GUIDE 
 
 TO A COURSE OF 
 
 especially of 
 
 MINEEALS AND FURNACE-PRODUCTS, 
 
 ILLUSTRATED BY EXAMPLES, 
 
 by 
 
 r c 
 
 - f f < 
 
 TUTION, MEMBER OF THE ACADEMY OF SCIENCE, BERLIN, &C. 
 
 TRANSLATED BY 
 
 J. TOWLER, M. D. 
 
 PROFE380K OF CIVIL BKGIKEEBING, CHEMISTRY, TOXICOLOGY, &C, HOBABT COLLB4B 
 AND GENEVA MEDICAL COLLEGE. 
 
 GENEVA, N.,Y. 
 
Entered according to Act of Congress, in the year 1871, by 
 J. TOWLER, in the Office of the Librarian of Congress, at 
 
 WASHINGTON. 
 
TO 
 
 |l fits iisi 
 
 WHO, BY THEIR UNTIRING PERSEVERANCE 
 
 IN 
 
 SETTING AND DISTKIBUTHSTG TYPE, 
 
 HAVE 80 ESSENTIALLY CONTRIBUTED TO ITS PUBLICATION, 
 
 THIS TRANSLATION 
 
 IS WITH AFFECTION 
 
 i 
 
 BY 
 
 THE TKANSLATOK, 
 
PREFACE. 
 
 The following work has been translated and is now pub- 
 lished expressly for those students in chemistry, whose time 
 and other studies in colleges do not permit them to enter 
 upon the more elaborate and expensive treatises of Freseni- 
 us and other authors. It is the condensed labor of a master 
 in chemistry and of a practical analyst. In order still more 
 to condense the translation and to diminish the expense as 
 much as possible, the entire part on ASSAYING, as well as the 
 chapter on the ANALYSIS of commercial drugs and metallur- 
 gical products, has been omitted ; the former, because that 
 does not properly belong to chemical analysis ; and the lat- 
 ter, because the reader is generally referred to the preceding- 
 pages for a description of the analysis of the substances un- 
 der consideration. 
 
 The translator has made no other changes in this work ; 
 even the old chemical equivalents have been allowed to re- 
 main unaltered, from the fact, that the results of computa 
 tions in all analytical investigations are precisely the same 
 whether the old or the new values be assigned to the diffe 
 rent elements respectively. 
 
 The only apology, that is deemed necessary to be made 
 in reference to the present publication, is for the want of ex- 
 perience and polish displayed in the typographical execu 
 tion ; the translator is as yet but a novice in the Hack art ; 
 but perfection is the aim of his aspirations, from which cir- 
 cumstance hopes may be entertained for something better 
 in the future. 
 
 J. TOWLEK. 
 
 Hobart College, 
 May 22, 1871. 
 
CONTENTS. 
 
 Page 
 
 I. Introduction 1 
 
 EL General Manipulations in Analytical Operations 10 
 
 III. Vessels of Platinum and Silver , 36 
 
 IV. Balance and Weight 37 
 
 V. Analysis by Measure 40 
 
 Analysis of separate Classes of .Compounds, 
 
 I. Metallic Alloys 47 
 
 A. Alloys soluble in NitricAcid 49 
 
 1. Silver and Copper 49 
 
 2. Copper and Zinc 51 
 
 3. Copper, Zinc and Nickel 54 
 
 4. Lead and Zinc 55 
 
 5. Copper and Bismuth 55 
 
 6. Lead and Bismuth 56 
 
 7. Copper, Lead and Zinc 57 
 
 B. Alloys not completely soluble in Nitric Acid 57 
 
 1. Tin and Copper 57 
 
 2. TinandLead 5 
 
 3. Tin, Lead and Zinc 5J 
 
 4. Tin, Copper, Lead and Zinc 59 
 
 5. Tin, Lead and Bismuth 50 
 
 6. Tin, Lead, Bismuth and Cadmium 60 
 
 7. Tin, Lead, Bismuth and Mercury 61 
 
 8. Antimony and Lead do 
 
 9. Antimony and Tin 63 
 
 O. Metallic Oxides 64 
 
 1. Analysis of Iron 64 
 
 2. Brown Hematite, Limonite 65 
 
 3. Bog Iron ore 65 
 
 4. Magnetic Iron ore, Magnetite 68 
 
 5. Analysis of the Oxide of Manganese. . . .< 69 
 
 6. Psilomelane 73 
 
 7. Oxides of Lead 75 
 
 8. Tin Stone, Tin Ore 76 
 
 in. Sulphur Metals 78 
 
 A. Analysis of Sulphur Metals by Aqua Regia 79 
 
 1. Sulphur and Iron 83 
 
 2. Sulphur, Copper and Iron 83 
 
 3. Sulphur, Copper, Iron and Zinc 84 
 
 4. Sulphur, Zinc, Iron and Cadmium 85 
 
 5. Sulphur, Antimony and Iron , 86 
 
 B. Analysis of Metallic Sulphides by Chlorine 87 
 
 1. Sulphur, Antimony and Silver 91 
 
 2. Sulphur, Antimony and Lead '. do 
 
 3. Sulphur, Antimony, Lead, Copper and Iron 92 
 
 4. Sulphur, Antimony, Arsenic, Copper, Iron and Zinc 92 
 
 5. Sulphur, Lead, Iron, Copper, Silver, Zin<% Nickel and Antim >:iy 94 
 
 IV. Arsenical Metals and their Sulpliicl ;,s '. 95 
 
 1. Arsenic and Iron 95 
 
 2. Sulphur, Arsenic and Iron 98 
 
 3. Arsenic, Nickel, Cobalt, Iron, Copper and Antimony 99 
 
VIII 
 
 Page- 
 
 4. Sulphur, Arsenic, Antimony, Nickel, Cobalt and Iron 103 
 
 5. Lead and Copper Mixtures 104 
 
 6. Arsenic and Antimony 10ft 
 
 7. Arsenic and Tin 107 
 
 V. Sulphates t ' 108 
 
 1. Sulphate of Baryta containing Strontia 110 
 
 2. Sulphate of Strontia containing also Baryta and Lime Ill 
 
 3. Alum 113 
 
 4. Mansfeld Black Vitriol 115 
 
 VI. Phosphates 116- 
 
 1. Phosphate of Lime with Chloride and Fluoride of Calcium 117 
 
 2. Bone Black 119- 
 
 3. Acid Phosphate of Lime 120 
 
 4. Phosphate of the oxide of Lead, Chloride of Lead with Arsenic &c 121 
 
 5- Phosphate of the protoxide of Iron, and Water 122 
 
 6. Phosphate of the protoxide of Iron, protoxide of Manganese, Lithia, &c. 123. 
 
 7. Phosphoric Acid, Alumina, oxide of Iron, Magnesia 126 
 
 VI. Carbonates 128 
 
 1. Lime, Magnesia, Carbonic Acid 131 
 
 2. Protoxides of Iron and Manganese, Carbonic acid, Lime and Magnesia. . 134 
 
 3. Oxide of Lead, Carbonic Acid, Water 135 
 
 VII. Borates do 
 
 1. Lime, Soda, Borucic Acid and Water 137 
 
 2. Magnesia, Boracic Acid, Chlorine, Water 138 
 
 IX. Silicates do 
 
 A. Silicates decomposable by acids 145 
 
 1. Silicate of the oxides of Iron, sometimes with Manganese, Lime, &c 151 
 
 2. Silicate of Alumina, Lime, Soda, Potassa and Water 125 
 
 3. Silicate of the protoxides of Iron and Manganese, Alumina, Lime, &c. . . 165 
 
 4. Silicate of the protoxide of iron, Alumina, Lime, Magnesia, Lead, &c... 158 
 
 5. Silicate of the oxide of Zinc with Water 159 
 
 6. Silicate of Alumina and Soda, containing Sulphide of Sodium 160 
 
 B. Silicates not decomposable by Acids 162 
 
 A. Analysis of Silicates with Carbonate of Soda 164 
 
 1. Silicate of Lime and Magnesia 170 
 
 2. Silicates of Alumina, protoxides of Iron and Manganese, Lime, &c do 
 
 3. Silicate of Alumina and sesquioxide of Iron together with Water do- 
 
 4. Silicates containing Fluorine do 
 
 5. Silicates containing Boracic Acid 17a 
 
 B. Analysis of Silicates with Carbonate of Lime and Chloride of Ammonium. 174 
 
 C. Analysis of Silicates with Hydrofluoric Acid 175 
 
 1. Silicates containing no Magnesia 179 
 
 2. Silicates containing Magnesia do 
 
 B. Decomposition of Silicates by means of Sulphuric Acid 181 
 
 E. Determination of the Degree of Oxidation of Iron in Undecomposable Sil : 182 
 
 F. Analysis of Mixed Silicates 183 
 
 1. Basalt 186 
 
 X. Alumina and Aluminates 190 
 
 1. Aluminate of Magnesia and Protoxide of Iron 19 j 
 
 2. Chrome Iron Ore 192: 
 
 XI. Compounds of Chromium do 
 
 1. Chromate of Potassa do 
 
 2. Chromate of the Oxide of Lead 193 
 
 3. Chrome Iron Ore 194 
 
 XII. Compounds of Titanium 196 
 
 1. Titanic Iron do 
 
 XIII. Compounds of Fluorine 198 
 
 Cryolite ; ' 199 
 
 XIV. Stcechiomotrical Computation of the Analysis do- 
 
 XV. Tables for tin- Compulation of Analyses 224 
 
INTRODUCTION. 
 
 The analysis of every inorganic product, whether a mine- 
 ral, some artificial compound or a product of the furnace, 
 is two-fold : qualitative and quantitative. The qualitative a- 
 nalysis, which teaches how to find the constituents of the sub- 
 stance to be analyzed, must always precede the quantitative 
 analysis, not alone because it is necessary to know in the first 
 place what a body contains, before it can be thought of deter- 
 mining the quantity of each ingredient, but because by means 
 of the qualitative investigation we learn what solution or de- 
 composing material will be the most appropriate for the ana- 
 lysis of the body in question, a circumstance which is always 
 dependent on the nature of its constituents. Thus the qualita- 
 tive examination of a silicate, whether a mineral or a slag, tells 
 us whether it is soluble in acids or not, the qualitative analy- 
 sis of metallic sulphides can decide whether the investigation 
 shall commence by dissolving the substance preferably in ni- 
 tric acid or in aqua regia and so on. 
 
 Another advantage derived from the qualitative analysis 
 consists in enabling the chemist at the same time to pronounce 
 which are the principal constituents in any substance, since 
 those ingredients, which form the largest proportion of a com- 
 pound, will be the most prominent in the examination ; and 
 a moderately experienced look will soon recognize them as 
 such. This must be kept in view in the quantitative analysis ; 
 for it is no rare occurrence that one and the same substance 
 is separated with advantage from the rest in the compound 
 in quite a different way, according as it exists in the combi- 
 nation in a larger or smaller proportion. 
 
 The quantitative analysis of inorganic bodies, as regards the 
 1 
 
2 
 
 operation, is sometimes easy and sometimes more or less dif- 
 ficult. Even if, as is the case in the present work, which is 
 designed for elementary instruction, the more rare substances' 
 are totally excluded, the number of metallic and non-metallic 
 bodies which occur more frequently, is still large enough to de- 
 mand various modes of separation. We shall not be able in the 
 following pages to present all these single cases and illustrate 
 them, by examples, but we shall lay before our readers the 
 most important, that is, those which occur most frequently. 
 
 If the different methods of quantitative modes of separa- 
 tion be examined more minutely, it will be found that only 
 very few of them can lay claim to great accuracy. Among these 
 may be reckoned, for instance, the determination of sulphuric 
 acid by means of the salts of barium, of silver by hydrochloric 
 acid, of lime by oxalic acid, and the reverse. In many cases 
 a precipitate is not quite insoluble, or the presence of one body 
 determines another to be precipitated partially by a reagent, 
 by which, when taken alone, it is easily and completely dis- 
 solved. In this way, for example, magnesia is frequently de- 
 termined by phosphate of ammonia, although the double phos- 
 phate of magnesia and ammonia is a substance somewhat more 
 soluble in water. Alumina is slightly soluble in potassa, ox- 
 ide of iron is not ; but if a mixture of the two is boiled with 
 ever so large an excess of potassa, there remains always a 
 trace of alumina with the oxide of iron. 
 
 The first rules, therefore, to be observed by beginners in 
 the art of chemical analysis are accuracy and care. Much de- 
 pends also on skill in the mechanical operations ; the operator 
 must make himself quite at home in weighing, drying, dis- 
 solving, precipitating, filtering, evaporating <fec ; even the 
 smallest loss, which might arise by spurting, pouring over the 
 side and such like, must be avoided with the the utmost care ; 
 and if such a one has occurred, the work must be immediately 
 recommenced ; for an uncertain result is worse than none at 
 all, because, in order to arrive so far, time and labor have been 
 sacrificed in vain. No work must be hurried, because accura- 
 
3 
 
 cy must thereby suffer. On this account many precipitates 
 can be filtered only a long time after the precipitation, be- 
 cause the latter operation is slow. Precipitates, that are to 
 be raised to a red heat, must previously be thoroughly dried 
 in the air, otherwise a portion may be ejected from the cru- 
 cible when strongly heated. Evaporations have to be effect- 
 ed always by a gentle heat ; for if the fluid should be raised to 
 a boiling temperature, a portion of it would inevitably be lost 
 by spurting. 
 
 Another very important circumstance IB -the appropriate 
 choice of vessels in which the analytical operations have to be 
 conducted. A principal rule to be observed in quantitative a- 
 nalyses is the following : to avoid every unnecessary trans- 
 ference of a substance from one vessel into another, because 
 it is necessarily accompanied with a small loss. Every act of 
 pouring a fluid from a beaker glass into a dish, or vice versa, 
 diminishes the accuracy of the result ; besides this, the water 
 required to rince out the empty vessels increases the quanti- 
 ty of fluid, which must frequently be diminished by evapo- 
 ration ; thus it would be a great fault to filter a fluid contain- 
 ing copper, which is to be precipitated by potassa, into a beak- 
 er, knowing that that precipitation must be effected at a 
 boiling temperature, and that such operations can not well 
 be performed in beaker glasses, or to do the same with a flu- 
 id, in which an alkali has to be determined by evaporation, 
 for which purpose, as in the first case, an evaporating dish 
 must be employed. To be desirous of effecting the solu- 
 tion of a metallic sulphide by means of nitro-hydrochloric a- 
 cid in such a vessel,or the decomposition of a gelatinizing si- 
 licate in a retort, would be no less unsuitable, because in the 
 former a loss might ensue in consequence of the spurting of 
 fluid during the evolution of the gas, 'and in the latter case 
 there would be considerable difficulty in general to obtain the 
 last traces of the silicic acid deposited on the parietes of the 
 vessel. These few remarks will be sufficient to cause beginners 
 to exercise their minds as to the choice of proper vessels before 
 
they undertake any analytical operation, and always to aim to 
 make such a selection, that all subsequent operations may be 
 performed without the necessity of rectifying the course of 
 proceeding by a change of vessels. 
 
 The components of a substance are generally separated from 
 one another by adding a body which produces with one of the 
 components a combination, which, owing to its insolubility, 
 can easily be isolated from the rest. Precipitates, on this ac- 
 count, perform an important part in chemical analysis. Never- 
 theless the means of separating a body and of its determination 
 are not always the same. In many instances we are obliged to 
 determine the components of a substance in a form different 
 from that in which they exist in the substance itself, because 
 we either can not separate the components at all as such, or at 
 least not determine their weight in their isolated condition. 
 Every body must be determined in a form, in which it i^s least 
 liable to change from the influence of air and heat, conse- 
 quently in which its weight can be taken with the utmost ac- 
 curacy. 
 
 Thus sulphuric acid must be weighed in the form of sul- 
 phate of baryta, the protoxide of iron in the form of the sesqui- 
 oxide, and copper in that of the oxide of copper. This natural- 
 ly presupposes, that the constitution of the sulphate of baryta, 
 of the sesquioxide of iron and of the oxide of copper is accu- 
 rately known, which from the investigations of chemists is the 
 case. From this consideration it becomes necessary to make a 
 computation, the principles of which are laid down hereafter. 
 
 It is not however always feasible to separate a body immedi- 
 ately in the form in which its determination must be effected. 
 Supposing, for example, we had a combination of sulphuric 
 acid, oxide of copper and oxide of zinc to analyze, and know 
 that the oxide of copper is to be determined as such. If we 
 were, without any further consideration, to precipitate the 
 oxide from the solution of the substance, we should commit 
 a great error, because a part of the oxide of zinc would be 
 thrown down at the same time, although when taken alone 
 
the latter oxide is soluble in potassa. In this case the copper 
 must be separated by hydrosulphuric acid from the acidified 
 8olution. The precipitate, although pure sulphide of copper, 
 cannot again be employed for the determination of the amount 
 of copper, because, during the act of drying and weighing, a 
 part of the sulphide will become oxidized in the air, so that 
 what we are weighing, represents an uncertain mixture, whose 
 amount of copper can not be computed. The sulphide of cop- 
 per, therefore, must first be converted into the oxide of cop- 
 per; this latter substance may be dried and heated to a red heat 
 without undergoing decomposition ; its weight can be taken 
 with accuracy, and its unchangeable constitution presents to 
 us a means of determining its amount with great nicety. This 
 example shews sufficiently, that the form, in which one sub- 
 stance is separated from another, and that, in which it is de- 
 termined, is not always the same. 
 
 If we have to analyze an alloy of tin and lead, we shall de- 
 termine the tin as oxide, and the lead in the form of sul- 
 phate, because these are the combinations in which the two 
 metals are obtained in the most unchangeable form. In a com- 
 bination of carbonic acid with the protoxide of iron, protoxide 
 of manganese, lime and magnesia the bases are determined for 
 similar reasons in the form of sesquioxide of iron, the double 
 oxide and sesquioxide of manganese, the carbonate of lime and 
 phosphate of magnesia. 
 
 Every analytical investigation must be as simple as possible; 
 for as the number of operations increases, accuracy decreases. 
 On this account those analyses are the most accurate in which 
 the substance to be analyzed remains in one and the same ves- 
 sel, so that it depends almost solely on the weighing how 
 great the accuracy shall be. The analysis of those metallic ox- 
 ides which are reducible by hydrosulphuric acid, as for instance 
 of the oxides of copper, of zinc and of lead, yields therefore in 
 proportion very accurate results; and it is such simple methods 
 as these which chemists especially employed when their aim 
 was to obtain the atomic weights, on which all other compu- 
 
6 
 
 tations are founded. Unfortunately, however, but very tew 
 analyses can be performed in so simple a manner, most mi- 
 nerals and products of the furnace, having a much more com- 
 plicated constitution, and thus requiring several different op- 
 erations, which, notwithstanding all conscientiousness in the 
 operator, are in proportion more prejudicial to the accuracy 
 of the iinal results. 
 
 As soon as an analysis is ended, all the constituents of the 
 body under examination determined, and from the form in 
 which they were determined, their respective amounts in the 
 form in which they existed in the original compound compu- 
 ted, it will scarcely ever happen that the sum of the separate 
 quantities obtained will be equal to the weight of the body 
 analyzed, or, as is customary for the sake of comparison, to 
 compute results in ..the parts of 100, this number being the 
 numerical value of the substance under examination, the sum 
 of the constituents found will scarcely ever amount to 100. 
 It is only now and then that such an accident may happen. 
 On the contrary either a loss or OM excess will shew itself. 
 Such a result will take place with the work of the most ex- 
 perienced analyst, and it is unavoidable, since all empiric- 
 operations are more or less imperfect. 
 
 The amount of loss in an analysis is in itself no certain criteri- 
 on as to the accuracy of the work; for it can not be maintained, 
 for example, that an experienced operator ought never to ex- 
 ceed one per cent loss in his analysis. The judgment on the con- 
 trary is formed from the nature of the constituents to be se- 
 parated and from their number. If the former are easily de- 
 termined, that is, if they can be definitely separated from 
 each other, and weighed in an unchangeable form, and if their 
 number is only small, the loss with good work may not a- 
 mount to more than that of one per cent. On the contra- 
 ry a carefully performed analysis of a body, consisting of six 
 or more constituents, and amongst these such as can not be 
 accurately separated, the loss may amount to one, one and a 
 half to two per cent, and yet in proportion be just as accurate 
 
I 
 
 as the other where the loss was the half of one per cent. A 
 more intimate consideration .of the nature of the substances to 
 be separated and of their number is, therefore, the only crite- 
 rion as to the accuracy of the operation, and not the magni- 
 tude of the loss alone. . . 
 
 If, notwithstanding great care in the analysis of a substance 
 whose constituents are few in number and such as are easily 
 determined, a loss of several per cent is manifest at the end 
 of the operation, this indicates a constituent not discovered in 
 the qualitative investigation, such, for instance, as water, an 
 alkaline compound, fluorine, ammonia &c. 
 
 It is not a rare occurrence to meet with an excess of weight, 
 which is caused with beginners very generally by imperfect 
 washing of the precipitates. Frequently it arises from the 
 fact, that, instead of an oxide, a basic salt was thrown down, 
 whose amount of acid was also weighed. Take the case where 
 alumina is prcipitated by ammonia, the precipitate frequent- 
 ly contains potassa, if the latter is present in the fluid, which 
 is no rare ; occurrence. Sometimes, however, the excess in 
 weight is only apparent, since metals, for instance, are deter- 
 mined and computed in the form' of a higher grade of oxida- 
 tion than that found in the substance examined, in which case 
 oxygen is the cause of the excess in question; and this frequent- 
 ly gives the first indication of the lower grade of oxidation. 
 
 We have remarked above, that very frequently a body is 
 determined in form of a combination of this selfsame body 
 with another, and that its amount is computed from the a- 
 mount of this combinaton according to its well known com- 
 position. This composition is not at all indifferent, and has a 
 remarkable influence on the accuracy of the result. The 
 smaller the quantity of the material sought in the combina- 
 tion with one or more other substances obtained by the ana- 
 lysis, the smaller will be each error in the determination, or 
 in fine the greater the accuracy, and vice versa. 
 
 An example will illustrate this. : Suppose we had to deter- 
 mine sulphur in a combination, and that we receive it, by 
 
8 
 
 reason of the analytical method adopted, in the form of sul- 
 phate of baryta, that is, in combination with barium and oxy- 
 gen, so that, in order to find its amount, we must refer to the 
 tables to ascertain how much sulphur is contained in 100 parts 
 of sulphate of baryta. We find that 100 parts of sulphate 
 of baryta contain 13.73 parts of sulphur. Here then the quan- 
 tity of the body to be determined is only one-tenth of the 
 combination in which we separate and weigh it. If now we 
 had made an error in the determination of the sulphate of 
 baryta" (and no work, even the most accurate, is absolutely 
 accurate), and for instance had received one per cent less than 
 we ought to have received, this error will be found, it is true, 
 in the amount of sulphur computed therefrom, but very 
 much diminished. If the amount of sulphate of baryta receiv- 
 ed were 0.99 instead of 1.00, the amount of sulphur by com- 
 putation would be 0.1359, whilst it ought to be 0.1373. The 
 error in the sulphate of baryta is 0.01 = one centigramme, 
 whilst that in the sulphur = 0.0014, only about one milli- 
 gramme. In other words : the error in the determination of 
 the sulphate of baryta is diminished to about one-tenth of its 
 absolute amount in the weight of the sulphur, or, in general 
 terms : the error in the determination of a body in the form 
 of a combination is diminished for this body in the ratio of 
 the weight of the compound to that of the body itself. 
 
 If on the contrary the quantity of sesquioxide of iron in 
 a fluid is determined by first reducing it by means of hy- 
 drosulphuric acid to the protoxide form and then computing 
 from the amount of precipitated sulphur the quantity of ses- 
 quioxide of iron, the method is less favorable to accuracy. 
 
 For since for every atom of sesquioxide of iron (=80 parts), 
 which is reduced to the protoxide, one atom of sulphur, that 
 is, only 16 parts are received, or the amount of the body to 
 be computed is five times greater than that of the one from 
 whose quantity it is determined ; thus it will easily be seen, 
 that if, for instance, we should receive 0.01(one centigramme) 
 of sulphur too little, the sesquioxide of iron would be compu- 
 
9 
 
 ted by 0.05 (five centigrammes) too little. In such a case the 
 error is multiplied as many times as the weight of the body 
 found is contained in the amount corresponding to it of the 
 body to be computed. 
 
 Can the amount of the oxygen or of the hydrogen in a combination be determined re- 
 latively with more accuracy if we separate and weigh each of these two substances in 
 the form of water? 
 
 It sometimes happens that some one or more components 
 in an analysis are not directly separated and determined by 
 absolute weighing. There are substances whose accurate de- 
 termination is so difficult, that it is better to determine the 
 remaining bodies in combination with them, and to deduct 
 their sum from the total weight of the body, in which case 
 the difference of the quantities will represent the substance 
 in question. In this way the quantity of water in many salts, 
 as well as acids, such as carbonic acid, nitric acid, phosphoric 
 acid and others, is determined from the amount of loss, but 
 of course only when the remaining bodies can be determined 
 with the greatest exactitude, because, as is quite evident, all 
 errors in the determination of these components must affect 
 the one determined indirectly. 
 
 As soon as the analysis of a body is ended, it frequently 
 becomes important to test the separate substances qualitatively 
 as to their purity. This is necessary, partly in order to assure 
 one's self that the separation of the substances has been ef- 
 fected with sufficient accuracy, that one ingredient does not 
 contain a portion of another and vice versa, in which case the 
 correct sum-total of the weights of the same would not give a 
 correct determination ; partly in order to hunt for other sub- 
 stances in minute quantities, which may have been overlook- 
 ed in the qualitative analysis, especially when but a small 
 quantity of the given material has been employed in the in- 
 vestigation. These tests naturally presuppose much caution 
 and knowledge, since it is necessary to know for each indivi- 
 dual case what materials, during the progress of the analysis, 
 may possibly be contained in the bodies separated. The blow- 
 pipe on such occasions, especially with metallic substances, 
 2 
 
10 
 
 will be found of great service. 
 
 If the body submitted to examination was a chemical com- 
 bination, and not a mixture, the result will correspond with 
 the laws of chemical equivalents, which present an important 
 control as to the correctness of an analysis. It is our inten- 
 tion to lay before our readers in full detail in one of the 
 last articles the method of computing stcechiometrically the 
 result of an analysis, and consequently we refer them to this 
 chapter. 
 
 II, General Manipulations in Analytical Opera- 
 tions. 
 
 There is a certain number of operations, which recur in ev- 
 ery or in almost every analysis, and whose execution requires 
 a variety of little knacks, that can be learned only by prac- 
 tice. The most important of these will be found in what fol- 
 lows. 
 
 1. General Preparations, Selection, Pulverization, Elu- 
 triation, Drying. 
 
 The substance to be analyzed is not always found in pure 
 fragments or crystals and so on, but frequently surrounded by 
 other bodies and mixed with them, from which it has to be, 
 if possible, with care mechanically separated, because, as 
 will easily be comprehended, the analysis of a mixture leads 
 to no result. Thus it is that most of the minerals occur em- 
 bedded in a matrix and in company with other minerals ; in 
 slags and stones from metallurgical processes we generally 
 find interspersed, here and there, small quantities of me- 
 tals in the reguline condition, which by reason of the rapid 
 congelation did not have time to settle. In all such cases the 
 mechanical separation of the substance from its companions is 
 an important circumstance ; and the result is altogether de- 
 pendent on its accurate execution. 
 
11 
 
 It is often much more difficult than is generally supposed 
 to liberate a substance from all adhering foreign matter in 
 order to prepare it for an analysis ; for the mineral substances 
 occur not only in a very line state of division in some foreign 
 matrix, but very freqiiently pieces, which to all appearance 
 are the purest, will be found, on a more minute examination, 
 surrounded by foreign materials. We have only to take into 
 consideration the formation of minerals by crystallization, in 
 order to conceive, that even the well-formed crystals of a sub- 
 stance are not a perfect guarantee for their absolute purity. 
 For, if a body crystallizes out of a fluid which contains other 
 bodies also in solution, it always happens that the incipient 
 crystals contain a trace of the mother liquor, be the latter an 
 aqueous solution or a melted mass*. The quantity of the fo- 
 reign matter thus enclosed increases with the size of the crys- 
 tals, so that it often happens that, in the artificial production 
 of salts, small crystals are produced intentionally, because 
 they are more pure (saltpetre, epsom salts, alum, &c.) If the 
 mother liquor is colored by some foreign material, the impu- 
 rity of the crystals is recognized in like manner by their co- 
 lor, if when pure they are white. 
 
 From this circumstance we meet in nature with substances, 
 whose proper color is modified by some admixture : quartz 
 in its purest varieties, as rock crystal, forming transparent 
 crystals, is found very frequently of a yellow, red, or brown 
 color, arid owes this color to certain metallic oxides with 
 which it is impregnated ; the same thing occurs with calc- 
 spar, heavy spar and feldspar &c. 
 
 It is evident, that it is more easy to judge as to the purity 
 of transparent substances, than of those which are opaque. If 
 quartz crystals were not originally transparent, we should 
 seldom become aware of the foreign bodies, such as chlorite, 
 rutile, specular iron ore &c, which the}' frequently contain. 
 When substances are opaque we have seldom any means of 
 testing the degree of purity of their internal parts, as long as 
 the eye either alone or by the aid of a lens has not been able 
 
12 
 
 to discover any foreign material. If larger crystals, apparent- 
 ly quite pure, are broken to pieces, it sometimes happens, 
 that a small nucleus of foreign substance is found within them. 
 Thus analcime sometimes contains a nucleus of apophyllite, 
 zircon one of carbonate of lime, leucite a crystal of augite &c. 
 
 From all this we may deduce, that small crystals, whether 
 colorless or slightly colored, and such as have grown on 
 the surface and combined for the formation of druses, are 
 the best adapted for the analyses of transparent substances, 
 because we can more legitimately predicate as to their puri- 
 ty, than we could in reference to larger opaque bodies, that 
 are at the same time colored and disseminated in a surround- 
 ing matrix ; indeed sometimes compact, foliated masses, even 
 crystals of the same substance are preferable for an analytical 
 investigation. 
 
 Eut we have not always the choice ; for many minerals, 
 for instance, are found only in one state of formation. No 
 means, therefore, mist be neglected to free, as perfectly as 
 possible, the specimen intended for analysis from all foreign 
 matter. The composition, consequently, of many minerals is 
 very uncertain ; their formulas, notwithstanding numerous 
 excellent analyses, are still too complicated and, therefore, 
 doubtful, because these bodies do not occur naturally in a 
 pure condition, and because a separation of that, which does 
 not essentially belong to the substance, by mechanical and 
 chemical means is impossible. 
 
 In general we are in the habit of regarding those bodies 
 that are found in the analysis to exist only in small quanti- 
 ties, as unimportant, and in the summing up of the results 
 their weights are deducted. But this is not always correct. 
 For such bodies frequently do not constitute in themselves 
 the impurity, but exist in chemical combination with one or 
 more of the essential constituents. If, for instance in the an- 
 alysis of a silicate consisting of lime, magnesia and protoxide 
 of iron, a small quantity of alumina should be found, it is ve- 
 ry improbable that this alumina should be regarded as a soli- 
 
13 
 
 tary admixture ; on the contrary it is much more probable 
 that it should exist likewise as a silicate. But now the ques- 
 tion arises, how much silicic acid belongs to the alumina, and 
 and must not also a part of the remaining bases be added in 
 the computation ? In general it is quite impossible to decide 
 this question ; and it is only when a substance is found in 
 small quantity and belonging to one accompanying the body 
 to be analyzed whose constitution, from other sources, is 
 known with certainty, that we are permitted to compute the 
 admixed quantity of the latter and to deduct it from the re- 
 maining constituents. 
 
 Suppose, for instance, we had to analyze a specimen of 
 copper pyrites, which occurs incorporated with galena, and, 
 after the most careful possible mechanical separation, we still 
 find some lead together with copper, iron and sulphur, there 
 is no doubt that this admixture of lead is to be attributed to 
 the presence of fine particles of galena interspersed in the ore. 
 In this case the composition of galena is known, and it can 
 easily be computed how much sulphur is to be reckoned as 
 belonging to the quantity of lead found. 
 
 We may pursue the same course in the analysis of geologi- 
 cal specimens, that is to say, of mixtures of several minerals, 
 when one of them is known. If, for instance, a doleritic stone 
 is to be examined, consisting of a white feldspar part, and of 
 a black augite or hornblende part, the diminutive size of the 
 ajngle particles often precludes the possibility of separating 
 them mechanically. But if we know, however, from the mi- 
 neralogical investigation, from its deportment with the blow- 
 pipe &c, that the feldspar part of the compound is labradorite, 
 or is supposed to be such with a considerable degree of cer- 
 tainty, the whole substance is subjected to an analysis. From 
 the alkaline and aluminous contents we may compute the a- 
 mount of labradorite in accordance with the well-known for- 
 mula of this mineral, that is, we may deduct the requisite a- 
 mount of silicic acid and lime ; if there remains a residue, in 
 which the silicic acid contains a quantity of oxygen double of 
 
u 
 
 that contained by the remaining part of the lime and the mag- 
 nesia and protoxide of iron found in the analysis, this residue 
 is in fact augite. 
 
 We have above remarked that, when a certain material ]V 
 found by the analysis to exist in minute quantity, it may of- 
 ten be regarded as unimportant, and on this account may be 
 deducted. But this can not take place, if it is isomorphou* 
 with one of the remaining constituents, that is, when it dis- 
 places an equivalent of one of them. For instance, suppose 
 \ve find in a carbonate of lime and magnesia (bitter spar, 
 brown-spar) a small per-centage of protoxide of iron, this can 
 not be deducted, since it is isomorphous with those bases. 
 We may ascertain whether it must be added or not from this 
 circumstance, namely, that it must be added only when the 
 amount of oxygen in the bases is found to be exactly half as 
 great as that in the carbonic acid. We must be guided, there- 
 fore, by the proportions, before we decide whether a substance 
 is important or not. 
 
 In general it is very difficult to pronounce judgment on con- 
 stituents, whose quantity is small, whether they are import- 
 ant or not. Comparative analyses of the material from diffe- 
 rent localities, in different forms of its occurrence, as well a& 
 regard to the nature of the substances in company with it. 
 can alone solve the question. 
 
 After this digression we now return to the methods by 
 which a substance to be analyzed may be separated from all 
 foreign materials accompanying it, and then submitted to the 
 analytical operation. In order to effect this separation the sub- 
 stance is folded \i)> in a piece of paper and then broken up 
 into small pieces, from which all the pure particles are select- 
 ed carefully by means of a pair of forceps. This labor is fre- 
 quently very troublesome, but it is absolutely necessary. 
 
 If metallic or magnetic iron oreas the accompanying body, 
 it is removed with a magnet. 
 
 Sometimes the separation may happen to be of such a na- 
 * re as to be capable of removal by chemical means, if the 
 
15 
 
 accompanying substance, for instance, is soluble in acids, 
 whilst the material itself is not at all affected by it. Thus 
 many minerals may be freed from adhering calcspar by pour- 
 ing over the roughly pulverized mass a quantity oi'dilwie hy- 
 drochloric acid and allowing it to stand some time, after which 
 the acid portion is decanted and the substance itself is washed 
 thoroughly clean and dried. Concentrated acids, however, 
 must not on any account be used for this purpose, since they 
 are almost always sure to attack the substance to be analyzed. 
 
 PULVERIZATION. Most substances must be pulverized 
 
 in order to be prepared for an analysis, so as to render them 
 soluble or capable of separation (silicates must be fused with 
 n alkali ). At one time the powder need not l>e very fine, 
 as for instance, of the carbonates, which are to be dissolved in 
 an acid ; at another it must be very fine, for example, of the 
 silicates and metallic sulphides and arsenides. In all cases 
 the substance is folded up in paper, broken up with a ham- 
 mer into rough particles, and then comminuted in an agate 
 mortar to a fine powder, taking care to operate with a small 
 portion at a time in the mortar, which facilitates the commi- 
 nution. The pulverizing is continued, if the powder is to be 
 very fine, until a pinch of the material between the finger 
 and the thumb is no longer sensitive to the touch. 
 
 Hard substances, especially if the materials are rare and 
 all loss is to be avoided, are first pulverized in a steel mortar, 
 and then comminuted in an agate mortar. Nevertheless by 
 this process fine particles of steel and iron are sure to get in- 
 to the powder, which may be shown by the magnet. On 
 this account the powder must be digested in hydrochloric a- 
 cid and thoroughly washed. From this it is evident, that such 
 substances as are attacked by this acid, must not be pulverized 
 in a steel mortar. If, however, they are so hard as to attack 
 the agate and in this way to become impure by reason of the 
 silicic acid in the mass, there is no alternative, the pulveriza- 
 tion must be effected in the steel mortar ; and if a very fine 
 powder is required, it is obtained by passing the pulverized 
 
16 
 
 material through a fine cambric sieve. Formerly it was 
 customary with very hard substances to determine the increase 
 in weight, which the powder received by being comminuted 
 in an agate mortar; but naturally this method could not be 
 accurate. 
 
 ELUTRIATION. In order to obtain powder in the state of 
 
 the greatest fineness, it is sometimes elutriated. This is ef- 
 fected by first pulverizing the substance in a mortar and then 
 pouring water upon it in a beaker glass, stirring it up tho- 
 roughly, and, after allowing the heavier particles to settle, 
 the turbid fluid, containing the elutriated particles, is decanted 
 into another glass. The rough particles are again rubbed up 
 in a mortar, again treated with water, and the operation so long- 
 repeated, until all or nearly all of the substance is fine. It is 
 sometimes exceedingly convenient to pulverize a substance in 
 a mortar containing a little water, from the fact that the rough- 
 er and heavier particles are thus brought continually under 
 the pestle, and in this way the operation is hastened. The 
 elutriated substance is allowed to settle for some time in the 
 beaker glass, and, when the liquid is clear, decanted ; after that 
 the powder is dried at a gentle heat either in the oven or on 
 the water bath &c. 
 
 The operation of elutriation has another advantage, which 
 allows one to easily discern the presence of another heavier 
 body in the last portions of the washed material, if it is one 
 of those substances sufficiently distinguishable by their colors. 
 Sometimes we resort especially to this plan of washing in or- 
 der to separate one substance from another, which naturally 
 is the less successful the smaller the difference of density in 
 the two bodies. 
 
 Sometimes the wash-water is not thrown away, but is al- 
 lowed to evaporate over the elutriated powder, in order not 
 to lose the portion that may have been dissolved by the wa- 
 ter. 
 
 FILING. FLATTENING. In order to render alloys easily 
 soluble, they are filed with a fine clean file, or, if malleable, 
 
17 
 
 are l)ea>ten out into thin foil under the hammer, and then cut 
 up into small pieces. 
 
 DRYING. Many bodies attract moisture from the air; 
 and this has to be removed before we proceed to the analysis. 
 This is particularly the case with powders ; and on this ac- 
 count they must be previously freed from all hygroscopic im- 
 purity. But in the performance of this operation we must 
 bear in mind that some bodies contain water of crystalliza- 
 tion, which we must not mistake for the water attracted from 
 the atmosphere ; the latter has to be removed, not the former. 
 
 Many salts contain so much water of crysallization, that the 
 common temperature of the atmosphere is capable of remov- 
 ing one part, whereby they are said to effloresce. In this ca- 
 tegory may be reckoned, for instance, carbonate, sulphate and 
 phosphate of soda and several metallic salts &c. To remove 
 all hygroscopic moisture and whatever moisture may be en- 
 closed between the crystalline lamellae of such salts, artificial 
 heat is not to be applied ; but the powder is folded up in filt- 
 ering paper and pressed compactly between several folds of 
 rough blotting paper by placing it under a board loaded with 
 a heavy weight, or between two boards capable of adjustment 
 by means of screws. After pressure the papers are all chan- 
 ged and fresh ones put in their place ; and the operation is 
 repeated until the moisture is entirely removed. The powder, 
 as soon as it is dry, is transferred at once to a covered cruci- 
 ble and weighed. 
 
 If the substance containing water does not effloresce, the 
 powder is placed in a watchglass or in an open platinum cru- 
 cible and then transferred to the desiccating apparatus, that 
 is, to the bellshade standing on aboard or on a plate of glass; 
 here it is left for several hours or days over sulphuric acid. 
 But it must be observed here, that several salts that do not 
 effloresce in the open air, are apt to do so in the dry air of 
 the desiccator. Such salts must be taken out before the efflo- 
 rescence commences, or they must be dried afterwards as 
 above described. It is well to weigh the substance before it 
 3 
 
18 
 
 is dried, as also to place it in the desiccator and, after allow- 
 ing it to remain there some time, to weigh it again. 
 
 If the water, which a substance contains, can not be remo- 
 ved by a temperature less than that of boiling water, the sub- 
 stance has to be dried either on the water bath or in a drying 
 oven. If the former be used, the substance is placed in a co- 
 vered crucible and this is fixed deeply imbedded in a dish 
 filled with sand or shot, and then the latter is placed on the 
 water bath ; but if the drying oven is to be employed, the 
 crucible is transferred at once to the oven and there hung up ; 
 the drying oven is then maintained at a constant temperature 
 of 212F. In both cases the crucible and its contents are weigh- 
 ed after the expiration of a short time and again placed in the 
 oven or on the bath ; and the drying is continued as long as 
 there is any diminution of weight. 
 
 The drying oven is an airbath, in which substances may be 
 exposed for any length of time to the most varied tempera- 
 tures. It consists of a copper cylinder, closed at the bottom, 
 from 3 to 4 inches high and about 3 inches in diameter ; there 
 is a lid with a narrow rim round its circumference which fits 
 loosely on the top of the cylinder. Through an opening in 
 this lid, near the edge, there is a thermometer fixed in a cork ; 
 and half way from the top in the inside there are three pins 
 to support the wire triangle which is to hold the. crucible. 
 The bulb of the thermometer is protected with wire gauze, 
 and is placed as near the crucible as possible. As soon as the 
 substance is placed in the crucible, and the latter is covered 
 up and fixed in position, the cylinder is heated by the flame 
 of a lamp to the proper temperature, and the flame is so re- 
 gulated as to preserve a constant thermal condition. After 
 having been maintained at the temperature required for some 
 time, the flame is removed, and the cylinder is allowed to cool 
 until it can be handled with impunity ; the crucible is then 
 taken out and weighed. The operation is repeated, either at 
 the same temperature as before, or still higher, until the con- 
 secutive weights remain the same. 
 
19 
 
 It is evident, that the drying oven may be employed for 
 substances which bear a higher temperature. A temperature 
 between the range 248F and 266F is very properly applied 
 to several bodies, although they may contain no water, but 
 which, on account of volatile contents or such as change by 
 heat can not be raised to a red heat, unless previously so 
 dried, as for instance the carbonates of the earths, the oxides 
 of the metals, protoxide of iron compounds, peroxides, &c. 
 
 Frequently a body, which contains no water chemically com- 
 bined with it, may be dried in the sandbath, over the lamp 
 at a moderate heat, or even by being raised to a red heat. 
 
 2, Analytical Operations, Dissolving, Fusing with flux- 
 es, Digesting, Evaporating, Boiling, Precipitating, Neu- 
 tralyzing, Filtering, Decanting, Washing, 
 
 DISSOLVING. The substance to be analyzed must previ- 
 ously be dissolved either in water or in an acid. Which of 
 these solvents is to be used depends upon the nature of the 
 substance, and has already been determined by the qualitative 
 analysis. Among the acids hydrochloric acid is the most ge- 
 neral solvent ; it is employed in the concentrated condition 
 and slightly fuming. On the contrary if a sulphur compound 
 has to be analyzed, either concentrated nitric acid or nitro- 
 hydrochloric acid has to be employed, because hydrochloric 
 acid either will not dissolve these compounds, or gives rise to 
 the liberation of hydrosulphuric acid, which carries away with 
 it a part of the sulphur. But it is found to be most conveni- 
 ent to produce the nitro-hydrochloric acid during the course 
 of the analysis, that is, in the first place to ozidize the substance 
 by means of nitric acid, and then as soon as the first operation 
 is over, and the mixture has been allowed to stand for some 
 time, hydrochloric acid is added and the digestion is conti- 
 nued. In the analysis of such sulphur compounds as might 
 easily evolve hydrosulphuric acid with nitric acid, such as sul- 
 phide of iron, sulphide of zinc and sulphide of manganese, we 
 employ a mixture of equal parts of nitric and of hybrochloric 
 acid, which is heated until chlorine begins to escape, and is 
 
20 
 
 then poured gradually upon the substance. 
 
 Fuming nitric acid is more employed, as for instance to ei- 
 feet the oxidation of sulphide of lead. 
 
 Alloys require not too concentrated nitric acid, but it must 
 be free from chlorine, especially in the presence of silver, lead 
 and bismuth. Acid, that is too strong, dissolves such metal? 
 with great difficulty, because their nitrates are insoluble in 
 free nitric acid. 
 
 A general rule and one, that cannot be too much impressed 
 on beginners, is to use no more acid in the solution of a sub- 
 stance than is necessary , for an unnecessary excess is so far 
 prejudicial to accuracy as to require afterwards for its neutra- 
 lization a larger quantity of an alkali, and consequently extra 
 washing and volatilization, and, especially if ammonia has 
 been employed, in order to effect small losses owing to the so- 
 lubility of many precipitates in ammoniacal solutions. 
 
 The solution of substances is effected in dishes, beaker glass- 
 es or flasks, and depends on the nature of those substances and 
 of the solvents employed. 
 
 We shall not give any rules for special cases here, but shall 
 recur to them afterwards in the description of the analytical 
 course in individual cases, and shall limit ourselves at this 
 stage to general remarks. 
 
 All solutions in acids, which produce an evolution of gas, 
 must be made in glass flasks, .which are heated over a spirit 
 lamp. If a substance dissolves only very slowly, it is digest- 
 ed on the sand bath or in the wire net over the lamp ; the 
 flask is slightly inclined on one side, and a small funnel is 
 placed in the neck to obviate all loss from spurting It is ve- 
 ry rare that -a fluid is required to be raised to boiling, for this 
 would cause a loss, as for instance in the solution of the sul- 
 phides ; for, when the fluid is concentrated, there is danger 
 
 of losing by volatilization some of the sulphuric acid that has 
 
 
 been produced. 
 
 On the contrary it would not be proper to use a flask for 
 the decomposition of the silicates by means of acids, from the 
 
21 
 
 fact that silicic acid, as it separates, adheres firmly to the 
 sides of the vessel, and can not easily be removed from a flask 
 or retort. In such cases beaker glasses or evaporating dishes 
 are preferable, because the contents can be digested on a sand 
 bath until the decomposition is effected, and during the ope- 
 ration the vessels may be covered with a glass plate (or still 
 preferably with a large watch glass,- a glass evaporating dish 
 or the bottom of a broken vessel). 
 
 FUSION BY FLUXES. Iii analytical chemistry this opera- 
 tion is employed in the solution of silicates or aluminates, 
 which are not acted upon by acids ; it consists in first fusing 
 them by the aid of the alkalies or their carbonates, the alkaline 
 earths, or with bisulphate of potash. Most of the natural and 
 artificial silicious compounds (slags, glasses), which contain 
 *o much silicic acid as to give a ratio of oxygen double or tre- 
 ble that of the bases, that is to say, most of the bisilicates and 
 trisilicates are scarcely, if at all, acted upon by acids, and con- 
 sequently can not be analyzed in this way, whilst those com- 
 pounds which are less rich in silicic acid, as for instance the 
 singulo-silicates, are decomposed by acids in such a manner, 
 that the silicic acid separates, and the bases become soluble. 
 
 Now the former, whose number is by far the greater of the 
 two, are fused with carbonate of soda, thus giving rise to sili- 
 cate of soda, from the fact that the silicates present in the com- 
 pound give up a part of their acid to the alkali, and displace 
 the carbonic acid which passes off. The more basic compound 
 thus produced is now in all cases easily decomposed by acids. 
 Bisulphate of potassa is the best flux for the aluminates 
 and similar bodies (spinel, corundum, chrome iron) ; it ope- 
 rates only by virtue of the excess of acid, that is, as sulphuric 
 acid alone ; but sulphuric acid can not be used as a substitute, 
 since the decomposition of those bodies requires a temperature 
 higher than this acid can bear. 
 
 Further particulars will be found when we come to the ana- 
 lysis of the separate compounds. 
 
 DIGESTING consists in submitting a fluid for a long time to 
 
a temperature below its boiling point. This operation has fre- 
 quently to be resorted to in analytical chemistry in order to 
 effect the solution or decomposition of bodies ; and the same 
 sort of vessels is used as those in effecting solutions. Further- 
 more, fluids are digested together with precipitates, in order 
 that the latter may completely subside ; such operations can 
 be performed on the digester, over the lamp, or in a hot-air 
 pipe of an ordinary stove. 
 
 EVAPORATION. Fluids are evaporated either partially 
 
 or completely (evaporation to dry ness). The first takes place 
 in order to remove volatile sudstances, for instance, hydrosul- 
 phuric acid from fluids from which metals have been precipi- 
 tated by means of this gas, or in order to concentrate fluids in 
 those cases where, in consequence of washing, their volume 
 has been increased to such an extent as to render it necessary 
 to diminish this volume for the subsequent operations. In 
 this case it is well to evaporate the wash- waters separately, 
 and then to add the residue to the remaining fluid. Frequent- 
 ly also such a solution has to be evaporated to a certain point 
 in order to get rid of an excess of free acid, at least in a great 
 measure, but in such cases the operation must be continued 
 until the acids begin to volatilize, and this does not take place 
 before most of the water has been expelled by evaporation. 
 
 The object of evaporating to dryness is to determine the 
 amoiint of a non-volatile substance, as for instance, of an al- 
 kali in the form of a chloride of a metal or of a sulphate, 
 which happens to be in the fluid, or to render certain sub- 
 stances insoluble, for example, silicic acid set free by acids 
 from the silicates. 
 
 Evaporation is always effected in evaporating -dishes or in 
 crucibles ; it requires much care, and particularly so, not to 
 allow the solution to boil during the whole operation, other- 
 wise there would be a loss by spurting. The heat, therefore, 
 must never be allowed to rise so high as to produce this re- 
 sult. As soon as the work is nearly through, or if the solu- 
 tion contains fixed and especially gelatinous bodies, the mass 
 
is very apt to bubble up even with a moderate temperature ; 
 when this is the case it will be necessary to stir the substance 
 all the while or to finish the evaporation on the water bath. 
 The most difficulty occurs when solutions have to be evapo- 
 rated to dryness in platinum, or porcelaine crucibles, because 
 the small quantity of water is easily raised to the boiling point. 
 This is obviated by placing the crucible on a triangle, furnish- 
 ed with a handle, over a lamp with a very small name until 
 it is properly heated, and then to withdraw the crucible from 
 the flame and oscillate it gently so as to get the solution into 
 a circular motion, and then promote the evaporation by blow- 
 ing with the mouth ; then to bring it again over the lamp and 
 so proceed until the mass appears dry ; the lid is now put on 
 the crucible, and the whole is again exposed to a gentle heat. 
 Spontaneous evaporation in a dry place is also to be recom- 
 mended ; but it frequently happens in such cases, that the salts 
 effloresce, that is, creep up to the edge of the crucible and ev- 
 en over to the outside, a circumstance that must be carefully 
 guarded against. 
 
 BOILING. The operation of boiling in quantitative ana- 
 lytical chemistry is to be avoided as much as possible on eve- 
 ry occasion, because of the loss which so easily is produced by 
 it. Nevertheless it is sometimes indispensable, and in such ca- 
 ses has to be performed in retorts, beaker glasses or evapora- 
 ting dishes. Thus a substance from which the oxide of cop- 
 per is to be precipitated by potassa, or the oxide of zinc by 
 carbonate of soda, must be boiled with these reagents. If too 
 the amount of oxide of iron is to be determined by metallic 
 copper, the whole must be boiled for some time ; solutions of 
 the protosalts of iron can be oxidized easily and completely 
 by nitrate of potassa by boiling alone. 
 
 PRECIPITATING. As a general rule the quantitative de- 
 termination of a body is founded on the fact of its producing 
 with another body (the precipitating reagent) an insoluble 
 compound, from whose weight and known constitution its 
 amount can be calculated. In this way baryta is precipitated 
 
'24- '. ,..: .- ,, 
 
 by sulphuric acid, lime by oxalic acid, silver by hydrochloric 
 acid; the sulphate of baryta, the oxalate of lime and the chlo- 
 ride of silver being insoluble substances, so that by an ade- 
 quate addition of this precipitating agent and a proper atten- 
 tion to all the necessary conditions every trace of baryta, lime, 
 or silver disappears from a solution ; and all that is required 
 to be known is the weight of the resulting precipitate, in or- 
 der to ascertain how much of the three ingredients was pre- 
 sent. 
 
 But it happens just as frequently, that the precipitant se- 
 parates the substance to be thrown down from its combi- 
 nation and takes its place. This is, for instance, always the 
 case when an earth or a metallic oxide is precipitated by an al- 
 kali, the latter uniting with the oxide which kept those bases 
 in solutions, and consequently separates them, since they are 
 insoluble in aqueous and alkaline fluids. 
 
 Again it does not always happen that a body is determined 
 in the form in which it is separated from others by means of a 
 precipitating reagent. Thus, for instance, all metals may be 
 precipitated completely by hydrosulphuric acid, if their solu- 
 tions are in the necessary condition (acid, or alkaline, accord- 
 ing to the nature of the metal), but not all of them can be 
 weighed, that is, determined directly in the form of sulphides, 
 partly because free sulphur is frequently intermingled, and 
 partly because they are partially oxidized in drying, for ex- 
 ample, the sulphides of iron, manganese, zinc and copper. 
 On the contrary mercury, cadmium, antimony and arsenic- 
 can be determined as sulphides. 
 
 In order to produce a precipitation in the proper manner, 
 it is necessary to know all the circumstances, by which it can 
 be completely effectuated. It must be known, whether the 
 fluid to be precipitated may be or must be acid, alkaline, or 
 neutral, and this varies with each body. If, for instance, ar- 
 senic is to be precipitated by hydrosulphuric acid, its solution 
 must be acid ; if silver is to be thrown down by hydrochloric 
 
 acid, the solution may be acid ; if, on the contrary, zinc is to 
 
 ' 
 
25 
 
 he precipitated by liydrosiilplmrie acid, or lime by oxalic acid, 
 its solution must lx j alkaline. 
 
 Furthermore it is not at all indifferent, whether the solu- 
 tion to be precipitated is in a concentrated or dilute condition. 
 If oxide of copper is to be precipitated by potassa, the solution 
 of the former must not be too concentrated, otherwise it will 
 dissolve a portion of the oxide of copper. If on the contrary, 
 oxide of iron and alumina have to be separated by potassa, the 
 solution must be concentrated, otherwise all of the alumina 
 will not be dissolved by the potassa. 
 
 The temperature of the solution to undergo precipitation is 
 another point to be taken into consideration, from the fact that 
 many precipitations must be made in cold solutions (that is to 
 say, at the ordinary temperature), and for others the solutions 
 must be warmed or even raised to the boiling temperature. 
 Tims the oxide of copper is precipitated by potassa, and the ox- 
 ides of zinc and nickel by carbonate of soda at a boiling heat. 
 
 The time, too, required to produce a complete precipitation, 
 must not be lost sight of. Thus when solutions of lime are to be 
 decomposed by oxalic acid, or the salts of magnesia by phos- 
 phate of ammonia, they must beset aside for at least twenty 
 IK >urs, before filtration is attempted, because the separation of 
 the oxalate of lime, and that of the double phosphate of mag- 
 nesia and ammonia take place gradually, and would be only 
 imperfect, if filtration were to take place soon after the pre- 
 cipitants had been added. 
 
 In like manner it often happens, that a longer digestion at 
 a gentle heat is necessary in order to produce a perfect sepa- 
 ration of a body, it being a general rule never to filter until 
 the precipitate has thorouyJily subsided, and the supernatant 
 fluid has become clear. This is necessary partly in order to 
 test the solution by the addition of a small quantity of the 
 precipitant, whether the precipitation is complete, and part- 
 ly in order to facilitate the filtration, since many precipitates, 
 as for instance the sulphate of baryta, easily pass through the 
 filter when they have not thoroughly settled, and thus render 
 
26 
 
 the liltrate turbid. Sometimes, too, a complete separation of 
 the precipitate is only effected by digestion at a moderate 
 heat, as with antimony and arsenic, when solutions of these 
 metals have been treated with hydrosulphuric acid. 
 
 As regards the quantity of the precipitant to be used, it is 
 a rule to add it in slight excess, in order to be certain that 
 the precipitation has been completely effected. But this ex- 
 cess must be small, as just remarked, for an analysis will be 
 the more accurate, the less the excess of reagents. If an acid 
 fluid is to be precipitated by an alkali, or an alkali by an acid, 
 the reaction itself is the surest sign when this excess is attain- 
 ed ; for if the precipitant has a distinct odor, the manifesta- 
 tion of this odor indicates at once the complete precipitation, 
 as for example, when ammonia or hydrosulphuric acid has 
 been used. Nevertheless it is necessary here to guard against 
 a possible deception, which arises from the fact that the space 
 above the fluid, not the fluid itself, may give rise to this odor. 
 It is therefore in this case, as in all cases of precipitation, in- 
 dispensable to stir the mixture continually with a glass rod, to 
 remove the vapor from the vessel by blowing with the breath, 
 and above all things, which is always the safest plan, to add 
 a small quantity of the precipitant after the precipitate has 
 properly settled and the fluid has become clear. 
 
 There are cases in which practice alone can determine the 
 right amount of the precipitant to be used. For instance in 
 the separation of the oxide of iron and alumina it is not suffi- 
 cient to add potassa just enough to produce an alkaline reac- 
 tion, on the contrary an excess of it is necessary in order to 
 produce a complete separation of the two bodies. 
 
 Finally as regards the -vessels to : be used for precipitations, 
 it may be remarked, that when the operations have to be per- 
 formed in the cold or at a moderate temperature whilst di- 
 gesting, beaker glasses are most appropriate; but when the 
 temperature of boiling water is to be applied, evaporating 
 dishes or flasks are more suitable, although with the latter 
 there is the difficulty of removing the last traces of the pre- 
 
27 
 cipitate. 
 
