← •
| º， ，
º ： * ,•••••• • • • . … . . . ."
* *ſ º
ſ ºvº ***■ ■
s ſ●Ō，
„ ， ， ， , ， ， ， ， , ， • • • •
، ،
≡ º tº º ºs• • • • • • • • • •
：！
- ：
• • • • i•ſjſaeſae·
º.º.º.º.º.' "ſ-^ , ^ '^^^ ºººººoºo ， el s a f ， ！• O • • • • • C • № C • • • •
··n-~... • • •：··
@，|－•
• º £~· · · * *º ：· -
• • • • • • • • • • • • • • ••••••••••••••••¿•º• • • №r:„…！！！！… ••*******�. …. . . . … -- ~~~~ … ~~~~ ~ ~ ~~~~ ~~~~ ~~~~--~~~~ ~~~************
· ：• ſ：،·C∞■， ！،a º º
* . . . . . . • • • • • • • • • • • •••••••••••••• • • *~*, 's'| - • • • • • • • •••••• • • • • • • • • •n.•， ！. . . . . .
∞
g * *
„…..……. …..…..…-, …--~~~~ ~~~~ ~~~~=== ••• • • • • • •******** * *
wae~∞e ！∞• • •： e • • • • • •
C-
• • • • • • ** **~~ ~~~~ ： ~~~~ ~~~~~ ~= <！--
ſj
e- →→→ • • • • • • • • •
·∞|－
f
, ， ， ， ， º º
• • • • •
<?！!！
º „ºſae:
∞ . .：، ، ،■
Cº.“, º „ºs · · · ·:·º·:·º.º.ººº.
[]** º * *
ºo： ***
vºº.º.）?. -
\, ,aeſae），§）,
ſſſsſſſſſſſſſſſſſ
ºiſſſſſae
， ， … → • • • • • •••••• • • • • • • • • • • • • •
º • • • • • • • • • • • • • • • • • • • • • • • • • •
：' .* ;
#########
{{#####
[×..
*





EI
H!!!!Nº.
EITTTTTTº ||
. . . . .
:
IIBRARYºº of THE
tºvºltiºn
f zººx- § . . . -
-: , ºr -
- *- *
NX
s
---
-
*
!
§
* * ~6. - w º
*! A. SN …” §
- * Sºś *: ... "
- tºº, - }.” '. -
*y . . . . º !}3:] t} y Tº º --- -
º, , º, .N
-- ... . . . . Y. |: | *: - . . . . .
! |\ si §º !
º’ -º, -., - * “º \\
ºr ſ *. - . --
\ *---
º
#: @"\\ \º:
fºx.
- - - . . º SS-
+
4
22.
;?
, t
- * *n
-
-
:
s
–
É
E
f-- --
l l ;
-- ITUT. T. CTD ſº C. CITAT C C C C & I
|-vſºn ºf . | * R - , ,
!!!...] till . . . . . . 1: i-l

The D. Van Nostrand Company
intend this beek to be sold to the Public
at the advertised price, and supply it to
the Trade on terrns which will not allow
of discount.
--—---8,

G.W.
"15
L^0%
E5
ao?
TECHNICAL METHODS OF
CHEMICAL ANALYSIS
WORKS BY GEORGE LUNGE, Ph.D., Dr.Ing.
Technical Methods of Chemical Analysis. English 7 ranslation
by CHAs. ALEX. KEANE, D.Sc., PH.D.
Vol. I., in Two Parts, royal 8vo, Io24 pp., not sold separately. Just
published Price £2, 12s. 6d.
Vols. II. and III. (each Volume sold separately). In active preparation.
Technical Chemists' Handbook. Tables and Methods of Analysis
for Manufacturers of Inorganic Chemical Products, being a
thoroughly revised and extended edition of Alkali Makers' Hand-
book, indispensable for the Technical Chemist. Just published,
crown 8vo, 280 pp., bound in leather. Price 10s. 6d. net.
The Manufacture of Sulphuric Acid and Alkali. A Theoretical
and Practical Treatise.
Vol. I. Sulphuric Acid, in Two Parts, 1200 pp., not sold separately.
Third and much enlarged Edition. Price £2, 12s. 6d.
Vol. II. Salt Cake, Hydrochloric Acid, and Leblanc Soda, 900 pp. Second
Edition, Revised and Enlarged. Price £2, 2s.
Vol. III. The Ammonia-Soda and various other Processes of Alkali-
Making, 800 pp. Second Edition, Revised and Enlarged.
Price £2, 2s.
Coal Tar and Ammonia. Third and enlarged Edition, 900 pp.
Price £2, 2s.
s”, A Fourth Edition, much enlarged, and brought up
to date. In active preparation.
Handbook of Technical Gas-Analysis. By CLEMENT WINKLER, ,
Ph.D. Translated by Professor LUNGE. Second English Edition.
Price 10s. 6d. met.
T E CHNICAL METHODS


CHEMICAL ANALYSIS
EDITED BY GEORGE LUNGE, Ph.D., DR.ING.
EMERITUS PROFESSOR OF TECHNICAL CHEMISTRY, FEDERAL POLYTECHNIC SCHOOL, ZURICH
IN COLLABORATION WITH
E. ADAM. F. FISCHER. W. JETTEL. J. PASSLER.
F. BARNSTEIN. F. FRANK. H. KöHLER. O. PFEIFFER.
T. BECKERT. H. FREUDENBERG. P. KREILING. O. PUFAHL.
O. BOTTCHER. E. GILDEMEISTER. K. B. LEEIMANN. H. RASCH.
C. COUNCLER. R. GNEHM. J. LEWROWITSCH. O. SCHLUTTIG.
K. DIETERICH. O. GUTTMANN. C. J. LINTNER. C. SCHOCH.
K. DüMMLER. E. HASELHOFF. E. O. v. LIPPMANN. G. SCHÚLE.
A. EBERTZ. W. HERZBERG. E. MARCKWALD. L. TIETJENS.
C. v. ECKENBRECHER. D. HOLDE. J. MESSNER. K. WINDISCH.
L. W. WINKLER.
ENGLISH TRA NSLATION
FROM THE LATEST GERMAN EDITION, ADAPTED TO ENGLISH
CONDITIONS OF MANUFACTURE
EDITED BY
CHARLES ALEXANDER KEANE, D.Sc., PH.D.
PRINCIPAL AND HEAD OF THE CHEMISTRY DEPARTMENT
THE SIR JOHN CASS TECHNICAL INSTITUTE, LoNDoN
IN COLLABORATION WITH
T. L. BAILEY. J. T. CONROY. J. HÚBNER. * H. R. PROCTER.
C. O. BANNISTER. C. F. CROSS. W. J. LAMIBERT. H. J. L. RAWLINS.
E. J. BEWAN. G. J. FOWLEIR. J. LEWECOWITSCH. W. F. REID.
W. BURTON. A. G. GREEN. A. R. LING. IP. SCHIDROWITZ.
W. A. CASPAIRI. O. GUTTMANN. C. A. MITCHELL. A. SMETHAM.
H. G. COLMAN. A. D. HALL. F. B. POWER. W. THOMASON.
V O L U M E | . – P A R T |.
N E W Y O R K
D. VAN NOSTRAND COMPANY
23 MURRAY AND 27 WARREN STREETS
I908
Printed in Great Britain.
PRE FACE
A revised edition of Dr F. Böckmann's Chemisch-technische Unter-
suchungsmethoden, edited by Professor G. Lunge, was published in 1899,
and very soon became recognised as a standard work on the subject.
A considerable time had elapsed since the last edition by Dr Böckmann
was issued, and the interval had been one of rapid progress in all
branches of chemical industry. Professor Lunge's revised version
appeared therefore as practically a new book, extended in material and
in scope, and adapted to modern methods of work. In 1904 a second
edition was called for, in which the whole of the subject matter was
thoroughly revised and brought up to date.
The plan of the book consists in the treatment of the technical
methods of analysis, applicable to specific industries, in separate sections;
all the important inorganic and organic industries are dealt with, and
each section has been written by a contributor fully and practically
acquainted with the subject. The chief methods of analysis are given
in all cases, but full details are described only for those methods which
have proved good and reliable in the hands of the writers of the various
sections, or which are otherwise recognised as efficient. So far as
possible, each section is similarly arranged, the raw materials being
treated of first, then intermediate products and methods of control in
working, and then final products. A general introduction on analytical
methods is included in Volume I., in which the more common operations
and apparatus are dealt with.
The English translation has been made from the second German
edition.
As the industrial processes, methods of analysis, and legislative
conditions are, in many respects, different in this country from those
obtaining abroad, almost all the sections have been submitted to experts
for revision, so that the English translation has thus been thoroughly
adapted to English conditions of manufacture, by specialists fully
familiar with the methods of work in use in this country. The contents
of each section have been brought well up to date by this revision, and
special care has been taken to include references to English work and
literature.
W
l94.950
vi PREFACE
This revision has been made as complete as possible, without
materially extending the subject matter of the several sections. In a
few instances portions of the German text have been omitted, owing
to their being inapplicable to the conditions of work here; in such cases
the omission is noted in the translation, and a reference is given to the
original text. Such further additions as were necessary to bring the
contents of the book up to date, since the publication of the second
German edition, have been added to the English translation by Prof.
Lunge and by the Editor.
The book will be issued in three volumes, and a full index will be
provided with each volume. A bibliography of books of reference is
appended to each section, and all important tables, in addition to being
printed in the text, are also printed for reference at the end of each
volume.
With the exception of those cases in which empirical factors are
employed in technical work, all the numerical data are calculated from
the table of atomic weights for 1908, issued by the International
Committee, with O = 16 as the basis. The numerical data for gases and
for the weights of substances to be taken for analysis so as to corre-
spond to definite volumes of gases, are all calculated from the real litre
weights according to the most reliable determinations, not from the
calculated values. All temperatures are given in centigrade degrees,
except where otherwise stated.
The Editor desires to record his thanks to Dr W. E. Adeney for the
description of his apparatus for the extraction and analysis of the gases
dissolved in water, and for analytical data of the dissolved gases in water,
included in the section on “Drinking Water and Water Supplies”; to
Dr R. T. Glazebrook, F.R.S., Director of the National Physical
Laboratory, for permission to include the methods and conditions
adopted by the National Physical Laboratory for the testing and
calibration of glass vessels, included in the section on “General Methods
used in Technical Analysis”; to Dr W. H. Sodeau for the description of
his apparatus for gas analysis, and for kindly reading the section on
“Technical Gas Analysis”; and to the Engineering Standards Com-
mittee for their permission to include the data from their “British
Standard Specification for Portland Cement,” and to reproduce the
illustrations of the forms of apparatus recommended, in the section on
“Calcareous Cements.” The Editor is also indebted, for the loan of
blocks, to Messrs Brady & Martin, Newcastle; Messrs Griffin & Sons,
London; and to Dr H. Seger and E. Cramer of Berlin.
CHARLES A. KEAN.E.
LONDON, August 1908.
LIST OF CONTRIBUTORS TO THE GERMAN
EDITION, AND OF REVISERS OF THE
ENGLISH TRANSLATION IN VOL. I.
General Methods used in Technical Analysis,
By Prof. G. LUNGE, Ph.D., Dr.Ing., Zürich.
Revised by C. A. KEANE, D.Sc., Ph.D., Principal and Head of the Chemistry
Department, The Sir John Cass Technical Institute, London.
Technical Gas Analysis.
By Prof. F. FischER, Ph.D., Göttingen.
Revised by C. A. KEANE, D.Sc., Ph.D., Principal and Head of the Chemistry
Department, The Sir John Cass Technical Institute, London.
Fuel Analysis.
By Prof. F. FISCHER, Ph.D., Göttingen.
Revised by T. L. BAILEY, Ph.D., Glassford Creek, Queensland, Australia;
formerly Lecturer on Metallurgy, The University, Liverpool.
Sulphurous Acid, Nitric Acid, and Sulphuric Acid.
By Prof. G. LUNGE, Ph.D., Dr.Ing., Zürich.
Translated by J. T. Conroy, B.Sc., Ph.D., The United Alkali Co., Liverpool. .
Saltcake and Hydrochloric Acid.
By Prof. G. LUNGE, Ph.D., Dr.Ing., Zürich.
Translated by J. T. Conroy, B.Sc., Ph.D., The United Alkali Co., Liverpool.
Sodium Carbonate.
By Prof. G. LUNGE, Ph.D., Dr.Ing., Zürich.
Translated by J. T. Conroy, B.Sc., Ph.D., The United Alkali Co., Liverpool.
The Chlorine Industry.
By Prof. G. LUNGE, Ph.D., Dr.Ing., Zürich.
Translated by J. T. Conroy, B.Sc., Ph.D., The United Alkali Co., Liverpool.
Potassium Salt8.
By L. TIETJENs, Ph.D., Chief Chemist to the United German Potash
Syndicate, Leopoldshall.
wii
viii LIST OF CONTRIBUTORS
Cyanogen Compounds.
By H. FREUDENBERG, Ph.D., Frankfurt.
Revised by J. T. Conroy, B.Sc., Ph.D., The United Alkali Co., Liverpool.
Clay.
By P. KREILING, Berlin.
Revised by W. BURTON, Director of Pilkington's Tile and Pottery Co.,
Clayton, Lancashire.
Clay Wares, Harthenware, and Glazes.
By K. DüMMLER, Berlin.
Revised by W. BURTON, Director of Pilkington's Tile and Pottery Co.,
Clayton, Lancashire.
Aluminium Salts and Alumina.
By Prof. G. LUNGE, Ph.D., Dr.Ing, Zürich.
Glass.
By Prof. E. ADAM, The Art-Trade School of the Royal Austrian Museum of
Art and Industry, Vienna.
Revised by W. THOMASON, Messrs Doulton & Co., Lambeth Pottery, London.
Calcareous Cements.
By C. ScHocH, Ph.D., The Royal Technical High School, Berlin.
Revised by W. F. REID, London.
Drinking Water and Water Supplies.
By Prof. L. W. WINKLER, Ph.D., Budapest.
Revised by G. J. FowlER, D.Sc., Lecturer in Bacteriology in the Public
Health Department, The University, Manchester.
Feed Water for Boilers and Water for other Technical Purposes.
By Prof. G. LUNGE, Ph.D., Dr. Ing, Zürich.
Sewage and Eiffluents.
By E. HASELHoFF, Ph.D., Director of the Agricultural Experiment Station,
and Lecturer at the University, Marburg.
Revised by G. J. FowlFR, D.Sc., Lecturer in Bacteriology in the Public
Health Department, The University, Manchester.
Soils.
By E. HASELHOFF, Ph.D., Director of the Agricultural Experiment Station,
and Lecturer at the University, Marburg.
Revised by A. D. HALL, M.A., Director of the Rothamsted Experiment Station.
Air.
By Prof. K. B. LEHMANN, Ph.D., Würzburg.
Revised by C. A. KEANE, D.Sc., Ph.D., Principal and Head of the Chemistry
Department, The Sir John Cass Technical Institute, London.
LIST OF ABBREVIATED TITLES OF JOURNALS
Journal.
American Chemical Journal . e e º
Analyst . e e e e g ſº
Annalen der Chemie . { } ſº o tº
Annalen der Physik . ſº e
Annales de Chimie et de Physik
Archiv der Pharmacie tº e g
American Journal of Science . e e tº * *
Berichte der deutschen chemischen Gesellschaft e tº
Biedermann's Centralblatt für Agricultur Chemie
Brewer's Journal ſº tº * ſº & gº
British and Colonial Druggist . § & ſº © wº
Bulletin de l'Association Chimique de Sucre et de Distillerie
Bulletin de la Société Chimique de Belgique
Bulletin de la Société Chimique de Paris . tº tº
Bulletin de la Société Industrielle du Nord de la France
Bulletin de la Société Industrielle de Mulhouse
Chemical News. e * e G tº
Chemical Trade Journal
Chemiker Zeitung . tº tº e º 'º º
Chemiker Zeitung Repertitorium . gº o
Chemische Industrie . g ſº º © e te
Chemische Revue tiber die Fett-und Harz-Industrie.
Chemisches Centralblatt . s & e e
Chemist and Druggist . . ſº tº ſº & ſº
Comptes-Rendus hebdomadaires des Séances de l'Acadamie
des Sciences . e e o
Dingler's polytechnisches Journal s
Electrician e ge tº gº © e &
Electrochemical and Metallurgical Industry º
Engineer . e ſº gº º tº tº ©
Engineering e tº º &
Engineering and Mining Journal º
Färber-Zeitung. te
Fischer's Jahresbericht e e e & * º
Gazzetta Chimica Italiana. . e e e G t
Gerber . tº dº e e
Gummi-Zeitung e dº
India-Rubber Journal tº e e tº ©
Journal de Pharmacie et de Chimie . tº & ©
Journal für Gasbeleuchtung und Wasserbesorgung .
Journal für praktische Chemie . tº ſº o
Journal of Gas Lighting .
Journal of Physical Chemistry . & e
Journal of the American Chemical Society
Journal of the Chemical Society . . . e e tº
Journal of the Chemical Society, Abstracts tº tº tº
Journal of the Chemical, Metallurgical, and Mining
Society of South Africa . . . . . . .
Journal of the Franklin Institute . e Q & g
ABBReviation.
Amer. Chem. J.
Analyst
Annalen
Ann. Physik
Ann. Chim. Phys.
Arch. Pharm.
Amer. J. Sci.
Ber.
Biedermann's Centr.
Brewer's J.
Brit. and Col. Drug.
Bull. Assoc. Chim. Sucr.
Bull. Soc. Chim. Belg.
Bull. Soc. Chim.
Bull. Soc. Ind. Nord
Bull. Soc. Ind. Mulhouse
Chem. News
Chem. Trade J.
Chem. Zeit.
Chem. Zeit. Rep.
Chem. Ind.
Chem. Rev. Fett-Ind.
Chem. Centr.
Chem. and Drug.
Comptes rend.
Dingl. polyt. J.
Electrician
Electrochem. Ind.
Engineer
Engineering
Eng. and Min. J.
Färber-Zeit.
Fischer's Jahresber.
Gazz. chim. ital.
Gerber
Gummi-Zeit.
India-rubber J.
J. Pharm. Chim.
J. Gasbeleucht.
. prakt. Chem.
. Gas Lighting
. Phys. Chem.
. Amer. Chem. Soc.
. Chem. Soc.
. Chem. Soc. Abstr.
J. Chem. Met. Soc., S. Africa
J. Franklin Inst.
J
J
J
J
J
J
ix
X LIST OF ABBREVIATED TITLES OF JOURNALS
JoURNALs.
Journal of the Institute of Brewing . e e ſº
Journal of the Institution of Mechanical Engineers .
Journal of the Society of Arts . gº º
Journal of the Society of Chemical lndustry
Journal of the Society of Dyers and Colourists.
Leather Trades Review . e & © e e t
Mittheilungen aus dem königlichen Materialprüfungsamt
zu Gross-Lichterfelde West ge e * tº
Mittheilungen aus der Centralstelle für wissenschaftlich-
technische Untersuchungen gº o tº e o
Mittheilungen des technischen Gewerbemuseums in Wien.
Monatshefte für Chemie der kaiserlichen Akadamie der
Wissenschaften, Wein e e e * gº
Moniteur Scientifique º tº © tº e e tº
Paper and Pulp.
Papier-Zeitung .
Petroleum Review tº †
Pharmaceutical Journal . ſº
Pharmaceutisches Centralblatt . º e
Philosophical Magazine and Journal of Science
Philosophical Transactions of the Royal Society . .
Proceedings of the American Electrochemical Society
Proceedings of the American Institute of Mining Engineers,
and also Bulletin tº e tº tº © º e
Proceedings of the Faraday Society . . . . .
Proceedings of the Institution of &vil Engineers e
Proceedings of the Institution of Mining and Metallurgy .
Proceedings of the Royal Society . . G tº ©
Revue Générale des Matieres Colorantes . g tº
Receuil des travaux chimiques des Pays-Bas et de la
Belgique . º e g ſº * º
Scientific American . &
Stahl und Eisen
Tonindustrie Zeitung ſº º tº º & tº *
United States Consular Reports tº º gº e tº
Zeitschrift der analytischen Chemie .
Zeitschrift für angewandte Chemie
Zeitschriſt der anorganischen Chemie e g
Zeitschriſt des Vereins der deutschen Zucker-Industrie
Zeitschrift des Vereins für deutsche Ingenieure. e
Zeitschrift für chemische Apparatenkunde
Zeitschrift für das gesammte Brauwesen . *
Zeitschrift für Elektrochemie . tº gº
Zeitschrift für Farben-und Textil-Chemie.
Zeitschrift für physikalische Chemie . .
Zeitschrift für Spiritusindustrie . cº © sº e wº
Zeitschrift für Untersuchung der Nahrungs-und Genuss-
mittel e e Qº & gº e º e
Zeitschrift für Zuckerindustrie in Böhmen. tº
West hindian Bulletin e g *
Wochenschrift für Brauerei ſº
ABBREVIATIONs.
J. Inst. Brewing
J. Inst. Mech. Eng.
J. Soc. Arts
J. Soc. Chem. Ind.
J. Soc. Dyers and Col.
Leather Tr. Rev.
Mitt. k. Materialprüf.
Mitt. Centralst. Wiss.-tech.
Unters.
Mitt. techn.-Gew. Museums
Monatsh.
Monit. Scient.
Paper and Pulp
Papier-Zeit.
Petrol. Rev.
Pharm. J.
Pharm. Centr.
Phil. Mag.
Phil. Trans.
Proc. Amer. Electrochem. Soc.
Proc. Amer. Inst. Min. Eng. ;
Bull. Amer. Inst. Min. Eng.
Proc. Faraday Soc.
Proc. Inst. Civ. Eng.
Proc. Inst. Min. and Met.
Roy. Soc. Proc.
Rev. Gen. Mat. Col.
Rec. trav. chim.
Scient. Amer.
Stahl. u. Eisen
Tonindustrie Zeit.
U.S. Cons. Reps.
Z. anal. Chem.
Z. angew. Chem.
Z. anorg. Chem.
Z. Ver. deut. Zuckerind.
Z. Verein. deutsch. Ingen.
Z. für chem. Apparatenkunde
Z. ges. Brauw.
Z. Elektrochem.
Z. Farb.-u. Text.-Chem.
Z. physik. Chem.
Z. Spiritusind.
Z. Unters. Nahr. u. Genussm
Z. Zuckerind. Böhm.
West Ind. Bull.
Woch. f. Brau.
CONT ENTS
PR EFACE º © * > º
PAGE
tº ſº V
LIST of CoNTRIBUTORs To THE GERMAN EDITION, AND of REVISERs of THE
ENGLISH TRANSLATION
ABBREVIATED TITLES OF JOURNALs
vii
P A R T I
GENERAL METHODS USED IN TECHNICAL ANALYSIS
PAGE
Introduction . tº e º © • 3
Historical development of technical
methods of chemical analysis. .. 3
Objects of technical methods of
chemical analysis . . . • 5
Classes of work comprised in techni-
cal methods of chemical analysis . 6
I. General Operations tº e e • 7
The taking of samples 7
Average samples tº º G ... 8
A. Materials in Large Pieces . • 9
Mechanical samplers . . IO
Filling of sample into bottles. I I
B. Raw Materials in the form of
Powder, Dross, etc. . o • I 2
C. Chemical Products in the form of
Powder . º { } º • I 3
Use of augers . tº ſº ſº • I 3
D. Liquids . o tº gº e • I4.
B. Gases . . . . . . 16
The collection and storage of samples 16
II. General Laboratory Operations . . 17
A. Grinding of Substances . . . I7
B. Weighing ſº ſº ſº ... • I9
Hand-scales . wº g • I9
The use of tares in weighing . . 20
Hase's balance . * . . 2C
Methods of weighing , 2I
PAGE
II. General Laboratory Operations—Continued.
C. Solution; Fusion; Evaporation . 22
Methods for preventing loss from
spirting . e ſº tº e
Funnels and other apparatus used in
Evaporation . tº o • 23
22
D. Precipitation, Washing, and Filtra-
tion of Precipitates . tº • 23
The settling of precipitates . . 23
Use of reagents of definite strength 24
Washing precipitates . se • 24
Withdrawal of clear solutions by
means of pipettes . O • 24
Filtration . tº gº ge • 25
The Gooch crucible © © • 25
The filtration of hot solutions . 26
Centrifugation of precipitates . 27
E. Drying and Ignition te . . 27
Ignition of wet precipitates . . 27
Weighing on dried filter paper . 28
Ignition of precipitates . . . 29
F. Heating Apparatus . g & . 29
Heating by gas . e ... • 29
Other methods of heating . . 31
G. Volumetric Analysis . . • 32
The calibration and standardisation
of apparatus used in volumetric
analysis . to tº tº . 32
The unit for standardisation; the
true litre . tº e © • 33
Mohr's litre . ſº e © • 34
xi
xii
CONTENTS
PAGE
G. Volumetric Analysis—Continued.
Standard temperature for calibration
Calibration tables for correcting for
temperature and pressure . .
Methods and conditions adopted by
the National Physical Laboratory
for the testing and calibration of
glass vessels . º
Limits of error •
Method of delivery in calibrating .
Calibration with the Ostwald pipette
Calibration of gas-burettes . G
Corrections for temperature . .
Apparatus for Volumetric Analysis .
Burettes tº Q ſº e e
The reading of burettes.
Floats . e & C e
Screens . & wº e g e
Cleaning measuring vessels
Burette-stands C. tº e ſº
Arrangements for filling burettes .
Overflow burettes .
Pipettes. tº º e
Vessels used for titration g
Quality of the glass employed in
volumetric analysis . gº o
Indicators for Acidimetry and Aléal-
imetry wº tº e g
Classification of indicators
Sensitiveness of indicators
Theory of indicators
Methyl orange
Litmus . © {º e is
Phenolphthalein . ſº & ſº
Table of the basicity of acids to-
wards different indicators .
Other indicators te *
Test papers . . tº º º
Normal Solutions tº &
General considerations . tº ge
Standard acids . © e &
Substances used as a basis for Acid-
imetry and Alkalimetry –
Sodium carbonate . tº gº
Sodium oxalate . º gº o
Preparation of normal hydro-
chloric acid . º g
Other substances used as a basis for
acidimetry . tº * tº e
Gravimetric methods for fixing the
strength of normal hydrochloric
acid . tº te ©
Use of oxalic acid .
Potassium tetroxalate
Potassium bi-iodate e tº o
The strength of normal acids. .
35
35
36
4 I
4 I
42
44
45
46
46
47
47
49
5O
5O
5.I
52
53
54
54
55
55
57
58
6I
67
7I
75
75
79
8o
8o
82
83
85
86
88
90
9I
92
93
93
Mormal Solutions—Continued.
The equivalent value of standard
PAGE
acids . ſº tº sº o 94.
Normal oxalic acid * 94.
Standard alkalis . g e • 95
Sodium hydroxide . . . . 96
Potassium hydroxide . de • 99
Barium hydroxide. tº gº • 99
The equivalent value of standard
alkalis o º ſe ſe • 99
Potassium Permanganate . e . 99
Preparation of a standard potassium
permanganate solution . . Io:3
Fixing the strength of the solu-
tion:—
Oxalic acid method . tº • IO3
Iron method . tº º • IOS
Ferrous ammonium sulphate
method . e wº wº . Io9
Hydrogen peroxide method . Io&
Gas-volumetric or nitrometer
method . e o . IoS
Volhard's iodometric method . Io9
Applications of potassium perman-
ganate solution . • II 2
Jodimetry . e e tº & ... II 2
Standard iodine solution tº . I I 3
Preparation of starch solution • II.4.
Standardisation of iodine solution . I 14
Applications of iodimetry . I I6
Sodium thiosulphate solution. . 117
Standardisation with pure iodine . 118
Other methods of standardisation . 12o
Arsenious Acid. e * º • I 22
Silver Mitrate and Ammonium Zhio-
cyanate . e t & & • I23
General Remarks on Volumetric
Analysis e º . I24
A. Gas-volumetric Analysis. . I 25
Zhe Azotometer. e e . I 25
Dietrich's table for the absorption
of nitrogen e e e . I28
Dietrich's table for the weight of
I c.c. of nitrogen • I29
Zhe Wºtrometer . tº º o . I 32
Manipulation wº g • I 34
The nitrometer for the analysis of
saltpetre and other nitrates . I36
The decomposition bottle . • I 37
Reduction of volume of gas to
normal volume . e wº . I 38
Applications of the nitrometer . I38
Zhe Gas-Volumeter . e o . I38
Adjustment and manipulation • I 39
Clamps . . tº • • - I 42
CONTENTS
xiii
PAGE
The Gas-Volumeter—Continued.
Gas-volumeter, with attached de-
composition vessel . © • I43
Graduation to give percentage
directly . © e tº e
Use of specific weights of sub-
stance to give percentage results
directly . e e ſº . I46
Other forms of the gas-volumeter . 148
Apparatus for the Determination of
Carbon Dioxide . © ſº . I48
Lunge and Marchlewski's apparatus 149
Lunge and Rittener's method. . 153
The Universal Gas-Volumeter . • I54
General Remarks on Measuring
Vessels for Gases . tº e . I 55
I45
J. Determination of Specific Gravity . 156
General considerations . ſº . I56
Fleischer's densineter . . I56
Twaddell's hydrometer . {e . I57
Baumé's hydrometer . © . I57
Gerlach's scale . & ge . I 57
The “rational” Baumé hydrometer 158
Comparison of various Baumé
hydrometers, for heavy liquids,
with specific gravities. e . I 59
Table for the conversion of specific
e - e. I5°. gº
gravities at # into “rational"
Baumé degrees . . º . I60
Cartier's hydrometer . . . I62
Beck's hydrometer. . . . I62
Baumé's hydrometer for liquids
lighter than water . . . I62
Comparison of various hydrom-
eters for liquids lighter than
Water . © e tº o . I62
Differential hydrometers e . I63
Volumemeters º e o . I63
The areo-pyknometer . º . I63
PAGE
Z. Determination of Specific Gravity—Contd.
Resolutions regarding hydrometers,
adopted by the International
Congress of Applied Chemistry. 164
A. Measurement of Pressure and of
Draught: Pressure Gauges and
Anemometers . e ſº . I65
Simple pressure gauges te . I65
Péclet's pressure gauge . . . 166
Seger's differential manometer . 167
König's differential manometer . 168
Langen-Lux manometer tº . I69
Calculation of the velocity of a
current of gas . º tº . I69
Péclet's differential anemometer . 17o
Fletcher-Lunge anemometer. . I 7O
Tables for the reduction of ane-
mometer readings to velocity of
Current ſº © © © . I7I
L. Measurement of Temperature . . I72
Thermometry tº º ſe . I73
Pyrometry . o © © • I74
High-temperature thermometers . 174
Expansion pyrometers . e . I75
Air pyrometers . tº g . IZ5
Fusion-point pyrometers g . I76
Optical pyrometers • . . I'77
Wanner's pyrometer . . . I77
Féry's radiation pyrometer . . I78
Calorimetric pyrometers . . I78
Electric resistance pyrometers ... I'79
Callendar's resistance pyrometer . 179
Thermo-electric pyrometers . . 180
Le Chatelier's pyrometer . . 181
M. Calculation of Analytical Results . I83
Use of tables. º ſº tº . I83
Limits in calculation . . . . I84
Atomic weights . º cº . I84
Formulation of analytical results . I86
Literature on Volumetric Analysis . . 186
SPECIAL METHODS OF TECHNICAL ANALYSIS
PAGE
Technical Gas Analysis.
FAGE
Apparatus for Gas Analysis–Continued.
Methods of sampling . . . I89
Automatic apparatus for the examination
of chimney gases . tº º • IQO
Non-analytical methods for testing chimney
gases • . . . . gº • I9 I
The collecting of samples . . . I91
Apparatus for storing samples of
gases . . . a tº • , 192
Apparatus for Gas Analysis . tº • I 94
Winkler's gas burette . e tº • I94
Bunte's gas burette . . . • I95
Honigmann's gas burette . ſº • IQ7
Fischer-Orsat apparatus . ſº . I98
Lunge's modification of the Orsat ap-
paratus . e º e te • 2CI
W. H. Sodeau's modification of the
Orsat apparatus o g º • 203
Hempel's apparatus . © e • 205
F. Fischer's apparatus for the analysis
of generator gas © tº 0 • 215
xiv CONTENTS
PAGE PAGE
4%aratus for Gas Analysis—Continued. Raw Materia's—Continued.
W. H. Sodeau's apparatus for accurate II. Sulphur from Gas Residuals . 270
work. e o o o tº º
Absorption coefficients of the commoner
2I9
gases in water . tº O e , 225
Zhe Calorific Value of Gaseous Fuel . . 226
Junkers' gas calorimeter . . . 226
Boys' gas calorimeter . . . . 230
Table for correcting the volume of gas
used to 15° C. and 760 mm. . 23.I
Fischer's gas calorimeter . . . 236
The loss of heat from chimney gases . 237
Determination of the total sulphur in
heating gases . & tº • , 239
Literature e {} o e e • 239
Fuel.
Sampling e gº e e • , 24.I
Determination of moisture º ſº • 242
Determination of ash e e • . 242
Residual coke or fixed carbon . • 244
Sulphur . . . . . • 245
Arsenic, phosphorus, nitrogen tº º
Determination of carbon and hydrogen
Zhe Determination of the Calorific Power of
Fuel . g e te g {}
Dulong's method of calculation . .
Fischer's calorimeter . C * e
The Berthelot bomb calorimeter . .
The Mahler bomb ſº tº e G
Parr's calorimeter. e e tº o
Lewis Thompson's calorimeter . .
W. Thomson's calorimeter . © º
Modifications of the Thomson calor-
247
. 248
25O
25O
25O
254
256
256
26o
262
imeter by Rosenhain and by Darling. 262
Comparative value of results obtained
by different calorimeters . . .
Literature tº o ſº tº O e
262
263
Manufacture of Sulphurous Acid, Nitric
Acid, and Sulphuric Acid.
Sulphurous Acid . . . . .
A’aw Materials :-
I. Sulphur (native sulphur) . .
Commercial varieties of Sicilian
sulphur . © G tº 0
Degree of fineness . . tº 0
Chancel’s sulphurimeter e G
Determination of:—
Ash, moisture, bituminous sub-
stances . tº tº o g
Arsenic, selenium e tº e
Sulphur . º g tº e
Specific gravity of solutions of sul-
phur in carbon bisulphide .
264
264
264
265
265
266
267
268
. 269
Determination of the available
sulphur . tº e º . 27O
III. Pyrites and other metallic Sul-
phides e © * wº • 272
Determination of:—
Moisture, sulphur . . 272
Wet extraction method . e . 272
Removal of the iron . ſº . 273
Precipitation of the barium sul-
phate. o e o ſº . 274
Dry extraction methods. . . 275
Estimation of available sulphur . 277
Volumetric methods for the deter-
mination of the sulphuric acid . 277
Titration with benzidine hydro-
chloride . gº g . . 280
Determination of:—
Arsenic . ſº o gº . 281
Copper . © tº tº . 285
Lead . º ſº tº . 286
Zinc . o tº § gº . 286
Carbonates of the alkaline earths 288
Carbon . C to e . 288
Distinction of ordinary from mag-
netic pyrites . o O . 289
IV. Zinc Blende . e . 289
Determination of :—
Total sulphur, zinc . wº . 289
Lead, calcium and magnesium,
arsenic, carbonic acid . • 292
Fluorine . º e º • 293
Available sulphur . . • 293
Control of Working Conditions e • 294.
I. Burnt Ore (residues after calcina-
tion) . ſº o * tº • 294
Crude or native sulphur g • 294
Gas residuals Q º • 294
Iron pyrites . . o • 294
Estimation of sulphur . • 294.
Copper, iron, zinc blende . . 298
II. Examination of the Kiln Gases . 299
Reich's method for the estimation
of sulphur dioxide . • , 3OO
Lunge's method for the estimation
of total acids . wº tº • 3O2
Examination by the specific gravity 304
III. Finished Products . © • 3O4.
Solutions of sulphurous acid . . 304
Liquid sulphur dioxide . . . 305
Sulphites . • • * - 305
Sulphite liquors for the manufacture
of wood-pulp . . . . .306
CONTENTS XV
PAGE
Nitric Acid Manufacture . . . .306
Chili Saltpetre . e Q e . 3C6
Commercial assay of saltpetre . . 3O7
Complete analysis, with the exception
of the determination of nitrate and
perchlorate . e e . . .308
The estimation of :—
Moisture, soluble matter, chlorine,
sulphuric acid, calcium, mag-
nesium, sodium carbonate . . 308
Potassium . e o º . 3O9
The estimation of nitrate :-
Summary of methods of estimation 309
Ulsch's method as modified by Böck-
Iſlall Il * o o tº • 3 II
Lunge's nitrometric method . . 315
The Schlösing - Grandeau - Wagner
method . e e e º . 3 I 7
Method by ignition with potassium
chromate or with silica . wº . 318
The estimation of perchlorate . • 3 I9
Estimation of sodium chlorate . • 32I
Control of Working Conditions o . 323
Examination of bisulphate º . 323
Milric Acid . º gº © e • 323
Properties o o e e • 323
Specific gravity e º ... • 323
Influence of hyponitric acid on the
specific gravity . . . . 324
Table of the specific gravity of nitric
acid . o o o • 325
Temperature correction for specific
gravity . . º . 327
Estimation of nitrogen peroxide (hypo-
nitric acid) . . . . . 328
Total acidity . e e e . 328
Tests for :—
Fixed residue o © e &
Sulphuric acid, chlorine, heavy
metals and alkaline earths, iron,
iodine © e 0 & • 33O
Examination of nitrating and spent acids.
(Mixtures of sulphuric acid, nitric
acid, etc.) . tº e º • 33I
Determination of:—
Total acidity, lower nitrogen acids,
total nitrogen acids, sulphuric
acid . © tº e º • 332
329
Sulphuric Acid Manufacture . . .333
Raw Materials . º e e • 333
Control of Working Conditions o • 333
Examination of the gases . º . 333
Inlet gases . . . . . .333
Chamber gases . . . . 334
Exit gases—Oxygen . . . 334
PAGE
Control of Working Conditions—Continued.
Examination for acids . O . 336
Nitric oxide . . . . . .338
Nitrous oxide e e º • 34O
Formula for calculation of loss of
sulphur. o º e º . 34O
Examination of the works' acids • 34 I
Nitrosity of the drips . • 34 I
Yield of chamber acid . º , 342
Examination of the Glover acid . 342
Examination of the Gay-Lussac
acid (nitrous vitriol) . o • 342
Table for the estimation of nitrous
acid in nitrous vitriol. º , 344
Total nitrogen acids . . . .345
Finished Product—Sulphuric Acid. . .346
Properties . e o º , 346
Specific gravity . o . . 347
Table for determining the specific
gravity from the percentage
Content . © e o . 348
Specific gravity of sulphuric acid
solutions according to Lunge,
Isler, and Naef . gº c . 350
Temperature correction for specific
gravity . . . • , 354
Influence of impurities on the
specific gravity. 0. . . .356
Melting points of sulphuric acid . 358
Boiling points of sulphuric acid . 358
Qualitative examination of sulphuric
acid for impurities . e • 359
General examination for gaseous
impurities . º • , 359
Sulphurous acid, hydrochloric
acid, nitrogen acids e • 359
Nitrous acid . º º . 360
Hydrofluoric acid, ammonia . 361
Solid impurities, iron, selenium . 362
Detection and approximate estimation
of arsenic . o o e . 362
The Marsh-Berzelius test . . 363
Electrolytic form of the Marsh-
Berzelius test . © º . 368
Other forms of the test . º • 37I
Reinsch's test e o G • 374
Gutzeit's test Q º • • 374
Bettendorf's test . . o . 376
Quantitative estimation of sulphuric
acid and of its impurities . . 377
Free sulphuric acid, sulphurous “.
acid ; nitrous acid, colorimetric *-
determination . . . • 377
Nitric acid, colorimetric deter-
mination . o e o • 379
Lead, iron . º o © • 38o
xvi CONTENTS
PAGE PAGE
Finished Product—Continued. D. Saltcake—Continued.
Colorimetric estimation of iron . 381 Sodium chloride . e • 404
Hydrochloric acid, arsenic . . .383 Iron, insoluble matter, calcium . . 405
Fuming sulphuric acid (anhydride, oleum) 385 Magnesium e Q º . 406
Properties . e . . . 385 Aluminium . º - . . 4O7
Melting points . . . . 385 Sodium sulphate . . . . 408
Boiling points º . . 385 | E. Hydrochloric Acid . . 4.09
Specific gravity . . . . 386 Control of working conditions . . 409
Quantitative analysis º . 387 Examination of exit gases. . . 409
Sampling . . . . . 387 Examination of the gases in the Har-
Weighing out of the sample—Clar greave's process . º • . 4II
and Gaier's bottle . . . 388 Properties of hydrochloric acid . . 4II
Winkler's tap-tube . . . 389 Specific gravity. . º • * 4 I2
Lunge and Rey's bulb-tap pipette. 390 Temperature correction for specific
Influence of sulphurous acid . . .391 gravity . . • . . . 413
Method employed at Freiberg . 392 Detection of impurities:—
Calculation of results . • , 393 Residue, sulphuric acid . • 4 I 3
Knietsch's table for calculation . 394 Arsenic. e º & . 4I4
Setlik's method of analysis . . 394 Iron, sulphurous acid . . 416
Grünhut's table . . . . .395 Chlorine, selenium . . 4I7
Rabe's method of analysis . . 396 Estimation of hydrochloric acid . 4 I?
Literature. © o e . . .396 || Literature o º º . 418
The Manufacture of Saltcake and of
Hydrochloric Acid.
A. Brine and Salt- Works . © e . 397
Specific gravity of brine . . 397
Analysis of brines and mother liquors 398
Determination of chlorine, sulphuric
acid, oxides of iron, aluminium, cal-
cium, and magnesium, bicarbonates
of ferrous iron, calcium and mag-
nesium . º © º e . 398
B. Sodium Chloride (Rock Salt) • , 399
Impurities tº c • 399
I. Ordinary Salt . • 399
Sampling . . o • 399
Qualitative analysis . o • 399
Quantitative analysis. . 400
Determination of :—
Moisture © º . 4OO
Combined water . • 40 I
Total chlorine . e . . 401
Sulphuric acid . • 40I
Insoluble matter, calcium, and mag-
nesium . e e e . 402
Calculation of results • 402
Direct determination of magnesium
chloride e O º o . 4O2
II. Pure Sodium Chloride for analytical
The Manufacture of Sodium Carbonate.
I. The Leblanc Soda Process .
• 419
Raw Materials . º e o • 4 IQ
I. Saltcake . o e e e • 419
2. Calcium Carbonate . e º . 42O
3. Mixing Slack . . & . . 42O
Determination of:—
Moisture º º º • 42O
Coke . o º o º • 42O
Ash © o º º . 42 I
Sulphur º tº tº e • 42 I
Nitrogen . Q e • 422
Control of Working Conditions o • 422
I. Black Ash O o - te • 422
Determination of:—
Free lime . o • 423
Total lime, total available alkali,
caustic soda o e • 424
Sodium sulphide, sodium chloride 425
Sodium sulphate . . 426
Carbonating test . . . 426
2. Wat Liquor . º e tº . 426
Specific gravity. . . . . .426
Chemical analysis . . . . .426
Determination of:—
Sodium ferrocyanide . • 427
Silica, alumina, and ferric oxide . 429
Carbonating test . e e
purposes tº 0 • . . 403
C. Sulphuric Acid . • . . 404
D. Saltcake . º e º • 404
Free acid . . e • 404
• 429
3. Carbonated Liquors. . c • 429
Determination of bicarbonate . • 429
4. Soda Mother Liquors • * 429
* CONTENTS .
xvii
PAGE
Control of Working Conditions—Continued.
Estimation of sulphide, sulphate,
sulphite, and thiosulphate, when
present together . 43O
5. Wat Waste & º o • 435
A. Non-oxidised alkali waste • 435
Determination of :—
Available soda, total soda, total
sulphur, oxidisable sulphur . 456
B. The Chance- Claus sulphur re-
covery process. . . . 437
Determination of :
Sulphide-sulphur, carbonic acid . 437
Soda, lime, thiosulphate . . 438
Lime-kiln gases. º º . 438
Gas from the gas-holder . . 438
Exit gases from the Claus kiln . 439
II. The Ammonia-Soda Process • 447
A’aw Materials . tº e e . 447
Control of Working Conditions e . 448
III. Manufacture of Caustic Soda . 440
A. Causticised Liquor . e e • 44 I
B. Fished Salts . e e g • 44 I
C. Caustic “Bottoms”. o tº • 44 I.
D. Lime Mud . e e º • 442
IV. Electrolytic Alkali Liquors. • 443
V. Crude Fused Soda of the Cellu-
lose Industry . . e tº • 443
Chemical analysis. . . . . 443
Calculation of results . . . • 445
VI. Finished Products of the Soda
Industry . . . . • 447
A. Soda Ash Q ſº © e • 447
Specific gravity of solution • 447
Temperature correction for specific
gravity . e © tº tº • 449
Available alkali. . . . • 45O
Table of alkalimetric degrees . • 452
General examination of soda ash :—
Specific gravity or density . . 453
Clearness of the aqueous solution . 454
Fineness, insoluble matter, sodium
chloride, sodium sulphate . • 455
The complete analysis of soda ash. 456
Physical and qualitative tests for
the lower grades of Leblanc ash. 459
Examination of chemically pure
sodium carbonate . o . 460
B. Soda Crystals . . . . . 462
C. Caustic Soda . . º º • 463
Specific gravity of solutions . . 463
PAGE
C. Caustic Soda—Continued.
Sampling . ſº & • * . 465
Analysis . . e e & . 466
Caustic salts . e e e . 467
D. Sodium Bicarbonate & º . 467
Properties. . ſº © º . 467
Qualitative analysis . . . º . 468
Quantitative analysis º . 468
Estimation of the total carbon dioxide 469
The Examination of Commercial Liquid
Carbon Dioxide e e © . 472
Literature . 475
The Chlorine Industry.
A’aw Materials. . © . 476
I. Manganese Ore . 476
Determination of moisture, active
Oxygen . o e G • 477
Carbon dioxide, hydrochloric acid
necessary for decomposition . 481
II. Limestone . te º . 481
Determination of insoluble matter,
lime, magnesia . . 482
Iron . tº e º . 483
III. Lime . º & e º . 483
A. Quick lime . & º . 483
Determination of free lime, carbon
dioxide . o e c . 483
A. Slaked lime. e . 484
Content of milk of lime in caustic
lime . G is o º . 484
Control of Working Conditions . . . 484
I. Manufacture of Chlorine by means of
native Manganese Ore . º . 484
II. The Weldon Process . . . 485
The examination of Weldon mud . 485
Determination of manganese dioxide. 485
Total manganese . © C. . 486
The “base” . e e to . 486
III. The Deacon Process º . 487
Ratio between the free chlorine and
the unchanged hydrochloric acid
gas º o e e • 487
Composition of gases before entering
the decomposer . e & . 490
Carbon dioxide. o ſº º • 49 I
Moisture . . • . . 495
IV. The manufacture of Bleaching
Powder . e tº 495
Determination of the chlorine in the
chamber atmosphere . . . 495
V. The manufacture of Potassium
Chlorate - © º o - 497
Determination of chlorate . • 497
Calcium chloride . © - - 498
Free chlorine and hydrochlorite . 498
xviii
CONTENTS
PAGE
Finished Products . & º • 499
I. Bleaching Powder • 499
Sampling . & & e . . 499
“Degrees” of available chlorine . 499
Specific gravity of bleach solutions . 5oo
Estimation of available chlorine:–
Gay-Lussac's arsenious acid
method . e © g . SOO
Graham's ferrous sulphate method 501
Bunsen's iodometric method . . 5OI
Bunsen's distillation method. . 5O2
Pontius' potassium iodide method. 503
Penot's arsenious acid method . 503
Lunge's gas-volumetric method . 504
PAGE
Finished Proaucts—Continued.
Vanino's table for the weight of I c.c.
of chlorine . º e o . 507
II. Bleaching Solutions and Electrolytic
Liquors. º e º e . 509
Determination of :—
Available chlorine . o e . 509
Free hypochlorous acid . . . 510
Chlorates . o e e . 5 II
Chloride-chlorine . ſº e • 5 I4
Carbonic acid © sº © . 514
Bases . e e & © . 516
Alkali hydroxide and carbonate . 516
III. Potassium Chlorate . º . 5 I?
IV. Liquid Chlorine . . . . 518
Vanino's hydrogen peroxide method 506 | Literature e Q tº e e . 519
P A R T I I
º PAGE PAGE
Potassium Salts. C. Manufactured Products—Continued.
General Methods for the Determination of 2. Potassium sulphate . e • 535
Aºotassium . º tº © e • 52O General considerations • 535
The potassium platinichloride method . 520
The perchloric acid method . . 522
Other methods . & e O • 524
Modifications of the platinichloride
method . e & e , - 525
Special Methods of Analysis © &
I. Stassfurt Salts. o tº
A. Crude Salts — Carnallite,
kieserite, kainite,
. 528
. 528
rock
Sylvinite,
“Hart-salz” . © tº . 528
Determination of potassium by the
platinum method e . 528
Determination by the perchloric
acid method . e . . 53O
Determination of magnesium
chloride, total magnesium, sul-
phuric acid e º e . 53I
Complete analysis of the crude
salts . e e e & • 53I
B. Intermediate Products of Manu-
facture e e e . 533
C. Manufactured Products . 533
I. Potassium chloride . . 533
Determination of potassium by
the platinum method . . 533
By the perchloric acid method . 533
Determination of:—
Sodium chloride . e . 534
Magnesium salts, water, in-
soluble matter . . . 535
Determination of:—
Sulphate e e . . .536
Potassium by the platinum
method . . . . 536
By the perchloric acid method 537
3. Potassium salt manures . • 537
II. Nitre . º e e º . 538
A. Raw Materials . e e . 538
I. Chili saltpetre . . 538
2. Potassium chloride . . 538
B. Intermediate Products. . 53O
C. Final Product—Nitre . º • 539
Determination of:—
Moisture, chlorine, insoluble
impurities, perchlorate, potas-
sium, chlorate, iodine. • 539
III. Potassium Carbonate . . . 540
A. Raw Materials º • 540
I. Potassium chloride . • , 540
2. Beet ashes o g • 540
B. Intermediate Products. • 543
Carbonated liquors . • 543
C. Final Products and By-products . 543
I. Potassium carbonate. © • 543
2. Impure potassium carbonate,
molasses carbonate, and beet
ash carbonate. . • 545
3. Hydrated potassium carbonate. 545
CONTENTS
xix
PAGE
Cyanogen Compounds.
I. Simple Cyanides . e o . 546
I. Gravimetric methods of analysis . 546
2. Volumetric methods of analysis:—
Liebig's method e e & • 547
Fordos and Gélis' method • 547
Hydrocyanic acid. . 548
Cherry laurel and bitter almond waters 548
Potassium cyanide e tº . 549
Estimation of cyanogen by Liebig's
method . iº gº * gº iº
The Fordos and Gélis method for the
estimation of cyanogen . tº £
55O
Sodium cyanide . • 554
Ammonium cyanide . º . 554
Cyanides of the alkaline earths 555
Mercuric cyanide . . 555
II. Double Cyanides . 555
I. Potassium ferrocyanide, yellow prus-
siate of potash . © . 555
Examination of “spent-oxide.” • 555
Determination of:—
Moisture . . 556
Prussian blue . 556
Knublauch's method . 557
Zulkowsky's method . 558
Nauss’ method 558
Moldenhauer and Leybold's method 560
Drehschmidt's method . . 560
Intermediate products of manufacture;
calcium ferrocyanide. . 56I
Commercial products . 562
Potassium ferrocyanide. . 562
Potassium ferricyanide . . 563
Sodium ferrocyanide . o . 563
2. Other cyanogen double salts . 563
III. Thiocyanates . gº . 564
Ammonium thiocyanate . 564
Potassium thiocyanate. , 567
Literature . 567
Clay.
General properties . 568
Physical Examination . 569
I. Mineral Mixtures and their State of
Division . * © g . 569
II. Plasticity e • . . . 57 I
III. Behaviour of Clays on Drying . 572
IV. Behaviour of Clays on Firing • 573
Evenness of colour 573
Colour • 574
Density • 575
Vitrification . ſº tº & • 575
Firing contraction . ſº . . 576
Refractoriness . . ſº . . 576
PAGE
Physical Examination—Continued.
Methods of determination :-
Bischof's method . . 577
Seger cones . e . 577
Kreiling's method. . 578
Chemical Analysis . e ſº © . 58o
I. Total or Centesimal Analysis . . 580
I. Fusion with alkali carbonate . 58o
Determination of :—
Silica, titanic acid . 581
Alumina, oxide of iron o . 582
Manganese, calcium, magnesium 583
2. Treatment with hydrofluoric acid. 583
3. Estimation of other constituents:—
Sulphuric acid, pyrites, carbon,
dioxide, loss on ignition. . 584
II. Rational Analysis . e ge . 585
Lunge and Millberg's method . . 585
Kreiling's method gº . 586
Other methods . . 587
Literature . 588
Clay Wares used for Building Purposes;
Earthenware and Glazes
Physical Examination . . 89
I. Density :—
I. Specific volume ę * . 590
2. Specific gravity & º e 590
3. Specific porosity, or water-absorption 591
4. Permeability to water . de . 59 I
II. Structure and Hardness:—
I. Structure of the material . ſº , 592
2. Flaws in the glaze . 592
3. Hardness. • 593
III. Resistance to Wear :—
I. Resistance to friction • 593
2. Toughness . o • 594
IV. Durability:—
I. Resistance to pressure . º • 594
2. Testing pipes for external pressure . 595
3. Testing pipes for internal pressure . 595
4. The breaking strength of roofing
tiles, etc. e • wº tº • 595
V. Durability towards Weathering:—
I. Examination for detrimental en-
closures g * º • 595
2. Permanence towards frost . 596
VI. “Fire-resisting” Qualities . 597
Chemical Analysis . 598
I. General Analysis:—
I. Determination of the composition of
fired articles . . 598
2. Examination for lead • 599
3. Estimation of alumina . * . 600
XX
CONTENTS
PAGE
Chemical Analysis—Continued.
II. Estimation of Soluble Salts . 6oo
III. Stability towards Acids . . . 601
IV. Stability of Glazes towards Weather-
ing o ſº º e . . 602
The Examination of Roofing Slates . 602
Imbibition test © º . 603
Weathering test . 603
Chemical analysis . 603
Literature tº . 604
Aluminium Salts and Alumina.
A’aw Mažerials : —
Kaolin . . 605
Bauxite tº * > sº . 605
Control of Working Conditions :—
Residue º * . 607
Aluminate solution . 607
Filter-press cakes. . 608
Finished Products . º G & sº
608
I. Aluminium Sulphate o . 608
Qualitative examination . . 608
Quantitative analysis . 61 o
Determination of : —
Alumina . 61 o
Iron . o * e . 614
Free acid . . 615
Other constituents. . 617
II. Potassium, Sodium and Ammonium
Alums . g . 617
III. Sodium Aluminate . 617
IV. Alumina . 619
V. Aluminium Salts used in Dyeing,
etc. (Acetate, Thiocyanate, etc.). 620
Literature © * tº ſº . 62o
Glass.
I. Raw Materials. . 621
I. Silica and silicates . 62 I
Sand . 622
Quartz . . 623
Flint tº . 623
Natural silicates . e o . 623
2. Boric acid and borates . 624
Boric acid . 624
Borax . . 625
Natural borates tº te . 625
3. Alkaline fluxes. e t e . 627
Potassium carbonate . 627
Sodium carbonate. . 627
Sodium sulphate . . 627
4. Lime ſº tº e tº º . 628
5. Lead oxide . © tº º e
628
PAGE
I. Raw Materials—Continued.
6. Other oxides and fluxes . . 629
Zinc oxide and baryta . . 629
Fluor-spar . . 629
Cullet . o e . 630
7. Decolorising agents. . 630
Manganese dioxide . 630
Nickel oxide, selenium and selen-
ium compounds, arsenious oxide
and nitre, compressed oxygen . 631
8. Colouring agents . e . . 631
Iron compounds . 631
Compounds of manganese, copper,
cobalt, chromium, nickel, ura-
nium, antimony . . . . 632
Silver, gold, selenium, cadmium . 633
Organic substances o . 633
Substances which produce opacity:—
Tin oxide, arsenic, bone ash, guano,
cryolite, cryolite substitutes . 633
Fluor-spar . 634
II. The Composition and Testing of
Glass tº e . 634
I. The Composition of Glass . . 634
Alkali-lime glass . 634
Alumina-lime glass . . 635
Boro-silicate glass . 635
Lead glass ſº . 635
II. The Testing of Glass for Resistance
against the Action of the Weather,
Water, and Chemical Agents . 635
Influence of the composition of the
glass. tº ſº º
te . 635
Action of acids, alkalis, and other
chemical agents. º Q . 636
Methods of testing, and results. . 638
III. The Analysis of Glass . . . 641
Qualitative Examination tº . . 64I
Examination for all constituents
except silica, boric acid, and
fluorine © e . 641
Silica, boric acid, and fluorine . 642
Quantitative Analysis . © e . 644
Preparation of the substance for
analysis te tº e gº . 644
Decomposition . e ſe & . 644
I. Leadless alkali-lime and alumina-
lime glass . . . . . . 644
Fusion with sodium carbonate . 644
Decomposition with hydrofluoric
acid . wº e * tº . 647
Decomposition with calcium car-
bonate and ammonium chloride. 649
CONTENTS
xxi
PAGE
Quantitative Analysis—Continued.
2. Lead glass . . 650
Semi-crystal glass. . 650
Flint glass . 65I
3. Glass containing boric acid . . 652
4. Glass containing fluorine . . 653
5. Glass containing phosphoric acid. 654
6. Coloured Glass . . 655
7. Glass enamels, metallic enamels,
and enamel colours . 656
Literature . 657
Calcareous Cements.
Raw Materials. e º gº º 658
I. Lime . . 658
A. Limestone º . 658
I. Preliminary tests, gas-volumetric
determination of the calcium
carbonate . e g º . 659
The Scheibler-Dietrich apparatus . 659
The Baur-Cramer apparatus. . 660
Volumetric determination of the
calcium oxide or carbonate. . 663
Determination of magnesium oxide 663
Clay . e © . 664
2. Complete analysis º . 665
Determination of :—
Moisture . © o . 665
Loss on ignition, carbon dioxide,
undecomposed matter (clay
and sand), silica, alumina, and
ferric oxide . . 666
Calcium . e G . . 667
Magnesium, sulphuric acid . 668
B. Quicklime e . 669
I. Keeping properties . 669
2. Slaking power . 669
3. Increase of volume on slaking . 670
Michaëlis volumenometer . 670
C. Lime . g © . 671
Tensile strength test. . 671
Determination of lime in mortar . 672
Frühling's method e . 672
Holmblad’s method . tº . 673
II. Hydraulic Admixtures . . . 674
A. Matural Puzzuolanas: Puzzuolana
proper, Santorin Earth, and Trass 674
Composition . . . . . 675
I. Determination of hygroscopic
water and of water of hydration ;
loss on ignition . o . 676
2. Degree of fineness . . . 677
3. Needle test o o e . 677
4. Tensile and crushing strength . 678
PAGE
B. Artificial Puzzuolamas . 679
Blast-furnace slag . . º . 679
Composition. º & & . 679
Strength º e º º . 68o
III. Hydraulic Cements . . . 68o
A. Watural Hydraulic Cements . . 68o
I. Hydraulic Lime . . o . 68 I
Analysis e O º º . 681
Determination of :—
Silica and gangue . p . 68 I
Alumina and ferric oxides . . 682
Calcium and magnesium, iron,
sulphuric acid, bitumen . . 683
The examination of marls . . 683
Differentiation of hydraulic lime
and Roman cement . d . 684
Permanency of volume . . . 684
2. Roman cement . º & . 685
3. Dolomitic or magnesian cement . 687
B. Artificial Hydraulic Cements, Port-
land Cement, etc. e e . 688
Raw Materials:—
Clay . º & . . 688
Estimation of the sand in clay,
elutriation . o e . 689
Nobel's apparatus . . . 690
Schöne's apparatus . . , 690
Portland cement e & e . 691
Chemical analysis . . . . 692
Determination of :—
Silica, sesquioxides . º . 692
Calcium oxide, magnesium oxide,
sulphuric acid, sulphur . . 693
Alkalis . © e º , 694
The composition of Portland
Cement gº Q o . 695
Specifications for Portland cement. 695
Clinker. o e e Q . 697
Ash and slurry . . 697
Examination of the physical pro-
perties of Portland cement—
Specific gravity . . . . 698
Degree of fineness of cement and
sand . . 7OI
The flourometer . 703
Hydraulicity. . 704
Time of setting . e . 706
Tests for permanency of volume . 706
Accelerated tests for permanency
of volume:—
Heintzel's calcining test . 709
Roasting test . tº º . 709
Boiling test . tº º . 7 Io
Tetmajer's investigations on the
tests for permanency of volume 7 Io
xxii
CONTENTS
PAGE
Physical Properties of Portland Cement—
Continued.
Increase of volume on mixing with
sand . e . 7 Io
Permeability to water . 7 II
Resistance to frost . 7 I 3
Abrasion test 7I4
Mechanical tests for Portland cement 714
The preparation of the test pieces. 715
The sand test . 718
The tensile strain . . 72 I
The crushing strain • 724
The adhesion tº e . 725
The bending strain e e . 726
Adulterations of Portland cement . 727
IV. Gypsum . . 729
Literature e e tº e . 73 I
Drinking Water and Water Supplies.
Rain water, well water . tº . 732
Spring water, river water . 733
Impurities in natural waters . . 733
Sampling º e º • 733
Physical Examination of Water • 734
Determination of temperature • 734
Clearness, colour, smell and taste. • 734
Chemical Examination of Water . 736
Statement of results . 736
Residue on evaporation . 736
Alkalinity . 737
Acidity. . 738
Hardness e e e . 739
Clark’s soap test . . º 74O
Reducing power or “oxygen absorp-
PAGE
Chemical Examination of Water—Continued.
tion ” g . 742
Silica, sulphuric acid . 748
Chlorine . . 750
Iodine . . 75 I
Nitric acid . 75I
The Schulze-Tiemann method of
determination . 753
Ulsch's method e . 753
Reduction by the zinc-copper couple 753
The Marx-Trommsdorf method • 754
Colorimetric determination with
brucine. º e º º . 756
Colorimetric determination with
phenolsulphonic acid . 758
Nitrous acid. º e e • 759
Phosphoric acid, carbonic acid • 762
Calcium and magnesium . 764
Sodium and potassium . 764
Ammonia . e º © e . 765
Nessler's reagent . º e º • 766
Albuminoid ammonia . . 769
Proteid ammonia . . 770
Iron © e . 772
Colorimetric estimation . 773
Lead e º e e º . 775
The determination of dissolved gases . 776
Nitrogen, oxygen, methane . . . 777
Adeney's apparatus for the extraction
and analysis of the gases dissolved
in water e e e . 78o
Winkler's iodometric method of analysis 782
“Rate of decay” of water . . 785
Solubility of oxygen and nitrogen in
Water e e * cº o . 785
Hydrogen sulphide . 786
AMicroscopic Examination . . 789
Bacteriological Examination . 789
Preparation of nutrient gelatine . 790
Counting the bacteria . e º . 79 I
Interpretation of Results . 793
Literature ſº . 797
Feed Water for Boilers and Water for
other Technical Purposes.
Hardness. ... 799
Alkalinity e . 8OO
Total alkaline earths e . 8oo
Sulphates, magnesium, iron, silica . . 8O1
Chlorides. º . 8o2
Acid waters º º tº o . 802
Special tests to determine the amounts of
lime and of sodium carbonate neces-
sary for the treatment of water . . 8O2
Pfeifer and Wartha's method . 804
Determination of the magnesium . . 804
Blacher's method. . 805
Literature . 806
Sewage and Effluents.
Constituents of effluents . . 807
Sampling. e tº . 809
Chemical Examination . º 8II
A. Preliminary Examination on the spot 811
B. Examination in the Laboratory .. 812
I. Residue on evaporation and loss
on ignition e e e . 8I2
2. Suspended and dissolved organic
matter and inorganic matter .. 813
3. The determination of colloidal
matter o e º ... 814
4. Oxidisability or oxygen-consuming
power e & . 815
5. Alkalinity . 8I6
6. Free acids . . e º ... 816
CONTENTS
xxiii
PAGE
Chemical A'zamination—Continued. |
7. Nitrogen . º e O ... 817
A. Total nitrogen, the Kjehdahl-
Jodlbaur method . 817
A. Ammonia . 818
C. Nitrous acid . e o º
82O
D. Nitric acid . 82I
The zinc-iron method . 822
The Schulze-Tiemann modifica-
tion of the Schlösing method 822
B. Suspended and dissolved organic
nitrogen and ammonia . . 824
F. Combined organic nitrogen
(albuminoid ammonia) . . 824
8. Sulphuretted hydrogen and sul-
phides . º e e . 825
9. Chlorine º . 825
Io. Other mineral substances . 826
II. Albuminoid compounds, sugar,
starch, and yeast . 826
I2. Oxygen . 827
Müller's “Tenax" apparatus . 827
I3. Organic carbon . e Լ . 829
A. Organic carbon in the filtered
Water . e e O . 830
A. Organic carbon in the suspended
matter . e o e . 83 I
I4. Carbonic acid . 832
I5. Detection of excreta . 832
I6. Detection of coal-gas products . 833
17. Tests for putrescibility; fermen-
tation experiments on effluents 833
Microscopic and Bacteriological Examination 834
Criteria as to the Contamination. Due to
Affluents and their Injurious Affects . 835
I. Effects on fish-culture . o . 835
A. By examination of the fauna and
flora of the contaminated water. 835
B. By examination of the water and
by experiments on its injurious
effect on fish . te . 836
2. Injurious effect on live stock . . 837
3. Injurious effect for industrial purposes 838
4. Injurious effect on the soil . 839
5. Injurious effect on plants . . . 839
6. Injurious action on ground and well
WaterS . e e tº & . 840
Literature © o © tº © . 840
Soils.
Classification of soils © tº º • 84I
PAGE
General considerations regarding the ex-
amination of soils. . 842
I. Sampling . . 843
II. Mechanical Examination . 844
III. Physical Examination . 849
Determination of :—
Capillary attraction . . 849
Capacity for water . ſº & 849
Absorptive power for salts in solution
849
Evolution of heat on moistening . 85I
IV. Chemical Examination . 85 I
Treatment with :—
Hot concentrated hydrochloric acid . 852
Citric acid º & . 852
Water containing carbonic acid. . 853
Concentrated hydrochloric acid . 854
Concentrated sulphuric acid . . 855
Hydrofluoric acid . • & . 856
Zhe Determination of the Individual Con-
stituents of Soil :-
Hygroscopic water, chemically combined
water, or loss on ignition, humus . 857
Carbon dioxide . 859
Total nitrogen, ammoniacal nitrogen,
nitric-nitrogen . © & . 861
Organic nitrogen, chlorine, sulphur . 862
Ferrous iron, copper, and lead . . 864
Zinc . © tº . 865
Literature . 865
Air.
The constituents of air . © tº . 866
I. Estimation of the Mormal Constituents of
Air . º ſº e & e . 867
I. Oxygen . 867
2. Nitrogen . e e © e . 869
3. Carbon dioxide e tº tº . 869
Pettenkofer's method . . 870
Gas-volumetric methods, Haldane's
method . © © tº c . 873
Approximate methods, Lunge and
Zeckendorf's method . º . 875
Wolpert's method . . 876
4. Aqueous vapour . . 877
5. Hydrogen peroxide . . 879
6. Ozone . . e º . . 881
7. Ammonia s º . 882
8. Nitrous (and nitric) acid . e . 882
II. Estimation of the Impurities in the Air
of Towns and of Confined Spaces . . 883
Methods of sampling . 883
I. Sulphurous acid . . . . 884
xxiv.
CONTENTS
PAGE
II. Estimation of the Impurities in the Air,
etc.—Continued.
2. Hydrogen sulphide . . . . 888
3. Hydrochloric acid . . . . 888
4. Carbon monoxide . tº tº . 889
The palladious chloride test . . 889
The spectroscopic test . . . 890
Haldane's colorimetric test . 892
Colorimetric test after the addition
of reagents e tº G . 893
The flame-cap test . . . 894
The quantitative determination of
carbon monoxide ſº tº . 894
5. Hydrocarbons:—
Methane, acetylene e g . 897
Benzene, petroleum vapour . . 898
6. Organic matter ſº . . . 898
7. Soot and smoke . * . . 899
8. Dust e * * > e g • 90I
III. The Estimation in Air of Impurities PAG
associated with Specific Inaustries • 903
I. Chlorine, bromine, and iodine; 2.
Phosphorous trichloride e • 903
3. Hydrofluoric acid; 4. Cyanogen and
hydrocyanic acid; 5. Hydrogen
phosphide; 6. Hydrogen arsenide 904
7. Mercury vapour; 8. Carbon bisul-
phide . © e • . 905
9. Ether vapour; Io. Mercaptan . 906
II Aniline . tº gº tº tº . 907
Effects of the Impurities in Air tº • 909
Tobacco Smoke. º e tº tº . 910
Literature . 91 I
APPENDIX tº * & tº º o • 9I3
INDEX . tº & tº ge gº . 977
GENERAL METHODS USED IN
TECHNICAL ANALYSIS
GENERAL METHODS USED IN TECHNICAL
AN ALYSIS
By Professor G. LUNGE. English translation revised by CHARLES A.
KEANE, M.Sc., Ph.D.
INTRODUCTION
TECHNICAL methods of chemical analysis have always included many
general analytical methods which were introduced as they gradually
developed for scientific purposes, as well as certain practical tests which
originated for the most part either accidentally or in imitation, on the
small scale, of the manufacturing processes of chemical industry. On
the other hand, whole classes of operations which now form a consider-
able part of general analytical chemistry were originally devised and
looked upon as purely “technical" methods. This is particularly true
of volumetric analysis, which, founded as early as 1795 and 1806 by
Descroizilles, was first used as a method for controlling and estimating
the value of technical products such as acids and alkalis. Gay-Lussac's
methods for chlorimetry (1824), alkalimetry (1828), and for the estima-
tion of silver (1832), as well as the permanganate method of estimating
iron introduced by Margueritte (1846), fall in the same category, and
were all devised to meet the requirements of technical work, and were
at first employed exclusively for this purpose.
Up to the middle of the last century, and still later in many
instances, “volumetric’’ analysis was unrecognised in the course of
instruction given in university laboratories: it was regarded with a
certain degree of contempt as a “technical” method, the study of which
could only be disadvantageous to the attainment of Scientific accuracy
by students. This view could, however, no longer be upheld after it
was shown by Bunsen, in his paper on volumetric analysis by means of
iodine (1853), that the accuracy of this process was greater than that
of most gravimetric methods, and after other investigators, especially
H. Schwarz (1850) and Friedrich Mohr (from 1855), had extended and
develºped the province of this branch of analysis.
4 GENERAL METHODS USED IN TECHNICAL ANALYSIS
Text-books of general analytical chemistry and introductory works
on the subject were at that time the only sources of information avail-
able for technical analysis; these were restricted, however, to the gravi-
metric methods of mineral analysis, elementary organic analysis, and a
few volumetric processes. Methods for the examination of products
of chemical industry and of associated manufactures were, if published
at all, scattered in all kinds of books and journals, and were only
accessible at the cost of considerable labour, and then with but very
indifferent success.
The publication by Bolley, in 1853, of his text-book of technical
methods of chemical analysis was accordingly heartily welcomed; it
was almost sold out within a year, but a second edition was not
published until 1861. Nearly half of this book of 475 pages is taken up
with a short sketch of general analytical operations, methods, and
apparatus, and with qualitative and quantitative mineral analysis,
treated on the same lines as in every work on analytical chemistry of
that period; the second and rather larger part of the book dealt with
the analysis of gunpowder, bleaching materials, soils, colouring matters,
fuels, fats, oils, illuminants, soap, beer, wine, sugar, milk, tea, coffee,
textile fabrics, tanning materials, manures, specific gravity determina-
tions, etc.
In spite of its modest scope, Bolley's book was one of great import-
ance and value at the time of its publication, insufficient as it appears to
our present conceptions. It, moreover, indicated a new field of work,
and introduced a fundamental principle which must still be recognised,
and which was expressed by Bolley in the following words in the preface
to the second edition of his book:—
. . . “On the other hand, no artifice, however empirical or wanting in
scientific basis, should be despised which, according to reliable informa-
tion or from personal experience, presents marked characteristics. In
fact, it lies in the nature of the case, and therefore I do not consider it
necessary to defend it, that certain methods of analysis are made use of
which are in some respects untrustworthy. I am not unaware of the
objection, which however I consider quite unfounded, that the propaga-
tion of such methods is detrimental to science. The latter will certainly
not allow itself to be thereby diverted from the search for more rational
and accurate methods of analysis, whilst all chemists who are concerned
with the investigation of commercial preparations or of industrial
products must in very many cases welcome every method of examina-
tion, even if it is only of a tentative nature.”
This view, that methods may and must be employed in “technical
analysis” which cannot be justified from a purely scientific standpoint,
still holds; the use of such methods is only justified as a temporary
expedient, and therefore only until they are replaced by better and
INTRODUCTION 5
more rational processes. It is scarcely necessary to add that the efforts
of all concerned should always be directed towards the attainment of
this object, and a great deal has been accomplished in this respect
during the half century which has elapsed since the publication of the
first edition of Bolley's book. But even now and in the future, very
many analytical methods will always be in use for technical purposes
which do not find, and ought not to find, a place either in text-books of
scientific chemical analysis or in the methods taught in university
laboratories. The methods of technical analysis have developed inde-
pendently in accord with the unavoidable specialisation that has arisen
in all branches of science, and they cannot, for the most part, be fitted
into the ordinary course of study which long experience has established
with tolerable uniformity as the most fitted for instruction in analytical
chemistry. This is especially so in respect to the technical examination
of organic products and processes; purely empirical methods are often
quite unavoidable, although many others with a thoroughly scientific
basis are also employed.
The necessity of controlling and regulating the practical operations
of works by means of chemical analysis also calls for special considera-
tion. In this case scientific accuracy is usually of no importance what-
ever; even were it attainable, it would have no greater value for the
control of the working than an approximate estimation. It is, on the
contrary, of far greater importance to obtain the analytical results
as rapidly as possible, in order that the working conditions may
be regulated accordingly. In such cases, which are of very frequent
occurrence in technical work, lengthy analytical methods are quite
excluded; even if a whole army of highly-trained scientific analysts were
employed, were this practicable, their reports would come “after the
fair,” and would therefore be quite worthless, whilst simple tests rapidly
carried out by an empirical “tester” give the manager all needful infor-
mation. Any great degree of accuracy is usually neither attainable nor
necessary in such instances. There are, however, some cases in which
rapidity and sºme degree of accuracy must be combined, and this is
successfully attained by “testers” through constant practice; the
accuracy attained by workmen exclusively devoted to one branch of
testing is often such as to command the admiration of the trained
chemist whose work is conducted at leisure in his laboratory. This is
the case, for instance, in iron-works, and the degree of certainty in the
quality of the product attainable nowadays in this industry is in great
part to be ascribed to the development of technical methods of analysis,
which combine extreme rapidity with astonishing accuracy.
In the analytical work concerned with the purchase of raw materials
and with the sale of manufactured products, the chemist has usually
more time, although by no means unlimited time, at his disposal. Here
6 GENERAL METHODS USED IN TECHNICAL ANALYSIS
again, the nature of the product often necessitates the employment of
more or less rough empirical methods, but if better and more exact
methods are available they should certainly be preferred. An accuracy
and certainty in analysis such as would be regarded even in purely
scientific laboratories as very good work, is often required of a works
chemist, by whom it can only be attained as the result of constant
practice.
The technical methods of chemical analysis comprise the following
classes of work:—
I. EXAMINATION OF RAW MATERIALS
Accurate methods of analysis are frequently employed. In other
cases, rougher tests may be sufficient within limits, and in part must
suffice in default of better methods; physical, microscopic, or other
external methods of examination are also often used.
2. CONTROL OF WORKING CONDITIONS
Chemical analysis, with the addition, in many cases, of other methods
of a non-chemical character (e.g., specific gravity determinations, pressure
relations, etc.), are adopted. The methods usually employed are those
which can be carried out with the greatest possible rapidity, and are
consequently often lacking in accuracy.
3. ExAMINATION OF FINAL PRODUCTS
Accurate methods of analysis are sometimes employed, especially
when a guarantee of composition is required; frequently, however,
easily observable external characteristics suffice. -
In all three classes different methods of examination may be neces-
sary, namely:—
(a) Qualitative tests, which are in most cases concerned with the
detection of an impurity.
(b) Quantitative determination of a principal constituent, which
is the subject of a guarantee in buying or selling, or which is a leading
factor in the working of a process.
(c) Quantitative determination of secondary constituents, in which
various considerations demand attention, namely:-
(a) Secondary constituents of value which form the subject of a
guarantee, such as the percentage of carbón in steel.
(8) Secondary constituents which are deleterious impurities, and
which ought not to exceed a certain maximum, e.g., phos-
phorus in steel, chlorides in potassium nitrate.
(y) Secondary constituents the estimation of which serves for the
indirect determination of the amount of the principal constituent
present, as in the valuation of Chili saltpetre.
TECHNICAL METHODS OF CHEMICAL ANALYSIS 7
(d) Quantitative determination of several constituents, usually a
combination of the two classes ó and c, i.e., the estimation of the chief
constituent and of one or more important impurities, as the estimation
of the alumina, iron, and free acid in the analysis of aluminium sulphate,
and the determination of the iron, sulphur, and phosphorus in the
examination of iron ores.
(e) Examination of certain quantitative or qualitative effects
produced by the substance, e.g., the tinctorial power of dyes, the
viscosity of lubricating oils, the flash-point of petroleum.
(f) Examination of certain external properties demanded by the
trade, such as, colour, lustre, density, strength, etc.
(g) Complete chemical analysis is only very exceptionally required
for technical purposes; it may be necessary in such instances as a new
discovery of ores and the like.
These various classes of analytical work are all dealt with in this
book; they are considered under the several sections dealing with the
various industries into which the subject matter is divided, according to
their nature. A general knowledge of analytical chemistry, and more espe-
cially of mineral analysis and of elementary organic analysis, is premised.
The description of special apparatus and methods of work is in all cases
directed to the conditions that are requisite or useful for the attainment
of the results that are required in technical work, even in those instances
in which general methods of analysis are employed. The greater part
of the book is, however, naturally devoted to the methods which have
been specially worked out for technical purposes, and which are but
seldom if ever practised in scientific laboratories.
The work is divided into a general and a special part. In the
former, the methods and apparatus described are of a general character,
such as are used for a variety of purposes, particular stress being laid
upon those considerations that serve to facilitate the work of technical
laboratories. The more specific applications of these general processes
is dealt with in the special part of the book in which the analytical
methods are grouped, under the headings of the products concerned
with the various branches of chemical industry.
I.—GENERAL OPERATIONS
THE TAKING OF SAMPLES
The taking of samples is a problem of some difficulty, and is one
which is practically outside of the domain of the purely scientific
analyst, who often fails to recognise its importance; the technical chemist
who has not had a scientific training is still more frequently, perhaps,
ignorant of its difficulties, so that serious errors and heavy losses
8 GENERAL METHODS USED IN TECHNICAL ANALYSIS
ensue. The preparation of an average sample, really representative of
the quality of the material to be investigated, requires special care and
precautions. It has happened, for example, that a piece of marl from
a quarry was given to an analyst in order to ascertain its suitability
for the manufacture of cement, and that an accurate analysis of the piece
was made, and conclusions drawn from the results as to the suitability of
the whole material in the quarry for this purpose !
Even in much simpler cases the difficulty of obtaining a fair average
sample is often not sufficiently realised by those concerned. How easy
it is, for example, to make a considerable error in taking a sample of
coal, if a large piece of pyrites finds its way into the sample—or con-
versely. But even in taking samples of powdered materials, serious
errors or intentional deception may occur, if all necessary precautions
are not taken.
The preparation of a really average sample is most difficult in the
case of materials occurring in large pieces; it is considerably easier with
more finely grained materials, still easier with powdered materials,
easiest of all, as a rule, with liquids, and more difficult with gases.
Special difficulties occur when contact with air, whilst the sample is
being taken and ground, may alter its condition by evaporation, absorp-
tion of water, oxidation, etc. In such cases very special precautions
are necessary, which will be fully dealt with subsequently.
The most accurate analysis is, of course, worthless, and endless mis-
takes and disputes must arise, if the samples are not properly taken ;
it is therefore essential to expend as much care on sampling as on the
analytical work in the laboratory.
As already stated, sampling is most difficult in the case of materials
occurring in lumps, and this is especially the case when the valuable con-
stituent occurs only in small amount and very unequally distributed,
as in ores of the noble metals, or when certain very deleterious impurities,
also unequally distributed, are present. The disputes that have arisen
from defective sampling have led to the drawing up of definite rules for
sampling; also, it has become the custom, in all important cases, for the
operations of taking samples, grinding them, and making them into a
form ready for handing to the chemist, to be carried out in the presence
of representatives of both parties concerned, or in the presence of a
mutually appointed referee. Usually several bottles are simultaneously
filled with the prepared average sample, and closed with the seal of both
parties.
The following directions lay no claim to being a permanent solu-
tion of the difficult problem of sampling; they can only represent an
approximation, but they are at least as satisfactory as any other plan
that is in use. They are founded essentially on the directions contained
in The Alkali Makers' Handbook, by Lunge and Hurter (second edition,
THE TAKING OF SAMPLES 9
1891, p. 172 et seq.), which were drawn up with the co-operation
of a number of experienced manufacturers, and are amplified by the
results of further experience.
In many cases, particularly in the industries connected with
organic chemistry, but also, e.g., with caustic soda, fuming sulphuric
acid, etc., the samples must be taken according to special methods
applicable to each individual case, owing to the peculiar nature of the
substances in question. The necessary information on these points is
given under the individual sections.
Detailed rules for sampling in metallurgical works have been given
by Juon; the remarks on sampling by Hintz” do not call for any
special comment.
A. MATERIALS IN LARGE PIECES
This class includes coal, metallurgical ores, pyrites, pyrolusite, etc.,
which are usually transported in shiploads, canal boats, or in railway
waggons. In case of water-transport, the check sampling takes place as
a rule at the port of arrival, during the transfer to the railway waggons,
or, in cases where the factory or the foundry receives the goods at its
own landing-place, samples are taken there during the unloading into
trucks or other transport vessels. In other instances the sample is
taken on arrival of the train at the works, either before or during the
unloading of the material. In all these cases the sampling can be
advantageously combined with the weighing. The taking of samples
from a previously unloaded large heap of coarse material is always a
very uncertain procedure, and should be avoided whenever possible.
Apart from the general difficulties which unavoidably occur in the
above-mentioned conditions of sampling, two additional difficulties arise
in the last case; on the one hand, changes in the amount of moisture
may occur, owing to evaporation, showers of rain, downward percolation
of moisture, etc., and secondly, the larger pieces are apt to roll forwards,
and thus make the mass still less homogeneous than before.
The larger and the less uniform the pieces of the material in question
are, the larger must be the samples taken. It is most important to
secure a due ratio between the larger pieces and the finer powder, which
is practically always present, since there is often an essential difference
in quality between the two.
If the pieces do not exceed a billiard ball in size, and are roughly of
the same dimensions, it is sufficient to take a sample from each unit-
load 8 by means of a scoop of about 5 kilos (12 lbs.) capacity. If the
* Z, angew. Chem., 1904, 17, 1544, * Z. f. §ffentl. Chem., 1903, No. 21.
* This term designates the load raised by a crane, carried by a truck, or any other conveyance,
by means of which the material is transported from the hold of the ship, etc., to the weighing
machine, or to the landing-place.
10 GENERAL METHODS USED IN TECHNICAL ANALYSIS
material is in larger pieces, and especially if not uniform in size, it is
preferable to empty a whole unit-load at intervals, e.g., the tenth or
twentieth load on the weighing machine, on to a separate place, from
which the whole average sample is collected. The greatest possible
care must be taken, in Sampling by this method, to have the ratio
between coarse and fine material represented as accurately as possible
in the average sample. - -
The average sample thus obtained is first crushed to pieces about
the size of a walnut. This is done either by hand or by means of a
mechanical arrangement such as studded rollers, according to circum-
stances, care being taken to break down the whole of the coarse pieces,even
although it may be troublesome to do so. The roughly broken material
is then thoroughly mixed by repeated scooping backwards and forwards,
and is then spread out in a flat heap and a smaller sample taken. This
is best effected by cutting two stripes crossing each other at right
angles out of the whole heap, and adding to this four smaller quantities
taken from the middle of each remaining quadrant.
The weight of the sample thus prepared should be at least IO-12
kilos (+ cwt.); if the material is very uneven in character, considerably
more must be taken and special attention must be paid to the precautions
described above for obtaining a true average
sample. In such cases it is often necessary
to repeat the above operation by Scooping
the first large sample together in a conical
heap, spreading it out flat again, and then
again cutting two stripes intersecting at right
angles, and adding further material from the
quadrants as above.
Several mechanical samplers have been
constructed in order to make this operation
simpler and more reliable. Of these the
“rapid sampler” of P. Clarkson" is largely
employed here (Fig. 1). The principle of
the machine is to cut out a number of
Miſſºlºiſſºl.III.iiiliºſitiºliifºliº sections of the material as it passes through
FIG. 1. the apparatus in an annular rotating stream ;
the relative amounts can be regulated at
discretion. This machine can be employed for liquids, powders, and for
substances in grains or in large fragments. It is also very suitable for
uniformly distributing a sample into bottles.
The reduced sample thus obtained is then further broken up. A
mechanical arrangement may also be employed at this stage, but only
when it is so arranged that it can be thoroughly cleaned after every
* For a full description, see J. Soc. Chem. Ind., 1894, I3, 2I4.

MATERIALS IN LARGE PIECES 11
operation. The more common practice is to effect the grinding
by hand, usually in a large iron mortar. This is, however, not the
best method, because it is not very easy to completely remove the
powder from the mortar without loss, and the uniformity of the sample
is thus adversely affected; also, some of the substance may easily remain
behind unnoticed in the mortar, and contaminate the next sample. (For
arrangements for grinding, cf. p. 17.)
To overcome the disadvantages of an ordinary mortar, a flat cast-
iron plate, specially constructed for the purpose, O-8-I metre square,
with a vertical rim 5-IO cm. high, interrupted at one point in order
to allow of the powder being easily run out, may be substituted ; these
plates should be 20-25 mm. thick, and solidly embedded in a horizontal
position to prevent breakage. The sample, previously reduced to pieces
not larger than walnuts, is further broken up on the plate with a sledge-
hammer, shaken from time to time through a sieve of 3 mm. mesh, and
the residual coarse powder repeatedly ground up until it has all passed
through the sieve. In this way complete cleanliness can be observed
with much greater certainty than by the use of a mortar.
The IO-12 kilos (3 cwt.) thus obtained are then spread out flat, and
a further decreased sample of 1-2 kilos (2 to 4 lbs.) prepared from it by
thorough mixing and removal of intersecting stripes, etc., in the same
way as before. This is subjected to a further mixing, and the separate
samples for analysis are then taken from it, preferably as follows.
Three, four, or more wide-mouthed sample bottles holding IOO-2OO c.c.
are placed close together on a sheet of paper; a handful of the sample is
taken, and the hand moved over the bottles in succession, so that some
of the substance falls into each. This is repeated until the bottles are
quite filled. Manipulation by hand is more reliable than the use of a
small scoop, etc.; in the latter case the coarser particles always roll forward,
so that too large a proportion is delivered into the foremost bottles.
When the bottles are full, they are closed immediately with tight-
fitting corks, which are cut off straight above the necks of the bottles
and carefully sealed. When control samples are taken, the seals of both
parties are affixed in such a way that the cork cannot be removed
without injuring them.
The powdering, mixing, and filling of the bottles should be carried
out as quickly as possible, to prevent the evaporation of water from
moist products, or, conversely, to prevent the absorption of moisture (under
special circumstances also of oxygen or of carbon dioxide) from the
atmosphere.
The analyst, on receiving the sealed bottles, notes the seal, which
must, of course, be intact, and the affixed label, opens the bottles, shakes
out the contents on to glazed paper, and mixes the sample quickly.
If the moisture is to be estimated, a sufficiently large sample, up to
12 GENERAL METHODS USED IN TECHNICAL ANALYSIS
100 g., is taken for this purpose without further grinding. The remainder
is ground up until the whole passes through a sieve of 1 mm. mesh.
Porcelain or steel mortars are used for the grinding, according to the
hardness of the substance. The former would not be suitable for pyrites,
for example, because they are attacked to some extent by substances of
this degree of hardness, and the sample might be contaminated; on
the other hand, iron vessels should not be used for reducing pyrolusite
to a fine powder, because some metallic iron might thus find its way into
the sample.
Finally, a few grams of the substance which has passed through
the 1 mm. sieve is taken, after a further thorough mixing, and reduced
by means of an agate mortar to the degree of fineness requisite for
analysis; in special cases a steel mortar may be employed.
Since some change in the amount of moisture is almost unavoidable
during the final grinding, the final sample is either dried in a drying
oven or exsiccator, and then weighed out in the dry condition for
analysis, or the moisture is determined separately in another sample and
the results calculated on the dried substance. The actual percentage of
moisture in the sample is taken from the determination made with
the coarsely powdered substance as above, and not from this last
determination.
B. RAW MATERIALS IN THE FORM OF POWDER, DROSS, ETC.
This class includes ores in the form of smalls or slimes, common salt,
potassium salts, etc., as well as many other inorganic and organic raw
materials. A simplified method of sampling can usually be employed for
products of this character, samples being taken from each “unit-load” by
means of a scoop of about ; kilo (I lb.) capacity; in the case of a railway-
waggon load, several samples from different parts, e.g., the front, middle,
and back are taken. The separate samples are placed together in a cask
and covered over. After the sampling is finished, the contents of the
cask are emptied out on a level, clean, hard surface, spread out flat, and
the mass scooped together into a cone in the centre by working
regularly round the heap with a spade ; the heap is again spread out
flat and a sample of about a quarter of the mass taken by removing two
intersecting stripes with the scoop, and adding to this some material
from the middle of each of the remaining quadrants. This operation is
repeated with the sample thus obtained, until finally not more than
2 kilos (4 lbs.) is left, which, after being well mixed, is divided among
the sample bottles required for the analysis, as described.
In the case of raw materials which arrive in boxes, casks, sacks, etc., the
procedure described under C can usually be followed; in fact, the final
products of one industry are very frequently the raw materials of another.
PRODUCTS IN THE FORM OF POWDER 13
C. CHEMICAL PRODUCTS IN THE FORM OF POWDER.
If these are loaded in loose condition, that is in ships' holds, railway
waggons, or carts, the sample may be taken as described under B; but
it is better to employ an auger, especially if the samples have to be taken
from stores or from casks or sacks. In the latter case a sample is taken
from every fifth, tenth, or twentieth cask or sack, according to the size
of the consignment and the probability of inequality in the material;
it is to be borne in mind however, that, owing to the action of the
air (or possibly owing to deceitful manipulation (), the condition of the
outside and superficial portions of the material may differ
more or less from that of the interior. It is, therefore, always
more reliable to use the auger.
The usual form of this instrument is shown in Fig. 2. It
consists of a long borer of stout sheet-iron, with a hollow
interior and a longitudinal slit on one side, provided with a
handle a at the upper end, and with the lower end beaten
out to a sharp point à. -
The use of the auger renders it possible to obtain a sample
throughout the whole layer of substance. If a sample is to
be taken from sacks or casks, the instrument is used only
once for each vessel, but if, on the other hand, it is to be
taken from a large heap stored in a warehouse or the like,
the auger should be inserted in from six to twelve different
places according to the size of the heap. Care must be taken
to turn the instrument on its axis when it is lowered into the
heaps, casks, etc., for drawing the sample. -
This simple form of the auger has several defects. For
example, in the sampling of sugar it is quite satisfactory for FIG. 2.
the purer primary products and even for well separated
secondary products, but it is not reliable for moist, syrupy cane sugar,
because the column of sugar, on being drawn out, is apt to break off
at the lower end of the auger. Consequently, according as the sack
is erect or inclined, the sample may represent either the moistest part
of the contents containing most syrup, or the best-drained portion.
Again, in taking samples of granulated or crystal sugar, it often
happens that none of the sugar remains in the sampler.
In order to avoid these drawbacks, Gawalowski' has constructed an
auger which Böckmann recommends from long personal experience.
This (Fig. 3) consists of a metal sheath which has a metal bearing
suitably soldered to it at B, and in which a quadrangular or triangular
steel rod provided with a handle (C), slides. The valve at the lower end
of the sampler is movable by means of a hinge, in such a way that in
* Oest. Zeit. Zuckerind, 1888, Part V.

14 GENERAL METHODS USED IN TECHNICAL ANALYSIS
position D, it allows the substance to be sampled, to enter from below
upwards, whilst in position D it prevents the sample from falling out
of A. When the auger is pushed into the
material, the rod C slides sideways in such
a way that the flap is fixed in position D1,
and on with drawing it the rod C releases
the flap and the latter is closed by the
pressure of the substance in the sack on the
attached flange.
Gawalowski's auger works automatically,
and samples of material can therefore be
taken quickly and safely; it is made of
copper or other suitable metal.
Angerstein's' auger, a tubular instru-
ment, with the lower part half-open and
running to a point at the extreme end, is
also to be recommended ; it allows of samples
being taken from the interior of a vessel
without admixture with the upper, altered
layers. In whatever way the separate
samples may have been taken, they should
first be emptied into a box, or, in the case
of alterable substances, into a large bottle,
which must be kept corked between each
addition. The whole contents of the vessel
are then thrown out on to a large sheet of
paper, thoroughly mixed, any lumps that
may be present crushed with a spatula, and
then if the quantity is too large a smaller average sample taken, as
described above. The samples are then placed in bottles, corked and
sealed.
In the case of substances which alter quickly in the air, such as
bleaching powder, potassium carbonate, etc., the sample bottles must be
kept well closed, and only opened when a new portion of the sample is
to be added; similarly, the mixing and distribution of the sample into
the bottles must be carried out as quickly as possible.
FIG, 3.
D. LIQUIDS
The sampling of mobile liquids is a very simple operation, but that
of viscous liquids, and especially of sticky, syrupy substances and pasty
mixtures, is more difficult. When ordinary liquids are contained in
large receptacles, the different layers very often differ in composition.
1 D.R.P. 2668o.

SAMPLING OF LIQUIDS - I 5
If, owing to shaking during transport, or from other circumstances,
it can be taken for granted that the contents of each individual vessel
are uniform, a small sample is taken from every fifth, tenth, or twentieth
cask or carboy, etc., by means of a syphon, or as otherwise desired,
and an average sample prepared by mixing and shaking. The taking
of samples when smaller vessels used for transport are being emptied
into a larger receptacle, is very simply
effected, by placing a small flask, held by
means of a wire (Fig. 4), under the bung-
hole or stopcock of the cask, or the mouth
of the jar.
In the case of large receptacles, where
the different layers might vary in quality,
the sample is taken by means of a glass
tube of suitable length, contracted above and below, the equivalent of
a pipette, which is gradually lowered into the liquid, so that a section
through the whole depth of the vessel is obtained. In many cases an
iron tube, provided with some kind of valve, may be employed for this
purpose.
If a sample of a very large quantity of liquid has to be taken,
e.g., from a tank-cistern holding IOOOO-15,OOO kilos (IO to 15 tons),
or from a larger reservoir holding IOO,OOO-2OO,OOO kilos (IOO to 200
tons), the sampling with the glass pipette or long iron tube provided
with a valve at the lower end, must be repeated several times.
Gawalowski' has constructed a suitable pipette for liquid and semi-
liquid substances, which consists of two iron cylinders, one inside the
other, closed at the lower ends, and each having a fairly wide
longitudinal slit; they are so connected, by means of a bayonet-slot,
that the two longitudinal slits can be made to coincide to form a
single closed hollow cylinder, by a simple rotation. The instrument,
which is sufficiently long to be able to penetrate all the layers of a large
cask, and is provided with a handle at the top, is introduced into the
liquid in the closed position and then opened, so that liquid enters
uniformly from all levels; the cylinder is then closed, withdrawn, and
the contents delivered into a shallow vessel.
For taking samples of the liquid intermediate products of a
working process, the specific apparatus is provided with stopcocks or
valves, by means of which samples can be taken as often as desired. If,
however, an average sample, consisting of a number of single samples
taken seriatim without a break, is desired, so-called “dropping-
bottles” are employed. These are large jars, holding about 50 litres,
into which the liquid drops uninterruptedly. The rate of flow is
regulated by means of a valve, and the dropping is controlled by means
* Oest, chem. u. tech. Zeitschrift, 6, 197,

16 GENERAL METHODS USED IN TECHNICAL ANALYSIS
of a glass tube attached by a rubber connection, carrying a narrow
rubber ring inside at the top, in which a small glass tube drawn
out to a fine point is fixed.
Satisfactory average samples of effluents and the like can be
obtained by employing a water-wheel revolving in the liquid, by means
of which small samples of the liquid are transferred continuously to a
reservoir, from which the average sample is taken.
Very concentrated liquids, in the analysis of which possible errors
in measuring would have a considerable effect, are often not employed
directly for analysis, but are first diluted (e.g., I, Io, or 20 c.c. to
IOO, 200, 500, or IOOO c.c.), and an aliquot part of the diluted liquid
taken. In many cases, however, it is simpler to measure out such liquids
with an accurate pipette, e.g., a I c.c. pipette graduated in T}o c.c.
E. Gases
The sampling of gases, which may be either final products, as with
illuminating gas, or by-products of industrial processes, as is more
frequently the case, is described in the section on “Technical Gas
Analysis,” and under the individual industries concerned.
THE COLLECTION AND STORAGE OF SAMPLES
In works where the processes go on uninterruptedly, three vessels
should always be kept ready for each product of manufacture under
analytical control, one of which is allotted for the night-shift and the
other two for the day-shift.
The taking of samples in the works is, as a rule, carried
out by a workman specially appointed for the purpose, in addition
to his other duties. In the morning after six o'clock, he fetches the
sample bottles collected during the night, which are distributed at
different parts of the works, or he takes the samples, if this has not
already been done. All the bottles are put into a wooden “sample
box,” which contains a number of compartments, and taken into the
laboratory. Suppose, for instance, it is seven o'clock in the morning of
December 6th. The laboratory boys begin testing the night samples
from December 5th to 6th which have just been brought in, as well as
the day samples brought in on the evening of December 5th before six
o'clock. In the evening the sample collector fetches away the sample
bottles brought in the morning as well as the day sample bottles brought
in twenty-four hours previously, whilst at the same time he brings the
day samples of December 6th.
These samples may be kept on a suitable stand provided with
several compartments. Thus over night only the day samples brought
COLLECTION AND STORAGE OF SAMPLES 17
in the evening, on the following day the night samples brought in the
morning as well, would be accommodated in the stand.
This system of taking samples has the drawback that the results for
the day samples are usually obtained fully twenty-four hours after the
sampling, whereby in many cases timely measures for correcting the
working of the process in question are rendered impossible. An advan-
tageous alternative method is to take the day samples only from six
o'clock in the morning till two o'clock in the afternoon, and the
night samples from then up to the next morning at six o'clock. The
results of the previous night's work are then got at midday, and those
of the same day in the evening. This system of taking samples has
also its disadvantages, however, because almost half the day is reckoned
as night, and therefore irregularities in the working occurring in the
afternoon cannot be so easily controlled and located.
These shift - samples, being temporary, are not retained for
more that twelve or twenty-four hours respectively. The samples
of raw materials and of final products, on the other hand, are
usually stored carefully for one or more months (not infrequently
sealed), so that they can be referred to in case of dispute. Date of
sampling, name of the seller or buyer, percentage of the constituents or
of the important constituents found, weight, and method of packing
(sacks, casks, etc.), and the number should be entered on the labels as
well as in a special laboratory book.
II.-GENERAL LABORATORY OPERATIONS 1
A. GRINDING OF SUBSTANCES
The reduction of a large average sample to a smaller one has
already been discussed; the grinding is effected by very diverse
methods, and is carried to very different degrees of fineness according
to circumstances. -
For the coarse powdering of hard substances, such as ores and
analogous products, iron mortars holding from I-20 litres and upwards
are suitable, provided contact with iron is not disadvantageous.
A good arrangement for grinding is shown in Fig. 5. Instead of the
pestle, there is a heavy almost spherical block, carried on a pivot at the
bottom of the mortar, which is moved round by a handle, and which can
readily be lifted out.
For substances which form a fine dust, the twofold evil of incon-
venience to the operator and loss of material is avoided by fixing a
kind of sack round the rim of the mortar; the sack is contracted at the
* Cf. H. Benedikt, Z. angew. Chem., 1902, 15, 78. This paper comprises some interesting con-
siderations, on the non-chemical side, of the arrangement of works’ laboratories.
B
18 GENERAL METHODS USED IN TECHNICAL ANALYSIS
top, where it is firmly bound round the pestle; this precaution is
specially to be recommended for poisonous substances.
It is almost always necessary to pass the material in the mortar
- through a sieve from time to time,
and to powder the coarse residue
separately. -
For very hard ores and the like,
an ore-crusher can be employed with
great advantage, such as that shown
with the cover open in Fig. 6.
For grinding small amounts of
material, the ordinary steel mortar
is used ; when dust is to be avoided,
it is provided with a brass cover,
which is screwed on.
In default of a steel-mortar, re-
course may be had to the following
plan. The pieces of material are
wrapped in tough paper of close texture, and broken up on a hard
support with a heavy hammer; a little of the substance is lost in this
manner, owing to its being beaten into the paper, from which it cannot
easily be completely detached.
Less hard substances are usually powdered in porcelain or earthen-
ware mortars; these would give up some of their material to harder
FIG. 6.
substances. Apparatus constructed on the principle of coffee-mills,
and made either of earthenware or of iron, are also used.
In large works where one and the same kind of substance has often
to be powdered, mechanical grinding-machines are employed, such as
small ball-mills, plain-, ribbed-, or toothed-rollers, apparatus similar to
coffee-mills, or small crushers with vertical rollers.
The breaking up of tough substances and of materials of an excep-
tionally irregular character often requires special apparatus; these are


WEIGHING 19
described in the respective sections, but it may be mentioned that
cutting-machines working on the lines of a hay or straw chopper, coffee-
mills, toothed-crushers, and the like, may often be adapted for this
purpose. -
All forms of grinding apparatus should be accessible in all their
parts for thorough cleaning, and should be cleaned regularly.
Reduction to a fine powder for analysis is effected in agate, porcelain,
or steel mortars, according to the nature of the substance, and is, as a
rule, combined with rubbing (“bagging”) through silk gauze.
B. WEIGHING
A laboratory balance when adjusted should give readings to within
o, I mg. ; the more delicate assay balances are treated of subsequently.
The principles of weighing, calibration of weights, etc., are described in
the ordinary text-books on analytical chemistry. Rough balances,
usually without a glass case, are used for weighing out large amounts of
substance. For many purposes an intermediate quality of balance,
reading to 1 mg., and provided with a glass case, is very convenient,
particularly for weighing out fairly
large quantities of material, which
are then dissolved and made up to a
definite volume, of which an aliquot
part is taken for the analysis.
For less accurate work, drug-
gist's hand-scales, with horn scale
pans suspended by silk cords, are
very useful, the substance being
weighed directly into the scale pan.
A handy form of this balance is ºw
shown in Fig. 7. One arm is Y- FIG. 7.
graduated in IOO divisions and
provided with a sliding weight, so that from hundredths of a gram up
to I gram can be weighed out directly, whilst larger quantities are
balanced by ordinary weights placed in the scale pan.
These hand-scales are usually only employed for loads up to 30 g.
and to an accuracy of IO mg.; they are also made to carry IOO g. and over.
In using them, they are held above at the handle, so that the
silk cords are taut and the scale pans just rest on the table;
by then gradually raising the balance a little it is easy to see if
it is in equilibrium, and if not, equilibrium is established by means
of small pieces of paper, shot, garnets, etc. The substance to be
weighed is always placed directly on the pan. If the physical con-
dition of the substance does not allow this to be done, it must be

20 GENERAL METHODS USED IN TECHNICAL ANALYSIS
weighed on a simple chemical balance. Hand-scales are not suitable
for weighing into beakers, watch-glasses, etc. Their chief advantage
is the rapidity gained in the weighing out of dry materials.
For calibrating litre flasks, and for many other purposes, a balance is
required which can be loaded up to 2 kilos and turns with 50 mg.; it
can be used for many other analytical purposes also, e.g., in the estimation
of moisture in coal, coke, salt, etc., where IOO g. or more of substance is
weighed out. No matter what kind of balance is employed, the
greatest possible rapidity in weighing is always to be aimed at ; the
necessarily numerous daily analyses cannot possibly be made if five or
ten minutes are taken up for each weighing.
Many simple artifices are accordingly made use of to effect
this. In weighing on Sensitive balances, crucibles, watch - glasses,
beakers, etc., the weight of which is known to I-2 mg., are used, and this
weight is written either on the glass itself, or the vessels are marked
with successive numbers, and a note of the weights corresponding to the
individual numbers recorded. In the weighing of platinum crucibles,
etc., which are in frequent use, the actual weighing is really only a check,
involving a correction of the weight within the limits of a few mg.
If one and the same vessel is very frequently used in weighing, it is
very convenient to cut a tare for it out of copper, nickel, or aluminium
foil, which is placed in the other scale pan. With suitable precautions
this procedure, which saves much time, may be employed in weighing
platinum crucibles for gravimetric analysis.
When the weighing out is not done on the horn scale pan itself,
- small beakers, watch-glasses, etc., are used,
and in the case of alterable substances,
stoppered weighing bottles, preferably such
as can stand on the pan. Scale pans made
of glass, aluminium foil, etc. (Fig. 8), from
which the substance can easily be delivered,
are very convenient.
Glazed paper is also often used. In all cases the appropriate tare,
which should be checked, is placed ready on the other scale pan.
In works laboratories it is exceedingly convenient to have a single
weight, made of aluminium or nickle foil, and adjusted by filing down,
as the tare for any very frequently occurring quantities. In order, e.g.,
in the analysis of bleaching-powder (cf. p. 22), to be able to read off
percentages of available chlorine directly on the burette without calcu-
lation, 7.O90 g. Of the sample must be weighed out; for this purpose a
piece of lead of exactly this weight, which is clearly marked on it, is
prepared and kept ready in the balance case.
R. Hase" has constructed an analytical balance, which, by means of
* Z. angew. Chem., 1898, II, 736.
FIG. 8.

