 2-3967
181
819
15 4-5212
225
775
19 1-3772
180
820
15 3-5018
224
776
19 0-3578
179
821
15 2-4823
223
777
18 11-3383
178
822
15 1-4629
222
778
18 10-3189
177
823
15 0-4434
221
779
18 9-2994
176
824
14 11-4239
220
780
18 8-2799
175
825
14 10-4045
219
781
18 7-2605
174
826
14 9-3851
218
782
18 6-2410
173
827
14 8-3656
217
783
18 5-2216
172
828
14 7-3461
216
784
18 4-2021
171
829
14 6-3267
215
785
18 3-L827
170
830
14 5-3072
214
786
18 2-1632
169
831
14 4-2878
213
787
18 1-1438
168
832
14 3-2683
212
788
18 0-1243
167
833
14 2-2489
GOLD-VALUING TABLE.
XXXI
FINE
GOLD
ALLOY
VALUE
FINE
GOLD
ALLOY
VALUE
s. d.
s. d.
166
834
14 1-2294
121
879
10 3-3530
165
835
14 0-2099
120
880
10 2-3345
164
836
13 11-1905
119
881
10 1-3151
163
837
13 10-1710
118
882
10 0-2956
162
838
13 9-1516
117
883
9 11-2761
161
839
13 8-1321
116
884
9 10-2567
160
840
13 7-1127
115
885
9 9-2372
159
841
13 6-0932
114
886
9 8-2178
158
842
13 5-0738
113
887
9 7-1983
157
843
13 4-0543
112
888
9 6-1789
156
844
13 3-0349
111
889
9 5-1594
155
845
13 2-0154
110
890
9 4-1399
154
846
13 0-9959
109
891
9 3-1205
153
847
12 11-9765
108
892
9 2-1010
152
848
12 10-9570
107
893
9 1-0816
151
849
12 9-9376
106
894
9 0-0621
150
850
12 8-9181
105
895
8 11-0427
149
851
12 7-8987
104
896
8 10-0232
148
852
12 6-8792
103
897
8 9-0038
147
853
12 5-8598
102
898
8 7-9843
146
854
12 4-8403
101
899
8 6-9649
145
855
12 3-8209
100
900
8 5-9454
144
856
12 2-8014
99
901
8 4-9259
143
857
12 1-7819
98
902
8 3-9065
142
858
12 0-7625
97
903
8 2-8870
141
859
11 11-7430
96
904
8 1-8676
140
860
11 10-7236
95
905
8 0-8481
139
861
11 9-7041
94
906
7 11-8287
138
862
11 8-6847
93
907
7 10-8092
137
863
11 7-6652
92
908
7 9-7898
136
864
11 6-6458
91
909
7 8-7703
135
865
11 5-6263
90
910
? 7-7509
134
866
11 4-6069
89
911
7 6-7314
133
867
11 3-5874
88
912
7 5-7119
132
868
11 2-5679
87
913
7 4-6925
131
869
11 1-5485
86
914
7 3-6730
130
870
11 0-5290
85
915
7 2-6536
129
871
10 11-5096
84
916
7 1-6341
128
872
10 10-4901
83
917
7 0-6147
127
873
10 9-4707
82
918
6 11-5952
126
874
10 8-4512
81
919
6 10-5758
125
875
10 7-4318
80
920
6 9-5563
124
876
10 6-4123
79
921
6 8-5369
123
877
10 5-3929
78
722
6 7-5174
122
878
10 4-3734
77
923
6 6-4979
XXX11
GOLD-VALUING TABLE.
FINE
GOLD
ALLOY
VALUE
FINE
GOLD
ALLOY
VALUE
s. d.
s. d.
76
924
6 5-4785
38
962
3 2-7392
75
925
6 4-4590
37
963
3 1-7198
74
926
6 3-4396
. 36
964
3 0-7003
73
927
6 2-4201
35
965
2 11-6809
72
928
6 1-4007
34
966
2 10-6614
71
929
6 0-3812
33
967
2 9-6419
70
930
5 11-3618
32
968
2 8-6225
69
931
5 10-3423
31
969
2 7-6030
68
932
5 9-3229
30
970
2 6-5836
67
933
5 8-3034
29
971
2 5-5641
66
934
5 7-2839
28
972
2 4-5447
65
935
5 6-2645
27
973
2 3-5252
64
936
5 5-2451
26
974
2 2-5058
63
937
5 4-2256
25
975
2 1-4863
62
938
5 3-2061
24
976
2 0-4669
61
939
5 2-1867
23
977
1 11-4474
60
940
5 1-1672
22
978
1 10-4279
59
941
5 0-1478
21
979
1 9-4085
58
942
4 11-1283
20
980
1 8-3890
57
943
4 10-1089
19
981
1 7-3696
56
944
4 9-0894
18
982
1 6-3501
55
945
4 8-0699
17
983
1 5-3307
54
946
4 7-0505
16
984
1 4-3112
53
947
4 6-0310
15
985
1 3-2918
52
948
4 5-0116
14
986
1 2-2723
51
949
4 3-9921
13
987
1 1-2529
50
950
4 2-9727
12
988
1 0-2334
49
951
4 1-9532
11
989
11-2139
48
952
4 0-9338
10
990
10-1945
47
953
3 11-9143
9
991
9-1750
46
954
3 10-8949
8
992
8-1556
45
955
3 9-8754
7
993
7-1361
44
956
3 8-8559
6
994
6-1167
43
957
3 7-8365
5
995
5-0972
42
958
3 6-8170
4
996
4-0778
41
959
3 5-7976
3
997
3-0583
40
960
3 4-7781
2
998
2-0389
39
961
3 3-7587
1
999
1-0194
GOLD-VALUING TABLE.
XXX111
To convert MINT VALUE into BANK VALUE when the Standard is
expressed in Thousandths.
Thousandths
Value in Pence
Thousandths
Value in Pence
1
001636
6
009816 .
2
003272
7
011352
3
004908
8
013088
4
006544
9
014724
5
008180
To illustrate the use of the above table, gold of ^^ths fine
may be taken. As in the Table for finding the Bank value of
gold when the standard is reported in carats, &c., the amounts
in pence, as above, are to be deducted from the prices attached
to corresponding standards in Table No. 2. Thus, the minus
value of ^ 5 O ths is '00818 of a penny ; therefore, the minus value
of T 5_o_o_ths is *818 of a penny, which amount must be deducted
from the Mint price of gold at the above standard. On refer-
ring to the Table it will be found to be 2 2s. 5-7272d. per oz.
Now, if -818 be deducted, the' remainder will be 2 2s. 4'9092d.,
representing the Bank value of 1 oz. of gold of the fineness just
mentioned.
3o
XXXI V
ASSAY TABLE.
TABLE III.
ASSAY TABLE, showing the Amount of GOLD or SILVER, in Ounces,
Pennyweights, and Grains, contained in a Ton of Ore, &c.
from the Weight of Metal obtained in an Assay of 200 Grains
of Mineral.
If 200 Grains of
One Ton of Ore
If 200 Grains of
One Ton of Ore
Ore give of
will yield of
Ore give of
will yield of
FINE METAL
FINE METAL
FINE METAL
FINE METAL
Gr.
Oz. Dwts. Grs.
Gr.
Oz. Dwts. Grs.
001
036
031
516
002
6 12
032
5 4 12
003
9 19
033
5 7 19
004
13 1
034
5 11 1
005
16 8
035
5 14 8
006
19 14
036
5 17 14
007
1 2 20
037
6 20
008
1 6 3
038
643
009
1 9 9
-039
679
010
1 12 6
040
6 10 16
Oil
1 15 22
041
6 13 22
012
1 19 4
042
6 17 4
013
2211
043
7 11
014
2 5 17
044
7 3 17
015
290
045
770
016
2 12 6
046
7 10 6
017
2 15 12
047
7 13 12
018
2 18 19
048
7 10 19
019
321
049
8 1
020
358
050
838
021
3 8 14
051
8 6 14
022
3 11 20
052
8 9 20
023
3 15 3
053
8 13 3
024
3 18 9
054
8 16 9
025
4 1 16
055
8 19 16
026
4 4 22
056
9 2 22
027
484
057
964
028
4 11 11
058
9 9 11
029
4 14 17
059
9 12 17
030
4 18
060
9 16
ASSAY TABLE.
XXXV
If 200 Grains of
One Ton of Ore
If 200 Grains of
One Ton of Ore
Ore give of
will yield of
Ore give of
will yield of
FINE METAL
FINE METAL
FINE METAL
FINE METAL
Or.
Oz. Dwts. Grs.
6V.
Oz. Dwts. Grs.
061
9 19 6
105
17 3
062
10 2 12
106
17 6 6
063
10 5 19
107
17 9 12
064
10 9 1
108
17 12 19
065
10 12 8
109
17 16 1
066
10 15 14
110
17 19 8
067
10 18 20
111
18 2 14
068
11 2 3
112
18 5 20
069
11 5 9
113
18 9 3
070
11 8 16
114
18 12 9
071
11 11 22
115
18 15 16
072
11 15 4
116
18 18 22
073
11 18 11
117
19 2 4
074
12 1 17
118
19 5 11
075
12 5
119
19 8 17
076
12 8 6
120
19 12
077
12 11 12
121
19 15 6
078
12 14 19
122
19 18 12
079
12 18 1
123
20 1 19
080
13 1 8
124
20 5 1
081
13 4 14
125
20 8 8
082
13 7 20
126
20 11 14
083
13 11 3
127
20 14 20
084
13 14 9
128
20 18 3
085
13 17 16
129
21 1 9
086
14 22
130
21 4 16
087
14 4 4
131
21 7 22
088
14 7 11
132
21 11 4
089
14 10 17
133
21 14 11
090
14 14
134
21 17 17
091
14 17 6
135
22 1
092
15 12
136
22 4 6
093
15 3 19
137
22 7 12
094
15 7 1
138
22 10 19
095
15 10 8
139
22 14 1
096
15 13 14
140
22 17 8
097
15 16 20
141
23 14
098
16 3
142
23 3 20
099
16 3 9
143
23 7 3
100
16 6 16
144
23 10 9
101
16 9 22
145
23 13 16
"102
16 13 4
146
23 16 22
103
16 16 11
147
24 4
104
15 19 17
148
24 3 11
3 o 2
XXXVI
ASSAY TABLE.