 .NEUTRALIZING. - An exact neutralization of an acid flu- 
 id by an alkali, and of an alkali by an acid is sometimes ne- 
 cessary and can be attained by adding gradually the neutrali- 
 zing material in small quantities at a time, until litmus paper 
 indicates the proper condition of neutralization of the fluid. 
 In this way an acid solution of oxide of iron is neutralized 
 with ammonia in order afterwards to be precipitated by suc- 
 cinate of soda. The smaller the amount of fluid, the more 
 difficult becomes the neutralization, and the greater the atten- 
 tion required in the operation. If carbonic acid is contained 
 in the fluid, the solution must be warmed in order to drive 
 it off, for this gas has of itself an acid reaction. 
 
 FILTERING.-- - There is no operation in analytical chem- 
 istry that is so often repeated as that of filtration. The re- 
 quisite apparatus for this purpose are funnels, filtering paper 
 and filtering stands. The size of the filter depends entirely 
 upon the amount of the precipitate, and not upon that of the 
 tin id ; and it is just as necessary to avoid too large as too small 
 filters. 
 
 Good filtering paper must allow the fluids to pass quickly 
 and clear through them, and contain the smallest amount of 
 ashes. Quick filtering paper is especially necessary in those 
 easels in which either the precipitate or the fluid attracts oxy- 
 gen or carbonic acid from the air (filtering of sulphide of cop- 
 per, sulphide of zinc and sulphide of nickel). But the absence 
 of inorganic matter hi the substance of the paper is of much 
 more importance, when the filters have to be burned; for this / 
 remains as ashes and thus increases the weight of the preci- 
 pitate. This inorganic matter consists in a great measure of 
 sand, sulphate of lime and alumina. If the amount is small 
 it can be determined by repeated experiments for the diffe- 
 rent sized filters and then deducted. Filtering paper, the 
 weight of whose ashes for the ordinary sized filters amounts 
 to more than a few milligrammes, must be rejected, since it 
 is not possible to remove the inorganic matter from the paper 
 
by washing with acids, &c. The so called Swedish paper doe> 
 not contain less foreign material than many other sorts of fil- 
 tering paper and besides this it is so thin as to allow fine pre- 
 cipitates to pass through its fibres. In order to prepare a fil- 
 ter, the first thing to be done is to select a funnel of the pro- 
 per size. A piece of paper is then cut out, square in shape, 
 and folded together so that two opposite sides come in contact. 
 Another fold is made so that the remaining two sides overlie 
 each other. In this way the folds all proceed from the centre 
 to the middle of the four sides. The next thing is to fold the 
 square into the form of a triangle, beginning at, the central 
 point again and extending to the opposite angle. Place the 
 the triangle in the funnel, central point downwards, and 
 mark how much has to be cut oif from the top ; this mark 
 must be about the one-eighth of an inch lower than the rim 
 of the funnel. The upper part is now cut oft' with a pair of 
 scissars in the form of an arc of a circle, the sides of the tri- 
 angle being the radii. The filter is now opened into a quad- 
 rant and pressed against the inside of the funnel. If the in- 
 clination of the sides of the funnel is not sixty degrees, the fil- 
 ter will not fit ; in this case it must be expanded or contract- 
 ed by making a slight change in the inner folds. In every 
 case the filter must be pressed into contact with the sides of 
 the funnel so as to form projecting folds; the upper edge 
 must not reach the rim of the funnel, but must be equally 
 distant from it (a condition attained only when the filter forms 
 a perfect circle when opened out and laid flat) ; and the point 
 of the filter must be in the middle of the tube of the funnel. 
 A filter not answering to these requirements must, in quan- 
 titative operations, be absolutely rejected. It is ready as soon 
 as it has been once more firmly pressed into contact with the 
 sides of the funnel and uniformly moistened with water. 
 
 Filters are also cut out by means of discs of wood or tin 
 plate of the proper size tor the funnels in general use ; or 
 for each size a quadrant of tin is cut out with turned up ed- 
 ges, in which the papers are laid after having been folded three 
 
>& 
 
 times; a similar quadrant then slides over this, and by this 
 fortn the periphery of the filter is cut. 
 
 All precipitates, which <!<> not admit of being ignited, be- 
 cause by this operation the v would undergo some change, but 
 which are to be dried only and in this condition weighed, re- 
 quire dried and weighed Jiffarz. Jn this category may be reck- 
 < >ned sulplmr, chloride of si 1 ver. hydrofiuosilicate oi baryta, the 
 double chloride of platinum and potassa, and the sulphate of the 
 oxide of lead. A filter must never be weighed just as it is, since 
 paper, being a hygroscopic substance, always contains a varia- 
 ble amount of moisture. It must, therefore, first be dried. For 
 this purpose it is placed folded in a crucible, which is then cov- 
 ered with its lid and put in the drying apparatus already de- 
 scribed, which is slowly heated to the temperature of 248F. 
 and kept at this temperature for at least a quarter of an hour. 
 Xurnerous experiments sbow that a filter, which has been ex- 
 posed at longest from 25 t< > '>< > n i inutes to a temperature of 248, 
 no longer loses in weight, if it is again placed in the drying 
 apparatus and kept a longer time at this temperature or 50F. 
 higher. In order to test the accuracy of the drying, all that 
 is required is to repeat the operation; but a diminution of 
 \veight will be observed only in those cases when the drying- 
 has been too much hurried. Since in general the precipitate 
 is afterwarde dried at the same temperature at which the fil- 
 ter was dried, it is only proper to fix upon a higher tempera- 
 ture than 212F., which exceedingly hastens the result. As 
 soon as the drying apparatus lias cooled down to 86 or 104F., 
 the crucible is taken out and placed immediately on the ba- 
 lance ; it is weighed accurately and afterwards the empty cru- 
 cible is weighed, if its weight is not already known. The 
 difference is the weight of the filter. 
 
 Filters, on which sulphur is to be dried, must be dried at a 
 temperature of 212F., beca use sulphur melts at a greater tem- 
 perature. But in this case the drying must be continued lon- 
 ger, at least half an hour, and again repeated, after the cruci- 
 ble has been weighed, until the weight remains constant. 
 
30 
 
 It has already been remarked, that each filter must be moist- 
 ened with water (alcohol) before its use. In the operation .of 
 filtering the supernatant clear fluid is first poured upon the 
 filter, holding a glass rod on the edge of the vessel to guide the 
 fluid upon the filter and to prevent any portion from running 
 over the edge and outside. Should this accident take place, 
 the vessel is placed on a glass plate, and the wet spots are af- 
 terwards cleaned off with the wash-bottle. 
 
 Care must be taken, whilst filtering, that there is no spurt- 
 ing of the fluid; this can be avoided by directing the stream 
 of the fluid by means of a glass rod against the sides of the 
 filter, and by placing the end of the rod almost in apposition 
 with the paper. 
 
 The fluid must never rise as high as the edge of the filter, 
 in order to prevent particles of the precipitate from creeping 
 over the edge by capillary attraction aad thus reaching the 
 glass. 
 
 As soon as the clear fluid has all been poured upon the fil- 
 ter, the precipitate is gently stirred up and in like manner is 
 poured a little at a time until it is all placed upon the filter ; 
 finally the vessel is thoroughly rinsed by means of the wash- 
 bottle and a feather, whose fibres have been cut short, or a 
 thin piece of india-rubber tube on the end of a glass rod. 
 
 The filtrate, that is, the fluid that has passed through the 
 paper, is received in a beaker glass, an evaporating dish, &c, 
 according to the nature of the next operation ; for the same 
 instructions are valid here in reference to the right selection 
 of vessels as before mentioned. The vessel is placed so that 
 the inside wall is in contact with the tube of the funnel, in 
 order that the fluid, as it passes through, may not fall freely 
 in single drops or in a continuous stream, which might easily 
 cause loss by spurting over. 
 
 The filter must be covered with a glass plate during the fil- 
 tration in those cases where the contents are liable to be affect- 
 ed by contact with the air ; as, for instance, during the filtra- 
 tion of metallic sulphides, which easily oxidize, or of ammo- 
 
iiiacal llnids containing involution baryta, strontia, or lime, 
 from which, when exposed to the air, the corresponding car- 
 bonates are thrown down, and thus become mixed with the 
 original precipitate. In all such cases the operation of filter- 
 ing must be performed a* quickly ax possible* and consequent- 
 ly the filter must never be left empty. Furthermore the fil- 
 ter must also be kept covered when alcoholic solutions are 
 filtered, as, for example, when potassa or ammonia has to be 
 determined by chloride of platinum, owing to the volatility 
 of those solutions. And in general, when the filtering or 
 washing has of necessity to be postponed for some time, the 
 same precautions must be observed, in order that the edges 
 of the filter may not dry, a circumstance that renders it dffi- 
 cult to wash the filter thorougly, as is easy to observe with 
 colored metallic solutions. 
 
 It 'sometimes happens that the fluid, that has been filtered. 
 is turbid, this takes place when the particles of the precipi- 
 tates are exceedingly fine ; for instance, with sulphate of bary- 
 ta, the double phosphate of ammonia and magnesia, and the 
 carbonate of protoxide of manganese ; but it never occurs with 
 viscous or gelatinous precipitates. If such an occurrence is 
 observed, the first turbid portions are poured back into the ori- 
 ginal vessel, and the receiver is changed for another as soon 
 as the fluid begins to run clear. Particular precautions will 
 be recommended further on, in reference to the treatment of 
 special substances. 
 
 DECANTING. It is only when the supernatant fluid is 
 
 perfectly clear, and when its volume is considerably larger 
 than the consistent precipitate that has subsided beneath it, 
 that the greater part of the fluid portion may be removed by 
 decantation, instead of passing it through the filter. But this 
 is seldom possible, because as a general rule some of the pre- 
 cipitate has become deposited on the parietes of the vessel or 
 accumalated as a thin scum on the surface of the otherwise 
 clear liquid beneath. The only time when it is practicable is 
 for example, when sulphur has separated owing to the oxida- 
 
tion of the metallic sulphide in the form of a single or seve- 
 ral adherent specks, to remove the fluid by decanting, and sub- 
 sequently, to wash off the former in several waters, and finally 
 to rinse them in a small porcelain dish. 
 
 WASHING . Every precipitate, collected on a filter, must 
 
 be washed, lest any of the remaining substances in the fluid 
 should be retained by it, and thus increase its weight. 
 
 Precipitates are washed either with cold or hot (boiling 
 water, with alcohol, or with annnoniacal solutions, according 
 to the nature of the precipitates : and since it is intended that 
 these bodies should gradually remove, and take place of, the 
 original fluid, a fresh addition must never be made upon the 
 filter before the first quantities have thoroughly drained off. 
 This can be effected, at least when water is used, as in most 
 cases, by means of the wash-bottle, which at the same time 
 stirs up the precipitate and thus brings every part of it equal- 
 ly into contact with the fluid. In every case it requires cau- 
 tion, however, not to produce any clefts in the precipitate, 
 through which the water would quickly flow away, without 
 penetrating the larger mass beneath. 
 
 Voluminous precipitates may be stirred up from time to 
 time with a glass rod, taking great care not to injure the fil- 
 ter. The edges of the filters must be very carefully washed, 
 for it is here that the original fluid with its contents is retain- 
 ed the longest. 
 
 The a/ttwwnt of wash-fluid to be poured on the precipitate 
 each time must not be very large, becaiise better results are 
 obtained by small quantities oft repeated. Beginners are apt 
 to fail very frequently in this respect, and thus soon accumu- 
 late large quantities of fluids, which the vessels cannot hold, 
 and which have to be evaporate* 1 afterwards, all which may 
 easily result in loss, a thing that must be avoided. But if a 
 precipitate is difficult to wash, and on this account the a- 
 mount of fluid becomes very great, the wash- water must be 
 filtered alone, evaporated to a small and convenient volume 
 and thus added to the filtrate t< be concentrated. 
 
38 
 
 The length of time required to wash a precipitate depends 
 upon the nature of the latter; for many precipitates (the pul- 
 verulent and the crystalline) are easily and quickly washed, 
 whilst others ('the gelatinous) are with difficulty washed. In 
 general the result is more quickly attained with hot water, 
 unless there should be some particular reason why this may 
 not be used. 
 
 The proof of a precipitate's having been thoroughly washed 
 can be ascertained by pouring a few drops of the draining 
 wash -water upon a smooth piece of platinum or into the pla- 
 tinum spoon, and then by allowing them slowly to evaporate. 
 If there is any deposit left on the metal, the operation of wash- 
 ing must be renewed. But several precipitates are not always 
 so completely insoluble as not to allow a speck to remain on 
 the platinum, only in this case the deposit will be exceeding- 
 ly small. A knowledge of the nature of different precipitates, 
 and practice will determine what is best to be done in such 
 cases. 
 
 3, Drying and Igniting of Precipitates, Reduction of 
 Filtres to Ashes by Burning, 
 
 The filters, together with their properly washed precipi- 
 tates, must now be set aside to dry. This can be best effected 
 by placing the funnels, together with the filters, covered with 
 paper, in the drying chamber of the digester on its sand bath, 
 or in the hot-air chamber of an ordinary stove, supported on 
 t H j x xls or on beaker glasses without bottoms. A frame of tin 
 standing on four legs and pierced with holes on its top, and 
 >f a size to be placed in the hot-air chamber of a stove will 
 answer the purpose quite well, in case a digester were not at 
 hand. , 
 
 The precipitates, that have thus been dried by hot air, must 
 now be either completely dried in the drying apparatus, in 
 those cases where they rest on weighed filters, or they must 
 be -it/nited, and in this case the filter has to be burned to ashes. 
 
 In the first case the filter together with its contents is placed 
 
 s 
 
84 
 
 in a platinum crucible, covered loosely with its lid, and heat- 
 ed in the drying apparatus to the temperature at which the 
 iilter had previously alone been dried (generally 248 F.) at 
 which temperature it is maintained uninterruptedly for half 
 an hour at least. After the crucible has sufficiently cooled, 
 the lid is firmly closed and transferred direct from the drying 
 apparatus to the balance, and the operation is repeated until 
 the weight remains constant. 
 
 But if a precipitate, on an unweighed filter, has to be ig- 
 nited, the (platinum or porcelain) crucible, in which the igni- 
 tion is to be performed, is placed on a sheet of smooth paper, 
 whose edges have been cut smooth ; the contents of the iilter 
 are then emptied into the crucible, and the particles of the 
 precipitate adhering to the filter are removed as much as pos- 
 sible by rubbing the folds of the paper together. The filter 
 is now folded together, first cutting otf the white portions, 
 when the precipitate is colored, and placed on the lid of a pla- 
 tinum crucible supported on a triangle ; in this way it is 
 brought over the flame of a lamp and burned. The paper is 
 first allowed to take fire, and then the flame of the lamp is di- 
 minished as much as possible and afterwards increased again 
 as soon as the filter has become charred. At last the plati- 
 num cover is raised to a red heat, and a piece of platinum foil 
 half an inch wide is placed on the edge of the cover in order 
 to institute a draft of air and thus promote the ignition of the 
 carbon of the filter. The burned gases that arise out of the 
 flame, carbonic acid and nitrogen, surround, so to say, the ig- 
 nited substance, to which by means of the platinum as it 
 were on a bridge, the oxygen of the atmosphere is drawn. 
 The mass is stirred from time to time with a piece* of thick 
 platinum wire, in order that all the carbonized particles may 
 be completely burned. As soon as there are no longer any 
 charred particles to burn, and the cover is completely cool, 
 the ashes are put to the rest ot the precipitate and brushed 
 off thoroughly from the lid either by means of a feather or a 
 fine pencil. The crucible^ containing the precipitate, is now 
 
35 
 
 covered and gradually heated to a red heat ; it is now tilted 
 a little nn one side, the lid being pushed back so as to expose 
 the contents, and the piece of platinum foil is again placed 
 mi the edge of the crucible, in order that whatever fibres of 
 paper may have been abraded from the filter and now con- 
 tained in the precipitate, may burn, or if, as is the case with 
 many metallic oxides, a reduction has taken place at different 
 points, the reduced metal must again be oxidized. After a 
 propei' amount of ignition, the temperature and continuation 
 of which depend upon the nature of the precipitates, the cru- 
 crucible is closed up tight, and, as soon as it has cooled, it is 
 placed in the balance and weighed. This is the more ne- 
 cessary from the fact, that all ignited pulverulent substances 
 are in the highest degree hygroscopic ; and, if the crucible 
 can not be weighed right away, in all exact experiments, it 
 must be placed under the bell shade in a vacuum over sul- 
 phuric acid. 
 
 'When filters are burned on the cover of a platinum cruci- 
 ble, there is always the possibility of losing some of the ashes 
 by reason of the draft of air; and this is particularly the case 
 with light, tiocculent precipitates, as, for example, with dried 
 silicic acid. Tn such cases as soon as the greatest part of the 
 precipitate has been poured into the crucible, it is much safer 
 to put the filter also in, to cover it well up and then to raise 
 the temperature gradually until the filter is carbonized. 
 
 When this has taken place the lid is shoved slightly on one 
 side so as to open the crucible partially, and the charred paper 
 is burned with the aid of the platinum foil laid over the edge 
 of the crucible to institute a current of air. 
 
 The same process is followed, too, when the amount of pre- 
 cipitate is so small as not to be capable of being removed from 
 the filter, the latter together with the adhering precipitate 
 being treated entirely in the crucible. 
 
 It is frequently not easy to drive the carbon from the char- 
 red filter, especially when the temperature is raised too high, 
 in which case the carbon is less combustible, and the access of 
 
36 
 
 air is too feeble ; or when the filter contains precipitates which 
 are not quite insoluble, and are thus protected from burning 
 from the fact that the fibres are surrounded with a til in of 
 the precipitate, as is the case with the double phosphate of 
 ammonia and magnesia. 
 
 It has already been observed, that filtering paper must con- 
 tain as little ash as possible, in order that its weight need not 
 be taken into consideration, or that its. amount may be deter- 
 mined beforehand for the different sized filters. 
 
 EL Vessels of Platinum and Silver. 
 
 The accuracy of analytical experiments has been greatly 
 increased by the application of crucibles, dishes, tfec, of plati- 
 num. Owing to the expense, however, of such apparatus, it 
 becomes necessary to avoid all circumstances by means <>f 
 which these vessels might be injured. 
 
 Platinum crucibles are indispensable in quantitative ana- 
 lyses for the drying and burning of filters, for the igniting 
 and weighing of precipitates, for the evaporation of fluids in 
 which alkalies are to be determined, and for the heating ami 
 igniting of residues after evaporation, finally for the fusion of 
 silicates, aluminates, &c, by means of fluxes of the alkaline 
 carbonates, bisulphate of potassa, fluoride of ammonium, ifec. 
 
 Platinum dishes are used with great advantage for evapo- 
 ation, to heat or boil substances with strong acids (sulphuric 
 acid) or with solutions of the caustic alkalies, for dissolving 
 by means of hydrofluoric acid, possessing as they do the ines- 
 timable property, in all accurate operations, of not being act- 
 ed upon by these bodies. 
 
 On the other hand the following bodies must not be treat- 
 ed in platinum vessels : 
 
 1. The caustic alkalies as well as baryta and strontia and 
 their nitrates at a red and a melting heat. 
 
 2. Fusible metallic sulphides or mixtures of sulphates 
 with carbon ; thus in particular the alkaline sulphides. 
 
 3. Phosphates and carbon, which at a red heat give rise 
 
37 
 
 to phosphide of platinum. 
 
 4. Metals that are easily fused, or mixtures of their oxides 
 with substances which reduce them (carbon). These form 
 with platinum fusible alloys. 
 
 5. Chlorine and all mixtures which liberate chlorine either 
 in the cold or by heat, for instance, fusible mixtures of a me- 
 tallic chloride and a nitrate, of sulphate of ammonia and chlo- 
 ride of ammonium. Furthermore manganates, permanga- 
 nates and hydrochloric acid. 
 
 Platinum vessels are cleaned with borax or bisulphate of 
 potassa, which is fused in them, they are afterwards digested 
 in dilute acid, or scoured with tine round sand (sea sand) 
 and water. 
 
 Silver crucibles are scarcely used for any thing else than 
 for the fusing of the hydrate of potassa or of soda (separation 
 of alumina from the oxide of iron, &c). They must be capa- 
 cious and strong, and they require much care when raised to 
 a red heat, on account of the easy fusibility of the metal, 
 which naturally must be quite pure. If they are brought 
 into contact with dilute hydrochloric acid, a certain amount 
 of silver is dissolved. 
 
 IV, Balance and Weight, 
 
 One of the most indispensable instruments for the chemist 
 is an accurate balance. For the analysis of inorganic substan- 
 ces a balance, when loaded with thirty grammes on either side 
 and still turns with one milligramme, is sufficiently sensitive. 
 
 It is placed firmly in position where it can be protected 
 both from dust and moisture. In order to fix it in a horizontal 
 position it is furnished with adjusting screws and a spirit level. 
 
 When properly adjusted, the oscillations of the beam, which 
 may be observed on a graduated index fixed behind the apex 
 of the tongue, must be equal on either side of the zero point. 
 
 Amongst the precautions to be observed in reference to the 
 use of a balance, may be mentioned, that neither a weight nor 
 a vessel must ever be placed on the pan of the balance or 
 
8i 
 
 removed from it, as long as the beam 1*8 not fixed (that is, the 
 motion of the beam is arrested, which generally is effected by 
 some contrivance attached beneath the pans). Furthermore 
 all violent agitation, every current of air must be avoided 
 whilst weighing. A hot w.v.W must never be placed in the 
 balance, because the weight in tin's case would not be accurate. 
 
 The weights, in general use among analytical chemists, are 
 French and belong to the decimal system. The standard unit 
 of these weights is the gramme (grm.), that is, the weight of 
 a cubic centimetre of pure water at its greatest density, that 
 is, at a temperature of 39.39F. Its equivalent weight is- 
 15.434 grains (Apothecary's weight). 
 
 A gramme is divided into 10 decigrammes, 1 decigramme 
 into 10 centigrammes, and 1 centigramme into 10 milli- 
 grammes. 
 
 All exprssions of weight are put down in the form of a de- 
 cimal, in whicn the number of units before the period indi- 
 cates the number of grammes, whilst the first unit figure to 
 the right of the period expresses the decigrammes (tenths of 
 a gramme), the sec6nd the centigrammes (hundredths of a 
 gramme), and the third the milligrammes (thousandths of a 
 gramme). 
 
 For instance, if a substance weighs 14 grammes 9 deci- 
 grammes 3 centigrammes and 5 milligrammes, it is express- 
 ed as follows : 14.935 grm. Tf one of the intermediate parts 
 is wanting, its place is occupied, of course, by a cipher. 
 
 It is not at all necessary that the weights should be abso- 
 lutely what they stand for, that is. that the gramme should 
 be in reality equal in weight to a cubic centimetre of water, 
 but relativel/y to each other they must be most accurate, and 
 in due correspondence throughout, from which it is to be un- 
 derstood, that any weight, either slightly more or less than 
 a gramme may be assumed as the stsndard unit, and that this 
 must be equal to 10 other weights exactly, each of which is 
 mutually equal, <&c. 
 
 When gases have to be weighed, it is then absolutely lie- 
 
cessary to make use of the real French gramme as a standard, 
 because the weight of gases is based on certain lixed volumes. 
 
 It is of great importance to be careful in writing down the 
 ascertained weights ; for a single fault in. this respect render* 
 the whole analysis useless: and there are no means of eon- 
 trolling it. 
 
 For the analysis of a substance it is not customary to weigh 
 out exactly one or two grammes, because this method would 
 take up very much time, and it is almost just as easy to com- 
 pute the result of the analysis with any other quantity. 
 
 It is no easy matter to determine in general the quantity 
 of substance to be used in a given analysis ; as a rule from one 
 to two grammes will be sufficient. It is only when one of the 
 constituents in a given substance exists in a very small quan- 
 tity, and from its nature a further investigation becomes ne- 
 cessary, that a larger quantity is to be operated upon. Most 
 substances are weighed out in vessels, for which purpose pla- 
 tinum vessels are the best, since the substances generally have 
 been previously reduced to a tine pow^der and in this state 
 dried, and in consequence are apt quickly to attract more or 
 less moisture. But such bodies, where this is not to be appre- 
 hended, and which in general are presented in more compact 
 masses, as for example alloys, may be weighed alone in the 
 balance. 
 
 In the weighing of substances contained in a 'vessel, as for 
 instance, in a platinum crucible, it is a general rule tirst to 
 weigh the substance and the vessel together, and then when 
 the substance is transferred to another more suitable f ()1 ' the 
 further analysis, to weigh the crucible alone, because it fre- 
 quently happens that a small quantity of the substance ad- 
 heres to the crucible and would thus increase its weight. 
 The difference of the two weights gi ves accurately the quan- 
 tity used in the analysis. But, if the substance is easily so- 
 luble in the solvent to be employed, as for instance, a salt in 
 water, the crucible is first weighed, afterwards the crucible 
 ;md substance together; the latter is then poured out, and 
 
40 
 
 the crucible is washed out with the solvent, which is added 
 ro the contents. 
 
 V, Analysis by Measure, 
 
 VOL u \r KT KIC M KT HOD. 
 
 The alkalimeter of Descroizelles is the type of all the me- 
 thods of analysis, according to which the amount of a body is 
 determined from the volume of a solution of some reagent of 
 known strength, which liberates the body either in the form 
 of gas, or of an insoluble combination, or gives rise to a cer- 
 tain color in consequence of oxidation or reduction. Thi> 
 branch of analytical chemistry owes its development especi- 
 ally to (.Tay-Lussac (alkalimetry, chlorimetry, the determina- 
 tion of silver), to Marguerite (analysis of iron by means of 
 the permanganate of soda), and to Bunsen (the determinina- 
 tion of iodine), and has already arrived at such a degree of 
 perfection, that the volumetric determinations of substances 
 are employed not alone for technical, but also for the most 
 accurate scientific analyses, and are in a very high degree to 
 be recommended owing to the shortness of time required in 
 the operations. 
 
 The first and most important requirement is the prepara- 
 tion of a solution of the reagent, which contains an accurate- 
 ly known quantity of the same in a given volume of the sol- 
 vent. The standard volume in use is the cubic centimetre 
 (c. c.), which is divided into halves (0.5), or into fifths (0.2), 
 or into tenths (0.1). The volume occupied by 1000 c. c. is 
 called a litre. The amount of the effective reagent contained 
 in a given volume of fluid is called its standard strength, and 
 the fluid itself is denominated the standard solution. The 
 preparation of a standard solution depend* upon the nature 
 of the reagent and may be made in two different ways. By 
 the first method an accurately weighed amount of the reagent 
 is dissolved in a so-called volume- vessel (flask, cylinder), that 
 is, in a vessel containing a given volume, one litre or more, 
 up to a mark on the neck or side of the vessel. This mode 
 
41 
 
 of determining the volume is sufficient as far as regards tech- 
 nical operations: but for scientific purposes, and particularly 
 in those cases where the substance can not be weighed with 
 exactness, the strength has to be determined by special expe- 
 riments much more accurately. (Vide Bunseii's process for 
 determining the amount of iodine). This method consists in 
 preparing a tiiiid whose strength is not fixed, or merely ap- 
 proximatively known, and then in arriving at accurate results 
 by special experiments (vide determination of the standard 
 acid in alkalimetry, and of the permanganate of potassa in 
 the iron analysis). 
 
 The strength must not change in the course of time. But if 
 such a change does take place, although to a trifling amount, 
 it has again to be determined from time to time; but if the 
 change is more appreciable, the adjustment must be made ei- 
 ther before or after each analysis. 
 
 The final reaction is denominated the visibly perceptible 
 appearance indicated by the completion of the reaction, that 
 is, the point at which any further addition of the reagent must 
 cease. A change of color is the most appropriate for this pur- 
 pose, that is, either the commencement of a color or its dis- 
 appearance ; consequently reagents are used whose colors are 
 definite and intense (permanganate of potassa), or a substance 
 is added which produces a colored combination with the small- 
 est possible excess of the reagent (starch paste for iodine, lit- 
 mus for acids and alkalies). 
 
 The vessels, in which the standard solution of the reagent 
 is contained, and from which it is allowed to pass out either 
 in a stream or drop by drop, when required, are long narrow 
 cylinders of glass drawn out at the bottom into a point and 
 (Called burettes. They are graduated, naturally, into cubic 
 centimetres and their fractional parts, and are in connection 
 at the top with a larger reservoir, from which they can easi- 
 ly be filled with fiuid in some convenient way. Those, which 
 are supplied at the lower end with compressor stopcocks, are 
 (Called compressor burettes, instruments first introduced by 
 
12 
 
 Molir; they allow the passage to he dosed by com pressing a 
 small tube of indiarubber attadied to the elongated narrow 
 orifice at the bottom of the tube; there are others which are 
 supplied with glass stopcocks and other modes of dosing the 
 lower orifice and regulating the outlet of fluid. But all the^e 
 forms have their faults; and stopcocks of infallible perfection 
 remain yet to be invented. The rending of the height of the 
 fluid, before and after the experiment, is effected by placing 
 the eye on a level with the surface of the fluid ; but more ac- 
 curately by means of a small glass tube, hermetically scaled. 
 and containing a small quantity of mercury which causes the 
 tube to sink in the standard solution. This plan was first 
 suggested by Erdmann, and the tube has a horizontal line 
 inscribed around its surface, and the coincidence of this line 
 with one of the graduated lines on the outer tube is very ea- 
 sily observed. Such burettes are called swim-burettes. 
 
 1, Alkalimetry and Acidimetry, 
 
 The object of this analysis is to obtain the amount of alka- 
 li in aqueous solutions and in their carbonates, as also the 
 amount of free acid in acid liquids, in all cases in which one 
 equivalent of base unites with, one equivalent of acid to form 
 a neutral combination. The following reagents are required : 
 
 1. Sulphuric acid. One part of English sulphuric acid 
 is diluted with ten or twelve of water, and in this condition 
 is kept in stock. 
 
 2. Solution of soda. It must be free from carbonic acid, 
 and preserved in well stoppered bottles to which carbonic 
 add can not get access. It must be diluted and have a spe- 
 cific gravity of about 1.1. 
 
 3. Tincture of litmus. The aqueous extract of litmus, 
 which after filtration has been freed from any excess of hase 
 by means of a little acetic acid, until the solution assumes a 
 violet color; it is kept in an open vessel. 
 
 Before use we must determine the strength of the sulphu- 
 ric acid, and then the relative strengths of the sulphuric acid 
 
48 
 
 and the soda solution. This is determined in the following 
 manner: Pour a certain volume (from 20 to 50 c.c.) of the 
 acid into water colored with a few drops of the tincture of 
 litmus, and then place the vessel under the soda burette, and 
 allow so much of the soda solution to flow r into it until the 
 last drop changes the red color to blue. (It is very conveni- 
 ent in practice, although not necessary, to make the strength 
 <>f each fluid equal, that is, that one cubic centimetre of the 
 acid shall neutralize one c,c. of the soda solution ; this is ea- 
 sily accomplished by diluting either one or the other of these 
 two fluids.) 
 
 In order to determine the absolute standard of the dilute 
 sulphuric acid, from one to two grammes of pure carbonate 
 of soda are placed in a covered platinum crucible and raised 
 to a red heat (not fused), it is then weighed, dissolved in a flask 
 in water, colored with a few drops of the tincture of litmus, 
 and then placed under the acid burette, until the fluid be- 
 comes red. The fluid is now heated almost to boiling in or- 
 der to expel any free carbonic acid, and then a trifling excess 
 of acid is added to produce a permanent color. This is ef- 
 fected by placing the solution under the soda burette, from 
 which so much soda solution is admitted as to reproduce the 
 blue color. The amount of soda solution used gives that of 
 the excess of acid, and by deducting this from the total 
 amount of acid used, there remains the quantity which was 
 required to neutralize the carbonate of soda. These experi- 
 ments must be several times repeated, in order to obtain 
 accurate results. 
 
 One equivalent of sulphuric acid(=40) saturates one equi- 
 valent of soda (=31), or one equivalent of hydrate of soda 
 (40), or <me equivalent of protocarbonate of soda (=53). 
 
 Supposing it has been ascertained, that one gramme of 
 carbonate of soda requires ten cubic centimetres of acid, then 
 100 cubic centimetres of the acid will saturate ten grammes 
 of carbonate of soda, or 7.547" grammes of hydrate of soda, or 
 5.K5 grammes of anhydrous soda. 
 
44 
 
 Since the equivalent of potassa is 47, of the hydrated po- 
 tassa 56, and of the carbonate of potassa 69, then .100 cubic 
 centimetres of this acid will saturate 13.019 grammes of the 
 protocarbonate of potassa, or 10.560 of the hydrate of potas- 
 sa, or 8.87 of the anhydrous potassa. 
 
 From this the strength of the dilute acid in anhydrous acid, 
 or the hydrated acid (SO 3 -f HO =49) is determined as follows : 
 53 : 40:: 10 : 7.547* 
 53 : 49:: 10 : 9.245 
 
 Tims 100 cubic centimetres of the acid are found to contain 
 7.547 grammes of the anhydrous acid, or 9.245 grammes of 
 the hydrated acid. 
 
 2, Volumetric Analysis with Permanganate of Potassa, 
 
 This reagent is a dilute solution of permanganate of potas- 
 sa, the best being that which is pure and crystallized. It 
 must not contain a trace of free potassa, nor too much chlo- 
 ride of ammonium. It owes its application to its powerful 
 oxidixing properties and its intense color. 
 
 In order to obtain the standard strength of the solution, 
 from one to two grammes of pure soft iron, free from rust, 
 are dissolved (protected from all access of air, for instance in 
 a flask provided with an India rubber valve,) in hydrochloric 
 acid, to which previously a few T pieces of marble have been 
 added. This solution is poured into from 20 to 30 litres of 
 water, and then so much of the chameleon solution is added, 
 until the last drop communicates a slight but distinct red co- 
 lor to the whole solution. This experiment is repeated seve- 
 ral times. 
 
 Instead of metallic iron a weighed quantity of the double 
 sulphate of iron and ammonia (AmS 8 +FeS 8 +6 Aq) maybe 
 used. This is dissolved in water, and, as in the preceding 
 experiment, is rendered quite dilute ; it is then made slightly 
 acid with hydrochloric acid placed beneath the burette con- 
 taining the solution of permanganate of potassa. The salt 
 
 * Dilute sulphuric acid contains exactly the same amount of anhydrous sulphuric, 
 acid as of the hydrate of soda, their equivalents being equal (=40). 
 
+5 
 
 must have been recently prepared, pulverized and quite dry. 
 Seven parts of this salt contain exactly one part of iron. 
 
 Ol)*. The solution of permanganate of potassa may be 
 kept unchanged for a pretty long time ; nevertheless it is 
 well, after the lapse of some time, to determine its contents 
 again. India rubber must not be used to close the aperture 
 < >f the burette. 
 
 3, Bunsen's Volumetric Analysis with Iodine, 
 
 This excellent method is used for determining the amount 
 of oxygen, chlorine, &c. ; it consists in allowing the latter di- 
 rect, the former in the form of an equivalent of chlorine, to 
 net on a solution of iodide of potassium, by which one equi- 
 valent of iodine is set free. 
 
 KI + C1=K()1 + I 
 
 The amount of iodine is determined by an aqueous solution 
 of sulphurous acid. 
 
 8O,+I + HO=8O a +HI 
 
 But since such a solution is constantly changing in strength, 
 the latter must be determined for each analysis. 
 
 The requisite reagents are as follows : 
 
 1 Sulphurous acid. From TO to 80 cubic centimetres of 
 H saturated aqueous solution are diluted to ten litres. 
 
 2. Solution of iodine. Twenty grammes of pure dry io- 
 dine and 200 grammes of iodide of potassium are dissolved 
 in six litres of water. One cubic centimetre of this solution 
 contains therefore 0.005 grammes of free iodine. But as io- 
 dine is not always pure and dry, and can not easily be accu- 
 rately weighed by reason of its volatility, the amount of the 
 iodine in the solution must be controlled in an especial man- 
 ner (for scientific purposes) as follows. 
 
 Pure recrystalliz'ed chlorate of potassa is pulverized to a fine 
 powder and dried at a teniperture of 392F. An accurately 
 weighed quantity (about 0.1 grin.) is placed in a flask con- 
 nected with tubes with two Wolfe's bottles. The first and 
 the larger of the two contains a solution of 3 grammes of 10- 
 
46 
 
 dide of potassium, tilling it with this solution up to one-fourth 
 or one-third of its volume ; and the tube dips a few lines into 
 the fluid. The second bottle contains only a small quantity 
 of the solution of iodide of potassium. The flask is most con- 
 veniently closed with a stopper of indiarubber through which 
 the narrow glass tube passes. As soon as all the connections 
 are made, dilute hydrochloric acid is poured upon the chlo- 
 rate of potassa, and the aperture is closed a% quickly as pos- 
 sible. The solution is now heated, but quite gradually, lest 
 the chlorine should be evolved too rapidly ; it is afterwards 
 made to boil, and the boiling is continued until all the chlo- 
 rine has been driven off. Since the fluid in the first bottle 
 becomes heated by the steam eliminated in the flask, the bot- 
 tle is placed in a vessel of cold water. The iodine set free 
 dissolves in the excess of iodide of potassium and gives rise to 
 a brownish red solution. The contents of the second bottle 
 remain colorless, if the evolution of chlorine is not too rapid. 
 In order to complete the experiment, the stopper is taken 
 out of the flask whilst the contents are still boiling ; and the 
 solution of iodine is placed in a beaker glass, to which is 
 added sulphurous acid from a burette, stirring all the while, 
 until the fluid has lost all its color, after which about 5 cubic 
 centimetres more of the sulphurous acid (up to the next gra- 
 duated line) are added. It is now mixed with a little thin 
 starch paste and then placed under the iodine burette from 
 which a sufficient quantity is allowed to flow so as to produce 
 a distinct blue color. 
 
 After the completion of this experiment the relative 
 strengths of the two solutions, namely, of sulphurous acid and 
 <f iodine are determined as follows: a fixed quantity, say 50 
 cubic centimetres of the former are poured into water, to this 
 a few drops of starch solution are added, and the solution is 
 finally placed beneath the iodine burette as before. 
 
 In this w r ay the strength of the iodine solution is determined. 
 
 From one equivalent of chlorate of potassa ( = 122.5) six 
 equivalents of chlorine are evolved, which set free 6 equiva- 
 
47 
 
 lents of iodine (= 702). If one gramme of the salt has been 
 used, 0.622 grin, of iodine will be set free. 
 
 Supposing the result of the second experiment shows, that 
 5 cubic centimetres of sulphurous acid were equal to 1 cubic 
 centimetre of the solution of iodine, and that there were in 
 the first experiment 630 cubic centimetres of sulphurous acid 
 and 1 cubic centimetre of the iodine solution, the excess of 
 the former would have been equal to 5 cubic centimetres, the 
 free iodine therefore would have required only 725 cubic 
 centimetres of sulphurous acid. Since now 
 
 : : I:: 625 : 125 
 
 therefore, 125 cubic centimetres of the iodine solution con- 
 tain just as much iodine as the quantity obtained by means 
 of 0.1 gramme of chlorate of potassa. that is, 0.622 gramme. 
 iSince now 
 
 125 : 0.622: MOO : 0.4975 
 therefore 
 
 100 c.c. solution of iodine contain 0.4975 gramme. 
 
 1 0.004975 " 
 
 of iodine, instead of 1 c.c. " containing .005 " as it should 
 have been. 
 
 The solution of the sulphurous acid accordingly contains 
 only 0.025 per cent of the latter. According to Bimsen its 
 strength must not be higher than from 0.03 to 0.04 percent* 
 
 The strength of the iodine solution must be repeatedly 
 tried in order to attain the greatest certainty. 
 
 I. Metallic Alloys, 
 
 Under this head are comprehended all substances, compo- 
 sed of metals in the reguline condition, with the exception 
 of arsenic, whose combinations with other metals will he- 
 treated of in a special section further on. 
 
 Of minerals, therefore, will be comprehended here : amal- 
 gams, antimonial silver, meteoric iron ; to a much greater ex- 
 
m 
 
 tent, however, all tlie artificially prepared alloys, as gold and 
 silver coins, brass, bronze, cannon and bell metal, German 
 silver, soft solder and other fusible combinations, type-metal, 
 as well as those metals produced by metallnrgic processes 
 which, as a rule, are accompanied with others, copper, tin. 
 zinc, lead, silver, nickel, fec.* 
 
 These bodies we divide into two classes in reference to 
 their comportment with nitric acid : First, those which are 
 completely soluble in nitric acid; and, secondly, those which 
 are perfectly insoluble in this acid. To the latter belong- 
 all alloys containing tin, antimony and gold. 
 
 The most suitable solvent for alloys is nitric acid, which 
 converts almost all the metals into soluble nitrates ; but gold 
 platinum and a few other metals of more rare occurrence are 
 not dissolved by this acid. Tin and antimony, however, al- 
 though they are oxidized by this acid, give rise to insoluble 
 oxides ; tin being converted into the insoluble tannic acid, 
 whilst antimony, according to the strength of the acid and 
 the continuance of its action, is converted into the teroxide 
 of antimony, antiinonite of antimony (formerly, antimonious 
 acid), and aiitimonic acid, of which one part is soluble, whilst 
 the other remains behind, thus rendering the accurate deter- 
 mination of the metal very difficult. 
 
 The nitric acid to be used as a solvent must I>e free from 
 chlorine if the alloys contain any gold, silver, lead, tin or 
 antimony. In other cases its presence does no harm ; on the 
 contrary, it sometimes happens that aqua regia must be used 
 for the solution of certain alloys. 
 
 Alloys containing much silver or lead are not so easily af- 
 fected with a very concentrated nitric acid, as with a more 
 dilute acid, from the fact, that the nitrates of these metals 
 are insoluble in the acid itself. 
 
 * The analysis of crude iron, bar iron and f teel will be found in another section.. 
 
49 
 
 A. ALLOYS, WHICH ARK SOLUBLE IN NITRIC ACID, THAT 
 
 IS, WHICH CONTAIN NEITHER GOLD, TIN NOR ANTIMONY. 
 
 1. SILVEE AND COPPER. 
 
 (SILVER COINS.) 
 
 Place a suitable quantity of the substance (from one to two 
 grammes) in a small flask or beaker glass, arid cover it with 
 H moderately strong pure nitric acid, warm the flask over the 
 lamp, or the beaker glass, covered with a piece of concave 
 glass, in the sand-bath, and continue the digestion until the 
 solution is complete. The solution is then diluted with wa- 
 ter and poured into a beaker glass if the operation had been 
 performed in a flask, otherwise it is continued in the same 
 beaker glass. In the former case the flask is thorougly wash- 
 ed with water, which is added to the solution. 
 
 The silver is precipitated out of the solution by means of 
 hydrochloric acid, which is added drop by drop, stirring all 
 the while, as long as a precipitate is produced. The solution 
 is then set aside in a gentle heat until all the precipitated 
 chloride of silver has settled to the bottom and the superna- 
 tant fluid is quite clear. In the meanwhile a filter sufficiently 
 large for the precipitate is placed to dry at a temperature of 
 24-8 F. (instructions for doing this are given on page 29), and 
 through this the clear solution is filtered into a capacious dish 3 
 having first tested it with a few drops of hydrochloric acid to 
 ascertain the complete separation of silver. The fluid por- 
 tion is poured first on the filter, and the chloride of silver is 
 stirred up with a little water and poured next ; the beaker 
 glass is thoroughly cleaned by means of the washing bottle 
 and a feather, and the wash-water filtered. It is weH to add 
 a few drops of nitric acid to the wash- water in order to pre- 
 vent any chloride, of silver from passing through the filter 
 and thus* rendering the filtrate turbid. The same addition 
 of nitric arid is to be recommended when the mass of chlo- 
 ride of silver on the filter is washed with hot water, but only 
 at the beginning ; afterward* pure water alone ie used, and 
 t 
 
50 
 
 the washing is continued until the filtrate no longer shows an 
 acid reaction, nor leaves any deposit when evaporated to dry- 
 ness. 
 
 Vide page 23 concerning precipitating, page 27 about fil- 
 tering, and page 32 in relation to washing. 
 
 The washed filter together with the chloride of silver is 
 dried (vide page 33), then placed in a platinum crucible and 
 put to dry in the drying apparatus (vide page 34), and the re- 
 sidue is then weighed, and from this the quantity of the sil- 
 ver is computed. 
 
 Add carbonate of soda to the copper solution in the dish 
 until it has only a faint acid reaction. The operation is per- 
 formed as follows: add the crystallized salt in small portions 
 at a time gradually and stir the solution after each addition 
 until the salt is dissolved, and then add fresh crystals. Cover 
 the vessel with a glass plate in order to prevent loss by the 
 evolution of carbonic acid; the glass cover is filially washed 
 with the washing bottle. As soon as the point has been at- 
 tained when the fluid has just a slight acid reaction, and the 
 incipient precipitate again dissolves on stirring, the dish is 
 heated over the lamp, and when the fluid is hot, a solution 
 of caustic potassa is added until it has a distinct alkaline reac- 
 tion; it is then raised to a boilng temperature, stirring it all 
 the while, until the precipitate becomes brown-black. When 
 sufficiently cool the contents of the dish are filtered, the dish 
 is carefully cleaned from all adhering oxide of copper, and 
 the latter is thoroughly washed on the filter with hot water. 
 
 The filtrate from the oxide of copper is tested with hydro- 
 eulphuric acid for any remnant of copper. Sometimes a small 
 quantity is found. In this case the precipitate is placed in 
 a covered vessel to settle, the fluid portion is poured off, and 
 
 the residue is filtered and washed in a small filter, dried and 
 
 . 
 
 and raised to a strong red heat under free access of air. It 
 is now oxide of copper which is weighed and added to the 
 larger quantity. 
 
 OBSERVATION. The addition of carbonate of soda is intended to remove the 
 
 largest portion of the free acid, in order not to require so much potassa, which is rarely 
 
51 
 
 ever pun-, and which consequently produces a larger quantity of potassa salts, that ara 
 not so easily removed as the corresponding soda salts. The oxide of copper must never 
 be precipitated out of very concentrated solutions, because it is somewhat soluble in a 
 concentrated solution of potassa and adheres tenaciously in part to the dish. In such a 
 case it may be removed by a few drops of hydrochloric acid and a dilute solution of po- 
 kissu with the aid of heat. It requires considerable time before the washing is complete 
 
 As soon as the oxide of copper has been well dried in the 
 air (vide Drying, page 33), it must be heated to a red heat 
 according to instruction, page 34, taking especial care to ad- 
 mit free access of air, because a part is easily reduced by the 
 charcoal to the state of the dinoxide, and it must be weighed 
 in the covered crucible as soon as it is cold. From this 
 weight, the amount of copper is computed. 
 
 2, COPPEE AND ZINC. 
 
 ('KKASS, SIMILOK, PINCHBECK, TINSEL OK FALSE GOLD LEAF, 
 AND BRONZE FREE FROM TIN.) 
 
 The substance is dissolved in the same manner precisely as 
 in the preceding example, only here be it observed, that it 
 is no disadvantage whatever if the nitric acid should contain 
 chlorine. In fact the alloy may first be covered with hydro- 
 chloric acid, a little nitric acid then added, the whole heated, 
 and then more of the latter acid may be added until the solu- 
 tion is complete. The resulting compounds in this case are 
 chloride of copper and chloride of zinc- instead of nitrates of 
 the same metals. 
 
 The acid fluid is diluted with sufficient water in a beaker, 
 and a current of hydrosulphuric acid is passed through it, ta- 
 king care to cover the vessel with a plate of glass. This ope- 
 ration is continued until all the copper is precipitated, a fact 
 recognized by the strong smell of hydrosulphuric acid after 
 stirring the mixture and blowing away the air above the flu- 
 id with the breath from the mouth, or still more certainly 
 by allowing the precipitate to settle, and the fluid to become 
 clear, when the latter will appear quite colorless, and on the 
 addition of a few drops of a solution of hydrosulphuric acid 
 no more precipitate is produced. 
 
 As soon as all the copper has thus been completely preci- 
 
52 
 
 pitated, the glass tube is disconnected from the indiarubber 
 connector of the hydrosulphuric-acid-generating apparatus, 
 and left in the beaker glass ; the sulphide of copper is then 
 filtered and the filtrate is received in an evaporating dish. 
 
 The filtering of sulphide of copper (as also of several other 
 metallic sulphides, for example, of zinc, iron, nickel, cobalt) 
 requires many precautions. Like them it is endowed with 
 the property, in the moist condition, of attracting oxygen from 
 the air, whereby it becomes sulphate of the oxide of copper. 
 
 To obviate this the filtration must be performed with ra- 
 pidity, and without access of air. Therefore not only must 
 the filter be covered with a plate of glass, but the beaker glass, 
 which contains the precipitate, must be similarly protect- 
 ed. As soon as the contents of the filter have been passed 
 through, a fresh supply must be immediately poured upon it, 
 and the funnel kept full until the precipite itself is completely 
 on the filter. Particles, which adhere to the beaker glass, to 
 the covers or conducting tube must be removed with the wash- 
 bottle and a feather and also placed on the filter, and the re- 
 sidue is finally washed for a short time with cold water con- 
 taining a few drops of hydrosulphuric acid. 
 
 If the necessary precaution be neglected, the portions of 
 fluid, that pass through last, will produce a turbidity in the 
 filtrate, arising from the fact that these last portions carry 
 down a solution of the sulphate of the oxide of copper produ- 
 ced by the oxidation of the sulphide, which, coming in con- 
 tact with an excess of hydrosulphuric acid in the filtrate, a- 
 gain produces a precipitate of the sulphide of copper. In this 
 way the operation is a failure, and must be repeated. 
 
 Since the precipitate of sulphide of copper, on account of 
 its liability to change when exposed to air and warmth, is not 
 appropriate for the direct determination of the copper, it must- 
 be converted into a sulphur compound corresponding to a prot- 
 oxide, which is unchangeable. The precipitate is dried, rub- 
 bed from the filter and poured .into a weighed porcelain cm 
 cible; the filter together with any adhering particles of the 
 
precipitate is burned on the cover of the crucible, and the 
 ashes are then added to the sulphide in the crucible; now add 
 a few centigrammes of the flowers of sulphur and mix it with 
 the rest. The crucible is then covered with a lid to which 
 is attached a bent porcelain tube, intended tV conduct dry 
 hydrogen gas into the crucible, which is placed over a lamp 
 and gradually heated to redness as soon as the atmospheric 
 air has been removed by the hydrogen. The contents of the 
 crucible are allowed to cool beneath an atmosphere of hydro- 
 gen. In this way pure disulphide of copper. Cu^S, is obtain- 
 ed, and this is weighed and the copper therefrom computed. 
 We now come to the fluid separated by filtration from the 
 sulphide of copper, containing the zinc in solution. This is 
 first evaporated in order to concentrate it, and to remove all 
 excess of hydrosulphurie acid. ('hryxtaUized carlxmote of 
 xoda is now added gradually, stirring the mixture all the while, 
 until all effervescence ceases; a precipitate is thus produced 
 and the fluid has an alkaline reaction. During this operation 
 the vessel is covered with a plate of glass to prevent loss by 
 the spurting of the carbonic acid, and the plate itself is after- 
 wards cleaned with the washing bottle. The fluid is next 
 raised to a boiling temperarure, the carbonate of zinc which 
 is precipitated is then removed by filtration and washed with 
 boiling water. The precipitate is dried and heated to a high 
 red heat, in order to remove the carbonic acid and thus pro- 
 duce pure oxide of zinc. The burning of the filter to ashes 
 is performed in the same way as with that containing oxide 
 of copper, only there is less fear of the reduction of the me- 
 tal. The oxide of zinc is then weighed and the amount of 
 zinc computed. 
 
 OBSERVATION 1. The fluid, filtered from the precipitated carbonate of ziue. 
 
 is tested by a few drops, of sulphide of ammonium to see if it is Tree from /inc. If a. pre- 
 cipitate is produced, it is allowed to subside in a covered beaker <;lass: the clear super- 
 natant fluid is poured off. and the residue is filtered and washed in a funnel covered witli 
 a Blasts plate; it is afterwards ditjesled together with the filter, whilst still moist, with 
 a little hydrochloric acid in a beaker i>Ia*(*. until the smell of hydrosulphuric acid has 
 been removed; the solution is then diluted, filtered intoadish and ajrain precipitated a-- 
 before with carbonate of soda. 
 
 OBSERVATION 2. A portion of the zinc oxide, after it has been weighed, i* 
 
54 
 
 tested for the presence of any carbonic acid, which might still remain if the carbonate 
 had not been heated long enough. It is moistened with water and then hydrochloric 
 acid is added which ought to produce no effervescence. 
 
 Sometimes it happens that the oxide of zinc has a yellow color ; this arises in general 
 from a very small quantity of iron which is seldom absent in such alloys. If it is desir- 
 ed to determine the amount of iron, the oxide of zinc that has been heated to a red heat 
 and weighed, is placed in a beaker glass, dissolved in dilute hydrochloric acid and su- 
 persaturated with ammonia. The oxide of iron is precipitated, and is separated by fil- 
 tration, thoroughly Avashed, dried, heated to a red heat and weighed. From this the 
 amount of iron is computed arid deducted from the preceding weight of the oxide of zinc, 
 which in general may be neglected. 
 
 3. COPPEE ZINC AND NICKEL. 
 
 (GERMAN SILVER, ARGENTAN, PACKFONG.) 
 
 The operation at flu- beginning is precisely the same as in 
 preceding example, the copper is precipitated from the acid 
 solution by hydrosulphuric acid, and converted into the disul- 
 phide. The fluid portion filtered from the sulphide of copper 
 containing the zinc and nickel is. in like manner, concentra- 
 ted by evaporation, in order to separate the remaining traces 
 of hvdrosulphuric acid. It is next supersaturated in a flask 
 with pure hydrated potassa, and afterwards sufficient hydro- 
 cyanic acid is added so as to convert the whole into a yellow 
 fluid. From this solution of the double cyanides the zinc is 
 precipitated as sulphide of zinc by the common sulphide of 
 potassium (prepared from the sulphate of potassa by means of 
 charcoal) ; the mixture is set aside to digest and settle ; it IB 
 then filtered in a covered filter into a flask; the precipitate 
 is washed in cold water containing a little sulphide of potas- 
 sium, and dissolved in a covered beaker glass by digestion 
 in hydrochloric acid, until the smell of hydrosulphuric acid 
 has been entirely driven off; the diluted solution is filtered 
 into a dish ; and the oxide of zinc precipitated by carbonate 
 of soda as in the preceding example. 
 
 The filtrate from the sulphide of zinc is boiled in the flask 
 with aqua regia. until all smell of hydrocyanic acid and hy- 
 drosulphuric acid has been expelled and does not return on 
 the addition of acid : the solution is then supersaturated in a 
 dish with hydrate of potnssa and kept boiling for some mi- 
 nutes. The hydrated oxide of nickel is then filtered from 
 
the solution, washed in hot water, dried, heated to 
 
 and weighed ; the weight of the nicke! is computed from the 
 
 amount of the oxide. 
 
 The direct determination of the nickel may be omitted ; in 
 this case the Huid separated from the sulphide of copper, af- 
 ter all hydrosulphuric acid lias been removed, is treated witli 
 an excess of carbonate of soda at a boiling temperature (as in 
 the precipitation of the zinc); the mixed oxide of zinc and of 
 nickel on the filter is washed, dried, heated to a red heat and 
 weighed. It is next placed in a flask and dissolved in hy- 
 drochloric acid, potassa is then added, afterwards hydrocyanic 
 acid and so on, as before. The amount of the oxide of nic- 
 kel is then obtained by the difference of weights. 
 
 4. LEAD AND ZINC. 
 
 The alloy is dissolved in moderately strong nitric acid, and 
 to the solution is added the proper amount of sulphuric acid; 
 the mixture is evaporated until the excess of sulphuric acid 
 begins to pass off, it is now allowed to cool, is then diluted 
 with water containing one-fourth its volume of alcohol and 
 filtered on paper dried at 248F. The precipitated sulphate 
 <>f lead is thoroughly washed and dried. As soon as it is dry* 
 it is placed, together with the filter, in the drying apparatus 
 at a temperature of 24-8 F., and left there until its weight be- 
 comes c( distant. 
 
 The filtrate collected in a dish is evaporated until all the 
 alcohol is driven off, and then the oxide of zinc is obtained 
 according to instructions, page 53. 
 
 5, OOPPEE AND BISMUTH. 
 
 The solution in nitric acid is evaporated to a small volume, 
 mixed with hydrochloric acid and diluted with much water. 
 by this the basic chloride of bismuth ( l>i( 'l^-f^l'iOg) is sepa- 
 rated completely. The complete separation depends solely on 
 'the right pn. port ion between the ucid and the water. The 
 greater the quantity of acid present, the gi-cjiler the .jiiantitv 
 
56 
 
 of water to be added. The deposit, therefore, is allowed to 
 settle, and the supernatant fluid is tested with an addition of 
 water.* The precipitate is either placed on a weighed filter, 
 so long washed with cold water until the draiiiings no longer 
 show any acid reaction and finally dried at the temperature 
 of boiling water, or it is collected on an unweighed filter, and 
 after drying melted in a covered porcelain crucible with five 
 times its weight of cyanide of potassium (the filter being pla- 
 ced on the top and covered with cyanide of potassium). As 
 soon as the crucible is cold, it is immersed in a dish of warm 
 water ; and the bismuth, that has been melted into fine glo- 
 bules, is washed first in water, afterwards in alcohol, it u 
 then dried and weighed. (If a part has been reduced in the 
 form of a black powder, it is melted over again with cyanide 
 of potassium, and so on as before.) 
 