WEIGHING 21
a system of levers, at once gives the weight to a decigram or even to a
centigram, so that it is only necessary to complete the adjustment
by means of the rider; the complete weighing can thus be very rapidly
carried out. This is particularly advantageous for weighing out
definite amounts of substance, the desired quantity being obtained
directly from the indication of the pointer on the scale, without having
to bring the balance to rest.
For the actual weighing off, a small horn, porcelain, or glass spoon
is employed, and the number of spoonfuls taken to approximately reach
the required weight observed. Daily practice leads to very great skill
in this respect, so that the proper weight, to within a few mg., is often
hit off at the first attempt.
The weighing out of exactly O-5, 1, 2, 5, IO, 20, 50 g., etc., has the
advantage that the analytical results are easily calculated to percent-
ages; this method can be much recommended, and has been generally
introduced in works' laboratories.
In order to transfer the weighed substance completely from the
watch-glass, beakers, etc., to the vessel for the analysis, a fine brush with
a handle about the size of a penholder is employed; the greater part
of the substance is transferred by tapping with the brush, the remainder
by brushing the glass.
When the percentage of moisture is to be determined, the
substance is weighed directly into the same dish that is placed in
the drying oven, and subsequently weighed ; deep vessels, such
as beakers, are not suitable for this purpose, because they do not
sufficiently permit the evaporation of water by renewal of the air.
When a substance is to be ignited, it can be weighed directly in the
platinum or nickel basin to be used ; polished iron basins can be
employed if a less degree of accuracy is required.
As shown by J. Wagner," a correction for weighing in air is not
necessary, either in respect of the varying specific gravities of the
different substances used in volumetric estimations or of the brass or
platinum weights used, as the errors are inappreciable.
For very accurate work, weights made of rock crystal are employed;
their adjustment requires special precautions.”
One very useful method of weighing out which is often employed in
technical laboratories, is to weigh out such an amount of substance (or,
in certain circumstances a multiple of the amount) that the number
of c.c. of volumetric solution used, or, in gas volumetric analysis, the
amount of gas evolved, gives a percentage value either without any
calculation or by means of a simple mental multiplication or division.
Many examples of this are given subsequently, so that the following
* Massanalytische Studien, p. 37.
* Cf. Göckel, Z. ſ. chem. Apparatenkunde, 1925, 1, 76.
22 GENERAL METHODS USED IN TECHNICAL ANALYSIS
will suffice as a type of the method. In the analysis of bleaching
powder the “available” chlorine is required; Penot's arsenious acid
solution is made up to correspond to O. I g. molecules, or 3-545 g. chlorine
per litre, or O,OO3545 g. per c.c. Consequently, 20 × O-3545 g. = 7.O90 g.
bleaching powder are dissolved in a litre, and 50 c.c. = O-3545 g. taken for
a titration ; then each c.c. of the solution indicates I per cent. of avail-
able chlorine. For the estimation of calcium carbonate in à sample of
limestone by measurement of the evolved gas, since I c.c. carbon
dioxide at O’ and 760 mm. weighs 1.9766 mg., this corresponds to
4,497 mg. calcium carbonate; accordingly O-2248 g. Of the limestone
is weighed out, so that each c.c. of carbon dioxide obtained corresponds
to 2 per cent. of calcium carbonate in the sample.
C. SOLUTION. FUSION. EVAPORATION
In many cases the substance can be weighed directly into the
vessel in which it is to be dissolved or fused ; this saves time, and
avoids the possibility of loss during transference
from the weighing bottle to the vessel employed for
the solution or fusion. It is, however, very impor-
tant to select a vessel that completely fulfils its
purpose, even if such direct weighing is thereby
prevented. -
The vessels chosen should in almost all cases be
adapted for heating; if this is the only consideration,
beakers, porcelain basins, etc., suffice, but if there is
any risk of loss from spirting, suitable precautions
are necessary. The employment of
very large basins or beakers as a
means of avoiding loss from this
cause is deceptive, and leads in other
respects to both inconvenience and inaccuracy. Such
operations should be carried out in narrow beakers or
in Erlenmeyer flasks, which are kept covered during the
process, either by clock-glasses bored through the middle,
or by funnels cut off at the neck (Fig. 9), and which are
subsequently washed both inside and outside with a
few drops of water. Ordinary clock-glasses are not
suitable, since any considerable evolution of vapour may easily raise
them to one side, and thus permit a loss of liquid. The less the amount
of liquid used the better, as the evaporation is thus either shortened or
avoided altogether.
When there is not much liquid to evaporate off, the evaporation is
best carried out in the vessel used for solution, especially when previous

SOLUTION. FUSION. EVAPORATION 23
filtration is not necessary, and a precipitation, etc., is to be effected in
the same vessel. With larger amounts of liquid, and especially if con-
centration at the boiling point is too
rapid, vessels with a large surface
area, such as basins, are employed,
covered preferably by a funnel of the
form shown in Fig. IO; or a current
of air is caused to pass over the
surface of the liquid by placing the
basin in front of a slit leading into a
chimney (Fig. I I); an adjustable
glass plate is fixed over the slit in
such a way that it extends beyond
the basin, and a space of I-2 cm.
should be allowed between the glass
plate and the basin.
As a substitute for wire gauze
and sand baths, Fritsch and Venator *
have suggested plates of aluminium e
3-4 mm. thick, heated either by a FIG. I.1.
ring Fletcher burner or by a Bunsen
burner, according to the size; Lunge recommends these plates as
being very practical.
The question sometimes arises of avoiding contact with the Sulphur
dioxide and carbon dioxide formed during the combustion of the
flame; this can be effected by the method of heating described
below (p. 243).
iii.
L --> *
D. PRECIPITATION, WASHING, AND FILTRATION OF PRECIPITATES
Regarding precipitation, there is little to add to what is known from
general analysis. In works’ laboratories it is usual, and quite rightly so,
not to be so scrupulous about the long standing after precipitation and
before filtration as is often customary in scientific laboratories. Accurate
investigations led long ago to the conclusion that even in cases where
formerly it was recommended to allow a precipitate to stand twelve to
twenty-four hours, quite as good results are obtained if the liquid is
allowed to clear, which it does after from half an hour to two hours,
and then filtered immediately. For instance, in the determination of
phosphoric acid by molybdic acid, in the precipitation of sulphuric acid
by barium chloride, and even in the precipitation of phosphoric acid by
magnesia, etc., prolonged standing is unnecessary if definite conditions
are observed ; in the precipitation of sulphuric acid, for example,
* Chem. Zeit., 1900, 24, 286.

24 GENERAL METHODS USED IN TECHNICAL ANALYSIS
the solution must be kept boiling briskly, and the chloride of barium
added hot, without interrupting the boiling, in order that the
precipitate may form in a granular condition. In other cases the
Solution must be continually stirred during precipitation; this is now
much facilitated by the use of mechanical stirrers.
In scientific laboratories it is very convenient to prepare reagents in
solutions of definite strength, and this is still more important in technical
laboratories, as it expedites the work very greatly if it is known before-
hand how much of a solution is required for a precipitation or any
other reaction. It is naturally very important not to add too small
an amount of the precipitant, but an excess may also be deleterious if
the reagent is carried down by the precipitate, as is the case with
barium chloride in presence of barium sulphate, or if the latter is soluble
in an excess of the former, etc. This is avoided by the use of solutions
of known strength, and by always taking the same or approximately
the same weight of substance for analysis; a definite, measured volume
of the reagent, sufficient for complete precipitation but not very greatly
in excess, is added. In this way a single control test suffices, instead
of having possibly to make a number of check tests, which always give
rise to delay in stirring and allowing to settle.
In washing precipitates, a very great advantage is in most cases attain-
able by using almost boiling water. With proper manipulation, so that the
liquids are not allowed to get cold, filtration proceeds very much more
quickly than is otherwise the case, and much less wash-water is required.
By the observance of the above precautions a skilled technical chemist
can effect a complete and accurate analysis of pyrites, including fusion,
filtration from the residue, precipitation of the iron, of the barium
sulphate, drying and igniting the residue, in from two to three hours,
instead of requiring two or three days, as is often the case in scientific
laboratories. To avoid the filtration and washing of a precipitate,
when the filtrate only is required for further investigation, a method
more frequently adopted in technical than in scientific labora-
tories consists in making up the liquid to a certain volume and
pipetting off an aliquot part, or, as an alternative, pouring it through a
dry filter paper. This plan expedites the work very much, but it is to
be borne in mind that a certain error arises, owing to the volume
Occupied by the precipitate, which must be taken into account in accurate
work, and also that in many cases more of the soluble constituents are
retained, owing to “adsorption,” etc., than corresponds to the interstices
of the precipitate. Further, the filter paper may remove certain con-
stituents from the liquid by adsorption; in such cases filtration must be
omitted, and the clear liquid separated from the precipitate by careful
decantation. If the solution is acted upon by the air, it must be pipetted
directly from the graduated flask containing the precipitate. It should
PRECIPITATION, WASHING, AND FILTRATION 25
hardly be necessary to emphasise the importance of the pipette being
carefully calibrated in relation to the measuring flask used, but
errors arising from the omission of this precaution are of frequent
OCCUlrren Ce.
The funnels selected for filtration must be of the right angle (60°), so
that the filter paper can be accurately fitted, otherwise the liquid passes
through too slowly; filter papers should be of close texture, and allow
of rapid filtration. For analytical purposes, cut filter papers rendered
ash-free by treatment with hydrochloric and hydrofluoric acids are now
almost exclusively employed; pleated filter papers
are not to be recommended, even for technical labora-
tories, on account of the very great difficulty in wash-
ing them out, except when no washing is necessary,
as, e.g., if the liquid has been made up to a definite
volume and an aliquot part of the clear solution is
to be employed for analysis. *
Filter pumps are not generally used in technical .
laboratories except in cases in which the precipitates
are particularly difficult to deal with ; this is because
a large number of filtrations are usually carried on
simultaneously, and the various plans to avoid the tearing of the filter
paper are too troublesome under these conditions. Every laboratory
ought, however, to possess a filter pump, of which there are very many
forms; a thick-walled “filter flask” (Fig. 12) with side tube should be
used for collecting the filtrate, in order to avoid accidents
due to pressure.
For most purposes sufficient acceleration of filtration is
attained by connecting the funnel, by means of a short rubber
tube, with a “fall tube,” 20-25 cm. long, preferably provided
with a loop (Fig. 13); this device is obviously no help unless
the filter paper fits the funnel so well that the suction pro-
duced by the fall tube acts through the precipitate and no
air enters from the side. Practically the same object is
attained by the use of funnels with very long and fairly
narrow tubes.
The Gooch crucible (Fig. 14) is now used instead of the
ordinary funnel in many cases. As it has not yet been
adopted in all laboratories, at least not to the extent that it
deserves, some details regarding it may be useful. It is
made either of porcelain or platinum, and has the bottom
perforated by a number of small holes; if the precipitate has to be
ignited as well as dried, as in platinum determinations, the crucible is
fitted with a smooth dish-shaped tray. An asbestos filter is con-
structed on the perforated base of the crucible by first introducing a
FIG. 13.


26 GENERAL METHODS USED IN TECHNICAL ANALYSIS
layer of asbestos in long fibres and then short fibre asbestos, both of
which have been previously washed with concentrated hydrochloric
acid and subsequently with water; a perforated plate of platinum or
porcelain is placed over the asbestos, and more fine-fibre asbestos above
the plate. The crucible a is fixed in an adapter b by means of a thin-
walled wide rubber tube, and the neck of the adapter inserted in the
neck of a filter flask c, to which a filter pump can be applied at d. Dis-
tilled water is first passed through to wash away any loose threads of
asbestos, the crucible is then taken out, dried under the same con-
ditions as are subsequently adopted with the precipitate, and weighed;
the filtration is then carried out as with an ordinary funnel, the precipi-
tate being collected on the layer of asbestos. After filtering and wash-
ing, the crucible is again placed in the drying oven, and dried at the
desired temperature; if it is necessary to heat over a free flame, the
supporting tray must be employed. Many estimations can be carried
out successively with the same filling of asbestos, especially if the upper
part of the contents of the crucible is removed without injuring the
asbestos filter.
An improved form of the Gooch crucible has been described by
Vollers.1 * -
For filtering solutions whilst hot, the well-known hot-water funnels
are mostly used. A slightly modified form, arranged for use with water,
is shown in Fig. I 5 ; the form shown in Fig. I6, consisting of a coil of
* Z, angew. Chem., 1905, 18, 1272.

PRECIPITATION, WASHING, AND FILTRATION 27
thick-walled copper or lead tubing, heated by a current of steam, is also
very convenient.
Filtration of the precipitate is avoided by the use of centrifugation,
which saves both drying and igniting, and thus leads to -
a great saving of time, but the method is only applicable
in quite special cases. The precipitations are carried
out in vessels of the form represented in Fig. 17, which
have at the bottom a narrow tube closed at the lower
end, on which an empirically constructed scale is etched,
which shows the volume occupied by the precipitate of
a definite substance for a definite unit of weight when
the precipitate is reduced to the smallest possible
volume by means of centrifugal force. The scale shows either the
weight of the precipitate in mg., or directly the percentage of the con-
stituent sought, when a definite amount of the original sub-
stance is employed; the precipitation is effected in the vessel
itself, and the measurement is made whilst the precipitate
remains in contact with the liquid.
The centrifugal method has been applied to the estima-
tion of phosphorus, more especially in iron and steel, the
precipitate of phosphomolybdic acid being measured, to the
determination of urinary deposits, of cellulose in food stuffs,
etc., and to a still greater extent to the estimation of fat in
milk, of water in butter, and to the rapid separation of
liquids containing bacilli and the like; tubes of different
construction to the above are used for these latter purposes.
FIG. 16.
E. DRYING AND IGNITION
The drying of washed precipitates in the funnel gives no
trouble when a drying oven is available, but it involves some
delay, which is a consideration if the result is required
quickly. The time required for drying can be very considerably cur-
tailed if the funnel is placed on a lead cone or broken beaker, and heated
directly over an iron plate or wire gauze, but there is a risk of cracking
the funnel in this way. It is therefore preferable, when practicable, to
burn the precipitate wet in a platinum crucible, a method which is also
largely employed in scientific work.
For this purpose the filter paper is carefully removed from the
funnel and laid on a piece of coarse filter paper; by moving it
repeatedly on to fresh dry parts of the latter and pressing very gently,
the greater part of the adhering moisture can be removed in a
few seconds. The precipitate and paper are then transferred directly
into the crucible and ignited.


28 GENERAL METHODS USED IN TECHNICAL ANALYSIS
This method can be safely employed for such precipitates as barium
sulphate, magnesium ammonium phosphate, calcium oxalate, silicic
acid, etc.
The direct burning of the filter paper is at least as accurate as the
indirect method of drying and then igniting, as the loss of substance
which occurs in removing the dry precipitate with a feather or platinum
spatula, and in burning the paper in a spiral of platinum wire, or on the
lid of a crucible, is practically excluded; also, it has been shown that the
combustion of a wet filter paper is more complete than that of a dry one,
and avoids the formation of tar-like products of carbonisation.
If the external appearance of the ignited precipitate is suspicious, it
is generally possible to apply an experimental check. Thus ignited
barium sulphate, which is grey instead of white, should be treated with
sulphuric acid, evaporated, and again ignited ; similarly, if magnesium
pyrophosphate shows a small dark patch, it should be treated with nitric
acid (repeatedly if necessary) until the mass is pure white or greyish-
white. When the precipitate is burnt wet it rarely shows anything
abnormal in its condition.
If a precipitate has to be dried at Ioo”, I Io”, or other definite
temperature, it leads to great loss of time to first dry the empty filter
paper, and then the paper and precipitate, till the weight is constant.
The following method, which is much in vogue in technical laboratories,
is preferable.
Two filter papers, as nearly as possible of the same diameter, are
taken, one laid on each pan of a balance, and small pieces cut off from
the heavier one until they balance each other accurately. The precipi-
tate is then filtered on to one of the papers, dried first on filter paper, as
described above, and the two filters placed in the steam oven for from
one to two and a half hours, according to the nature and size of the
precipitate; in weighing, the empty filter paper is placed as a tare against
the one on which the precipitate has been collected.
It has been found 1 that a paper dried at 120° gains several mg. in
weight if subsequently washed with alcohol and dried at I IO’, as is usual
in the determination of potassium ; when 90 c.c. of alcohol are used for
washing, the increase of weight amounts to 5 mg, which must be
deducted from the weight of the potassium platinum chloride.
Rudorff’ recommends placing the filter paper in a cylindrical weigh-
ing bottle, and drying it for thirty minutes in the drying oven, after
removing the stopper of the weighing bottle; the stopper is then
replaced, the bottle allowed to cool in the air for thirty minutes, and
then placed in the balance case for ten minutes before weighing. After
the precipitate has been transferred to the filter paper, it is similarly
dried and allowed to cool before weighing.
'J. Stroof, Private communication. * Z, angeſt, Chem., 1890, 3, 633.
HEATING APPARATUS 29
A very clean but somewhat expensive drying cupboard with
removable porcelain plates is described by Treadwell."
In works' laboratories ignitions are done in platinum crucibles
for quantitative work; it is best to employ a solid tare in the weighing
(cf. p. 20). The ignition of larger amounts of substance when it is
unnecessary to weigh beyond centigrams, as in estimations of water,
etc., can be performed in polished iron basins. &
Except when a hygroscopic substance has to be dealt with, the
crucible may as a rule be allowed to cool on a pipeclay triangle or on a
block of marble instead of in a desiccator, and the cold crucible may be
handled with clean fingers instead of with crucible tongs.
Those forms of desiccator are to be recommended which allow a free
passage of air, and are accordingly provided with a calcium chloride
tube, and in some cases with a soda lime tube in addition.
To clean platinum crucibles, fine sand (silver sand) is stirred up
with commercial hydrochloric acid,
and, after the acid has repeatedly
been poured off, the crucible is gently
rubbed with a portion of the moist
sand ; in this way the cleaning is
rapidly and efficiently performed.
F. HEATING APPARATUS
Gas is the most convenient and
cleanest source of heat, and when
available is naturally preferable to anything
else. Beyond the ordinary Bunsen burner the
following more special burners are useful.
I. Burners with a horizontal bent tube
(Fig. 18), known in France as the Berthelot
burner. Burners of this form which turn on a
hinge are also made. They are advantageous
in cases where boiling over may readily occur,
and where there is not sufficient height for a
Bunsen burner. -
2. Müncke's burner (Fig. 19) is particularly
suitable as a substitute for heating in the
blowpipe flame. The temperature is sufficient
to melt sodium carbonate in a few minutes,
so that the fusion of silicates and similar
operations can be performed without a blast,
and there is less risk of substance being
carried away mechanically.
* Analytical Chemistry, vol. ii., p. 22.


30 GENERAL METHODS USED IN TECHNICAL ANALYSIS
3. The Tecluburner (Fig. 20) has similar advantages to the fore-
going, and by means of its various adjuncts permits of the regulation of
the heat. - - -
4. The multiple-flame burner, of which Fig. 21 shows one of the
FIG. 20.
many forms, serves for heating tubes as well as for the simultaneous
heating of several separate vessels. A very useful form of this burner,
devised by Wiesnegg, is
shown in Fig. 22, as used
for heating a muffle.
The various forms of
Fletcher gas-burners serve
more for general laboratory
purposes than for analytical
operations, and are very
useful in certain technical
tests. For high tempera-
tures up to a full white heat
gas ovens such as those of
Perrot (Fig. 23), Seger,
Rössler, and others are
employed.
When heating over the
naked flame is inadmissible,
wire gauze or asbestos boards
are used. In technical labo-
ratories, where many vessels have to be heated simultaneously, a flat
iron plate covered over with a few mm. of sand is very convenient ;
another and preferable arrangement consists of two pieces of sheet
FIG. 22.


HEATING APPARATUS 31
iron held about 5 mm. apart by rivets, thus forming a hot-air bath.
Aluminium plates (p. 23) can be similarly used.
FIG. 23.
These arrangements serve for heating to temperatures from 1 Io’ to
2OO° or higher. For heating up
to not more than IOO", a steam
bath or steam-heated sand-bath
can easily be fitted up in most
works; the latter are particu-
larly suitable for digesting with
low boiling inflammable liquids,
such as alcohol, benzene, ether,
etc.
Where gas is not accessible,
recourse to spirit-lamps, char-
coal furnaces, etc., was formerly
necessary. Now, other and more
convenient heating appliances
are available which give suffi-
ciently high temperatures, such
as the alcohol burner (Fig. 24),
FIG. 24.
Barthel's petroleum burner (Fig. 25), etc.; petroleum tube furnaces


32 GENERAI, METHODS USED IN TECHNICAL ANALYSIS
for combustions are also obtainable. These burners are free from
danger, and are said to give a higher temperature than gas furnaces.
Henri St Claire Deville
has designed a universal
furnace (made by Wiess-
negg, Paris) for use with
heavy mineral oil, which
serves as a muffle tube
furnace or crucible fur-
is nace, and gives a tempera-
H ture of 1300°.
< Still higher tempera-
- - - - tures are obtainable by
FIG. 25. means of coke furnaces
and electrically heated fur-
naces. A description of these forms of apparatus falls beyond the
scope of this section.
G. VOLUMETRIC ANALYSIS 1
Volumetric analysis includes “titration analysis” or volumetric
analysis in the restricted sense, which consists in the application of
standard solutions to the analysis of liquid and solid substances,
gas-volumetric analysis, which consists in the valuation of a liquid or
solid by the measurement of an evolved gas, and gas analysis. These
three branches of volumetric analysis are all very extensively employed
both for scientific as well as for technical work. Many of the methods
are equally serviceable for both purposes, whilst others have been
specially worked out for technical analysis. The latter are described
in connection with the respective industries; a special section is devoted
to technical gas analysis.
The Calibration and Standardisation of apparatus used
in volumetric analysis.
In all branches of volumetric analysis it is of primary importance
that the vessels employed should be accurately graduated, just as the
first requirement for gravimetric analysis is that the balance and
weights must be accurate. It has long been customary to check the
weights used in analytical work, but the calibration of volumetric
* The data on volumetric analysis given in this and subsequent sections comprise a number of
special methods which are of technical value, and which are not usually dealt with in the ordinary
text-books on quantitative analysis.
For the history of the subject, cf. L. L. de Koninck, Historiyue de la méthode titrimétrigue,
published by Havrez, Brussels, 1901; and, Bull. de l’Assoc. Belge de Chimistes, 1901, 15,
November and December.

CALIBRATION OF APPARATUS FOR WOLUMETRIC ANALYSIS 33
apparatus is far less general; the graduations and capacity of the
apparatus is too often taken on trust, especially by technical chemists.
The calibration of volumetric apparatus used in the analysis of
products which are bought or sold should on no account be omitted ;
less accurate apparatus suffices for use in the control of working
processes, but this should at least be submitted to a comparative
standardisation if not actually calibrated. Apparatus for volumetric
analysis can now be bought which has been tested and stamped by
the National Physical Laboratory or by the Imperial Normal
Standards Commission in Germany.
In the great majority of cases the apparatus tested in such institutes
may be relied upon as possessing the degree of accuracy correspond-
ing to the tests used and guaranteed by the certificate of examina-
tion. But this does not absolve the skilled chemist from the duty of
taking cognisance of the character of these tests, or of himself
making further tests in cases of very special importance or of doubt,
just as in similar circumstances he may require to make further tests
of weights adjusted by a mechanic. Also the cost incurred in the
official standardisation may, in case many instruments are used, induce
some to prefer doing the work themselves. This, like every other form
of manipulation, requires a degree of accuracy which is obviously much
greater in the hands of the officials of a standardising institute than in
those of the skilled chemist, but the latter should certainly be able to
do the work, and may prefer to do it in some cases. With this in
view, the following details of the methods employed are given.
The conceptions as to what is to be understood by standardised
vessels and what is to be required of them are very varied, and in some
respects obscure. There is a considerable amount of literature on the
subject" on which the methods and considerations that follow are
chiefly based ; the work of Schloesser” is especially important.
In the first place, the unit taken as a basis for the standardisation
must be clearly defined. From the scientific standpoint the litre is the
space which one kilo of water, weighed in a vacuum, occupies at its
maximum density. Since this corresponds to a temperature of 4°, the
work must either be always carried out at this temperature, which is
naturally not practicable, or at a suitable higher temperature to be
definitely determined, the corrections to be made being ascertained
either by calculation or from tables. This is essential in order to
ascertain the value of the “true litre” and its subdivisions, regard being
had to the temperature of the water and the buoyancy of the air, which
* Cf. J. Wagner, Massanalytische Studien, Leipzig, 1898; and Z. physik. Chem., 1899, 28, 193;
Göckel, Chem. Zeit., 1901, 25, Io84; 1902, 26, 159; and Z. angew. Chem., 1902, 15, 707; 1903,
16, 49 and 562.
* Z, angew. Chem., 1903, 16, 953, 977, and 1904; Chem. Zeit., 1905, 29, 509.
34 GENERAL METHODS USED IN TECHNICAL ANALYSIS
depends on the temperature, the pressure, and the degree of moisture
present. In order to avoid the calculations necessitated by the employ-
ment of the true litre, Mohr, as early as I 855, introduced as the unit
the volume which IOOO g. of water Occupies when weighed in air with
brass weights at 14° R. = 17°. 5 C. This value is called “Mohr’s litre,”
and the designation is also extended to the use of other temperatures
(15°, 20°, or higher). It is immaterial what this temperature is, so long
as all utensils employed in connection with one another are standardised
for the same temperature, and are subsequently used at the same or
about the same temperature. In order to ensure this, the standard
temperature is etched on all the utensils in question; this temperature
is, therefore, the same for the vessel and for its contents (water,
mercury, etc.). The “Mohr's litre" and its subdivisions were for
many years in almost general use for volumetric analysis both by
makers and by users of apparatus. Its convenience was then un-
deniably very great, since, in calibrating, all calculations which were
formerly not facilitated by tables could be dispensed with.
Unfortunately, however, the desired agreement is not obtained even
by calibrating at a definite temperature, since the density of the air at
the time of the calibration must also be taken into account. If, for
instance, we take the pressure of the air = 76O mm., and the temperature
17°.5, calculation shows that one “Mohr's litre”= IOO2.3 true c.c., and one
true litre = 997.7 “Mohr’s” c.c. If, however, the calibration were carried
out at a pressure of 700 mm., the difference due to the change in pres-
sure would make a difference in weight of almost O. I g., and the volume
if determined subsequently at the higher pressure would be found too
small by this amount. This source of error is also not avoided by fixing
a normal pressure of 760 mm. for the Mohr litre, since the buoyancy of
the air depends not only on the pressure, but also on the temperature
and the degree of moisture present.
Taking these considerations into account, the advantages which
Mohr and his successors claimed for the choice of an alternative to the
true litre disappear. The use of tables (p. 46) cannot be avoided, and
the true litre in conjunction with tables (pp. 37 to 38) is, therefore, just
as easily employed. It is, accordingly, very desirable that "the word
“litre” and its subdivisions should be restricted to the true (metric)
litre, and that all chemical measuring utensils should be based upon it.
At the same time it must be recognised that much apparatus standardised
according to Mohr’s system is still in use; the tables given below can be
used to determine the deviations of such vessels from true litre measures.
The calibration of measuring flasks should always be carried out with
the same kind of liquid as that with which they are to be filled when in
use, so that the conditions of wetting, the meniscus correction, etc., are
similar; apparatus to be used with mercury should always be calibrated
THE TRUE LITRE 35
with mercury, whilst burettes, pipettes, and measuring flasks are cali-
brated with water, which does not differ essentially in its behaviour from
the very dilute volumetric solutions generally used in these vessels.
To avoid all the errors mentioned above, the only means is the
employment of the true litre; as already stated, this is the volume
which I kilo of water at 4° occupies under standard pressure. If this
space is to be marked off, e.g. on a flask, the position of the mark will
depend on the temperature of the flask. The standard temperature to
which certificates are used by the National Physical Laboratory is 15°
C.; thus the statement that the volume of a one-litre flask is correct
means that at a temperature of 15° the volume of the contents of the
flask is the same as that of a kilo of water at a temperature of 4°. The
tables given below allow the adjustment to be made directly at any
desired temperature and pressure, the weights that must be placed on
the scale pan to effect this being given in each case.
Two true litre measures of glass, adjusted at different temperatures,
differ only by the difference in the expansion of glass between the two
temperatures, if water of the same temperature has been used in testing
them. Therefore it is only necessary to calculate the weight which is
in equilibrium with the weight of water occupying a true litre measure
for one normal temperature, e.g., I 5°. If the temperature of the air does
not greatly deviate from this, and the height of the barometer is not
very far from 760 mm., mean assumptions may be made for the factors
which influence the buoyancy of the air, pressure, temperature, and
degree of moisture, and the reductions thus obtained may be combined
with those due to the temperature of the water. The values in Table
I. (p. 37) can then be employed directly in order to find how the
volume of a true litre should be marked off on a flask. If, for example,
the air and the water have a temperature of 17°, the empty flask along
with a kilogram weight is placed on One scale pan and brought to
equilibrium by a tare on the other; the kilogram weight is then
removed, and on the same side (that is, along with the flask) weights to
the amount of 2.208 g. are placed ; equilibrium is then re-established
by filling the flask with water of 17°, and the volume occupied by this
weight of water marked on the neck of the flask.
In the case of greater deviations of the temperature of the air from
I5°, and of the atmospheric pressure from 760 mm., Table II. (p. 38) is
used to correct the values of Table I. If, e.g., the height of the
barometer is 720 mm., the temperature of the air 25°, that of the water
24°.3, the weight to be added for a litre is :— -
From Table I. tº iº * tº tº 3564 mg.
From Table II. o Q o & gº – 92 mg.
3472 mg.
36 GENERAL METHODS USED IN TECHNICAL ANALYSIS
The weight of the volume of water required to correspond to a true litre
for the flask at I 5° is therefore IOOO – 3:472 = 966.528 g.
For any other normal temperature (t) the magnitude (t–15) O-OOOO27
must be added to the above; thus for a normal temperature of 20° all the
values of Table I. must be increased by IOOO (2O— I 5) O-OOOO27 = 135 mg.
For a water temperature of 20°, 2699-H 135 = 2834 mg. must therefore be
added.
If a different value to the above be chosen for the mean expansion
coefficient of glass, then IOOO (a'-O-OOOO27) (t–15) must be added to
the values of Table I., a denoting the new coefficient of expansion of
glass, and t the temperature of the water. The amount of this correc-
tion is always very small, and is hardly likely to be employed by the
chemist in calibrating.
The Tables I. and II., pp. 37 and 38, were calculated by the Normal
Standards Commission, and communicated by Schloesser."
The following are the methods and conditions adopted by the
National Physical Laboratory for the testing and calibration of glass
vessels.”
Vessels are calibrated either gravimetrically or volumetrically.
When the highest accuracy is required, gravimetric calibration is
employed. Volumetric calibration, however, which is simpler and more
rapid, is in practice sufficiently exact for most purposes.
Gravimetric method.—I. Vessels to be tested for contents. The
vessel to be calibrated is weighed empty; it is then filled with distilled
water up to the mark, and weighed again, the temperature being care-
fully noted before and after the weighings, and the weighings reduced
to weighings in vacuo.
The difference between the weights is the weight of water which
fills the vessel at the temperature of the observation. The weight of
water which would fill the vessel under the given conditions of
temperature, if its volume were correct, can be calculated from a
knowledge of the density of water at different temperatures,
and of the expansion of the containing vessel by use of the Tables
I. and II.
2. Vessels to be tested for delivery. The contents of the flask or other
vessels are poured into a vessel of known weight, and the same
procedure is adopted.
Volumetric method.—The method to be employed depends on the
apparatus which it is required to test. For many purposes automatic
pipettes are found useful. These are first graduated by the gravimetric
* Z. angew. Chem., loc. cit.
* The Editor is indebted to Dr R. T. Glazebrook, Director of the National Physical
Laboratory, for these data. A pamphlet containing full details and charges is issued by the
Laboratory.
CALIBRATION TABLES
37
0868390IZ9[8][989.Igg? I916096909I†09† I091.86%08
63696 †Zg6#g06ý818$I98††ZSŤ161, †01.178Ť 15;9IZŤ6%
8Z689†899ý189ýIL9f;g8 giff6ggiº889 º1097I8 #ffggțý8Z
/36Ziff80ýſ;1.18%Zggiſ;4，857Z08 †117, †ZgŽÍ7ZZZŤZOZ†13
9%/, /, IŤZgIŤ13 IŤZOIį810ſ†g0Ť090 †900ýZ869S96893
9%†8680I689888Ź98968889 I99Ģ6480! 18Ziff/8†Z28QZ
†ZI0188/999999Z99860999899#999Źț98039886ý8† Z
8Z9159†gțgZgŤg0I7888881988Q†889888†08888Z88Z
ZZZ9Z8I†Z80ZZ866 [88.1, I899 I888I88II8860881.08ZZ
IZ890988088I08（ 866%81,6%69630ý6%IZ,6%Z06Z8883Iő
0%†98%gț8Z938%108Z68.1%I 14%892398/41.I.13669%0%
6II89Z9.99%gț9Z1.39%0I9Z869%9.29%69g%@j,9%9.89%6I
8I809ZI6ţ，† 1. ÞZ1,9Ť ZOýț%†ZŤZ,S07%Z68Z9.18%0.98%SI
/, I††84838Z8I8%86ZZ88ZZ89ZZ897. Z88ZZ8ZZZ803%! I
9I86 IZ8A, IZ† 9IZ09 IZ98 IZZZIZ80 IZ†60Z0803990Z9I
ÇI· 890Z0 #0Z1ZOZ#IOZI00%886 I916 I396 I676I186IgI
† IgZ6I£I6 II06 I688 I! 18 I998I898IIŤ8I08.SI6 18I† I
8I808I1,64 I981. Ig/, /, I†91, I89/, I8ț¢ 1.I881. I8ZZI8I/I8l
ZI80ſ. I869 I889 I81.9I† 99 I999I9ț9I189IQZ9I6][9][ZI
III 19I809 U969 I189I619 II 19I899 Iggg I1. ºg I689 III
0II89 I†ZgI1.I9L0I9 I809 I96ïI68ïI88ŤI!,/ ŽIIZ ŽI0I
6†9ţI89ȚIZgțI1țițIZŤțI98ţIIgțI9ZţIIZŤ， I/ If I6
8ZIŤ I80ýI†0ýI00ýI968TZ68I888 I#88I088 I9/8I8
4.Z/8I698I998 I898.I098 I898 I998 I†gĢIZ98 I098 I!
9’Sý8I978IgŤ8I††g I&#8IZŤ8IIý8I0ț8I688 I888 I9
9888I888I888I888I888 I888I888I6$$I()ț8 Il #9 I9
*đuſº I,6.8./*9.9.†.8.3.I-0.*duuº I,
Jo Qanssa. Id ue3.JN
".$ I ºu ſe 9ų). Jo aunqe 19du 19.L
4
‘uuuu O94 %1949Uuou eq Jo qų319 H （6.0%-,9 uuouſ ºlnų euadu.19) 19ņēAA
ºungeuaduI3, IeuuuoN ，O „o uad ZzOOOO-o=sseĮ3 Jo uoſsuedx9 go queſ0.ggºoo oſqno 9ų) { -3-9 OOOI JOJ pºppe aq og ºut uſ !ų313AA




ºuņįI 19d ‘3tu 3= Inodea Snoonbe
'I GITIĶIVL
• 9 I
3
I9 || 6 I 83 9% 09 8/ 98 66 ZII 9&I 83 I Z9 I 99T 6/I 86 I 90% - || I9
03 gl 63, Ziff 99 69 38 96 60I 3&I 99T 6?I 39T 9/I 06I 80% 03
6% II 93, 88 Zg 99 8/. 36 90I 6II Č3 I 97I 09 I 3/I 18|I 003, 63,
83 8 I6 98 8? 39 9/ 68 ZOI 9II 6&I giff I 19| 0/...I #8I /6I 86
A3 f 8I I9 g? 89 I/ 98 86 ZII 9&I 69 I 99T 99T 08I 86 I 16
92. I — #1 – | 82 – | If – gg — 89 - Z8 - 96 60I - || ZZI- 99T – Og|I – 39T – // I - 06 I - || 93
g3 | g + OI – | #2 /8 Ig g9 8/. 36 Q0T 6II 99T Zip I 09 I #/I /8I 23,
#2, Z + | 9 - || 0& 88 Af I9 9/. 88 30I 9 [I 03I ## I /9I IZI #8I fº,
8% II + | g – || 9 I 09 ## 89. Il 98 86 &II 9%I Of I #9T /9I I8I 36
33, #I + | T + || 3I — 92 Of #g 89 IS 96 60I 8&I A8L 09 I #9T All ZZ
Ić, 8I+ | p + | 6 – 83 – 19 — Ig — #9 – 8/. I6 – gOI – 6II – || 38L – ºf I – I9'I - || 71. I – I&
0% Ø, 8 + | g – | 6T – || 33 Ziff I9 #/. 88 30I 9II 03 I gi/I M.91 IZI 0&
6I 93, &I + | I – | g I – | 62, gif /9 I/ 98 66 3II /.3T. Of I #9 I 89 I 6I
8T 03 9I + | g + | II — gº, 69 99 /9 I8 96 60I 8&I /8T I9 I 99T 8T
AI #9 OZ + | 9 + | 8 – ZZ 98 09 #9 8/. 36 90I 0&I 88.1 Ziºl I9 I /...I
9I 83 + | PZ + | OL + | p – 8I – Z3 – 95 – || 09 #1 – 88 – ZOI — 9II – || Og|I – || Wºl - | 89 I- 9T
g|I Číž 8% #I O #I — 8% Číž /9 I/ 98 86 ZII A&T If I 99 I gT
#I 97 38 8I # + | OL — #3, 83 39 /9 I8 #6 80I 3&T 88.I 39T # I
3I 09 98 66, 8 + | 9 – 0% fg 6? 39 Al I6 Q0I 6II #3L 8?I 3I
&I 99 If 93 ZI-H | Z — 9I I9 gº 09 #/. S8 ZOI 9IT 18T g; I ZI
II 69 + | g F + | 08 + || 9 I + || 3 + ZI – A3 - If 99 - || 0Z – #8 – 86 - || ZII – ZZI- If I – II
OI 89 65 #3 0% 9 + 8 8% /8 Z9 99 08 96 60I #6, I 88.I 0 I
6 /9 39 88 #6, OI + # — 6 L 88 8? 39 Al T6 90I 0%I 98.I 6
8 &/ A9 37 8% #I + 0 GI 6% ## 89 8/. /8 30I 9II I31 8
4. 91. I9 Ziº 38 8T + 3 + II 93, Of 99 0/. £8 66 §II 8%I 1.
9 08 99 I9 A8 ZZ + 1 + A. 36, 99 Ig 99 08 Q6 60I #&I 9
o9 #8 + || 0 + | g g + | If + | 93 + II + 9 – 8I 29 – 17 – Z9 - || 11 - || I6 – 90I - || I&I - || 29
"duel, ‘dulo I,
UIIN *06/, *08/ "011, '09/ "Og/ “Ofſ, '08/ "031, ‘OI!. "001, *069 ‘OS9 *019 ‘099 °099 *UWI

“uu oë4-059 “ainssaid ouaudsouye Ioj 'o'o oool Ioy 3UI uſ 'I atav Lui senſea au, Jo uoupe.I.IOO

'O 219-.5 ‘IIe aul Jo ainleiadual pue
'II GITQIVIL



STANDARD METHODS OF CALIBRATION 39
method, and then form standards by the use of which other vessels can
be calibrated.
Methods of measurement.—(1) Flasks for contents. In the case of
flasks up to IOO c.c. inclusive, the temperature of the water is determined
after the weighing; in the case of larger flasks, both before and after the
weighing. If the neck of the flask is too narrow to admit the ther-
mometer, the temperature may be taken in an auxiliary vessel. The
adjustment of the meniscus is made after the first temperature
measurement.
(ii) Flasks for delivery. The flask, filled to the mark, is inclined, as
its contents are being poured into the tared glass, until, when the
continuous stream of water has ceased, it stands nearly vertical. In
this position it is allowed to drain for one minute into the glass, and
the last drop is removed from it by touching against the side. The
temperature is taken immediately before the adjustment of the
meniscus.
(iii) Pipettes with one mark. The pipette is clamped vertically, and
by means of a rubber tube is filled from the jet to about 1 cm. above
the mark. The suction tube is closed by the finger, the water-delivery
tube taken away, and the level of the meniscus adjusted to the mark.
Any drops adhering to the jet are removed. The water is then allowed
to run off into a tared glass, the point of the jet touching the wall of the
vessel. After the continuous outflow has ceased the pipette is allowed
to drain for fifteen seconds, and the jet is then stroked off the wall of
the vessel. The temperature is taken on a sample collected in a special
glass before the filling of the pipette, and on the water run out of the
pipette after the weighing.
(iv) Pipettes with two marks. The method of procedure is the
same as in the foregoing section, except that when the water has flowed
down to within 2-3 mm. of the second mark, the suction tube is kept
closed by the finger for fifteen seconds, and the adjustment to the lower
mark is then made by slightly raising the finger. The last drop on the
point is stroked off.
(v) Overflow Pipettes. These are tested like ordinary pipettes. The
tap must be kept full open during the outflow. Experience has shown
that accurate results can be obtained only when the pipette is filled and
emptied four or five times before the actual test.
(vi) Measuring Glasses to contain. The glass, together with as many
grams as it contains c.c., is counterpoised on the balance. The glass is
removed, filled to the smallest marked volume, and then replaced. As
many grams as the part to be tested contains c.c. are then taken off the
balance pan, and the balance again counterpoised. After this test, the
glass is filled up to the next mark and tested in the same way, and so
on until the complete volume is reached. Each mark is tested from the
40 GENERAL METHODS USED IN TECHNICAL ANALYSIS
zero. In calculating the error between two marks, if a be the error to
the upper mark and b the error to the lower mark, the error between
the marks is a-6.
(vii) Measuring Glasses to deliver. The measuring glass is filled up
to the lowest of the divisions which it is required to test. The water is
then poured into a beaker which has been counterpoised with its cover,
the same regulations applying here as in the case of flasks “to deliver.”
After this the glass is filled to the next numbered division, and the
operation repeated. The temperature is taken immediately before the
first and last filling.
(viii) Measuring Glasses with incomplete divisions. These are usually
tested at five points, viz., at the top, at the bottom, and at three inter-
mediate points. In other respects they are treated as other measuring
glasses.
(ix) Burettes. In all tests the water is run out continuously from
the zero marks. The burette is clamped vertically, and filled up to
5-7 mm. above the zero. The water is allowed to run out exactly to
the zero mark, and any drops adhering to the nozzle are removed. Care
must be taken that no air bubbles are present. The tap is opened
wide and the water run out into a counterpoised vessel; the flow is
stopped when the water is about 5 mm. above the mark to be tested.
After waiting two minutes the water is run out exactly to the mark,
and the last drop touched off against the side of the glass.
Burettes fitted with glass taps should have the tap lubricated
slightly with vaseline. .
(x) Gay-Lussac Burettes. The burette is first filled as exactly as
possible up to the mark. It is then inclined until the first drop flows
from the outflow tube and note is taken of how far the meniscus is
moved backwards. The burette is then filled to the zero, and gradually
emptied into a glass counterpoised with weights corresponding to the
volume to be tested, until the meniscus comes to the same point as
previously. It is then placed vertically, and two minutes allowed to
elapse before the exact reading is taken. Weights corresponding to
the content which is being tested are then removed from the balance,
and the weighing completed in the usual way. -
(xi) Measuring Pipettes. The pipettes are clamped vertically, and
fitted with a pinch-cock by means of a rubber tube. In other respects
they are treated as burettes.
Amongst the general conditions specified, it is pointed out that the
adjustment of the meniscus of the water to the mark of the vessel is
made so that the lowest point of the meniscus is in that plane which con-
tains the mark. In this adjustment the axis of the vessel should be verti-
cal, and the observer's eye placed in such a position that the front and
back halves of the mark coincide. A black background should be used.
STANDARD METHODS OF CALIBRATION 41
Outflow jets should be drawn out straight and the ends should be
smooth and even; when the tap is open full, water must run out
continuously.
The limits of error are as follows, expressed in c.c. :-
(i) Flasks.
For c.c. ſº • 50 IOO to 250 3OO to 500 5OO to IOOO 2OOO
C. C. ſ To contain . . O5 • I • I 5 •3 •5
* \To deliver . . .1 •2 •3 .6 I
(ii) Pipettes.
For c.c. inclusive . g ... 2 IO 3O 75 2OO
C.C. & º e © e •OI •O2 •O3 •O5 • I
(iii) Measuring glasses.
For c.c. inclusive . . IO 3O 5O IOO 2OO 5OO IOOO
C.C ſ To contain . º . .O4 •o6 • I •2 •5 I 2
*\To deliver . e . .o& I 2 • 2 •4 I 2 4.
For parts of the whole, less than half the total content, the limit of
error is half that for the total content.
The limit of error for measuring glasses with incomplete divisions
is the same as for measuring glasses with complete divisions.
(iv) Burettes and measuring pipettes.
For c.c. inclusive . © ... 2 IO 3O 5O IOO
C. C. e & tº * ... •OI O2 -O3 O5 • I
For parts of the whole, less than half the total content, the limit of
error is half that for the total.
Corresponding measurements of flasks for sugar determinations
and for viscosity measurements are also carried out by the National
Physical Laboratory."
There is still much difference of opinion as regards the best method
of delivery and the time interval that should be allowed. According to
J. Wagner, for burettes with glass stopcocks, a total time of outflow of
7o seconds is sufficient, and in the case of burettes with clips, 45 seconds;
it would be preferable to say that burettes should empty in 40-75 seconds,
according to the size. Pipettes can be emptied much more accurately
than the official instructions assume; an accurate although somewhat
tedious method consists in making a determination between two marks
on the pipette, and sufficient accuracy is attained also by blowing out.
As Göckel rightly points out, it is very wearisome to have to wait 2
minutes with measuring pipettes, and it would be better to prescribe an
interval of 4 minute. Ostwald prefers blowing out; others, touching
against the wall of the vessel.
* Note.—The conditions of official standardisation adopted in Germany are given in the German
edition, vol. 1, pp. 47-50. They are also described by Weinstein, Z. angew. Ghem., 1904, 17, 1745.
Cf. also Fischer's Jahresher., 1897, 43, II.46; and Schloesser, Z. angew. Chem., 1903, 16, 977.
e”
42 GENERAL METHODS USED IN TECHNICAL ANALYSIS
According to Schloesser, the method of delivery, whether free or
against the wall of the receiving vessel, is immaterial; the differences
arising therefrom are small. Free delivery with subsequent removal of
the drop seems preferable, however, in the case of burettes, because the
end reaction can be then more readily recognised; this applies also to
pipettes, because there is then no temptation to hold them obliquely,
which is always objectionable. Blowing out, on the other hand,
Schloesser absolutely rejects, an opinion in which Lunge concurs, since
it gives rise to too great differences between different observers. As
regards the time interval to be allowed for delivery, it is only necessary
to wait until the residual liquid has been reduced to a negligible small
amount ; in the case of pipettes, this can be taken as # minute.
Should it be desirable for some special reason to undertake the
calibration of apparatus, this can be done with sufficient accuracy,
and within the official limits of error, by employing a sufficiently
stable balance, turning with O.O5 g., for vessels holding up to 500
c.c., and an ordinary analytical balance and a weighing bottle, provided
with a glass stopper, for smaller vessels. The temperature of the
water employed in calibrating should be determined with an accurate
thermometer to within + O. I*.
The contents of ordinary pipettes are emptied at one delivery into
the weighing bottle, those of burettes and measuring pipettes in
portions of 2-IO c.c. at a time, according to the accuracy desired; the
conditions with regard to the holding of the pipette, removal of the
drop and the interval allowed for draining as practised at the National
Physical Laboratory being observed. If Mohr's system is used, the
weight for the normal temperature chosen must be calculated according
to the Table on p. 46. For the true litre the data given on p. 35 et seq.
are to be observed, and Tables I. and II. (pp. 37 and 38) used for
the necessary corrections. The error of the vessel is then given by
the difference between the calculated additional weight and that
actually required. -
If the vessel is calibrated for delivery, thorough cleaning must be
strictly attended to (cf. p. 50). -
At least two calibrations should be made, and more if there are
large differences in the determinations, and from the mean of the
results a table of corrections is drawn up. In the case of measuring
flasks and ordinary pipettes, it is preferable to alter the graduation
to the corrected position.
If a great many calibrations have to be carried out the delay due to
weighing may be avoided by the use of standard measuring instruments,
although this is less accurate than direct weighing. For this purpose
the Ostwald pipette (Fig. 26) is usually employed. These pipettes
are made to hold 2 or 5 c.c., according to the degree of accuracy
CALIBRATION WITH THE OSTWALD PIPETTE 43
required, and are tested as follows. The pipette is fitted on to the
lower end of the burette, as shown, the burette is filled with water of
the normal temperature, and then by opening a pinch-cock at a the
pipette is filled exactly up to the mark b. A weighing bottle is then
placed below the pipette, the pinch-cock at d opened, and the water
allowed to flow out until it just reaches the mark c, one or
preferably two minutes being allowed to elapse for com-
pletion of the delivery. It is unnecessary to connect the
side tube of the pipette with the jet of the burette, instead
of, as is shown here, with the rubber tubing, if this 1-2
minutes' interval is allowed. The mean of three weigh-
ings of the contents of the pipette, if they agree well
among themselves, is taken as the standard. To calibrate
the burette e, it is filled with water at any desired tempera-
ture, which, however, must remain constant during the
calibration ; then, by opening a, the water is allowed to
rise to c, and the burette filled to the zero mark. By
again opening a, the water is allowed to rise exactly to 6,
and a reading of the burette taken. The water in the
pipette is next allowed to run out to c by opening d,
and the pipette again filled up to the mark & by open-
ing a ; the position in the burette is again read off, and
so on until the burette is empty. An interval of a minute
is allowed to elapse before each reading in e, b, or c. The
mean of two series of experiments is taken, and the actual
capacity of each interval of the burette, corresponding to
the capacity of the pipette, is thus determined. If, for
example, the capacity of the pipette determined by the use of Table I.
has been found to be 2. 1234 true c.c., and if the first pipette delivery occu-
FIG. 26.
© e 2
pies 2.20 c.c. in the burette, each c.c. of the burette=; #,
If the reading of the burette after the second delivery is 4.35 c.c., each
4:35 – 2:2O
2, 1234
way a correction table for the burette readings can be constructed.
Cushman" mentions an improvement on the Ostwald pipette which
Ostwald himself had previously adopted, which consists in graduating
the upper narrow tube, care being taken that the capacity of the pipette
from the mark c to about the middle of the upper tube b is .
2 c.c. It is then not necessary to determine the capacity of the pipette
by a number of accurate weighings as above, but only to find the value
of the pipette Scale with regard to the burette scale by a few determina-
tions. In calibrating, 2 c.c. at a time are allowed to pass from the
* Chem. News, 1902, 85, 77.
= I.O36 C.C.
c.c. in this section contains = I.O.I.3 c.c., and so on. In this

44 GENERAL METHODS USED IN TECHNICAL ANALYSIS
burette into the pipette, and the height of the liquid in the latter noted;
the corrections of the burette can thus be calculated. This calibration
is of course only relative; if the absolute values of the graduations are
required, the capacity of the pipette must be determined in the ordinary
way. Or it may be found by making a few weighings, so as to ascertain
up to which of the graduations on the upper tube the pipette has to
be filled so that it holds exactly 2 c.c.; it is subsequently always filled
up to this mark, and used for calibrating according to the method
described above, whereby the corrections to be applied to the readings
of the burette can be found without much calculation. For calibrations
according to the true litre, attention must be paid to the details on
p. 35 et Ség.
In the calibration of measuring instruments for gases, a distinction
must be drawn between apparatus for the actual analysis of gases, for
which only relative measurements are required, and that employed
for gas-volumetric analysis, i.e., the estimation of a constituent of a
solid or liquid in the gaseous form, for which purpose absolute measure-
ments are necessary. A further distinction must be made, according as
whether the gas is confined over water, over mercury, or in the use of
nitrometers, over sulphuric acid, and the meniscus correction for the
confining liquid must be allowed for, in case measurements are
made at parts of the tube of unequal bore. Lastly, instruments
must not be calibrated in an inverted position compared with that in
which they are to be employed, since the menisci will then be inverted;
or if this is unavoidable, the necessary meniscus correction must be
ascertained.
The expansion of the glass is, in these cases, scarcely ever taken into
account. The temperatures of calibration should be stated as in the
I 5° 2O”
case of liquids, 4° Or 4° ' etc., since if the calibration has been effected
at I 5°, determinations made at this temperature and at 25° are not
equivalent.
Göckel points out that the plan of taking the mean of the readings
at the highest and lowest points of the meniscus, as the correct reading,
is decidedly to be condemned. L. W. Winkler regards the values
for the meniscus correction given by Bunsen” as too low; according to
more recent determinations by Göckel,” however, Bunsen's figures are
very near the correct values. As these corrections are not required for
ordinary volumetric apparatus, reference to the original papers will
suffice.
The time interval of two minutes before taking a reading, as recom-
mended in the case of burettes, is not applicable to gas measuring tubes,
* Z. anal. Chem., 1901, 40, 403; and Z, angew. Chem., 1903, 16, 718.
* Gasometrische Methoden, 2nd edition, p. 38. * Z, angew. Chem., 1903, 16, 49.
CORRECTIONS FOR TEMPERATURE 45
since it depends on their form and diameter. In most cases, the values
given below for the time taken by the water, used in gas burettes, to attain
its final position may be taken as correct.
For 25 c.c. 3 minutes. For I25 c.c. 7 minutes.
} } 5O } } 4. } } } % I5O } } 8 } }
} } 75 5 y 5 } } } } I75 } } 9 } }
3, IOO }; 6 } } } } 2OO , , IO , ,
The capacity of gas burettes graduated to hold exactly IOO c.c., or
any other definite volume between two glass stopcocks, such as Winkler's
burette, is hardly ever correct, and the actual capacity should always be
determined and the corrected value marked on the burette; it should
further be noted for which point of the meniscus the reading has been
calibrated, and whether the meniscus correction has been made for
water or for mercury.
Gas measuring tubes graduated in mm., such as the Bunsen
eudiometer, can of course be used at any temperature and with any
confining liquid ; they must be specially calibrated in each case, and a
corresponding table of corrections drawn up. Such measuring tubes are
scarcely ever used in technical analysis.
The temperature of the volumetric solutions should really be always
the same as that at which the burette was calibrated, and this tempera-
4°' 17°. 5’
point of view some prefer the calibration to be made at 17°.5 or 20°,
because these temperatures approximate more to the average laboratory
temperature. Small deviations from the normal temperature can be
overlooked, as their effect falls within the limits of experimental error,
but a considerable error is introduced by working at a temperature of
say 8° to IO* or even lower, as often happens in works’ laboratories
in winter, or, on the other hand, when the temperature rises to 25° or
more in summer, as may occur in the best laboratories. Under
such circumstances if the temperature differs from the normal by
more than 2°, the necessary corrections are imperative. In most cases
these corrections may be made with sufficient accuracy on the basis of
the expansion of distilled water; if greater accuracy is required, the
tables drawn up by A. Schulze" for a number of normal solutions may be
used. For correcting readings when the temperature of the water
exceeds 15°, use may be made of the following table, calculated by
Schloesser * for the expansion coefficient of glass=O-OOOO27, and
the values for the expansion of water given by the Physico-technical
Reichsanstalt. The figures given in the table are the number of c.c.
which must be subtracted from IOOO c.c. to give the volume of
ture should be indicated on the vessels (e.g., etc.) From this
* Z. anal. Chem., 1882, 21, 167. * Private communication.
46 GENERAL METHODS USED IN TECHNICAL ANALYSIS
distilled water which at tº fills a litre flask, calibrated at I 5°, so that it
Occupies a volume of IOOO c.c. at I 5°.
Temp. C.C. Temp. C.C.
I5° O-OOO 23° I.348
I6° O-I3O 24° I-563
17° o.272 25° 1.788
I8° O'42 26° 2-O23
19° o:58 27° 2.267
20° o,76 28° 2.52O
2 I ‘’. O-94 29° 2.782
22° I-14 30° 3-O53
This table holds for the apparent expansion of distilled water. The
values for N/5 and M/IO solutions differ so slightly from these, that the
table may also be used for these solutions. For N/I solutions, on the
other hand, the deviations are greater: for hydrochloric acid, the
expansion between I 5° and 25° may be taken as 2.42 per cent. ; for
Oxalic acid, 2.62 per cent. ; for sulphuric acid, 3.05 per cent. ; for sodium
hydroxide, 3. I 5 per cent, and for sodium carbonate, 3.03 per cent.
In using such solutions errors may also be introduced by working at
temperatures differing appreciably from that at which the solutions
were standardised. For instance, in the case of normal hydrochloric
acid, standardised at I 5°, readings made at IO’ must be divided by O.999.1
and readings at 20° by I.OOI I, in order to avoid an error of + o- I per
cent. ; for normal sulphuric acid the error is about I } times greater
than for distilled water.
APPARATUS FOR VOLUMETRIC ANALYSIS
The Ordinary apparatus used in volumetric analysis, such as
measuring flasks, measuring cylinders, pipettes and burettes, is familiar
to every technical chemist, and is fully described in text-books on
analytical chemistry. Attention may, however, be advantageously
directed to some points of particular importance for technical labora-
tory work. -
Burettes.
The cheapest and most convenient of the various forms of burette
is that shown in Fig. 27, in which the exit tube is closed with a glass
bead, thus dispensing with a metal pinch-cock.
By Squeezing the rubber tube outside the bead a, it takes the form
shown in Fig. 28, thus leaving a space at each side through which the
liquid flows out; with a little practice in the manipulation, the liquid
can be allowed to escape as quickly or as slowly as may be desired.
This arrangement lasts much longer than any of the various forms of
pinch-cock, with their accompanying rubber tubes, and can even be
used in gas analysis apparatus such as the Orsat apparatus, in place of
expensive and fragile stopcocks.
THE READING OF BURETTES 47
Burettes with glass stopcocks are, however, indispensable for
potassium permanganate and iodine solutions; those with stopcocks
at the side should be employed.
When used in the control work of manufacturing processes, no special
precautions in reading burettes are necessary. For more
accurate laboratory work, however, the following precautions
are necessary:— * ||É
In order to be fairly certain of reading burettes of 50 c.c. *
capacity, and graduated in ºr C.C. to within O.OI c.c., the use
of a pocket lens is advantageous; it is also necessary to avoid
two sources of uncertainty, namely, the indistinct boundary
between air and liquid, and the parallax error. The first is
generally avoided by taking the lower boundary line of the
black meniscus as the normal reading, but this cannot be
done in the case of dark liquids such as permanganate solu-
tion. The boundary line is rendered much sharper
by shutting off the light from below, by holding
a sheet of black paper behind the burette, with the
upper edge some mm. below the meniscus; as an
alternative, the fingers held close together may be used. It
is scarcely necessary to illuminate the upper part specially.
Instead of black paper, G. Bergmann proposed the use of a
blackened wooden clamp with a holder, which is fixed on #||
to the burette from the side, and on which a milk-glass plate |
is screwed. Schellbach recommends the use of burettes with
two narrow white longitudinal strips at the back; in this |
way a fine point, which can easily be read, is formed at |
the boundary of the meniscus.
The use of floats naturally obviates the difficulty arising
from the want of definiteness of the meniscus, provided that
they are accurately made and that their mark is finely etched.
The second source of uncertainty is due to parallax. If the eye is
not quite accurately in the plane of the meniscus, the reading may be
several hundredths of a c.c. either too high or too low. This is obviated
by having a circular mark for reading, which appears as a single
line when the eye is exactly in its plane. For this purpose Erdmann
devised the cylindrical float, which is, however, open to the objections
that it often sticks, especially when the burette is being filled, and that
the greater accuracy of reading hoped for from its use is rendered quite
illusory by its frequently taking an oblique position and by the unequal
capillary rise of the liquid between its exterior wall and the inside wall
of the burette. For these reasons most chemists have abandoned
its use.
FIG. 28.
#
FIG. 27.
* Z, angew. Chem., 1898, 11, 853.


48 GENERAI, METHODS USED IN TECHNICAL ANALYSIS
The spherical float of Beutell (Fig. 29), in which capillary effects are
obviated, is a considerable improvement on that of Erdmann. With
a good float of this kind, one which takes a true vertical position,
an experienced observer can read quite well to O.OI c.c. and in any case
with much greater certainty than without a float, when ordinary
burettes without circular marks are used.* These floats are often not
properly weighted, and consequently hang obliquely; they are then
quite useless.
To take readings to O.OI c.c. with a float, it is preferable to use a
magnifying glass, or for greater certainty, a reading telescope; but
the increased accuracy thus obtained is illusory if the errors of the
burette are too large, or if the graduations are too coarse or not
sufficiently uniform, as is often enough the case. The
graduation of Erdmann's float in ninths, as proposed by
Prinzl, although in principle undoubtedly an improvement,
has scarcely come into use, chiefly because it can only be
employed for cylindrical floats, the drawbacks of which
have been mentioned above.
The utility of using floats has been much disputed,
e.g., by Reinitzer, Andrews, J. Wagner, Kreitling, and
Schloesser. The majority of the objections raised, how-
ever, apply to the cylindrical Erdmann float, and include
the points referred to above. Lunge regards a properly
constructed Beutell float as reliable and satisfactory.
Ordinary floats cannot be used for fairly strong per-
manganate solution, on account of its dark colour. The
double spherical float proposed by Rey,” one sphere of
which projects above the liquid and carries the mark, for
such solutions, is unsatisfactory, on account of the difficulty
of getting floats of this kind which hang vertically and
also because the upper sphere is too small. The floats constructed by
Diethelm * are somewhat better, but the upper bulb is also unavoidably
too small.
Many chemists are prejudiced against the use of floats altogether,
because they are often faulty in construction and wrongly used. The
German Imperial Normal Standards Commission also condemns them
and desires to replace them by having the graduations carried either
right round, or three-fifths or half-way round the burette; the burette
must then be placed so that the graduations are on the left-hand side
of the observer, are visible both from the back and front, and appear
as a single line at the point of reading when the head is held in the
proper position. In this way parallax is certainly avoided, but the
* CAE. G. Lunge, 5th Congress of Applied Chemistry, 1903.
* Aer., 1891, 24, 2098. * Chem. Zeit., Io2, 26, 607.

GöCKEL'S SCREEN 49
uncertainty in the reading of the meniscus remains, and must be over-
come by the use of an illuminator; also, such burettes are expensive
(about Is. 6d. to 2s. dearer than ordinary burettes of the same accuracy),
and, according to the makers, they are more fragile than those with short
graduations. The objection has been made that these longer gradua-
tions are fatiguing to the eye, but this, and also their greater fragility,
are denied by Schloesser.
The problem of avoiding parallax and at the same time attaining
a sharper reading is, according to Lunge, most satisfactorily solved
by the use of the light screen constructed by Göckel' (Fig. 30),
which renders both floats and circular graduations unnecessary. It is
based on a principle similar to that of Bergmann's clamps (p. 47), and,
like these, is clamped on the burette 2-3 mm. below the lowest point of
the meniscus; owing to its diagonal form, it can be fitted to tubes of
from 9-20 cm. diameter with one and the same clamp. The black
clamp shades off the superfluous light, thus forming a very sharp, dark
boundary line. If necessary, especially when the light is not very good,
H20 Meniscus Hg. Meniscus
a sheet of white paper may be held behind the burette to give better
illumination on the scale, or it may be fixed to the clamp, thus acting
as a complete substitute for the fragile frosted-glass plate employed by
Bergmann. The new feature in Göckel's screen is the avoidance of
parallax, by making the opening of the Screen exactly at right angles
to its horizontal surfaces. Small metallic plates, screwed to the clamp,
act in such a way that on opening and closing the screen its motion is
always in the same plane. In order to avoid parallax, therefore, it is
only necessary to bring the eye into such a position that the front and
back edges of the upper surface of the screen coincide. The same
object can be attained for a mercury meniscus by placing the screen
2-3 mm. above it, as shown in the figure.
Lunge” regards Göckel's screen as thoroughly satisfactory, and is of
opinion that it makes the use of floats in ordinary burettes with short
graduations quite unnecessary; it may also be employed for avoiding
parallax in dealing with opaque solutions, such as permanganate,
although in such cases the meniscus is, of course, invisible.
* Chem. Zeit, 1903, 27, Io 36. * Z. angew. Chem., Igo4, 17, 198.
D


50 GENERAL METHODS USED IN TECHNICAL ANALYSIS
Dupré" uses a small glass rectangular mirror, 72 by 50 mm., for
reading burettes. The lower portion of the mirror is covered with a
piece of paper, fixed parallel to the side of the mirror, the right-hand
half of the paper being white and the left-hand half black. In taking a
reading the mirror is held against the back of the burette in such a
position that the black part of the paper gives the best definition of the
meniscus, and the head is so placed that the pupil of the eye is seen
reflected at the same height as and close to the division to be read.
Before being used, burettes, pipettes, and other measuring vessels
must be well cleaned, especially from fatty substances. The following
reagents are recommended for this purpose:–alcohol and ether; strong
Sodium hydroxide solution, followed by water, hydrochloric acid, and
then water again ; strong nitric acid ; acidified permanganate solution
for from one to two days. A mixture of sulphuric acid and potassium
bichromate is probably the most useful reagent; it can be kept for
several months, and in obstinate cases may be used warm.
The German Imperial Normal Standards Commission (1902) rightly
condemns the use of hydrofluoric acid for cleansing purposes, as is
sometimes recommended, and also the use of all other strongly acting
chemicals, which is going somewhat too far. They give the preference
to soap and water (concentrated and hot if necessary) to be applied
either with strong brushes, or by shaking with filter paper, according
to the shape of the vessel. Such mechanical cleaning is certainly
preferable to any other method.” y
Insufficiently cleaned burettes, etc., give rise to serious errors by the
adherence of drops to the dirty parts; instruments, which in spite of all
efforts cannot be satisfactorily cleaned, should only be used for rough
estimations.
Different liquids wet glass to a different extent, and consequently
give different volumes on delivery. According to J. Wagner (cf. p. 41),
who has made experiments on this point, the errors arising in this way
are, as a rule, negligible, but in the case of concentrated Sulphuric acid
the amount adhering to the glass is much greater than in the case of
aqueous liquids (O.7 per cent. instead of O.2 per cent.), and this must
be taken into account in the analysis of nitrous vitriol. ,
In technical laboratories, none of the more complicated forms of
burette stands and holders should be used, but only the simplest and
cheapest, more particularly those which hold a large number of burettes.
It is very necessary to choose forms in which no part of the scale is
covered by the holder. A very simple and inexpensive form is shown
in Fig. 31.” The burette fits into the semicircular opening in the wooden
support, and is held against it by the piece of gut a attached to the peg
* Private communication to Prof. Lunge. * Cf. Schloesser, Z. angew. Chem., 1903, 16, 962.
* Made by C. Desaga, Heidelberg. -
ARRANGEMENTS FOR FILLING BURETTES 51
à ; by turning b, a can be loosened, so that the burette can easily be
taken out or replaced.
Burettes are very largely employed in technical laboratories, which
have an arrangement for filling them from below by means of a side
tube, or by means of a T-piece introduced between the lower end of an
ordinary burette and the delivery
tube; a rubber tube is joined to
the horizontal arm of the T-piece
and leads to a stock-bottle at a
higher level (Fig. 32). The burette
is filled from below, a pinch-cock
or glass stopcock being attached
to the side tube to regulate the
supply of the solution ; an obvious
saving of time and gain in cleanli-
ness is thus effected. This method
of filling has also been suggested
for liquids which must not be
FIG. 32. FIG. 31.
brought in contact with rubber;" this may be done by making a
rigid glass connection between the burette and the stock-bottle, sealed
on to the burette at its lower end, and with the upper end dip-
ping into the bottle, like a syphon. This arrangement is, however,
much too fragile, as it may be easily damaged by shaking. It is
* Gawalowski, Z. anal. Chem., 1885, 24, 218.


52 GENERAL METHODS USED IN TECHNICAL ANALYSIS
important to bear in mind that whether burettes are filled from an
attached stock-bottle or by hand, it is always necessary to shake the
stock-bottles of normal solutions at least once a day, preferably at
the beginning of the day's work, in order to mix the water which
has evaporated and condensed on the upper empty part of the bottle
with the rest of the solution ; such shaking is obviously out of the
question with a rigid connection, such as that suggested above. Stock-
bottles, if connected for filling up the burette as described above, must
be shut off from the air by tubes con-
taining calcium chloride, caustic alkali,
etc., or, less conveniently, a tube may be
led from the stock-bottle into the top
of the burette, so as to prevent changes
by evaporation of water, absorption of
carbon dioxide, oxygen, etc. The large
stock-bottles in which the volumetric
solutions are kept in bulk, and which
often are of a capacity of 50 litres or
more, must be similarly shaken before
filling the smaller supply-bottles; this is
facilitated by making use of the ordinary
stands for carboys.
The apparatus constructed by O.
Knöfler, in which the burette is fixed
into the stock-bottle itself, and filled
from below by blowing in air with a
rubber ball, is very convenient for many
purposes. Various forms of this appa-
ratus are obtainable. Contact with
rubber can be avoided by this device
without making it impossible to shake
the bottle. -
An overflow arrangement can readily
be connected to burettes of this type,
as shown in Fig. 33; in this case it is not necessary to adjust to
zero, because this is regulated by the upper end of the burette
itself, since the liquid which overflows, when filling, runs back into the
bottle. - e
Göckel 1 has devised a burette with a long bent side tube for
titrating hot solutions; an apparatus for facilitating the titration
of a large number of solutions has been constructed by W.
Schmidt.”
* Z, für chem. Apparatenkunde, I905, I, 99. * Chem. Zeil., 1904, 28, 154.
}
;
(-


PIPETTES
Pipettes.
Graduated pipettes of large dimensions are seldom required ; the
most useful sizes are I c.c. (divided in rāgths), 2 c.c. (divided in ºoths),
and 5 c.c. or IO c.c. (divided in gºths or ſºoths). Since
they are mostly used for measuring out the solutions to
be analysed, great stress must be laid on their accuracy.
The method of delivery has been described above (p. 39).
The measuring vessels suggested by O. Bleier," of
which the measuring pipette is shown in Fig. 34, are
designed to deliver quantities up to say 50 c.c. by means
of one and the same pipette, according as the point à, c,
d, e, or f is taken as zero. The capacity from 5 to c, c to
d, d to e, and e to f, must always be exactly IO C.C., and
the connecting tubes must not be too wide, as otherwise
the practical value of the device is lost.
As regards the method of using ordinary pipettes,
reference may be made to p. 42. Pipettes which have the
graduation on the upper part only, and which accordingly
deliver their full content, are preferably used only for less
important work, such as in the control of a manufacturing
process; when the greatest possible accuracy in measuring
is desirable, it is preferable to use pipettes having a lower
graduation as well. The latter should be just above the
contracted part of the tube, at the point to which the
pipette empties itself when used for dilute solutions. The
º
}*
*
*
Kºsº
*E-
*
-
ºmºm
ya
FIG. 34.
tediousness referred to by J. Wagner in using a pipette with two marks
is then avoided, even when the necessary interval for draining is allowed,
which is best done as follows. The pipette is
FIG. 35. FIG. 36.
run out to within a few mm. above the lower
mark, an interval of 3 minute allowed, then run out
accurately to the mark, and finally the last drop
removed. Before adjusting to the upper mark, the
liquid wetting the lower part of the pipette on the
outside should always be washed off with water.
When one and the same volume of liquid has
often to be measured off, self-adjusting pipettes
(overflow pipettes) are very suitable, especially when
large amounts (50-IOO c.c.) have to be taken. There
are many forms of these pipettes, of which those
shown in Figs. 35 and 36 are examples.
It must not be overlooked that, even under
the most favourable circumstances, measuring in
pipettes is far less accurate than weighing. The procedure very
* Chem. Zeit., 1897, 2I, Io28.





54 GENERAI, METHODS USED IN TECHNICAL ANALYSIS
generally and advantageously adopted for technical work, of weighing
out a large amount of substance, dissolving it in a measuring flask
and withdrawing an aliquot part of the solution for analysis, is never
as accurate an analytical operation as the weighing out and direct
analysis of a small amount of substance, even if the measuring vessels
have been carefully calibrated. The advantages of the latter method
may, however, be more than counterbalanced by the difficulty of
obtaining a satisfactory average sample in small bulk.
Vessels used for Titration.
Beakers are employed as a rule, but if it is desirable to exclude air
as much as possible, ordinary or Erlenmeyer flasks are preferable ; a
current of carbon dioxide can be passed into the flask during the titra-
tion, if desirable. Dupré' uses a large bulb U-tube; when the colour
change of the indicator in use appears, in the portion of the solution
contained in the bulb into which the titrating solution is delivered, the
contents of the tube are mixed by gentle shaking. The titrations are
thus carried out rapidly and with safety.
Quality of the Glass employed in volumetric analysis.
Attention to the quality of the glass of which the measuring vessels,
beakers and flasks, employed in volumetric work, are made, is of importance.
It has long been known that many kinds of glass are gradually attacked
even by distilled water, and more readily by alkalis, with the liberation of
both silicic acid and of alkali. Glass is much more resistant towards acids.
Formerly, it often happened that burettes in which sodium hydroxide
had stood for a long time developed cracks, but this seems to be of less
frequent occurrence now ; dependence should, however, not be placed on
standard alkaline solutions which have stood for a long time in vessels
made of ordinary glass, without further standardisation.
Glass is attacked to a much greater extent on heating. For this
reason, the sharper change of colour shown by phenolphthalein and
litmus as compared with methyl Orange is, in many cases, quite deceptive,
because, when the first two indicators are used, long boiling is necessary
to drive out the carbon dioxide, and if this is done in vessels made of
ordinary glass, the amount of alkali may easily come out distinctly too
high. This can be guarded against by using porcelain vessels, or, though
by no means perfectly, by using specially made glass, such as Jena
“resistance glass.” - -
That glass giving a strongly alkaline reaction is still sometimes met
with in commerce, has been pointed out by C. Liebermann.” .
Lunge 3 states that on boiling normal sodium carbonate solutions in
* Private communication to Prof. Lunge. * Ber., 1898, 31, 1818.
* Z, angew. Chem., 1924, 17, 196.
INDICATORS FOR ACIDIMETRY AND ALKALIMETRY 55
vessels of ordinary Thuringian glass, for a long time, considerable
amounts of sodium and of silicic acid go into solution, calcium silicate
being precipitated, and that even Jena resistance glass is distinctly acted
upon under the same conditions.
INDICATORS FOR ACIDIMETRY AND ALKALIMETRY 1
Acidimetry and alkalimetry depend on the fact that when acid and
alkaline solutions, which, however, need not necessarily contain free
acids or alkalis, are brought together, a definite point is reached at
which the indicator, which has been added to one of the solutions,
changes colour. -
A very great number of indicators have been suggested, but the
majority have been found to be of but little value in practice; many
of them are quite useless for accurate work, and most of them are
superfluous. Only three indicators, which suffice for all purposes, are
in general use—litmus, methyl orange, and pheno/phthalein. In fact,
litmus, although the oldest indicator, and formerly almost exclusively
used, can also be dispensed with, and is now only used in many
laboratories in the form of litmus paper; but all eyes are not so
sensitive for the change of colour with methyl orange as with litmus, and
for this reason, as also from long custom, litmus is still widely used.
Indicators are usually divided into three classes—(1) those which are
only slightly or not at all sensitive to weak acids, such as carbonic acid,
sulphuretted hydrogen, silicic acid, boric acid ; (2) those of medium sensi-
tiveness; (3) those which are almost as sensitive to the weakest as to the
strongest acids. Indicators of the first class have a distinctly acid char-
acter; they combine with all bases, even weak ones, to form salts, which
have a definite colour and which are only decomposed by stronger acids,
with the production of a different colour pertaining to the free acid. They
are therefore sensitive to alkalis and to strong acids, but not to weak
acids. The most important of these is methyl orange; other azo-
colouring matters such as tropäolin OO, Congo red, and benzopurpurin
are also used, and further, lacmoid, cochineal, para-nitrophenol, and iodo-
eosin, the latter being particularly sensitive towards alkalis.
The third class has exceedingly weak acid characters, usually only
phenolic, and their salts are therefore decomposed even by the weakest
acids. These indicators are very sensitive to acids, even carbonic acid
and sulphuretted hydrogen, less sensitive to bases. The most important
indicator of this class is phenolphthalein ; others are, turmeric, rosolic
acid, flavescin, etc.
* A comprehensive account of indicators and their applications is given in the following mono-
graphs:—A. I. Cohn, Indicators and 7est Papers, 2nd edition, Igo2 ; F. Glaser, Indikaloren der
Acidimetrie und Alkalimetrie, Igoi.
56 GENERAL METHODS USED IN TECHNICAL ANALYSIS
The second class forms the transition between the other two ; it
includes indicators which are somewhat sensitive even to weak acids,
but less so than those of the third class, and for this reason there are
often gradual transitions when using them, in titrating, which do not
coincide with the formation of definite compounds. If the weak acids
set free during titration, such as carbonic acid and Sulphuretted hydrogen,
are removed by boiling, they act like the first class. Litmus is the
chief representative of this class; others are phenacetolin, alizarin,
haematoxylin, etc. -
Since strong acids, such as hydrochloric acid or sulphuric acid (Oxalic
acid has almost gone out of use in alkalimetry), and strong bases are
now universally used as normal solutions, the following conclusions
regarding the use of indicators may be accepted.
The strong mineral acids can be titrated sharply with all indicators,
but most reliably and, in the cold, only with those of the first class;
with the third and even with the second class, the normal alkalis must be
free from carbonic acid, or the titration must be carried out in the boiling
solution.
Organic acids of medium strength (Oxalic acid, lactic acid, tartaric
acid, etc.) cannot be titrated with indicators of the first class ; those of
the second class may be used, but it is preferable to employ those of
the third class, the same conditions being observed for the normal
alkalis as above.
Polybasic mineral acids of medium strength (phosphoric acid, sulphur-
ous acid, etc.) cannot be titrated at all with indicators of the second class,
but indicators of the first and third classes may be employed; different
degrees of saturation occur in their titration, as will be seen below.
Weak acids, both organic and inorganic, can only be titrated, if at
all, with indicators of the third class.
The strong bases (hydroxides of potassium, sodium, barium, calcium)
can be titrated sharply with all indicators.
Bases of medium strength (ammonium hydroxide, amine bases) can
be titrated sharply only with indicators of the first class and a few of
the second, not with those of the third class. -
Weak bases cannot be titrated sharply even with indicators of the
first class, and not at all with those of the remaining classes.
J. Wagner' rejects the above classification of indicators in favour of
a more scientific one, and divides them into the following groups:–
A., Indicators with a univalent characteristic ion (I. anion: 2. cathion);
B., those with a polyvalent ion (I. positive and negative ion = amphoteric
electrolytes: 2. univalent and divalent anion or cathion). With A there
are no intermediate colour changes, but these occur with B. Most
indicators belong to the group A. I., only a few to A. 2. and B. 2.
* Z, anorg. Chem., 1901, 27, I 38.
THE SENSITIVENESS OF INDICATORS 57
Methyl orange belongs to group B. I. This classification has not been
generally adopted.
Many investigations on the sensitiveness of indicators have been
carried out, but the results are frequently contradictory, and since the
sensitiveness is so greatly affected by various other influences, such as
temperature, dilution, presence of other substances (salts, alcohol, etc.),
no broad and general conclusions can be deduced. The following
remarks include the most reliable deductions. -
Only in quite rare instances is it advisable to titrate with solutions
of less than M/Io strength, because only very few indicators are sensitive
enough to give a sharp colour change with I-2 drops of the solution ;
iodo-eosin, in ethereal solution, is sufficiently sensitive to allow of the use
of N/IOO solutions. - - .
Even with M/10 solutions there are differences with the ordinary indi-
cators. In the absence of weak acids (carbon dioxide and sulphuretted
hydrogen) the transition with phenolphthalein or well purified litmus
is distinct, for most eyes, with a single drop ; with methyl orange, two
drops are necessary. This difference is illusory, however, since in
practice it is seldom possible to completely exclude carbonic acid, and
further, all indicators are less sensitive in hot solution.
With W/5 and stronger solutions, the sensitiveness of the three
principal indicators is such that a single drop suffices to bring about the
change of colour; other indicators are less sensitive.
Küster and Grüters' find that the orange (brownish) intermediate
colour of methyl orange is always produced with the same hydrogen ion
concentration, therefore the amount of acid required to effect this is
roughly proportional to the volume of liquid. In 50 c.c. of water, O.O3 c.c.
M/Io hydrochloric acid is required to bring about the change of colour;
in Ioo c.c., O.O6 c.c.; in 200 c.c., o. I2 c.c. Similarly, when chlorides are
present (that is, more chlorine ions), more acid is required.
They state further, that the point of neutralisation can be
determined most sharply by conductivity measurements, a method
particularly applicable when precipitates or colorations make the use
of coloured indicators impossible.
Solutions standardised with an indicator for hydroxyl ions, such as
phenolphthalein, are not strictly correct when employed with an indi-
cator for hydrogen ions, such as methyl orange (cf. p. 60).
Both alcohol and acetone lessen the sensitiveness of indicators, and
the presence of considerable amounts of these substances in solutions to
be titrated should, therefore, be avoided as far as possible; with
phenolphthalein particularly, they cause great irregularity. If an
addition of alcohol cannot be avoided, it is best to make a blank
experiment to find how much acid or alkali is required to produce
* Z, anorg. Chem., Igo3, 35, 454.
58 GENERAI, METHODS USED IN TECHNICAL ANAIYSIS
the colour change in pure water containing the same amount of alcohol,
and to allow for this in the titration. Alkaloids have sometimes to be
titrated in strong alcoholic solution ; in such cases lacmoid is the best
indicator.
Neutral salts, which themselves do not alter the colour of the
indicator, usually only affect the transition in titrating, when present in
great concentration; such excess can as a rule be easily avoided. This
is true more particularly of the alkali chlorides and sulphates, which
are themselves formed in titrations. Lunge and Lohöfer have shown
that a certain amount of sodium chloride is necessary to get accurate
results in titrating sodium silicate, sodium carbonate, and sodium
sulphate with phenolphthalein, but that if present in large excess, it
acts unfavourably, unless greatly diluted.
Warming often brings about a change of colour, even without the
addition of acid or alkali. If, for instance, sodium hydroxide is added
to sulphuric acid in the presence of methyl orange until a neutral
orange tint is produced and the solution heated, the colour changes to
a distinct yellow, and becomes Orange again on cooling ; therefore, in
titrating acid with alkali in hot solution, using methyl orange as
indicator, too little is required, and in titrating alkali with acid, too
much. On the other hand, phenolphthalein in hot solution changes
from red to colourless too soon. With methyl orange, the transition in
hot solution is, as a rule, not sharp, and this indicator should, therefore,
always be used at or only slightly above the ordinary temperature.
Both phenolphthalein and litmus are less sensitive in hot solution
than in cold, or rather the end-point is somewhat different, and due
attention must be paid to this, since it is usually necessary to
boil the solution to drive out the carbonic acid when using these
indicators. For accurate estimations with litmus and with phenol-
phthalein, the solution should therefore always be boiled in the presence
of a slight excess of acid, then allowed to cool, and the titration
completed.
As already mentioned (p. 54), the action of hot solutions on glass
must be taken into account and as far as possible avoided.
Theory of Indicators. The employment of indicators depends upon
the fact that they change colour at a definite point on the addition of an
acid or alkaline solution. This point may coincide with what is
commonly known as “neutrality,” but this is by no means always the
case. If solutions of hydrochloric acid and Sodium hydroxide are brought
together, the change of colour will, it is true, take place with all indi-
cators in practical use, when one molecule of sodium hydroxide is
present for one of hydrochloric acid, so that the compound sodium
chloride is formed. Since dilute solutions are always used, it is not to
* Z. angew. Chem., 1901, I4, II29; also, Küster, Z, anorg. Chem., 1897, 13, I45.
THEORY OF INDICATORS - 59
be assumed, according to modern views of solution, that only the com-
pound NaCl actually exists in the solution, but rather that the greater
part of it is dissociated into free Na’ and Cl’ ions, just as the sodium
hydroxide was previously, to a great extent, dissociated into Na' and
OH' ions, and the hydrochloric acid into H' and Cl’ions. Neutralisation
means rather that there is no appreciable amount of hydrogen ions
which give the acid reaction, or of hydroxyl ions which give the alkaline
reaction, in the solution, since the H ions of the hydrochloric acid have
united directly with the OH' ions of the sodium hydroxide to form
molecules of water.
In the presence of certain colouring matters even the smallest excess
of an acid or of a base, that is, very small amounts of hydrogen or of
hydroxyl ions, bring about a marked change of colour in one direction
or the other, and the point of neutralisation is thus ascertained.
Ostwald" was the first to advance a theory of indicators; further
contributions have been made by Küster,” Waddell,” Julius Wagner,"
Bredig and Winkelblech,” Vaillant," Roloff' and Stieglitz.”
In regard to the theory of indicators, it will suffice to state that
the indicators are or contain acids, or in some few cases bases, the undis-
sociated molecules of which have a different colour from that of their
ions.” Thus the undissociated molecules of the litmus acids are red, the
anions blue, and since these acids are weak, in other words slightly
dissociated, a mixture of the two colours results in aqueous solution.
On addition of hydrogen ions, that is, acids, the litmus acids become
practically undissociated, the blue anions disappear, and the red colour
of the undissociated molecules appears. To effect this change, however,
a certain minimum concentration of hydrogen ions is necessary, so that
very weak acids such as carbon dioxide and sulphuretted hydrogen have
practically no action ; acetic acid, on the other hand, acts quite dis-
tinctly, unless its ionisation is so much decreased by the addition of
sodium acetate that an insufficient amount of free hydrogen ions remains.
When bases are added, salts of the litmus acids are formed which are
largely dissociated, so that the blue colour of the ions is pronounced.
Phenolphthalein has a still weaker acid character; its undissociated
molecules are colourless and its anions red. The few free hydrogen ions
present in a solution are so greatly lessened in amount even by the
weakest acids (carbon dioxide, etc.) that the red ions completely dis-
* The Scientific Foundations of Analytical Chemistry, trans. by G. M“Gowan, 2nd edition, 1900
* Z. anong. Chem., 1897, I3, I27. 3 ſ. Phys. Chem. I898, 2, 171. -
* Z, anorg. Chem., 1901, 27, I42. * Z. Electrochem., 1899-1900, 6, 35.
* Comptes rend., 1903, 136, 1912. * Z. angew. Chem., Igo2, I5, 594 and 599.
* /. A mer. Chem. Soc., 1903, 25, III2.
* The colour changes of indicators are regarded by some chemists as due to a rearrangement
of the structure of the molecules, rather than to dissociation. Cf. A. G. Green and P. E. King,
J. Soc. Chem. Ind., 1908, 27, 4.
60 GENERAL METHODS USED IN TECHNICAL ANALYSIS
appear and only the colourless molecules are left. Even the neutral
aqueous solution shows no appreciable colour. The smallest trace of a
strong base suffices to remove the hydrogen ions in combination with
hydroxyl ions and to form a salt, the dissociation of which produces the
red colour of the phenolphthalein anions. With weak bases, on the
other hand, even with ammonium hydroxide, hydrolysis takes place, so
that the free base and free colourless phenolphthalein molecules are
formed ; in order to obtain a sufficient number of the red ions, more
than the minimum quantity of the base must be added, so that the salt
formation, and consequently the red coloration, is persistent.
Methyl orange is a fairly strong acid, the undissociated molecules of
which are red and the anions yellow; so many of the latter are present,
even in a neutral solution, that a mixed colour results. On the addition
of weak, in other words, slightly dissociated acids, the ionisation of
methyl orange is not completely overcome, so that the mixed colour
persists, the more so if a fairly large amount of the salt of the weak acid
is formed in the titration ; the hydrogen ions are then not sufficient to
cause the combination of the yellow anions of methyl orange to undis-
sociated molecules. This recombination is brought about, on the con-
trary, on adding very small quantities of strong mineral acids, which are
almost completely dissociated, and therefore yield sufficient hydrogen
ions to render the methyl orange almost completely non-ionised, and so
to produce the red colour.
More complicated cases are discussed under the individual indicators.
As a general rule, in titrating the same substances with the same
standard solutions but with different indicators, the results obtained are
not identical, because the changes of colour do not show identically the
same “point of neutralisation”; more hydrogen or hydroxyl ions may
be necessary in one case than in the other to produce a change of
colour. For example, a solution of hydrochloric acid which is exactly
one-fifth normal when titrated with pure sodium hydroxide, using
methyl orange as indicator, is not accurately one-fifth normal to phenol-
phthalein, even when the titration is carried out in the complete
absence of carbonic acid, with continual boiling and all necessary pre-
cautions. For this reason normal Solutions should as far as possible
always be standardised with the same indicator as they are subsequently
to be used with in practice. Even supposing it proved that, e.g., the
true point of neutralisation is obtained more correctly with phenol-
phthalein, all precautions being taken, than with methyl orange, never-
theless a standard acid to be employed subsequently for titrating in the
cold with methyl orange must be standardised with the latter, and
not with phenolphthalein; accordingly, a substance such as sodium
carbonate, which can be titrated with methyl orange, should be employed
for standardising the acid, whereas potassium tetroxalate, which can
METHYL ORANGE 61
only be titrated with phenolphthalein, ought not to be used if the
standardised solution is to be employed with methyl orange."
I. Methyl orange.
This indicator is the violet, crystalline p.dimethylaminoazo-
benzene p.sulphonic acid, SO3.H.C.H.I.N: N.C.H.I.N(CHs), or its
yellow sodium salt, which is readily soluble in water. The latter is
Sometimes adulterated with dextrin; if this is the case, it does not
form a clear solution, and should be rejected. The free acid cannot
easily be adulterated, and is to be preferred. The sodium salt was
manufactured for some time as a dye under the name of Poirrier's
Orange No. 3, and also under the name of Helianthin; after it
was recognised that it was useless as a textile dye, on account
of its want of fastness, both names were transferred to other
dyes, and for this reason ought no longer to be applied to the
indicator. Lunge, who was the first to introduce it for this purpose,”
suggested the name methyl orange, which is now generally adhered
to, both for the free acid and for the sodium salt.
Next to litmus, methyl orange is no doubt the most widely used
of all indicators, and in many cases it has completely displaced the
former. It is true that it cannot be used in all instances, but in conjunction
with phenolphthalein for certain purposes, it does in fact meet all the
requirements of an indicator for alkalimetry and acidimetry.
Methyl orange is to be preferred to litmus also on the score of
expense. According to Thomson's investigations,” O. 5 c.c. of a solution
containing O’ I 5 g. per litre is required to give the same reaction as o. 5 c.c.
of a litmus solution, which contains 20 g. of the dry extract in a litre,
corresponding to about 60 g. of good litmus ; the equivalent of I g.
methyl orange is, therefore, 4OO g. litmus. I kilo of methyl orange costs
I6s., and I kilo of Ia litmus, at the price per IOO kilos, 3s. ; inferior kinds
of litmus are nominally cheaper, but, when the colouring power is
taken into account, actually dearer. The 400 kilos of litmus which give
the same result as the single kilo of methyl orange cost, therefore, £60,
or seventy-five times as much as the equivalent amount of methyl
orange. The various “purified" litmus preparations, azolitmin, etc.,
are much more expensive. Phenolphthalein costs 30s. and lacmoid
5OS. per kilo.
For use, I g. of pure methyl orange is dissolved in I litre of water,
and about two drops of this solution is added in each titration; this
is sufficient to give a distinct yellow colour. If, owing to dilution during
the titration, the colour becomes too faint, another drop of the indicator
is added. Under no circumstances should too much of the indicator
* Cf. Lunge, Z, angew. Chem., 1904, 17, I99.
* Chem. Ind., 1881, 4, 348; Ber, 1878, II, 1994. * Chem, News, 1883, 47, 123.
62 GENERAL METHODS USED IN TECHNICAL ANALYSIS
be added at the beginning, since the transition is not sharp if the
colour is too deep; for this reason many prefer to use a more dilute
solution of methyl orange (about I to IO,OOO), so that the correct
amount is not so readily exceeded. Methyl orange solutions some-
times become less sensitive when kept for a long time, and should then,
of course, not be used.
It is imperative that the titrations should be made in the cold
Solution. In hot solutions the change of colour is only gradual, and
the end-point cannot be recognised with any degree of certainty. If
necessary, a temperature up to 30° is admissible, but it is better to
work at the ordinary temperature." According to Glaser,” the
sensitiveness of methyl orange is not affected simply by titrating
in hot solution, but only in the presence of neutral salts. Since,
however, these are always formed in course of titration, the foregoing
statement holds good. Large amounts of alcohol in the solution have
also a disturbing effect. The sensitiveness is also lessened by dilution
of the liquid.
The reddish-yellow colour of the indicator is changed to a pure
yellow by alkaline liquids; this is effected not only by free alkali,
but also by soluble carbonates, bicarbonates, sulphides, silicates, borates
salts of arsenious acid, salts of fatty acids, and in general by the salts
of all weak acids. The addition of a strong mineral acid (hydrochloric
acid, sulphuric acid, nitric acid) to Solutions of alkalis or of their salts
with weak acids does not change the colour of the indicator, in spite
of the fact that carbonic acid, Sulphuretted hydrogen, or other weak
acid is set free, until just before the formation of the normal salt of
the strong acid, NaCl, NagsOp or NaNO3, of course, with the exception
of the colour which appears at the point where the acid comes into
contact with the solution, and which disappears on stirring. Just before
neutralisation, the very faint yellowish colour changes to a deeper
brownish tint, which may be taken as the end-point of the reaction.
On the addition of a further drop of the normal acid, the solution
instantly becomes a decidedly red (carnation red) colour.
Küster (loc. cit.) has pointed out that even free carbonic acid in pure
water, or solutions of neutral salts of strong acids, turns methyl orange
red. Since, in the titration of carbonates as well as in most other cases
where the work has not been carried out with the intentional exclusion
of carbonic acid, the solution finally becomes Saturated with carbonic
acid, the titration should not be stopped at the brownish tint formerly
regarded as the neutral colour, but acid should be added until the Solution
is definitely red. As a comparative guide for the colour change, a
solution may be prepared by passing purified carbonic acid through
pure water coloured with two drops of the indicator.
* Cf. Lunge and Marmier, Z. angew. Chem., 1897, Io, 3. * Indikaloren, loc, cit., p. 51.
METHYL ORANGE 63
If this plan is followed, it must not be overlooked that in titrating
acid with alkali, the point of neutralisation will be different from that
observed with the inverse procedure, and it is therefore always
necessary to finish the titration in the same sense as that in which the
strength of the solution was adjusted, titrating back eventually, if
necessary.
Lunge is of opinion that the surest plan is to titrate, in both cases,
until the brownish transition colour appears, and then to test whether
the solution becomes decidedly red or yellow respectively on adding
another drop, the last drop not to be included in the titration. The
most trustworthy method would be to subsequently titrate back with
the other volumetric solution until the transition colour again appears,
and to subtract the amount required for this purpose, but this is not
necessary in ordinary daily work.
According to Treadwell, the most satisfactory plan is to titrate, in
the cold, until the solution begins to change colour, drive out the
carbonic acid by boiling, and then titrate the yellow solution, after
cooling, until the brown tint reappears; but this is only necessary for
M/IO acids, whilst with AW/5, AW/2, or N/I acids, the alkali carbonates can
be accurately titrated in the cold, by simply running in the acid until
the yellow colour changes to reddish-orange (brownish). In technical
analysis it is only necessary in quite exceptional cases to work with
acids or alkalis weaker than M/5, so that the plan given by Treadwell is
seldom called into consideration.
Lunge states that while carbon dioxide in distilled water does
change the colour of methyl orange to reddish, although by no means
to the same tint as that produced by a trace of hydrochloric acid or
similar substances, yet in a solution of pure sodium chloride of the same
strength as that usually formed in a titration, it gives rise to the brownish
“transition colour,” so that it is incorrect to titrate to a distinct red
colour.
When normal or semi-normal acids are used, a single drop suffices,
under all circumstances, to produce a change of colour which can be
recognised with absolute certainty, even in artificial light; with AW/5 acid,
even an unskilled worker can, after a few experiments and comparisons
with a standard colour, get results accurate to a single drop, but only in
fairly good daylight or in white artificial light, preferably incandescent
(Welsbach) light or acetylene light. A white surface should always be
placed, not only under the flask in which the titration is being made, but
also from 30 to 50 cm. round it; this applies to the use of all coloured
indicators. In dull weather, or with ordinary gaslight, the results may
easily be uncertain to one or two drops, and this uncertainty is corre-
spondingly greater with M/Io acids; but there is seldom any object in
* Analytical Chemistry, vol. ii., p. 427.
64 GENERAL METHODS USED IN TECHNICAL ANALYSIS
working with so weak an acid, even for scientific purposes, and in works’
laboratories they are never used.
Although the change of colour from blue to red with litmus and still
more, that from red to colourless, or conversely, with phenolphthalein,
can be more readily recognised than that from yellow to brownish-red
with methyl orange, these advantages are not appreciable with normal
or semi-normal acids, or even with M/5 acid ; as accurate results can be
got with methyl orange as with the other indicators, after some practice.
Since methyl orange has the very great advantage that the titrations
are carried out at the ordinary temperature, any disturbing effect due to
the glass being acted upon is eliminated; also, the estimation can be
obviously carried out in less time than is required with other indicators
with which boiling is necessary after each addition of acid.
With most other indicators the change of colour is sharper in cold
than in hot solution, and it is therefore a disadvantage, as compared with
the use of methyl orange, to have to carry out titrations with these in
hot solution ; also, because such titrations should, under no circumstances,
be carried out in glass vessels.
The chlorides and sulphates of the heavy metals such as ferrous
sulphate, ferrous chloride, copper Sulphate, cupric chloride, and zinc
Sulphate, which are acid towards litmus, are neutral towards methyl
Orange.
From the above it is clear that methyl orange is much to be preferred
to any other indicator for titrating bases, especially total alkali. On
addition of mineral acids it indicates an extremely small amount of
hydrogen ions, by the appearance of the red colour. It is equally
applicable to the determination of alkali hydroxides, alkaline earths,
ammonia, carbonates, and bicarbonates (for calcium carbonate and the
like, it is, of course, necessary to add excess of acid and titrate back with
alkali), to silicates, borates, salts of arsenious acid, and to Sulphides,
since sulphuretted hydrogen has no disturbing effect." The fatty acids
of Soaps, etc., are without action, so that the amount of alkali in solutions
of soap can be directly determined by titration.
Aniline, toluidine, and quinoline, which are indifferent towards
litmus and phenolphthalein, behave as bases towards methyl orange,
and can be fairly accurately titrated with this indicator;” more accurate
results are obtained, in fact, as accurate as those with ammonia, with
the aliphatic amine bases and with alkaloids in aqueous solution.
Aluminium, ferric, chromic, and zinc salts are neutral towards methyl
orange, so that any content of free mineral acid can be determined,
although only approximately.
Methyl orange is particularly suitable for the estimation of alkalinity
* Cf. Lunge and Löhofer, Z. angew. Chem., 1901, I4, II 25.
* Lunge, Dingl, polyt.J., 1884, 251, 40; Chem. Ind., 1893, 16, 490.
METHYL ORANGE 65
in water analysis, i.e., of the carbonates of alkalis, calcium, and
magnesium.
As normal acid for the titration of alkaline solutions, hydrochloric
acid, nitric acid, or sulphuric acid may be used, but not oxalic
acid.
Methyl orange is, further, the best indicator for titrating the strong
mineral acids, hydrochloric, nitric, and sulphuric; the titration is
continued until the red colour gives place to the brownish mixed
colour. As standard solutions, sodium or potassium hydroxide, which
need not in this case be guarded from the absorption of carbon dioxide,
or sodium carbonate may be used; the sensitiveness is not quite so
great when the carbonate is employed. Baryta water may naturally
also be used, although there is no object in doing so in this case.
Since the transition from red to brownish is not quite so marked for
many people as the inverse change, the addition, in doubtful cases, of
a drop of normal acid to bring back the red colour may be used as a
check.
Thiosulphuric acid behaves in the same way as the above strong
mineral acids.
Methyl orange cannot be used for the titration of organic acids,
whether strong (oxalic acid, tartaric acid, citric acid) or weak (acetic
acid, etc.); in these cases the colour change is only gradual, and takes
place before complete neutralisation.
Some mineral acids of medium strength behave peculiarly. With
sulphurous acid, the change of colour with methyl-Orange takes place
when the reaction SO, + NaOH = NaHSOs is just completed, and
corresponds to the formation of the acid sulphite." Therefore I c.c.
of M/I sodium hydroxide indicates, not one, but two equivalents
(I molecule) of SO2=OO64O6 g. Inversely, normal sodium sulphite
can be titrated with hydrochloric acid according to the equation
NagSOa-i-HCl = NaHSO,--NaCl. The colour change takes place with
phenolphthalein on completion of the reaction SO,--2NaOH =
Na2SOs--H2O, so that in this case I c.c. M/I sodium hydroxide is
equivalent to Oo3203 g. SOs. Litmus gives intermediate values, and
so cannot be used in this case.”
The thiosulphates are neutral to methyl orange, so that sodium
thiosulphate does not react with it, while, on the contrary, normal
sodium sulphite is alkaline, and the acid sulphite, neutral.
The tribasic acids, phosphoric acid and arsenic acid, give a neutral
reaction with methyl orange when one of the hydrogen atoms is
saturated, corresponding to the formation of the compound NaH2PO,
* Lunge, Dingl polyt.J., 1883, 250, 530; Z, angew. Chem., 1890, 3, 563.
* Cf. Lunge, loc. cit.; Thomson, Chem. News, 1883, 47, 123 and 184; Blalez, Comptes rend.,
1886, 103, 69.
E
66 GENERAL METHODS USED IN TECHNICAL ANALYSIS
etc.; they behave in this case as monobasic acids, whilst they are
dibasic towards phenolphthalein, that is, the change of colour takes
place on the formation of Nag HPO, etc. Litmus gives indefinite results.
Therefore, phosphoric acid and arsenious acid, like sulphurous acid,
may be titrated with methyl orange as well as with phenolphthalein,
but the alkali has double the saturation value with the former indicator
that it has with the latter.
The behaviour of nitrous acid towards methyl orange is specially
to be noted. The colouring matter is gradually destroyed by free
nitrous acid, but small amounts, such as occur in chamber acid and in
other commercial sulphuric acids, have scarcely any disturbing effect
on the titration. In titrating nitrous vitriol or ordinary nitric acid
containing oxides of nitrogen, on the other hand, the colour of the
indicator disappears during the operation. This difficulty can, however,
be very readily overcome in two ways. The indicator may either be
renewed during the titration, or added only just before neutralisation,
or else normal sodium hydroxide may be added in excess, then the
indicator and the Solution titrated back with normal acid. In other
respects nitrous acid behaves towards methyl orange like the strong
mineral acids; that is, sodium nitrite has a neutral reaction, and one
molecule of nitrous acid is saturated by one molecule of sodium
hydroxide."
Alumina stands midway between acids and bases. Aluminium
sulphate is neutral to methyl orange, but alumina itself (in statu
mascendi) behaves as a base. In titrating sodium aluminate, either by
itself or as a constituent of impure sodium carbonate, the change of
colour of methyl orange only occurs when all the sodium carbonate
and all the aluminate has been changed to normal sulphate. Suspended
or colloidal dissolved aluminium hydroxide reacts only slowly and
incompletely, in the cold, and the titration should be carried out at
about 40°, the change of colour taking place when the alumina has
been converted into Al2(SO)s.” The Na,C) can thus be determined in
sodium aluminate by first titrating with normal acid in the presence
of phenolphthalein (which is indifferent to alumina), and then the
alumina with methyl orange.
Methyl orange can sometimes be used in conjunction with a second
indicator, e.g. phenolphthalein, to estimate different constituents in
one and the same solution, as in the analysis of sodium aluminate.
Thus, the amount of sodium carbonate in caustic soda can be
determined with sufficient accuracy for technical purposes, by titrating
first with phenolphthalein till colourless and then with methyl orange
* Lunge, Z, angew. Chem., 1903, 16, 59. -
* Cº. Cross and Bevan, J. Soc, Chem, Ind., 1891, Io, 202; also, Lunge, J. Soc, Chem, Ind., 1891,
10, 3:4.
LITMUS 67
till the red colour appears; the first indicator gives all the sodium
hydroxide and half the sodium carbonate, corresponding to the
formation of NaHCOs, the second gives the other half of the
carbonate." In the analysis of soap solutions Filsinger and Elsner
first determine the sodium combined with fatty acids, with normal acid
and methyl Orange and then the sodium chloride with silver nitrate,
in the same solution, using potassium chromate as indicator.
A mixture of methyl orange and phenolphthalein is coloured red both
by acids and bases, but retains its clear yellow colour in neutral solutions.”
The mother substance of methyl orange, dimethylaminoazobenzene,
which is soluble in alcohol, has been recommended by B. Fischer and
Philipp * instead of methyl orange; according to both Lunge 4 and
Thomson,” it offers no advantage over the latter, and is less sensitive ;"
this is also true of the ethyl compound (ethyl orange), which has been
preferred by some to methyl orange.
2. Litmus.
Litmus is supplied in small bluish cubes, which contain the actual
colouring matter, the blue potassium salt of the red azolitmic acid,
mixed with varying amounts of calcium carbonate, gypsum, etc.,
and sometimes adulterated with indigo. Litmus usually contains
4 to 5 per cent. of the actual colouring matter.
The general method formerly used in making litmus solution
consisted simply in pouring distilled water over the cubes (preferably
without grinding them), pouring off the first extract, which contained
excess of potassium carbonate, and using the subsequent extracts. As
this solution required a fairly large amount of acid to change the blue
colour to the neutral violet tint, it was customary to divide it into two
equal parts, and to add hydrochloric or other strong acid carefully to one
half until it became distinctly red (brick-red) and then mix it with the
other half; the mixture was of a violet to wine-red colour and sensitive
both to acids and alkalis.
Attempts have been made to overcome the defects of the solution
prepared as above, especially its lack of sensitiveness, by various
improvements; they all add, however, to the cost of an initially
expensive indicator, and have accordingly not been adopted in technical
laboratories. . According to Mohr-Classen,” the litmus cubes should
first be boiled three or four times with 85 per cent, alcohol, to remove a
foreign colouring matter. The first aqueous extract is also to be
* Cf. Lunge, Z, angew. Chem., 1897, Io, 41.
* Gawalowski, Zeit, anal. Chem., 1883, 22, 397.
* Arch. Pharm., 1885, 23, 434. * /. Soc. Chem. Ind., 1887, 6, 196.
* Ber., 1885, 18, 3290. * Lunge and Marmier, Z. angew. Chem., 1897, Io, 3.
* 7 Zehrbuch der chemischen analytischen Titrirmethode, 7th edition, 1896, p. 75.
68 GENERAL METHODS USED IN TECHNICAL ANALYSIS
rejected and hydrochloric acid added, drop by drop, to the subsequent
extracts until they assume a violet colour. The most sensitive solution
is prepared by first adding an excess of dilute sulphuric acid, boiling
to expel carbon dioxide, and then adding baryta water, drop by drop,
until a violet colour is produced.
Lüttke' recommends treating the aqueous extracts with hydro-
chloric acid and then removing the latter by dialysis. According to
Stolba, the colouring matter should be transferred to linen and washed,
then extracted with alkaline water and the solution neutralised
with acid. Reinitzer” states that only certain kinds of commercial
litmus can be used for analytical purposes. These should be ex-
tracted with water and concentrated hydrochloric acid added, drop
by drop, to the boiling solution until, after seven to eight minutes'
boiling, a distinct red colour is obtained. The solution is then allowed
to cool and an equal volume of strong alcohol added.
A specially pure litmus colouring matter, soluble in water, has been
introduced under the name of Ago/itmin, it is scarcely more sensitive
than good litmus solution, and is too expensive for use in works.
According to Böckmann, no plan in which too much colouring matter is
lost can under any circumstances be used technically, since a single
works may often use Ico-150 kilos of litmus.” -
Litmus solution spoils when kept in closed vessels, doubtless owing
to the development of anaërobic microbes. It
can, however, be kept good as long as desired
by allowing free access of air, which may be
done by keeping it in bottles which are not quite
filled, the entrance of dust being prevented by
a loose cap or by a pellet of cotton wool. The
most convenient plan is to provide the bottle
with a loosely fitting Cork, in which a small
pipette, for taking out any desired quantity of
solution, is inserted (Fig. 37). The same method
of storage, but using a well-fitting cork, is also
strongly to be recommended for other indicators.
Since there is no objection to keeping litmus
solution with access of air, addition of phenol
and similar substances is unnecessary.
Properly prepared litmus solution, provided it
is neutral, is an excellent indicator for bases, which change the violet
colour to blue, as well as for strong acids, which give a pure red (yellow-
red) colour. If the solutions in question are absolutely free from carbon
1. Cf. Zehrbuch der chemischen analytischen Zitrirmethode, 7th edition, 1896, p. 97.
* Z. angew. Chem., 1894, 7, 547.
3 For a special method of preparing litmus solution, ºf Förster, Z, anal. Chem., 1889, 28, 428.