If 200 Grains of
One Ton of Ore
If 200 Grains of
One Ton of Ore
Ore give of
will yield of
Ore give of
will yield of
FINE METAL
FINE METAL
FINE METAL
FINE METAL
Gr.
Oz. Dwts. Grs.
Gr.
Oz. Dwts. Grs.
149
24 6 17
193
31 10 11
150
24 10
194
31 13 17
151
24 13 6
195
31 17
152
24 16 12
196
32 6
153
24 19 19
197
32 3 12
154
25 3 1
198
32 6 19
155
25 6 8
199
32 10 1
156
25 9 14
200
32 13 8
157
25 12 20
201
32 16 14
158
25 16 3
202
32 19 20
159
25 19 9
203
33 3 3
160
26 2 16
204
33 6 9
161
26 5 22
205
33 9 16
162
26 9 4
206
33 12 22
163
26 12 11
207
33 16 4
164
26 15 17
208
33 19 11
165
26 19
209
34 2 17
166
27 2 6
210
34 6
167
27 5 12
211
34 9 6
168
27 8 19
212
34 12 12
169
27 12 1
213
34 15 19
170
27 15 8
214
34 19 1
171
27 18 14
215
35 2 8
172
28 1 20
216
35 5 14
173
28 5 3
217
35 8 20
174
28 8 9
218
35 12 3
175
28 11 16
219
35 15 9
176
28 14 22
220
35 18 16
177
28 18 4
221
36 1 22
178
29 1 11
222
36 5 4
179
29 4 17
223
36 8 11
180
29 8
224
36 11 17
181
29 11 6
225
36 15
182
29 14 12
226
36 18 6
183
29 17 19
227
37 1 12
184
30 1 1
228
37 4 19
185
30 4 8
229
37 8 1
186
30 7 14
230
37 11 8
187
30 10 20
231
37 14 14
188
30 14 3
232
37 17 20
189
30 17 9
233
38 1 3
190
31 16
234
38 4 9
191
31 3 22
235
38 7 16
192
31 7 4
236
38 10 22
ASSAY TABLE.
XXXV11
If 200 Grains of
One
Ton of Ore
If 200 Grains of
One
Ton of Ore
Ore give of
wi
11 yield
of
Ore give of
will yield
of
FINE METAL
FINE METAL
FINE METAL
FINE METAL
Gr.
Oz.
Dwts.
Grs.
Gr.
Oz.
Dwts.
Grs.
237
38
14
4
281
45
17
22
238
38
17
11
282
46
1
4
239
39
17
283
46
4
11
240
39
4
284
46
7
17
241
39
7
6
285
46
11
242
39
10
12
286
46
14
6
243
39
13
18
287
46
17
12
244
39
17
1
288
47
19
245
40
8
289
47
4
1
246
40
3
14
290
47
7
8
247
40
6
20
291
47
10
14
248
40
10
3
292
47
13
20
249
40
13
9
293
47
17
3
250
40
16
16
294
48
9
251
40
19
22
295
48
3
16
252
41
3
4
296
48
6
22
253
41
6
11
297
48
10
4
254
41
9
17
298
48
13
11
255
41
13
299
48
16
17
256
41
16
6
300
49
257
41
19
12
301
49
3
6
258
42
2
19
302
49
6
12
259
42
6
1
303
49
9
19
260
42
9
8
304
49
13
1
261
42
12
14
305
49
16
8
262
42
15
20
306
49
19
14
263
42
19
3
307
50
2
20
264
43
2
9
308
50
6
3
265
43
5
16
309
50
9
9
266
43
8
22
310
50
12
16
267
43
12
4
311
50
15
22
268
43
15
11
312
50
19
4
269
43
18
17
313
51
2
11
270
44
2
314
51
5
17
271
44
5
6
315
51
9
272
44
8
12
316
51
12
6
273
44
11
19
317
51
15
12
274
44
15
1
318
51
18
19
275
44
18
8
319
52
2
1
276
45
1
14
320
52
5
8
277
45
4
20
321
52
8
14
278
45
8
3
322
52
11
20
279
45
11
9
323
52
15
3
280
45
14
16
324
52
18
9
XXXV111
ASSAY TABLE.
If 200 Grains of
One Ton of Ore
If 200 Grains
of One Ton of Ore
Ore give of
will yield of
Ore give of
will yield of
FINE METAL
FINE METAL
FINE MKTAL
FINE METAL
Gr.
Oz. Dwts. Grs.
Gr.
Oz. Dwts. Grs.
325
53 1 16
369
60 5 9
326
53 4 22
370
60 8 16
327
53 8 4
371
60 11 22
328
53 11 11
372
60 15 4
329
53 14 17
373
60 18 11
330
53 18
374
61 1 17
331
54 1 6
375
61 5
332
54 4 12
376
61 8 6
333
54 7 19
377
61 11 12
334
54 11 1
378
61 14 19
335
54 14 8
379
61 18 1
336
54 17 14
380
62 1 8
337
55 20
381
62 4 14
338
55 4 3
382
62 7 20
339
55 7 9
383
62 11 3
340
55 10 16
384
62 14 9
341
55 13 22
385
62 17 16
342
55 17 4
386
63 22
343
56 11
387
63 4 4
344
56 317
388
63 7 11
345
56 7
389
63 10 17
346
56 10 6
390
63 14
347
56 13 12
391
63 17 6
348
56 16 19
392
64 12
349
57 1
393
64 3 19
350
57 38
394
64 7 1
351
57 6 14
395
64 10 8
352
57 9 20
396
64 13 14
353
57 13 3
397
. 64 16 20
354
57 16 9
398
65 3
355
57 19 16
399
65 3 9
356
58 2 22
400
65 6 16
357
5S 64
401
65 9 22
358
58 9 11
402
65 13 4
359
58 12 17
403
65 16 11
360
58 16
404
65 19 17
361
58 19 6
405
66 3
362
59 9 12
406
66 6 6
363
59 5 19
407
66 9 12
364
59 9 1
408
66 12 19
365
59 12 8
409
66 16 1
366
59 15 14
410
66 19 8
367
59 18 20
411
67 2 14
368
60 2 3
4T2
67 5 20
ASSAY TABLE.
XXXIX
If 200 Grains of
One Ton of Ore
If 200 Grains of
One Ton of Ore
Ore give of
will yield of
Ore give of
will yield of
FINE METAL
FINE METAL
FINE METAL
FINE METAL
Gr.
Oz. Dwts. Grs.
Gr.
Oz. Dwts. Grs.
413
67 9 3
457
74 12 20
414
67 12 9
458
74 16 3
415
67 15 16
459
74 19 9
416
67 18 22
460
75 2 16
417
68 2 4
461
75 5 22
418
68 5 11
462
75 9 4
419
68 8 17
463
75 12 11
420
68 12
464
75 15 17
421
68 15 6
465
75 19
422
68 18 12
466
76 2 6
423
69 1 19
467
76 5 12
424
69 5 1
468
76 8 19
425
69 8 8
469
76 12 1
426
69 11 14
470
76 15 8
427
69 14 20
471
76 18 14
428
69 18 3
472
77 1 20
429
70 1 9
473
77 5 3
430
70 4 16
474
77 8 9'
431
70 7 22
475
77 11 16
432
70 11 4
476
77 14 22
433
70 14 11
477
77 18 4
434
70 17 17
478
78 1 11
435
71 1
479
78 4 17
436
71 4 6
-480
78 8
437
71 7 12
481
78 11 6
438
71 10 19
482
78 14 12
439
71 14 1
483
78 17 19
440
71 17 8
484
79 1 1
441
72 14
485
79 4 8
442
72 3 20
486
79 7 14
443
72 7 3
487
79 10 20
444
72 10 9
488
79 14 3
445
72 13 16
489
79 17 9
446
72 16 22
490
80 16
447
73 4
491
80 3 22
448
73 3 11
492
80 7 4
449
73 6 17
493
80 10 11
450
73 10
494
80 13 17
451
73 13 6
495
80 17
452
73 16 12
496
81 6
453
73 19 19
497
81 3 12
454
74 3 1
498
81 6 19
455
74 6 8
499
81 10 1
45(5
74 9 14
500
81 13 8
xl
ASSAY TABLE.
If 200 Grains of
One
Ton of Ore
If 200 Grains of
One
Ton of Ore
Ore give of
wi
11 yield of
Ore give of
will yield of
FINK METAL
FINE METAL
FINE METAL
FINE METAL
Gr.
Oz.
Dwts.
Grs.
Gr.
Oz.
Dwts.
Grs.