 The filtrate containing the copper is evaporated to a suita- 
 ble volume, then sufficient carbonate of soda is added to still 
 leave the fluid slightly acid, and afterwards a small quantity 
 of aqueous sulphurous acid. The mixture is set aside for a 
 tew hours either in a cold place or where it is gently warm- 
 ed. Hereupon the copper, reduced by this proceeding to the 
 state of protoxide, is precipitated by a concentrated solution 
 of sulphocyanide of potassium ; the precipitate is allowed to 
 settle, and the fluid is tested again with the reagent. The 
 precipitate of sulphocyanide of copper, Cu 2 CyS 2 , is collected 
 on a weighed filter, washed in cold water and dried at a tem- 
 perature of 212F. 
 
 6. LEAD AND BISMUTH, 
 
 The solution in nitric acid is reduced by evaporation to a 
 small volume, and then mixed with a small quantity of water 
 and with hydrochloric acid so as to retain all the bismuth and 
 the lead in solution, if the latter is not present in too great a 
 
 *Very acid solutions of bismuth mtwt be freed from their excess of acid either by eva- 
 poration., or by a partial neutralization with an alkali. When free nitric acid is present, 
 either chloride of sodium or chloride of poUMtsinm i iwed k> produce precipitation, in- 
 Htettd of hydrochloric acid. 
 
57 
 
 quantity, in which some chloride of lead is separated. In or- 
 der to ascertain the right quantity of hydrochloric acid, a lit- 
 tle of the fluid is mixed with water. If a single drop produ- 
 ces turbidity, more acid must be added. When this has been 
 done, the clear fluid is poured into a beaker glass (any chlo- 
 ride of lead that may have been separated may, according to 
 it8 quantity, be either dissolved in water by boiling, or be 
 placed on a weighed filter, washed with dilute alcohol and 
 dried at M8 F.), and the lead is precipitated by the addition 
 of moderately dilute sulphuric acid, and the mixture is set a- 
 side for some time to settle. After this, alcohol, spec. grav. 
 O.S, is added and the mixture is again allowed to settle; the 
 precipitated sulphate of the oxide of lead is then separated by 
 filtration, washed with alcohol containing a small quantity of 
 hydrochloric acid, finally with pure alcohol, aud treated after- 
 wards according to instructions, page 55. 
 
 The bismuth in the filtrate is precipitated as the basic chlo- 
 ride by means of a large quantity of water. Since the preci- 
 pitate contains traces of sulphuric acid, it is reduced by ve- 
 ry accurate manipulations with cyanide of potassium (vide 
 page 56). 
 
 7, COPPEK, LEAD, ZINC. 
 
 The nitric acid solution is evaporated according to instruc- 
 tions, page 55, with sulphuric acid ; and the lead is determi- 
 ned as sulphate of the oxide of lead. 
 
 The alcohol is expelled by heat from the filtrate in an evar, 
 porating dish ; and then hydrosulphuric acid is passed through 
 it. Vide, for the rest of the treatment, page 51. 
 
 B, ALLOYS, WHICH ABE NOT COMPLETELY SOLUBLE IN 
 
 TKIC? ACII). 
 
 1. TIK, COFFEE. 
 
 (BKONZK, UKLL-METAL, CANNON-METAL, <fec.) 
 8 
 
58 
 
 1. METHOD. 
 
 The alloy is first reduced into the smallest pieces possible 
 and then digested in moderately concentrated nitric acid. sp. 
 gr. 1.3, free from chlorine, until the metallic parts are oxidi- 
 fced; the fluid portion, together with the deposited stannic 
 acid is put into an evaporating dish and evaporated to dry- 
 ness on the water bath ; the dish is then placed over the flame 
 of a lamp, and the temperature is raised until a portion of the 
 nitrate of copper begins to be decomposed, liberating the 
 black oxide of copper ; it is now set aside to cool and then 
 moistened with nitric acid until the black oxide again disap- 
 pears ; after diluting the residue, the stannic acid is removed 
 by filtration, and washed in several waters until it no longer 
 shows an acid reaction. When perfectly dry the stannic acid 
 is rubbed from the filter and poured into a porcelain crucible ; 
 the filter itself is incinerated on the cover of the crucible, and 
 the ashes are added to the contents of the crucible, which are 
 now moistened with nitric acid, then heated and finally igni- 
 ted. The amount of tin is computed from the weight of the 
 stannic acid. 
 
 The copper is precipitated from the filtrate by means of 
 hydrosulphuric acid (vide page 51). 
 
 OBSERVATION. The stannic acid may contain a trace of copper. 
 
 2. METHOD. 
 
 Proceed from the commencement as before until the mass 
 is reduced by evaporation to dryness ; the residue is then treat- 
 ed with hydrochloric acid and completely dissolved by the 
 addition of water and the application of heat. The stannic 
 acid is then precipitated by dilute sulphuric acid and allowed 
 to settle ; it is then filtered, washed, dried and ignited as be- 
 fore indicated. After ignition a small piece of carbonate of 
 ammonia is placed within the crucible, which is covered up 
 and heated a second time. The operation is repeated, after 
 weighing, until the weight remains constant. 
 
 The copper is determined as already indicated. 
 
59 
 2. TIN, LEAD. 
 
 (SOFT SOLDER, BLOCK TIN, TIN FOIL ALLOYED WITH LEAD, &C.) 
 
 The alloy is first oxidized with nitric acid, evaporated to 
 dry ness on the water bath ; the residue is then heated at a 
 higher temperature, moistened with nitric acid ; and the stan- 
 nic acrid is separated as in the preceding example (first meth- 
 od). The filtrate is evaporated with the addition of sulphuric 
 acid ; and the lead is determined as sulphate of the oxide of 
 lead (vide page 55). 
 
 3. TIN, LEAD, ZINC. 
 
 Analysis the same as in the foregoing, combined with No. 
 4 (page 55). 
 
 4. TIN, COPPER, LEAD, ZINC. 
 
 (SEVERAL SORTS OF BRONZE AND OF BRASS.) 
 
 In such alloys the amount of tin and lead is generally very 
 small. The process is that of No. 1 (first method) and No. 
 2 ; after the separation of the stannic acid by filtration, the 
 lead is removed from the filtrate by sulphuric acid, then the 
 copper by hydrosulphuric acid, and finally the zinc by carbo- 
 nate of soda ( vide page 53). 
 
 TIN, LEAD, BISMUTH. 
 
 (FUSIBLE METAL, ROSE'S METAL.) 
 
 The alloy is first cut up into small pieces, and oxidized in a 
 capacious flask with nitric acid ; the solution is slightly dilu- 
 ted, saturated with ammonia in excess and digested for a con- 
 siderable time with concentrated sulphide of ammonium to 
 which is added a small quantity of sulphur, taking care to 
 shake the mixture frequently. In this way the tin is dissol- 
 ved as sulphide ; the sulphides of lead and bismuth are sepa- 
 rated by filtration and washed with cold water. 
 
 To the fluid, containing the sulphide of tin, is added dilute 
 sulphuric acid in excess; and the vessel, loosely covered with 
 
BO 
 
 paper, is submitted to quite a gentle heat, until the contents 
 no longer smell of hydrosulphuric acid. The yellow sulphide 
 of tin is washed with cold water on the filter, and is then .so 
 far dried as to allow the filter with its contents to be removed 
 from the funnel. The latter is then put. in the first place, 
 in a covered porcelain crucible, which is afterwards opened, 
 and exposed to a long continued gentle heat, until it no long- 
 er smells of hydrosulphuric acid. This point attained, the 
 contents of the crucible are moistened with a few drops of ni- 
 tric acid, and heated at first gradually with access of Mir, and 
 finally ignited. The sulphide of tin is next changed into 
 stannic acid, and afterwards treated with a little carbonate of 
 ammonia (vide page 58). 
 
 The sulphides of lead and bismuth are dried and rubbetl 
 off from the filter which is incinerated; these ashes and the 
 dried sulphides are digested with nitric acid (vide page 57). 
 in order to separate these two metals. 
 
 6. TIN, LEAD, BISMUTH, CADMIUM. 
 
 (WOOD'S FUSIBLE ALLOY.) 
 
 The process begins the same as in the preceding example. 
 'After separating the bismuth by water as basic chloride, and 
 evaporating the fluid to a suitable volume, the cadmium can 
 be determined by two methods. 
 
 A. Add carbonate of potassa in excess; the mixture is 
 
 then heated ; the carbonate of the oxide of cadmium is sepa- 
 rated by filtration, washed, dried, rubbed off from the filter 
 and placed in a porcelain crucible ; the filter itself is incine- 
 rated on the cover of the crucible, and the ashes are added 
 to the contents of the' crucible; these are ignited and main- 
 'iained at this temperature for some time in order to drive off* 
 the carbonic acid, which is no easy operation. The brown o\- 
 "Mo of cadmium is now weighed, then moistened with a few 
 drops of nitric acid, again ignited and weighed, in order to 
 ascertain whether all the carbonic acid has been driven of). 
 
 B. The cadmium is precipitated by hydrosulphuric ftc- 
 
m 
 
 id, the resulting sulphide is filtered on a weighed filter. wash- 
 ed with cold water, containing a little hvdrosulphnric acid 
 and hydrochloric acid, and finally dried at a temperature of 
 21 2 F., until the weight remains constant. 
 
 7. TIN, LEAD, BISMUTH, MERCUKY, 
 
 (FUSIBLE ALLOY.) 
 
 Proceed as in the two preceding examples. After the se- 
 paration of the bismuth, the mercury is converted into the 
 sulphide by hydrosulphurie acid: the latter is filtered on a 
 weighed filter, washed with cold water, and dried at a tem- 
 perature of 212F., until the weight remains constant. 
 
 8. ANTIMONY AND LEAD. 
 
 This alloy forms the metal denominated type-metal, but it 
 sometimes contains a little copper and bismuth. 
 
 Tf antimony and lead alone are present, the alloy is divid- 
 ed up into very small pieces, and then digested in a flask with 
 concentrated nitric acid, until the antimony is oxidized. Tin- 
 solution is then diluted ; and ammonia in excess is added, ami 
 afterwards a concentrated solution of the yellow sulphide of 
 ammonium to supersatu ration. It is necessary that the latter 
 reagent should he yellow."- because the solution of the sul- 
 phide of antimony takes place more quickly in the presence 
 of free sulphur. If, therefore, the sulphide ef ammonium lw> 
 been recently prepared, a small quantity of the flowers of sul- 
 phur must be added. The whole is digested for a long time 
 in the flask, and towards the end the mixture is heated near- 
 ly to the boiling point. The flask is finally removed from the 
 heat, corked and set aside until the precipitate has subsided. 
 The latter must be of a pure black color, whilst the superna- 
 tant fluid must be yellow: if this is not the <-ase with either 
 of them, too little <>f the sulphide <if ammonium has been em- 
 
 *Sulphido of ammonium is prepared by pitting a current of hydrosulphuric :)cij 
 through a mixture of one part: of ammonia and two parts of watvr, until the fluid u. 
 longer precipitates a solution of nulphate of magnesia. It become* yellow by kecpim. 
 through the formation of Riperaulphide of ammonium. 
 
62 
 
 ployed, consequent]} more must be added and the digestion 
 repeated. As soon as the solution is cold, the undissolved 
 i-mlphide of lead is removed by filtration from the solution of 
 antimony, washed thoroughly with cold water, dried and then 
 placed in a weighed porcelain crucible ; the filter is burned 
 .separately and the ashes are added to the sulphide; a quanti- 
 ty of sulphur is now mixed up with it and the whole is igni- 
 ted in a current of hydrogen, in the mariner in which sulphide 
 of copper was treated (vide page 53). As soon as the cruci- 
 ble containing the sulphide of lead has been weighed, more 
 sulphur is added and the ignition is repeated, in order to see 
 whether the weight remains constant. The amount of lead 
 is estimated from the quantity of sulphide of lead (PbS). 
 
 The solution of sulphide of antimony is decomposed by hy- 
 drochloric acid, adding it drop by drop and stirring the solu- 
 tion all the while, until it has a slight acid reaction, arid then 
 the vessel is covered with a plate of glass. By proceeding 
 in this manner, the sulphide of antimony is precipitated. The 
 mixture, is digested at a gentle heat, until all smell of hydro- 
 Aulphuric acid has disappeared, it is then filtered on a filter 
 that has been dried at a temperature of 248 F. ; the precipi- 
 tate is washed in several waters (cold), to which at the begin- 
 ning a few drops of hydrochloric acid have been added, in 
 order that the fluid may not be turbid as it passes through. 
 After it has been dried in the air, it is maintained at a tem- 
 perature of from 24:8 to 266F. in the drying apparatus, as 
 long as its weight changes. 
 
 From its weight, however, the amount of antimony can not 
 be computed in a direct manner, since it is not alone a mix- 
 ture of the two different sulphides of antimony, but in gene- 
 ral contains also free sulphur. In order to ascertain the am- 
 ount of antimony in the precipitate, as much as can be remo- 
 ved from the filter is placed in a weighed capacious porcelain 
 crucible, aiid the filter, together with that portion of the pre- 
 cipitate still adhering- to it, is dried again in order to ascertain 
 the amount which has been taken out. The latter is first 
 
moistened with a few drops of nitric acid, sp. gr. from 1.2 to 
 1.4, afterwards a larger quantity of acid, .sp. gr. l.r> is added : 
 the crucible is then covered with a watch glass, heated on a 
 water bath, and afterwards very gradually over the lamp to 
 ignition. By this means the sulphides are converted into the 
 combination of the two oxides, Sb() 8 Sb() 5 , which remains a* 
 a white mass, from which the amount of antimony is compu- 
 ted. The ignition is repeated in order to ascertain whether 
 the weight remains constant or not. 
 
 6, ANTIMONY, TIN, 
 
 The alloy is divided up into the smallest, possible pieces 
 and then covered with pure nitric acid in a beaker glass in 
 which it is digested, until all has been oxidized ; the contents 
 are then placed in a porcelain crucible and evaporated to dry- 
 ness; the residue is now heated, but not to ignition. It is 
 now mixed with an excess of hydrated soda, and fused in a 
 silver crucible and maintained at a red heat for some time. 
 After cooling the residue is immersed in a large quantity of 
 water containing one-third its volume of alcohol, sp. gr. ,S33. 
 By this procedure the stannate of soda is dissolved, whilst the 
 antimoniate of soda remains behind. The latter is removed 
 by filtration, and washed at first with equal parts of alcohol 
 (sp. gr. 0.833) and water, and afterwards with a mixture of 
 three volumes of alcohol and one of water. The funnel is 
 now fixed air-tight in the cork of a flask ; and a mixture o? 
 hydrochloric acid and tartaric acid is poured upon the salt : 
 the funnel is now covered up and set aside. As soon as the 
 contents are dissolved, the fluid is allowed to pass, through 
 the filter, which is washed with the same mixture and finally 
 with water. The antimony is next precipitated by menus of 
 liydrosulphuric acid; and the resulting sulphide of antimony 
 is treated as in the preceding example. 
 
 The solution of stannate of soda is gently wanned, in order 
 to drain off the largest portion of the alcohol, water is then 
 added, and the solution is rendered acid with hydrochloric 
 
04 
 
 acid. Finally the tin is precipitated with hydrochloric acid. 
 The yellow sulphide of tin in converted into stannic acid (vi- 
 de page 60). 
 
 ' H METALLIC OXIDES. 
 1. ANALYSIS OF IRON. 
 
 In order to determine the amount of iron in a fluid the vo- 
 lumetric process* with permanganate of potassa may be em- 
 ployed (vide page 44). If the iluid contains a salt of thejpro- 
 te/V/A the volumetric process may be applied at once ; if the 
 solution contains a salt of the xexq'uioa'ide, the latter must tirst 
 be reduced to the state of protoxide; and if finally salts of 
 both &&&%4rata present, two volumetric analyses are made: 
 t.he difference between the two is that amount of protoxide 
 which by computation must be changed into sesqnioxide. 
 A. Determination of the Sesqnioxide. 
 
 The solution in hydrochloric acid is diluted in a flask, if it 
 5* very acid it is partially neutralized with carbonate of soda, 
 *nd then reduced with zinc. This is effected by suspending 
 ,1 ball of zinc, attached to a platinum wire by casting, in the 
 tiuid ; the wire is supported by a loosely fitting cork. In this 
 condition the fluid is kept heated until it becomes colorless. 
 
 The ball is then taken out and rinced ; whilst the solution 
 is poured into a large quantity of water and submitted to the 
 action of permanganate of potassa from a burette. 
 
 B. -Determination of the Sesquioxide and the Protoxide. 
 
 Firstly (a). A weighed quantity of the substance is dissolv- 
 ed in hydrochloric acid, and, in order to prevent all access of 
 *ir, either a small piece of marble is dissolved with it. or the 
 solution is made in a small 1lask furnished with an india-rub- 
 ber, valve. The Holutaon 5* poinvd into vessel of pure water, 
 
65 
 
 and then submitted to the action of permanganate of potassa. 
 
 This experiment gives the amount of the protoxide. 
 
 Secondly (b). A fresh quantity of the substance is dissol- 
 ved in hydrochloric acid and treated as in article A. This 
 jives the whole of the iron in the form of protoxide; from 
 this the quantity found in the first part of B is deducted, and 
 the difference, when multplied by ten-ninths, will give the 
 amount of sesquioxide in the substance. 
 
 If in a substance containing both the oxides of iron the 
 whole amount of iron is determined, by precipitation with 
 ammonia, as sesquioxide, then, naturally the first analysis^) 
 is all that is required. For, if the amount of protoxide found 
 be multiplied by ten-ninths and the product be deducted from 
 the total amount of sesquioxide of iron precipitated by am- 
 monia, we obtain tne amount of protoxide originally present 
 in the substance axamined. 
 
 2, BKOWN HEMATITE, LIMONITE, 
 
 (SKSQUIOXIDK OF IRCN, CLAY, WATER.) 
 
 A. A portion reduced to rough powder is ignited, and 
 
 the loss, of weight is determined. This loss gives the weight 
 of the water. 
 
 j$. A. quantity reduced to a line powder is digested in 
 
 hydrochloric acid until all, with the exception of the clay it- 
 self, is dissolved. To the acid solution is added ammonia in 
 excess ; the precipitate is removed by filtration, washed with 
 hot water, dried, ignited and weighed as sesquioxide of iron. 
 
 In the filtrate may be found sometimes small quantities of 
 lime (by means of oxalic acid) and of magnesia (by phosphate 
 of soda). Vide Carbonates, for example, Spathic Iron Ore. 
 
 3, BOG IKON OEE, 
 
 Hydrated oxide of iron, and hydrated oxide of manganese, 
 mixed with silicic acid, phosphate, sulphate aud huminate of 
 the sesquioxide and of the protoxide of iron, as also with 
 
66 
 
 quartz-sand,"" 
 
 Bog ores are impure brown hematite, in which the sub- 
 stances above-mentioned are not always present. 
 
 A A weighed quantity in fine powder is dried over sulphu- 
 ric acid in the drying apparatus (under a bell-shade of the air- 
 *pump) in order to determine the amount of hygroscopic wa- 
 ter. The remaining results are computed in reference to the 
 weight of the dried substance. 
 
 B The powder, that has been dried in (A.), is heated in 
 a platinum crucible, then ignited with access of air. The W* 
 of weight consists of water and organic substances. 
 
 To determine the amount of each of these substances M-pa- 
 rately is of no use in practice; and besides this an organic el- 
 ementary analysis would be required; we shall therefore pro- 
 ceed no further with them. 
 
 C A weighed quantity of the. ore is reduced to an impal- 
 pable powder and digested in an evaporating dish with hy- 
 drochloric acid (if chlorine is evolved, it indicates the presence 
 of the oxide of manganese) and evaporated to dryness on the 
 water-bath. When cold the residue is moistened uniformly 
 with hydrochloric acid, after standing some time water is add- 
 ed, and the mixture is heated and then filtered. The insolu- 
 ble residue consists of the silicic acid from the silicates pre- 
 sent in the ore, and of quartz. After it has been thoroughly 
 washed, dried, ignited and weighed, it is added gradually to 
 a boiling concentrated solution of from 20 to 30 grammes of 
 carbonate of soda, and is kept boiling for some time ; the hot 
 fluid is afterwards diluted and filtered ; the filtrate contains 
 the silicic acid, which has been separated from the silicates, 
 whilst the residue is quartz ; this is washed, dried, ignited 
 and weighed. The amount of the silicic acid is obtained by 
 taking the difference of the two weights. The solution in hy- 
 drochloric acid is supersaturated w r ith ammonia. The preci- 
 
 
 
 *Sometimog also alumina, lime, magnesia, oxide of copper and arsenious add. The 
 presence of alumina together with phosphoric acid renders the analysis difficult. (Vi- 
 de Phosphates, example 7.) 
 
67 
 
 pitate is thoroughly washed, dried, ignited and weighed. It 
 contains beside the oxide of iron also oxide of manganese and 
 the phosphorus acids. In order to determine the amount of 
 these substances, the residue is nibbed to a fine powder, then 
 ignited and weighed; it is then mixed in a platinum crucible 
 with from three to lour rimes its weight of powdered dried 
 carbonate of soda. This mixture is maintained for a quarter 
 of au hour at a vivid red heat; the crucible, when cold, is pla- 
 ced together with its contents in water,. in which it is left to 
 soak. By the addition- of alcohol and the application of heat 
 the manganic acid produced is decomposed. The oxides of 
 iron and manganese are now removed by nitration, well wash- 
 ed, dissolved, without removing the filter, in a capacious dish 
 with hydrochloric acid to which a little alcohol is added; the 
 whole is now pretty well diluted and gently warmed, whilst 
 po wdered carbonate of soda is gradually added (keeping the 
 dish between whiles covered with a plate of glass), until a 
 part of the oxide of iron is precipitated, although the fluid is 
 still of a red color. A few grammes of acetate of soda are 
 now added and the mixture is heated to boiling. The oxide 
 of iron thrown down is filtered while hot, thoroughly washed, 
 again returned to the dish, dissolved in hydrochloric acid and 
 precipitated with ammonia. From the first filtrate manga- 
 nese is precipitated by boiling with an excess of carbonate of 
 soda. The carbonate of manganese is then obtained by filtra- 
 tion, washed, dried and ignited, with access of air, by which 
 it is converted into the protoxide. 
 
 The only thing, that now remains to be done, is to deter- 
 mine the amount of phosphoric acid in the first filtrate from 
 the oxides of iron and manganese. This filtrate is rendered 
 slightly acid with hydrochloric acid in a covered beaker glass, 
 and gently heated for some time in order to remove the car- 
 bonic acid, it is then supersaturated with ammonia and de- 
 composed with a clear solution of sulphate of magnesia, chlo- 
 ride of ammonium and ammonia (magnesia mixture). The 
 mixture is set aside for twelve hours. Hereupon the double 
 
68 
 
 phosphate of magnesia and ammonia is removed by nitration, 
 washed with water containing one-fourth its volume of am- 
 monia, dried, ignited and weighed as pyrophosphate of mag- 
 nesia, from which the amount of phosphoric acid is computed. 
 
 The sum of the three ingredients must be nearly equal in 
 weight to the amount of the precipitate by the ammonia u.sed 
 in the investigation. These are then computed in reference 
 to the original quantity of the latter. 
 
 D. If the iron ore contains sulphuric acid, a. special analy- 
 sis will be required for its determination. The fine powder- 
 is digested with hydrochloric acid, diluted and filtered : to t he 
 filtrate chloride of barium is added. The precipitated sul- 
 phate of baryta is washed with hot water containing at the 
 beginning a little hydrochloric acid ; it is then dried, ignited 
 and weighed. 
 
 E. The amount of the protoxide of iron present is deter- 
 mined by boiling the fine powder with hydrochloric acid, ta- 
 king care to exclude the air. The solution is then mixed 
 with a large quantity of water and then submitted to the vo- 
 lumetric analysis with permanganate of potassa. If such an 
 ore contains the oxide of manganese, too small a quantity of 
 the protoxide of iron will be indicated, and if the amount of 
 the manganese is so large as to give rise to the evolution of 
 chlorine when treated with hydrochloric acid, there will, na- 
 turally, be no protoxide of iron in the solution. 
 
 4. MAGNETIC IRON OKE MAGNETITE. 
 
 (SCALES OF IRON FROM THE FORGE.) 
 
 The combinations or mixtures of the sesquioxide and prot- 
 oxide of iron are examined either with the aid of the volu- 
 metric methods or alone afterwards. 
 
 In the first case the fine powder is heated in a flask with 
 hydrochloric acid. If there is a residue, it consists of for- 
 eign admixtures (rocky materials, &c.). The mixture is dilu- 
 ted and filtered. The residue ie thoroughly washed and it 
 
weight is determined. The .solution is completely oxidized 
 by heating it with a little chlorate of potassa until chlorine 
 begins to be manifest by the smell ; the iron is then precipi- 
 tated by means of ammonia, it is afterwards removed by fil- 
 tration, washed with hot water, dried and ignited. A fresh 
 
 quantity of the substance is dissolved in hydrochloric acid* 
 taking care to exclude all access of air; ahd the quantity of 
 protoxide of iron is determined by permanganate of potassa 
 by the volumetric method. The mode of computing it will 
 be found, page 63. 
 
 In the second case the process is exactly the same as the 
 one given, page 64 1. B. 
 
 5. ANALYSIS OF THE OXIDE OF MANGANESE. 
 
 (HAUSMANNITE, PYROLUSITE, BLACK OXIDE OF MANGANESE, 
 RED OXIDE OF DO.) 
 
 Amongst the natural oxides of manganese it is well known, 
 that pyrolusite(polianite) is the richest in oxygen, it is the bin- 
 oxide of mangnese, the so-called black oxide of manganese. 
 Nevertheless the black oxide of manganese of commerce i> 
 seldom pure pyrolusite ; and even the best sorts contain a cer- 
 tain amount of other intermingled ores of manganese, a cir- 
 cumstance occurring in a still higher degree with the inferior 
 sorts. These mixed ores of manganese (psilomelane, manga- 
 nite, hausmannite, brauuite) contain lower degrees of oxida- 
 tion of manganese, or consist of such, that is, they are always 
 less rich in oxygen than pyrolusite or the normal black oxide 
 of manganese. ISTow the value of the black oxide of manga- 
 nese is proportional to the quantity of oxygen over and above 
 that contained in the protoxide, and which is expressed by 
 the equivalents of chlorine evolved by means of this substance. 
 
 But the value of black oxide of manganese is naturally di- 
 minished also by the admixture of porphyry, quartz, heavy 
 spar, or by calcareous spar, bitter .spar (dolomite), red oxide 
 and brown oxide of iron, in the first place especially on ac- 
 count of the smaller amount of the binoxide of manganese: 
 
TO 
 
 and, secondly, if calcareous >par, bitter spar, that is, the car- 
 Ivonate of lime and magnesia, or the oxides of iron are present, 
 owing to the greater expense for the acids required in the 
 preparation of chlorine from such specimens of the black oxide. 
 
 Strictly speaking, the value of the oxides of manganese 
 inust be computed in reference to the quantity of acid requi- 
 red to produce a given quantity of chlorine. The two for- 
 mulas stand as follows: 
 
 MnOo-f 2H( }[= Gl,MnCl + 2HO 
 MiiaO s + 3 ITC1 = Cl,MnCl -I- 3HO, 
 
 and show, that a specimen of black oxide oi manganese, con- 
 taining only the protoxide, will require one time and a half 
 as much acid for the production of a given quantity of chlo- 
 rine, as a mineral consisting of the binoxide. Upon the whole, 
 however, this difference is scarcely taken into consideration. 
 
 The numerous analyses of the ozides of manganese, already 
 known, all aim to determine the amount of available oxygen 
 either alone or its equivalent in the form of chlorine, carbo- 
 nic acid, <fcc. They are partly volumetric analyses, and part- 
 ly analyses by weight. We give below the best. 
 
 A, ANALYSIS BY WEIGHT, 
 
 Oxalic acid, 2 O 3 , is converted by the higher oxides of man- 
 ganese, into carbonic acid, whereby these are reduced into the 
 state of protoxides. 
 
 MnOo + ( V J 8 =MnO + 2CO 2 
 Mn 2 O 8 -f 0*03= 2MnO 4- 2CO 2 . 
 
 Since MnO 2 is equivalent to MnO + O and Mn 2 O 3 to 2MnO 
 -f O, each atom of oxygen (8 at. wt.) will give rise to the ev- 
 olution of two atoms of carbonic acid (44 at. wt,), that is, one 
 part of oxygen corresponds to five parts and one half of car- 
 bonic acid, or one part of carbonic acid to 0.1818 of oxygen. 
 
 From one to two grammes of finely powdered black oxide 
 of manganese are placed in the flask of a Geisler's carbonic 
 acid apparatus, to this- arc added about two and a half times 
 its amount of simple oxalate of potassa and a little water ; the 
 
upper part of the apparatus is then filled with concentrated 
 sulphuric acid and adjusted in its place: the whole apparatus 
 is now accurately weighed in a proper balance. Sulphuric 
 acid is now allowed to flow drop by drop into the flask so that 
 the carbonic acid may be evolved quietly and uniformly un- 
 til the black color of the oxide has disappeared. A gentle 
 heat is now applied, the carbonic acid in the apparatus is dis- 
 placed with air and the loss of weight is determined. 
 One part of carbonic acid = 0.1818 of oxygen, 
 
 = 0.8068 of chlorine. 
 
 . 
 
 B, Volumetric ANALYSIS, 
 a) Bi XSKN'S ANALYSIS, 
 
 This method is applicable not alone for technical pin-poses 
 but also for the most scientific investigations. Tt has been 
 described, page 45. 
 
 From 0.1. to 0.5 grammes of finely powdered black oxide of 
 manganese are placed in a flask, which is two-thirds full of 
 tolerably concentrated hydrochloric acid; the chlorine evolv- 
 ed is made to pass through a solution of iodide of potassium ; 
 each equivalent of which liberates an equivalent of iodine, 
 the amount of which is computed exactly as indicated in the 
 analysis referred to. 
 
 Since the equivalents of iodine and chlorine are in the 'ratio 
 of 127 : 35.5 or of 1 : 0.28, and the equivalents of iodine and 
 oxygen as 127 : 8 or of 1 : 0.063 ; therefore for each part of 
 iodine indicated by the analysis, 0.28 parts of chlorine or 
 0.063 parts of oxygen must be computed. (One part of oxy- 
 gen =15. 875 parts of iodine ; and one part of pure MnO 2 = 
 0.372 parts of oxygen =5. 9055 parts of iodine. On this ac- 
 count an excess of black oxide of manganese is not required.) 
 
 b) A:N T ALYSIS WITH PERMANGANATE OF POTASSA. 
 
 Ten grammes of pure, dried, crystallized double sulphate 
 <)f iron and ammonia are dissolved in water, to this the weigh- 
 ed (about 0.5 gnu.), finely powdered hlack oxide of manganese 
 
72 
 
 is added, together with a little hydrochloric acid ; a gentle- 
 heat is applied so that the oxide is dissolved without setting" 
 chlorine free. The solution is then poured into a large quan- 
 tity of water, which is next placed beneath the burette con- 
 taining permanganate of potassa, in order to determine the 
 quantity of protoxide of iron not used for oxidation. 
 
 Since in this process each atom of oxygen (8) in the oxide 
 of manganese, which is set free, converts 2 atoms of the prot- 
 oxide of iron into the sesquioxide, it requires also 2 atoms of 
 the double salt, whose equivalent is seven times that of iron, 
 that is , 196, consequently 2 equivalents =392. Since 
 
 8 : 392:: 1 : 49 
 
 one part of oxygen (=4.4375 parts of chlorine) requires pre- 
 cisely 49 parts of the salt which is to be oxidized. 
 
 If 0.5 grin: of binoxide of manganese had been taken in the 
 experiment, this quantity containing 0.186 of disposible ox- 
 ygen, consequently 9.114 grms: of the salt would be used ; 
 hv the permanganate solution, therefore, 0.886 of the salt (= 
 0.12657 of the iron) would be indicated. If the strength of 
 the permanganate solution is made in reference to metallic* 
 iron, the amount of iron indicated by the number of centime- 
 tres used, has to be multplied by 7, in order to determine the 
 corresponding amount of the salt. This is deducted from the 
 quantity started with ; the differeiice(=d) is the oxidized por- 
 tion, the one-forty-ninth part of which is equal to the quant- 
 ity of oxygen used. 
 
 Ln order to compute the amount of chlorine, which is to 
 serve as standard for the goodness of the oxide of manganese. 
 the quantity of oxygen found has to be multiplied only by 
 4.4S75 (the ratio of the equivalents of oxygen and chlorine 
 being as 1 to 4.4375). 
 
 TESTING FOE MOISTUBE. 
 
 The powdered mineral is dried in the drying apparatus at 
 a temperature of from 130F. to 248F. 
 
73 
 6. PSILOMELANE. 
 
 We select psilomelane as an example for the complete ana- 
 lysis of an ore of manganese, because it either contains or may 
 contain besides manganese and oxygen, also the oxides of iron, 
 copper, nickel, cobalt, as well as baryta, lime and water. 
 
 From 5 to 10 grammes are levigated into an impalpable 
 powder and preserved in a stoppered vessel for future use. 
 
 A. -One part is dried in the desiccator in order to deter- 
 in ine the amount of hydroscopic water; the results are com- 
 puted in reference to the substance so dried. 
 
 ]>. -About two grammes are dissolved in hydrochloric 
 acid ; and the solution is diluted. If there is any residue, it 
 is removed by filtration, and its amount determined; it con- 
 sists of rocky material, quartz, or feldspar. To the solution 
 dilute sulphuric acid is added; the sulplrate of baryta re- 
 mains on the filter ; it is washed, dried, ignited and weighed $ 
 from this weight the amount of baryta is computed. To the 
 filtrate hydrosulplmrie acid is added; this separates a small 
 quantity of sulphide of copper, which is removed by filtration; 
 ir is ignited in a small porcelain crucible and weighed as 
 oxide of copper. Ammonia is now added in excess to the 
 filtrate, and afterwards the proper amount of sulphide of am- 
 monium ; the mixture is digested for some time and then set 
 aside tj settle; the fluid portion must have a slight yellow 
 tinge. The precipitate contains in addition to the sulphide 
 of manganese also sulphide of nickel and of cobalt. This, in 
 like manner with sulphide of copper, is filtered in a covered 
 filter; it is then washed quickly and without interruption in 
 cold water containing a few drops of sulphide of ammonium; 
 the filtrate is preserved in a dish. The filter is now remov- 
 ed from the funnel, and placed in a beaker glass, water is first 
 added and then hydrochloric acid ; the mixture is digested un- 
 til all smell of hydrosulphuric acid has been driven oif, then 
 filtered into a capacious dish. On the filter will be found the 
 undissolved sulphides of nickel and cobalt mixed with sulphur, 
 10 
 
74 
 
 these are dried and ignited ; and the residue is weighed. It 
 consists of the oxides of cobalt and nickel, whose separation, 
 owing to their small quantity, is of no avail, although it is ne- 
 cessary to show their presence qualitatively. The solution in 
 hydrochloric acid contains the protoxides of manganese and 
 iron. A small quantity of chlorate of potassa is added to the 
 solution which is now heated in order to oxidize the iron ; and 
 when this is effected, a small quantity of alcohol is added to 
 remove the free chlorine. The solution is now diluted with 
 much water, and the two metals are separated by carbonate 
 and acetate of soda (vide page 65). 
 
 OBSERVATION. The smaller the quantity of iron in proportion to that of the man : 
 
 ganese, by so much the more difficult it is to hit the right quantity of carbonate of soda 
 to be added. Several of the ores of manganese are quite free from iron ; in this case it 
 will, of course, not be necessary to add chlorate of potassa and alcohol ; the solution is 
 supersaturated in a dish with carbonate of soda and heated to boiling, and the carbonate 
 of the protoxide of manganese is converted by ignition with access of air into the red 
 oxide of manganese (=A^jiO-r-Mu2O3) from which the amount of manganese or of its 
 protoxide may be computed. 
 
 The filtrate from the sulphide of manganese is rendered 
 slightly acid with hydrochloric acid and concentrated by eva- 
 poration ; the sulphate produced is filtered off; the filtrate is 
 next supersaturated with ammonia and reduced with pure 
 oxalic acid. By this operation a small quantity of oxalate of 
 lime is precipitated, which, after standing for some time, is 
 removed by filtration, washed in hot water, dried and heate-i 
 to an intense heat, whereby it is converted into caustic lime. 
 
 The filtrate is evaporated to dryness ; and then weighed. 
 It consists of sulphate of lime, which contains a small quanti- 
 ty of sulphate of magnesia (also perhaps a trace of manga- 
 nese) . If it is intended to determine these metals, the mass 
 is treated in a crucible with water and carbonate of soda, the 
 mixture is then raised to boiling, after which the insoluble 
 portion is placed on a small filter. After ignition and weigh- 
 ing, it is allowed to stand in water containing a few drops of 
 nitric acid. The sesquioxide of manganese remains undissolv- 
 ed ; from its weight that of the magnesia is deducted, and by 
 computing both of them as sulphates of the protoxide of man- 
 ganese and of magnesia and deducting their amount from the 
 
75 
 
 entire amount, that of the sulphate of potassa is obtained. 
 
 C. DETERMINATION OF THE AMOUNT OF OXYGEN. In order 
 to ascertain the degree of oxidation of the manganese, or ra- 
 ther, how much oxygen is at the same time combined with 
 protoxide of manganese, proceed as directed in 5. A. or B. 
 
 I). DETERMINATION OF THE AMOUNT OF WATER. A srnalJ 
 quantity of the manganese mineral is weighed and put into 
 a retort in connection with a chloride of calcium tube, which 
 is expanded into a globe in front, into whose terminal tube 
 a perforated cork is fixed to receive the neck of the retort, and 
 which has previously been weighed; the retort is now heated 
 to redness, this drives the water forward into the chloride of 
 calcium tube, which, after the termination of the operation, 
 is again weighed. 
 
 OBSERVATION. If a mineral of manganese contains a carbonate (calc spar &c.), this 
 must be removed before the analysis by means of acetic acid or very dilute nitric acid, 
 or the amount of carbonic acid must be determined by a special process (vide Carbo- 
 nates). 
 
 7. OXIDES OF LEAD. 
 
 The oxides of lead occur dfiefly in the form of litharge in 
 an impure condition, as also of minium. 
 
 a. ANALYSIS OF LITHARGE. 
 
 Litharge is not a pure oxide of lead. It contains in gene- 
 ral silicic acid, carbonic acid, oxide of copper and oxide of ir- 
 on. The examination is as follows : 
 
 1. A small quantity of the fine powder is placed in a 
 porcelain crucible and dried in the desiccator at a tempera- 
 ture of boiling water, it is weighed and then ignited ; the diff- 
 erence of weight after ignition gives the amount of carbonic 
 acid. 
 
 2. Another quantity is also dried and then dissolved ID 
 dilute nitric acid ; water is added, and the mixture is filtered ; 
 the insoluble portion is silicic acid ; to the filtrate contained in 
 a dish, a proper quantity of sulphuric acid is added, and the 
 mixture is then evaporated until the fumes of sulphuric acid 
 begin to be evolved, and is then set aside to cool, after which 
 
76 
 
 it is filtered on a weighed filter, that has previously been 
 dried at a temperature of 248F. The sulphate of lead on 
 the filter is washed in cold water, then dried in the air first, 
 end afterwards in the drying apparatus until its weight re- 
 mains constant. 
 
 A current of hydrosulphuric acid is passed through the fil- 
 trate in order to throw down the copper. Its separation and 
 further treatment will be found described, page 51. The 
 fluid portion is evaporated to a small volume and heated with 
 a few drops of nitric acid in order to convert the iron that had 
 been reduced to the state of protoxide again into that of the 
 8esquioxide; the latter is then precipitated with ammonia. 
 
 OBSERVATION. Litharge generally contains also traces of silver, which, according to 
 Kersten, exists in the form of oxide and may be removed by repeated digestion with 
 ammonia. -The solution thus obtained is neutralized with hydrochloric acid, whereby 
 the silver is thrown down as chloride. Nevertheless the amount of silver in most speci- 
 mens of litharge is so small, as almost to be lost in the investigation. 
 
 The litharge formed at the beginning of the operation, the SCUM so called, generally 
 contains a considerable amount of antimony, probably of antimonious acid, which, when 
 treated with nitric, acid, remains behind. 
 
 For the mere determination of the amount of copper, the litharge in the state of a fine 
 powder may be digested in a solution of carbonate of ammonia, which dissolves the ox- 
 ide of copper. The solution is then evaporated to dryness ; and, after ignition, the resi- 
 due is weighed as oxide of copper. 
 
 b. ANALYSIS CF MINIUM. 
 
 This analysis is very simple, and consists in simple ignition. 
 by means of which the minium is reduced to oxide of lead. 
 Other admixtures, which generally are those of the oxide of 
 lead from which it was originally prepared, may be determi- 
 ned by the method given in the preceding article. Besides 
 this, the specimens of commercial minium always contain a 
 quantity of oxide of lead in an uncombined condition, which 
 may be determined by digesting the minium in a solution of 
 the neutral acetate of lead ; this dissolvss the free oxide and 
 gives rise to the subacetate of lead. By washing and drying 
 the remainder, the difference of weight will determine tl;< 
 Amount in question. 
 
 
 
 8. TIN STONE, TIN ORE. 
 
 The stannic acid, that is found native, the .so-called tin 
 
77 
 
 stone, generally contains a small portion of oxide of iron ami 
 silicic acid, and is not acted upon by acids. 
 
 A. The ore is converted into an impalpable powder, 
 and then a mixture of one part of this powder, three parts 
 of dried carbonate of soda and three parts of sulphur well mix- 
 ed together are placed in a porcelain crucible and melted over 
 a spirit lamp, taking care to cover the crucible. As soon as the 
 crucible is cold, it is placed in a vessel of water which is heat- 
 ed until the con tents (the tin as sulphide of tin) dissolves; the 
 iron (and perhaps also one or more of the other electropositive 
 metals) remains behind as the black sulphide, which is remo- 
 ved by filtration, washed, and, after drying in the air, is igni- 
 ted in order to convert it into the sesquioxide of iron. (If the 
 quantity is somewhat considerable, the sulphide of iron IB 
 dissolved in hydrochloric acid, oxidized and precipitated by 
 ammonia). Sulphide of tin in the diluted alkaline solution 
 is thrown down by dilute sulphuric acid (vide page 59). 
 
 B. There is another method of analyzing tin ore , which 
 may be regarded at the same time as an example of assaying, 
 it consists in placing a given weight of the line powder in a 
 weighed porcelain crucible and then reducing the ore by 
 means of a current of hydrogen ; but the heat in this case 
 must be a vivid red and continued for a long time. After 
 the contents are cold, they are weighed (the vessel being till- 
 ed with hydrogen gas), and the operation is repeated until 
 there is a certainty that the reduction is complete. From the 
 loss of weight the amount of stannic acid may be determined 
 (one equivalent of oxygen being equal to 4.673 parts of stan- 
 nic acid), when there is no other oxide present, which might 
 be reduced at the same time. Since, however, tin almost al- 
 ways contains the oxide of iron, a part of the loss must be 
 attributed to this; on this account the crucible together with 
 its contents (the reduced tin) is digested in a beaker gln^ 
 with hydrochloric acid. As soon as the solution is coniplet* . 
 it is diluted, and filtered into a capacious flask. The amount 
 of residue (silicic acid &c.) is then determined. Ammonia is 
 
78 
 
 added in excess to the filtrate, and then a sufficient quantity 
 of sulphide of ammonium containing so much of the flower* 
 of sulphur a,s to convert the protosulphide of tin into the sul- 
 phide, which must dissolve in the sulphide of ammonium. 
 The contents are now digested in a loosely stoppered flask un- 
 til the black oxide of iron is completely separated from the 
 yellow fluid ; they are then filtered in a covered funnel. The 
 residue is washed in water containing a little sulphide of am- 
 monium, then dissolved in hydrochloric acid, oxidized with 
 chlorate of potassa, and finally the sesquioxide of iron is. 
 thrown down by ammonia. 
 
 If it is desirable to determine the amount of tin right away, 
 it is precipitated as sulphide by dilute sulphuric acid, &c. 
 
 If the amount of oxide of iron is known, the weight of the 
 oxygen which it would lose in the reduction (10 parts of the 
 sosquioxide of iron = 3 parts of of oxygen) has to be deducted 
 from the whole weight ; the difference is the weight of the 
 oxygen of the stannic acid. 
 
 It is much easier by this method to determine the amount 
 <>f silicic acid <fcc. than by the method given in A., where it 
 is rendered soluble by fusion with carbonate of soda, or has- 
 to be sought for in the sulphide of iron (which may also con- 
 tain some undecomposed substance). 
 
 ;; : ; ( ; ';;; EL SULPHUR METALS. 1 
 
 [n general there are two distinct methods for the analysis 
 of sulphur metals : treatment with aqua regia, or with chlo- 
 rine. Although in principle they are both the same, they 
 differ in their execution. When sulphurets are treated with 
 nitro-hydrochloric acid, the metals are converted into chlo- 
 rides, and the sulphur into sulphuric acid ; the solution, there- 
 fore, contains at the same time metallic colorides and sul- 
 phates. But this complete conversion of sulphur into sulphu- 
 ric acid can be effected only when its amount is not very 
 or when a large quantity of aqua regia is employed to 
 
79 
 
 effect the decomposition of the substance and the digestion 
 is continued for a long time. In by far the most numerous 
 cases a part of the sulphur is separated as such, and can in 
 this form be determined ; whilst on the other hand the sulphu- 
 ric acid, that is produced during the operation, is precipita- 
 ted by a salt of barium, and from the amount of the sulphate 
 of baryta the oxidized sulphur is computed. 
 
 This method, however, can not be followed, when one of 
 the metals contained in the substance gives rise to a chloride 
 which is either with difficulty soluble or altogether insoluble 
 in the acid, as, for instance, silver and lead\ for in this ease 
 the chloride is then mixed with the separated sulphur. In 
 like manner the method can not be applied when a metal pro- 
 duces in combination with sulphuric acid a compound insolu- 
 ble in acids, as, for instance lead ; for in this case, when sul- 
 puret of lead is treated with nitro-hydrochloric acid there is 
 produced by, oxidation a mixture of sulphur, sulphate of lead 
 and chloride of lead in the residue. 
 
 In all these cases (and it is just these that occur so frequent- 
 ly, from the fact that a large number of the most important 
 minerals and furnace-products may be placed in this catego- 
 ry) dry chlorine gas is used for the decomposition, as taught 
 by Berzelius. If a sulphide is treated with chlorine, the ba- 
 ses are all converted into chlorides ; and since the chlorides 
 of sulphur, of antimony and of arsenic are volatile, they can 
 easily be separated by heat from the non- volatile chlorides of 
 silver, copper, lead &c. 
 
 A. ANALYSTS OF SULPHUR METALS BY AQUA REGIA. 
 
 The sulphur compound to be analyzed must be reduced into 
 an impalpable powder, in order to facilitate the decomposi- 
 tion. If this is neglected, the mineral will have frequently 
 to be digested a very long time, and especially this .is tbe 
 case owing to the tact, that the sulphur, as it separates, sur- 
 rounds the last remaining heavy parts of the powder and pro- 
 tects them from the action of the acid. 
 
80 
 
 It is always most convenient to make these solutions in 
 small flasks. 
 
 As a rule the substance is first covered with concentrated 
 mtric acid, added drop by drop, in order that the operation 
 may not be too violent, for which reason also it is generally 
 not advisable to \&v fuming nitric acid for this purpose. It 
 is only when the operation begins to cease, that heat is appli- 
 ed, arid continued for some time, still the solution must not 
 be made to boil. As soon as red vapors cease to be given off, 
 an equal volume or even double the volume of concentrated 
 hydrochloric acid is added gradually to the mixture ; the di- 
 gestion is continued until the sulphur appears on the surface 
 of the fluid in the form of one compact yellow mass, or of se- 
 veral yellow drops, nevertheless in all this operation a conti- 
 nued boiling must be avoided. Sometimes it happens, that 
 the sulphur remains all the while of a grey color, especially 
 in single places ; in this case the fluid is diluted, and the sul- 
 phur is allowed to settle completely to the bottom ; after 
 which the fluid portion is carefully separated as much as pos- 
 sible by decantation ; the' sulphur is again digested with aqua 
 regia, but in smaller quantity than before, until its color chan- 
 ges to pure yellow. 
 
 It is not advisable to dry the still impure sulphur without 
 this mode of purification, in order to determine its weight, 
 and afterwards to burn it ; for the difference in weight does 
 not determine its amount, because by ignition the remaining 
 part of the sulphur compound undergoes a change, a part of 
 its sulphur passing of as sulphuric acid, whilst the remainder 
 represents a mixture of metallic oxides arid basic sulphates. 
 
 A sulphur compound must never bejirst treated with hy- 
 drochloric acid, because this mode of treatment almost always 
 gives rise to the evolution of hydrosulphuric acid, which thus 
 causes a loss of sulphur. This is especially the case with those 
 sulphides, which are decomposed with diluted acids, as, for 
 instance, the sulphides of iron, zinc, manganese, even of nick- 
 el arid cobalt. With some of these, indeed, as with zinc and 
 
81 
 
 manganese, a quantity of hydrosulphuric acid is evolved, even 
 when nitric acid is employed first; on which account it is 
 well to prepare a mixture of equal volumes of concentrated 
 nitric and hydrochloric acids, and to heat them together un- 
 til the color changes to yellow, and chlorine begins to be 
 evolved. Similar precautions must be taken in the analysis 
 of the sulphides of the alkalies and the alkaline earths. The 
 removal of the sulphur, that has been separated from the 
 metallic solution, from the fluid after dilution, varies with its 
 coherence. If the sulphur is in the form of one or of several 
 globular shaped pieces, the fluid is carefully decanted, and 
 the residue is washed repeatedly with water, and is then 
 transferred to a very small porcelain crucible, in which it is 
 dried at a gentle heat in the air. If, on the contrary, the 
 sulphur is divided up into very small flocculent pieces, it is 
 collected on a filter, previously dried at a temperature of 212 
 F., and, as soon as the contents have been dried in the air, 
 they are transferred to the drying apparatus and submitted 
 to a temperature of 212 until, after repeated weighing, the 
 crucible containing the filter and its contents no longer indi- 
 cates any further loss of weight. 
 
 To be quite safe on this score it is' well to ignite the sul- 
 phur, thus obtained, in a small porcelain crucible. It some- 
 times happens that there is a residue rising from some unde- 
 composed sulphide. The quantity of this must not exceed a 
 few milligrammes up to a centigramme, if the determination 
 of the amount of sulphur is to be tolerably free from error. 
 Ets weight is deducted from that of the sulphur. The sub-- 
 stance itself is then dissolved in a few drops of nitro-hydro- 
 chloric acid, and the solution is added to that already obtain- 
 ed. It is much more common, however, for the residue, 
 
 obtained after all the sulphur has been burned off, to consist 
 of quartz, sulphate of baryta &c., in which case its weight 
 has to be deducted from that of the sulphur as also from that 
 of the substance originally started with. 
 
 The sulphuric acid in the acid is now precipitated by chlo- 
 * 11 
 
82 
 
 ride of barium, taking care to avoid any great excess of the 
 precipitant. As soon as the fluid has become perfectly clear, 
 it is filtered. There is less fear in this case than in others that 
 the filtrate will be turbid (vide the section on the Sulphates). 
 The sulphate of baryta is thoroughly washed in hot water and, 
 after drying, it is ignited and weighed. 
 
 Since the fluid in general contains nitric acid, especially 
 when the compound was one of those, that are not easily de- 
 composed, and on this account required a greater excess of 
 acid, it happens, that a small quantity of nitrate of baryta is 
 precipitated together with the sulphate of baryta, and is not 
 easily thorougly removed even by continued washing in hot 
 water; nor even is it removed, when the precipitate, after se- 
 paration by liltration, is digested for some time in a beaker 
 glass with hydrochloric acid. 
 
 The amount of barytic salt adhering is indeed so slight as 
 in many instances to be neglected; but in very accurate in- 
 vestigations this amount must be taken into consideration. 
 In this case the sulphate of baryta, after ignition in the cru- 
 cible, is covered with concentrated hydrochloric acid; the 
 crucible is then covered and heated to boiling ; the contents 
 are diluted with water'and submitted to filtration, after which, 
 the residue on the filter is thoroughly washed in hot water. It 
 will be always observed in such instances, that the filtrate 
 b.ecomes turbid on the addition of sulphuric acid, owing to 
 the fact, that by ignition the nitrate of baryta had been con- 
 verted into caustic baryta and carbonate of baryta, which 
 combine with effervescence with this acid. 
 
 The metallic sulphides, especially those, that easily evolve 
 hydrosulphuric acid when treated with acids, may be opera- 
 ted upon in the following manner : pulverized chlorate ofpo- 
 tassa is added to the fine powder in a long necked flask, to 
 which small quantities of hydrochloric acid are gradually 
 poured. The flask is set aside for twenty-four hours and af- 
 terwards placed in the digester and heated at a temperature 
 of between 86 and 104F. By this proceeding, as a rule, the 
 
 
 
83 
 
 whole amount of the sulphur is dissolved and converted into 
 sulphuric acid. Water is now added, and the mixture is heat- 
 ed, until the whole amount of chlorate of potassa is decom- 
 posed, which is shown to be effected when the addition of 
 inori: acid no longer evolves the smell of chlorine. When- 
 tliis is the case the sulphuric acid is precipitated with chlo- 
 ride of barium. 
 
 Now before the separation arid determination of the me- 
 tals can be effected, the excess of baryta contained in the fluid 
 must, be removed by the gradual addition of dilute sulphuric 
 acid, taking care also here to avoid a large excess. The fil- 
 tration can take place in this instance before the fluid has be-, 
 come quite clear, as soon as there is a certainty of having ad- 
 ded a sufficient quantity of sulphuric acid, because there is 
 no fear of turbidity caused by the particles of the precipitate 
 passing through the fibres. 
 
 We will now proceed to a few of the more frequently oc- 
 curring examples, to which the preceding described opera- 
 tions are supposed to have been applied. 
 
 1. SULPHUR AND IEON. 
 
 (IKON PYKITES, SPEAR PYRITES, MAGNETIC PYRITES.) 
 
 The iron is precipitated by means of ammonia. As regards 
 the determination of the sulphur, and the precipitation of the 
 sesquioxide of iron by means of ammonia, all this has already 
 been given in detail, it will, therefore, be unnecessary to re- 
 peat the processes here. 
 
 2. SULPHUE, COPPER, IRON. 
 
 (COPPER PYRITES, VARIEGATED COPPER, MOST OF THE SPECI- 
 MENS OF COPPER GLANCE, AND SEVERAL COPPER ORES. 
 
 A current of hydrosulphuric acid is passed through the ac- 
 id solution, with all the precautions indicated in reference to 
 the separation of copper and zinc (vide page 51). 
 
 The precipitated sulphide of copper is converted into the 
 
84 
 
 disulpliide (vide page 52). 
 
 The filtrate from the sulphide of copper is so far evapora- 
 ted in a dish, until the iron is again oxidized by the nitric 
 acid still remaining, or, if this be not the case, by the addi- 
 tion of a few drops of nitric acid or of chlorate of potassa. 
 The contends are then placed in a beaker glass, to which am- 
 monia is added in order to precipitate the sesquioxide of iron 
 which, after iiltration, is thoroughly washed, dried and ig- 
 iiited. 
 
 3. SULPHUR, COPPER, IRON, ZINC. 
 
 (SEVERAL VARIETIES OF COPPER ORE.) 
 
 The process is exactly the same as in No. 2. The filtrate 
 from the sesquioxide of iron is treated with sulphide of am- 
 monium, and then set aside to settle; it is then submitted to 
 a rapid filtration in a covered funnel ; the sulphide of zinc 
 is well washed in several waters, and then placed together 
 with the filter in a beaker glass, and digested with hydrochlo- 
 ric acid until all smell of hydrochloric acid has passed off. 
 The solution is much diluted and filtered into an evaporating 
 dish ; the residue is well washed on the filter, supersaturated 
 with carbonate of soda, taking care that there is no loss by 
 effervescence, and finally heated to boiling. The carbonate of 
 zinc is then obtained' by filtration, washed in hot water, dried 
 and ignited, whereby it is converted into the oxide of zinc. - 
 The filtrate is then tested with a few drops of sulphide of am- 
 monium to ascertain its perfect freedom from all zinc. 
 
 OBSERVATION. This method is not quite accurate, because, although the oxide of 
 
 sine alone is easily soluble in ammonia, a small quantity remains with the oxide of ir- 
 on. Those cases where greater accuracy is necessary, the separation is effected accord- 
 ing to No. 4. 
 
 Several copper ores contain small quantities of nickel (co- 
 balt). This is found in the course of the analysis with oxide 
 of zinc and is weighed in mixture with it as oxide. If the 
 quantity of both is large enough to admit of a quantitative, 
 separation, they are dissolved in hydrochloric acid and sepa- 
 rated according to the instructions given, page 54. 
 
85 
 
 If the copper contains lead, the analysis must be effected 
 with chlorine. 
 
 4. SULPHUK, ZINC, IBON, CADMIUM. 
 
 (ZlNC BLENDE.) 
 
 In the filtrate from excess of baryta a current of hydrosul- 
 phuric acid is passed in order to separate the cadmium. The 
 yellow residue is obtained from the fluid by filtration ; the 
 fluid itself is received in a dish and evaporated to dryness. 
 
 The sulphide of cadmium on the filter is thoroughly wash- 
 ed ; it contains some free sulphur owing to the reduction of 
 the sesquioxide of iron, and is treated exactly in the same 
 way as sulphide of zinc precipitatd by sulphide of ammoni- 
 um. The acid solution is precipitated with carbonate of po- 
 tassa as indicated, page 60. 
 
 The amount of cadmium is generally very trifling in zinc- 
 blendes ; its determination, therefore, is more simply effected 
 out of the sulphides. The latter is then filtered on a weighed 
 filter, which has recently been dried at a temperature of 248 : 
 it is thoroughly washed, and, the tube of the funnel being 
 closed with a small cork, it is treated with sulphide of ammo- 
 nuiin previously heated, and set aside, covered up with a plate 
 of glass for a quarter or half an hour. By this means the sul- 
 phur, mechanically mixed up with the solution, is separated 
 and drawn out. The cork is now removed and the fluid is 
 allowed to filter off; the sulphide of cadmium on the filter IB 
 well washed, dried in a warm place and afterwards in the dry- 
 ing apparatus, until it no longer undergoes any loss of weight. 
 