LITMUS 69 -
dioxide, free alkalis can be titrated with acids and free, strong acids with
bases (potassium, sodium or ammonium hydroxide), with the greatest
accuracy. Reinitzer has shown that the greatest sensitiveness is
obtained with litmus, as is the case with methyl orange, by titrating in
the cold—a condition, however, which is not readily compatible with
the complete absence of carbon dioxide, which is necessary with this
indicator. Lungel has shown that the statement of Reinitzer, that
litmus is eight times as sensitive as methyl orange, is greatly
exaggerated; under the most favourable circumstances the sensitiveness
is only double that of methyl orange, and the advantage of this greater
sensitiveness only arises in exceptional circumstances, under the ordinary
conditions of work.
Besides the three mineral acids, sulphuric, hydrochloric and nitric,
oxalic acid also gives an equally sharp end-reaction with litmus. It is
in all cases advisable to titrate until a distinct blue colour is produced
and not to stop at a violet coloration, because the transition from red
to violet is not sharp, whilst that from violet to blue can be determined
to a single drop.
Litmus behaves quite differently towards weak acids. Carbonic
acid causes a pale port-wine red tint which renders the transition from
the blue, produced by alkalis, to the clear red produced by acids,
indistinct. The direct titration of alkaline carbonates cannot therefore
be carried out in the cold, but the solution must be heated to boiling
as soon as it turns red; if it becomes blue again, as is generally the case,
more acid must be added and the liquid again boiled, and this must be
continued until the colour is permanent and distinctly red. This
makes the operation very tedious. The whole titration should be
carried out in the boiling solution, were this possible, but for obvious
reasons the burette cannot well be fixed above the boiling liquid itself.
Since every kind of glass, even the so-called “resistance glass” and hard
Bohemian glass, give up alkali to boiling alkaline liquids, porcelain
vessels should be employed for the titration; but in using them, it
should not be overlooked that on continued boiling in basins, small losses
by spurting may readily occur without being noticed.
Reinitzer has shown that litmus is much less sensitive in hot than in
cold solutions; also, that even lime water and baryta water, as well as the
aqueous litmus extract itself, contain sufficient carbonic acid to cause
inaccurate results on titrating in the cold.
For the above reasons, recourse must be had to an indirect method
for titrating carbonates, if accurate results are desired. This is attained
by adding an excess of standard acid, boiling the solution to drive out
all the carbon dioxide, and then titrating back with standard alkali;
the standard alkali must be free from carbon dioxide, and the absorption
* Z. angew. Chem., 1894, 7, 733.
70 GENERAL METHODS USED IN TECHNICAL ANALYSIs
of carbon dioxide during cooling and titrating must be avoided.
Reinitzer has suggested that the cooling should be accelerated by
means of a jet of water, but this is scarcely practicable with porcelain
basins and is likely to cause breaking with most kinds of glass. In
the majority of cases it is preferable to titrate the solution whilst hot,
even at the expense of some loss in sensitiveness. º
Since the omission of any of the above precautions does away with
the greater sensitiveness of the reaction with litmus, and since the
necessity of boiling the solution leads to great loss of time, it is not
surprising that litmus, which was formerly in general use for the
titration of carbonates, has been to a very large extent discarded in
favour of methyl orange. Phenolphthalein is open to exactly the same
objections as litmus and has accordingly in no way replaced it.
Sulphuretted hydrogen destroys the colouring matter of litmus ; in
estimating sulphides, therefore, it is necessary to add excess of acid,
boil and titrate back with alkali.
Boric acid, phosphoric acid, etc., do not give a sharp end-reaction,
and the same is true of sulphurous acid (cf. methyl orange, p. 65), so
that the salts of these acids cannot be titrated with litmus ; on the
other hand, silicates give a good end-reaction, as the precipitated silicic
acid has no action on litmus.
Weak bases, such as aniline, quinoline, and many alkaloids, have no
action whatever on litmus. .
It is much more difficult to recognise the change of colour with
accuracy in artificial light than in daylight. The suggestion to titrate in
monochromatic light (sodium flame) cannot be regarded as practicable.
To summarise, litmus can be used very satisfactorily for the titration
of free, strong acids, but if great accuracy is required, the titration must
be made in the cold and with alkalis (sodium and potassium hydroxides,
ammonia, baryta water) free from carbon dioxide. The material of
the vessel in which the titration is done is then not of much importance.
But even in this case litmus is only preferable to methyl orange when
M/IO alkalis are used, which is very seldom the case in practice and if
the precautions with regard to the exclusion of carbon dioxide are
carefully observed throughout the whole operation. This is, however,
very difficult, as even baryta water may contain sufficient carbon dioxide
in Solution to make the end-reaction quite inaccurate.
Litmus cannot be used for the titration of salts of sulphurous, boric,
and phosphoric acids.
Litmus may be used for the titration of carbonates, but is by no
means to be specially recommended. It can only be used if the
carbon dioxide is driven out by continued boiling, and even then an
accurate result is only obtainable if the instructions given above are
followed.
PHENOLPHTHALEIN 71
3. Phenolphthalein. -
Phenolphthalein is a triphenylmethane derivative, being the
anhydride of dihydroxy-triphenyl carbinol-carboxylic acid, and has
the structural formula—
C.H., . OH
C–CeBI, . OH
N
NC, H, . CO. O
|
It was introduced as an indicator by Luck." It is a white crystalline
powder, melting at 150°, sparingly soluble in water but readily in alcohol.
It is used in I per cent. Solution of 90-95 per cent. alcohol, about two
drops being added for each titration ; an excess is not harmful, as is the
case with methyl orange. The colourless solution of phenolphthalein
is reddened by hydroxyl ions, that is, even by the smallest trace of an
alkali hydroxide, and on this account it is one of the most sensitive of
indicators ; its use is, however, very greatly restricted, owing to its
being too sensitive to the weakest acids, more particularly carbon
dioxide, which decolorise the red alkaline solution, just as strong
acids do.
The transition from red to colourless, produced by acids, is not so
delicate as the inverse change, which is very distinct even in artificial
light.
Phenolphthalein cannot be employed for the titration of ammonia,
nor in the presence of ammonium salts; also, it cannot be used if any
considerable amount of alcohol is present (cf. p. 57).
The real utility of phenolphthalein is, exclusive of its use as a
qualitative test for alkalis and alkaline earths, in the titration of
the weaker acids, from chromic acid and Oxalic acid down to
acetic acid. Methyl orange is quite useless in these cases, and litmus
is less sensitive. It is preferable to use barium hydroxide solution for
the titration, adopting all necessary precautions; if potassium or sodium
hydroxide is used, they must be quite free from carbon dioxide,
otherwise the estimation is altogether inaccurate. It is not always
possible to titrate in hot solution so as to exclude carbon dioxide, e.g.,
with acetic acid, on account of its volatility, nor with citric acid, etc.,
according to Thomson, since when heated, in solution, a normal salt
alkaline to phenolphthalein is formed, so that inaccurate results are
obtained. The fixed organic acids can be titrated with sodium
hydroxide containing a little carbonic acid, if the estimation is made
in the boiling solution; this is best done by adding a slight excess of
the alkali, boiling until the solution has become permanently red,
and then titrating back with standard acid.
* Z, anal. Chem., 1877, 16, 3202,
72 GENERAL METHODS USED IN TECHNICAL ANALYSIS
This operation is always tedious, and there is no object in adopting
it for the titration of strong mineral acids or alkalis, which are so much
more conveniently titrated with methyl orange in cold solution.
The point of neutralisation is reached with phosphates and arsenates
when the compounds Na, HPO, and Nag HASO, are formed; with
sulphites the normal salt is neutral towards phenolphthalein.
The red phenolphthalein solution is decolorised by a number of
substances which are very weakly or scarcely perceptibly acid towards
litmus, such as boric acid, arsenious acid, potassium bichromate, and
even gum arabic.
Boric acid can only be accurately titrated with phenolphthalein in
the presence of glycerine;" chromic acid can be very accurately esti-
mated with this indicator, the change of colour taking place on the
formation of K.CrO,
For the estimation of the hydrotides of the alkalis and alkaline
earths, phenolphthalein is undoubtedly the best and most sensitive
indicator, if these are present alone in solution and free from carbonate;
as far as the alkalis are concerned, however, this is seldom the case in
practice. In the presence of alkaline carbonates, the formation of
bicarbonates must be taken into acount (cf. infra); with the alkaline
earths, the carbonates of which are practically insoluble in water, it is
possible, according to C. Winkler (confirmed by Küster and Lunge, cf.
infra), by the careful addition of hydrochloric acid, to strike the point
when all the calcium or barium hydroxide is just saturated. The small
amount of carbon dioxide which is set free from the carbonates by the
next drop or two of standard acid, decolorises the phenolphthalein and
shows the end of the reaction. This method can also be adapted for
the estimation of alkali carbonates in the presence of alkali hydroxides
by adding chloride of barium in excess to the solution, all the carbonic
acid being thus combined with the barium ; this is, in fact, the most
accurate method for the purpose, but the hydrochloric acid must be
added slowly to the mixture, drop by drop, with continual thorough
stirring or shaking.
For the titration of alkaline carbonates, a distinction must be made
between titrating in hot and cold solution. In hot solution, carbon
dioxide is removed, and accordingly phenolphthalein acts exactly like
litmus, but the solution requires longer boiling. The change of colour
is much more striking than with methyl orange and the sensitiveness is
also theoretically greater, but the necessity of long boiling makes the
advantage, with regard to sensitiveness, quite illusory when glass vessels
are used, owing to the unavoidable action of the hot solution on the
latter; also the titration takes much longer than with methyl orange.
There is, therefore, no object in using phenolphthalein in this case.
* Z, angew. Chem., 1896, 9, 561 ; 1897, Io, 5; Igo2, 15, 733.
PHENOLPHTHALEIN 73
In the cold, so long as the carbon dioxide formed by the reaction
Na,COs--2HC1=2|NaCl-i-CO2+ H2O, has more than sufficient sodium
carbonate at its disposal to form bicarbonate, the red colour persists;
when, however, it is all changed to NaHCOs and free carbonic acid is
formed, the colour disappears, since it is destroyed by free carbon dioxide.
It was formerly supposed that this method could be used directly for deter-
mining the amount of alkali in normal carbonates," each equivalent of
normal acid of course indicating two equivalents of alkali (Na2CO3 +
HCl = NaCl-i-NaHCO3), and the reaction is still made use of, especially
for estimating sodium carbonate in presence of sodium hydroxide, by
first titrating with phenolphthalein till the solution becomes decolorised,
when all the hydroxide and half the carbonate are saturated, and then
either completing the titration in boiling solution, or, more conveniently
(cf. p. 66), by adding methyl orange and finishing the titration cold.
The result of the second titration, multiplied by two, gives the amount of
Na,COs originally present; the amount of NaOH is obtained from the
difference between the first and second titrations.
It has been shown by Küster” that this estimation is not quite
accurate, and he has given an explanation of his results, based on the
electrolytic dissociation theory. Lunge” has shown that the method
is sufficiently accurate for the estimation of a small quantity of sodium
carbonate in the presence of much sodium hydroxide, e.g., in caustic
soda (cf. p. 66), but not for much carbonate, with a small quantity of
hydroxide. This point was subsequently more accurately investigated
by Lunge and Lohöfer.” It was found that, as previously shown by
Küster, the titration gives results nearer the correct value if sodium
chloride is added, the introduction of ions of the same kind hindering
the ionisation of the sodium carbonate; if the solutions are not too
concentrated, results accurate to o, I c.c. are obtained if the ratio of
Sodium chloride to sodium carbonate is at least I-75 to O. 5 molecules.
This holds with increasing amounts of sodium chloride, but when ten
molecules and upwards are present for O. 5 molecule of carbonate, it is
found that too little hydrochloric acid is used, which is in accord with
the ionisation theory. This explains why, in the titration of caustic
soda—that is, a large quantity of hydroxide and a little carbonate—the
latter can be accurately estimated by Warder's method, since in this
case a sufficient amount of sodium chloride is formed from the
hydroxide, before the action of the hydrochloric acid on the carbonate
begins.
Lunge has recently found that, in order to get accurate results, it is
not only necessary to observe the above condition (excess of sodium
chloride), but that the titration must be carried out in as concentrated a
* Warder, Amer. Chem. J., 1881, 3, 55. * Z, angew. Chem., 1897, Io, 41.
* Z, anorg, Chem., 1897, 13, 14.I. * /bid., 1901, I4, II 25.
74 GENERAL METHODS USED IN TECHNICAL ANALYSIS
solution and at as low a temperature (little over O’) as possible; all of
these conditions diminish the ionisation of the sodium carbonate.
Further, the titration must be done rapidly, so as to avoid escape or
absorption of carbon dioxide.
The estimation of sodium carbonate in the presence of sodium
bicarbonate, using phenolphthalein as indicator, is carried out in the
same way. The titration is done in the cold until the solution becomes
decolorised, which takes place when half the carbonate present is
changed into bicarbonate; the liquid is then heated to boiling, and
titrated till the colour again disappears. The number of c.c. of acid
required for the first decolorisation, multiplied by 2, gives the amount
of carbonate present, and the difference between this and the total c.c.
required gives the number of c.c. corresponding to the bicarbonate."
A simpler plan is to carry out the first titration in the cold solution with
phenolphthalein, and the second likewise in the cold, after the addition
of methyl orange.
Thomson, and subsequently Lunge and Lohöfer, have shown that
sodium silicate cannot be titrated alone, with phenolphthalein as
indicator, with any degree of accuracy; the change of colour is gradual
and cannot be sharply determined, and it always takes place long
before the completion of the reaction Na,SiO2+2HC1=2|NaCl-F H2SiO,
Sharp results are obtained, however, in accord with this equation if
much sodium hydroxide is present, from which sodium chloride can be
formed, or if sodium chloride is added directly, apparently because
sodium silicate is hydrolytically dissociated and acts towards phenol-
phthalein as a weak acid as long as it is present as hydrosol ; but in
presence of sodium chloride, the silicic acid passes into the hydrogel,
and has then no action on phenolphthalein.
Titration with phenolphthalein in alcoholic solution has been studied
by Hirsch’ and by Schmatolla.”
Potassium or sodium sulphide can be titrated with acid in the presence
of phenolphthalein. The colour disappears when half the sodium has
been changed to NaHS, so that the relations are completely analogous
to those with carbonates:—
HCl + NagS = NaHS + NaCl.
Aniline, toluidine, quinoline, and most alkaloids have no action on
phenolphthalein, and their salts can therefore be titrated with normal
sodium hydroxide like free acids.
Thomson 4 has collected many of the above facts in the following
table (altered in some particulars by Lunge). The figures give the
number of hydrogen atoms, in the acid molecule, replaced by sodium
* Cf. Kippenberger, Z. angew. Chem., 1894, 7, 495. * Zºid., 1902, 35, 3905.
* Ber., 1902, 35, 2374. * / Soc. Chem. Ind., 1887, 6, 198.
OTHER INDICATORS 75
when the solution becomes neutral ; cases in which the end-point cannot
be sharply recognised with the indicator are represented by a line in
the table.
Table of the Basicity of Acids towards different Indicators in
titrating with Sodium Hydroxide.
Methyl orange. Phenolphthalein. Litmus.
Acids.
In the Cold. In the Cold. Boiling. In the Cold. | Boiling.
Sulphuric Acid, H2SO4 .
Hydrochloric Acid, HCl .
Nitric Acid, HNO3 . e
Thiosulphuric Acid, H2S2O3
Carbonic Acid, H2CO3 .
Sulphurous Acid, H2SO3 .
Phosphoric Acid, HaPO4 .
Arsenic Acid, HaAsO4 .
Arsenious Acid, H3AsO3 .
Nitrous Acid, HNO2
Silicic Acid, H2SiO3 *
Boric Acid, HaBO3 .
Chromic Acid, H2CrO4
Oxalic Acid, H2C2O4
Acetic Acid, HC2H5O2 .
Butyric Acid, HC3H7O2 .
Succinic Acid, H2C4H4O4.
Lactic Acid, HC3H5O3 .
Tartaric Acid, H2C4H4O6.
Citric Acid, H2C6H5O1
i
1 (cf. p. 72)
1
(c
:/.
2
1
1
2
0
66) 1
0
2
1 (about)
1 } }
2
3 y
2
1
1
2
0
2
2
mºT:
The table holds for ammonium hydroxide as well as for sodium
hydroxide, with the exception of phenolphthalein, which, as stated, does
not give a sharp end-point with the former.
Other Indicators.
Very many other indicators are known, but they are, as a matter of
fact, superfluous for most purposes; their supposed advantages are in
part quite illusory and in part counterbalanced by drawbacks. The
following are the most important.
Indicators of the first class:–
Lacmoid, prepared and recommended as an indicator by Traub and
Hock" and by Förster,” is used in alcoholic solution. It has far more
analogy with methyl orange than with litmus, resembling the latter only
in being turned red by acids and blue by alkalis. It is considerably
more sensitive than litmus, especially when used as test paper. It is not
suitable for the titration of carbonates, sulphites, and sulphides, but can
be used for borates and silicates. Neutral chromates react strongly
alkaline to it, acid chromates neutral, so that it can be used, in the form
* Ber., 1894, 27, 2615. * Z, angew. Chem., 1890, 3, 163.
76 GENERAL METHODS USED IN TECHNICAL ANALYSIS
of test paper, to detect the presence of neutral chromate mixed with
acid chromate (litmus paper cannot be used for this purpose.) It is use-
less for titrating organic acids, and cannot be used in presence of nitrous
acid or of sulphuretted hydrogen.
According to Messner, lacmoid is the best indicator for titrating
cinchona alkaloids in alcoholic solution.
Blue lacmoid paper is also distinctly blue in artificial light, whilst
litmus paper is violet; the former is scarcely more sensitive than the
latter. Red lacmoid paper, which must be protected against access of
air, is more sensitive than red litmus, and retains its colour better.
Various ago-dyes behave like methyl orange, but are all inferior to it.
Reference has already been made to dimethylaminoazobenzene, and
ethyl orange (p. 67). Tropāo/in OO (orange IV., diphenylamine Orange)
was recommended by von Miller” even before the introduction of methyl
orange. Like methyl orange, it is turned red by the stronger acids and is
not altered by carbonic acid ; it is, however, so much less sensitive than
methyl orange that is must be regarded as quite useless for work of
any degree of accuracy. Diaminoagotoluenesulphonic acid, described by
Troeger and Hille,” seems to act very similarly to methyl orange.
Rubrescin is a new indicator discovered by Rosenfeld and Silber.” It
is prepared by melting together 50 g. resorcinol and 25 g. chloral
hydrate in an oil bath at 160°, the powdered mass being purified by
treatment with warm chloroform. The indicator has a strongly acid
character, and is much more sensitive towards alkalis than phenolphtha-
lein ; one drop of M/IO sodium hydroxide in IOO c.c. of water retains its
red colour with this indicator for an hour, whilst the red colour of
phenolphthalein disappears in a few seconds under the same conditions.
With borax and sodium carbonate it is equally sensitive. If the colour
is not very intense it disappears on addition of acid, if of greater intensity
it changes to yellow.
Iodo-easin, also known commercially as erythrosin or pyrosin, is
tetraiodofluorescein. It is not used in ordinary alkalimetry, but for the
detection of very small amounts of alkali, as, for example, that dis-
solved from glass by the action of water.” A solution of 2 mg. in
IOoo c.c. of carefully purified ether containing a little water, is used ; Io-
20 c.c. of the solution is added to 50-IOO c.c. of the solution to be tested
and the mixture shaken. In the presence of free alkali the lower layer
becomes red ; the colour disappears on addition of acid. This indicator
may also be used colorimetrically, and is particularly suitable for the
estimation of alkaloids. According to Glücksmann," it is indifferent
towards carbonic acid.
* Z. angew. Chem., Igo3, 16, 449. * Chem. Zeit. A’ep., 1902, 26, 130.
* Ber., I878, II, 460. * Mylius and Förster, Z. anal. Chem., 1892, 31, 248.
* /. prakt. Chem., 1903 [2], 68, 297. * Chem. Centr., Igo?, II, 476.
OTHER INDICATORS - 77
Ferric salicylate is recommended by J. Wolf as an indicator for strong
acids, in the presence of boric acid, and thus permits of the subsequent
determination of the latter. On Saturation with sulphuric acid, the
violet colour changes to an alizarin orange. According to Lunge,' this
indicator is in no way superior to methyl orange.
Congo red is turned blue by acids; the red colour is restored by
alkalis. Free carbonic acid colours it bluish-violet, acid salts have no
action, so that it might be employed, e.g., for estimating free acid in
aluminium sulphate. Although this indicator was at first regarded as
of considerable value, it was soon found to be useless for most purposes
and not very sensitive. According to Thomson,” it is unreliable,
both because the change of colour is not sharp and because the
uncertainty of the end-point is increased by the presence of salts,
such as sulphates, chlorides, and nitrates of the alkalis; also, it is
less sensitive for the detection of free acid than all the other indi-
cators that are suitable for this purpose. The former opinion, that
Congo red test paper is as good as or even preferable to litmus paper,
has been similarly disproved. Congo red cannot be used in hot solu-
tion, nor in the presence of nitrous acid, nor for the titration of weak
acids.
Para-nitrophenol has been recommended by many workers, especially
by Spiegel,” in place of methyl orange. It has a yellow colour in
alkaline solution, which disappears on neutralisation with strong acids.
It is only slightly acted upon by carbonic acid, but distinctly more so than
methyl orange; * Glaser (p. 55) therefore includes it among indicators
of the second class. For this reason it is by no means so useful as
methyl orange; further, the disappearance of the yellow colour or its
first appearance takes place so gradually that it can only be observed
with accuracy by a quick eye in clear daylight, and then only with
considerable effort. According to Glaser, it can be used advantageously
for indicating the continued alkaline reaction after the addition of soda,
etc., to the water in a steam boiler, because, unlike other indicators, it
is not decomposed even at I 50°.
Goldberg and Naumann” obtained better results by using a fairly
large amount of the indicator; they confirm its sensitiveness to carbonic
acid, however, as well as the observation that the change of colour is
much more difficult to recognise than with methyl orange.
Ortho-nitrophenol is, according to Lunge," less useful than the para-
compound.
Tincture of Cochineal, when properly prepared, is, according to
* Z. angew. Chem., Igo4, 17, 203. * J. Soc. Chem. Ind., 1887, 6, 175.
* Ber., 1900, 33, 2640; Z. angew. Chem., 1904, 17, 715.
* Lunge, Z, angew. Chem., 1903, 16, 560; 1904, 17, 202.
* Z. angew. Chem., 1903, 16, 644. * Ibid., 1904, 17, 202.
78 GENERAL METHODS USED IN TECHNICAL ANALYSIS
Dupré, almost as sensitive as phenolphthalein; it is but little affected
by carbon dioxide.
Indicators of the third class :-
Turmeric is only of importance as a test paper. The yellow paper
is turned red-brown by alkalis, but it is much less sensitive than litmus.
It is of practical value for the titration of dark-coloured solutions of
organic acids for which other indicators cannot be used, and in the
detection of boric acid.
Poirrier's blue, C 4B, would fill a great blank, if it could be more
generally employed.” Its colour is altered even by the very weakest
acids, which have no effect on any of the other indicators. Borax,
which is alkaline to all other indicators, has an acid reaction to
Poirrier's blue, which only disappears on complete neutralisation of
the boric acid. In the same way, monohydrogen salts of phosphoric
acid, which are alkaline to litmus and methyl orange and neutral to
phenolphthalein, give an acid reaction, and in this case also the
change of colour only takes place on complete neutralisation of the
phosphoric acid. Arsenic acid behaves in an exactly analogous way.
It has been proposed to use this indicator for the determination of alkali
hydroxide in the presence of alkali carbonate, but since this is only
practicable in concentrated solution, with any degree of accuracy, on
account of the great irregularity due to the hydrolysis of the indicator,
it is no real value for this purpose.
It is evident from the above that different indicators not infrequently
give different reactions with one and the same substance. Thus, neutral
potassium chromate and sodium acetate are neutral to phenolphthalein,
weakly alkaline to litmus and turmeric, whilst the latter is acid to
Poirrier's blue. Monohydrogen salts of phosphoric and arsenic acid,
which are alkaline to litmus and methyl orange, are neutral to phenol-
phthalein, and acid to Poirrier's blue. Borax is acid to Poirrier's blue,
but alkaline to all other indicators. Boric acid, which turns turmeric
paper red, similarly to alkalis, has no action on methyl orange, so that
the acid can be titrated with borax solution, using methyl orange as
indicator. Urine, which is usually acid to litmus and phenolphthalein,
is strongly alkaline to lacmoid. Fresh milk, which gives a neutral
or amphoteric reaction with litmus, is distinctly alkaline to lacmoid
and distinctly acid to phenolphthalein.
The sensitiveness of the more important indicators has been deter-
mined by precise physical methods both by Friendenthal” and by
Salm.4
1 Private communication to Prof. Lunge. -
* Engel and Ville, Comptes rend., 1885, Ioo, Io'ſ 3; Engel, Z. anal. Chem., 1888, 27, 30.
* Z. Electrochem., 1904, Io, II.3. - -
* Ibid., 1904, Io, 341 ; Z. physik. Chem., 1906, 57, 47 I.
TEST PAPERS 79
Test Papers.
These are ordinarily employed for indicating the acidic or basic
character of a liquid, and for this purpose are made, conversely,
either weakly basic or acidic. Fine white writing paper is usually
prescribed for their preparation (Mohr), because a drop of liquid
placed on it does not spread too much. According to Glaser, filter
paper is preferable for use with solutions which are coloured or which
contain salts, which disturb the reaction, since in such cases the drop
spreads out, leaving the colouring matter or salt in the centre, and a
sharper reaction is obtained at the edge. Paper used for this purpose
must previously be well washed and dried.
When writing paper is used, the solution of the indicator is spread
on it, whilst filter paper is saturated by dipping it into the solution;
the paper is dried by hanging it up in an atmosphere free from acid
and ammoniacal vapours. Papers prepared with indicators of the first
class are most suitable for weak bases, with those of the third class for
weak acids; for strong bases and acids they are all equally sensitive.
Glaser gives the following table, showing at what dilution the acid
or alkaline reaction just ceases to be noticeable:—

Sulphuric Sodium Ammonium
Acid. Hydroxide. Hydroxide.
Yellow Methyl orange paper . Tº M.
Congo Red paper . tº wº Twºwo AV.
Blue Lacmoid paper & g Twº wo AV. & © tº º tº
Red Lacmoid paper © tº º e gºw M. rºw Av.
Blue Litmus paper . gº dº Twº AV. e & e e tº º
Red Litmus paper . tº e tº º gºw M. sºon M.
Violet Azolitmin paper . g atºw AW. tº A. Fºrd Av.
Turmeric paper . ſº & © tº $ ºw M. ºw M.
Phenolphthalein paper . . tº gº tº gºw A'. gºw A/
He states, however, that the sensitiveness may, in individual cases,
be very different from the figures given above, owing to the method of
preparation of the test papers.
Test papers are mostly used for the qualitative detection of acids or
bases; for quantitative purposes they are used only with coloured
liquids, or with such as destroy indicators. gº.
Test papers should always be kept in well-closed vessels, and
protected from light, either in tightly fitting wood or metal boxes, or
in glass bottles covered with black paper; they are soon spoiled by
moisture.
Litmus paper, blue, red, and violet (neutral), is by far the most widely
used test paper. To prepare it, purified litmus Solution (cf. pp. 67-68),
or a solution of azolitmin (O. I g. in IOO c.c.) is used. These should
80 GENERAL METHODS USED IN TECHNICAL ANALYSIS
colour the paper a faint blue; if not, a trace of alkali may be added to
the solution. For red paper, a trace of acid is of course added, and the
neutral paper is prepared from a mixture of the two solutions. A
sample should always be made and dried in order to be sure of the
right colour, since a paper which is violet when wet becomes blue on
drying. Neutral azolitmin paper is used for titrating weak acids.
Both yellow and red methyl orange paper is used. The yellow,
prepared directly from methyl orange solution, should only be faintly
coloured; it is employed for the detection of free mineral acids in the
presence of weak acids, e.g., free sulphuric acid in acetic acid, aluminium
sulphate, zinc sulphate, etc., and in the titration of pyridine. The red
paper, prepared with an acidified solution of the dye, can be used for
the detection of bases, but it is not very sensitive.
Colourless phenolphthalein paper serves for the detection of free
bases, but it is not as sensitive as red litmus paper. Red phenol-
phthalein paper is so readily acted upon by the carbon dioxide of the
air and by light, that it is of no practical value.
Turmeric paper, prepared from the alcoholic extract of turmeric root,
turns brown with alkalis (free or as carbonates), and with alkaline
earths, but it is not very sensitive ; its real use is for the detection of
boric acid and salts of uranium.
Congo red paper, which is turned blue by alkalis and red by acids,
is not very sensitive.
Lacmoid paper has, for most purposes, scarcely any advantage over
litmus paper. It is used for the stronger organic acids and in testing
for acid chromates, which in the presence of neutral chromate turn red
lacmoid paper blue, and in the presence of free chromic acid turn the
blue paper red, the colours making their appearance after the coloured
chromate solution has been washed away. Red lacmoid paper has a
very great tendency to turn blue, even with the moisture of the
hand.
NORMAL SOLUTIONS
A large number of standard solutions are used in technical analysis;
those employed only for special purposes are treated of in the sections
in which their applications are discussed, the present section being
restricted to a description of the standard solutions in general use, and
which are employed in various branches of technical work.
General considerations.—Normal solutions are prepared either
according to equivalents or according to units of weight of the sub-
stance to be estimated, or if they readily alter in strength on keeping,
they are only approximately standardised and titrated anew for each
series of experiments.
The first method is generally adopted. By “normal solution” in
NORMAL SOLUTIONS 81
the narrower sense, that is an M/I solution, is to be understood a solution
of which one litre corresponds to a hydrogen equivalent, in grams, of
the substance to be examined ; ; normal, # normal, and I'm normal
solutions (M/2, M/5, AW/IO) are equivalent to proportionately smaller
amounts of the substance.
The ratio of the gram equivalents refers usually, but not always,
also to the composition of the normal solution itself. Thus an M/I
hydrochloric acid solution cqntains 36.461 g. HCl perlitre; an M/I sulphuric
acid which contains two replaceable hydrogen atoms = 49-O4 g. H2SO4,
and so on, but this is really only accidental. It is not a question of
what the normal solution contains, but what it corresponds to.
An AV/I hydrochloric acid solution must always correspond to a gram
equivalent of a base, e.g. to 40.06 g. NaOH or l;1 = 85.72 g. Ba(OH),
which will, of course, be the case when it contains a gram equivalent
HCl = 36.461 g. in the litre. But an AW/IO potassium permanganate
solution is not one which contains ºth equivalent KMnO, =
I 58.2
-is = i 5-82 g. per litre, but one which yields Hoth equivalent of
Oxygen =
zio-o.8 g. O per litre. Since, according to the equa-
tion :-
2 KMnO, +3H,SO = K,SO, + 2MnSO, +3H2O + 50,
two molecules KMnO, yield 5 × 16 = 80 O, an M/Io permanganate solu-
tº te O.8 2 × I 58.2
tion must contain 2 × I 58.2 × − =
8O IOO
permanganate, therefore only a fifth of the above amount of 15.82 g.
Normal solutions, prepared according to the chemical equivalents,
have the merit of being applicable, without further calculation, to all
cases in which they can be employed. Thus I c.c. of AW/I sodium
hydroxide corresponds to O.O.3646 g. HCl, or O.O4904 g. H2SO, or
O.O63O5 g. HNO, ; I c.c. N/I HCl, or N/I HASO, corresponds to O.O.3105 g.
Na2O, O.O4OO6 g. NaOH, O,O5305 g. Na,COs, O.O5616 g. KOH, or
O-O6915 g. K2COs, etc. An AV/IO permanganate solution corresponds to
O.OO5588 g. iron, or O.OO63O2 g. oxalic acid, etc. It is, therefore,
customary to prepare such standard solutions as are applicable for a
variety of purposes, in the ratio of their chemical equivalents.
An alternative method is preferable in special cases with solutions
of this character. If, for example, it is only necessary to determine
whether a commercial product is of a definite strength, it is more
convenient, and saves time, to make up a volumetric solution to corre-
spond to a simple quantity by weight of the substance in question, say,
I, 5, or IO g. per litre.
This mode of preparing volumetric solutions is particularly suitable,
and is to be generally recommended when the solution is employed
F
= 3. I64 g. of pure potassium
82 GENERAL METHODS USED IN TECHNICAL ANALYSIS
exclusively or almost exclusively for the estimation of a simple
substance, such as a silver nitrate solution for the determination of
sodium chloride in products of the Soda or salt industry, in Chili salt-
petre, in drinking water, in boiler feed-water, in sulphates, etc.; or,
conversely, when sodium chloride is used for estimating silver. In
such cases there is no advantage whatever in making up solutions
according to equivalents, as for example an AV/IO silver nitrate solution,
each c.c. of which corresponds to O-O 585 g. of sodium chloride, since a
calculation, which cannot readily be made mentally, is then necessary
for each titration; it is far preferable to prepare a solution containing,
e.g., 2.906 g. silver nitrate per litre, so that each c.c. then corresponds to
O.OOI g. NaCl.
A third method is to make up the volumetric solution, not to
correspond to chemical formulae at all, but according to empirical tests.
In the estimation of tanning materials with permanganate, according to
Löwenthal's method, for example, it would be quite useless to weigh
out an amount of permanganate yielding the oxygen requisite for
oxidation, basing the calculation on the chemical formula of pure
tannin; the amount of permanganate required is not the same under
all circumstances, and cannot be calculated directly from the formula as
with inorganic or simple organic substances, but depends upon how
the estimation is carried out, on the temperature, and on the rate at
which the oxidising substance is added ; the titre of the permanganate
solution, prepared by weighing out a simple, arbitrary quantity, must
accordingly be determined empirically with an actual tannin, skin
powder, etc.
STANDARD ACIDS
Hydrochloric acid, and for certain purposes ovalic acid, are the only
acids really required for use as standard solutions; the latter cannot be
used for alkalimetry. Sulphuric acid was formerly in general use
as a standard acid, but it has been very largely displaced by
hydrochloric acid, which has the threefold advantage that it is more
generally applicable, as, for example, in the titration of the alkaline
earths, that its strength can be controlled gravimetrically with silver
nitrate much more accurately than that of sulphuric acid with barium
chloride, and that it has a much greater avidity in the cold than sulphuric
acid ; further, as it is a monobasic acid, any disturbing influences such as
are caused by the hydrolysis of acid sulphates are avoided.
There is no advantage to be gained by using sulphuric acid as
a standard acid, and it can therefore be dispensed with altogether; it is
prepared in exactly the same way as normal hydrochloric acid.
AVormal mitric acid can also be substituted by hydrochloric acid,
STANDARD ACIDS 83
except for the estimation of chlorides, in an alkaline solution (for
example, crude caustic soda) with silver nitrate, using potassium
chromate as indicator; in this case an amount of standard nitric acid
just sufficient to neutralise the alkali is added, this amount having been
found from a previous titration with hydrochloric acid. For this purpose,
however, a standardised nitric acid is not really necessary if methyl
orange is used as indicator in the neutralisation; or, if an excess of acid
has been added, it can be neutralised by a slight excess of Sodium
hydroxide or sodium carbonate, without disturbing the subsequent
titration with silver nitrate.
Standard ovalic acid was brought forward by F. Mohr as a “basis
of alkalimetry,” but it has so many drawbacks that it is now but
seldom employed, except for the standardisation of solutions, in which
respect, however, it is much inferior to sodium carbonate. It is used
for the preliminary standardisation of potassium permanganate
solutions, and for estimating the “base” in Weldon mud.
Many different substances have been suggested as a basis or
standard for acidimetry and alkalimetry, of which the following are the
most important.
Sodium carbonate takes precedence of all the others, both on account
of its extensive use and of its actual merit. It can be prepared and
weighed out absolutely pure and free from water with the greatest ease,
and can be very accurately titrated with hydrochloric acid and methyl
orange, and, with proper precautions, with litmus or with phenolphthalein.
“ Chemically pure sodium carbonate,” free from any appreciable amount
of chloride, can be obtained commercially; it can also be readily
prepared from sodium bicarbonate which has been freed from chloride
and Sulphate, if present, by washing with small amounts of distilled
water. The sodium carbonate must be quite free from sodium oxide
as well as from water.
If pure commercial sodium carbonate is used as a standard, it must
dissolve in water to a perfectly clear solution, give no turbidity with
silver nitrate after neutralisation with nitric acid, and no reaction for
sulphates, after excess of hydrochloric acid has been added. In
applying these tests a sufficient amount must be dissolved, say 2 to 3 g.,
and the solution sufficiently diluted, in order to avoid precipitation of
barium chloride. A very slight opalescence with silver nitrate is often
noticed, but it is easy to judge if an amount of chloride sufficient for
quantitative estimation is present. If the sodium carbonate is pure, it
is dried by heating in a platinum crucible, with frequent stirring, for
twenty minutes, the temperature being so regulated that only the
bottom of the crucible is red hot; it should not form into hard lumps
during the heating, as this may indicate the formation of a little
sodium oxide. The ignited substance is shaken into a well-stoppered
84 GENERAL METHODS USED IN TECHNICAL ANALYSIS
weighing bottle whilst still warm, and kept in a desiccator; or the
platinum crucible is allowed to cool in the desiccator and the several
(three to five) portions for titration weighed directly into beakers, the
crucible being kept covered during the weighings.
In order to be quite certain that the sodium carbonate has been ignited
sufficiently but not too much, the crucible should preferably be heated
in a sand bath to 270° to 3OO", with continual stirring, until the weight is
constant ; half an hour is sufficient for this purpose, exclusive of the
time required to attain the temperature. The sand is packed round the
outside of the crucible to the same height as the sodium carbonate
inside, and a thermometer is placed quite close to, or even in, the crucible.
This method, which is quite as convenient as heating with the naked
flame, is recommended by Lunge.
Sodium bicarbonate can be obtained commercially of a sufficient
degree of purity for the preparation of the carbonate ; it should be
tested, as above, for the presence of insoluble impurities, chlorides and
sulphates, of which the two latter can be removed by washing, owing to
the slight solubility of the bicarbonate. The second equivalent of
carbon dioxide and the water can then be driven off by heating in a
platinum crucible, as in the case of sodium carbonate ; but a more
reliable result is attained by heating to a moderate temperature, which
can be regulated. Lunge" has shown that the bicarbonate is completely
converted into the carbonate by heating, for a few minutes, at 260° to
270°; if heated for from a half to one hour in a sand bath or in an air
bath at a temperature not over 300°, the carbonate can be relied on as
being free from bicarbonate, water, and sodium oxide.
Although sufficiently pure bicarbonate can readily be obtained
commercially, yet it may at times be necessary to purify an impure
product, so as to obtain a pure substance for standardising purposes.
Reinitzer” gives the following method of purification. About 250 c.c.
of distilled water is heated to 80° in a tall beaker (preferably of Jena
glass) and the bicarbonate added, in small quantities at a time, till the
solution is saturated, the liquid being continually stirred ; during this
process, part of the carbon dioxide is evolved, with effervescence. When
nothing further dissolves, the solution is filtered through a pleated filter
paper, preferably in a hot-water funnel, into a flask and cooled to Io’ to
I5°. In this way a considerable amount of a coarsely crystalline salt is
thrown down, which is a mixture of bicarbonate and trona,” and which
can be separated from the mother liquor by filtration through a funnel
fitted with a platinum cone or with a loosely fitting glass stopper; it is
preferable not to use a filter paper, so as to avoid contamination with
fibres of paper. The mother liquor is pumped off and the substance
* Z. angew. Chem., 1897, Io, 522. * /bid., 1894, 7, 551.
* # Carbonate. Cf. Z. angew. Chem., 1893, 6,446 and 573.
STANDARD ACIDs 85
washed several times with small quantities of cold water, which are
completely removed by suction after each addition. The salt is then
transferred to a platinum crucible, which is heated to such a temperature
that it is just visibly red hot in daylight; after cooling, it is powdered
and again dried. If this method is employed, it is difficult, when dealing
with large quantities, to avoid overheating some portions, and thus
producing traces of sodium oxide. It is therefore better simply to dry
the salt thoroughly, grind and mix together, and then to heat the small
quantities to be employed in standardising, in a sand bath or air bath, to
a temperature not exceeding 3OO", as above.
For standardising or controlling the strength of a normal acid, the
corresponding quantities of sodium carbonate must always be weighed
out separately; the plan of preparing a normal sodium carbonate
solution containing say 53 g. to the litre and withdrawing aliquot
portions for the different estimations, is unsatisfactory. Errors in
measuring, in the emptying of pipettes, etc., cannot be avoided, even
when standardised apparatus is employed, and the inaccuracy arising
from these causes is much greater than errors of weighing.
The possibility of preparing sodium carbonate quite free from
moisture and from Sodium hydroxide has, from time to time, been
questioned by various investigators." Ilunge” has, however, shown that
the product obtained by the method of ignition given above is quite free
from moisture, and that the quantity of sodium hydroxide formed is
negligible for all ordinary analytical work; in a special experiment to
test this point, he found only O.OO4 per cent. of the hydroxide in a
sample of carbonate dried at 270° to 300°.
Sörensen “has recommended the use of sodium oxalate as the basis
for alkalimetry instead of sodium carbonate; it has the advantage of
being applicable also as a standard for oxidation methods of volumetric
analysis. The salt can readily be obtained free from water of crystal-
lisation, according to Sörensen, and is not hygroscopic. As prepared
according to his directions, by Kahlbaum, by precipitation with alcohol
and drying at 240°, it may be employed for standardising purposes, either
directly or after a few hours' heating in the steam oven. An accurately
weighed quantity of the salt is carefully heated, for a quarter to half an
hour, in a covered platinum crucible, over a small gas flame, supported
in a hole in an asbestos card, so as to avoid contamination by sulphur
compounds from combustion, or over a spirit lamp," until the sodium
carbonate formed just begins to fuse; towards the end of the operation
* Cf. C. L. Higgins, J. Soc. Chem. Ind., 1900, 19, 958; Sörensen and Andersen, Z. anal. Chem.,
1905, 44, 156; B. North and W. Blakey, J. Soc. Chem. Ind., 1905, 24, 396; Sebelien, Chem.
Zeit., 1905, 29, 638.
* Z. angew. Chem., 1904, 17, 231 ; Igoš, 18, 1520.
* Z, anal. Chem., 1897, 36, 639; Igo3, 42, 333; 1905, 44, 156.
* Cf. Lunge, Z, angew. Chem., 1905, 18, 1520.
86 GENERAL METHODS USED IN TECHNICAL ANALYSIs
the lid is moved aside so as to cover about one-half of the crucible,
to promote the complete combustion of the carbon formed. A
mixture of sodium carbonate, containing a little sodium hydroxide, is
thus formed, which is not weighed, but transferred along with the
crucible to a tall beaker, moistened with water, treated with an excess of
the acid to be standardised and warmed on the water bath; the solution
is then poured into a conical flask, the crucible and cover and the
beaker rinsed out, ten drops of a solution of O.5 g. phenolphthalein in
50 c.c. alcohol--50 c.c. water added and boiled, till the carbon dioxide is
completely expelled, in a current of air freed from carbon dioxide, cooled
in cold water, and the excess of acid titrated back with W/Io sodium
hydroxide. The equivalent of sodium oxalate is ** =67.05, so that
Io c.c. of an AV/IO acid correspond to O.O67O5 g. of sodium oxalate; or O. I g.
of the salt = 14.91 c.c. of an M/IO acid. In a subsequent paper, Sörensen
gives accurate directions for the examination of commercial “pure”
sodium oxalate, for impurities.
Sörensen's investigations were carried out with great care, and they
show that very good results are obtainable with sodium oxalate, as a
basis for alkalimetry. Whether the ignition of the oxalate with com-
plete combustion of the carbon (regarded by Sörensen himself as
absolutely necessary) can be performed without loss by those without
special experience in the method, requires confirmation. The examina-
tion of the substance as regards absolute purity is more troublesome and
tedious than in the case of sodium carbonate, and this is true of the
whole process, even if the titration is not carried out with phenol-
phthalein in boiling solution as above, which is quite unnecessary, but
with methyl orange in the cold.” There is, therefore, no reason for
giving sodium oxalate the preference over sodium carbonate, but it can
be advantageously used as a control in doubtful cases.
To prepare an AV/I hydrochloric acid solution, pure hydrochloric acid
is diluted to a specific gravity of about 1.O2O, so as to obtain a pre-
liminary acid somewhat over the normal strength (36.46 g. HCl per
litre). This acid is then titrated against a freshly ignited sample of
sodium carbonate. If w is the weight of carbonate neutralised by a c.c.
zº)
O-O5305
formula the number of c.c. of normal acid that would be required is
of acid, 4 = will correspond to a normal solution. From this
calculated ; this number, y, is therefore = and a will be smaller
O-O5305’
than y if the preliminary acid is too strong, which is generally the case.
In order to find out how much the “preliminary” acid must be diluted
* Z. anal. Chem., 1903, 42, 512.
* Cf. Lunge, Z, angew. Chem., 1905, 18, 1520.
STANDARD ACIDS 87
º - * = & IOOO 2* © -
to obtain a normal solution, let u = ; u is then the number of c.c.
of the preliminary acid which has to be made up to IOOO c.c. with water
to obtain a normal acid.
If normal sodium hydroxide is available, it may be used in an
exactly analogous way to titrate the “preliminary” acid and to prepare a
normal Solution.
The diluted acid is then titrated against new samples of freshly
ignited sodium carbonate as a check, that is, to ascertain if a = y. A
further control, by determining the proportion of chlorine with silver
nitrate, is also very desirable; IO c.c. of the acid (=O.3646 g. HCl) should
give I-435 g. AgCl, -
Since 50 c.c. of M/I acid, the capacity of an ordinary burette, cor-
responds to a weight of 2.652 g of sodium carbonate, from 2.0 g. to 2.5 g.
should be weighed out for each titration, so as to reduce the error in
reading the burette, as far as possible. If only 20 c.c. of acid are used
for the titration, the effect of the error in reading is double that for 40
c.c., and if more than 50 c.c. is needed, so that the burette has to be
refilled, four instead of two readings are necessary, so that the degree of
accuracy is also lessened. The use of IOO c.c. burettes of the ordinary
form, in order to avoid the four readings, is also disadvantageous,
because they can only be graduated into fifths of a c.c. with accuracy,
and the error in reading is thus double that with a 50 c.c. burette gradu-
ated in tenths. If, however, a great many analyses requiring the use of
more than 50 c.c. of acid for each titration have to be made, burettes
with bulbs, constructed on the principle of Lunge's bulb nitrometer, may
be employed, in which the accurate graduations begin at 60 or 70 c.c.
In order to interpret the experimental results clearly, a coefficient
for the normal acid should be calculated from each observation, by
finding the amount of acid theoretically necessary to neutralise the
weighed quantity of sodium carbonate and dividing this by the number
of c.c. actually required; the quotient gives the number with which the
results of the titration have to be multiplied to find the amount of
sodium carbonate actually present. Suppose, for example, that 2.500 g.
of sodium carbonate have been weighed out, and that this amount
neutralises 47.00 c.c. of acid ; if the acid had been actually normal,
2-500
O-O5305
=47. 13 c.c. would have been required. The acid is, therefore,
slightly too strong, and the readings must be multiplied by # :=
I.OO28 to correspond to normal acid.
All normal solutions should be prepared at I 5°, and a correction
must be applied as detailed on p. 45, if they are subsequently employed
at a temperature differing by more than I" or 2° from 15°.
88 GENERAL METHODS USED IN TECHNICAL ANALYSIS
In order to be able to rely on the correctness of a normal acid, at
least three, but preferably four, experiments should be made, and the
greatest deviations from the coefficient, calculated as above, should not
exceed o-ooro, so that the mean of the results has a maximum error of
+ O.OOO5. A skilled worker can get more accurate results, but it is
not easy to get within the half of the above error, that is, an error of
+ O.O25 per cent. and + O.OI per cent, which may be regarded as the
greatest obtainable accuracy, can only be attained by taking very
special precautions, making the readings with a cathetometer, etc.
Since small errors may be introduced each time the normal solution is
used, the second decimal place per cent. is always uncertain in alkali-
metric and acidimetric observations, and even the correctness of the first
decimal place, that is one-thousandth of the total amount, is only to be
depended on when very great care is taken. It is, therefore, absurd to
calculate out analyses of this kind to more than two decimal places per cent.
In scientific laboratories it is usual to prepare a volumetric solution
of approximate accuracy and to correct the readings by means of a
coefficient, determined as above. In technical laboratories, on the other
hand, where large amounts of standard solutions are made up at once,
the use of a coefficient involves loss of time and may give rise to error;
it is, therefore, preferable to adjust the solutions so that no correction is
required. If the coefficient does not exceed # O.COO2, this deviation
from a normal solution can, in most cases, be neglected.
Of other substances which have been suggested for the standardising
of normal acids and alkalis, the following are the most important."
Grandeau,” and subsequently Pincus* and Fresenius,” suggested the
use of calcspar (Iceland spar), which they regard as chemically pure
calcium carbonate, but this is not always the case. Thiele and Richter 5
found deviations of O.2 per cent. from the true value, with this substance.
Hartley" and Neitzel 7 have made use of metallic sodium.
Borax is recommended by Salzer.” Rimbach and Worms,” Rich-
mond,” Buchanan,” and Perman and John.”
Ammonium chloride has been suggested by Reinitzer,” and by
Seyda and Weinig,” and ammonium sulphate by Knublauch.”
* Many of the following references are taken from a paper by Vanino and Seitter, Z. anal.
Chem., 1902, 41, 141, which contains a fairly complete summary of the literature on volumetric
solutions and substances used in standardising them. Cf. also B. North and W. Blakey, J. Soc.
Chem. Ind., 1905, 24, 396.
* Z. anal. Chem., 1863, 2, 426. * Ibid., 1896, 35, 338; 1897, 36,688.
* /bid., 1863, 2, 426. * Chem. Mews, 1895, 72, 5.
* Quantitative Analysis, 7th edition, vol. ii., * J. Soc. Chem. Ind., 1904, 23, Io93.
p. I94. * Chem. Wews, 1895, 71, 296.
* Z. angew. Chem., Igoo, 13, 486. * Z. anal. Chem., 1895, 34, 577; 1900, 39,
* Z. anal. Chem., 1873, 12, 89. 458.
7 Ibid., 1893, 32, I22. * Z. angew. Chem., 1892, 5, 204.
* Ibid., 1893, 32, 449. * Z. anal. Chem., 1882, 21, 165.
STANDARD ACIDS 89
An iodometric method, based on the reaction :-
5KI+ KIO,--6HC1 = 6I+3H,O +6KCl,
is described by Mohr, Kjehdahl, Gröger,” and Fessel.” v. Than 4
proposed potassium bi-iodate, which is also recommended by Meineke.”
Riegler" has suggested pure iodic acid. All the iodometric methods
for preparing normal acids and alkalis are indirect, and the sum of the
small unavoidable errors may easily exceed the permissible limit stated
on page 88. The basis of iodimetry is not more certain than that of
acidimetry, if pure sodium carbonate is used as the standard, but rather
the contrary, so that acidimetry may justifiably be employed as a basis
for iodimetry rather than the reverse (cf. p. 117).
The iodometric method is recommended by Petersen 7 for determin-
ing the amount of acid in coloured plant extracts, roots, beer, etc.
Morse and Chambers * base a method of standardisation on the
decomposition of neutral hydrogen perovide by potassium perman-
ganate, according to the equation :-
2KMnO,--5H2O, +3H2SO, = K,SO,--2MnSO,--8H3O+50,
In this reaction the volume of gas evolved may, of course, also be
measured. Another method, involving measurement of a gas, in which
potassium iodide, potassium iodate, sodium hydroxide, and hydrogen
peroxide are employed, has been proposed by A. Baumann.”
Hart and Croasdale” and Kohn ” have proposed the electrolysis of
copper sulphate as a basis for acidimetry. The advantage of this
proposal lies in the fact that it is dependent only upon the ratio of
copper to Sulphuric acid. The copper sulphate itself need not be
weighed, and therefore the difficulty of obtaining a substance, which
can be dried without decomposition or without loss of water of
crystallisation, is avoided. It is only necessary that the sulphate is
free from metals other than copper, and this degree of purity is effected
by two recrystallisations of the commercial salt, a little nitric acid
being added in the first recrystallisation to oxidise any ferrous sul-
phate present. From I to 2 g of copper sulphate is electrolysed in
aqueous solution, and the weight of sulphuric acid formed determined
from the weight of the equivalent of copper deposited. The method
serves only as a basis from which the strength of other solutions may
be determined, but not for preparing actual standard solutions for use,
as there is an obvious limitation to the quantity of solution that can
be conveniently prepared in this way. Dauvé” states that good
* Z. anal. Chem., 1883, 22, 327. 7 Z. anal. Chem., 1903, 42, 308.
* Z. angew. Chem., 1890, 3,353 and 385. * Ibid., 1898, 37, 183.
* Z, anorg. Chem., 1899, 23, 67; 38, 1904, 449. "Ibid., 1892, 31, 450.
* Annalen, 1891, 261, 358. * /. anal. Chem., 1890, 4,424; 1892, 6,421.
* Chem. Zeit., 1895, 19, 2. * /. Soc. Chem. Ina., 1900, 19, 962.
* Z. anal. Chem., 1896, 35, 308; 1899, 38, 250. ” /. Pharm. Chem., 1902, 16 [2], 65.
90 GENERAL METHODS USED IN TECHNICAL ANALYSIS
results cannot be obtained by this method, because if no free acid is
present, the deposited copper does not adhere to the electrode and
contains suboxide of copper, but no experiments are given to support
this conclusion. Meade” has described the preparation of standard
sulphuric acid, and indirectly of standard hydrochloric and nitric acids,
by the electrolysis of copper sulphate.
North and Blakey” recommend sodium bicarbonate for fixing the
strength of standard acids, but according to Lunge” it is not reliable.
Normal hydrochloric acid can of course be prepared and checked
gravimetrically, by precipitation with silver nitrate. Lunge finds that
with careful working, the results by this method agree with those
obtained by the use of sodium carbonate to within O.O5 per cent., some-
times even to within O.O2 per cent. ; he does not regard this gravi-
metric method as more accurate than standardisation with pure sodium
carbonate.
It is much more difficult to attain the same degree of accuracy with
normal sulphuric acid by precipitation with barium chloride, owing
to the well known difficulty of obtaining absolutely pure barium
sulphate.
Several authors have recommended evaporating down the acids
with ammonia, and weighing the ammonium salt formed.
Moody * has suggested the preparation of normal hydrochloric acid,
by passing hydrochloric acid gas into a known weight of water, and
determining the increase in weight. The method has since been
studied and improved by Higgins ; * very accurate results can be
obtained, but somewhat special apparatus is required which is hardly
suitable for technical work. Both Roth" and Raschig" have also
described this method.
A. Marshall 8 recommends the preparation of normal sulphuric acid
on the basis of its specific gravity. This proposal has been extended
to hydrochloric acid by Küster and Siedler,” but the examples given
show too great deviations from the actual strength, even in cases
regarded by them as correct up to O. 5 per cent, to allow of the recom-
mendation of the method, for accurate work. In a later communication,”
Küster and Münch state that the method is correct to within + o-OOOI
of the sp. gr. Tables of specific gravities of sulphuric acid and of
hydrochloric acid, between the limits of N/Io and M/I solutions, have
* /. A mer. Chem. Soc., Igor, 23, I2 and 343. * Chem Soc., 1897, 73, 658.
* / Soc. Chem. Ind., 1905, 24, 396. * /. Soc. Chem. Ind., 1900, 19, 958.
* Z. angew. Chem., 1905, 18, 1520.
* Z. angew. Chem., 1904, I7, 7Io; /. Soc. Chem. Ind., 1904, 23, 837.
7 Z. angew. Chem., 1904, I'7, 577.
* /. Soc. Chem. Ind., 1900, 19, 4; 1902, 21, 15II.
* Chem. Zeut., 1902, 26, Ioš5; Ber., 1905, 38, I50.
* Ber., 1905, 38, I $o; Z, an rg. Chem., 1905, 43, 373.
STANDARD ACIDS 91
also been prepared by Worden and Motion;' such tables may very
well be used for the approximate adjustment of volumetric solutions.
Quincke” has proposed a method, depending upon the measurement
of oxygen evolved from potassium ferrocyanide and alkaline hydrogen
perpxide. -
There are several methods, in which an acid substance is taken as
the basis, by means of which an alkaline solution is standardised and the
latter then employed for standardising a normal acid.
The most important of the compounds used in this way is oralic
acid, the value of which, as a basis for acidimetry, has been advocated by
Mohr. It is, however, very difficult to prepare oxalic acid quite pure,
and with the definite proportion of water corresponding to the formula
H.C.O.2 H.O. This difficulty is by no means overcome by Hampe's
suggestion * to use the extremely hygroscopic anhydrous acid, which
sublimes at IOO", instead of crystallised oxalic acid. Of the numerous
methods proposed for the preparation of pure Oxalic acid, that of
C. Winkler is most largely used.* 500 g. of commercial oxalic
acid are dissolved in an equal weight of boiling hydrochloric
acid of sp. g. I.O7, in a porcelain basin, the Solution placed in ice water
and allowed to crystallise, with continual stirring ; the finely crystalline
mass is transferred to a funnel, provided with a plug of glass wool,
allowed to drain, and washed several times with hydrochloric acid.
It is then redissolved in pure, boiling hydrochloric acid, again cooled
with continual stirring, the crystals filtered off, washed with a little water
and recrystallised by dissolving in just the requisite amount of boiling
water. The finely crystalline mass thus obtained is again washed
and the whole treatment is repeated two or three times. The final
product is left to dry in a cool place, the drying material being repeatedly
changed. Thus prepared, it is free from chlorine and mineral matter,
but still contains enclosed liquid, and must therefore be dehydrated
before use. This is best effected by drying for five to six hours at 60°,
as oxalic acid volatilises to a considerable extent at IOO’. The dry acid
having the formula H,C,C), must naturally be kept in tightly closed
bottles; when it is to be employed for standardising purposes, about
o.2 g is weighed out into each of several weighing tubes, again dried
for several hours at 60° to 80°, allowed to cool in the closed tubes, and then
accurately weighed.
This somewhat troublesome method of purification is likely to prove
too tedious for technical work, and the method possesses no advantage
over the use of pure sodium carbonate. Winkler recommends it only
for standardising purposes, and not for the preparation of normal oxalic
acid. To carry out the method, a normal alkali is first standardised with
* J. Soc. Chem. Ind., 1905, 24, 178. * Chem. Zeit., 1883, 7, 73 and Ioff.
* Z. anal. Chem., 1892, 31, I. * Uebungen in der Massanalyse, 3rd edition, p. 69.
92 GENERAL METHODS USED IN TECHNICAL ANALYSIS
the Oxalic acid and then used for adjusting the strength of the normal
acid; the method is, therefore, indirect, and accordingly liable to small
unavoidable errors which do not arise in direct standardisation. The
normal alkali used in the titration must either be free from carbon
dioxide, or the solution must be boiled, allowed to cool, and the titration
completed. Instead of an alkali hydroxide, barium hydroxide may be
employed, but the adjusting of the latter solution, so as to contain an
exact equivalent, is difficult (cf. infra).
Instead of oxalic acid, its acid salt, potassium tetroxalate, is regarded
by many as a more reliable basis for alkalimetry, as also for the titration
of potassium permanganate. The crystallised salt has the formula
KHC,O.H.C.O.2H2O ; its employment was first suggested by Kraut,”
and subsequently recommended by Ulbricht and Meissl,” and by
Meineke,” whilst Wells,” Hinmann,” and Dupré and Kupffer" have
criticised its value adversely. It can be used either with litmus or
preferably with phenolphthalein as indicator; as in the case of free
Oxalic acid, the presence of carbon dioxide in the alkali must be taken
into consideration. The tetroxalate has recently been again strongly
recommended by J. Wagner, who frees it from moisture by pressing
between blotting paper, instead of by standing over sulphuric acid, as
recommended by Meineke, and employs it in the air-dried state. It
can be obtained, free from the dioxalate which it is apt to contain, by
dissolving one gram-molecule of neutral oxalate in water and adding it to
a solution of rather more than three molecules of oxalic acid, in such a
way that excess of the latter is always present; the solution is warmed
for an hour on the water bath, and the product recrystallised. The
tetroxalate is then used for standardising barium hydroxide with
phenolphthalein as indicator, and the strength of the normal acid
adjusted from the hydroxide.” The same method is recommended by
Kühling.” Lunge” states that, as the result of a large number of experi-
ments, he has not succeeded in obtaining potassium tetroxalate with a
constant proportion of water corresponding to the formula given above,
and he therefore does not concur in recommending it in place of sodium
carbonate as a basis for alkalimetry, although it may quite well be
employed for the standardisation of potassium permanganate, via
sodium carbonate, hydrochloric acid, and barium hydroxide.
Objection must be made on principle to standardisation with oxalic
acid, potassium tetroxalate and all other substances, such as potassium
* Annalem, 1863, 126, 629. * Z, anal. Chem., 1893, 32, 453.
* Z. anal. Chem., 1887, 26, 350. * Ibid., 1894, 33, 456.
* Chem. Zeit., 1895, 19, 2. * Z, angew. Chem., 1902, 15, 352.
7 Proc. 5th /nternational Congress for Applied Chemistry, 1903.
* J. Wagner, private communication.
* Z, angew. Chem., 1903, 16, IoSo; Chem. Zeit., 1904, 28, 596 and 612.
* Z. angew. Chem., 1904, 17, 227; Chem. Zeit., 1904, 28, 7or.
STANDARD ACIDS 93
bi-iodate and cream of tartar, with which phenolphthalein is used as
indicator, when the normal solutions are subsequently used with methyl
orange (cf. p. 6O). -
Potassium bi-modate has been recommended as a basis for the direct
adjustment of normal sodium carbonate by Meineke' and by E. Crato.”
According to Lunge,” iodic acid behaves, towards methyl orange, as a
rather weaker acid than hydrochloric acid and similar acids; this is
naturally not the case with phenolphthalein, but still the results are not
very accurate, nor is it possible to establish the absolute purity of
the substance readily and quickly by qualitative analysis, as in the case
of sodium carbonate. J. Wagner also finds” that the commercial
“chemically pure” bi-iodate often does not correspond to the formula
KH(IOS), and that it cannot be satisfactorily dried by the method
recommended by Meineke. There is accordingly no reason for using
this salt in preference to sodium carbonate as a basis for alkalimetry,
either directly or indirectly. -
Potassium hydrogen tartrate has been recommended as a standard by
Bornträger ; * it appears to be used at Italian experimental stations.
The weak acidic character of tartaric acid and the consequent decrease
in the sharpness of the transition tints of the indicator, is an obvious
objection to its use, and it has no advantages over sodium carbonate.
Potassium bichromate has been suggested by Richter" as a substance
which can readily be obtained pure, but even if this be so, the weak
acidic character of chromic acid and the transition in colour from one
tint of yellow to one of a more reddish hue, are clearly unfavourable to
its use in alkalimetry.
The Strength of Normal Acids.
In works, AWI acid is used for controlling working processes and for
many laboratory purposes; the change of colour of the indicator with
acid of this strength is very sharp, under all circumstances. Its use
necessitates taking a correspondingly larger amount of substance for
analysis, and in practice, a sufficient degree of accuracy is attainable,
with less carefully calibrated weights and measuring vessels, than are
used for other work. It is, however, often necessary to work more
accurately, and there is then the choice of #, #, or 1%, M acid. These
are most readily prepared by dilution of the normal acid ; the strength
of the diluted acid must always be checked as directed on page 87, for
which purpose I.OOO to I-250 g. Of sodium carbonate is weighed out for the
M/2 acid, O-4 to O.5 g. for the AV/5 acid, and O-2 to O-25 g. for the AV/IO acid.
* Chem. Zeit., 1895, 19, 2. * Z. angew. Chem., 1904, 17, 225.
* Massanalytische Taſel. * Massanalytische Studien, p. 60.
* Chem. Zeit., 1881, 5, 519; Z, anal. Chem., 1886, 25, 333; 1892, 31, 56; I894, 33, 713.
* Z, anal. Chem., 1882, 21, 205; cf. also, Z, anorg, Chem., 1893, 3, 84. *
94 GENERAL METHODS USED IN TECHNICAL ANALYSIS
It is undesirable to have too many normal solutions in use, and usually
one other acid besides the normal acid suffices; AV/5 acid is the best
strength for the dilute acid. Very little is gained by using W/2 acid
instead of the normal, and N/IO acid is too dilute to use as the weaker
solution ; at least two drops of it are required to bring about a sharp
change of colour, so that it is really not more accurate than M/5 acid,
only one drop of which is required to effect the same result. Titra-
tions to within one drop of an AV/5 solution (O.O2 to Oo3 c.c.) are as
near as can be relied upon, taking into account the lack of sensitive-
ness of many indicators and other sources of inaccuracy.
In works, a large quantity of normal acid is usually prepared at once,
say 50 to 60 litres, the capacity of an acid carboy. For laboratory use it
must, of course, be transferred to smaller vessels, such as stoppered
bottles, holding at most 5 litres; this is preferably done as soon as the
acid has been standardised. If, however, a sufficient number of smaller
vessels is not available, or if there is no room for them, so that the
carboy has to be emptied gradually, care must be taken to stopper it
well, and to shake thoroughly before use (cf. p. 52). The smaller
bottles must also be well shaken once a day before use, since water may
evaporate and condense in the upper empty part of the vessel.
The equivalent of I c.c. of +, +, +, and fºr AW acids, in grams of the
substance to be estimated, is given in the following table:–


N/1 N/2 N/5 N/10
Potassium Hydroxide . 0-05616 0 02808 0 01123 0 °005616
Potassium Oxide . 0-04715 0-02357 0-00943 0-004715
Potassium Carbonate 0 °06915 0-03457 0 °01383 0 °006915
Sodium Hydroxide 0 °04003 0°020015 0 °00801 0 °004003
Sodium Oxide . 0 °03105 0-01552 0 °00621 0°003105
Sodium Carbonate 0 °05305 0 °02652 0 °01061 0 °005305
Barium Hydroxide 0-08572 0 °04286 0-01714 0°008572
Barium Carbonate 0 °09872 0 °0.4936 0-01974 0-0098.72
Strontium Hydroxide . 0 °06082 0 °0304] 0 °01216 0-006082
Strontium Carbonate 0 07382 0 °03691 0’014764 0-007382
Calcium Hydroxide 0-03707 0 °01853 0’007414 0-003707
Calcium Oxide 0 °02806 0 °01403 0 °005613 0 °002806
Calcium Carbonate 0-05006 0 °02503 0 °010013 0 °005006
Magnesium Oxide 0 °02018 0-01009 0 °004036 0 °002018
Magnesium Carbonate . 0 °04218 0 °02109 0 °008436 0 °004218
Normal Oxalic Acid.
Oxalic acid is quite superfluous for alkalimetry, but it is never-
theless still used by some chemists, and is required for a few other
purposes, such as the examination of Weldon mud. It cannot be
employed with methyl orange as indicator, and even for the estima-
STANDARD ALKALIS 95
tion of the hydroxides of the alkaline earths, in the presence of their
carbonates, it can be completely replaced by hydrochloric acid."
It can be readily prepared by dissolving 63.025 g. of chemically pure
oxalic acid in one litre for the normal acid, and correspondingly smaller
amounts for the weaker acids; since the preparation of absolutely
pure and dry Oxalic acid, according to the method detailed on p. 91,
is so troublesome, even C. Winkler, who holds to this acid as the basis
of alkalimetry, recommends that normal oxalic acid for use should be
prepared of approximate strength from the ordinary “pure” acid, and
the strength determined by titration with accurate normal alkali.
Only very weak acidic indicators, such as litmus and still better
phenolphthalein, can be used with oxalic acid. The solutions do not
keep well and decrease in stability with the degree of dilution. They
decompose gradually even in closed vessels and when kept in the dark,
but more rapidly when exposed to light. The more concentrated
solutions, including the AW/I acid, are, however, fairly stable; if sterilised,
they can be kept in the dark for some years without alteration in
strength.” Very dilute solutions such as the M/IOO acid, used in water
analysis for the determination of “organic matter,” keep only for a
very short time, and must be prepared fresh on each occasion. The
strength can, of course, readily be determined, at any time, by titration
with sodium hydroxide, using litmus or phenolphthalein as indicator;
in order to obtain accurate results, the solution should be boiled with a
slight excess of sodium hydroxide, cooled quickly, and titrated back in
the cold, in order to avoid error, owing to the presence of sodium
carbonate in the hydroxide.
STANDARD ALKALIS
The hydroxides of sodium, potassium, ammonium, and barium are
employed as standard alkalis; of these, the first is most largely used.
The only reason for using the more expensive potassium hydroxide is
that it has rather less action on glass; this is true for solutions of the
same percentage composition, but apparently not for solutions of the
same molecular concentration.” Mohr found that burettes filled with
sodium hydroxide sometimes cracked longitudinally, which was not the
case when potassium hydroxide was used ; this is, however, of so infre-
quent occurrence, that it cannot be regarded as a reason for limiting the
use of sodium hydroxide.
Ammonium hydroxide has been recommended as a standard alkali
because it does not readily absorb carbon dioxide from the air and has
* Küster, Z. anorg. Chem., 1898, 18, 127 ; Lunge, Z, angew. Chem., 1897, Io, 41.
* H. Beck, Inaugural Dissertation, Jena, Igo2.
* Lunge and Millberg, Z, angew. Chem., 1897, Io, 398.
96 GENERAL METHODS USED IN TECHNICAL ANALYSIS
no action on glass. These advantages, the first of which is only relative,
are, however, counterbalanced by the tendency of the solutions to give
up ammonia. Even W/2 and M/5 solutions alter in strength so rapidly that
they cannot be used for any length of time without control; such solu-
tions should not be allowed to stand, even for a short time, in burettes
which are not thoroughly shut off from the air. -
Barium hydroxide, which absorbs carbon dioxide from the air,
giving a precipitate of the carbonate, can only be used when all the
usual precautions are taken to obviate this source of error. It is
difficult to prepare a normal solution and still more difficult to keep its
strength unaltered, so that a correction factor has always to be used.
It is only in certain cases that the advantages of barium hydroxide are
sufficient to counterbalance these drawbacks, as, for example, in the titra-
tion of weak acids with phenolphthalein, since, unlike the hydroxides of
potassium and sodium, it can readily be obtained free from carbon
dioxide. When this is of no advantage, as in all cases in which methyl
orange is used as indicator, nothing is gained by its use.
Sodium hydroxide.—In technical laboratories and even for most
scientific purposes, the commercial white caustic soda may be used
directly for preparing this solution, since the impurities usually present,
such as chloride, sulphate, silicate, and aluminate, very seldom affect the
result. Sodium hydroxide in sticks, purified by alcohol, answers all require-
ments. The expensive product prepared from metallic sodium contains
not only sodium carbonate, but sometimes gives a flocculent residue on
solution. A similar residue is often obtained when ordinary caustic
soda is used, and a sample should therefore be selected which dissolves
in water to a clear solution; otherwise the solution must be allowed to
settle and the clear liquid decanted.
To prepare normal sodium hydroxide, clear transparent lumps of the
best white commercial caustic soda are picked out, any opaque parts on
the surface scraped off, and 50 g. weighed out for each litre. It is
then dissolved in water, made up to the litre at I 5°, and 50 c.c. of this
provisional solution titrated with accurately normal hydrochloric acid,
using methyl orange as indicator, till the brownish transition tint
appears (p. 62). If n c.c. of acid are used, the number of c.c. of the pre-
liminary alkali, which must be diluted to a litre, with pure water, to
obtain accurately normal sodium hydroxide, is 5ox rooo.
The solution must, of course, be titrated again to check its accuracy;
weaker solutions are similarly prepared.
Bousfield and Lowry' have shown that standard solutions of sodium
hydroxide can be prepared from metallic sodium. Küster” has also
| Phil. Zºrans., 1905, 204, 253.
* Z, anorg, Chem., 1904, 41, 472. J. Soc. Chem. Ind., 1904, 23, Io27 ; 1906, 25, 982.
STANDARD ALKALIS 97
described a method for the preparation of pure sodium hydroxide solu-
tions from the metal.
As is well known, sodium hydroxide solutions should not be kept in
bottles with ground-glass stoppers, as these easily become fixed in the
neck of the bottle. This difficulty is to some extent overcome by
greasing the stoppers with a little paraffin or vaseline, or by tying a
thread round the stopper, or by covering it with a piece of wide rubber
tubing; a rubber stopper may also be used. Corks cannot be used, as
they soon decay and pieces fall into the solution.
It is not advisable to prepare very large amounts of sodium
hydroxide at once as a stock solution, because it slowly attacks the
glass and alters a little in strength. In order to obtain accurate results
when litmus or phenolphthalein are used as indicators, the hydroxide
must be prepared free from carbon dioxide, and when in use, especially
in the burette, must be carefully guarded from contact with this gas.
To effect this, it has been suggested that the sodium or potassium
hydroxide should be boiled with milk of lime, till a filtered sample
evolves no carbon dioxide when poured into dilute hydrochloric acid.
This mode of procedure is, however, insufficient to remove the carbonate
completely. The method suggested by Küster," to dissolve sodium in
boiling alcohol, displace the alcohol by water, free from carbon dioxide,
and dilute with the latter to normal strength, would certainly give more
satisfactory results, but the method is inconvenient and expensive, and
is not likely to be used in technical laboratories. Moreover, Lunge
found that solutions prepared in this way, with all possible precautions,
still contained considerable amounts of carbon dioxide.
Küster and Grüters * recommend that barium hydroxide should
always be used for accurate determinations with phenolphthalein,
because potassium and sodium hydroxides give uncertain results, owing
to the variable amounts of contained carbonate.
There is, further, no object in preparing an alkali hydroxide solution
free from carbon dioxide, unless its re-absorption from the air is carefully
guarded against, when in use. To effect this, the solutions should
always be prepared and diluted with boiled, carbon dioxide-free water,
the titration should be made, as far as possible, without access of air, and
the burette and supply-vessel should be guarded against the absorption
of the gas. The method generally adopted to protect the solution is to
provide the supply-bottle with a stopper with three holes; in one is
placed a soda-lime tube open to the air, the second is connected with
the top of the burette, and a syphon tube, reaching to the bottom of the
bottle, is placed in the third hole and connected by means of a T-piece
with the lower part of the burette, above the pinch-cock. There are con-
* Z. anorg. Chem., 1897, 13, 134; cf. Treadwell, Analytical Chemistry, vol. ii., p. 441.
* Zbid., 1903, 35, 459.
G
98 GENERAL METHODS USED IN TECHNICAL ANALYSIS
siderable objections to this arrangement, however, since although the
carbon dioxide is excluded, it renders the daily shaking of the supply-
bottle, which is absolutely necessary, owing to evaporation or condensa-
tion of water in the upper part of the latter, difficult or impossible. If
the connections, as is preferable in other cases, are made, as far as
possible, of glass, and are therefore rigid, shaking is out of the question; if
rubber tubing is used instead, the alkali becomes contaminated,
especially with sulphur, by prolonged contact. It is therefore preferable
to close both burette and stock-bottle with a soda-lime tube and not to
connect them, but to fill the burette at the top in the usual way, after
shaking the bottle; no appreciable error arises from the momentary
contact of the alkali with the air, certainly less than when the burette
and flask are rigidly connected. Some such mode of connection as
that described above is, however, essential in using barium hydroxide
solutions.
Glaser" and others contend that the errors due to the small quan-
tities of carbonate contained in solutions of the alkali hydroxides and to
the carbon dioxide absorbed from the air during titration, are practically
negligible when litmus or phenolphthalein are used as indicators, and
may be disregarded. Experiments by Lunge” and by Raschig,” how-
ever, show definitely that this is not the case. In a series of determina-
tions by the former, 25 c.c. of M/5 hydrochloric acid required exactly 25 c.c.
of N/5 sodium hydroxide, prepared from metallic sodium and using methyl
orange as indicator, in each of three independent determinations; in six
determinations with phenolphthalein, titrated cold, the minimum value
was 25-54 c.c., the maximum 25.63 c.c., an average of 25.59 c.c., or 2.32
per cent, too much. In a separate set of determinations, with different
solutions, the coefficient for the alkali, using methyl orange, was I.OOO6 to
I.OOO8 in three experiments, whilst with phenolphthalein the value in five
experiments was O.9799 to 0.9804, or again more than 2 per cent out,
an error due to the colour change taking place when half of the carbon
dioxide of the contained carbonate is disengaged, with the formation of
bicarbonate.
For comparatively rough work, complete protection against the
absorption of carbon dioxide is unnecessary when litmus or phenol-
phthalein are used as indicators; if methyl orange is used, such protec-
tion is not required, even for accurate work. With the latter indicator
it might be thought that sodium hydroxide could advantageously be
replaced by sodium carbonate, by dissolving 53.05 g. in a litre of water
at 15° for an M/1 solution; this is not to be recommended, however,
because a solution of sodium carbonate also gradually attacks glass, even
when cold, taking up a small amount of alkali, and further, the solu-
1 Indikatoren der Acidimetrie und Alkalimetrie, p. 28.
* Proc. 5th Congress Applied Chemistry, 1903. * Z. angew. Chem., 1903, 16, 820.
STANDARD ALKALIS 99
tion creeps up and evaporates at burette jets and on the neck of flasks,
thus giving rise to serious errors, unless this action is continuously
watched.
Potassium hydrotide solution is made up similarly to sodium
hydroxide, allowance being made for its higher molecular weight and
also for the greater proportion of impurities that are generally contained,
in weighing out the quantity required for the preliminary solution.
Barium hydroxide solution is usually made of empirical strength. The
following method is given by Stutzer." Thirty-five g. barium hydroxide
and 5 g. barium chloride are dissolved in I litre of water. To obtain
a clear solution on filtration, the bottle for receiving the solution is
washed out with very dilute hydrochloric acid, a layer of petroleum about
IO to 12 mm. deep placed on the bottom, and a funnel, with a long exit
tube reaching down to the layer of petroleum, placed in the neck; a
filter paper is carefully fitted into the funnel by moistening it with
distilled water, and is then also partially filled with petroleum. The
flask, containing the solution to be filtered, is placed at a higher level
than the funnel and its contents syphoned off through the latter. The
filtered solution is shaken by giving the contents of the bottle a rotatory
motion, whereby contact with the air is reduced as much as possible.
In the following table the analytical value of +, +, +, and ºr normal
solutions of alkali are given, expressed in grams per c.c. of solution:—

N/l N/2 N/5 N/10
Hydrochloric Acid" { } 0 °03646 0-01823 0-007292 || 0-003646
Nitric Acid ". . . . 0 °06305 0-003152 0-01.261 0-006305
Sulphuric Acid . 0 °04904 0 °02452 0-00981 0°004904
Phosphoric Acid.”
(a) With Methyl Orange 0-07100 0 °03550 0°01420 0-00710
(b) , , , , Phenolphthalein 0 °03550 0-01775 0-00710 0°00355
Oxalic Acid 8 º, º & 0 °06302 0-031501 0 °012604 || 0-006302
Citric Acid 8 0-07502 0-037501 0-015005 || 0-007502
Acetic Acid " 0-069802 0 °034901 0°01396 0°006.980
0-06003 0-030015 0-012006 || 0-006003
1 Towards all indicators.
2 Cf. pp. 65 and 72; the values given are calculated for the anhydride P2O5.
& Towards litmus and phenolphthalein; the values given are calculated for the crystallised acids, includ-
ing the water of crystallisation for oxalic and citric acids.
POTASSIUM PERMANGANATE
A solution of potassium permanganate is used for oxidising purposes,
in which it gives up five-eighths of its oxygen, according to the
equation:-
2KMnO, +3H2SO, = K,SO, + 2 MnSO, +Os + 3H2O.
* Private communication to Dr Böckmann.
100 GENERAL METHODS. USED IN TECHNICAL ANALYSIS
Referring to the explanation given on p. 81, it will be seen that a
normal solution would yield O.OO8 g. available oxygen per c.c., and
would be obtained by making up 3 I-64 g. pure potassium permanganate
to a 1-litre solution. A solution of this strength would, however,
be liable to crystallise at low temperatures, owing to the slight solu-
bility of the salt; hence solutions stronger than N/2 are never pre-
pared, and much more dilute solutions are very frequently used, e.g.,
M/10, W/20, N/50, or even W/IOO. The use of such dilute solutions is, in
this case, consistent with accuracy, since an indicator, in the ordinary
sense of the term, is not used ; on the other hand, when an indicator
is specially added to show the end of a reaction, by means of a
colour change, the latter is brought about by mass action between
the ions, and hence the addition of a certain excess of the titration
liquid is always necessary. In the case of permanganate solution, the
very intense colour of the MnO, ion is independent of the corresponding
positive ion H or K', etc. The liquid will, of course, remain colourless
as long as the MnO, ion is destroyed and converted into other
colourless ions, according to the above reaction; at most, the
solution will assume a slight coloration, owing to the formation of
ions from other substances taking part in the reaction, e.g., the formation
of tervalent from bivalent iron. In the dilute solutions used for these
titrations such colorations are extremely slight, and disappear alto-
gether when compared with the colour of the permanganate ion. An
excess of even one drop of M/IOO permanganate solution in IOO c.c. or
more water, produces the intense purple, or in small quantities, light-red
colour of the permanganate ion, and thus indicates the end of the
reaction. -
The main reason for using the standard solution fairly strong is, that
otherwise, in most cases, an inconveniently large volume of the solution
would be required to complete the titration. Another reason why
very dilute permanganate solutions should only be used in
exceptional cases, is that permanganate acts on the impurities which
are present even in ordinary distilled water, and its available oxygen is
thus decreased. A freshly prepared permanganate solution should
therefore never be used immediately, but should be left to stand for at
least three or four, or better for eight to ten days, before it is standard-
ised; the impurities in the water will then be completely oxidised.
The standardisation of a very dilute permanganate Solution cannot, how-
ever, be made very accurate by any method, since the weight of substance
or volume of solution to be titrated must necessarily be small, and thus
the unavoidable experimental errors greatly influence the result. The
obvious artifice of first preparing a concentrated solution, standardising
it accurately and then diluting, leads of course to inaccuracies, owing to
the action of permanganate on the water used for dilution, and it is not
POTASSIUM PERMANGANATE 101
easy to determine the extent of these errors and to allow for them.
These difficulties can be surmounted by adding a little permanganate to
the water to be used for dilution, and leaving it to stand for from eight to
fourteen days; when it is only just possible to detect a pale red colora-
tion on looking through a considerable depth of this water, it may be used
for diluting stronger permanganate solutions without appreciable error.
Again, no appreciable error is incurred if a permanganate solution is
diluted with ordinary distilled water and used immediately, and this
again leads to the conclusion that dilute permanganate solutions should
not be kept in stock; strong solutions (at least I/IO normal) should be
prepared and accurately standarised and diluted to M/2O, AV/50, or
N/100 for immediate use, according to requirements.
The permanganate solutions formerly used were unstable, because
pure crystallised potassium permanganate was hardly obtainable, and
the solutions contained manganate and deposited manganese dioxide.
At the present time there is no reason why permanganate solution
should not keep, provided it is protected from evaporation, dust, etc.
Oddy and Cohen' state that permanganate solutions lose 2 to 3 per
cent. of their oxidising value in four months, whether exposed to light
or not. Their observations, however, are insufficient in number and in
accuracy. Treadwellº found that a well-protected permanganate
solution lost only O. I7 per cent. of its oxidising value, after keeping for
eight months; he recommends that the solution be re-standardised
every two or three months, if great accuracy is required. Lunge has
observed a decrease of O.2 per cent, after three months, and agrees with
Treadwell's recommendation.
For the sake of certainty, however, a check is desirable in this case,
as with every standard solution which has been kept for lengthy periods;
it is impossible to know to what extent small errors, due to some
unknown source, may accumulate and finally produce an error which
would vitiate results. If this condition is attended to, there is no
reason why the value of the solution should be made empirical, neces-
sitating the use of a factor every time a titration is to be calculated, as
used formerly to be directed. On the contrary, solutions may be
prepared in definite molecular proportions, just as in the case of
standard acid or standard iodine solution; or, if estimations of only one
particular kind are to be made, e.g., iron estimations, the permanganate
solution may be so prepared that I c.c. of the solution corresponds to
(say) O'OIO g. iron.
Permanganate solutions are, of course, subject to change if in contact
with organic matter; such change is made visible by a separation of
brown manganese dioxide. Hence the solutions cannot be filtered
through filter paper and must be protected from all forms of dust.
* /. Soc. Chem. Ind., 1890, 9, 17. * Analytical Chemistry, vol. ii., p. 481.
102 GENERAL METHODS USED IN TECHNICAL ANALYSIS
They should be stored in bottles with well-fitting, ungreased glass
stoppers. It is desirable to protect them from direct sunlight. Pinch-
cock burettes must not be used for permanganate titrations, on account
of the rubber, although the error thus caused is, according to de
Koninck, much smaller than is usually assumed, and is in fact negli-
gible in most cases. Gay-Lussac or other delivery burettes were
formerly in use; burettes with glass stopcocks are now exclusively
used, preferably without grease. Lunge” states that a small quantity
of vaseline on the stopcock does not influence the results.
Permanganate solutions often deposit a small quantity of man-
ganese dioxide on keeping; this should of course not be shaken up
and mixed with the liquid, as such a procedure would obviously vitiate
the results. On the other hand, the stock-bottle should be shaken
before use, to mix the solution with the water which has evaporated and
condensed on the interior of the empty part of the bottle; hence, as
soon as such a deposit is noticed, the contents of the bottle should be
left to settle for at least twenty-four hours and then carefully decanted
into another bottle, neglecting the last portions containing the precipi-
tate. The solution must then, of course, be re-standardised. For
accurate work, it is better, however, only to use solutions which have not
precipitated any manganese dioxide.
On account of the deep colour of concentrated solutions, the upper
meniscus of the solution should be read ; dilute solutions, on the other
hand, allow of the lower level of the meniscus being read as usual. The
floats (p. 48) designed for use with permanganate solution have dis-
advantages, and it is better to work with a reading screen (p. 49).
Permanganate titrations are generally carried out in Sulphuric acid
solution. Hydrochloric acid has the great disadvantage of acting on
potassium permanganate, even in very dilute solution, chlorine being
liberated; hence solutions containing chlorides cannot be titrated with-
out special precautions. This point was investigated by Kessler, and
subsequently by Zimmermann, who showed that perfect accuracy is
attained, provided a fairly large quantity of manganous Sulphate is
added to the liquid to be titrated; a solution containing 200 g. of the
crystallised salt per litre is used, and about 20 c.c. of this solution are
added.” Even with this precaution available, the presence of chlorides
should be avoided or limited as far as possible, since the marked colour
of ferric chloride hinders the recognition of the end-reaction.
Pure, dilute nitric acid is, like sulphuric acid, indifferent to per-
manganate; the lower oxides of nitrogen, however, reduce perman-
ganate, and this action affords the best method for their estimation.
* Private communication to Prof. Lunge. * Z. angew. Chem., 1904, 17, 197.
* For an explanation of this action, cf. J. Wagner, Massanalytische Studien, p. 77; Gooch and
Peters, Z. anorg. Chem., 1899, 21, 185; Skrabal, Z, anal. Chem., I903, 42, 359.
POTASSIUM PERMANGANATE 108
A normal permanganate solution cannot be prepared by simply
dissolving a weighed quantity of potassium permanganate and diluting
to a definite volume, partly on account of the above-mentioned action
on distilled water, and partly because the “chemically pure” commercial
salt generally contains a little potassium sulphate, chloride, nitrate, or
Other impurities; it cannot be regarded as IOO per cent. potassium
permanganate, although Gardner, North, and Taylor" have recently
asserted the contrary. For these reasons, a little more than the calcu-
lated amount of the salt is weighed out, about 16 g. for N/2 solution, 3.2 g.
for N/Io solution, etc.; this is made up to I litre of solution at 15°, and left
to stand for about a week before standardisation. It is not advisable
to dissolve the salt in the litre flask, as the presence of undissolved
crystals may easily be overlooked, owing to the dark colour of the solu-
tion. The weighed quantity of the salt is preferably dissolved in warm
water in a beaker, the solution poured into the litre flask, and any of the
salt that may have remained behind in the beaker then dissolved in
more Water,
Very many methods have been suggested for fixing the strength of a
permanganate solution; only a limited number of these can be referred
to, and only a few will be described in detail.
The oxalic acid method was formerly regarded as the most reliable,
provided the purity of the oxalic acid is ascertained by referring it
back to pure sodium carbonate. Sörensen has, however, shown that
sodium oxalate can be obtained quite pure and free from hygroscopic
water, and his recommendation of this salt for the standardisation of
permanganate solutions has been fully confirmed by Lunge (cf. infra).
I. Oxalic Acid method. — This method, proposed by Hempel, is
based on the oxidation of oxalic acid, in presence of sulphuric acid, to
carbon dioxide and water:—
5C, H2O, + 2KMnO, +3H,SO, = 10CO3 +8H3O+ K,SO, + 2MnSO4.
The reaction proceeds slowly in the cold, rapidly at 40° to 50°;
even at this temperature the first drops of permanganate are not
immediately decolorised. When once the reaction has started, however,
it proceeds rapidly, owing to catalytic action on the part of the
manganese salt; the drops of permanganate are then decolorised the
instant they reach the liquid, and the final colour change, from
colourless to light red, is sudden and very sharp. -
The above equation shows that 5 × 90.O3=45O. I 5 parts anhydrous
Oxalic acid require 80 parts oxygen; hence one part anhydrous Oxalic
acid corresponds to O-17772 parts oxygen. This refers to anhydrous
oxalic acid, whilst normal solutions of Oxalic acid are prepared from the
crystallised acid C.H.O. 2H2O, molecular weight 126.05, equivalent
* / Soc. Chem. Zna., I903, 22, 73I.
104 GENERAL METHODS USED IN TECHNICAL ANALYSIS
63.025. As Riegler” has pointed out, standard Oxalic acid solutions
may be rendered more permanent by an addition of sulphuric acid and
by protecting them from light; nevertheless, these precautions by no
means ensure absolute permanence of the solution, and this fact, quite
apart from the absence of any guarantee of the purity and dryness of
the oxalic acid, renders it impossible to apply the solution directly
to the standardisation of permanganate.
If it be, nevertheless, desired to use this direct method, results of
any practical value can only be obtained by using perfectly anhydrous
oxalic acid, purified by the somewhat tedious method described above
(p. 91), and not the pure commercial acid. The acid must be finally
dried for six to eight hours at 60° to 80° in the weighing-bottle itself,
and the bottle must be kept tightly closed during weighing; it is then
dissolved in about 20 c.c. of water and strongly acidified with sulphuric
acid. - - - --
But even when the method is thus modified, it is impossible to be
certain that the results will be absolutely reliable. On the other hand,
Lunge” has shown that accurate results are obtained by preparing a
solution from pure, ash-free oxalic acid and determining its strength by
acidimetry; thus the substance to which the permanganate is
primarily referred is not a weighed quantity of oxalic acid, the
amount of water in which can never be known with certainty, but
sodium carbonate, which can be obtained in a condition of undoubted
purity and dryness (cf. p. 83). In this method hydrochloric acid, best
about W/5, is first standardised against sodium carbonate, using methyl
orange as indicator, and this used to standardise a barium hydroxide
solution of about AV/5 strength (p. 92); the latter is then used to titrate
the oxalic acid solution, with phenolphthalein as indicator, any excess
of baryta being titrated back with W/5 hydrochloric acid. In this way
the true amount of Oxalic acid in the solution can be found, with a
possible error of + goºd, provided all necessary precautions are taken,
viz., the use of calibrated burettes and pipettes, the introduction of
corrections for temperature, etc. Fifty c.c. oxalic acid are then measured
out, using the same pipette as was employed in the above acidimetric
determination of the acid, 5 c.c. sulphuric acid and 200 c.c. water, at about
70° added, and the strength of the permanganate solution deter-
mined; I c.c. N/5 barium hydroxide solution corresponds to O'oi 2605 g.
C.H.O. 2 H2O, or O,OI 118 g. Fe, or O.OOI6 g. available oxygen.
The permanganate, thus standardised, should be used at once to
determine the value of a stock of iron wire by method No. 2 described
below. When it is subsequently required to standardise permanganate
it is more convenient to titrate against this iron wire than against
* Z, anal. Chem. I896, 35, 522; J. Chem. Soc. Abstr., 1896, 70, 676.
* Z. angew. Chem., 1904, 17, 268.
POTASSIUM PERMANGANATE 105
oxalic acid, since the permanence of the latter cannot be relied upon,
mor can it be secured by addition of sulphuric acid, since this would
exclude its standardisation by the above acidimetric method.
Kraut 1 has recommended the substitution of potassium tetrovalate
(p. 92) for oxalic acid, as it can be prepared in a pure state and is
permanent in the air. Ulbricht and Meissl, Meineke, and others
recommend the use of this salt; as already stated, however (p. 92), its
composition, whether dried at the ordinary or at a higher temperature, is
far too uncertain. It should therefore only be employed when its value
has been ascertained by the sodium carbonate, hydrochloric acid,
barium hydroxide method.” -
Rüst” gives the preference to manganese oralate, prepared by
precipitation and dried between filter paper; it is then said to exactly
correspond to the formula with two molecules of water of crystallisation.
This, however, would appear to be hardly reliable.
Barbieri and Neppi 4 employ ferrous oxalate.
Sörensen (cf. p. 85) recommends the use of chemically pure sodium
oralate, prepared by Kahlbaum from his directions; for accurate work
the salt must be placed for a few hours in a drying oven at IOO’ and left
to cool in a desiccator, over calcium chloride. It is then dissolved
in water in a flask, and titrated in sulphuric acid solution. Ten c.c. M/IO
potassium permanganate solution correspond to 67.05 mg. Sodium
oxalate, or O. I g. of the latter to 14.91 c.c. of the former. Lunge"
confirms the accuracy of this method, and recommends it, in preference
to all others, for the standardisation of permanganate.
2. Iron method.—Very fine soft iron wire is used as the basis of this
method; it should be cleaned with emery paper and rubbed with writing
paper before weighing. It is, of course, not pure iron and may contain
O.3 per cent, or more of impurity. As Treadwell has pointed out, this does
not imply that the value of the wire is equal to that of the contained
iron; on the contrary, it may exceed IOO per cent., owing to its
content of carbon, silicon, phosphorus, etc. To avoid errors, which
may be considerable, its value must be determined as described below.
To check an N/2 permanganate solution, a piece of wire weighing
about 1.25 g., for an AV/IO solution a piece weighing about O.25 g., is
weighed to within O.OOOI g. The wire is then stretched out, its length
measured, and a number of pieces of the same length cut off; these can
then be very rapidly weighed, as their approximate weight is known.
Each piece will require nearly, but not more than, 50 c.c. of perman-
ganate Solution. -
* Annalen, 1863, 126, 629.
* Cf. Reinhardt (communication to Meineke), Z. Öffentl. Chem., 1898, 4, No. 13.
* Z. anal. Chem., 1902, 41, 606; / Soc. Chem. Ind., 1902, 2I, I413.
* Rend. Soc. Chim. Åoma, 3, 16.
* Z. angew. Chem., 1924, 17, 230 and 269; 1905, 18, 1520.
106 GENERAI, METHODS USED IN TECHNICAL ANALYSIS
The weighed wire is then dissolved in hot dilute sulphuric acid, in
absence of air; this may be effected in a flask fitted with a Bunsen valve
(Fig. 38), or in any similar form of apparatus; or a flask pro-
vided with a tube, bent twice at right-angles and dipping
into a beaker containing sodium bicarbonate solution, may
be used. If the flame is removed when the iron has dis-
solved, the bicarbonate solution is sucked back and cools the
contents of the flask, whilst the carbon dioxide evolved keeps
out the air." The device (Fig. 39), designed by Contat” and ſº
improved by Göckel,” is much more convenient than any of º
the above arrangements. It consists of a bulb provided with Wºº,
a syphon overflow tube, which is fitted into the flask in which .
the iron is dissolved. The amount of water or sodium bi-
carbonate solution, placed in the bulb at the beginning of the experi-
ment, should be such that the end of the longer arm of the syphon
just dips below its surface, and this level
is not altered until boiling is finished;
a Saturated sodium bicarbonate solution
is then poured into the bulb, which will
be sucked into the flask as the contents
cool, until the pressure of the carbon di-
oxide balances that of the air; the solu-
tion remaining in the bulb protects the
contents of the flask from the air.
The standardisation of permanganate
by means of iron has been subjected to
detailed investigation by Skrabal.4 He
found that the apparent value of a sample
of iron wire, when determined by per-
manganate, was IOO-O3 to IOO-21 per
cent, whereas the actual amount of iron
contained was only 99.38 per cent. (to-
gether with O.O3 P, O.31 Mn, o. 185 Cu,
O-O23 S, O.O9 C). He therefore rejects
the method as unreliable, and attributes its inaccuracy to the fact that
although the carbide in the wire is not decomposed, even by boiling with
permanganate, it is decomposed when warmed with permanganate for
a long time in presence of a ferrous salt, since it acts as an “acceptor.”
in the oxidisation of the ferrous sulphate. For the standardisa-
'Jahoda, Z. angew. Chem., 1889, 12, 87.
* Chem. Zeit., 1898, 22, 298. -
* Z. angew. Chem., 1899, 12,620; J. Soc. Chem. Ind., 1899, 18, 708. The apparatus is made by
Alt, Eberhard, & Jaegar, Ilmenau, and by Göckel & Sauer, Berlin.
* Z, anal. Chem., 1903, 42,359.