501
81
16
14
545
89
8
502
81
19
20
546
89
3
14
503
82
3
3
547
89
6
20
504
82
6
9
548
89
10
3
505
82
9
16
549
89
13
9
506
82
12
22
550
89
16
16
507
82
16
4
551
89
19
22
5C8
82
19
11
552
90
3
4
509
83
2
17
553
90
6
11
510
83
6
554
90
9
17
511
83
9
6
555
90
13
512
83
12
12
556
90
16
6
513
83
15
19
557
90
19
12
514
83
19
1
558
91
2
19
515
84
2
8
559
91
6
1
516
84
5
14
560
91
9
8
517
84
8
20
561
91
12
14
518
84
12
3
562
91
15
20
519
84
15
9
563
91
19
3
520
84
18
16
564
93
2
9
521
85
1
22
565
92
5
16
522
85
5
4
566
92
8
22
523
85
8
11
.567
92
12
4
524
85
11
17
568
92
15
11
525
85
15
569
92
18
17
526
85
18
6
570
93
2
527
86
1
12
571
93
5
6
528
86
4
19
572
93
8
12
529
86
8
1
573
93
11
19
530
86
11
8
574
93
15
1
531
86
14
14
575
93
18
8
532
86
17
20
576
94
1
14
533
87
1
3
577
94
4
20
534
87
4
9
578
94
8
3
535
87
7
16
579
94
11
9
536
87
10
22
580
94
14
16
537
87
14
4
581
94
17
22
538
87
17
11
582
95
1
4
539
88
17
583
95
4
11
540
88
4
584
95
7
17
541
88
7
6
585
95
11
542
88
10
12
586
95
14
6
543
88
13
19
587
95
17
12
544
88
17
1
588
96
19
ASSAY TABLE.
xli
If 200 Grains of
One Ton of Ore
If 200 Grains of
One Ton of Ore
Ore give of
will yield of
Ore give of
will yield of
FINE METAL
FINE METAL
FINE METAL
FINE METAL
Gr.
Oz. Dwts. Grs.
Gr.
Oz. Dwts. Grs.
589
96 4 1
633
103 7 19
590
96 7 8
634
103 11 1
591
96 10 14
635
103 14 8
592
96 13 20
636
103 17 14
593
96 17 3
637
104 20
594
97 9
638
104 4 3
595
97 3 16
639
104 7 9
596
97 6 22
640
104 10 16
597
97 10 4
641
104 13 22
598
97 13 11
642
104 17 4
599
97 16 17
643
105 11
600
98
644
105 3 17
601
98 3 6
645
105 7
602
98 6 12
646
105 10 6
603
98 9 ]9
647
105 13 12
604
98 13
648
105 16 19
605
98 16 8
649
106 1
606
98 19 14
650
106 3 8
607
99 2 20
651
106 6 14
608
99 6 3
652
106 9 20
609
99 9 9
653
106 13 3
610
99 12 16
654
106 16 9
611
99 15 22
655
106 19 16
612
99 19 4
656
107 2 22
613
100 2 11
657
107 6 4
614
100 5 17
658
107 9 11
615
100 9
659
107 12 17
616
100 12 6
660
107 16
617
100 15 12
661
107 19 6
618
100 18 19
662
108 2 12
619
101 2 1
663
108 5 19
620
101 5 8
664
108 9 1
621
101 8 14
-665
108 12 8
622
101 11 20
666
108 15 14
623
101 15 3
667
108 18 20
624
101 18 9
668
109 2 3
625
102 1 16
669
109 5 9
626
102 4 22
670
109 8 16
627
102 8 4
671
109 11 22
628
102 11 11
672
109 15 4
629
102 14 17
673
109 18 11
630
102 18
674
110 1 17
631
103 1 6
675
110 5
632
103 4 12
676
110 8 6
xlii
ASSAY TABLE.
If 200 Grains of
One
Ton of Ore
If 200 Grains
of One
Ton of Ore
Ore give of
will yield
of
Ore give of
will yield
of
FINE METAL
FINE METAL
FINE METAL
FINE METAL
Gr.
Oz.
Dwts.
Grs.
Gr.
o*:
Dwts.
Gra.
677
110
11
12
721
117
15
6
678
110
14
19
722
117
18
12
679
110
18
1
723
118
1
19
680
111
1
8
724
118
5
1
681
111
4
14
725
118
8
8
682
111
7
20
726
118
11
14
683
111
11
3
727
118
14
20
684
111
14
9
728
118
18
3
685
111
17
6
729
119
1
9
686
112
22
730
119
4
16
687
112
4
4
731
119
7
22
688
112
7
11
732
119
11
4
689
112
10
17
733
119
14
11
690
112
14
734
119
17
17
691
112
17
6
735
120
1
692
113
12
736
120
4
6
693
113
3
19
737
120
7
12
694
113
7
1
738
120
10
19
695
113
10
8
739
120
14
1
696
113
13
14
740
120
17
o
697
113
16
20
741
121
14
698
114
3
742
121
3
20
699
114
3
9
743
121
7
3
700
114
6
16
744
121
10
9
701
114
9
22
745
121
13
6
702
114
13
4
746
121
16
22
703
114
16
12
747
122
4
704
114
19
17
748
122
3
11
705
115
3
749
122
6
17
706
115
6
6
750
122
10
707
115
9
12
751
122
13
16
708
115
12
19
752
122
16
12
709
115
16
1
753
122
19
19
710
115
19
8
754
123
3
1
711
116
2
14
755
123
6
8
712
116
5
20
756
123
9
14
713
116
9
3
757
123
12
20
714
116
12
9
758
123
16
3
715
116
15
16
759
123
19
9
716
116
18
22
760
124
2
16
717
117
2
4
, -761
124
5
22
718
117
5
11
762
124
9
4
719
117
8
17
763
124
12
11
720
117
12
764
124
15
17
ASSAY TABLE.
xliii
If 200 Grains
of One Ton of Ore
If 200 Grains of
One Ton of Ore
Ore give of
will
yield
of
Ore give of
will yield
of
FINE METAL
FINE METAL
FINE METAL
FINE METAL
Gr.
Oz.
Lwts.
Grs.
Gr.
0*.
Dwts.
Grs.
765
124
19
809
132
2
17
766
125
2
6
810
132
6
767
125
5
12
811
132
9
6
768
125
8
19
812
132
12
12
769
125
12
1
813
132
15
19
770
125
15
8
814
132
19
1
771
125
18
14
815
133
2
8
772
126
1
20
816
133
5
14
773
126
5
3
817
133
8
20
774
126
8
9
818
133
12
3
775
126
11
16
819
133
15
9
776
126
14
22
820
133
18
16
777
126
18
4
821
134
1
22
778
127
1
11
822
134
5
4
779
127
4
17
823
134
8
11
780
127
8
824
134'
1!
17
781
127
11
6
825
134
15
782
127
14
12
826
134
18
6
783
127
17
19
827
135
1
12
784
128
1
1
828
135
4
19
785
128
4
8
829
135
8
1
786
128
7
14
830
135
11
8
787
128
10
20
831
135
14
14
788
128
14
3
832
135
11
8
789
128
17
9
833
136
1
3
790
129
16
834
136
4
9
791
129
3
22
835
136
7
16
792
129
7
4
836
136
10
22
793
129
10
11
837
136
14
4.
794
129
13
17
838
136
17
11
795
129
17
839
137
17
796
130
6
840
137
4
797
130
3
12
841
137
7
6
798
130
6
19
842
137
10
12
799
130
10
1
843
137
13
19
800
130
13
8
844
137
17
1
801
130
16
14
845
138
8
802
130
19
20
846
138
3
14
803
131
3
3
847
138
6
20
804
131
6
9
848
138
10
3
805
131
9
16
849
138
13
19
806
131
12
22
850
138
16
16
807
131
16
4
851
138
19
22
808
131
19
11
852
139
3
4
xliv
ASSAY TABLE.
If 200 Grains of
One
Ton of Ore
If 200 Grains of
One Ton of Ore
Ore give of
will yield
of
Ore give of
will
yield
of
FINE METAL
FINE METAL
FINE METAL
FINE METAL
Gr.
Oz.
Dwts.
Grs.
Gr.
Oz.
Dwts.
Grs.
853
139
6
11
897
146
10
4
854
139
9
17
898
146
13
11
855
139
13
899
146
16
17
856
139
16
6
900
147
857
139
19
12
901
147
3
6
858
140
2
19
902
147
6
12
859
140
6
1
903
147
9
19
860
140
9
8
904
147
13
1
861
140
12
14
905
147
16
8
862
140
15
20
906
147
19
14
863
140
19
3
907
148
2
2
864
141
2
9
908
148
6
3
865
141
5
16
909
148
9
9
866
141
8
22
910
148
12
16
867
141
12
4
911
148
15
21
868
141
15
11
912
148
19
4
869
141
18
17
913
149
2
11
870
142
2
914
149
5
17
871
142
5
6
915
149
9
872
142
8
12
916
149
12
6
873
142
11
19
917
149
15
12
874
142
15
1
918
149
18
19
875
142
18
8
919
150
2
876
143
1
14
920
150
5
8
877
143
4
20
921
150
8
14
878
143
8
3
922
150
11
20
879
143
11
9
923
150
15
3
880
143
14
16
924
150
18
9
881
143
17
22
925
151
1
16
882
144
1
4
926
151
4
22
883
144
4
11
927
151
8
4
884
144
7
17
928
151
11
11
885
144
11
929
151
14
17
886
144
14
6
930
151
18
887
144
17
12
931
152
1
6
888
145
19
932
152
4
12
889
145
4
1
933
152
7
19
890
145
7
8
934
152
11
1
891
145
10
14
935
152
14
8
892
145
13
20
936
152
17
14
893
145
17
3
937
153
20
894
146
9
938
153
4
3
895
146
3
16
939
153
7
9
896
146
6
22
940
153
10
16
ASSAY TABLE.
xlv
If 200 Grains of
One Ton of Ore
If 200 Grains
of One
Ton of Ore
Ore give of
will yield of
Ore give of
will yield of
FINE METAI,
FINE METAL
FINE METAL
FINE METAL
Gr.
Oz.
Dwts.
GTS.
Gr.
Oz.
Dwts.
Grs.