 The filtrate from the sulphide of cadmium is mixed with 
 a small quantity of sulphuric acid and evaporated until the 
 hydrochloric acid, and especially the nitric acid, has all passed 
 off. By this means the iron, which by the hydrosulphuric 
 acid had been reduced to the state of protoxide, will be again 
 oxidized (towards the end of the operation it may be found 
 necessary to add a little nitric acid to aid in the oxidation). 
 
80 
 
 The residue is now highly diluted and then the oxide of iron 
 is separated from the oxide of zinc by means of carbonate of" 
 soda and acetate of soda, as already indicated in reference to 
 oxide of iron and protoxide of manganese (vide page 67). 
 
 5. SULPHUR, ANTIMONY, IEON. 
 
 (ANTIMONY GLANCE, GRAY ANTIMONY, BERTH IEKITE.) 
 
 The substance is oxidized in the usual way in a flask by 
 means of nitro-hydrochloric acid, As soon as the separated 
 sulphur appears pure, a small quantity of a solution of tarta- 
 rie acid, and then water are added ; the sulphur is then remo- 
 ved, the quantity of sulphuric acid determined, and the ex- 
 cess of baryta again separated by means of sulphuric acid. 
 The introduction of the tar taric acid is intended to prevent the 
 precipitation of the solution of antimony, which otherwise 
 would take place. In this case the sulphate of baryta has to 
 be filtered off as quickly as possible ; it is afterwards thorough- 
 ly washed with boiling water, ignited and examined in the 
 manner recommended, page 82. 
 
 A continuous current of hydrosulphuric acid is now passed 
 through the solution in order to throw down the antimony, 
 until the fluid has a strong smell of the gas : the solution is 
 then allowed to digest at a gentle heat for some time. The sul- 
 phide of antimony is received on a filter, previously dried at a 
 temperature of 248F. and weighed ; it is then washed tho- 
 roughly in cold water, dried in the air and finally in the dry- 
 ing apparatus at a temperature of 248F. to 266F. until it 
 remains constant. 
 
 The amount of antimony is determined according to the 
 method given, page 62. 
 
 Or a part of the dried sulphide of antimony is placed in a 
 weighed porcelain crucible, which is then covered and main- 
 tained at a temperature of 300F. to 446F. until the weight 
 no longer undergoes any change. By this means the excess 
 of sulphur has been volatilized, and pure SbS 8 remains in the 
 -rueible. 
 
87 
 
 The fluid filtered from the sulphide of antimony, contain- 
 ing iron, is supersaturated with ammonia and then precipita- 
 ted with sulphide of ammonium. Since the sulphide of iron, 
 by reason of the presence of tartaric acid, is partly suspended 
 in the fluid in a very fine state of division giving it a greeji 
 color, it must be digested in the covered vessel in the sand- 
 bath, until the precipitate has thoroughly settled, and the so- 
 ution is clear and of a yellowish color. The sulphide of iron 
 is then filtered as quickly as possible, taking care to cover 
 the funnel during the filtration, in order to obviate the oxi- 
 dation of the precipitate when exposed to the air ; the resi- 
 due is washed for only a short time, and is then allowed to 
 dry in the funnel. After this the dry filter is folded toge- 
 ther, placed in a porcelain crucible and heated with a small 
 quantity of tartaric acid, which in general still adheres to the 
 paper, until both are carbonized, taking care however to avoid 
 an intense ignition. The mass is now digested in a small 
 flask with a little aqua regia, with which the crucible had 
 been previously washed out. As soon as the solution is ef- 
 fected, it is diluted, and then filtered in order to remove the 
 charred residue; after which the sesquioxide of iron is preci- 
 pitated with ammonia. 
 
 OBSERVATION 1. - If a compound of this kind contains at the same time zinc and 
 manganese, these metals are precipitated, together with the iron as sulphides. In thie 
 case the process is the same as in the preceding example, dissolving the sulphides in ni- 
 tro-hydrochloric acid, and evaporating the solution together with sulphuric acid, and eo 
 on. Thus, by means of carbonate of soda, a mixture of carbonate of zinc and carbonate 
 of the protoxide of manganese is obtained, which is then dissolved in an excess of 
 acetic acid ; from this solution the zinc is precipitated by hydrosulphuric acid. The sul- 
 phide of zinc is treated as indicated, page 54 ; and carbonate of soda throws down the 
 manganese from the filtrate, after all traces of hydrosulphuric acid have been removed. 
 
 OBSERVATION 2. - Native sulphide of antimony generally con tains sulphide of lead. 
 In this case an accurate determinatoin can be attained only by analysis with chlorine. 
 
 B. ANALYSIS OF METALLIC SULPHIDES BY CHLORINE. 
 
 We are indebted to Berzelius for this excellent 
 which finds an application not only when combinations of me- 
 tallic sulphides containing lead or silver. but also when am> r 
 nic and antimony are to be analyzed. It requires, however. 
 attention and foresight 
 
88 
 
 The requisite apparatus consists in the first place of a capa- 
 cious flask containing a mixture of two parts of black oxide 
 of manganese, three parts of common salt and dilute sulphu- 
 ric acid (from one and one-half to two parts of water to one 
 of acid). The success of the analysis depends upon the uni- 
 form evolution of gas, and this on the proper dilution of the 
 acid. If the latter is too concentrated, the mixture becomes 
 very frothy and is apt to foam over, and at all events produ- 
 ces such a boisterous evolution of gas as to cause apart of the 
 volatile chloride to be lost. If, on the other hand, the mix- 
 ture is too dilute, the current of gas is too slow and the fluids in 
 the adjoining two-necked bottle are apt to rush back ; to ob- 
 viate which the mixture in the flask has to be heated so in- 
 tensely as to cause hydrochloric acid to be evolved at the same 
 time with the chlorine. The mixture must be so constituted 
 as to generate, without the application of heat, a regular cur- 
 rent of gas for one or several hours, the bubbles of gas pass- 
 ing through the fluid in the first bottle at the rate of one per 
 second, and with the slightest addition of heat much more 
 rapidly. 
 
 A cork is accurately fitted to the neck of the flask through 
 j which passes the shorter leg of a tube bent twice at right ari- 
 gies ; the longer leg passes through a cork in one of the aper- 
 tures of a two necked bottle and dips down into a layer of con- 
 centrated sulphuric acid at the bottom ; this acid shows the rate 
 lit which the gas is being evolved and also partially dries it. 
 Another tube, bent at right angles at the top and once at the 
 bottom, passes just through the cork in the second neck of the 
 Wolfe's bottle and at the bottom enters the cork of a chloride 
 of calcium apparatus. The gas, after it leaves this vessel, en- 
 ters the bulb-tube which contains the compound to be analy- 
 zed. The bulb-tube is made of glass which does not fuse at 
 a very low temperature. The bulb is capacious enough for 
 the material to be analyzed, the tube at one side is wider 
 and longer than the other. The bulb-tube is first weighed, 
 and then from one to two grammes of the impalpable powder 
 
89 
 
 to be analyzed are introduced through the wider tube into the 
 bulb. The tube is now cleaned with a feather and then the 
 whole is weighed. The narrow and shorter tube is connected 
 with the chloride of calcium apparatus ; the longer tube is bent 
 at right angles and connected with a Liebig's potassa apparatus 
 of rather large dimensions, but its legs must be parallel. This 
 vessel is tilled with dilute hydrochloric acid, to which tarta- 
 ric acid is added when the mineral contains antimony. A 
 tube proceeds from the second leg of this vessel through a 
 cork in a two-necked bottle and dips into a similar solution 
 at the bottom of the bottle from which another conducts the 
 chlorine away out of the laboratory. 
 
 As soon as all the the connections are accurately made, chlo- 
 rine gas is evolved and a current is kept up for one or seve- 
 ral hours. Gradually the apparatus becomes filled with chlo- 
 rine ; in general the temperature of the substance rises con- 
 siderably owing to the decomposition effected; and white va- 
 pors appear in the tube and first receiver. The evolution of 
 gas can be so great as to cause the first two-necked bottle to 
 be filled with vapor, and a small portion may pass into the 
 last ; but this latter event must be avoided, otherwise a part 
 might easily escape into the air. 
 
 The bulb of the bulb-tube is heated with an ordinary small 
 spirit lamp taking care not to commence this operation before 
 all the parts are tilled with chlorine. But at this point the 
 iToatest attention is required in reference to the evolution of 
 gas ; for it can easily happen to the inexperienced, that the 
 fluid can be drawn over out of the first bottle into the flask, 
 or that by a sudden formation of aqueous vapor a part of the 
 volatile chloride may pass without absorption through the re- 
 ceivers. All this is a matter of experience to know how to 
 avoid such mishaps ; for, when the flask is heated, the gas is 
 quickly absorbed and thus by condensation the fluid may rush 
 back wards; and the application of a higher temperature in 
 order to prevent this misadventure easily causes too much va- 
 por to rise, which is carried along with the gas. The aim i 
 
90 
 
 always to generate the gas so that the bubbles are perceived in 
 the two receivers at considerable intervals of time between 
 them. 
 
 The volatile chlorides, that collect in the first branches of 
 the bulb-tube, are driven forwards by means of a small spirit 
 lamp, taking care that the tube does not get stopped up 
 during the operation, a circumstance that easily occurs unless 
 the bore of the tube is pretty large. If after continuous heat- 
 ing of the bulb, (and this heat must never rise so high as a, 
 visible red heat, because it might easily happen that a part of 
 the chlorides of lead, copper &c. might be volatilized) there 
 is no longer any deposit in the adjacent parts of the tube (if 
 the substances contain iron, the chloride of iron will be con- 
 tinually formed and this is not so easily volatilized, and its 
 complete volatilization is not to be expected), and the subli- 
 mate has been driven forward by means of the blowpipe a 
 far as the cork of the receiver, the tube is scratched with a 
 sharp file a little distance above the cork and there broken 
 off by means of a point of burning charcoal. The aperture 
 of the piece remaining in the cork, however, must be closed 
 immediately with a small cork held in readiness for this pur- 
 pose. 
 
 The chloride of sulphur, that has arrived as far as the wa- 
 ter in the receiver, is decomposed by giving rise to hydrochlo- 
 ric acid, sulphur and sulphurous acid which by the free chlo- 
 rine is converted into sulphuric acid. It frequently happens 
 also, that the whole amount of sulphur is dissolved as sulphu- 
 ric acid. 
 
 After the termination of the experiments the two bottles are 
 set aside for about 24 hours, in order that the chlorides, that 
 remain in the broken tube, may dissolve. If this does not 
 take place of itself, the small cork is removed from the tube, 
 the tube itself also is drawn out of the larger cork, and then, 
 letting it drop into the receiver, the aperture is immediately 
 closed with a large cork held in readiness, in order that 
 none of the very volatile vapors may escape. After the time 
 
91 
 
 above-mentioned the contents of the receivers are poured into 
 a beaker glass, the receivers themselves are well rinced with 
 water, and the combined fluids are digested at a gentle tem- 
 perature, until the smell of chlorine has passed off. If there 
 is any tree sulphur present, it is placed on a weighed filter, 
 if not. the sulphuric acid is precipitated with chloride of bari- 
 um ; the sulphate of baryta thus produced is well washed in 
 hot water, and after ignition, if the fluid contain tartaric acid, 
 it is treated as indicated, page 83, after which the excess of 
 baryta is removed by dilute sulphuric acid. 
 
 These operations are the same in all analyses of this kind; 
 the remaining operations in reference to the separation and 
 determination of the metals will be given in detail in the fol- 
 lowing examples, in which the further treatment of the con- 
 tents is described under A. and that of the volatile chlorides 
 under B. 
 
 1. SULPHUR, ANTIMONY, SILVEE. 
 
 (RUBY SILVER, BKITTLE SILVER ORE, MIARGYRITE.) 
 
 These compounds are the most easily decomposed of all 
 others. 
 
 A. Since the bulb in this case contains only chloride 
 
 of silver, all that is necessary after the termination of the op- 
 eration is to weigh the bulb-tube together with the piece that 
 has been cut off, as soon as they have been thoroughly dried, 
 in order to ascertain the amount of chloride of silver. 
 
 B. The antimony is precipitated by hydrosulphuric ac- 
 id ; the further process is the same as indicated on page 62. 
 
 OBSERVATION. If the substance contains small amounts of copper and iron, the 
 
 mode of proceeding is that given (Grey Copper) on the next page. 
 
 2. SULPHUE, ANTIMONY, LEAD. 
 
 (ZlNKENITE, PLAGIONITE, JAMESONITE, BOULANGEBITE, FEA- 
 ER ORE.) 
 
 The analysis is exactly the same as in 1. The bulb-tube 
 ia weighed with the chloride of lead. 
 
92 
 
 If there are traces of copper and iron present, the contents 
 of the bulb are treated with a few drops of hydrochloric acid 
 and hot water, with the aim of removing them from the glass ; 
 they are then poured into a dish and heated to boiling, after- 
 wards filtered on a weighed filter whilst still hot; and the fil- 
 trate is received in a second dish containing dilute sulphuric 
 acid. The residue is boiled repeatedly with water slightly 
 acidulated, as long as chloride of lead is dissolved, it is then 
 thoroughly washed in hot water on the filter; and its weight 
 is determined. It may consist of undecornposed matter, rock- 
 y substance, or chloride of silver, which must be examined 
 more intimately. The filtrate is evaporated to dryness, the 
 lead is determined as sulphate of lead (vide page 55.); the 
 filtrate is then precipitated by hyclrosulphuricacid ; the small 
 quantity of sulphide of copper is ignited, computed as oxide, 
 and finally the filtrate, containing still a small portion of iron, 
 is added to that from the sulphide of antimony. 
 
 As soon as the antimony lias been precipitated by hydro- 
 sulphuric acid, the fluid is supersaturated with ammonia, and 
 then sulphide of ammonium is added. The sulphide of 
 iron having been completely separated by continued digestion, 
 is removed by filtration, raised to a low red heat, dissolved 
 in aqua regia, and precipitated by ammonia as oxide of iron. 
 (Compare page 87). 
 
 3. SULPHUR, ANTIMONY, LEAD, COPPER, IRON. 
 
 (BOUKNONITE.) 
 
 The analysis of this material is the same as in the preced- 
 ing example. The amount of copper, however, being some- 
 what considerable, the sulphide precipitated by hydrosulphu- 
 ric acid is converted into the disulphide of copper, as indica- 
 ted, page 53. 
 
 4. SULPHUR, ANTIMONY, ARSENIC, COPPER, 
 
 IRON, ZINC. 
 
 (GREY COPPER.) 
 
93 
 
 Chloride of silver, chloride of copper and the largest part* 
 of the chlorides of iron arid zinc will be found in the bulb of 
 the bulb-tube. The chlorides of sulphur, of antimony, of ar- 
 senic, as well as a part of the chlorides of iron and zinc are 
 volatilized. 
 
 A. Analysis of tlie Non-volatile Chlorides. The con- 
 tents of the bulb are transferred by the aid of a little hydro- 
 chloric acid and water to a beaker glass, the chloride of sil- 
 ver is then placed in a weighed filter, and the copper is pre- 
 cipitated by hpdrosulphuric acid, as indicated, page 52. The 
 filtrate from the sulphide of copper is evaporated in a dish to 
 a small amount, and afterwards added to the solution to be 
 obtained from the remaining part of the iron and zinc by the 
 B analysis. 
 
 B. Analysis of the Volatile Chlorides. As soon as the 
 
 sulphuric acid is precipitated by chloride of barium, and the 
 excess of baryta has been removed again by sulphuric acid, 
 according to processes already given, a current of hydrosul- 
 phuric acid is passed through the fluid for some time, in or- 
 der to precipitate antimony and arsenic at the same .time. 
 Since arsenic is thrown down later than antimony and only 
 slowly, the current has to be continued longer, and when the 
 fluid exhales a strong smell of the gas, it is heated on a 
 sand-bath for some time, it is then allowed to cool and after- 
 wards filtered, after which the two sulphides are thoroughly 
 washed in cold water. The filtrate may again be treated 
 with hydrosulphuric acid, and afterwards heated in order to 
 ascertain whether any more sulphide of arsenic can be obtain- 
 ed. If this should be the case, the small quantity of sulphide 
 of arsenic has to be placed on a separate filter, which is then 
 dried and weighed ; from this the amount of arsenic is com- 
 puted, taking it for granted that the arsenic exists here as 
 AsS 3 , which may be regarded without hesitation for so small 
 a quantity to be the case. 
 
 The mixture of sulphide of antimony and sulphide of arse- 
 nic, previously dried in the air, is placed together with the 
 
94 
 
 filter in a flask ; here it is oxidized either by means of aqua 
 regia, or chlorate of potassa and hydrochloric acid. If the 
 sulphur has thoroughly separated, a small quantity of tartaric 
 acid is added, then water, and the fluid is filtered. A consi- 
 <ierable quantity of chloride of ammonium is added to the fil- 
 trate, which is neutralized with ammonia in excess, the fluid 
 remaining clear all the while ; the magnesia mixture (vide 
 page 67) is now added which precipitates the arsenic acid a& 
 the double arsenate of ammonia and magnesia. The mixture 
 is now set aside in a cool place to settle, after which the res- 
 id no is placed on a dried and weighed filter, and washed with 
 old water to which one quarter of its volume of ammonia 
 lias been added. After drying in the air, it is dried at a tem- 
 perature of boiling water until its weight remains constant, 
 ft is then represented by the symbol: 2 MgO -f NH 4 O -f As & 
 4-Aq. 
 
 The filtrate is now neutralized with hydrochloric acid in 
 .vxcess ; and the antimony Is precipitated by hydrosulphuric 
 acid and treated as sulphide of antimony (vide page 62). 
 
 The filtrate from the sulphide of antimony and sulphide of 
 arsenic is supersaturated with ammonia and then precipitated 
 with sulphide of ammonium. As soon as the iron and zinc 
 have been precipitated, and by digesting with heat have tho- 
 roughly settled from the yellowish fluid, they are placed in 
 a covered filter and washed as quickly as possible with cold 
 water, to which a few drops of sulphide of ammonium are 
 added every time fresh water is used. The filter is now pla- 
 ted in a beaker glass, covered with hydrochloric acid and di- 
 gested in the sand-dath, until all is dissolved, and all smell of 
 hydrosulphuric acid has passed off. The solution is slightly 
 diluted with water then filtered and added to the filtrate ob- 
 tained from A. The iron is now oxidized by means of a lit- 
 tle chlorate of potassa and then separated from the zinc ac- 
 cording to the process given on page 67. 
 
 5. SULPHUK, LEAD, IEON, COPPEB, SILVEB, 
 ZINC, NICKEL, ANTIMONY. 
 
95 
 
 (LEAD STONE, LEAD AND COPPER STONE.) 
 Furnace products of this kind contain of metals either lead 
 and iron (lead stone) or copper, lead and iron (copper and 
 lead stone). The remaining metals exist in these products 
 to a very small extent. 
 
 The decomposition of such electro-positive sulphides by 
 means of chlorine is much slower than that of the preceding 
 compounds ; the process is given in the 2nd and 3rd examples; 
 the non-volatile chlorides contain lead, copper, a part of the 
 iron, and zinc. The volatile chlorides in the receivers (no 
 tartaric acid need be added to the fluid) are treated with hy- 
 drosulphuric acid as soon as the sulphur has been separated. 
 as also the sulphate of baryta and the excess of the chloride 
 of barium ; this treatment precipitates a small quantity of an- 
 timony; the filtrate from this is combined with that obtained 
 from the sulphide of copper, and serves to determine the a- 
 mount of iron and zinc, the latter of which sometimes con- 
 tains nickel. 
 
 IV, AKSENIOAL METALS 
 and their Sulphides, 
 
 Of the compounds of arsenic with metals those of iron, nick- 
 el, and cobalt are in general of the most frequent occurrence, 
 and their analysis is of most importance. 
 
 There are several methods of analyzing these compounds ; 
 the best mode of proceeding, however, is to oxidize them 
 with aqua regia, thus giving rise to arsenates of metallic ox- 
 ides, and then to decompose these at a red heat with carbo- 
 nate of soda, or to precipitate the arsenic with hydrosulphu- 
 ric acid. 
 
 1. AKSENIC AND IBON. 
 
 (LEUCOPYRITE.) 
 The substance is first reduced to an impalpable powder 
 
and then covered in a flask with concentrated nitric acid ; as 
 soon as the violence of the action has ceased, the contents are 
 digested for some time, then mixed with hydrochloric acid 
 and heated until the solution is effected (quarz &c. being left 
 undissolved). Or the substance is oxidized with chlorate of 
 potassa and hydrochloric acid, and maintained at a gentle heat, 
 until all is dissolved ; afterwards it is diluted with water and 
 lieated in an evaporating dish as long as there is any smell 
 of chlorine. 
 
 if it were desirous to treat the solution directly with hy- 
 Irosiilphuric acid in order to separate the arsenic, the separa- 
 tion would proceed very slowly, because arsenic is precipita- 
 ted quite gradually, and a fluid, containing arsenic acid and 
 thoroughly saturated with hydrosulphuric acid, can still retain 
 in solution much of this metal ; and a continuous current of 
 gas oven for several days, digesting in heat, and repeated fil- 
 ? ration would be required to effect the object in view. It is 
 much easier, however, to precipitate solutions containing ar- 
 son i<nis acid; in consequence of this, as was first proposed by 
 Woohler, arsenic acid is reduced to arsenious acid, and then 
 precipitated with hydrosulph uric acid. This reduction is best 
 effected with sulphurous acid. But since free nitric acid, on 
 account of its oxidizing properties, is antagonistic to this re- 
 duction, it must be removed from the solution. 
 
 The solution, therefore, of arsenical iron is placed in a dish 
 cind evaporated to a small volume, adding towards the last a 
 *mall quantity of hydrochloric' acid ; it is then diluted with 
 water and so much acid sulphite (potassa, soda) is added, un- 
 til it has a strong smell of sulphurous acid, but still has an 
 acid reaction. It is then set aside for several hours, after 
 which it is heated to drive off this smell, and placed in a beak- 
 er glass ; a current of hydrosulphuric acid is now caused to 
 pass through the solution for -some time. 
 
 An aqueous solution of sulphurous acid may also be used 
 instead of the alkaline sulphite, or the gas itself may be made 
 to pass through the fluid. < )n account of the injurious action 
 
9T 
 
 of nitric acid, as a reducer, it is frequently preferable to 
 dissolve the arsenical mineral by means of chlorate of potas- 
 sa and hydrochloric acid. Ilydrosulphuric acid is then pass- 
 ed through the fluid until the latter has a strong smell of this 
 gas: it is, then set aside uncovered in a gentle heat until the 
 smell has ceased ; the sulphide of arsenic is then removed by 
 filtration. The gas is again passed through the iiltrate, which 
 is also heated as before, in order to ascertain whether any ar- 
 senic still remains in the solution. The precipitate is wash- 
 ed with cold water; it is then dried in air and oxidized either 
 by means of aqua regia, or chlorate of potassa and hydrochlo- 
 ric acid, from which the amount of arsenic may be computed 
 in the form of the double arsenate of ammonia and magnesia 
 by the method described, page 94. 
 
 There now remains still the fluid containing iron, h'ltered 
 from the sulphide of arsenic. This iiltrate is evaporated in 
 the dish to a small bulk, oxidized with nitric acid, diluted, 
 poured into a beaker glass and precipitated with ammonia. 
 From the amount of precipitated sesquioxide of iron, after a 
 thorough washing, drying and ignition, the amount of iron 
 may be computed. 
 
 OBSERVATION 1. Arsenical iron ore generally contains a small quantity of sulphur, 
 la this case it i best to use a fresh quantity and oxidize it in like manner with nitro- 
 hydrochloric acid, and then precipitate the diluted solution with chloride of harium. 
 
 OBSERVATION 2. Tf the arsenical iron ore contains so much nickel or cobalt as to 
 make it necessary to determine its amount, their separation from the iron is effected by 
 the method described in the following examples. 
 
 The following is. a second method of analysis: a given quan- 
 tity of the substance, reduced to an impalpable powder, is 
 mixed in a porcelain crucible with three times its weight of 
 dried carbonate of soda, as also three times its weight of sul- 
 phur. The crucible is then covered, and the mixture is melt- 
 od over the lamp. In this experiment the crucible is not at- 
 tacked. As soon as the crucible is cold, it is immersed in 
 water, which is then heated to effect a solution. The sulphide 
 of iron is then filtered from the solution containing sulphar- 
 *oiiide of sodium. The latter is concentrated by evaporation*, 
 placed in a flask, and carefully supersaturated with hydrochlo- 
 
98 
 
 f ric acid to which chlorate of potassa is gradually added in or- 
 der to oxidize the separated sulphide of arsenic, the process 
 being the same as that given before, page 94. After the ad- 
 dition of chloride of ammonium the arsenic is precipitated 
 as the double arsenate of ammonia and magnesia. 
 
 The sulphide of iron is dissolved in hydrochloric acid, oxi- 
 dized by a small quantity of chlorate of potassa, and the sesqui- 
 oxide of iron is precipitated by ammonia. 
 
 It is still a simpler method to heat the solution of the arsen- 
 ical iron ore in nitro-hydrochloric acid (or in chlorate of potas- 
 sa and hydrochloric acid) with an excess of the hydrate of po- 
 tassa to boiling. By this means the sesquioxide of iron is se- 
 parated, which is then removed by filtration, thoroughly wash- 
 ed, dissolved in hydrochloric acid, and again precipitated by 
 ammonia. The alkaline fluid is made acid with the same acid, 
 excass of ammonia is then added, and finally arsenic acid i,s 
 precipitated in the form of the double arsenate of ammonia 
 and magnesia. (Nevertheless arsenic acid can be separated 
 in this manner from no other oxides except those of iron and 
 copper.) 
 
 If it is required to determine the amount of iron alone in 
 any arsenical iron ore, the tine powder is mixed in a porcelain 
 crucible with sulphur and then ignited with hydrogen gap. 
 This experiment is repeated. The proto-sulphide of iron, 
 Fe S, then remains. 
 
 2. SULPHUB, AESENIC, IRON. 
 
 (MlSPICKEL, ARSENICAL IRON). 
 
 The analysis of such a compound, in which the sulphur forms 
 an essential component, is made in the same manner as that 
 of arsenical iron. It is dissolved in nitro-hydrochloric acid, 
 or in hydrochloric acid and chlorate of potassa ; the amount 
 of separated sulphur, as well as the amount of sulphuric acid 
 that has been produced, is determined in the manner indica- 
 ted on page 79 for the metallic sulphides ; the excess of ba- 
 ryta is removed, and the iron is separated from the arsenic; 
 
99 
 
 oi- a separate quantity of the substance may be devoted to the 
 special determination of sulphur itself. 
 
 The amount of iron also may be determined by heating the 
 instance, mixed with sulphur, in hydrogen gas (vide above). 
 
 3. ARSENIC, NICKEL, COBALT, IRON, COPPER, 
 ANTIMONY, 
 
 (GOPPKK NICKEL, WHITE NICKEL, SMALTINE, NICKEL GLANCED 
 COBALT GLANCE, COMMERCIAL NICKEL.) 
 
 These substances differ only in respect to the quantities of 
 their components. In copper nickel and white nickel, in 
 nickel and cobalt glance, and in nickel the latter predomi- 
 nates, and this is the case in several specimens of smaltine; 
 whilst in the other forms the amount of cobalt is very near- 
 ly equal to that of nickel or even exceeds it. Iron is always 
 or almost always present. Arsenic, as well as iron (and copper) 
 exists in exceedingly small quantities in commercial nickel. 
 In the following table will be seen at a glance the amounts 
 of these components in such substances, as determined by 
 the analyses up to the present time : 
 
 Arsenic. Nickel. Cobalt. Iron. 
 Topper nickel 5056 4045 0.10.3 02.7 
 
 White nickel 7072 2130 04 013 , 
 
 Smaltine 5080 035 324 119 
 
 Glances (speise) 2644 555 023 010 
 Nickel (commercial) 05 7590 36 
 
 Nickel and cobalt are probably always found in these ores ; 
 and if in the analyses, that have hitherto been made, only 
 one of them is reported as found, it is probable that the other 
 also is present in very small quantity, but has been over- 
 looked. 
 
 1. Method. The compound, as in No 1., is reduced to an 
 impalpable powder and then oxidized either by means of ni- 
 tro-hydrochloric acid, or by chlorate of potassa and hydro- 
 chloric acid; and the arsenic is precipitated and determined 
 by the method indicated. 
 
100 
 
 The filtrate from the sulphide of arsenic is heated in order 
 to get rid of all hydrosulphuric acid. The iron is then oxi- 
 dized by means of chlorate of potassa; the chlorine is expell- 
 ed by the application of heat and the addition of a little alco- 
 hol ; the fluid, highly diluted, is placed in a dish and mixed 
 with carbonate of soda, so that a little oxide of iron (as a 
 basic salt) begins to separate, but a part of the iron remains 
 still in solution, or at least so far, until the fluid, in conse- 
 quence of the formation of a basic salt of the sesquioxide of 
 iron, assumes a dark red brown color; it is then mixed 
 with acetate of soda and heated to boiling (vide page 67). 
 By this proceeding the iron is wholly precipitated as a basic 
 salt of the oxide, whilst the oxides of nickel and cobalt re- 
 main in solution; and the fluid, according as either one or 
 the other predominates, assumes either a greenish or a light 
 red color. The precipitate of iron is filtered hot, thoroughly 
 washed with hot water, and then again placed in the dish to- 
 gether with the filter; here it is dissolved in dilute hydrochlo- 
 ric acid, and, after filtration, the iron is thrown down as the 
 sesquioxide by means of ammonia. (No precipitate must be 
 produced, nor any discoloration by the addition of sulphide 
 of ammonium.) 
 
 To the filtrate, in a dish containing nickel and cobalt, hy- 
 drate of potassa is added in excess, it is then heated to boil- 
 inT, filtered and washed whilst hot. The mixture of the two 
 oxides is then dissolved in acetic acid. A concentrated solu- 
 tion of nitrite of soda (if it contains any free potassa, you 
 must see that there is an excess of acetic acid) is added to 
 this solution, and then the mixture is set aside for twenty- 
 four hours. By this means a yellow precipitate is formed of 
 the double nitrite of the sesquioxide of cobalt and potassa ; 
 tjiis is removed by filtration and thoroughly washed by a sa- 
 turated cold solution of chloride of potassium. Nitrite of po- 
 tassa is again added to the filtrate, which is allowed to stand 
 ff r a long time, in order to be certain that there is no longer 
 any more precipitate formed. The yellow precipitate is now 
 
101 
 
 digested with hydrochloric acid with which it is dissolved by 
 the evolution of chlorine ; the fluid is next diluted and filter- 
 ed into a dish ; and the oxide of cobalt is precipitated by boil- 
 ing with an excess of hydrate of potassa. The precipitate i^ 
 thoroughly washed whilst hot, then dried and strongly igni- 
 ted in a porcelain crucible in hydrogen gas, by means of which 
 < 4 obalt is reduced to the metallic form. After cooling in n 
 current of the gas it is weighed in the covered crucible. Since 
 is is very difficult, however, to remove every trace of potassa 
 from the oxide of cobalt by washing, the reduced metal (if tilt- 
 analysis is to be quite accurate) must be washed with water 
 and then again ignited in hydrogen gas. 
 
 The filtrate from the precipitated cobalt contains nickel* 
 which is thrown down by boiling with hydrate of potassa. 
 After careful washing, in order to remove the alkali, drying 
 and igniting, pure oxide of nickel is obtained, from which the 
 amount of nickel may be computed, 
 
 The weight of the two metals together may also be deter- 
 mined, the cobalt then precipitated, and the amount of nick- 
 el will be the difference of the two weights. In this case the 
 two oxides are dried (the filter is incinerated, and the ashes 
 added thereto) and reduced in hydrogen gas, then weighed ; 
 and, when great accuracy is required, the reduced metal is 
 carefully washed, once more ignited in hydrogen gas and again 
 weighed. It is finally dissolved in nitric acid ; the solution 
 is diluted and then treated with nitrite of potassa. 
 
 2. Method. The substance is reduced to a very fine powder 
 and then placed in a covered porcelain crucible, mixed with 
 sulphur, and heated to ignition. This operation is repeated 
 once or oftener until no longer sulphide of arsenic volatilizes. 
 By this mode of proceeding all the arsenic, with the excep- 
 tion of a mere trace, is volatilized, and there remains a mix- 
 ture of metallic sulphides. The volatilization of the arsenic- 
 takes place the more slowly the greater the proportion of co- 
 balt contained in the substance, and, on this account this me- 
 thod is adopted neither for cobalt proper (speisekoba It), nor 
 
102 
 
 when the substance contains antimony. The metallic sul- 
 phides are dissolved in nitro-hydrochloric acid, or in chlorate 
 of potassa and hydrochloric acid ; and the separation of iron r 
 nickel and cobalt is effected by methods already given. The 
 iinount of arsenic is represented by the amount of loss. 
 
 Several of the substances above-mentioned contain other 
 components, but in very small quantities, besides those alrea- 
 dy given. 
 
 Sulphur is found in almost all of them. To determine its 
 Minount it is better to employ a separate quantity of the sub- 
 tance ; the process is the same as that in No 2., or according 
 to page 79. 
 
 Silica occurs in commercial nickel and is separated by dis- 
 solving the latter in nitric acid, &c., as silicic acid, of which y 
 however, a small portion remains in solution. The whole is, 
 therefore evaporated to dryness on the water bath ; the resi- 
 due, after cooling, is moistened with acid, then treated with 
 water, after which the silicic acid is separated by filtration ; 
 r,he filtrate requires further treatment. 
 
 Antimony is found in nickel and cobalt ores, and also in 
 gray nickel. This substance is precipitated together with 
 the arsenic by hydrosulphuric acid, and its separation is effec- 
 ted according to the method given on page 93. 
 
 COPPER, BISMUTH, LEAD. -T- When these metals are present 
 in small quantities, a' current of hydrosulphuric acid is pass 
 (3d through the solution a short time. They are thrown down 
 sooner than the arsenic, but always to a small extent accom- 
 panied with this metal. The solution is filtered, and then 
 a current of hydrosulphuric acid is passed through the filtrate 
 until the whole of the arsenic has been separated. The pre- 
 cipitated metallic sulphides are placed, while still moist, tor 
 Aether with the filter itself in a flsisk, and digested at a pret- 
 ty high temperature with concentrated yellow sulphide of 
 
103 
 
 After the solution is cool, it is diluted and filtered in a CMV 
 yered funnel ; the residual sulphurets are washed well with 
 water mixed with a few drops of sulphide of ammonium ; th<s 
 filtrate is neutralized with a slight excess of hydrochloric acid, 
 and the precipitated sulphide of arsenic is separated by filtra- 
 tion and afterwards added to the principal mass already ob- 
 tained. The sulphides are separated from each other, if their 
 quantities will permit it, by methods indicated on pages 55 
 and 56. 
 
 If the quantities of such metals are considerable, which ih 
 the case with several of the furnace-products designated ae 
 mixtures (Speise), their mode of analysis will be given in n. 
 subsequent article. (Example 5.) 
 
 4. SULPHUR, ARSENIC, ANTIMONY, NICKEL, 
 COBALT, IRON. 
 
 (NICKEL GLANCE, COBALT GLANCE, MIXTURES SPEISE), 
 
 The analysis of such substances, as contain a large amount 
 of sulphur, is effected on the whole as in the preceding ex- 
 ample. The estimation of all the components is determined 
 either from one and the same amount of substance, that is to 
 say, the sulphur that is set free by the solution is first dot or- 
 mined, the sulphuric acid is precipitated by chloride of bari- 
 um, the excess of baryta is then removed and finally the me- 
 tals are separated. Or two separate quantities of the sul> 
 stance are employed, one for the determination of the amount 
 of the sulphur, and the other for that of the metals, the pro- 
 cess to be used being that given in example 3.* 
 
 If arsenic and antimony are both present (several speci- 
 mens of nickel glance) the determination of both of them 5* 
 Tnade according to instructions found on page 93. 
 
 If small quantities of copper are present, the remarks to 
 be found on page 102 are valid here. 
 
 *Nickel glance and mifipickel comport thenreelves with regard to nitric acid likt Th< 
 eimple metallic sulphides, the metal is first oxidized, and then the aulphur, apart; ttf the 
 latter being set free. On the contrary cobalt glance dissolves in nitric acid without H- 
 derating enlphur. > 
 
104 
 
 The method adduced in the preceding example for decom- 
 position by means of sulphur (2.) can be adopted here only 
 in those cases where the substance contains no antimony and 
 but little or no cobalt. Naturally the method in question is 
 intended only for the determination of electropositive metals. 
 According to Rose* all such substances, not containing anti- 
 mony, may be converted indeed by means of acid sulphate of 
 ammonia into sulphates tree from arsenic : but the crucibles 
 will be thereby acted upon. 
 
 5. LEAD AND COPPER MIXTURES. 
 
 The substances designated by the names Lead-speise and 
 Oopper-speise belong to the most complicated compounds,, 
 that are found in the furnace-prod nets after the smelting of 
 lead and copper ores containing arsenic and antimony; for 
 those contain copper, lead, iron, nickel, cobalt, zinc, bismuth r 
 silver, arsenic, antimony and sulphur. 
 
 A. Mixtures containing little or no lead or antimony. 
 
 One part is set aside for the determination of the sulphur; 
 another is dissolved, either in nitro-hydroohloric acid, or in 
 hydrochloric acid and chlorate of potasMi. The silver gene- 
 rally exists in these mixtures in such a trifiing quantity as 
 not to leave any precipitate of chloride of silver. If any 
 Chloride of lead has separated, it is again dissolved in boiling- 
 water. The solution is then treated with hydrosulphuric a- 
 viid, which precipitates lead, copper, bismuth, antimony and 
 arsenic as sulphides; these are separated by filtration. Hy- 
 irosulphuric acid is again passed through the filtrate, previ- 
 ously heated, in order to be certain that all the arsenic hats 
 been completely precipitated. The metallic sulphides are 
 digested with concentrated yellow sulphide of ammonium in 
 order to dissolve the sulphides of antimony and arsenic, add- 
 ing so much the more of the reagent and continuing the di- 
 gestion (even to boiling) so imi<-h the longer, according to the 
 greatness of the amount { vide page W'2). After filtration the 
 
 * Poggend. Ann.. Pae4ttf>. 
 
105 
 
 two sulphides are precipitated by hydrochloric acid and then 
 separated according to instructions given on page 93. 
 
 The undissolved sulphides of copper, lead and bismuth are 
 dried in the air and rubbed off from the filter into the dish; 
 the filter itself is ignited, and the ashes are dissolved alone 
 in a small porcelain crucible in nitric acid. The mixture of 
 sulphides in the dish is also dissolved in nitric acid, from 
 which the impure sulphur is removed, and digested with heat 
 in nitric acid. These three solutions are mixed together. 
 (A small quantity of sulphate of lead may remain undissolv- 
 ed; this is to be collected and weighed on a weighed filter). 
 The solution of the three metals in nitric acid is concentrated 
 by evaporation, and then lead is separated first according to 
 the process given on page 56. From the filtrate the bismuth 
 is next precipitated, and finally the copper (vide page 56). 
 
 In the filtrate from the precipitate made by hydrosulphu- 
 ric acid are found still iron (in the form of protoxide), nickel, 
 cobalt, and zinc. It is evaporated in order to get rid of the 
 hydrosulphuric acid and most of the free acid; the iron is 
 then oxidized by means of a little chlorate of potassa ; the flu- 
 id is diluted ; the free chlorine is liberated by heat and a few 
 drops of alcohol; and finally the sesquioxide of iron is sepa- 
 rated from the rest according to the method given, page 66. 
 To the filtrate raised to boiling an excess of carbonate of soda 
 is added ; the precipitate is removed by filtration and is dis- 
 solved without washing in an excess ot acetic acid. A cur- 
 rent of hydrosulphuric acid is now passed through the diluted 
 fluid ; this precipitates the zinc alone. The sulphide of zinc 
 is filtered off, dissolved in hydrochloric acid, and precipitated 
 a-s carbonate of zinc by means of carbonate of soda. The fil- 
 trate from the sulphide of zinc is concentrated by evapora- 
 tion ; to this is added nitrate of potassa (and, when necssary, 
 free acetic acid) to separate cobalt and nickel (vide page 99), 
 
 OBSERVATION. The separation of zinc from nickel and cobalt by the method de- 
 scribed succeeds only in the presence of the requisite amount of free acetic acid. More- 
 over the precipitation of the sulphide of zinc must not be continued too long, other- 
 wise a small quantity of sulphide of nickel and of sulphide of cobalt will be precipitat- 
 ed at the same time. The following is another method of separation : the three metals 
 
 14 
 
106 
 
 are thrown down as carbonates ; these are filtered, washed hot, dried and ignited. The 
 oxides are reduced to an impalpable powder, mixed intimately in a porcelain crucible 
 with an excess of sulphur and then heated in hydrogen gas (vide page 53). The sulphides 
 thus produced are treated in the cold with highly diluted hydrochloric acid, which dis- 
 solves the sulphide of zinc alone, and leaves as a residue the sulphides of nickel and 
 cobalt, which are dissolved in nitric acid, after which the two metals are separated. 
 
 B. Mixtures containing much lead or antimony. 
 
 The substance, previously reduced to an impalpable powder, 
 is decomposed by heat in chlorine, as described on page 88. 
 The decomposition is not effected, as is the case with the me- 
 tallic sulphides, in the cold; it is only when the substance is 
 heated that the decomposition takes dlace, and then so sud- 
 denly and energetically as to require the greatest caution. 
 The bulb contains, after the termination of the operation,the 
 non-volatile chlorides of lead (silver),bismuth, copper, nickel, 
 cobalt, and partly those of iron and zinc. They are dissolved 
 by treating them with water and hydrochloric acid, leaving 
 as a residue undecomposed material and chloride of silver, 
 which are collected on a weighed filter, and after weighing 
 are separated by ammonia. The solution in hydrochloric 
 acid is filtered into a dilute solution of sulphuric acid and e- 
 vaporated ; lead is then separated as sulphate of oxide of lead, 
 proceeding as indicated on page 55. A current of hydrochlo- 
 chloric acid is passed through the fluid ; this precipitates cop- 
 per and bismuth which are separated as indicated on page 56. 
 The filtrate, containing iron, zinc, nickel and cobalt, is treat- 
 ed as above prescribed. 
 
 The sulphuric acid is first separated from the volatile chlo- 
 rides by means of chloride of barium, the excess of baryta 
 being in its turn removed by sulphuric acid, then the anti- 
 mony and arsenic are thrown down by hydrosulphuric acid 
 and afterwards separated as indicated on page 94. The fil- 
 trate contains still iron and zinc, the treatment of which will 
 be found on page 95. 
 
 6. ABSENIC, ANTIMONY, 
 
 The substance, reduced to a fine powder, is heated in a 
 
107 
 
 bulb-tube in a current of dry carbonic acid. As soon as the 
 whole arrangement is filled with this gas, heat is applied to 
 the substance and continued as long as arsenic is volatilized. 
 
 This metal is removed from the tube by applying heat gra- 
 dually along the tube to the end. In this operation a high 
 temperature must be avoided ; and it is well to perform it in 
 a place where there is a good draft of air. By weighing the 
 residual quantity of antimony, the amount of arsenic is de- 
 termined. 
 
 If it be desired to determine both the metals directly, dis- 
 solve the substance in aqua regia, or in hydrochloric acid and 
 chlorate of potassa ; tartaric acid is then added to the solution, 
 and afterwards chloride of ammonium ; the solution is then 
 neutralized with ammonia in excess. The arsenic acid is fi- 
 nally precipitated by salt of magnesia as the double arsenate 
 of ammonia and magnesia, after which the antimony in the 
 filtrate is determined as prescribed on page 94. 
 
 7. AESENIC, TIN. 
 
 The alloy is divided up as much as possible, then intimate- 
 ly mixed with five times its weight of carbonate of soda and 
 five of sulphur ; the mixture is then melted at a low heat 
 over the lamp. As soon as it begins to flow gently, it is 
 kept ignited for some time. When cool the mass, consisting 
 of the double sulphide of arsenic and sodium and the double 
 sulphide of tin and sodium, is completely dissolved in water 
 (traces of electro-positive metallic sulphides, for instance of i- 
 ron, copper, lead, remain undissolved) ; the solution is highly 
 diluted, then hydrochloric acid is added in excess ; and the 
 mixture is kept digesting at a very gentle heat, until all smell 
 of hydrosulphuric acid has passed off. The sulphides of tin 
 and arsenic, that have separated, are collected on a weighed 
 filter, washed thoroughly with cold water, containing a few 
 drops of acid, dried in the air, then in the drying apparatus 
 at a temperature of boiling water, until their weight remains 
 constant. 
 
108 
 
 A part of the dried sulphides is placed in a weighed bulb- 
 tube, as already described for the analysis of metallic sul- 
 phides by means of chlorine, page 88 ; the remaining part, 
 together with the filter, is again dried in order to ascertain 
 the quantity in the bulb. The wider tube attached to the 
 bulb is bent at an obtuse angle and dips just below the sur- 
 face of a quantity of ammonia placed in a two-necked Woulfe's 
 bottle. The narrow tube of the bulb is connected with an 
 arrangement through which dried hydrosulphuric acid is ge- 
 nerated and passed. As soon as the whole arrangement is 
 filled with gas, the bulb is heated, as also the wider tube, by 
 which the sulphide of arsenic is driven forward into the am- 
 monia. As soon as there is no longer any formation of the 
 yellow sulphide of arsenic, the bulb is allowed to cool in a 
 current of gas ; and the projecting part of the wide tube is 
 broken off, taken out of the cork and digested in warm wa- 
 ter in order to wash off all adhering substance; the wash- 
 water is then added to the ammoiiiacal solution in the Woulfe's 
 bottle. This fluid is next supersaturated in a flask with hy- 
 drochloric acid, to which chlorate of potassa is added, in or- 
 der to oxidize the sulphide of arsenic by digestion at a gentle 
 heat. To the filtrate from the sulphur ammonia is added in 
 excess, which precipitates the arsenic as the double arsenatc 
 of ammonia and magnesia. 
 
 The sulphide of tin in the bulb, a mixture of protosulphide 
 of tin, SnS, and a small quantity of sulphur, is shaken into 
 a weighed porcelain crucible, moistened with nitric acid, and 
 carefully heated, in order to convert it into stannic acid, (vi- 
 de page 60). 
 
 If arsenic is not to be determined by a direct method, the 
 metallic sulphides are treated with hydrosulphuric acid in a 
 porcelain crucible, in which then the sulphide of tin remains 
 nnvolatilized. 
 
 V. SULPHATES. 
 
309 
 
 In the analysis of the sulphates the process varies accord- 
 ing as they are soluble or insoluble in water or in acids. 
 
 Soluble in water are the sulphates of the alkalies, of mag- 
 nesia, of alumina, and of most of the metallic oxides, if they 
 are neutral. 
 
 Soluble in acids are the basic sulphates of alumina and of 
 most of the metallic oxides. Hydrochloric acid is their most 
 appropriate solvent 
 
 Insoluble or with difficulty soluble both in water and acid* 
 are the sulphates of baryta, strontia, lime, the oxide of lead 
 and of a few rare oxides. These are decomposed by ignition 
 with carbonate of soda, giving rise to sulphate of soda and a 
 carbonate, which may be separated by water. 
 
 General determination of sulphuric add. This is always 
 effected by means of chloride of barium, the fluid having been 
 previously rendered slightly acid by hydrochloric acid in or- 
 der partly to decompose any carbonate that may be present 
 and secondly to promote the deposition of the sulphate of ba- 
 ryta. Notwithstanding all this, it sometimes happens, even 
 when the precipitate has settled properly, that a part of it 
 passes through the filter, rendering the filtrate turbid, whilst 
 another part fills up the pores of the paper to such an extent, 
 although the filtrate passes clear, yet the filtration is exceed- 
 ingly slow, and if this be relieved by washing, the fluid be- 
 comes again turbid. 
 
 In order to obviate this, the clear supernatant fluid is first 
 poured upon the filter, the deposit is then stirred up with 
 boiling water, this is allowed to settle, and the clear fluid is 
 again poured upon the filter. This operation is repeated ; fi- 
 nally the deposit itself is placed on the filter. 
 
 The sulphate of baryta is washed with hot water and dried 
 in the air, it is then removed as much as possible from the 
 filter into a platinum crucible, the filter itself being incine- 
 rated on the lid of the crucible and the ashes thrown into the 
 crucible with the rest. 
 
 The contents are now ignited, taking care to place a email 
 
110 
 
 atrip of platinum across the edge in order to promote the in- 
 cineration of any small fibres of paper y as also, if by means of 
 the carbon there should be formed a trace of sulphide of ba- 
 rium r to reduce this again to the state of oxide. 
 
 1. SULPHATE of BAKYTA containing STUOFHA. 
 (HEAVY SPAR,) 
 
 The substance is first reduced to an impalpable powder and 
 then placed in a weighed platinum crucible, it is now mixed 
 gradually with three times its weight of dried carbonate of 
 soda, also reduced to a fine powder ; the mixture is stirred 
 up intimately with a glass rod at each addition. The cruci- 
 ble is now covered and heated over a lamp until the contents 
 fuse. As soon as the crucible has cooled, it is placed in a 
 dish of water and heated in order to loosen the contents and! 
 get them out of the crucible. 
 
 This being effected, the contents are filtered, and the resi- 
 due, consisting of carbonate of baryta and strontia, is tho- 
 roughly washed. The filter and its contents are now transfer- 
 red to a beaker glass, covered with dilute acetic acid and di- 
 gested, until the carbonates are dissolved, taking care to co- 
 ver the vessel with a concave watch glass, in order to avoid 
 all loss by evaporation. The filter is now taken out and wash- 
 ed by means of a washing bottle, and then the fluid is filtered ; 
 in case there is any undissolved substance left on the filter, it 
 is dried, ignited and weighed ; the amount is deducted from 
 that originally started with. If the solution is somewhat di- 
 lute, it must be concentrated by evaporation in an evaporat- 
 ing dish ; it is then mixed with an excess of hydrofluosilicic 
 acid and a quarter of its volume of alcohol, and set aside for 
 twelve hours, during which time the silicofluoride of barium 
 is gradually deposited. At this stage the fluid is filtered on 
 a weighed filter, and dried at a temperature of 248 F ; the re- 
 sidue is washed for a short time in cold water containing a 
 little alcohol, it is then dried in the air, and afterwards in the 
 drying apparatus at a temperature ranging between -248 F, 
 
Ill 
 
 ?md 266F., until its weight no longer changes. 
 
 The filtrate in a dish (preferably of platinum) is mixed with 
 at little sulphuric acid and evaporated to dryness ; the residue, 
 consisting of sulphate of strontia, is placed in a platinum cru- 
 cible and ignited at a low temperature. Or after the addi- 
 tion of the sulphuric acid, a volume of tolerably strong alco- 
 hol equal to that of the fluid is mixed with it ; the sulphate 
 of strontia is then removed by filtration, washed with dilute 
 alcohol, dried, ignited and weighed. 
 
 The sulphuric acid in the alkaline fluid, to which hydro- 
 fluoric acid has been added in excess, is determined in ge- 
 neral by methods previously given. 
 
 2. SULPHATE of STKONTIA containing also 
 BAEYTA and LIME, 
 
 (CELESTINE.) 
 
 The substance is reduced to an impalpable powder, as in 
 the preceding example, then fused with carbonate of soda ; the 
 fused residue is next washed in water in order to dissolve all 
 that is soluble ; the mixture is filtered and the sulphuric acid 
 is determined in the alkaline filtrate. 
 
 The carbonates of the three earths remaining on the fil- 
 ter are transferred to a dish, and whatever adheres to the filter 
 is washed with water containing a few drops of hydrochloric 
 acid, and added to the rest A sufficient quantity of dilute 
 sulphuric acid is now added, and the whole is evaporated un- 
 til all excess of acid has been volatilized. A solution of carbo- 
 nate of ammonia or of bicarbonate of potassa is next added to 
 the sulphates; and the mixture is set aside for at least twen- 
 ty-four hours at a temperature Mow <J8F., taking care to stir 
 it frequently. By this treatment the sulphate of baryta re- 
 mains unchanged, whilst the sulphates of strontia and lime 
 are converted respectively into carbonates. The mixture is 
 filtered, washed with a very dilute solution of carbonate of 
 potassa, then with pine cold water, until the wash-water by 
 the addition of hydrochloric acid no longer indicates the pre- 
 
112 
 
 gence of chloride of barium. The whole is then treated in 
 the cold with dilute hydrochloric acid ; and the sulphate of 
 baryta is removed by filtration, washed, dried and ignited. 
 
 The solution of strontia and lime is evaporated to dryness 
 on the water bath ; the free acid is driven off; the two chlo- 
 rides are then dissolved in the smallest quantity of water pos- 
 sible, to which is added a solution of about fifty times its vo- 
 lume of sulphide of ammonium ; the mixture is then set aside 
 in the cold for twenty-four hours, stirring it frequently ; or 
 it may be boiled for some time, taking care to replenish the 
 water lost by evaporation. By this means the sulphate of 
 strontia, which is insoluble in solutions of the alkaline sul- 
 phates, is completely separated, whilst the sulphate of lime 
 remains dissolved. The former is separated by filtration and 
 washed with a concentrated solution of sulphate of ammonia 
 until the wash-water no longer becomes turbid with oxalate 
 of ammonia. To the filtrate, diluted with water, ammonia 
 and oxalic acid are added, taking care that the fluid still re- 
 tains no alkaline reaction ; the mixture is then set aside at a 
 gentle heat, until the oxalate of lime has entirely settled. A 
 few drops of oxalic acid are added in order to be certain that 
 the whole of the lime has been precipitated ; the latter is then 
 separated by filtration, washed with hot water, dried and ig- 
 nited, at first over the lamp with access of air, then for a few 
 minutes over a gas-blast, by which it is converted into pure 
 lime.* 
 
 Instead of treating the sulphate with a solution of carbo- 
 nate of ammonia or of carbonate of potassa in the cold, it may 
 be boiled with a % solution of carbonate of potassa to which a- 
 bout one-third of sulphate of potassa is added. The solution 
 is allowed to settle ; the clear part is filtered, and the filtrate 
 is boiled with a solution of the two salts. The mixture is 
 washed from the sulphate of baryta and from the carbonates 
 of strontia and lime; the subsequent process has been given. 
 
 If the amount of lime exceeds a few decigrammes, or if the gas-blast IP not at hand, 
 the oxalate of lime is converted toy gentle ignition, &c., into carbonate of lime. (Vide- 
 page 119). 
 
113 
 
 3. ALUM. 
 
 (SULPHURIC ACID, ALUMINA, POTASSA, AMMONIA, WATER.) 
 
 a. Estimation of the sulphuric acid. This is effected by 
 the general method already prescribed. 
 
 b. Estimation of the alumina and of the potassa. A 
 fresh quantity of the salt is dissolved in warm water in a beak- 
 er glass, to this is added a sufficient quantity of ammonia as 
 to give the solution an alkaline reaction. The fluid is filter- 
 ed into a dish ; the residue (alumina) on the filter is repeat- 
 edly washed with boiling water, it is then dried and ignited. 
 The filtrate is evaporated to dryness ; the residue is placed 
 in a platinum' crucible, and the evaporating dish is rinsed out 
 with the smallest possible quantity of water, which is placed in 
 the crucible with the residue. The residue is again evapora- 
 ted- to dryness, taking care, however, not to cause the mix- 
 ture to boil ; as soon as the contents are dry, they are ignited 
 over the lamp at a pretty high temperature. 
 
 Since the sulphate of potassa, which is thus obtained, has 
 an acid reaction from the sulphuric acid derived from the al- 
 umina-salt, and since this excess of acid, or at all events a 
 large portion of it, is but slowly driven off by ignition, the 
 latter is performed in an atmosphere of carbonate of ammo- 
 nia, which materially expedites the volatilization. 
 
 At the termination of the operation a strip of platinum is 
 hung over the edge of the crucible; the lower end of the leg 
 of this strip within the crucible is bent slightly upwards so 
 as to form a sort of stirrup on which is placed a small piece 
 of carbonate of ammonia ; the crucible is now covered with 
 its lid, quickly heated and raised to vivid ignition. After a 
 few minutes the lid is taken off in order to see if there are 
 still fumes of sulphuric acid given off, which are easily made 
 visible by simply breathing into the crucible. If this should 
 be the case, the operation is repeated with a second piece of 
 carbonate of ammonia ; finally the strip of platinum is with- 
 drawn, and the crucible with its cover on is allowed to cool, 
 whereupon it is immediately weighed. The amount of po~ 
 15 
 
114 
 
 tassa is computed from the amount of neutral sulphate of po- 
 tassa found. 
 