POTASSIUM PERMANGANATE 107
tion Skrabal uses either chemically pure iron, or iron wire of
known composition, i.e., of definite value; the iron is dissolved in
sulphuric acid, 15 c.c. M/IO permanganate added to the warm solution,
which is then left to stand over-night in a warm place, whereby the
carbides are completely oxidised ; concentrated hydrochloric acid is then
added, the solution heated to boiling, reduced with zinc, and titrated
with permanganate. These directions are doubtless good, but uniform
and correct results can be obtained without this complication, provided
due care is taken to adhere to the same conditions of dilution, heating,
etc., both in standardising and in the subsequent use of permanganate.
The cold iron solution is titrated with permanganate until a faint
rose-red coloration is obtained, and remains permanent for thirty
seconds; it is not necessary that it should be permanent for longer. If
the number of c.c. used is divided into the number which would have
been required had the permanganate been accurately normal, a coefficient
is obtained by which every reading must be multiplied; it is preferable,
however, to dilute the solution to the normal value. Supposing 4o-oo
c.c. of a correct normal solution would have been required, whilst in
the experiment only 39-00 c.c. were used, then the coefficient is
º:- 1.0256; hence the solution will be corrected by the addition of
25.6 c.c. of water to every IOOO c.c. of the solution.
In order to avoid the uncertainty due to the composition of the iron
wire, Classen has proposed the use of electrolytically deposited iron.
The iron is deposited on a platinum dish, by electrolysing a solution
of ferrous ammonium sulphate containing ammonium oxalate; the
deposited iron is washed with water and alcohol, dried at 80° to 100°,
weighed, dissolved in sulphuric acid, and the solution titrated with
permanganate whilst still in the dish. Even supposing that the electro-
lytically deposited iron is pure, it is questionable whether oxidation
during drying and dissolving is not a source of greater errors than are
incurred when iron wire is dissolved in a flask provided with a valve,
assuming the actual percentage of iron to be 99.8.
Treadwell" also recommends the electrolytic method, which he
modifies by depositing the iron on a small cylinder of platinum foil and
dissolving it in air-free sulphuric acid, in a current of carbon dioxide.
Skrabal” found o. IO to I-47 per cent. impurities, including hydrogen,
in electrolytic iron obtained by Classen's method. He gives an alterna-
tive method for preparing pure electrolytic iron, but according to
Classen,” the iron thus obtained is not pure.
Avery and Dales* have shown that iron, deposited electrolytically
from ammonium oxalate solutions, contains O-2 I to O-42 per cent. of
'Analytical Chemistry, vol. ii., p. 81. * /bid., 1903, 42, 516.
* Z, anal. Chem., 1903, 42, 359. * Berichte, 1899, 32, 64 and 2233.
108 GENERAL METHODS USED IN TECHNICAL ANALYSIS
carbon, and that a small quantity of iron, on an average O-35 per cent,
is always left in solution. Whilst Verwer and Groll" contend that the
deposition is complete and free from carbon, the results of Avery and
Dales have been fully confirmed by Kohn, Brislee, and Froysell.” Accord-
ing to their experiments, the electrolytic method gives results from O2 to
o.3 per cent, too high, calculated on the weight of iron deposited, and its
apparent accuracy is due to the compensating errors of incomplete
deposition and the presence of carbon; the percentage of the latter
varies from O-43 to O-74, according to the length of time of the electro-
lysis, whilst O. I to O.2 per cent of iron is left in solution.
Ledebur * states that he has never found the value of iron wire to
exceed IOO per cent., and that it is safe to use the percentage of pure
iron present, as a basis; to determine this percentage, however,
accurately to within o. I per cent, is a matter of great difficulty, and
in any case carbides, etc., notoriously act as reducing agents.
According to Lunge,” electrolytic iron cannot be relied upon as pure ;
on the contrary, its use may lead to errors of several units in the first
place of decimals. It is very much better to determine the value of
a large stock of iron wire by means of permanganate, after having
previously standardised the latter by oxalic acid, referred to pure
sodium carbonate (p. IO4). Since sodium oxalate is free from all the
uncertainties connected with the use of iron, it forms a far more reliable
basis for standardisation than the latter.
3. Ferrous Ammonium Sulphate method. — The double salt
FeSO,. (NHA),SO,.6H,O, sometimes called Mohr’s salt, is more stable
than ferrous sulphate, and was therefore recommended by Mohr for
standardising permanganate. The dissolving of the iron is avoided and
the errors arising from the uncertainty, due to the variable amount of
carbon in iron wire are eliminated. Still greater uncertainty is, however,
introduced, owing to the presence of water of crystallisation and of im-
purities in the salt. The commercial salt is quite unreliable, and it is
difficult to prepare a perfectly pure salt containing exactly the right
amount of water of crystallisation. Where errors of some tenths per
cent. are not of consequence, this salt may be used for standardising;
it cannot be recommended for accurate work. The value of the salt
is 14.29 per cent. of that of iron, i.e. almost exactly one-seventh.
Graeger, and subsequently Biltz, proposed the use of ferrous
sodium sulphate, NagsO4. FeSO4.4H2O, in place of the ammonium
salt, but it has not come into general use. -
Meineke" also considers ferrous ammonium sulphate to be quite
unreliable, and recommends the use of iron alum ; this is reduced with
* Berichte, 1899, 32, 806. * British Association A'eports, 1900, p. 174.
* Stahl und Eisen, 1902, 22, 1242 ; cf. also, Thiele and Deckert, Z. angew. Chem., 1901, I4, 1233.
* Z, angew. Chem., 1904, 17, 265. * Z. Öffentl. Chem., 1898, 4, No. 13.
POTASSIUM PERMANGANATE - 109
stannous chloride, mercuric chloride finally added, and the solution
then titrated with permanganate, following Reinhardt's directions. The
amount of iron in the “chemically pure” iron alum must be checked
by precipitation with ammonia. WJowiszewski” has proposed a
method in which he prepares “chemically pure ferric oxide"; this is
reduced with stannous chloride, and then used to standardise perman-
ganate. Gintl” uses a spiral of palladium wire, Saturated with hydrogen,
as reducing agent, and prefers this to all other methods of reduction.
4. Volhard’s “Iodimetric method.—If a permanganate solution is
added to a solution of potassium iodide, acidified with hydrochloric or
sulphuric acid, a quantity of iodine is liberated, which is equivalent to
the amount of available oxygen in the permanganate. The iodine
dissolves in the excess of potassium iodide solution, and its amount
can be determined by adding excess of sodium thiosulphate and titrating
back with standard iodine solution:-
KMnO, + 5.KI+8HCl = 51 + 6KCl + MnCl2 + 4H2O.
This process, though theoretically quite correct, has the disadvantage
of requiring two additional standard solutions, on the accuracy of which
the process is dependent; the unavoidable summation of small errors
may easily lead to a final error of several tenths per cent.
Volhard gives a series of examples, according to which his method
gives similar results to those obtained with sodium oxalate, iron
ammonium alum and iron wire. The data, however, show discrepancies
up to O.3 per cent, and it must be borne in mind that both iron
ammonium alum, and iron wire may lead to errors of several units in
the first place of decimals, when used for the primary standardisation
of permanganate, so that Volhard's examples cannot be considered as
proof of the accuracy of his method. F. Dupré" states that the method
gives accurate results.
5. Titration with Hydrogen Peroxide.—Morse and Chambers" add
neutral hydrogen peroxide to a measured volume of standard sulphuric
acid; permanganate solution is then run in from a burette, until only a
small excess of hydrogen peroxide remains, and the free acid is then
titrated back. The decomposition is represented by the equation —
2KMnO, + 5H2O3 + 3H2SO4 = K,SO, + 2MnSO, + 5O2+8H2O.
6. Gas-volumetric, Hydrogen Peroxide, or Nitrometer method.”—
This method is not used as frequently as it deserves, partly because all
laboratories are not provided with the necessary apparatus, although it
* Stahl und Eisen, 1884, 4, 704; Chem. Zeit., 1889, 13, 323.
* Stahl u. Eisen, 1901, 21, 816. * Annalem, 1879, 198, 333.
* Z. angew. Chem., 1902, I5, 431. * Z. angew. Chem., 1904, 17, 815.
* Amer. Chem. J., 1896, 18, 236; Z. anal. Chem., 1898, 37, 183.
7 Lunge, J. Soc. Chem. Ind., 1890, 9, 21. Cf. also, H. v. Jüptner, Osterr. Zeit für Berg. und
Aillenwesen, 44, 371 ; Vannino, Z, angew. Chem., 1890, 3, 8o ; / Soc. Chem. Ind., 1890, 9, 552.
110 GENERAL METHODS USED IN TECHNICAL ANALYSIS
is applicable to a number of other purposes, and partly because it
belongs to a sphere foreign to the usual applications of permanganate.
It has certain fundamental advantages over all other methods. It
requires neither balance nor weights, a point of interest rather than of
practical importance; more important is the fact that it dispenses with
the use of any primary substance of undoubted purity, the requirement
which puts such great difficulties in the way of other methods. In
addition, none of those detours are involved which lead to an accumu-
lation of small errors. The available oxygen is measured directly, with
the qualification that its volume is doubled by an equal volume of
oxygen derived from the hydrogen peroxide, which halves the error
of measurement. The gas is measured over mercury, so that the
measurement can quite easily be made accurate to I/40 c.c., and an
accuracy to I/IOO c.c. can be attained by taking special precautions, such
as reading the volume of gas with a cathetometer, etc.; since the volume
of gas measured exceeds IOO c.c., the error in measurement is reduced
to 1/4000 or I/IOOOO, provided an accurately calibrated nitrometer is used,
a degree of accuracy which cannot be attained with burettes, etc. The
use of an accurate thermometer and barometer, or as an alternative, of a
calibrated “gas-volumeter” and an accurately adjusted “reduction
tube,” are, of course, assumed. This degree of accuracy will not be
attained in the standardisation of dilute permanganate solutions, since
-the amount of liquid which can be treated in the nitrometer is limited ;
but this applies to all other methods, since the standardisation of dilute
solutions is always relatively less accurate than that of strong solutions
(p. IOO).
The method is based on the following reaction between hydrogen
peroxide and a solution of potassium permanganate, strongly acidified
with sulphuric acid:—
2KMnO, +3H2SO, + 5H2O3= K,SO, + 2MnSO, +8H,O + 502.
Thus, whichever of the reagents is not in excess gives up the whole of
its available oxygen, and this is increased by an exactly equal volume
of oxygen derived from the other reagent.
The available oxygen in hydrogen peroxide can be determined by
decomposing it with an arbitrary amount of permanganate, provided the
latter is in excess, and similarly the available oxygen in permanganate
can be estimated by using an excess of hydrogen peroxide; it is then
only necessary to measure the volume of oxygen evolved, to reduce
this volume to O’ and 760 mm., and halve the result. The reaction
proceeds quite smoothly and is quantitative; the solutions retain no
oxygen, since they were previously saturated with air. Certain simple
precautions must be observed: viz., a large excess of sulphuric acid must
be present, but not too large an excess of hydrogen peroxide; the
POTASSIUM PERMANGANATE 2 111
apparatus must be adjusted immediately the solutions have been mixed,
and attention must be paid to those conditions which must be observed
in all gasometric methods, such as uniformity of temperature, etc.
The latter is easily attained, since the reaction involves no appreciable
rise of temperature.
Any form of apparatus for gasometric analysis can be used for the
determination; to ensure great accuracy, the evolved gas should be
measured over mercury. Lunge's nitrometer with attached bottle
is particularly suitable for the decomposition, and by employing it or
similar apparatus, in conjunction with a “gas-volumeter,” the calculation
of the volume of oxygen to standard conditions is saved. Both
instruments will be described subsequently (pp. 132 and I38); the
following description is therefore restricted to their application to the
valuation of permanganate solutions.
Since I litre of dry oxygen at O’ and 760 mm. weighs 1.429 g. in
these latitudes, and half of the evolved oxygen comes from the per-
manganate, every C.C. of the measured volume corresponds to O-7145
mg. available Oxygen in the permanganate. Hence I c.c. of an accu-
rately semi-normal permanganate solution will give 5-597 c.c. of oxygen
(N.T.P.); an AV/IO solution will give I. I 19 c.c. Oxygen per c.c. of solution,
and so on. It is not practicable to decompose much more than 20 c.c. of
permanganate Solution at a time, so that it is best to take exactly
20 c.c. An M/2 solution will then give about I 12 c.c. gas (N.T.P.), an N/Io
solution about 20 c.c.; ordinary conditions of temperature and
pressure will increase these volumes by from 5 to 15 per cent, so that for
an M/2 solution a nitrometer with a 150 c.c. bulb tube should be used ;
for an M/IO solution, one with a 50 c.c. straight tube will suffice.
The valuation is conducted as follows: 2O c.c. of permanganate
solution are accurately measured into the outer compartment of
the decomposition bottle and 30 c.c. dilute sulphuric acid (1 : 5)
added. For AW/2 permanganate, 15 c.c. ordinary commercial hydrogen
peroxide are then introduced into the inner vessel of the decomposition
bottle, by means of a pipette; for weaker solutions less peroxide is
taken. A preliminary experiment should be made, in a beaker, to ensure
the presence of an excess of hydrogen peroxide; any considerable
excess is unnecessary and may constitute a source of error, but the
results are quite accurate even when double the theoretical amount
of hydrogen peroxide is taken. The decomposition bottle is then
closed by the stopper, and the resulting excess pressure relieved by
momentarily opening the tap of the measuring tube; after making
sure that the mercury fills this tube right up to the tap, the latter is
turned so that the measuring tube communicates with the decomposi-
tion bottle. The solutions in the bottle are then mixed, by tilting, and
the bottle shaken for exactly three minutes; shorter shaking is not
112 GENERAL METHODS USED IN TECHNICAL ANALYSIS
sufficient, and longer shaking may vitiate the results. The level of the
mercury in the pressure tube and the measuring tube is then adjusted
and the stopcock closed. The decomposition bottle should be held by the
rim when being shaken, and the usual precautions must be taken to avoid
any warming of the measuring tube and bottle. For accurate work, it
is preferable to place the bottle, before the decomposition, in a vessel
containing water at the temperature of the room, and to replace it in
this vessel after shaking and before adjusting the pressure tube, so that
the temperature is the same at the beginning and at the end of the
experiment.
The volume of gas in the measuring tube is then calculated to N.T.P.,
allowance of course being made for the tension of aqueous vapour;
or the correction may be made by the gas volumeter. Each c.c. of the
corrected volume equals O.7145 mg. available oxygen in the perman-
ganate used ; supposing that 20 c.c. of the latter gave II 2.75 c.c. gas,
II 2.75 × O-7 I45
2O
able oxygen. This corresponds to a solution a little stronger than
semi-normal, and having the factor I.OO77.
The Oxalic acid and iron methods require an appreciably longer time
to carry out than this hydrogen peroxide method, and are, moreover,
dependent upon the use of pure substances. The deviations in these
methods do not exceed I/IOOO, which is really the limit of possible
accuracy in individual observations with them all; this has been shown
both by Lengfeld" and by Lunge.” Comparative experiments by
Vanino,” using the azotometer and the iodimetric method, confirm this
conclusion; this author gives a table for the weight of I c.c. of oxygen
at different temperatures and pressures.
The applications of permanganate are extremely numerous, and
include the estimation of iron, oxalic acid and oxalates, nitrous acid and
nitrites in sulphuric acid solution, potassium ferrocyanide, tanning
materials, and hydrogen peroxide; it is also used indirectly, for titrating
back excess of reducing agents, e.g., ferrous oxide in the estimation of
nitric acid, nitrates, chlorates, manganese dioxide, peroxides, bleach, etc.
then I c.c. permanganate would contain =4-O3 I mg. avail-
IODIMETRY
The iodimetric methods of volumetric analysis were introduced by
Bunsen, and are now very extensively used ; they are amongst the most
accurate known, owing to the extreme sensitiveness of the indicator,
starch solution, that is employed. -
Iodine oxidises many substances indirectly, by combining with the
* Z, angew. Chem., 1890, 3, II. * Ibid., 1904, 17, 270.
* /bid., 1890, 3, 8o ; / Soc. Chem. Ind., 1890, 9, 552.
IODIMETRY 113
hydrogen of water, liberating the oxygen, which then combines with
other compounds present. In other cases it combines directly with the
hydrogen or sodium in the substance to be estimated, as in the deter-
mination of hydrogen sulphide and of sodium thiosulphate.
The “re-sublimed iodine” of commerce is a very pure product, which
may be used for many purposes without further purification; it only
requires to be dried in the exsiccator to free it from water, for the
preparation of normal solutions. For accurate standardisation, however,
the iodine must be further purified (p. 118).
An M/Io iodine solution is generally used; it is prepared as follows:
12.7 g., or a little more, pure re-sublimed iodine are weighed out on a
rough balance and shaken into a litre flask containing a solution of 15
to 18 g. potassium iodide in about 30 g. water; the flask is closed,
shaken till the iodine is completely dissolved, and then filled up to the
mark. The solution is then titrated against N/Io arsenious oxide or
sodium thiosulphate Solution, previously standardised by perfectly pure
iodine (p. I 18).
For the estimation of small quantities of sodium sulphide a special
solution, prepared by dissolving 3.252 g. pure iodine and 5 g. potassium
iodide and diluting to I litre, is sometimes used ; I c.c. of this solution
=O.OOI g. sodium sulphide. -
The use of N/IOO iodine solutions has been recommended, on the
ground that it can be used in pinch-cock burettes; experiments by
Lunge" have shown, however, that such dilute solutions readily attack
rubber. Further, a comparz ely large excess is necessary to give a dis-
tinct blue coloration with starch, though this disadvantage can be
lessened by using a large excess of potassium iodide.” Iodine solutions
should be kept in well-stoppered bottles and stored in a cool place; the
strength should be checked about once a month, as the contained iodine
is distinctly volatile. For this reason the solution, when in use, should
be kept out of contact with the air as much as possible, and the titrations
should be conducted in bottles or narrow-necked flasks, not in beakers or
dishes; for the same reason, the iodine should be run into the solution
to be titrated, whenever possible, and not vice versa. This does not
apply to the titration of sulphites, since the action of atmospheric oxygen
on the sulphite must be avoided (cf. infra). Burettes with glass taps
should be used and the taps greased with vaseline. The top of the
burette should be kept well closed, and only be opened sufficiently, when
in use, to allow the solution to flow from the burette. The volatility of
iodine must also be borne in mind when gases containing Sulphur
dioxide or sulphuretted hydrogen are bubbled through iodine solution,
in order to absorb these gases.
In conjunction with the iodine solution, starch Solution is used as
* Z, angew. Chem., Igo4, 17, 197. * Treadwell, Analytical Chemistry, vol. ii., p. 513.
H
114 GENERAL METHODS USED IN TECHNICAL ANALYSIS
the indicator, and either sodium thiosulphate or arsenious oxide solutions,
of corresponding strength, as the reducing solution.
Starch solution may be prepared as follows:–3 g. potato starch are
ground with a little water to a uniform paste, and gradually added to
3OO g. boiling water in a porcelain dish; boiling is continued until an
almost clear solution is obtained. This is allowed to settle in a deep
vessel, and the clear liquid filtered and saturated with common salt. The
solution keeps for a considerable time if kept in a cool place; it should
be discarded as soon as fungoid growths are observed in it. The
addition of mercuric iodide" preserves the solution.
Zulkowsky's soluble starch is a more convenient reagent; it is
obtained in the form of a thick paste, which should not be allowed to
dry up ; a small quantity is taken out, on the end of a glass rod, as
required for use.
De Koninck” recommends grinding 2 g of potato starch into a paste
with cold water and adding this to a litre of boiling water; the whole is
then boiled for two or three minutes, and 8 c.c. of a 10 per cent.
Solution of potassium iodide, saturated with mercuric iodide, added.
The solution is allowed to cool, poured into a tall glass cylinder, and the
clear portion decanted, after standing for one or two days. Starch
Solutions prepared in this way keep perfectly well for years.
Wroblewski% uses potassium hydroxide for preparing soluble starch,
Förster," hydrochloric acid, and Syniewski,” sodium peroxide. Another
preparation is known as “ozone starch " (manufactured by C. Conrad,
Kyritz); this is best rubbed up with cold water, poured, with stirring,
into boiling water and rapidly cooled. The solution keeps for a
considerable time, but mould develops after three or four weeks.
Iodine solution is generally standardised by an accurate solution of
Sodium thiosulphate or of arsenious oxide, the strength of which has
been fixed by pure iodine (p. I 18). It may also be directly standardised
by a gasometric method, using the nitrometer; in this case the following
directions must be accurately adhered to. Fifty c.c. of the iodine solution
are placed in the outer compartment of the decomposition bottle, and a
freshly prepared, cold mixture of 6 c.c. of 2 per cent. hydrogen peroxide and
8 c.c. of 50 per cent. potassium hydroxide are placed in the inner vessel.
Any considerable excess of hydrogen peroxide must be avoided, and
its strength must not exceed 2 per cent. After connecting with the
measuring tube, a circular motion is first communicated to the bottle
without allowing the solutions to mix; the bottle is then suddenly
tilted through 90°, so as to effect an instantaneous and intimate mixing
of the two solutions. After shaking for not more than a few seconds
* Mutniansky, Z, anal. Chem., 1897, 36, 220. * Chem. Zeit., 1898, 22, 375.
* Private communication to Prof. Lunge. * Ibid., 1897, 21, 41.
* Ber, 1897, 30, 24I5 ; 1898, 31, 1791.
IODIMETRY 115
the pressure is adjusted and the volume of gas read off. One c.c. oxygen
corresponds to O.OI 133 I g. iodine according to the equation:—
I2 + H2O2 = 2 HI+ Og.
Kalmann" recommends sodium sulphite for the standardisation of
iodine. A solution of arbitrary strength is run into the iodine solution
until it is just decolorised, when the following reaction takes place:–
Na2SO44-2I + H2O = 2 HI+ NagSO4.
Thus the acid liberated is exactly equivalent to the iodine. The acid is
then estimated by standard potassium hydroxide, using methyl orange as
indicator. The method is independent of the amount of water and of
sulphate in the sodium sulphite, provided the latter is perfectly free from
bisulphite, thiosulphate, and carbonate. The decomposition goes best
when the solution is not too dilute. One c.c. M/Io alkali hydroxide corre-
sponds to O.OI 2697 g. iodine.
The principle of this method is perfectly sound, but the difficulty of
obtaining sulphite free from bisulphite, thiosulphate, or carbonate makes
it practically useless for standardisations; “chemically pure” sodium
sulphite generally contains some carbonate. A way out of this difficulty
would be, to add phenolphthalein and M/5 hydrochloric acid till decolor-
ised, to decompose the carbonate present, forming sodium bicarbonate,
then a further quantity of hydrochloric acid equal to that already used,
thus converting the bicarbonate into sodium chloride and obtaining
a perfectly neutral solution of sodium sulphite; but this method is not
practicable at ordinary temperatures, experiments by Lunge” having
shown that under these conditions the colour change is gradual and
uncertain. e
S. W. Young” recommends anhydrous sodium thiosulphate for
standardising iodine solutions; tartar emetic has been proposed by
Metzl,” and potassium bichromate by Bruhns' and others.
Kalmann's method can, of course, be conversely applied to the
determination of sulphurous acid, but it offers no advantage over
direct titration with iodine, and is, in fact, less accurate; further, it may be
used for determining sulphite, in presence of thiosulphate, by running the
latter from a burette into a measured volume of standard iodine until just
decolorised, and then titrating with M/Io sodium hydroxide and methyl
orange. The sulphite is given by the amount of iodine equivalent to the
alkali used ; by subtracting this amount from the total iodine used, the
iodine corresponding to the thiosulphate is obtained.
Experiments by Finkener, Volhard, and others have shown that the
reaction between iodine and sulphurous acid or sodium sulphite is only
quantitative, provided the solution of the latter is added drop by drop,
| Ber., 1887, 20, 568. * /. A mer. Chem. Soc., 1904, 6, Io28.
* Z, angew. Chem., 1904, 17, 235. * Z, amorg. Chem., 1906, 48, I56.
* Ibid., 1906, 49, 277.
116 GENERAL METHODS USED IN TECHNICAL ANALYSIS
with stirring, to the iodine solution; in the reverse process too little
iodine is used. This has been confirmed by Raschig."
Berg” has pointed out that this is not due to the formation of hydro-
sulphurous acid, as Volhard supposed, but rather to the action of
atmospheric oxygen on sulphur dioxide and to volatilisation of the
latter; by following the above directions this effect is more easily
avoided than when the iodine is added to the Sulphurous acid solution.
Rupp * states that sulphur dioxide can be accurately titrated in
alkaline solution containing bicarbonate. This can only be done,
however, by adding an excess of iodine, leaving it to react for fifteen
minutes and then titrating back, as the oxidation velocity of sulphites
is appreciably smaller than that of free sulphurous acid. According to
Ruff and Jeroch,” the correct results obtained by this method are due to
a fortuitous compensation of errors. Accurate results can be obtained
by omitting the back-titration, and by excluding the action of oxygen,
by the addition of mannitol and working in an atmosphere of carbon
dioxide.
The applications of iodine solution, known as iodimetry, generally
in conjunction with a corresponding sodium thiosulphate or arsenious acid
solution, are very numerous, and include the following estimations:—
Sulphurous acid, either as gas, or in very dilute aqueous solution, or
dissolved in water, to any concentration, in the form of salts; if free
sulphurous acid has to be estimated, it need only be neutralised by
sodium carbonate. The sulphite solution must be added to the iodine,
not vice versa (p. I I5).
Thiosulphates are converted into tetrathionates (cf. infra).
Free Chlorine and Bromine are estimated by liberating iodine from
potassium iodide. Similarly, the chlorine of hypochlorous acid and
of hypochlorites may be determined; Lunge has proposed the following
method for their simultaneous estimation. When potassium iodide acts
on hypochlorous acid, half as much alkali is liberated as when it acts on
hypochlorites:—
HOC1+ 2KI = KOH + KCI+I,
NaOCl 4-2KI+ H2O = 2KOH + NaCl 4-I,
Hence free hypochlorous acid liberates two equivalents of iodine to one
equivalent of potassium hydroxide, and any excess of alkali above this
is a measure of the amount of hypochlorite present.
Bleach may be estimated by iodimetry, but it is better to titrate it
directly with arsenious acid solution (p. P22).
Sulphuretted hydrogen and Sulphides react as follows:—
H.S +I, + 2 HI+S.
This decomposition is quantitative when the solutions are dilute, and are
* Z, angew. Chem., 1904, 17, 579. * Ber, 1902, 35, 3694.
* Aull. Soc. Chim., 1902 [3], 27, Ioz7; Chem. Centr., 1903, I, 249. * Ibid., 1905, 38,409.
SODIUM THIOSULPHATE 117
protected from the oxygen of the air, . Hence also, all those substances
may be thus estimated which liberate sulphuretted hydrogen on
decomposition with hydrochloric acid, the evolved gas being absorbed in
iodine solution.
Arsenious oride :-
AsgOs + 2 H2O+4I = As,C), +4.H.I.
Manganese dioxide and other Perovides.
Chloric acid, Chromic acid, and their salts may be decomposed by
hydrochloric acid, the chlorine liberated absorbed in potassium iodide,
and the resulting free iodine estimated. Many methods may be included
under this head, in which chromate is added in excess and the excess
titrated back.
Iodimetry may also be made use of in connection with Acidimetry;
the following reaction ensues on acidifying a mixture of potassium iodate
and excess of potassium iodide :-
KIO, + 5.KI+6HC1 = 614-6KCl +3H,O.
The free iodine is estimated by thiosulphate or arsenite solution, and
thus the accuracy of iodimetry may be conveyed to acidimetry. This
reaction may also be made the basis for the estimation of free alkali, by
adding an excess of standard acid to the solution and then estimating
the excess of the latter, as above."
Many ascribe a much greater accuracy to iodimetry than to alkali-
metry, and this is correct in so far that the iodine-starch reaction is much
more sensitive and obvious than are the colour changes of alkalimetric
indicators. But when the volatility of iodine and the recognised
uncertainties in calibrating apparatus and in taking readings are taken
into account, an accuracy of + rºotſ of the total may be considered as
good work; this degree of accuracy can also be attained or even exceeded
with litmus or methyl orange in acidimetry, provided every care is taken.
SODIUM THIOSULPHATE
An AV/IO solution is prepared by dissolving 24.83 g of the pure
crystallised salt Nags,CA. 5H2O in water, which has been completely
freed from air by boiling; the solution is diluted to I litre with cold
air-free water.
Meineke” states that pure sodium thiosulphate can be obtained by
repeatedly recrystallising the “chemically pure” commercial salt from
water and then rubbing it to a fine powder with strong alcohol. The
resulting magma is thoroughly washed on the filter pump with pure
ether, and dried on the pump until the ether is almost completely
evaporated; the powder is then removed from the filter, covered over
with paper, left to dry in the air for twenty-four hours, and then trans-
* Gröger, Z. angew. Chem., 1890, 3, 353 and 385. * Chem. Zeit., 1894, 18, 33.
118 GENERAL METHODS USED IN TECHNICAL ANALYSIS
ferred to a tightly closed bottle. According to Meineke, the salt
prepared by this method is pure enough (99.99 per cent.) to be used
for standardising iodine solution, but this is not the case, because the
percentage of water in the salt is by no means reliable.
Pure sodium thiosulphate should dissolve in water without the least
turbidity, should give no precipitate with barium chloride (a precipitate
would indicate the presence of sulphate, sulphite, or carbonate), and
should not redden phenolphthalein, (presence of carbonate). It should
also be free from sulphide and chloride.
Iodine and sodium thiosulphate react with the formation of sodium
tetrathionate, the iodine combining with half the sodium :—
2NagS,Os-- 2 I = 2 NaI+ NagSOs.
Whenever possible, the iodine is added to the thiosulphate solution;
the latter should be neutral or faintly acid, because if more strongly
acid free thiosulphuric acid is formed, which decomposes on standing
into water, sulphur dioxide, and free sulphur. A slight yellow coloration
is observed as soon as an excess of iodine is present. The change is
sharper in presence of starch solution, but the blue coloration is only
obtained at ordinary temperatures. Starch solution must never be
added whilst a large excess of iodine is present; for instance, if the
thiosulphate or arsenic solution is run into the iodine the starch is
decomposed by the iodine, forming variously coloured compounds.
The starch should be added towards the end of the reaction, when
only very little iodine is present, if the titration is done in this
way.
When the iodine solution has been accurately standardised, the
thiosulphate solution may be prepared from the ordinary “pure”
commercial salt; it is then compared with the iodine, and corrected.
The following method may be used as a further check on the
thiosulphate solution and serves as a means for its standardisation,
and indirectly for that of iodine solution. A small quantity of pure,
dry iodine is prepared, by powdering together o. 5 g. pure iodine and
O. I g. potassium iodide and heating the powder in a small porcelain
dish on a sand-bath, or on an asbestos card, until vapour is copiously
evolved; the dish is then covered with a dry watch-glass, and the
greater part, but not the whole of the iodine, is sublimed on to the
latter. The watch-glass is then covered with a second, accurately fitting
glass and the whole weighed, the weight of the two watch-glasses
of course being known. The watch-glasses are placed gently in a
solution of I g. potassium iodide in Io g. water; after waiting for a
moment, to allow the iodine to dissolve, the solution is diluted with
IOO c.c. of water and titrated with the thiosulphate solution. The
potassium iodide must be free from potassium iodate; a test for its
SODIUM THIOSULPHATE 119
purity in this respect consists in adding hydrochloric acid to a not
too concentrated solution of the salt, when no immediate yellow colora-
tion develops, in absence of iodate. When the colour has been reduced
to a pale yellow, a little starch solution is added and the titration
continued until the blue colour just disappears. If the thiosulphate
solution is of right strength, the amount used in c.c. multiplied by
O-oſ 2697 equals the weight of iodine taken.
According to Lunge, this method of preparing, weighing, and dis-
solving pure iodine is perfectly accurate in the hands of a fairly
experienced worker, and when due attention is paid to the obvious
necessity of rapid work, to guard against the volatilisation of the iodine.
The following modification of the method recommended by Treadwell!
is easier to carry out. Two to 2.5 g. pure potassium iodide and not
more than O. 5 c.c. water are placed in each of two or three small
weighing tubes with tightly fitting, ground stoppers; the tubes are
closed and weighed, and then O-4 to O-5 g. pure iodine added to each,
and weighed again. The iodine dissolves almost instantly in the
concentrated potassium iodide solution. One of the tubes is then
placed in the neck of a 500 c.c. Erlenmeyer flask, held in an inclined
position and containing 200 c.c. of water and about I g. potassium iodide.
The tube is dropped carefully to the bottom of the flask, the stopper
being removed just as it begins to fall, and allowed to follow it. In
this way no iodine can be lost. The examples given by Treadwell
show deviations of about O.2 per cent. from each other; these deviations
can be reduced to O. I per cent., when using the first method given
above, after considerable practice.
Winkler has proposed a method for preparing pure iodine in quantity,
which consists in adding 5 per cent, potassium iodide and IO per cent.
quicklime (to retain water) to the crude iodine; the mixture is sublimed
from a vessel covered with a flask filled with water, on which the sub-
limed iodine condenses. A funnel, the tube of which is loosely plugged,
serves equally well to collect the iodine.
De Koninck’s ” method of preparing pure iodine for standardisation
is as follows:—Potassium iodide is freed from potassium cyanide by
recrystallising in presence of a little hydriodic acid ; one part of the
recrystallised salt is then intimately mixed with I? parts previously
fused potassium bichromate, both having been finely powdered and
dried, and the mixture placed in a small retort, the neck of which
fits tightly into a small Erlenmeyer flask. The latter is previously
weighed along with a blown glass bulb, which subsequently serves as
a stopper; the weight of the condensed iodine can thus be determined.
The retort is gradually heated to incipient redness, the neck being
sufficiently warmed to prevent iodine condensing in it; the flask must
* Analytical Chemistry, vol. ii., p. 507. * Chem. Zeit., 1903, 27, 192.
120 GENERAL METHODS USED IN TECHNICAL ANALYSIS
be protected from the heat by an asbestos card. The yield of iodine
exceeds 95 per cent. The reaction is:—
The assumption that the decomposition is represented by the
equation,
6KI+ K2Cr,C) = 6I+ 4.K.O + Cr;Os,
is quite erroneous."
Lean and Whatmough * consider that the purest iodine is obtained
by heating cuprous iodide.
When iodine is kept in a desiccator, the latter should contain calcium
chloride and not sulphuric acid, and the cover should not be greased ;
neglect of these precautions leads to contamination of the iodine, since
sulphuric acid is likely to be taken up by the iodine and grease is
attacked by iodine vapours, forming hydriodic acid.
Meineke” prepares chemically pure iodine, for standardisation pur-
poses, by precipitating a Solution of equal parts potassium iodide and
iodate with sulphuric acid. He subsequently recommended 4 the use
of potassium bi-iodate, which had been previously suggested by C. v.
Than;" the commercial form of this salt is said to be extremely pure
and to be quite permanent, even in dilute solution. In acid solution
the following reaction takes place:–
KH(IOS), + IoRI+ II HCl = 12 I-II KCl + 6H,O.
Hence 390. I parts of pure potassium bi-iodate correspond to 1523-6
parts iodine.
This method appears very convenient, particularly for checking
thiosulphate solutions, from time to time, by means of a solution of
the bi-iodate, but the purity of any commercial bi-iodate cannot be
relied upon, as the salt may contain either too much or too little iodic
acid. Experiments by Lunge,” with samples from the very best sources,
gave results which would by no means warrant their use for direct
standardisations, and their reliability was not improved by recrystallisa-
tion, either with or without the addition of free iodic acid. It was also
observed that desiccation over sulphuric acid at ordinary temperatures,
as recommended by J. Wagner, is insufficient, and that a temperature of
Ioo” is required to dry the salt, as stated by Meineke; no appreciable
decomposition results from this treatment. Treadwell" also states
that commercial bi-iodate cannot be relied upon for direct standardisa-
tion. Hence, if it be desired to use potassium bi-iodate, its value must
first be determined; this cannot be done directly by acidimetry (cf.
* Chem. Centr., 1903, I, I435. :* º I9, is
* /. Chem. Soc., 1898, 73, I48. . anal. Chem., 1877, 17, 477.
* Chem. Zeit., 1892, 16, 1219. * Z, angew. Chem., 1904, 17, 233.
7 Analytical Chemistry, vol. ii., p. 509.
SODIUM THIOSULPHATE 121
p. 89), so that sodium thiosulphate solution, standardised in some other
way, must be resorted to. This negatives the value of bi-iodate for
direct standardisations, and restricts its application to the occasional
checking of thiosulphate, by means of a prepared solution of bi-iodate,
the strength of which has been determined. Titration against pure
iodine, as described on p. I 18, is to be emphasised as the best method
of standardisation, in spite of the many other proposals that have been
put forward.
Riegler" has recommended crystallised iodic acid for standardising
thiosulphate, but Walker * has pointed out objections to the method.
Kratschmer * has suggested sodium bromate, Zulkowsky,” and subse-
quently Crismer,” normal potassium chromate, but Meineke" has indi-
cated the disadvantages of this last reagent. J. Wagner" considers the
use of potassium bichromate, under perfectly definite conditions, preferable
to other reagents. Dietz and Margosches * use potassium chlorate,
which, on addition of hydrochloric acid, potassium bromide, and potas-
sium iodide, liberates an exactly equivalent amount of iodine.
M/IOO thiosulphate is occasionally required, especially when using
M/IOO iodine; the former must always be freshly prepared, as it does not
keep long. Neither is M/Io thiosulphate permanent under all conditions.
Carbon dioxide acts upon it, in presence of oxygen and sunlight, forming
sulphurous acid and free sulphur, and fungoid growths also develop.
Topf,” after subjecting these changes to exact investigation, recommends
that the solution be prepared with well-boiled water, free from carbon
dioxide and stored in a cool place, protected from direct sunlight; a
small quantity of potassium carbonate (not ammonium carbonate)
should be added, but this necessitates addition of free acid to the
iodine solution, when titrating. He further recommends that the solu-
tion be covered with a layer of petroleum ether, and that communication
with the air be permitted only through an arrangement for absorbing
carbon dioxide, the solution being withdrawn by means of a syphon.
This proposal is probably rarely adopted, as it will be found preferable
to check the strength of the solution about every two months.
Treadwell 19 states that all difficulties, due to the presence of carbon
dioxide, etc., in preparing and storing thiosulphate solutions, may be
overcome by dissolving the “pure” commercial salt in distilled water
(even if it contains carbon dioxide) and allowing it to stand for at least
a week (eight to fourteen days) before standardising. The carbon
dioxide thus exhausts its action, a corresponding amount of sulphur
* Chem. Centr., 1897, I, II69. * Chem. Zeit., 1895, 19, 3.
* Z. amorg. Chem., 1898, 16, 99. 7 Z. angew. Chem., 1898, II, 951.
* Z, anal. Chem., 1885, 24, 546. * Ibid., 1903, 16, 317; 1905, 18, 1516.
*J. prakt. Chem., 1868, Io:3, 351. * Z. anal. Chem., 1887, 26, 150.
* Ber., 1884, 17, 642. "Analytical Chemistry, vol. ii., pp. 506 and 510.
122 GENERAL METHODS USED IN TECHNICAL ANALYSIS
separates, and the solution can then be kept for months without under-
going appreciable change. Ammonium carbonate should on no account
be added to the solution.
Plimpton and Chorley' have recommended the use of barium thio-
sulphate for iodimetry; it is a very difficultly soluble salt. Mutniansky”
has pointed out that a solution of this salt, saturated at 17°. 5 C., is
exactly centinormal.
ARSENIOUS ACID
An alkaline solution of arsenious oxide is unaffected by oxygen,
but reacts with chlorine as follows:—
As2O3 + 4C1 + 5H2O=4HCl + 2 HaAsO4.
Bromine and iodine act similarly, in alkaline solution. Hence, in
many cases, arsenious acid Solution may be used in place of sodium
thiosulphate in conjunction with iodine; it has the advantage of being
quite permanent. Its chief application is to the estimation of available
chlorine in bleaching powder and in alkaline hypochlorites, the end-
point of the reaction being ascertained by spotting on starch-iodide
paper (cf. Bleaching powder, p. 503).
Pure, powdered, commercial arsenious oxide is used in preparing
this solution ; it should always be tested by subliming a small
quantity from a dish on to a watch-glass; the first portion of the sub-
limate should not be yellowish, which would indicate the presence of
the more easily volatile arsenious sulphide, and the whole should
volatilise completely on heating more strongly. The oxide should be
dried in the exsiccator over sulphuric acid for some time before use; it
may then be weighed out without special precautions, as it is not
hygroscopic.
To prepare a decinormal solution, exactly 4,950 g. arsenious oxide
are weighed out and boiled with Io g. pure sodium bicarbonate and
2OO c.c. water, till completely dissolved ; IO g. of sodium bicarbonate
are then added, the solution allowed to cool and diluted to I litre.
Reuter and Petriccioliº consider this method to be inaccurate, owing
to the formation of sodium carbonate, which absorbs iodine, but Lunge 4
has shown that this objection lacks foundation, and that there is no
reason for altering the method of preparing the solution as described.
The solution keeps perfectly; I c.c. is equivalent to O.OO3545 g.
chlorine, or O-OI 2697 g. iodine.
This solution will be of correct strength, provided pure dry arsen-
ious oxide is used. It may be standardised, however, against pure iodine,
exactly as described on p. I 18, in connection with sodium thiosulphate,
* J. Chem. Soc., 1895, 67, 315. * Z. angew. Chem., 1901, 14, II81.
* Zeit, anal. Chem., 1897, 36, 220. */bid., 1901, 14, 1298.
SILVER NITRATE AND AMMONIUM THIOCYANATE 123
and this check should not be omitted when a large quantity of
solution is prepared.
Lunge" has shown that sodium arsenite and sodium thiosulphate
give identical results and are equally applicable.
SILVER NITRATE AND AMMONIUM THIOCYANATE
Silver nitrate solution is employed for the volumetric estimation of
chlorides, by precipitation as silver chloride, using potassium chromate or
sodium arsenate as indicator, in those cases in which rapidity is of
greater importance than extreme accuracy; it is also applied to the
estimation of cyanides. The method was introduced by F. Mohr.
The preparation of this solution is extremely simple, as commercial
crystallised silver nitrate is dry and chemically pure, and only requires
to be dried in the desiccator, as a precaution, before being weighed out.
A decinormal solution is obtained by dissolving 16.994 g. silver nitrate
and making up to a litre ; I c.c. of this solution is equivalent to
O.Oo3545 g. Cl, O.OO3645 g. HCl, O.OO5850 g. NaCl, etc.
A solution equivalent to O.OOI g. NaCl per C.C. is sometimes
preferred ; this is obtained by dissolving 2.906 g. silver nitrate per
litre of solution.
The solutions to be estimated must not contain any free acid ; if this
is present, it must be neutralised by sodium carbonate. A slight excess
of carbonate is immaterial, whilst the least excess of acid interferes with
the reaction. Four to five drops of a cold, saturated solution of normal (yel-
low) potassium chromate are added to the chloride solution, and the silver
solution is then run in, with vigorous stirring, until the white precipitate
has acquired a reddish colour, which does not disappear on stirring ; the
colour is seen even better by artificial light than in daylight, as the
yellow colour of the solution is not apparent, whilst the red of the precipi-
tate persists. With a decinormal silver nitrate solution an excess of
about O.2 c.c. is required ; this amount must be subtracted from the
quantity of solution used. The colour change is sharper when sodium
arsenate is used as indicator, and a correction for excess is unnecessary.
A titration which has been carried too far can be corrected by addition
of W/io sodium chloride solution.
Volhard's" method permits the use of acid solutions. A solution of
ammonium thiocyanate is required in addition to M/IO silver nitrate; this
is prepared by dissolving 7.5 to 8 g. of the salt, which is always moist, in
I litre of solution and accurately adjusting it to the silver nitrate
solution. Once standardised, the titre remains perfectly constant.
A solution of iron alum (ferric ammonium sulphate), saturated
at the ordinary temperature, serves as indicator; 5 c.c. of this are
* Z. angew, Chem., 1904, 17, 233. * Annalem, 1878, I90, 49.
124 GENERAL METHODS USED IN TECHNICAL ANALYSIS
added to from 200 to 300 c.c. of the solution to be titrated. To fix the
strength of the thiocyanate solution, IO or 20 c.c. silver nitrate solution
are diluted with 200 c.c. water, and 5 C.C. iron alum solution are added ;
if a colour develops, nitric acid is added until it disappears. The
thiocyanate solution is then run into the silver solution, with continual
stirring ; the appearance of a light-brown coloration indicates the end-
point of the reaction, and its observation is facilitated by the coagulation
of the silver thiocyanate. Henriques has pointed out that the thio-
cyanate must always be run into the silver Solution, not vice versa. The
thiocyanate solution thus tested is corrected and made up to decinormal
strength, as described on p. 86. For chloride estimations, the solution
to be analysed is treated with AW/IO silver nitrate until all the chlorine is
precipitated as silver chloride and excess of silver nitrate is present;
iron alum is then added and the excess of silver titrated back with M/Io
thiocyanate. The amount of chloride is given by the difference between
the volumes of silver nitrate and of thiocyanate solutions used.
GENERAL REMARKS ON VOLUMETRIC ANALYSIS
Correct illumination is important in almost all volumetric work, to
ensure the recognition of colour changes such as are produced by the
addition of a single drop of solution. Direct sunshine may be as
harmful as the insufficient light of a dull winter's day, or of an unsuitable
position in the laboratory. Wherever possible, the burettes should be
placed opposite a window, preferably with a north light, even if only to
facilitate the reading. Colour changes, especially that of litmus, are
generally less distinct in artificial light; yellow light is particularly
objectionable in this respect, but incandescent gas light or electric light
are nearly as good as daylight, after some experience has been gained
with these illuminants.
In all cases a pure white background should be placed immediately
under the vessel in use and spread over a considerable area of the
bench. The nicest and cleanest arrangement is a tiled bench, but a
sheet of white paper is almost as serviceable.
In order to be sure, in doubtful cases, whether a colour change has
really occurred or not, the flask should be observed by both reflected
and transmitted light, and it should also be looked at from the side and
from above. A beginner will frequently find it advantageous to have
two other vessels, of the same size as the titration vessel ready to hand,
each filled with a solution coloured to the same intensity as the solution
to be titrated, and in one of which the colour change has been effected.
These considerations apply to most colorimetric work.
It is generally advisable, in cases of uncertainty in colour changes,
to read the burette and then to add a further drop of the solution; if
GAS-VOLUMETRIC ANALYSIS 125
only the intensity of the colour changes and not its shade, this last drop
should not be reckoned in the titration. It is frequently more certain
to titrate, drop by drop, until the change has become quite definite, and
then to determine the exact end-point by titrating back with the
appropriate standard solution.
The liquid should be agitated during the titration, especially towards
the end of the reaction. It is better to shake the vessel well than to
stir with a glass rod, as the falling drops may easily splash off the
latter. A beaker can be shaken in this way without fear of loss,
provided it is not more than half full, but it is better to use an
Erlenmeyer flask with a short, wide neck. Narrow-mouthed flasks are
not convenient, as drops from the burette may fall outside.
The importance of the quality of glass has been previously discussed
(p. 54).
When a solution has to be boiled in a porcelain dish, the dish
should on no account be more than half full; otherwise the spitting
of the solution may easily be a source of loss.
H. GAS-VOLUMETRIC ANALYSIS
Gas-volumetric analysis, as distinct from gas analysis, comprises
those operations in which a constituent of a solid or liquid substance is
determined by the generation and measurement of a gas. The gas is
usually evolved by the decomposition of the substance under examination;
for instance, carbon dioxide from limestone, nitrogen from ammonium
salts, nitric oxide from nitrates, oxygen from peroxides, etc.; some-
times, however, it is partially or entirely generated from the reagent
used, as in the valuation of hydrogen peroxide by potassium per-
manganate or of zinc-dust, by means of hydrogen liberated from water.
In some instances, air displaced by the generated gas is measured
instead of the latter.
The description of the methods and forms of apparatus proposed
for gas-volumetric work is, in this section, restricted to such as are
employed for a number of different purposes and are in general
laboratory use. Those forms of apparatus which are either used only
for a special purpose, or in connection with a particular industry, are
described in the corresponding sections.
The calibration of gas-volumetric apparatus has already been
considered (p. 44).
The Azotometer.
This instrument was designed by Knop" for the determination of
ammonia, by the decomposition of its salts with sodium hypobromite,
* Chem. Centr., 1860, 5, 244. 2. anal. Chem., 1874, 13, 383; 1876, 15, 250.
126 GENERAL METHODS USED IN TECHNICAL ANALYSIS
and measurement of the liberated nitrogen. . It has been improved in
several respects, especially by P. Wagner, and can be used for other
gas-volumetric purposes in which measurements are made over water,
as has been recommended by A. Baumann (1890) and others.
The apparatus is shown in Fig. 40. A is a generating vessel for
holding the hypobromite solution. This is prepared as follows:—IOO g.
sodium hydroxide are dissolved in water and diluted to 1.25 litres, 25 c.c.
bromine are added, with external cooling by water, the whole well
shaken, and finally cooled. The solution is kept in the dark in a
well-stoppered bottle. The glass cylinder a, of about 20 c.c. capacity,
is fused on to the bottom of A, and serves to hold the solution of the
ammonium salt. The large glass vessel B, which has a capacity of
about 4 litres, is for cooling the generating vessel to its original
temperature, after the decomposition. The neck of the generating
vessel is rough-ground inside, so as to give a good hold for the rubber
* Z, anal. Chem., 1870, 9, 225; 1875, I4, 247.

THE AZOTOMETER 127
stopper, which must be pressed in tightly, and must not slip. C is a
tall cylinder filled with water (to which a little hydrochloric acid is added
to prevent the growth of moulds), and closed by a cork in which the
communicating burettes, c and d, are fixed, and carrying a small ther-
mometer. The burettes and their supply vessel h are filled with water,
which may be slightly coloured.
By following the subjoined instructions of P. Wagner' carefully,
ammoniacal nitrogen can be very accurately determined with this
instrument. IO c.c. of the ammonium salt solution under examination
are introduced into the inner vessel a, and 50 c.c. of hypobromite
solution into the outer vessel A, by means of a funnel. The rubber
stopper is then tightly inserted and the vessel placed in B, which is
filled with water. The glass tap f is loosened slightly, water intro-
duced into the burettes c and d by squeezing the rubber ball i and
opening the tap g, and then run off through g down to the zero mark
of the burette. After about ten minutes the cock f, which is slightly
greased, is inserted tightly, but left open so that A and c are in com-
munication. If, after an interval of five minutes, the liquid in c has
risen, the cock f is again loosened, re-inserted tightly, and the apparatus
again allowed to stand for five minutes. If the liquid has remained at
the zero mark, the temperature of the generating vessel and contents
has become equal to that of the surrounding water, and no more carbon
dioxide, from the enclosed air, is being absorbed by the hypobromite
solution.
To carry out the decomposition, 30 to 40 c.c. of water are first
withdrawn from the burette through g, and the generating vessel
removed from B and tilted, so that a small portion of the contents
of a flow into the hypobromite solution, with which they are mixed
by gentle shaking. This is repeated until the greater portion of the
ammoniacal liquid has been poured out and decomposed. The tap f
is then closed, the generating vessel thoroughly shaken, f again opened
to allow any liberated nitrogen to pass to the burette, and this repeated
until the liquid in c remains at a constant level; shaking three times
is usually sufficient. The generating vessel A is then replaced in B;
fifteen to twenty minutes suffice for this vessel and its contents to
acquire the temperature of the surrounding water and for the gas in c to
reach that of the water-jacket C, which is indicated by the thermometer.
The tap g is then opened until the water is at the same level in c and d.
and the volume of liberated nitrogen, the temperature of the water in
C, and the barometric pressure noted. The weight of nitrogen may
then be calculated from the following tables prepared by Dietrich (pp.
I28-130).
* Zehrbuch der Düngerfabrikation, 1877, p. 16I.
128 GENERAL METHODS USED IN TECHNICAL ANALYSIS
I.—Dietrich's Table for the Absorption of Nitrogen.
In 60 c.c. generating liquid (50 c.c. ſaypobromite solution and IO c.c. water),
the hypobromite solution having the sp. gr. I. I, and of such strength
that 50 c.c. correspond to 200 mg. M when from I to IOO c.c. of gas
are evolved.

Liberated . . . 1 2 3 4 5 6 7 8 9 10
Absorbed . . . 0-06 || 0-08 || 0-11 || 0°13 || 0°16 || 0°18 || 0°21 || 0°23 || 0°26 || 0°28
Liberated . . | 11 12 || 13 14 15 | 16 || 17 | 18 || 19 20
Absorbed . . . 0-31 || 0 33 || 0°36 || 0 °38 || 0°41 || 0°43 || 0°46 || 0 °48 || 0°51 || 0°53
Liberated . . . 21 | 22 || 23 || 24 || 25 | 26 27 | 28 29 || 30
Absorbed . . . 0-56 || 0:58 || 0-61 || 0-63 || 0-66 || 0-68 || 0-71 || 0-73 || 0-76 || 0-78
Liberated . . 31 || 32 || 33 || 34 || 35 | 36 || 37 38 || 39 40
Absorbed . . . 0-81 || 0°83 || 0-86 || 0 °88 || 0-91 || 0 °93 || 0°96 || 0 '98 || 1:01 || 1:03
Liberated . . . 41 || 42 43 || 44 45 46 || 47 || 48 || 49 50
Absorbed . . 1-06 || 1-08 || 1-11 || 1-13 || 1 °16 || 1 °18 1°21 1°23 || 1 °26 || 1 °28
Liberated . . || 51 || 52 || 53 || 54 || 55 | 56 57 58 59 || 60
Absorbed . . 1-31 || 1:33 || 1:36 || 1:38 || 1:41 || 1:43 || 1:46 || 1:48 || 1:51 || 1:53
Liberated . . . 61 | 62 | 68 64 65 | 66 | 67 | 68 69 || 70
Absorbed . . | 1.56 | 1.58 | 1.61 | 1.63 | 1.66 | 1.68 || 1-71 1-73 || 1:76 | 1.78
Liberated . . . 71 72 73 74 75 76 77 78 79 80
Absorbed . . 1-81 | 1.83 1-86 1-88 1°91 || 1 '93 || 1 '96 || 1 '98 || 2°01 || 2:03
Liberated . . . 81 82 83 84 85 86 87 88 89 90
Absorbed . . . 2°06 || 2:08 2-11 || 2°13 2°16 2°18 || 2*21 || 2°23 || 2°26 || 2:28
Liberated . . . 91 92 93 94 95 96 97 98 99 || 100
Absorbed . . 2-31 || 2:33 2-36 || 2:38 || 2°41 2°43 || 2°46 || 2°48 2°51 2-53
§
#8T60. I Igg30. I 83980. I g3380. I 33610. I | 61910. I 918/10. I | 81010. I |OI/90. I | 1,0790. I | 70I90. I | I0890. I | 66790. I 9&
S0/60. I | 70760. I 00I60. I 96.180. I 86780. I | 68L80. I 988/0. I I8910. I 11310. I | 81690. I 69990. I |99890. I I9090. I jº,
91z01.1 11660.1 99960.1 | 19860.1 |92060.1 19180.1 |9%0.1|IV180.1|98820-1||18q10.1|96610. I [I3690. I|91990.1 33,
6880I.. I | 8890I.. I | ZZZOI. I | I&660. I | ?I960. I 80860. I 60060. I 96980. I | 06880. I £8080. I | 84/10. I | 3/720. I | 991/0. I 33
863TT.I 980 II. I 6//0ſ. I 31.70ſ. I 99TOI. I | 19860. I 0.9960. I 87360. I 98680. I 63980. I | 33380. I |g|IO80. I | 80/10. I T&
376II. I g39II. I | 138T.I.. I 8IOII. I | 0I/0I.. I 30701. I | 7600I.. I 98/60. I 8/760. I | 02I60. I 39880. I 79.980. I 97380. I 03
#8%I. I g/IZI. I | 998II. I | 199T.I.. I 8%II. I | 8860L. I | 6690I.. I | 0330L. I | II00I.. I 30260. I | 36860. I |88060. I 71280. I 6I
gz08I. I gLºgI. I g0%I. I g60&I. I g8!II. I g/7II. I | 99 III. I | ?980I.. I | 7790I.. I | #830I.. I | 73660. I | WI960. I | 70860. I 8T
Z998.I.I Igz21. I | 076ZI.T | 669&T.I | 11831. I 900&I.I. g69II. I W88TI-I 810T.I.. I | 19101.1|0g/01-1 |68101.1 | 86860. I AI
géOWI. I Z813I. I 01781. I | 8918T.I g783.I.I 889&I. I | 03631. I |806T.I.. I |969 II. I | 88&II. I I/60I.. I | 8990T.I | 9780T.I 9T
0.29%I. I | 90951. I g668I. I | 0893.T. I | 9988.I.. I | 8908I. I | 68/3.I.I 93%I. I | 8TI&I. I | 661 II. I 987 II. I | 3/III. I | 6980I.. I g|I
g?IgT. I | 0985 I. I gT9?I. I | TO&#I.T |9888I. I 3/98.I.. I | 1938 I. I &#6&I. I | 839&I. I 3T3&I. I | 666II. I | #891.I.. I 698 II. I #I
899q1..I 8v891. I | 680gſ. I 911WI. I | TOWI. I |980WI. I |6918T.I., WGV81. I |88181.1 |&686.I.I 909&I. I 161&T.I glSII. I 3I
11191.1 | 098q1.1 | 87991. I | 1zzgr. 1 |016VT. Iſ 869WI.I.'ſ 11&WI. I |09681. I|87981. I|9&881. I | 01081. I | 869&I.I |918&I.I &T
9699I. I | 8/39T. I | 09091. I | 37/9I. I | 7379[.. I | 1019 [.. I | 68/7][. I | I/77T. I | 8917ſ. I g888I. I | | Iggl. I | 6618T. I I88&I. I II
802.I.I | 6889T. I | 0199T. I | Igó9I. I | 6869 I. I | 8E991. I | 7639T. I g167T. I 99.97I. I | 188?I. I | 8IO7I. I | 6698.I.I 08881.1 OI
## 1. Žiž/ Of Z 89/. 98/. #8/. 38/. 08/. 83/ 93/. fº/ 36/, 06/, 'apelaguao
sooiáop
‘Sanauſſ|[I]N up ‘dual

J Sz of OI 240.4/
saungpaaduta Z pup 42n.2/277 ſo satzautº/ſ4/7 O44 of Oz.4 attoºſ sounssaad Zv ‘sutolºſſeſſ ºf
‘ueño.141N Jo O'o I go quare/YA au: Joy equ L S.uol.1421CI—'II

8.108T. I | 01/&I. I | 197&I. I 791&I. I I98II. I | 899 II. I ggðII. I Z960I.. I 6790I.. I | 97801. I 8700T. I | 07/60. I | 19760. I 92,
6998I. I | 9988[.. I Ig()8][. I ºf 13 I. I gp%I. I | 68 IZI. I g38II. I Z39T.I.. I 83%II. I | 7360I.. I | 0390 I. I 9I80I.. I | 3IOOI. I #2
I?67I. I 9868L. I I898T. I | 9388 I. I IZO8T. I 9I/3I. I II;2|[.. I | 90IZI. I IO8II. I 967 II. I I6III. I 9880T. I ISgOI. I 33,
81871. I &I971. I 96371. I | 0068L. I 7698.I.I |8838L. I | 386&I. I g193T.I 69831. I | 890&I. I | 1.9/II. I | Ig?II. I g?III. I 3%
9839T. I 8109T. I | IAPI. I #957 I. I | 191?I. I | 0983.I.. I | 8%gg.T. I | 9333 I. I 836&I. I I39%I. I | ?[8&I. I | 100&I. I |00/II. I I6
8769 I. I | 0%991. I Z8891. I | 720g L. I 9II?I. I 807?I. I | 6668I. I | I613I. I | 8878.I.. I g/Ig|I. I | 1983.I.. I | 699&I. I | I93&I. I 0%
f099T. I | 96.19ſ. I 98.89T. I 91.99T. I | 1939T. I | 896?I. I | 679%I. I | 078?I. I | 03071. I I3/8L. I ZIP3I. I 30IgT. I | #61&I. I 6 I
990/I. I 97/9I. I | 93.79[.. I | 93I9I. I 9I89T. I | 9099T. I 96Ig|I. I |988?I. I 919?I. I 9967T. I 9963.I.. I g?98I. I |9888I. I 8I
8091.I.. I | 163/I. I | g869T. I | #199 I. I g989T. I | 3909T. I | Iſ/gL. I | 6′379T. I | 8IIgſ. I | 1087 I. I | 96771.I |98IWI. I | 31881. I II
99T8I. I | 7781.T. I | [89/.T. I | 6IZZI. I | 9069T. I | #699 I. I 3839T. I | 6969 I. I | 1999T. I jºgg|I. I Z809T. I | 0317T. I | 1.07% I. I 9I
#698I. I | [88SI. I 1908I. I | #9/AI. I | Off/I. I | 13 IZI. I | FI89T. I |0099 I. I | 1819I. I | 8489T. I | 0999.I.I 1739T. I 38671. I 9I
78.36T. I | 0368I. I 90981. I | I638I. I | 916/I.T | I991.I.. I | 1781.T. I | 380/I. I | 8T191. I |8079[.. I | 8809T. I | 71191. I | 6979 I. I #I
891.6L. I 39,761.. I 98.I6T. I | 0388T. I g098I. I | 68I8I. I | 8/8/I. I | 899/.T. I 373/.I.I 936.9L. I II99T. I g639T. I 61.69 I.I. 8T
f630%. I 1166 I. I 09961. I | #736T. I | 1306 I. I OI/8L. I | #688I. I | 1.108T. I | 09//I. I | ???! I. I | 1.3L/I. I OI89T. I 3679 I. I ZI
63803. I IIg()3. I | 36|I03. I | g 1861. I | 1996I. I | 6836T. I IZ68I. I 3098I. I |9838L. I | 896/I. I | 0991.I.. I | 688/I. I | #101.I.I II
998.I.Z. I 98013. I II/03. I | 86803. I 61003. I | 09/6T. I | If?6I. I | 3&I6I. I 8088T, I | V878.I.I 99.18L. I | 978/I.T | 139/I. I OI
0// 89/. 99/ 79/. 391. 09/. 89/. 99/. jºgſ, 39/. 09/ 87/. 971. 'ape.ISguao
saa13ap
“sa.I]auIIIIHW uţ ‘dua L

"penunuoo—uašo.4?AV ſo 22 I ſo 1/5:2AM 2/1 40/2/421 sº.42*CT


THE AZOTOMETER 131
Example of calculation :
Volume of liberated nitrogen . º 22 C.C.
Temperature . º o tº & I6°
Barometric pressure e g º 756 mm.
Then, according to Table I., O. 58 c.c. nitrogen remain dissolved in
the generating liquid, and according to Table II., the weight of I c.c.
of nitrogen at 16° and 756 mm. pressure is I. I 5969 mg. The weight
of nitrogen in the substance examined is therefore :—
(22.0 + o-58) × I-15969 = 26.19 mg.
These tables are given as they are in very general use; they do not,
however, possess any special advantages, nor are they based on the
correct value for the weight of I c.c. of nitrogen (cf. infra).
Lunge" has shown that it is unnecessary to use Table I. to
correct for the so-called “absorption of nitrogen,” which is really
Émit
|||||||||||||||
º
due to incomplete reaction, and that it suffices to add 2.5 per cent, to
the volume of gas found. This calculation may be simplified by using
* Chem. Ind., 1885, 5, 165.

132 GENERAL METHODS USED IN TECHNICAL ANALYSIS
the increased factor O.OOI 558 to convert each c.c. of nitrogen at o' C.
and 760 mm. (dry) into grams of ammonia. This factor is calculated
from the observed density of nitrogen.
A. Baumann recommends the simplified form of Wagner's Azoto-
meter shown in Fig. 4I, for technical laboratories, in which the
measuring tube is replaced by an ordinary pinch-cock burette; the
manipulation of the apparatus is in all respects similar to that described
above. If many analysis have to be performed it is well to keep two
or three such simplified azotometers in use, so as to save time whilst
the original temperature is being attained, after the decompositions.
It is particularly important, when using the azotometer, that the
temperature in the generating vessel and measuring tube should remain
constant during the whole experiment. If the capacity of the former
is only 150 c.c. a temperature variation of 1° C. causes an error of O-5
c.c., and one of 2°C. of about I c.c. Such errors influence the result
of the analysis considerably, especially when small quantities of gas
are liberated, amounting, for instance, to 5 to IO per cent. Of the total
volume when IO c.c. of gas are evolved.
The Nitrometer.
The name “Nitrometer” was given by Lunge" to the apparatus
he devised for the determination of nitrogen acids, by decomposition
with sulphuric acid, in presence of mercury. This reaction was made
use of long ago by W. Crum ” for the analysis of nitrates and of
gun-cotton, and subsequently, in an improved form, by Frankland
and Armstrong” for the determination of nitrates in water, and by
G. E. Davis + for the valuation of nitrous vitriol, but manipulative
difficulties restricted its applicability; the nitrometer has rendered the
reaction thoroughly practicable, and has thus been the means of
bringing it into general use.
In the method as originally devised, the mercury over which the
gas is collected also takes part in the reaction. It has since been
extended to many other purposes, in which the mercury does not take
part in the decomposition, but serves simply as the confining liquid for
the liberated gas, and also to determinations in which the decomposi-
tion is effected in a separate vessel and the evolved nitric oxide
collected and measured, over dry mercury, in the nitrometer. Various
alterations have thus been made in the original apparatus from time to
* Ber., 1878, 11, 174; 1885, 18, 1878 and 2030 (New applications); 1888, 21, 376. Ding/.
polyt.J., 1879, 231, 522 (Tables); 1882,245, 171 (Analysis of dynamite). Chem. Ina, 1881, 4, 346;
1885, 8, 161 (New applications); 1886, 9, 273 (Analysis of explosives). J. Soc. Chem. Ind., 1882,
1, 15; 1885, 4, 495; 1886, 5, 82; 1888, 7, 232 ; 1892, II, 778. Sulphuric Acia and Alkali,
2nd edition, 1891, p. 182. Cf. also A. H. Allen, J. Soc. Chem. Ind., 1885, 4, 178.
* Phil. Mag., 1847, [3], 30, 426. * Chem. Soc., 1868, 21, IoI.
* Chem. News, 1878, 37, 45. -
THE NITROMETER * 133
time; the literature is given in the above references. The “gas-
volumeter,” which was developed by Lunge from the nitrometer, will be
described subsequently.
The very extended applicability of the nitrometer is mainly due to
the following reasons. In the first place, most gases, even those
which are soluble to a considerable extent in water, can, by its means,
be measured over mercury without the use of a mercury trough, which
previously was always necessary for this purpose, and the apparatus is
simple, requires but little mercury, and is easy to manipulate and to
shake; secondly, it is equally suitable, either for those cases in which a
gas is evolved and measured in the purc state, or for those in which it
is liberated in a separate vessel and the displaced air measured. It can
also be used for ordinary gas analysis, and for determinations in which
a gas, generated from a solid or liquid substance, has to be separated
from other gases, as in the determination of carbon dioxide by Lunge
and Marchlewski's method, etc.
Böckmann,” in describing the nitrometer, says:—
“Lunge's nitrometer, in its various modifications, is an exceptionally
valuable apparatus, and is therefore very largely used for technical
work. There is scarcely another ap-
paratus for technical analysis which has
so many practical applications and is
at the same time so easily handled.
Those of its applications which have
been published are numerous ; those
which are unpublished, and which are
carried on in almost every technical
laboratory, are certainly still more
numerous.”
The nitrometer is shown in Fig. 42
in its original form, as still used in
most cases for the analysis of nitrous
vitriol in the sulphuric acid industry,
and for many other purposes. The
tube a has a capacity of 50 c.c.; it is
drawn out at the bottom and is gradu-
ated in ſº C.C. The graduations start
immediately below the three-way tap
which terminates the tube at the top. This tap may either have a
vertical and an axial passage, as in Winkler's or Bunté's gas-burettes,
or it may be a Friedrich-Greiner tap with two oblique passages,
as shown in its three positions in Fig. 43. The latter form closes more
tightly and is more easily manipulated than the former, and is there-
- * Third edition, I, 63-64.
FIG. 42.

134 GENERAL METHODS USED IN TECHNICAL ANALYSIS
fore usually attached to the newer forms of the apparatus. Above the
tap, a beaker and a side tube d are attached ; the latter replaces the
axial passage of the older form of tap. In position A the measuring
tube communicates with the side tube d, in position B with the beaker,
and in position C the tap is closed.
The measuring tube a (Fig. 42) is connected by means of thick-walled
rubber tubing with the pressure tube b. The latter is a simple cylindrical
glass tube of the same diameter as the measuring tube, and drawn out at
the bottom for the attachment of the rubber tubing. Both tubes are
held by clamps, in which they can be moved.
To use the apparatus, for instance for the assay of nitrous vitriol,
the tube b is placed so that its lower end is somewhat higher than the
tap on a, and mercury is poured in through 5, the tap on a being
open, until it enters the beaker on a ; as the mercury enters a from
below, no air-bubbles are formed on the sides of the tube. The tap
§lºi is
is then closed, the mercury in the beaker allowed to flow out through
the side passage of the tap, the pressure tube lowered, and the tap
closed.
The nitrous vitriol is then introduced into the beaker from a I c.c.
pipette divided into Tâm c.c.; in the case of very strong nitrous vitriol
O. 5 c.c. are used, in other cases 2 to 5 c.c. The pressure tube 6 is then
lowered, the tap carefully opened, and the acid drawn into the
measuring tube, care being taken not to admit any air. The beaker
is then rinsed with 2 to 3 c.c. of sulphuric acid, free from nitrogen
acids, which is similarly drawn into the measuring tube, and the
washing repeated with I to 2 c.c. more of sulphuric acid. The decom-
position is then started by removing the tube a from its clamp, and
thoroughly mixing the acid and mercury by repeatedly holding the
tube almost horizontally, taking care that no acid gets into the
rubber tubing, and then sharply raising it to a vertical position. The
tube is shaken for one to two minutes until no more gas is evolved.
The two tubes are then placed so that the mercury in b is as

THE NITROMETER 135
much higher than that in a as is necessary to compensate for the layer
of acid in the latter; I mm. mercury is allowed for every 6% mm. of
acid. After the temperature has become equalised the pressure is
exactly adjusted by pouring a little acid into the beaker and opening
the tap cautiously. If the gas is under diminished pressure, which is
preferably the case, acid will flow into a ; the tap is at once closed
before air can enter, and the operation repeated after raising the
pressure tube very slightly. If, on the contrary, the enclosed gas tends
to force its way through the acid, the tap is closed, the pressure tube
lowered slightly, and the tap again opened. With a little care these
manipulations can always be successfully carried out. The volume of
the gas, the barometric pressure, and the temperature are then read;
the latter by means of a thermometer, the bulb of which is placed
close to the measuring tube and near the middle of the column of
gas.
When the determination is finished the measuring tube a is lowered,
so that no air may enter on opening the tap, and the gas expelled
by raising the pressure tube. The acid is expelled, either through the
side tube d (Fig. 43), or through the axial passage in the older form
of tap, and the last traces removed with filter paper. The nitrometer
is then ready for the next determination.
It is always necessary to make sure that the tap of the nitro-
meter fits tightly, and this is best secured by greasing it with a little
vaseline; care must be taken that no vaseline gets inside the tap and
So comes into contact with the acid, as this causes the formation of
froth, which only settles very slowly.
A glass tap can hardly be expected to keep perfectly closed, for
any length of time, when exposed to considerable variations of pressure;
the tap can, however, be regarded as satisfactory if no air enters the
measuring tube during an interval of two hours after it has been com-
pletely filled with mercury and the pressure tube lowered. Details with
regard to the calculation of the amount of nitrogen compounds from
the volume of gas obtained are given in the section “Sulphuric Acid.”
The analysis of nitrates or of nitrites soluble in water, to determine
the total nitrogen, is carried out similarly, as is described in connection
with the analysis of sodium nitrate. In such cases, where a solid,
soluble in water, is to be analysed, the weighed substance is introduced
into the beaker, dissolved there in a very small quantity of water, the
solution drawn into the measuring tube, the beaker rinsed with con-
centrated sulphuric acid, and the decomposition carried out as described.
In the analysis of substances insoluble in water but soluble in
strong sulphuric acid, more especially dynamites and pyroxylins, for
which purpose the nitrometer is now almost always used, the solution
in sulphuric acid is also carried out in the attached beaker. In this
136 GENERAL METHODS USED IN TECHNICAL ANALYSIS
case, in order to avoid loss of nitrous fumes, the device shown in
Fig. 44, due to Lunge," is used. The beaker of the nitrometer is closed
by a rubber stopper provided with an S-tube, end-
ing above in a small funnel. The substance is
placed in the beaker, and concentrated sulphuric
introduced through the funnel. The lower bend
of the tube naturally remains full of sulphuric acid,
which prevents the escape of nitrous fumes, and
which flows into the beaker, as the acid in the
latter is drawn into the measuring tube.
It is immaterial whether an insoluble powder,
such as kieselguhr in the case of dynamite, or some
undissolved saltpetre, etc., remains, as this is sucked
into the measuring tube with the liquid; in the
analysis of pyroxylin it is,
however, better to wait till
it has completely dissolved
in the beaker.
Nitrates and esters of nitric acid, such as
nitroglycerine and nitrocellulose, can be analysed
in the apparatus shown in Fig. 42, but a high
degree of accuracy is not obtainable, as not
more than 40 c.c. of gas can be measured. By
using a modified form, the “nitrometer for salt-
petre,” shown in Fig. 45, an accuracy of O. I per
cent., which is not surpassed by that of any
other method, may be obtained in these deter-
minations. In this form a larger space for gas
is provided without making the apparatus in-
conveniently long, by means of a bulb of nearly
IOO c.c. capacity, below which the graduations
extend from IOO c.c. to 130 c.c.
In Fig. 46 a form of nitrometer is shown
which may be used for determinations in which
either a small or a large volume of gas is
evolved. As it cannot be made so short as
the forms shown in Figs. 42 and 45, it is not
so suitable for shaking, but it is well adapted
FIG. 44.
400
A
#:
j
rt
*
wº
sº
*
É=s
# = 3
-:
-:
.-:
Hºmº
... =:
-:
**
*-
4-
='
sº-sº
sº
*
-->
isºm
.*
430
!.
t
FIG. 45.
*
FIG. 46.
for use in conjunction with a decomposition bottle or as a gas-volumeter
(cf. infra).
When nitrous and nitric acid compounds are decomposed by shaking
in the measuring tube itself with mercury and sulphuric acid, a layer of
sulphuric acid finally remains between the gas and the mercury, which
* Chem. Ind., 1886, 9, 274.





THE NITROMETER - 137
has to be allowed for in adjusting the pressure for the final measure-
ment, as stated above. This causes no special difficulty as a rule, but
when considerable quantities of substance are decomposed, much
frothing sometimes occurs; on dilution with water, which is, in some
cases, unavoidable, Sulphate of mercury is precipitated, and in the
analysis of dynamite the kieselguhr floats on the top of the acid, etc.
These factors make the final adjustment and measurement uncertain
and inexact. In such cases, for instance in the analysis of saltpetre
and of explosives, it is therefore advisable to use a separate decomposition
vessel, with an attached pressure tube, as in the gas-volumeter. The
gas is then always measured over a sharp meniscus, and the volume
can be read with the greatest accuracy.
The same considerations apply to all those methods of analysis in
which the nitrometer is used in conjunction with an attached decom-
position bottle, as shown in Fig. 47. The manipulation is then
identical with that of the azotometer. The substance to be decomposed
is placed in the outer annular space, and the decomposing agent in
the inner vessel, which is fused on to the bottom of the bottle.
After replacing the stopper, the bottle is connected with the tap of the
nitrometer, the measuring tube of which has been completely filled with
mercury; the stopper is again loosened, to make sure that there is no
excess of pressure in the generating bottle, and the mercury in the
nitrometer re-adjusted to zero, if necessary, by moving the pressure
tube. The tap of the nitrometer is then turned so as to connect the
measuring tube and the generating bottle, and the latter tilted, so that
the liquid in the inner vessel flows into the surrounding space. In
doing this care must be taken that the bottle is not warmed by the
hand, and this applies also to the subsequent shaking, to promote the
disengagement of the gas; it is safest, especially if heat is evolved, as
in decompositions by the sodium hypobromite method, to place the
generating bottle up to the neck in a beaker of water, before and after the

138 GENERAL METHODS USED IN TECHNICAL ANALYSIS
reaction. As the mercury in the measuring tube is depressed by the
evolved gas, the pressure tube is löwered accordingly, to prevent any
excessive pressure; it is sometimes advantageous to lower the pressure
tube considerably towards the end of the reaction, in order to facilitate
the removal of the gas. After the original temperature has been attained
the mercury is brought to the same level in both tubes, and the volume
of the gas read, together with the temperature and pressure, as
described on p. I 35.
In reducing to 760 mm. pressure, it must be borne in mind that the
gas is moist, and not dry, as is usually the case in working with a
nitrometer; since it is always generated from dilute solutions, the
tension of the water-vapour may be considered equal to that of pure
water at the same temperature.
The reduction of the volume of gas to normal volume is calculated
most easily by means of the Tables VI. to VIII., appended to this volume,
which were specially compiled for use with the nitrometer by Lunge;
the calculation can also be done by reversing the formulae on p. 140.
The numerous applications of this form of the nitrometer include
the valuation of potassium permanganate solution (p. 109), the
analysis of bleaching powder, manganese dioxide, potassium ferri-
cyanide, lead peroxide, nitrous acid, hydrogen peroxide, the deter-
mination of carbon dioxide in carbonates (preferably by Lunge and
Marchlewski's method), of nitrogen in ammonium salts, in urea (ureo-
meter), and in diazo-compounds, the control of the strength of acids
by liberation of carbon dioxide from carbonates, the valuation of zinc-
dust, the examination of chromates” and the titration of iodine
solutions” (cf. p. I I4).
The nitrometer can also be used as an absorptiometer,” as an
apparatus for reducing the volume of gases to O’ and 760 mm. pressure,”
although the gas-volumeter is far preferable for this purpose, and for
most gas-analytical work;" also for the collection and analysis of
gases dissolved in water, etc."
The Gas-Volumeter.”
This name was given by Lunge to an apparatus, by means
of which the reduction of a volume of gas, either wet or dry, to any
required normal conditions, usually, of course, to O’ and 760 mm.
pressure, is effected without observation of the temperature and
* Riegler, Z. anal. Chem., 1897, 36, 665.
* A. Baumann, Z. angew. Chem., 1891, 4, 135, 198, 339, 392.
* Z, angew. Chem., 1891, 4, 203, 328, 450. * Ibid., 1885, 8, 163.
* Chem. Ind., 1885, 8, 162. * Ibid., 1885, 8, 169.
7 Zbid., 1885, 8, 17o; Z. anal. Chem., 1886, 25, 309.
* Lunge, Ber., 1890, 23, 440; 1892, 25, 3157. Z. angew. Chem., 1890, 3, 139; 1891, 4, 197,
4Io; I892, 5, 677. / Soc. Chem. Ind., 1890, 9, 547.
THE GAS-VOLUMETER 139
pressure, and without calculation or reference to tables. This
operation, which is necessary in all gas-volumetric work, can be per-
formed by means of this instrument, in less than a minute, in a purely
mechanical manner, and is applicable, not merely for relative measure-
ments, but also for absolute measurements, such as are required
in the analysis of solid and liquid substances, by gas-volumetric
methods. -
The principle of the apparatus is to enclose a known volume of air,
at such a pressure that it takes up exactly the volume which it would
occupy at O’ and 760 mm. pressure. If
the same pressure and temperature are
then applied to another volume of gas,
this will also take up the volume which
it would occupy at O’ and 760 mm.
pressure. This condition is attained
by confining the known volume of air
in a “reduction tube” to which a pres-
sure tube is attached, and placing the
latter in such a position that the gas
in the reduction tube is brought to the
volume it would occupy under normal
conditions; the reduction tube is also
connected with the gas measuring tube
by means of a T-piece, and by adjusting
the level of the mercury in the two
tubes to the same height, the volume §.
correction is applicable directly to the §
gas in the measuring tube.
The apparatus is shown in Fig. 48.
The three tubes, A, B, and C, are con- :
nected by means of a T-piece and suffi- . ; 3.
ciently long thick rubber tubing, and : *
are held to a stand by clamps, in which ; W :
they can be moved in a vertical direc- ! \ .
tion. The tube A is the measuring } w
tube, and may be a nitrometer of any sº
form, a Bunté burette, or any other | ?
apparatus for the measurement of gases. *A lº
The second tube, B, the so-called reduc-
tion tube, is enlarged to a bulb at the
top; the first graduation is below the widened portion of the tube, and
indicates a volume of IOO c.c., and the lower cylindrical portion of the
tube is graduated in To c.c. for another 30 to 40 c.c. The tube is set,
once for all, by observing the temperature and pressure, calculating the
FIG. 48.


140 GENERAL METHODS USED IN TECHNICAL ANALYSIS
volume which IOO c.c. of dry air would occupy under the observed
conditions, bringing the mercury to the corresponding division, and
+7 −7 closing the tap above the bulb. If the tap is air-tight
the reduction to O’ and 760 mm. pressure has been
| permanently made; an alternative is to provide
| |. the upper end of the tube with a capillary, which
|
|

<-- - - -> . can be sealed off, after the volume has been correctly
adjusted.
This form of reduction tube was subsequently
modified by Lunge and replaced by the cylindrical
| form shown in Fig. 52 (p. 144), in order to bring the
volumes of enclosed gas as far as possible into a
| parallel position. Both the simple tap shown in Fig.
º / 48 and a sealed capillary have certain disadvantages,
| which are avoided by the use of the tap devised by
| Jea. Göckel." This is a ridged horizontal tap, sealed by
tº mercury; Lunge states that it is in every way Satis-
|
|
|
§
CN
+.
A T
factory and generally applicable.
§ The pressure tube, C, is now made in the form
§ shown in Fig. 49, which saves a great deal of mercury.
If moist gases are to be measured in the measuring
tube, a small drop of water must be introduced into
the reduction tube; if dry gases are to be measured,
e.g., nitric oxide generated over sulphuric acid in the
J-—- ordinary nitrometer, a drop of concentrated sulphuric
Ü acid is introduced into the reduction tube, but never
in sufficient amount to more than cover the top of
the mercury meniscus. The gases must be measured either quite dry
or saturated with moisture. *
|
|
|
|
ſ
|
|
FIG 49.
Formula for dry gases VI = Vo *
e * (273+ £) 760
Formula for moist gases VI = Vo 273(B —f)
Where VI = Volume of gas required.
Vo = Normal volume of gas.
# = Observed temperature.
B = Observed barometric pressure.
f = Tension of water-vapour at observed temperature.
To adjust the reduction tube, the temperature is observed by means
of a thermometer suspended by the side of the tube, and the barometric
pressure read. The volume V, which IOO c.c. of gas at normal tempera-
ture and pressure V, will occupy at the observed temperature and
pressure, either dry or moist, is then calculated, either by means of the
* Z. angew. Chem., 1900, 13, 961 and 1238. This tap is made by Alt, Eberhard, & Jaeger,
Ilmenau, Thuringia.
THE GAS-VOLUMETER 141
above formulae or from the Tables VI. to VIII. appended to this volume.
The pressure tube is then adjusted, so that the mercury in the reduction
tube is at the calculated graduation, and the tap closed. The instru-
ment is then ready for use.
Dry gases can, if desired, be measured with a wet reduction tube, or
vice versa. In the former case the temperature is observed, the corre-
sponding vapour-tension f in mm. found from Table VIII., and the
mercury in the measuring tube placed f mm. higher than in the
reduction tube, after the volume in the latter has been adjusted to IOO
c.c. This is facilitated by using measuring tubes in which each c.c. of
the graduation corresponds almost exactly with IO mm. length, as the use
of a measuring scale is thereby avoided. When a dry reduction tube
is used for moist gases, the mercury in the measuring tube must be
placed f mm. lower than in the reduction tube, in which the volume
has been adjusted to IOO c.c.
An alternative plan for comparing dry gases with a moist reduction
tube is to moisten the gas in the measuring tube by drawing in a drop
of water; this is preferably done before the gas is collected, but it can
also be done afterwards with careful manipulation ; for the reverse
comparison a drop of concentrated sulphuric acid is used. In neither
case must so much liquid be added that it more than covers the top of
the mercury meniscus.
The three tubes, A, B, C (Fig. 48), are connected with the T-piece
D by very thick-walled rubber tubing of 13.5 mm. external and 4.5
mm. internal diameter. Thick tubing of this character withstands the
pressure of the mercury without being distended and does not need bind-
ing with wire, especially if the ends of the tubes are enlarged a little; it is
easily drawn over tubes of IO mm. or greater diameter. All three tubes
are held in strong clamps, so that they may be moved vertically with
friction, but do not drop down by their own weight. If desired, A and
B may be fitted with water-jackets, in which case the upper part of
B must be cylindrical (as in B, Fig. 52, p. 144), and larger clamps must
be used. This addition is quite unnecessary for technical analysis, as
the two tubes are placed close together and will only be at different
temperatures in case of careless manipulation; also the fairly large mass
of mercury greatly assists the equalisation of all temperature differences.
When any gas-analytical or gas-volumetric operation has been
carried out in A, the volume of gas is not read in the Ordinary manner,
after adjusting the mercury in A and C to equal levels. This adjust-
ment is only made as a preliminary step when an auxiliary generating
bottle is used, as in decompositions by sodium hypobromite or hydrogen
peroxide, or in carbon dioxide determinations, etc.; in these cases the
mercury levels in A and C must first be equalised, in order to bring the
gas in A to the ordinary atmospheric pressure. The tap of A is then
142 GENERAL METHODS USED IN TECHNICAL ANALYSIS
closed, without reading off the volume of contained gas. If the gas has
been generated in A itself, or has been transferred from another vessel, this
adjustment is of course unnecessary. Before actually reading the
volume of gas, in A, the three tubes are so adjusted that the mercury in B
stands at the IOO c.c. graduation and the levels in A and B are the same.
The gas volumes in A and B then correspond to such temperature and
pressure that their volumes are equal to those which they would occupy,
when dry, at O’ and 760 mm. pressure, since this condition has been
ensured once for all in the case of B, and the gas in A is at the same
temperature and the same pressure.
The required adjustment is most quickly and easily carried out in
the following manner. The tube A is fixed in its clamp whilst B and C
are raised, C to such an extent that the mercury in B rises to the
graduation IOO C.C.; B and C are then lowered simultaneously, so that
the difference of level of the mercury in the two tubes is maintained,
until the mercury stands at equal levels in A and B, in the latter still at
the IOO c.c. graduation. The adjustment will usually be found to be not
quite exact, and a slight movement of B will probably be necessary to
complete it, but this double adjustment only requires a few seconds
longer than the ordinary adjustment of the pressure tube of a gas-
burette. The equalisation of pressure in A and B can be aided by the
usual means, such as making the observation against a horizontal wall-
line, a window-frame, a special horizontal rod provided with a spirit-
level," or in any other manner. -
The simultaneous movement of two tubes filled with mercury, as
described above, is rather difficult if strong Spring clamps are used, and
these, moreover, wear out in time.
This difficulty is overcome by the
use of the forked clamp shown in
Figs. 50 and 5 I.
This consists of two cork-lined
clamps supported by a cast-iron
fork. The smaller clamp a holds
the reduction tube (below the IOO
c.c. graduation), and the larger, b,
the pressure tube. The fork is
fastened to the stand by means of
an ordinary boss c, shown on a larger scale in Fig. 51, which may be
strengthened by a spring.” By this means the reduction and pressure
tubes are held in a common clamp, and can be moved together. After
the gas-analytical operation is finished, the adjustment is made by first
bringing the reduction and pressure tubes to the approximate level of
FIG. 50. FIG. 51.
* Cſ. Lunge, Æer, 89, 24, 3948.
* These clamps are made by Desaga, Heidelberg, and other firms.

THE GAS-VOLUMETER 143
the mercury in the measuring tube. The pressure tube is then moved
in its clamp b, until the mercury in the reduction tube stands exactly
at the IOO c.c. graduation. The forked clamp is then moved as a whole,
by means of the boss c, until the mercury levels in the reduction and
measuring tubes are equal. The whole adjustment only takes a few
seconds, and is much easier than when separately movable clamps are
used.
In cases where, in addition to the mercury, another liquid is intro-
duced into the measuring tube, the pressure which it exerts must be
taken into account. Thus, in nitrogen determinations by Dumas'
method, a mark is made below the IOO c.c. graduation of the
reduction tube, corresponding to ºr of the height of the potassium
hydroxide solution in the measuring tube, the specific gravity of which
may be taken as I-36, i.e., ſo of that of mercury. In reading off the
volume of nitrogen the mercury in the reduction tube is brought to the
IOO c.c. graduation, and that in the measuring tube to the level of the
special mark below, whereby the height of the potassium hydroxide solu-
tion is allowed for.
It is obvious, from the above, that by the use of the gas-volumeter all
thermometric and barometric readings, and also all reductions by calcu-
lation or by special tables, are avoided ; the volume of gas is read off
directly under conditions corresponding to the normal pressure and
temperature. As already stated (p. 140), the reduction tube must be
arranged for dry or moist gas according to the character of the gas-
analytical operation.
The measuring tube may be of any of the forms described on pp.
I 33 to I36. For evident reasons it is, however, not so simple and easy
to shake the mercury and sulphuric acid in the tube itself, when a reduc-
tion tube is used, and a special danger is that the gas may inadvertently
get into the reduction tube. The gas-volumeter is accordingly more
applicable for work in which a separate decomposition bottle is used, and
for nitrometric analyses, in which the gas is generated in an ungraduated
vessel, and then transferred to the volumeter for measurement.
This arrangement of apparatus is shown in Fig. 52. A, B, and C
represent the volumeter, as in Fig. 48, p. 139; E is the reaction vessel,
and has a capacity of about IOO c.c. when used for the analysis of
nitrous vitriol, and of about 200 c.c. for the analysis of saltpetre,
dynamite, nitrocellulose, etc.; it is provided with a tap, a beaker c, and
pressure tube F, arranged as in the ordinary nitrometer. E is best
supported by a ring and F by a clamp; E can, of course, be replaced by
a Hempel burette or other similar apparatus.
To carry out a determination, F is first raised until the mercury reaches
the end of the capillary a, which is then closed with a ground-glass or
rubber cap b, to prevent the escape of mercury on shaking, and the tap
144 GENERAL METHODS USED IN TECHNICAL ANALYSIS
of E also closed. The nitrous vitriol or other material is introduced, as
usual, through the beaker c, the reaction induced by shaking, and the
whole allowed to stand until it has acquired the atmospheric tempera-
ture, E and A are then brought to the same height, as shown in the
figure, and the mercury in the measuring tube of the volumeter A driven
over to the end of the thick rubber tube attached to the capillary d. The
cap 6 is then removed, the
capillary a slipped into the
rubber connecting tube
until glass touches glass,
F raised and C lowered, the
two connecting taps opened,
and the gas syphoned from
E into A. The tap on A is
closed as soon as the acid
WI from E reaches the bottom
| of the capillary tube e.
|C The levels in A and B are
then adjusted, and the mer-
cury in B brought to the
IOO c.c. graduation, as de-
scribed above. This method
of transferring the gas is
advantageous, not only
because the pressure due
to the sulphuric acid has
not to be allowed for, but
also because the unavoid-
able dirtying of the ap-
paratus is confined to the
reaction vessel E, which is
easily cleaned. Care should
be taken that the bores of
the capillaries a and d are
not enlarged at the ends,
but, if anything, slightly
narrowed, so that no bubbles of gas adhere on connecting up. A
single-bore tap suffices for the volumeter, and the attached beaker can
be dispensed with, as shown in Fig. 52, and this applies also to the use
of the apparatus in conjunction with an auxiliary generating bottle, or
with a gas-burette. The measuring tube A may be either cylindrical,
and of 50 or 100 c.c. capacity, or a bulb-tube graduated from IOO to
140 c.c., etc., according to the special purpose of the apparatus. For
general work the form of measuring tube shown in Fig. 46 (p. 136),
|
:
|
|
i




THE GAS-VOLUMETER 145
with a central bulb, and graduations above and below, is to be
recommended.
By using several reaction vessels, many more determinations can be
carried out, than with the ordinary nitrometer, in the same time. The
contents of the reaction vessels are allowed to cool, after shaking, until
the evolved gas is ready for measurement, a new determination started
in a second vessel, and so on. In analysing explosives, the funnel of the
reaction vessel is connected with the sulphuric acid trap described on
p. I 36, and shown in Fig. 44.
A modification of the Lunge gas-volumeter, in which all the connec-
tions are made of glass, has been described by Gruskiewicz."
Lunge” has devised a mechanical stand, which is extremely con-
venient for manipulating the gas-volumeter, which is described below
(p. 154 et seq.); it is used by Lunge and Marchlewski with their
apparatus for the determination of carbon dioxide and of carbon, and
is applicable to all other forms of the nitrometer.
Special forms of the gas-volumeter have been designed by Lunge,”
for the determination of nitrogen in elementary organic analysis, and
by Lunge and Neuberg “for the determination of vapour densities.
When the gas-volumeter is in daily use for a specific determination,
the graduations on the measuring tube may be made to correspond
with definite weights in mg. of the constituent determined, or to
percentages, when a definite weight of initial material is taken for
analysis. Such graduations may either replace the divisions into c.c.,
or may be placed alongside them. For nitrogen determinations by
Dumas' method, intervals of O-7997 c.c. are marked off and divided
into tenths. As I c.c. of nitrogen under normal conditions weighs
1.25o; g., each graduation corresponds to I mg. of nitrogen, and the
weight of the gas can thus be read directly, as soon as the tube has
cooled, after the combustion. For the azometric determination of
ammoniacal nitrogen with sodium hypobromite, either the same
graduation can be employed, or divisions of O-658 c.c. = 1 mg. NHs,
since I c.c. of nitrogen under normal conditions corresponds to I-52O
mg. NHS. If, for instance, O.2 g of ammonium salt are used, then
each such division represents, in the former case, O.5 per cent. N,
and in the latter, O.5 per cent. NHs.
For the determination of calcium carbonate in bone, charcoal,
marl, and the like, that is for use as a calcimeter, each c.c. CO, under
normal conditions corresponds to I-9766 mg. CO2, or 4:497 mg. CaCO3.
lf, therefore, a calcimeter is to indicate I mg. of CaCOs for each
graduation, these must correspond to O.222 c.c., and if O.5 g. substance
is used for analysis, then each division will indicate O.2 per cent.
* Z, anal. Chem., 1904, 43, 85; Chem. Centr, 1904, I., II67.
* Z. angew. Chem., 1892, 5 678. * Ber, 1890, 23, 446. * Ibid., 1891, 24, 729.
K
146 GENERAL METHODS USED IN TECHNICAL ANAI.YSIS
CaCOs. The absorption of carbon dioxide by the generating liquid
must, of course, be taken into account, unless it is excluded by using
the form of generating apparatus shown in Fig. 53, p. 15O. -
Such special graduation of the measuring tube is unnecessary, if the
quantities given in the following table are used for analysis; an
ordinary measuring tube, divided into fºr c.c., can then be employed.
Table for the Conversion of Volumes of Gas, obtained in the
Gas-Volumeter, into Weights of Active Constituents.
1 2 3 4 5
G 1 c.c. Gas =
Substance Analysed. Active Constituent. | Method of Analysis. E.d #: Con-
- * stituent in mg.
Organic substances Nitrogen Dumas' method N2 1°2505
Ammonium salts Do. Wi. sodium N2 1.2818 1
ypobromite
Do. Ammonia Do. N2 1-5582 1
Urine Urea Do. N2 2.956 1
Bone-charcoal, marl, etc. Carbon dioxide { D º1On } CO2 1 ‘976
Do. Calcium carbonate O. CO2 4°497
Pyrolusite Manganese dioxide With H2O2 O2 3°885
Bleaching powder Chlorine Do. O2 1°6095
Potassium permanganate Oxygen Do. O2 0°7146
Chili saltpetre Sodium nitrate In nitrometer NO 3°7986
Nitrous vitriol 2O3 Do. NO 1°6975
Do. - ... HNQ, Do. NO 2°8144
Do. {sº} Do. NO 5-3771
Do. Sodium nitrate Do. NO 3-7986
Nitroglycerine, dynamite, etc. Trinitroglycerine Do. NO 3°358
Do. Nitrogen Do. NO 0-6257
Nitrocellulose, pyroxylin Do. Do. NO 0-6257
* These values are corrected for the “nitrogen absorption”; 2.5 per cent, in the case of
ammonium salts, and 9 per cent. in the case of urea. (Cf. p. 131.)
The table comprises a number of substances, specified in the first
column, which are frequently examined by gas-volumetric methods.
The second column gives the “active constituent” of the substance
analysed ; the third column, the analytical method employed; the fourth,
the nature of the evolved gas; the fifth, the weight of “active con-
stituent” in mg. corresponding to I c.c. of gas at O’ and 760 mm.
pressure. If one hundred times the quantity of substance given in
column 5 is taken, then the number of c.c. of gas obtained is numerically
equal to the percentage of the active constituent. In some cases it is
necessary to use a larger quantity of substance, and then I c.c. of gas
represents, of course, a correspondingly smaller percentage.
The data given in this table are based on the observed densities of
THE GAS-VOLUMETER 147
the gases' concerned, not on the values calculated from the molecular
weights, which do not take into account the deviations from Boyle's
law.
When the gas-volumeter is used as an azotometer, an addition of
2.5 per cent of the volume of gas obtained must be added, to allow
for the incompleteness of the reaction, the so-called “nitrogen absorp-
tion” (cf. p. 131). With this correction, I c.c. of nitrogen under normal
conditions corresponds to I-281.8 mg. ammoniacal nitrogen or I-5582 mg.
ammonia, so that O. I 282 g. or O. I 558 g. of ammonium salt must be
taken for analysis, if I c.c. of gas is to correspond to I per cent. of N
or of NHs respectively.
Japp * has shown, that by suitably adjusting the position of the
reduction tube, the volume of gas in the measuring tube may be made
to indicate the weight of active substance present, without further
calculation and without Special graduation. As an example, suppose
that a reduction tube is used, the normal volume of which is 25 c.c.,
that is, the gaseous contents occupy this volume at O’ and 760 mm.
pressure; the weight of I c.c. of nitrogen under these conditions is
o.OOI 2505 g., and of the 25 c.c. O.OOI 2505 × 25 = O.O3 I26 g. If the
reduction tube is adjusted so that the volume of contained gas is
31.3 c.c., each C.C. of gas in the measuring tube will then correspond to
I mg. of nitrogen. Q
According to Lunge,” 25 c.c. is too small a normal volume to adopt
for accurate work, and it is preferable to use an ordinary reduction tube,
or one of 90 to I 50 c.c. capacity, which is adjusted to contain IOO c.c.
under normal conditions. To convert the volume of gas into its
equivalent in mg., the mercury in the reduction tube is adjusted so
that the volume is equal to IOO c.c., multiplied by the weight of gas
per litre, thus:—
For atmospheric air tº e g I 29.3 C.C.
, Oxygen e e ſº * I42.9 C.C.
, nitrogen & © © & I 25-O5 C.C.
, nitric oxide . gº º ſº I 34-O2 C.C.
The corresponding figure for carbon dioxide being 197.66 c.c., which
would extend beyond the graduations, the reduction tube is adjusted to
98.8 c.c. in this case, so that each c.c. in the measuring tube corresponds
to 2 mg. of gas. This principle may be extended so that the volumes
of gas found correspond, not to weights of gas, but to weights of the
active constituent of the substance analysed. Thus, in determining the
* The densities are taken from the Physikalisch-chemische Zabellen of Llandolt-Börnstein, 3rd
edition, 1905.
* J. Chem. Soc., 1891, 59,894; cf. also Lunge, Ber., 1891, 24, 1656 and 3491.
* Ber., 1892, 25, 3162.
148 GENERAL METHODS USED IN TECHNICAL ANALYSIS
amount of active chlorine in bleaching powder, by means of hydrogen
peroxide, oxygen is obtained instead of chlorine, volume for volume; by
adjusting the reduction tube to # of the weight of a litre of chlorine,
that is, to Io/.3 c.c., each C.C. of gas in the measuring tube corresponds to
3 mg. Cl. In order to read off weights of CaCOs instead of CO, the
volume in the reduction tube is adjusted to I97.0 × IOO#. *= 112.3 C.C. ;
I c.c. of CO2 in the measuring tube then represents 4 mg. CaCOs.
Japp's proposal, even with the above extensions, seldom offers,
however, any advantage over the method of working described above
(p. 145 et seq.), and set out in the table, according to which the
quantity of substance taken for analysis is such, that on adjusting the
reduction tube to IOO c.c., percentages may be read directly in the
measuring tube.
Other forms of apparatus for measuring gases, without the necessity
for barometric and thermometric readings, have since been constructed
by Hempel, Bleier.” Bodländer,” and others.
The barothermoscope devised by F. Salomon “ is provided with a
scale on which the volume, corresponding to the temperature and
pressure of the surrounding gas, can be read off directly. The
instrument is most ingenious, and is based on the new thermometric
system proposed by Salomon ; it is, however, rather troublesome to
adjust and difficult to transport, and has accoºingly not been adopted
for practical work.
A. Wohl” has described an apparatus for measuring the volume of
gases by determining the pressure, a vessel of constant volume being
employed for the purpose. Wohl and Poppenberg" have extended this
method, especially to the determination of nitrogen in nitrates and
nitric esters such as gun-cotton, etc.
Apparatus for the determination of Carbon Dioxide.
This section is restricted to a description of the methods for the
gas-volumetric determination of carbon dioxide generated from solid
and liquid substances. The methods for the determination of carbon
dioxide in gaseous mixtures are dealt with under “Technical Gas
Analysis”; the various forms of apparatus for the determination of
carbon dioxide, by loss of weight, in calcimeters, in which the gas is
expelled by means of a stronger acid, are also omitted, as they never
give very accurate results. *
Both the azotometer (p. 125) and the nitrometer (p. 132) may be
used for the gas-volumetric determination of carbon dioxide, but they
* Z, angew. Chem., 1894, 7, 22. * /öid., 1893, 6, 376; 1894, 7, 686.
* Ber., 1897, 30, 2733; 1898, 31, 236. * Aer, 1902, 35, 3493; 1903, 36,674.
* Z. angew. Chem., 1895, 8, 49. * Ibid., 1903, 36, 676.
DETERMINATION OF CARBON DIOXIDE 149
involve difficulties, on account of the solubility of the gas in the liquid
present in the decomposition vessel and in the water over which it
is collected, in the case of the azotometer.
Many suggestions have been made to overcome these difficulties,
either by the use of a special confining liquid, over which the gas is
collected, as in the apparatus patented by Baur," which is merely a
simplified and less effective form of azotometer, or by the use of
correction tables, as in Scheibler's apparatus, which is used chiefly in
sugar works, in those of Dietrich and of Michaëlis, which are largely
used in cement works, and in those of Rumpf, Thörner, Fuchs, and
many others.
Satisfactory results were first obtained with the apparatus devised
by Pettersson,” in which the carbon dioxide is finally displaced
by hydrogen, generated from iron or from aluminium wire, the gases
collected over mercury, and the carbon dioxide absorbed in an Orsat
apparatus. The expulsion of the carbon dioxide is, however, only
complete after three or four repetitions of the operation, and the volume
of gas obtained must be corrected to normal temperature and pressure
in the usual way; the method is therefore unsuitable for technical
laboratories.
The successful solution of the above difficulties has been effected
by Lunge and Marchlewski.” Their method resembles that of
Pettersson, but the gas is completely expelled by a single boiling, and
no readings of the thermometer and barometer or calculations are
required, as the apparatus is so arranged that weights or percentages
may be read off directly. Although the apparatus appears, at first sight,
as complicated as many others devised for the same purpose, its mani-
pulation is very simple; a determination can be performed in a few
minutes, with greater exactitude than can be obtained by any other
process, including the titrimetric and the volumetric methods, as it
avoids certain sources of error which are inherent to the latter.
The apparatus is shown in , the actual size in Fig. 53; Fig. 54
shows an improved form of the generating flask on 4 scale. For solid
substances, a flask of 30 c.c. capacity is used ; for solutions and for
other special purposes, such as the determination of carbon in iron, a
large flask is employed, to which a small reflux condenser is attached.
The flask A is closed by a good soft rubber stopper, in which a
dropping-funnel a and a capillary tube ba', bent at right angles, are
fixed. The latter is 30-35 cm. long, and is connected with
* Z. anal. Chem., 1884, 23, 371. * Ber., 1890, 23, 1402.
* Z, angew. Chem., 1891, 4, 229 ; 1893, 6, 395. / Soc. Chem. /nd., 1891, Io, 658.
* Cf. Stahl und Eisen, 1891, II, 666; 1893, 13, 655; 1894, I4, 624; also, Z. angew. Chem.,
1891, 4, 412. (Determination of carbon in iron and steel, and of carbon dioxide in aqueous
solutions.)
150 GENERAL METHODS USED IN TECHNICAL ANALYSIS
the lateral capillary e of a gas-volumeter B. The measuring tube of
the volumeter ends in a bulb of about IOO c.c. capacity, below which
the tube is graduated in ſº c.c. for another 50-60 c.c. The form of
tube, described below, in connection with the “Universal Gas-volumeter”
may also be used, but it is not so suitable.
The glass tap f has two oblique passages (cf. p. 134), one of which
connects through the tube e to d, and the other through the tube g, and
the two-way tap h to the
“absorption vessel” E, which
is rather more than half-filled
with a solution of I part of
sodium hydroxide in 3 parts
s. ... sº...:
FIG. 54.
of water; its outer limb is
surmounted by a tube i, filled
with soda-lime. B communi-
A tº cates below, by means of a
FIG. 53. T-piece and thick rubber
- tubing, with the reduction
tube C (p. 139), which is graduated below the bulb between 100-130 c.c.,
and with the pressure tube D, which also serves as a pump; the two
tubes are held by the forked clamp l. (p. 142), so that they can be moved
either separately or simultaneously, B, C, and D are filled with mercury.
Instead of the reduction tube ending in a capillary, as shown in
the figure, it is better to use one fitted with a Göckel's mercury-
tap (p. 140). The adjustment of the reduction tube is made for
moist air as described above (p. 140). If no barometer is available
-
#
#1
E.