941
153
13
22
985
160
17
6
942
153
17
4
986
161
22
943
154
11
987
161
4
4
944
154
3
17
988
161
7
11
945
154
7
989
161
10
17
946
154
10
6
990
161
14
947
154
13
12
991
161
17
6
948
154
16
19
992
162
12
949
155
1
993
162
3
19
950
155
3
8
994
162
7
1
951
155
6
14
995
162
10
8
952
155
9
20
996
162
13
14
953
155
13
3
997
162
16
20
954
155
16
9
998
163
3
955
155
19
16
999
163
3
9
956
156
2
22
1
grain
163
6
16
957
156
6
4
2
326
13
8
958
156
9
11
3
490
959
156
12
17
4
653
6
16
960
156
16
5
816
13
8
961
156
19
6
6
980
962
157
2
12
7
1143
6
16
963
157
5
19
8
1306
13
8
964
157
9
1
9
1470
965
157
12
8
10
1633
6
16
966
157
15
14
11
1796
13
8
967
157
18
20
12
1960
968
158
2
3
13
2123
6
16
969
158
5
9
14
2286
13
8
970
158
8
16
15
2450
971
158
11
22
16
2613
6
16
972
158
15
4
17
2776
13
8
973
158
18
11
18
2940
974
159
1
17
19
3103
6
16
975
159
5
20
3266
13
8
976
159
8
6
21
3430
977
159
11
12
22
3593
6
16
978
159
14
19
23
3756
13
8
979
159
18
1
24
3920
980
160
1
8
25
4083
6
16
981
160
4
14
26
4246
13
8
982
160
7
20
27
4410
983
160
10
3
28
4573
6
16
984
160
14
9
29
4736
13
8
xlvi
ASSAY TABLE.
If 200 Grains of One Ton of Ore
If 200 Grains
of One
Ton of Ore
Ore give
of will
yield of
Ore give of will yield of
FINE METAL FINE METAL
FINE METAL FINE METAL
Grs.
0*:
Dwts.
Grs.
Grs.
Oz.
Dwts.
Grs.
30
4900
74
12086
13
8
31
5063
6
16
75
12250
32
5226
13
8
76
12413
6
16
33
5390
77
12576
13
8
34
5553
6
16
78
12740
35
5716
13
8
79
12903
6
16
36
5880
80
13066
13
8
37
6043
6
16
81
13230
38
6206
13
8
82
13393
6
16
39
6370
83
13556
13
8
40
6533
6
16
84
13720
41
6696
13
8
85
13883
6
16
42
6860
86
14046
13
8
43
7023
6
16
87
14210
44
7186
13
8
88
14373
6
16
45
7350
89
14536
13
8
46
7513
6
16
90
14700
47
7676
13
8
91
14863
6
16
48
7840
92
15026
13
8
49
8003
6
16
93
15190
50
8166
13
8
94
15353
6
16
51
8330
95
15516
13
8
52
8493
6
16
96
15680
53
8656
13
8
97
15843
6
16
54
8820
98
16006
13
8
55
8983
6
16
99
16170
56
9146
13
8
100
16333
6
16
57
9310
101
16496
13
8
58
9473
6
16
102
16660
59
9636
13
8
103
16823
6
16
60
9800
104
16986
13
8
61
9963
6
16
105
17150
62
10126
13
8
106
17313
6
16
63
10290
107
17476
13
8
64
10453
6
16
108
17640
65
10616
13
8
109
17803
6
16
66
10780
110
17966
13
8
67
10943
6
16
111
18130
68
11106
13
8
112
18293
6
16
69
11270
113
18456
13
8
70
11433
6
16
114
18620
71
11596
13
8
115
18783
6
16
72
11760
116
18946
13
8
73
11923
6
16
117
19110
ASSAY TABLE.
xlvii
If 200 Grains
of One
Ton of Ore
If 200 Grains of One
Ton of Ore
Ore give of
will yield
of
Ore give of
will yield
of
FINE METAL
FINE METAL
FINE METAL
FINE METAL
Grs.
a*.
Dwts
Grs.
Grs.
02.
Dwts
Grs.
118
19273
6
16
160
26133
6
16
119
19436
13
8
161
26296
13
8
120
19600
162
26460
121
19763
6
16
163
26623
6
16
122
19926
13
8
164
26786
13
8
123
20090
165
26950
124
20253
6
16
166
27113
6
16
125
20416
13
8
167
27276
13
8
126
20580
168
27440
127
20743
6
16
169
27603
6
16
128
20906
13
8
170
27766
13
8
129
21070
171
27930
130
21233
6
16
172
28093
6
16
131
21396
13
8
173
28256
13
8
132
21560
174
28420
133
21723
6
16
175
28583
6
16
134
21886
13
8
176
28746
13
8
135
22050
177
28910
136
22213
6
16
178
29073
6
16
137
22376
13
8
179
29236
13
8
138
22540
180
29400
139
22703
6
16
181
29563
6
16
140
22866
13
8
182
29726
13
8
141
23030
183
29890
142
23193
6
16
184
30053
6
16
143
23356
13
8
185
30216
13
8
144
23520
186
30380
. 145
23683
6
16
187
30543
6
16
146
23846
13
8
188
3U706
13
8
147
24010
189
30870
148
24173
6
16
190
31033
6
16
149
24336
13
8
191
31196
13
8
150
24500
192
31360
151
24663
6
16
193
31523
6
16
152
24826
13
8
194
31686
13
8
153
24990
195
31850
154
25153
6
16
196
32013
6
16
155
25316
13
8
197
32176
13
8
156
25480
198
32340
157
25643
6
16
199
32503
6
16
158
25806
13
8
200
32666
13
8
159
25970
INDEX.
ABS
ABSOLUTE heating power of fuel,
154
Acid, boracic, 221
coloured flame, 295
nitric, 213
oxalic, 174
phosphoric, coloured flame of, 296
tartaric, 174
Acids, 3
Addison's process for the estimation of
carbon in steel, 367
Adhesive paste, 115
Agate, 250
mortars, 20*
Alabaster, 261
Alkalies, action of, on galena, 503
caustic, 181
in iron ores, 355
Alkaline persulphides, 190
Alloys, metallic, capable of direct cupel-
lation, 729, 730
incapable of direct cupellation,
729, 733
of antimony and lead, assay of, 565
copper, 434
- gold, 740, 741, 768, 769
silver, platinum, and copper,
759
standard of, 768
lead, assay of, 530
silver, 600
and copper, assay 1 of, 627, 632
- zinc, assay of, 574
Almandine, 884
ruby, 257
Almond meal, 114
Alum, 288
Alumina crucibles, 130
in iron ores, 341, 349
Aluminium blowpipe support, 212
- silicate, 192
silicates, 291
Amalgam, assay of, 737
Amalgamation of silver, 626
Amethyst, 249, 890
Ammonium fluoride, 222
Analysis, colorimetric, 307
ARS
Analysis of fuel, 150
iron ores, calculation of results,
349
platinum ores, 800
volumetric, 298
Anglesite, 271, 289
Anthracite, 171
Antimonial silver, assay of, 735
or arsenical silver ores, 282
silver ore, 284, 286
ores, 276
Antimonides of silver, 633
Antimonium crudum, assay of, 557
Antimoniuretted from arseniuretted
hydrogen, to distinguish, 564
Antimony and arsenic, separation of tin
from, 564
and lead, assay of alloys of, 565
sulphides, 277
assay of, 556
coloured flaine, 295
in' iron ores, 356
sublimates, detection of, 563
regulus, assay of, 558
separation of, from bismuth, 815
sulphide, 186, 189, 263, 284
assay of, 557
sulphides, 282
Anvil, 14
Apatite, 258, 289, 290
Apparatus for silver cupellation before
the blowpipe, 717
Appendix, i.
Aquamarine, 254, 892
Aqua regia, parting with, 766
Aragonite, 260, 288
Argentiferous tin, assay of, 734
Argol, 199
action of, on galena, 503
reducing power of, 607
Armentiferous zinc, assay of, 735
Arnold's apparatus for analysis of iron
and steel, 373
process for estimation of chromium
in iron and steel, 827
Arquerite, assay of, 737
Arseniate of copper, 434
3p
INDEX.
ARS
Arseniate of iron, 287
Arsenic, 264, 284
- and antimony, separation of tin
from, 564
assay of, 831
colour of flame, 295
in copper, 499
iron ores, 356
separation of, from bismuth, 814
sulphide, 264, 286
sulphur, nickel and cobalt, assay
of, 851
Arsenical nickel, 284
-268
or antimonial silver ores, 282
pyrites, 265
silver ore, 284, 286
ores, 276
Arseniuretted from antimoniuretted
hydrogen, to distinguish, 564
Asbestos, 252
card, 213
Ash in fuel, 70
estimation of, 162
pit, 58
Assay balances, 28
before blowpipe, 235
complete, of iron ores, 332
crucible, of silver ores, 605
English copper, 435
furnace, gas, 87
in scorifier, 615
Lake Superior copper, 455
mode of taking from ingot, 687
of antimonial silver, 735
antimony, 556
argentiferous iron and steel, 737
mercury, 737
tin, 734
arsenic, 831
auriferous copper ores, 775
bell metal, 734
bismuth, 809
brass, 733
bronze, 734
chromium, 820
copper, electrolytic, 480
in the wet way, 461
fuel, 150
galena, 521
gold, 740
coin and bullion, 769
gun metal, 734
iron, 308
in the dry way, 309
wet way, 321
lead, 502
- by fusion with black flux, 510
- iron, 511
potassium carbonate,
604
sodium carbonate or
black flux and metallic iron, 514
BER
Assay of lead, roasting and reducing,
515
with standard solutions, 531
sulphuric acid, 519
manganese, 833
mercury, 587
blowpipe, 598
nickel and cobalt, 839
platinum, 781
ores, 795
pure tin oxide, 539
pyrites for gold, 773
saltpetre, 177, 178
silver, 600
and copper, 739
before the blowpipe, 715
- by the wet way, 634, 647, 674
sulphur, 859
in wet way, 861
telluric silver, 735
zinc, 567
volumetrical, 575
' pound,' 694
spectroscopic of gold, 779
table for gold, xxxiv
volumetric of bismuth, 817
chromium, 827
mercury, 595
Asterias, 256
Atkinson's modification of Penny's
volumetric assay of iron, 325
Atomic weights, 2, 3, 6
Attwood's blowpipe assay of mercury,
598
Aufrye and d'Arcet's muffle furnace, 63
Augite, 252, 289, 290
Azurite, 434
BALANCE, the, 26
theory of the, 28
Balas ruby, 257
Balling's process for titration of silver
in galena, 714
Bank value into mint value, table to
convert, xxxiii
Bankers or bullion balance, 27
Bars of furnace, 59
Baryta, coloured flame of, 296
Barytes, 262
Basic cinder in manufactured iron, 414
Beale's cement, 120
Bears from smelting furnaces, assay of
737
Becker's process for assay of antimony
563
Beech -wood ash, 143
Bell metal, assay of, 734
tin in, 549, 550
Berthier's assay of mercury, 590
method for assay of fuel, 155
Beryl, 254
blue, 889
INDEX.