 Instead of carbonate of ammonia, a saturated solution of 
 sulphide of ammonium, recently prepared, is almost to be 
 preferred as a precipitant of the alumina. 
 
 c. Estimation of the ammonia. A fresh quantity of the 
 salt is dissolved in a flask in water and then boiled with a so- 
 lution of either soda or potassa. The ammonia set free is 
 caused to pass through a tube bent twice at right angles into 
 a Woulfe's bottle and to dip into a solution of dilute hydro- 
 chloric acid filling the bottle about one-third full ; the bottle 
 itself is kept cool by being placed in a vessel of cold water ; 
 it is also connected with a second two-necked bottle by means 
 of a tube reaching down but a few lines into the same 
 dilute acid. The gas is caused to be evolved regularly .but 
 slowly by boiling the mixture for some time. As soon as it 
 may be presumed, that all the ammonia has been expelled, 
 the flask is disconnected from the tubes and the lamp remov- 
 ed. The fluids in the two bottles are poured into an evapo- 
 rating dish, and evaporated in this nearly to dryness and 
 then transferred to a platinum crucible and evaporated to dry- 
 ness on the water bath, The residual chloride of ammoni- 
 um is completely dried at a temperature of 21 2 F. and then 
 weighed. 
 
 d. Estimation of the water. If the amount of water is 
 not to be determined by the loss, the salt must be ignited. 
 Since, however, the sulphates of the weaker bases, as for in- 
 stance of alumina, also lose sulphuric acid by ignition, this 
 operation alone would produce an erroneous result. The salt, 
 therefore, is first mixed with oxide of lead, and then ignited ; 
 this retains the sulphuric acid. The pulverized salt is weigh- 
 ed and placed in a capacious porcelain crucible. In another 
 crucible about six times the amount of powdered oxide of 
 lead is placed, this is heated, but without fusion ; it is then 
 allowed to cool in the covered crucible, and afterwards it is 
 accurately weighed. The powdered salt, in its crucible, is 
 
115 
 
 now mixed with this oxide of lead by means of a glass rod, 
 the last portions being poured upon the salt by shaking. Fi- 
 nally the crucible, which previously had contained the oxide 
 of lead, together with any adhering particles, is weighed, in 
 order to ascertain with accuracy the quantity added to the 
 salt. The mixture is heated for some time at rather a high 
 temperature, and as soon as it is cold, it is immediately weigh- 
 ed. Subtracting the weight of the contents from that of the 
 salt and oxide of lead together will give the weight of the 
 water. 
 
 4. MANSFELD BLACK VITRIOL. 
 
 The bases of this salt, which crystallizes in dark blue- violet 
 cyrstals, having the same form as those of sulphate of iron, 
 out of the mother liquor of roasted copper-ores from which 
 sulphate of copper has already been extracted, are : oxide of 
 copper, protoxide of iron, protoxide of manganese, oxide of 
 cobalt, oxide of nickel, oxide of zinc and magnesia. 
 
 In order to undertake the analysis of this complicated 
 compound, a weighed portion of the salt is dissolved, a little 
 hydrochloric acid is added, and then the copper is precipita- 
 ted by hydrosulphuric acid. The filtration and the further 
 treatment of the sulphide of copper are given in detail, page 
 52. The filtrate from the sulphide of copper is received in a 
 dish and concentrated by evaporation ; the protoxide of iron 
 is then oxidized by nitric acid, and the evaporation is conti- 
 nued until all excess of acid is expelled. The fluid is then 
 very much diluted with water, and the sesquioxide of iron is 
 separated from the remaining bases by means of carbonate and 
 acetate of soda (vide page 67). 
 
 The fluid filtered from the basic salt of iron is treated with 
 an excess of carbonate of soda, and heated to boiling. By 
 this means all the bases present are thrown down ; these are 
 separated by filtration, w r ashed with hot water, and dissolved 
 in excess of acetic acid. A current of hydrosulphuric acid is 
 now passed through the dilute solution, which precitates the 
 
116 
 
 zinc, as sulphide, from the rest of the metals (vide page 105). 
 After the expulsion of the fumes of hydrosulphuric acid by 
 heat, carbonate of soda is added, which precipitates the ox- 
 ides of manganese, nickel and cobalt and also magnesia. The 
 dried and ignited residue is mixed in a porcelain crucible with 
 sulphur, and the mixture is ignited in hydrogen gas. The sul- 
 phide of manganese and the magnesia are now dissolved out 
 by means of dilute hydrochloric acid, and finally nickel and 
 cobalt, remaining undissolved as sulphides, are separated 
 from one another by the method given on page 100, if their 
 
 amount is large enough. 
 
 ~ ,\ 
 
 Acetate of soda is added to the solution containing manga- 
 nese and magnesia ; the mixture is heated, and a current of 
 chlorine is passed through it. The dark-colored fluid is now 
 supersaturated with ammonia and boiled in order to drive off 
 most of the free ammonia. The separated oxide of manga 
 nese is converted by ignition into the sesquioxide of manga- 
 nese (the double protoxide and binoxide) ; whilst the magne- 
 sia is precipitated from the filtrate by means of phosphate of 
 soda. 
 
 OBSERVATION. The filtrates from the precipitates produced by carbonate of soda 
 
 are tested with ammonia, sulphide of ammonium and phosphate of soda, to ascertain 
 whether all the metals and magnesia have been removed. 
 
 A separate quantity of the vitriol is required for the deter- 
 mination of the amount of sulphimc acid. 
 
 '*. ;< VI. PHOSPHATES, .;;- 
 
 The combinations of phosphoric acid with bases belong in 
 general to a class of substances that is more difficult than the 
 rest to analyze. For not only is the determination of the acid 
 itself never quite accurate when precipitated by means of 
 a base (lime, baryta, and oxide of lead), with which it forms 
 an insoluble combination, owing to the solvent action exert- 
 ed by other salts, particularly that of ammonia, on the preci- 
 pitated phosphate, but it is also very difficult in many cases 
 to separate the phosphoric acid from the base in combination 
 
117 
 
 with it, and more especially so from the alumina and magne- 
 sia ; and these difficulties increase if the substance contains 
 other components, as for instance, fluorine or silicic acid. 
 
 Those metallic phosphates are the easiest to analyze, whose 
 bases can be precipitated from an acid solution by means of 
 hydrosulphuric acid, as for instance the phosphates of the ox- 
 ide of copper, of silver and of lead, &c. They are dissolved 
 in an acid; the metal is then precipitated by hydrosulphuric 
 acid from the dilute solution ; and the amount of phosphoric 
 acid in the filtrate is determined from the sulphide. The 
 phosphates of silver and lead may be examined by separating 
 the bases by means of hydrochloric acid or sulphuric acid ; the 
 same operation, too, may be applied to the phosphates of ba- 
 ryta, strontia and lime. 
 
 1, PHOSPHATE of LIME with CHLORIDE and 
 FLUORIDE of CALCIUM. 
 
 (APATITE.) 
 
 a. ^Estimation of the lime and phosphoric acid. The mi- 
 neral is reduced to an impalpable powder and then dissolved 
 by digesting it in hydrochloric acid. If there is a small quan- 
 tity of insoluble residue (of quartz, &c.), the solution is dilu- 
 ted and then filtered ; the residue on the filter is well wash- 
 ed, dried, ignited and weighed ; the amount is deducted. 
 The acid solution is mixed with a proper amount of sulphu- 
 ric acid and with three or four times its volume of alcohol. 
 By this expedient the whole of the lime is precipitated as sul- 
 phate. The mixture is set aside for twelve hours, and then 
 filtered through paper moistened with alcohol; the residue is 
 washed with alcohol. As soon as the residue is dry, it is ig- 
 nited and weighed; from the weight the amount of lime is 
 computed. 
 
 The quantity of hydrochloric acid must not be greater than 
 what is required to effect the solution. The specific gravity 
 of the alcohol need in this experiment to produce the total 
 precipitation of the sulphate of lime is 0.80. The precipitate 
 
118 
 
 Laving thoroughly settled, the clear liquid is tested with al- 
 cohol and a little sulphuric acid to ascertain whether all the 
 lime has been removed. 
 
 The filtrate from the sulphate of liine is diluted with wa- 
 ter, heated for some time in order to drive off the alcohol, 
 and then supersaturated with ammonia. This ought to pro- 
 duce no precipitate ; but should there be any (phosphate of 
 lime), it must be removed by filtration, dissolved in a few 
 drops of hydrochloric acid and decomposed by means of alco- 
 hol and sulphuric acid. To the alkaline fluid an appropriate 
 quantity of a solution of sulphate of magnesia, chloride of 
 ammonium and ammonia (magnesia mixture) is added ; the 
 vessel is then set aside for a few hours at a temperature of 
 86F. ; after which the double phosphate of ammonia and 
 magnesia is separated from the fluid by filtration and then tho- 
 roughly washed with a mixture of one part of ammonia and 
 three parts of water applied warm. As soon as this substance 
 is dry, it is rubbed off from the filter into a platinum cruci- 
 ble ; the filter itself is folded together and placed on the top ,, 
 the crucible is loosely covered and gently heated for some 
 time and then the temperature is gradually raised to ignition, 
 the cover in the meanwhile being pushed back over half the 
 entrance and a strip of platinum foil placed over the edge. 
 (If the precipitate is heated too quickly, it will remain grey.) 
 By ignition the salt is changed into pyrophosphate of magne- 
 sia,- from which the phosphoric acid may be computed. 
 
 b. Estimation of the chlorine. The fine powder is digest- 
 ed with nitric acid, taking care to apply only a very gentle 
 heat, in order not to expel any of the chlorine. The fluid is 
 then diluted and filtered ; and a solution of nitrate of silver 
 is now added, as long as there is any precipitate. The vessel 
 is finally set aside in a warm place in order that the precipi- 
 tate may completely settle. The fluid is now filtered through 
 a weighed filter and dried at a temperature of 248F. ; the 
 chloride of silver is thoroughly washed, then dried, in tho 
 the first place in the air, afterwards in the drying apparatus- 
 
119 
 
 at a temperature of 248 266F., until it ceases to lose any 
 weight. From this, its final weight, the amount of chlorine is 
 computed. 
 
 The fluorine can not be determined by any direct method. 
 Its amount is computed, by first computing, from the phos- 
 phoric acid found, the amount of 3CaO + PO 5 and, from the 
 amount of chlorine found, the quantity of CaCl ; the rest of the 
 calcium is to be regarded as originally in combination with 
 fluorine, forming fluoride of calcium, 
 
 2. BONE BLACK. 
 
 (LAMP BLACK.) 
 
 A mixture of phosphate and carbonate of lime, carbon and 
 sometimes also phosphate of lime containing hygroscopic wa- 
 ter, sand and other impurities. The small quantities of 
 magnesia, soda, fluorine and sulphuric acid, occurring in the 
 organic mass of bones, are not taken into consideration here. 
 
 a. A quantity of the powder is dried at a temperature of 
 248F., until its weight remains constant. 
 
 b. Another quantity is decomposed in a carbonic acid ap- 
 paratus (vide Carbonates) by means of hydrochloric acid. 
 The loss of weight of the apparatus determines the weight 
 of the carbonic acid. The contents are diluted, placed on a 
 weighed filter, and the residue is washed with hot water, it 
 is then dried at a temperature of 248F., ignited in a current 
 of air in order to burn off the carbon, and weighed. The 
 substance thus obtained generally consists of sand. The dif- 
 ference gives the amount of carbon. 
 
 The solution in hydrochloric acid is next mixed with am- 
 monia in order to neutralize the largest portion of the free 
 acid, but it still must retain a slight acid reaction ; acetate 
 of soda is now added and then the whole amount of lime is 
 precipitated by oxalate of ammonia. As soon as the preci- 
 pitate has completely settled, it is removed by filtration, wash- 
 ed with hot water and dried. The dry residue is now placed 
 in a platinum crucible, and the filter is incinerated on the 
 
120 
 
 lid and the ashes are poured to the rest in the crucible.. The 
 mass is now covered up with the lid and heated very slowly y 
 finally the lower part of the crucible is made red hot and re- 
 tained at this temperature for some time. By this means the 
 oxalate of lime is converted into carbonate of lime, which is 
 weighed. Since it is possible, however, that by ignition it 
 may have lost a small portion of carbonic acid, it is moisten- 
 ed with a few drops of a solution of carbonate of ammonia, 
 dried and heated as before. If the weight now does not cor- 
 respond with the first, the operation is repeated until the 
 two last weights remain exactly equal. The amount of lime 
 is computed from the quantity of carbonate found. 
 
 Ammonia is added in excess to the filtrate from the oxa- 
 late of lime. (If a precipitate IB hereby formed, it consists 
 of the double phosphate of ammonia" and magnesia, which 
 has to be removed by filtration and further treated as indica- 
 ted on page 118. The amount of magnesia and phosphoric 
 acid are computed from the quantity of the ignited salt.) 
 The phosphoric acid is precipitated by the magnesia mixture 
 as prescribed on page 118. 
 
 3. ACID PHOSPHATE OP LIME. 
 
 This name is given, though wrongly, to a technical prepa- 
 ration, which is used as a manure, and prepared by decom- 
 posing the substance of bones by means of acids. Its value 
 is in proportion to the quantity of free phosphoric acid. It 
 is a mixture of sulphate of lime, chloride of calcium and free 
 phosphoric acid with water, containing almost always a small 
 quantity of undecomposed phosphate *of lime, also carbon, 
 sand, &c. 
 
 Our task is simply to determine the amount of free phos- 
 phoric acid. An extract is made from about ten grammes 
 by means of hot water, which is added until the filtrate has 
 no longer an acid reaction. The filtrate is now diluted to a 
 litre. One tenth part of this (100 c.c.) is now taken for the 
 analysis. A sufficient quantity of chloride of calcium is add- 
 
121 
 
 ed to the solution, and then an excess of ammonia. By this 
 mode of proceeding all the phosphoric acid is precipitated as 
 the phosphate of lime 3CaO-hPO 5 , which is separated in a 
 close vessel by filtration, thoroughly washed, dried, ignited 
 and weighed. From this weight the amount of phosphoric 
 acid is computed. 
 
 4, PHOSPHATE of the oxide of LEAD and CHLOKIDE 
 of LEAD together with AESENIO acid and LIME, 
 (SEVERAL GKEEN AND BKOWN LEAD OKES.) 
 
 The analysis of such a compound is on the whole the same 
 as that of apatite (1. example). 
 
 a. Estimation of the oxide of lead, the lime, the arsenic 
 acid and the phosphoric acid. 
 
 The fine powder is dissolved in nitric acid ; and a current 
 of hydrosulphuric acid is then passed through the solution. 
 As soon as the precipitate shows an admixture of yellow, 
 it is submitted to filtration. It contains all the lead and a 
 small quantity of arsenic. Hydrosulphuric acid is again pass- 
 e*d through the filtrate ; the precipitate now is sulphide of arse- 
 nic alone ; the mixture is heated on a sand bath and then fil- 
 tered. The operation of passing a current of hydrosulphurie 
 acid through the filtrate, of heating and filtering is repeated 
 until no longer any more sulphide of arsenic is thrown down. 
 
 The mixture of sulphide of lead and of sulphide of arsenic 
 is placed in a small flask by the aid of the wash-bottle and 
 digested whilst still moist with sulphide of ammonium, until 
 all the insoluble substance is of a pure black color ; the sul- 
 phide of lead is then separated from the solution by filtra- 
 tion, after which proceed as directed on page 104. The sul- 
 phide of lead is treated with sulphur in hydrogen as indicat- 
 ed, page 62 ; and the sulphide of arsenic is precipitated from 
 the filtrate by means of hydrochloric acid and afterwards con- 
 verted into the double arsenate of ammonia and magnesia 
 (vide page 94.) 
 16 
 
122 
 
 The filtrates from arsenic and lead contain still lime and 
 phosphorus. In the first place the hydrosulphuric acid is ex- 
 pelled by evaporation, ammonia is then added to give the 
 fluid a slight alkaline reaction, after which acetic acid is add- 
 ed in slight excess ; the lime is now precipitated by oxalate of 
 ammonia, and the phosphoric acid by magnesia salt (vide 
 page 118). 
 
 b. Estimation of the Chlorine. The process is the same 
 as in example 1. 
 
 5, PHOSPHATE of the protoxide of IEON, and WATEE, 
 
 (BLUE IKON ORE. YlVIANITE.) 
 
 The substance is reduced to an impalpable powder and 
 placed in a platinum crucible, and then mixed with three or 
 four times its weight of dry carbonate of soda, also reduced 
 to a fine powder, the latter being added gradually, and then 
 brought into intimate contact with the powdered ore by 
 means of a glass rod. The mixture is ignited in the flame of 
 a lamp, after which the melted mass when cold is digested 
 with water, until all that is soluble is in a liquid state. Tito 
 alkaline fluid, which contains phosphate of soda f is filtered 
 from the residual (containing an alkali) oxide of iron ; the 
 latter is thoroughly washed, and then placed, together with 
 the filter, in a beaker glass and digested with hydrochloiio 
 acid. The fluid is diluted with water, and filtered ; it is now 
 heated with chlorate of potassa in order to oxidize whatever 
 portion of the protoxide of iron may be present. As soon 
 as the fluid is cold, ammonia is added, which precipitates 
 the sesquioxide of iron, which is separated by filtration, wash- 
 ed, dried and ignited. 
 
 The alkaline filtrate is rendered acid in a flask with hydro- 
 chloric acid, and then heated for some time in order to get 
 rid of all the carbonic acid. Ammonia is now added in ex- 
 cess and the phosphoric acid is precipitated by the magnesia 
 mixture (vide page 118). 
 
 Estimation of the Protoxide iron. By the method just 
 
123 
 
 given the entire amount of oxide of iron is obtained in the 
 form of sesquioxide. But since the original substance con- 
 tains both the oxides of iron, one of them has to be sepa- 
 rately determined. To accomplish this object the compound 
 is dissolved in a close vessel by means of hydrochloric acid; 
 and the solution is then submitted to the volumetric test 
 with permanganate of potassa. 
 
 The amount of water is obtained from the loss. 
 
 OBSERVATION. The phosphate of iron may also be analyzed by treating the so- 
 lution with ammonia and sulphide of ammonium ; the precipitated iron is then remov- 
 ed by filtration ; and the phosphoric acid is determined in the filtrate. Nevertheless it is 
 no easy matter to separate the sulphide of iron by filtration ; it is easily oxidized, and 
 the filtrate will naturally contain iron. On this account this method is less to be recom- 
 mendeb here than with other metallic phosphates. 
 
 6, PHOSPHATE of the protoxide of IKON, protoxide 
 of MANGANESE, LITHIA and SODA, 
 
 (TRIPHYLINE.) 
 
 The compound contains, moreover, small quantities of 
 magnesia and potassa. 
 
 1. Method. The substance in line powder is dissolved by 
 heat in concentrated hydrochloric acid. The solution is di- 
 luted, and whatever residue there may be (quartz, felspar, 
 tfcc.) is removed by tiltration ; the filtrate is then boiled in a 
 disli for some time with hydrate of baryta. Baryta is add- 
 ed until the fluid has a strong acid reaction, and a film of car- 
 bonate of baryta begins to form on the surface. The fluid is 
 filtered hot, and the residue is washed for some time with 
 hot water. The baryta precipitates the phosphoric acid, the 
 protoxides of iron and manganese and the magnesia ; in the 
 solution there remain lithia andsoda, together with the small 
 quantity of potassa. 
 
 The precipitate together with the filter is digested with 
 dilute hydrochloric acid, mixed with water and so much sul- 
 phuric acid as is necessary to precipitate any adhering bary- 
 ta ; after which the sulphate of baryta is separated by filtra- 
 tion. Ammonia is added in excess to the filtrate, and then a 
 sufficient quantity of sulphide of ammonium, so that by heat- 
 
ing well, the black precipitate thoroughly settles, and the su- 
 pernatant fluid has a yellow color. The residue, consisting 
 of sulphide of iron and sulphide of manganese, is filtered hot, 
 taking care to hasten the operation as much as possible and 
 to cover the funnel ; it is then thoroughly washed with water 
 containing a few drops of sulphide of ammonium, dissolved in 
 hydrochloric acid, digested until all smell of hydrosulphuric 
 acid has passed off, and finally the solution is submitted to 
 filtration. The iron is then oxidized by means of chlorate 
 of potassa ; and the separation of iron and manganese is ef- 
 fected by means of carbonate and acetate of soda (vide page 
 67). The filtrate from the metallic sulphides is rendered 
 slightly acid with hydrochloric acid, then heated in order to 
 expel all the hydrosulphuric acid and filtered ; finally the 
 phosphoric acid is precipitated by means of ammonia and the 
 magnesia mixture (vide page 118). 
 
 The alkaline filtrate from the precipitate of baryta is mix- 
 ed with carbonate of ammonia and pure ammonia and warm- 
 ed for some time in order to separate the baryta that had 
 been dissolved; this is then removed by filtration and wash- 
 ed with hot water. Hydrochloric acid is added in excess to 
 the filtrate, which is evaporated to dryness on the water-bath ; 
 the residue is placed in a platinum crucible and heated, but 
 not to ignition, the crucible being well closed all the while. 
 The chlorides of lithium,s odium and potassium, thus obtain- 
 ed, are then weighed. They are afterwards placed in a well 
 stoppered vessel containing equal parts of concentrated alco- 
 hol and ether, and set aside for several days, taking care to 
 shake the mixture frequently. By this means the chloride 
 of lithium is dissolved. The mixture is then submitted to 
 filtration on a weighed and covered filter ; the chlorides of 
 sodium and potassium on the filter are washed with alcohol 
 and ether, then dried at a temperature of 248F. and weigh- 
 ed. The difference in weight gives the amount of chloride 
 of lithium. 
 
 If the small amount of potassa is also to be determined, 
 
125 
 
 the separation is effected by chloride of platinum, as described 
 hereafter (vide Silicates of alumina, lime, soda and potassa). 
 
 OBSERVATION. This method has the same trouble as that mentioned in the prece' 
 
 <ling example , the somewhat difficult precipitation and separation of the sulphide of 
 iron and the sulphide of manganese. 
 
 2. Method. The substance reduced to an impalpable pow- 
 der is dissolved in nitric acid ; pure mercury is added to the 
 solution, which is then evaporated to dryness on the water- 
 bath. The evaporating dish is covered during the evapora- 
 tion with a concave glass ; and the amount of mercury taken 
 must be sufficient to leave a little excess in the dry mass. If 
 the latter still smells of nitric acid, it is moistened with wa- 
 ter and again evaporated to dryness ; and this operation must 
 be repeated, until no more acid vapors are evolved, that is, 
 until a glass rod moistened with ammonia no longer gives rise 
 to white vapors when it is held over the dish. The residue 
 is now covered with water, warmed, filtered and then wash- 
 ed until the wash-water, when evaporated to dryness on a 
 piece of glass, no longer leaves any residue. The filter now 
 contains all the phosphoric acid combined with protoxide of 
 mercury, mixed with the nitrates of the protoxide and of the 
 oxide of mercury and metallic mercury ; in addition to these 
 it contains a large portion of the oxide of iron ; and in the 
 filtrate are found the nitrates of all the bases previously in 
 combination with the phosphoric acid. 
 
 The filter is dried together with its contents ; the latter are 
 then shaken into a capacious platinum crucible arid mixed 
 with about four times their weight of pulverized and dry car- 
 bonate of soda. The filter, containing still some adhering 
 particles, is rolled together into a round ball and placed in 
 the middle of the powder. Finally the whole is covered with 
 a layer of carbonate of soda, and the crucible is covered up 
 and heated in a place where the noxious vapors of mercury 
 may escape, at first a long time at a gentle temperature, be- 
 low ignition, taking care not to fuse the mass.* At last, 
 however, the contents are ignited and thoroughly fused. As 
 
 *If the mass is not dry and is quickly heated, the crucible will be much corroded. 
 
soon as the crucible is cold, it is placed in water and warm 
 ed. By this treatment all is dissolved with the exception of 
 a quantity of sesquioxide of iron, which is removed by filtra- 
 tion and for the present set aside. Hydrochloric acid is now 
 added to the filtrate to give it an acid reaction ; and the phos- 
 phoric acid is precipitated by ammonia and the magnesia 
 mixture. The filtrate from the bases, which contains also a 
 certain amount of the nitrates of the protoxide and of the ox- 
 ide of mercury, is evaporated to dry ness ; the residue is heat- 
 ed in order to volatilize the mercurial salts (observing the- 
 precautions above-mentioned for the escape of the mercurial 
 vapors) ; hereupon it is mixed with crystallized oxalic acid 
 and a little water, concentrated, gradually heated and finally 
 ignited ; that which is soluble is extracted with hot water. 
 
 The residue consists of the sesquioxide of iron, the oxide 
 of manganese and magnesia ; whilst the alkalies as carbonates- 
 will l)e found dissolved in the fluid. These, together with 
 the oxide which had been previously set aside, are dissolved 
 in hydrochloric acid ; the diluted solution is then mixed with 
 acetate of soda and heated, whilst a current of chlorine is pass- 
 ed through it to saturation. Ammonia is now added in ex- 
 cess; and the mixture is then boiled to expel all the free am- 
 monia. The oxides of iron and manganese are removed by 
 filtration; dissolved in hydrochloric acid to which a little al- 
 cohol has been added, then separated by carbonate and ace- 
 tate of soda (vide page 67). Magnesia is precipitated from 
 the filtrate by means of phosphate of soda (vide page 137). 
 
 The solution of the carbonates of the alkalies are decom- 
 posed by hydrochloric acid and evaporated to dryness ; and 
 as soon as the weight of the chlorides has been determined* 
 they are separated by the method already given. 
 
 7. PHOSPHORIC acid, ALUMINA, OXIDE of IKON, 
 MAGNESIA. 
 
 There is no salt in which phosphoric acid so easily escape* 
 detection as in the phosphate of alumina, from the fact that 
 
5't comports itself in regard to several reagents like pure alu- 
 mina. Especially does its analysis present difficulties, which 
 increase in proportion according as the oxide of iron, lime 
 and magnesia are also present. These constituents occur not 
 only in a few rare minerals (Kal ait, Wavellite, Lazuli te, Chil- 
 drenite, *fee.), but also in soils; and several bog iron ores con- 
 tain, besides the constituents mentioned on page 65, alumina. 
 
 The substance is mixed in a platinum crucible with an equal 
 quantity of pure silicic acid and six times its weight of carbo- 
 nate of soda, it is then gradually heated, and finally ignited.* 
 As soon as the mass is cold, it is digested with water, in or- 
 der to dissolve out all that is soluble, and then filtered. The 
 filtrate is heated with an addition of chloride of ammonium, 
 whereby silicic acid and alumina still remaining are precipita- 
 ted ; the precipitate is also removed by filtration. The fil- 
 trate is first neutralized with hydrochloric acid, and then 
 with ammonia ; after which the phosphoric acid is precipita- 
 ted by the magnesia-mixture. The residue on the two fil- 
 ters, consisting of silicic acid, alumina, oxide of iron, lime 
 and magnesia, are brought together and dried; the dried 
 mass is placed in a dish (also the ashes of the two filters), di- 
 gested in hydrochloric acid, and evaporated on the water-bath 
 to dryness. The dry mass, when cold, is moistened with acid 
 and, after standing for some time, it is heated with water 
 and filtered. 
 
 As soon as the silicic acid has been removed, chloride of 
 ammonium is added to the filtrate in proportion to the mag- 
 nesia present ; the fluid is then neutralized with ammonia in 
 excess and boiled until all fumes of ammonia disappear. Af- 
 ter filtration, the oxide of iron and alumina remain on the 
 filter, these are washed with hot water, dried, ignited and 
 weighed. The simplest method of determining the relative 
 proportion of each of these two substances consists in redu- 
 
 *If the materials are found In solution of hydrochloric acid, carbonate of soda is add- 
 ed to neutralize the solution ; this is then evaporated to dryness. The residue i* placed 
 in the crucible and mixed with the silicic acid and four times its weight of carbonate of 
 soda. 
 
128 
 
 cing a weighed quantity to an impalpable powder and them 
 to heat it with hydrochloric acid and zinc, and finally to esti- 
 mate the amount of the chloride of iron volumetrically. 
 The difference of weight is that of the alumina. 
 
 Ammonia and oxalic acid precipitate the lime from the fil- 
 trate (vide page 131), whilst magnesia is thrown down with 
 phosphate of soda (vide page 133). 
 
 VE. CARBONATES. 
 
 Estimation of the carbonic acid. The amount of carbonic 
 acid is frequently computed from the loss incurred when the 
 bases in combination with it are determined directly, in case 
 the substance contains no water, or when the degree of oxi- 
 dation of the compound is already known. 
 
 If, however, the estimation is to be made by a direct meth- 
 od, this can be effected essentially in two ways : 
 
 1.) by igniting the substance, whereby the carbonic acid is- 
 expelled, and its weight determined by the amount of loss ; 
 
 2.) by means of an acid, which in like manner drives off 
 the carbonic acid ; the amount of the latter, too, is found in 
 a similar manner, 
 
 1. Estimation of the carbonic acid by ignition. 
 
 This method is far from being applicable to all carbonates. 
 It presupposes that carbonic acid may be expelled by a mo- 
 derate red heat, which is not the case, for instance, with the 
 salts of the alkalies, of baryta, strontia and with difficulty 
 with carbonate of lime; and furthermore that the remaining 
 base undergoes no change, is not raised to a higher degree of 
 oxidation, a circumstance that does not take place with the 
 salts of the protoxides of iron and manganese and of the ox- 
 ide of cobalt. If nothing of this sort does occur, the ignition 
 is effected in a small crucible over the lamp, taking care to- 
 repeat the operation, after the weight has been determined r 
 in order to be certain that the ignition the first time had not 
 
129 
 
 been stopped too soon, which is known by the fact that the 
 weights in both cases remain the same. 
 
 The direct determination of carbonic acid, in general, is 
 seldom undertaken, and especially is this the case with those 
 carbonates, whose bases suffer some change when ignited in 
 the air. When such a determination has to be made, the ig- 
 nition must be performed in some other indifferent gas, for 
 instance, in carbonic acid itself and not in air. 
 
 If the carbonate, whose acid is to be determined by igni- 
 tion, contains also water, it is ignited in a small glass retort 
 or some other similar apparatus, attached to a receiver con- 
 taining chloride of calcium. 
 
 2. Estimation of the carbonic acid by another acid. 
 Since the stronger acids liberate carbonic acid from all its 
 combinations in the form of gas, they are had recourse to in 
 all those cases in which the ignition of the substance is not ap- 
 plicable. This method is two-fold : either the volume of the 
 evolved gaseous acid is determined over mercury in a gradu- 
 ated tube ; or the loss of weight of the apparatus, which is 
 caused by the evolution of the gas, is ascertained. The first 
 method is not so frequently used, because it presupposes that 
 the mercurial apparatus is always at hand, and especially a 
 practical experience in analytical investigations with, gases. 
 
 The second method is the one in general use. Although 
 in principle it is exceedingly simple, yet certain precautions 
 must be taken, especially such as aim to prevent the escape, 
 along with the carbonic acid, either of water or of a part of 
 the acid employed in producing the decomposition in the 
 form of vapor. On this account the whole arrangement must 
 be of such a nature as to be easily placed on a sensitive ba- 
 lance and weighed. 
 
 Among the various forms of apparatus for this purpose we 
 give preference to that of Geissler. It consists of three parts : 
 the flask, the vessel for the acid, and the drying apparatus. 
 The first is a small flask with a flat bottom, in which the sub- 
 stance is weighed and decomposed by the acid It has two 
 17 
 
130 
 
 necks ; through one of these a glass tube passes and extends 
 nearly to the bottom of the flask, terminating here with a 
 small orifice ; but on the outside of the flask this tube is wi- 
 dened out into a cylindrical shape for the reception of the 
 *acid employed to produce the decomposition. In order that 
 the acid may be made to flow or drop at will into the flask, a 
 long narrow glass tube, open at both ends, leads through it, 
 the raising and sinking of which regulates the flow of the a- 
 cid. The drying apparatus, consisting of two small cylind- 
 ers, is fitted into the second neck ; it is partly filled with con- 
 centrated sulphuric acid and so arranged, that the carbonic 
 acid, as it is evolved, has to pass through it and is thereby 
 dried. The separate parts of this apparatus are fitted into 
 each other air-tight by grinding. 
 
 When about to be used, the flask is weighed alone, then the 
 substance in rough powder is poured into the flask, which is 
 again weighed; the difference is the weight of the substance. 
 A small quantity of water is now poured upon the powder ; 
 acid (in general hydrochloric acid) is next introduced into 
 the vessel intended for its reception ; this acid, although con- 
 centrated, must not emit any fumes ;* the acid vessel is pla- 
 ced on the flask ; the drying apparatus is adjusted in its place ; 
 and finally the whole arrangement is accurately weighed. 
 By letting air into the glass tubef a small quantity of acid 
 is allowed to flow into the flask gradually so as not to give 
 rise to a too rapid current of gas, and this is continued un- 
 til there is no longer any effervesence, and the carbonate is 
 dissolved, with the exception perhaps of a small insoluble 
 residue of some foreign material. The apparatus is now fill- 
 ed with carbonic acid ; and the fluid in the flask is saturated 
 with the gas. The flask therefore is heated but never to 
 boiling in order to expel all the carbonic acid from the appa- 
 ratus ; this is effected by taking the cork out of the tube and 
 
 *The same apparatus is used for the analysis of black oxide of manganese described, 
 page 70 ; but in this case sulphuric acid is used to produce the decomposition. 
 
 tEven when the terminal point of the acid receiver dips into the fluid, it is necessary 
 to close the tube previously with a cork,&c. 
 
131 
 
 drawing through a current of air at the extremity of the 
 drying apparatus either with the moiith or by means of the 
 aspirator ; a tube filled with lumps of potassa is attached also 
 to the glass tube. The loss of weight which the apparatus 
 has suffered is owing to the carbonic acid that 'has been ex- * 
 pelled from the carbonate. 
 
 The fluid in the flask may in general be employed for the 
 determination of the bases. 
 
 1. LIME, MAGNESIA, CAEBONIC ACID. 
 
 (BiTTEKSPAK, DOLOMITE.) 
 
 A weighed quantity of the substance in powder is placed 
 in a flask and dissolved in hydrochloric acid, applying a gen- 
 tle heat until all the carbonic acid has been driven off, andj 
 in case there should be any insoluble residue, on the ad- 
 dition of a few drops of acid there should be no more gas eli- 
 minated. The fluid is then diluted ; the insoluble residue 
 is removed by filtration ; it consists chiefly of quartz, which 
 is to be well washed. 
 
 Ammonia is added to the filtrate until an alkaline reaction 
 is produced on litmas paper, or until there is a decided smell 
 of this alkali ; then a solution of oxalic acid is added gradu- 
 ally in order to precipitate the lime, and until no longer any 
 precipitate is produced by this reagent. But even at the end 
 of the operation the fluid must be alkaline, otherwise a part 
 of the lime would remain in solution, on which account it 
 may be necessary towards the conclusion to add more ammo- 
 nia. The vessel is set aside for a number of hours in order 
 that the precipitate may settle completely and the fluid be- 
 come quite clear ; the latter is then tested with a few drops 
 of the precipitant to ascertain whether it contains any more 
 lime. 
 
 The fluid is poured upon a filter and afterwards the oxa- 
 late of lime ; the vessel that had been used is then thorough- 
 ly rinsed out by means of a feather and the wash-bottle ; and 
 the residue on the filter is washed with hot water until a few 
 
132 
 
 drops of the wash-water, when evaporated to dryness on a 
 piece of platinum foil, leave no observable marks. The re- 
 sidue is finally dried in the air and afterwards at a low red 
 heat in order to convert it into carbonate of lime. 
 
 But since the residue, if submitted to a more intense red 
 heat, is apt to lose a part of the carbonic acid, and the weigh- 
 ing of such a mixture of carbonate of lime and caustic lime 
 would in every case give an incorrect result, it is necessary 
 to employ several precautions whilst heating the precipitate. 
 
 As soon as the dried oxalate of lime has been shaken olf 
 from the filter as much as possible into a platinum crucible, 
 the filter itself is burned on the lid of the crucible, and the 
 ashes are added to the rest in the crucible, which is now 
 thoroughly covered, and gradually heated over the lamp un- 
 til the lower part of the crucible is raised to a low red heat, 
 at which temperature it is kept for some time. After it is 
 cool, it is weighed. The contents are then moistened with 
 a few drops of a solution of carbonate of ammonia and then 
 they are heated again as before. If the two weights exactly 
 correspond, it is evident that no carbonic acid has been lost 
 in the first experiment ; but should not this be the case the 
 correction and heating are repeated until the two weights are 
 exactly the same. From the carbonate of lime thus received, 
 the weight of the lime in the substance is computed. 
 
 The convertion of the oxalate of lime into caustic lime is 
 much more quickly effected by the use of a powerful gas- 
 blast. As soon as the precipitate has been already strongly 
 ignited in the air, the covered crucible is held in the gas- 
 blast from a few minutes to a quarter of an hour according 
 to the quantity of lime. The crucible is weighed, and the 
 ignition repeated. The lime at this stage, when first moist- 
 ened with water, must dissolve in acids without effervescence. 
 If the amount of lime is pretty large, it is advisable to moist- 
 en the contents of the crucible with water after ignition for 
 the first time and then to apply a gentle heat, before it is a 
 second time ignited. 
 
133 
 
 There remains now only the magnesia to be determined 
 m the filtrate from the oxalate of lime. A rather large quan- 
 tity of phosphate of soda is added to the fluid, and dissolved 
 by the application of heat ; then ammonia is added to the flu- 
 id which is now set aside for twenty-four hours. By this 
 means the double phosphate of ammonia and magnesia is gra- 
 dually thrown down. The residue is then removed by filtra- 
 tion and washed with a cold mixture of one part of ammo- 
 nia (spec. grav. 0.96) and three parts of water, (because it is 
 soluble to a slight extent in pure water alone) until the last- 
 drops, when evaporated to dry ness on a strip of platinum, 
 leave no marks behind. The contents of the filter are now 
 dried on the filter, and afterwards ignited ; the filter itself is 
 incinerated, an operation which, owing to the saturation of 
 the fibres of the paper with the salt is rather slow ; the ash- 
 es are added to the contents of the crucible. At the begin- 
 ning the crucible is covered, but afterwards the cover is re- 
 moved, and a strip of platinum foil is placed over the edges 
 in order to institute a draft of air. By this ignition of the 
 precipitate ammonia and water are expelled, and a rapid 
 extinction of flame is produced ; finally there remains py- 
 rophosphate of magnesia (2MgO-hPO 5 ), from which the 
 amount of magnesia is computed. 
 
 The filtrate is again mixed with phosphate of soda, is again 
 set aside, and as a general thing is never thrown away, until 
 in this way a certainty is obtained that all the magnesia has 
 been precipitated. If there should be still any precipitate, 
 it is placed with that, previously obtained, on the filter. 
 
 It often happens, that a carbonate of lime and magnesia 
 contains also iron, either in mixture as oxide, or as a carbo- 
 nate of the protoxide in isomorphous mixture. If the amount 
 of iron is very small, it is either neglected, or it is found in 
 the lime, which is dissolved in an acid and supersaturated 
 with ammonia. The precipitation and filtration of the ses- 
 quioxide of iron must take place in a closed vessel (vide the 
 following example). When the quantity of iron is large, the 
 
134 
 
 process is the same as that in example second. 
 
 If when ammonia is added, at the beginning of the analy- 
 sis, in excess to the solution there should be produced a light 
 colored precipitate, it is magnesia, and it is caused by a de- 
 ficiency of ammoniacal salt. The whole is then supersatura- 
 ted with hydrochloric acid and again neutralized with am- 
 monia. 
 
 2, PEOTOXIDE of IKON, protoxide of MANGANESE, 
 CAKBONIC acid, LIME and MAGNESIA. 
 
 (SPATHIC IRON ORE, BITTERSPAR.) 
 
 The substance in powder is digested in a flask with hydro- 
 chloric acid, until by the addition of a few drops, of acid 
 no more change is observable, or until all is dissolved. The 
 protoxide of iron (protochloride) is converted into the ses- 
 quioxide (chloride) by adding to the fluid, still hot, a small 
 quantity of chlorate of potassa, until it assumes a yellow color 
 and begins to smell of chlorine. The solution is now heat- 
 ed, and, in case there should be any insoluble substance 
 (quartz, &c.), the latter is thoroughly washed, dried, ignit- 
 ed and weighed. 
 
 The solution is made very dilute in a capacious dish, and r 
 being covered with a plate of glass, it is gently heated and 
 so long mixed with powdered carbonate of soda, which is ad- 
 ded in small portions at a time, until the color becomes a 
 dark-red, and a part of the sesquioxide is precipitated as a 
 basic salt. (The greater the quantity of iron compared with 
 that of the manganese, the greater will be the amount of pre- 
 cipitate of the former, but the fluid must still always have a 
 red color . When the amount of iron is small, in order not 
 to go beyond the right point, carbonate of ammonia, at the 
 last preferably in solution, must be added only as long as the 
 fluid becomes colored.) Then a quantity of acetate of soda 
 (from 2 to 10 grammes), corresponding to the proportion of 
 iron, is added and the solution is heated to boiling. By this 
 means all the sesquioxide of iron is precipitated as a basic salt, 
 
135 
 
 whilst protoxide of manganese, lime and magnesia remain in 
 solution. The latter is filtered whilst hot, and the residue is 
 washed with hot water ; but whatever adheres firmly to the 
 dish is left there, and in this dish the filter, containing the 
 washed residue, is placed and digested with dilute hydrochlo- 
 rjp acid. This solution is filtered, and to the filtrate ammo- 
 nia is added, this throws down the sesquioxide of iron which 
 is washed with hot water, dried, ignited and weighed. 
 
 The filtrate, which contains the remaining bases, is again 
 mixed with acetate of soda, gently heated and saturated with 
 chlorine. Ammonia is added in excess, and the mixture is 
 then boiled until the excess has all been expelled. The oxide 
 of manganese is then removed by filtration, washed with hot 
 water, dried and ignited at an intense red heat, by which 
 it is converted into the protoxide. The filtrate is neutraliz- 
 ed with ammonia, the lime is then precipitated with oxalic 
 acid, and the magnesia with phosphate of soda (vide exam- 
 pie 1.). 
 
 3. OXIDE of LEAD, CABBONIC ACID, WATEE. 
 
 (WHITE LEAD.) * 
 
 The purest white lead contains only traces of the acetate, 
 sulphate and chloride of lead. But the common sorts con- 
 tain heavy spar (which by solution in acetic acid remains un- 
 dissolved), chalk, &c. 
 
 A part is ignited in a porcelain crucible. By this means 
 a portion of the oxide of lead is obtained. Another part is 
 used for the determination of the carbonic acid, for which 
 purpose nitric acid is used as the solvent. 
 
 The amount of water is obtained by the loss of weight. 
 
 VH. BORATES. 
 
 In the analysis of borates either the base is alone determi- 
 ned, the acid computed out of the loss, or it is also determi- 
 ned in a direct manner. 
 
136 
 
 In the first case the method to be selected depends on the 
 nature of the base : if this is a metallic oxide proper, it is se- 
 parated by hydrosulphuric acid ; if it is the oxide of lead, 
 baryta, strontia or lime, the separation is effected by sulplju- 
 ric acid (and alcohol) ; and if it is magnesia and the compound 
 is easily dissolved in an acid, the base is precipitated by phos- 
 phate of ammonia. In like manner almost all the borates 
 both of the alkaline earths and of the metals may be decom- 
 posed by ignition with carbonate of soda ; when treated with 
 water the borate of soda is dissolved, and the base remains 
 behind either free or in the form of carbonate. 
 
 But if the borate is a salt of an alkali, or if the compound 
 is with difficulty dissolved in acids, it is converted by means 
 of concentrated hydrofluoric acid into a mixed fluoride and 
 borofluoride, which is heated with sulphuric acid, and then 
 the base is determined either as a sulphate, or may be sepa- 
 rated from this in an especial manner. I have used this me- 
 thod with great advantage in the analysis of the native bo- 
 rate of magnesia, that is, boracite.i .Qj 
 
 The direct determination of boracic acid by precipitation 
 with a salt of one of the alkaline earths or of a metal, as may 
 be done with so many other acids, does not succeed on ac- 
 count of the solubility and easy decomposition of such borate 
 precipitates. It. is only possible in the form of borofluoride 
 of potassium, a method which was first proposed by Berzeli- 
 us and afterwards improved by A. Stromeyer. 
 
 To the solution of boracic acid, which must not contain 
 any other bases except the alkalies, half as much again of 
 hydrated potassa as the amount of boracic acid, is added ; after 
 this pure hydrofluoric acid is added to the mixture in a pla- 
 tinum dish in excess, so that the fluid has not only an acid 
 reaction, but exhales when heated acid vapors. The mixture 
 is evaporated to dry ness on the water-bath. A solution of 
 one part of acetate of potassa in five parts. of water is poured 
 upon the residue, which is set aside for a number of hours at 
 the common atmospheric temperature, stirring it frequently ; 
 
13Y 
 
 the borofluoride of potassium, that is thus thrown down, is 
 separated by filtration on a weighed filter and washed with a 
 similar solution, until the wash-water is no longer precipitat- 
 ed by chloride of calcium, and finally with concentrated al- 
 cohol. The salt is dried at a temperature of 212F., and from 
 this the amount of boracic acid is computed. 
 
 If the solution of the boracic acid contains many salts, es- 
 pecially salts of soda or sulphate of potassa, it is well to heat 
 the dried mass with the solution of acetate of potassa and to 
 set it then aside from twelve to twenty-four hours, before it 
 is filtered. 
 
 1. LIME, SODA, BOEACIO ACID and WATEE. 
 
 (BOKONATROCALCITE, TIZA.) 
 
 The mineral contains, in addition, small quantities of po- 
 tassa, chlorine and sulphuric acid. 
 
 a. It is heated in a platinum-vessel with a mixture of hy- 
 drofluoric acid (or fluoride of ammonium) and sulphuric acid; 
 it is then evaporated to dryness ; and the residue is heated 
 to long ignition. When treated with water a large portion 
 of sulphate of lime remains undissolved ; this is removed by 
 filtration, afterwards thoroughly washed, dried, ignited and 
 weighed. The remaining part of the lime in the filtrate is 
 thrown down by means of ammonia and oxalic acid (vide page 
 131) ; the fluid is evaporated to dryness ; the sulphate of soda 
 is placed in a platinum crucible, ignited, and rendered neu- 
 tral by means of carbonate of ammonia (vide page 113). If 
 it be required to determine the small amount of potassa, 
 which it contains, it is weighed and then converted into 
 chlorides ; the two alkalies are separated by chloride of pla- 
 tinum (vide Silicates, analysis by means of hydrofluoric acid), 
 b. The amount of water is determined by igniting a part 
 of the substance either alone or with a weighed quantity of 
 oxide of lead. 
 
 c. For the determination of the chloride and the sulphu- 
 ric acid separate portions of the substance are respectively 
 18 
 
138 
 
 employed ; the former is determined by means of nitrate of 
 the oxide of silver, the substance having been previously 
 dissolved in cold dilute nitric acid (vide page 118) ; the lat- 
 ter by chloride of barium in a dilute solution of hydrochlo- 
 ric acid. 
 
 The amount of boracic acid is derived from the loss. If it 
 is to be determined directly, the substance is dissolved in hy- 
 drochloric acid; the lime is precipitated with carbonate of 
 potassa ; and the nitrate is treated as directed, page 136. 
 
 2. MAGNESIA, BOEAOIO acid, CHLOEINE, 
 WATEE. 
 
 (STASSFURTHITE.) 
 
 a. The substance previously reduced to a fine powder is 
 fused with four times its amount of pure carbonate of potassa, 
 the mass is then digested in hot water and filtered ; magne- 
 sia remains behind and is washed in hot water, dried, igni- 
 ted and weighed. The boracic acid in the filtrate may be de- 
 termined directly according to instructions given on page 136. 
 
 b. Another portion is dissolved in dilute nitric acid from 
 the solution of which the chlorine is determined as in the 
 preceding example. 
 
 c. If the water is to be determined directly, it is effected 
 by ignition with oxide of lead (vide page 114). 
 
 IX. SILICATES. 
 
 The compounds of silicic acid, comprehending as they do 
 a large portion of the minerals as well as all the slags and 
 vitreous products of fusion, are in analytical chemistry of 
 particular importance, and so much the more so from the fact, 
 that the separation of the silicic acid from the bases in com- 
 bination with it requires different methods in the different 
 cases, and in general is frequently no easy matter, For al- 
 though alone, when ignited, it may be regarded as insoluble 
 
139 
 
 in water and in almost all acids, yet it occurs, when its salts 
 have been decomposed, in a condition in which this property 
 of insolubility is no longer so distinct a characteristic, so that 
 a certain smaller amount remains behind in acid solutions, 
 and is separated at the same time with the precipitated bases, 
 thus rendering the separation and accurate determination of 
 this substance a difficult operation. 
 
 All silicates, by reason of their comportment with acids 
 (hydriodic, nitric, sulphuric) are divided into two classes : 
 
 A. Those which are decomposed by them, and 
 
 B. Those which are not decomposed by them. 
 When a silicate is decomposed by an acid, the silicic acid 
 
 is separated, whilst the bases, previously in combination with 
 it, are dissolved in the stronger acid. But the silicic acid thus 
 precipitated has not always the same appearance ; sometimes 
 it is thrown down in a gelatinous form, so that after diges- 
 tion in the proper amount of acid, the silicic acid forms on 
 cooling a stiff sort of jelly. Silicates of this description are 
 said to gelatinize. In other cases the silicic acid, when it is 
 precipitated, assumes a slimy, flocculent or pulverulent ap- 
 pearance. 
 
 The silicates, that thus undergo decomposition in acids, 
 form the smaller class of the two ; belonging to this class may 
 be reckoned many minerals, especially those containing water, 
 and designated as the family of the Zeolites (Mesotype, Ap- 
 ophyllite, Analcime, Chabasite, Stilbite, Heulandite, &c.), 
 and besides these many others. In this class also may be 
 comprehended many of the artificially formed silicates, espe- 
 cially several slags, such as iron slags and artificial ultrama- 
 rine, &c. 
 
 With the majority of these substances, yet not with all, 
 especially with the hydrous class, the decomposition by means 
 of acids is either entirely prevented by ignition or consider- 
 ably impaired, and it thus frequently happens, that the 
 analysis can not be continued with the same portion which 
 had been ignited for the estimation of the water. That this 
 
140 
 
 is not the case with those which owe their formation to a fu- 
 sing process, is self-evident. 
 
 On the other hand it happens, that a silicate of the second 
 class, which is not decomposed in acids or only with difficulty 
 decomposed, may ~be decomposed by acids after ignition or 
 fusion and as a rule so completely as to gelatinize. This in- 
 teresting fact was observed by Fuchs, von Kobell and Mag- 
 nus in regard to Epidote, Garnet and' Yesuvian ; and I have 
 also confirmed the fact in reference to Axinite. This change 
 is allied to the transition of the silicate from the crystalline 
 to the amorphous condition, and is accompanied with a dimi- 
 nution of its specific gravity. On this account ignition or 
 fusion may be with propriety used in the analysis of such 
 minerals ; for the process of those, which undergo decompo- 
 sition by means of acids, is more simple than of those in the 
 opposite condition. 
 
 Nevertheless it must be acknowledged, that the line of de- 
 marcation of these two classes of silicates in reference to their 
 comportment with acid, is not absolutely fixed, for, strictly 
 speaking, there is not a single one in the pulverized condi- 
 tion that can resist the continued action of a concentrated a- 
 cid, especially when heated; and the felspar and micacious 
 minerals, hornblende and augite, which belong to the second 
 class, suffer in this case, according to the duration of the ac- 
 tion, a more feeble or intense decomposition. If, therefore, 
 our classification is to be regarded as correct to a certain ex- 
 tent in accordance with science, we must in every case compre- 
 hend in the category of silicates which suifer decomposition 
 by means of acids only those, whose decomposition proceeds 
 easily and quickly r , although this takes place in very diffe- 
 rent degrees among the different substances placed in this 
 class, so much so indeed that sometimes a silicate, which may 
 be accurately enough decomposed by means of an acid for a 
 qualitative investigation, must not be treated by the same 
 means fora quantitative analysis, but has to be first fused 
 wii an alkali. 
 
m 
 
 It must not, however, be supposed when a silicate is de- 
 composed by an acid, that all the silicic acid is set at liberty, 
 nor that that, which is precipitated, is always pwre silicic 
 ,acid. As regards the first point it is to be observed, that 
 there remains in the acid fluid no inconsiderable part of sili- 
 cic acid in solution, which is subsequently thrown down to- 
 gether with the bases present. On this account it becomes 
 necessary, after these precipitated bases have been ignited 
 .and weighed, to separate the small amount of silicic acid mix- 
 up with them ; this is effected by digesting them respective- 
 ly with an acid (hydrochloric acid), which leaves the silicic 
 acid undissolved. But again here are the following trou- 
 bles : on one side a portion of the latter is again dissolved ; 
 .and on the other some of the ignited bases (alumina, sesqui- 
 oxide of iron) are with difficulty soluble in acids, a circum- 
 stance which renders the separation any thing but easy and 
 certain. Or the decomposed silicate is evaporated to dryness, 
 so that the residue appears dusty ; it is then treated with wa- 
 ter, which dissolves the bases in combination with the acid 
 used to produce the decomposition, but leaves the silicic acid 
 undissolved. But in this case also a part of the latter is dis- 
 solved in the acid, although the quantity is small, and which 
 in many cases may be neglected. But there is another diffi- 
 culty when a decomposed silicate is evaporated to dryness, it 
 3s this, the weaker bases (alumina, sesquioxide of iron) are 
 easily converted into basic salts or partly thrown down in the 
 basic form, and in this condition are insoluble in water, con- 
 sequently they remain in mixture with the silicic acid. It is 
 therefore indispensable to moisten the dry mass, left by eva- 
 poration, uniformly with hydrochloric acid before it is treat- 
 ed with water, and to allow it to stand for an hour or so, in 
 order that those bases may again combine with the acid which 
 they had parted with. With silicates rich in iron it will, how- 
 ever, be easily observed, that, if the evaporation had been 
 carried so far as to produce the red color of the sesquioxide 
 of iron in the mass, even although treated with the acid, the 
 
142 
 
 silicic acid can no longer be obtained free from oxide of iron, 
 On this account it becomes necessary always to pay great at- 
 tention to the temperature employed in the evaporation, and 
 in cases like that just alluded to, it is safest to complete the 
 work on a water-bath, only not forgetting, as already observ- 
 ed, to continue the heat so far as to produce a residue perfect- 
 ly dry and of a dusty appearance. 
 
 As regards the second point, namely, that the precipitate 
 produced by treating a silicate with an acid must not be, at 
 once, looked upon as pure silicic acid, this point is of great 
 importance in analysis, and it is only lately that it has been 
 accurately recognized. Even when a mineral apparently has 
 been so perfectly decomposed by an acid, that the silicic acid 
 separates in the gelatinous form, it can not ahvays be assumed 
 with certainty, that this is pure silicic acid. But because this 
 has been the rule hitherto, it is quite possible that the deter- 
 mination of the silicic acid in many minerals is not right. 
 In order to be convinced that the silicic acid, obtained by de- 
 composing a silicate by means of an acid, was pure, the ac- 
 id so obtained was ignited and weighed, then boiled with a 
 concentrated solution of carbonate of sod a; this dissolved the 
 silicic acid; the residue was regarded as undecomposed sili- 
 cate and was deducted from the amount of silicic acid as 
 well as from the amount of substance originally employed. 
 
 This, may certainly be right, if the silicate employed for 
 the analysis had not been previously reduced to a sufficiently 
 fine powder. But even when the pulverization has been 
 performed with care, and the mass has been decomposed 
 with hydrochloric acid so well that it is not possible to find 
 any undecomposed, hard and therefore gritty particles by 
 means of a glass rod, it often happens, that silicic acid when 
 boiled with carbonate of soda will leave an undissolved resi- 
 due. It is thus that H. Rose* discovered, that silicic acid, ob- 
 tained by the decomposition of titanite by means of hydro- 
 chloric acid, and treated with a solution of carbon ate of soda r 
 
 . ANK. Vol. 62, page 253. 
 
U3 
 
 leaves as much as thirty per -cent of residue. The amount of 
 tliis residue is at one time larger and at another smaller, al- 
 though powder of the same degree of fineness is employed. 
 If the silicate is easily decomposed by means of hydrochloric 
 acid, and the silicic acid separates in a gelatinous form, the 
 latter contains more of insoluble residue if the powder of the 
 silicate is treated with highly concentrated hydrochloric acid, 
 and the two are not intimately mixed together by continual 
 stirring so that the gelatinous substance is soon formed. On 
 the contrary the residue left, after boiling the silicic acid in 
 a solution of carbonate of soda, is small if the powdered sili- 
 cate has been stirred about for some time in a more dilute acid 
 whereby the formation of the gelatinous substance is prevent- 
 ed from taking place, and especially if the acid is left to act 
 on the powder as long as possible. 
 
 The combination of the residue in question is by no means 
 that of the silicate; and, on this account, its weight in this 
 state must not be deducted. It is only when the substance 
 has not been previously reduced to a sufficiently fine powder, 
 that a part may remain undecomposed and mixed with the 
 silicic acid. Otherwise, however, this undissolved residue in 
 the solution of carbonate of soda consists in general of silicic 
 acid and small particles of the bases, and the amount of the 
 former in the zeolites, according to H. Hose, is from 90 to 
 97 per cent ; the remaining part is in general alumina and 
 lime. 
 
 Of all the silicates those, which contain either titanic 
 acid or zirconic acid, appear to produce the largest quantity 
 of such residues. As already mentioned, H. Rose has called 
 particular attention to this circmstance in the analysis of tita- 
 nite, and I have confirmed this result in the analysis of eudi- 
 alite a silicate, containing zirconic acid, sesquioxide of iron, 
 lime and soda in which the residue separated from the pre- 
 cipitated silicic acid appears to be at the same time a distinct- 
 compound,* as also of that of cerite, the half of whose silicic 
 
 *PooKND. ANN. Vol. 63, page 142. 
 
144 
 
 acid set free by hydrochloric acid consists of bases. * 
 
 If in the analysis of a silicate the amount of such insoluble 
 residue is only very inconsiderable, the error indeed is not 
 very great if it be regarded as undecomposed substance, and 
 still smaller if it be regarded as silicic acid. But the error 
 becomes more considerable in the examination of rocky ma- 
 terials consisting of a silicate, that may be decomposed by 
 means of acids, and of another that resists their action. If 
 the whole substance is treated with an acid, the insoluble 
 portion consists of the silicate that can not be so decom- 
 posed and of the silicic acid of that portion that can be so de- 
 composed. If the insoluble residue is now boiled in carbo- 
 nate of soda, in order to dissolve out the silicic acid, a part 
 of the latter remains in combination with small quantities of 
 the bases, and this consequently will be the cause, in the a- 
 nalysis of the residue, of finding more silicic acid than pro- 
 perly belongs to the silicate not decomposible by acids. 
 