DETERMINATION OF CARBON DIOXIDE 151
for this purpose, the apparatus itself may be used to determine the
atmospheric pressure, as follows. The tap on C is momentarily closed,
all air expelled from B by raising D, the tap f closed, and D lowered
by about o-5 m. to make sure that no air leaks in through f; D is
then lowered for another O.3 m. approximately, so that the mercury
just leaves the tap f and the vertical distance between the mercury
levels in B and D, which represents the barometric pressure, measured.
It is advisable to bind all the rubber tubes to the glass with fine
wire, with the exception of the connection between d and e, which has
to be separated for each determination; the apparatus should always
be tested before use, to make sure that it is air-tight.
To carry out a determination, the flask A is cleaned, care being
taken to remove all traces of acid, and a weighed quantity of the
substance for analysis introduced, together with about O.O8 g. of
finest aluminium wire. The best plan is to weigh Out O.O.8 g. of the
stock of aluminium wire, measure its length, cut off a number of similar
pieces and make each into a little roll, for use in each determination;
this amount of aluminium should generate about IOO C.C. of hydrogen.
After the introduction of the substance and of the aluminium, the stopper
is replaced, d connected with e, and the air evacuated from A by
lowering D as far as possible and then expelling the air, drawn into B,
by raising D until the mercury reaches the tap f; this operation is
repeated two or three times. A few c.c. of pure hydrochloric acid,
diluted with three times its volume of boiled distilled water, are then
added through a, and the flask A warmed gently, with a spirit-lamp,
for about two minutes; a further quantity of dilute acid is then added,
the flask again warmed for two minutes, and this operation repeated.
In this way the carbon dioxide is liberated and the hydrogen generated
from the aluminium at the same time; all the carbon dioxide dis-
solved in the liquid is expelled by the hydrogen, assisted by the
diminished pressure in the apparatus. Meanwhile D is placed so
that the contained mercury is at a lower level than in B. If the
evolution of gas becomes too vigorous, the heating is discontinued.
When all the aluminium is dissolved and the liquid is quite clear, dilute
acid is added through a until the flask A and the tubes b and d are
quite full, almost as far as the tap f ; after closing f the apparatus is
allowed to stand for ten minutes for the temperature to become
equalised. Meanwhile A and d may be removed, cleaned, and the
substance for the next determination weighed out.
To prepare the pipette E for the absorption of the carbon dioxide,
air is blown in through i until the sodium hydroxide solution flows
out from the outer tube, leading from the tap h, which is then closed.
C and D are then adjusted until the mercury in C is at the IOO c.c.
graduation and also at the same height as in B. The volume of gas
152 GENERAL METHODS USED IN TECHNICAL ANALYSIS
in B is read off, and then transferred to E, three times in succession, to
absorb the carbon dioxide, after turning the tap h so that E and g
communicate, and opening the tap f: The solution in E is finally
syphoned back to the tap h, C and D adjusted as before, and the
volume of gas in B again read off. If large quantities of carbon
dioxide have been absorbed, the apparatus should again be allowed
to stand for a few minutes so that the temperature may become
equalised.
Any small quantity of water which may get over into B, on heating
A, may be disregarded, the volume of gas in B being read from the
top of the mercury meniscus, both before and after the absorption of the
carbon dioxide.
The difference in the readings before and after the absorption of
the carbon dioxide represents the volume of the gas at O’ C. and 760 mm.
pressure. If it amounts to n c.c. it corresponds to n x 1.976 mg.
CO2, or m × 4,497 mg. CaCOs. If p g. of substance were used for
m × O. IQ76
analysis, then the percentage of CO2 is , and the percentage of
CaCOs ***4497. This calculation may be avoided by the following
method, which enables percentages to be read off directly.
Suitable amounts of the material under examination are weighed
out; if percentages of CO, are required, the following amounts are
taken :—
For substances which contain much CO2 O. 1976 g. ; I c c = I per cent. CO2
For substances which contain less CO2 O-3952 g. ; I c.c. = O-5 per cent. CO2
For substances which contain little CO2 I-976 g. ; I c.c. = O-I per cent. CO2
If percentages of CaCOs are required, the following amounts are
taken —
For pure limestone ſº tº ſº . O. 1799 g. ; I c.c. = 2.5 per cent. CaCO3
For marls rich in calcium carbonate . O.2248 g. ; I c.c. = 2.0 per cent. CaCO3
For marls poor in calcium carbonate . O'4497 g. ; I c.c. = I.O per cent. CaCO3
Substances containing very little cal-
cium carbonate, cements, etc. . 2.248 g. ; I c.c. = O.2 per cent. CaCO3
Examples.
A. Marl rich in calcium carbonate. Weight of substance: O.2248 g.
First reading: 145.3 c.c. Second reading after absorption of CO, :
Io9-7 c.c. Difference: 41.6 c.c. 41.6 × 2 = 83.2 per cent. CaCO3.
B. Portland cement. Weight of substance: 2,248 g. First reading:
I 16.9 c.c. Second reading: I IO-4 c.c. Difference: 6.5 c.c. 6.5 × 0.2 = 1.3
per cent. CaCOs.
Or:-Weight of substance: 1.976 g. First reading: I 16. I. c.c.
DETERMINATION OF CARBON DIOXIDE 153
Second reading: I IO-4 c.c. Difference: 5-7 c.c. 5-7 x 0. I = O. 57 per cent.
CO2:
The above method for the estimation of carbon dioxide has recently
been considerably simplified by Lunge and Rittener," and rendered
applicable to the determination of this gas in the presence of chlorine or
of sulphuretted hydrogen. The decomposition of the salt is conducted
exactly as in the Lunge-Marchlewski method, and a similar flask
is used for this purpose, but the volumeter and absorption pipette
are replaced by an ordinary cylindrical Bunté burette. A saturated
brine solution is used, instead of mercury, as the confining liquid.
The method and apparatus is described in detail in connection
with the determination of chlorine and carbon dioxide (p. 514,
Fig. I46). To carry out a determination, the decomposition bottle
is first connected by means of the capillary tube (à d, Fig. 53) to
the capillary exit-tube of the burette, and the whole exhausted by
attaching an ordinary pump to the lower tap of the burette; sufficient
exhaustion is obtained in two to three minutes, and the burette tap is
then closed and connected with the pressure bottle in the usual way.
As a check on any possible leakage at this tap, a little of the brine
solution is introduced from the pressure bottle, so as to just fill the
capillary portion of the bottom of the nitrometer, and the tap again
closed. After the completion of the decomposition, the burette is
disconnected, allowed to stand for twenty to twenty-five minutes, and
the volume of gas read, after adjusting the pressure, and corrected for
temperature and pressure. The contained carbon dioxide is then
absorbed by introducing a solution of sodium hydroxide (I : 2) into
the burette, through the attached funnel; the contents of the burette
are well shaken, to promote the absorption of the carbon dioxide, and
further successive portions of sodium hydroxide are introduced until
there is no further decrease in volume. The final reading is then taken
and corrected for temperature and pressure as usual. The difference
between the two readings, before and after the absorption by sodium
hydroxide, gives the volume of carbon dioxide obtained. In correcting
the volume of gas, the vapour tension of the salt solution may be taken
as 80 per cent. of that of pure water. The solubility of carbon dioxide
in the brine solution is negligible, as the gas is only in contact with a
small surface of the solution and is not shaken up with it. Results
obtained by the authors with pure sodium carbonate, with cement and
with marl, agree well with those got by the Lunge-Marchlewski method.
The application of the method to the determination of carbon dioxide
in presence of sulphuretted hydrogen is described on p. 439, and in
presence of chlorine on p. 514.
* Z. angew. Chem., 1906, 19, 1849.
154 GENERAL METHODS USED IN TECHNICAL ANALYSIS
The Universal Gas-Volumeter."
The manipulation of the Lunge-Marchlewski apparatus is greatly
facilitated by the use of the mechanical stand mentioned above
º
FIG. 55.
(p. I49). This is shown apart from its base in Fig. 55, which
* This apparatus with all necessary fittings is supplied by C. Desaga, Heidelberg, and by
A. Jöge, Zurich. -

THE UNIVERSAL GAS-VOLUMETER 155
represents the Universal Gas-Volumeter, as arranged by Lunge for all
gas-volumetric work. A is the measuring tube, B the reduction tube,
and C the pressure tube, as in the ordinary gas-volumeter. E is
the reaction or shaking vessel for the analysis of saltpetre, dynamite,
etc. (p. 144), with pressure tube F, and auxiliary bottle D (p. 137).
The parts a to 2 serve for determinations of carbon dioxide, as in
Fig. 53. The mechanical stand permits the rough adjustment of the
forked clamp l by means of the steel wire rope m ; this is wound on to
the drum m, which is provided with a spiral groove, and with a brake o
acting on a cogged wheel, by means of the handle p. Five or six turns
of the handle suffice to move the forked clamp from the highest to the
lowest position (e.g., for evacuating). The brake o works automatically
when the clamp is being raised ; whilst lowering, it is thrown out of
action by means of the small clutch shown in the figure. The exact
adjustment of the clamp / is effected by the screw s, connected with
the support of the drum ; this screw acts as a fine adjustment, by
pressing the wire rope forwards out of its direct position, thus increas-
ing the tension on the wire.
Both the reduction tube B and the pressure tube C can, of course, be
moved independently, as is always necessary for the final adjustment;
much relative adjustment of these two tubes is not usually required, as
the volume of air in the reduction tube is generally kept in the
neighbourhood of loo c.c.
A straight-edge (p. 142) with spirit-level t, and supported by an
extension of the clamp Z, may be used for the final adjustment;" it
should reach across to the measuring tube A.
The measuring tube A is suitable for all classes of analyses. It has
a capacity of about 140 c.c. (I 50 c.c. would be preferable, were it not for
the added length), and is graduated from the top, in ºr c.c., from O to
3O C.C.; this portion of the tube is followed by a bulb, below which the
graduation is continued from IOO to 140 c.c. The measurement of
both small and large quantities of gas is thus provided for. The tube
can be moved in its two clamps, but this is not usually necessary, as
all adjustments are effected by means of the clamp l.
General Remarks on Measuring Vessels for Gases.
As emphasised on p. 44, it is essential that accurate apparatus
should be used for measuring gases; it should, therefore, be either
calibrated, or controlled and stamped by a recognised authority. In
the case of ordinary gas analyses, the main requirement is that the
volumes marked as c.c. should be actually equal to one another, but
in the case of gas-volumetric work, it is important that the graduations
* Cf. Aer, 1891, 24, 3948.
156 GENERAL METHODS USED IN TECHNICAL ANALYSIS
should correspond to actual c.c., as the weight of gas is deduced from
its volume. In calibrating the measuring vessels the following points
must be observed:—I. The adhesion of the confining liquid used ; in
the case of mercury this is zero, but it must be allowed for with water,
aqueous solutions, petroleum, etc. 2. The meniscus correction, which
varies, of course, with the nature of the liquid employed, and which
must be specially taken into account when the calibration by weigh-
ing water or mercury is done in an inverted position to that in
which the vessel is subsequently used. Göckell is quite justified in
contending that the graduations of gas measuring vessels should be
clearly stated, e.g., “Corr. for H2O’’ or “Corr. for Hg.”
I. DETERMINATION OF SPECIFIC GRAVITY
The determination of the specific gravity of liquids plays an
extremely important part in chemical technology. It is almost
exclusively measured by means of hydrometers; pyknometers, the
Westphal balance, and other well-known forms of apparatus for the
determination of specific gravity, are only used occasionally. It is,
however, very desirable that a works' laboratory should possess some
instrument with which to control the readings of hydrometers, as the
latter, especially the cheaper kinds, are very often inaccurate.
A very reliable method of control consists in determining the specific
gravity of a suitable liquid, by weighing it in a calibrated litre flask, to
within O. I g., and then testing it with the hydrometers
under examination, using tables for the reduction of
their readings to specific gravities (cf. infra). The

ſº
B Fº © &
3. 2 temperature for which the hydrometers are calibrated
\ must, of course, be adhered to. Hydrometers must
4, 42 not be greasy, soiled, or damp, before introduction
| into the liquid; they must be immersed carefully, and
must be read from below, at the height of the meniscus
of the liquid, that is, at Al A, (Fig. 56), and not at B, B2.
Hydrometers which indicate the specific gravity directly are fairly
often used for liquids lighter than water, and it would be advan-
tageous if they were used more in other cases, instead of the various
scales of Baumé, Beck, etc.; they are practically never used in works
for specific gravities above I.O. Fleischer's “Densineter,” the gradua-
tions of which represent multiples of O.OI, and are convertible into
actual specific gravities by addition of I, have unfortunately almost
entirely failed to be adopted in practice; on this scale, 7° D = 1.07;
25° = I-25, etc.
Densineter degrees are rather large. In the Twaddell hydrometer,
| Cf. Chem. Zeit., 1902, 26, 159.
FIG. 56.
DETERMINATION OF SPECIFIC GRAVITY 157
which was used in this country long before the introduction of
Fleischer's densineter, and which is still in general use for technical
work, the degrees are half of the above value. Each degree corresponds
to O.OO5 sp. gr., prefixed by unity. Thus, 7° Tw. = 1.035; 20° Tw. = I. IOO;
IOO’Tw. = 1.5oo sp. gr., etc. No table is therefore required for the
conversion of degrees Twaddell into ordinary specific gravities, and
the value of a degree Twaddell is quite definite. It is remarkable
that this rational and practical hydrometer, the scale of which is usually
distributed over six spindles, should be almost universally used in
this country, which is so very conservative and unpractical with regard
to weights and measures, whilst the continental nations, which have
all adopted the metric system of weights and measures (with the excep-
tion of Russia), have not yet adopted a rational hydrometer.
In the United States, the Baumé hydrometer is used, for which a
special scale, different to that employed in other countries, has been
adopted.
During the last generation, France, Germany, and other continental
countries have adopted the Baumé scale almost universally, to the
exclusion of others; the advantage thus gained is, however, very
limited, as there is no exact and recognised interpretation of the signi-
ficance of a Baumé degree. The hydrometers were originally graduated
by calling the point up to which they sank in pure water O’, and that
up to which they sank in a IO per cent. Salt solution at 17°. 5 C., IO’B.
The intermediate space was then divided into IO equal parts, and these
“degrees” continued upwards, at equal intervals, so that the scale value
of each degree was the same, but in no way corresponded to equal
changes in specific gravity. The scale was also affected by the fact
that the salt used was seldom pure and dry, so that the specific gravity
of a so-called IO per cent. salt solution varied considerably. Deter-
minations by Gerlach gave the value I.O73 II at 14° R. (17°. 5 C.), and
from this he calculated a table comparing Baumé degrees with specific
gravities." It is doubtful whether hydrometers, according to Gerlach's
scale, are made now ; their lower degrees do not agree with those
ordinarily used, and the higher ones show extraordinary differences.
Thus, according to Gerlach's values, 66° B. = sp. gr. 1.8171, which would
correspond with a sulphuric acid of 89.5 per cent. H2SO, whereas the
real value of 66° B. is 93 per cent. H2SO,. 2.
The modern practice is to mark the hydrometers 66° B., at
the point to which they sink in “English sulphuric acid” (i.e.,
sulphuric acid of 93 to 95 per cent. H2SO4) at 17°. 5 C., and to
divide the space between this and O’ into 66 equal parts. But
uncertainty is introduced in the definition “English sulphuric acid.”
Pure acid was certainly never employed, as is proved by the fact that
' Dingl, polyt.J., 1870, 198, 315.
158 GENERAL METHODS USED IN TECHNICAL ANALYSIS
the specific gravity I-842, which is usually considered the equivalent
of 66° B., is certainly greater than that of pure sulphuric acid at
17°. 5 C., even at its point of maximum density (97.5 per cent. H2SO.).
Apart from this, there is the uncertainty as to the percentage composi-
tion of the acid. Accordingly, many and very varying scales of the
Baumé hydrometer were placed on the market; a test made by C. F.
Chandler (1881) disclosed thirty-four different scales, none of which
were correct. About twenty-five years ago, Kolb, in France, tried to
remove this confusion by the construction of a “rational” Baumé hydro-
meter, the scale of which was based on the following perfectly sound
physical principle.
If a hydrometer sinks in water to O’, and in a liquid D, of specific
gravity d, to n°, then the weight of the volume of displaced liquid is, in
each case, equal to the weight (G) of the hydrometer, thus:—
Weight of water displaced by hydrometer = G
Weight of an equal volume of liquid D = d G
Weight of water displaced by n scale divisions = n
Weight of an equal weight of liquid D = nd
As the weights d G and G differ by na,
d G – G = n.d.; therefore, d = gº
_ nd
and G - Z-T.
If, according to Kolb, 66° B. is taken to correspond to sp. gr. I-842 at 15°C, then
when n=66, d- I-842. This gives G = I-443, and
d = - 1443–.
I 44-3-71
The formulae in the subjoined table are calculated from the values
v, for the specific gravity of a Io per cent. Salt solution.


Temperature in Liquid heavier Liquid lighter
degrees C. 77. than Water. than Water.
O e * _ 145'88 _ 145'88
146°3 146'3
°-0 C. e d = —t:-- -
15°-0 C 1.073350 146°3 -- 71 136°3+ m
O e e _ 146-78 146-78
17°-5 C. 1-073110 146-78 – n T 136-78. In
Kolb's hydrometer, though based on an inexact specific gravity for
pure sulphuric acid, i.e., I-842, is more concordant, especially for the
higher degrees, than the Gerlach hydrometers, based on the IO per cent.
salt solution. It has not, however, been generally adopted even in
BAUMF, HYDROMETERS 159
France, but it is used to a considerable extent in Germany. Two other
“rational” Baumé scales are in use, namely, that used in Holland, in
which d-–tº– , and that adopted in America, in which d =–143–
I44–? I45 – ?
In the following table the specific gravities corresponding to the
“Rational,” Gerlach and American Baumé scales, are given ; * the
table issued by the German Normal Standards Commission in 1904, for
tº º - tº I 5° . e y 3 f
the conversion of specific gravities at ; into “rational * Baumé
degrees is also given.
Comparison of various Baumé Hydrometers, for Heavy Liquids,
with Specific Gravities.


Rational Baumé's American Rational Baumé's American
Hydrometer. Hydrometer Scale. D Hydrometer. H. Scale.
* | a_1448 tº..., | a__145 |*||a- 1448 | dº;, g_ 145
144".3 — m. Scale. 145 – m. 144°3–7. Scale. 145 – m.
I 1°007 1-0068 1°005 34 1°308 1°3015 1°309
2 1*014 1*0138 1*011 35 1 *320 1°3131 1-317
3 1*022 1*0208 1-023 36 1 °332 1°3250 1°334
4 1*029 1*0280 1*029 37 1°345 1 °3370 1°342
5 1-037 1*0353 1*036 38 1°357 1°3494 1°359
6 1*045 1*0426 1*043 39 1 °370 1 °3619 1°368
7 1*052 1*0501 1*050 40 1 °383 1°3746 1°386
8 1*060 1*0576 1.057 41 1 397 1°3876 1°395
9 1*067 1*0653 1*064 42 1°410 1°4009 1°413
10 1-075 1-0731 1-071 43 1 “424 1°4134 I •422
11 1-083 1*0810 1-086 44 1°438 I '4281 1°441
12 I •091 1*0890 1°093 45 1°453 1°44'21 1°451
13 1 100 1-0972 1 100 46 1°468 1°4564 1°470
14 1 *108 1°1054 1-107 47 1 “483 1°4710 1°480
15 1*116 1 1138 1*114 48 1°498 1°4860 1°500
16 1-125 1°1224 1-122 49 1°514 1°5012 1°510
17 1*134 1*1310 1*136 50 1°530 1°5167 1°531
18 1°142 1°1398 1°143 51 1°540 1°5325 1°541
19 1*152 1 J487 1°150 52 1°563 1°5487 1°561
20 1*162 1°1578 1*158 53 1-580 I ‘5652 1°573
21 1-171 1°1670 1-172 54 1°597 1°5820 1°594
22 1-180 1°1763 1-179 55 1.615 1°5993 1-616
23 1-190 1°1858 1*186 56 1 *634 I-6169 1-627
24 I 200 1*1955 1 °201 57 1 *652 1 *6349 1°650
25 1°210 1-2053 1 °208 58 1-671 1°6533 1 *661
26 1 220 1°2153 1°216 59 1°691 1-6721 1°683
27 1°231 1*2254 1°231 60 1-711 1 *6914 1-705
28 1°241 1°2357 1-238 61 1°732 } •7111 1-727
29 1.252 1-2462 1°254 62 1°753 1°7313 1-747
30 1.263 1°2569 1°262 63 1-774 1 °7520 1-767
31 1°274 1-2677 1-269 64 1-796 1-7731 1-793
32 1 °285 1°2788 1 -285 65 I '819 1-7948 1°814
33 1.297 1°2901 1°293 66 1°842 1°8171 1°835
* The figures for the American Baumé scale are taken from the Chemical Trade Journal, 1887,
2, 183.
o
973. I? | I]. I. If 1,60. Iy 830. If 676.07 || 91.8.07 IO8.07 || 131-07 | 899. Of 8/9.07 | 68
809.07 || 837.07 | 898.07 || 8/3.07 303.07 |8&I.07 | 890.07 || 8/6.68 || 306.63 | 938.68 || 88
09/.68 || 7/9.68 869.68 || ZZ9.68 977.68 || 0/3.68 || 763.68 || AI3.68 || Of I.63 | 890.68 18
986.88 606.88 || 388.88 gg/.88 || 8/9.88 || IO9.88 || 739.88 977.88 | 898.83 || 063.88 || 93
3T3.88 || 78.I.88 990.83 || 11.6.18 868. 18 || 038. 18 I?/- 18 || 399. 18 || 889. 18 || 709. 18 98. I
967. 18 || 978-13 || 193. 18 18I. 18 || 10T, 18 || 830. 18 || 876.98 || 898.98 || 881.98 || 801.98 || 78
839.98 || 879.98 || 199.98 || 988.98 g08.98 || 733,98 || 8?I-98 || 390.98 || IS6.98 || 668.98 || 38
AI8-98 || 98/.98 || 899.98 I/9.98 || 687.98 || 107.98 || 938.98 || 873-98 || 09I.99 || 110.98 || 38
766.78 II6.7g | 838.78 gp1.78 399. £8 || 619.7g g67.78 IIſ. W8 || 133.jg | 87%. 78 I8
69 I-78 91.0. 79 I66.88 106.88 || 338.88 || 13/.38 399.88 || 199.88 || 387.88 168.88 || 08. I
&I8.88 133.88 || IWI-88 990.88 696.38 || 888.33 16/.38 III. Žg | 739.38 199.38 63
09:7.28 | 898.68|913.68 | 68I.Zg | ZOI.33 |g|IO.33 836. Ig 078. Ig | Zg/. Ig | 799. Ig | 8%
9/9. Ig | 887. Ig | 007. Ig | II8. Ig | 333. Ig | 88.I. I3 || 770. Ig | 996.08 || 998.03 || 1//- 08 || ||3
889.03 || 669.03 609.08 || 6 I?.08 6Z8.08 || 68Z.08 || 67 I.08 || 890.08 196.6% 918.6% 93
98/.6% | #69.6% 309.6% &I9.6% | 037.6% 633.6% | 183.6% | 97 I.6% | 890.6% | I96.8% g3. I
698.8% 1//.8% 789.8& I69.8% 867.8% g07.8% ZI8.8% | 613.8% gåI.8% I30.8% 7%
/86. 16 || 878. 1% | 671. 13 || 999. 13 || 099.1% |997. 13 || 018. 1% 913. 13 || 08I. 13 980. ZZ | 3%
066.9% | 968.9% | 661.9% 801.9% 109.9% | IIg.9% 917.9% 618.9% | ZZZ. 13 || 93I.9% 3%
830.9% I96.93 || 738.9% | 934-93 || 889.93 || 079. g3 || 377.9% 778.9% 9%. g3 || 87T. g3 I&
670.93 || 096.jø | I98.f3 || 39/.7% | 899.iº, I jºgg. P& 997.f6, 998.7% 99%. 5% gg.T. #3 || 0%. I
990.7% |ggó.8% 798.8& | 891.8% 399.83 || Igg.8% | 097.8% | 678.8% li/Z.8% g?I.8& 61
870.8% | IW6.3& | 688.6% 131.3% V89.3% Igg.ZZ | 837.3% gå8.3% | 333.33 ||61|I.ZZ | 8T
9I0.3% II6. I& | 1,08. I& 801. I& | 669. IZ | #67. I& | 688. I& 78%. I& 6/I. IZ | #10. I& I
696.0% | 898.0% 191.0% I99.0% | 979.03 || 637.0% | 888.0% | 933.0% | 6TI.0% &IO.0% 9T
g06.61 | 861.6L | 069.6L | 389.6L | 717.6L |998.6L | 893.6L |671.6L | 070.6L | 186.8I g|I. I
Č8.8I 8II.8L | #09.8L | 767.8L | 788.8T | p 1%.8I 79[.8I 790.8L | 876. I Z88. II Pl
I&A. AI | 019. AI 867. AI 988. AI | 7/3. AI | 39T. II | 090. II | 886.9L | 938.9L | 3I/.9L | 8 I
669.9L 987.9L | 818.9L 69%.9T g?I.91 || Ig().9T | II6.9 I 308.91 889.91 || $19.9L ZI
897. g|I|| 878. g|I|| 833.9 I | ZII.9 I | 966. FI | 088. FI #9/..? I | 159. WI Ogg. WI | 3Liſ. FI II
96%. 71 || 61, I. WI 390. WI 776.8 I | 938.8L | 802.3T | 069.8L | I].7.8'I 398.8T | 333.8L OI. I
fºLI.3.I | 766. ZI | #18.3.I fºg/.3I 789. &I | PIG. ZI 869. &I 3/3. &I Igl. &I | 030. &I 60
606. II | 1.8/. II 999. II | 879. II | I37. II : 86%. II | 9./I. II || 390. II | 636.0I g08. OI 80
I89.0 I | 199. OT | 88ſ. OI | 608.0T | 78T. OT || 690. OT || 786.6 || 608.6 || 789.6 | 899.6 /0
387.6 908.6 6/I-6 || 390.6 | 936.8 || 86/.8 I/9.8 gi’9.8 || gliº. 8 || 183.8 90
89 I-8 | 660.8 006. / | IAA. A 379. A &I9. 1 || 388. 1 || 393. A 3&I. A | I66.9 90. I
| 098.9 63/.9 || 869.9 997.9 || 788.9 || 303.9 || 0/0.9 || 136.9 | #08.9 || I/9.9 #0
189.9 307.9 || 693.9 g8I.9 IOO. g . 998. W Ig/. W 969. V | [97. W gag. # 90
68I.j | 890. 7 || AI6.8 || 08/.8 || 879.8 || 909.8 198.8 633.8 I60.9 || 396.3 30
#18.3 g/9.3 989.3 | 1.68. & | 1.9%. 3 || 1 II. & 9/6. I 988. I | 769. I ggg. I [0
&I?. I 0/3. I | 8&I. I | 986.0 | 878. 0 || 001.0 | 1.99.0 | ? If.0 0.1%. 0 | 9&I.0 00. I
8I0.0 - 66.0
6. 8. 1. 9. 9. #. 3. 3. 1. 0. §
!
|
‘sear:30p 9tuneq (, Ieuope?I,
17
Oluſ ‘ā 4t ‘soº!Aejo olypods Jo uois.19Auoo out, Joj aſqa L
SISXTVNV TVoINHO'HL NI (IHSn SCIOHL'HW Tvad Nāo 091
BAUME HYDROMETERS
161
Table for the Conversion of Specific Gravities—Continued.
Sp. e º º º © º e g º &
gr. 0 I 2 3 4 5 6 7 8 9
1 “40 41-318 41-392 || 41-466 || 41539 || 41-612 || 41°685 || 41°758 || 41-831 || 41-904 || 41'977
41 42-049 || 42°122 || 42°194 || 42°266 || 42°338 || 42°410 || 42°482 42°554 || 42-626 42°698
42 42-769 || 42-840 || 42°912 42°983 || 43°054 || 43°125 || 43-196 || 43°267 || 43-338 || 43°409
43 43-479 43°550 43-620 || 43°690 43°760 || 43-830 || 43.900 || 43-970 || 44-040 || 44*110
44 44-179 || 44'248 44°318 44-387 44'456 || 44°525 || 44°594 || 44-663 44-732 44'801
1 °45 44-869 44-938 || 45.007 || 45-075 45-143 || 45-211 || 45-279 || 45-347 || 45-415 || 45°483
46 45-551 45-619 || 45 °687 45°754 45'821 || 45°888 45°955 46-022 || 46'089 || 46°156
47 46.223 46-290 || 46°357 || 46'423 || 46°489 || 46-555 46.621 46-687 || 46°753 46°819
48 46'885 || 46.951 || 47:017 || 47-083 || 47-148 || 47-213 || 47-279 || 47-344 47-409 || 47°474
49 47-539 47-604 || 47-669 47-734 47-799 || 47-863 47-928 47 '992 || 48:05.6 48*120
1°50 48.184 48'248 || 48-312 || 48-376 || 48-440 || 48-503 || 48-567 || 48-631 || 48-694 || 48°757
51 48.820 || 48.884 || 48-947 49-010 || 49-073 || 49.136 || 49-199 || 49-262 || 49:325 || 49-387
52 49-449 || 49-512 49°574 49-636 || 49-698 || 49.760 49°822 || 49°884 49°946 50°008
53 50-069 || 50-131 || 50-193 50-254 50-315 || 50-376 50°437 50°498 || 50:559 50-620.
54 50-681 || 50-742 50-803 || 50°864 50°924 || 50°984 || 51-045 || 51°105 || 51-165 || 51*225
1 *55 51 °285 51°345 || 51 “405 || 51°465 51525 51-584 || 51-643 || 51-703 || 51-763 || 51°822
56 51°881 || 51°940 || 51°999 || 52.058 || 52°117 || 52°176 52°235 | 52°294 || 52°353 || 52°411
57 52°469 52-528 52.587 52-645 52-703 || 52-761 52°819 52.877 52.935 | 52°993
58 53-051 53-109 || 53°167 || 53-225 53°282 || 53-339 53-397 || 53-454 53°511 53°568
59 53-625 || 53-682 53°739 53-796 || 53-853 || 53-909 || 53°966 || 54-023 54-079 || 54°135
1 '60 54 * 191 54°248 || 54°304 54'360 54'416 || 54'472 54°528 || 54°584 54°640 54°696
61 54°751 || 54'807 || 54°863 || 54°918 54°973 || 55-028 || 55°083 || 55°138 55-193 || 55 '248
62 55-303 || 55 °358 || 55-413 55-468 55-523 55-577 55-632 55-687 55.742 55-796
63 55'850 | 55-904 || 55'958 56-012 || 56-066 || 56°120 56°174 56°228 56-282 56°336
64 56-389 || 56-443 56°497 56-550 56-603 || 56-656 || 56-709 || 56-763 56-816 || 56°869
1 *65 56-922 56.975 57-028 57-081 57-134 || 57-186 57-239 57 292 57-344 57°396
66 57-448 || 57-501 || 57-553 || 57-605 || 57-657 || 57-709 || 57-761 || 57-813 57.865 57.917
67 57 °968 || 58-020 58'072 58'124 || 58-175 || 58-226 58°278 || 58-329 || 58-380 || 58°431
68 58°482 58°533 || 58°584 58-635 58-686 || 58-737 58°788 58-839 58-890 58'940
69 58°990 || 59:04.1 59'092 || 59'142 59-192 || 59°242 59°292 || 59°342 59°392 || 59°442
1-70 59°492 || 59°542 59°592 59-641 59-691 || 59-741 59°791 || 59-840 59-890 59°939
71 59°988 || 60-038 || 60-087 | 60° 136 60-185 || 60°234 || 60°283 || 60°332 60-381 60°430
72 60°478 60°527 | 60-576 (20-625 | 60-673 || 60-721 60-770 60-818 60-866 60°914
73 60°962 61-010 || 61-058 61°106 || 61°154 || 61.202 || 61°250 | 61°298 || 61°346 || 61°394
74 61°441 61°489 || 61°537 61-585 61-632 || 61-679 61-727 | 61-774 61-821 | 61°868
1-75 61-915 61.962 62-009 || 62.056 62.103 || 62-150 | 62°197| 62-244 62.291 | 62-337
76 62-383 || 62°430 62°477 62°523 62°569 || 62-615 62-662 62-708 || 62-754 62-800
77 62-846 62°892 || 62.938 || 62.984 || 63-030 || 63-075 63°121 | 63-167 63.213 || 63-258
78 63°303 || 63-349 63-395 || 63-440 | 63-485 || 63-530 | 63°576 63-621 | 63-666 || 63-711
79 63-756 63-801 || 63-846 63-891 || 63.936 || 63-980 || 64°025 | 64-070 64-115 64°159
1-80 64-203 || 64'248 || 64°293 || 64’337 | 64'381 || 64°425 | 64°469 || 64°514 | 64°558 || 64 -602
81 64°646 64-690 | 64*734 || 64-778 || 64'822 || 64-866 || 64°910 || 64-954 64°998 || 65-041
82 65°084 || 65°128 65-172 || 65.215 65’258 || 65-301 || 65°345 || 65°388 65.431 || 65'47.4
83 65°517 65-560 || 65 '603 || 65-646 65-689 || 65 731 || 65-774 || 65°817 || 65-860 || 65-902
84 65°944 65°987 | 66-030 | 66-073 | 66°115 | 66-157 | 66°200 | 66.242 | 66.284 | 66-326
1-85 66°368

162 GENERAL METHODS USED IN TECHNICAL ANALYSIS
Other forms of hydrometer include the following:—
Cartier's hydrometer, in which 21° corresponds to 22° B., on the instru-
ment for liquids lighter than water; above and below this point, 16° B.
= I 5° Cartier.
Beck's hydrometer, in which the point to which the instrument sinks
in pure water is taken as zero, and that to which it sinks in a liquid of
specific gravity O.850 as 30°; one-thirtieth of the intermediate distance is
taken as I?, and graduations of this value are made both above and
below the zero. -
In Baumé’s hydrometer for liquids lighter than water, the point to
which the instrument sinks in a solution of one part of salt to nine parts
of water is taken as O’, and that to which it sinks in pure water as IO’.
The distance between these points is divided into ten equal parts, and
the graduations continued to the top of the spindle, the zero being
placed at the bottom of the stem.


Degrees, Baumé. Cartier. Beck. Degrees. Baumé. Cartier. Beck.’
Baumé, Baumé,
Cartier, - Cartier,
and Beck. Sp. gr. Sp. gr. Sp. gr. and Beck. Sp. gr. Sp. gr. Sp. gr.
O 1*0000 36 0°8488 0°8439 0 °8252
1 0*994.1 37 0 °8439 0-8387 0°8212
2 0°9883 38 0 °8391 0 °8336 0-8173
3 0-9826 39 0 °8343 0 °8286 0°8133
4 0-9770 40 0 °8295 tº º º 0°8095
5 0°9714 41 0°8249 - - - 0°8061
6 0-9659 42 0-8202 - e. e. 0-8018
7 0°9604 43 0-8156 • * > 0°7981
8 0°9550 44 0-8111 - - e. 0°7944
9 e e e e e Q 0-9497 45 0-8066 - - - 0-7907
10 1*0000 • * * 0*9444 46 0-8022 • * * 0-7871
II 0 °9932 1°0000 0°9392 47 0-7978 • * * 0°7834
I2 0 °9865 0 °9922 0*9340 48 0 °7935 - - - 0-7799
13 0 °9799 0 °9846 0 °9289 49 0-7892 e tº e 0-7763
14 0°9733 0°9764 0-9239 50 0°7849 tº e º 0-7727
15 0 °9669 0°9695 0-9189 51 0°7807 - - - 0 °7692
16 0*9605 0-9627 0°9139 52 0-7766 tº e e 0°7658
17 0*9542 0°9560 0-9090 53 0-7725 - * > 0 7623
18 0'9480 0*9493 0 °9042 54 0°7684 tº º º 0-7589
19 0*9420 0°9427 0-8994 55 0.7643 - * * 0-7556
20 0-9359 0°9363 0°8947 56 0-7604 - © e 0-7522
21 0 '9299 0°9299 0-8900 57 0-7565 - © tº 0-7489
22 0 °924.1 0°9237 0 °8854 58 0-7526 tº º º 0-7456
23 0 °9183 0-91.75 0°8808 59 0.7487 - - - 0-7423
24 0 °9125 0°9114 0-8762 60 0.7449 e - e. 0-7391
25 0 °9068 0*9054 0-87.17 61 e - e. - e. e. 0°7359
26 O'9012 0 °8994 0-8673 62 tº º & tº º e 0-7328
27 0.8957 0 °8935 0 °8629 63 • * * - - tº 0-7296
28 0-8902 0.8877 0.8585 64 • * * ... " 0-7265
29 0 °8848 0 °8820 0 °8542 65 tº º e - tº º 0°7234
30 0-8795 0-8763 0-8500 66 tº º º • * tº 0-7203
31 0-8742 0-8707 0-8457 67 © tº e * - e. 0.7173
32 0-8690 0-8652 0°8415 68 tº º e & a tº 0°7142
33 0°8639 0°8598 0°8374 69 e tº º • * * 0-7112
34 0.8588 || 0:8545 0 °8333 70 tº e & tº 0 tº O'7083
35 0°8538 0°8491 0°8292
DETERMINATION OF SPECIFIC GRAVITY 163
A comparison of these hydrometers with the specific gravity, for
liquids lighter than water, at 12°.5, is given in the preceding table.
If the strength of a liquid, containing a dissolved substance, is to be
determined by means of its specific gravity, the nature of all the consti-
tuents present must be known, and no foreign matter, which could affect
the density, should be present. A qualitative chemical examination
should therefore be made first, if necessary.
Hydrometers in actual use should be checked by comparison with
reliable standard instruments, graduated in real specific gravities; if con-
siderable exactitude is required, the range of specific gravities on the
standards should be distributed over an extended series of up to nine-
teen spindles.
G. Müller tried to avoid this by constructing a “differential hydrom-
eter,” which is provided with glass weights, by means of which the
value of the graduations is altered in a definite manner; Lunge is of
opinion that instruments of this character do not give satisfactory
results.
For the determination of the density of solid substances, the
volumemeters of Schumann,” Michaëlis, Thörner, and others are
used.”
A very simple and quick method for the determination of the
specific gravity of liquids consists in weighing the quantity delivered
by an accurate pipette; the liquid must, of course, first be brought to
the desired temperature. Ruber 4 weighs the liquid in the pipette for
this purpose; the pipette is not graduated for delivery, but so that it
contains exactly 20 c.c. from the point to a mark on the stem. More
exact results are said to be obtained with pipettes surrounded by a
Weinhold-Dewar vacuum jacket, but the necessary adjustment and
manipulation is then less simple.
An instrument for determining the specific gravity of small
quantities of liquid, called an Areo-pyknometer, has been devised by
Eichhorn.” It is a hydrometer which contains a bulb, provided with a
stopcock, between the spindle and the gravity-bob ; the bulb holds
Io c.c., and is filled with the liquid under investigation. If the instru-
ment is then immersed in distilled water at 17°. 5 C., the specific gravity
is indicated directly on the scale. A similar instrument has been
described by Rubenstorff."
Special hydrometers, such as alcoholometers, saccharimeters, acetom-
eters, etc., have long been used in various industries, some of which
are subject to official control. Similar instruments for the hydrometric
examination of mineral acids, alkalis, and salts, made according to the
* Cf. P. Fuchs, Z. angew. Chem., 1898, II, 505. * Chem. Zeit., 1903, 27, 94.
* Z, anal. Chem., 1884, 23, 1887. * Ger. Pat. 49683.
* Cf. “Cements,” p. 698. * Chem. Zeit., 1904, 28, 889.
164 GENERAL METHODS USED IN TECHNICAL ANALYSIS
most reliable tables, have lately been introduced by G. Müller;" they are
not, however, in very general use. Detailed temperature correction
tables for these hydrometers, for nitric acid, sulphuric acid, hydrochloric
acid, and ammonia, and also tables for a large number of liquids and
solutions, have been compiled by P. Fuchs.” The tables of the Imperial
German Standards Commission will be referred to in the section on
“Sulphuric Acid" (p. 347).
The 2nd International Congress of Applied Chemistry (1896)
passed the following resolutions regarding hydrometers:—”
I. The scales should be either in specific gravities, or in degrees
Baumé, Brin, Balling, etc. The relation of these degrees to the specific
gravities to be determing d by an international committee.
2. For liquids with different capillary properties, special hydrometers,
graduated accordingly, should be used; otherwise corresponding correc-
tions must be made.
3. Readings should, as a rule, be taken at the intersection of the
surface of the liquid with the stem of the hydrometer, without regard to
the meniscus (cf. p. I 56, Fig. 56). In the case of opaque liquids, where
this is impossible, the readings of a hydrometer, which has not been
specially calibrated for such liquids, must be corrected to give readings
corresponding to the actual surface of the liquid.
4. Hydrometers should be provided with a centigrade thermometer,
graduated from the zero-point.
5. One end of the spindle should be provided with a mark coinciding
with the final graduation of the scale, in order to be sure that the scale
does not change its position. -
6. The total errors of a hydrometer should not be greater than one
graduation of the scale.
Further regulations were adopted by the 6th International Congress
of Applied Chemistry (1906), which bear chiefly on the construction of
hydrometers; these include, as a general rule, that if hydrometers are
not used at the normal temperatures (15° or 20°C.), the results must be
reduced to the normal temperatures, by reference to the special tables
prepared by the International Congress.”
Göckel" is of opinion that hydrometers should be marked, not only
with the temperature at which they are to be used, but also with the
temperature of the water taken as standard, and also with a statement
as to whether the specific gravities refer to normal pressure (760 mm.)
or to vacuum. The German Standards Commission calibrates hydrom-
eters with reference to water at maximum density and at normal
1 Chem. Zeit., 1898, 22, IO4.
* Z. angew. Chem., 1898, II, 745 and 909. * /ahresber. f. chem. Zech., 1896, II67.
* These tables have not yet been published.
* Z, angew, Chem., 1903, 16, 562.
MEASUREMENT OF PRESSURE - 165
pressure, for instance, #. (760). The weights of hydrostatic balances
such as Mohr's and Westphal's, are, on the contrary, almost invariably
adjusted for #. (760), and cannot, therefore, agree with officially con-
trolled hydrometers. Both hydrometers and pyknometers should also
always be marked with an indication whether they are to be read in
the plane of the surface of the liquid or at the top of the capillary
elevation.
K. MEASUREMENT OF PRESSURE AND OF DRAUGHT : PRESSURE
GAUGES AND ANEMOMETERS
The determination of the pressure of a current of gas is often
necessary in the course of technical work, especially those relatively small
differences of pressure which occur in chimneys, gas-generators, sulphuric
acid chambers, and similar plant; indeed, many chemical processes can
only be rationally conducted with the aid of regular measurements of
this nature. Observations of the differences of pressure, registered by
the instruments in use, are sometimes sufficient, whilst in other cases
the velocity of the current of gas and the quantity of gas traversing the
system in a given time are deduced from the observations by calculation.
Instruments which simply measure the statical difference of pressure
between two adjacent spaces, usually between the inside of an apparatus
and the external atmosphere, are called
Pressure Gauges or Manometers. Those
which measure dynamic differences, that
is, which register the velocity of the
current of gas directly, are called
Anemometers. Only those forms of t |
anemometer which depend on mano-
metric principles will be considered ; for |
the theory of the subject the reader is
referred to text-books on Physics and
Mechanics.
The simplest pressure gauges, which
are for many purposes quite satisfactory FIG. 57.
for measuring small differences of pres-
sure, consist of glass U-tubes (Fig. 57) provided with a scale C; one
limb of the tube is in communication with the current of gas by means
of the tube D, which is fixed through the wall E, whilst the other
limb B is open to the outside air. If the pressure on the left-hand side
of the wall is greater than on the right, the liquid will stand at a higher
level in B than in A, and conversely.


166 GENERAL METHODS USED IN TECHNICAL ANALYSIS
Presssure gauges with a simple vertical tube, as shown in Fig. 57, are
not sufficiently sensitive for the measurement of the draught in flues,
etc., as a pressure of only O, I mm. of water corresponds to a velocity of
1.23 m. per second. The simplest apparatus for measuring such small
differences of pressure, with greater accuracy, is the pressure gauge, with
an inclined side tube, Fig. 58, designed by Péclet. It consists of a wide
vessel A, communicating with a slightly inclined side tube B, provided
with a corresponding scale. The tube is fixed to a vertical board, which
is supported on a horizontal stand provided with levelling screws and a
spirit-level, by means of which it can be brought into a horizontal
position; the angle of inclination of the tube B is then easily deter-
mined. The liquid, either water or preferably alcohol, forms a very long
meniscus in the tube B, the summit of which, if the tube has a diameter
of 2-3 mm., is vertical to the cross-section of the latter, so that the
readings are very easily made. If the tube has an inclination of I in 25,
FIG. 58.
then each mm. on the inclined tube represents a difference of pressure
of # mm. in A; as the scale can be read to O.5 mm., pressures of O.O2
mm. can be measured.
Alcohol is to be preferred to water as the registering liquid, both on
account of its smaller coefficient of friction, and also because it returns
to its original position much more quickly; the pressures must then, of
course, be corrected in accordance with the specific gravity of the
alcohol. If alcohol of O.800 sp. gr. is used, I mm. corresponds to O.8
mm. of water pressure.
The upper portion of the inner walls of the tube should be moistened
before each experiment by inclining the instrument.
As the reduction of the readings from the angle of inclination in B
to vertical pressure is only correct if the upper inner curvature of B is
true, which is seldom the case with glass tubes, it is better to choose A
of such a form that its internal diameter can be accurately measured,
and then to calibrate B empirically. This is effected by delivering
successive equal quantities of liquid, from a pipette, into B and noting
the level after each addition; the height that each reading would
correspond to in A can be easily calculated, knowing the volume of the

DIFFERENTIAL MANOMETERS 167
liquid, and the diameter of A and the value of the scale on B, in vertical
pressure, is thus obtained.
The position of the tube, connected with the pressure gauge, in the
flue, and the angle at which it is inclined to the current of gas, are
very important in all manometric measurements. The statical pressure
is correctly indicated only when the opening of the connecting tube
is at right angles to the direction of the current, as shown in Fig. 57;
the pressure also varies at different points in the cross-section of the
flue.
Accurate determinations of small differences of pressure can also
be made by means of the Differential Manometer. One form of this
instrument, designed by Seger, for the measure-
ment of draught and of pressure, is shown in
Fig. 59.
This consists of a calibrated glass U-tube A,
both limbs of which end in large glass tubes B
and C, of equal diameter. The tube B, A, C is
fastened to a board, which also supports a sliding
scale D, fixed parallel with one limb of A, and
adjustable by means of slits a a and screw-pins & b.
The U-tube is filled with two non-miscible liquids,
of almost the same specific gravity, in such a way that
their point of contact is near the zero of the scale.
Water and aniline or light petroleum spirit and
dilute alcohol may be used, and it is preferable to
colour one of the liquids. For the measurement
of pressures greater than that of the atmosphere,
the limb B, which is provided with a rubber stopper
and glass tube, as shown, is connected by means
of rubber, glass, or metal tubing with the flue or
chimney to be tested, C being open to the atmo-
sphere; if a pressure less than that of the atmo-
sphere is to be measured, C is connected with the
flue and B left open.
The apparatus is fixed vertically on a wall near to where the
measurements are to be taken, and the scale adjusted so that its zero
graduation is at the surface of contact of the two liquids; on then
connecting B or C with the flue, etc., to be tested, a slight difference
of level of the liquids in the upper wide portions of the two limbs will
be produced. This difference is multiplied in the movement of the
surface of contact of the liquids in A, in the ratio of the sectional areas
of B and C, and that of the narrow tube A.
If, for instance, the sectional area of B and C is twenty times
that of A, then a pressure-difference of 1 mm. will cause a displacement

168 GENERAL METHODS USED IN TECHNICAL ANALYSIS
of the surface of contact X of 20 mm., so that very small differences
of pressure are easily observed. . '
The divisions on the scale may be either empirical, if relative
measurements only are required, or they may be made to correspond
to Water-pressure in mm., for more exact determinations; in the latter
case, the graduations are calculated from the above ratio, and the
specific gravities of the liquids used.
The sensitiveness and accuracy of the instrument increases with
the ratio of the sectional areas of the upper vessels to that of the
lower tube, and with the nearness in specific gravity
of the two liquids employed.
König's differential manometer, Fig. 60, depends
on the same principle as the above. The two limbs
of the tube are concentric, which facilitates the read-
ings. A colourless liquid (mineral oil) and a red
liquid (coloured alcohol) are supplied with the appar-
atus. It is filled by removing the inner tube, intro-
ducing about IO c.c. of the red solution from a pipette,
care being taken to avoid wetting the sides of the
tube, until it reaches the point a, and then carefully
pouring on the colourless liquid till it reaches b,
without stirring up the red liquid below. The inner
glass tube, to which a piece of rubber tubing has
been attached, is then inserted as far as c, filled up
to d with the upper liquid, by suction, and then
pushed down into its final position, the liquid being
meanwhile kept at the same height, by keeping the
end of the attached tubing closed. To adjust the
Cº-He instrument, air is carefully blown in from above, so
– that one or two drops of the colourless liquid pass
into the lower red solution, until the surface of con-
tact of the two liquids stands at e. Should this point
be passed, the upper tube must be liſted out as far as e and the
operation repeated. The exact adjustment to the zero of the scale is
effected by means of a stud, after fixing the instrument in its stand.
Neither the instrument nor the stock solutions should be exposed to
direct sunlight. The apparatus is cleaned by rinsing with concentrated
Sulphuric acid, washing with water, and drying carefully. As the
point of contact of the liquids changes with the temperature, it must
be brought to zero before each determination. It is advantageous to
fit the apparatus with a three-way tap, which is ordinarily turned so
that the interior of the apparatus communicates with the external
atmosphere; the instrument is first adjusted to zero with the tap in
this position, before each observation, and then turned through a right
FIG. 60.

DIFFERENTIAL MANOMETERS 169
angle to establish connection with the space to be examined, the
reading taken, and the tap finally turned back to its original position.
Lunge states that this instrument works very well, and that it is
more convenient than Seger's apparatus.
Langen's manometer, modified by Lux, Fig. 61, also depends on
the principle, that if two tubes of unequal diameter communicate, then,
on disturbing the equilibrium, the alterations in
level of a liquid in the two tubes are inversely
proportional to their cross-sections. When the
difference in the cross-sections is sufficiently great,
the alteration of level in the wide tube may be
neglected, and the pressure measured by observing
the narrow tube only. Five modifications of this
apparatus are made, which are arranged for the
measurement of high and low pressures and of
reduced pressures; it is a convenient instrument,
but it is not suitable for measuring very small
differences of pressure.
The velocity of a current of gas can be calcu-
lated from the pressure indicated by a manometer.
According to Péclet,” for a gas at a small excess
of pressure E, the velocity,
where E is the pressure, in terms of a column of
water, indicated by the manometer, B the baro-
metric pressure, also in terms of a column of water
(i.e., the height of the mercury barometer × 13.59)
f, the temperature, and 6, the density of the gas.
The value of p varies with the form of the aperture through which
the gas escapes. For example, for atmospheric air at O’ and 760 mm.
barometric pressure (i.e., B = IO-334 m.) the relations between the
velocity V, in metres per second, and the excess of pressure E, in
metres are :-
E. I O- I O-OI O-OO I O-OOOI O-OOOOI 11].
V 1 17.61 38.763 I 2.283 3.895 I-232 O-3895.2 m.
An excess of pressure of Tºra mm. indicates, accordingly, a velocity of
almost O-4 metres per second.
For holes in thin plates the value for p is O-65; for cylindrical aper-
tures, O.83; for convergent, conical apertures, O.83 to O-65, according
to the angle; and for apertures which first contract a little and then
'J. Gasóeleucht, 1890, 33, 217 ; 1891, 34, 288. * Ser. physique industrielle, 1, 249.

170 GENERAL METHODS USED IN TECHNICAL ANALYSIS
expand towards the outside (Fig. 62), the relation to the angle a is as
follows:—
O O O
Value of a o I 3 5° 7° 9° 20° 30° 50°
53 ºf I-OO I-24. I-70 2.25 2.45 I-95 I-3o 1.18 1-o;
O
Thus the maximum is at 7° C., and is four times as great as the
velocity from an opening in a thin plate.
When a manometer is connected to the interior of the current of gas
by a cylindrical tube, with an opening placed vertically in the direction
of the current, as is very usually the case in practice, the value of q,
may be taken as = I, and accordingly the velocities, in metres per
Second, are given directly by the above table.
It is to be borne in mind that the velocity varies in different parts of
the cross-section of the chimney, and therefore, in order to obtain a
º
|
|
FIG. 62. - FIG. 63.
mean value, a series of readings should be taken by moving the tube D
(Fig. 57, p. 165) backwards and forwards; from such determinations
the volume of gas passing up the chimney can be calculated, by
multiplying the average velocity by the Sectional area.
The dynamic method of measuring differences of pressure, as
initiated by Péclet in his differential anemometer, consists in placing
both limbs of the manometer in communication with the current of gas,
the opening of the one connecting tube being placed vertically to, and
that of the other parallel to, the direction of the current. In this
case it is not the pressure-difference between the current and the outside
air which is measured, but that between the static and dynamic pressure
in the current itself. The method is very delicate, and serves for the
direct measurement of the velocity of the current.
Fletcher has constructed an anemometer on this principle, which has
been improved and simplified by Lunge.


ANEMOMETERS 171
It consists of two glass or brass tubes a and b (Fig. 63), which are
fixed air-tight, by means of a cork, into a suitable opening in the
channel or chimney, in which the velocity of the gas is to be measured,
so that their ends are rather less than one-sixth of the distance of the
diameter of the flue from its inner wall. The tube a, with the straight
end, must be fixed as exactly vertical as possible to the direction of the
current, whilst the tube b must be so fixed that the gas blows straight
into its open end. These tubes communicate, by means of rubber
tubing, with the U-tube ca, which is half-filled with ether. The current
of gas produces suction at a and pressure at b; consequently the ether
rises in d and falls in c. The difference in level between the ether
columns in c and d is measured by means of a millimetre scale and
vernier; this is called the “anemometer reading.” By turning the
sliding disc e through 180°, a is connected with c and b with d, the
currents being thus reversed, an equal difference of levels, but in the
opposite direction, should be recorded, which serves as a check on the
first reading and should be identical with it.
The following tables serve for the calculation of the velocity of the
current from the anemometer readings.
I.—Table for the Reduction of the Anemometer Readings to
Velocity of Current, expressed in feet per second.
Column a gives the anemometer readings in inches; column & the velocity in feet

per second at a temperature of 15° C. =60° F., and barometric pressure
760 mm. = 29-92 inches.
0. b. 0. b. 0. b. 0. b.
Inches. Feet per sec. Inches. | Feet per sec. Inches. Feet per sec. I Inches. | Feet per sec.
•01 2°855 •16 11°42 •32 16°15 •95 27-83
•02 4°038 •17 11-77 "34 16*65 1°00 28°55
•03 4°945 • 18 12°11 •36 17-13 1°25 31°93
•04 5°710 •19 12°45 •38 17-60 1°50 34°97
•05 6°384 •20 12-77 *40 18°06 1°75 37-77
•06 6°993 •21 13-08 *45 19°15 2°00 40°37
•07 7°554 •22 13°39 •50 20*18 tº e tº e e e
•08 8-075 •23 13-70 •55 21-17
•09 8°565 •24 13°99 •60 22°12
*10 9°028 •25 14°28 '65 23°02
•11 9°469 •26 14°56 •70 23-89
•12 9°891 •27 14°84 •75 24-73
•13 10-29 •28 15-11 •80 25°54
*14 10°68 •29 15°38 •85 26°32
•15 11°06 •30 15°64 ‘90 27-08
172 GENERAL METHODS USED IN TECHNICAL ANALYSIS
II.-Table for the Reduction of the Anemometer Readings to
Column a gives the anemometer readings in millimetres ; column & the velocity in
Velocity of Current, expressed in metres per second.
metres per second at a temperature of 15° C., and barometric pressure 760 mm.


0. b. 0. b. Q. b. (t. b. Ol. b. 0. b.
II]. In IIl. In Iſle II]. In In II]. In Iſl. II] . Ill Iſl. II] . III Iſh. IIl.
O-1 0-575 || 1:4 || 2:040 || 2-7 || 2-833 || 5-0 || 3:855 10-0 5'452 || 19-0 7.5.15
0-2 || 0-771 || 1 °5 2-111 || 2:8 2°885 || 5-2 || 3'931 || 10-5 5'586 || 20-0 7-710
0-3 || 0-944 1-6 || 2:181 || 2:9 || 2:935 || 5-4 || 4-006 || 11-0 || 5’718 21 7-900
0°4 1-090 || 1-7 2.248 || 3-0 | 2.986 5-6 || 4-080 11-5 5°846 22 8-086
0°5 1-205 || 1 '8 2-313 || 3-2 || 3:077 || 5-8 4°152 12-0 5°972 || 23 8°268
0°6 1-341 || 1 °9 || 2:376 || 3-4 || 3-179 || 6-0 || 4-223 12-5 6°095 || 24 8°446
0-7 1°442 2°0 2.438 || 3-6 || 3:271 || 6-5 || 4-395 13-0 6°216 || 25 8-620
0-8 1°560 || 2:1 2°498 || 3°8 || 3-361 7-0 || 4°561 13°5 6°334 30 9 *443
0-9 1-636 || 2:2 || 2:557 || 4-0 || 3:448 || 7-5 || 4-721 || 14-0 6°450 || 35 10-199
1-0 1-724 || 2–3 || 2-615 || 4-2 || 3:569 || 8-0 || 4°876 15-0 6'677 || 40 10*903
1°1 1-808 || 2°4 2-671 || 4-4 || 3-616 || 8°5 || 5°026 16-0 6'896 || 45 11°565
1.2 1°889 || 2°5 || 2:726 || 4-6 || 3:698 9-0 || 5-172 17-0 7-108 || 50 12:190
1-3 1°966 2°6 2-779 || 4-8 3-777 || 9-5 5-314 18-0 7-314 º e ſº tº tº º
III.-Table for Correction for the Temperature at which the
Anemometer Readings are made, to 15° C. = 60° F.
Column a gives the observed temperature ; column 6, the factor by which the values
in column 6 of Tables I. and II. must be multiplied, to give the correct
velocity.
0. b. e b. 0. b. (l. b. 0. b. a. b.
tº C. tº C. tº C. tº C. tº C. tº C.
– 10 1*046 18 0°995 42 0-956 66 0-922 || 140 || 0-835 | 260 || 0-735
– 5 1*036 20 O'991 44 0 °953 || 68 0-919 || 150 0°825 || 270 0-728
0 1-027 22 0 °988 46 0-950 || 70 0-916 || 160 | 0°815 || 280 || 0-721
2 1*022 24 0°985 48 0 °947 75 0-912 || 170 0-806 || 290 || 0-715
4 1*020 26 0-981 50 0 °944 80 O'903 || 180 0-797 || 300 0.709
6 1*016 28 0-978 52 0 °941 85 0-899 || 190 0-788 || 320 | 0-697
8 || 1 °012 30 0.975 54 0 °938 90 || 0 890 || 200 || 0 °780 || 340 0-685
10 1°009 || 32 0-972 || 56 0 °935 | 95 0-884 || 210 || 0-772 || 360 0.676
12 1°005 34 0-968 58 0 °933 || 100 O'878 || 220 || 0-764 || 400 || 0-654
14 1°003 36 0-965 60 0 °930 || 110 0-867 || 230 || 0-756 || 450 || 0-631
15 1°000 8 0°962 62 0 °927 | 120 0°856 || 240 || 0-749 || 500 || 0-603
16 0-998 40 O'959 64 0 °924 || 130 0.845 250 || 0-742 gº tº º $ tº º


L. MEASUREMENT OF TEMPERATURE
The treatment of this subject is restricted to certain general con-
siderations and to the description of a few typical instruments for the
measurement of temperature.
For further details the reader
is
MEASUREMENT OF TEMPERATURE 173
recommended to consult text-books on Physics and the following
works:—
High Temperature Measurements, H. le Chatelier, trans, by G. K.
Burgess, 2nd edition, 1904; Introduction to the Study of Metallurgy,
Sir W. Roberts-Austen, 5th edition, 1902; the article on “Thermom-
etry” in the Dictionary of Applied Chemistry, by Thorpe and Muir;
Chemische Technologie der Brennstoffe, F. Fischer (1880-1897), and
Chemisch-technische Analyse, Post; the subject is also dealt with in the
section on “Clay,” p. 568.
Temperature measurements may be divided into the two groups,
Thermometry and Pyrometry; the first includes the measurement of
temperatures for which a mercury thermometer can be used, the
second, high temperature measurements. Recent developments of
technology may now also necessitate the measurement of temperatures
below the scales of the mercury or alcohol thermometer.
Thermometry.—The mercury thermometer is almost invariably
used for thermometric purposes in technical work; alcohol thermometers
are used occasionally, but necessarily only for temperatures up to about
50°, and they are not to be regarded as very exact. With ordinary mercury
thermometers, the limit of temperature that can be measured is at
most 300°, owing to the unavoidable change in the capacity of the
bulb, as the boiling point of mercury (360°) is approached. By filling
the upper space of the thermometer tube with nitrogen, under pressure,
and by the selection of suitable glass, mercury thermometers are made
which can be used up to 550°; very long, ordinary thermometers, placed
vertically, so that the mercury column exerts considerable pressure,
may be used up to and above 360°.
Thermometers are very seldom checked by physical methods, in
technical laboratories; it is usually sufficient to compare them with
one or more standard instruments; such comparison is especially
desirable when thermometers have been exposed to high temperatures
for long periods. Corrections for the unexposed portion of the mercury
Column are not usually applied in technical work.
It is often necessary to employ special forms of thermometers for
technical purposes, particularly instruments in which the bulb is
situated at some distance from the scale; thermometers with bent
tubes, etc., are also necessary at times. The bulb and tube must
often be protected by a sheath, so as to guard against mechanical
injury or fracture through sudden heating. If a perforated metal tube
or other similar arrangement is used for this purpose, the thermometer
will indicate the temperature fairly accurately, but if an unperforated
sheath has to be used, the readings are always too low, even if the
sheath is filled up with mercury, copper filings, etc., as should always
be done. Some mechanical protection is afforded by wrapping asbestos
174 GENERAL METHODS USED IN TECHNICAL ANALYSIS
round the exposed portions of the thermometer, but this considerably
decreases the sensitiveness.
In order to ascertain that the temperature to be measured is not
too high for the use of a mercury thermometer, a very practical
suggestion, due to Weinhold, is to first place a piece of lead at the spot
where the temperature measurement is to be made; the melting point
of lead is 323°, and if the test-piece does not show any softening at the
edges, even after some time, the mercury thermometer can be safely
used. *.
In cases where frequent readings are not required, and no small
rapid variations of temperature need to be observed, Weinhold
recommends the use of the thermometer block (Fig. 64).
4% This is made of cast iron or of copper, with an obliquely
ºf; -
bored cavity h, which is half-filled with mercury, and
§ : > provided with a perforated knob #; it is introduced into
A the flue, etc., to be tested, through a suitable opening, by
N. means of wire passed through the slit in the knob. The
block is left in position for ten to fifteen minutes, then
quickly withdrawn, and a small thermometer, which has
previously been warmed to about the required tempera-
ture, at once placed in the cavity h; the reading is taken
when the thermometer reaches the highest point and
* the mercury begins to fall.
Many forms of recording thermometers, for automatically registering
temperatures in drying chambers and the like, at a distant point, are in
use; these all depend on the electrical transmission of the readings.
Pyrometry.—The following methods are in practical use for high
temperature measurements:—
I. High temperature Thermometers. Mercury thermometers, filled
with nitrogen under pressure, may be used up to 550°, as stated
above. . - *
Kühn has drawn attention to the fact that works' thermometers,
made for use up to 550°, which are often several metres long, are
graduated for immersion to a definite position, which should always
be adhered to in making observations. Also, that they may register
up to 30° too high, or from 50° to 200° too low, after being used for
some time; the former error is due to a contraction of the bulb,
arising from the instrument having been insufficiently “aged ” after
manufacture. Low readings arise from the expansion of the glass,
which may be caused by overheating, due to the thermometer not
having been properly adjusted to the right position, when in use. When
thermometers are inserted to other points than those for which they are
calibrated, a correction must be applied. Thermometers in which the
* Chem. 2eit., 1903, 27, 54.


PYROMETERS 175
mercury is replaced by a low-melting alloy of potassium and sodium
are also used for high temperature measurements, but they are
expensive, and very breakable.
2. Eapansion Pyrometers. The expansion of solids is mainly made
use of in metallic pyrometers. One of the most usual forms, which can
be used up to 900°, is that of Gauntlett and Desbordes," in which the
difference in expansion of a cast-iron and of a copper tube is measured.
Steinle and Hartung's graphite pyrometer is a conjunction of graphite,
which is assumed not to expand on heating, and metal; it is graduated
up to 1200°, and was formerly generally used in Germany.
The coefficient of expansion of metals alters after frequent and
prolonged heating, and therefore no expansion pyrometers are really
reliable; they must accordingly always be compared with a standard
instrument at frequent intervals.
The expansion of vapours and the measurement of the resulting
pressure is made use of in the Thalpotasimeter devised by Klinghammer.
This consists of a narrow tube, bent into the form of an S above, and
closed below. The whole of the shorter, and two-thirds of the longer
limb are filled with a liquid, the vapour pressure of which is measured
on a manometer, thus indicating the temperature. The liquids used
are liquid carbon dioxide for temperatures between —65° and 12°.5;
liquid sulphur dioxide, for from — Io” to + IOO’; dry ether, for from
+35° to 120°; distilled water, for from Ioo” to 226°; high-boiling
petroleum oil, for from .216° to 360°; and mercury, for from 357°
to 780°.
3. Air Pyrometers. Many forms of air pyrometer are in use, in
which the pressure caused by the expansion of air is measured. The
instruments used by physicists are much too complex for technical
purposes, for which, however, special forms have been devised, notably
those of Wiske,” F. Fischer.” Knöfler,” and Heisch and Folkard.”
A distinct variety of air pyrometer comprises those in which the
difference of pressure between ordinary and heated air is measured, as in
the instruments of Wiske" and of Wiborgh." The latter is provided with
a correction for the atmospheric temperature and pressure, and gives
accurate results, but a special, though simple manipulation is necessary
before each observation, so that it is not suitable for continuous
readings.
Another form of air pyrometer is that of Victor Meyer and Riddle,”
in which the air is expelled from a heated platinum or porcelain vessel
* Made by Schäffer & Budenberg, Manchester. * Z. angew. Chem., 1890, 3, 591.
* Ger. Pat. 9681. * Ger. Pats. 4oo81 and 43603.
* Cſ. Hurter, J. Soc. Chem. Ind., 1886, 5, 634. * Ger. Pat. Ioof 5.
* Ibid., 43958; J. Soc. Chem. Ind., 1889, 8, 214; Stahl, u. Aisen, 1891, 11, 915.
* Ber, 1893, 26, 2443 and 3Ioo'; 1894, 27, 3129.
176 GENERAL METHODS USED IN TECHNICAL ANALYSIS
by means of hydrochloric acid gas, and measured, after absorption of the
latter.
4. Fusion point Pyrometers. Temperature determinations, by obser-
vation of the fusion of alloys, were first made by Prinsep in 1828, and
have been extended by others, many special forms of apparatus having
been constructed for this purpose. For high temperatures, the non-
oxidisable metals, silver, gold, platinum, palladium, and their alloys, are
used, the melting points of which have been determined by Erhard and
Schertel, Holborn and Wien, Violle, Heycock and Neville, Harker, and
others.
The alteration of clay by shrinking is the basis of the oldest pyrom-
eter, that of Wedgwood (1782), which for nearly a century was the
only guide for the measurement of high temperatures; it is now com-
pletely replaced by more reliable methods, of which the standard
“cones” introduced by Seger! (1886) are the most important. These
are known as “Seger's fusible comes,” and are largely used, especially in
the ceramic industries. They are small triangular pyramids, composed
essentially of quartz, sand, felspar, calcium carbonate, and kaolin, the
proportions of which are adjusted so as to give a series of fusion points'
at intervals of 20° to 30°. The fusion points of the mixtures, as deter-
mined by Hecht” by comparison with the Le Chatelier electric pyrom-
eter, are given in the following table:—

• Tem- * Tem- Tem- - Tem- * Temp.
. perature. . perature. ..". perature. ... perature. ... I)
°C. °C. °C. °C. °C.
022 590 010 950 3 1190 15 1430 27 1670
021 620 09 970 4. 1210 16 1450 28 1690
020 650 08 990 5 1230 17 1470 29 1710
019 680 07 1010 6 1250 18 1490 30 1730
018 710 06 1030 7 1270 19 1510 31 1750
0.17 740 05 1050 8 1290 20 1530 32 1770
016 7.70 || 04 1070 9 1310 21 1550 33 1790
015 800. 03 1090 10 1330 22 I570 34 1810
014 830 02 1110 11 1350 23 1590 35 1830
013 860 01 1130 12 1370 24 1610 36 1850
012 890 1 1150 13 1390 25 1630 - © & ſº tº º
011 920 2 1170 14 1410 26 1650
The standard cones, in sets of fifty-nine, which cover temperatures
between 590° and 1890°, are made by the Royal Berlin Porcelain Works.
Each cone is stamped, and a selection, according to the temperature to
be measured, is placed in sequence, on a tile, at a point in the furnace
which can be observed from the outside. The order of fusion is then
noted, the final point being taken as the last of the series which is
* / Soc. Chem. Ind., 1886, 5, 489; 7honindustrie Zeit., 1886, Io, 135 and 229.
* 7 honindustrie Zeit., 1896, No. 18.
PYROMETERS 177
sufficiently fused for the summit of the cone, which being the thinnest
portion is always softened first, to bend over so that it touches the tile;
the fusion point of this cone is taken as the temperature of the furnace.
Kochs and Seyfert" have given “estimated" values for the tempera-
tures of fusion of Seger cones, which are considerably higher than those
of Hecht for the higher temperatures, but these can hardly be considered
to be so reliable. Heraeus * determined the fusion point of cone
number 36= 1785° and number 37 = 1800° C.; that of corundum =
1865°.
Cone 26 is only fused by the highest temperatures which have
so far been attained in technical furnaces, with the exception of electric
furnaces; it corresponds to the minimum melting point of “fire-proof”
clays. The cones 27-36 go up to the melting point of platinum, and
are mainly used for determinations of the fire-resisting properties of
clays.
Wiborg's Thermophone * is constructed upon an entirely different
principle. It consists of an explosive, surrounded by a fire-proof sub-
stance which is a bad conducter of heat, and is made up in the form of
cartridges. If one of these is introduced into a furnace, the time which
elapses before it explodes will vary with the temperature of explosion of
the contained substance, the thickness of the covering, and the surround-
ing temperature. The latter is estimated by observation of the time
interval between the introduction and explosion of the cartridge, and
comparison with an empirically prepared table.
5. Optical Pyrometers.” Optical phenomena have been made use of
in a variety of ways for pyrometrical purposes—either roughly, by
observation of the colour of incandescent matter and comparison with a
standard glass, or more accurately, in association with the spectroscope
or polariscope. The “lunette pyrométrique” of Le Chatelier is fairly
frequently used in France, but it cannot be regarded as an accurate or
reliable instrument.
Hempel” has constructed a spectroscopic apparatus for high
temperature measurements, depending on the principle that for one
and the same substance the length of the spectrum increases with the
temperature. -
Wanner's optical pyrometer" depends on the photometric comparison
of polarised light from a small electric incandescent lamp and the
* Z, angew. Chem., 1901, 14, 721. * Chem. Zeit. Rep., 1902, 303.
* Eng. Pat. 6284, 1895; J. Soc. Chem. Ind., 1896, 15, 377. The cartridges are supplied by Dr
H. Geissler, Nachfolger, Bonn.
* A valuable and full monograph on Optical Pyrometry, by C. W. Waidner and G. K. Burgess,
has been published by the U.S. Bureau of Standards, Bulletin No. 2, 1905, pp. 189-254.
* Z, angew. Chem., 1901, 14, 237.
* Ibid., 1902, 15, 715; cf. also, W. Feld, Chem. Ind., 1903, 23, 256. The pyrometer is supplied
by Messrs Townson & Mercer, London. -
M
178 GENERAL METHODS USED IN TECHNICAL ANALYSIS
furnace, etc., under observation. The instrument is arranged like a
telescope; it is easy to manipulate, and is strongly to be recom-
mended for estimating temperatures above I5OO° C.; Le Chatelier's
instrument is useless above this temperature, and the Wanner pyrom-
eter has the further advantage, that it is better adapted for observing
temperatures at points situated at some distance from the walls of a
furnace. -
In the Féry radiation pyrometer" or thermo-electric telescope the
radiation from an incandescent body is focussed, either by a mirror or
by a system of lenses, on to a small and sensitive thermo-couple; the
electromotive force generated by the raising of the temperature of the
couple is recorded by a sensitive potential galvanometer placed in series
with the couple, the scale of which is divided so as to give direct
temperature readings. The pyrometer is in the form of a telescope, and
is supplied to cover various ranges of temperature from 900° up to 3500°.
The readings are independent of the distance of the pyrometer from
the furnace, and are uninfluenced within wide limits by the size of the
hot body or observation hole under examination. -
6. Calorimetric Pyrometers. Of the many forms of calorimetric
pyrometers for gases, that of Bradbury is designed to measure the
temperature of a hot-air blast, by mixing the gas with nine times its
volume of air; the temperature of the mixture is determined with a
thermometer and multiplied by ten. Values obtained by this method
can only be approximately correct. -
Methods in which a solid is exposed to the source of heat, and the
absorbed heat determined calorimetrically, are much more important.
The method, if correctly employed, gives good results, and it was
formerly used for the control of other forms of pyrometer previous to
the introduction of thermo-electric pyrometry. .
The principle of the methods is as follows:—A small cylinder of iron,
copper, or preferably of platinum, is placed in the furnace in a suitable
receptacle. When it has remained there long enough to acquire the
temperature of its environment, it is removed and dropped as quickly as
possible into a definite quantity of water, the rise of temperature of
which is measured with all the precautions necessary in calorimetric
work. The temperature, T, of the furnace is then found from the
equation:— *
T = a + A*-0
pe
where t = the initial temperature of the water; t = the final tempera-
ture of the water; p’= the “water-equivalent” of the calorimeter, includ-
* Comptes rend, 1902, 134, 977. The pyrometer is supplied by the Cambridge Scientific
Instrument Co., Cambridge.
PYROMETERS 179
ing the contained water; A = the weight of the metal cylinder, and c= its
specific heat at the temperature T.
Various forms of these pyrometers are in use for technical purposes,
in which p and p' are constant, so that a simple factor can be employed
to facilitate the calculations. In Siemens' copper pyrometer there is a
special scale, which gives the temperature corresponding to the
readings on the thermometer directly. Instruments graduated in this
way are, however, only suitable for rough determinations; for more
exact measurements more accurate apparatus must be used, and the
results calculated as above.
7. Electric Resistance Pyrometers utilise the changes in the electric
resistance of metals, with the temperature for pyrometric measurements.
In the Siemens resistance pyrometer, a platinum wire of a very
high degree of purity is wound spirally, in a corresponding groove, round
an earthenware cylinder which is contained in a strong cast-iron tube,
closed below, and is insulated from the tube by means of asbestos. The
spiral wire has a diameter of O-4 mm., and is joined to three thicker
wires, which are of equal thickness; one of these is fused to
one end, and two to the other end, of the spiral wire. These thicker
wires are protected by thin earthenware tubes, which are placed in the
further end of the same iron tube that carries the spiral; only the
portion of the tube containing the spiral is placed in the furnace. The
thick platinum wires are attached by copper wires to a differential
voltameter, which is connected with a fairly strong galvanic battery.
The current of the latter is divided between two circuits, of which each
passes through one of the voltameters, one of them through a known
resistance in addition, and the other through the resistance to be
measured. As the currents in the two circuits are inversely propor-
tional to the resistances, and the quantities of gas generated in the
voltameters are directly proportional to the current strengths, the ratio
of the quantities of electrolytic gas is the inverse ratio of the resist-
ances. From this comparison the temperature is determined by
reference to a specially constructed table.
This pyrometer was for a long time regarded as a very reliable and
standard instrument, and its utility was widely recognised. As the
result, however, of special investigations, it has had to forego this
claim, it having been fully established that it is liable to changes of
zero.1
A modified resistance pyrometer, in which due precautions are taken
to eliminate the sources of error of the Siemens pyrometer, has since
been devised by H. L. Callendar.” In this instrument, the platinum
wire is wound on a mica frame and connected by stouter wires, either to
* Cf. Report by Committee of the British Association, Reports, 1874, 242.
* Phil. Trans., 1887, 178, 161. Phil. Mag., 1891, 32, Ioq ; 1892, 33, 220.
180 GENERAL METHODS USED IN TECHNICAL ANALYSIS
an indicator on which the temperature can be read directly, or to a
recording drum. The resistance wire is enclosed in a porcelain, steel,
or brass tube, according to circumstances, to protect it against fumes;
a zero method is adopted for measuring the resistances with a galvan-
ometer. Rapidly varying temperatures can be readily followed and
measured, and the instruments are constructed for varying ranges of
temperature, between low temperatures and 1400°."
ſº-
FIG. 65.
This instrument has been largely adopted for many forms of
technical pyrometric determinations.
Resistance pyrometers, either with or without their accessory
apparatus, are verified by the National Physical Laboratory.
8. Thermo-electric Pyrometers. Thermo-electricity was suggested
for the measurement of high temperatures by Becquerel (1830) and
* The Callendar pyrometer is made by the Cambridge Scientific Instrument Company,
Cambridge. -


PYROMETERS 181
by Pouillet (1836), but Le Chatelier' (1887) was the first to construct a
really reliable and practical instrument, which can be used for the
measurement of very high furnace temperatures, and also for the
extremely low temperatures which have recently been obtained.
The working portion of the Le Chatelier pyrometer (Figs. 65 and 66)
consists of a thermo-electric couple, formed of a wire of pure platinum,
a, O-6 mm. thick and 1.5 m. long, and a wire, b, of the same dimensions,
made of an alloy of nine parts platinum to one part rhodium. The
junction of the two wires, the temperature of which determines the
strength of the current, must be made without the use of any foreign
material; Le Chatelier unites the wires by twisting them together,
whilst Heraeus and others adopt the more reliable method of fusing
them together. Both wires are insulated by means of porcelain tubes,
c and d, I m. long, which are protected in their turn by the iron tube
ee, as shown in Fig. 66. In a reducing flame, a platinum carbide is
§ (2
S
%
§ gº: §s, &
%
i.e.----—--- 4000 Inm >
FIG. 66.
formed which would alter the electromotive force; the porcelain tubes,
which are glazed outside, prevent this action. The wires are connected
up outside with platinum wires leading to a d’Arsonval galvanometer
(Fig. 65), the needle of which gives readings on two scales. One of
these indicates the electromotive force in volts, and the other the
corresponding temperatures as determined by comparison with an air
thermometer; the readings of the instrument on the former scale can,
therefore, be easily controlled by comparison with a Clark's standard
cell.”
The National Physical Laboratory undertakes the verification of
thermo-junction pyrometers at temperatures between 300° and I3OO”.
Instead of the arrangement of porcelain and iron tubes, shown in
Fig. 65, which is suitable for large furnaces, other forms can easily be
made, adapted for use with crucibles, tubes, and other laboratory
apparatus. One of the great advantages of this instrument is that it
can be used equally well for the measurement of the temperature of
* Comptes rend., 1886, Ioz, 819; Bull. Soc. Chim., 1887, 47, 42.
* The Chatelier and other forms of thermo-electric pyrometers are made by the Cambridge
Scientific Instrument Company, Cambridge.




182 GENERAL METHODS USED IN TECHNICAL ANALYSIS
either large or small masses of material, as it is only necessary to
expose the thermo-junction ab to the temperature to be measured.
The galvanometer may be placed at a considerable distance from
the furnace, etc., without the necessity of making any allowance for the
resistance of the conducting wires—for instance, in the office of the
works-manager—but it must be on a support free from vibration, such
as a wall-bracket. If it is at such a distance from the thermo-couple
that the external resistance of the circuit is essentially greater than
I ohm., this must be allowed for in the calibration; for distances up to
IOO m., insulated copper wire of 2 mm. diameter may be used as the
conducting wire. The thermo-junction and the leads to the galvanom-
eter should be practically at the atmospheric temperature; this can
be effected, either by having a sufficiently long thermo-couple, or by
special cooling. .
Full precautions must be observed in the adjustment and use of the
galvanometer; it must always be accurately centred and perfectly
horizontal, as otherwise the needle will not come to rest at the zero-
point of the scale. If the temperature of the cold junction is higher
than that at which the instrument has been calibrated, the galvanom-
eter scale must be moved from right to left by the number of degrees
corresponding to this difference. Thus, if the junction is at 50° C., and
the calibration has been effected at O’, the temperature usually adopted
in the verification of thermo-junction pyrometers by the National
Physical Laboratory, the needle will, after correct adjustment, come to
rest at the 50° graduation of the scale. As an alternative, this temper-
ature difference may simply be added to the temperature found, with
the galvanometer adjusted to the zero of the scale.
The thermo-couple should be chosen of such length that the
terminals are so far from the furnace that their temperature does not
much exceed that of the atmosphere. For technical purposes the
relatively small difference in temperature is then usually disregarded,
and the galvanometer readings as calibrated are considered to be
sufficiently accurate. The thermo-couple should only be used
without a protective sheath when contact with substances which might
affect the incandescent platinum is excluded ; any luminous flame has
an injurious effect. If either of the wires of the thermo-couple breaks
off, a satisfactory contact can be made by twisting the loose ends
tightly together for a distance of about I cm. The thermo-junction of
the two wires may be repaired similarly. For most technical purposes
the thermo-couple should be mounted in a porcelain tube; one wire of
the couple is then held in a narrow porcelain tube so that the thermo-
junction projects just outside, whilst the other wire passes back along
the outside of the tube, thus ensuring the insulation of the two wires
from one another. The tube may, of course, be made up of several
PYROMETERS 183
sections. A wider porcelain tube, glazed outside and closed at the
bottom, serves to protect the couple from the action of furnace gases,
etc. To protect the outer porcelain tube against mechanical injury and
violent fluctuations of temperature, it is often advantageous to wrap it
in asbestos fibre, and to place it in an outer tube of iron, nickel, or
fire-clay. If metallic protecting tubes are used, care must be taken
that they are not bent by the heating, and thus cause breaking of the
porcelain tubes inside.
If the couple is to be left in the furnace for some time, this method
of protection should always be adopted. No further precautions are
necessary at temperatures below IOOO". Above this temperature the
iron becomes white-hot, and may bend and break the porcelain tube
within ; the couple should in such cases be only left in the furnace for
a short time, or the iron sheath should be supported on a fire-clay slab,
or coated with fire-clay, or protected by a fire-clay tube.
The length of time necessary for a thermo-couple to attain a
temperature of 700° is five minutes if protected by a porcelain tube
only, and ten minutes if it is further wrapped in asbestos fibre and
placed in an iron sheath. If practically immediate readings are required,
the thermo-junction is left unprotected, and the main length of the wires
protected as above; the thermo-junction becomes brittle in time, how-
ever, with this treatment, and must be re-joined.
Rapid fluctuations of temperature should be avoided; the tube
should be slowly warmed before being placed in the furnace, and
should be allowed to cool slowly after removal.
O. Pfeiffer" is of opinion that the fragile and expensive outer
porcelain tube d (Fig. 66, p. 181) may be omitted altogether and the two
wires of the couple passed through thin-walled porcelain tubes I.2 m.
long, 4 mm. external, and 2 mm. internal diameter. These are placed in
an iron gas-pipe I cm. wide, and a little shorter than the porcelain tubes,
so that the junction projects at one end, and the free space is filled in
with asbestos fibre.
M. CALCULATION OF ANALYTICAL RESULTS
Means which shorten the calculations of analytical results, especially
if they also contribute to the avoidance of error, greatly facilitate
the work of a technical laboratory. Four-figure logarithms are not
sufficiently accurate for most work, as the fourth place is uncertain,
but five-figure logarithms almost always suffice. Ready-reckoners are
preferred by many, and the slide-rule can be very advantageously used
for many technical calculations. Tables specially calculated for chemical
purposes, such as those given in books on Analytical Chemistry, and
* Z. angew. Chem., 1901, I4, 390.
184 GENERAL METHODS USED IN TECHNICAL ANALYSIS
reference tables, are a still more convenient adjunct for analytical
calculations."
Published tables are naturally not adapted for every class of work,
and in many cases it is convenient to prepare special tables of factors,
for specific calculations such as those given in the section on
“Sulphuric Acid,” or for use with the nitrometer (p. 146).
It is a general rule, in both scientific and technical analysis, to
calculate to one, and only one figure, beyond that which can be regarded
as significant. If a method is reliable to within one-tenth per cent, the
percentage results should be given to not more than two places of
decimals. If, however, an ore is valued in percentages only, it is
often advisable to give the analytical result in whole percentages, or
at most to one quarter or one-tenth per cent., so as to avoid confusion
in the sales.
There are two quite different cases which must not be confused
in this connection. If the main constituent of a substance, representing
perhaps 50, 80, or 90 per cent. of the total, has to be determined, the
percentage will practically never be given to more than two places of
decimals, as even the first place will only be reliable in exceptional
cases. In most other cases, also, it would be useless, or even misleading,
to give the percentage to more than two places of decimals. But
where a constituent has to be determined which only occurs in very small
amount, such as gold in quartz, phosphorus in steel, ammonia in water,
and the like, it is necessary to give the result to a hundredth or even to a
thousandth per cent. of the original substance, provided the analytical
methods employed are sufficiently accurate and a knowledge of the
exact quantity is important.
The choice of the atomic weights to be used in calculating the
results of analyses is an important consideration. New determinations
are constantly being made, and within recent years many fundamental
figures, e.g. the ratio H : O = I : I 5-96, which had long been considered
absolutely reliable, have been found to need revision; nor can work in
this direction be regarded as final. Many chemists are accordingly of
the opinion that it would be an advantage to use approximate atomic
weights for technical purposes, as these are less likely to be affected by
subsequent determinations; others maintain that it is wrong in
principle to use numbers which are really inaccurate, such as I97 instead
of 197.2 for gold, which is certainly too low.
The rounded-off values were used in the former (German) edition
of this work, largely because considerable differences of opinion existed
as to the exact atomic weights of many of the elements. The table
* Lunge recommends Zogarithmische A’echentaſeln ſir Chemiker, by F. W. Küster, 6th
edition, 1906, which have been used throughout for the calculation of the figures given in this
book, except for the volumes of gases, which have been calculated from the observed densities.
ATOMIC WEIGHTS . 185
prepared by the International Committee on Atomic Weights, which is
given below, has almost completely removed this objection to the use
of the values, now regarded as the most accurate. The atomic weights
are all given on the basis O = 16, as this is the only basis likely to be
adopted internationally for analytical purposes, and a universal agree-
ment of this character is of far greater importance for technical purposes
than any theoretical advantages of the basis H = I.
International Atomic Weights (1908).
O = I6. O = 16.
Aluminium tº © . Al 27. I Molybdenum . º . Mo 96.O
Antimony . tº g . Sb I2O.2 Neodymium . te . Nd I43-6
Argon tº ſº & . A 39.9 Neon. tº tº o . Ne 20
Arsenic . e tº . As 75-O Nickel º e o . Ni 58.7
Barium . tº g . Ba I37.4 Nitrogen . ë º . N I4-OI
Bismuth . tº e . Bi 208-o Osmium . & tº . Os 191
Boron te sº e . B II Oxygen . g tº ... O 16-oo
Bromine . tº e . Br 79.96 || Palladium . g ſº . Pd 106.5
Cadmium . tº * > . Ca II2.4 Phosphorus tº e . P 3I.O
Caesium . º ſº . Cs I32.9 Platinum . tº tº ... Pt 1948
Calcium . g tº . Ca 4O. I Potassium . g g . K 39. I5
Carbon . & ſº . C I2-OO Praseodymium . e . Priao.5
Cerium . tº tº . Ce 140-25 | Radium . e g ... Rd 225
Chlorine . tº © . Cl 35-45 | Rhodium . & ſº . Rh IO3,O
Chromium. g e . Cr 52. I Rubidium . o (* ... Rb 85.5
Cobalt gº • * . Co 59:O Ruthenium ſº ſº . Ru IoI-7
Columbium e g . Cb 94 Samarium . e & . Sm I50.3
Copper . tº e . Cu 63-6 Scandium . e e . Sc 44' I
Dysprosium . © . Dy 162.5 Selenium . e * . Se 79.2
Erbium . tº e . Er I 66 Silicon te e e . Si 28.4
Europium . tº º . Eu 152 Silver (* º º . Ag IO7.93
Fluorine . tº & . F I9'O Sodium . e tº . Na 23-OS
Gadolinium * ſº . Gd 156 Strontium . gº º . Sr 87.6
Gallium . . . tº . Ga 7O Sulphur . wº & . S 32-O6
Germanium e tº . Ge 72-5 Tantalum . ge tº . Ta I8 I
Glucinum . gº te . Gl 9. I Tellurium . e o . Te 127.6
Gold . te * * . Au IQ7.2 Terbium . tº & . Th I 59.2
Helium . tº (º . He 4-O Thallium . * e . Ti 204. I
Hydrogen . o te . H 1.008 || Thorium . ſº © . Th 232.5
Indium . . . . In II 5. Thulium . . . . Trn 171
Iodine i.e. tº e . I 126-97 Tin . te tº tº . Sn I IQ-o
Iridium . * & . Ir I93-O Titanium . & º . Ti 48. I
Iron . º {} º . Fe 55.9 Tungsten . g * . W 184
Krypton . tº tº . Kr 81.8 Uranium . & o . U 238.5
Lanthanum e ſe . La I 38.9 Vanadium . tº e . V 5 I-2
Lead . . . . . Pb 206.9 Xenon . . . . Xe 128
Lithium . . . . Li 7.03 || Ytterbium . . . . Yb 173-o
Magnesium e & . Mg 24-36 || Yttrium . $º tº . Yt 89-o
Manganese e º . Mn 55-O Zinc . ſº tº © . Zn 65.4
Mercury . . . . Hg 200-O Zirconium . . . . Zr 90.6
186 GENERAL METHODS USED IN TECHNICAL ANALYSIS
When tables are employed, the use of exact figures, which is
certainly of advantage, takes practically no more time than that of
approximate numbers. If the latter are preferred, as being often
sufficient for “practical” purposes, the values and factors given are
easily rounded.
With regard to the formulation of analytical results, W. Fresenius'
considers that it is most desirable that the formulae of all constituents
should be appended to their names; that the name of an acid should
be understood to indicate the acid itself, and not its anhydride or
characteristic ion; that if results are expressed according to the old
dualistic notation, they should be written, “Sulphuric acid (anhydride)
SOs,” or “Sulphuric acid, in terms of anhydride, SOs,” and so on; and
that if expressed in ions, the names of the metals should be used, or in
case of various stages of oxidation, such terms as “Mercurous-ion,”
“Mercuric-ion,” and the anions expressed as “Sulphate-ion SOs,” etc.
LITERATURE ON VOLUMETRIC ANALYSIS
CLASSEN, A.—Ausgewöhlie Methoden der analytischen Chemie, 1901-3.
FRESENIUS, R.—Quantitive Chemical Analysis, 7th edition. Vol. i. translated by
A. Vacher, 1876; vol. ii. translated by C. E. Groves, 1900.
MoHR-CLASSEN.—Lehrbuch der chemischen analytischen Titrirmethode, 7th edition,
1896.
SUTTON, F.—Volumetric Analysis, 9th edition, 1904.
TREADwell, F. P.-Analytical Chemistry, translated from the 2nd German edition
by W. T. Hall, 1904.
* Fifth International Congress of Applied Chemistry, 1903; Z. angew. Chem., 1903, 16, 539.
SPECIAL METHODS OF TECHNICAL
ANALYSIS
TECHNICAL GAS ANALYSIS
By Professor F. Fisch ER, Göttingen. English translation revised by
CHARLES A. KEANE, M.Sc., Ph.D.
THE analysis of gases is undertaken chiefly in connection with the
examination of gaseous fuels and of combustion products. It also
constitutes a most important method of analysis in the alkali industry,
in the manufacture of illuminating gas, and in many other branches of
chemical manufacture; the special methods that are employed in
these cases are described under the respective sections.
In all branches of gas analysis the proper taking of the sample is
quite as important as the method of analysis adopted. In the case of
chimney gases, samples may be regarded as having a fairly constant
composition if the charging is done by means of a hopper, or by some
continuous mechanical process, but with ordinary intermittent charging
the composition varies considerably: this is illustrated in the following
data obtained in the firing of a boiler furnace:—
1 Minute after 12 Minutes
Stoking. later.
Carbon dioxide & & 13.5 per cent. 4.0 per cent.
Carbon monoxide te e O }} O 33
Oxygen e ſº * 5-5 }} 16-5 99
Nitrogen g tº sº 8I.O , 79°5 , ,
Soot . o te & Present. O
29
In such an instance the analysis of a single sample is seldom of
value. So-called average samples are equally unsatisfactory, because
even when a constant method of suction is used for taking the
sample, it is hardly possible to collect such a fraction for analysis as
really represents the composition of the whole. Even were such a
method reliable, analyses made from single samples obtained in this way
would seldom be of use, and would, moreover, lead to wrong con-
clusions in respect to the working of a furnace. Such analyses, for
instance, led Scheurer-Kestner" and Schwackhöfer” to the entirely
* Cf. Fischer, Chemische Technologie der Brennstoffe, vol. i., p. 220.
* Z. angew. Chem., 1887, 216; 1893, 6,398.
189
190 TECHNICAL GAS ANALYSIS
erroneous assumption that at least two to three times the volume of
air, theoretically necessary, is required for complete combustion.
A reliable criterion of the course of a combustion can only be
obtained by the analysis of a series of separate samples taken at
successive short intervals; the effect of the stoking and similar
influences can then be gauged. If the nature of the fuel is fairly well
known, the possible presence of appreciable quantities of carbon
monoxide and of other combustible gases can be judged from the total of
carbon dioxide and oxygen that are present; in such cases the deter-
mination of the two latter gases suffices."
In order to determine the small quantities of combustible gases
which may occasionally be present in normal chimney gases, even
accurate volumetric methods are insufficient, and gravimetric methods
of analysis must be employed; these are carried out by drawing the
sample for a sufficient length of time through a suitable absorption
apparatus, whilst at the same time a gas-volumetric test is made every
five or ten minutes, in order to follow the course of the combustion.”
An old method that serves as a rough criterion of the course of a
combustion consists in inserting a burning splinter in the effluent gases;
if it continues to burn, a large excess of oxygen is indicated. Schäffer
and Budenberg,” and also Hempel,” have suggested placing a gas flame
in the exit gases, as a similar indication.
Several forms of apparatus have been designed for the automatic
examination of chimney gas. Of these the “Ados” ” or “Sarco"
apparatus, in which the carbon dioxide is determined by absorption, is
extensively used. Other automatic forms of apparatus for the deter-
mination of the carbon dioxide in chimney gases are the Simmance-
* Since C+O2=CO2, the 21 per cent of oxygen in air gives 21 per cent. of carbon dioxide;
for the combustion of hydrogen every 2 kilos use up 1 I-15 cb.m. of oxygen. In the case of a coal
having the composition :-
Carbon tº e * * g 84 per cent.
Hydrogen tº gº o e gº 4 »
Oxygen tº tº © d * 8 m
Ash, water, etc. g o ſe © 4 »
the burning of I kilo involves the consumption of air required for o-84 kilo carbon and o.o.3 kilo
hydrogen, the contained oxygen of the fuel sufficing for the combustion of the rest of the hydrogen.
For every #-7 x 22.3 = 156. I cb.m. of carbon dioxide formed, #x II-15 =o-75 x 22.3 = 16.75
cb.m. of oxygen disappear from the gaseous products, owing to the combustion of the hydrogen.
If, therefore, the chimney gases contain 14 per cent of carbon dioxide, the average free oxygen,
when combustion is complete, will be 2 I – ( I4 — *...*)- 5.5 per cent. After a fresh stoking
- \ {}
the proportion of free oxygen decreases to 4-5 per cent, but increases to 6-7 per cent, when the
charge is spent. The analytical results should be in accord with these data, and should always
be carefully checked if this is not the case.
* Cf. Ding!. polyt.J., 1884, 251, 323; Fischer's Jahresöer, 1885, 31, 1295.
* Z. angew. Chem., 1888, 1, 167. * Chem. Ind., 1886, 9, 98.
* Z, angew. Chem., 1905, 18, 1231; Ger. Pat. 160288. The apparatus is supplied by Sanders,
Rehders & Co., IoS Fenchurch St., London, E.C.
THE COLLECTING OF SAMPLES - 191
Abady “Combustion Recorder "" and the “Autolysator,” of Strache,
Johoda, and Genzken. H. le Chatelier * has suggested a rapid
method for testing chimney gases, in which the ends of two insulated
conducting wires are connected by a deposit of copper oxide, which is
placed in the path of the chimney gases; if the air supply to the fuel
is insufficient for complete combustion, the copper oxide is reduced to
metallic copper, and an electric circuit is completed, either to a
galvanometer or to an electric bell; if there is sufficient air, no
reduction takes place and no current passes. F. Haber," has designed
an instrument by means of which the composition of chimney gases is
estimated by the observation of the refractive power: it can be adapted
for photographic registration as well as for direct reading, and is
sensitive to differences of from O-2 to O-25 per cent. of carbon dioxide.
It has frequently been proposed to deduce the content of carbon
dioxide from the specific gravity of the gases. Uehling and Steinhardt."
have suggested the determination of the rate of effusion for this
purpose; Lux and Precht" pass the gas through a hollow ball, which is
subsequently weighed. Pfeiffer" and Arndt * draw the products through
a suitable holder suspended on a balance, and Siegert” passes the gas
through a hollow ball, which is arranged to Swing on a balance, so that
the pointer indicates the percentage of carbon dioxide directly. Results
obtained by such methods can only be regarded as approximate values
which may be of use as a check on the ordinary working of boilers and
the like, because the specific gravity of the gases is not only dependent
upon the carbon dioxide, but also upon the water-vapour, unburnt gases,
etc., and is, moreover, variable with the pressure. A proper chemical
examination is accordingly to be regarded as the only satisfactory
method of analysis.
THE COLLECTING OF SAMPLES
The collecting of samples is effected by withdrawing the gases
from the flue or chimney, through connecting tubes, into a suitable
containing vessel. The containing vessel may either form part of the
gas-analysis apparatus, as is generally the case when the sample
is submitted at once to analysis, or it may be an independent vessel
suitable for transport. If the gas to be sampled is under pressure the
filling of the vessel is easily effected, otherwise an aspirator must be
* Eng. Pat. 18680, 1906. The apparatus is made by Alexander Wright & Co., West-
minster, London.
* Z. f. chem. Apparatenkunde, 1907, 2, 57. The apparatus is made by the Ver Fabriken für
Laboratoriumsbedarf, Berlin.
* Bull. Soc. d’Ancourag., 1904, IOG, 471 ; / Soc. Chem. Ind, 1904 23, 951.
* Z. Electrochem., 1906, I2, 519. * Fischer's Jahresber, 1896, 42, 1166.
* /öid., 1893, 39, I2O5. 7 D. R. P., No. 78612.
* Ibid., Nos. 70829, 125470, and 129163.
* Z. Verein, deutsch. Ing., 1888, 32, Io90; 1893, 37, 595.
192
TECHNICAL GAS ANAI,YSIS
employed; small indiarubber aspirators, such as that shown in Fig. 76
C (p. 199), and various forms of glass aspirators and water-pumps, are
generally used.
Previously to collecting the sample, it is important to be certain that
FIG. 67.
all air is removed from the connecting tubes and any other
intermediary apparatus. The material of the connecting
tube must be selected so as to resist the prevailing
temperature, and it must be unaffected by the con-
stituents of the gas. Glass tubes are used whenever
possible; they can be safely employed for temperatures
up to 500° or 600°. Above this, porcelain tubes are used ;
these necessitate gradual heating up, owing to their
liability to crack. Iron and other metal tubes should be
avoided ; in presence of oxygen, iron tubes effect a con-
siderable absorption of this gas, even at low temperatures,
and give it off to reducing gases. It has been found, for
instance, that in passing generator gases through iron
tubes, heated to a dull red heat, the percentage of carbon
dioxide increased from I-5 to 26 per cent." The connect-
ing tubes are held in position by a perforated cork or
rubber stopper, or by means of a suitable cement. Should
there be any doubt that a well-mixed sample of the gases
has not been obtained, further samples should be taken at
Some other point in the furnace or gas pipe.
To collect the gases from a Bessemer converter,
F. Fischer” uses a porcelain tube (Fig. 67), placed inside
an iron tube b, the two being held together at a by a
collar of fireclay, and the porcelain tube projecting about
5 cm. beyond the iron tube. The porcelain tube is con-
nected with the glass tube 3 at c, by a joint made with a
mixture of clay and water glass; the glass tube is con-
nected by rubber tubing g, to a series of bulbs m, the last
of which is attached to a small rubber aspirator. To
collect the sample, the tube is placed in an upright posi-
tion over the mouth of the converter, so that it is in the
centre of the current of gas; the sample is then aspirated
through the bulbs, one of which is sealed off every two
minutes.
Apparatus for Storing Samples of Gases.—If the gas
sample is not drawn directly into the gas-analysis appar.
atus, transport vessels are necessary. Glass bulbs of about IOO c.c.
capacity (Fig. 68) are used, which are attached to the collecting tubes
".-
s
.
.
.
* Fischer, Dingl. polyt.J., 1879, 234, 528.
* Z. Verein. deutsch. Ing., 1902, 46, Ioof and 1367.

THE COLLECTING OF SAMPLES - 193
as described above (Fig. 67). After being filled, they are sealed off at a,
on either side. Should it be necessary to seal off the tubes in the open,
a small oil lamp (Figs. 70 and 71, in half
actual size) may be advantageously used. It a—ſ 4 \-4—
is provided with a wickholder d, and a metal \ /
Cover B, having air-holes at the bottom c, a FIG. 68.
larger opening at the top, and a round hole e
at the side, through which the blowpipe flame, blown from the blowpipe n,
(to which a rubber tube and mouthpiece are attached) is directed.
FIG. 69.
If the samples are only kept for a few hours before examination,
light glass cylinders of about IOO c.c. capacity, provided with well-
2- ~ ground stopcocks at either end, may be used (Fig. 69). A
/ convenient form of tube, in which a sample of gas can be
- sealed and subsequently withdrawn readily for analysis, is
€ == described by Sodeau i
The gas sampler devised by Stead 2 (Fig. 72) is a very
J'u't A useful form of apparatus, by means of
| ò N which the sample can be aspirated and
4 subsequently readily transferred for ana-
• Eº. —l lysis. The gas-collector H is provided
FIG. 70. FIG. 71. with four stopcocks, of which C is attached
to the reservoir A, by means of a stout piece of rubber tubing. The
reservoir is filled with mercury or water, according to the nature of the
sample to be taken. By placing the reservoir
on the shelf B, and opening the taps C and D,
the air in H is expelled and replaced by the
mercury. The taps C and D are then closed,
the connecting tube from the flue or chimney
connected to E, and the reservoir lowered ; by
now opening D and E, the gas is drawn through
the collecting tube and the upper tubes of the
aspirator, thereby completely removing the
contained air. D is then closed, the reservoir
placed under the tap F, and the latter opened,
when the mercury flows out, thereby drawing
in the gas, the rate of entrance of the gas being
regulated by the rate of flow of the mercury.
After the sample has been collected, the taps are closed and the gas
* /. Soc. Chem. Ind., 1903, 22, 190. * Jöid., 1889, 8, 176.
N
FIG. 72.


194 TECHNICAL GAS ANALYSIS
transferred for analysis, by placing the reservoir on B and opening the
stopcocks C and D, sufficient gas being first allowed to escape to clear
the connecting tubes. If the sample is to be kept for some time, it is
well to put the gas under slight pressure, by opening C, whilst the
reservoir is on the shelf, after the sample has been collected. A similar
form of sampler has been described by Kent Smith and Towers."
APPARATUS FOR GAS ANALYSIS
The apparatus for carrying out the analysis of gases may be
divided into two groups:–
I. APPARATUS IN WHICH THE MEASURING TUBE SERVES ALSO
FOR THE ABSORPTION.
2. APPARATUS IN WHICH THE MEASURING TUBE IS APART FROM
THE ABSORPTION APPARATUS.
The first group includes the Bunsen” eudiometer, and the gas
burettes designed by Winkler, Bunté, and Honigmann. Of these the
eudiometer is not employed in technical gas
analysis.
Winkler's Gas Burette.”—This form of ap-
paratus is one of the earliest forms devised for
technical gas analysis; although still used, it has
been largely replaced by later and more con-
venient forms of apparatus. It consists of two
communicating tubes (Fig. 73), held in position
by clamps and connected at the bottom by a
rubber T-piece K, the free end of which is closed
by a pinch-cock, and which is made use of for
cleaning out the apparatus. The graduated
measuring tube is provided with a simple tap B
at the top, and a three-way tap A at the bottom ;
the capacity between the two taps, which must be
determined by an exact measurement, is about
a 100 c.c. The pressure tube is provided with a
º; tap C for levelling purposes, and is closed at the
tº top by a rubber cork D, to which a rubber tube is
attached. The stand is provided with a movable
holder for the tubes, so that they can be placed
either in a horizontal or in a vertical position. In the manipulation
of the apparatus, the measuring tube is first filled with the sample of
* /. Soc. Chem. Ind., 1897, 16, 400.
* Gasometrische Methoden, R. Bunsen, 2nd edition, 1877; English translation by Sir H. E.
Roscoe of Ist edition, 1857.
* / prakt. Chem., 1872, 6, 303.
FIG. 73.