li
BER
Beryl or emerald, 293
Bettell's process for estimating cinder
in iron, 415
titanium, 432
Binary substances, 3, 5
Bioitite, 869
Bismuth, 263, 284, 286
assay of, 809
cupellation of silver in, 730
effect of, on ductility of silver, 699
ores, 809
purification of, from arsenic, 814
antimony, 815
copper, 815
sulphur, 816
refining crude, 813
volumetric, assay of, 817
Bisulphite of ammonium for reduction
of iron salts, 328
Black band iron stone, 308
cupric oxide, 274
flux, 197
assay of lead by fusion with, 510
lead, 258
crucibles. 123
or graphite, 170
Blast furnace, 60
cinder, 310
Blende, 187, 269, 285, 287, 289
assay of, 573
cupriferous assay of, 573
Bloodstone, 250
Blossom's process of assaying gold ores,
741
Blowpipe and its uses, 202
assay of coal,165
mercury, 598
nickel, 849
silver, 715
operations, 234
reagents and fluxes, 213
Blue copper carbonate, 275
flames, 294
Boeckmaiin's assay of sulphur, 866
Boiler cement, 120
Boiling or evaporating furnace, 110,
990
Bone-ash, 717
for cupels, 141
ashes, 224
Boracic acid, coloured flame, 295
- vitrified, 221
Borate of lead, 177, 201
Borax, 193, 218
bead, colour of, 246
Boron, test for, 222
Bournonite, 434
Brasque crucibles, 312
Brass, assay of, 584, 733
on glass cement, 114
Braunite, 833
Brazil-wood paper, 213
Brick muffle furnace, 64
CAR
Bricks, magnesia, 130
Brightening of silver, 620
Britton's burette, 305
modification of Eggertz's estimation
of combined carbon in iron and steel,
374
Bromine, coloured flame, 295
Bronze, assay of, 584, 734
tin in, 550
Bronzite, 869
Brown ochre, 267
Brown's gas assay furnace, 87
volumetric assay of copper, 479
Brunton's automatic sampling machine,
11
Bruyeres cement, 120
Buisson's volumetric assay of lead, 533
Bullion, assay of, 769
silver, assay of, 631
Bunsen's gas burner, 105, 113
process for assay of manganese ores,
834
platinum ores, 782
Burette, 302, 636, 637
Burse's process for assay of tin in
bronze, 550
Busteed's process for assay of silver in
Indian mints, 687
,250
\J Calamine, 270, 289
Calcination, 42
Calcining furnace, 55
Calcite, 259
Calcium, coloured flame of, 297
fluoride, 195, 223
silicate, 192
sulphate, 223
Gale-spar, 259, 288
Calculation of results, 164
Caoutchouc, 117
Carbon as a reducing agent, 170
combined, in iron and steel, estima-
tion of, 371
in iron, inorganic standards for the
colorimetric test, 381
and steel, 363
estimation of minute
quantities of, 376
ores, 351
steels, estimation of, 367
Carbonaceous matter in iron ores, 351
Carbonate of copper, 287, 289
iron, 268
lead, 271
sodium, 214
zinc, 270
Carbonates, alkaline, 187
of copper, blue and green, 275
potassium and sodium, 181, 195
Carbonic acid in iron ores, 349
3 p 2
Hi
INDEX.
CAR
Carbonisation of fuel, volatile products
of, 161
Carnelian, 250
Carnelly's colorimetric assay of copper,
470
Caustic alkalies and carbonates, 187
Cement for brass on glass, 114
mending pestles, 115
waterproof, 116
Cementation, 47
Ceruse, 201
or white lead, 176
Cerusite, 288, 271, 289
Chalcedony, 250
Chalcopyrite, 434
Chapman's process for detection of
antimony in sublimates, 563
detection of copper in iron pyrites,
497
Charcoal. 182
crucibles, 126
for blowpipe, 210
furnaces, 70
fuel, examination of 161, 182
lined crucibles, 312
Chemical characters of minerals, 244
Chiastolite, 291
Chimney, 56, 59
Chisel, cold, 16
Chloride of silver, 224, 277
reduction of, 684
sodium, 196
process for assay of silver, 687
Chlorine, blue flame of, 295
Chlorite, 251, 290
Chlorcspinel, 257
Chromate of copper, 434
Chrome diallage, 869
iron ore, assay of, 820
Chromic iron, 267, 285, 287
Chromium, assay of, 820
in iron and steel, estimation of, 827
ores, 317
Chromometer, new form of, 380
. Chrysoberyl, 257, 293, 879
Chrysolite, 252, 291, 891
Chrysoprase, 892
Cinnabar, 189, 272, 282, 284, 286, 287
in ore assay of, 592
Cinnamon stone, 886
Clarke's process for assay of chrome
iron ore, 824
Claus's method of decomposing osmi-
ridium, 803
Clay iron stone, 308
Cleaning platinum crucibles, 136, 138
Cloud's colorimetric assay of copper, 467
Coal, 258
assay of, before the blowpipe, 165
valuation of, for the production of
illuminating gas, 167
Coating retorts, 119
Cobalt and nickel, separating, 839, 846
COP
Cobalt assay of, 839
for traces of, in nickel, 849
bloom, 268, 287
glance assay of, 851
in iron ores, 344
nitrate, 221
ores, 839, 846
speiss, assay of, 843, 851
tin-white, 269
Coin, gold, assay of, 769
Cold chisel, 16
Colorimetric analysis, 307
copper assay, 461
Coloured flames, 294
Colour of borax bead, 246
minerals, 239
Coke, 171
for furnaces, 70
from fuel, examination of, 161
Compactness of fuel, 152
Condensation, 52
Copper, 213, 284
alloys, 434
and lead speiss, assay of, 855
silver, assay of, 739
alloys, assay of 627, 632
arseniate, 434
arsenic in, 499
assay colorimetric, 461
in the wet way, 461
Lake Superior, 455
of, 434
volumetric, 476
sulphide, argentiferous, 600
carbonate, 287
carbonates, blue and green, 275
chromate, 434
coloured flame of, 296
electrolytic, assay of, 480
from zinc, separation of, 584
glance, 434
gold alloys, assay of, 756
grey, 274, 282, 283, 284
in iron ores, 356
pyrites, 497
native, 273
nickel, 831.
ores, 434
assay of auriferous, 775
oxide, 181, 201, 221
red, 285
and black, 287
phosphate, 434
pyrites, 273, 282, 434 .
assay of, for sulphur, 859
494, 496
selenide, 277
silicate, 434
silver and platinum alloys, assay of,
633
alloys, cupellation of, 732
separation of, from bismuth, 815
sulphate, 181, 288
INDEX.
liii
COP
Copper sulphide, 185
titration of silver in presence of,
713
vanadate, 434
vitreous, 282
zinc and nickel alloys, assay of, 853
Cornish crucibles, 120
Corundum, 256, 293
Coruscation of silver, 620
Covellite, 434
Cream of tartar, 199
Crucible assay of gold ores, 742
perforated, 361
Crucibles, cupels, &c., 120
for assay of iron ores, 311
calcination, 42
mounts for, 107
supporting, 66
Crystalline form, 237
Cube, 237
Cupel bottoms, assay of, 525
moulds, 142
or muffle furnace, 62
Cupels, 141, 618
crucibles, &c., 120
Cupellation, 54, 617, 720
- loss, 727
loss in assay of copper and silver
alloys, 630
of gold, 749
and silver, lead to be employed
in, 759
silver before the blowpipe, 716
in lead or bismuth, 730
Cupric oxide, black, 274
Cupriferous bismuth, assay of, 810
blende, assay of, 573
Cuprous oxide, red, 274, 285, 287
Cyanide of potassium, 214
Cyanite, 888
Cyanosite, 434
Cymophane, 879
DANIEL'S pyrometer, 144
D'Arcet and Aufrye's muffle fur-
nace, 63
Day standard colours, 382
Day's process for assay of chrome iron
ore, 825
Debray's process for assay of platinum
ores, 795
Decinormal solutions, 300
De Clauby's process for volumetric
assay of tin, 550
Desulphurising agents, 181
Deutecom's assay of sulphur, 864
Deville's furnace, 62
process for assay of platinum ores
795
tin in gun and bell
metal, 549
Diamond, 257, 292, 294, 867
FER
Dichroite, 889
Diehl's assay of lead, 534
)isthene, 888
Distillation, 51
of mercury, 587
zinc, 568
Distilled water, 52
Dodecahedron, 237
Dolomite, 260, 288
Domeykite, 434
Draught crucible gas furnace, 91
Dressing, washing, or vanning, 23
Drill for silver ingots, 687
[)rip proof gas burner, 112
Dry assay of iron, 309
distillation, 51
Dumonte's assay of lead with standard
solutions, 531
Dyce's process for separating silver from
base metals, 633
T7ARTHENWARB retorts, 53
Jj Effects produced by wind and
blast furnaces, 72
Egg and lime lute, 114
Eggertz's colour test for combined car-
bon in iron and steel, 372
process for estimating sulphur in
iron and steel, 388
the estimation of combined
carbon in iron and steel, 371
graphite in iron and
steel, 369
Electrolytic assay of copper, 480
mercury, 592
Elements, 1, 2, 3
Elutriation, 22
Emerald, 254, 292, 293
yellow, 881
green, 892
Emery, 256
Endemann's colorimetric assay of
copper, 467
English copper assay, 435
Epidote, 290
Epsomite, 288
Eschka's process for assay of mercury,
591
Escosura's [L. de la] electrolytic assay
of mercury, 592
Essonite, 886
Evaporating furnaces, 57
FjUHLERZ, 434
argentiferous, 600
Faraday's blast furnace, 60
blowpipe directions, 206
Fat lute, 113
Fatty oils, 172
Felspar, 255, 290, 895
Ferric oxide, 266
liv
INDEX.