 There is a very simple and easy means of ascertaining the 
 purity of the silicic acid obtained in analyses, which also may 
 be employed with advantage in the subsequent investigation 
 of the residues in question. It consists in the employment 
 of hydrofluoric acid. If the silicic acid is covered with this 
 acid in a platinum dish, and the fluid evaporated on a water- 
 bath to dryness, there ought to be no residue, if the silicic 
 acid were pure, because it volatilizes as fluosilic acid. If, on 
 the contrary, there is any residue, it consists of the silico- 
 fluorides of the metals of the bases combined with the silicic- 
 acid. Finally sulphuric acid is added, which converts the 
 silico-fluorides by heat into sulphates, which admit of being 
 easily analyzed. Fluoride of ammonium may, with equal 
 success, be substituted for hydrofluoric acid, which is mixed 
 in a platinum crucible with the silicic acid and a little wa- 
 ter and then heated. 
 
 We will now proceed to the special analytical processes 
 for the two classes of silicates. 
 
 *Pooamn>. ANN. Vol. 107, page 631. 
 
145 
 
 The bases contained in these substances are in general the 
 following : 
 
 Potassa, 
 
 Soda, 
 
 Lime, 
 
 Magnesia, 
 
 Alumina, 
 
 Protoxide and Sesquioxide of Iron, 
 
 Protoxide and Binoxide of Manganse. 
 These bases rarely occur all together in one and the same si- 
 licate, but they are so common, and the mode of their separa- 
 tion so simple, that a qualitative analysis is seldom underta- 
 ken beforehand, and it is only necessary when there happens 
 to be any doubt whether iron exists in the silicate in the form 
 of protoxide or sesquioxide, or in both conditions at the same 
 time, as also when it is of importance to ascertain the presence 
 of such bases as occur only in certain silicates and for the se- 
 paration of which the analytical process has to be modified. 
 Of this nature are the following : 
 
 BARYTA (Ilarmotome, Brewsterhite.)* 
 STKONTIA (Brewsterhite.) 
 
 BERYLLIA )Helvine, Gadolinite, Orthite ? Tschewkinite.) 
 YTTBIA (Gadolinite, Orthite, Tschewkinite.) 
 THOKINA (Thorite.) 
 
 PROTOXIDE of CERIUM, OXIDE of LANTHANUM, OXIDE of 
 DIDIMIUM (Cerite, Orthite, Gadolinite, Tschewkinite, Mo- 
 eandrite.) 
 
 ZlRCONIA (ElJDIALITE, WHOELERITE.) 
 
 OXIDE of ZINC (Calamine, Silicious oxide of Zinc.) 
 OXIDE of LEAD (Thorite, Calamine, Lead Slags.) 
 OXIDE of BISMUTH (Eulytine, Silicate of Bismuth.) 
 OXIDE of COPPER (Diaptase, Chrysocolla, Allophane, Cop- 
 per Slags, &c.) 
 
 OXIDE of URANIUM (Thorite, &c.) 
 
 Of Electronegative Bodies the following sometimes occur: 
 
 *In this list are only those silicates that can be decomposed by means of acids 
 19 
 
146 
 
 TITANNIC Aero (Titanite, Tschewkinite.) 
 TANTALIC ACID and NIOBIC ACID (Whoelerite.) 
 BORACIC ACID (Datolite, Botryolite, Bliodicite.) 
 PHOSPHORIC ACID (Ledererite ? Eulytine ?) 
 SULPHURIC ACID (Hauyne, Noseane, Lapis Lazuli, Sko- 
 lopsite, Ittnerite, Ultramarine.) 
 
 CARBONIC ACID (Cancrinite, Stroganowite.) 
 FLUORINE (Chondrodite, Apophyllite, Scapolite.) 
 CHLORINE (Sodalite, Eudialite, Pyrosmalite, Hauyne, La- 
 pis Lazuli.) 
 
 SULPHUR (Helvine, Hauyine, ]N"oseane, Lapis Lazuli, Itt- 
 nerite, artificial Ultramarine and Slags.) 
 
 Water is sometimes a constituent of the silicates of this 
 class, and sometimes it is not. ; 
 
 When fluorine exists as a constituent of the silicate, certain 
 precautions have to be observed in the determination of the 
 silicic acid and the water ; the same may be said also when 
 titanic acid and other rare oxides are in combination with si- 
 licic acid. The determination of the water, too, is modified 
 when chlorine, sulphur, sulphuric acid and carbonic acid are 
 present at the same time. 
 
 GENERAL BULES for the ANALYSIS of SILICATES, 
 
 Pulverization. Every silicate, previous to the commence- 
 ment of its analysis, must be carefully reduced to an impalpa- 
 ble powder, so that neither under the pestle of the mortar 
 nor between the fingers it is possible to distinguish any grit- 
 tiness. It is only when this degree of fineness has been at- 
 tained, that a complete decomposition is to be expected, and 
 it is necessary that the decomposition must be complete, be- 
 cause the amount of that which is not decomposed and which 
 remains in mixture with the precipitated silicic acid, can 
 never be easily and accurately determined. The silicates, 
 that undergo decomposition in acids are never so hard as to 
 require to be first elutriated ; this operation, too, must be 
 avoided in order not to have to dry the powder at a higher 
 
147 
 
 temperature, by which it easily parts with the water in chem- 
 ical combination with it, and becomes more difficult of de- 
 composition in acids, 
 
 Drying. Since all impalpable powders attract moisture 
 from the air, it is necessary to dry the powdered silicate. 
 This is effected quite easily by placing it a short time in a mo- 
 derately warm place, as for instance, in the drying chamber 
 of a stove, previous to weighing it. A method just as good, 
 and in many respects better, is to dry it in the desiccator a 
 few hours over sulphuric acid. 
 
 Decomposition by Acid. Concentrated hydrochloric acid 
 is always employed for this purpose. (If the silicates contain 
 much oxide of lead, it becomes necessary to use nitric acid.) 
 The powder is first weighed in a platinum crucible (in gene- 
 ral from one to two grammes is the quantity operated upon) ; 
 this is then placed in a deep dish (not with a flat bot- 
 tom), a platinum one is the best ; it is then covered with a 
 little water and intimately mixed with it by means of a glass 
 rod, to which hydrochloric acid is now added. The dish is 
 now heated over the flame of a lamp, taking care to stir 
 the mixture quickly all the time, and to leave no part of the 
 powder on the bottom of the dish unstirred. If the decom- 
 position of the silicate is easily effected, the gritty sound of 
 the powder, as the glass rod moves over it, soon ceases, the si- 
 licic acid separates, and the whole gelatinizes in many cases 
 (either immediately or on cooling), and the decomposition 
 is complete. But if the silicate is not so easy of decomposi- 
 tion, or if the powder is not uniformly impalpable, the heat 
 must be continued to incipient boiling. But in every case 
 now the decomposition must be complete. By proceeding 
 in this way the silicic acid is prevented from aggregating a- 
 round undecomposd portions of the silicate ; if this is not ob- 
 viated, it is not possible to produce a perfect decomposition 
 of the silicate. The mixture is now evaporated to perfect 
 dryness on the water-bath, stirring it well towards the last 
 
 *Frequently foreign substances in admixture remain undissolved, causing the gritti- 
 ness still to continue. 
 
148 
 
 in order to disintegrate the rougher portions. If the silicate 
 contains neither alumina nor oxide of iron, the drying can be 
 completed over the lamp, still this must be done with great 
 care and only at a very gentle heat. In all cases, as soon as 
 the residue is cold, it is moistened uniformly with hydro- 
 chloric acid ; the dish is then covered with a plate of glass 
 and set aside for at least an hour. Water is then added ; 
 the mixture is heated ; after which the silicic acid is remov- 
 ed by filtration, washed with hot water, exceedingly well 
 dried and then ignited. 
 
 Of course the filter is also ignited together with the silicic 
 a'cid in the crucible, both being covered up and gradually 
 heated ; as soon as the filter is perfectly carbonized, the lid . 
 is pushed half on one side, and by means of a strip of plati- 
 num a draft is instituted and the carbon is burnt up. If the 
 filter were to be heated in the open air, the gases and the 
 draft produced by the lamp would carry off a considerable 
 amount of the acid in fine dust into the air. As soon as the 
 crucible, covered up tight, has cooled, it is weighed immedi- 
 ately; or the crucible is set aside to cool in the desiccator 
 over sulphuric acid, before it is weighed ; for silicic acid af- 
 ter ignition is exceedingly hygroscopic. . 
 
 Purity of the Silicic Acid. We have already called atten- 
 tion to the circumstance (vide page 141), that the silicic acid 
 obtained in this manner is not pure, but that it contains a 
 smaller or a larger amount of the remaining constituents of 
 the silicate. Although in most cases the quantity if this ad- 
 mixture is only trifling, it is permited only in analysis, whose 
 aim is more technical than scientific, to regard the silicic acid 
 at once as sufficiently pure ; in all other cases it is necessary 
 to examine the silicious residue still further, which, as alrea- 
 dy mentioned, essentially can take place by three methods, 
 each possessing its own peculiar advantages: 
 
 1. J3y Ignition with Carbonate of Soda. To the silicic 
 acid in the platinum crucible are added gradually three times 
 its weight of finely powdered dried carbonate of soda, rubbing 
 
U9 
 
 the mixture as intimately together as possible by means of a 
 ^lass rod, and placing the crucible during the operation on 
 a sheet of paper in order to avoid all loss; the glass tube, al- 
 so, is cleaned from all adhering particles by means of a feath- 
 er or camel's hair pencil. The mixture is then heated to in- 
 tense ignition over the lamp or in a blast furnace ; the cold 
 mass is then detached from the crucible by slightly pressing 
 the sides of the crucible, and placed in a beaker glass, over 
 which is poured a sufficient quantity of dilute hydrochloric 
 acid so as to cover it ; hydrochloric acid is also poured into 
 the crucible in order to dissolve any adhering portions; this 
 solution is poured to the rest in the crucible which is now co- 
 vered up with a concave piece of glass. The mixture is di- 
 gested, and acid is added in drops as long as there is any effer- 
 vescence ; in this way the digestion is continued until the se- 
 parate pieces of the mass have become totally decomposed 
 into light transparent flakes of silicic acid. The mixture is 
 now poured into an evaporating dish and so long evaporated 
 either by continual stirring, or still more safely on the water 
 bath until there is no longer the slightest moisture, but on 
 the contrary the whole appears to be a dry powder. After 
 the powder is cool, it is uniformly made moist with a few 
 <lrops of hydrochloric acid and set aside for an hour ; w r ater 
 is then added, after w T hich the silicic acid is removed by fil- 
 tration ; so obtained the silicic acid is quite pure. After 
 washing, drying and igniting it is weighed; and from its 
 weight we shall be enabled to determine whether small por- 
 tions of the bases had before been present or not!, If the 
 weight is now less than it was before, this indicates that the 
 silicic acid had been mixed with these bases, which have 
 now to be removed from the acid solution in the following 
 manner, that is to say, sesquioxide of iron and alumina by 
 ammonia, lime by oxalic acid, and magnesia by phosphate 
 of soda, proceeding as more fully explained in their determi- 
 nation in the following examples. 
 
 If now, after one or more of these bases have been found 
 
150 
 
 and estimated, there should still be a small quantity to be 
 accounted for in reference to the original weight of the silicic 
 acid, this can only consist of an alkali, which can not be sought 
 for by this method. But such an event is of rare occurrence ; 
 and the amount of alkali in silicic acid especially is always 
 very small, generally amounting to less in the whole silicate 
 itself than the remaining constituents. 
 
 2. By Boiling with a solution of Carbonate of Soda. 
 
 From the fact, that the silicic acid obtained from the de- 
 composition of silicates is amorphous and also after ignition is 
 soluble in solutions of the alkaline carbonates, we have here 
 a means for testing its purity. About 30 to 50 grammes of 
 crystallized carbonate of soda are dissolved in a deep dish r 
 preferably of platinum or silver, in a small quantity of watei% 
 so as to obtain a moderately concentrated solution, which is 
 then raised to the boiling point ; a small quantity of silicic 
 acid, previously reduced to an impalpable powder in an ag- 
 ate mortar, is taken up on the point of a small spatula and 
 tjhrown into the boiling solution ; the boiling is continued 
 until the silicic acid is dissolved; a fresh quantity is now ad- 
 ded and so on repeatedly until the whole has been thrown 
 in. The fluid is now quite concentrated by boiling, after 
 which it is diluted with water, again made to boil for a mo- 
 ment and immediately poured upon a thin filter, which al- 
 lows the fluid to pass quickly through it ; any residue, that 
 map be left on the filter, is washed with boiling water. If 
 the amount of silicic acid had been pretty considerable, or 
 that of the carbonate of soda not in sufficient quantity, or if 
 the filtration should proceed slow r ly thus allowing the fluid 
 to get cool, it frequently happens that the silicic acid sepa- 
 rates in the gelatinous form; this must be avoided. The in- 
 soluble residue is ignited and weighed. The nature of this 
 has already(vide page 143) been described in detail. 
 
 3. By Hydrofluoric Acid. This is the best method by 
 which all admixtures of silicic acid admit of being easily re- 
 cognized ; to this we have already alluded (vide page 144). 
 
151 
 
 Estimation of the Water. This requires a special analy- 
 sis and in general is a work of no difficulty. The best plan is 
 to break the material up into small pieces or into a sort of 
 rough powder, since if it were in the form of fine powder, it 
 would be apt to be carried off in the operation. The igni- 
 tion is effected in closed platinum crucibles, it is continued 
 for about a quarter of an hour and may be repeated after 
 weighing, in order to be quite certain that every trace of 
 water has been removed. 
 
 We will now give by examples in the following article* 
 the analytical mode of proceeding in each of the more fre- 
 quently occurring cases, presupposing an acquaintance with 
 the general rules that have just been evolved and which may 
 be applied to all of them. 
 
 1, SILICATE of Sesquioxide and Protoxide of IEON, 
 sometimes with traces of Manganese, Lime and Magnesia, 
 (SLAGS in the refining of metals, FORGE SLAGS, STEEL 
 
 SLAGS, &G.) 
 
 The powdered substance is decomposed by hydrochloric a- 
 cid in the manner already given (vide page 14Y). The silicic 
 acid separates in the gelatinous form. The fluid portion is 
 filtered into a beaker glass, and is then heated ; in this con- 
 dition small portions of chlorate of potassa are added gradu- 
 ally until the smell of chlorine begins to be very distinct ; 
 by this mode of proceeding the protoxide of iron is convert- 
 ed into the sesquioxide. As soon as it is cold, ammonia is 
 added gradually, stirring the mixture all the while, until an 
 alkaline reaction becomes manifest, and the smell of ammo- 
 nia begins to predominate. The sesquioxide, thus thrown 
 down, is separatod by filtration, thoroughly washed, dried, 
 ignited and weighed. From this weight the amount of iron 
 is computed. 
 
 Sometimes the substance contains only in addition very 
 small quantities of lime and magnesia. To search for these 
 and to estimate their amount the filtrate from the sesquioxide 
 
152 
 
 of iron is decomposed with a small quantity of oxalic acid and 
 then set aside for twenty-four hours. If by this time a small 
 quantity of lime has been precipitated, it is removed by ni- 
 tration, washed, dried, intensely heated, until all the cor- 
 bonic acid has been expelled, which is easily effected when 
 the quantity is small, and finally weighed as pure lime. The 
 filtrate is mixed with phosphate of soda and also set a- 
 side for the same length of time as the precipitate. What- 
 ever amount of the double phosphate of magnesia and ammo- 
 nia is thus separated, it is treated as prescribed, page 133. 
 
 If there is only a trace of manganese, it will be found with 
 the iron. When larger quantities are present the separation 
 of the sesquioxide of iron, the protoxide of manganese, the 
 lime and the magnesia must be effected in the manner pre- 
 scribed, page 134. 
 
 In order to determine the relative proportion of the two 
 oxides of iron, another portion of the substance is dissolved 
 in a small flask with hydrochloric acid in a close vessel ; the 
 dissolved mass is then diluted with water and submitted to 
 the volumetric process with permanganate of potassa (vide 
 .page 71). By this means the amount of protoxide of iron 
 is obtained; and if the amount be computed as sesquioxde <>1 
 iron and deducted from the total amount obtained before. 
 we get the quantity of sesquioxide of iron in the silicate. 
 
 2, SILICATE of ALUMINA, LIME, SODA, POTAS- 
 SA and WATEE, 
 
 (ANALCIME, MESOTYE, STILBITE, HEULANDITE, CHABASITE, 
 THOMSONITE, &C.) 
 
 The amount of water is determined by ignition. 
 
 The substance in fine powder, but not ignited, is decom- 
 posed by hydrochloric acid as indicated on page 147 ; the si- 
 licic acid is removed by filtration, and further examined ac- 
 cording to instructions found on page 150. 
 
 The acid filtrate is supersaturated with ammonia* This 
 precipitates the alumina, which must be filtered oft' as quick- 
 
153 
 
 ly as possible in a closed filter, because the fluid attracts car- 
 bonic acid, and carbonate of lime is precipitated. (If such an 
 occurrence should happen owing to the exceedingly slow fil- 
 tration, the filter together with the alumina, and without any 
 previous washing, is placed in a beaker glass and digested 
 with the requisite amount of dilute hydrochloric acid; the 
 filter is then taken out, well rinsed with water, and the alumi- 
 na is precipitated as before with ammonia.) The alumina is 
 thoroughly washed with boiling water, until a drop of the 
 filtrate when evaporated on platinum foil scarcely leaves any 
 mark behind. As soon as the substance has been dried in 
 the air, it is ignited and weighed, 
 
 Since the fluid, from which the alumina has been precipi- 
 tated, attracts carbonic acid from the atmosphere,, thus giv- 
 ing rise to a separation of carbonate of lime only when it 
 con tains free ammonia, this may be avoided by boiling the 
 mixture as soon as the ammonia has been added, until all 
 smell of the latter has been expelled. If already some car- 
 bonate of lime had been formed, it would be decomposed by 
 the chloride of ammonium present, and the lime would again 
 be dissolved. The alumina is removed by filtration whilst 
 the fluid is still hot ; but the washing must not be continued 
 too long as the residue is apt to clog the pores of the filter. 
 On this account the substance is allowed to dry, until it cracks 
 up into separate parts, which however are still moist in the 
 inside, and the washing with hot water is completed. 
 
 Oxalic acid, quite free from any admixture of alkali, is ad- 
 ded to the filtrate. (The solution of oxalic acid must leave 
 no residue when heated in a platinum crucible.) By this 
 means the lime is precipitated, but completely only when 
 the fluid has an alkaline reaction, and has been set aside for 
 at least twelve hours after the addition of oxalic acid. The 
 oxalate of lime is collected on a filter, then washed with hot 
 water, dried and converted into carbonate of lime according 
 to the rules laid down on page 132. 
 
 The filtrate is evaporated to dryness, towards the last in a 
 20 
 
154: 
 
 water-bath ; the residue is carefully removed from the dish 
 and placed in a platinum crucible loosely covered ; it is then 
 heated as long as ammoniacal vapors are given off; after- 
 wards a few drops of hydrochloric acid are added so as to 
 moisten it thoroughly ; the lid is now firmly closed, and the 
 mass is again heated, and towards the last gently ignited. In- 
 tense ignition, but especially opening of the crucible during 
 this operation, is apt to produce a loss of alkaline chloride. 
 After cooling, the residue is immediately weighed ; it con- 
 sists of a mixture of chloride of potassium and of chloride of 
 sodium. Whilst still in the crucible it is covered with wa- 
 ter and heated ; this dissolves all, with the exception of a 
 few flakes of silicic acid, which is removed by filtration on a 
 small filter, washed, dried and ignited. This weight is add- 
 ed to that previously received, but deducted from that of the 
 chlorides. The fluid portion, which had been received in a 
 dish, is mixed with a sufficient quantity of a solution of chlo- 
 ride of platinum so as to convert the chlorides into double 
 chlorides, and evaporated on the water-bath almost to dry- 
 ness. The residue, whilst still moist, is allowed to cool, and 
 is then mixed with moderately strong alcohol (spec. grav. 
 0.83), containing from one-sixth to one-fifth its volume of e- 
 ther ; the precipitated double chloride of platinum and po- 
 tassium is separated by filtration on a filter, previously dried 
 at a temperature of 248F. and weighed, it is then washed 
 with the mixed alcohol and ether, and dried, towards the 
 last at a temperature from 248F. to 256F. until its weight 
 remains constant. 
 
 The amount of chloride of potassium is computed from the 
 weight of the double chloride of platinum and potassium ; 
 this is next deducted from the total weight of the two chlo- 
 rides ; the difference is the weight of the chlorides of sodi- 
 um. The respective amounts of potassa and soda are easily 
 computed from those of their chlorides. 
 
 If it be desired to make a direct determination of the a- 
 mount of soda, an operation which is less accurate, but still 
 
155 
 
 all right and proper in order to be convinced of its presence 
 and to test its purity, a concentrated solution of chloride of 
 ammonium is added to the alcoholic fluid to precipitate the 
 largest proportion of the platinum in excess. The double 
 chloride of platinum and ammonium is received on a filter 
 and washed with dilute alcohol. The filtrate is evaporated 
 to clryiiess in a dish; the residue is then placed in a porce- 
 lain crucible and heated until the ammoniacal salt is volati- 
 lized ; as soon as it is cold, it is moistened with a solution 
 of oxalic acid and evaporated to dryness. The dry mass is 
 heated to a high temperature, and afterwards the chloride 
 of sodium is dissolved leaving the reduced platinum undis- 
 solved. 
 
 In order to acertain whether the double chloride of plati- 
 num and potassium is pure, it is folded up in the filter and 
 moderately heated in a porcelain crucible in order to carbo- 
 nixe the filter. Hydrogen gas is then introduced, while the 
 salt is being heated, finally to ignition. "Water dissolves the 
 chloride of potassium leaving the platinum undissolved, 
 which is then weighed. Its amount must be forty and one- 
 half per cent of the double salt. Furthermore the solution of 
 the chloride of potassium can easily be tested as to the ab- 
 sence of soda. 
 
 On the other hand the purity of the chloride of sodium 
 and especially its freedom from chloride of potassium may be 
 ascertained by evaporationg its solution when mixed with 
 sulphuric acid in a platinum crucible to dryness ; the dry mass 
 is intensely heated, and afterwards dissolved is water. This 
 solution is placed on a watch glass and set aside to evaporate 
 spontaneously. As soon as the fluid has dried, the crystals 
 of the sulphate of soda soon begin to effloresce, and may be 
 with care blown off, whilst, if there be any potassa, it may 
 be easily distinguished by the presence of hard and transpar- 
 ent crystals of sulphate of potassa which may be furthermore 
 recognized by their shape. 
 
 OBSERVATION. In the analysis of silicates containing alkalies there are generally 
 
 found in the chloride of sodium very minute quantities of alumina, lime, OP magnesia, 
 
156 
 
 which in accurate operations have to be determined. 
 
 3. SILICATE of Protoxide of IEON, Protoxide of 
 MANGANESE, ALUMINA, LIME and MAGNESIA. 
 
 (SEVERAL SLAGS.) 
 
 The substance, reduced to a fine powder, is decomposed 
 by means of hydrochloric acid, and silicic acid separates. 
 
 If the amount of manganese is small, the process for the 
 separation of the bases is that found on page 127, that is, the 
 solution is mixed with chloride of ammonium, then ammonia 
 is added in excess, and the mixture is then boiled, d?c. The 
 small quantity of manganese then remains in the oxide of 
 iron. 
 
 If the sesquioxide of iron and alumina are to be determin- 
 ed by a direct process, the ignited and weighed precipitate 
 of the two in mixture is reduced in an agate mortar to an 
 impalpable powder, and then all that can be shaken out of 
 the mortar is again ignited and weighed. (Since this is not 
 the entire original quantity, from the fact that a portion is 
 lost, the final results are computed in reference to the origi- 
 nal quantity.) This is rubbed up in a silver crucible togeth- 
 with a small quantity of carbonate of soda, pure hydrate of 
 potassa is then added in sufficient quantity ; the covered cru- 
 cible is then heated very cautiously (on account of the inci- 
 pient foaming) over the lamp. Finally the temperature of 
 the melted mass is raised to ignition, and maintained at this 
 heat for about a quarter of an hour. As soon as the crucible 
 is cold, it is placed in water, to which (for the reduction of 
 the manganic acid) a small quantity of alcohol is added ; the 
 mixture is now heated to boiling, filtered and washed with 
 hot water. The residue on the filter, consisting of sesquiox- 
 ide of iron, is digested, together with the filter, in hydro- 
 chloric acid ; the solution is diluted and filtered ; and the ox- 
 ide of iron is precipitated with ammonia. The solution of 
 the alumina in potassa is slightly acidified with hydrochloric 
 aeid (taking care that nothing is lost by the effervescence of 
 
157 
 
 the fluid), and then sulphide of ammonium is added ; this pre- 
 cipitates the alumina, which is washed with hot water, dried, 
 ignited and weighed. The sum of the weight of this and of 
 that of the sesquioxide of iron must correspond to the amount 
 of the precipitate used in this experiment. 
 
 OBSERVATION. If the hydrate of potassa, as is so frequently the case, is not quite 
 
 free from silicic acid and alumina, the direct determination of the alumina in the solu- 
 tion of potassa is omitted, and its amount is obtained by taking the difference. 
 
 If the amount of the manganese is more considerable, per- 
 haps greater than that of the iron, as for instance in many 
 slags from the smelting furnaces, the process is as follows: 
 to the fluid filtrate from the silicic acid chloride af ammo- 
 nium and a small quantity of acetate of soda are added ; the 
 mixture is then heated and a current of chlorine is passed 
 through it to saturation ; ammonia is added and afterwards 
 the whole is boiled in order to expel all the free ammonia. 
 The precipitate, consisting of sesquioxide of iron, binoxide of 
 manganese and alumina, is separated by filtration and tho- 
 roughly washed with hot water ; it is then dissolved in hy- 
 drochloric acid, to which a small quantity of alcohol is added, 
 the solution is diluted, but not filtered, and then carbonate 
 of baryta levigated with water is added in sufficient quantity 
 so as to leave a slight excess undissolved. The mixture is 
 now set aside for about an hour, being frequently stirred up 
 during the time. The sesquioxide of iron and the alumina, 
 together with the carbonate of baryta in excess, are now re- 
 moved by filtration, washed with cold water, and dissolved 
 in dilute hydrochloric acid ; sulphuric acid throws down the 
 baryta, which is left by filtration on the filter ; sesquioxide 
 of iron and the alumina arc precipitated by ammonia; after 
 ignition and weighing, these substances are separated as be- 
 fore indicated. The fluid, containing manganese, is treated 
 in like manner with sulphuric acid which removes the dis- 
 solved baryta, afterwards with carbonate of soda and boiled 
 (vide page 67) ; and in this way the manganese is estimated. 
 
 From the filtrates of the three bases lime is thrown down 
 by oxalic acid, and magnesia by phosphate of soda according 
 
158 
 
 to methods already given. 
 
 OBSERVATION. Small quantities of sulphur iu the form of metallic sulphides, ai* 
 
 also of the alkali, are found in several slags. The first is recognized during the decom- 
 position of the silicate by the evolution of hyclrosulphuric acid. To estimate its amount,. 
 a separate quantity of the fine powder is mixed in a small flask with chlorate of potassa 
 and a small quantity of water ; hydrochloric acid is added to the mixture, which is di- 
 gested until the sulphur is converted into sulphuric acid ; the mixture is then diluted 
 and submitted to filtration, which removes the silicic acid : the sulphuric acid is then 
 determined by means of chloride of barium. The amount of sulphur is computed from, 
 the sulphate of baryta. 
 
 In order to determine the amount of potassa or soda the process, previously given, is 
 followed, excepting that acetate of soda is left out ; and, after the separation of the 
 Ume, the magnesia is separated from the alkali by means of hydrate of baryta, or carbo- 
 nate of ammonia as will be shown further on (vide : Silicates that are not decornposiblo 
 by acids and which contain magnesia and alumina). 
 
 4, SILICATE of the Protoxide of IKON, ALUMINA, 
 
 LIME and MAGNESIA, with small quantities of oxide of 
 
 LEAD or of COPPER, 
 
 (LEAD OB COPPER SLAGS.) 
 
 The substance is decomposed by hydrochloric acid accord- 
 ing to the general rules already given (vide page 147). It" 
 lead is one of the constituents, the silicic acid must be care- 
 fully washed with boiling water, in order that no chloride of 
 lead may remain mixed up with it. 
 
 A current of hydrosulphuric acid is passed through the ni- 
 trate in order to precipitate the lead or copper : as soon as 
 the precipitate is complete, the residue is filtered off as quick- 
 ly as possible in. a covered funnel to prevent all access of air; 
 the filtrate is received in a dish. 
 
 If the precipitate consists of lead, it is treated as directed 
 on page 61. If on the contrary it consists of sulphide of cop- 
 per and its amount is very small, it is raised to intense ig- 
 nition in a current of air ; by this means it is converted into 
 the oxide of copper which is weighed. When the amount 
 is large, it is' treated, on the contrary, according to the meth- 
 od given on page 52. 
 
 The filtrate from the metallic sulphide is concentrated to 
 H small volume by evaporation, it is then oxidized by chlo- 
 rate of potassa aiid diluted ; from the dilute solution the ba- 
 ses are separated as in the preceding example. 
 
159 
 
 5, SILICATE of the oxide of ZINC with Water, 
 (CALAMINE OR SILICIOUS OXIDE OF ZINC.) 
 
 Although pure calamine is nothing more than a hydrated 
 silicate of the oxide of zinc, yet it occurs generally accompa- 
 nied with zinc spar (carbonate of the oxide of zinc), brown 
 hematite (hydrated oxide of iron) and limestone, &e. <lal- 
 rney is frequently thus constituted, from which zinc is ex- 
 tracted. The substance therefore contains oxide of zinc, ox- 
 ide of iron, lime, magnesia,* silicic acid and water. 
 
 A quantity of the substance (sufficient for all the experi- 
 ments) is reduced to an impalpable powder. 
 
 a) In one part the amount of carbonic acid is determined- 
 according to the method given on page 129. 
 
 b) Another portion is decomposed by hydrochloric acid. 
 The silicic acid is removed by filtration and tested as to itw 
 purity ; the filtrate is diluted, moderately warmed, and then 
 carbonate of soda is added until the fluid begins to assume a 
 dark-red color, or in fact until a portion of the basic salt of 
 the oxide of iron is precipiated, whereupon acetate of soda is 
 added and the mixture is made to boil. By this means the 
 sesquioxide of iron is separated from the oxide of zinc and 
 earths here present as also from the protoxide of manganese 
 and the oxides of copper and cobalt. (Compare pages 67 
 and 133.) 
 
 A current of hydros ulphuric acid is passed through the fil- 
 trate from the sesquioxide of iron (the filtrate must contain 
 a sufficient quantity of acetate af soda). This precipitates 
 the zinc. The sulphide of zinc is allowed to settle complete- 
 ly, it is. then filtered in a covered funnel as quickly as possi- 
 ble, washed with water containing in solution hydrosulphu- 
 ic acid, and digested with hydrochloric acid. The solution 
 is diluted and raised to a boiling temperature ; carbonate of 
 soda is then added, and the mixture is filtered ; the residue 
 is washed, dried and ignited. 
 
 The filtrate from the sulphide of jzinc is freed from hydro- 
 
 "Contained in dolomitic limestone. 
 
160 
 
 sulphuric acid by heating ; ammonia is then added in excess.. 
 and afterwards oxalic acid (vide page 131). After the remo- 
 val of the oxalate of lime by filtration, the magnesia is pre- 
 cipitated by phosphate of soda (vide page 133). 
 
 c) The amount of water in the substance is computed ei 
 ther from the loss, or by igniting a separate portion intense- 
 ly, which drives off at the same time both the water and the 
 carbonic acid, and then by deducting the amount of the lat- 
 ter (found in a) from the loss thus produced by heat. 
 
 6, SILICATE of ALUMINA and SODA, containing 
 
 SULPHIDE of SODIUM, 
 (ULTRAMAJRINE, LAPIS LAZULI, HAUYNE, &c.) 
 
 Ultramarine contains, besides a silicate of alumina and^so- 
 da, one or more sulphides, whose nature is not yet accurate- 
 ly known, nor can they be determined by the analysis, for 
 this is limited to finding the amount of sulphur. In gene- 
 ral also are found small quantities of lime, potassa, chlorine- 
 and sulphuric acid. 
 
 A. The substance, previously reduced to a fine powder r 
 is decomposed by hydrochloric acid. By this treatment it 
 loses its color, evolves hydrosulphuric acid and forms a com- 
 plete gelatinous substance. After evaporation, &c., the sili- 
 cic acid is separated, dried, ignited and weighed. It was ac- 
 companied with sulphur, which volatilized when heated. 
 It is tested in reference to its purity (vide page 148). 
 
 Chloride of barium is added to the filtrate as long as sul- 
 phate of baryta is precipitated, which is allowed to settle r 
 then filtered, thoroughly washed and ignited. From its 
 weight the amount of sulphuric acid may be computed. 
 
 Dilute sulphuric acid is cautiously added to the filtrate in 
 order to throw down all excess of chloride of barium ; after 
 which the alumina is precipitated by ammonia. The preci- 
 pitate is washed with hot water, dried, ignited and weighed. 
 (The alumina frequently contains a small quantity of oxide 
 of iron.) 
 
161 
 
 Oxalic acid (free from alkali) is added in small quantity to 
 the filtrate from the preceding precipitate, in order to preci- 
 pitate the lime, proceeding in this respect as directed on 
 page 132. 
 
 After this the fluid is evaporated to dryness ; the residue 
 is placed in a platinum crucible, heated but not to ignition, 
 until all vapors from ammonical salts have passed off, it is then 
 intensely ignited, thus converting the sulphate of soda (and 
 sulphate of potassa) by the method given in detail on page 
 113 into neutral sulphate, which is weighed. 
 
 If it be desirable to determine the small amount of potas- 
 sa which is frequently contained in the residue, the sulphates 
 must first be converted into chlorides. This is effected by 
 mixing them in a porcelain crucible with chloride of ammo- 
 nium, and heating the mixture as long as any vapors of am- 
 moniacal salts are given off, taking care that the temperature 
 is not so high as to fuse the mass. This treatment with chlo- 
 ride of ammonium must be repeated until the weight of the 
 residue remains constant. The chlorides thus produced are 
 separated by means of the chloride of platinum according to 
 the rules given on page 154. 
 
 I>. -In order to determine the whole amount of sulphur, 
 the process to be followed will be found on page 158, taking 
 care that no part of the sulphur remains unoxidized. Instead 
 of filtering immediately after the solution is diluted, the whole 
 may be evaporated to dryness, then treated with water and 
 filtered ; to the filtrate add hydrochloric acid and afterwards 
 chloride of barium. If the sulphur be computed from the 
 amount of sulphate of baryta, that portion must be deducted 
 which is contained in the sulphuric acid found under arti- 
 cle A. 
 
 As regards the analysis it is a matter of indifference, but 
 as regards the constitution of the ultramarine and similar bo- 
 dies it is interesting to determine how much sulphur is set 
 free "by decomposition by means of hydrochloric acid and 
 how much in the form of hydrosulphuric acid. For this pur- 
 21 
 
162 
 
 pose therefore a fresh quantity is dissolved by acid, the gel- 
 atinous silicic acid, together with the sulphur, is removed by 
 filtration, thoroughly washed, dried and then treated with 
 chlorate of potassa and hydrochloric acid. As soon as the 
 sulphur has been completely oxidized, the solution is diluted 
 and filtered, after which the sulphuric acid thus produced is 
 precipitated by chloride of barium. By deducting the a- 
 mount of sulphur, computed from the sulphate of baryta, 
 from the total weight of the substance, w T e obtain that part 
 which is evolved in the form of hydrosulphuric acid. 
 
 C. In order to estimate the amount of chlorine, a sepa- 
 rate quantity of the substance is treated with nitric acid at 
 a very gentle temperature ; the silicic acid is removed by fil- 
 tration ; to the fitrate nitrate of silver is added to precipitate 
 the chlorine. If the chloride of silver should contain any 
 sulphide of silver, it is separated after drying and weighing 
 by digestion with ammonia. 
 
 Here also the bases, which generally occur, are the same 
 which are met with in the silicates that can be decomposed 
 by acids (vide page 145), namely : the oxides of iron and man- 
 ganese, magnesia, lime, soda and potassa. Of those which 
 occur less frequently we find the following : 
 
 LITHIA (Petalite, Spodumene, Mica, Tourmaline.) 
 
 BARYTA (Felspar.) 
 
 GLTJCINA (Beryl and Emerald, Phenakite, Euclase, Chryso- 
 beryl, Leucophane.) 
 
 OXIDE OF ZING "] Slags and Glasses ; 
 LEAD j^ small quantities 
 COPPER J in some minerals. 
 
 OXIDE OF NICKEL (Pimelite, Saccharite, Chrysoprase, Ol- 
 ivine.) 
 
 OXIDE OF COBALT (Smalt.) 
 
 OXIDE OF CHROMIUM (Chrome ochre, Garnet, Pyrope, Ser- 
 pentine, Miloschine, Chrome Mioa, &c.) 
 
 Of electronegative bodies are the following : 
 
 Stannic acid (Glasses and Enamels.) 
 
163 
 
 Titannic acid (Acmite, Several sorts of Mica, Hornblende, 
 Arc.) 
 
 ZIRCONIC ACID (Zircon.) 
 
 BOBACIC ACID (Axinite, Tourmaline, Several Glasses and 
 Enamels.) 
 
 PHOSPHORIC ACID (Sordawalite, Mica.) 
 
 FLUORINE (Topaz, 1 Leucophane, Mica, Hornblende, Sca- 
 polite, Slags, &c.) 
 
 Water is sometimes a constituent of this class of silicates. 
 
 Judging from the number, this division is the larger of the 
 two ; and we find in it not only the most important and the 
 most widely scattered minerals, such as felspar, mica, horn- 
 blende, augite, garnet, tourmaline, &c., but also the differ- 
 ent sorts of glass. 
 
 In order to decompose these compounds, the following 
 methods in particular present themselves, namely : 
 
 A.-* Ignition with Carbonate of Soda. 
 
 B. Ignition with Carbonate of Lime and Chloride of Am- 
 monium. 
 
 (j. Solution in Hydrochloric acid. 
 
 D. Decomposition by means of Sulphuric Acid. 
 
 Other methods, for instance the application of carbonate 
 of baryta, of caustic potassa,. or of fluor spar and sulphuric 
 acid, are scarcely any longer made in reference to the analysis 
 of silicates. The first mentioned is used, however, in the 
 following way: Decomposition by means of carbonate of soda, 
 a process which is most frequently employed, serves to deter- 
 mine all the constituents of a silicate with the exception 
 of the alkalies ; if these are absent, this method naturally 
 completes the analysis. The use of carbonate of lime and chlo- 
 ride of ammonium is sufficient for the determination of all 
 the constituents of a silicate, even of the alkalies ; whilst by 
 means of hydrochloric acid all the constituents, with the ex- 
 ception of silicic acid, may be determined. 
 
 A complete description of these processes, together with 
 examples for each, will be found in the following articles. 
 
164 
 
 A. ANALYSIS of SILICATES with CARBONATE 
 of SODA. 
 
 If a silicate, which is decomposed by an acid with difficul- 
 ty or not at all, is ignited with carbonate of soda, carbonic acid 
 is evolved, with the formation of silicate of soda at the ex- 
 pense of a part of the acid taken from the bases, thus giving 
 rise to a mixture or a combination of basic salts, that can ea- 
 sily be decomposed by a strong acid. The efficacy of carbo- 
 nate of soda consists therefore essentially in the conversion of 
 silicates of this class into those of the former, consequently 
 the further steps to be taken in the analysis are in fact pre- 
 cisely similar to those already described. 
 
 GENERAL RULES for the ANALYSIS. 
 
 Pulverization. We have already remarked, that this ope- 
 ration must be carefully performed, and that those silicates, 
 that are very rare and that can with difficulty be decomposed, 
 must be subjected also to elutriation. (Compare page 140.) 
 
 Drying and Weighing of the Substance. The impalpable 
 powder must first of all, in all accurate investigations, be 
 dried. If the substance contains water, it is ignited at a low 
 'temperature in a covered platinum crucible and weighed. 
 If, on the contrary, we have to deal with a hydrated si- 
 licate, the drying of the powder must be effected in the dry- 
 ing chamber of a common stove, or in the drying apparatus at 
 a temperature of between 86F. and 104F., or in the desic- 
 cator over sulphuric acid. 
 
 OBSERVATION. With some porous silicates it is very difficult to determine how 
 
 much hygrometric water and how much water in chemical combination they contain, 
 since a portion of the latter appears to be given off by a gentle heat also over sulphuric 
 acid, in the same manner as several chrystallized salts effloresce under similar circum- 
 stances owing to the loss of their water of crystallization. 
 
 The powder, having been dried and weighed in a small 
 platinum crucible, is shaken into a more capacious crucible; 
 the small crucible, together with any adhering powder, is 
 again weighed in order to ascertain with accuracy the quan- 
 tity of powder about to be used in the analysis. 
 
165 
 
 Admixture with Carbonate of soda. Dry carbonate of 
 soda, to the amount of the substance, is taken and rubbed to 
 a fine powder in an agate mortar ; the platinum crucible con- 
 taining the substance to be analyzed, is then placed on a 
 sheet of smooth paper, and about half the carbonate of soda 
 is poured into the crucible and, by means of a glass rod with 
 a smooth round end, the two substances are intimately mix- 
 ed together until they form one homogeneous substance. 
 More carbonate of soda is now added and again mixed as be- 
 fore, and so on until nearly all the soda has been poured in. 
 Finally the glass rod is brushed thoroughly clean with a 
 feather, and the remaining part of the soda is poured into the 
 crucible and is allowed to form a layer to cover the rest. 
 
 Ignition. This operation is performed either in the blast 
 furnace or over the lamp according as the quantity of the 
 mixture is large or small. When a silicate is ignited with 
 carbonate of soda, the mass is either completely fused or has 
 only undergone a sort of semi-fusion. In order to obtain a 
 perfect decomposition, fusion is by far to be preferred ; but 
 all silicates rich in alumina and magnesia, as a general thing, 
 can be reduced only to a semi-fused mass. On this account 
 it is necessary in these cases, that the powder used for the 
 analysis should be exceedingly fine, even sometimes elutriat- 
 ed ; whilst in those cases where the fusion is complete, this is 
 not necessary to the same extent, since in the melted fluid 
 mass all the parts come in contact with each other. 
 
 If a blast furnace be used, the platinum crucible well cover- 
 ed up is placed in a Hessian or French crucible, which is al- 
 so covered, and at the bottom of which had been previously 
 placed a layer of either calcined magnesia or its carbonate, in 
 this the platinum crucible is embedded and prevented from 
 coming in contact with the sides of the greater crucible. 
 
 The ignition in all cases must be maintained for at least 
 half an hour, reckoning from the time when the mass had 
 attained the temperature of ignition. 
 
 Decomposition of the Ignited Mass Ijy Hydrochloric Acid. 
 
165 
 
 As soon as the crucible is cold, the melted mass is loosened 
 from the sides of the crucible by gentle pressure and then 
 transferred to a beaker glass. This transference is sometimes, 
 quite complete, especially if the mixture has undergone on- 
 ly a sort of semi-fusion. If some parts, however, adhere firm- 
 ly to the crucible, water is poured upon them and heat ap- 
 plied; and if this does not effect their complete separation 
 a few drops of hydrochloric acid are added, and the crucible 
 is covered up and set aside for some time. If, however, the 
 fused contents of the crucible adhere so firmly as scarcely to 
 be affected by pressure, then the crucible itself together with 
 the cover (if any of the fused substance adheres to it) is pla- 
 ced in the beaker glass and digested with the dilute acid, re- 
 collecting, however, that, if the mass has a green color and 
 consequently contains manganic acid, chlorine will be set free- 
 and will act upon the crucible itself (vide page 37). In such 
 a case as this alcohol is first put in the crucible before the: 
 acid comes in contact with the platinum. 
 
 The mass in the beaker glass is well covered with water,, 
 the beaker glass being furnished with a concave plate of glass 
 as a cover ; hydrochloric acid is then added gradually and 
 cautiously, in order that nothing may be lost by effervescence. 
 
 The fluids, that were used for cleaning the crucible, are 
 likewise added either now or later. The mixture is now di- 
 gested either on the sandbath, in the digestorium or in fact 
 in any moderately warm place, taking care to break up th& 
 larger pieces by means of a glass rod, because they easily be- 
 come covered with the silicic acid as it separates and are thus- 
 protected from the action of the decomosing agent. In case- 
 of need also more acid is added. The silicic acid must sepa- 
 rate in the form of light flakes (sometimes the whole becomes 
 a gelatinous mass), and when the digestion is finished and 
 the glass rod is rubbed over the bottom of the glass no hard 
 sandy particles must be observed producing a gritty sound. 41 " 
 
 *It is necessary to guard against being deceived, for, when the fluid is highly con- 
 centrated, crystals of chloride of sodinm are easily formed ; but these dissolve immedi- 
 ately on the addition of hot. water. 
 
167 
 
 Although now the largest portion of the silicic acid ha? 
 separated, no inconsiderable portion remains dissolved in 
 the fluid, which has to be converted into the insoluble form 
 by evaporation. To effect this the whole is placed in a mo- 
 derately deep dish (a platinum dish is the best), and is con- 
 centrated by evaporation and constant stirring to dryness. 
 Since, however, the thick fluid is very apt to bubble up and 
 spirt, it is by far the best to evaporate it on the water-bath, 
 continually stirring the mixture in order to break up the in- 
 crustation of chloride of sodium, which forms on its surface, 
 for this prevents the vapors from passing oif. The evapo- 
 ration is continued, until the mass has assumed a dry dusty 
 appearance, when the vessel may be placed over an open fire, 
 or over the flame of a lamp for a few minutes, constantly 
 stirring the mixture, taking care, however, with silicates 
 rich in iron not to raise the temperature so high as to sepa- 
 rate some of the oxide of iron, which manifests itself by com- 
 municating a red color to the mixture ; for if this should 
 take place, the oxide is not again dissolved by the acid, and 
 consequently remains with the silicic acid. 
 
 Towards the end of the evaporation the weaker bases, such 
 as sesquioxide of iron and alumina, have lost a portion of 
 their hydrochloric acid, and, if the mass were treated imme- 
 diately with water in order to remove all that is soluble, a 
 part of these bases would remain insoluble with the silicic 
 acid. Therefore in order to reinstate that portion of acid 
 which has been volatilized, the mass, as soon as it has cooled, 
 is moistened uniformly with a few drops of hydrochloric 
 acid and set aside for at least half an hour. Water is then 
 added and the mixture is placed on a filter ; after which the 
 silicic acid is thoroughly washed with boiling water, and 
 afterwards dried spontaneously ; it is then ignited and weigh- 
 ed. The precautionary measures recommended in regard to 
 this ignition on page 148, are to be observed here. 
 
 In the acid filtrate are found all the bases contained in the 
 silicate, and these have to be separated by means described 
 
168 
 
 in the following example. This fluid contains also still a 
 small quantity of silicic acid, which is precipitated at the 
 same time with the bases, but the difficulty of its separation 
 from these is in some measure very great, and its amount so 
 minute as not to make any essential change in the result as 
 a general thing. At the very most it might be advisable to 
 eliminate that portion which remains in solution in the fluid 
 after the bases have been precipitated. The filtrate then is 
 evaporated to dryness ; and water is added to dissolve what 
 is soluble ; after which the small quantity of silicic acid is 
 left behind. 
 
 The silicic acid is pure, if the decomposition is complete. 
 It is customary to test it, as to its purity, by projecting it in 
 small quantities at a time into a boiling concentrated solution 
 of carbonate of soda (vide page 149). 
 
 In general it is the custom to regard the undissolved por- 
 tion as undecomposed material and to deduct it both from the 
 amount of the original quantity started with and from that 
 of the silicic acid ; there might be just as little propriety in do- 
 ing this here as with those silicates which are soluble (vide 
 page 142), for the residue consists, as experience has taught 
 us, of silicic acid and a variable amount of certain bases, as 
 for instance, alumina and magnesia ; and such silicic acid not 
 altogether quite pure is what is received by the observance 
 of all the precautionary measures of analysis, that is, even 
 when the powder was quite impalpable, and the fusion and 
 decomposition of the fused mass had been quite complete. 
 This occurs especially : firstly, with silicates which contain, 
 more than fifty per cent of silicic acid and are rich in magne- 
 sia, for instance with tremolite ; the silicic acid in this instance 
 always contains magnesia ; secondly, with the silicates rich 
 in, alumina, and poor in silicic acid (cyanite, andalusite, stau- 
 rolite) ; in this case the silicic acid alw r ays contains alumina. 
 
 If the amount of the residue from the solution in carbonate 
 of soda is very small, it is best to regard it as silicic acid; if 
 it is considerable, or if the greatest accuracy is required in the 
 
169 
 
 analysis, the residue is either ignited again with carbonate of so- 
 da, &c., after which the silicic acid and the hasps are estima- 
 ted, or, after it has been dried and the filter has been inci- 
 nerated, it is covered with hydrofluoric acid, &c. (vide page 
 144), or it is heated with fluoride of ammonium and sulphu- 
 ric acid, by which means the bases are retained as sulphides. 
 
 In all accurate analyses, the aim of whose results is pure- 
 ly scientific, it is best to heat right away the ignited and weigh- 
 ed silicic acid with hydrofluoric acid (or with fluoride of am- 
 monium), finally to treat it with sulphuric acid and to conti- 
 nue the analysis of the remaining sulphates of alumina, ses- 
 quioxide of iron, magnesia and lime. 
 
 The great difficulty in the analysis of silicates lies in the se- 
 paration of the silicic acid from the bases. If the mass, which 
 has been decomposed by means of acid be simply evaporated 
 to-dryness on the water-bath, the silicic acid is obtained in a 
 purer condition, but it is also in a condition to be more easi- 
 ly dissolved afterwards together with the bases in which it 
 lias to be sought as also in the last filtrate. But if the mass, 
 after it has been evaporated to dryness on a water-bath, be 
 heated over an open flame, the reverse takes place. 
 
 Estimation of the Water in Silicates. This operation will 
 be found on page 151. The ignited product, after reduction 
 to an impalpable powder, may afterwards be employed for 
 the subsequent analysis, in case there is a lack of material. 
 If the silicate contains protoxide of iron, it will be oxidized 
 by ignition, and consequently an anhydrous silicate of this 
 nature when ignited will increase in weight. With hydra- 
 ted silicates, therefore, the amount of water is obtained by ad- 
 ding to the loss incurred by ignition the amount of oxygen 
 which was necessary to convert the protoxide into sesquioxide 
 of iron. In such cases the color of the silicate is generally 
 changed into a yellow or reddish. 
 
 The operations just described are those to be followed in the 
 analysis of every single silicate, so that in the examples which 
 
 22 
 
170 
 
 follow we shall take into consideration alone that which IB 
 necessary for the further determination of the bases. 
 
 1. SILICATE of LIME and MAGNESIA. 
 
 As soon as the silicic acid has been separated by the meth- 
 ods already given, the acid fluid is neutralized with ammo- 
 nia slightly in excess ; after which the lime is precipitated 
 with oxalic acid (vide page 131), and the magnesia with phos- 
 phate of soda (vide page 133). ^ 
 
 2, SILICATE of ALUMINA, PEOTOXIDE of IEON, 
 PEOTOXIDE of MANGANESE, LIME and MAGNESIA, 
 
 The filtrate from the silicate is analyzed according to in- 
 structions on page 155. 
 
 3, SILICATE of iLUMINA and Sesquioxide of IEON 
 
 together with Water, 
 (CLAY.) 
 
 The purest specimens of clay contain besides these essen- 
 tial constituents small quantities of lime, magnesia and alka- 
 lies. Moreover they .are frequently mixed with quartz or the 
 remains of those minerals to the decomposition of which they 
 owe their formation (felspar). The analysis of clay contain- 
 ing these admixtures, with the exception of the determina- 
 tion of the alkalies, is effected as in the preceding example, 
 which consult. 
 
 4, SILICATES containing FLUOEINE, 
 
 The analysis of the native silico-fluorides, given on page 
 263, is no easy matter ; and the accurate estimation of the 
 fluorine is scarcely possible. The different steps of the ana- 
 lysis are not always the same, depending as they do on the 
 different amounts of fluorine in the substance. 
 
 As it appears the fluorine may be driven completely from 
 the silicates by a sufficiently high temperature in the form 
 
171 
 
 of fluoride of silicon ; and, when the silicates contain no wa- 
 ter, this is the simplest method of determining the amount 
 of the fluorine ; nevertheless this method can be applied 
 with ease only to those silicates which contain but a small 
 amount of fluorine. The ignition is best effected in double 
 platinum crucibles, placed in Hessian, &c., crucibles, and by 
 means of a very high temperature in a blast-furnace; by 
 this treatment all the fluorine in micas containing but small 
 quantities, as also in tourmalines, &c., is entirely expelled. 
 Even from the silicates the richest in fluorine, topaz and pyk- 
 nite, Forchhammer, Deville and Fouque have received, by 
 intense ignition, residues entirely free from fluorine ; but 
 such results require the most intense white heat, to produce 
 which the necessary means are not always on hand.* 
 
 If such a silicate (mica) contains water, which at the same 
 time requires a high temperature for its complete expulsion, 
 the determination of the fluorine can not be effected by ig- 
 nition. 
 
 In silicates containing only minute quantities of fluorine, 
 this substance has been frequently overlooked, and not only in 
 those which may be decomposed by acids (apophyllite) but al- 
 so in those which can not be decomposed. In the ordinary 
 course of their analysis, there is always a corresponding loss 
 of silicic acid which passes off as fluoride of silicon. 
 
 A. Analysis of silicates containing only minute quanti- 
 ties of Fluorine, that is, of such as contain at most but a 
 small percentage of fluorine. The tine powder is ignited 
 with four times- its weight of carbonate of soda, The semi- 
 fused or completely fused mass is digested or boiled in a pla- 
 tinum or silver dish with water ; the residue is washed with 
 boiling water for some time (a thorough extraction by wash- 
 ing is impossible). The tiltrate, containing in solution fluo- 
 rine in the form of a fluoride of sodium, and likewise a cer- 
 tain amount of silicic acid and alumina is decomposed with 
 
 *Pyknite in my experiments in the blast-furnace, heated with coke, lost only fifteen 
 per cent, that is, 10.75 of fluorine ; whereas, according to Forchhammer, it must have 
 loet 25.8 per cent, that is, 18.48 of fluorine. 
 
172 
 
 solid carbonate of ammonia and warmed. By this means the 
 alumina and part of the silicic acid are precipitated, and then 
 removed by nitration ; the nitrate is concentrated by evapo- 
 ration and then neutralized as accurately as possible with ni- 
 tric acid; a few drops of dilute acetic acid are added, then 
 the mixture as evaporated to dry ness on a water-bath. The 
 dry mass is now treated with cold-water, which leaves silicic 
 acid undissolved ; . this is removed*by, filtration, after which 
 the iiltrate is again evaporated to dryness and dissolved. At 
 this stage the fluorine is precipitated by means of a solution 
 of chloride of calcium and by raising the mixture to boiling ; 
 the latter is then allowed to settle, and the supernatant fluid 
 filtered. The residue is again boiled with water, and allow- 
 ed to settle; the fluid is decanted and filtered; and the ope- 
 ration is repeated until all that is soluble in the residue in 
 this way has been pretty thoroughly removed by washing. 
 The latter finally is collected on the filter, dried, ignited 
 and weighed. # 
 
 Observation. In order to determine the presence of very minute quantities of fluo- 
 rine, the ignited precipitate is gently heated with a little concentrated sulphuric acid in 
 a crucible covered with a plate of glass prepared for the purpose. As soon as the pre- 
 sence of fluorine is shown by the corrosion of the glass plate, the crucible is closed with 
 its lid and heated, at last to low ignition; the sulphate of lime, thus formed, is then 
 weighed andi from its weight the purity of the fluoride of calcium is determined. (The 
 equivalents' of -'fluoride of calcium and of* sulphate of lime have the ratio of 39 to 68 or of 
 1 to 1.7436.) 
 
 The first residue, as well as also those obtained from the - 
 filtrate containing fluorine, are dried, then separated from 
 the filters and digested together. with--the-'ashes v of:the^ filters 
 with dilute hydrocnToric acid, by which the mass is gelati- 
 nized. The gelatine is evaporated to complete dryness on 
 the water-bath, after which the silicic acid is separated ac- 
 cording to the method; found on page 167. The bases of the 
 silicate are to be sought for in the filtrate. 
 
 B. Analysis of Silicates Rich in Fluorine. If the at- 
 tempt be made to determine the amount of fluorine, in the 
 m: inner described, in silicates containing ten per cent of fluo- 
 rine or more and at the same time not much silicic acid, the 
 
173 
 
 whole amount of fluorine can not be obtained because of the 
 want of a sufficient quantity of silicic acid. Forclihammer 
 has shown that this defect can be remedied by the addition 
 'of more silicic acid. Therefore one part of the fine powder 
 of the silicate is mixed with from one-half to one part of re- 
 cently ignited pure silicic acid and from 4 to 6 parts of carbo- 
 nate of soda. The mass is then heated, at the beginning it 
 melts and then again concretes, after which it is raised to 
 -an intense ignition, proceeding in other respects precisely a 8 
 previously described. If the amount of silicic acid had been 
 .accurately weighed, all that is needed is to deduct it from the 
 whole amount in order to ascertain that portion contained in 
 the silicate. 
 
 5, SILICATES containing BORACIC ACID, 
 
 (ToiJRM ALINE, AXINITE.) 
 