BUNTÉ'S GAS BURETTE I95
gas, either by suction, or by pressure, through the tap A, the tap B
being kept open during the filling; after the contained air has been
completely displaced, the taps are closed and the sample of gas
brought under atmospheric pressure by momentarily opening A or B.
The absorption liquid is then introduced into the pressure tube, and
any gas below the tap A driven out, by making the connection between
A and the outside air; the tap is then turned into the position con-
necting the two tubes. The liquid at once begins to enter the measur-
ing tube, and its access may be facilitated by blowing through the
rubber tube attached to D ; after sufficient liquid has been introduced,
the tap A is turned and the absorption hastened by alternately placing
the tube vertically and horizontally. When no more absorption occurs
the level of the liquid in the two tubes is adjusted, either by means of
the tap C or of the pinch-cock at K, and the reading taken. Carbon
dioxide is absorbed by potassium hydroxide, Oxygen by alkaline
pyrogallol, and carbon monoxide by cuprous chloride solution.
Winkler" recommends an ammoniacal cuprous chloride solution, whilst
others prefer a hydrochloric acid solution; both are reliable if a fresh
solution is used for each determination, but this is necessary, as old
solutions are apt to give off carbon monoxide, and may thus cause an
increase of volume after absorption (cf. p. 2 IO).
Bunté's Gas Burette.”—This burette is a development of a simpler
form of apparatus due to Raoult.”
It consists (Fig. 74) of a burette A provided with taps a and 6, of
which b is a three-way cock; the burette is fitted with a water-jacket to
protect it against rapid changes of temperature. The space between the
taps is rather more than IOO c.c., and is divided into fifths of a c.c. The
IOO c.c. division coincides with the bottom of the tap b, and the zero is
6 or 8 c.m. above the lower tap a ; the divisions are marked for IO c.c.
below the zero. The funnel c is provided with a mark, and all
measurements are made at atmospheric pressure, plus the pressure of a
column of water, contained up to this mark in the funnel. The reservoir
C, which can be attached by a rubber tube to the bottom of the burette,
as shown, serves to fill the latter with water; an ordinary funnel may be
used for the same purpose.
The sample of gas to be analysed is introduced by first filling the
burette with water from the reservoir C.; the taps a and 6 are then
closed and the rubber tube from C detached. The three-way tap 6 is
then connected with the gas supply, and the sample drawn into the
burette by running water out of the bottom tap a. Rather more than
IOO c.c. of the gas are drawn into the burette, and the adjustment to
the zero mark is then made by means of the aspirating bottle S, which
* Z. anal. Chem., 1889, 28, 269. * Dingl. Aolyl. J., 1878, 228, 529.
* Comptes rend., 1876, 82, 844.
196 TECHNICAL GAS ANALYSIS
serves for forcing water into the burette, or for withdrawing it therefrom.
In both cases the rubber tube s is filled with water by blowing at t, so
>}} as to remove all air, and
º whilst the water is still
running out, it is at-
tached to the tip of the
burette at a. For the
adjustment, the gas in
the burette is com-
pressed to about 95 C.C.
by forcing water in from
the bottle S. The tap a
is then closed, s de-
tached, and the water
cautiously run out,
through a, exactly to the
zero mark. The gas is
now under excess pres-
sure, which is corrected
by filling the funnel cup
to the mark with water,
and then momentarily
opening the tap 6, when
the excess of gas escapes
through the water. The
burette now contains ex-
actly IOO C.C. of gas at
the atmospheric pres-
sure, plus the pressure
of the column in the
funnel c. In case the
burette is filled by draw-
ing the sample of gas
through it, with an as-
pirator, until all the air
has been expelled, the
adjustment is made by
compressing the con-
tained gas, by attaching
the vessel C to the
burette, and then ad-
justing as above.
In order to absorb any one of the constituents of the gas, the
absorbing liquid is introduced into the burette. This is effected by first
!\\\\\\\\
.
1.
º



BUNTÉ's GAs BURETTE 197
drawing off the water below the zero mark by means of S, the rubber
tube of which, s, is attached to the burette, after first filling it with water,
as described. The tap a is then opened, and by applying suction at t,
the water is readily withdrawn to a point a little above the stopcock.
The suction bottle is then disconnected, the burette removed from the
clamp, and the end inserted in a cup containing the absorption liquid,
when on opening a the liquid at once rises into the burette. The tap
is then closed and the contents of the burette well shaken, to facilitate
the absorption of the gas. A further quantity of liquid is then admitted
as before, and this is repeated, the burette being shaken after each
absorption, until the level of the liquid in the burette remains constant.
The pressure is adjusted by filling the funnel c with water and opening
the tap 5, when a little water will enter the burette; the adjustment for
pressure is completed by filling c to the mark with water and then
closing b, when the reading can be taken. *-
Owing to the adhesion of the absorption liquids to the walls of the
burette, it is preferable to replace them by water, before the final
readings are taken. This may be effected by removing the absorption
liquid by means of the suction flask, and replacing it by water, intro-
duced through the funnel c; by repeating this two or three times, the
liquid is completely replaced by water. Another method is to wash out
the burette by a steady stream of water, introduced through the funnel
c, and run out through a, the two taps a and 6 being opened during
the washing ; when the absorption liquid has been fully removed, the
pressure is adjusted as before, and the reading taken. Further absorp-
tions are effected successively by suitable absorbents. Carbon dioxide
is absorbed by potassium hydroxide, Oxygen by alkaline pyrogallol,
and carbon monoxide by ammoniacal cuprous chloride. The composition
of the solutions employed is given on pp. 209-2IO.
After each analysis the burette should be thoroughly cleaned, as
otherwise errors are likely to arise, especially in the determination of
carbon dioxide, which may be either absorbed or produced by dirt
present in the burette." -
A further description of the Bunté burette and of its application to
the analysis of illuminating gas is given under the section “Gas
Manufacture,” Vol. II.
Honigmann's Gas Burette.—This burette, which is of a very simple
character, is especially intended for the estimation of carbon dioxide in
the carbonating gases of the ammonia-Soda process. It consists of a
measuring tube A (Fig. 75), of IOO c.c. capacity, closed at the top by the
tap a, and open at the bottom, where a stout rubber tube is attached
at b. The absorbing liquid is contained in the cylinder C. The sample
* Cf. F. Fischer, Z. angew. Chem., 1890, 3, 599; Foster and Haldane, The Examination of Mine
Air (1905), p. 98.
198 TECHNICAL GAS ANALYSIS
of gas is introduced by an aspirator, through a, and after all air in the
burette has been thus expelled, the tap is closed and the burette immersed
to the zero mark in C, which is filled with a solution of potassium
- hydroxide; the tap a is then opened momentarily, and
thus exactly IOO C.C. of gas, at the atmospheric pressure,
are obtained for the absorption. This is effected by
first immersing the burette somewhat lower in the solu-
r tion, so as to wet the sides with the absorbent, and then
A. drawing it out, so that the bottom of the rubber tube
remains in the solution, whilst the burette is raised over
the edge of the cylinder. The potassium hydroxide at
H once begins to enter the burette, and by agitating a few
º times the absorption is rapidly completed. The burette
is then placed in the liquid, so that the inner and outer
levels are the same, and the reading taken. The method
is not to be regarded as one of great accuracy, but it is
extremely simple and quick.
APPARATUS IN WHICH THE MEASURING TUBE IS
APART FROM THE ABSORPTION APPARATUS.
The forms of apparatus included in this group com-
bine both greater rapidity and more convenience in
working than those described above. The earliest gas-
analysis apparatus of this type was devised by Regnault
and Reiset" (1849); it has been succeeded by a very
great variety of improved forms, of which those due to Orsat, Hempel,
F. Fischer, and Sodeau will be described in this section.
Of other forms used in technical gas analysis, those of Pfeiffer, Jäger,
and Drehschmidt are described in connection with the analysis of
illuminating gas (Vol. II.).
Orsat's apparatus.--This form of apparatus is primarily due to
Schlösing and Rolland;” it has been modified and improved by Orsat,”
Salleron, F. Fischer,” Lunge, Sodeau, and others, and is generally known
as the Orsat gas analysis apparatus.
Fischer's modification consists of a measuring tube or burette A
(Fig. 76), surrounded by a water-jacket, which is attached at the top to
absorption vessels, generally called pipettes, D and E, by means of
capillary tubes, and at the bottom to a pressure bottle L. The whole is
fixed, by suitable supports, in a case, which can be closed by sliding
shutters for transport. The burette has a capacity of IOO C.C., the lower
portion being graduated in tenths and the upper in whole c.c. The
capillary tube, leading to the pipettes, is provided with a tap at each
* /. Chem. Soc., 1853, 6, 128. * Annal. des Mines, 1875, 8 [vii.], 485 and 501.
* Ann. Chim. Phys., 1868, I4 [iv.], 55. * Fischer's Jahresber, 1880, 26, 330.


FISCHER-ORSAT APPARATUS 199
junction, as shown, and there is a mark ºn, on each of these, below the tap ;
the tap c is a three-way stopcock which allows of connection between
the gas supply and the tube B, or between B and A, or between A and
the outside air. The U-tube B is filled with cotton wool, which serves
to retain any soot or dust from the gas sample. The pipettes are filled
with bundles of glass tubes, so as to provide a large surface for contact,
between the gas and the absorbent; the open end of each pipette is
closed by a rubber stopper carrying a glass tube, to which a thin rubber
ball is attached, the object of which is to protect the contained liquids
from contact with the air.
To prepare the apparatus
for use, the pressure
bottle L is filled with
water and then raised, so
as to drive out the air in A,
through the tap c. The
pipette D is then filled
with potassium hydroxide
solution, for the absorp-
tion of carbon dioxide,
and the pipette E with
pyrogallol solution, for the
absorption of oxygen; a
third pipette, charged with
cuprous chloride solution,
can be employed for the
absorption of carbon mon-
oxide. The method of
preparing these solutions
is given on pp. 209-2IO, in
connection with Hempel's
gas - analysis apparatus.
To introduce the absorb-
ing liquids, they are poured
in from the back, so as to fill about half of the pipette, and then drawn
up to the mark m, by opening the respective tap and lowering the pres-
sure bottle L; when the liquid reaches the mark, the tap is closed and
the rubber ball inserted at the back of the pipette. Before proceeding
with an analysis, the apparatus should first be tested, to see whether it
is air-tight. This is done by filling the burette with water, then closing
the tap c and lowering the pressure bottle, when it will at once be seen,
from any change in level of the water in A, whether there is any leak.
To introduce the sample of gas, A is again filled with water, the tap
c turned, so as to communicate with B and the outside air, and the

200 TECHNICAL GAS ANALYSIS
rubber aspirator C, which is connected with the gas supply, then
attached to the exit tube of B; after clearing out the air in B, by
aspirating, the tap c is turned, so as to connect the gas supply with the
burette, and the sample introduced by lowering the pressure bottle L.
A little more than IOO C.C. of gas are syphoned into A, and the tap c
then closed. To obtain exactly IOO c.c. for analysis, the gas is
compressed to the zero mark by raising L, the rubber tube s, then
closed, either by the pinch-cock or by the fingers, and the tap c moment-
arily opened to the outside air, so as to establish the atmospheric
pressure on the contained gas. -
To determine the carbon dioxide, the tap connected with D is
opened and the gas in A transferred into the pipette, by raising the
pressure bottle; this operation is repeated several times, by alternately
lowering and raising L, until the absorption is complete, the level of the
liquid in D being finally adjusted to the mark m, the attached tap
closed, and the reading in the burette taken, the pressure on the gas
being adjusted to that of the atmosphere by raising L, so that the
height of the contained liquid and that in the burette is the same.
The reading obtained gives the percentage of carbon dioxide directly.
The determination of the oxygen is similarly effected in the pipette E.
In respect to the Subsequent determination of carbon monoxide, it is to
be observed that cuprous chloride solution becomes unreliable after
having been used for a short time (p. 2 IO), and Fischer prefers to
omit this determination in the analysis of furnace gases. Other forms
of apparatus are, of course, available for the determination of this
gas, and, as already pointed out (p. 190), its quantity can also be
approximately gauged from the percentage of carbon dioxide and
Oxygen contained in gases of this character. Carbon monoxide does
not occur nearly so frequently in furnace gases, in presence of free
oxygen, as has been assumed to be the case, an assumption that is
largely the result of inaccurate analyses.
When the analysis is completed, the residual gas is cleared out of
the burette and the apparatus is then again ready for use; an analysis
can be completed in five minutes, with an accuracy of O.2 per cent. To
re-charge the pipettes, when exhausted, the contents are removed by
means of a small syphon, and thoroughly washed out with water, before
introducing the fresh reagent. In case any of the absorption liquid
should have got into the capillary tube, above the stopcocks of the
pipettes, the tubes should be thoroughly washed out with water through
the tap c, and the water in the burette and pressure b ttle renewed. It
is important to grease all the stopcocks carefully before the apparatus is
put aside after use. -
As the result of experience with more than four thousand analyses,
Fischer regards this form of apparatus as the most suitable for the
LUNGE-ORSAT APPARATUS 201
examination of the gases from boiler furnaces," heating apparatus, brick
kilns, pottery kilns, ultramarine kilns, black-ash furnaces,” and the like,
and also as advantageous for the examination of blast-furnace gases,
cupola-furnace gases,” the gases from puddling furnaces,” the saturation
gases in the manufacture of sugar, and for the study of petroleum
lamps and of gas engines.”
Lunge's modification" of the Orsat apparatus (Fig. 77) is .
arranged for the determination of hydrogen, in addition to that of carbon
FIG. 77.
dioxide, oxygen, and carbon monoxide. The pipettes b, c, d, are provided
for the absorption of the latter gases, and are charged with potassium
hydroxide, alkaline pyrogallol, and cuprous chloride respectively; the
* Dingl. Aoyt, J., 1878, 229, 130; 1879, 232, 346; 233, 183. Fischer's Jahresöer, 1881, 27,
146, Io;5 and loso; 1882, 28, 131 ; 1883, 29, 1289; 1885, 31, 1295. -
* Fischer's Jahresber, 1878, 24, 431; 18 so, 26, 274; 1881, 27, 35 and Ioaz.
* Zbid., 1879, 25, 71; 1884, 30, 37; 1885, 31, 37. * Ibid., 1883, 29, 1229.
* Ibid., 1881, 27, 35. * Dingl. polyt.J., 1882, 245, 512.

202 TECHNICAL GAS ANALYSIS
manipulation of the apparatus is in every way similar to that of the
Orsat-Fischer form. The determination of hydrogen is effected by
combustion over a palladium asbestos fibre, placed in the capillary tube
f; the tube is connected to the pipette h, which is filled with water, and
through the tap e, to the capillary connecting tube of the apparatus.
The small spirit lamp g, carried by the movable rod i, is for heating
the palladium asbestos. To carry out the combustion, the residual gas,
after the absorption of the carbon dioxide, oxygen, and carbon
monoxide, is first mixed with air in the burette a ; this is admitted
through the tap A, by lowering the pressure bottle, and sufficient air is
added to bring the total volume up to about IOO c.c. If the gas is very
rich in hydrogen, it is advisable, either to add a further proportion of air
after the first combustion has been completed, or to repeat the combus-
tion with the residual mixture, or to substitute oxygen for air in the
first instance. The reading of the total volume of gas and air (or
oxygen) is first taken, and the mixture passed through the tap e, over
the palladium asbestos, previously gently heated; the temperature of
the tube should be such as to allow of its being touched, for a moment,
without burning the fingers. The progress of the combustion is at once
indicated by the glowing of the asbestos thread at the end where the
gas enters, and by carefully syphoning the mixture backwards and
forwards several times, complete combustion is effected ; this is checked
by the volume remaining constant. If carbon monoxide is combusted
with the hydrogen, the residual gas must be passed into the pipette b,
to remove the carbon dioxide formed, and the total contraction measured
after this absorption.
Example of Analysis.
IOO c.c. of Generator Gas gave the following readings on the burette :-
1. After absorption of Carbon dioxide . & . . 3.2
2. After absorption of Oxygen . . . e tº te 3.2
3. After absorption of Carbon monoxide e gº c 24.2
4. After addition of Air . & * tº e e O.9
5. After combustion te º tº ge tº { } IO-8
6. After absorption of Carbon dioxide formed in 5 e § I I-4
The percentage volumes of carbon dioxide and oxygen are obtained
directly from the above data. The total contraction is I 1.4–0.9 = Io. 5;
of this II.4— IO-8 = O-6 is due to the formation of carbon dioxide, and
since carbon monoxide forms its own volume of carbon dioxide, on
combustion, this represents the volume of carbon monoxide present
and not previously absorbed by the cuprous chloride solution. The
total carbon monoxide is, therefore, 24.2 – 3-2 + O.6=2 I-6. Since both
2 volumes of hydrogen and 2 volumes of carbon monoxide require
I volume of oxygen for combustion, two-thirds of the total contraction
represents the hydrogen and unabsorbed carbon monoxide, and by
SODEAU-ORSAT APPARATUS
203
subtracting the latter, the volume of hydrogen present is obtained—
(II.4–0.9 × 3)—O-6= 6.4. The composition of the gas is therefore:
Carbon dioxide
Oxygen
Carbon mono-vide
Aydrogen ſº g
Mitrogen and some Methane .
3-2
O-O
2 I-6
6.4
68.8
I OO-O
W. H. Sodeau's modification of the Orsat apparatus is designed
for the analysis of mixtures containing small proportions of combustible
gases, and is especially applicable to the examination of chimney gases.
It can be readily used in the
Stokehold, and in combination
with a suitable high tempera.
ture thermometer it affords a
ready means of determining the
total chimney losses, including
those due to unburnt hydro-
gen, which are usually ignored ;
the efficiency of boilers can
thus be ascertained. The dis-
tinguishing feature from the
ordinary Orsat, is the replace-
ment of the cuprous chloride
pipette and palladium Com-
bustion tube, by an adaptation
of the Winkler combustion
pipette (p. 214). The apparatus
is shown in Fig. 78.
The combustion pipette
may consist of two separate
vessels, the combustion bulb O
and the aspirator CR, as shown
in the figure, or it may be
shaped somewhat like an ordi-
iſſm"Tí
=s
- -, -------º-º-º:
-:---------. ::::::::::::::===::::::-----------. .
FIG. W.S.
nary Hempel pipette for solid absorbents (Fig. 84, p. 206), provided with
a straight instead of with a bent capillary tube.
The bulb O is pro-
vided with a spiral of platinum wire O-25 to O-3 mm. in diameter,
attached to a pair of unlacquered brass electrodes, and is heated by a
current of about 5 amperes, obtained either by means of a 4-volt accum-
* Chem. Mews, 1904, 89, 61 ; Eng. Pat. 12225, 1966.
The editor is indebted to Dr W. H.
Sodeau for this description of his apparatus; it is made by Messrs Brady & Martin, Newcastle-
on-Tyne.

204 TECHNICAL GAS ANALYSIS
ulator, using a short length of iron wire as a controlling resistance, or
from a lighting circuit, connected through a few lamps in parallel;
water is used as the confining liquid. A lens L, mounted on a fitting
which slides on a vertical rod, gives increased accuracy in reading,
owing to the elimination of errors of parallax. The pipettes K and P
are for the determination of carbon dioxide and of oxygen respectively.
Analysis of chimney gas. After determining the carbon dioxide, as
usual, by means of the potash pipette K, the gas is passed into the com-
bustion pipette and the current turned on for one to two minutes, when
the gas is syphoned back and the contraction measured. The carbon
dioxide produced by the combustion is then determined by a one-
minute absorption in the pipette K; the decrease in volume gives the
percentage of carbon monoxide. Half the volume of the carbon dioxide
produced represents the contraction due to the combustion of the
carbon monoxide; this value is deducted from the total contraction on
combustion, and two-thirds of the difference equals the percentage of
hydrogen present. Only one absorbent is thus required for the analysis.
The determination of these three constituents can be completed within
fifteen minutes, and by means of appropriate curves the results can be
stated as lbs. of air used per lb. of fuel, and loss of heat as unburnt
gases, provided the composition of the fuel is approximately known.
For ordinary practical purposes, the determination of oxygen is, there-
fore, unnecessary. It may, however, be carried out by means of the
pyrogallol (or phosphorus) pipette P, either after the other constituents
have been determined, in which case the amount consumed in the
combustion is added, or else immediately after the first determination
of carbon dioxide, in which case it is necessary to draw in a little air
and to re-measure the total volume before combustion.
Rapid joint estimation of carbon monoxide and hydrogen. As these
two gases are, volume for volume, of nearly identical calorific value,
it often suffices to make a joint determination of the two, after
absorption of the carbon dioxide. For this purpose it is necessary to
employ a combustion pipette with a movable reservoir, CR, as shown
in Fig. 78. The measurement after combustion is omitted, in order to
save a little time. The large stopcock S is closed as soon as the gas
has been driven over into the combustion pipette O. After combus-
tion, the tap of the potash pipette K is opened, and the gas directly
transferred, by raising CR and finally closing the tap of O ; after
allowing one minute for absorption, the stopcock S is opened, the
residue drawn back into the measuring tube, and the tap of K
closed. Two-thirds of the contraction equals the percentage of “com-
bustible gases,” i.e., carbon monoxide and hydrogen.
A simpler and more compact form of this apparatus is also made
which is specially suitable for the work of marine engineers.
HEMPEL’S APPARATUS 205
A modified Orsat apparatus, specially designed for the analysis of
generator gas, has been devised by Meyer." The measuring burette
has a capacity of 120 c.c., and a platinum spiral combustion pipette is
provided.
Hempel's Apparatus. – This is, in part, a
development of the Ettling gas pipette; it is a
much used and very adaptable form of appar-
atus.
The burette (Fig. 79), consists of a measuring
tube à, graduated into IOO C.C. with
. O. 2 c.c. divisions, and the pressure
tube a, each of which is contracted at
the lower end and bent at right angles,
so as to fit into slots in weighted
º wooden supports. The measuring
FIG. 80. tube is also contracted at the top, and
terminates in a capillary tube 3 to 1 mm. in diameter and 3 cm. long.
The tubes are connected by a piece of rubber tubing about IOO cm.
* Z, angew. Chem. 1905, 18, 446. -



206 TECHNICAL GAS ANALYSIS
in length, which is preferably divided in the middle, and the parts
joined together by a short piece of glass tubing, to facilitate the cleaning
of the burette.
A short piece of thick-walled rubber tubing, provided with a
pinch-cock, is attached at f
A modified form of burette, which has many advantages in practice,
has the rubber connection and pinch-cock at f replaced by a three-way
tap, of the form shown in Fig. 80.
Another modification, generally known as the modified Winkler
burette, is shown in Fig. 81. This is provided with a simple gas tap at
a and with a three-way tap at 5; its chief advantage is in the analysis
of mixtures containing an easily soluble gas, since the initial volume
of gas can be measured without bringing it into contact with water,
which is always used as the confining liquid in the various forms of
Hempel burette employed for technical analysis.
Any of these forms of burette may be provided with a water-jacket,
if desired, in order to minimise the errors due to changes of temperature.
The pipettes, in which the various reagents for the absorption of the
gases are contained, consist of two bulbs, a and & (Fig. 82), connected
º
.FIG. 82. FIG. S3. FIG. 84.
by the tube d, of which b is attached to a capillary tube c of 3 to 1 mm.
diameter; a white porcelain plate m, is fixed at the back of c, to render
the thread of liquid in the capillary more visible. The capacity of the
bulb 6 is about 150 c.c., that of a about 100 c.c.; this allows of a
sufficient quantity of the absorbent being left in b, after the introduction
of IOO C.C. of gas. The pipettes are mounted either on wooden
(Figs. 82 and 86) or on adjustable iron frames (Figs. 83, 84, 85, and 87).
Pipettes for use with solid absorbents, such as phosphorus, are provided
with a cylindrical bulb, tubulated at the bottom (Fig. 84), and are closed
by a good rubber stopper; double pipettes (Figs. 85 and 86) are used

HEMPEL’S APPARATUS 207
for absorbents which require to be protected from the air, such as
pyrogallol, or which evolve irritating fumes, such as bromine.
A good form of simple pipette, in which a small second bulb, filled
with glass beads, is placed above the absorption bulb, is shown in Fig. 87.
FIG. S5.
The simple pipettes are filled by pouring the absorbent through the
tube attached to the bulb a (Fig. 82), whilst the air is sucked out from
b; sufficient of the liquid is introduced to entirely fill the bulb b, the
capillary tube c, and to leave a small quantity in a. To fill the double
pipettes, the best plan is to attach a glass tube a
metre long, and provided with a funnel, to the
top of the capillary tube, and to pour the absor-
bent liquid in through this, until the first bulb is
completely filled; to effect this, without the
liquid overflowing to the protecting bulbs c
and d (Fig. 86), it is necessary to blow out the
air contained in a through the funnel, several
times successively, after a portion of the absor-
bent has been introduced. After a and the
lower part of & have been filled, water is intro-
duced into d and c, through the tube attached
to d, after first blowing over a portion of the R-
liquid in a into b, so as to remove the air in the FIG. S7.
latter.
The burette and pipettes are connected by a piece of thick-walled
capillary tubing of I mm. internal diameter and 18 cm. in length, as
shown in Fig. 88, the attachment to the pipette and burette being made
by means of stout rubber tubing.
To introduce gas, for analysis, into the burette, the pressure tube B
(Fig. 88) is filled with water and then raised, so as to completely drive


208 TECHNICAL GAS ANALYSIS
out the air in the measuring tube A, and the pinch-cock or tap at f
closed; the gas supply is then attached at f, and the sample syphoned
in, by opening f and lowering the pressure tube. A little more than
IOO c.c. of gas are drawn in and then f is closed When the Winkler
modification of the burette (Fig. 81) is used for the analysis of a
mixture containing a very soluble constituent, the sample is introduced
through the three-way tap b, at the bottom of the burette. To obtain
exactly IOO c.c. of gas for analysis, the pressure tube is raised gradually
and the connecting rubber tube closed, by pressing it between the thumb
and finger, when the liquid stands exactly at the IOO c.c. graduation;
f is then momentarily opened to
establish the atmospheric pressure,
and the volume verified by bringing
the water in the two tubes of the
burette to the same level.
ºff The burette and the pipette are
then connected, as shown in Fig. 88.
To obviate the error due to the air
| in the capillary tube E, it is first con-
nected to the pipette, and the absor-
bent liquid blown over from the latter
until it begins to run out at the free
end of the capillary, when it is at
once attached to the burette; it is
convenient to attach a piece of rubber
tubing to the tube of the upper bulb
of the pipette for this purpose. The
transference of the gas, from the bur-
ette to the pipette, is effected by
raising the pressure tube and open-
ing f; sufficient water is driven over
to wash out the capillary tube of the
pipette, as far as the top of the absorption bulb, so that the gas is
confined between two surfaces of liquid, and f is then closed. The
absorption, is promoted by gently shaking the burette; some recom-
mend the removal of the capillary E for this purpose, but sufficient
movement can be given to the liquid without disconnecting. After
absorption, the gas is syphoned back into the burette by lowering B,
the liquid in the pipette being drawn over as far as the pinch-cock
(or stopcock) at f; after allowing two minutes for the water in the
burette to drain, the liquid in the two tubes of the burette are brought
to the same level and the reading taken. The absorption is repeated
until a constant value is obtained, and the decrease in volume, with each
reagent, represents the percentage of absorbed gas present in the
n
#:
|
t
|
|
ſ
#
ºłºwº. I
º'. "I l
1. -
| vº
| #.
|| ||}}} " .
#º" || ||
# , , . . . . . . . . . . .
iſy/ ||
| - *w- |
| |
ſº |
ºlº ºf
|
Hill
FIG. SS.




















HEMPEL’S APPARATUS 209
mixture. The capillary connecting tube E must be thoroughly cleaned
after each successive reagent.
As the Hempel apparatus is not used for work of high accuracy,
some workers prefer to avoid introducing the absorbents into the
connecting capillary tube, and in transferring the gas from the burette
to the absorption pipette, and in syphoning it back, only bring the
confining liquid as far as the tap of the burette and the top of the
capillary of the pipette respectively.
The Hempel apparatus can be used for the analysis of mixtures
containing carbon dioxide, Oxygen, carbon monoxide, hydrogen,
methane, nitrogen, olefines, and hydrocarbon vapours. The methods
adopted for the determination of the two latter will be described
in connection with the analysis of illuminating gas (Vol. II.); the other
gases are determined in the order given, either by absorption or by
combustion. »
Carbon dioxide. A solution of potassium hydroxide, containing
one part potassium hydroxide to two parts water, is used for the
absorption.
Oxygen. Either alkaline pyrogallol or phosphorus is used as the
absorbent. The pyrogallol solution is made up by dissolving 20 g. of
pyrogallol in 500 c.c. of potassium hydroxide solution (I : 2); this large
excess of alkali is necessary, as otherwise an appreciable quantity of
carbon monoxide may be formed, when gases rich in oxygen are passed
into pyrogallol." A double pipette (Figs. 85 and 86) is used with this
reagent.
Phosphorus is a very energetic absorbent for Oxygen, but its
action is hindered by even traces of certain gases and vapours, such as
the unsaturated hydrocarbons, phosphoretted hydrogen, ammonia, and
alcohol vapour; also the absorption is very slow below 14°. This
reagent is used in the form of thin rods, surrounded by water, in the
pipette for solid absorbents, shown in Fig. 84. As phosphorous oxide
fumes are evolved in the absorption, these are removed by passing the
gas into a potassium hydroxide pipette, after the absorption of oxygen
is complete. The water in the pipette should be renewed, from time
to time, as it becomes saturated with phosphorous acid, and the
pipette should be kept in the dark, when not in use, to avoid the
formation of the inactive amorphous phosphorus.
Sodium hydrosulphite has also been proposed as an absorbent for
oxygen ; * it is cheaper than pyrogallol, and absorbs oxygen equally
readily, at low and at higher temperatures. The solution is prepared
by dissolving 50 g. of the salt in 250 c.c. of water and adding 40 c.c. of
a sodium hydroxide solution, containing 500 g. of alkali to 700 c.c. of
* Cſ. Hempel, Ber., 1885, 18, 278; Clowes, J. Soc. Chem. Ind., 1896, 15, 170.
* Qſ. H. Franzen, Ber, 1906, 39, 2069.
O
210 TECHNICAL GAS ANALYSIS
water, and is used in a pipette filled with rolls of iron-wire gauze; the
absorption takes place according to the equation : -
Na,S.O., + H2O+O = 2 NaHSOs.
Carbon monoride. Cuprous chloride, either in ammoniacal or in
hydrochloric acid solution, is used as the absorbent, and double pipettes
must be employed. The former solution is prepared by dissolving Io g.
commercial cuprous chloride in 250 c.c. hydrochloric acid (concen-
trated), and reducing the cupric chloride present, by the addition of
copper gauze, until the solution is colourless; the acid liquid is then
poured into 1500 c.c. of water, the water decanted, after allowing to
stand for twenty-four hours, and the cuprous chloride dissolved in 250
c.c. ammonium hydroxide of sp. gr. O.91. Another method is to prepare
a stock solution, by dissolving 200 g. of cuprous chloride in a solution
of 250 g. of ammonium chloride in 750 c.c. of water; this solution must
be kept in stoppered bottles and made up for use as required, by the
addition of one-third of its volume of ammonium hydroxide of sp. gr.
O'905. The hydrochloric acid solution is prepared as above, by dissolv-
ing commercial cuprous chloride in hydrochloric acid, in presence of
copper gauze. Sodeau recommends dissolving IOO g. cupric chloride
(cryst.) in 500 c.c. hydrochloric acid (conc.) and 500 c.c. water, and
placing this solution in a number of small narrow-mouthed bottles, of
about I 50 capacity, containing rolls of copper gauze ; the reagent is
ready for use as soon as it has become colourless, and the bottles are
filled up each time that a portion of the contents is withdrawn. It is
to be borne in mind that cuprous chloride also absorbs oxygen,
ethylene, and acetylene, so that these gases must be previously
removed, if present.
The absorption of carbon monoxide by cuprous chloride is much
slower than that of carbon dioxide and oxygen by their respective
reagents, and it has further been established by Drehschmidt" and by
Winkler” that cuprous chloride solutions, after they have absorbed a
considerable volume of carbon monoxide, give up a portion of the
absorbed gas, when shaken with an indifferent gas, as in the ordinary
course of analysis. Hence, an increase of volume, instead of a decrease,
may occur after the absorption. This source of error is avoided by
using two pipettes successively, the first for the absorption of the bulk
of the gas and the second for the final residue ; after the former is
sufficiently exhausted, it is re-charged and the order of the two pipettes
reversed. Drehschmidt's experiments show that the ammoniacal solu-
tion of cuprous chloride is preferable to the hydrochloric acid solution,
in respect to the liberation of absorbed carbon monoxide.
* Ber., 1887, 20, 2572; 1888, 21, 21.58.
* Z, anal, Chem., 1889, 28, 269; J. Soc. Chem, Ind., 1889, 8, 570,
HEMPEI,’S APPARATUS 211
Hydrogen. This gas is practically always determined by combustion
with oxygen, either over palladium asbestos or by explosion. The
former method is carried out by replacing the connecting capillary tube
E (Fig. 88), by a similar tube containing a thread of asbestos fibre,
coated with metallic palladium, which is heated as described in connec-
tion with the Lunge-Orsat apparatus (p. 202).
To prepare the fibre, I g. of palladium is dissolved in aqua regia, the
solution evaporated to dryness on the water-bath, and the residual
chloride dissolved in a little water; the solution is then reduced by the
addition of a few c.c. of a cold saturated solution of sodium formate and
sufficient sodium carbonate, to render the whole strongly alkaline.
About 1 g of long-fibred, soft asbestos, which should suffice to absorb
the whole of the liquid, is soaked in the solution, and the fibre dried at
a gentle heat; it is then carefully washed to remove adhering salts and
again dried, when it is obtained as a dark-grey coloured product,
containing about 50 per cent. of metallic palladium. The capillary
combustion tube is 15-16 cm. long, 1 mm. bore, and 5 mm. external
diameter. To introduce the fibre, a few loose strands are moistened
with water and then rolled into a fine straight thread, which is slid into
the capillary tube, whilst held vertically and before it has been bent, for
attachment to the burette and pipette. The thread is easily brought
into position in the centre of the tube; the water is then drawn off, by
means of filter paper, the moisture removed by drying, and the ends of
the tube bent at right angles for a length of 3.5 to 4 cm.
The capillary tube is attached on the one side to the burette in the
usual way, and on the other, to a simple pipette filled with water. To
carry out the combustion, not more than 25 c.c. of the gas in the burette
is first passed over into the pipette, the necessary quantity of air (about
75 c.c. should suffice) then introduced into the burette, its volume
measured and also the volume of gas-Fair, after syphoning back the
former from the pipette, as a check. The mixture is next passed into
the pipette, which is gently shaken so as to mix the whole, and then
passed backwards and forwards over the heated palladium asbestos till
no further contraction occurs; two-thirds of this contraction is due to
the hydrogen. The level of the confining liquid is brought to the
bottom of the capillary combustion tube, in syphoning the gas to and
from the burette. For the introduction of the air or oxygen, the
burette with the three-way tap at the top (Fig. 80) is very advantageous,
as it avoids disconnecting the capillary tube.
This method is very valuable as a means of determining hydrogen,
in presence of methane; the latter is not combusted under the above
conditions, if the temperature does not exceed 500°." Carbon monoxide
* Cf. H. G. Denham, J. Soc. Chem. Ind., 1905, 24, 1202; O. Brunck, Z. angew, Chem., 1903,
16, 695; J. Soc. Chem. Ind., 1903, 22, 925. - -
212 TECHNICAL GAS ANALYSIS
can be similarly determined by combustion over palladium asbestos, the
carbon dioxide formed being subsequently absorbed by potassium
hydroxide. The determination of hydrogen, by explosion with oxygen,
is carried out in the Hempel explosion pipette (Fig. 91).
The determination of hydrogen by absorption with palladium.
Hempel has worked out the conditions under which the well known
power possessed by palladium of absorbing hydrogen can be made use
of for the determination of the gas. -
Pure palladium is indifferent towards a mixture of hydrogen,
methane, and nitrogen, but if it contains a small proportion of palladious
oxide, a partial combustion of the hydrogen occurs, and the heat thus
generated is sufficient to ensure the
absorption of the rest of the hydro-
gen present; the process is accord-
ingly in part a combustion and in
part an occlusion of the hydrogen.
The palladium is prepared by
heating 4 to 5g of palladium sponge,
in portions of I g. at a time, on the
lid of a platinum crucible, until it
nearly glows, and then allowing it to
cool slowly, whereby a thin film of
oxide is formed upon the surface of
the metal. For use, 4 g. of this oxi-
dised sponge is placed in a U-tube,
of 4 mm. internal diameter and
20 cm. total length ; the tube is
placed in a beaker, and kept at a
temperature of 90° to IOO’ by hot
Fio. 89. - - water, as shown in Fig. 89.
To carry out the determination,
the U-tube is attached on the one side to the burette, and on the other
to a pipette charged with water, and the gas syphoned backwards and
forwards three times; the beaker H is then replaced by one containing
cold water, and the gas again passed twice through the tube, so as to
cool it to the original temperature. The volume is finally adjusted by
syphoning the liquid in the pipette to i, and the reading taken; the
decrease in volume represents the absorbed hydrogen and the volume
of oxygen originally contained in the U-tube. The latter is determined,
once for all, by closing one end of the U-tube with a glass stopper
attached by a piece of rubber tubing, and then placing it in a beaker
of water at 9°; the open end of the tube is then connected with the
burette, previously filled with water, and the tube heated in the beaker
to IOO’. The increase of volume, as measured in the burette, is that due

HEMPEL's APPARATUS 213
to an increase of 91° in temperature, and therefore equal to one-third of
the volume of gas ; from this value the volume of oxygen in the U-tube
is obtained. The palladium is regenerated, after use, by first passing
air over it, when it gets quite hot; any drops of water that may collect
are removed, the palladium then shaken out of the tube and oxidised
as before.
Carbon dioxide, oxygen, and carbon monoxide, heavy hydrocarbons,
and vapours of hydrochloric acid and of ammonia, prevent the deter-
mination of hydrogen by this method, and it is but seldom used in
technical gas analysis. It has, however, the advantage that since no air
is added, there is no restriction as to the volume of the gas that can be
used for the analysis.
Methane is determined by combustion with oxygen, either by
explosion or by a heated platinum wire.
The determination of methane by explosion. The Hempel explosion
pipette is used for this method. The earlier form of this pipette is
shown in Fig. 90; it is an ordinary simple pipette, provided with
FIG. 90.
electrodes at # and a stopcock at d, and is filled with water. An
improved form of pipette, in which mercury is used as the confining
liquid, is shown in Fig. 91 ; the second bulb of the ordinary pipette is
replaced by a levelling bulb, attached by a stout piece of rubber tubing.
It is advantageous to have a small stopcock fitted to the top of the
capillary tube at c, so that the mixture of gas and air or oxygen can be
fired, without risk of loss by leakage.
To carry out the determination, the residual gas, after the deter-
mination of the hydrogen, is syphoned over into the explosion pipette,
the necessary quantity of air or oxygen added, in the same manner as
in the combustion of hydrogen over palladium asbestos, the total volume
then measured as a check, and the whole passed over into the pipette,

214 TECHNICAL GAS ANALYSIS
where it is mixed by gentle shaking ; the water in the burette is
syphoned over until the capillary tube of the burette is filled, when the
tap at c is closed. The pressure on the mixture is reduced slightly,
before firing, by lowering the pressure bulb a ; the tap d is then closed
and the gases fired by attaching the terminals of a small Ruhmkorff
coil to the electrodes of the pipette. After the combustion, the gas is
syphoned back into the burette, and the carbon dioxide formed deter-
mined by absorption with potassium hydroxide.
In respect to the volume of air or oxygen to be added for the
combustion, the ratio of the total combustible to incombustible gases
must be between the limits of 26 to 64: IOO, in order to ensure no
nitrogen being simultaneously oxidised ; ' a ratio of 50 : IOO is a good
working proportion, it being borne in mind that the oxygen required
theoretically for the combustion of the methane, according to the
equation :
CH4+ 20, - CO2 + 2 H2O,
is included in the combustible gases, the excess forming part of the
incombustible gases. The choice of air or oxygen will accordingly be
decided by the volume of the gas to be dealt with and its content of
methane; oxygen from the ordinary Oxygen cylinders can be con-
veniently used, the percentage of the contained oxygen having been
previously determined. One-third of the total contraction, after the
absorption of the carbon dioxide, represents the volume of methane
present.
Methods for the simultaneous determination of two or three
combustible gases, such as hydrogen, carbon monoxide, and methane,
are described in connection with the analysis of illuminating gas
(Vol. II.). The residual gas, after combustion, contains the nitrogen and
the excess of air or oxygen; the nitrogen, in the original sample, is
taken by difference. *
Combustion of methane by a heated platinum zwire. Methane is
completely combusted, without
—r explosion, by passing it, mixed
DE —I RY_ with Oxygen, over a heated sur-
- face of platinum. Winkler *
g designed a modified form of
FIG. 92. Hempel pipette for this purpose,
in which a small platinum spiral,
enclosed in the bulb of the pipette, is heated by an electric current.
A much more convenient device for this method of determination is
the Drehschmidt * capillary tube shown in Fig. 92. This consists of a
seamless platinum tube, 20 cm. long, 2 mm. thick, and O-7 mm. internal

E-l
* Bunsen, Gasometrische Methoden, 2nd edition, p. 72.
* Z. Anal, Cheun., 1889, 28, 269; J. Soc. Chem. Ind., 1889, 8, 570. * Aer., 1888, 21, 32.45.
FISCHER'S APPARATUS 215
diameter, containing three or four pieces of platinum wire; pieces of
brass tubing are soldered to the ends of the platinum tube, which serve
for connecting to the measuring tube and pipette. Two small cooling
cylinders, also made of brass, are fixed on to the brass tubes, just above
the bend. The tube must be tested before use, to make sure that it is
air-tight.
To carry out the combustion, the capillary tube is attached by
rubber connections to the burette and to a pipette charged with
potassium hydroxide, in the usual way, and the mixture of gas and air
(or oxygen) obtained, as described above; the tube is then heated to a
bright red heat, preferably by a flat-flamed Bunsen burner, and the
gaseous mixture passed backwards and forwards twice, over the heated
platinum. The reading is taken after a final absorption in the
potassium hydroxide pipette, and one-third of the total contraction
represents the methane present.
F. Fischer's Apparatus.*--In this apparatus (Fig. 93), either mercury
or water is used as the confining liquid ; it has been designed especially
for the analysis of producer gas, “mixed ” or semi-water gas, and water
gas. A similar form of apparatus, of a more complex character, was
previously designed by Frankland and Ward,” and subsequently modi-
fied by McLeod,” and by Thomas.”
The essential parts consist of the bulb or laboratory vessel A, in which
the absorptions and combustions are effected, the measuring tube M,
surrounded by a water-jacket, and the levelling tube D. The upper
part of A must be sufficiently wide to prevent the adherence of drops of
liquid ; the measuring tube is graduated either in millimetres or in
cubic centimetres, and must be calibrated by mercury if the former
method of graduating is adopted. Pressure bulbs, F and L, supported
on wooden blocks and movable in the notches of the frame H, are
attached to A, and to the tube V, which is connected with the
measuring and levelling tubes respectively. The combustion spiral g
(Fig. 94) consists of a nickel tube in which a nickel wire v, insulated
by a glass tube, is fixed, the ends of which are connected by a spiral
of platinum wire; the free ends of the wire project outside the bulb
A, and are attached to binding screws. In order to keep the platinum
wire at the right temperature, and so prevent its becoming detached by
excessive heating, an adjustable resistance should be placed in the
circuit, to regulate the current; three Grove cells will keep the spiral
red hot for one and a half to two minutes, which is sufficiently long for
a combustion to be completed. The tubes attached to the measuring
tube and the laboratory vessel are connected at v, by rubber tubing and
brass clips (the distance between v and d is somewhat greater than
* Z, angew. Chem., 1890, 3, 591. * /ötd., 1869, 22, 313.
* /. Chem. Soc., 1853, 6, 197. * /bid., 1879, 35, 213.
216 TECHNICAL GAS ANALYSIS
FIG. 93.

FISCHER'S APPARATUS 217
represented in the figure); d. and n are three-way taps, and h a simple
straight-bore tap.
To introduce the gas sample, g is drawn down to the bottom of the
laboratory vessel, and A, M, and D filled with mercury by raising
F and L; the taps d and h are then closed and the gas syphoned into
A, through the three-way tap d, by turning it so as to connect the gas-
supply with the laboratory vessel. If the gas is to be drawn from a
sealed bulb (Fig. 68, p. 193), this is attached to the rubber tube
connected with the tap d, and the end broken inside the tubing;
the other end is broken off under mercury, in a suitable containing
vessel, and the gas drawn into A by lowering the pressure bottle F.
The gas is then syphoned from A into the measuring tube, any excess
being removed through d, by making the connection with the outside
air. After reading the volume of the gas in M, O-3 to O. 5 C.C. of
potassium hydroxide solution are introduced into A,
through the funnel t, for the absorption of carbon dioxide.
The gas is then transferred from M into A, and syphoned
back to a mark just in front of the tap d, after the absorp-
tion. The volume of the connecting capillary tube up to
this mark must be ascertained, in the calibration of the
measuring tube, and allowed for. For the absorption of
oxygen, O. I c.c. of a I : 3 pyrogallol solution is similarly
introduced into the laboratory vessel. After the determi-
nation of these gases, A must be thoroughly cleaned, before
the hydrogen, carbon monoxide, and methane are com-
busted; this is effected by washing it out two or three
times with IO-20 c.c. of water, which is introduced through
the funnel t, and expelled through the tap d. If any of Fig. 94.
the absorbent has inadvertently got beyond d, or into the
measuring tube, it is removed by first clearing the air out of A, syphon-
ing water from the laboratory vessel into M, and then drawing back
the gas and liquid into A, until the mercury reaches d'; the gas is next
syphoned back into M, and the wash-water expelled, as before,
through d.
The combustion of the remaining gases is effected by drawing in
the necessary quantity of air or oxygen, through the tap d, into the
measuring vessel; the volume of the mixture is then measured, the
tube carrying the platinum spiral pushed into position in the upper
part of A (as shown in Fig. 93), the electric current turned on, and the
gas transferred into A by raising the pressure bottle L. The combustion
generally takes place quietly, without any explosion; at most, a slight
and quite harmless explosion occurs. The gas is syphoned back, after
combustion, into M, and the contraction measured ; the carbon dioxide
formed is determined by absorption with potassium hydroxide, as
s} ºf
.



218 TECHNICAL GAS ANALYSIS
described above. Finally, the excess of oxygen can be determined by
absorption with pyrogallol and the residual nitrogen measured.
The volume of air required for every IOO volumes of producer gas,
semi-water gas, and water gas is I2O volumes, I 50 volumes, and 350
volumes respectively. The volume of air required for the combustion
of water gas restricts the volume of gas that can be used for the
determination to about 25 c.c.; to secure greater accuracy, a mixture of
oxygen and air can be used in the proportion of 35 c.c. of air and
25 c.c. of oxygen for 40 c.c. of water gas. It is preferable to combust
this mixture in successive portions of about 30 c.c. at a time, rather
than all together, especially if the worker is inexperienced with the
apparatus; also, it is advisable to effect the combustion under slightly
reduced pressure.
The oxygen can be generated in a small tube of the form shown in
Fig. 95; about 2 g of potassium chlorate are heated in the tube, which
is attached by a to the rubber connection of the tap d
(Fig. 93). After the contained air has been cleared out
through e, which is placed under water, the taps d and n
are opened to first clear out the capillary tube with
oxygen, n then closed, and the Oxygen introduced
into A by lowering F.
In order to obtain scientifically accurate results with
this apparatus, the tubes M and D are graduated in
mm., and readings taken of the barometer, the height
of the columns of mercury in M and D, and the tem-
perature of the water in the jacket round M, at each
observation during the analysis. The volumes are then corrected to O’
and IOOO mm. pressure, by the formula:
v (B – b – e)
Iooo [1+(o'Oo366t)]
FIG. 95.
in which v is the volume of gas, B the height of the barometer (reduced
to oº), b the difference of level between M and D, and e the tension of
aqueous vapour at the temperature of the experiment £. By adjusting
the columns of mercury in M and D to the same level, after each
determination, the correction for pressure is avoided ; the correction for
temperature is only required in cases where special accuracy is looked
for, as the changes of temperature with a jacketed measuring tube are,
as a rule, but slight during the interval of an analysis.
The calculation of the proportions of hydrogen, carbon monoxide, and
methane follows from the volumetric changes according to the equations:
H2+O = H,C) CO+O = CO, CH, + 20, = CO2 + 2H,O.
The contraction in the case of hydrogen (w) is =#, in the case of
carbon monoxide (c) = }, and in that of methane (m) = 2 : both carbon

FISCHER'S APPARATUS 219
monoxide and methane form their own volume of carbon dioxide.
Taking V as the total volume of combustible gases, n as the total
Contraction, and £ as the total carbon dioxide formed, the resulting
equations are:—
V = w -- c + m. n = #w + #c + 2m. A = c + m.
Or,
Hydrogen (w) = V – A.
Carbon monoxide (c) = , k + V – #m.
Methane (m) = ## – V + šn.
Example of Analysis.
Volume of Gas taken, 60 c.c. *
After &bsorption of Carbon dioxide, 56.1 ; therefore 3.9 c.c. Carbon dioxide.
After addition of Air, II6. I ; therefore 60 c.c. Air, containing 47.4 c.c. Mitrogen
added.
After combustion, IoI-2; therefore n = 14.9.
After absorption of Carbon dioxide, 87.6; therefore k = 13.6.
After absorption of Oxygen, 85-7. .
Of this Mitrogen, 47.4 added as Air; therefore 38.3 c.c. Mitrogen,
and V = (56.1 – 38.3) = 17.8.
A/ence—
Volume of Hydrogen = 17.8 — I 3-6 = 4.2 c.c.
Volume of Carbon monoxide = 4.5+ 17.8 – 10 = 12.3 c.c.
Volume of Methane = 9.1 – 17.8+ IO = 1.3 c.c.
and the composition of the Gas—
Carbon dioxide . e & 3.9 or 6.5 per cent.
Aydrogen tº te & 4-2 or 7-O ,
Carbon monoxide ſº . I2-3 or 20.5 ,
AMethane . tº tº e I-3 or 2-2 ,
AVitrogen . iº . 38'3 or 63.8 ,
6O-O or IOO-O per cent.
A simple and very efficient form of gas-analysis apparatus, in which
mercury is used as the confining liquid, has been recently described by
W. A. Bone and R. V. Wheeler.1
W. H. Sodeau's Apparatus.”—This is based upon the gas-analysis
apparatus of Macfarlane and Caldwell,” to which very many modifica-
tions and additions have been made, so that but little of the original
arrangement remains. It can be used for the analysis of practically all
ordinary gaseous mixtures, and is capable of giving results of great
accuracy, with but moderate expenditure of time and trouble. Rapidity
of working is obtained mainly by the adoption of (a), a ready, though
highly accurate method of transference; (b), pipettes in which the gas
can be thoroughly agitated with the absorbent; (c), a device showing at
* /. Soc. Chem. Ind., 1908, 27, Io.
* /hid., 1903, 22, 187. The editor is indebted to Dr W. H. Sodeau for this description of his
apparatus; it is made by Messrs Brady & Martin, Newcastle-on-Tyne.
* /. West Scotland Iron and Steel /nstitute, 1892, vol. i., No. 2.
220 TECHNICAL GAS ANALYSIS
a glance the correction for temperature and pressure. Contamination of
the measuring tube is carefully guarded against.
The measuring tube M (Fig. 96) is graduated in ºr c.c. divisions up
to 50.5 c.c., so that no trouble arises if rather more than 50 c.c., the
normal sample, should be drawn in. For special work, in which a con-
siderable proportion of the gas remains unchanged, it is convenient to
employ a measuring tube having a bulb and narrow graduated stem.
A reliable verification of the graduations is essential. The top of the
measuring tube is joined to a
three-way, oblique bore, capil-
lary stopcock N, arranged
so that the capillary K may
be placed in communication
either with the interior of M,
or with the detachable bent
tube U containing water, and
having a short length of
rubber tubing connected to
its Outer end by means of a
* perforated rubber stopper
º (not shown in the figure).
The zero point of the gradua-
tion is at the outer side of the
plug of the stopcock N. The
level tube L communicates
with the measuring tube by
means of a side branch, bent
So as to exclude any air-
SL11.63S bubbles entangled in the
º mercury. Its lower end is
jº s: connected to a T-piece, one
3. ºn branch of which is provided
Ş º
FIG. 96. with a stopcock and leads
e to the mercury reservoir,
whilst the other is prolonged across the table to a point near the reading
telescope, where it terminates in a piece of strong rubber tubing, the
compression of which, by a broad screw clip, affords a means of accu-
rately adjusting the level of the mercury, whilst looking through the
reading telescope. In order to render the apparatus more compact,
the reading telescope is placed on the opposite side of the gas-analysis
table, instead of on a separate support; the graduations are consequently
on the side opposite to that from which the stopcocks are manipulated.
The illuminator is attached to a tubular fitting which slides on the
rod P. It consists of a piece of ground glass having its upper half
t—————1–1–1–
Şı"
50
-



SODEAU’S APPARATUS 221
rendered opaque, and is adjusted so that only a narrow band of light is
visible above the mercury. A IOO-volt 2% candle-power lamp is
attached.
The mercury reservoir can be adjusted without using both hands, and
does not usually require clamping in position.
The “Aezy principle” correction tube C is so called because, as in the
“Kew "barometer, the disturbance of the level of the liquid is allowed
for in the graduation, instead of being adjusted before each reading. It
consists of a cylindrical bulb, closed above by a stopcock, and attached
below to a U-tube graduated on the inner limb, and filled with water up
to about the level of the zero mark; the outer end of the stopcock is
attached to a bent glass tube, terminating in a short length of rubber
tubing (not shown in the figure). At the beginning of an analysis the
water is brought exactly to the zero mark by manipulating the rubber
tube, and the stopcock is then closed. The volume of air contained in
the instrument is such, that the water is displaced to the extent of one
small division, by changes of temperature and atmospheric pressure
which will cause a gas to alter its volume to the extent of O. I per cent."
These small divisions are viewed by means of a mounted lens sliding
on the rod R, and are further subdivided by eye estimation. The
corrections are thus read directly, in percentages, as easily as the
temperatures would be read by means of a thermometer. The correc-
tions may be kept very small, if the water in the jacket is syphoned off
through the tube S before commencing an analysis, and brought to the
probable average temperature of the room.
The absorption pipette is shown in Fig. 96. The absorption bulb E
is inclined to the extent of about I in 20, in order to facilitate the return
of the unabsorbed gas, and usually contains about 20 c.c. of the
absorbent, confined over mercury. The bulb F contains clean mercury,
and like the bulb E can be placed in communication with the capillary
G by means of the three-way stopcock H. Rubber tubing is attached
to the tubes, projecting upwards, from D and F, in order that
the pressure in these bulbs may be varied by blowing or sucking, as
required. Each pipette is mounted on a wooden support, which is slid
into a very loosely fitting foot when in use, and afterwards replaced in
a rack, inclined at an angle of not less that 35°, in order to relieve the
pressure on the stopcock whilst out of use. -
It is important that all capillary tubing employed in the apparatus
should be of I.O to I-5 mm. bore, and that the external diameter of the
* A small division equals o-o; c.c.; hence, external changes which would necessitate a I per
cent. correction must cause o-5 c.c. displacement of water. If this movement disturbs the level of
the water to the extent of N mm., and if X c.c. = the volume of air in the instrument, it
follows that
I-orz– (x + o-5) ( + IOI N
N -
jº) and therefore z = so + 206.7 - 2 N'
222
TECHNICAL GAS ANALYSIS
capillary G should be practically the same in all the pipettes. The
external end of G must be quite flat, and at right angles to the axis of
the tube.
The eaglosion pipette, Fig. 97, consists of a cylindrical bulb T, of
about 220 c.c. capacity, connected to a mercury reservoir V. Wires are
i.
2:F
Fº
{
W
wº---
l
i
Hempel apparatus (p. 2
$º
i
sealed in, near the top, for connection to a small
induction coil, and a strip of paper, bearing
rough graduations, is fixed to the back of the
pipette. The arrangement lettered F, G, H, is
common to all the pipettes.
The combustion pipette, Fig. 98, is used
when the proportion of combustible gases is
not sufficient for a determination by explo-
sion. Two nickel wires pass through the
rubber stopper closing the bottom of the
pipette, one being insulated from the mercury
by means of a glass tube, and the two are
connected above by means of a small spiral of
platinum wire, W, about 0.25 mm. in diameter.
This wire is heated by means of a 4-volt accum-
ulator, the current being regulated by means
of a short length of iron wire about O. 5 mm. in
diameter.
The phosphorus pipette, Fig. 99, which it is
sometimes convenient to use for the absorption
of oxygen, has one bulb filled with thin sticks
of phosphorus immersed in water, as in the
O9). It is clamped in a canister of water, which
affords protection from light and also provides the means for readily
raising the temperature of the phosphorus to about 20°.
FIG. 98.
FIG. 99.
A small flap
in the cover of the canister is removed whilst the gas is being returned
to the measuring tube. No mercury is used in the bulb F.



SODEAU’S APPARATUS 223
The following description of an analysis of a mixture of carbon
dioxide, oxygen, carbon monoxide, hydrogen, methane, and nitrogen,
will serve to illustrate the mode of working, when great accuracy is
desired. -
Introduction of the sample. The tube containing the gas is con-
nected to a suitably bent capillary tube, preferably filled with mercury,
which is in turn joined to the measuring tube capillary K (Fig. 96), by
means of red rubber pressure tubing, after turning the stopcock N so that
K is in communication with the tube U. The contents of the capillary
are driven into U, and the cock N is then reversed, so that the required
volume of gas can be drawn into the measuring tube M, by lowering the
mercury reservoir. The cock N is then again reversed, the mercury
roughly levelled, and a puff of air sent down the tube S, in order to stir
up the water in the jacket.
Attachment of a pipette. The potash pipette is placed on the stand
and the wetted end of the capillary G made to meet that of K, within
the rubber tube. A little water is then sucked from U into the bulb F,
and mercury allowed to run back and fill the capillary tubes.
Measurement of the sample. The stopcock leading to the mercury
reservoir having been closed, and the level tube L being open to the
atmosphere, the mercury is accurately levelled by means of the screw
clip, and the volume of the gas read, by means of a reading telescope.
The water in the correction tube is then adjusted to the zero mark.
Absorptions. To determine the carbon dioxide, the gas is driven
over into the absorption pipette, followed by sufficient mercury to clear
the capillaries, and the pipette is then rapidly rocked to and from the
operator for two minutes, without disconnecting it from the measuring
tube, so as to thoroughly churn up the absorbent; absorption is practi-
cally complete in twenty seconds, if the shaking is sufficiently vigorous.
A little more mercury is then run over, so as to clear away the alkali from
the bottom of the capillary attached to the absorption bulb, and the
cock N reversed, so that the mercury in the capillaries runs into the tube
U, instead of mixing with the clean mercury in the measuring tube.
If any gas should be accidentally allowed to pass through the cock N
into U, it is retained by the water, and can of course be drawn back
without loss. The cock N is then again turned and the gas passed
slowly into the measuring tube, the rate being controlled by the cock
H, which is reversed as soon as the absorbent has filled the passage
through its plug, so that the gas may be swept out of the capillaries by
means of the clean mercury from the bulb F. The cock N is closed as
soon as the mercury reaches it, and reversed after closing H, in order
to facilitate the removal of the pipette. The next pipette (alkaline
pyrogallol) is then connected up, the measurement of the residue
carried out as described above, and a reading of the correction tube
224 TECHNICAI, GAS ANALYSIS
taken. It is convenient to enter the readings throughout the analysis in
the following manner:—

Correction. Reading. Corrected Reading. Difference. Percentage.
•00 50-01 cc. 50°01 c.c. tº ſº tº e tº º
+ °06 per cent. 33°24 c.c. 33°26 c.c. 16°75 c.c. CO, 33:50
The residual gas is then similarly shaken, for three minutes, with
alkaline pyrogallol, at a temperature not below 16°. The decrease gives the
amount of oxygen. To determine the amount of carbon monoxide, the
residue is shaken for two minutes with cuprous chloride solution (taking
care that the liquid is not allowed to absorb more than its own volume
of carbon monoxide before being replaced), and then transferred
directly to a second pipette, containing fresh cuprous chloride solution.
After again shaking for two minutes, it is sent into a pipette containing
about 2 c.c. of water, and then back into the measuring tube. It is a
good practice to wash out the measuring tube during the absorption of
carbon monoxide, in case any trace of potassium hydroxide has been
introduced, as it would prematurely absorb part of the carbon dioxide
subsequently formed by explosion. Air is drawn in after the water
used in the washing has been expelled, and the mercury then allowed
to re-enter very slowly, in order that the film of water may be properly
Squeezed off the sides of the tube. After the carbon monoxide has
been determined, oxygen, stored over mercury in one of the ordinary
absorption pipettes, is added in excess, and the mixture measured. If
the oxygen is prepared by electrolysis it should be freed from traces
of hydrogen, e.g. by treatment in a combustion pipette kept for that
purpose.
Aydrogen and methane are estimated by means of the combustion
pipette (Fig. 98), if only small amounts are present, but the explosion
pipette (Fig. 97) is employed when the proportions are sufficient. The
mode of transference to these pipettes is the same as with the
absorption pipettes, but in returning the gas to the measuring tube,
mercury from the bulb V (instead of from the bulb F) may be used
for driving the gas out of the capillary. When the combustion pipette
is employed, the current is switched on as soon as the top of the glass
insulating tube has become exposed, and the coil is maintained at a
very bright red heat for two or three minutes, in order that complete
combustion may be effected.
Before passing a spark through a mixture in the explosion pipette,
the gas must be expanded (by lowering the mercury reservoir), to such
an extent that no appreciable amount of nitrogen will be oxidised ; if,
SODEAU’S APPARATUS 225
however, the expansion is carried too far, the combustible gases will not
be burnt completely. This point merits more attention than it appears
to have generally received. If the expanded volume is markedly less
than ten times the volume of hydrogen, measured under atmospheric
pressure, plus twenty times that of the methane, the amount of nitrogen
oxidised is usually sufficient to cause the condensed moisture to have a
pronounced action upon the mercury; in the absence of very definite
information, it seems best to expand to an ample extent and to subse-
quently employ the combustion pipette to ascertain whether any
traces of unburnt combustible gas remain.
After combustion or explosion, the residue is measured and the
resulting carbon dioxide absorbed. From the volume of carbon dioxide
produced and the total contraction, the proportions of the methane and
hydrogen are calculated. The nitrogen is determined by difference.
Finally, the excess of oxygen is determined by means of pyrogallol or
phosphorus, as a check upon the other measurements.
Aösorbents which attack mercury, such as fuming sulphuric acid,
can usually be employed in the ordinary absorption pipettes, an
appropriately filled guard tube being attached to the bulb D. Of
course, no mercury is used to confine these liquids, and the mercury in
the capillaries is very carefully driven into the bulb F before transfer-
ring the gas. After absorption, the gas may be transferred directly to
the potash pipette, and thence to the measuring tube. -
Other accessories. Almost any form of absorption or combustion
device can be employed in conjunction with this apparatus, if it is fitted
with a horizontal capillary stopcock, or preferably with the usual
arrangement of the bulb and three-way stopcocks, F, G, and H.
ABSORPTION COEFFICIENTS OF THE COMMONER GASES
IN WATER
In the following table the absorption coefficients of the commoner
gases in water at 15°, are given, as determined by Bunsen (propylene by
v. Than). These values must be taken into consideration in all cases
in which gases are collected or analysed over water.
Propylene . e c . O-237 Carbon monoxide e . O-O24
Ethylene . © e . o. 162 | Hydrogen . tº © . O.OI9
Methane . - © . O.O39 Oxygen tº- * º . O.O3O
Carbon dioxide . e . I.OO2 Nitrogen . & e . O.OI 5
More recent determinations by L. W. Winkler give the following
values for oxygen, nitrogen,” hydrogen,” and carbon monoxide; * the
* Ber., 1891, 24, 3609. * Ibid., 1891, 24, 99.
* Ibid., 1891, 24, 3606. * Z. physikal. Chem., 1892, 9, 173.
P
226 e TECHNICAL GAS ANALYSIS
coefficients are stated for every degree between o’ and IOO", in the case
of oxygen and of nitrogen, in the original papers.
gº; 10° 12° 14° 16° 1S* 20°
Oxygen . e 0-038 0.036 0-035 0-033 0°032 0-031
Nitrogen . te 0-018 0.018 0-017 0-016 0-016 0-015
Hydrogen . e 0-020 0-019 0-019 0-019 0.018 0-018
Carbon monoxide 0 028 tº º º * c º tº tº & tº e & 0-023
THE CALORIFIC VALUE OF GASEOUS FUEL
The calorific power of heating and of illuminating gas can be
calculated from that of its constituents, as determined by analysis,
but the method involves a complete gas
analysis, and is unsatisfactory on account of
the difficulty of obtaining an accurate deter-
mination of the individual hydrocarbons
(cf. Vol. II., Gas Manufacture). A direct
####: determination of the calorific value is
- 24 accordingly advantageous. The calorimeters
devised by Junkers' and by Boys are gener-
| ally used for this determination; that of
F. Fischer” will also be described in this
: section.
a li Junkers' Gas Calorimeter.—In this ap-
paratus, the heat generated by the flame of
§ { burning gas is transmitted to a current of
# j ñ, water flowing at a constant rate, and measure-
- - ments are made of the volume of gas burnt,
the quantity of water heated, the difference
- in temperature of the water on entering and
** - ". . . © à leaving the apparatus, and the quantity of
*** . . . . . water condensed from the products of the
combustion. The gas is burnt in a large
Bunsen burner (Fig. IOO), in the combustion
chamber a, which consists of an annular
copper vessel, the annular space being tra-
versed by a number of copper tubes b,
connecting the top with the bottom of the
chamber. The products of combustion pass through these tubes in a
downward direction, the condensed water being collected at 0, and the
1. Cf. H. Kühne, J. Soc. Chem. Ind., 1895, I4, 63.I.
* Taschenbuch für Peuerungstechniker, 5th edition, 1883, p. 5.


JUNKERS’ GAS CALORIMETER 227
waste gases escaping through the side tube d. The current of water
enters at e, where it passes through a sieve h, and enters the calorimeter
through g and i, thus circulating in an opposite direction to that of the
products of combustion, whereby complete cooling of the latter, to the
temperature of the water, is effected. The water, after passing through
the annular chamber, goes over a series of baffle plates at m, to ensure
thorough mixing, and then passes out through the overflow 2, where
it is collected and measured. The pressure of water is kept constant
by the two overflows at h and 2, the water being run into h at a greater
rate than it passes through the calorimeter; the rate of flow is regu-
lated by the cock i. The temperature of the entering water is taken
by the thermometer Æ, that of the exit water by the thermometer
above m. The whole instrument is enclosed in a cylindrical air-jacket
of polished or plated copper, to prevent radiation. -
The whole apparatus, as arranged for actual experiment, is shown in
Fig. IoI. The gas to be tested is measured by a small wet meter,
provided with a thermometer, and is passed through a governor, to
ensure a steady pressure of gas at the burner; a water-pressure
gauge is attached to the governor, which indicates the pressure, above
that of the atmosphere, at which the gas is measured. As the water
supply may vary considerably in temperature, even within a short
period of time, if taken directly from a tap, it is preferable to have
a large reservoir for its storage. Pfeiffer" recommends a sheet-zinc
tank of 60 litres capacity (H, Fig. IOI), provided with a pipe w, which
serves both for filling and discharging, a glass gauge, and an over-
flow pipe u ; the pipe w is connected with the pressure box of the
calorimeter, through the tap v. The experiment should be so regulated
that the temperature of the exit gases at s (Fig. IOI) is the same as that
of the laboratory, so that they are only saturated with the quantity of
moisture corresponding to this temperature. Instead of measuring the
water collected from c, it may be weighed in the bottle F, of about IO
litres capacity, the weight of the empty bottle having been previously
determined.
The condensed water from the calorimeter is collected in the measur-
ing cylinder d (Fig. IOI), which is placed in position after the gas-meter
index has completed a revolution, from the start of the experiment, and
the collecting continued for a corresponding period, after the bottle F
has been removed.
To carry out a test, the water is first turned on, and when it is
running from both overflows, the gas is lighted and the speed of the
water regulated by the cock i, so that there is a difference of from Io”
* Mote.—These recommendations, in respect to the use of the Junkers' calorimeter and the
accompanying figure, are incorporated from Dr Pfeiffer's contribution on Gas Manufacture in
Vol. II. of the German edition.—C. A. K.
228 TECHNICAL GAS ANALYSIS
to 20° between the inlet and outlet thermometers; illuminating gas
is burnt at the rate of from IOO to 300 litres (4 to II cubic feet) per
hour, producer gas at the rate of from 400 to IOOO litres (14 to 35
cubic feet). When the difference in temperature of the two thermom-
eters is nearly constant and the condensed water is dripping
regularly into d, the test can be begun. This is done by waiting
until the gas-meter index passes any desired point, and then quickly
introducing the outlet water-tube c, into a suitable measuring vessel;
a series of readings of the thermometers are taken whilst about 28
litres (one cubic foot) of gas is burnt, and the volume of gas and of
collected water measured. The inlet thermometer k should be read
every minute, and the outlet thermometer (T, Fig. IoI) every half
minute.

JUNKERS’ GAS CALORIMETER 229
The calorific value is,
_ Weight of water x temperature difference
e- Volume of gas consumed y
the volume of gas being corrected for temperature and pressure.
The value thus obtained represents the calorific power of the gas,
when the water formed by the combustion is converted into the liquid
state in the calorimeter. To obtain the calorific value when the water
formed escapes as steam, the latent heat of the condensed water must
be deducted; the quantity of heat evolved by each c.c. of condensed
water may be taken, with sufficient accuracy for most purposes, as
O.6 big calories (the amount of heat required to raise IOOO g. of water
1° C), and this value, multiplied by the number of c.c. of water con-
densed and divided by the gas consumed, will give the total heat to
be deducted from the gross or “higher” calorific value, in order to give
the net or “lower” calorific value of the gas, with the water formed in
the combustion remaining as steam. The result in calories, multiplied
by 3.97, gives the calorific value in British thermal units.
No further corrections are necessary, and the results are in every
respect satisfactory if the experimental conditions are properly carried
out. It is important that the determinations should be made in a room
free from any considerable variations of temperature.
If a large flame is used, the combustion is apt to be incomplete; it
is accordingly recommended to ascertain, by a preliminary experiment,
that the waste gases contain several per cent. of oxygen, as otherwise
the determination of the calorific value may be 5 per cent, or more too
low (F. Fischer).
Example of Determination.
Volume of gas consumed (5 revolutions of meter index,
at 3:06 litres per revolution) tº º e . I 5.3 litres.
Weight of water * e * 53IO g.
Temperature of inlet water, constant & 13°.8
Temperature of outlet water: 28.1 – 28.15 – 28.15 – 28.1
28. I — 28-I – 28. I — 28. I — 28. I — 28. I © . Mean : 28°. II
Temperature difference tº e e & ge I4°-31
Volume of condensed water . e º e & I4 C.C.
Temperature of room . ſº e g © ſº 22°
Barometer tº º ſº g ge {º tº 766 mm.
Gross calorific value per cubic ). 5.3.Iox 14:31 ×273-22 760
metre, at 15° C. and 760 mm. O-OI 53 288 766
= 5082.
Net calorific value per cubicl_ (5.3.10 × 14:31)—o.6× 14 ×373:32, 760
metre, at 15° C. and 760 mm. JT O-OI 53 288 766
= 4.533.
Owing to the small quantity of condensed water collected in a
230 TECHNICAL GAS ANALYSIS
single experiment, Pfeiffer" regards the above method of arriving at
the net calorific value as somewhat inaccurate, and as the result of his
own observations, is of the opinion that it can be best calculated from
the gross calorific value on the basis of independent experiments in
which 50 to 70 c.c. of condensed water are collected. From numerous
determinations, he finds that the net value is II. I per cent, less than the
gross value, in the case of gas made from English or from Westphalian
coal and, as an average, 8.6 per cent. less in the case of carburetted
water gas. From these data, the net calorific value may be calculated
directly from the experimental determination, without reference to the
volume of condensed water formed.
To facilitate the correction of the calorific value for the volume
occupied by the gas at 15° and 760 mm., Pfeiffer has constructed the
subjoined table (p. 231), in which the figures are given, which must be
added to or subtracted from the value:—

Weight of water x temperature difference
Volume of gas consumed o
The table is calculated for gases having a calorific value of 50oo
calories per cubic metre; the corrections given in the table are
sufficiently near to those required for gases, with a calorific value within
+ 2CO calories of this limit, to be applicable also within this wider
range. The correction, for instance, at 22° and 750 mm. for gases of
48OO and 52OO calories per cubic metre respectively, is O. 18 and O. 196,
against O. I89 in the table. -
Calculations in cubic feet and corrected to 60°F. and 30-inch pressure
are similarly made, the tables used in the determination of the illuminat-
ing power of coal gas being employed for the latter corrections.
Boys' Gas Calorimeter.” — This apparatus is prescribed by the
Metropolitan Gas Referees (1906), for testing the calorific value of
illuminating gas. It has been designed with the object of providing
ample space for the circulation of the stream of gases, so that they
pass slowly and freely through the instrument, and are thus effectively
exposed to the cooling surfaces. The water content of the instrument
is reduced to the smallest quantity, so that the outflowing water attains
its ultimate temperature very quickly after the gas is lighted. The
whole of the circulating water takes the same course continuously,
being debarred from any parallel or alternative routes, and thus unequal
heating and the attendant irregularity of the temperature at the outflow
are avoided. The small content of water suffices to abstract the whole
of the heat from the slowly travelling stream of gases, owing partly to
* Zoc. cit.; and J. Gasóeleucht, 1904, 44, 684.
* Proc. Roy. Soc., 1906, 77, Series A. 122. The apparatus is made by J. J. Griffin & Sons,
Kingsway, London.
º
89 I | If I 9&I 60I 36 #/. 89 0? 33 g II & | 9% 29 OS 96 || $11 || Ogi 871 | 991 || 381 | 944
99T | 87I Z3I 9II | 66 I8 j9 Af 63 &I # 23 | 89 || Zg | #4 || 16 || 20I göl I?I | 891 || 94I | 74%
III | #91 | 68 I ZZI | 90I | 18 0/. 89 93 8I I 9I & | Ig | 89 g8 || 301 | 6TH | 9&I | 891 || 691 | 844
8/I | 39 I gºl | 8&I | ZII | 86 11. 69 Číž #2, 8 6 93 | ?? | 19 || 64 g6 || 311 | 631 | 9%r | 891 || 311
58L | 891 Ig| | #3 I | 8II | 66 38 99 8% 08 AI 9 Ià | 89 99 || $4 || 06 || 901 | 721 || 0%I | 491 | IA!
Z61 g/I | 89 I | Of I | 93 I 90I 06 8/. gg A8 33 || 8 ŽI | 29 6% 99 || $8 || 00I 9II | #91 || Ogſ || 01/
661 38I g0I | 1.7L Igl 8 II | 1.6 6/. 39 ## 08 0I 9 gº 2% 6g | ZZ | 86 OII | Z21 7%I | 691
90% 88I III | 89 I | 68I | 6 II || 30I 98 A9 IQ 98 || AI | T 03 || 99 || #g | IA. | 89 ŽOI 231 | 891 | 891
ZIZ | 96 I 8/I | 09L g?I 9&I | OII | 86 9/. 89 If | 93 || 9 &I | 83 || 4% | 89 || T8 || 46 gTI | 391 || 191
61% 30% 98T | 891 || 39 I | 78.I | 1 II | 66 38 #9 87 || 03 || 3I | 9 22 || 69 || 49 || 7 | 16 | 407 gºſ | 994
93% 80% | 36|I | 7/I | 89 I | Of I | 3&I gOI 88 0/. fºg 99 || 6 || I 9I | #9 || Ig | 89 g8 || 3OT || 6 II | 991.
383 || 9 I& 66I | I8I g9I 97T | 03] | ZII | 76 Al 09 || 3p || 93 || 8 6 Z3 | ?? | Ig | 64, g6 || @II | 79/
0% | 333 || 903 || 88I | 3/I | 39 I 23 I | 6′II IOI 78 /9 || 09 || 38 || 9 I & I? | 49 99 || 34 | 68 goſ | 891,
9% 633 ZIZ || 76T | 8/I 69T | Zſ, I gº, I | 101 || 06 j/ 99 || 68 IZ | 8 ŽI Ig | Sº 99 || 98 || 00I 391
89% 983 || 6 IZ | IOZ | 98I 99T || Og| | 89 I | WII Z6 I8 || 39 97 || 8& | 0 | | 9 % | 0% | 69 || AA || 96 || I91.
093 87% | 93% 80% 86I | #1. I 191 || Of I | 3&I 90I 88 || 0 | 89 gg | 8I O AI 79 &g | OZ | 98 || 091,
A93 || 093 || 78% 913 || 003 || ISI | #9T | 9% I | 6&I III | 76 | 1.4 || 69 Zī | #3 | 1. OI 4& | 9% £9 || 0& 69/
813 || 293 OW& | 333 || 903 | 1.8L | OZI 39L g8 I | ZII | 00I | 88 g9 || 8p || 08 gL | 7 & | 69 || 49 || 7 | 89A,
I83 || 79% | 1.7% 67% 81& 76 I | 1/I 69 I | If I | 73L | 1.0 [ | 06 || Z/ | g g | 18 || 03 || & GI £9 || Og | S9 A9/.
883 || IA% | #93 983 || 033 IOZ | #8I | 99 I | 8? I | Ig I j II | 16 || 6/ | 39 || 7p || 13 || 6 S’ 9% 9% 09 || 99/
96% 81% I9% gić | 83% 803 || 36|I | PZI 99 I | 88I 3&I WOI 18 || 69 Zg #8 / I I O2 99 || 79 || 99/
308 || 783 | 193 || 673 | #83 gT3 || 86I | 08I 39 I | ??I 8&I OII | 36 g/ | Sg Of £3 | 9 ŽI Ig | 8% | #9/
608 || I6% 7.1% | 1.93 || IWG | 333 g03 || 18T | 69T | Iglſ g8T | III | 66 Z8 #9 || 17 | 6% &I 9 % | 0% | 89/
9I3 | 66% 38% | #9% | 85% | 663 || ZIZ || 761 | 9/. I | 89 I | If I | #31 || 90I | 68 II. gº 98 || 6 || || 0 &I | 79 || 394
838 || 108 || 68Z || 31% gga, 93% 6IZ | IOZ | 88I g9I 8?I Igl g|II g6 || 1 || 09 || 87 || 93 | 1 OI 4& I9/.
088 || 818 g6Z | 11& IQZ | ZWZ gzz | 10a | 68L | IAI | #g I | 18T | 6II | IOI | #8 99 || 67 || Ig | 8L | 7 & | 09/
Z88 || 038 808 || 78% Z93 || 6% Z32, PIZ | 96 I 8/I | a 9T | PyL | 1.3L | 60 [ ] I6 | f | | 99 || 68 || 0% & QI 67/
fºg | 1.38 || 013 || I6% 7/Z | 193 || 0% ZZZ | 70% 98I 69T | Ig| | 73I 9II | 86 IS | 89 97 13 || 6 S’ 85/.
393 || 738 813 | 668, IS3 | #93 || 253 || 63% 0IZ || 36|| || 9/I | 89 I | 07I | 8&I gOI 88 || 01 | 89 || 78 || AI I /f/
698 || 378 || 938 || 103 | 68% I/Z | #93 98% 8IZ || 003 || 38I g9I | 1.7L | 08I ZII | 76 || 1 || 69 | If | 8% 9 95/.
998 || 878 || Igg | 8 Ig g6% | 1.1% | 093 || 37% 73% 90% | 68 I | III | 89 I | 93 I | 8TI 00I | 88 99 || 17 | 66 &I 97/.
69 I og I of I o? I e6][ oll oOI Ul VI
608
e66
68%
oló
o9%
o93
ofº,
olò
606
o6I
oSI
'senſea snului are sain:39 poulouſ eul : anſea oglioIeo poureliadse out on poppe aq on Junoue au, oat:3 saināg au.L
'uu Og/, put ‘O SI or ‘aijau orgno Jed
‘(SoHoſeo Ooz:F) serioleo OOOS Jo anie A orglioleo e 5ulatºu sases) Jo auntoA out, 3U1339.I.Ioo Joy aſqa L




232 TECHNICAL GAS ANALYSIS
the avoidance of parallel routes, but mainly by reason of its flow
through a pipe of small diameter, the heat-collecting power of which is
greatly increased by attached wires.
The calorimeter is shown in vertical section in Fig. IO2. It consists
of three parts, which may be separated, or which, if in position, may be
turned relatively to one another about their common axis. The parts
are (1), the base R, carrying a pair of burners B, and a regulating tap.
The upper surface of the base is covered with a bright metal plate, held
in place by three centring and lifting blocks C. The blocks are so
placed as to carry (2), the vessel D, which is provided with a central
copper chimney E, and a condensed water outlet F. Resting upon the
rim of the vessel D are (3), the essential parts of the calorimeter, attached
to the lid G. At the centre, where the outflow is situated, there is a
brass box, which acts as a temperature-equalising chamber for the outlet
water. Two dished plates of thin brass, K, are held in place by three
scrolls of thin brass, LLL. The lower or pendent portion of the box is
kept cool by water, circulating through a tube which is sweated on to
the outside of the bell. Connected to the water channel at the lowest
point, by a union, are six turns of copper pipe, such as is used in a motor
car radiator, of the kind known as Clarkson's. In this, a helix of copper
wire threaded with copper wire is wound round the tube, and the whole
is sweated together by immersion in a bath of melted solder. A second
coil of pipe, of similar construction, surrounding the first, is fastened to it
at the lower end by a union. This terminates at the upper end in a
block, to which the inlet water box and thermometer holder are secured
by a union, as shown at O. An outlet water box (P) and thermometer
holder are similarly secured above the equalising chamber. The lowest
turns of the two coils N are immersed in the water, which in the first
instance is put into the vessel D.
Between the outer and inner coils N is placed a brattice (Q), made
of thin sheet brass, containing cork dust, to act as a heat insulator. The
upper annular space in the brattice is closed by a wooden ring, and this
end is immersed in melted rosin and beeswax cement, to protect it from
any moisture which might condense upon it. The brattice is carried by
an internal flange, which rests upon the lower edge of the casting H.
A cylindrical wall of thin sheet brass, a very little smaller than the
vessel D, is secured to the lid, so that when the instrument is lifted out
of the vessel and placed upon the table, the coils are protected from
injury. The narrow air space between this and the vessel D also serves
to prevent interchange of heat between the calorimeter and the air of
the room.
The two thermometers for reading the water temperatures, and a
third for reading the temperature of the outlet air, are all near together
and at the same level. The lid may be turned round into any position,
BOYS GAS CALORIMETER 233
-
º
ſ§
† |
P
Zº Z2 g d
RNA- % w = ,
% O
{T} ſ
Z/
R ºl.
N %. § lº
|; ##
%
º 62 %
SYºt SES & Y. M.
N S.11ſ, SS㺠Zºº
|iºlº
t
%
S.
-à
> º
%
2}
R
ŠN
%
à
º
:
§
;
$.
g
sº
?
º
#
:º
5.
t:
|
A.
§
%
:
:
o
;
º-
2.
:*
§
S
S.
g
§
*3.
i
S
wº-
Ş
2.
:
g.&wº
i
|S.
†
É.
i
tº:
s
& ſº G |
9-º-º-L-L-ll Ełżzza–C
JZ ſº ZZZZZZZZZZZZZZZ
E -3
|
ZZZZ 22>. "
% GRIFFIN LONDON
- 2×
Fig. 102 (§ Nat. Size).



















































234 TECHNICAL GAS ANALYSIS
relatively to the gas inlet and condensed water drip, that may be con-
venient for observation, and the inlet and outlet water boxes may them-
selves be turned, so that their branch tubes point in any direction.
The general arrangement of the apparatus, as set up for a test, is
shown in Fig. 103, the gas being first passed through a meter and
balance governor, before being led to the calorimeter. The gas supply
is connected up to the central tube at the back of the meter and thence
to the governor, preferably by means of composition piping. The pipe
leading from the governor should terminate in a tap with a nozzle, to
which a short length of rubber tubing is attached for connecting to the
calorimeter.
A regular supply of water is maintained, by connecting one of the
two outer pipes of the overflow funnel, shown in Fig. IO3, to a small tap
over a sink. The overflow funnel is fastened to the wall about one
metre above the sink, and the other outer pipe is connected to a tube, in
which there is a diaphragm with a hole 2.3 mm. in diameter. This tube
is connected to the inlet pipe of the calorimeter. A piece of stiff rubber
tubing, long enough to carry the overflow water clear of the calorimeter,
is slipped on to the outflow branch, and the water is turned on so that a
little escapes by the middle pipe of the overflow funnel, and is led by a
third piece of tube into the sink. The amount of water that passes
through the calorimeter in four minutes should be sufficient to fill the
graduated vessel shown in Fig. IO3 to some point above the lowest
division, but insufficient to come above the highest division in five
minutes. If this is not found to be the case, a moderate lowering of
the overflow funnel or reaming out of the hole in the diaphragm will
effect the necessary correction.
The thermometers for reading the temperature of the inlet and
outlet water are divided into tenths of a degree, and are provided with
reading lenses and pointers that slide upon them. The thermometers
are held in place by corks, fitting the inlet and outlet water boxes. The
thermometers for reading the temperature of the air near the instrument
and of the outlet gases are divided into degrees.
The flow of air to the burners is determined by the degree to which
the passage is restricted, at the inlet and at the outlet. The blocks C
(Fig. IO2), which determine the restriction at the inlet, are made of metal,
about 5 mm. thick, while the holes round the lid which determine the
restriction at the outlet, are five in number and are I6 mm. in diameter.
The thermometer used for finding the temperature of the effluent gas is
held by a cork in the sixth hole in the lid, so that the bulb is just above
the upper coil of pipe. -
The calorimeter should stand on a table by the side of a sink, so that
the condensed water and hot-water outlets overhang and deliver into
the sink, as shown in Fig. IO3. A glass vessel is provided of the size
BOYS' GAS CALORIMETER 235
w
.

236 TECHNICAL GAS ANALYSIS
of the vessel D, which should be filled with water, in which sufficient
carbonate of soda is dissolved to make it definitely alkaline. The
calorimeter, when not in use, is lifted out of its vessel D and placed in
the alkaline solution, and left there until it is again required. The
liquid should not come within two inches of the top of the vessel, when
the calorimeter is placed in it. The liquid must be replenished from
time to time and its alkalinity maintained.
The measuring vessel, shown in Fig. Io9, carries a change-over
funnel, which is placed beneath the short tube attached to the hot-
water outlet. One side of the funnel delivers into the sink, and the
other delivers into a tube, which directs the water into the vessel.
Full details for carrying out a test are given in the Notification of
the Metropolitan Gas Referees for 1906; 1 the above description of the
apparatus is taken from this notification.
Fischer's Gas Calorimeter.”—This form of calorimeter (Fig. IO4),
consists of a wooden vessel B, in which nickel-plated copper vessels A
- and C are contained. A is suspended
from the top of B as shown, and C is
fitted into B by the water-tight junction
at v, which should be greased before
use. The detachable cover D is pro-
vided with an attachment at # for insert-
ing a thermometer, and to carry the
stirrer R ; the inner vessel C is expanded
into three lenticular chambers, each of
which contains a sheet of metal n, Ser-
rated round the edge, the object of
which is to promote intimate contact
between the products of combustion and
the inner surface of the calorimeter.
§P The whole vessel is supported on the
i feet F. The burner E is supported by
i the socket f and the arm m, and is so
i arranged that it can be readily adjusted
\\
\
Zy 22 S
Z * -
4
N
§
N
N
WN
N
N
N
into position, by means of the tongue a
and a pin below f. A small inverted
FIG. 104. cone of nickel gauze is placed in the
tube of the burner, to prevent the gas
striking back; a cone, preferably of platinum gauze, with its apex
upwards, may also be fitted to the top of the burner tube. Air enters
at the inclined plate s, the direction of which serves to retain the con-
densed water in the calorimeter, and the waste gases leave through 6.
1 Published by Wyman & Sons, Fetter Lane, London, E.C. Price Is. 6d.
* Fischer's Jahresher., 1901, 47, 50.

FISCHER'S GAS CALORIMETER 237
To carry out a determination, the required quantity of water is
placed in A, the inner vessel C is placed in position, the cover D put
on, and a thermometer inserted at t The water is agitated by the
stirrer until the temperature is constant, and the burner drawn down
below the body of the calorimeter and attached to the gas supply; the
gas should be passed through the meter for some time previous to the
determination, in order to saturate the contained water. The burner is
then lighted, and quickly introduced into the position shown ; the
flame is adjusted for complete combustion, this adjustment being best
determined by a preliminary experiment so as to ensure an excess of
5 per cent, of oxygen in the waste gases. In testing semi-water gas or
producer gas, the air-holes at the bottom of the burner are closed; in
the case of difficultly combustible gases, such as blast-furnace gas, it is
recommended to supply oxygen (about one-fifth of the volume of the
gas) through a tube fixed in the centre of the burner. The volume of
gas required for a determination is I litre of coal gas, I # to 2 litres of
water gas, and 3 litres of producer or of semi-water gas. When the
desired volume of gas has been burnt, the rubber tube connected with
E is closed by pressing, and the temperature of the water in the
calorimeter read, after agitating with the stirrer for two minutes. º
The increase of temperature, multiplied by the heat equivalent of
the calorimeter, gives the calorific value of the total gas burnt, which
can be calculated to the gross value per cubic metre at normal pressure,
and I 5°C, as described above. To obtain the net calorific value of
the gas, the water in C, in which vessel the greater part of the water
formed in the combustion is condensed, is weighed; this is effected by
first emptying the water out of A, and then detaching and weighing C,
after carefully drying it outside, the weight of the empty vessel having
been previously ascertained. A determination can be made in a few
minutes, and the volume of gas used is much less than that required for
the Junkers calorimeter; the apparatus may accordingly be advantage-
ously used as a check on working conditions, and also when the gas
sample cannot be tested in situ.
An apparatus for the determination of the calorific value of small
quantities of gas (2 to 3 litres) has been described by Hempel."
THE LOSS OF HEAT FROM CHIMNEY GASES
The loss of calorific effect in the combustion of fuel, owing to
the retention of heat by the chimney exit gases, can be calculated
from the following considerations. If the average percentage com-
position of the chimney gas is # parts carbon dioxide, o parts
oxygen, and n parts nitrogen, and that of the air for the com-
*J. Soc. Chem. Ind., 1901, 20, 880; Z. angew. Chem., 1901, 14, 713.
238 TECHNICAL GAS ANALYSIS
bustion, a parts oxygen and 8 parts nitrogen, then the ratio v of the
quantity of air used, to that theoretically required is expressed by the
equation :- - -
% 72 - . 2 I
Z) = x – (20: m) Or 11 – (20:2) or zi – (790 : m)
with air containing 2 I per cent of oxygen. Since I kilo carbon gives
1.854 ch.m. of carbon dioxide on combustion, I kilo of fuel, containing c
per cent. of carbon (after deducting any unburnt carbon left in the ash
or slag), gives 1854×;=K cb.m. of carbon dioxide (at o’ and 760
mm.); the resulting chimney gases will include, in addition, Kx + =O.
cb.m. of oxygen, and Kx}=N. cb.m. of nitrogen. The total weight
of water-vapour W, contained in the chimney gases, consists of that
derived from the moisture present in the fuel, as ascertained by
analysis, that formed by the combustion of the hydrogen, and that
contained in the air used in the combustion. With s as the percentage
of sulphur in the fuel, the total volume of the products of combus-
tion is:—
k1. K(0.4 m) 2S W
4–’ + 286-4" o'So; cb.m., at O’ and 760 mm.
(286.4= weight of I cb.m. of sulphur dioxide; O.805 = weight of I cb.m.
of water-vapour).
The loss of heat is calculated by multiplying the volume of each gas
by its specific heat, and by the difference between the temperature of
the air supply and that of the exit gases; the total of these factors gives
the heat loss in calories. The loss due to incomplete combustion can
be calculated from the calorific value of the unburnt coal in the furnace
residue and of the combustible constituents of the chimney gases, carbon
monoxide, methane, hydrogen, and soot."
For practical purposes, a shortened formula proposed by Lunge,”
dependent upon the determination of carbon dioxide only, is to be
recommended. The proportion of oxygen and nitrogen in the waste
gases is taken by difference. If n = the volume per cent. of carbon
dioxide, then IOO– m = the volume per cent. of oxygen plus nitrogen;
and since I kilo of carbon forms I-854 ch.m. of carbon dioxide on com-
bustion, the volume of oxygen plus nitrogen in ch.m. per kilo of
carbon is:—
I OO – 72
72
= I '854 ×
* For further details, cf. Fischer, Zaschenbuch für Feuerungstechniker, 5th edition, 1885.
* /. Soc. Chem. Ind., 1889, 8, 531 ; and Z. angew. Chem., 1889, 2, 240.
TOTAL SULPHUR IN HEATING GASES 239
If t'=temperature of the chimney gases, t = temperature of the air,
c= specific heat of I cb.m. carbon dioxide (O-41 up to 150° and O-43 to
O-46 between 150° and 350°), and c = specific heat of I cb.m. of nitrogen
(about O-31), then the loss of heat (L) per kilo of carbon burnt, expressed
in calories, is :—
L = I-854(t’ – t)c + 1854(e-): “H” c'.
Since the calorific value of I kilo of carbon is 8.08 calories, then
IOO x L.
8'O8
= Percentage loss of heat by the waste gases.
DETERMINATION OF THE TOTAL SULPHUR IN HEATING GASES
This determination, which is occasionally desirable, can be carried
out by means of the apparatus shown in Fig. IO5 (ºr actual size).
The gas supply, after passing the experimental
meter, is burnt in the small Bunsen burner B, at
the bottom of the adapter C; the latter is fitted
into the bulb tube n, at v, either by a cork or
by an asbestos ring. The tube is surrounded by
a condenser K, the water supply of which enters
at a and leaves at n. The water formed in the
combustion condenses in the tube mm, and there
absorbs the sulphurous and sulphuric acids, the
liquid products being finally collected in F. The
sulphur content is determined by titration with
M/IO alkali, after first oxidising the sulphurous acid
with hydrogen peroxide; or a test-tube, containing
ammonium hydroxide, may be placed by the side of
the burner, during the combustion, in order to ensure
a more complete condensation of the sulphur dioxide,
and the sulphuric acid determined gravimetrically.
The current of gas is regulated so that there is an
excess of 4-6 per cent. of free oxygen, as can be
determined by taking a sample of the exit gases
at O; with good condensation, about 50 c.c. of liquid
are collected per 50 litres of gas burnt.
FIG. 105.
LITERATURE
BUNSEN, R —Gasometriche Methoden, 2nd edition, 1887; English translation of 1st
edition by Sir H. E. Roscoe, Gasometry, 1857.
DITTMAR, W.-Quantitative Chemical Analysis, 1887.

240 TECHNICAL GAS ANALYSIS
DowsoN, J. E., and LARTER, A. T.–Producer Gas, 1906.
FiscHER, F.–Die chemische Technologie der Brennstoffe, 188o-19o1.
FISCHER, F.–Taschenbuch für Feuerungstechniker, 5th edition, 1885.
HEMPEL, W.–Gasanalytische Methoden, 3rd edition, 19oo; English translation of 3rd
edition by H. Dennis, Methods of Gas Analysis, 1902.
NEUMANN, B.–Gasanalyse und Gaszolumetrie, 19oI.
SUTTON, F.– Volumetric Analysis, 9th edition, 1904.
TRAvERs, M.–The Experimental Study of Gases, 19oI.
WINKLER, C.–Anleitung zur chemischen Untersuchung der Industriegase, 1877-79.
WINKLER, C.–Lehrbuch der Technischen Gasanalyse, 2nd edition, 1892; English
translation of Ist edition, with a few additions, by G. Lunge, Handbook of
Technical Gas Analysis, 1885.
FU E L ANALYSIS
By Professor F. FISCHER, Göttingen. English translation revised by
T. LEWIS BAILEY, Ph.D.
THE term Fuel, strictly speaking, is applied to the following sub-
stances:—Wood, peat, lignite, coal, and their products of carbonisation,
viz., wood charcoal and coke.
The analysis of wood, considered as a fuel, is seldom carried out,
and if done, is usually confined to the estimation of moisture. The
moisture of freshly felled timber varies from 15 to 45 per cent. ; ; further
investigation has no significance, as the composition of dry wood varies
but little. The calorific power of dry wood is about 4500 calories.”
Peat, on the other hand, varies considerably, not only in the amount
of moisture it contains, but also in the percentage of ash and in the
chemical composition of the peat substance proper;" consequently the
calorific power is variable. Peat is, however, used as fuel in very
few industries, so that its analysis is seldom required; when necessary,
the same method that is employed for coal is adopted.
Sampling.—Before proceeding with the analysis, it is necessary to
obtain an average sample, for if the sample be not representative, the
whole analysis naturally becomes worthless. In taking the sample, it
is to be borne in mind, that the composition of the coal of some pits
is very variable, whereas, in other cases, the variation is only slight.
For instance, in a piece of Deister coal, weighing 5 kilos, Fischer found
the ash to vary from 4 to 31 per cent. in different portions of the
sample. In twenty-four samples from the Unser Fritz mine, the ash
varied only between the limits 1.1 and 7.8 per cent.* Stein" found
considerable variations in the composition of coals from Saxony. It
is to be noted, too, that the rough-hewn coal often contains a different
* Cf. F. Fischer's Taschenbuch f. Feuerungstechniker, 5th ed., p. 57.
* Z, angew. Chem., 1893, 6, 575; 1899, 12, 334.
* Fischer's Jahresber., 1894, 7 ; Stahl u. Bisen, 1901, 21, 749.
4 /öld.
* Stein, Untersuchung der Steinkohlen Sachsens, 1857; Fischer, Chemische Technologie der
Arunº, vol. i., p. 503.
241
Q
242 FUEL ANALYSIS
amount of ash from the cobbles, so that, in sampling, relative propor-
tions of each grade of the material must be taken.
As an example of sampling, suppose that an average sample of the
fuel used for a boiler trial, extending over ten to twelve hours, is required.
A shovelful is taken from each consecutive barrow-load, or from every
second load, and put into a covered box; the coal is subsequently
broken down, mixed over, spread out in the form of a square, and
divided by the diagonals into four parts. Two opposite quarters
are thrown out, the remaining two mixed, broken down further, again
quartered, and so on, until only about 2 kilos (4 lbs.) remain. This
sample is then bottled and sealed. For accurate work, it is advisable
to prepare a second sample from the first half of the material thrown
out, and to use this for an independent analysis. As there is danger
of a loss of moisture during sampling, smaller samples, of about 50 g.
each, are taken from time to time and put into weighed stoppered
bottles, to serve for the moisture determination.
Preparation for analysis.--The samples are first coarsely crushed
and well mixed ; 200 g. are then finely powdered, care being taken
that none of the more difficultly crushed portions are thrown out.
Moisture.--This determination must not be carried out in open
vessels, as many fuels oxidise slowly in the air, on being warmed.” For
technical purposes, 5 to IO g. of the fuel are warmed between two watch-
glasses, or in a covered crucible, for a couple of hours, at a temperature
of IOS’ to I IO’, in an air-bath, then allowed to cool in a desiccator, and
Subsequently weighed. In the case of the 50 g. samples, previously
mentioned, the moisture is ascertained by weighing the stoppered
bottles with their contents before and after drying.
An alternative method is to effect the drying over concentrated
sulphuric acid, in a vacuum desiccator, for twenty-four hours.
Ash.-This estimation is especially necessary in the case of coal and
coke. From 2 to 5 g. of the finely powdered sample, according to
the probable quantity of ash present, are heated in a platinum dish.
Ti.e heating must at first be gentle, and the temperature must be
increased gradually, so as to prevent rapid formation of gas and con-
sequent mechanical loss of ash, in the form of dust; finally, a
moderate red heat is employed until the operation is complete. During
the later stages, the ash is occasionally stirred with a platinum wire.
The gradual heating is more especially advisable in the case of caking
coal, a small flame being employed at first, best after about two
hours' previous heating in an open dish, in an air-bath, at I2O to 150°;
this prevents sintering of the coal.
Some kinds of coke, and also furnace clinker, are difficult to burn
completely to ash, unless they are very finely powdered, and they
* Z, angew. Chem., 1899, 12, 465 and 766.
DETERMINATION OF ASH 243
require frequent stirring during the combustion. Excessive heating
is to be avoided, owing to possible loss of alkalis.
G. Lunge's method of determining ash may be strongly recom-
mended. A circular hole is cut in a piece of asbestos board and the
crucible is placed in this, as shown in Fig. IO6; the flame thus comes
º rºos. ----- rº 107.
in contact with the lower part of the crucible only, and the fuel itself
is in free contact with the air. If many ash determinations have to
be carried out, shallow rectangular trays of platinum (Fig. Ioz) may
FIG. 108.
be conveniently employed; samples of only I to 2 g. are taken, which
are heated gradually in a muffle furnace (Fig. 108). In this way a
number of determinations can be carried out simultaneously.