FER
Ferric oxide, brown and red, 287
in iron ores, 341
Ferro-manganese, manganese in, 426
Ferrous oxide in iron ores, 346, 347
Fire lute, 113
Fish-eye, 895
Flame for blowpipe, 205
Flasks, measuring, 307
Fleck's volumetric assay of copper, 476
Fleitmann's process for assay of traces
of cobalt in nickel. 849
Fletcher's draught crucible furnace, 91
drip proof gas burner, 112
gas reverberatory furnace, 90
injector gas furnace, 92
muffle gas furnace, 91
petroleum furnace, 94
safety Bunsen burner, 113
solid flame gas burner, 111
universal gas furnace, 84
Flint, 250
Fluoride of ammonium, 222
calcium, 223
Fluor spar, 195, 259, 290
Flux, Turner's, 222
Flux, white, black, and raw, 197
Fluxes, 191
blowpipe, 213
for iron ores, 309, 310
metallic, 201
reducing power of, 200
Fluxing crucibles, 122
Forbes's blowpipe assay of silver, 715
charcoal, 210
directions, 203
lamp, 206
charcoal borers, 211
glass and platinum forceps, 50
lime crucibles, 129
soda-paper, 233
Forceps, glass and platinum, 50
Fracture of fuel, 152
minerals, 238
Franklinite, 308
Fresenius's assay of copper pyrites, 496
Fuel, assay of, 150
external appearance of, 152
for furnaces, 70
Furnace, blast, 60
calcining, 55
evaporating, 57
fusion, 57
operations, 66
universal, 65
wind, 57
Furnaces, 55
Fusion, 48
Fusion furnace, 57
Fusibility of minerals, 244
GALENA, 189, 270, 282
action of oxygen on, 502
GOL
Galena action of metallic iron on, 502
alkalies and alkaline carbo-
nates on, 502
potassium nitrate on, 502
argol on, 502
assay of, in the wet way, 521
separation of silver from, 627
titration of silver in, 714
Galletti's process for assay of zinc, 575
Garnet, 253, 290, 291, 293, 884
noble, 886
Garrett's burette, 340
Gas assay furnace, 87
burner, Bunsen's, 105
solid flame, 111
furnaces, 74
illuminating, valuation of coal for
the production of, 167
reverberatory furnace, 90
Gay-Lussac's burette, 304
Gems, 291
and precious stones, discrimination
of, 867
Genth's process for assay of chrome
iron ore, 820
German silver, assay of, 733
Gibbs's process for assay of chrome iron
ore, 823
separating nickel and cobalt,
846
of assaying platinum ores, 804
Glanzkobalt, 831
Glass, 193, 290
analysis of various kinds of, 194
and artificial gems, 896
forceps, 50
of antimony, 557
lead, 201
retorts, 53
to brass, cementing, 114
Globules of silver, weight of, 723
Gooch's perforated crucible, 361
Gore's gas furnace, 95
Gozdorf's method of estimating, weight
of spheres of, 753
Graphite, 170, 258, 285
crucibles, 123
in iron and steel, estimation of, 369,
370
Green copper carbonate, 275
flames, 294
Grey copper, 274, 282, 283, 284
Greenearth, 267, 287
Griffin's oil furnace, 74
reverberatory gas furnace, 100
Grinding and powdering ores, 332
Gold, 275, 284
and palladium, rhodium, silver, and 1
mercury, 740
platinum, 799
alloys for pyrometry, 149
silver, melting by gas, 92
separating, 767
INDEX.
Iv
GOL
Gold assay of pyrites for, 773
weights for, 35
assays, apparatus for boiling, 763
coin, assay of, 740, 769
copper, alloys, assay of, 756
cupellation of, 749
estimating weight of minute spheres
of, 753
in copper ores, assay of, 775
minerals, detection of traces of, 777
leaf (false), assay of, 584
silver, platinum, and copper alloys,
759
standard of alloys of, 768
valuing tables, ii, xx, xxi, xxxiii
Gum, 174
Gun metal, assay of, 734
tin in, 549, 550
Guyard's process for assay of platinum
residues, 799
Gypsum, 223, 261, 290
HADOW'S process for separation of
cobalt and nickel, 839
Hallet's process for assay of tin, 548
Hammers, 15
Hard cement, 117
Hardness of gems, 291
iron and steel, 433
minerals, 240
Haswell's process for titration of iron,
330
Hausmannite, 833
Heat, production and application of, 55
Heating power of fuel, estimation of,
154
Heavy spar, 262, 290
Heine's colorimetric copper assay, 461
Hematite, 266
red and brown, 308
Herpin's electrolytic assay of copper
and nickel, 853
Hessian crucibles, 120
Holland's assay of sulphur, 865
Hood, 57
Hornblende, 252, 289, 290
Horn silver, 277, 289
Hornstone, 250
Houzeau's assay of sulphur, 864
Hyacinth, 254, 884
Hydrogen gas, 169
reduction by, 48
Hyposulphite of sodium for titration of
iron, 330
TNCINERATING precipitates, 41
JL Ingot mould, 69
Injector gas furnace, 92
Inquartation, 761
Iodide of silver, 230
Iodine, coloured flame of, 296
tincture of, 229
KUN
Iridium, 782, 785
and platinum, 797
Iron, 182, 284
arseniate, 287
assay of, 308
assay of lead by fusion with, 511
and silver, assay of, 737
steel, assay of chromium in, 827
estimation of silicon in, 407,
413
hardness of, 433
estimation of sulphur in, 388
carbonate, 268, 289
cement, 119
chromic, 285, 287
crucibles, malleable, 133
dry, assay of, 309
estimation of carbon, sulphur, sili-
con, phosphorus, &c., in, 363
in ores, estimation of, 338
magnetic, 266, 285
manganese in, 422
manufactured, basic cinder and oxide
in, 414
metallic, as a reducing agent, 175
mortar, 17
native, 265
new colorimetrical process for esti-
mating sulphur in, 393
on galena, action of, 502
ores, estimation of sulphur in, 338,
405
list of, 308
manganese in, 426
phosphoric acid in, 334
oxides, 201
oxide, brown, 285
pyrites, assay of, for sulphur, 859
peroxide, 181
phosphate, 287
phosphorus in, 416
pyrites, 190, 265, 280
magnetic, 278, 28?-
retorts, 53
specular, 285
sulphate, 181, 288
sulphide, 183
titanic, 285
titanium in, 430, 432
wire, 213
TACQUELAIN'S colorimetric assay of
tl copper, 472
Jargon, 879
Jasper, 250
Juptner's (von) method of separating
gold and silver, 767
TTIMBERLITE, 870
JV. Klaproth's process for assay of tin,
546
Kunzel's process for assay of zinc, 577
Ivi
INDEX.
LAD
T ABLE, wrought iron, 69
JJ Lake Superior copper assay, 455
Laws of combination, 1, 5
Lea's process for assay of platinum
ores, 790
Le Blay's colorimetric assay of copper,
467
Lead alloys, assay of, 530
and antimony, alloys of, assay of,
565
sulphides, 277, 282
assay of, 502
by fusion with black flux, 510
iron, 511
potassium carbonate,
504
sodium carbonate, or
black flux and metallic iron, 514
with standard solutions, 531
with sulphuric acid, 519
remarks on, 517
borate, 177, 201
carbonate, 271
assay of, 523
colour of flame, 295
cupellation of silver in, 730
fumes, assay of, 525
in iron ores, 356
metallic, as a reducing agent, 176
nitrate,
nitrate, 17, 188
ores, 502
phosphate, 271
proof, 224
roasting and reducing, assay of,
515
selenide, 277
silicate, 201
silicates, 177
slags, assay of, 525
speiss, assay of, 855
sulphate, 181, 188, 201, 271
assay of, 527
sulphide, 187, 271, 284
to be employed in cupellation of
gold and silver, 759
Lens pocket, 622
Lenssen's process for volumetric assay
of tin, 551
Lersen's process for separation of zinc
from copper, 584
Level's assay of lead with potassium
^ferrocyanide and cyanide, 519
Liebig's condenser, 52
process for separating nickel and
cobalt, 846
Lime and egg lute, 114
and its silicate, 192
coloured flame of, 297
crucibles, 128
in iron ores, 341
Lined crucibles, 127
Linseed meal, 114
MET
Litharge, 176, 182, 201
action of, on sulphides, 183
assay of, for silver, 608
assay of, 523
silver assay with, 602
Lithia, coloured flame of, 296
Litmus paper, 213
London crucibles, 120
Lowe, Mr., apparatus for boiling gold
assays, 763
Luckow's electrolytic assay of copper,
487
Lustre of minerals, 239
Lutes and cements, 113
Luting vessels for furnace operations,
118
Lyte's assay of lead, 529
ll/TACKINTOSH on estimating phos-
ITJ_ phorus in iron and steel, 417
Magnesia crucibles and bricks, 130
in iron ores, 341
Magnesite, 260,289
Magnesium silicate, 192
Magnetic iron, 266, 285
pyrites, 279, 282
ore, 308
Malleable iron crucibles, 133
Manganese, assay of, 833
in iron, 422
ores, 317,341
speigleisen, 426
ores, 268, 285, 287
assay of, 833
peroxide, 180
sulphide, 183
Manganite, 833
Markus on lead assay, 517
Mascazzini's assay of lead, 529
Matrix of diamond, 868
Measuring flasks, 307
the heat of a furnace, 144
Mechanical treatment of iron ores, 332
Meerschaum, 251
Melting arrangement (gas) for gold
and silver, 92
points of metals, 147
Mercury and gold, 740
assay of, 587
argentiferous, 737
silver alloys containing, 686
blowpipe, assay of, 598
electrolytic, assay of, 592
native, 272
selenide, 277
sulphide, 272
titration of silver in presence of,
714
volumetric assay of, 595
Merrick's assay of pyrites for gold, 773
Metallic fluxes, 201
INDEX.