 The great difficulty of the direct determination of boracie 
 sicid has hitherto been the reason of estimating its amount in- 
 directly from the loss, after the most accurate separation of 
 the silicic acid and the bases, as computed from their quan- 
 tity. The greater the number, however, of the bases, the 
 more complex will be the course of the analysis, and the less 
 inevitable will be the losses ; the computation, therefore, of 
 the boracie acid will be inaccurate, especially, when besides 
 the earths and the oxides of iron and manganese, there are pre- 
 sent also the alkalies, as in tourmaline, whose determination 
 requires a special analysis. The direct determination of bo- 
 racie acid, however, may be effected by aid of the method 
 .given on page 136, if all the requisite precautionary measures 
 are observed. 
 
 The substance, reduced to a fine powder, is ignited with 
 four times its weight of pure carbonate ofpotassa; the mass 
 is then ignited and boiled with water (to which alcohol is ad- 
 dled, if the fluid becomes green by the formation of manganic 
 acid); the mixture is then filtered, and the residue is washed 
 for some time with hot water. The filtrate contains the bo- 
 
174 
 
 racic acid as borate of potassa, and in addition a small quan- 
 tity of silicic acid and alumina. In order to separate both 
 of these, chloride of ammonium is added, and the mixture is- 
 heated for some time ; the precipitate thus formed is separa- 
 ted by filtration ; and the operation is repeated with the fil- 
 trate until there is no further separation of any thing (the 
 riuid may also be evaporated to dryness on the water-bath), 
 the fluid is then heated with a small quantity of carbonate of 
 potassa in order to destroy the undecomposed chloride af am- 
 monium, after which hydrate of potassa is added in excess. 
 The boracic acid is now determined, according to instructions 
 given on page 136 as borofluoride of potassium. 
 The residue containing the silicic acid. and the bases, as well 
 as the precipitate of silicic acid and alumina obtained after- 
 wards, are dried and rubbed off from the niters. These are 
 incinerated, and the ashes are added to the preceding resi- 
 dues, which are decomposed with dilute hydrochloric acid ; 
 in this way the silicic acid is separated from the bases, after 
 which these are separated from one another by methods al- 
 ready known. 
 
 B. Analysis of Silicates by means of Carbonate of Lime 
 and Chloride of Ammonium, 
 
 This method, as well as the following for the solution of 
 the silicates by means of hydrofluoric acid, is applied in the 
 analysis of silicates containing the alkalies, for which pur- 
 pose the preceding naturally could not be used for their de- 
 termination at the same time with the rest. 
 
 The silicate is previously reduced to an impalpable pow- 
 der, or even elutriated, if the decomposition is to be success- 
 ful. This powder is mixed in a platinum crucible with an e- 
 qual weight of ammonium and five times its weight of carbo- 
 nate of lime (prepared by precipitation from pure chloride of 
 calcium with carbonate of ammonia) ; the crucible is then clo- 
 sed and heated slowly, just raising the temperature of the 
 bottom of the crucible to a low red heat and maintaining it 
 so for some time. After the expiration of half an hour the- 
 
175 
 
 temperature is raised to intense ignition, which reduces the 
 contents to a semi-fused condition, but does not fuse them. 
 The process with the contents, which can easily be removed 
 from the crucible, is the same exactly as that pursued in the 
 fusion of a silicate by carbonate of soda (vide page 165); the 
 silicate is first estimated, and then the bases in the filtrate, 
 resorting partly to the methods already given (page 152), and 
 partly to such as are described hereafter in the decomposi- 
 tion by means of hydrofluoric acid. 
 
 Although this method has the advantage of determining 
 at the same time both the silicic acid and the alkalies, it is 
 subject for all that to some disadvantages. The great amount 
 of lime, w^hich thus enters the analysis and which has to be 
 separated before we can search for the bases, entails a very 
 -considerable amount of precipitate of oxalate of lime. If the 
 silicate itself also contains lime, the amount of carbonate <>f 
 lime added, which must be well dried, must be accurately 
 weighed; and the conversion of the oxalate of lime, obtained 
 afterwards, into carbonate of lime, as also the operation of 
 weighing it, in order to obtain the amount of lime in the si 
 'licatc by deducting this weight, is on account of its large :i 
 mount a difficult piece of work. 
 
 Furthermore the silicic acid, after it has been weighed, has 
 to be tested as to its purity ; and if, after it has been boiled 
 with a solution of carbonate of soda (vide page 150), it should 
 leave a residue, this has to be deducted without any more to 
 do as undecomposed substance, which possibly may give a 
 false result. 
 
 C. Analysis of Silicates with Hydrofluoric Add. 
 
 There is no method for the accurate determination of the 
 alkalies and the remaining bases in silicates equal to this one 
 first introduced into analytical chemistry by Berzelius. It 
 'is founded upon the fact, that when a silicate is heated with 
 hydrofluoric acid, the silicic acid is partly volatilized as fluo- 
 rite of silicon, and the bases are converted into silico-fluorides. 
 Adding sulphuric acid at the same time, the silico-fluorideR 
 
176 
 
 lose all the fluoride of silicon and pass into sulphates, whose- 
 bases then admit of being separated. 
 
 Since, however, when a silicate is decomposed by hydro- 
 fluoric acid the amount of silicic acid is not found by a direct 
 method, this determination is used only in those cases where 
 a deficiency of material precludes the special determination 
 of silicic acid. In all other cases a separate quantity of sili- 
 cate is fused with carbonate of soda, in order to compare the 
 quantity of silicic acid got by this process with that from the 
 solution with hydrofluoric acid, which to be correct must be 
 equal (not comprehending in the loss the water or other vo- 
 latile or undetermined constituents such as fluorine or bora- 
 eic acid). Besides it is always highly to be recommended 
 not to confine the analysis with carbonate of soda to the de- 
 termination of silicic acid, but to extend it to the determina- 
 tion of all bases, excepting of course the alkalies, so that the 
 two analyses in this respect may be a check on one another. 
 
 The different silicates are not decomposed by hydrofluoric 
 acid with equal facility; those, which in general are decom- 
 posed by acids, are most easily decomposed by hydrofluoric 
 acid. But those silicates not decomposible by the com- 
 mon acids (hydrochloric acid) show considerable difference 
 in their comportment with hydrofluoric acid. The amorph- 
 ous silicates, for instance all sorts of glass, in the form of a 
 rough powder are easily decomposed ; several of the crystal- 
 line silicates, for instance the felspars, must be reduced lo- 
 an impalpable powder and allowed to stand a long time in 
 contact with the acid, or to be submitted to heat ; finally se- 
 veral, such as zircon and tourmaline, can be decomposed un- 
 der the same conditions only with difficulty. The more ea- 
 sily a silicate can be decomposed and the more concentrated 
 the acid, the more violent will be the action, that is, the 
 elevation of the temperature of the mixture, which may even> 
 cause the fluid to boil. 
 
 It is not advisable to use a very concentrated, that is, a, 
 highly fuming hydrofluoric acid ; the decomposition frequent- 
 
177 
 
 ly takes place more easily when the acid is more dilute, be- 
 sides there is this consideration, when the acid is fuming, it 
 is exceedingly injurious to the health. 
 
 According to the authority of A. Mitscherlich, the decom- 
 position of a silicate is effected more easily when it is heated 
 with a mixture of hydrofluoric acid and two parts of dilute hy- 
 drochloric acid almost to boiling, which frequently produces 
 a perfect solution. With those silicates which are more diffi- 
 cult of decomposition the fluid must be kept boiling for some 
 time, taking care to renew the acid as it volatilizes. 
 
 If hydrofluoric acid is ready at hand, the object in view is 
 most quickly attained. But the acid must be pure, that is, 
 free from all non-volatile substances (lead, lime, alkalies). 
 
 The presence of silicic acid, more properly of hydroflouosili- 
 cic acid, is not injurious. If the acid has to be prepared, either 
 fluor spar or kryolite is decomposed by means of concentrated 
 sulphuric acid ; the latter of these is to be preferred on ac- 
 count of its purity. The distillation is performed in a strong 
 cylindrical vessel of lead with a loose-fitting lid which has to 
 be luted with gypsum ; to this lid is attached a platinum-tube 
 growing wider and wider to conduct the vapors into a plati- 
 num dish containing water, into which the end dips through 
 about two-thirds of this fluid. The platinum dish is covered 
 with a concave piece of platinum foil, and placed in a vessel 
 tilled with quite cold water. The acid is caused to be evol- 
 ved and passed through the water in the dish until this wa- 
 ter has become sufficiently saturated with it. 
 
 The manner of submitting a silicate to the action of hydro- 
 fluoric acid is two-fold. Either the tine powder, previously 
 weighed, is shaken into the prepared acid, or it is placed in 
 the platinum dish, covered with water, and then the vapors 
 of the acid are conducted upon it through the platinum tube 
 to saturation. The latter method in general is to be prefer- 
 red. Whichever way, however, be followed, it is always 
 advisable afterwards to place the mixture aside for at least 
 twelve hours.. 
 
178 
 
 In order to volatilize the fluoride of silicon and to convert 
 the fluorides into sulphates, the fluid must be evaporated to 
 dryness with sulphuric acid. The evaporation is effected on 
 the water-bath, and as regards the success of the operation, 
 it appears to be indifferent whether the sulphuric acid (not 
 quite concentrated) is added at the beginning or after the eva- 
 poration to dryness. In every case the dish is kept over the 
 water-bath as long as acid vapors are evolved, when these 
 cease, it is placed over the flame of a lamp and heated to a 
 higher temparature in order to drive oif all excess of sulphu- 
 ric acid ; finally the temperature is raised until the bottom 
 of the dish is of a low red heat. As soon as the dish has cool- 
 ed, hydrochloric acid is poured upon the mass, which is di- 
 gested for some time ; after which a large quantity of water 
 is added and the mixture is heated to boiling. It is now that 
 we can judge, whether the silicate has been completely de- 
 composed or not ; if completely, the solution must be clear. 
 It is only with silicates rich in lime, that the largest part 
 of the sulphate of lime remains undissolved. But it happens 
 not unfrequently, that there is a small quantity of residue, 
 this is removed by filtration, it consists either of undecompo- 
 ed substance or sulphate of baryta (strontia, lime) or titanic 
 Acid, and requires, therefore, to be accurately investigated. 
 
 The acid solution of the bases is then examined for their 
 determination according to the rules already given. 
 
 H. Rose recommends, instead of hydrofluoric acid, the 
 employment of the acid fluoride of ammonium (HFl-f AmFl), 
 which naturally must be quite free from all impurities fixed 
 in the fire. One part of the finely pulverized silicate is mix- 
 ed in a platinum dish or in a capacious crucible with seven 
 parts of fluoride of ammonium ; water is then added and the 
 mixture is stirred up into a pulpy consistence and cautiously 
 heated until the mass is dry. The temperature is then rais- 
 ed to dark-red ignition and maintained at this heat as long 
 as any vapors are evolved. Sulphuric acid is poured upon 
 the residue, which is heated in order to drive off all excess 
 
179 
 
 of sulphuric acid. The further treatment, digestion with 
 hydrochloric acid, &c., has been given already. If there 
 should remain any residue of undecomposed substance after 
 solution, it must be heated again with fluoride of ammonium. 
 
 The separation of the bases is performed in general as with 
 the silicates decomposable in acids, with the exception that 
 the presence of magnesia modifies the course. It must, how- 
 ever, never be forgotten, that thereby no material, fixed in 
 the fire, must be introduced into the analysis, and conse- 
 quently the reagents must be quite free from such. 
 
 1. SILICATES containing NO MAGNESIA. 
 
 If the bases are alumina, lime, soda and potassa, the pro- 
 cess is that on page 152. If to these are added protoxide of 
 iron (protoxide of manganese), the processes on page 152 and 
 156 are combined. 
 
 2, SILICATES containing MAGNESIA, 
 
 BASES : alumina, lime, magnesia, soda and potassa. As 
 soon as the alumina has been precipitated by ammonia, the 
 lime by oxalic acid, the filtrate evaporated to dryness and 
 the mass freed by heating in the crucible from the ammoni- 
 acal salt, the separation of the magnesia may be effected in 
 two different ways : 
 
 a.) The residue, consisting of the sulphates of magnesia 
 and of the alkalies, is dissolved in the smallest quantity of 
 water, to which is added a sufficient quantity of the concen- 
 trated solution of carbonate of ammonia (each litre contain- 
 ing 230 grammes of solid carbonate of ammonia and 180 
 grammes of ammonia of the specific gravity of 0.92). The 
 mixture is well stirred together and then set aside for twenty- 
 four hours. By this means the magnesia is precipitated com- 
 pletely as double carbonate of ammonia and magnesia. The 
 precipitate is removed by filtration, washed with a quantity 
 of the cold precitant, dried, ignited and weighed as pure 
 magnesia. The filtrate is evaporated to dryness; and the 
 
180 
 
 residue is placed in a weighed porcelain crucible and heated 
 cautiously and repeatedly with chloride of ammonium (vide 
 page 161) in order to convert the sulphates into a mixture of 
 chloride of potassium and chloride of sodium. This is weigh- 
 ed, dissolved and treated with chloride of platinum, &c., to 
 effect the separation of potassa and soda (vide page 154). 
 
 OBSERVATION. The precipitate of the doable carbonate of ammonia and magnesia 
 
 contains a small quantity of the corresponding alkaline salt, if this alkali happens to 
 preponderate. After ignition the magnesia is then mixed with a little alkaline carbo- 
 nate, which is extracted by means of hot water, after which the magnesia is weighed ; 
 and the alkaline extract is added to the filtrate. 
 
 b.) The residue is dissolved in a large quantity of water ; 
 to the solution is added crystallized hydrate of baryta or a 
 recently prepared solution of this substance ; the solution is 
 boiled for a few minutes and then filtered ; finally the residue 
 on the filter is washed with hot water, it consists of sulphate 
 of baryta, magnesia and carbonate of baryta, whilst the fil- 
 trate, containing in addition to the alkalies an excess of ba- 
 ryta, becomes covered even during filtration with a film of 
 carbonate of baryta. 
 
 The filter is digested with dilute hydrochloric acid, to this 
 immediately a sufficient quantity of sulphuric acid is add- 
 ed so as to precipitate the baryta that has been dissolved, 
 and then filtered. Ammonia is added in excess to the fil- 
 trate, and the magnesia is precipitated with phosphate of soda 
 (vide page 133). 
 
 The filtrate containing the alkalies is digested with a mix- 
 ture of ammonia and carbonate of ammonia ; the carbonate 
 of baryta is separated by filtration; the filtrate is evapo- 
 rated to dry ness ; finally it is placed in a platinum crucible, 
 to which hydrochloric acid is added in slight excess, and then 
 the mixture is evaporated to dryness. The alkaline chlorides 
 are ignited in a covered crucible at a low red heat, then they 
 are weighed, and finally separated by chloride of platinum. 
 
 If the oxides of iron (manganese) are also among the base*, 
 the course of the analysis is changed only in the manner al- 
 ready given. 
 
181 
 
 D. Decomposition of Silicates by means of Sulphuric 
 
 acid. 
 
 Some silicates, upon which hydrochloric acid has scarcely 
 any action at all, for instance magnesian mica, are complete- 
 ly decomposed by sulphuric acid. The substance is reduced 
 cither to an impalpable powder or in general to a fine state 
 of division, then covered with a mixture of equal parts of 
 concentrated sulphuric acid and water in a capacious plati- 
 num crucible, which is then covered and heated to boiling 
 and maintained at this temperature until the mass is dry, af- 
 ter which it is raised gradually to a low red heat. As soon 
 as it is cool, it is again moistened with sulphuric acid, and 
 the evaporation and ignition are repeated. After these oper- 
 ations it is covered with a little hydrochloric acid, and diges- 
 ted ; water is added ; the mixture is heated ; and the silicic a- 
 cid is removed by filtration, and has afterwards to be tested 
 as to its purity. The bases are separated according to the 
 known methods, only bearing in memory, that they are part- 
 ly in combination with sulphuric acid. 
 
 If this is employed to silicates rich in lime, it is possible 
 that a part of the sulphate of lime separates at the same time 
 with the silicic acid, which is afterwards converted by boil- 
 ing with a solution of carbonate of soda into carbonate of 
 lime. 
 
 According to A. Mitscherlich, such silicates as do not ad- 
 mit of being decomposed at all or imperfectly by this method., 
 may be dissolved by heating a mixture of one part of the sub- 
 stance in question with two parts of water, and six parts of 
 sulphuric acid in a glass tube hermetically sealed in an air- 
 bath whose temperature ranges between 392F. and 464F. 
 The possibility of an explosion requires, however, a special 
 apparatus for this experiment. 
 
 Several silicates are not easily dissolved after fusion with 
 bisulphate of potassa ; and the separated silicic acid can not 
 be obtained pure by washing. 
 
182 
 
 E. Determination of the Degree of Oxidation of Iron 
 in Undecomposable Silicates. 
 
 It is an easy matter in silicates, that can be decomposed 
 by acids, to show the presence either of the protoxide of iron; 
 >r of the sesquioxide of iron or of both together, and to de- 
 termine the amount quantitatively (vide page 152); but it is* 
 just the reverse in silicates that can not be decomposed by 
 acids. 
 
 T. METHOD. The fine powder is mixed in a platinum 
 erucible with six times its weight of powdered calcined borax ; 
 a layer of this is then laid over the top of the mixture ; the- 
 platinum crucible is placed in a larger crucible and embed- 
 ded in magnesia; both are then covered and heated either 
 in a gas-furnace (in which case a Hessian crucible will be a 
 sufficient protection for the platinum crucible) to as intense 
 a heat as possible and until the silicate has been fused in the- 
 borax (from a quarter to half an hour). The inner crucible 
 together with its contents, is then weighed ; after which the 
 fused mass is loosened from the crucible by bending and pres- 
 sure upon the sides, it is then reduced to a rough powder r 
 weighed portions of which are used for repeating the expe- 
 riment. The substance is dissolved in dilute hydrochloric 
 acid, taking care to exclude all access of air, after which the- 
 amount of protoxide of iron is determined by the volumetric 
 process with a solution of permanganate of potassa (vide 
 page 65). 
 
 This method gives too small a quantity of iron, owing 
 probably to the oxidation during the fusion with borax. If 
 therefore the amount of protoxide of iron found by the volu- 
 metric iron test*be compared with the amount of iron found 
 by means of the analysis and computed from the sesquioxide, 
 there turns out to be a small amount of sesquioxide of iron, 
 even in those cases where the latter does not exist. 
 
 II. METHOD. According to A. Mitscherlich the silicate 
 in fine powder obtained by elutriation is exposed in an ami- 
 
183 
 
 ratoly covered platinum dish with a mixture of hydrochloric 
 and hydrofluoric acids to a temperature of nearly 212F, un- 
 til all is either dissolved or decomposed, after which the flu- 
 id is highly diluted with water, and submitted to the vol- 
 metric test for the protoxide of iron. 
 
 III. METHOD. According to the same authority the si- 
 licate, reduced to a tine powder by elutriation, is decompo- 
 sed in the way described on page 181 in a glass tube herme- 
 tically sealed, by means of sulphuric acid, after which the 
 fluid is submitted to the volumetric iron test as before. If a- 
 ny sesquioxide of iron is present, it is reduced by zinc in a 
 second experiment (vide page 64) ; and in this way the total 
 .amount of iron is determined. 
 
 F. Analysis of Mixed Silicates. 
 
 Several stones in a crystalline condition consist of such si- 
 licates, of which one can easily be decomposed by acids, tho 
 other with difficulty, if at all. Since it is very important for 
 our knowledge of stones to know intimately the minerals of 
 which they are composed, and since no certainty in this res- 
 pect can be drawn from an observation of them in their 
 varied finely granulated forms, an analysis is of inestimable 
 value in determining the nature of their constituents. Me- 
 teorites also (meteoric stones) are constituted in a great mea- 
 sure like the mixed rocky formations; and with these too 
 the analysis under favorable conditions leads to a knowledge 
 of their constituents. The principle of the analysis of such 
 a mixture is to decompose one silicate by treatment with an 
 ticid, and then to determine the bases in combination with it, 
 und which are now in the acid fluid. The residue, consisting 
 of the silicic acid separated from the silicate, as well as of the 
 whole amount of indecomposable silicate, is boiled with a so- 
 lution of carbonate of soda in order to separate this silicic a- 
 rid (whose amount is found either directly out of its solution 
 or indirectly) ; after which the indecomposable silicate is an- 
 alyzed by itself. 
 
184 
 
 According to the nature of the silicate such a chemical op 
 oration can never be quite accurate, on account of its decompo- 
 sability by acids, a circumstance alluded to (vide page 141) r 
 possessing properly speaking no sharp line of demarcation, 
 hut being merely one of degree, so that, the more concentra- 
 ted the acid employed, the longer the time of its action, and 
 the higher the temperature, so much the more of the so-call- 
 ed undecomposable silicate will undergo decomposition at the 
 same time with the decomposable part. The result will be 
 the more distinct and definite, the more opposite the two si- 
 licates are with regard to each other in their respective de- 
 composabilities ; for in this case a highly concentrated acid 
 will not be employed. This is, for instance, the case, if the 
 mixture consists of a silicate that gelatinizes with acids (ne- 
 pheline, hauyne, olivine rich in iron, even anorthite) as w 
 shown by my analyses of a pseiidomorpous* specimen of leu- 
 cite consisting of nepheline and vitreous felspar, and of the 
 meteoric stones consisting of anorthite and augite by Juve- 
 nas and Stannern.f 
 
 Unfortunately, however, this favorable circumstance is re- 
 latively rare. If the decomposable silicate requires a con- 
 centrated acid (olivine containing but little iron), the other 
 (augite, labrador) will be considerably attacked, and the re- 
 sult loses in definite character. 
 
 A further difficulty lies in the presence of other minerals 
 in addition, which are met with sometimes either in minute 
 quantities (accessory admixtures) or in larger quantities (as 
 essential admixtures). If they are not silicates, but com- 
 pounds soluble in acids, such as magnetic iron ore, titanife- 
 rous iron, sulphuret of iron or metallic iron '(in meteorites 
 combined with nickel, &c.), apatite, titanite, they may be 
 sometimes removed by the magnet from the powdered mate- 
 rial, or, in case their recognition is in any way possible, their 
 amount may be computed from the quantity of one of the 
 constituents peculiar to them as found by means of the 
 
 *Poggend. Ann. 98, 53. 
 tPoggencl, Ann. &3, 585. 
 
185 
 
 analysis. Pyrites or magnetic pyrites in this way is compu- 
 ted from the amount of sulphur, apatite from the phosphoric 
 acid, titanite from the titanic acid, and magnetic iron from 
 the sesquioxide of iron. But it will easily be comprehended, 
 that these expedients can not always be employed. 
 
 But it is quite impossible to make a sure computation of 
 the analysis, if there are more than two silicates in the 
 mixture, so that either the decomposable or the undecompo- 
 sable part, or indeed both of them consist of several silicates. 
 Thus for instance in several phonolites the undecomposable 
 part, it is true, consists only of vitreous felspar, w r hilst the 
 decomposable part is a mixture ; in the lavas of Vesuvius 
 the former is only augite, the latter, however, consists of a 
 mixture of leucite, nepheliue and other minerals ; in several 
 meteoric stones the earthy matrix as regards the decomposa- 
 ble part is only olivine, but the undecomposable part consists 
 probably of several silicates ; finally in basalt both parts are 
 of a mixed nature. The computation of the analysis on the 
 presupposed determined minerals therein is merely hypothe- 
 sis ; it indicates sometimes with equal probability two quite 
 different minerals, and oftentimes the result is any thing but 
 satisfactory. 
 
 These operations have especial reference to the analysis 
 of mixed silicates, which do not admit of being separated by 
 an acid, and which are analyzed as a whole, and then the at- 
 tempt is made by computation to derive a solution as to their 
 probable mixed constituents. Sometimes this can be done 
 and with success, for instance if only two silicates are present, 
 if their bases are not exactly the same, and the condition of 
 the whole renders the presence of either one or the other of 
 the constituents very probable. Suppose for instance we 
 have a stone, which is either dolorite or a doloritic lava, and 
 we analyze the whole, it will be easy enough to compute the 
 relative quantities of labrodorite and augite, taking it for 
 granted that augite is free from alumina, and that labrodo- 
 rite contains no iron, which strictly speaking is not correct, 
 24 
 
186 
 
 still the computation will not be essentially modified. As- 
 cribing so much lime to the alkalies found as to give an am- 
 ount of oxygen equal to one-third of that of the alumina, and 
 so much silicic acid as to represent an amount of oxygen dou- 
 ble that of the alumina, we thus get the amount of labro do- 
 rite ; and the rest must be so constituted, if it is to be augite, 
 that the oxygen of the protoxide of iron, of the magnesia and 
 of the remaining lime must be half as much as that of the 
 remaining silicic acid. 
 
 If it should happen in these mixed silicates that either of 
 the constituents should present itself in large and separate 
 particles, it would naturally be of essential service in leading 
 to the denomination of the compound, and especially so if 
 these separate particles can be analyzed alone. 
 
 The accurate analysis of compound silicates, among which 
 may be reckoned also the meteorites, is indisputably a diffi- 
 cult task, in which we must be careful not to overlook seve- 
 ral rare substances (oxides of chrominm and titanium). 
 
 BASALT, 
 
 We will take basalt as an example of a mixture of decom- 
 posable and uudecomposable silicates, and will suppose, that 
 the modification to be analyzed, as is frequently the case, 
 shows distinct particles of olivine and augite, and that the 
 qualitative analysis has indicated the presence of carbonic 
 acid, water and phosphoric acid together with small quanti- 
 ties of chlorine. Furthermore we will presuppose, that the 
 magnet has extracted from the powdered stone black parti- 
 cles that turn out to be magnetic iron ore free from titanium. 
 
 A larger quantity of the stone, as much as will be nececsa- 
 ry for the various experiments to be undertaken, is reduced 
 to a fine powder, which is intimately mixed together and 
 preserved in a close vessel. 
 
 a.) One part is dried in the desiccator, as long as its 
 weight undergoes any diminution. This gives the amount 
 of the hygrometric water. The dried powder is ignited in 
 
1ST 
 
 a crucible whilst dried carbonic acid is conducted through a 
 hole in the cover to the contents, in order to prevent the ox- 
 idation of the protoxide of iron (of the magnetic iron, olivine 
 and augite). The loss of weight gives the amount of water 
 chemically combined. If the basalt should contain carbonic 
 acid, that is, carbonate of lime, a part of it may escape, and 
 this part may not be restored whilst cooling in a current of 
 this gas. In this case we estimate the amount of carbonic a- 
 cid in the ignited powder and compare it with the quantity 
 obtained in Z>. 
 
 OBSERVATION When carbonate of the protoxide of iron is present the determina- 
 tion of the amount of water is not quite accurate, because this salt when ignited leaves 
 behind ferroso-f erric oxide, and evolves carbonic oxide together with carbonic acid. 
 
 b.) A part is decomposed in the carbonic acid apparatus 
 (vide page 129) by means of hydrochloric acid ; this gives the 
 amount of carbonic acid. 
 
 6'.) A larger quantity (from six to nine grammes) is mix- 
 ed intimately in a deep dish with a little water, then cover- 
 ed with a sufficient quantity of moderately concentrated hy- 
 drochloric acid, and the whole is heated, stirring the mix- 
 ture all the while, until it becomes gelatinous. (In this ex- 
 periment it is well either to weigh or measure the amounts 
 of water and of acid, in order, in case the experiment should 
 have to be repeated or for the volumetric analysis of iron, to 
 be able to use the same proportion of substance for both expe- 
 riments.) The mass is then diluted with water and filtered, 
 and the gelatinous residue is washed with hot water. We 
 have now: A. The acid solution of the bases of the decom- 
 posable silicates and of the carbonates, together with the so- 
 lution of magnetic iron ore and of apatite ; B. The silicic ac- 
 id of the former together with the undecomposable silicates. 
 
 A. This solution contains a certain amount of silicic acid; 
 it has therefore to be evaporated to dry ness, according to page 
 127, on the water-bath, &c. After the separation of the sili- 
 cic acid a current of hydrosulphuric acid is passed through it, 
 by which meant* small quantities of copper, tin, &c., are fre- 
 quently indicated. As soon as the gas in excess has been dri- 
 
188 
 
 ven off by heat, and the iron that has been reduced to the 
 state of protoxide has again been oxidized by means of hydro- 
 chloric acid, ammonia is added in excess, and the bulky pre- 
 cipitate is removed by nitration and washed with hot water. 
 It consists of sesquioxide of iron and of alumina, but it con- 
 tains also a part of the magnesia, and also of the lime, besides 
 protoxide of manganese and all the phosphoric acid. The 
 lime exists, partly at least, as carbonate of lime, which has 
 been formed during the nitration. 
 
 As soon as the precipitate has been dried, ignited and 
 weighed, it is triturated, heated and weighed once more (be- 
 cause there has been incurred a trifling loss) and mixed with 
 about as much pure silicic acid and six times its amount of car- 
 bonate of soda. The determination of the relative amounts 
 of sesquioxide of iron and alumina is effected either as there 
 indicated, or by fusion with hydrate of potassa (vide page 
 156). If the amount of manganese is so great as to render 
 a separate determination apparently preferable, the process 
 to be followed is found on page 179. 
 
 The nitrate from the precipitate by ammonia contains lime, 
 magnesia and both the alkalies. It is analyzed as indicated 
 on page 179. B. In the next place the free silicic acid, 
 which belongs to A. in the residue is separated from the un- 
 decomposable silicates ; to effect this separation there is the 
 choice of two methods. 
 
 Either the contents of the filter while still moist are trans- 
 ferred gradually into a boiling moderately concentrated solu- 
 tion of carbonate of soda, to which also at the last the filter 
 is added ; the mixture is then treated according to instructions 
 on page 150. The residue is washed with hot water (if the 
 filtrate is turbid, it is collected by itself, and as soon as it has 
 settled, it is again filtered on a separate filter), dried, ignit- 
 ed* and weighed. It now represents the undecomposable 
 silicates of the basalt. The alkaline solution is cautiously aci- 
 dified with hydrochloric acid, warmed for some time and 
 
 *There is a slight inaccuracy produced here from the fact that the protoxide of iron 
 of the augite is raised to a higher state of oxidation. 
 
189 
 
 evaporated to dryness on the water-bath ; the mass is now rais- 
 ed to a higher temperature, and all that is soluble is then ex- 
 tracted with water, and the separated silicic acid is collected 
 on a filter. The weight of this is added to that which was 
 received in A. 
 
 Or the residue is dried, ignited and weighed, and then 
 transferred according to instructions given on page 150, into 
 ji boiling solution of carbonate of soda. The amount of the 
 .insoluble portion alone being determined from the weight, 
 the difference will be the amount of the free silicic acid. 
 
 OBSERVATION. The second method is by far the shorter of the two, but it has this 
 
 disadvantage : the silicic acid after ignition is no longer as easily soluble as before. Af- 
 ter boiling and diluting, therefore, the fluid is allowed to settle, the clear liquid is pla- 
 ced on the filter, and the residue is again boiled with carbonate of soda and water, &c. 
 But even after repeated boiling such a residue always loses a little in weight, when the 
 operation is undertaken afresh ; this must not be forgotten in the computation. 
 
 The undecornposable silicates contain as bases, alumina, 
 the oxides of iron and manganese, lime, magnesia, soda and 
 potassa. Their amount must be large enough for two analy- 
 ses to be made with them, One part is for the determina- 
 tion of the bases, with the exception of the alkalies; it is fu- 
 sed with carbonate of soda and treated according to instruc- 
 tions found on page 155, for the separation of the bases. Tin* 
 residue is decomposed with hydrofluoric acid, and the bases 
 are determined (vide page 179) as regards iron and manga- 
 nese. 
 
 d.) One part is decomposed in a flask, taking care to ex 
 elude the air, and to pay attention to the observations in o. in 
 reference to the quantities; a large quantity is then added, 
 and then the mixture is submitted to the volumetric test with 
 permanganate of potassa ; by this means the amount of prot- 
 oxide of iron in A. is obtained. 
 
 .) One part is decomposed at the ordinary temperature 
 with nitric acid, spec. grav. 1.2; the mixture is then diluted 
 and filtered ; after which the chlorine, derived from the apa- 
 tite, is precipitated \vith nitrate of silver. (The fluorine in 
 the apatite does not admit of being determined directly, it 
 merely be computed by assuming the phosphoric acid and 
 
190 
 the chlorine in the formula for apatite as a base. 
 
 < X. ALUMINA and ALUMINATES. 
 
 Corundum and the compounds of alumina with strong ba- 
 nes (ammiiiates) are not acted upon by acid. They form a 
 part of the spinel group, to which belong spinel, pleonaste 
 (ceylonite, gahnite and chrome iron ore ; whilst another part, 
 in which the oxide of iron is substituted for alumina, (mag- 
 netic iron ore, magnoferrite, franklinite) is soluble in acids. 
 The first mentioned are for the most part isomorphous mix- 
 tures, whose strong bases are magnesia, oxide of zinc, prot- 
 oxide of iron, protoxide of manganese (perhaps also protox- 
 ide of chromium, whilst alumina is .represented partly by 
 the oxide of iron and of chromium, perhaps also the oxide of 
 manganese) as a substitute. 
 
 These minerals, especially corundum, are so hard as to at- 
 tack agate-mortars, and to become impregnated with silicic 
 wid when pulverized therein. Steel mortars are, therefore, 
 used for this purpose ; arid the powder is sifted through 
 gauze ; and the iron particles obtained by trituration are 
 removed by hydrochloric acid. 
 
 The best method of dissolving corundum and the alumi- 
 nates consists in fusing them with bisulphate of potassa. We 
 are indebted to H. Rose tor this plan. For the solution of 
 chrome iron ore it is alone less adapted. A mixture is made 
 in a capacious platinum crucible of one part of the substance 
 in fine powder and four parts of sulphate of potassa also in pow- 
 der, to this is added a sufficient quantity of concentrated sul- 
 phuric acid so as to produce a sort of thick gruel when stirred 
 together with a stout platinum wire ; the crucible is covered, 
 with the exception of a small opening in the lid to allow the 
 contents to be observed, and then cautiously heated, especial- 
 ly at the begin n ing, because it frequently happens that the 
 contents are apt to foam over. Afterwards the heat is raised, 
 the mass is ignited, the crucible being now completely 
 
 .: 
 
191 
 
 covered up, and the ignition is kept up until the content* 
 eease to give out any bubbles, and flow smoothly. Since the 
 solvent action of the bisulphate of potassa depends solely on 
 that part of the acid which converts the sulphate into the bi- 
 sulphate, and since this part is volatilized by ignition, though 
 slowly, and furthermore since the sulphate of alumina in like 
 manner loses its acid under the same circumstances, it is fre- 
 quently advantageous to moisten the cooled mass with sul- 
 phuric acid and to repeat the operations of heating and igni- 
 tion for a short time. 
 
 The mass contains the bases as sulphates, it is digested 
 when cold for some time with hydrochloric acid and after- 
 wards mixed with a sufficient quantity of water and boiled, 
 By this means all is dissolved, unless a portion should hap- 
 pen not to have been decomposed. In this case the unde 
 composed substance is removed by filtration and deducted. 
 
 The separate constituents 'are precipitated from the solu- 
 tion in the ordinary way. 
 
 According to A. Mitscherlich both corundum and the alii- 
 minates may be dissolved in sulphuric acid by exposing a 
 mixture consisting of one part of the elutriated mineral, two 
 parts of water and six parts of concentrated acid, hermet- 
 ically sealed in a glass tube, to a temperature of about 410 
 :F. for several hours. 
 
 1, ALUMINATE of MAG-NESIA and PROTOXIDE 
 of IRON. 
 
 (PROTOXIDE OF MANGANESE.) 
 
 I. METHOD. The acid solution of the fused mass is de- 
 composed by means of chloride of ammonium in quantity 
 proportionate to that of the magnesia. Ammonia is added 
 in slight excess ; and the mixture is then boiled until the va- 
 pors of ammonia have passed off. The magnesia is precipi- 
 tated from the filtrate by means of ammonia and phosphate 
 of soda. The precipitate is dried, ignited and weighed. It 
 consists of oxide of iron (a trace of oxide of manganese) an d 
 
192 
 
 of alumina. In order to determine botli of these, a portion 
 is triturated to an impalpable powder, weighed and dissolved 
 in a flask in hydrochloric acid ; zinc at the same time being- 
 added to effect the reduction of the oxide of iron, after which 
 the protoxide of iron is determined by the volumetric test 
 with permanganate of potassa. (The precipitate, too, may 
 be fused with hydrate of potassa according to the instruc- 
 tions given 011 page 156.) 
 
 II. METHOD. Ammonia is added in slight excess to the 
 solution, after which it is filtered, and the residue is washed 
 with hot water. The magnesia is determined in the filtrate. 
 The precipitate, which also contains magnesia, is fused, after 
 it has been ignited and weighed, with hydrate of potassa (vi- 
 de page 156) ; the alumina is either determined by the differ- 
 ence of weight, or precipitated from the alkaline solution ac- 
 cording to the method found on page 156. The portion, that is 
 insoluble in potassa, is dissolved in hydrochloric acid ; the solu- 
 tion is then highly diluted and boiled (vide page 134) with 
 carbonate and acetate of soda ; and in this way the oxide of 
 iron is separated from the magnesia. 
 
 If the quantity of manganese is more considerable, the* 
 methods found on pages 135 and 157 are employed. 
 
 2, OHEOME IKON ORE, 
 
 (Vide the following article). 
 
 3 XI. COMPOUNDS of CHROMIUM. 
 
 Among the compounds of chromium we may mention chro- 
 mate of potassa, chr ornate of the oxide of lead and chrome 
 iron ore as particularly of importance; on this account we 
 give the methods of their analysis here. 
 
 1. OHKOMATEofPOTASSA. 
 
 A weighed quantity of the salt is dissolved in a capacious 
 flask in a small quantity of water, to this a proper amount of 
 hydrochloric acid and double the quantity of alcohol are 
 
193 
 
 added ; the mixture is then boiled until the fluid appears of 
 an intense green color. The chromic acid is by this opera- 
 tion reduced to the oxide of chromium, which remains in so- 
 lution as chloride, together with chloride of potassium ; whilst 
 chloride of gethyle escapes.' The fluid is placed in a dish ; 
 water is added, and heat is applied until the alcohol has all 
 passed off. The oxide of chromium is then precipitated by 
 neutralization with ammonia and afterwards sulphide of am- 
 monium. The grey-green precipitate of the hydrated oxide 
 of chromium is separated by nitration, washed, dried and ig- 
 nited. It is then pure oxide of cromium, from whose weight 
 the amount of chromic acid may be computed. The fluid is 
 evaporated to dryness ; the residue is placed in a platinum 
 crucible and heated until the chloride of ammonium has been 
 volatilized ; the residual chloride of potassium is then ignit- 
 ed, keeping the crucible completely covered and avoiding too 
 intense a heat. After it has been weighed, it is dissolved 
 in water, and a small quantity of oxide of chromium, which 
 sometimes occurs, is separated by filtration ; this is weighed 
 and added to the quantity previously obtained, but deducted 
 from the weight of the chloride of potassium. From the lat- 
 ter we obtain the quantity of potassa in the salt. 
 
 OBSKRVATION. The commercial neutral chromatc of potassa often contains no in- 
 considerable amount of sulphate of potassa. In this case the mass, remaining after the 
 chloride of ammonium has been volatilized, is to be decomposed with sulphuric acid, 
 again evaporated to dryness, intensely ignited and afterwards neutralized by means of 
 carbonate of ammonia (vide page 113). 
 
 Although the amount of sulphuric acid may be computed, if the amount of the chro- 
 mic acid and of the potassa is known, it may also be determined directly, with the re- 
 servation, however, not after the reduction of the chromic acid by the method above 
 given, because of the easy formation of sulphuric ether, which can not be precipitated 
 by any known base. On this account a fresh quantity of the salt is dissolved in a flask 
 with water, to which hydrochloric acid is then added, it is then boiled for some time, 
 by which a large proportion of the chromic acid is reduced and chrome is evolved. The 
 fluid is then diluted with water, and the sulphuric acid is precipitated by chloride of 
 barium ; the precipitated sulphate of baryta is then washed on the filter with boiling 
 water. If, after ignition, the residue should have a yellowish tinge (containing a little 
 chromate of baryta), it is digested with a small quantity of hydrochloric acid and alco- 
 hol, after which it is again washed, dried, ignited and weighed. 
 
 2. CHEOMATE of the OXIDE of LEAD, 
 
 (ItUD LEAD OJtK, CHROME YELLOW, CHROME ORANGE, 
 CHROME KKI), COLOGNE YELLOW.) 
 
 25 
 
194 
 
 The substance, previously reduced to fine powder, is di- 
 gested in a flask with hydrochloric aicd and alcohol, until the 
 fluid has assumed an intense green color, whilst the residue 
 is a pure white. The former contains chloride of chromium; 
 the latter is chloride of lead. This is collected on a filter 
 (previously dried at a temperature of 248F. and weighed), 
 well washed with alcohol, and dried in the air and afterwards 
 in the drying chamber, until it no longer loses any weight. 
 From this dry chloride the quantity of oxide of lead is com- 
 puted. 
 
 The alcoholic fluid is evaporated, after it has been mixed 
 with water, until all vapors of alcohol have passed oft"; am- 
 monia and sulphide of ammonium are then added, as in the 
 preceding example, which precipitate the hyd rated oxide of 
 chromium; a small quantity of oxide remains in solution, 
 which is obtained by evaporating the filtrate, ignition of the 
 residue and treatment with water. 
 
 OBSERVATION. Commercial chrome yellow frequently contains other substances, 
 
 (and this is the case also with white lead) with which it is adulterated, partly in order 
 to make it cheaper, and partly in order to modify its color. To this class of substances 
 belong sulphate of baryta, sulphate of lime, sulphate of the oxide of lead, alumina, &c. 
 These substances remain as a deposit with the chloride of lead, although it is difficult to 
 determine their amount with complete accuracy, since, for example, the sulphate of 
 lead is also partly converted into chloride of lead. The latter is boiled with water, as 
 long as any thing dissolves ; after which the amount of the residue is determined. Sul- 
 phate of baryta may in this way be determined with accuracy. 
 
 3. OHEOME IRON OEE. 
 
 Chrome iron ore consists of the oxides of iron and chrome 
 as constant constituents, and of alumina and magnesia as 
 sometimes failing. If is not remarkably acted upon either by 
 acids or by the alkaline carbonates as fluxes. If it is ignited 
 with a mixture of carbonate of potassa and hydrated potassa, 
 or with the carbonate and nitrate of an alkali (in a silver cru- 
 cible), there is certainly a decomposition, oxide of chromium 
 being converted into chromic acid, and the iron into oxide of 
 iron, so that water dissolves the former together with the a- 
 lumina as compounds of potassa, whilst the oxide of iron and 
 magnesia remain behind ; but the decomposition, even when 
 
195 
 
 the ore has been elutriated, is always only partial ; and it is 
 difficult to determine the undecomposed part, of which be- 
 sides we can not know with certainty whether the constitu- 
 tion is the same as that which has been decomposed or not. 
 
 But, on the other hand, chrome iron ore may be fused with 
 bisulphate of potassa according to the method given on page 
 190. But very little of this is soluble in water from the fact, 
 that a double sulphate of chrome and potassa has been form- 
 ed which is insoluble in acids. On this account the mass, 
 when cold is covered in the crucible with a mixture of carbo- 
 nate of soda and saltpetre ; the mixture is fused into a fluid 
 mass (this operation slightly attacks the crucible). When cold 
 it is thoroughly extracted with boiling water, then filtered ; 
 and the residue is washed with hot water. 
 
 The alkaline solution contains all the chromic acid and a 
 part of the alumina. A considerable quantity of chlorate of 
 potassa is added, it is then acidified with hydrochloric ac- 
 id and evaporated down until it becomes somewhat consist- 
 ent, adding from time to time a little chlorate of potassa. It 
 is then completely dissolved in water ; and the alumina is 
 precipitated by ammonia. The filtrate is boiled for some 
 time with hydrochloric acid, is allowed to cool ; alcohol is 
 added, in order to complete the reduction ; after this the al- 
 c.ohol is distilled oif ; and the oxide of chromium is precipi- 
 tated by ammonia and sulphide of ammonium (vide page 193). 
 The residue consists of iron, alumina and magnesia, which 
 are determined according to instructions on page 191. 
 
 According to the authority of A. Mitscherlich, chrome ir- 
 on ore is soluble in sulphuric acid, if the mixture is exposed 
 in an hermetically sealed tube to the temperature of 210F. 
 (vide page 191). 
 
 If chrome iron ore is ignited in a current of hydrogen gas, 
 its weight is very little diminished as proved by Moberg's 
 experiments as well as my own. According to Rivot, on the 
 contrary, if the temperature is sufficiently high, all the iron 
 in reduced to the reguline condition, which may then be dis- 
 
196 
 
 solved by dilute nitric acid. This, however, requires con- 
 firmation ; and a better method of analysis is highly desirable. 
 
 XH. COMPOUNDS of TITANIUM. 
 
 Since certain minerals containing titanium, such as tita- 
 nite and titanic iron, occur in many stones, and small quan- 
 tities of titanic acid especially are frequently found in seve- 
 ral minerals (for instance in hornblende, acmite), it can not 
 be said, that titanium is a rare metal. 
 
 In the analysis of undecomposable silicates by means of car- 
 nate of soda, the titanic acid can be found partly in the sili- 
 cic acid and partly in the ammoniacal residue (alumina and 
 oxide of iron). If we have to deal with a silicate, contain- 
 ing an alkali, and this has been dissolved by hydrofluoric ac- 
 id, the residue which remains when the sulphates are dissol- 
 ved (vide page ITS), may also contain titanic acid, as well as 
 the precipitate by ammonia. 
 
 The titanic acid is formed and determined in such residues 
 and precipitates by fusion with bisulphate of potassa and by 
 boiling the solution, as will be described more in detail in the 
 following article. 
 
 1. TITANIC IRON. 
 
 The constituents of titanic iron are titanic acid, the oxide 
 of iron, manganese and magnesia. Sometimes besides these 
 are found small quantities of chromium, tin and lead. 
 
 A. General Analysis. The substance reduced to a fine 
 powder is fused (vide page 190) with bisulphate of potassa. 
 For -one part of the powder from six to eight parts of sul- 
 phate of potassa and the necessary quantity of sulphuric acid 
 are taken and maintained in fusion for some time ; the treat- 
 ment, too, with sulphuric acid is repeated. As soon as the 
 mass is cold, the crucible, together with its contents, is pla- 
 ced in a dish over which is poured a considerable quantity of 
 
197 
 
 water; the whole is then allowed to stand, stirring frequent 
 ly, until all is dissolved. No heat must be applied, because 
 this would cause a turbidity or a precipitation of titanic 
 acid. If any thing remains undissolved, it is either* unde- 
 composed titanic acid or silicic acid in admixture, which iv 
 removed by filtration and deducted. The clear solution is 
 decomposed with a solution of sulphurous acid or with the bi- 
 sulphite of an alkali, in order to convert the oxide of iron in- 
 to the protoxide, and boiled for some time (about an hour), 
 replenishing the water lost by evaporation, and 'adding also 
 from time to time a little sulphurous acid. The titanic acid, 
 that has been precipitated, is collected on a filter and washed 
 with hot water. The nitrate is again heated to boiling, in 
 order to be certain by a continuation of the boiling that no 
 more titanic acid is precipitated. The titanic acid is dried, 
 ignited and weighed. 
 
 OBSERVATION. Notwithstanding every precaution the titanic acid frequently con- 
 tains a small quantity of oxide of iron. The acid can be separated from the oxide by 
 allowing it to settle after precipitation by boiling, by decanting the clear fluid, by dis- 
 solving the residue (in a platinum dish) in sulphuric acid, and by evaporating the solu- 
 tion, until the largest part of the acid in excess has been expelled. When this is cold, 
 a small quantity of water is added gradually so as to avoid raising the temperature too 
 high, the solution is-then highly diluted and sulphurous acid is added; the mixture is 
 then boiled as before prescribed. The small quantity of iron then remains in solution. 
 
 Any sulphurous acid that may remain in the filtrate is dri- 
 ven off by heat, chlorate of potassa is then added in order to 
 oxidize the iron, afterwards chloride of ammonium and am- 
 monia are added in excess. In this way the oxide of iron (to- 
 gether with a little oxide of manganese) is precipitated. In 
 the filtrate are estimated lime and magnesia. The oxide of 
 iron may be also separated from the earths by means of car- 
 bonate of soda and acetate of soda (vide page 134). If the 
 amount of manganese is considerable, and has to be determi- 
 ned, proceed as indicated on page 135. 
 
 B. Determination of the Degree of Oxidation of the Iron. 
 
 This is best effected by the volumetric process, for which 
 there are two methods : 
 
 a.) The substance reduced to the most impalpable pow- 
 der possible is boiled in a vessel from which air is excluded 
 
198 
 
 tor some time wth hydrochloric acid, until it is dissolved,. 
 that is, until the black color has disappeared. If there is- 
 much |itanic acid present, a small quantity separates. The 
 solution is diluted, and is then poured into a large quantity of 
 water in which the protoxide of iron is determined by per- 
 manganate of potassa in the volumetric process. 
 
 b.) The titanic acid is dissolved as before in hydrochloric 
 u-id. The solution, as soon as it is cold, is placed in the 
 same flask which forms a part of the apparatus described on 
 page 45, and is mixed with an accurately measured quantity 
 of a dilute solution of chlorate of potassa, of a given strength 
 as described on page 45. 
 
 A quantity of this solution is taken containing an amount 
 of the chlorate sufficient to oxidize all the iron found in A. r 
 if present in the form of protoxide. The solution is placed 
 in a pipette and is allowed to flow into the flask, at the same 
 time the cork with the bent glass tube is quickly placed where 
 it belongs, whilst the other end of the tube dips into the so- 
 lution of iodide of potassium. By heating and boiling, the 
 chlorine is driven out, which releases an equivalent quantity 
 of iodine, which is determined by means of sulphurous acid 
 and a standard solution of -iodine. By this means we ascer- 
 tain how much chlorate of potassa has not been used for the 
 oxidation of the protoxide of iron. The difference gives the 
 quantity used, and from this is derived the quantity of the 
 protoxide of iron. 
 
 1 equivalent of KO + 01O S =6 01 = 6 O 
 
 KO = C1O B = 123.5 12 FeO=432. 
 
 Therefore one part of chlorate of potassa converts S.5265& 
 parts of protoxide of iron into the sesquioxide. 
 
 Xm, COMPOUNDS of FLUORINE. 
 
 The mode of determining fluorine, when the metallic fluo- 
 rides are in combination with the phosphates ur silicates, ha^ 
 
199 
 
 already been shown on pages 191 and 170, At present we 
 liave to deal with their analysis, when they occur alone. 
 
 The most common fluorides are fluor spar (fluoride of calci- 
 mm) and cryolite (the double fluoride of aluminum and sodi- 
 mn). We select the latter as an example. 
 
 1, OEYOLITE. 
 
 The substance in fine powder is heated in a platinum cru- 
 cible with concentrated sulphuric acid, until the decomposi- 
 tion is complete and also the largest part of the acid in excess 
 has been driven off. The mass when cold is digested with a 
 small quantity of hydrochloric acid, and the sulphates of alu- 
 mina and soda are dissolved in water. The first of these i> 
 -precipitated by ammonia or sulphide of ammonium ; the fil- 
 trate is evaporated to dryness; after which the residual sul- 
 phate of soda is ignited and rendered neutral by carbonate of 
 ammonia (vide page 113). 
 
 The amount of fluorine is found by the loss. 
 
 XIV. STOECHIOMETKICAL COMPUTATION 
 
 OF THE ANALYSIS. 
 CONSTEUOTION of a FOEMULA. 
 
 As soon as the analysis is completed, and the quantities of 
 the constituents of the substance analyzed have been compu- 
 ted, by aid of the tables, from the amount of the precipitates, 
 <fcc., found, it is customary for the sake of comparison to as- 
 certain the percentage of each component of the substance 
 investigated. This is the empirical result. 
 
 If all the constituents of a body have been determined with- 
 out exception in a direct manner, it will invariably happen 
 that either a low or an excess will be the result. The former 
 is inevitable and is to be attributed to the different 'manipu- 
 lations, and increases with their number; on this account, as 
 already remarked (vide page 3), every unnecessary manipu- 
 
200 
 
 lation, for example, every unnecessary change of vessel must 
 be carefully avoided. The loss, therefore, seems to stand in 
 a certain ratio with the number of the constituents; never- 
 theless this must be taken only in a general sense, since the 
 loss depends essentially on the nature of these constituents,, 
 from the fact that many bodies admit of being determined 
 with greater accuracy than others. 
 
 If a substance contains only two or three constituents* 
 the loss by the analysis at the highest ought not to amount. 
 to more than 0.5 per cent ; whereas if the substance is a very 
 complicated mixture or consists of constituents that can with 
 difficulty be determined, the loss may amount to 2 per cent 
 without the implication of any great error in the analysis. 
 
 In every case, in the summation of the result, the loss must 
 be given with all honesty and uprightness; for the method 
 of dividing it up among all the constituents in proportion to 
 their respective quantities, in order to make it disappear, as 
 is sometimes the practice, is to be avoided altogether. For 
 it is easy to conceive, that the loss maybe attributed in pre- 
 ference to either one or other of the constituents ; and it is- 
 evident, that the one first separated in all probability will bo 
 more correct, than those determined afterwards. 
 
 An excess in analyses is either real or only apparent. The 
 first occurs with beginners just as frequently as the contrary, 
 a loss, and arises in general from the imperfect washing of 
 the precipitates, a manipulation that has to be performed 
 with great care. It is no rare occurrence to find that a pre- 
 cipitate, which is taken for a pure oxide, is in reality either 
 a basic salt or a double salt, and that thus either the acid or 
 the base is the cause of the excess in question. For instance^ 
 suppose that magnesia is precipitated by an alkaline carbo- 
 nate, it may happen under certain circumstances, that a dou- 
 ble carbonate will be thrown down ; alumina, that has been 
 precipitated from a solution containing potassa and sulphuric 
 r-irid, easily retains a part of either base ; and sulphate of ba- 
 ryta, thrown down from solutions continuing nitric ucid. al- 
 
201 
 
 ways contains nitrate of baryta. All such circumstances, 
 that are liable to give rise to error of the kind in question, 
 require to be known ; and, on this account frequent mention 
 has been already made of them. In every case, in which 
 the excess is to be attributed to this cause, the analysis has 
 to be repeated. 
 
 An apparent excess always occurs in analyses, which in o- 
 ther respects are quite accurate, if a constituent exists in the 
 compound either in the reguline condition or in the form of 
 a lower oxide, which the direct result oxidizes or presents as 
 a higher oxide ; as for instance, in a compound containing 
 iron in the form of protoxide, which is always determined as 
 sesquioxide in the estimation of iron. Sometimes this excess 
 may serve as the simplest proof of the presence of the protox- 
 ide, in those cases where (with silicates not decomposable by 
 acids) it is difficult to arrive at any positive evidence, by 
 means of a qualitative examination, of the presence of either 
 the protoxide or the sesquioxide. In compounds containing 
 fluorine an excess is sure to be produced, if the electro-posi- 
 tive constituents are altogether determined in the form of ox- 
 ides. This is the case with the different forms of topaz : si- 
 licic acid, alumina and fluorine ; and the best analyses indi- 
 cate an excess varying between 6 and 8 per cent, arising from 
 the oxygen of those quantities of silicic acid and alumina 
 which existed in the compound as silicon and aluminium 
 with fluorine. 
 
 It has already been observed (vide page 9) that, as soon as 
 an analysis is complete, it becomes necessary to test the re- 
 spective constituents obtained, as to their purity. 
 
 STCECHIOMETEICAL COMPUTATION of an ANA- 
 LYSIS, 
 
 As it is only with purely chemical compounds, and not 
 with mixtures, that such a computation is possible, it is ev- 
 ident that it can not be applied to every analysis. The ana- 
 lytical investigation of mixtures is indeed very important, al- 
 though in general it is of value only for technical purposes, 
 
202 
 
 as is the case with slags, glasses, alloys, &c; nevertheless 
 such mixtures frequently approximate so nearly to purely 
 chemical compounds, as to allow them to be submitted to a 
 stoechiometrical computation, and to deduce from their ana- 
 lysis a chemical formula, approximatively at least, corres- 
 ponding to them. 
 
 The aim of the stoechiometrical computation is to search 
 for the relative and the absolute number of equivalents 
 in which each constituent exists in the compound. Since 
 now every chemical compound consists of a fixed number of 
 constituents, that is, since it can contain its constituents on- 
 ly in accordance with the ratio of weights, which corresponds 
 to one or several equivalents (atoms) of these constituents, it 
 is evident that, when an analysis of a substance shows such a 
 ratio of weights, the substance itself must be a chemical com- 
 pound, and not a mixture. 
 
 The stoehiometrical computation of an analysis shows whe- 
 ther the substance under investigation is a chemical com- 
 pound or not. But supposing a substance has been analyzed, 
 whose characters in other respects indicated beforehand such 
 a compound (for example the crystalline condition), and for 
 all that the result is found not to accord with the laws of de- 
 finite proportions, that is, the weights of the constituents do 
 not stand in the same ratio to each other as the chemical eq- 
 uivalents or atomic weights or their multiples do to each oth- 
 er ; this is a proof that either the analysis is erroneous, or 
 the substance is not pure (crystals may contain foreign ingre- 
 dients enclosed in their substance). In such cases and for 
 such reasons the computation exercises an important control 
 as regards the accuracy of the analysis. 
 
 THE GENERAL MODE of making the COMPUTATION. 
 
 The analysis has already furnished us with the relative a- 
 mounts of weight of the constituents, and the question now 
 presents itself: What is the relative number of atoms (che- 
 mical equivalents) of these constituents ? 
 
203 
 
 Divide the weight of each constituent by the chemical equi- 
 valent (atomic weight) of this constituent. The quotient ex- 
 presses the number of chemical equivalents or atoms. 
 
 Example. The analysis of a specimen of zinc blende by 
 Arfvedson gave the following : 
 
 Zinc 66.34 
 
 Sulphur 33.66 
 
 10000 
 
 The question is: How many equivalents of these elements 
 constitute the combination ? 
 