244. FUEL ANALYSIS.
After weighing the ash, it may be further examined; an alkaline
ash, or one which produces effervescence with hydrochloric acid, will
attack acid refractory materials; the ash may also be tested for
sulphates and phosphates, and its fusibility, etc., ascertained.
Residual Coke or Fixed Carbon.—From 1 to 1.5 g. of the sample
is heated in a covered platinum crucible, placed on a platinum
wire triangle. The heating is effected by a good Bunsen flame, the
crucible being completely enveloped by the flame; the bottom of
the crucible should be about 3 cm. above the top of the burner.
Heating is continued until flame no longer issues from under the
crucible lid."
According to Constam and Rougeout,” the results obtained by this
method are from I to 3 per cent. too high, and they give preference to
the following method, due to Broockmann.” The sample is placed in a
crucible, between 22 and 35 mm. deep, provided with a well-fitting
cover, in the centre of which is a hole 2 mm. in diameter; the bottom of
the crucible is placed about 6 cm. above the top of the burner, the total
height of the flame above the burner exit being 18 cm. The heating is
continued until there is no longer any flame burning at the hole in the
crucible lid.
A modified form of Henrich's method * has been adopted in the
United States, as the result of a report by a Special Committee on Coal
Analysis.” An ordinary platinum crucible, supported by a platinum
triangle, is used, placed 6 to 8 cm. above the top of the burner, and the
heating is continued for seven minutes. The determination is made on
the undried, finely powdered sample.
Gruner" recommends a short form of examination of coal, which
consists in ascertaining the yield of coke (reckoned ash-free), and
the percentage of moisture; the difference gives the amount of
volatile matter given off on coking. He considers that those coals
which give the smallest amount of coke have the lowest heat-pro-
ducing value on combustion, a statement which is not altogether
COrrect.
Von Jüptner's method" is similar ; his method of calculation of the
calorific power is open to criticism.
* Q/ Fischer, 7echnologie der Brennstoffe, vol. i., p. III ; also, Muck, Entwicklung der Stein-
Kohlenchemie, p. 18.
* Glückauf, 1906, 42, 48: ; / Soc. Chem. Ind., 1906, 25, 631.
* Muck, Chemie der Steinkohle, p. 32.
* Z. anal. Chem., 1869, 8, 133.
* J. Amer. Chem. Soc., 1899, 21, 1122; cf. also, Somermeier, ibid., 1906, 22, Ico2.
* Ping!, polyt, J., 1876, 219, 178.
' Die Bestimmung des Heizwertes von Brennmaterialien, by H. von Jüptner, in Sammlung
chemisch-techn. Vorträge, edited by Ahrens, vol. ii., pt. 12, 1898; cf. Z. angew. Chem., 1901,
I4, I26o. -
DETERMINATION OF SULPHUR 245
The appearance of the crust of coke, produced in the crucible from
finely powdered coal, is differentiated as follows:—
ſ
Porous all over or to
| near the edge . . I. Sand-coal (long flame dry coal).
Roug Black. Fritted hard, porous in
ough, tº
fine-grained. centre only . . • 2. Sintered sand-coal.
Fritted hard all over . 3. Sinter-coal (long flame fat coal).
Grey and hard, with blistered
\ appearance e º e . 4. Caking Sinter-coal.
Smooth, lustrous, and hard. e e º . 5. Caking coal proper.
Schondorff' gives the following data as characteristic:—
Class of Coal. Yºke. Character of Coke.
1. Long flame dry coal º . 50 to 60 | Pulverulent or sintered.
2. Long flame fat coal (gas coal) . 60 to 68 Fused, but considerably fissured.
3. Fat coal proper (furnace coal) . . . 68 to 74 | Fused ; tolerably dense.
4. Short flame fat coal (coking coal). 74 to 82 | Fused; very dense; few fissures.
5. Poor anthracitic coal . & e 82 to 90 Sintered or powdery.
In the case of caking coal, the increase in volume should be taken
into consideration.
Sulphur.—Sulphur occurs in coal as iron pyrites; it is seldom
combined with other metals. It is present also in the form of organic
compounds” and as sulphates. The determination of the character
and amount of these various compounds serves no useful purpose from
a technical standpoint; the important factor to know, is the amount of
sulphur present and the proportion which is detrimental to the process
in which the fuel is to be used.
During the distillation of coal, either in gas retorts or in coke ovens,
the sulphates are reduced to sulphides, pyrites forms ferrous sulphide,
and oxide of iron may be converted into Sulphide, by interaction with
the organic sulphur compounds. How much of the sulphur passes off
in the gas and how much remains in the coke can only be ascertained by
direct experiment and analysis.
When the coal is to be used for ordinary firing, only the “volatile
sulphur’ need be considered, that is to say, the Sulphur which passes
off in the products of combustion as sulphur dioxide and partly as
sulphur trioxide. Neither the sulphur remaining in the ash nor the
total sulphur give any useful information. On the other hand, in the
case of those smelting processes in which the fuel comes into direct
contact with the material, as in black-ash furnaces, blast-furnaces, etc.,
the estimation of the total sulphur is necessary.
The estimation of Total Sulphur is usually carried out by Eschka's
* Z. f. Berg-, Hutten-, und Salinenwesen, 1875, 149. * Cf. Z. angew. Chem., 1899, 12, 766.
246 FUEL ANALYSIS
method. About I g. of the finely powdered sample is mixed, in a
platinum crucible, with 2 g. of a mixture composed of two parts of well-
calcined magnesia and one part of anhydrous sodium carbonate. The
crucible, without the lid, is placed in a hole cut through an asbestos
board (Fig. IO6, p. 243), and is heated in such a way that the lower
half only of the crucible attains a red heat. Heating is continued for
about an hour, the mixture being frequently stirred with a thick platinum
wire. As soon as the grey colour of the mixture has given place to a
uniform yellowish, reddish, or brownish tint, the heating may be con-
sidered to be finished. After cooling, the crucible is placed in a beaker
and covered with water ; this is raised gradually to boiling, and ex-
traction is completed by successive additions of water. Bromine water
is then added until the liquid becomes slightly yellow, and the
solution warmed to effect the oxidation of any sulphides present. The
whole is then filtered, the filtrate acidified with hydrochloric acid, and
barium chloride added to the boiling solution. From the amount of
barium sulphate obtained, the sulphur is calculated in the usual way.
According to the proposal of Rothe,” the roasting of the mixture
may be carried out in a porcelain crucible, placed in a muffle furnace.
The duration of the operation may be shortened by carefully adding
sodium peroxide, little by little, to the roasting mass, until Oxidation is
complete.
C. Sundström,” as also Pennock and Morton,” use sodium peroxide
alone; Konek" determines the sulphuric acid, in the solution obtained
in the estimation of the calorimetric value of fuel, by Parr's method
(p. 256). A modification of this method is described by Schillbach."
With the object of avoiding the possibility of error, from sulphur
formed in the combustion of the gas used for heating, Brunck’ has
proposed a method for the determination of sulphur in coal, which
consists in heating I g. of the sample, mixed with 2 g. of a mixture of
2 parts of cobalt oxide and I part of sodium carbonate, in a current of
oxygen. The combustion is completed in fifteen minutes. The pro-
ducts of the oxidation are extracted with water, the cobalt oxide
filtered off, and the sulphuric acid estimated gravimetrically. With
certain coals, the oxide of cobalt may retain sulphur as a basic sulphate,
which must be tested for and determined, if present.
The Volatile Sulphur can be ascertained by subtracting the sulphur
of the ash from the total sulphur, or it may be determined directly, by
combustion of the coal in a current of oxygen. This latter determina-
tion is carried out in the same way as an organic analysis, but a larger
* Z. anal. Chem., 1878, 17, 497; J. Soc. Chem. Ind, 1889, 8, 361.
* Mitſ, techn. Versuchsanst, Berlin, 1891, Ioſ. * Z. angew. Chem., 1903, 16, 516.
* /. A mer. Chem. Soc., 1903, 25, 184. * Zbid., 1903, 16, Io&o.
* Ibid., 1903, 25, 1265. g 7 /bid., 1905, 18, 1560.
DETERMIN ATION OF ARSENIC, PHOSPHORUS AND NITROGEN 247
weight of sample is necessary (O.8 to I g.), and the combustion tube
should contain a short length of platinum clippings at n (Fig. I 12),
instead of the usual copper oxide. The sulphur dioxide and trioxide
produced are passed into hydrogen peroxide, and the sulphuric acid
thus produced is either precipitated as barium sulphate or is titrated
with M/IO potassium hydroxide.
Arsenic has been found in certain samples of English coal and
coke;" methods for its estimation have been described by Thorpe,” and
by M“Gowan and Floris.”
Phosphorus.--The estimation of phosphorus is necessary only for
metallurgical purposes. One to two g. Of the ash are digested with con-
centrated hydrochloric acid, the whole taken to dryness, and the residue
extracted with water containing a little hydrochloric acid. This solu-
tion, after the addition of nitric acid, is again taken almost to dryness
and precipitated with ammonium nitro-molybdate in the usual manner.
Nitrogen.—This determination is not often required. In the
Kjeldahl method," O.8 to I g. of the very finely powdered coal is mixed
in a flask of well-annealed potash glass, with I g. of finely ground
yellow mercuric oxide and 20 c.c. of concentrated sulphuric acid ;
the mixture is heated in a fume chamber for about an hour, on a hot
plate, until the more energetic part of the reaction is past, and is kept
subsequently at the boil for from one to two hours, on a wire gauze,
until the solution is colourless. All classes of coal, even anthracite,
dissolve in the time stated. After cooling, the contents of the flask
are transferred to a second flask of about # litre capacity, containing
a little water, and standing in cold water; I2O to 140 c.c. of pure
sodium hydroxide solution of sp. gr. 1.26 to I-28 are added (made by
dissolving about 3 IO g. of caustic soda sticks in water and making up
to a litre) and 35 C.C. of a solution of yellow sodium sulphide, con-
taining about 40 g. to the litre, together with a small piece of zinc,
to prevent bumping during boiling. The whole is then distilled for
from twenty to thirty minutes into a receiver containing 30 c.c. of N/2O
sulphuric acid. By employing a tall distillation flask of considerable
capacity, the introduction of a bulb between the flask and condenser,
for the purpose of catching alkaline liquid mechanically carried from
the flask, is rendered unnecessary. The sulphuric acid in the receiver
is finally titrated with N/2O baryta water, or with M/IO sodium
hydroxide solution, using rosolic acid as indicator.
According to investigations by Fischer,” the inconvenience of using
sodium sulphide and mercury may be avoided as follows:—About I g.
* /. Soc. Chem. Ind., 1901, 20, 437. * J. Chem. Soc., 1903, 83, 969.
* /. Soc. Chem. Ind., 1905, 24, 265.
* Cf. Z. anal. Chem., 1886, 25, 3I4, and subsequent papers.
* Fischer's Jahresher., 1894, 6.
248 FUEL ANALYSIS
of peat, powdered lignite, or very finely powdered coal, is boiled with 20
c.c. of concentrated sulphuric acid (free from nitrogen compounds), and
8 to Io g. of potassium sulphate, using the same precautions as de-
scribed above; decomposition is complete in about two hours, after
which the ammonia may be distilled over and estimated as usual.
The determination of Carbon and of Hydrogen. — Since wood,
in a finally divided condition, and more particularly peat, absorb
moisture very readily, after having once been dried, the accurate
weighing of such samples for ultimate analysis or for calorific power
determinations is a matter of some difficulty. This difficulty, as well
as the danger of error due to the oxidation of coal during drying
(p. 242), may be avoided by weighing out O-3 to O-4 g. of very finely
powdered coal, or about O.5 g. of finely rasped air-dried wood, and
compressing it in a cylindrical mould.
The kind of mould employed for the purpose is shown in Fig. IOQ.
A steel ring a is clipped to the bottom plate m, by means of a hex-
agonal nut n. The sample is filled in, and is compressed
(2, by the small ram s ; a strong copying-press, with iron
rº base-plate, serves well for the purpose. The nut n is
then unscrewed with the aid of a spanner, the inner
= Nº ring a placed on a circular plate, and the sample pushed
out by means of the ram s. The pressed cylindrical
samples are then put into stoppered weighing bottles
and dried at IoS’ to I Io°. -
Towards the end of the operation, completion of the drying may
be hastened by drawing air through the weighing bottle, or, in case
several samples are being dried simultaneously,
through the drying oven. To ensure that the
samples are not broken down by the too rapid
evolution of water-vapour during drying, the tem-
perature is raised gradually to IOS’ to 108°; this
precaution is especially necessary in the case of
wood and peat. For the drying operation Fischer
uses the apparatus shown in Fig. I IO.' The cylin-
drical case is made of sheet copper and is IO cm. in
diameter; the lower portion is rounded as shown,
and is provided with a brass tube c, to the lower
end of which is attached a short calcium chloride
tube, for drying the air to be admitted to the apparatus. The cover t is
provided with an asbestos sheet on its under side. An asbestos shield,
resting on a ring r, surrounds the whole, at a distance of about 5 mm.
from the outside of the case. The gas tube 6, is bent in the form of a
ring and is pierced with four holes, thus providing four tiny gas flames.
* Z. angew. Chemie, I894, 7, Ig.
S
FIG. 109.
FIG. I.10.

THE DETERMINATION OF CARBON AND HYDROGEN 249
Two perforated trayss, ensure uniform heating of the samples contained in
the weighing bottles g. The amount of gas required is extremely small.
The determination of the carbon and hydrogen, in the dried sample,
is carried out in the same way as the ordinary ultimate analysis of
organic compounds. Fischer" recommends the following modifications
in the process. The two
end-plates, b and p, of the
combustion furnace (Fig.
I I I) are joined at their
lower ends to the base-
plate, and are connected
above by the two iron
rods u. The tiles s stand
in a shallow channel run-
ning along each side of the
furnace, and their upper
ends rest against the rods u.
The burner flame, protected
from draughts by the side- -
plates a, is thus caused to pass completely round the combustion tube,
which lies in a semicircular sheet-iron bed o. A combustion tube of
hard glass is used; aa (Fig. I 12) are two asbestos plugs, wrapped in
very thin platinum foil, and between these is a layer of granular cupric
oxide n, which in the case of coal containing much sulphur is replaced
FIG. I.12.
by a shorter layer of coarse lead chromate. After the introduction of
the platinum boat m, containing the sample, the open end z, of the
tube is connected to an oxygen supply, and the calcium chloride tube c,
fixed directly into the other end zv.
Before beginning the combustion proper, the tube is laid on the
bed o, the tiles are placed against the rods u (Fig. I I I), and the
copper oxide is heated to redness by means of three or four flat-flame
burners B (a portion of a is omitted in the figure so as to show the
burners). A single burner suffices for heating the other portion of
the tube. A current of air, washed by means of potassium hydroxide
and dried by sulphuric acid, is first passed for about ten minutes, and
the tube is then allowed to cool. The stopper u is then removed,
the platinum boat containing the weighed sample introduced and the
* Chemische Zechnologie der Brennstoffe, vol. i., p. 127.


250 - FUEL ANALYSIS
stopper replaced. The calcium-chloride tube, etc., are attached, and
the combustion is then carried out in the usual manner, in a current
of oxygen. -
Such modern improvements of the combustion process, as the use
of electrically heated furnaces and of Dennstedt's 2 simplified method
of organic analysis, are of course applicable to the estimation of carbon
and hydrogen in fuel.
THE DETERMINATION OF THE CALORIFIC POWER OF FUEL
The calorific power of fuels may be calculated from their elementary
composition, or it may be determined by direct experiment. The
formula of Dulong serves for the calculation; referred to liquid water
at O’C. as product of combustion, it is:—
[8 IOO c-- 3422O (h—g)+25OO s]+ IOO ;
or, referred to water-vapour at 20°C., which is more practical,
[8 IOO c-H 2880O (h–?)+25OO s—600 wi-- IOO,
in which c, h, o, and s represent the percentages of carbon, hydrogen,
Oxygen, and sulphur respectively.
The value in B.T.U. is obtained by multiplying the value in calories
by 3-97.
Dulong's formula gives fairly good results in some cases, whilst in
others the results are from 2 to 6 per cent. out; in the case of wood,
peat, and lignite especially, the results are considerably too low.”
Many modifications of this formula have been proposed, but they are
all only approximations. The true calorific power of fuels can only
be reliably determined by means of direct calorimetric measurements.
Numerous forms of apparatus have been proposed for this purpose,”
of which those in which the combustion is effected in oxygen, either
at atmospheric pressure or in the Berthelot bomb, are the most
satisfactory.
Fischer's Calorimeter.—For carrying out the combustion at
atmospheric pressure, Fischer" makes use of a combustion vessel of
silver or of nickel-plated copper, which is clipped to the bottom of a
nickel-plated copper cooling vessel B (Fig. I 13), by means of three
bent feet f. In the tube d of the cover is fixed a tube e, which is
widened out into a plate at its lower end and carries a cylindrical
vessel ?, made of platinum or of nickel. Attached to the ring l is a
small basket s, made of either platinum or nickel gauze, in which the
* Cf. Blount, Analyst, 1905, 30, 29; Morse and Taylor, Amer. Chem. J., 1905, 33, 591.
* Ber., 1906, 39, 1623.
* Z. angew. Chem., 1893, 6, 397 and 575 ; Fischer's Jahresöer, 1893, I.
* Fischer, Chemische Zechnologie der Brennstoffe, vol. i., pp. 129, 148, and 530.
* Fischer's Jahresher, 1885. I2O7. Z. angew. Chem., 1892, 5,542; 1894, 7, 19; 1901, 14,444.
FISCHER'S CALORIMETER 251
sample is placed. If the sample is small, it is advisable to use a round
basket. The opening in the bottom of the vessel p is covered with
a piece of platinum gauze c, and underneath is a plate v. Oxygen
is conducted into the apparatus through the glass tube a, and the
products of combustion generated in s are carried downwards by the
pressure of the oxygen, through the gauze c, and against the plate v,
So that complete mixing and combustion is attained. The gases then
pass through the outlet i into the
flattened vessel ce (which may consist *—tº r
of a spiral tube), through the tube g 7?? /
and its glass extension b, to waste;
or they may be collected in a gas-
holder, the oxygen being subsequently
freed from carbon dioxide, and used
for sulphur estimations and the like. #==
The cover n of the apparatus is -
made in two halves. One half is
fastened down by means of a screw,
and through it pass the rods of the
stirrer rm and a thermometer t; the
thermometer scale is graduated to
read to #1", so that differences of
++, can be read by means of a
cathetometer.
The space between B and D is
loosely packed with dry swan's-down,
and the joint between the edge of the tºpy— Sºlº
copper vessel B and the wood case D . §
is coated with shellac varnish, to pre- §§
vent absorption of moisture. If the FIG. 113.
apparatus be kept dry when not in
use, the heat loss is very small, amounting to O'ooz5°, per degree differ-
ence of temperature, per minute. This value and the water equivalent
of the apparatus are determined, either by the method of mixture, or
the water equivalent may be ascertained by weighing, thus:–
77/
Tºv
§
}
-&*;
/n):
.
Y
{
- * |I||
3|:/E/Ejº,
N
Grams. Sp. heat. Cal.
Weight of containing-vessel with stirrer . 916.5 O-O95 = 87. I
Weight of silver calorimeter . . . . 278.4 O.OS6 = 15-6
Allowance for platinum extras, thermometer, etc. . © . 2-3
IO5
The value found by means of water, in the above instance, was 112
cal. ; it is evident, therefore, that a certain amount of heat is communi-
* Fischer's Jahresher., 1885, 1208.

252 FUEL ANALYSIS
cated to the part of the apparatus surrounding B, so that about 5 cal.
should be added, on an average, to the value obtained by weighing.
The ignition of the charge is effected in the usual way, by the
introduction of a splinter of red-hot charcoal about 1.5 mg. in weight;
the allowance necessary for this is 12 cal. The electrical ignition con-
trivance described on p. 215 may also be used; in this case about 2 mg.
of gun-cotton or cotton wool are placed in the platinum spiral beforehand.
With some practice, it is quite easy to carry out a combustion so
that the whole of the material is completely burnt; an examination
of the gases passing from 6 is then unnecessary. There should be no
deposit or smell noticeable on opening the calorimeter, after a com-
bustion. Should it be deemed desirable, however, to examine the
products of combustion, the gases escaping from 6 are passed through
a bottle containing potassium hydroxide solution, so that the velocity
and pressure of the exit gases may be observed; following this is a
Soda-lime tube 3 cm. in diameter and 20 to 25 cm. long, a similar
tube containing calcium chloride, and then a combustion tube 15-2O
cm. in length and containing a 4 cm. layer of copper oxide, placed
between pieces of coarse platinum or nickel gauze. To the end of
this tube a calcium chloride tube, 2.5 cm. in diameter, and a similar
tube containing soda lime, are attached. From the increase in weight
of the two latter tubes, the amount of the incompletely burnt gases
can be determined.
The manipulation is as follows:—A compressed sample of the
fuel is taken, by means of forceps, from the weighed drying bottle, the
weight of the sample being ascertained by difference; this is placed
in the small basket s, p is then fixed on to the end of d, the cover
fixed firmly on to the calorimeter, as shown in Fig. I 13, a small
quantity of a mixture of melted caoutchouc and vaseline being used
to make it tight. The calorimeter is then placed in B, the tube a
connected up to the oxygen supply, the necessary quantity of water
poured into B, the cover n laid on, the thermometer t inserted, and
the temperature observed until constant. The oxygen supply is then
gently turned on, a small piece of glowing charcoal thrown in through
a, and the cap on a quickly replaced; at the same time the oxygen
Supply is increased. In a couple of seconds or so it will be seen
that combustion has begun, and then oxygen is supplied, at such a rate,
that about 2 litres pass in the course of half a minute. As the energy
of the combustion slackens, the oxygen supply is reduced to about
one-third. The combustion occupies only three-quarters of a minute.
In the course of a further couple of minutes, if the calorimeter is of
silver, or of from three to four minutes, if the calorimeter is of nickel,
the thermometer should show a maximum reading, which indicates the
completion of the experiment.
FISCHER'S CALORIMETER - 253
Fischer uses two oxygen gasometers, with at least half a metre
water-pressure, for the combustion; these are connected to a glass
T-piece, and thence to the calorimeter. Assuming c to be connected
to the calorimeter, a and 6 are connected to the oxygen cylinders by
rubber tubes provided with screw clips. The clip on a is arranged
to allow of the passage of about I litre of oxygen in a quarter of
a minute, and a second easily removable clip is used for completely
cutting off the supply. On the tube connected to 6 is
a screw clip, by means of which the oxygen supply can b
be readily increased, as soon as ignition has taken place. : —:
On the first sign of combustion the removable clip is
taken off a, so that an excess of oxygen is at once available. When
the vigour of the combustion shows signs of decreasing, a is again
closed, and the combustion is finished by means of the supply through b.
It is well to pass about 2 litres of air through the complete apparatus
in half a minute before proceeding to a determination, so as to ascer-
tain that all connections are sufficiently open. No mechanical loss, even
in the case of peat, can occur if this test is satisfactory.
Fischer uses I600 g. of water with this apparatus, so that the total
calorimeter value is I600+ I IO = 17 Io. The heat loss of the calorimeter
for 1° temperature difference, per minute, is only 4 cal., if the packing be
kept dry. As an experiment lasts scarcely three minutes, this correction
is small, being only 15 to 20 cal. If the cooling water is 1° to I°-2
below the temperature of the room, this correction may be neglected.
The following figures will serve as an example:—
Temperature rise of water at end of I minute . I°.2
25 35 2 minutes . 2°-5
33 22 3 25 2°.9
53 23 4 32 3’-o
The mean temperature for the individual minutes may be taken
a.S —
I minute . o°.6 corresponding to 2-4 Cal.
2 minutes I°.8 22 7°2 2,
3 99 2°.7 35 IO-8 ,
4 93 2°-9 55 I I-6 is
32-O cal.
For technical purposes, therefore, intricate calculations are un-
necessary; it is quite sufficient to take two-thirds of the temperature
rise as a mean, in this case 2°; thus 2 × 4 × 4 = 32. The correction is
usually considerably less than this.
The gases escape at a temperature 1° to 2° higher than that of the
cooling water, according to the rapidity of the oxygen current; that is
to say, about 4° warmer than the oxygen employed. If 3 to 4 litres
254 - FUEL ANALYSIS
of oxygen are used, the loss of heat from this source is about 6 cal.,
which is practically the same as the heat produced by the charcoal used
for the ignition. These two corrections, therefore, practically neutralise
each other.
The greater part of the water formed in the calorimeter is con-
densed. If the calorific power is referred to liquid water as product of
combustion, 6. I cal. must be added for every IO mg. of water-vapour
given off; but if calculated to water-vapour at 20°, then 6. I cal. must
be subtracted for every IO mg. of water condensed. The amount of
water condensed can be ascertained by drying the calorimeter externally,
and weighing, after removing a and b; it is then washed out with
distilled water, to remove any sulphuric acid that has been formed,
dried by warming, and again weighed.
The Berthelot Bomb Calorimeter."—The bomb (Fig. I 14), con-
sists of an iron tube, lined with platinum, to which a strong bottom
IO mm. thick and a cover of about 30 mm. thick-
ness, are screwed and soldered. It has a capacity
of about 250 c.c., and should be tested to 50 atmo-
spheres. It is closed by the headpiece A, pro-
vided with a screw plug a.
At 66, a flange (not shown) can be fastened on
by screw bolts. An iron rod c is screwed into A,
% which also carries the insulated rod d, for electrical
% connection. Platinum wires f and g, of about
% O.8 mm. thickness, are screwed and fastened firmly
% into c and d respectively; these carry a small fire-
%
%
%
%
clay tray e. The insulation of d is effected by a
piece of thin-walled rubber tubing i, pushed over
the thickened portion h, of the rod; by passing
the tubing through the slightly conical hole in A,
then pulling the tube and at the same time sharply
pushing the rod d, a satisfactory fixing is effected.
The lower portion of the rubber tubing is cut off, so that it shall extend
about I cm. beyond the thickened portion of d, whilst at its upper end
it projects somewhat beyond the hole. To prevent burning of the rubber
inside the hole, its lower portion is packed with asbestos. The com-
pressed sample of coal is put into electric connection with f and g by
wrapping it round with platinum wire. Sheet lead is used to make
the screw-plug and the headpiece quite gas tight. When the headpiece
has been screwed into position, the apparatus is filled with oxygen.
For this purpose connection is made with a cylinder of compressed
oxygen, provided with a pressure gauge. In order to fill the bomb, the
screw-plug a is eased to the extent of a full turn, and the tap of the
- * Comptes rend, 1892, 115, 20I. -
º ſ
22,223.2% €,
FIG. 114.



THE BERTHELOT BOMB CALORIMETER 255
oxygen cylinder carefully opened. As soon as the pressure in the
bomb has reached 6 atmos., the cylinder, tap is closed, and the oxygen
is allowed to escape from the bomb, by somewhat slackening the
flange bolts in 66; the escaping oxygen carries most of the nitrogen,
originally contained in the bomb, with it. Oxygen is then again
admitted and the pressure is allowed to reach 12 atmos., when the
supply is cut off, the screw-plug closed, and the bomb placed in the
calorimeter vessel, as shown in Fig. I 15. -
This consists of a covered metal vessel G, suspended in the vessel
H, at a distance of about 2 cm. from its walls; G contains a litre
of water. K is a thermometer, reading to
Tºo", and N a stirrer. The stirrer consists
of a circular disc attached to a couple of rods
and worked by means of a string passing
through a ring, as shown in the figure.
Wires pass from the mercury contacts i &
and AE (Fig. I I4), to a battery. When all J
is in readiness, the apparatus is allowed to !
stand until two successive readings of the
thermometer, taken at an interval of five - +
minutes, show that the temperature is con- -
stant. The electric circuit is then completed, AM
and the ignition of the sample thus effected.
Stirring of the water is continued until the
temperature has attained a maximum and
begins to fall. The initial and final tempera-
tures are noted.
Kroecker' furnishes the upper portion of
the bomb with two passages, capable of being
absolutely closed. One is intended for the
introduction of the oxygen, the other is con-
tinued inside, to the bottom of the bomb, by
means of a platinum tube; this latter is
intended for the removal of the gaseous products of combustion and
the condensed water, after the combustion is finished. Both openings
can be closed by plugs. In addition to these two passages is a third,
through which an insulated platinum wire is passed.
Of other modifications of the Berthelot bomb the following may be
mentioned :—
Scheurer-Kestner and Meunier-Dollfus” use a platinum-lined steel
bomb. W. Hempel” makes use of the Berthelot principle, and inserts a
platinum wire into the sample, for the purpose of electrical ignition.

—
|
§
N
FIG. I.15.
* Z, angew. Chem., 1897, Io, 327; 1898, 11, 865; 1901, 14,444.
* Bull. Soc. Ind. Mulhouse, 1891, 577. * Z, angew. Chem., 1892, 5, 389; 1896, 9, 486.
256 - FUEL ANALYSIS
The Mahler bomb' is a cylinder of mild steel 8 mm. thick, and of
about 650 c.c. capacity; it is nickel-plated on the outside and enamelled
inside, so as to withstand the corrosive and oxidising action of the
products of combustion. The shell narrows somewhat at the top, and
the cover, which carries the tray for the sample, is screwed on the
outside of this. The cover is provided with a lead ring, fitting into a
groove in the edge of the bomb. The gas is admitted through a
conical nickel-steel valve in the centre of the cover. The calorimeter
itself is of thin brass, and the amount of water employed is 2.2 kilos,
so that error due to evaporation is eliminated; the correction for loss
of heat during the experiment is almost negligible.
The Mahler-Donkin * bomb calorimeter is constructed of non-
corrosive metal, and is gold-plated inside.
Lunge” remarks that the Mahler bomb is costly, if one is to employ
all necessary appurtenances. Including the delicate thermometer,
stirrer, bomb, etc., the cost approaches £60, and accurate results can
only be expected if the apparatus is used in a special room, of constant
temperature. The use of highly compressed oxygen and the other
operations entailed in connection with all bomb apparatus require
considerable skill, so that the apparatus is, in his opinion, unsuitable for
inexperienced workers.” One of the chief drawbacks is the liability of
the enamel lining to injury, such injury not being repairable.
Langbein” also points this out, and is in favour of lining the bomb with
platinum.
Holman and Williams" replace the enamel lining of the Mahler
bomb, by electroplating the inside with gold, and have also improved
the methods of making the apparatus tight.
The Mahler bomb has been successfully used by J. Pfeiffer," for the
determination of sulphur in petroleum and other fuels, side by side with
the estimation of the calorific value. From 2.0 to 2.5 g. are taken for
the determinations, and the combustion is effected under a pressure of
30 atmos. of oxygen; the total sulphur remains in the bomb, after the
combustion, in the form of sulphuric acid, and is determined gravi-
metrically.
Parr's Calorimeter.—In this apparatus the sample is burnt with
sodium peroxide. A nickel-plated vessel A (Fig. I I6), of about 2
litres capacity, serves as the calorimeter. This is placed inside a
wooden vessel C, which itself stands in a similar vessel B. The two
air-spaces, a and 6, and the double cover G, containing an air-space g,
1 Cº. Mahler, Aitudes sur les Combustibles, 1903; also, J. Soc. Chem. Ind., 1892, 11, 840.
* Cf. Dowson and Larter, Producer Gas, .* Ibid., 1900, 13, 1235; 190I, I4, I260.
p. 22I. * Cf. Gill, Gas and Fuel Analysis, Igor,
* Z, angew. Chem., 1901, I4, 794. p. 57.
* Cf. ibid., 1898, 11, 865. Bull. Soc. de Sciina, 1900, 8, No. 6.
PARR'S CALORIMETER 257
and the wood material of the vessels, provide sufficiently for all practical
purposes, against loss of heat. The reaction vessel D (Fig. I 17) consists
of a strong nickel-plated brass cylinder, the ends of which are firmly
screwed on, and provided with leather washers; it has a capacity of
about 35 c.c. The base J rests on a conical support F, forming part of
the inner vessel E ; from the upper
end of D passes the tube H, which
extends right through the cover G,
and is provided with a pulley P.
A series of four small propeller T
blades, h/, is attached by spring -
clamps to D. D is rotated by =# (?
means of a water turbine, and by • Ży
employing a sufficiently high speed, 8%i || A | A fief
the water can be kept in circula- zila a
tion through E in the direction of FN
the arrows; a thorough mixing is 7|D|7.
thus produced, and an even tem-
perature maintained. The blades h
are so fixed as to cause the current
to flow downwards next to the
cylinder D and upwards outside E,
when the rotation is right-handed. Inside the tube H (Fig. I 17) is a
narrower tube L, notched on One side and continued downwards to a
conical valve K. By the action of a spiral spring M, connection
with the interior of D is cut off, unless the tube L is depressed by
pressure at N ; any escape of gases, during combustion, is thus pre-
vented. Ignition of the charge is brought about by dropping a
red-hot piece of iron wire into L, then sharply depressing N, when
the wire drops through K into D. A delicate thermometer, held in
position by a thick rubber ring and reaching about half-way down the
vessel A, is passed through a hole 8 to 9 mm. wide in the cover G.
The double vessel B C is placed on a firm table, within reach of the
necessary motive power. Two litres of water are poured into A before
it is placed into position, so as to avoid water being splashed into C ;
for if the outer side of A or the inner portion of C should be moist
this would ultimately produce an error in the results, owing to its
evaporation. The temperature of the water should be about 2° below
the temperature of the room. A is carefully set into C, and is then
ready to receive the reaction vessel D.
The latter is carefully dried internally and externally, which is best
done by gently warming on a sand-bath or hot-plate, the base J is
screwed on and tightened with a key. About Io g. of sodium peroxide
(passed through a sieve of 1 mm. mesh) are then introduced. This
R
FIG. 116.

258 FUEL ANALYSIS
reagent is best kept in a wide-mouthed stoppered bottle, along with a
small measuring vessel, supplied with the instrument; the measuring
vessel, which holds Io g. of the reagent, is made of nickel-plated brass,
and is provided with a small handle.
After introducing the sodium peroxide into D, O-5 to I g. of the
coal sample, together with any other necessary reagents, is put in, the
cover screwed on, and the whole well shaken, K being kept closed
by an upward pressure of the finger on N, otherwise a small quantity
of the mixture may be shaken into L. The vessel is then tapped, to
bring the whole of the mixture to the bottom, the working of K
tested to see that it acts smoothly, and the blades h clamped on ;
D is then placed inside A so that it rests on the conical peg at the
bottom. -
Lunge" recommends the following method for carrying out the
determination :-
The coal to be tested is powdered, and should be passed through
a wire sieve of O.3 mm. mesh ; in the case of hard coals and anthracite,
a more finely divided sample should be used, which may be obtained by
placing a piece of bolting cloth under the wire sieve. Lignite must be
previously dried, for about an hour, at a temperature of IO5° to I Io’;
exactly I g. of this substance is then weighed out and shaken into
the reaction vessel, previously charged with the Io g. of sieved sodium
peroxide, after which the vessel is closed and shaken for one to two
minutes, as described above. Bituminous coals do not require previous
drying, unless the moisture exceeds 2 to 2% per cent. In the case of
coal, exactly O-5 g is weighed out, and to this is added exactly or 5 g.
of finely powdered pure tartaric acid; this mixture is then added to
the usual IO g. of the peroxide. The addition of potassium per-
sulphate, which has been suggested as necessary for the combustion of
hard coals and anthracite, is regarded by Lunge and Grossmann” as
superfluous and as the source of discordant results.
The cover G having been placed in position, the pulley P is attached,
and the thermometer inserted in the cover, as shown in the figure.
The turbine is then set in motion so that the pulley revolves in the
direction of the hands of a watch. A speed of about I 50 revolutions
per minute is kept up until the thermometer shows a constant tem-
perature, which takes about three minutes; this temperature is noted,
and the motor is kept running at the same speed till the conclusion of
the experiment.
Ignition is effected by means of a piece of red-hot iron wire, 2.5
mm. in diameter and IO mm. long; a piece of this size should weigh
about O'4 g., and the same piece can be used a good many times before
* Z. agnew. Chem., 1901, I4, 794 and I27o ; I903, I6, 911 ; Igoš, 18, 1249.
* Ibid., 1905, 18, 1249.
PARR'S CALORIMETER 259
its weight falls much below O-4 g. Experience has shown that soft iron
is more suitable for the purpose than nickel, copper, silver, or platinum.
It does not melt, as the first three of these do, forming peroxides, but
it becomes covered, when first used, with a hard, black, adhesive coating
of the magnetic oxide, which protects it for a considerable time.
Using the above weight and at a temperature of, say 700°, the heat
introduced by the iron amounts to O-4 × O. I2 × 700 = 33.6 calories, cor-
responding to a temperature rise in the calorimeter, of O'oï6. As the
readings are carried out to O’.OO5, a constant deduction, fixed at O’ or 5
is made. The wire, held pointing downwards with a pair of bent
tweezers, is heated to redness and dropped through N ; N is then
sharply depressed with the tweezers, so that the iron passes through
K, without permitting any escape of gas through L. A hissing noise,
lasting for several seconds, denotes the combustion of the charge,
and the thermometer rises, at first rapidly, then more slowly; after
about four or five minutes the maximum temperature is attained, which
remains constant for about five minutes, when the reading is taken.
The experiment is then finished. The motor is stopped, the thermo-
meter, pulley, and cover are taken off, the vessel A, with contents, is
taken out bodily, and the agitators hå are removed from D, which is
taken to pieces and placed in a basin of warm water, so as to dissolve
out the contents. On neutralising the solution so obtained, with
hydrochloric acid, the presence of any unburnt fuel can be detected;
should any be present, the results are worthless. -
Calculation of results. The water value of the instrument was 123.5
g. ; including the water, therefore, the total value is 2123.5 g. Of the
heat tº —t produced, tº being the final and t the initial temperature of
the calorimeter, 72.5 per cent. is regarded as due to the combustion
proper, and 27.5 per cent, to the reaction of the products of combustion
with the sodium peroxide and oxide. When I g. of the fuel is burnt,
as is the case with lignite, the number of calories produced is thus
o,725 x 21:23.5 (t’—t)= 1540 (t’—t). It is simply necessary therefore to
subtract O’.OI5 from the recorded value, t’—t, on account of the heat
introduced by the iron wire, and to multiply by 1540; the result gives
the calorific power in centigrade units. With coal, only O. 5 g. is used,
so that the recorded temperature difference is multiplied by 1540 x 2 =
308O. It is necessary first, however, to subtract the heating due to the
combustion of O.5 g. of tartaric acid, and that due to the iron; careful
experiments have shown this to be o’.832. If the sodium peroxide
contains an excessive amount of moisture, as may be the case if a
considerable quantity of peroxide is kept and frequently opened for use,
the results will come out too high. It is then necessary to make a blank
test, using O-5 g. of tartaric acid, and about two-thirds of a measure
(about 7 g.) of sodium peroxide. If the calorimeter temperature rises
260 - FUEL ANALYSIS
more than o’.832, it will be necessary, in future use of this stock, to
take such extra rise into consideration, by allowing o". I5 for every
o°. I excess rise obtained in the blank experiment (o’. 15 being, of
course, the allowance for a full measure of Io g, of peroxide).
Parr's method of determining calorific power is more convenient,
cheaper, and far simpler in manipulation, than any of the processes
involving the use of a bomb, but it has certain obvious defects. In the
first place, the combustion may be either too violent or incomplete, if the
proportion of added materials is not appropriately adjusted; secondly,
the heat generated by the absorption of the water and carbon dioxide,
formed in the combustion by the sodium peroxide, has to be determined
empirically for each instrument; and thirdly, the value of the coefficients
given above does not hold for all kinds of coal, and still less for other
fuels, such as lignite, petroleum, etc. These points have been specially
investigated by Lunge and Offerhaus," and by Lunge and Grossmann.”
The latter conclude that the method is unreliable for coals having a
calorific value less than 7500 cals. For coals having a higher calorific
value, the method is of practical value, under the conditions given above.
Both the coal and the sodium peroxide must be finely powdered and
intimately mixed, and the peroxide must be of good quality. The
heat value of the tartaric acid, together with that of the ignition wire,
to be subtracted from t' -- t, should be O’.832, as given above, instead
of oº.85 previously accepted, and the coefficient should be 1540 as
stated, in place of 1550 as formerly used. The latter value holds, of
course, only for a calorimeter of the water value 2123.5 when charged.
Langbein” discards Parr's process as altogether unreliable, but this
conclusion is not justified. .
Lewis Thompson's Calorimeter, in which the fuel is oxidised by
a mixture of potassium chlorate and nitre, is largely used for technical
work. The determinations can be made very rapidly, and although the
readings necessitate an empirical correction, the results are sufficiently
accurate for many practical purposes.
The apparatus (Fig. I 18) consists of a narrow copper cylinder or
furnace B, a wider copper cylinder F, with small perforations round the
base, and provided with a long narrow tube terminating in a tap and
a glass calorimeter vessel A, graduated for a content of 2000 g. and of
29,0Io grains. F is supported by the base C, in which it is held by the
springs E.; D is a socket to carry the furnace B.
The determination is made as follows:–2 g of an average sample
of the fuel, finely ground and sieved, are thoroughly mixed with about 20
to 24 g. of a well-dried mixture of three parts of potassium chlorate and
one part of nitre; this is shaken down into the furnace, previously
placed in the socket D, and ignited by a short fuse, consisting of a piece
* Z. angew. Chem., 1903, 16, 911. * Ibid., 1905, 18, 1249. * Ibid., 1903, 16, Io'ſ 5.
LEWIS THOMPSON'S CALORIMETER 261
of ordinary wick soaked in potassium nitrate solution and subsequently
dried, the end of the wick being embedded in the mixture. The cover
F is then immediately fixed over the furnace, with the tap closed and
the whole lifted and placed in A, previously charged with 2000 g. of
water. Combustion soon commences and proceeds rapidly: the gases
generated pass out through the holes in the base of the cover, and rise
through the water. When the combustion is ended, which should
occupy about two minutes, the tap on F is opened and the contents of
the furnace thoroughly mixed with the water, by using the cover as a
H-2000 grims.
iſiºn grs
t
FIG. I.18.
stirrer. The increase of temperature is read to within o’.os. The
initial temperature of the water in A should be from 3° to 4° below
the temperature of the room.
As the result of experiment, it has been shown that the average
total of loss of heat in using the apparatus, due to that carried off by
the escaping gases, that taken up by the apparatus itself, and that
absorbed by the added salts and the products of combustion, is about
one-tenth of the total, and this empirical correction is added to the
observed rise of temperature, in each determination.
By taking 3O grains of fuel and placing 29,0IO grains of water in A,
the evaporative power, in British units, is obtained directly, if the
temperature change is measured in Fahrenheit degrees, since the latent
heat of vaporisation of water is 967 B.T.U. and 967 x 30 = 29,0Io grains.
J. W. Thomas' has given a considerable amount of information
regarding the capabilities of the Lewis Thompson calorimeter, and has
* Chem. Wews 1881, 34, 135.




262 FUEL ANALYSIS
shown that the more bituminous a coal is, the better are the results; he
does not appear, however, to have estimated the unburnt coal, of which
there is frequently a considerable amount. W. Thomson" has pointed
out most of the objections to the instrument. Scheurer-Kestner” has
compared it with a Favre and Silbermann's calorimeter, and states
that with 15 per cent, addition for losses, the results agree within a
maximum of 4 per cent.
W. Thomson's Calorimeter* in some respects resembles the pre-
ceding, but combustion is carried out by means of oxygen, and the
escaping gases, during their passage through the water, are baffled by
a series of gauze rings. The addition of similar rings on the combustion
chamber of the Lewis Thompson apparatus is of distinct advantage
when using that apparatus.“
An improved form of the Thomson calorimeter has been devised by
Gray,” in which the gauze baffles are replaced by thin perforated brass
discs, and which is provided with an arrangement for the electrical
ignition of the coal. -
Rosenhain" has also modified the Thomson calorimeter in the
details of construction, to render it more accurate for technical work,
including its application to the calorimetry of volatile liquids.
Darling" has described an easily made apparatus, similar in principle
to that of W. Thomson, but providing more intimate contact of the
products of combustion with the cooling water.
The comparative value of the results obtained by the Lewis
Thompson, the William Thomson, the Fischer, and the Mahler calor-
imeters, has been studied independently by Brame and Cowan,” and by
Gray and Robertson.9 These authors agree, that in order to obtain
accurate results, a bomb calorimeter should always be employed. The
only objection to its use is that of expense; at the same time it is not
an instrument suited to the more expeditious requirements of chemical
or of engineering works. The results obtained with the other calor-
imeters specified are always too low, owing to the difficulty of effecting
complete combustion. Brame and Cowan state that the amount of
unburnt coal should always be determined, in using the Lewis
Thompson calorimeter; it then gives fair comparative results for
bituminous coals. They found the calorific value from I-8 to 6.2 per
cent. lower than with the Mahler bomb; Gray and Robertson obtained
1 / Soc. Chem. Ind., 1886, 5, 58.I.
* Bull. Soc. Ind. Mulhouse, 1888, 506.
* /. Soc. Chem. Ind., 1886, 5, 58.I.
* Brame and Cowan, ibid., 1903, 22, 123.I.
* Ibid., 1906, 25, 409.
* Phil. Mag., 1902 [iv.], 457; J. Soc. Chem. Ind., 1906, 25, 239.
7 Engineering, 1902, 801 ; / Soc. Chem. Ind., 1902, 2I, 927.
* /. Soc. Chem. Ind., 1903, 22, 1230. * Ibid., 1904, 23, 704.
W. THOMSON'S CALORIMETER 263
results from O-9 to 13-o per cent, lower. The results obtained by
Brame and Cowan with the Fischer calorimeter were from 2.6 to 5.7
per cent. lower than with the Mahler calorimeter; the results were
concordant with the same coal. With the W. Thomson calorimeter
the results were from 1.8 to 6.9 per cent, lower than with the bomb;
the comparative values found by Gray and Robertson were from O-7
to 2.9 per cent, lower.
LITERATURE
ABADY, J.-Gas Analysts’ Manual, 1902.
DOWSON, J. E., and LARTER, A. T.-Producer Gas, 1906.
FISCHER, F.—Die chemische Technologie der Brennstoffe, 1880-1901.
FISCHER, F.—Taschenbuch für Feurungstechniker, 5th edition, 1885.
GILL, A. H.-Gas and Fuel Analysis for Engineers, 1901.
MAHLER, P-Études sur les Combustibles, 1903.
POOLE.-Calorific Power of Fuel, 2nd edition, 1903.
ROBERTS-AUSTEN, Sir W.-An Introduction to the Study of Metallurgy, 5th edition,
I902.
SEXTON, A. H.-Fuel and Refractory Materials, 1897.
MANUFACTURE OF SULPHUROUS ACID,
NITRIC ACID, AND SULPHURIC ACID
By Professor G. LUNGE ; translated by JAMES T. Conroy, B.Sc., Ph.D.
SULPHUROUS ACID
FORMERLY sulphurous acid was produced almost entirely for con-
version into vitriol by the lead-chamber process; to-day it is made
in considerable quantity in connection with the sulphite-cellulose
industry, and to a lesser extent for conversion into liquid sulphur
dioxide. On this account it is accorded separate treatment.
RAW MATERIALS
I. SULPHUR (NATIVE SULPHUR)
Sicilian native sulphur is shipped in the form of blocks, each weigh-
ing from 28 to 30 kilos; owing to the friable nature of the sulphur,
these blocks are usually shattered during transport into lumps of
varying size and to some extent to powder. The best quality (“firsts,”
“prima Lercara’’ or “prima Licata”), consists of large, shining pieces
of amber-yellow colour. The second quality (“seconds,” “seconda
vantaggiata”) is also of a good yellow colour, but not quite so brilliant
in appearance. The bulk appears in commerce as a third quality
(“thirds,” “terza vantaggiata”); its colour is dull and not pure yellow.
Nevertheless the ash is frequently as low as per cent, and seldom
exceeds 2 per cent. ; in exceptional cases it may rise to 4 per cent.
or more. The fourth grade is of a greyish-yellow colour, and may
contain up to 25 per cent. of earthy matter. “Zolfo ventilato” is
sulphur which has been ground, separated by an air blast, and sieved.
Commercial Sicilian sulphur contains only very small quantities of
volatile matter and of bitumen, and generally is free from, or contains
only traces of, arsenic and selenium. The sulphur from the Solfatara
at Naples, which Phipson found to contain II. 162 per cent. of arsenic
and O. I64 per cent of selenium, is of no importance as a commercial
product.
264
NATIVE SULPHUR 265
The degree of fineness of ground sulphur must be determined when
the sulphur is intended for application in vine culture to destroy the
grape-disease Oidium. This examination is effected by means of
Chancel's sulphurimeter, shown in Fig. I 19. The sulphurimeter consists
of a cylindrical glass tube, 23 cm. long and 15 mm. diameter, sealed
below, fitted with a glass stopper above, and graduated from the bottom
upwards into IOO degrees. Each degree equals 4 c.c.; the IOO degrees
(25 c.c.) occupy a length of IOO mm. The ground sulphur is shaken
with ether in the tube, and allowed to settle, when it forms a layer the
height of which bears a relationship to the fineness of the grinding.
The sulphur to be examined is first passed through a sieve of I mm.
mesh, in order to break down the lumps formed during storage. Five
g. of the sieved sulphur are placed in the tube, which is
then half-filled with anhydrous ether as nearly as possible
at a temperature of 17°. 5. The sieved sulphur in the
tube is broken down still further by thorough shaking,
after which ether is added until the level stands I c.c.
over the IOOth division; the whole is again well shaken
and the tube then fixed in a vertical position. When the
sulphur layer has come to rest, its position on the scale
is read off; the reading gives the fineness in “Chancel
degrees.” As a rule, ground sulphur shows 50° to 55°,
finer qualities 70° to 75°, Zolfo ventilato 90° to 95°. For
use in viniculture, a minimum of 60° and frequently of
75° Chancel is demanded.
According to H. Fresenius and P. Beck," the dimen-
sions of the sulphurimeter are of importance and must be
exact; they recommend the instrument made by J. Greiner,
Munich. This differs in its dimensions from the French
instrument described above; the length of the tube to
division IOO is 175 mm., the length of the straight portion between
IO’ and IOO is I 54 mm., and the internal diameter 12.68 mm. The
ether employed should be distilled over sodium, and all disturbance
should be avoided after the shaking, the instrument being immediately
clamped in a stand and immersed in water at 17°.5.
It may further be necessary to determine whether the fine sulphur
intended for vine-dusting is really flowers of sulphur, or whether it
consists of a mixture of ground sulphur with flowers of sulphur. This
is best done by microscopic examination, the ground sulphur being
distinguished by its crystalline appearance from the partially amorphous
flowers of sulphur. The method is not, however, absolutely trustworthy,
since flowers of sulphur change in course of time almost completely into
the crystalline condition (Alpha sulphur). The partial insolubility of
* Z, anal. Chem., 1903, 42, 21.
FIG. 119.

266 MANUFACTURE OF SULPHUROUS ACID
flowers of sulphur in carbon bisulphide may also be employed as a
qualitative test, but with old samples the percentage of insoluble sulphur
may occasionally be very slight. This reaction cannot be applied as
a quantitative method, since the ratio between soluble and insoluble
sulphur varies considerably even in freshly prepared flowers of sulphur.
Opinions differ very considerably as to whether crude sulphur
or flowers of sulphur is the more efficient in viniculture. H.
Fresenius and Beck recommend the former, maintaining that the
crystalline ground sulphur adheres more firmly to the leaves than does
the amorphous variety. In France, on the other hand, the latter form
is preferred, owing to its finer state of division, as compared with
the ground Sulphur (which is frequently employed on account of its
cheapness), and also because of the very active destructive action of
the adhering acid on the oidium.
The chemical examination of native sulphur includes the following
determinations:— -
I. Ash.--This is determined by burning off about Io g. of sulphur
in a tared porcelain crucible and weighing the residue.
2. Moisture.—In the case of lump sulphur this determination is
frequently unnecessary, but it should be made if adulteration by
spraying is suspected, or if the sulphur has been exposed to rain.
The estimation is, however, always misleading, owing, on the one hand,
to the difficulty of obtaining a genuine average sample, and on the
other hand to the loss of moisture which almost inevitably occurs in
breaking down the sample to suitable fineness for laboratory work.
The sample should not be crushed further than to a coarse powder, and
this should be done as speedily as possible. At least IOO g. should
be taken for the estimation. In the case of ground sulphur, it is of
course much easier to obtain an average and more finely divided
sample. According to Fresenius and Beck, the drying operation
should only occupy a short time and the temperature should not
exceed 70°.
Difficulties in the moisture determination arising from the presence
of gypsum have been experienced by F. B. Carpenter" in the examina-
tion of Mexican sulphur. He consequently recommends making the
estimation by drying the sample in vacuo over Sulphuric acid. A
portion of the sample is freed from gypsum by boiling with dilute
sulphuric acid, the residue dried and weighed, and the contained
sulphur estimated, either by burning-off or by solution in carbon
bisulphide (cf. infra).
3. Bituminous substances.—Any appreciable quantity is evidenced
by the bad colour of the sulphur, but in most commercial sulphurs the
amount present is too small to render a determination necessary.
* /. Soc. Chem. Ind., 1902, 21, 832.
NATIVE SULPHUR 267
Sulphur recovered from gas residuals is, however, frequently quite
black from the presence of this impurity.
H. Fresenius and Beck drive off the sulphur by heating to slightly
over 200°, weigh the residue, ignite for ash and re-weigh, and regard
the difference between the two weighings as representing the organic
matter present.
4. Arsenic.—This impurity may occasionally occur in very minute
quantity in Sicilian sulphur, to a greater extent in the Solfatara
sulphur (see p. 264), and especially in sulphur recovered from pyrites or
from alkali-waste. The arsenic may be present either as arsenious
sulphide, As,Sº, or as arsenious oxide, As2O3; exceptionally also as
calcium or iron arsenite. J. T. Conroy (private communication) finds
the sulphur recovered from alkali-waste by the Chance-Claus process
particularly free from arsenic; tested by the Hager method, only very
minute and doubtful traces are indicated. Sulphur recovered from
pyrites does not show such freedom.
For a qualitative examination for arsenic, Hager's method may be
employed. It is carried out by shaking I g. of sulphur with 15 drops
of ammonium hydroxide and 2 c.c. of water, filtering after half an hour,
and adding to the filtrate contained in a test-tube 30 drops of hydro-
chloric acid and I 5 drops of a solution of Oxalic acid. A strip of clean
brass is then placed in the solution and this is warmed to a temperature
of 60° to IOO’; in the presence of arsenic an iron-coloured to black film
forms immediately on the brass. A blank test should always be made
with the reagents used, and to ensure uniformity of temperature it is
advantageous to heat the two test-tubes side by side, in a beaker of
Water.
Sulphide and oxide of arsenic may be extracted by digesting the
sulphur with dilute ammonia at 70° to 80°; they remain behind on
evaporating the ammoniacal extract. It is better, however, to add excess
of hydrochloric acid to the ammoniacal extract and then precipitate by
Saturation with sulphuretted hydrogen; any arsenic originally present
as AS,Ss naturally comes down before passing in the gas. The arsenic
present is best estimated by Schäppi's method,” which consists in
exactly neutralising the ammoniacal solution with nitric acid, diluting,
and then titrating with M/IO silver nitrate solution, using as indicator
neutral potassium chromate, which is coloured brown by a drop of the
solution after all the arsenic has been precipitated.
The salts of arsenious acid remain in the insoluble residue on
extracting sulphur with carbon bisulphide; this residue must be
digested with aqua regia and examined for arsenic in the usual manner,
as is described more fully under “Sulphuric Acid.”
5. Selenium may be detected by oxidising the sulphur, best by
* Pharm. Centr., 1884, 265 and 443. * Chem. Ind., 1881, 4, 409.
268 MANUFACTURE OF SULPHUROUS ACID
deflagrating with nitre; on dissolving the melt in hydrochloric acid and
treating with sulphurous acid, selenium will, if present, be precipitated as
a red powder.
In America, according to Reed, the examination for selenium is
carried out as follows:–0.5 g. of the sulphur is boiled with a solution of
o. 5 g. potassium cyanide in 5 c.c. of water, the solution filtered, and the
filtrate acidified with hydrochloric acid ; after standing for an hour no
red coloration due to selenium should result. A pale yellow coloration,
due to perthiocyanic acid, may be neglected. The test is made still
more sensitive by boiling I g. of the sulphur for an hour with a solution
of 2 g of potassium cyanide, then adding a further O.5 g. potassium
cyanide and continuing the boiling for half an hour. Any iron present
will, of course, react with the thiocyanate formed. -
6. Determination of Sulphur.—Macagno” recommends the direct
method of extraction with carbon bisulphide, and has calculated a table
for this purpose. The method has been recently investigated by
Pfeiffer 8 and the specific gravities involved determined with considerable
exactitude. To estimate sulphur by this process, a weighed quantity
of the powdered sample is shaken in a well-stoppered bottle with at least
four times its weight of pure carbon bisulphide, accurately weighed, and
the specific gravity of a clear sample determined at a known tempera-
ture (+15° C. and higher). -
The specific gravity thus determined is if necessary reduced to that
at 15° C. by means of the formula: -
I + at + 6t” + . .
Specific gravity at I s” – spec. grav. at t” x
p g y 5 pec. g I + a x 15 + b x 15°
where a = O.OOI I398 and b = O-OOOOOI37O.
The percentage of pure sulphur in the sample examined is deduced
from the formula :
% s = ***
C
)
where—
a = the weight of sulphur dissolved by IOO g. carbon bisulphide, accord-
ing to the following table,
&= the weight of carbon bisulphide used in the extraction,
c= the weight of substance. -
For an approximate estimation, Marcilli “ places 5 g. of the sample
in a Chancel tube (Fig. I 19, p. 265), fills the tube with carbon bi-
sulphide, and, after inserting the stopper, shakes well and allows to
settle ; the volume of the undissolved portion is then read off.
Ceruti" uses aniline warmed to 75° as a solvent.
* Chem. Zeit. Alep., 1897, 21, 252. * Chem. Wews, 1881, 43, 192.
* Z. anorg. Chem., 1897, 15, 194; also, P. Fuchs, Z. angew. Chem., 1898, II, II89.
* Z, angew. Chem., 1906, 19, 99. * Chem. Centr., 1904, II., 615.
NATIVE SULPHUR 269
Specific gravities of solutions of Sulphur in Carbon Bisulphide
with the corresponding weights of Sulphur dissolved by IOO
parts by weight of pure Carbon Bisulphide at 15° C., compared
with Water at 4° C.

Specific Sulphur Specific Sulphur Specific Sulphur Specific Sulphur
gravity. dissolved. gravity. dissolved. gravity. dissolved. gravity. dissolved.
1:2708 0-0 1 2999 6-4 I 3263 12-8 1.3507 19-2
1.2717 0-2 1°3007 6-6 1.3271 13-0 1°3514 19°4.
1.2726 0°4 1°3016 6-8 1-3279 13-2 1°3521 19:6
1.2736 0. 1°3024 7:0 1.3287 I 3-4 1-3529 19-8
1:2745 0-8 1°3032 7.2 1.3295 13-6 1-3536 20-0
I-2754 1-0 1°3041 7-4 1-3303 13-8 I-3543 20:2
1-2763 I-2 1°3050 7-6 1-3311 14-0 1-3550 20-4
1-2772 1°4 1°3058 7-8 1°3319 14-2 1-3557 20-6
1.2782 1°6 1°3066 8-0 1°3326 14'4 1°3564 20:8
1.2791 1°8 1:3074 8-2 1°3334 14-6 1°35'71 21-0
1-2800 2°0 1°3083 8’4 1-3342 14-8 1.3577 21.2
1°2809 2°2 1°3091 8-6 I-3350 15-0 1°3584 21'4
1-2819 2°4 1°3100 8:8 1.3357 15-2 1°3591 21-6
1-2828 2’6 1°3108 9:0 1-3365 15°4 1°3598 21-8
1.2838 2°8 1°3116 9°2 1.3373 15-6 1-3605 22-0
12847 , 3'0 1.3125 9°4 1.3380 15-8 1-3612 22.2
1-2856 3'2 1-3133 9:6 1.3388 16-0 1-3619 22°4
1.2866 3°4 1°3142 9°8 1.3396 16.2 1°36'26 22-6
1.2875 3°6 1-3150 10-0 1°3403 16-4 1°3633 22.8
1°2885 3'8 1°3158 10°2 1°3411 16-6 1°3640 23:0
1°2894 4°0 1°3166 10°4 1°3418 16-8 1°3646 23.2
1°2903 4'2 1-3174 10'6 1°3426 17:0 1°3653 23°4
1-2912 4'4 1°3182 10°8 1°3433 17.2 1°3660 23-6
1.2920 4°6 1°3190 11:0 1°3441 17.4 1°3667 23.8
1.2929 4°8 1°3199 11°2 1°3448 17.6 1-3674 24°0
1°2938 5'0 1°3207 11°4 1°3456 17-8 1-3681 24-2
1.2947 5°2 1°3215 11'6 1°3463 18:0 1°3688 24°4
1°2956 5*4 1.3223 11.8 1.3470 18-2 1-3395 24'6
1°2964 5°6 1°3231 12-0 1.3478 18°4 1-3702 24°8
1-2973 5°8 1-3239 12.2 1°3485 18-6 1-3709 25-0
1-2982 6-0 1.3247 12°4 1°3492 18:8 tº º 0. e & ©
1°2990 6-2 1-3255 12.6 1'3500 19-0
In regard to the examination of refined sulphur—
Roll sulphur is generally to all intents chemically pure; it may be
examined for ash, arsenic, and Selenium (see p. 266).
Flowers of sulphur are not quite so pure; if not carefully washed,
they are especially liable to contain free acid, more particularly
sulphurous acid, sulphuric acid, and also thiosulphuric acid, which
may be tested for by the usual methods. If intended for use in the
manufacture of fireworks, the sulphur must be free from all trace of
acidity. According to determinations made by Janda,” the residue
on ignition amounted to O-O63 per cent, as an average of thirty
samples, the maximum value being O-283 per cent. The solubility
in boiling sodium hydroxide solution of I-2 sp. gr. led to a mean
value of 98.04 per cent, with a maximum of 99.99 per cent, and a
* Jahresöer, der chem. Tech., 1897, 421.
270 MANUFACTURE OF SULPHUROUS ACID
minimum of 88 per cent.; in one case, however, only 68 per cent, was
dissolved.
According to Domergue, only those products should be classed as
“flowers of sulphur" which, when fresh, contain at least 33 per cent.
insoluble in carbon bisulphide.
II. SULPHUR FROM GAS RESIDUALS
The spent purifying materials from gasworks frequently contain
over 50 per cent. of free sulphur.” They are first freed from ammonium
salts by extraction with water, and then subjected to a special treat-
ment for the recovery of cyanogen compounds. The residue remaining
after this treatment forms a valuable raw material for the production
of sulphurous acid. It consists essentially of free sulphur and oxide
of iron admixed with sawdust, tarry substances, etc., and may also
contain varying quantities of lime, etc. Any lime present will, on
burning, fix part of the sulphur, and on this account a method of
analysis is adopted which takes account of available sulphur only.”
The spent oxide is burnt by the aid of platinised asbestos, the resulting
gases are passed into a solution of potassium hydroxide and potassium
hypobromite, and the sulphuric acid contained in the resulting solution
estimated by precipitation with barium chloride.
The burning is effected in a combustion tube, 60 cm. in length
(Fig. I2O), narrowed at a, and drawn out at one end into a downwardly
FIG. 120.
directed and not too thin narrower piece IO cm. long. A layer of
platinised asbestos, 20 to 25 cm. long, is placed between a and b, and
7 to Io cm. behind this a porcelain boat containing about O'4 g. of the
sulphur. The free end of the tube is connected at # with a gasholder
containing Oxygen. Two I4 cm. high bulb U-tubes c and d and the
tube e, filled with glass wool, are employed for absorbing the evolved
gases. They are charged with a solution prepared by dissolving 180 g.
of potassium hydroxide (purified from sulphate by alcohol) in water,
* Chem. Zeit. Rep., 1905, 29, 19. .
* The examination of these residuals for their more important constituents will be described
in the section on Gas Manufacture, Vol. II.
* Zulkowsky, Dingl polyt.J., 1881, 241, 52.

SULPHUR FROM GAS RESIDUALS 271
adding to this IOO g. of bromine drop by drop, cooling as required,
and diluting to I litre; thirty c.c. of this solution are sufficient for
the estimation of O. 5 g. sulphur. The glass wool in the tube e should
also be moistened with the solution.
A current of moist oxygen is passed through the tube, and the
portion between a and b heated to redness; the boat is then pushed in
and the heating extended from right to left, until finally the tube is
heated to the point f. The current of gas must be much stronger
than that employed in ordinary organic combustions, in order to
prevent sulphur from passing away unburnt; on the other hand, it
must not be so strong that the sulphur dioxide escapes unabsorbed.
Any deposit formed at / must be driven forward by means of a Bunsen
burner; the combustion is complete when such deposit ceases to form,
and this is usually the case after one hour. The absorbing vessels are
then removed, emptied, and washed out ; any sulphuric acid remaining
in the tube at h is recovered by repeatedly drawing up water by suction
applied at #. The solutions and washings are combined, hydrochloric
acid added in sufficient excess to neutralise the alkali and decompose
the hypobromite, the whole heated and if necessary concentrated, and
the sulphuric acid precipitated as barium sulphate by addition of a hot
solution of barium chloride.
This method is also applicable to the determination of available
sulphur in pyrites ; in this case the platinised asbestos is omitted, and
the combustion tube, drawn out and bent round as above, need not be
more than 40 cm. long. -
Hydrogen peroxide may be used with advantage instead of potassium
hypobromite solution, titration of the sulphuric acid by standard
alkali being in this case substituted for the gravimetric determination.
Allowance must, of course, be made for any acidity in the hydrogen
peroxide employed. This modification is much more rapid than
Zulkowsky's method, and does away with the necessity of employing
absolutely sulphate-free potassium hydroxide. The reaction is simple:
H2O2+SO2 = H2SO4.
Similar methods, generally intended for the estimation of available
sulphur in pyrites, have frequently been recommended ; these are
described below.
The estimation of the sulphur by oxidation with hydrogen per-
oxide always gives too high results, and some tar is also dissolved,
Pfeiffer' burns I g. of the sample in a bottle filled with oxygen, on
the bottom of which a solution of sodium hydroxide is placed. The
solution is finally oxidised with I c.c. of 30 per cent. neutral hydrogen
peroxide, and titrated.
* J. Gasóeleucht, 1905, 48, 977; Chem. Centr., I905, II., 1831.
272 MANUFACTURE OF SULPHUROUS ACID
III. PYRITES' (and other Metallic Sulphides)
Pyrites is generally received in sealed duplicate samples of 200
to 300 g., the average sample being prepared according to the rules
given on p. 9. -
The most important estimation for the acid manufacturer is
naturally the determination of the sulphur content. In addition to
this, moisture, less frequently copper, arsenic, zinc, and carbonates are
determined. The complete analysis of pyrites is very seldom under-
taken, generally only in examining the first consignments of this raw
material. *
I. Moisture.—The coarsely ground sample is dried at Ios”, until
the weight becomes constant. For the following tests, the finely
divided average sample preserved in a well-sealed bottle, not the dried
sample, is employed.
The analytical results are calculated on the dry pyrites, for which
reason a special moisture determination in the average sample is
necessary.
2. Sulphur.—It must be decided whether the total sulphur, or that
rendered soluble by aqua regia, or finally that recoverable by heating
in a current of air, is to be determined. Sulphur present as heavy spar
or galena would be included in the first case, which is useless for
the production of sulphurous acid. For this reason and also because
of its greater rapidity, and the avoidance of the fusion, which strongly
attacks the platinum crucible, the wet method, as recommended by
Lunge,” is almost universally used. By publication in the Alkali
Makers' Handbook the method has attained the rank of a standard
process, and is generally employed in buying and selling.
The third group of methods in which the available sulphur is
estimated has been dealt with under “Gas Residuals” (p. 270); it is
not as a rule employed for pyrites.
As a preliminary to any method of determination it is essential to
reduce the sample to a very fine powder, first in a steel mortar and
finally in an agate mortar. A porcelain or Wedgwood mortar should
not be used, as they yield quite appreciable quantities of foreign matter
to the sample. The complete sample is then sifted through very
fine silk gauze, without elimination of any residue.
Lunge's method for the wet extraction of pyrites is as follows:—o. 5
g. of the pyrites is treated with about 10 c.c. of a mixture of 3 volumes
nitric acid of 1.4 sp. gr. and I volume strong hydrochloric acid (both
acids must be tested for complete freedom from sulphuric acid), with
occasional heating, and care being taken to avoid spurting (cf. p. 22).
1 Cf. G. Lunge, Sulphuric Acid and Alkali, 3rd edition, 1903, p. 34.
* Z, anal. Chem., 1881, 20, 419 ; and, Z. angew. Chem., 1889, 2, 473.
PYRITES 273
In exceptional cases some free sulphur may separate; if so, it may
be oxidised by the cautious addition of potassium chlorate. The
mixture is evaporated to dryness on the water-bath and this evaporation
is repeated after the addition of 5 c.c. of hydrochloric acid, after which
nitrous fumes should cease to be evolved. A further I c.c. of concen-
trated hydrochloric acid is added to the residue on the water-bath,
and after a few minutes IOO C.C. of hot water are added, the whole
filtered through a small filter paper and the insoluble matter washed
with hot water." The residue may be dried, ignited, and weighed. It
may contain in addition to silica and silicates, barium, lead, and
occasionally calcium sulphate, the sulphuric acid of which being
absolutely non-available is purposely neglected. If the residue is not
to be estimated, the filtration may be neglected and the precipitation
with ammonia proceeded with as soon as all the nitric acid has been
driven off.
To remove the iron, the filtrate and washings are treated with
ammonia in not too large excess and heated to 60° to 70° for ten to
fifteen minutes, but the solution should not be heated to boiling; it
should still smell strongly of ammonia, otherwise the precipitate may
contain basic ferric sulphate.
According to Treadwell,” the separation of basic ferric sulphate and
consequent retention of sulphuric acid during washing need not be feared
if the solution is supersaturated with ammonia in the cold, instead of
as described above, and then heated almost to boiling with constant
stirring. The separated ferric hydroxide is filtered off and washed. The
operation may be carried out in shorter time (one-half to one hour) by
observing the following precautions:—I. Filtering hot and washing on the
filter with hot water, taking care to avoid channels by thoroughly disturb-
ing the whole precipitate with each washing. 2. Using a sufficiently dense
but rapid filtering paper. 3. Employing a correctly made funnel with
an angle of 60°, the tube of which fills completely with the filtrate.
The washing is continued until about I c.c. of the washings yields
no turbidity with barium chloride after standing several minutes. In
doubtful cases it is advisable to prove the complete absence of basic
sulphate by fusing the dried ferric hydroxide precipitate with pure sodium
carbonate and examining the aqueous extract of the melt for sulphuric
acid.
Hundreds of check tests, carried out in the manner described, have
shown (Lunge) that by observing the above precautions even novices
almost always obtain a ferric hydroxide free from sulphuric acid, and
that with experienced workers this is always the case. Küster and
* Cf. Dennstedt and Hassler, Z. angew. Chem., 1905, 18, 1137 and 1562; also, G. Lunge,
ibid., 1905, 18, 1656.
* Analytical Chemistry, vol. ii., p. 282.
274 MANUEACTURE OF SULPHUROUS ACID
Thiel, who incorrectly assume that the ferric hydroxide cannot be
completely freed from sulphuric acid by washing, recommend as an
alternative method to precipitate with barium chloride without previous
removal of the ferric hydroxide but to subsequently remove this by pro-
longed digestion with hydrochloric acid, or to add a considerable amount
of ammonium oxalate so as to altogether prevent the precipitation of the
iron. Both methods occupy much more time than that described above
and are in no way more exact.”
The combined filtrate and washings should not exceed 300 c.c.; if
they do they must be concentrated by evaporation. The solution is
acidified with pure hydrochloric acid, avoiding any considerable excess,
heated to distinct boiling, and after removal of the flame a solution of
barium chloride, also at the boiling temperature, added all at once
with constant stirring. By adding the hot barium chloride solution
all at once to the solution to be precipitated, instead of drop by drop,
an error is introduced owing to the absorption of some barium chloride
by the barium sulphate, but this just compensates the loss caused by
the solubility of the barium sulphate in the solution, and the error due
to the unavoidable co-precipitation of double sulphates of barium and
ammonium. The presence of copper introduces no error in the de-
termination ; zinc if present in small quantities has also no effect, but
if present in larger proportions, as in the analysis of blende, the error
is greater, owing to the solubility of barium sulphate in the ammonium
salts necessary to keep the zinc in solution. This error is, however,
also compensated, as stated above, by adding the barium chloride
solution all at once.” Twenty c.c. of a Io per cent. barium chloride
solution diluted with IOO c.c. of water, will suffice for g. of pyrites.
A larger excess of barium chloride is to be avoided, or the results will
come out too high. The whole is allowed to stand for forty minutes
after precipitating, by which time the filtrate should have cleared com-
pletely and be ready for further treatment without the necessity of first
waiting several hours, as in the method previously described. The
clear liquid is decanted as completely as possible through a filter, IOO
c.c. of boiling water poured on to the precipitate and the whole stirred ;
after two to three minutes the solution should have cleared again and
be ready for a further decantation. The washing by decantation is
repeated three or four times until the liquid ceases to show an acid
reaction; the precipitate is then washed on the filter, dried and ignited.
It should be perfectly white and should not cake. The filter paper,
after separating the precipitate, is either burnt in a platinum spiral, or
* Z, amorg, Chem., 1899, I9, 97 ; 1900, 22, 424.
* Cº. Lunge, Z, amorg, Chem., 1899, 19, 454; Herting, Z, angew, Chem., 1899, 12, 274; also,
Chem. Zeit., 1899, 23, 768.
* Lunge and Stierlin, Z, angew, Chem., 1905, 18, 1921.
PYRITES 275
the filter paper and precipitate are burnt wet in the crucible. In the
latter case it is necessary to convert the resulting barium sulphide into
sulphate, by gentle ignition with the crucible in a slanting position.
One part BaSO, is = O. I 373 parts of sulphur.
With practice, the total volume of solution to be precipitated with
barium chloride will not exceed 300 c.c., so that evaporation, before
precipitating, should not be necessary. If it is unavoidable it must be
carried out on the water-bath, or in such manner that any sulphurous
combustion products are kept away from the solution, e.g. by heating
on an asbestos or aluminium plate.
Other methods for removing the iron are described by Heidenreich 1
by Herting and Lehnhardt,” and by Gyzander.”
The previous precipitation of the iron is necessary, since, if present,
it gives rise to the formation of a double sulphate of iron and barium
in the barium precipitate, which double sulphate loses part of its sul-
phuric acid on ignition.* The resulting error may, according to Lunge,
reach a maximum of 3 per cent. in the case of pyrites (this result
differing from that of Thiel, who lost up to 7 per cent); it is, however,
avoided by precipitating the iron as described on p. 273.
The injurious effect of a large excess of barium chloride, owing to
absorption by the barium sulphate precipitate, has long been known,
and Lunge has drawn special attention to this point. Richards and
Parker * have carefully investigated the subject, but the methods of
purification suggested by them, in addition to being extremely trouble-
some, fail in their purpose unless the solubility of barium sulphate in the
solution for precipitation be taken into account. Richards states that
under ordinary conditions the two errors exactly compensate each
other, and a result very near the truth may consequently be expected
if the details given above are carefully followed.
Such methods as solution of the pyrites in fuming nitric acid, or
in hydrochloric acid and potassium chlorate or sodium chlorate,” or in
hydrochloric acid saturated with bromine, are less satisfactory than that
described above, owing to the liability to separation of sulphur.
Of the dry methods of analysis, that of Fresenius," though some-
what tedious, is probably the most exact. O.5 g. of pyrites is fused with
IO g. of a mixture composed of two parts of sodium carbonate and one
part of potassium nitrate, the melt is extracted with water, and after
precipitation of the lead by a slow current of carbonic anhydride, the
* Z, anorg. Chem., 1899, 20, 233. * Chem. Zeit., 1899, 23, 768.
* Chem. Mews, 1906, 93, 213.
* Cf. Jannasch and Richards, / prakt. Chem. I889, 39, 321 ; also, correction by these authors in
consequence of communication from Lunge, ibid., 1889, 40, 236. See also Lunge, Z, angew. Chem.,
I889, I2, 473.
* Z, amorg, Chem., 1895, 8,413. * Noaillon, Z. angew. Chem., 1897, Io, 35I.
* Z, anal. Chem., 1877, 16, 335.
276 MANUFACTURE OF SULPHUROUS ACID
solution is boiled with sodium carbonate solution, filtered, and washed
with hot water. The combined filtrate and washings are acidified with
hydrochloric acid, evaporated several times with additions of hydro-
chloric acid to remove the nitric acid, the residue taken up with dilute
hydrochloric acid and precipitated by barium chloride, the barium
sulphate precipitate after ignition being finally purified by boiling with
hydrochloric acid. This is a lengthy process, and has been replaced
by Lunge's method even in Fresenius' laboratory.
Other dry methods are generally restricted to the determination of
sulphur in burnt ore (see p. 294), to a preliminary testing of iron pyrites,
or to the analysis of some special pyrites for which the wet method is
less suitable. The wet method is always the best for pyrites, which, like
the Spanish, contain only about 4 per cent. of copper. The method
used in Freiberg is to heat I g. of powdered pyrites to a red heat in a
muffle furnace with 2 g. of sodium carbonate and 2 g of potassium
nitrate; the melt is extracted with hot water, filtered into excess of
hydrochloric acid, and titrated with barium chloride solution by Wilden-
stein's method (cf. infra). According to M. Liebig," twenty to twenty-four
such tests can easily be performed in six to eight hours. Böckmann
heats O. 5 g. pyrites with 25 g. of a mixture of six parts sodium carbonate
and one part potassium chlorate gradually until fusion has occurred
and oxygen ceases to be evolved, then extracts and precipitates with
barium chloride (cf. p. 295). According to Treadwell, this method gives
the same results as that of Fresenius; its use is practically confined to
the estimation of the sulphur content of pyrites cinders (cf. p. 294).
The best dry methods of treatment appear to be those in which
sodium peroxide is employed, as, for example, those recommended by
Hempel and by Höhnel and Glaser.” List” recommends the following
simplification of a method proposed by Fournier:-O-5 g. of the finely
divided pyrites (or burnt ore) is intimately mixed with 5 to 6 g.
of powdered sodium peroxide in a wrought-iron crucible, which is then
covered and gently heated in a Bunsen flame. Reaction quickly sets in
and is complete in one minute. As soon as the mass is fused the flame is
removed and the crucible plunged into I5o to 200 c.c. of warm water, in
which the melt quickly dissolves. The contents of the crucible are
washed out and the solution approximately neutralised with Io c.c.
concentrated hydrochloric acid, filtered from the insoluble ferric
hydroxide, which is washed as recommended on p. 273, and the
barium sulphate precipitated as there described. List states that in
absence of sulphate of lead or of alkaline earth metals, the results agree
with those obtained by Lunge's wet method. The iron crucibles ar
cheap and will stand fifty fusions. :-
* Post, Chemisch-technische Analyse, 2nd edition, vol. i., p. 677.
* Chem. Zeit., 1894, 18, 1448. * Z, angew. Chem., 1903, 16,414.
•PYRITES 277
Clark 1 heats with sodium bicarbonate and magnesia, and estimates
the resulting sulphuric acid gravimetrically. Fahlberg and Iles” fuse
with a large excess of potassium hydroxide in a silver crucible.
Of methods intended for the estimation of available sulphur, that is,
of such sulphur as is converted into Sulphur dioxide on roasting, that of
Zulkowsky has been described under Gas Residuals (p. 270). Similar
methods have been previously described by Mixter,” Brügelmann,” and
Sauer; * more recently by Jannasch." These methods have not been
adopted in general practice for the examination of iron pyrites; they
are, however, useful in the case of mixed pyrites and the like.
Other methods depend on the reduction of the FeS2 to FeS, which is
then decomposed by hydrochloric acid, the sulphuretted hydrogen so
evolved being passed into iodine solution and estimated by titrating
back the excess of the latter. Such methods have been described by
Gräger, Treadwell,” and Eliasberg,” but none have found much accept-
ance in practice.
The sulphuric acid formed in all the foregoing instances is estimated,
in all exact work, by precipitation with barium chloride, as described on
p. 274.
Silberberger" has proposed to replace barium chloride in the gravi-
metric method by precipitating the sulphuric acid in alcoholic solution
with strontium chloride. This suggestion has been opposed by Lunge,”
and the method has been rejected by the Analysis Committee of the
International Congress of Applied Chemistry.”
Numerous proposals have been made to shorten the determination
of fixed sulphuric acid by employing a volumetric method. Such
methods, however, in spite of the contrary view of Teschmacher and
Smith,” do not possess the exactness of the gravimetric determination,
and further, the saving of time, unless many determinations are made
simultaneously, is not considerable. At the same time it is to be borne
in mind that the gravimetric determination of sulphuric acid as barium
sulphate is by no means one of the most exact analytical operations, and
may, in spite of all precautions, give an error amounting to gººd of the
whole, or many times that involved in the estimation of chlorine as silver
chloride.
The more important volumetric methods for the estimation of fixed
sulphuric acid are appended, although, as stated, they are not
sufficiently accurate to justify their use in the case of pyrites.
* / Soc. Chem. Ind., 1885, 4, 329 and 724. * Aer., 1891, 24, IQ37; 1892, 25, 2377.
* Ber., 1878, II, 1187. * Z. anal. Chem., 1889, 28, 240.
* Amer. Chem. J., 1880–81, 2, 396. * Ber, 1903, 36, 2755, 4259.
* Z. anal. Chem., 1873, 12, 32. * Ber., 1903, 36, 3387. Z. angew. Chem.,
* Ibid., 1876, 15, 175. 1904, 17, 913, 949; 1905, 18, 449.
* / prakt. Chem., 1889, 40, 239; 1890, 41, 566. * Report, 1906, p. 344.
7 Dingl, polyt.J., 1883, 248, 53. * Chem. Wews, 1871, 24, 61, 66, and 140.
278 - MANUFACTURE OF SULPHUROUS ACID
Wildenstein' titrates with a normal solution of barium chloride, and
determines the end-point by testing a small filtered sample as follows:–
The solution is placed in an inverted bell-jar A (Fig. 121), through the
neck of which passes a bent tube 6, which is closed below
by a spring clip 3 and ends above in a funnel f bent
downwards, which is closed with two thicknesses of
filter paper covered and held in position with muslin.
The level of acidified solution must extend above the
funnel. The barium chloride solution is run into this
with good stirring, and after each addition a filtered
sample is withdrawn by first allowing several c.c. to flow
through f and pouring them back into A, and then
drawing a further sample and testing this with a drop
of barium chloride solution. If a precipitate results, the
sample is poured back into A, more barium chloride
added, and a further test made. Should the end-point
be overstepped, a few c.c. of normal sulphuric acid are added to the
solution and the testing completed, the volume of acid added being
subtracted from the volume of barium chloride used
Wilsing” exactly neutralises the sulphate solution, using phenol-
phthalein as indicator, heats to boiling in a porcelain dish, and adds
a measured quantity of a 4 per cent, barium chloride solution together
with phenolphthalein, and then titrates with a 2 per cent. solution
of sodium carbonate until the red coloration is obtained. The volume
of carbonate solution required gives the excess of barium chloride
added.
Similar methods have been described by Carl Mohr. Knöpfler,
Monhaupt, Blacher and Koerber,” and others.
Andrews' precipitates the Sulphuric acid by adding excess of a
hydrochloric acid solution of barium chromate to the boiling solution
of the sulphate; the excess of the precipitating reagent is removed by
addition of ammonia or of calcium carbonate to the boiling solution,
the whole filtered, and washed with hot water. The filtrate, which
contains alkali chromate corresponding to the sulphate originally
present, is cooled, concentrated hydrochloric acid added, followed by
potassium iodide, and the liberated iodine titrated with sodium thio-
sulphate. According to Reuter,” only 5 c.c. of concentrated hydro-
chloric acid should be added, the whole washed for five minutes after
the addition of the potassium iodide, the beaker being kept covered
and a current of carbon dioxide passed over the surface of the solution.
FIG. 121.
* Z. anal. Chem., 1862, I, 432. * Chem. Ind., 1886, 9, 25.
* Cf. Mohr-Classen, Zehrbuch der Chem. analyt. 7 itrirmethoden, 7th edition, 1896, pp. 150-154.
* Annalem, 1885, 230, 260. * Chem. Centr., 1905, I., 296. * /bid., 1905, II., 511.
7 Chem. Zeit. Rep., 1889, 13, 39. * Chem. Zeit., 1898, 22, 357.