Ivii
MET
Metallic iron as a reducing agent, 175
lead as a reducing agent, 176
Metals, melting points of, 147
Method of weighing, 36
Mica, 251, 255, 286, 290
Micacsous iron, 266
Microcosmic salt, 219
Millerite, 277
Minerals, determination of, 279
discrimination of, 237
Minium, assay of, 523
Mint value, to convert, into bank value,
xxxiii
Mispickel, 265, 283
Mohr's burette, 302
process for assay of manganese ores,
835
- zinc, 582
volumetric assay of copper, 476
Moissenet on the English copper assay,
435
process for assay of tin, 548
Molybdenite, 263, 282, 286
Monger's process for assay of cupri-
ferous blende, 573
Mortars, 17
Mould, ingot, 69
Mounts for crucibles, 107
Muffle gas furnace, 81, 91
or cupel furnace, 62
Muir's test for bismuth, 818
Mundic, 265
\TEPHRITE, 251
ll Nickel and cobalt, separating, 839,
846
arsenical, 268, 284
assay for traces of cobalt in, 849
of, 839
commercial assay of, 851
metallic assay of, 845
crucibles, 140
glance, assay of, 851
in iron ores, 344
ores, 839, 845
oxalate, 221
pyrites, assay of, 851
- sulphide, 277, 283
white, 278, 284
zinc, and copper alloys, assay of,
853
Nipples, blowpipe, 204
Nitrate of cobalt, 221
lead, 180, 188
potassium, 220
sodium, 180
Nitrates of potassium and sodium, 177
Nitre, 188, 196, 220, 261, 288
oxidising power of, 607
Nitric acid, 213
Nomenclature, chemical, 1
Normal solutions, 300
PER
OCTAHEDRON, 237
Ohl's assay of nickel speiss, 854
Oil and gas furnaces, 74
Oil of cassia, testing gems in, 873
Oils, fatty, 172
Olassen's assay of zinc, cobalt, and
nickel, 850
Olivine, 252, 869
O'Neill's process for assay of chrome
iron ore, 822
Onyx, 250
Opal, 250, 291
Ores of antimony, 556
arsenic, 831
gold, 740
lead, 502
manganese, 833
mercury, 587
platinum, 781
silver, 600
principal of copper, 434
Oriental topaz, 256
amethyst, 256
emerald, 256
ruby, 256
Osmiridium, 795, 796, 803
Oudemans's process for titration of
iron, 330
Oxalate of nickel, 221
Oxalic acid, 174
Oxidation before blowpipe, 209
Oxide of copper, 181, 221
red and black, 287
iron, brown and red, 287
iron in manufactured iron, 414
- tin, 537
, red cuprous, 285
Oxides, 3, 4
of copper and iron, 201
Oxidised ores of copper, 434
Oxidising agents, 176
Oxychloride of zinc cement, 120
Oxygen on galena, action of, 502
PALLADIUM, 782, 798
JL and gold, 740
titration of silver in presence of,
714
Parnell's process for assay of arsenic,
832
Parting assay, 771
of gold and silver, 761
Paste, adhesive, 115
Pattinson's process for assay of man-
ganese ores, 836
Paull's assay of manganese ores, 835
Pearson's assay of copper pyrites, 494
process for assay of sulphur, 861
volumetric assay of bismuth, 817
Penny's process for assay of iron in wet
way, 321
Peridot, 891
Iviii
INDEX.
PER
Peridotite, 871
Perofskite, 869
Peroxide of iron, 181
Perry's process for assay of platinum
ores, 807
Personne's process for volumetric assay
of mercury, 595
Pestle and mortar, 17
Pestles, c., mending, 115
Peters on the Lake Superior copper
assay, 455
Petroleum furnace, 94
Phosphate of copper, 434
iron, 287
lead, 271
sodium and ammonium, 219
Phosphoric acid, coloured flame of, 296
estimation of, in iron ores, 334
Phosphorus in iron, 317
_ and steel, 363, 416
Pinchbeck, assay of, 584
Pipette, 306
Pitchblende, 278, 286, 287
Plaster of Paris, 114
Platinum, 275, 284
and gold alloys for pyrometry, 149
iridium, 797
silver, assay of alloys of, 632
assay of, 781
crucibles, 133
preservation of, 134
forceps, 50
ores, analysis of, 782, 790
silver and copper alloys, assay of, 633
gold, and copper alloys, 759
spoons, 212
wire for blowpipe, 211
Plattner's detection of nickel before
blowpipe, 849
Pleonast, 257
Pliers, 622
Plumbago crucibles, 123
Pokers, 67
Polybasite, 600
Porcelain crucibles, 122
mortar, 17
Porosity of fuel, 152
Potash, 181
Potassium binoxalate, 199
bisulphate, 221
bitartrate, 199
carbonate, 181, 195
coloured flame of, 297
cyanide, 214
nitrate, 177, 188, 196, 220
on galena, action of, 503
Pound assay,' 694
Precipitates, incinerating, 41
weighing moist, 39
Price on a source of error in sulphur
estimations, 860
Prism, 238
Proof lead, 224
ROS
Prospecting for gold, &c., 26
Psilomelane, 833
Pulverisation, 18
j Pyrites, arsenical, 265, 831
assay of, for gold, 773
sulphur, 859
copper, 273
assay of, 494, 496
iron, 190, 289
magnetic iron, 278, 282
Pyrolusite, 833
Pyrometers, 144
Pyrometric heating power of fuel, 154
power of fuel, 160
Pyromorphite, 271, 288, 289
Pyrope, 869
AUARTATION, 761
U Quartz, 291, 292, 874
Quartz and silicates, 248
-violet, 890
yellow, 884
Quicksilver, 272
T)AMMELSBERG'S process for assay
XL of tin ores, 554
Raw flux, 197
Realgar, 831
Red copper, Malachite, 434
cuprous oxide, 274
flames, 295
ochre, 266
Reduction, 46
- before blowpipe, 209, 216
of iron salts by stannous chloride,
327
sulphite of ammonium,
328
zinc, 327
Reducing agents, 169
power of various agents, 175
fluxes, 200
Regulus of antimony, assay of, 558
Resins, 172
Resinous or hard cement, 117
Retorts, 53
Reverberatory furnace gas, 90, 100
Rhodium, 782, 785, 799
and gold, 740
Rhombohedron, 238
Riley's manganese in speigleisen, 424
process for estimating titanium in
iron, 430
Roasting, 44
gold ores, 745
Roasting-test, 44
Rock crystal, 249
salt, 261
Roman cement, 114
Rose quartz, 250
Rose's method of separating gold and
silver, 767
INDEX.
lix
ROS
Boss's aluminium support, 212
Kubicelle, 257
Kuby, 293
silver, 600
spinel, 886
Rumpf's and Scherer's process for
assay of manganese ores, 834
Ruthenium, 782, 785
Rutile, 278, 287
SALAMANDER ' brand of crucibles,
121
Salt, 196
- common, 288
normal solution of, temperature cor-
rection, 651
of sorrel, 199
standard solution of, 636, 641
Salts, 3, 4
Saltpetre, 177, 188
Sample, preparation of the, 8
Sampling machine, 11
automatic, 11
steel and iron borings, 364
Sapphire, 256, 293
white, 876
yellow, 879
red, 886
- blue, 889
violet, 890
green, 891
water, 889
Sard, 250
Sardonyx, 250
Scale of hardness of minerals, 240
Schaffner's process for assay of zinc,
577
Scherer and Rumpf 's process for assay
of manganese ores, 834
Schwarz's process for assay of zinc, 581
Schwartz's volumetric assay of lead, 532
Scorification, 54, 610
assay of gold, 747
Scorifier, assay in, 615
Scorifiers, 144
Scully, effect of bismuth on ductility of
silver, 699
Sef Strom's blast furnace, 61
Selenides of mercury, lead, silver, and
copper, 277
Selenite, 261
Selenium, 280
coloured name, 295
Sell's process for volumetric assay of
chromium, 829
Serpentine, 251, 290
Sexton on arsenic in copper, 499
Shears, 17
Shimer's process for assay of metallic
iron and steel, 363
Siemens's pyrometer, 148
Sieve, 20
SIZ
Sifting, 22
Silica, 192, 222
in iron ores, 341
Silicate of copper, 434
lead, 177, 201
zinc, 279, 289
assay of, 572
Silicates of aluminium, 291
calcium, magnesium, aluminium,
192
Silicon, estimation of, in iron and steel,
407, 413
in iron and steel, 363
Silver, 276, 284
a blowpipe assay of, 715
alloys containing mercury, assay of,
686
and copper, assay of, 739
alloys, assay of, 627, 632
gold separating, 767
iron, assay of, 737
mercury, assay, of, 737
platinum, assay of alloys of, 632
antimonial, assay of, 735
assay of, 600
- by the wet way, 634, 647, 674
litharge for, 608
weights for, 35
with litharge, 602
bullion, assay of, 631
chloride, 77, 224,
reduction of, 684
copper alloys, cupellation of, 732
crucibles, 140
effect of bismuth on ductility of, 699 1
from galena, separation of, 627
glance, 600
gold, platinum, and copper alloys,,
759
horn, 289
in lead or bismuth, cupellation of,.