 The equivalent (atomic wt.) of zinc is 32.5, and that of sul- 
 phur is 16. 
 We have therefore 
 
 66.34 _ 33.66 
 
 = 2.041, and _ = 2.10 
 
 The quotients 2.04 and 2.10 or 204 and 210 express, there- 
 fore, the relative number of equivalents comprehended in 
 the formation of zinc blende. These two numbers are in the 
 ratio of 1 : 1.03 or of 0.97 : 1, that is, so near to the ratio of 
 1 : 1, that the difference may be attributed to some error in 
 the analysis. 
 
 Such deviations from simple ratios occur in the computa- 
 tion of every analysis, and are the consequence of the errors 
 attached to every empirical operation. It is on this account, 
 that stoechio metrical computation aims to ascertain the 
 right chemical proportions, and in accordance with these to 
 apply afterwards the correction to the quotients found. The 
 smaller the correction to be applied, the more correct will be 
 the analysis. If a proportion is found lying between two 
 simple ratios, it will remain doubtful, which is the right one. 
 In sucli a case the analysis is of no use for the stcechiometri- 
 cal computation ; and the substance was either a mixture, or 
 the numerical results of the analysis are not correct. 
 
 If the analysis of the zinc blende in the preceding example 
 had shown, that zinc and sulphur exist in the compound at- 
 om for atom, or in the ratio of single equivalents, theformu- 
 
204: 
 
 la or symbol will be ZnS. 
 
 The most intelligible way of exhibiting the degree of ac- 
 curacy of an analysis consists in placing the computed compo- 
 sition side by side with that which has beenfound. The for- 
 mer is obtained by the addition of the chemical equivalents 
 of the formula and by the reduction of the sum to 100. 
 1 equivalent of zinc = 32.5 = 67.01 
 1 sulphur = 16.0 = 32.99 
 
 4&5 100.00 
 
 48.5 : 32.5:: 100 : x x = 67.01 
 
 48.5 : 16.0:: 100 : y y = 35.99 
 
 If we now place the composition of the compound, as de- 
 rived from the analysis, side by side with that computed, we 
 have: 
 
 Found : Computed : 
 
 Zinc 66.34 67.01 
 
 Sulphur 33.66 32.99 
 
 100.00 100.00 
 
 From this it will be observed, that there is either 0.67 zinc 
 too little, or 0.67 sulphur too much. 
 
 OBSERVATION. If, as in the preceding example, the analysis gives the exact 
 
 sum of 100 parts, one of the two constituents was not determined by a direct method, 
 but was simply deduced by subtraction of its weight from the 100 parts employed. This 
 was the case with the sulphur. It may, therefore, be said, that the analysis has given 
 0.67 too much sulphur. 
 
 This is the mode of proceeding with the computation of ev- 
 ery analysis. By means of the (empirical) formula the com- 
 puted (theoretic) constitution of a compound is sought for, 
 and this is then placed side by side with that which has been 
 found. 
 
 Copper Pyrites. H. Rose found in a crystallized speci- 
 men of this substance : 
 
 Sulphur 35.87 1 equivalent 16 
 Copper 34.40 = 31.7 
 
 Iron 30.47 = 28 
 
 100.00 
 
 The division of the respective parts of 100 by their chemi- 
 cal equivalents gives : 
 
205 
 35.87 34.40 30.47 
 
 The three quotients 2.24, 1.09 and 1.09 have the ratio of 
 3.05 : 1 : 1, or so near to 2 : 1 : 1, that this ratio is undoubt- 
 edly the true one ; from this we see that copper pyrites con- 
 sists of two atoms of sulphur, one atom of copper and one at- 
 om of iron, which is expressed by the empirical formula: 
 Chi + Fe 4- 2 S, or numerically : 
 
 2 equivalents of sulphur = 32,0 
 1 copper = 31.7 
 
 1 iron = 28.0 
 
 9L7 
 which, reduced to parts of 100, are as follows : 
 
 91.7 : 32.0:: 100 : x x = 34.90 
 
 91.7 : 31.7:: 100 : y y = 34.57 
 
 91.7 : 28.0:: 100 : z z = 30.53 
 
 The composition, therefore, of copper pyrites is : 
 
 Computed : Found : Differences : 
 
 Sulphur 34.90 35.87 + 0.97 
 
 Copper 34.57 34.40 0.17 
 
 Iron 30.53 30.47 0.00 
 
 100.00 190.74 
 
 From the differences annexed it will be perceived, that too 
 much sulphur and too little copper and iron have been found. 
 ^and that the sulphur is the least accurate element in the com- 
 pound. The difference of sulphur is almost six times as great 
 as that of copper, or sixteen times as great as that of iron. 
 Since 
 
 34.90 : 35.87:: 100 : 102.8 
 34.57: 34.40:: 100: 99.5 
 39.53 : 30.47:: 100 : 99.8 
 we may say, that 
 
 the excess of sulphur = 2.8 per cent, 
 loss copper = 0.5 
 iron = 0.2 
 The excess of the sum of the substances found by the ana- 
 
lysis over the quantity employed in the operation, namely 
 0.74, is to be attributed evidently to the excess of sulphur 
 found, 
 
 Bitterspar. Hirzel found a crystalline specimen of this- 
 .substance to consist of: 
 
 carbonate of lime 55.06 
 
 magnesia 44.55 
 
 99.61 
 
 The chemical equivalent of lime =28, that of carbonic acid 
 = 22, therefore one equivalent of carbonate of lime=50. 
 Furthermore, since one equivalent of magnesia = 20, one e- 
 quivalent of carbonate of magnesia=45, from these we get: 
 
 55.06 44.55 
 
 = 1.10 = l.Of) 
 
 50 42 
 
 The quotients 1.10 and 1.06 have the ratio of 1.0877 : l r 
 thus again so near a ratio of equality, that is of 1 : 1, as to 
 permit the conclusion, that the specimen of bitterspar em- 
 ployed in the analysis is evidently a compound of single equi- 
 valents of the two carbonates, which is expressed by the for- 
 mula OaO 2 +MgOC 2 . 
 
 The constitution numerically expressed and reduced t<* 
 parts of 100 is as follows : 
 
 L equivalent of carbonate of lime = 50 = 54.45 
 
 \ magnesia = 42 = 45.65 
 
 92 100.00 
 
 A comparison of these numbers with those found by the an- 
 alysis shows that there has been found 0.71 per cent too much 
 carbonate of lime and 1.1 per cent too little carbonate of mag- 
 nesia. 
 
 Sulphate of Copper. The crystallized salt contains accord- 
 ing to Berzelius : 
 
 Oxide of copper 82.18 
 
 Sulphuric acid 31.57 
 
 Water 86.80 
 
 100.00 
 
 The chemical equivalents are: ( 1 uO 8t>.7, SO 8 = 40, 
 
207 
 IIO 9 ; we have therefore 
 
 32.13 31.57 36.30 
 
 -Wi = a809; loo- = - 789 ' To- = 
 
 Dividing the quotients by the smallest, they have the ra- 
 tio of 1.055 : 1 : 5.111, thus evidently = 1:1:5, that in. 
 sulphate of copper is a compound of one equivalent of oxide 
 of copper, one of sulphuric acid and five of water. The cor 
 responding formula is CuSO 3 +5Aq., from which we derive 
 by reduction : 
 
 1 equivalent of oxide of copper 39.7 = 31.84 
 1 sulphuric acid 40.0 = 32.08 
 
 5 water 45.0 = 36.08 
 
 124.7 100.00 
 
 The differences are consequently 
 
 Oxide of copper -f 0.29 
 Sulphuric acid 0.51 
 
 Water + 0.22 
 
 Silicate of the Oxide of Zinc. In a specimen of this min- 
 eral were found : 
 
 Chem: Equiv: Quotients. 
 
 Silicic acid 25.30 30.0 ^ = 0.845 
 
 Oxide of zinc 67.74 40.5 ^7/ = 1.672 
 
 Water 7.54 9.0 ^ji =0.838 
 
 The three quotients are to each other as 1 : 2 : 1 ; the sili- 
 cate of the oxide of zinc is consequently a compound of one 
 equivalent of silicic acid, two equivalents of the oxide of zinc 
 4ind one equivalent of water, corresponding to the formula 
 Zn 2 O 2 SiO 2 +Aq., which numerically gives: 
 
 1 equivalent of silicic acid = 30 = 25.0 
 
 2 oxide of zinc = 81 = 67.5 
 1 water = 9 = 7.5 
 
 120 ~ "TOO 
 
 Here the differences of the analysis are : 
 Silicic acid -f 0.30 
 Oxide of zinc -f- 0.24 
 Water ~f- 0.04 
 
208 
 
 Sodalite. A specimen of sodalite from the TTralian moun 
 tains gave : 
 
 Chlorine 7.10 
 
 Silicic acid 38.40 
 Alumina 32.04 
 
 Soda 24.75 
 
 102.29 
 
 The excess of 2.29 percent is only apparent, since the chlo- 
 rine is in combination with sodium in the compound, whilst 
 the latter is computed as soda. 
 
 The chemical equivalents are : 01 = 35.5, SiO 2 = 30. 
 A1 2 O S = 51.4. NaO = 31 ; the quotients therefore are : 
 
 71.0 38.40 32.04 24.75 
 
 0.2, = 1.28, = 0.624, = S 
 
 35.5 30 51.5 31 
 
 These quotients, when divided by 0.2, have the ratio of 
 1 : (). 1 : 3.1 : 4, that is, of 1 : 6 : 3 : 4, so that sodalite is a 
 compound of one equivalent of chlorine, six equivalents of 
 silicic acid, three equivalents of alumina and four of soda. 
 Since now one equivalent of chlorine requires one equiva- 
 lent of sodium in order to produce one equivalent of chlo- 
 ride of sodium, sodalite contains, properly speaking, three 1 
 equivalents of soda, or it consists of one equivalent .of chlo- 
 ride of sodium arid a silicate of alumina and soda, correspon- 
 ding to the formula : 
 
 NaCI + 3 NaO + 3 A1 2 O 3 -f 6 SiO 2 
 = NaCI -I- 3 (NaO A1 2 O 8 + 2 SiO 2 ). 
 
 Numerically and reduced to the ratio of 100, this formula 
 gives : 
 
 1 equivalent of chlorine = 35.5 7.31 / 
 
 1 sodium = 23.0= 4.78 f : NaO!12 - 4 
 
 (> silicic acid = 180.0 = 37.00 
 
 3 ,, alumina = 154.2 31.75 
 
 3 soda = 94.0 = 19.15 
 
 485/T 1 00.00 
 
 In order to he able, however, to make a direct comparison 
 between the computation and the analysis, the 4.73 for sodi- 
 
209 
 
 um must be converted into 6.38 for soda, which with the 
 19.15, already for soda, gives together 25.53 per cent for so- 
 da. The analysis therefore ought to have given : 
 Chlorine 7.31 
 
 Silicic acid 3T.06 
 Alumina 31.75 
 
 Soda 25.53 
 
 101.65 
 
 in which the excess, 1.65 per cent, corresponds to the oxy- 
 gen required to convert sodium into soda. 
 The differences of the analysis are : 
 
 Chlorine - 0.21 
 
 Silicic acid + 1.34 
 
 Alumina + 0.29 
 
 Soda - 0.78 
 
 Ruby Silver. Bonsdorff found in a specimen of dark ruby 
 silver from Andreasburg : 
 
 Equiv : Quotients. 
 
 Sulphur 17.78 16 ^^- = 1.11 
 
 Antimony 23.26 120.3 fJv 2 .! = - 19 
 
 Silver 58.96 108 4^ - - 55 
 
 100.00 
 
 The three quotients have the ratio of 6.4 : 1 : 2.9, which 
 =6:1:3, from this it is seen that there are present 6 equi- 
 valents of sulphur, 1 equivalent of antimony and 3 equivalents 
 of silver. This is expressed by the empirical formula 3 Ag 
 + Sb + 6 S, which numerically gives: 
 
 6 equivalents of sulphur = 96 = 17.77 
 1 antimony = 120.3= 22.26 
 
 3 silver = 324 = 59.97 
 
 540.3 100.00 
 
 The differences are : 
 
 Sulphur + 0.01 
 
 Antimony 1.00 
 Silver 1.01 
 
 27 
 
210 
 
 Computation of the Proportion of Oxygen (Sulphur) in 
 the, Acid and the Base. 
 
 The stoechiometrical computation of the secondary combi- 
 nations is obtained by first getting the relative weight of the 
 common constituent in the two primary combinations, of 
 which the secondary combination consists. This is known 
 to be always definite and in general very simple. 
 
 This mode of computation is therefore applied firstly to the 
 great class of oxygen salts, then to the sulphur salts and the 
 haloid double salts. We will employ the examples already 
 given, as far as they refer to this subject, in illustration of 
 our remarks. 
 
 /Sulphate of Copper. Oxide of copper contains 20.14 per 
 cent of its weight of oxygen, sulphuric acid |-, and water -|. 
 We have therefore : 
 
 100 : 20.14 :: 32.13 : x x = 6.47 
 
 5 : 3 :: 31.57 : y y = 18.94 
 
 9 : 8 :: 36.30 : z z = 32.27 
 
 The numbers 6.47 : 18.94 : 32.27 are = 1 : 2.9 : 5. The 
 next simple ratio is 1 : 3 : 5. Thus the acid in sulphate of 
 copper contains three times as much oxygen, and water five 
 times as much oxygen as the base, and since one equivalent 
 of sulphuric acid contains 3 equivalents of oxygen, and one 
 equivalent of water contains 1 equivalent of oxygen, it is ev- 
 ident the same result is obtained by this mode of computa- 
 tion as by the preceding, namely, that sulphate of copper 
 consists of 1 equivalent of oxide of copper, 1 of sulphuric a- 
 cid and 5 equivalents of water. 
 
 Calamine. Since silicic acid contains 53.3 per cent of ox- 
 ygen, and the oxide of zinc 19.75 per cent, we find that : 
 25.30 silicic acid = 13.48 oxygen 
 67.74 oxide of zinc 13.38 
 7.54 water = 6-70 
 
 The quantities of oxygen have the ratio to each other 
 of 2 : 1.997 ll, consequently as 2:2: 1, which, again leads 
 to the formula 2 ZnO + SO 2 + Aq. 
 
211 
 
 /SodaliteThe equivalents of chlorine and oxygen have 
 the ratio to each other of 35.5 : 8 :: 4.4376 : 1 . alumina con- 
 tains 46.8 per cent, and soda 25.8 per cent of oxygen. The 
 quantities of oxygen therefore corresponding to 
 
 7.10 chlorine 1.60 oxygen 
 
 38.40 silicic acid 20.47 
 
 32.04 alumina 14.99 
 
 24.75 soda 6.38 
 
 The quantity of sodium in combination with chlorine cor- 
 responds to 1.6 of oxygen, and there remains 4.78 for that 
 quantity of oxygen which goes to the formation of soda. The 
 ratios 4.78 : 14.99 : 20.47 = 1 : 3.1 : 4.5 , which are nearly 
 = 1 : 3 : 4, the silicate is NaO + A1 2 O 3 -f 2 SO 2 , and since 
 1.60 : 4.78 :: 1 : 3, we see that 1 equivalent of chloride of so- 
 dium is in combination with 3 equivalents of the silicate. 
 
 Observation. Also here we see quite distinctly, that the 
 analysis has given too much silicic acid, so much indeed as 
 to raise the question whether the proportions of oxygen, in- 
 stead of being as 1 : 3 : 4, might not be as 1 : 3 : 4.5 = 1 : 
 6 : 9 = 4 : 12 : 18, corresponding to the formula 4 NaO-f 
 4 A] 2 O 3 -f 9 SiO 2 , if this were not in opposition to the re- 
 sults obtained by other analyses of sodalite. 
 
 With the sulphur salts this mode of computation is the 
 more important inasmuch as the element common to both the 
 acid and the base, namely sulphur, is determined directly, 
 and the computed quantity will approximate the more near- 
 ly to that which is found, according as the analysis is more 
 correct. 
 
 Ruby Silver. If we assume, that antimony is represent- 
 ed by SbS 3 , and silver by AgS, and that the former contains 
 28.5, the latter 12.9 per cent of sulphur, we have : 
 71.5 : 28.5:: 23.26 : x x = 9.27 
 87.1 : 12.9:: 58.96 : y y = 8.73 
 
 18.00 
 
 The quantity of sulphur found was 17.78 per cent ; if there- 
 fore the two metals had been estimated correctly, the quan- 
 
212 
 
 tity found would be too little by only 0.22 per cent. The 
 proportions of sulphur, required by antimony and silver, and 
 in simple relation to each other, are as 1.06 : 1, evidently 
 therefore as 1 : 1 ; and since 1 equivalent of sulphate of anti- 
 mony contains 3 equivalents of sulphur, and 1 equivalent of 
 sulphide of silver contains 1 equivalent of sulphur, it follows 
 that the combination consists of 1 equivalent of sulphide of 
 antimony and 3 equivalents of sulphide of silver, correspon- 
 ding to the formula 3 Ag S + Sb S 8 . 
 
 Cryolite.. The analysis by Berzelius of this substance, 
 which consists of fluorine, alumininium and sodium, gave : 
 Sodium 32.93 
 
 Aluminium 13.00 
 
 Fluorine (loss) 54.07 
 100.00 
 
 Assuming that the two fluorides correspond to the known 
 oxides of the metals, the fluoride of sodium will consist of 1 
 eq: of sodium and 1 of fluorine, the fluoride of aluminium 
 of 2 eq: of aluminium and 3 equivalents of fluorine. Since 
 the eq : of ]STa=23, of Al=13.65 and of Fl=19 we have : 
 23.0 : 19:: 32.93 : x x = 27.20 
 
 27.3 : 57:: 13.00 : y y = 27.14 
 
 54.34 
 
 The quantities of fluorine are evidently equal,, therefore 
 cryolite consists of 3 equivalents of fluoride of sodium in com- 
 bination with 1 equivalent of fluoride of aluminum, corres- 
 ponding to 
 
 This formula numerically represented gives : 
 3 equivalents of sodium = 69 = 32.81 = ISTaFl 59.91 
 2 aluminium = 27.3 = 12.98 = A1 2 F1 3 49.99 
 6 fluorine =114.0 = 54.21 = 
 
 210.3 100.00 
 
 Computation of isomorphous Mixtures. An isomorphons 
 mixture consists of two or several isomorphous compounds, 
 which as a rule are analogously combined and in general hold 
 
213 
 
 in common either one or the other more proximate constitu- 
 ent. The computation of its formula is two-fold: general 
 and special. 
 
 The General Computation regards it as a single compound, 
 in which one part of the one constituent is, as it were, sub- 
 stituted by an equivalent quantity of an isomorphous body. 
 The respective numbers constituting 100 are divided by their 
 corresponding chemical equivalents, the quotients of the i 
 somorphous bodies are added together and, the sum is com 
 pared with the remaining quotients. 
 
 Smaltine. A specimen of this substance analyzed by 
 Eiechelsdorf gave : 
 
 Chem : Equiv : Quotients. 
 
 72 64 
 Arsenic 72.64 75 0.969 
 
 20 74- 
 
 Nickel 20.74 29 0.715 
 
 29 
 
 o q/r 
 
 Cobalt 3.37 39 0.112 o.<>4;t 
 
 o9 
 
 q OK 
 
 Iron 3.25 28 0.116 
 
 100.00 
 
 If we denote the electro-positive metals nickel, cobalt and 
 Iron by R, the sum of their equivalents has a ratio to that of 
 arsenic as 1 : 1.03, which is evidently=l : 1 ; the general for- 
 mula for the mixture, II As, tells us that each of the three iso- 
 morphous bodies contained in the mineral consists of one equi- 
 valent of the two constituents, that is, NiAs, Co As and FeAs. 
 
 The Special Computation of an isomorphous mixture aims 
 to determine the number of equivalents of the separate iso- 
 morphous compounds, in order to evolve from this the spe- 
 cial formula. It consists in eliciting the ratio in which the 
 separate quotients of the isomorphous bodies stand with re- 
 gard to each other, a ratio which in truth is always definite, 
 but not always as simple as that of the components in a real 
 compound. 
 
214 
 
 In thepreceding example the equivalents of iron, cobalt and 
 nickel have the ratio to each other of 11 6 : 112 : 715 : = 
 1:1: 6.4. Let us assume the ratio to be 1 : 1 : 6, that is r 
 one equivalent of iron, one of cobalt and six of nickel, or 
 that the mixture consists of one equivalent of arsenide of iron* 
 one of arsenide of cobalt and six of arsenide of nickel. In; 
 this case the special formula will be : 
 
 6 Ni As + Co As + Fe As 
 which numerically gives: 
 
 8 equivalents of arsenic = 600 = 72.11 
 6 nickel = 174 = 20.92 
 
 1 cobalt = 30 = 3.60 
 
 1 iron = 28 = 3.37 
 
 ~832~ 10000 
 
 In such a special computation we are frequently in the ha- 
 bit of representing the mixture as a single compound, for in- 
 stance of the most predominant constituent, of reducing the 
 formula and computation to 1 equivalent of the -same, and of 
 specifying the ratio in 'which the isornorphous constituents 
 rftand in regard to each other. 
 
 In the case just given, of the 8 equivalents 6 equivalent* 
 6. s. n i c kel, |- cobalt and iron. The compound is there- 
 fore represented as arsenide of nickel, KiAs, in which -J- of 
 the nickel is displaced by cobalt, and |by iron; the formula 
 then stands : 
 
 f Ni \ 0.75 Ni , 
 
 iCoVAs 0.125 Co [As 
 
 I Fe J 0.215 Fe ' 
 
 and by computation and reduction we have ; 
 
 1 equivalent of arsenic = 75 = 72.11 
 f nickel = 21.75 = 20.92 
 
 | cobalt = 3.75 = 3.60 
 
 -J- iron = 3.5 = 3.37 
 
 104.00 100.00 
 Naturally 104-J of 832. 
 ff the form nl ji of isoinorphoiis mixtures are written in this 
 
215 
 
 abbreviated way, the different symbols of the isomorphous 
 constituents are regarded altogether as one equivalent (unless 
 there should be a number behind them). 
 
 If in an isomorphous mixture one or more of the isomor 
 phous bodies should exist in very minute quantities when 
 compared with the principal component, these small quanti- 
 ties are converted, in the computation of the analysis, into 
 an equivalent quantity of the principal constituent, and con- 
 sequently the mixture itself into a compound of this consti- 
 tuent, and from this the equivalents of the two constituents 
 are computed. 
 
 Since the equivalents of iron, nickel and cobalt are in the 
 ratio of 28 : 29 : 30, we have: 
 
 30 : 29:: 3.37 : x x = 3.26 nickel. 
 
 28 : 29:: 3.25 : y y = 3.37 
 
 6.63 
 
 The whole, regarded as pure arsenide of nickel, would there 
 fore become : 
 
 Quotients 
 
 Arsenic 72.64 ^^ = 0.969 
 
 75 
 
 Nickel 27.37 ^ = 0.944 
 
 2/u 
 
 which, since each constituent is present as a single equiva- 
 lent, will be represented by the formula NiAs ; and, since 
 6.63 is nearly equal to of 27.37, of the nickel is dis- 
 placed by cobalt and iron, whose quantities again are nearly 
 equal to each other. 
 
 Zinc Blende. A specimen of black zinc from South Am- 
 erica gave : 
 
 Chem : Equiv : Quotients. 
 
 qq /rq 
 
 Sulphur 33.73 16 ^lr = 2.108 
 
 16 
 
 Zinc 51.95 32.5 5 -l^ = 1.6 
 
 Iron 14.32 28 <-. 0.51 
 
 28 
 
216 
 
 Since 2.108 : 2.11::1 : 1, the whole mixture = RS, R be- 
 ing regarded as = ZnFe. 
 
 Again since 9.51 : 1.6:: nearly 1 : 3, that is, there are 
 
 present 1 equivalent of iron and three equivalents of zinc ; 
 the whole is an isomorphous mixture of 1 equivalent of sul- 
 phide of iron and 3 equivalents of sulphide of zinc. 
 
 which numerically gives : 
 
 4 equivalents of sulphur = 64 
 3 zinc = 96.5 
 
 1 iron = 28 
 
 1881 
 
 The same result will naturally be obtained by computing 
 the quantity of sulphur required by each of the two metal& 
 in order to form RS. Thus 
 
 32.5 : 16:: 51.95 : x x = 25.58 sulphur 
 
 28.0 : 16 : : 14.32 : y y = 8.18 
 
 33.76 
 
 The two amounts of sulphur have the ratio nearly of 3:1- 
 From the preceding it follows, that the special formula of 
 this zinc blende may be written : 
 
 | Zn , 0.76 Zn , 
 
 iFeT 0.25Fe| S 
 
 Ghaiybite. In a specimen, a variety of this mineral, from 
 Siegen were found : 
 
 Chem : Equiv : Quotients. 
 
 Carbonic acid 39.19 22 1.98 
 
 Protoxide of iron 4T.96 36 L'33 \ 
 
 manganese 9.50 35 0.27 1 1.75 
 
 Magnesia 3.12 20 0.15 ' 
 
 99/77 
 
 Since 1.75 : 1.78::1 : 1, chalybite is a mixture of isomor- 
 phous carboiuit.es consisting of one equivalent of the base and 
 one equivalent of the acid, RO CO 3 . and since the equiva 
 lents of magnesia, protoxide of manganese and protoxide of 
 iron are pretty nearly as 1 .: 2 : 9, the mixture would bo 
 
217 
 denoted by the formula : 
 
 9 FeO CO 2 +2 MnO CO 2 +MgCCO 2 
 which when computed gives : 
 
 12 equivalents of carbonic acid = 264 = 
 9 protoxide of iron 324 = 
 
 2 manganese = 70 = 
 
 1 of magnesia = 20 = 
 
 "678" 
 
 Observation. According to the analysis the atomic rela- 
 tion of magnesia, the protoxide of iron and of manganese is 
 properly as 15 : 27 : 1335 : 9 : 45, and we should be ob- 
 liged to regard the formula as : 
 
 45 FeO CO 2 +9 MnO CO, + 5MgO CO 2 
 numerically as : 
 
 59 equivalents of carbonic acid = 1298 = 38.94 
 
 45 protoxide of iron = 1620 = 48.69 
 
 9 manganese = 315 = 9.45 
 
 5 magnesia = 100 = 3.01 
 
 3333 100.00 
 
 If the computation is compared, in accordance with both 
 formulas, with the numbers found, the differences for the 
 two formulas will be as follows : 
 
 First formula. Second formula. 
 Carbonic acid - 0.25 - 0.25 
 
 Protoxide of iron -0.17 + 0.64 
 
 Protoxide of manganese - - 0.82 4- 0.05 
 
 Magnesia - 0.17 + 0.11 
 
 Whilst the two formulas are almost equally valid for car- 
 bonic acid and magnesia, the second is less favorable for iron 
 from the fact that it corresponds better with the quantity of 
 manganese found. Therefore the second formula has no dis- 
 tinct advantage over the first ; consequently here, as in all 
 similar cases, the simpler ratio will be preferred. 
 
 The computation by means of the quantities of oxygen 
 will furnish the following results, if 
 
218 
 
 11 carbonic acid = 8 oxygen 
 
 9 protoxide of iron = 2 
 
 35 protoxide of manganese = 8 
 5 magnesia = 2 
 
 then 
 
 39.19 parts of carbonic acid contain 28.50 of oxygen. 
 47.96 protoxide of iron 10.66 \ 
 
 9.50 manganese 2.17 - 14.08 
 
 3.12 magnesia 1.25 ) 
 
 The oxygen of the bases and that of the acid = 1:2; the 
 mixture consequently = HO CO 2 . 
 
 Furthermore the oxygen of the magnesia, of the protoxide 
 of manganese and of the protoxide of iron = 1 : 1.7 : 8.5 = 
 1.2 : 2 : 10 = 1 : 1.8 : 9, or nearly = 1:2:9. 
 
 Olivine. Olivine from meteorites, commonly called Pal- 
 las meteorite, contains according to Berzelius : 
 
 Chem : Eq : Quotients. 
 
 Silicic acid 40.86 30 1.36 
 
 Magnesia 47.35 20 3.37 
 
 9 7i) 
 Protoxide of iron 11.72 36 0.33 c 
 
 99.93 
 
 The equivalents of the acid and of the bases are as 1 : 
 2 ; olivine is therefore an isomorphous mixture of subsilica- 
 tes, 2RO, SiO 2 , that is, since 0.33 : 2.37 ::1 : 7, of 1 equiva- 
 lent of the silicate of the protoxide of iron and 7 equivalents 
 of the silicate of magnesia, thus : 
 
 7(2 MgO,Si0 2 )+2 FeO,SiO 2 . 
 Computed numerically : 
 
 8 equivalents of silicic acid = 240 = 40.54 
 14 magnesia = 280 = 47.30 
 
 2 protoxide of iron = 72 = 12.16 
 
 592 100.00 
 
 The computation by means of the oxygen ratios is : 
 15 silicic acid = 8 oxygen 
 
 5 magnesia = 2 
 
 9 protoxide of iron = 2 
 
219 
 
 Therefore we have the quantities of oxygen in 
 Silicic acid 21.79 
 
 Magnesia 18.94 \ 
 
 Protoxide of iron 2.60 ) 
 
 They are consequently equal in the acid and the bases, so 
 that the mixture is denoted by the formula 2 EO,SiO 2 ; and 
 since 2.60 : 18.94=1 : 7.3, that is nearly = 1:7, the pre- 
 ceding formula is corroborated. 
 
 Garnet. A variety of garnet from Sala is constituted ac- 
 cording to Bredberg as follows : 
 
 Chem : Eq : Quotients. 
 Silicic acid 36.62 30 1.220 
 
 Alumina 7.53 51.3 0.147 1 
 
 Protoxide of iron 22.18 80 0.277) 
 
 Lime 31.80 28 1.136 |h 1-658 
 
 Magnesia 1.95 20 0.098 f 
 
 100.08 
 
 The sum of the equivalents of the bases, and that of the a- 
 id have the ratio to each other of 1 : 1.36=3 : 4.08. If we 
 assume this ratio to be as 3 : 4, it is evident that garnet is 
 an isomorphous mixture containing three equivalents of silicic 
 acid for every four equivalents of the bases. But the latter 
 are different in their conditions, two of them, for instance a- 
 lumina and the oxide of iron, being isomorphous as sesquiox- 
 ides (RaOa), whilst lime and magnesia are isomorphous as 
 protoxides (EO). The mixture, therefore, is constituted 
 of isomorphous double silicates of the two kinds of bases. 
 Since now the equivalents E 2 O 3 andEO have the ratio of 0.424 
 : 1.234=1 : 3, the mixture is a compound of 3 equivalents 
 of EO, 1 of E 2 O 3 and 3 of SiO 2 : 
 
 From the special formula we derive as follows: the equiva- 
 lents of alumina and the sesquioxide of iron = 0.147 : 0.277 
 = 1 : 1.9, which we must assume as 1 : 2. Those of mag- 
 nesia and lime = 0.098 : 1.136 = 1 : 11*6, for which we as- 
 
220 
 
 sume 1 : 12. According to this view -J R^s = alumina, f 
 = sesquioxide of iron ; ^ of 3 RO = magnesia, ||- = lime, 
 from which numerically we have the following : 
 
 3 equivalents of silicic acid = 90 = 37.11 
 
 ^ alumina = 17.1 = 7.05 
 
 f sesquioxide of iron = 53.3 = 21.98 
 
 ff lime = 77.5 = 31.96 
 
 -fg magnesia = 4.6 = 1.90 
 
 242.5 100.00 
 The special formula would be : 
 
 i Al A 
 I FeA 
 
 The following are used as a basis in the computation by 
 means of the proportions of oxygen : 
 
 15 silicic acid = 8 oxygen. 
 
 100 alumina = 46.8 
 
 10 sesquioxide of iron =3 
 
 7 lime . = 2 
 
 4 magnesia =2 
 
 From which we receive the following proportionate quan- 
 tities of oxygen : 
 
 36.62 silicic acid = 
 
 7.53 alumina = 
 
 22.18 sesquioxide of iron = 
 
 31.80 lime = 
 
 1.95 magnesia = 0.78 
 
 From this computation it follows, that the oxygen of the 
 protoxides and of the sesquioxides has the ratio of 1 : 1, al- 
 so that the oxygen of the bases together, and that of the acid 
 have the same ratio of 1 : 1, which again leads to the same 
 formula as before : 
 
 3RO + RA-f3SiO 2 
 
 Furthermore 3.52 : 6.65=1 : 1.9 and 0.78 : 9.09=1 : 11.6 
 for which we may substitute 1 : 2 and 1 : 12. 
 
 Grey Copper. In a variety of this mineral from Wolfach 
 II. Rose found the following : 
 
221 
 
 Chem: Eq : 
 
 Quotients. 
 
 Sulphur 
 
 23.52 
 
 16 
 
 1.147 
 
 Amtimony 
 
 26.63 
 
 120.3 
 
 0.221 
 
 Copper 
 
 25.23 
 
 63.4* 
 
 0.398 
 
 Silver 
 
 17.71 
 
 108 
 
 0.164 
 
 Iron 
 
 3.72 
 
 28 
 
 0,133 
 
 Zinc 
 
 3.10 
 
 32.5 
 
 0.095 
 
 99.91 
 
 The sum of the equivalents of all the metals, and that of 
 the sulphur have the ratio to each other of 1.011 : 1.47=1 : 
 1.45=9 : 13. 
 
 If we denote the electro-positive metals, whose sulphides 
 contain. 1 equivalent of sulphur, with R, the equivalents of 
 antimony and E, will stand in the ratio of 0.22 : 0.79=2 : 7, 
 or those of antimony, R and sulphur will have the ratio of 
 2 : 7 : 13 ; and since 7 equivalents of antimony require 6 of 
 sulphur, the formula for grey copper would be : 
 
 7RS + 2SbS 3 
 
 The chemical equivalents of Zn : Fe : Ag : Cu 2 furthermore 
 have the ratio to each other as 1 : 1.4 : 1.7 : 4.2. Let us as- 
 sume for them the following ratio of 1 : 1J : If : 4J-=6 : 9 
 : 10 : 25, and then compute them numerically : 
 
 13 equivalents of sulphur = 208.0 = 23.52 
 2 antimony = 240.6 = 27.20 
 
 7 x |f = J copper = 221.9 = 25.09 
 
 7 x ff = f silver = 151.2 = 17.10 
 
 7xA=ff iron = 35.3= 4.00 
 
 7 x -fa = fj. zinc = 27.3 = 3.09 
 
 884.3 100.00 
 
 The special formula would have to be written : 
 0.50 Cu 2 ] 
 
 0.2 Agi S + 2SbS 8 
 0.18 Fe f 
 0.12 Zn j 
 
 *Since the sulphide of copper, CU2S, in this and similar minerals is isomorphoue 
 with the remaining RS, 2 equivalents of copper must be employed'in the calculation. 
 
222 
 
 The second mode of computation assumes the following 
 data : 
 
 71.5 antimony = 100 sulphide of antimony, SbS 3 , 
 79.80 copper = 100 sulphide of copper, Cu 2 S. 
 87.1 silver = 100 sulphide of silver, AgS. 
 
 7 iron = 11 sulphide of iron, FeS. 
 
 67 zinc = 1.00 sulphide of zinc, ZnS. 
 
 These require therefore : 
 
 26.63 antimony = 10.62 sulphur 
 
 25.23 copper 6.36 j 
 
 17.71 silver 2.62 j 
 
 3.72 iron 2.13 j 
 
 3.10 zinc 1.53 
 
 23.26 sulphur. 
 
 Now since 23.52 parts of sulphur were found, the differ- 
 ence is + 0.26 per cent. 
 
 The numbers 12.64 and 10.62 must stand in a simple ratio 
 to each other having a whole number for the first, when the 
 latter 3, or a multiple of this number (since 1 equivalent of 
 sulphide of antimony contains 3 equivalents of sulphur). 
 They have the ratio of 3.5 : 3 = 7 : 6, that is, grey copper is 
 represented by the formula : 
 
 7RS-f2SbS 3 . 
 
 Furthermore the quantities of zinc, iron, silver and copper 
 have the ratio of 1 : 1.4 : 1.7 : 4.1, as was found before. 
 
 If we examine more intimately the results of the analysis 
 that have been here submitted to computation, it will be 
 observed, that these results do not furnish quite simple ra- 
 tios of composition, as we ought to expect. For the separate 
 isomorphous sulphur salts, which ^ enter into the formation 
 of grey copper, are so constituted as to exhibit 7 equivalents 
 of base with 2 equivalents of acid, or the amounts of sulphur 
 in the two as 7 : 6. The following formulas: 
 4 RS + SbS 3 and 3 RS + SbS 3 
 
 would undoubtly be much simpler, but they do not corres- 
 pond with the results of the analysis. For although the 
 
223 
 
 accurate analysis of a specimen of grey copper is fraught with 
 many difficulties, the name of the author of the analysis in 
 question is sufficient guarantee against any very serious error, 
 And the only question, that might possibly arise, would be 
 whether the far from simple ratio of 7 RS + 2 SbS s might not 
 be regarded as 
 
 (4 RS + SbS 3 )-f(3 ES + SbJSa), 
 or, 
 
 (6 RS + SbS 3 ) + 5 (3 KS + SbS 3 ) 
 
 that is, whether each separate member of the grey copper 
 mixture might not be looked upon as a double salt, which 
 would furnish very simple ratios between the base and the 
 acid in the latter case (2 : 1 and 1 : 1). But this would be 
 mere hypothesis. 
 
 Cases like this in the analysis of minerals are not rare. A 
 single analysis, even when it is quite accurate, leaves a doubt 
 as to the true composition of the mineral, because of the possi- 
 bility that the mineral was not quite pure, fresh and unchan- 
 ged. It is thus absolutely necessary to make a number of 
 analyses of the varied specimens of a mineral as found in dif- 
 ferent localities, and to comqare the results with one another. 
 This leads to the theory of the Constitution of chemical com- 
 pounds, the determination of their atomic weights and their 
 rational formulas, a branch of chemistry not comprehended 
 in the aim of the present work, but belonging to the pro- 
 vince of general theoretic chemistry. 
 
224 
 
 TABLES 
 for the 
 
 COMPUTATION OF ANALYSES. 
 
 
 Chem: Eq: 
 
 
 ALUMINIUM 
 
 Al=13.65 
 
 
 Alumina 
 
 A1 2 O 3 =51.3 
 
 46.8 pCt of oxygen 
 
 ANTIMONY 
 
 Sb=120.3 
 
 
 Antimonious acid 
 
 2 SbO 4 = 304.6 
 
 79.0 antimony 
 
 Sulphide of antimony 
 
 SbO 3 = 168.3 
 
 71.5 antimony 
 28.5 sulphur 
 
 ARSENIC 
 
 As =75 
 
 
 Arseiiious acid 
 
 AsO 5 =115 
 
 34.78 oxygen 
 
 Arsenate of ammonia 
 
 2MgO,AsO 5 + 
 
 60.53 arsenic acid 
 
 and magnesia 
 
 NH 3 +2aql90 
 
 39.47 arsenic 
 
 
 
 60.98 arsenic 
 
 Sulphide of arsenic 
 
 AsS 3 =123 
 
 39.02 sulphur 
 
 BARIUM 
 
 Ba 68.55 
 
 
 Baryta 
 
 BaO=76.55 
 
 10.45 oxygen 
 
 Sulphate of baryta 
 
 BaSO 4 =116.55 
 
 65.68 baryta 
 
 Silico-fluoride of 
 
 BaFl-fSiFl s 
 
 
 barium 
 
 =139.55 
 
 54.85 ,, baryta 
 
 BORON 
 
 B=ll 
 
 
 Boracic acid 
 
 B0 3 =35 
 
 68.57 ,. oxygen 
 
 Boro-fluoride of potas m 
 
 KF1 + BF1126 
 
 27,78 boracicacid 
 
 BROMINE 
 
 Br=80 
 
 
 Bromide of silver 
 
 AgBr=198 
 
 41.55 oxygen 
 
 Caesium 
 
 08=123.3 
 
 
 Oxide of caesium 
 
 OsO=131.3 
 
 6.09 oxygen 
 
 CALCIUM 
 
 Ca=20 
 
 
 Lime 
 
 OaO=28 
 
 28.57 oxygen 
 
225 
 
 
 Chem: Eq: 
 
 per cent 
 
 Carbonate of lime 
 
 CaO,CO 2 = 50 
 
 56 of lime 
 
 Sulphate of lime 
 
 CaO,SO 3 =68 
 
 41.18 lime 
 
 
 
 (7 lime =2 oxygen) 
 
 BISMUTH 
 
 Bi=104 
 
 
 Oxide of bismuth 
 
 Bi 2 O 3 =232 
 
 10.34 oxygen 
 
 Basic chloride of 
 
 Bi 2 Cl 3 + 2Bi 2 
 
 6.16 oxygen 
 
 bismuth 
 
 O 3 =778.5 
 
 80.15 bismuth 
 
 
 
 89.4 oxide of do 
 
 CADMIUM 
 
 Cd=56 
 
 
 Oxide of cadmium 
 
 CdO=64 
 
 12.5 oxygen 
 
 
 
 (8 oxide of cadmium 
 
 
 
 =1 oxygen) 
 
 Sulphide of cadmium 
 
 CdS=72 
 
 22.2 sulphur 
 
 
 
 (9 sulphide of cadmi- 
 
 
 
 um =2 sulphur) 
 
 CARBON 
 
 C=6 
 
 
 Carbonic acid 
 
 C0 2 =22 
 
 72.72 oxygen 
 
 
 
 (11 carbonic acid=i 
 
 
 
 8 oxygen) 
 
 Oxalic acid 
 
 C 2 O 3 =36 
 
 66.66 oxygen 
 
 
 
 (3 oxalic acid=2 ox n ) 
 
 Carbonate of lime 
 
 CaO,CO 2 = 50 
 
 44 carbonic acid 
 
 CERIUM 
 
 Ce=46 
 
 
 Oxide of cerium 
 
 CeO=54 
 
 14.8 oxygen 
 
 CHLORINE 
 
 Cl=35.5 
 
 
 Chloride of silver 
 
 AgCl= '143,5 
 
 24-74,, chlorine 
 
 CHROMIUM 
 
 Cr=26 
 
 
 Oxide of chromium 
 
 Cr 2 O 3 :=76 
 
 31.58 oxygen 
 
 
 
 131.6 chromic acid 
 
 Chromic acid 
 
 CrO 3 =50 
 
 48 oxygen 
 
 Chromate of the oxide 
 
 PbO,Cr0 3 = 
 
 30.96 'chromic acid 
 
 of lead 
 
 161.5 
 
 23.53 chrome 
 
 COBALT 
 
 Co=30 
 
 
 Oxide of cobalt 
 
 CoO=:38 
 
 21.05 oxygen 
 
 COPPER 
 
 Cu=31.7 
 
 
 Suboxide of copper 
 
 Ou jl O=71.4 
 
 11.2 oxygen 
 
 Oxide of copper 
 
 CuO=39.7 
 
 20.14 oxygen 
 
 Subsulphide of copper 
 
 Cu 3 S=79.4 
 
 79.86 copper 
 
 29 
 
226 
 
 
 Chem: Eq: 
 
 per cent 
 
 Subsulphide of copper 
 
 
 20.14 sulphur 
 
 
 
 100. oxide of copper 
 
 DlDYHlIUM 
 
 Di=48 
 
 
 Oxide of didymium 
 
 DiO=56 
 
 14.28 oxygen 
 
 
 
 (7ox e ofdid m =loxy D 
 
 FLUORINE 
 
 Fl=19 
 
 
 Fluoride of calcium 
 
 CaFl=39 
 
 48.72 fluorine 
 
 GLUCTNUM 
 
 G=7 
 
 
 Oxide of glucinum 
 
 G 2 3 =38 
 
 63.19 oxygen 
 
 GOLD 
 
 Au=196 
 
 
 HYDROGEN 
 
 H=l 
 
 
 Water 
 
 H0=9 
 
 88.89 oxygen 
 
 
 
 (9 water =8 oxygen 
 
 Ammonia 
 
 NH 8 =17 
 
 32'35 nitrogen 
 
 
 
 153 oxide of am 
 
 
 
 monium 
 
 Oxide of ammonium 
 
 AmO=r26 
 
 30.77 oxygen 
 
 Chloride of ammonium 
 
 AmCl=53.5 
 
 26.17 nitrogen 
 
 
 
 31.77 ammonia 
 
 
 
 48.6 oxide of am- 
 
 
 
 monium 
 
 IODINE 
 
 1=127 
 
 
 Iodide of silver 
 
 Agl=235 
 
 54.04 iodine 
 
 Iodide of palladium 
 
 Pdl=180 
 
 70.55 
 
 IRIDIUM 
 
 Ir=99(?) 
 
 
 IRON 
 
 Fe=28 
 
 
 Protoxide of iron 
 
 FeO=36 
 
 (9 protoxide of iron= 
 
 
 
 2 oxygen) 
 
 Sesquioxide of iron 
 
 Fe 2 O 3 =80 
 
 30 oxygen 
 
 
 
 ( sesquioxide of iron 
 
 
 
 =3 oxygen) 
 
 
 
 90 protoxide of iron 
 
 
 
 (10 sesquioxide of 
 
 
 
 iron =9 protoxide) 
 
 Sulphide of iron 
 
 FeS=.44 
 
 36.36 sulphur 
 
 
 
 (11 sulphide of iron 
 
 
 
 =4 of sulphur) 
 
 LANTHANUM 
 
 La=46.5-47 (?) 
 
 
 LEAD 
 
 Pb=103.5 
 
 
227 
 
 
 Chem: Eq: 
 
 per cent 
 
 Oxide of lead 
 
 PbO=111.5 
 
 7.175 oxygen 
 
 Sulphide of lead 
 
 PbS = 119.5 
 
 86.6 lead 
 
 
 
 13.4 sulphur 
 
 Sulphate of the oxide 
 
 PbO,SO 3 = 
 
 73.6 oxide of lead 
 
 of lead 
 
 151.5 
 
 68.32 lead 
 
 Chloride of lead 
 
 PbCl=139 
 
 74.46 lead 
 
 LITHIUM 
 
 Li=7 
 
 
 Lithia 
 
 LiO=15 
 
 53.33 oxygen 
 
 
 
 (15 lithia=8 oxygen) 
 
 Chloride of lithium 
 
 Li Cl =42.5 
 
 35.29 lithia 
 
 MAGNESIUM 
 
 Mg=12 
 
 
 Magnesia 
 
 MgO=20 
 
 40 oxygen 
 (5 magnesia = 2 oxy n ) 
 
 Pyrohosphate of 
 
 Mg 3 OP,0 5 = 
 
 36.03 magnesia 
 
 magnesia 
 
 111 
 
 
 MANGANESE 
 
 Mn=27 
 
 
 Protoxide of mangan 6 
 
 MnO=35 
 
 22.86 oxygen 
 
 Sesquioxide of do 
 
 Mn 2 O 3 =78 
 
 30.77 do 
 
 Hausrnannite 
 
 MnOMn 2 O 3 = 
 
 71.68 manganese 
 
 
 113 
 
 92.92 protoxide of 
 
 
 
 manganese 
 
 
 
 103.54 sesquioxide 
 
 
 
 of manganese 
 
 Binoxide of manganese 
 
 MnO 2 =43 
 
 62.8 manganese 
 
 MERCURY 
 
 Hg=100 
 
 
 Oxide of mercury 
 
 HgO=108 
 
 7.4 oxygen 
 
 Sulphide of mercury 
 
 HgS=116 
 
 86.2 mercury 
 
 
 
 13.8 sulphur 
 
 Chloride of mercury 
 
 Hg 2 Cl=235.5 
 
 84.92 mercury 
 
 MOLYBDENUM 
 
 Mo =46 
 
 
 Molybdic acid 
 
 Mo0 3 =70 
 
 34.28 oxygen 
 
 NICKEL 
 
 Ni=29 
 
 
 Oxide of nickel 
 
 NiO=37 
 
 31.62 do 
 
 NIOBIUM 
 
 Nb=49 (?) 
 
 
 Nitrogen 
 
 N=14 
 
 
 Nitric acid 
 
 NO 5 =54 
 
 74.07 do 
 
 OSMIUM 
 
 Os=100 (?) 
 
 
 OXYGEN 
 
 0=8 
 
 
228 
 
 
 Chem : Eq : 
 
 per cent 
 
 PALLADIUM 
 
 Pd=53 
 
 
 PHOSPHORUS 
 
 P=31 
 
 
 Phosphoric acid 
 
 PO 5 = 71 
 
 56.34 oxygen 
 
 Pyrophosphate of 
 
 Mg 2 0,P0 5 = 
 
 63.97 phosphoric 
 
 magnesia 
 
 111 
 
 acid 
 
 Phosphate of lime 
 
 3Cao,PO 5 =155 
 
 45.80 do 
 
 PLATINUM 
 
 Pt=99 
 
 
 Potassio-chloride of 
 
 KCl+PtCl 2 = 
 
 40.49 platinum 
 
 platinum 
 
 244.5 
 
 
 Ammonio-chloride of 
 
 AmCl + PtCl 2 
 
 6.26 nitrogen 
 
 platinum 
 
 = 223.5 
 
 7.6 ammonia 
 
 
 
 11.63 oxide of am- 
 
 
 
 monium 
 
 For 100 of platinum 
 
 
 14.14 nitrogen 
 
 
 
 17.17 ammonia 
 
 
 
 26.26 oxide of am- 
 
 
 
 monium 
 
 POTASSIUM 
 
 K=39 
 
 
 Potassa 
 
 K0=47 
 
 W oxygen 
 
 Chlorid e of potassium 
 
 KC1=74.5 
 
 52.35 potassium 
 
 
 
 63.1 potassa 
 
 Sulphate of potassa 
 
 KO,S0 3 =87 
 
 54 potassa 
 
 Potassio-chloride of 
 
 KCl+PtCl 2 = 
 
 19.22 potassa 
 
 platinum 
 
 244.5 
 
 30.47 chloride of 
 
 
 
 Dotassium 
 
 RHODIUM 
 
 Rh=52 
 
 
 RuBiDium 
 
 Rb=85.36 
 
 
 Rnbidia 
 
 RbO=93.36 
 
 8.57 oxygen 
 
 RuTHENIUm 
 
 u=52 (?) 
 
 
 SELENIUM 
 
 Se=39.5 
 
 
 SILVER 
 
 Ag=108 
 
 ( 
 
 Oxide of silver 
 
 AgO=116 
 
 6.9 do 
 
 SODIUM 
 
 STa=23 
 
 
 Soda 
 
 N"aO=31 
 
 25.8 do 
 
 Chloride of sodium 
 
 ISTaCl=58. 
 
 39.32 sodium 
 
 
 
 53 soda 
 
 Sulphate of soda 
 
 ]STaOSO 3 =71 
 
 43.66 soda 
 
 STRONTIUM 
 
 3r=44 
 
 
 Strontia 
 
 SrO=52 
 
 5.4 oxygen 
 
229 
 
 
 Chem : Eq : | per cent 
 
 Sulphate of strontia 
 SULPHUR 
 
 SrOSO 3 =92 
 8=16 
 
 56.52 strontia 
 
 Sulphuric acid 
 Sulphate of baryta 
 
 SO 3 =40 
 
 BaO,SO 3 = 
 116.55 
 
 60 oxygen 
 (5 sulphuric acid 3 
 oxygen) 
 34.32 sulphuric 
 acid 
 
 TANTALUM 
 TELLURIUM 
 
 Ta=(l) 
 
 Te=64 
 
 13.73 sulphur 
 
 THALLIUM 
 
 () 
 
 
 THORIUM 
 
 Th=59.16 
 
 
 TIN 
 
 Sn=58.8 
 
 
 Stannic acid 
 TITANIUM 
 
 SnO s =74.8 
 Ti=24 
 
 21.4 oxygen 
 
 Titanic acid 
 
 TiO 2 =40 
 
 40 do 
 (5 titanic acid 2 
 
 TUNGSTEN 
 
 W=92 
 
 oxygen 
 
 Tungstic acid 
 URANIUM 
 
 WO 8 =116 
 U=60 
 
 20.7 oxyge 
 
 VANADIUM 
 
 V=68.5 
 
 
 YTTRIUM 
 ZINC 
 
 Y=35(?) 
 Zn=32.5 
 
 
 Oxide of zinc 
 Sulphide of zinc 
 ZIRCONIUM 
 
 Zr=45 
 
 19.75 do 
 33 sulphur 
 
 Zirconic acid 
 
 ZrO 2 =61 
 
 26.23 oxygen 
 
230 
 
 INDEX 
 
 A, 
 
 Acidimetry, 42. 
 Alkalimetry, 42. 
 Alloys, 47. 
 Alum, 113. 
 Alumina 7 113. 
 Ammonia, 113. 
 Analcime, 152. 
 Analysis by measure, 40. 
 Analysis by weight, 71. 
 Antimony, 61, 
 Apatite, 117. 
 Argentan, 54, 
 Arsenic, 92. 
 Axinite, 163. 
 
 B. 
 
 Basalt, 186. 
 Baryta, 111. 
 Bell metal. 57. 
 Berthierite, 86. 
 Bismuth, 55. 
 Bitterspar, 131. 
 Black oxide of manganese. 
 Black vitriol, 115. 
 Blende, 85. 
 Block tin, 59. 
 Blue iron ore, 122. 
 Bog iron ore, 65. 
 Bone black, 119. 
 Boracic acid, 137. 
 Borates, 135. 
 Boronatrocalcit % 1 37. 
 Boulangerite, tvi. 
 Bournonite, 92. 
 
 Brass, 51. 
 
 Brittle silver ore, 91 . 
 
 Bronze, 57. 
 
 Brown hematite, 65. 
 
 Brown lead ore, 121. 
 
 Bunsen's analysis, 71 . 
 
 C. 
 
 Cadmium, 60. 
 Calamine, 159. 
 Cannon metal, 57. 
 Carbonates, 128. 
 Celestine, 111. 
 Chabasite, 152. 
 Childrenite, 127. 
 Chlorides, 93. 
 
 ,, of ammonium, 174 r 
 ., of calcium, 117. 
 of lead, 121. 
 Chromates, 193. 
 Chrome iron ore, 192. 
 
 orange, red, yellow, 193. 
 mica, 163. 
 69. ochre, 162. 
 Chrysoberyl, 162. 
 Chrysoprase, 162. 
 Clay, 65. 
 Cobalt glance, 99. 
 Compounds of fluorine, 198. 
 of titanium, 196. 
 Computation of analyses, 202. 
 Construction of formulas, 199.. 
 Copper, 49, &c. 
 
 glance, 83. 
 nickel. 99. 
 
Copper pyrites, 83. 
 slags, 158. 
 stone, 95. 
 Corundum. 190. 
 Oyolite, 199. 
 
 D. 
 Dolomite, 131. 
 
 F. 
 
 False gold leaf, 51. 
 Feather ore, 91 
 Fluorine, 170. 
 Foil, 59. 
 Forge slags, 68. 
 Franklinite, 190. 
 Fusible alloys, 59. 
 
 G. 
 
 German silver, 54. 
 Green lead ore, 121. 
 Grey antimony, 86. 
 Grey copper, 92. 
 
 H. 
 
 Hausmannite, 69. 
 Hauyne, 160. 
 Heavy spar, 110. 
 Hematite, 65. 
 Heulandite, 152. 
 
 I. 
 Iron, 64, &c. 
 
 J. 
 Jamesonite, 91. 
 
 K 
 Kalait, 127. 
 
 L. 
 
 Lamp black, 119. 
 Lapis lazuli, 160. 
 Lead, 55, 56, 57, 59, fcc, 
 Leucopyrite, 95. 
 Lime, 111, &c. 
 Limonite, 65. 
 Litharge, 75. 
 
 231 
 
 Lithia, 123. 
 
 M. 
 
 Magnesia, 131, &c. 
 . Magnetite, 83. 
 Magnoferrite, 190. 
 Manganese, 69. 
 Mesotype, 152. 
 Mercury, 61. 
 Miargyrite, 91, 
 Mispickel, 98. 
 
 N. 
 Nickel, 54, 99, 103. 
 
 O. 
 Oxides, 64, of lead, 75. 
 
 P. 
 
 Packfong, 54. 
 Phosphates, 116, &c. 
 Phosphoric acid, 126, 
 Pinchbeck, 51. 
 Potassa, 113. 
 Psilomelane, 73. 
 Pyrolusite, 69. 
 
 K. 
 
 Red lead ore, 193. 
 Ruby silver, 91, 209. 
 
 S. 
 
 Silicates, 138. 
 Silver coins, 49. 
 Similor, 51. 
 Soda, 123. 
 Soft solder, 59. 
 Spathic iron, 134, 
 Spear pyrites, 83. 
 Speise, 103. 
 Spinel, 190. 
 Stassfurthite, 138, 
 Steel slags, 151. 
 Stilbite, 152. 
 Sulphides, 87, &c. 
 Sulphur metals, 78, 79, <fce. 
 
s. 
 
 Thomsonite, 152. 
 Tin, 57, 59, 60, &e. 
 Titanic iron,. 196. 
 Tiza, 137. 
 Tourmaline, 173. 
 Triphyline, 123. 
 Type metal, 61. 
 U. 
 [Jltramarine, 160. 
 
 232 
 
 V, 
 
 Vitriol, 115. 
 Yivianite, 122; 
 
 W. 
 
 Wavellite, 127. 
 White nickel, 99, 
 White lead, 135. 
 
 Z. 
 
 Zinkenite, 91. 
 Zinc, 92, 94, 159, 
 
THIS BOOK IS DUE ON THE LAST DATE 
 STAMPED BELOW 
 
 AN INITIAL FINE OF 25 CENTS 
 
 WILL BE ASSESSED FOR FAILURE TO RETURN 
 THIS BOOK ON THE DATE DUE. THE PENALTY 
 WILL INCREASE o SO CENTS ON THE FOURTH 
 DAY AND TO $1.00 ON THE SEVENTH DAY 
 OVERDUE. ' 
 
 9 1938 
 
 JUN 2 8 2003 
 
 -ftOfr 
 
 
 
 
 J1)L6 
 
 LD 21-95rn-7,'37