730
iodide, 231
lead, concentration of, 717
native, assay of, 633
ore, antimonial, 284
arsenical, 284
ores, arsenical and antimonial, 276
or antimonial, 282
ore, arsenical and antimonial, 286
ores, 600
platinum, and copper alloys, assay
of, 633
preparation of pure, 685
selenide, 277
separation of, from base metals, 633
sulphide, 276, 283
telluric, 735
titration of, in presence of, 714
copper, 713
with ammonium sulphocyanide,.
711
Size and shape of fuel, 152
Ix
INDEX.
SKE
Skey's method of detecting traces of
gold in minerals, 777
Skittle pots, 120
Slag, iron, 310
Smaltine, 269, 284
Smaragdite, 869
Smith, J. Lawrence, on phosphorus in
iron and steel, 419
Smithsonite, 270
Smoky quartz, 250
Soap, 199
Soda, 181
coloured flame of, 296
- paper (Forbes's), 233
Sodium ammonio-phosphate, 219
biborate, 218
bisulphate, 220
carbonate, 181, 195, 214
chloride, 196
coloured flame of, 297
nitrate, 177, 180
sulphate, 181
Soft cement, 114
Solution, 49
Sonstadt's solution, 242
for testing gems, 873
Spathic iron ore, 308
Specific gravity of fuel, 153
iron ores, 360
minerals, 241
heating power of fuel, 160
Spectacles, neutral tint, 68
Spectroscopic assay of gold, 779
Specular iron, 266, 285
Spiegeleisen, manganese in, 424, 426
Speiskobalt, 831
Sphene, 278, 291
Spheres of gold, estimating weight of,
753
Spinel, 256, 292, 294
ruby, 886
Spoon platinum, 212
Srubescite, 434
Standard colours, day, 382
night, 386
of alloys of gold, 768
solution of silver, 640
solutions, 302, 300
Standards, inorganic, for colorimetric
carbon test, 381
Stannous chloride, reduction of iron
salts by, 327
Starch, 173
: Stead on the estimation of minute quan-
tities of carbon in iron and steel,
376
Steel and iron, estimation of silicon in,
407, 413
argentiferous assay of, 737
estimation of carbon, sulphur, sili-
con, phosphorus, &c., in, 363
sulphur in, 388
hardness of, 433
SUT
Steel mortar, 19, 332
phosphorus in, 416
Steinbeck's electrolytic assay of copper,
480
Stromeyer's process for volumetric assay
of tin, 551, 554
Streak of minerals, 239
Stones, precious discrimination of, 867
Storer's assay of galena, 521
Stourbridge clay crucibles, 122
Strontia, coloured flame of, 296
Sublimates, detection of antimony in,
563
Sublimation, 54
Sugar, 173
Sulphate of calcium, 223
lead, 188, 201, 271
sodium, 181
Sulphates of lead, copper, and iron,
181
zinc, iron, and copper, 288
soluble, 261
Sulphide of antimony, 189, 263, 284
assay of, 557
arsenic, 264, 286
copper, argentiferous, 600
lead, 271
mercury, 272
nickel, 277, 283
silver, 276
zinc, 285
assay of, 573
Sulphides, action of litharge on, 183
alkaline, 190
lead and antimony, 282
of lead and antimony, 277
Sulphite of ammonium for reduction of
iron salts, 328
Sulphocyanide of ammonium, titration
of silver with, 711
Sulphur, 189, 262, 286
arsenic, cobalt, and nickel assay of,
851
assay of, 859
in wet way, 861
in fuel, estimation of, 162
iron, 317
and steel, 363
estimation of, 388
new colorimetrical process for
estimating, 393
ores, estimation of, 338, 405
ores, 859
, separation of, from bismuth, 816
Sulphuric acid, assay of lead with, 519
in iron ores, 348
Sulphurising agents, 189
Sulphuretted ores of copper, 434
Sulphurous earth, 859
Supporting crucibles in gas furnace, 80
Supports, blowpipe, 210
Button's process for assay of antimony,
563
INDEX.
SYM
Symbols, chemical, their employment
and uses, 7
STABLE, assay for gold, xxxiv
JL of cupellation loss, 728
the corrections of the standard
solutions of common salt, 654
weights of colourless stones in
air and water, 878
yellow stones in air and
water, 883
brown, 885
- red, 887
- blue, 890
violet, 891
green, 893
chatoyant, 895
for conversion of mint
value into bank value, xxxiii
Tables of quantity of fine gold in alloys
and the mint value, ii
Talbott's process for assay of tin in
presence of tungsten, 546
Talc, 250, 290
Tallow, 172
Tamm's process for assay of bismuth
ores, 809
Tantin's process for estimating phos-
phorus in iron and steel, 416
Tartaric acid, 174
Telluric silver, assay of, 735
Tellurium, coloured flame of, 296
graphic and foliated, assay of, 740
Temperature correction for normal
solution of salt, 651
Tennantite, 831
Terreil's process for separating nickel
and cobalt, 846
Testing minerals, requirements for,
248
Tetrahedron, 237
Thompson's, Dr., scheme for the deter-
mination of minerals, 279
Thurach's process for purify ing bis-
muth, 816
Tin, assay of, 537
argentiferous, 734
by wet way, 546
volumetrically, 550
- foil, 224
from antimony and arsenic, separa-
tion of, 564
in gun and bell metal, assay of, 549,
550
ore, 262, 285, 291
assay of, by fusion, 553
reduction with hydro-
gen, 552
ores, 287
containing arsenic, sulphur, and
tungsten, 543
oxide, 537
VOL
Tin oxide, assay of, 539
mixed with silica of, 542
slags, assay of, 553
Tinstone, assay of, 539
Tin-white cobalt, 269
Titanic acid in iron ores, 358
- iron, 267, 285
Titaniferous iron ore, 308
Titanium in iron, 430, 432
ores, 317
Titration of iron with potassium bi-
chromate, 321
sodium hyposulphite, 330,,
silver with ammonium sulpho-
cyanide, 711
Tongs, 67, 68
Topaz, 253, 292, 293
false, 250
- white, 876
yellow, 880
red, 887
blue 888
Tosh's process for the estimation of
graphite in iron and steel, 370
Touch-needles, 757
Touchstone, 757
Tourmaline, 252, 292
yellow, 880
brown. 886
- red, 887
blue, 889
violet, 890
green, 892
Turner's flux, 222
Turmeric paper, 213
Turner's process for estimating the
hardness of iron and steel, 433
estimation of silicon in iron
and steel, 413
the detection of boracic acid,
295
Turquoise, 889
Type metal, assay of, 565
TTNIVERSAL furnace, 65
U gas furnace, 84
Ure's calorimeter, 158
method, for assay of fuel, 157
TTANADATE of copper, 434
V Vanning, washing, or dressing,
23
Varvacite, 833
Vegetation of silver, 620
Vermeil garnet, 884
Vitreous copper, 273
282, 283
Volatile products of carbonisation of
fuels, 60
Volhard, titration of silver with am-
monium sulphocyanide, 711
Ixii
INDEX.
VOL
Volumetric analysis, 298
assay of bismuth, 817
copper, 476
chromium, 829, 827
iron, 321
mercury, 595
silver, 634, 647, 674
- tin, 550
zinc, 575
Von Hubert's colorimetric assay of
copper, 472
WAEKEN'S process for assay of
cobalt speiss, 843
Washing, dressing, or vanning, 23
Water, distilled, 52
in fuel, estimation of, 152
iron ores, 351
minerals, 296
Waterproof cement, 116
Wax, yellow, 114
Wedgwood's pyrometer, 144
Weighing, 26, 36
moist precipitates, 39
Weight of minute silver globules, 723
spheres of gold, 753
Weights, 34
assay for silver and gold, 35
Wet assay of iron, 321
process for assay of silver, 634, 647,
674
_ l Z inc, 571, 572, 573
White flux, 197
lead, 176
and oil lute, 114
nickel, 278, 284
Wiborgh's process for estimating sul-
Shur in iron, 393
lis's lute for crucibles, 119, 124
Wilson's pyrometer, 147
Wind furnace, 57
Winkler's process for separating tin
from antimony and arsenic, 564
ZIR
Wohler.'s method of decomposing
osmiridium, 803
Wolfram, 278, 287
Wolfsbergite, 434
Wood charcoal, 171
Wood tin, 537
Wright's calorimeter, 159
process for assay of sulphur, 861
TTELLOW flame, 295
JL ochre, 267
wax, 114
7EOLITES, 255, 289
/J Zinc, 213
alloys, assay of, 574
argentiferous assay of, 735
assay of, 567
- by the wet process, 571, 572.
573
carbonate, 270
cobalt, and nickel, assay of, 850
coloured flame of, 296
copper and nickel alloys, assay of,
853
distillation of, 84, 568
from copper, separation of, 584
in iron ores,. 344
ores, 567
oxy chloride of cement, 120
reduction of iron salts by, 327
silicate, 270, 289
assay of, 572
sulphate, 288
sulphide, 187, 285
assay of, 573
volumetric assay of, 575
Zircon, 254, 292, 293
white, 875
yellow, 879
brown, 884
fpE
(fmVEBSr
'^
PRINTED BY
SPOTTISWOODB AND CO., NEW-STREET SQUARE
LONDON
UNIVERSITY OF CALIFORNIA LIBRARY
BERKELEY
Return to desk from which borrowed.
This book is DUE on the last date stamped below.
MAY 20 1948
LD 21-100m-9,'47(A5702sl6)476
YC 68390
UNIVERSITY OF CALIFORNIA LIBRARY