PYRITES 279
(The analytical results given in the paper do not show any very high
degree of accuracy.)
Marboutin and Moulinié" describe a similar method in which
equivalent solutions of barium chloride and of potassium chromate
are employed. The sulphuric acid is precipitated in acid solution
by barium chloride, and the excess of this reagent precipitated by
addition of potassium chromate; the excess of the latter is then
determined by iodine or by arsenious acid. An almost identical
method is recommended by Telle.” He employs (1) a solution con-
taining 12.07 g. BaCl22H2(2H2O)C per litre (= 4 g. SO); (2) a solution
of potassium bichromate equivalent to this, and containing 7.3635 g.
per litre; (3) a solution containing 38 g. sodium thiosulphate per litre;
and (4) a 10 per cent. Solution of potassium iodide. The thiosulphate
equivalent of the bichromate solution is obtained by treating Io c.c.
of solution No. 2 with 5 c.c. of solution No. 4, and titrating the
liberated iodine with solution No. 3 after addition of distilled water
and excess of hydrochloric acid. One molecule of potassium bi-
chromate (294-54) liberates six molecules of iodine, and thus corre-
sponds to 6× 248.3 of thiosulphate, or 2 x 244-37 of BaCl, 2H,O,
according to the equation :-
2 BaCl, + K2Cr,C), + 2NH3+ H2O=2|BaCrO, + 2KC1+ 2NH,Cl.
Grossmann” decomposes the neutral solution of the sulphate with
barium hydroxide, filters off one-half, precipitates the excess of baryta
with carbon dioxide, boils, cools, dilutes to 500 c.c., and titrates one-
half of the clear solution with normal acid. Employing the ºr
equivalent of sodium sulphate=3:554 g., the percentage content in
Na,SO, is equal to eight times the number of c.c. of normal hydro-
chloric acid required ; a correction of O.4 per cent. must be made for
the volume occupied by the barium precipitate. Finally, there also
remains a constant loss of 1.3 per cent, the cause of which has not
been determined, but which must nevertheless be allowed for ; this
renders the process useless for exact determinations.
Nikaido * precipitates the sulphate in alcoholic solution by addition
of N/Io lead nitrate solution, using potassium iodide as indicator. The
method is only applicable in special cases.
Riegler" precipitates the sulphate with a known quantity of barium
chloride, treats with iodic acid, filters off the barium iodate, and arrives
at the quantity of barium used by measuring, in the nitrometer, the
nitrogen evolved on addition of hydrazine sulphate.
Recently, methods first proposed by Vaubel," and based on the use
1 Chem. Centr., 1898, I., 218. * /. A mer. Chem. Soc., 1902, 24, 774.
* /bid., 793. * Z, anal. Chem., 1902, 41, 17.
* Chem. Wews, 1880, 41, 114; Ber, 1880, 13, 824. * Ibid., 1896, 35, 821.
280 MANUFACTURE OF SULPHUROUS ACID
of benzidine, have attracted attention. W. A. Müller 1 has found that
fixed sulphuric acid can be quantitatively precipitated by benzidine
hydrochloride when the latter is present in sufficient excess, and that.
such excess can be directly determined by titration with alkali. This
is due to benzidine being such a very weak base, that the hydrochloride
is sufficiently dissociated hydrolytically in solution to behave like free
hydrochloric acid towards phenolphthalein. The decomposition corre-
sponds to the equation :-
R,SO, + C19H19N3.2HC1 = 2 RCl+C19H19N3.H.S.O.
The solution of benzidine hydrochloride employed contains about
30 g. per litre, and is standardised against barium or sodium hydroxide
by means of normal acid, using phenolphthalein as indicator. (This
indicator should be avoided. Cf. p. 97.) The decomposition must be
carried out in warm solution, since in the cold the separated sulphate
carries down some of the hydrochloride. The heating is done in a 250
C.C. flask on the water-bath, and the solution maintained hot for some
minutes; in the presence of volatile acids the flask should be stoppered.
The solution is rapidly cooled and made up to a definite volume,
filtered through a dry filter paper, and the excess of hydrochloride
determined by titrating an aliquot portion with standard alkali and
phenolphthalein.”
Raschig” has modified this method as follows. Forty g. of benzidine
are ground up with 40 c.c. of water, washed with about 750 c.c. of water
into a litre flask, 50 c.c. of concentrated hydrochloric acid added, and then
water to the graduation mark. On shaking, the whole or the greater
part dissolves to a brown solution, which is filtered if necessary, and
then diluted to twenty times its volume. One hundred and fifty c.c. of
this weak solution are used to precipitate O. I g. sulphuric acid. The
sulphate solution to be examined is allowed to flow, with continual
stirring, into the benzidine solution, and the benzidine sulphate which
rapidly separates is filtered off at the pump. A 200 c.c. funnel, fitted
with a Witt's filter plate, is used for the filtration. The plate, which is
of 40 mm. upper diameter, is covered by two moistened filter papers
of 46 mm. diameter, the extending edges of which are, after applying
the suction, pressed by means of a sharp-angled glass rod to form a
circular pad, and so make the filter perfectly tight. Any precipitate
adhering to the walls of the beaker or other vessel used for the
precipitation is washed off with a portion of the clear filtrate,
and as the last drop of mother liquor drains off from the filter, the
precipitate is washed twice successively with from 5 to IO c.c. of water.
* Ber., 1902, 35, 1587.
* CAE also, Müller and Dürkes, Z. anal. Chem., 1903, 42, 477; Müller, Z. angew. Chem., 1903,
I6, IoI7.
* Z. angew. Chem., 1903, 16, 617 and 818.
PYRITES 281
By thus washing with a minimum quantity of water, the objection
raised by W. A. Müller and others to the method on the grounds of
the solubility of benzidine sulphate in the wash-water, is overcome.
The funnel is removed from the filter-flask, placed on a watch-glass
of 50 to 60 mm. diameter, inverted, the filter-plate and precipitate trans-
ferred to the watch-glass by the aid of a glass rod, the filter-plate
removed, the filter pressed together in the fingers and dropped into
an Erlenmeyer flask of 250 c.c. capacity and 30 mm. wide at the neck.
The watch-glass and funnel are washed with not more than 25 c.c.
of water, and the washings added to the contents of the flask, which
is then closed with a rubber stopper and well shaken until there results
a uniform magma of precipitate and disintegrated filter paper quite
free from small lumps of benzidine sulphate. The mixture is then
warmed to 50°, and after addition of phenolphthalein titrated with M/Io
sodium hydroxide solution, finally heating to boiling to remove from
the solution all carbon dioxide, since it affects the indicator. An
accidental excess of alkali solution may be titrated back with normal
acid. This method is applicable to the estimation of sulphuric acid
present either in the free state or as copper or ferrous sulphate ; it
is not directly applicable to ferric sulphate, as in this case the iron
must be precipitated as described on p. 273. The method is also
inapplicable in many other cases, as for example in the presence of
certain organic compounds.
The benzidine method may, moreover, be used even in the presence
of iron when not more than I atom of sulphur is present per I atom of
iron, provided that (as first recommended by M. Schlötter) ferric salts
are first reduced to the ferrous state. For this purpose, Raschigi
employs a dilute solution of hydroxylamine hydrochloride, and gives the
following details for the determination of sulphur in pyrites by this
method. Exactly O.8 g. of the finely divided sample is weighed into
a 200 c.c. conical flask, 5 C.C. of fuming nitric acid added, and the flask
with a small funnel in the neck heated on the water-bath. After half
an hour, 30 c.c. of water are added, the whole of the contents rinsed
into a 100 c.c. flask, made up to the mark and well shaken. Twenty
c.c. of this solution are then placed in a 600 c.c. beaker, Io c.c. of a
I per cent. solution of hydroxylamine hydrochloride added, and 500
c.c. of benzidine hydrochloride solution prepared as above. The
precipitation and subsequent treatment of the benzidine sulphate is
then carried out as described. The filtrate from the sulphate should
always be tested with barium chloride, to make sure that sufficient
benzidine solution has been used for complete precipitation.
3. Arsenic.—The estimation of arsenic in pyrites is somewhat
laborious, and is consequently not carried out in works as frequently as
* Z. angew. Chem., 1905, 18, 331 ; 1906, I9, 331.
282 MANUFACTURE OF SULPHUROUS ACID
its importance would demand. The arsenic in pyrites not only finds its
way into the vitriol, but also into the hydrochloric acid prepared there-
from, and may, in this and in many other cases, prove very harmful.
The methods employed for its determination lead, unfortunately, to
widely differing results.
According to the method devised by Reich, as modified by M'Cay,"
O. 5 g. pyrites is treated with strong nitric acid in a porcelain crucible,
the free acid evaporated without taking the mixture to dryness, 4 g. of
sodium carbonate added, and the whole taken to dryness on the sand-
bath, after which 4 g. of potassium nitrate are added, and the mixture
heated to gentle fusion for ten minutes. The melt is extracted with
hot water, filtered, and the filtrate, after acidifying with a little nitric acid,
heated for some time to remove all carbon dioxide, and after addition
of silver nitrate, cautiously neutralised with dilute ammonia. The
precipitate, which contains all the arsenic as silver arsenate, is dissolved
in dilute nitric acid, and the silver either estimated by titration with
ammonium thiocyanate according to Volhard's method, or the
solution evaporated in a platinum dish and the residue weighed.
One part Aga ASO, =O. I62 IAs. If it be desired to precipitate the
arsenic as arsenic pentasulphide, the solution is transferred to a flask,
acidified with hydrochloric acid, the flask almost filled with air-free
water and saturated with sulphuretted hydrogen. The stopper is then
securely fastened and the flask heated for an hour on the water-bath,
whereby a precipitate of the pentasulphide, uncontaminated with free
sulphur, is obtained.
Other methods give results which differ rather widely from those
obtained as above.
Blattner and Brasseur” quote an instance in which different
analysts found from O. 19 to O. 57 per cent. of arsenic, and another in
which the figures varied from O-O5 to O-39 per cent. Records of still
greater differences may be found in the journals. They recommend the
following process:—
I. Wet treatment. Ten g of pyrites is added gradually with gentle
warming to aqua regia, prepared from 125 C.C. nitric acid of sp. gr.
1.37, 250 c.c. hydrochloric acid of sp. gr. I. I 5 to I. I7, and IOO c.c. water,
in a litre flask. The bulk of the nitric acid is driven off by evaporation
after addition of hydrochloric acid; since all the arsenic is present as
arsenic acid, no loss through evolution of arsenious chloride is said to
take place. After the addition of IOO c.c. of water, the solution is
cooled and filtered, ammonia is added until a slight precipitate of ferric
hydroxide is formed, sulphurous anhydride passed through the cold
solution until the reduction of the iron to the ferrous state is complete,
* Chem. Mews, 1883, 48, 7. Amer. Chem. J., 1886, 8, 77; 1887, 9, 174; 1888, Io, 459.
* Bull. Soc. Chim., 1897 [3], 17, 13.
PYRITES 283
the excess of sulphur dioxide removed by warming, and after cooling
to 60°-70°, the arsenic precipitated by passing in sulphuretted hydrogen
for six to seven hours. The solution is then allowed to stand for
twelve hours, and filtered. The precipitate is washed with water con-
taining hydrochloric acid and sulphuretted hydrogen until all the iron
has been removed, and finally with distilled water.
It is then dissolved by digesting the precipitate and filter paper with
ammonium carbonate, and the solution filtered. The filtrate is rendered
strongly acid with hydrochloric acid warmed to 50°-70°, and the arsenic
precipitated as arsenious Sulphide by passing in sulphuretted hydrogen
for an hour. The arsenic in the precipitate may be determined
either as magnesium ammonium arsenate or as silver arsenate. In
the former case the precipitate is dissolved in strong ammonia,
evaporated to dryness on the water-bath, taken up with Io c.c. nitric
acid, and after a further slight evaporation, rendered ammoniacal, and
a small quantity of alcohol, followed by magnesia mixture, added ; after
standing for twelve hours, the precipitate is collected on an ash-free
filter paper, washed with a solution of one part of ammonium chloride
and one of alcohol in three of water, dried, and ignited after separation
from the filter paper. The filter paper is ignited separately after
addition of a small quantity of ammonium nitrate. One hundred parts
Mg3As, O, correspond to 48.28 parts As.
2. Dry treatment. Two g. of pyrites are thoroughly mixed in a
platinum crucible of 30 c.c. capacity with IO to I2 g. of a mixture of
equal parts of potassium nitrate and sodium carbonate, and covered
with a layer of 2 g. of the same mixture. The crucible is then covered
and heated over a Bunsen flame 3 cm. high. When the reaction is
complete, the mass is allowed to cool, the contents of the crucible
placed in 70 c.c. of boiling water, and after solution, the whole
filtered and the residue washed with boiling water. All the arsenic is
then in the filtrate as arsenate. The solution is acidified with nitric
acid, heated to boiling, allowed to cool, rendered exactly neutral by
ammonia, and after the addition of one drop of nitric acid, again
neutralised until a drop of the solution only produces a blue tint on red
litmus paper after some seconds. Silver nitrate is then added, drop by
drop, until no further precipitation occurs; the silver arsenate is collected
on a filter and washed with cold water until the filtrate no longer gives
a turbidity with hydrochloric acid. The silver arsenate is dissolved
on the filter in very dilute nitric acid, 5 c.c. of a sulphuric acid—nitric
acid—iron solution added as indicator, and the solution titrated with
M/IO ammonium thiocyanate solution to the appearance of a rose
coloration. Each I c.c. of the thiocyanate solution corresponds to
o.OO25 g. As. This method is rapid, and the results obtained agree
very closely with those obtained by the first method, and also with
284 MANUFACTURE OF SULPHUROUS ACID
those of Clark, who employs a mixture of magnesia and sodium
hydroxide for fusing the ore (cf. infra).
This second method of Blattner and Brasseur is to be preferred to
the first, since a loss of arsenic by volatilisation as arsenious chloride
may occur in the latter.
The same authors have also described a method 1 based on the
precipitation of the arsenic as tri-iodide and the subsequent volumetric
estimation of the latter.
List” heats 2 g. pyrites with IO g. sodium peroxide in a crucible, as
described on p. 276. By this treatment all the arsenic is converted
to arsenic acid, which may be estimated as in the method of Blattner
and Brasseur.
Clark “has devised the two following methods for the estimation of
arsenic in pyrites:—
I. Precipitation method. About 3 g. of the ground sample is mixed
with four times its weight of a mixture of equal parts of freshly cal-
cined magnesia and pure sodium hydroxide previously ground together
in a porcelain mortar, and the whole heated for about ten minutes in an
uncovered platinum crucible; the mixture contracts slightly, but does
not fuse. The product is extracted with hot water, and the filtrate
acidified with hydrochloric acid, whereby sulphuretted hydrogen is freely
evolved. On boiling the nearly colourless solution thus obtained for a
few minutes, the arsenic separates as Sulphide. After saturation with
sulphuretted hydrogen to ensure complete precipitation, the precipitate
is filtered off, washed, the arsenic trisulphide dissolved in ammonia, and
the resulting solution evaporated to dryness on the water-bath. The
residue is dissolved in a small quantity of strong nitric acid and the
arsenic estimated as magnesium ammonium arsenate, or precipitated as
silver arsenate, and its weight calculated from the silver found, as
determined by titration by Volhard's method, or by cupellation accord-
ing to Richter. -.
The method gives very accurate results, and is applicable even when
the arsenic is only present in very small quantity.
2. Distillation method. About 1.7 g of the finely powdered pyrites
mixed with six times its weight of the magnesia-sodium hydroxide
mixture is heated over a Bunsen flame of medium height in an un-
covered platinum crucible for one hour, by which time oxidation
should be complete. The contents of the crucible are transferred
to a flask, moistened with water, and dissolved in about 70 c.c. of
strong hydrochloric acid. It is necessary to heat towards the end,
until all action ceases. The flask is fitted with a funnel tube which
dips below the surface of the liquid, and it is further connected with a
* Z. angew. Chem., I904, I'7, 1727. * /bid., 1903, 16, 415.
* /. Soc. Chem. Ind., 1887, 6, 352.
PYRITES 285
small glass cooling-worm, to the end of which a straight calcium chloride
tube is attached. A considerable excess of reducing agent, dissolved in
strong hydrochloric acid, is then admitted through the funnel tube. The
reducing agent employed is cuprous chloride, which forms a readily
soluble double salt with sodium chloride, and reduces quite as readily
as a ferrous salt. A mixture of cuprous and ferrous chlorides, obtained
by dissolving copper in ferric chloride, also forms an excellent reducing
agent. The contents of the flask are then slowly distilled into water
for an hour, after which about 30 g. of strong hydrochloric acid are
added and the distillation continued for a further half-hour. All the
arsenic should now be in the receiver, but it is advisable to check this
by adding more hydrochloric acid, changing the receiver, and testing
the fresh distillate collected. The arsenic so recovered may be precipi-
tated as sulphide and collected on a tared filter paper, or it may be
titrated with iodine in the usual manner.
The distillation method requires less time than any other, and gives
equally exact results even for small amounts of arsenic. It may also be
employed with advantage in estimating the arsenic in metallic copper.
According to H. Fresenius, the fusion method offers no advantages
over that of direct distillation in the presence of ferrous chloride, where
the solution has been treated hot with chlorine or with hydrochloric acid
and potassium chlorate.
The method of determining arsenic proposed by Kühn and Säger.”
which consists in obtaining and weighing an arsenic mirror, is too
troublesome for technical work.
The methods for detecting and estimating traces of arsenic are
described on pp. 362 to 376.
4. Copper.—The following process is employed at the Duisburg
copper works, and was first described in the Alkali Makers' Hand-
book. Five g. of the pyrites, ground and dried at IOO", are gradually
dissolved in 60 c.c. nitric acid of sp. gr. 1.2 in an inclined Erlenmeyer
flask. As soon as the violence of the reaction is over, the flask is
warmed and evaporation allowed to proceed until sulphuric acid vapours
are given off. The dry residue is dissolved in 50 c.c. hydrochloric acid
of sp. gr. I. 19, and after the addition of sodium hypophosphite (2 g
sodium hypophosphite, NaH2PO, dissolved in 5 c.c water) the solution
is boiled for a time to remove arsenic and to reduce the ferric chloride.
An excess of strong hydrochloric acid is then added, the solution diluted
with about 300 c.c. of hot water, sulphuretted hydrogen passed in, and
the precipitate filtered off and thoroughly washed. A hole is made in
the filter by means of a glass rod, the precipitate washed back into
the precipitating flask, and the metallic sulphides adhering to the filter
* Z. anal. Chem., 1888, 27, 34; cf. also, Nahnsen, Chem. Zeit., 1887, II, 692.
* Ber., 1890, 23, 1798. -
286 MANUFACTURE OF SULPHUROUS ACID
paper as well as the main bulk of precipitate, brought into solution by
treatment with nitric acid; the solution in the flask is then evaporated
to dryness on the water-bath, and the residue again taken up with nitric
acid and water, neutralised with ammonia and dilute sulphuric acid
added in slight excess. When cold, the solution is filtered to remove
lead sulphate and other insoluble matter, the flask and the filter being
washed with water containing sulphuric acid, 3 to 8 c.c. of nitric acid
of sp. gr. 1-4 added to the filtrate, and the copper deposited electro-
lytically. From the weight of copper found, O.OI per cent, is deducted
for the simultaneously deposited bismuth and antimony.
Nahnsen estimates the copper in pyrites as follows. The pyrites
is very finely ground, dried, and 12.5 g. of the dried material are
covered with Io c.c. of water and I c.c. of strong sulphuric acid in a
thin-walled beaker about 17 cm. deep. The beaker is covered with
a porcelain dish and nitric acid of sp. gr. 1-4 added in successive
small portions until frothing ceases to occur. The solution is then
boiled over a fairly large flame; after several minutes the porcelain
dish, which is sufficiently washed by the acid which condenses under-
neath, is removed and the gradually thickening solution is allowed to
boil vigorously with constant agitation, until the liquid almost adheres
to the glass and yellow particles of Salt begin to separate. The magma
is then quickly brought into solution by the addition of warm water.
With some practice the whole operation may be performed in from
ten to fifteen minutes. The cold solution is transferred to a 250 c.c.
flask, made up to the mark and filtered through a dry filter paper. Two
hundred c.c. (= Io g. pyrites) of the solution, which is now free from
silica and lead, are treated for some hours with a good current of
sulphuretted hydrogen, until the precipitate becomes flocculent and
the solution appears clear. The solution is then filtered and the
precipitate continually pressed out by means of a glass rod and washed
on the filter paper with water. It is unnecessary to wash copper
sulphide with sulphuretted hydrogen water when, as in this case, the
sulphuretted hydrogen has been allowed to act until the precipitate
settles well and leaves a clear supernatant liquid. The portion of the
precipitate remaining on the filter is washed back into the main bulk
with the least possible quantity of hot water, and concentrated sodium
sulphide solution added in such quantity that no sulphur remains
undissolved after the solution has been brought to, and maintained for
several minutes at, boiling point. After diluting with water and allowing
to settle in a warm place, the solution, which contains the arsenic and
antimony, which is now quite free from copper, is filtered, and the sulphide
of copper washed with hot water. Traces of iron sulphide, which always
adhere to the sulphide of copper (in two cases this amounted to O.O2
* Chem. Zeit., 1887, II, 692.
PYRITES 287
per cent on the pyrites), are extracted by washing with hot water
acidified with a few drops of hydrochloric acid; the precipitate is then
washed until free from chlorine, and the copper estimated as cuprous
sulphide. -
Cadmium and bismuth sulphides, present in the cuprous sulphide,
are determined by dissolving the precipitate in nitric acid, and treating
the warm solution for some time with ammonium carbonate. The
resulting precipitate is weighed as oxide and allowed for.
Should 25 g. be taken for analysis instead of 12.5 g., the method,
according to the author, is but slightly more troublesome and longer.
When the pyrites contains 3 to 5 per cent. of copper, 5 g. of the sample
will suffice.
A very complete account of the estimation of copper in pyrites,
and more particularly of the “Cornish assay,” has been published by
Westmoreland."
List” burns the pyrites in a porcelain crucible of special shape,
extracts with hydrochloric acid, neutralises with ammonia, adds
sulphurous acid, and precipitates the copper as cuprous thiocyanate;
this is oxidised with a mixture of sulphuric and nitric acids to Sulphate,
and the copper determined by electrolysis.
5. Lead.—Lead remains in the residue in the form of sulphate
when the pyrites is treated with aqua regia, as described on p. 272.
The lead sulphate is extracted from the insoluble residue by warming
with a concentrated solution of ammonium acetate; the resulting
solution is evaporated with addition of a little pure sulphuric acid, and
the residue finally dried and ignited in a porcelain dish or crucible.
One part PbSO, =0.6829 Pb.
6. Zinc.—The zinc in pyrites is occasionally determined, according
to Hassreidter and Prost, since it is scarcely possible to recover the
sulphur combined with the zinc. Schaffner's method, as described
under Zinc Blende on p. 289, is not available, owing to the large pre-
ponderance of iron, and on this latter account a gravimetric method must
be resorted to. One g. Of the pyrites is dissolved in aqua regia, as de-
scribed on p. 272, the excess of nitric acid evaporated off, and the residue
taken up with 5 c.c. of strong hydrochloric acid; water is then added
and the solution treated with sulphuretted hydrogen to remove any
metals yielding sulphides insoluble in acid solution. Any precipitate
formed is filtered off, and the filtrate freed from sulphuretted hydrogen
by boiling and oxidised with aqua regia. The oxidised solution is
treated, when cold, with ammonium carbonate, until the resulting
precipitate only redissolves very slowly; ammonium acetate is then
added, the solution boiled for a short time, and filtered. The basic
ferric acetate so precipitated carries down some of the zinc, and must,
* J. Soc, Chem. Ind., 1886, 5, 49 and 277. * Z, angew, Chem., 1903, 16, 416.
288 MANUFACTURE OF SULPHUROUS ACID
therefore, be dissolved in hydrochloric acid, and again precipitated as
described, the operation being repeated so long as zinc can be detected
in the filtrate. The combined filtrates are concentrated if necessary,
and the zinc precipitated in the warm solution by sulphuretted
hydrogen ; the whole is then allowed to stand for twenty-four hours,
the clear liquor decanted, and the zinc sulphide filtered off and washed.
The precipitate and filter are treated with dilute hydrochloric acid, and
the resulting solution, after boiling till free from sulphuretted hydrogen,
filtered, precipitated with sodium carbonate, the zinc carbonate well
washed, dried, and converted by ignition to ZnO of which one part =
O.8034 Zn. For very exact determinations any silica, oxide of iron, and
alumina precipitated with the zinc oxide must be estimated and allowed
for ; this, however, is seldom necessary.
Mayer and Lösekann" have worked out a method for estimating
the zinc as zinc ammonium phosphate.
7. Carbonates of the alkaline earths are occasionally estimated,
since when present they cause sulphur to be retained, owing to the
formation of the corresponding sulphates. Since their quantity is always
small, the carbonic acid cannot be determined accurately by loss, and
consequently a direct method must be used. The carbonic acid is
liberated by addition of strong acid and absorbed in soda lime, taking
care to retain moisture and any acid carried forward ; either the
apparatus of Fresenius” or that of Classen & may be employed for the
determination. Greater rapidity and accuracy can be attained by
measuring the volume of evolved gas in the Lunge and Marchlewski
calcimeter ‘(p. 149).
8. Carbon.—It is occasionally necessary to estimate the carbon
present in pyrites, especially in the products sorted out from coal and
known under the name of “coal brasses.”* This is done, according to
Treadwell and Koch,” by combustion in a porcelain boat, as in the
ordinary method of organic analysis, employing a 30 cm. layer of copper
Oxide and 25 cm. of lead peroxide to retain the Sulphur; or more
conveniently and in shorter time by oxidising with a mixture of
chromic and sulphuric acids. For the latter process they follow the
method of Corleis," passing the gas liberated in the decomposing flask
through a Io cm. layer of glowing copper oxide, followed by IO cm. of
Solid chromium trioxide, and after this through two small U-tubes each
containing 3 c.c. of a solution of chromium trioxide in concentrated
sulphuric acid. After leaving the U-tubes the gas is passed through a
* Chem. Zeit., 1886, Io, 729. * Mohr's Titrirmethoden, 7th edition, p. 641.
* Quantitative Analysis, vol. i., p. 449. * Z. angew. Chem., 1891, 4, 229.
-* * Cº. Lunge, Sulphuric Acid and Alkali, 3rd edition, 1903, p. 4I.
* Z, angew. Chem., 1903, 16, 173.
7 Stahl u. Eisen, 1894, 14, 587 ; cf. Vol. II., “Iron.”
ZINC BLENDE 289
tube filled with glass beads moistened with sulphuric acid, then through
a couple of calcium chloride tubes, and is finally absorbed in two
weighed tubes filled with soda-lime. The results obtained in this
manner were identical with those of the combustion method (26.26
per cent. carbon).
9. It is necessary in some instances to distinguish between Ordinary
and magnetic pyrites (Pyrrhotite), Fe,Ss, more especially in the case of
American ores. To effect this, the mineral is ground till it will pass
through a sieve of sixty meshes to the inch (24 per cm.), but not finer,
and the powder spread on a sheet of glazed paper. The magnetic
pyrites is then picked out by the aid of a magnet and freed from any
ordinary pyrites adhering mechanically, by gently tapping the magnet,
after which the armature is fixed to the magnet and the magnetic ore
brushed off. This operation is repeated five or six times and the
sulphur separately estimated in the two portions.
IV. ZINC BLENDE
I. Total Sulphur.—o.5 g. of the very finely ground sample is
covered with about 20 c.c. of a mixture of three parts strong nitric acid
and one part strong hydrochloric acid, or with hydrochloric acid saturated
with bromine, allowed to stand covered overnight, evaporated almost to
dryness, and a few C.C. of hydrochloric acid and 50 c.c. of water added ;
the solution is filtered hot and precipitated with barium chloride, after
removal of the iron by ammonia, as described on p. 273, should this be
present in any quantity.
According to Thiel,” the estimation of sulphuric acid in the presence
of large quantities of zinc always leads to low results, owing to the
formation of complex salts on precipitation with barium chloride (with
equivalent quantities of zinc the results are O.33 per cent, too low).
This source of error may be avoided if, before the addition of the barium
chloride, the zinc is exactly precipitated as hydroxide by addition of
ammonia and the hydroxide subsequently dissolved out of the mixed
precipitate with a slight excess of hydrochloric acid before filtering.
2. Zinc.—According to the modified Schaffner method now employed
at zinc works,” 2.5 g. of the blende, dried at IOO’ and finely ground, are
treated with 12 c.c. of fuming nitric acid in an Erlenmeyer flask of 200
c.c. capacity, first in the cold and then with gentle heating to disappear-
ance of the red fumes, when 20 to 25 c.c. of strong nitric acid are added
and the whole evaporated to dryness on the sand-bath. Five c.c. of hydro-
chloric acid and a little water are then added and the mixture warmed
until as much as possible has dissolved, after which the Solution is
* Cone, J. Amer. Chem. Soc., 1896, 18, 404. * Z, anorg. Chem., 1903, 36, 85.
* Communicated by V. Hassreidter and E. Prost.
T
290 MANUFACTURE OF SULPHUROUS ACID
diluted with 50 to 60 c.c. of water and the whole heated to 60° to 70°,
until only gangue and separated Sulphur remain undissolved. A
moderate current of sulphuretted hydrogen is next passed into the
solution and 50 to IOO c.c. of cold water added, little by little, with
continuous stirring until the precipitation of lead and cadmium sulphides
is complete, which may be recognised by the transparency of the
bubbles as they rise through the solution. Excessive dilution and too
slow a current of sulphuretted hydrogen should be avoided. The
solution is filtered and the precipitate washed with IOO c.c. of sulphur-
etted hydrogen water, to which 5 c.c. of hydrochloric acid have been
added, until a drop of the filtrate ceases to give the reaction for zinc with
ammonium sulphide. The filtrate and washings, which amount in all
to about 300 c.c., are boiled to remove the sulphuretted hydrogen,
the complete removal of which is checked by lead acetate paper, and
the ferrous iron then oxidised by addition of 5 c.c. of strong nitric acid
and Io c.c. of hydrochloric acid. After partial cooling the solution is
poured into a 500 c.c. flask, IOO C.C. of ammonia of sp. gr. O.9 to
O-91 and IO c.c. of a cold saturated solution of commercial ammonium
carbonate added, and the whole well shaken and allowed to cool.
In the meantime an ammoniacal solution zinc of known content
is prepared, by dissolving in a 500 c.c. flask chemically pure zinc,
approximately equal in quantity to the zinc content of the ore, in 5
c.c. of nitric acid and 20 c.c. of hydrochloric acid diluted with about
250 c.c. of water. After solution, IOO c.c. of ammonia and IO c.c.
of a saturated solution of ammonium carbonate are added, the whole
shaken, and allowed to cool. In the presence of manganese, IO c.c. of
hydrogen peroxide are added before the ammonia. When quite cold,
both flasks are filled with water to the graduation mark, and the solution
prepared from the ore passed through a dry, pleated filter paper. For
titration, IOO c.c. are pipetted from each solution into a thick-walled
glass cylinder or beaker, and each portion diluted with about 200 c.c.
of water. The standard solution for titration consists of a concentrated
solution of commercial sodium sulphide crystals diluted with IO to 20
times its volume of water, so that each c.c. corresponds to O.OO5 to O.OIO
g. of zinc. The sodium sulphide solution is allowed to flow alternately
from two adjacent 50 c.c. burettes into each of the two solutions, at
first in 2 or 3 c.c. at a time, and later in such smaller quantities as may
appear necessary; the solutions are stirred and a drop of each solution
placed simultaneously, by means of thin glass rods, on a strip of Sensitive
lead acetate paper. After ten to fifteen seconds have elapsed, the drops
are washed off by the aid of a small wash-bottle, and the titration with
the sulphide solution continued until the drops taken from each of the
solutions, after an equal lapse of time, give rise to weak but distinctly
perceptible brown stains of equal intensity. Should the volume of liquid
ZINC BLENDE 291
withdrawn for the tests be appreciable, the operation must be repeated ;
in all cases the end-reactions in the two solutions must be reached at the
same time and the readings made to within O-O5 c.c.
If the quantity of pure zinc taken for the standard solution be
denoted by a, the number of c.c. of the sodium sulphide solution required
per IOO c.c. of the standard by b, and the number of c.c. required per
IOO c.c. of the solution of the ore (=O-5 g. ore) by c, then the percentage
4O &C
&
In very exact determinations a quantity of ferric chloride, corre-
sponding to the iron content of the ore, should be added to the standard
solution, to compensate for the zinc carried down with the ferric
hydroxide precipitate.
F. Meyer’ treats or 5 g. of the ore, in an ordinary flask, with IO c.c.
of aqua regia (made up of two parts of hydrochloric to one part of nitric
acid), starting cold and then warming, evaporates off the excess of acid,
takes up the residue with IO c.c. of sulphuric acid (I : 2), and evaporates,
the solution, to precipitate the lead, until dense white fumes are evolved.
He then allows the solution to cool, dilutes with 60 to 80 c.c. of hot
water, precipitates copper, cadmium, etc., by the addition of IO C.C. of
Sodium thiosulphate solution (1:8), boils until the solution is nearly
clear, and, after filtering from gangue, lead sulphate, etc., and boiling
the filtrate with 5 c.c. of nitric acid until the sulphur has collected into
globules, precipitates the iron and aluminium with 30 c.c. of ammonia
of sp. gr. O.92; finally, when nearly cold, the manganese is thrown
down by the addition of 20 c.c. of bromine water. After standing for
Some time the bromine is expelled by boiling, the solution filtered into
a thick-walled beaker, the flask and filter paper partially washed, the
contents of the filter dissolved in 5 c.c. of aqua regia and collected in
the flask, the iron, aluminium, and manganese precipitated as before by
20 c.c. of ammonia solution and 20 c.c. of bromine water, the filtrate
added to the preceding and the two made up to 500 c.c. The solution
is then allowed to stand overnight and titrated with sulphide of sodium
Solution, of which I c.c. corresponds to about O.OI g. zinc, using a burette
graduated in Po c.c. The preparation of the standard solution and the
titration are carried out as in Hassreidter and Prost's method described
above.
Jensch" has drawn attention to the occurrence of certain siliceous
zinc ores which obstinately resist the ordinary methods of treatment.
A critical review of the various volumetric methods of estimating
zinc has been published by O. Brunck.”
Herzº estimates the zinc gravimetrically by precipitation with
of zinc in the ore is given by the formula
* Z, angew. Chem., 1894, 7, 391. * Chem. Zeit., I903, 27, 339.
* Ibid., 155. - * Z, anorg, Chem., 1901, 26, 90.
292 MANUFACTURE OF SULPHUROUS ACID
dimethylamine, but according to Herting this method is not to be
recommended.
Herting" is of opinion that the gravimetric determination, by pre-
cipitating as basic carbonate and weighing as oxide, is preferable to
any volumetric method.
As a rule, it is necessary to separate the zinc from the other metals.
The ordinary method consists in precipitating as sulphide with sulphur-
etted hydrogen in acetic acid solution, or by adding ammonium sulphide.
In either case great difficulty is experienced in the filtration of the
sulphide, which readily passes through the filter paper. This difficulty
may be quite overcome if Murmann's suggestion” of adding a solution'
of mercuric chloride before precipitation be followed; the mixed pre-
cipitate of the sulphides of zinc and mercury settles in a few minutes,
and yields a clear solution on filtration. On ignition, the mercuric
sulphide is completely eliminated, whilst the sulphide of zinc is con-
verted into the oxide, even in the absence of a current of oxygen.
Herting strongly recommends this mode of procedure. He also con-
firms the observation of Flath,” to the effect that zinc may be completely
separated from iron by a double precipitation of the latter with
ammonia, especially after the addition of ammonium acetate.
3. Lead.-The metallic sulphides precipitated as above (2) are, if
necessary, digested with a moderately concentrated sodium sulphide
solution, the resulting solution diluted, filtered, and the insoluble matter
washed. The residue is dissolved in dilute nitric acid, the solution
filtered and evaporated with excess of sulphuric acid, to obtain the lead
in the form of sulphate. One part PbSO, =0.6829 Pb.
4. Calcium and Magnesium fix sulphur during calcination, and
must consequently be estimated. Two to 5 g. of the blende are digested
by warming with 50 c.c. of dilute hydrochloric acid (I : IO), the solution
decanted, and the operation repeated once or twice. The residue is
washed, the filtrate freed from sulphuretted hydrogen by boiling,
oxidised by bromine water, and precipitated with ammonia free from
carbonate. The calcium in the filtrate is precipitated in the usual
way by addition of ammonium oxalate and, after strong ignition,
weighed as oxide; the magnesium is determined by the addition of
ammonium phosphate to the filtrate from the oxalate. (Cf. infra under
Sulphate.)
5. Arsenic, as described above, p. 281.
6. Carbonic Acid may be estimated as described under Pyrites (p.
288). This estimation is of some importance, since blende Occasionally
contains spathic iron ore and calamine, in addition to calcium and
magnesium carbonate.
* Chem. Zeit., 1903, 27, 987. * Monalsh., 1898, 19, 404.
* Chem. Zeit., 1902, 26, 6II.
ZINC BLENDE 293
7. Fluorine is determined, according to Prost and Balthaser, by
mixing 5 g. of the blende with powdered quartz, heating with sulphuric
acid, condensing the liberated silicon fluoride in water, filtering off
the separated silica, precipitating the hydrofluosilicic acid with potassium
chloride and alcohol, and weighing on a tared filter paper. This method
gives results from O-6 to O.8 per cent. too low.
Bein” estimates the fluorine indirectly, from the silica separated as
above. This method is, however, inexact.
Bullnheimer" obtained the best results by the following modifica-
tion of the method of Fresenius.* An Erlenmeyer flask of 300 to
350 c.c. capacity is fitted with a rubber stopper, bored in three places,
for a thermometer and for inlet- and outlet-tubes respectively. To the last
is connected a U-tube filled with glass wool, and beyond this a Winkler
cooling worm, which is cooled by immersion in water; a Drehschmidt
wash-bottle filled with 80 c.c. of potassium chloride solution is attached
to the worm. The inlet-tube is intended, as in the Fresenius method,
for admission of air, previously dried and purified by sulphuric acid
and soda-lime. About 2.5 g. of the blende are intimately mixed with
from 3 to 5 g. of powdered quartz, and treated in the flask with 20 g.
of chromic acid and IOO c.c. of concentrated sulphuric acid. Should
the chromic acid not be absolutely dry, it must be mixed with the
sulphuric acid before adding it to the ore. The flask is quickly
stoppered, shaken, and a current of air passed through, at first in the
cold, and subsequently with gradual heating to about 80°. When the
evolution of oxygen begins to slacken, further heat is applied, until
the temperature reaches I 50° to 160°; so long as Oxygen is coming
off, the current of air may be discontinued. The reaction proceeds
sometimes quietly, at other times violently. After three hours' heating,
all the silicon fluoride should have been driven over into the receiver,
and nothing more should be evolved from the absorption flask. The
contents of the latter are then emptied out, an equal volume of alcohol
added, the mixture allowed to stand for a time, and then titrated with
N/Io sodium hydroxide solution, using phenolphthalein as indicator.
The titration must be done rapidly, and the reading taken at the first
reddening of the indicator.
8. Available sulphur.—From the total sulphur, estimated as directed
on p. 289, there must be deducted —
for I part Pb found in No. 3 : O. I 595 parts S,
» I CaO y) 4 : O. 57 I2 32 S,
» I MgO yy 4 : O'7943 y) S.
The quantity of sulphur remaining after these deductions is that
* Z, angew. Chem., 1901, I4, IoI. * Z. anal. Chem., 1887, 26, 733.
* Z, angew. Chem., 1901, I4, Io9.
* Quantitative Analysis, 6th edition (1876), vol. i., p. 328.
294 MANUFACTURE OF SULPHUROUS ACID
*
available for the production of sulphuric acid. Any sulphur present as
heavy spar remains in the insoluble residue from the original extraction.
CONTROL OF WORKING CONDITIONS
The objects to be aimed at in the production of Sulphurous acid
are (i) the fullest possible utilisation of the raw material employed, and
(ii) the correct admixture of air. The former is controlled by the
analysis and the appearance of the cinders, and the latter by examina-
tion of the products of combustion. Anemometers find but little appli-
cation at this stage, but may be of great importance during the
subsequent conversion of sulphurous acid to Sulphuric acid.
I. BURNT ORE.—(Residues after Calcination.)
Crude or native Sulphur.—An analysis is scarcely necessary,
since with the small quantity of fixed residue usually present, it is
apparent to the eye whether combustion has been complete or not.
Should it be deemed desirable, the residue is examined by ignition in
a porcelain dish, or by Oxidation with aqua regia.
Gas Residuals.-The residues in this case contain iron oxide and
frequently lime. The presence of the latter renders a determination
of the total sulphur of little value; the available sulphur must be
estimated by the method given on p. 270.
Iron Pyrites.—The sulphur in pyrites cinders may exist either as
sulphide or as sulphate. In the case of pyrites containing lead or zinc,
it is impossible to completely decompose the sulphates of these metals
in the kiln, whilst with cupriferous pyrites the complete burning-off of
the sulphur is not desirable in view of the subsequent working up of
the burnt ore for copper by the “chlorinating roast.” In the latter
case the burnt ore may contain, according to circumstances, from 3 to
5 per cent. of sulphur, and in the case of mixed pyrites even larger
quantities. On the other hand, with pure iron pyrites, the sulphur
may be reduced to I per cent., at all events in the case of “smalls.”
As a rule, only the total sulphur is determined in the burnt ore;
in the case of cupreous pyrites the copper is also determined, and less
frequently the iron is estimated as well. -
Estimation of sulphur. The wet method (p. 272) is seldom employe
for this determination, partly because it is more troublesome than in
the case of the ore, and also because rapid dry methods which are
available are sufficiently accurate (cf. p. 276). If the cinders are in-
tended for use in blast furnaces, Jene 4 states that the dry methods of
determination should always be used, as the wet methods do not give
the total sulphur; this has been confirmed by Gottlieb.” Many of
* Chem. Zeit., 1905, 29, 362. * Ibid., 1905, 29, 688.
BURNT ORE 295
these are, however, untrustworthy or not sufficiently convenient. This
is the case, for example, with the method of Pelouze and also with
Kolb's method of ignition with sodium carbonate and oxide of copper."
The method proposed by Böckmann is as follows:—A sample of
the burnt ore from the kilns is taken every twelve hours, each kiln
being provided with two sample boxes, one large and one small,
numbered to correspond to the kiln. The samples are in the first
instance collected in the smaller receivers and brought in these to the
laboratory, where they are transferred to the corresponding larger
boxes. At the end of the week's working a good average sample is
drawn from the latter, broken down in an iron mortar, the coarsely
ground sample so obtained divided by quartering on a piece of stout
paper, and one of the quarters finely ground and sieved. The ground
samples are filled into bottles, also numbered to correspond to the kilns.
Finally, several grams of each sample are ground in successive small
portions in an agate mortar until no grittiness is felt on rubbing between
the fingers.
For the sulphur determination I-5 to 2 g of the very finely ground
material are weighed off, and mixed with about 25 g. of a mixture of six
parts sodium carbonate and one part potassium chlorate ; this mixing
is done in a large platinum dish, by the aid of an agate pestle fixed to a
wooden handle. The mixture is then fused over the blowpipe. The
melt is allowed to cool till only just warm, then covered with hot water,
heated to boiling, and both the solution and the insoluble residue
washed into a 250 c.c. flask. The contents are cooled under the tap,
made up to the 250 c.c. mark, and 4 of the solution filtered through a
pleated filter paper into a 200 c.c. flask. The insoluble portion of the
melt (oxides of iron, copper, etc.) only occupies a small volume, and it
is unnecessary to make any correction for this. The solution is acidified
with hydrochloric acid and the sulphuric acid estimated in the usual
II] a 1111621.
The method devised by Watson * is much more rapid than that of
Böckmann, but it is unsatisfactory in its original form, owing to the
difficulty of preventing the very finely divided oxide of iron from
passing through the filter and thereby interfering with the end-reaction
of the titration. Lunge * has modified the process so as to overcome
this difficulty, whereby the method has been rendered convenient and
rapid, and accurate to within O.2 per cent. The Watson-Lunge method
is as follows:—
Exactly 2 g of sodium bicarbonate of known alkalinity are
intimately mixed by means of a flattened glass rod with 3.2 g of the
ground cinders in a nickel crucible of 20 to 30 c.c. capacity; the mixture
* Cf. Lunge, Sulphuric Acid and Alkali, 3rd edition, 1903, vol. i., p. 74.
* J. Soc. Chem. Ind., 1888, 7, 305. * Z. angew. Chem., 1892, 5, 447.
296 MANUFACTURE OF SULPHUROUS ACID
is heated for ten minutes over a small gas flame the tip of which just
touches the bottom of the crucible. The crucible must be covered
during this preliminary heating, and the contents must not be dis-
turbed during this period lest some of the material be carried away by
the gas evolved. Finally the heating is continued for fifteen minutes
over a larger flame; the mixture is stirred frequently during this
period, and the temperature must not be allowed to rise sufficiently high
to cause fusion.
To prevent absorption of sulphur from the combustion products of
the gas, it is best to support the crucible in a hole cut in an obliquely
placed asbestos card, as shown in Fig. IO6, p. 243. The arrangement
proposed by D. Pfeiffer" is perhaps preferable. By the action of
atmospheric oxygen the sulphur is converted into sulphate at the cost
of a corresponding quantity of the bicarbonate. The contents of the
crucible are emptied into a porcelain dish, the crucible washed out with
water and the solution boiled for ten minutes, with the addition of a
saturated solution of sodium chloride; without this addition it is difficult
to prevent some oxide of iron passing through the filter paper at a later
stage. The insoluble residue is then filtered off and washed until free
from alkali, the filtrate cooled, and the unchanged sodium carbonate
titrated with normal hydrochloric acid (of which I c.c. =O.O5305 g.
Na,CO3, corresponding to O.OI603 g. S), using methyl orange as in-
dicator. If 2 g. bicarbonate require a c.c. of normal acid and the final
solution requires à c.c., then the percentage content of sulphur–º
The method is not applicable to pyrites containing much zinc
(cf. p. 298), nor to zinc-free cinders which contain 6 per cent. or more of
sulphur. For the latter the following mixture must be substituted for
that given above, in order to ensure complete oxidation and at the same
time to guard against fusion:*—1.603 g of the sample, 2.0 g. sodium
bicarbonate, 4 g. potassium chlorate, and 2 to 3 g. ferric oxide, free from
sulphur. The determination is otherwise conducted as above; the
sulphur content is = a – 6.
The sodium chloride employed must be quite pure, that is, not only
react perfectly neutral, but also be free from calcium and magnesium
chlorides, which, if present, would naturally fix a corresponding quantity
of Sodium bicarbonate.
At the Duisburg copper works the same method is employed for
estimating the sulphur in burnt ore as that described for pyrites (p.
276). In dissolving the substance, however, only a few drops of hydro-
chloric acid are added to the nitric acid, since with larger amounts of
hydrochloric acid sulphuretted hydrogen may occasionally be evolved.
List applies the sodium peroxide method (p. 276) to burnt ore.
* Chem. Zeit., 1904, 28, 38. * Lunge, Z, angew. Chem., 1906, 19, 27.
BURNT ORE 297
The following process is employed in one of the largest French
works for the regular daily control of the calcination; it enables a large
number of sulphur tests in burnt ore to be carried out simultaneously,
and the results are obtained in a comparatively short time. The
method depends upon the fact that at a red heat hydrogen will decom-
pose all sulphur compounds of iron with the formation of sulphuretted
hydrogen. The sulphuretted hydrogen so liberated is passed into a
standard solution of silver nitrate, and is estimated by titrating back
the excess of this reagent according to Volhard's method (p. 123). The
test is carried out as follows:–Several porcelain tubes, closed at both
ends by non-vulcanised rubber stoppers carrying glass tubes, are
arranged in a furnace heated by a six-flamed Bunsen burner. The
glass tubes are connected at the inlet end with a hydrogen generator
by means of a corresponding number of lengths of rubber tubing
furnished with screw clips; the outlet pieces are connected with
vertically bent tubes which dip into small test-glasses. The hydrogen
used must be freed from any accompanying sulphuretted hydrogen by
washing with silver nitrate solution. Each set is numbered, and the
porcelain tubes, stoppers, inlet- and outlet-tubes and test-glasses are
marked to correspond.
A preliminary test is first made to ensure that no precipitate results
when the hydrogen is passed into the silver nitrate solution in the test-
glasses. When this has been ascertained, exactly I g. of the finely
ground cinders is weighed into a numbered porcelain boat, the boat
pushed to the marked position in the porcelain tube by means of a
glass rod, and 25 c.c. of silver nitrate solution (containing IO-6O4 g.
AgNO, corresponding to 3.65 g. NaCl per litre) placed in each test-
glass. When all the weighed samples are in position a current of
hydrogen is passed through the tubes and regulated by means of the
screw clips to two to three bubbles per minute. After ten minutes,
when all the air has been expelled, the furnace is heated, at first
gently, and then gradually raised to a red heat. After one and a half
hours' heating, the elimination of the sulphur is complete. This is
shown by no further cloudiness appearing in the silver nitrate solution,
and by the improved and satisfactory settling out of the black pre-
cipitate of silver sulphide. The gas supply to the furnace is then
gradually reduced, and the hydrogen current interrupted. The test-
glasses are removed in turn, and without filtering off the precipitate,
I c.c. of iron-indicator (2.5 g. ferric nitrate dissolved in IOO c.c. nitric
acid of sp. gr. I-38) is added, and the solution immediately titrated
with ammonium thiocyanate solution to permanent redness. The
thiocyanate solution contains 4,752 g. per litre, and should
correspond exactly with the silver nitrate solution. If the number
of c.c. of ammonium thiocyanate required be denoted by a, then
298 MANUFACTURE OF SULPHUROUs Acid
the percentage of sulphur in the pyrites cinders is given by the
: ... 25 - d.
expression tº
IO
Copper is estimated in the Duisburg works as described on p. 285,
except that only I g. of the sample is employed, and solution effected by
hydrochloric acid, to which a few drops of nitric acid have been added.
Further, no deduction for bismuth and antimony is made from the
weight of copper obtained by electro-deposition."
List” ignites 5 g. Of the cinders with 5 g. of sodium bicarbonate in
an iron crucible, a treatment which renders the iron oxide readily
soluble in strong hydrochloric acid. The copper in the solution thus
obtained is estimated as described under pyrites (p. 285).
Iron. O.5 g. of the burnt ore is taken, and the iron brought into
solution by prolonged warming with strong hydrochloric acid ; the
boiling solution is then reduced by zinc free from iron, or, more con-
veniently, by stannous chloride, any excess of which is removed by the
addition of a small quantity of mercuric chloride. The ferrous chloride
solution so obtained is poured into 500 c.c. of water, to which have been
added about 2 g. of manganese sulphate, coloured pink by one or two
drops of potassium permanganate solution. The iron content is arrived
at by titration with M/IO potassium permanganate solution, of which
each I c.c. equals O.OO56 g. Fe, or, on O. 5 g. of burnt ore, I. I2 per cent.
of iron.
Zinc Blende.—Experiments by Lunge have shown that the Watson-
Lunge method (p. 295) for the determination of sulphur in burnt ore
is not applicable to roasted blende; subsequent investigations by
Lunge and Stierlin * have, however, proved that the oxidation can be
satisfactorily effected by the addition of potassium chlorate to the
sodium bicarbonate. 3.206 g. of the finely powdered and sieved sample
are mixed with 2 g of sodium bicarbonate and 2 g of finely powdered
potassium chlorate and heated for thirty minutes in a nickel crucible
over a flame 3 to 4 cm. high, the tip of which is 2 to 3 cm. from the
bottom of the crucible; after this interval the heating is continued for
a further twenty minutes with a bigger flame, so that the bottom of
the crucible is reached by the tip, and finally for ten minutes more with
a still stronger flame, so that the bottom of the crucible is red hot,
but care being taken that the contents do not fuse. The mass is then
extracted and titrated as described under the Watson-Lunge method
(p. 295). The method can also be applied to unburnt (green) zinc
blende, if 2 g of sulphur-free oxide of iron be added to the above
mixture for the decomposition.
As an alternative the sulphur may be estimated gravimetrically,
*. Cf. Fresenius, Z. anal. Chem., 1877, 16, 338. * Z, angew. Chem., 1903, 16, 416.
* Z, angew. Chem., Igoó, 19, 2I.
EXAMINATION OF THE KILN GASES 299
following the method described under raw blende (p. 289). The
precipitation of the iron by ammonia may be omitted in the ordinary
routine testing, at works.
Hassreidter and van Zuylen reduce the sulphates formed with
stannous chloride, and estimate the sulphur from the sulphuretted
hydrogen obtained, by passing it into brominated hydrochloric acid.
A rough guide to the progress of the calcination consists in treating
the calcined ore in a flask with IO c.c. of hydrochloric acid (I : 2) and
testing the evolved gas with neutral or slightly alkaline lead acetate
paper; the extent of the calcination is judged by the degree of
darkening obtained.” This test is employed by the foreman in the
furnace room. -
Zinc is estimated as on p. 289.
II. EXAMINATION OF THE KILN GASES
This examination extended formerly only to the determination of
the percentage volume of Sulphur dioxide in the gases, and was usually
carried out according to Reich's method,” the apparatus recommended
by Lunge being as a rule employed (cf. infra). It has, however, been
proved that the kiln gases always contain sulphuric anhydride, which
may either be useful, as in the manufacture of sulphuric acid, or
harmful, as in the preparation of sulphite liquor for use in the wood-
pulp industry. The quantity of sulphur trioxide present may reach Io
per cent of the sulphur content of the gas, and as it is not estimated
in Reich's method of testing, the conclusions drawn as to the proper-
ties of the gas from the determination of the sulphur dioxide alone,
may be quite erroneous, as has been shown by F. Fischer “ from other
considerations.
Lunge” consequently recommends the estimation of the total acidity
in the gases, that is, SO2+SOs, by passing them through a solution of
sodium hydroxide, coloured by phenolphthalein, until the colour dis-
appears, instead of merely passing them through iodine solution. Other
indicators are not suitable; phenolphthalein is the only one which
indicates the simultaneous formation of sodium sulphite and sodium
sulphate from the mixed gases (cf. pp. 65 and 72). In the manufacture
of sulphuric acid the Lunge test is generally sufficient for works
purposes; Reich's test may, however, if desired, be employed in
* Z. angew. Chem., 1905, 18, 1777; 1906, 19, I37.
* Meyer, Z. angew. Chem., 1894, 7, 7392.
* F. Reich, “Die bisherigen Versuche zur Beseitigung des schädlichen Einfäusses des Aiitten-
rauches bei den fiskalischen Hüttenwerken zu Freiberg, Freiberg, 1858.” Special reprint from
Berg and Hütten Zeitung; cf. also Winkler, Industriegase, vol. ii., p. 350; and, Lunge, Sulphuric
Acid and Alkali, vol. i., p. 402.
* Dingl. polyt.J., 1885, 258, 28. * Z. angew. Chem., 1890, 3, 563.
300 MANUFACTURE OF SULPHUROUS ACID
addition. Both tests should be carried out in the preparation of
sulphite liquors, and also when the kiln gases are worked for sulphur
trioxide and only the residual sulphur dioxide goes into the lead
chambers.
Reich's method for the estimation of Sulphur Dioxide.—The
kiln gases are drawn by means of an aspirator through water to which a
measured quantity of iodine and starch solutions have been added.
The end of the reaction is indicated by the disappearance of the blue
colour. The volume per cent. of sulphur dioxide in the gases is cal-
culated from the quantity of iodine solution employed and the volume
of water aspirated, this latter being the equivalent of the non-absorb-
able portion of the gas.
The arrangement of the Reich apparatus, as modified by Lunge, is
shown in Fig. I22.
The bottle A has a capacity of about 200 c.c., B a capacity of
IOOO c.c. After removal of the rubber stopper d, A is filled with dis-
tilled water to about a quarter of its capacity, and starch solution, Sodium
bicarbonate, and sufficient iodine solution to produce a strong blue
coloration added. The sodium bicarbonate increases the absorptive
capacity of the iodine solution and at the same time overcomes any
irregularity likely to occur in the course of the reaction. The bottle B
is filled almost completely with water, and the cork c is then fitted in the
hole specially made in the gas-main for the purposes of the test.
It is first necessary to test whether the apparatus is perfectly tight.
With this object the spring clip m is closed and the clip i opened; the
water should then only flow for a very short time. Should the water
continue to flow the apparatus is not air-tight, and the leak must be
located and repaired.
The rubber tubing 6 and the tube a are then filled with the kiln gas
to be examined. To effect this the clip m is opened and then the clip i
sufficiently to allow the water to flow slowly out and the gas to enter in
single bubbles through a and rise to the surface of the liquid. As soon
as decoloration takes place the clip i is closed and Io c.c. of M/IO iodine
Solution added through the opening d.
Before starting a test, the gas to be examined must first be syphoned
to the lower end of the tube a, so that the air contained in the
vessel A may be at the same pressure as that obtaining in the subse-
quent operation. This is done by cautiously opening the clip i until
the kiln gas is drawn to the desired point. This clip is then closed
quickly, the water collected in C thrown away, and the measuring
cylinder again placed in position. The clip i is then opened with the
one hand and the flask A shaken with the other until the colour of the
Solution disappears, when the clip is closed and the volume of the water
collected is read.
EXAMINATION OF THE KILN GASES - 301
The reaction proceeds according to the equation:—
2I-SO, +2H,O = 2 HI+H,SO,.
Consequently the IO c.c. of iodine solution employed (=o. 12697 g. I)
FIG. 122.
correspond to O.O32O3 g. SO2, or to IO-93 c.c. of SO, measured at o°
and 760 mm. Assuming that 128 c.c. of water have been collected in

302 MANUFACTURE OF SULPHUROUS ACID
any one determination, this corresponds to 128 c.c. of non-absorbable
gas drawn through the iodine solution. The total volume of gas
aspirated is, therefore, 1284- IO-93 = I 38.93 C.C., and the percentage
volume of SO2 –
Io.93 × IOO == 8 per Cent.
138.93
This calculation is avoided by making use of the following table :-
Cubic Centimetres | W*: º cent. | Cubic Centimetres | Volum? 5. cent.
Water collected. in Kiln Gas. Water collected. in Kiln Gas.
80 -1 12 125-7 8
84*1 11 °5 134°8 7-5
88°4 11 145 °2 7
93°2 10-5 157.2 6'5
98 °4 10 171-2 6
104 °l 9°5 187-8 5'5
110 °5 9 207-8 5
117-7 8'5 gº tº e tº º ſº

In this table no account is taken of variations in temperature and
barometric pressure; should it be desired to correct for these, the
volume read off is reduced to O’ and 760 mm. by Tables Nos. VI. to
VIII. (Appendix), and the percentage volume of sulphur dioxide taken
from the corrected volume by the above table. The necessary addition
for the volume of the absorbed gas is included in the table.
Reich's method is not applicable to gases containing more than
very small quantities of nitrogen acids. It cannot, for example, be
employed in the case of gases leaving the chamber system, since the
nitrogen acids contained in these would re-oxidise the hydriodic acid.
Lunge's method for the estimation of Total Acids (SO2+ SO3).-
The above apparatus may be employed, but the glass inlet-tube to the
absorption flask should preferably be closed at the
bottom, and the portion immersed in the solution
pierced with a number of small holes to subdivide
the stream of gas (Fig. I23). The gases are con-
ducted through a solution of M/IO sodium hydroxide
solution coloured with phenolphthalein until the
colour is just discharged, the flask being continuously
shaken during the absorption. The total acidity is
usually calculated to Sulphur dioxide by means of
the table given above."
Errors may arise with both methods under certain
conditions, owing to the presence of arsenious acid,
which accumulates in the absorption tube; they are best guarde
against by filtering the gases through asbestos. . -
* For further details, cf. Lunge, Z, angew. Chem., 1890, 3, 563.

s
*E=L
q
=
*=mºss-msm-mm-
E:#
:
===
FIG. 123.
EXAMINATION OF THE KILN GASES 303
Lunge employs the absorption bottle shown in Fig. I23, both for
this test and for that of Reich. It has a capacity of 4 IO c.c., and is
charged with 230 c.c. of water, IO c.c. of N/IO sodium hydroxide solution,
and 3 drops of phenolphthalein solution, for the total acidity determina-
tion. The gas is not aspirated continuously, as described above, but
intermittently in such fashion that a volume of gas, much smaller
than the free space in the bottle, is syphoned over, allowed to stand
over the solution, shaken for about half a minute, then a further
quantity of gas drawn over, and so on. A longer period of shaking is
especially necessary towards the end of the
absorption, and it is advisable to place a white
ground below the flask to facilitate the observa- «A
tion of the colour change. <!.
In the Thirty-fourth Report of the Inspector ſº
for Alkali Works (1897, p. 22), an absorption º
flask is described which gives good results even º
in the most difficult cases, such as for example,
the absorption of acid fog. It cannot be used
for iodine solution, owing to the presence of
rubber, but it is very suitable for the Lunge
test, for the estimation of hydrochloric acid gas,
and in many other cases. The apparatus is
shown in Fig. 124 in half the actual size. The
flask is fitted with a rubber stopper provided
with an inlet- and exit-tube as shown. The
former is 8 mm. wide, closed at the bottom,
and pierced with a number of small holes,
through which the gas passes to the double
bulb, which is attached by means of a rubber
stopper. The upper bulb is filled with small ====
cuttings of rubber tubing, which are kept in rºw.
motion by the stream of gas, which is thus
brought into very intimate contact with the absorbing solution; the
lower bulb is open at the bottom. The success of the apparatus
depends largely on the correct dimensions being adhered to. The
lower opening of the double bulb is 6 mm. in diameter, the lower bulb
I5 mm. and the upper 18 mm. in diameter, and the upper opening
through which the inlet-tube passes I 3 mm. wide. The gas passes
from the bulb into the flask through several small holes, and finally
leaves it through the exit-tube, which is narrowed below and widened
above to form a cylindrical chamber; the lower, narrow portion is
filled with the rubber rings and the upper, wider portion with glass
wool. When used for the absorption of acid vapours, the exit-tube
is moistened with water coloured by methyl orange, which serves

304 MANUFACTURE OF SULPHUROUS ACID
to indicate whether complete absorption is being effected in the
bottle. w
Lunge" has proposed to determine the acidity of kiln gases by
measuring their specific gravity, a method which might be developed
to give a continuous graphic record of the operation of the furnace.
Differences of I per cent of sulphur dioxide by volume affect the value
of the specific gravity in the second decimal place, those of O. I per cent.
in the third decimal place. Such measurements might be made by
a modification of the Lux gas balance, which, however, as at present
constructed is not suitable for use with acid gases, or by the aid of
F. C. Müller's% method of determining the density of gases. The
“Ados” apparatus (cf. p. 190) for the examination of flue gases might
also be employed ; this proposal, however, does not appear to have
been worked out. -
III. FINISHED PRODUCTS
These comprise, firstly, solutions of Sulphurous anhydride in water,
which, although seldom appearing on the market, are frequently
obtained as intermediate products in many works, as for example, in
Hänisch and Schröder's process for the preparation of liquid sulphurous
acid; secondly, liquid anhydrous sulphur dioxide; thirdly, the “sulphite
liquor" of the wood-pulp industry, a Solution of calcium or magnesium
sulphite in aqueous sulphurous acid containing more or less sulphuric
acid.
Solutions of Sulphurous Acid.—The determination of free sulphur-
ous acid, or of that present as bisulphite, may be made acidimetrically
by titration with normal alkali hydroxide solution, bearing in mind the
behaviour of the indicator employed. Of the Ordinary indicators, litmus
is quite unsuitable, phenolphthalein gives the colour change to red, when
the normal salt, Na,SOs, has been formed, whilst with methyl orange
the change to yellow corresponds to the formation of the acid sulphite
NaHSO, (cf. p. 65). In the absence of other acids (for example,
sulphuric acid, which, if present, may be detected by barium chloride
in hydrochloric acid solution), each I c.c. of normal alkali used will thus
correspond to O.O32O3 g. SO2, when phenolphthalein is used as indicator,
and to O'o64O6 g. SO, when methyl orange is employed. Should free
sulphuric acid or other strong mineral acid be present, the sulphurous
acid may be determined along with the stronger acid, by titrating two
portions, the one with methyl orange, and the second with phenol-
phthalein as indicator. Less normal alkali will be used in the first than
in the second case, in accordance with the behaviour of the two
indicators towards sulphurous acid. The difference between the
number of the c.c. used in the two cases, multiplied by O.O64O6, gives
* Z, angew. Chem., 1890, 3, 567. • * . * Ibid., 513. -
FINISHED PRODUCTS 305
the quantity of free sulphur dioxide in grams; the remaining alkali
corresponds to the free strong mineral acid.
An alternative method of estimating free sulphurous acid in the
presence of other free non-reducing acids, consists in running the acid
into iodine solution, observing the precautions given on p. I I 3. One
c.c. N/Io iodine solution corresponds to O.Oo32O3 g. SO, Air-freed
water must always be used for dilution in this method.
Pure aqueous solutions of sulphurous acid may also be tested by
determining the specific gravity; further information on this point
will be found in Lunge's Sulphuric Acid and Alkali.' The following
table, published by Giles and Shearer,” gives the specific gravities of
solutions containing from O-99 to I3-O9 per cent. SO2:


Specific Temp. Per cent. Specific Temp. Per cent.
gravity. ° C. SO2. gravity. ° C. SO2.
1°0051 15 °5 0-99 1*0399 15 °5 8:08
1°0102 15°5 2°05 1*0438 15°5 8°68
1°0148 15°5 2.87 1*0492 15 °5 9°80
1*0204 15°5 4°04 1*0541 15°5 10'75
. 1-0252 15-5 4 "99 1*0597 12-5 11.65
1-0297 15°5 5'89 1*0668 11 13°09
1*0353 15°5 7:01 e tº & e tº e tº G o
Liquid sulphurous acid may be examined in exactly the same way
as solutions of sulphurous acid. The contained sulphuric acid may
sometimes amount to as much as 20 per cent.”
Properties of liquid sulphurous an/hydride : *—

At 0°. At 15°. At 30°.
Specific gravity. 1°434 1°391 1°349
Vapour tension . e 1°5 2-7 4-5 Atmos.
One kg. corresponds to a gas volume of 348 litres at O’ and 760 mm.
Critical temperature, 155°.4; critical pressure, 78°.9; boiling point, at
760 mm. pressure, — IO’; melting point, — 79°.
Sulphites.—The normal sulphite and the acid sulphite present in
sulphites (solid or in solution), may be estimated as follows. The total
sulphurous acid is determined by titration with iodine, and the acid
sulphite by titration of a second portion with standard alkali and
phenolphthalein.
NaHSOs--NaOH = Na,SO, +H.O.
One c.c. normal sodium hydroxide solution corresponds to Oo32O3 g.
SO2 present as acid sulphite.
* Vol. i., p. 149. *J. Soc. Chem. Ind., 1885, 4, 503. * Papier Zeit., 1892, No. 62.
* According to A. Lange, 7aschenā, d. Berl. Bezirksvereins deutscher Chemiker, 1898-99, p. 82.
U
306 MANUFACTURE OF NITRIC ACID
If the quantity of sulphurous acid present in a solution exceeds that
necessary to form bisulphite, the solution is first titrated with normal
sodium hydroxide and methyl orange until the red colour changes
phenolphthalein is then added, and the titration continued till the
coloration is formed. Each I c.c. of the alkali required with methyl
orange corresponds to O.O64O6 g. SO2 in the free condition, and each
further c.c. required for the subsequent titration with phenolphthalein
to O'o64O6 g. SO2 present as NaHSOs, or half this quantity = O.O32O3 g.
as “half-free” sulphurous acid. -
Sulphite Liquors for the manufacture of Wood-pulp.–The
following estimations are necessary:-
1. Total content of sulphurous acid. The sulphite solution, which
usually contains about 50 g. of Sulphur dioxide per litre, is preferably
diluted to twice its volume and the diluted solution allowed to flow
from a burette into 25 c.c. of an acidified M/IO iodine solution until the
colour is discharged (cf. p. I I6). The volume of the iodine solution
employed oxidises O.O8OO75 g. SO2, consequently the volume of sulphite
liquor used for the titration contains this total amount of free and fixed
sulphur dioxide.
2. Content of half-free and of free sulphurous acid, that is, all sulphur
dioxide over and above that required for the formation of calcium
sulphite and which is present in the half-free condition as acid sulphite
or as free sulphur dioxide in solution, in the liquor. The quantity
present is arrived at by titration with normal sodium hydroxide solution,
using phenolphthalein as indicator. Each I c.c. alkali used corresponds
to O.O32O3 g. SO, in the half-free or free state.
NITRIC ACID MANUFACTURE
Chili saltpetre is the only raw material employed in this industry,
which is dealt with in this section. Sulphuric acid and the testing of
the various waste acids from the manufacture of nitrobenzene, nitro-
glycerin, pyroxylin, etc., which are frequently used in the nitric acid
industry, are described in a subsequent section.
CHILI SALTPETRE
Chili saltpetre should contain at least 95 per cent, and the better
qualities 96 to 98 per cent. of sodium nitrate. In addition to this there
may occur potassium nitrate,” Occasionally up to 9 per cent., sodium
chloride, sodium sulphate, Sodium iodate, sodium perchlorate, and
* Cf. Lunge, Chem. Ind., 1886, 9, 269 ; Hagen, Chem, Zeit., 1891, 15, 1528; and, Z. angew.
Chem, 1893, 6, 495 and 698,
CHILI SALTPETRE - 307
insoluble matter; also, in exceptional instances, sodium carbonate,
magnesium sulphate, and salts of the heavy metals.
Of these substances the following are usually only tested for qualita-
tively.
Potassium detected by platinic chloride.
Jodates. According to Beckurts," the solution is acidified with
nitric acid and potassium iodide starch solution added; the method
will detect r}t; mg. of iodic acid in I g. nitre. Or the iodic acid may
be reduced by zinc and the iodine liberated by heating with strong sul-
phuric acid, the solution diluted and shaken with carbon bisulphide,
which will extract any iodine present, giving a rose-red solution.
Bromine, under similar conditions, colours the carbon bisulphide
yellow to reddish yellow.
Sodium perchlorate, the presence of which was first recognised by
Beckurts,” and has since been repeatedly confirmed, is, according to van
Breukeleeven,” most easily detected by a micro-chemical test. This is
carried out by adding a little rubidium chloride to a few drops of a con-
centrated filtered solution of the saltpetre placed on a glass slide.
Permanganate is added to the solution to the production of a wine-red
colour, and the solution evaporated till single crystals form, when
the object glass is placed under the microscope and examined to
see whether, in addition to the colourless crystals of sodium nitrate,
the reddish-violet crystals of rubidium perchlorate are also present.
H. Fresenius and Bayerlein 4 recommend this method.
COMMERCIAL ASSAY OF SALTPETRE
In commercial dealings in Chili saltpetre an indirect method of
valuation has been generally employed until quite recently, and is still
used to a considerable extent for the large shipments to England and to
Hamburg.
According to this method only the percentages of moisture, sodium
chloride, sodium sulphate, and matter insoluble in water are determined,
the sum of these being called the refraction ; it is assumed that all the
rest is actual nitrate. This method * cannot in any way be looked upon
as accurate, and may lead to considerable error.
For example, the method quite overlooks any potassium nitrate
which may be present (cf. p. 306), and which, even in refined nitre,
may reach several per cent. Since the nitrate is employed for the
manufacture of nitric acid, explosives, etc., the higher nitrogen content
of sodium nitrate as compared with that of the potassium compound is
* Pharm. Centr., 1886, 233. * Chem. Centr., 1898, I., 960.
* Fischer's Jahresöer., 1886, 305. * Z, anal. Chem., 1898, 37, 501.
* Cf. Alberti and Hempel, Z. angew. Chem., 1892, 5, IoI.
308 MANUFACTURE OF NITRIC ACID
the chief consideration, and it cannot be a matter of indifference to the
manufacturer whether he, for instance, receives 9 per cent. potassium
nitrate in place of 9 per cent. Sodium nitrate, or whether a lot which,
according to the analytical results shows 96 to 97 per cent. of sodium
nitrate, on testing with the nitrometer is found to contain only 94 to 95
per cent. In addition, the “refraction method" also includes as nitrate
any perchlorate which is present.
COMPLETE ANALYSIS, WITH EXCEPTION OF THE DETER-
MINATION OF NITRATE AND PERCHLORATE
Moisture.—O.8 g. of the well-mixed and coarsely powdered sample
is weighed into a platinum crucible, and cautiously heated over a
small flame, so as to just fuse the nitrate; with a little practice it is
easy to hit off this point exactly without it being necessary to observe
the temperature. The crucible and contents are then allowed to cool
in a desiccator and weighed ; the heating is repeated at the same
temperature, to make certain that constant weight has been reached.
Or Io g. of the nitrate are dried in the air-bath at 130°, until the
weight is constant.
Insoluble Matter.—Fifty g. of the saltpetre are weighed into a
beaker on a balance turning to O.O5 g. The sample is dissolved in
water and filtered through a filter paper previously tared against a
similar paper (p. 28); the washed residue and the tare are then dried
together and weighed.
Should the insoluble residue appear to contain appreciable quantities
of organic matter, the latter is determined approximately by igniting the
filter paper and insoluble residue. The previous drying of the residue
is, in this case, preferably done at a higher temperature, Say I2O to I3O”,
since otherwise there will always be a small difference between the
weight obtained by drying at IOO’ and that got by ignition, even in
samples perfectly free from organic matter.
Chlorine, Sulphuric acid, Calcium, Magnesium, and Sodium car-
bonate.—Five g. of the sample are placed in a filter paper standing
over a 500 c.c. flask, and dissolved by addition of boiling water; any
sand remaining on the filter after washing is complete, is ignited and
weighed. The filtrate when cold is made up to 500 c.c., and of this
solution 50 c.c. are taken for the determination of chlorine by titration
or precipitation with silver nitrate (p. 123), the result being calculated
to sodium chloride. A further 50 c.c. are heated to boiling, and pre-
cipitated with barium chloride, the resulting barium sulphate filtered
off, weighed, and calculated to calcium sulphate. To estimate the
calcium and magnesium, 20 g. of saltpetre are dissolved in IOOO c.c. of
boiling water, and the calcium estimated by the addition of ammonium
CHILI SALTPETRE 309
oxalate to 500 c.c. of this solution, the magnesium in the remaining
5OO C.C. by addition of ammonium phosphate. The sodium carbonate
is obtained either by difference, or by dissolving 20 g. of the sample
in water, making up to IOOO C.C., and treating IOO c.c. of this solution
with sulphuric acid, evaporating to dryness, and igniting until the
weight of the residue is constant. The sodium carbonate is calculated
from the weight of sulphate so obtained, after making allowance for
the calcium and magnesium sulphates present. Any potassium present
must also be allowed for.
Potassium.—Potassium was formerly estimated indirectly by con-
verting the bases into Sulphates, but the method leads to very uncertain
results when, as in this case, one of the constituents is present in
only very small proportion.
A much more exact method consists in repeated evaporation of
the saltpetre with strong hydrochloric acid and precipitation with
platinic chloride, as described in the section on “Potassium Salts.”
The potassium is calculated to potassium nitrate, one hundred parts
of which are equivalent to 84.09 parts NaNO3.
*
THE ESTIMATION OF NITRATE
Numerous methods, of very unequal value, have been proposed for
the estimation of nitric acid in saltpetre. These are grouped together
in the following list as arranged by Böckmann, with several additions
of recent date; it is essentially a summary, and does not make any
pretence to be complete.
I. Methods depending on Reduction to Nitric Oxide.
A. By measurement of nitric oxide. I. Lunge's nitrometric method
(p. 3 I 5 ; Literature, p. 132).
2. The method of Schlösing-Grandeau' as improved by P. Wagner”
(p. 317).
B. Volumetric determination of the nitric acid obtained on oridising
nitric oxide with hydrogen peroxide. Wilfahrth's method.”
C. Reduction by ferrous salts and subsequent titration with perman-
gamate solution. Method of Pelouze and Fresenius.
II. Reduction to Ammonia in Alkaline Solution.
There are many modifications of this method, of which the following
may be mentioned :—
I. Stutzer's method.” Reduction by sodium hydroxide and aluminium
* Agriéulturchem. Analyse, p. 31. * Z, anal. Chem., 1888, 27, 411.
* Chem. Zeit., 1883, 7, 1710; 1884, 8, 475. * Z. angew. Chem., 1890, 3, 695.
310 MANUFACTURE OF NITRIC ACID
wire. Aluminium prepared by the old method of manufacture is the
only kind suitable, and this has ceased to be an article of commerce.
2. Sievert's method. Reduction with zinc dust and iron powder
in alcoholic potassium hydroxide solution.
III. Reduction to Ammonia in Acid Solution.
1. Ulsch's method.” Reduction by ferrum reductum, and dilute
sulphuric acid with subsequent distillation, after rendering alkaline
with sodium hydroxide solution. The method is described on p. 3 II.
2. Schmitt's method.” Reduction by a mixture of zinc and iron
dust in acetic acid solution.
3. Hildesheimer's modification * of Jodlbauer's method. Reduction
with phenolsulphuric acid, zinc dust, and mercury.
4. Förster's method.” Reduction with sulphosalicylic acid, sodium
thiosulphate, and mercury.
5. Ulsch's electrolytic method."
6. Ingham's electrolytic method.”
IV. Reduction to Nitrous Acid.
Gantter's methods Reduction by phosphorous acid, and measure-
ment of the nitrogen evolved on heating the resulting ammonium
nitrite.
V. Estimation of Nitric Acid by means of the Hydrogen Deficit.
Ulsch's method." A measured quantity of sulphuric acid is allowed
to act on strongly coppered iron alone and, separately, upon the couple
together with the nitrate solution; the hydrogen evolved in the two
cases is measured in the nitrometer and the volumes compared.
VI. Decomposition with Hydrochloric Acid.
I. Förster's method.” Two to 3 g. of the saltpetre, dried at 150°, are
evaporated with 25 c.c. of 19 per cent. hydrochloric acid three times
successively in a roomy porcelain crucible on the water-bath, and the
resulting chloride dried at 150° and weighed. The nitrogen equals the
loss in weight of the dried saltpetrex O. 5283.
2. Gowan's method.” Decomposition with hydrochloric acid and
* Cf. Fricke, Z. angew. Chem., 1891, 4, 240.
* Chem. Centr., 1890, II., 926; 2. angew. Chem., 1891, 4, 24.I.
* Chem. Zeit., 1890, 14, 14 Io; Z. angew. Chem., 1891, 4, 240.
* Sullwald, Chem. Zeit., 1890, 14, 1674. 7 /. Amer. Chem. Soc., I 904, 26, 1241.
* Chem. Zeit., 1889, I3, 229 ; 1890, I4, 1674. * Z, anal. Chem., 1895, 34, 25.
* Z. Elektrochem., 1897, 3, 546. * /öid., 1891, 30, 175.
* Chem. Zeit., 1890, I4, 509.
* Chem. Mews, 1891, 63, 245; Z. angew. Chem., 1891, 4, 557.
CHILI SALTPETRE 311
absorption of the resulting nitrosyl chloride and chlorine in potassium
iodide :-
HNO3 + 3HCl= Cl, + NOCl + 2H2O. (I HNO3=3Cl).
3. Bohlig's method." Decomposition with sulphuric and hydrochloric
acids, absorption of the liberated chlorine in potassium ferrocyanide,
and subsequent titration with permanganate.
4. Bensemann's method.”
VII. Ignition Methods.
I. Chromate method These methods are described on pp. 318
2. Silica method } and 319.
VIII. Combustion Methods with Soda Lime and other additions
(analagous to Will and Varrentrapp's method).
I. Arnold's method.” Ignition with soda lime, sodium thiosulphate,
and sodium formate.
2. Houzeau's method.” Ignition with soda lime, sodium thiosulphate,
and sodium acetate.
3. Boyer's method.” Ignition with soda lime, calcium oxalate, and
sulphur. -
IX. Precipitation by “Nitron.”
Busch's method.” “Nitron * is the name given to the base I : 4
diphenyl, 3:5 endanilodihydrotriazole; it forms a nitrate which is
almost insoluble in water and which is suited for the detection and
determination of nitric acid.
This summary of the better-known methods for the determination
of nitric acid is included on account of the importance in industrial and
agricultural chemistry of an exact determination of the nitrate in salt-
petre. More detailed accounts of Ulsch's method, Lunge's nitrometric
method, Wagner's improved form of the Schlösing-Grandeau method,
and of the methods depending on ignition with potassium chromate and
quartz respectively, are given below.
DETAILED DESCRIPTION OF THE CHIEF METHODS FOR THE
DETERMINATION OF THE NITRATE IN SALTPETRE
Uisch's Method, as modified by Bockmann.
The modifications relate to the size of the flask used for the
reduction and for the subsequent distillation; to more effective air-
* Z. anal. Chem., I900, 39, 498. * Z. angew. Chem., 1906, 19, 471.
* Chem. Zeif., 1885, 9, 715. * /hid., 998. * Comples rend., 1891, II.3, 503.
* Ber., 1905, 38,856 and 861 ; J. Soc. Chem. Ind., 1905, 24, 291; Gutbier, Z. angew. Chem.,
1995, 18, 494.
312 MANUFACTURE OF NITRIC ACID
cooling, and avoidance of other methods of cooling in the distillation;
to dispensing with the special bulb-tube intended to prevent alkaline
solution being carried into the distillate; to rendering a back titration
unnecessary; to the construction of the absorption vessel; to the quantity
of the absorbing solution, and to the strength of the Sulphuric acid
used for titration. -
Also, the weights of sample taken for analysis, of sulphuric acid for
decomposition of iron powder, and of sodium hydroxide solution em-
ployed by Ulsch and Fricke are doubled. These latter changes had
previously been recommended by Alberti and Hempel." The essential
features of Ulsch's method remain unchanged.
The following solutions are required for the determination — I
Dilute sulphuric acid (1:2). 2. Sodium hydroxide solution of sp. gr.
1.25, absolutely free from nitrogen compounds. 3. W/5 sulphuric acid
or N/5 hydrochloric acid prepared as described on p. 86.
Twenty g. of each of the well-mixed average samples, ground
somewhat further in a porcelain mortar, are weighed out, dissolved in
water, and the solutions made to IOOO c.c.; 50 c.c. (= I.O. g. saltpetre)
of each solution are then run into a round-bottomed flask of at least
I litre capacity. The neck of the flask should be at least IO c.c. long,
and of an internal diameter of 23 to 3 cm. One g. ferrum reductum
is added to the contents of each flask, and then 20 c.c. of the
diluted sulphuric acid (I : 2); as a precaution, an ordinary glass
funnel is placed in the neck of each. The funnel is not absolutely
necessary, since, with flasks of the dimensions given, any loss by
spurting during the violent evolution of gas is almost impossible.
When the gas evolution has either ceased altogether or has become
very slow, it is accelerated by placing the flasks on sheets of asbestos
and heating by a small flame, so that the solutions reach a quiet
boil in about four minutes; boiling is then maintained for about a
further six minutes, at the end of which time the reduction should be
complete.
The flasks are then allowed to cool slightly, in order that the
subsequent addition of water shall not crack the funnels, and after the
addition of about 150 c.c. of water, the solution is made alkaline by
the addition of 50 c.c. of the sodium hydroxide solution prepared as
above. At the same time a rapid rotary motion is given to the contents
of the flask, and any syrupy material adhering to the upper walls of the
flask is washed down by the aid of a wash-bottle. A pinch of zinc dust
is then added to each flask, and distillation carried out as follows:—The
distilling flask is placed on a sheet of asbestos card, supported on an
ordinary tripod, and the neck fitted with a rubber stopper, through
which a bent delivery tube passes for the escape of the gas; Lunge
* Z, angew. Chem., 1891, 4, 398.
CHILI SALTPETRE 313
recommends the use of a bulb-tube. The vertical portion of this tube
should have a length of 14 to 15 cm., of which a good IO to I2 cm.
- is outside the stopper, whilst the downward portion leading to the
absorption dish should be about 7 cm. long, and connected by a piece
of rubber tubing about IO cm. long to a second tube which dips into
the absorption dish, care being taken that the two tubes are held flush
together. This second tube is continued for about 7 cm. in the same
direction as the downwardly bent portion of the exit-tube from the
distilling flask, and is then bent at right angles and continued for a
length of 50 to 60 cm., when it dips into the absorption dish. This
is an ordinary earthenware dish, of about 8oo c.c. capacity; it is
charged with about 250 c.c. of distilled water to which about two drops
of a somewhat strong Solution of litmus have been added ; the solution
should be tinged a rather faint but distinct blue. A drop of the
standard sulphuric acid is then added to the contents of the absorption
vessel from a burette standing over the latter, and the solution is tested
for neutrality by a piece of sensitive litmus paper (p. 79). When the
solution is neutral, which may necessitate the addition of a further
drop of acid, the burette is filled with acid to the zero mark and the
flask heated ; the contents, although thick and coloured greenish
brown, owing to separation of ferrous and ferric hydroxides, distil
quietly and without bumping. Four distillations can be easily carried
out simultaneously, and due attention be given to the necessary
additions of acid from the burette. Should the nitrate content be
approximately known, as is frequently the case in works, so that 90
per cent. of the acid required for titration may be added at the start,
six distillations can be attended to at the same time.
Instead of the Böckmann absorption vessel, Lunge prefers a Peligot
or other similar tube, in which the gas inlet-tube does not dip into the
acid used for absorption (cf. Fig. 125, p. 315); this avoids the risk of the
acid being drawn back into the distilling flask.
As soon as the acid solution has been neutralised by the ammonia
evolved in the distillation, any further ammonia that comes over is clearly
indicated by blue stream lines rising from the mouth of the inlet-tube
through the reddish-coloured solution. The latter soon assumes a bluish-
red colour, unless fresh acid is added from the burette. Even should this
take place, there will be no loss of ammonia, provided that fresh acid be
added at once, always taking care to maintain the solution as nearly
neutral as possible by testing with sensitive litmus paper. Towards
the end of the operation, the solution in the absorption vessel becomes
appreciably warm, but no loss of ammonia need be feared on this
account. Finally, when no further addition of standard acid from the
burette has been necessary for several minutes, the inlet-tube is removed
from the absorption vessel and, after making certain by the smell that
314 MANUFACTURE OF NITRIC ACID
the steam coming off is quite free from ammonia, the distillation is
interrupted. º
With a distilling flask of the dimensions given, the surface of the
liquid is about 18 cm. below the gas exit-tube, and there is consequently
no danger of alkaline liquor being carried forward from the flask; in
fact the water condensing on the inner end of the exit-tube always
shows a neutral reaction, as may readily be proved at the end of the
distillation. -
Using N/5 sulphuric or hydrochloric acid, each I c.c. corresponds to
O-oi 702 g. NaNO3, O.O.2024 g. KNOs, O.OIO81 g. N2Os, or O,OI261 g.
HNOs. One g. pure NaNO3 consequently requires 58.75 c.c. of M/5
acid.
Böckmann recommends the use of burettes of IOO c.c. capacity
graduated to Tºm c.c.; this, however, necessitates a very inconvenient
length. His proposal to use burettes of this capacity and graduated
to ºn c.c. is quite unpractical. Any great error in gauging the quantity
of sulphuric acid to be added may be avoided by making the other
determinations (refraction determination and potassium estimation, if
necessary) prior to the nitrogen test. Should these indicate that more
than 50 c.c. of acid are requisite, the full burette content of 50 c.c.
may be run into the absorption vessel before beginning the distillation,
the burette refilled, and the distillation started. -
If the burette be read to O-O5 c.c., or preferably to O.O25 c.c., the
total error will not exceed 4: O. I per cent. of the nitrate if due pre-
cautions are taken. This degree of accuracy can, however, only be
attained after considerable practice.
Vogtherr' recommends the apparatus described below (Fig. 125) for
carrying out this test. It was primarily designed for nitrogen deter-
minations by Kjeldahl's method, in which the digestion with sulphuric
acid is carried out without the use of a fume chamber, and in the same
apparatus as is used for distilling off the ammonia, and serves also for
the estimation of ammonia in ammonium sulphate and analogous
determinations.
A Jena glass Kjeldahl flask is fitted above with a glass bell accur-
ately ground in so as to be gas-tight, the bell being continued as a glass
tube bent first in a slightly inclined downward direction and then verti-
cally. The vertical portion is connected by a cork or rubber stopper with
a bulb-tube the lower open end of which dips below the solution in the
absorption flask. The long neck and the glass bell prevent spurting of
the liquid from the reaction flask, the pear-shaped bulb prevents the
sucking back of the liquid used for absorption, whilst the cork or rubber
connections employed are reduced to a minimum.
A drawback to this otherwise very satisfactory method lies in the
* Chem. Zeit., 1902, 27, 988.
CHILI SALTPETRE - 315
fact that commercial ferrum reductum is not always sufficiently pure,
and according to Brandt," errors amounting to as much as O.8 per cent.
may arise from this cause.
FIG. 125.
Lunge's Nitrometric method.—The principle of this method, which
is due to W. Crum, was first rendered practically available by the intro-
duction of Lunge's “nitrometer” (p. 132). The method depends on
the fact that nitric acid, nitrous acid and their salts and esters, when
brought into intimate contact with mercury and a large excess of sul-
phuric acid are decomposed quantitatively, so as to yield the whole of
their nitrogen in the form of nitric oxide, the volume of which is
measured. The reaction is:—
2HNO3 + 3H,SO, +6Hg = 2NO+3Hg,SO, +4.H.O.
In the analysis of saltpetre the bulb nitrometer (Fig. 45, p. 136) may be
employed. It is preferable, however, to carry out the decomposition of
the nitre with mercury and sulphuric acid in a special decomposition
vessel, as shown in Fig. 52 (p. 144), so that the gas measurement can be
made over dry mercury. The decomposition may be effected in the
* Chem. Zeit., 1899, 23, 22.

316 - MANUFACTURE OF NITRIC ACID
ordinary simple nitrometer, taking note of temperature and barometric
pressure, and reducing the gas volume obtained to normal (O’ and 760
mm.), but it is more convenient to employ the gas volumeter (p. 138),
which allows the corrected gas volume to be read directly.
The quantity of nitre taken for the determination must be such that,
under the conditions of temperature and barometric pressure obtaining
at the time, not less than IOO C.C. and not more than I2O c.c. of
nitric oxide are evolved. This corresponds to about O. 35 g., a quantity
which has given rise to much adverse criticism of the method. So far
as such criticism relates to the accuracy of the actual determination it
may be dismissed, since a gas volume of over IOO c.c. is obtained, and
this volume can be read off with ease to O.O5 c.c. or even less, which
means an accuracy of at least I in 2000. This degree of accuracy is
not obtainable by any other method. There is more reason, however,
in the contention that a correct average sample cannot well be obtained
when so small a quantity as O-35 g. is taken for the analysis. This
objection may be overcome if, as recommended by J. Stroof, 20 g. of the
nitre, dried at I IO’, are thoroughly ground, and the exact average sample
taken from this again dried to constant weight.
When the bulb nitrometer without a decomposing vessel is em-
ployed, the operation is carried out as follows. The well-mixed finely
powdered sample, obtained as above described, is filled into a narrow,
marked weighing tube, and the tube corked and weighed. When filled
to the mark, the tube should contain about O-35 g. substance.
The contents are then shaken into a previously prepared “saltpetre
nitrometer,” that is, one of at least 130 c.c. capacity (p. I36), so that
the powder falls as far as possible on to the bottom of the glass cup.
During this operation the three-way cock must be in such a position
that none of its passages are open either above or at the side. About
half a c.c. of water is then poured in, and allowed to stand a short time
until the nitre is completely or nearly dissolved, when the solution and
crystals are drawn into the measuring tube by cautiously opening the
tap and lowering the levelling tube; the cup is washed out with , or at
most I c.c. of water, and about 15 c.c. of strong pure Sulphuric acid
admitted to the tube. If too much water is employed, that is, more
than 1% c.c. in all, the sulphuric acid is rendered too dilute and a
persistent froth containing basic mercuric sulphate, is formed which
causes difficulty in reading. The reaction is brought about by
thoroughly shaking the acid solution with the mercury. The levelling
tube is then placed approximately at its correct height so as to avoid
any great difference of pressure and consequent error owing to leakage,
and the apparatus allowed to cool for at least half an hour. The tubes
are then accurately levelled, allowing one division of mercury for each
six and a half divisions occupied by the acid solution in the measuring
CHILI SALTPETRE 317
tube, and the gas volume read off. To test whether the gas is actually
under atmospheric pressure, a few drops of sulphuric acid are poured
into the cup, and allowed to flow into the measuring tube by cautiously
opening the tap as described on p. 135.
The temperature and barometric height are read at the same time
and the gas volume reduced to O’C. and 760 mm. pressure by means of
the tables. Each I c.c. NO corresponds to o-Oo37986 g. NaNO3 ; the
total volume a divided by the weight a taken and multiplied by IOO
gives the percentage content, which is thus-9379**
The nitrometer must, of course, be accurately graduated, and the
operator should satisfy himself that the readings are correct.
By this method it is easy to obtain results agreeing to within O.2 per
cent. Still greater accuracy may be attained with practice, certainly
to at least O. I per cent, by using a separate decomposing vessel and
transferring the gas to the measuring tube as described on p. 143.
The manipulation is much cleaner when working in this way; also
several decomposing vessels may be operated at the same time and all
the gas volumes read off in a single measuring tube.
If a gas volumeter be employed, the reduction tube must, of course,
be set for dry gas, or if set for moist gas, the operation must be carried
out as described on p. 141, most conveniently by sucking a tiny drop of
water into the gas measuring tube.
Baskerville and Miller" have stated that mercury is attacked at
ordinary temperatures by sulphuric acid of sp. gr. I-84 with the formation
of sulphur dioxide, and that this may give rise to errors in nitrometric
work. Pitman * has shown that this objection is groundless, and this
has since been admitted by the above authors; * they now state that
such action only occurs with acid of 99 per cent. strength and not with
acid of 94 to 95 per cent, such as is generally used for laboratory work.
The view that the solubility of nitric oxide in the resulting solutions
may lead to appreciable error has been disproved by Lunge, Nernst
and Jellinek,” Tower,” and by Newfield and Marx.7 -
The Schlösing-Grandeau-Wagner Method.
This method depends upon the reduction of nitric acid to nitric
oxide, which is collected and measured.
The apparatus required for the determination consists of a flask of
250 to 300 c.c. capacity, closed by a double-bored rubber stopper and
furnished with a tap funnel of 15 c.c. capacity. The lower end of the
* Chem. Centr., 1898, I., 85. * /bid., 1898, I., 709. * /bid., 1898, II., 89.
* Ber., 1885, 18, 1391; 1886, 19, III. Chem. News, 1886, 53, 289.
* Z. anorg. Chem, 49, 219 and 236. * /öid., 50, 382.
7 J. Amer. Chem. Soc., 1906, 28, 877.
3.18 MANUFACTURE OF NITRIC ACID
funnel-tube, which is fused together so as to contract the outlet, reaches
to the middle of the flask, but not to the solution. A tube, suitably
bent to dip into a glass dish containing well-boiled water, passes
through the second opening in the stopper and serves as the exit-tube
for the liberated gas. The measuring tube, which is preferably pro-
vided with a glass tap at its upper end and graduated from the top
downwards in Po C.C., is supported above the dish in a clamp. Forty
c.c. of ferrous chloride solution, containing 400 g. per litre, and an equal
volume of well-boiled IO per cent. hydrochloric acid, are introduced
into the flask and the contained air expelled from the apparatus by
continued boiling, care being taken that the funnel-tube always contains
hydrochloric acid. When all air has been expelled the gas exit-tube
is placed in position below the measuring tube and IO c.c. of a normal
saltpetre solution, containing exactly 33 g. pure Sodium nitrate per litre,
is introduced into the funnel-tube. The tap is then turned so as to
allow this solution to flow drop by drop into the boiling ferrous chloride
solution. When all but a small volume has been introduced the funnel
is washed out twice with IO per cent. hydrochloric acid, which, like the
nitrate solution, is added drop by drop to the boiling solution in
the flask. The operation is finished when the evolution of nitric oxide
ceases. About 90 c.c. of nitric oxide should result under the foregoing
conditions. The measuring tube is removed and, supported by a small
dish containing water, cautiously transferred to a correspondingly tall
glass cylinder filled with water. A second measuring tube is then
brought into position and, without interrupting the boiling, a measured
volume of a solution of the nitre to be tested is introduced and treated
as above described, and so on with a third, fourth, and further tests as
required. When the several tubes have attained the temperature of
the cooling water the volume of nitric oxide in each is read off, after
raising the tube so as to bring the water inside and outside to the same
level. The content of each solution is obtained by comparing the gas
volumes evolved from the sample and from the standard solution re-
spectively. When carried out in this manner the method is very simple,
and may be recommended; it requires, however, careful manipulation.
According to Liechti and Ritter," who have thoroughly investigated
Schlösing's method, good results are obtained even in the presence
of organic matter, for which purpose it was originally worked out; this
has, however, been disputed by R. Pfeiffer.”
Method by Ignition with Potassium Chromate or with Silica.
To the dried O.8000 g. of substance used for the moisture determina-
tion (p. 3O8) are added approximately 3 g. of a previously fused and,
after cooling, finely ground mixture of equal parts of bichromate and
* Z, anal. Chem., Igo3, 42, I. * Ibid., 612.
CHILL SALTPETRE 319
neutral chromate of potassium, and the whole heated, at first gradually
and finally more strongly till the mass melts uniformly. The loss in
weight=N,Os. This method can only be used if the sample is quite
free from alkaline carbonates. An alternative method is to ignite a
mixture of 2 g of the salt with about seven times its weight of ignited
silica free from carbonates, for from two to four hours at a good red
heat.
Pauli 1 recommends this method, but Alberti and Hempel” together
with most other authorities do not consider it sufficiently trustworthy.
The method cannot be employed in the presence of perchlorate.
The conclusions drawn by Alberti and Hempel, and which have
been amply corroborated, are:—
I. The indirect method of examining Chili saltpetre (“refraction ”
determination), at present very generally employed, should be abolished.
The results obtained are extremely inaccurate, and the method cannot
be justified from the chemical standpoint.
2. The direct methods of determining the nitrogen in Chili salt-
petre by means of Lunge's nitrometer, by the process of Schlösing
and Grandeau as improved by Wagner, and by Ulsch's method, give
accurate results, and the introduction of one of these methods is to be
advocated.
3. In the case of Chili saltpetre intended for agricultural use, the
results should be expressed on the basis of the nitrogen content only ;
calculation to sodium nitrate is incorrect, and should only be made
exceptionally and under reserve.
4. In saltpetre intended for industrial use, the impurities should be
estimated in addition to the nitrogen, and an exact analysis, taking
account of the potassium compounds present, is necessary.
ESTIMATION OF PERCHLORATE
Sodium perchlorate is directly injurious in many of the applications
of saltpetre. It occurs practically in all shipments of Chili nitre, and its
presence must always be taken into account. A method for its qualitative
detection has been described on p. 307.
All methods for the estimation of the perchlorate are based on a
determination of the chlorine present as chloride in the sample, the
conversion of perchlorate to chloride in a second portion of the material,
followed by a determination of the total chlorine then present as chloride;
the difference between the two tests represents the perchlorate.
The conversion of the perchlorate to chloride may be effected by
simply heating for a sufficient length of time, whereby the oxygen of
any perchlorate present is driven off.
* /. Soc. Chem. Ind., 1897, 16, 494. * Z. angew. Chem., 1892, 5, Io3.
320 MANUFACTURE OF NITRIC ACID
Sjollema' and Freytag” have given detailed instructions for carrying
out the estimation on this principle.
According to Selckmann,” the method is both slow and uncertain,
and he recommends fusing 5 to Io g. of the nitre, the chloride content of
which has been determined, with three to four times its weight of lead
shavings in a porcelain crucible of 40 to 50 c.c. capacity, gradually
increasing the heat, and stirring with a copper wire; finally, after ten
to fifteen minutes heating, raising the temperature until the bottom
of the crucible is at a dark red heat, at which it is maintained for one to
two minutes. The melt is then soaked in hot water, the lead chloride
decomposed by warming with sodium bicarbonate or hydroxide solution,
the solution filtered, and the chlorine estimated gravimetrically as silver
chloride.
Hönig 4 finds that good results may be obtained with finely divided
iron (ferrum limatum). Two to 3 g of the iron are added to 5 to Io
g. of the fused nitre, in a nickel crucible, and the mixture heated for
half an hour without bringing the temperature of the crucible to
appreciable redness.
Erck" proposes to first decompose the chloride by boiling with
nitric acid of sp. gr. I-4, and alcohol, and then to convert the perchlorate
into chloride by ignition.
Winteler" has pointed out various sources of error in the earlier
methods. He finds that fuming nitric acid will quantitatively reduce
perchlorate to chloride at 200° in a sealed tube. Any chloric acid
originally present must, however, be removed by evaporation with
hydrochloric acid.
O. Förster" recommends the following method as universally appli-
cable. Ten g. of the nitre, the chloride content of which is known, are
mixed with an equal quantity of sodium carbonate, free from chlorine,
and heated in a covered platinum dish, or in a roomy porcelain crucible
over the full flame until the melt becomes perfectly fluid and only small
bubbles are given off; this should be effected in ten minutes. The
product is then dissolved in nitric acid, and the chloride estimated in
the usual manner.
Blattner and Brasseur” first estimate the chlorine present as chloride,
and then heat a mixture of 5 to IO g. of the nitre and 8 to 15 g. of calcium
hydroxide, free from chlorine, over a Bunsen burner for fifteen minutes,
allow to cool, dissolve in nitric acid, and estimate the total chlorine by
any suitable method. Calcium hydroxide is to be preferred to an alkaline
carbonate or hydroxide, on account of the greater ease in manipulation.
* Chem. Zeit., 1896, 20, IOO2. * Ibid., 1897, 21, 21.
* Chem. Centr., 1898, I., I2O3; Z angew. Chem., 1898, II, Io21. ° Ibid., 75.
* Z. angew. Chem., 1898, II, IoI. 7 Ibid., 1898, 22, 357.
* Chem. Zeit., 1903, 27, 32. * Ibid., 589.
CHILI SALTPETRE 321
Various authorities recommend an addition of native peroxide of
manganese (pyrolusite) for the better decomposition of the perchlorate.
This was first proposed by Hellich.' Ahrens and Hettº heat the nitre
with sodium carbonate and chlorine-free manganese dioxide, to facilitate
the decomposition of the perchlorate, to fusion and redness, treat the
solution of the melt with nitric acid and potassium permanganate to
permanent coloration, and estimate the chloride by Volhard's method.
C. Gilbert, in a brochure entitled Methoden gur Bestimmung des
Perchlorats (Tubingen, 1899), gives the following directions for carrying
out the pyrolusite method:—Twenty-five g. of nitre are dissolved in
water, made up to 250 c.c., and the chloride determined by titrating 50
c.c. of the filtered solution with silver nitrate (I c.c. = O.OI g. NaCl) and
potassium chromate (see p. 123). A second 25 g. Of the nitre, preferably
after the addition of 2.5 g. of the purest powdered pyrolusite (Merck's),
is heated for half an hour at 540° in an air-bath in a nickel crucible
fitted with a deep concave lid, and having a capacity of 70 c.c. The
aqueous solution of the melt is then made up to 250 c.c., filtered, and the
chlorine again determined as above. The air-bath recommended is that
designed by Lothar Meyer; it will hold from three to six crucibles, and
should be fitted with a gas regulator and a Le Chatelier pyrometer or
thermometer reading up to 570° (p. 174). Dupré” proceeds in essentially
the same manner, omitting, however, the pyrolusite, but heating 20 g.
of the sample for one hour at 545° in an air-bath of the type recom-
mended by Gilbert. Instead of a pyrometer, a covered nickel crucible
may be used in which is placed a platinum crucible, resting on a layer
of asbestos paper, and containing about I g. of potassium perchlorate.
By observing, after the completion of the heating, whether the residue
in the platinum crucible fuses without any appreciable evolution of gas,
a gauge of the temperature employed is obtained. The chloride may
also be determined by Volhard's method (p. 123), or gravimetrically.
Should iodate be present, Ahrens and H. Gilbert effect the fusion
with addition of sodium carbonate. Twenty g. of the dried nitrate are
thoroughly moistened with 2 to 3 c.c. of a concentrated solution
of sodium carbonate, I g. of pyrolusite added, the mixture dried and
maintained in a state of fusion at a low red heat for fifteen minutes.
The solution of the melt is then oxidised with potassium permanganate
solution, and the chlorine determined by Volhard's method.
Recently, sodium chlorate as well as perchlorate has been detected
in Chili saltpetre, and methods for its estimation have been devised.
In the examination of IO7 samples of nitre, Mărcker * found the per-
centage of perchlorate present to vary from O-27 to 5-64, with a mean
value of O.94 per cent, and in addition O. I to I.O per cent. of chlorate, as
* Chem. Zeit., 1894, 18, 485. * Chem. Centr., 1898, II, 558; Z. angew. Chem., 1898, II, Io20.
* / Soc. Chem. Ind., 1902, 21, 825. * Chem. Centr., 1898, II., 925.
- X
322 MANUFACTURE OF NITRIC ACID
determined by the method then in vogue. The presence of perchlorate
is regarded as objectionable in nitre intended for agricultural purposes,
and although the practical investigation into the extent of the injury it
may cause has not yet been concluded, it has been suggested that the
maximum quantity to be allowed should not exceed I? per cent.
Mennicke' estimates a mixture of chloride, chlorate, and perchlorate
in the following manner. (a) The nitre is ignited with addition of alkali
hydroxide or carbonate to convert all the chlorine compounds to chloride,
which is determined. (b) Chloride and chlorate are estimated by gently
boiling 5 g. of the nitre with IO g. of zinc dust (free from chlorine), and
I 50 c.c. of I per cent. acetic acid for half an hour, filtering and determin-
ing the chloride. (c) The chloride originally present in the nitre is
determined directly. This method has shown the presence of appreci-
able quantities of chlorate.
Blattner and Brasseur” treat a solution of 5 to IO g. of the nitrate
with excess of sulphurous acid, either as gas or in solution, thereby
reducing only the chlorate and not the perchlorate, drive off the excess
of sulphur dioxide by boiling, and saturate the warm solution with
calcium carbonate. The chloride is estimated in the cold, filtered
solution in the ordinary manner, and the chlorine present in the sample
as chlorate is obtained by deducting from this value that of the chloride
originally present in the nitre. The total chlorine is determined by
igniting the saltpetre with calcium hydroxide as described above, and
the perchlorate arrived at by difference.
Arnould “ has described the following method, which is employed in
the French Government laboratory. The chloride originally present is
precipitated by the addition of neutral silver nitrate solution and filtered
off; the chlorate in the filtrate is reduced by warming this to 90°, and
adding an excess of lead nitrite, prepared by simply shaking lead nitrite
with water and without filtering; any cloudiness due to the nitrite
is removed by the addition of a few drops of dilute nitric acid. If any
opalescence persists after this treatment, chlorate is present, and the
amount is estimated by comparison with similarly prepared solutions
to which known quantities of chlorate have been added. For the
French powder factories it is specified that the nitre must contain less
than O-OI per cent, chloride, O.OI per cent. chlorate, and O. I per cent.
perchlorate. -
Lemaitre * reduces the perchlorate by fusion with sodium sulphite;
this method has been improved by Fschernobojew,” who has shown
that chlorate can be determined simultaneously.
* Chem. Zeit. Rep., 1898, 22, II7. * Monit. Scient., 1904, 18, 253; Chem. Centr.
* /bid., 1900, 24, 793. I904, I., I427.
* Mémorial des poudres et saloétres, 1902. * Chem. Zeit, 1906, 29, 442.
CHILI SALTPETRE 323
CONTROL OF WORKING CON DITIONS
The process control in the manufacture of nitric acid consists essen-
tially in the determination of the yield and of the quality of the nitric
acid. In addition, care must be taken that no nitrogen oxides escape
into the chimney or into the atmosphere; the tests applied are similar
to those employed in the sulphuric acid industry (cf. p. 336).
The bisulphate drawn from the retort must also be examined ; this
is done in the following manner – -
I. Free acid is estimated by titration with normal sodium hydroxide
solution. In the presence of considerable amounts of iron Oxide or of
alumina no indicator is added, and the titration is considered finished as
soon as the first signs of a permanent flocculent precipitate appear.
2. Nitric acid may be determined in the nitrometer or gas volumeter
as in the nitrometric method for the estimation of saltpetre. The acid
sulphate is dissolved in the minimum quantity of water in the cup above
the tap, and decomposed with a large excess of sulphuric acid (pp. I37
and 139). Since the quantity of nitric acid present in acid sulphate is
always small, a nitrometer either with a narrow measuring tube as used
in the examination of acids, or with a central bulb (Fig. 46, p. 136),
should be employed.
NITRIC ACID
Pure, absolute nitric acid, HNO, is very difficult to prepare, and
scarcely keeps any length of time, owing to the liberation of oxygen
with formation of hyponitric acid, which latter causes the colourless acid
to turn yellow or, with larger quantities of nitrogen peroxide, red. The
C
boiling point is 86°, the specific gravity at * slightly above I-52.
The hydrates of nitric acid have been especially studied by Küster
and Kremann." -
The strongest commercial acid has, when pure and almost free from
nitrogen peroxide, a sp. gr of 1.50 or slightly higher, corresponding to
94 to 95 per cent. HNO3. It begins to boil a little above 86°, the boiling
point rising as distillation proceeds, owing to more acid than water
distilling over, until 126° is reached, at which temperature an acid con-
taining 68.9 per cent. HNO3, and having a sp. gr. 1-42 distils over
unchanged. Weaker acids have again a lower boiling point, and yield
on distillation, with a continual rise in temperature, more water than
acid until the boiling point 126° is attained, when acid of the above
composition distils over. º
Lunge and Rey’s” investigations into the relationship between the
* Z, anorg, Chem., 1904, 40, I. * Z, angew. Chem., 1891, 4, 165.
324 MANUFACTURE OF NITRIC ACID
specific gravity and percentage content of nitric acid solutions have led
to the older tables, such as that of Kolb, being superseded. The
following table is based on their researches. The specific gravity
determinations made by Winteler" have been shown to be unre-
liable by Lunge” by Veley and Manley,” and by Pützer." Values
agreeing with those of Lunge and Rey within, at the outside sºw (gºw
in one case only), have been obtained by Veley and Manley,” who add
that greater concordance is scarcely to be expected between different
observers employing different methods.
An additional table (p. 327) for correcting the specific gravity
obtained at higher or lower temperatures to the normal value, viz. the
specific gravity of the acid at I 5°, compared with water at 4° as the unit,"
is also given. -
This table has been worked out for pure acids, and the values do
not apply absolutely to observations made on commercial acids, which
are never pure. In the cases of sulphuric and hydrochloric acids, except
for the highest strengths, such differences are not great; they may,
however, be considerable in the case of commercial nitric acid, owing
to the presence of lower oxides of nitrogen, generally calculated to
hyponitric acid. The influence of hyponitric acid on the specific gravity
has been observed by many investigators," but no method has been
Specific Gravity of Initial Acid I-496O at I 5"/4° (in vacuo).

N2O4 Alteration of the N2O4 Alteration of the N2O4 Alteration of the
per cent. Spº per cent. Sp º per cent. sºlº
0 °25 0 00050 4°50 0.02875 8 75 0-05825
0-50 0-00075 4-75 0-03050 9-00 0 °06000
0-75 0-00150 5°00 0-03225 9°25 0-06160
1-00 0-00300 5°25 0 °03365 9°50 0 °06325
1°25 0-00475 5°50 0 °03600 9-75 0 °06500
1°50 0-00675 5°75 0.03775 10:00 0°06600
1°75 0'007.75 6:00 0-03950 10°25 0-06815
2°00 0 °01050 6°25 0-041.75 10 °50 0-06975
2-25 0 01250 6'50 0 °04300 10-75 0-07135
2-50 0 °01425 6'75 0-04475 11 *00 0 07300
2-75 0-01625 7:00 0-04650 11-25 0-07450
3-00 0-01800 7.25 0-04720 11-50 0-07600
3-25 0 0 1985 7:50 0-05000 11-75 0-07750
3-50 0-02165 7.75 0-051.65 12°00 0.07850
3-75 0-02350 8:00 0-05325 12-25 0°08050
4°00 0 °02525 8'25 0-05500 12°50 0 °08200
4 °25 0.02690 8°50 0-05660 12-75 0°08350
* Chem. Zeit., 1905, 29, 689 and IOO9. * Ibid., 1905, 29, 689 and Ioy2. '
* Ibid., 1905, 29, 1270. * Ibid., 1905, 29, 1221. * / Soc. Chem. Ind., Igo3, 22, 1228.
* Cf. Fuchs, who has published a detailed table, Z, angew. Chem., 1898, II, 747.
7 Loring Jackson and Wing, Chem. Zeit. Rep., 1887, II, 293; and, R. Hirsch, Chem. Zeit,
I888, I2, 91.I.
NITRIC ACID
325
Table of the Specific Gravity of Nitric Acid of various strengths
at 15° C. referred to Water at 4° C.

§: 100 Parts by weight contain 1 Litre contains kg.
l’8,V. e –
* | ##| #3
15 | #5 | #: Acid Acid | Acid Acid Acid | Acid
4. F# | P = | N2O5. HNO3. of -of Of N2O5. HNO3. of of Of
(in E-4 36° B. 40° B. 48$” B. 36° B. 40° B. |48}” B.
vacuo).
1 °000 || 0 0 0-08 || 0-10 || 0-19 0°16 0 "10 || 0-001 || 0-001 || 0-002 || 0-002 || 0-001
1-005 || 0-7 1 0-85 || 1:00 1-89 1 *61 1*03 || 0-008 || 0-010 || 0-019 || 0-016 || 0-010
1-010 || 1 °4 2 1-62 1-90 3-60 3-07 1 ‘95 || 0-016 || 0-019 || 0-036 || 0-031 || 0-019
1 °015 || 2:1 3 2°39 || 2-80 || 5:30 4 °52 2'87 || 0-024 || 0:028 || 0-053 || 0-045 || 0-029
1-020 || 2:7 4 3-17 | 3.70 || 7-01 5°98 3-79 || 0-033 || 0-038 || 0-072 || 0-061 || 0-039
.1-025 || 3-4 5 3-94 || 4 -60 8-71 7-43 4-72 || 0-040 || 0-047 || 0-089 || 0-076 || 0-048
1 *030 || 4°1 6 4*71 || 5°50 | 10:42 8-88 5-64 || 0-049 || 0-057 || 0-108 || 0-092 || 0-058
1.035 || 4-7 7 5 °47 || 6′38 || 12:08 || 10:30 6°54 || 0-057 || 0-066 | 0°125 || 0-107 || 0-068
1-040 || 5-4 8 6-22 || 7-26 || 13-75 | 11-72 7'45 || 0-064 || 0-075 || 0-142 || 0°121 || 0-077
1*045 || 6-0 9 6'97 || 8'13 || 15-40 || 13-13 8 °34 || 0-073 || 0-085 || 0-161 || 0-137 || 0-087
1*050 || 6-7 10 7-71 8-99 || 17-03 || 14-52 9°22 || 0-081 || 0-094 || 0°178 || 0°152 || 0-096
1*055 || 7-4 11 8-13 9'84 | 18-64 15-89 || 10-09 || 0-089 || 0-104 || 0-197 || 0-168 || 0-107
1*060 || 8-0 12 9 "15 10°68 20°23 || 17-25 || 10-95 || 0-097 || 0-113 || 0:214 || 0-182 || 0-116
1.065 || 8-7 13 9-87 || 11 °51 21 '80 18'59 || 11 °81 || 0-105 || 0-123 O-233 || 0-198 || 0-126
1-070 || 9°4 14 || 10:57 | 12:33 23:35 | 19.91 | 12-65 || 0-113 || 0-132 || 0-250 || 0-213 || 0-135
1'075 || 10-0 15 || 11 °27 | 13 "15 24-91 || 21 24 || 13°49 || 0°121 || 0-141 || 0°267 || 0°228 || 0-145
1-080 || 10:6 16 || 11-96 || 13°95 26°42 22°53 || 14-31 || 0°129 || 0-151 0.286 || 0°244 || 0-155
I •085 || 11°2 17 | 12-64 || 14-74 27-92 || 23-80 || 15-12 || 0-137 || 0-160 || 0°303 || 0°258 || 0-164
1-090 || 11-9 18 || 13-31 | 15 °53 || 29°41 || 25-08 || 15'93 || 0-145 || 0-169 || 0-320 | 0°273 || 0-173
1°095 || 12-4 19 || 13-99 || 16-32 || 30-91 || 26°35 | 16-74 || 0-153 || 0-179 || 0-339 || 0-289 || 0-184
1 100 || 13-0 20 || 14-67 || 17-11 || 32°41 27-63 || 17:55 || 0-161 || 0-188 || 0-356 || 0-304 || 0-193
1°105 || 13-6 21 || 15-34 || 17-89 || 33-89 || 28-89 || 18-35 || 0-170 || 0-198 || 0-375 0-320 | 0:203
1*110 || 14 °2 22 || 16-00 | 18-67 || 35-36 || 30-15 | 19:15 || 0-177 || 0.207 || 0-392 || 0:335 | 0-212
1 115 || 14-9 23 || 16-67 | 19°45 || 36-84 || 31°41 | 19.95 || 0-186 || 0:217 | 0°411 || 0-350 || 0-223
1*120 || 15 °4 24 || 17-34 20°23 38-31 || 32-67 || 20-75 || 0-195 || 0-227 || 0-430 || 0°366 O-233
1-125 | 16°0 25 | 18-00 || 21-00 || 39°77 || 33-91 21-54 || 0-202 || 0-236 || 0:447 || 0-381 0-242
1 - 130 | 16 °5 26 || 18-66 21 '77 || 41 °23 35-16 || 22-33 || 0.211 || 0-246 || 0:466 || 0-397 || 0-252
1-135 | 17-1 27 | 19:32 22°54 || 42’69 || 36-40 || 23-12 || 0-219 || 0°256 || 0°485 || 0°413 || 0°263
I •140 || 17-7 28 || 1998 || 23°31 || 44°15 || 37-65 || 23-91 || 0-228 || 0°266 || 0'504 || 0°430 || 0°273
1°145 || 18°3 29 || 20-64 24-08 || 45-61 || 38-89 || 24.70 || 0-237 || 0:276 || 0°523 || 0°446 || 0°283
I-150 | 18-8 30 || 21°29 || 24°84 || 47-05 || 40°12 || 25-48 || 0:245 || 0°286 || 0°542 || 0°462 || 0°293
1-155 19°3 31 || 21-94 25-60 || 48-49 || 41-35 | 26*26 || 0-254 || 0°296 || 0-561 || 0:478 0-304
1-160 || 19°8 32 || 22.60 |26-36 || 49'92 || 42.57 27-04 || 0-262 | 0.306 || 0-580 || 0:494 || 0-314
1°165 20 °3 33 || 23-25 | 27'12 51°36 || 43-80 || 27-82 || 0-271 || 0-316 || 0°598 || 0-510 || 0-324
1-170 || 20°9 34 || 23.90 27°88 52°80 || 45.03 || 28'59 || 0 °279 || 0-326 || 0-617 | 0°526 || 0-334
1 °175 || 21 “4 35 | 24°54 28°63 || 54°22 || 46°24 29°36 || 0-288 0°336 || 0°636 || 0°543 || 0°345
I-180 22°0 36 || 25°18 29-38 55'64 || 47°45 || 30-13 || 0-297 || 0-347 || 0-657 || 0-560 || 0-356
1*185 22:5 37 H 25.83 || 30-13 || 57-07 || 48-66 || 30-90 || 0-306 || 0-357 || 0-676 || 0 °577 || 0-366
1°190 || 23-0 38 26-47 || 30-88 || 58°49 || 49-87 || 31.67 || 0.315 || 0°367 || 0-695 || 0°593 || 0-376
1-195 || 23°5 39 || 27-10 || 31-62 59'89 || 51:07 || 32°43 || 0-324 || 0°378 || 0-715 || 0-610 || 0-388
1*200 24°0 40 || 27-74 || 32-36 61°29 || 52.26 33-19 || 0-333 0-388 || 0-735 || 0-627 | 0-398
1°205 || 24°5 41 || 28°36 || 33-09 || 62’67 || 53°23 || 33-94 || 0-342 || 0-399 || 0°755 || 0°644 || 0-409
1°210 || 25-0 42 28-99 || 33-82 64-05 || 54-21 | 34-69 || 0-351 || 0°409 || 0°775 || 0:661 || 0°419
1.215 || 25-5 43 || 29-61 34-55 65.44 55-18 35-44 || 0-360 || 0:420 || 0-795 || 0-678 || 0:431
1 220 || 26-0 44 || 30°24 || 35-28 || 66'82 56°16 || 36-18 || 0°369 || 0°430 || 0°815 || 0-695 || 0°441
1*225 || 26°4 45 || 30-88 || 36°03 | 68°24 || 57-64 || 36-95 || 0-378 || 0-441 || 0°835 | 0-712 || 0°452
1°230 || 26-9 46 || 31°53 || 36-78 || 69'66 59-13 || 37.72 || 0-387 0°452 || 0-856 || 0 °720 || 0-466
1°235 || 27°4 47 || 32-17 || 37-53 || 71-08 || 60-61 || 38-49 || 0-397 || 0-463 || 0-877 || 0-748 || 0-475
1-240 || 27.9 48 || 32°82 38-29 || 72°52 61-84 || 39-27 || 0:407 || 0:475 0-900 || 0°767 || 0-487
1*245 28°4 49 || 33-47 || 39-05 || 73-96 || 63-07 | 40-05 || 0°417 | 0°486 || 0°921 || 0-785 || 0-498
1°250 || 28°8 50 || 34°13 || 39°82 75°42 || 64-31 | 40°84 || 0°427 | 0°498 || 0°943 || 0:804 || 0-511
969
CIIC)V OIRIJIN. SIO ºſłIſ),LOWSHI) NVN
177. I 813.3 || 319.3 II?. I | Olć. I | 09.96 96. Ig|I 6I.8/I 60.76 g0.08 I. Si, 009. I
W07. I II3.3 869.3 698. I #1. I. I g6.86 86. If I | 85.31. I 09. I6 Zg.8/. 8. 17 | 967. I
698. I 99 I-3 || 839. & 983. I Wiſ I. I | 06. I6 || 0/.. pp. I | 69.69 I | 09.68 || 08.9/. f. If I 0.6%. I
988. I | 80I. & 997. & 303. I | 9 II. I 96.68 | 89. If I | 60.99 I | 0/./8 || 8 [.9/, I. Zij || 987. I
108. I | 890.3 || 8 IW. & 713. I 360. I | 93.88 | 16.88I | 16.39L | 90.98 || 91.8/. 8.9% 087. I
816. I &IO. & | 098.6 |973. I | 890. I 39.98 || 63.93I 76.69|| || 95.78 || 63.3/ j.97 g/?. I
096. I 696. I 608. & 6 IZ. I g;0. I 30.98 || 88.93 I | 00. 191 || 06.38 90. I/. I.97 || 0/f. I
fö6. I | 136. I | 69%. 3 || 36|I. I 830. I | [9.88 || 65. Ig|I|| 03.jg|I| Zſ. I8 || 61.69 8.9f Q97. I
86 I. I 988. I &I3-3 | 89 I. I IOO. I 80.38 || LI.6%I | 1.f. Ig|I 86.6/ | 99.89 j. Gº || 097. I
81 I. I | 878. I | 1.9L. & 77I. I | [86.0 || 39.08 || 76.9%I 98.8%I | 09.8/ | 88.19 I. g7 || 99%. I
09 I. I | 0I8. I | 8&I. 3 || I&I. I | IQ6.0 || 93.6/ | I3. #8, I 98.9%I | 8%. 11 #3.99 8. Wiſ I 097. I
9&I. I | 811. I | 080. & 860. I | IW6.0 || 86.11 | I]. ZZI | 06.8%I 86.9/ | 8 [.99 #. ff. I giff. I
80I.. I 981. I | 180.3 |910. I | I36.0 || 69.9/ | IQ.0%I | 77. If I | 89.7/ | IO.j9 I. Fif I Off. I
080. I IO/. I | 966. I | 890. I 306.0 || 13.9/, Zg.8II | 66.83T | 68.8/ | I6.29 8.3% 937. I
890. I | 199. I | 996. I | 380. I | 988.0 || 30.7/, gg.9II | 89.99 I | 1.I.Z. 98. I9 f.3i Ogiń. I
180. I | 889. I g|I6. I | II0. I | 198.0 || 08.3/, 39.5II 8ſ. fg|I|| 86.01 || 78.09 I. 37 || Q&#. I
9 I0. I | 009. I | 1.18. I I66.0 | 678.0 || 69. IZ 3/.3.II 6I.33T || 08.69 || 38.69 /... Zī 0&f. I
966-0 | 899. I | 688. I | II.6.0 Z88.0 | 68.01 | #8.0II | 86.6%I | 89.89 || 38.89 3. Ziº || 9 If. I
916-0 | 189. I 80S. I | 396.0 |9|I8.0 || 83.69 I0.60I | 78.1%I | 09. A.9 98.19 0.37 || 0 Tf. I
196-0 || 109. I 191. I | 886.0 || 008.0 || 0 I-89 | #3. 10I g/.go, I | 07.99 || 36.99 9. If I GOf... I
186-0 || 917. I L81. I | WI6.0 | 88.1.0 16.99 || 97.90I | 19.8%I | 03.99 || 16.99 3. If || 007. I
6 I6.0 | 177. I | 169. I 968.0 | 891.0 || 06.99 || 91.30L | 89. IZI 93.79 || 10.99 8.07 || 968. I
606-0 || 037. I | 999. I 6/8.0 39.1-0 || 98.79 || ZI.30I g/.6II | 8&.89 || 0%. 79 G.07 || 063. I
£88.0 | 668. I | 889. I | 398.0 | 681.0 || 78.89 || Ig.00I 88./II | 73.39 gg.89 I. Of 988. I
618-0 | 888. I 339. I | 19.8.0 g8/-0 || Ig.89 00.00I 13. II | 36. I9 80.89 0.07 ||3888. I
898-0 | 998. I 809. I 978-0 gö/- 0 || 78.39 g6.86 70.9 II | 1.3. I9 Zg. &g 8.69 || 088. I
098.0 | 688. I | 819. I | 668.0 | III. 0 || 98. I9 || 83. 16 || 0%. VII | 08.09 || 69. Ig f.68 || 9/8. I
988. 0 || 7Ig. I | 8f 9. I | WI8.0 | 869. 0 || I6.09 || I6.96 || 87.%II | 63.69 || I6.09 0.68 || 0/8. I
8I8-0 | 68%. I 819. I 861-0 | #89.0 || 86.69 ##.76 g/.0II | 87.89 gl.09 9.88 || 998. I
808.0 993. I 887. I 881.0 I/9.0 || 90.69 | 16.36 30.601 | 19. 19 g8.6% 3.89 || 093. I
881-0 || 0%. I gg?. I | 891.0 | 899.0 || II.89 Ig. I6 | Ig. 101 | 99.99 || 19.8% 8. A8 || 998. I
614-0 || 9 I&. I | 137. I | 891.0 979. 0 || 33./9 || OI.06 || 19.90I 6/.99 || 38. Af jº. A8 || 098. I
891.0 861. I 007. I | 681.0 889.0 || 78.99 || I/.88 #0.j0I 86.79 |80. If 0.18 || 958. I
iſ 71.0 | III. I | 818. I | 93.1.0 IZ9.0 || 97.99 || Zg. 18 I#.30L | 10.79 gg.9% 9.99 || 073. I
861-0 || 87ſ. I 978. I | 0IA.0 609.0 | 89.79 g6.98 || 08.00I 33.89 || 39.97 3.98 || 988. I
361.0 | 181. I 888. I | 701-0 || 809.0 | g I. 79 || 13.98 || 00.00I | 08.39 9%. g7 0.98 (9:388. I
9 Il-0 || 9%I. I 038. I 169.0 | 169.0 || IA-39 89.f3 6I.66 | 18.39 || 68. ff. 8.98 || 08:8. I
I01-0 || 30L. I £63. I | 889-0 || 989.0 || 98.39 || 33.88 || 09. 16 | 89. Ig II. Wy j.93 + 938. I
989.0 | 080. I | 893. I | 699. 0 || 8/9. 0 || IO.39 || 06. I8 g0.96 || IA.09 || Zj.g? 0.98 || 038. I
8/9.0 690. I | 87%. I 999.0 399.0 || || I. Ig | Zg.08 || 67.76 | 68.67 91.3; 9.j9 || 9 Ig. I
699.0 830. I 813. I 879.0 Igg. 0 || 38.09 gå.6/ | 76.36 || 10.6% 90.37 3. fºg 0Ig. I
979.0 | | I0. I | 86 I. I | 089.0 079.0 || Og. 67 | #6. // | 0%. I6 93.8f 18. I? /.38 Q08. I
889.6 | 166.0 | 691. I | LI9.0 | 639.0 || II.87 || 0/.91 || 76.68 || 67.17 | I/.0% 9.88 || 003. I
I39.0 | 116.0 | 97 I. I 909-0 | 619.0 || 36. If gº.g/ | 87.88 || 3/.9% g0.0% 8.33 || 963. I
809.0 | 1.96.0 8%I. I 869-0 | 80%. 0 || 3I./7 IZ. PA 30, 18 96.97 | 68.68 jº. 33 0.6%. I
969.0 886.0 | 00I.. I I89.0 | 867. 0 || W9.9f 96.3/, /.g.98 || 8 I.gif £1.89 0.69 || 98%. I
$89.0 || 8 I6.0 | Z10. I | 899.0 | 187. 0 || 99.97 &A. I./ | II. W8 If...?? | 10.88 9. [8 || 086. I
0.9.0 | 868.0 | 790. I | 999.0 | 1.1%. 0 || 91.7% | 85.0/, g9.38 || 79.87 | If. 19 I. Ig | 9/3. I
899.0 | 618-0 | [80. I 799.0 1957.0 | 16.87 93.69 || 0%. I8 || 18. Zip 91.98 9.08 || 013. I
179.0 || 098.0 600. I 889-0 | 197.0 || 8I.87 | 66, 19 || W1.6/ | 0I.Zī 60.9% 3.03 || 99%. I
V89.0 I?8.0 | 186.0 | I&g. 0 || 177.0 | Op.37 || 91.99 || 03.8/ | #8. If |ff.gg A.6% || 093. I
339.0 || 338.0 996.0 | 609.0 | 187.0 || 39. If | f g.g.9 || 98.91 | 89.0% 81.f3 9.6% 99%. I
gº * ‘(omopa
“I:SW 8.0% 8.98 |. • (? '91 ºf Sf "QI s(); “I:08 || * H tº wº)
JO J0 JO || “ONEH | “O&N | Jo JO 39 |*ONH | *O'N | g g | #g of
plow proV plow plow prov ppoW §§ : º ºf
- ## #3 qu
= * S*" on "Atº.1%)
"3XI supgquoo olqſſ I uſt 4u.00 quâgo A Aq S4 It'J 00I ‘oodS
'pontlºtt00–342 “prov op.131N Jo A31Ae 10 oupads oug Jo aſqa L
NITRIC ACID
327
Table of the Specific Gravity of Nitric Acid, etc.—Continued.

Spec. 100 Parts by weight contain 1 Litre contains kg.
Grav. ºn 3 * = -- - - -
àº, # = 3.3
!. #3 | # Acid Acid | Acid Acid Acid Acid
4 ää | # | N2O5. HNO3. ...of Of Of N2O5. HNO3. of Of Of
(in 36° B. 40° B. 484° B. 36° B. 40° B. 48}* B.
wacwo).
1°501 81-09 || 94-60 179°16 152-78 97.03 || 1:217 | 1.420 2-689 2-293 || 1 °456
1°502 81°50 95-08 || 180-07 || 153'55 97.52 || 1:224 1-428 2*704 || 2:306 || 1:465
1°503 81-91 95.55 | 180-96 || 154-31 || 98-00 || 1:231 || 1:436 || 2:720 || 2:319 || 1:473
1 °504 ... ... 82°29 96-00 | 181-81 | 155-04 || 98-46 || 1:238 1-444 2-735 || 7-332 1'481
1°505 || 48°4 101 || 82-63 96-39 182°55 155-67 98-86 || 1:244 || 1:451 |2-748 || 2°343 || 1 °488
l"506 ... ... 82-94 96-76 | 183°25 | 156°27 99-27 | 1.249 | 1.457 || 2:759 || 2:353 || 1:494
1 °507 ... 83-26 97-13 | 183'95 | 156-86 99-62 || 1:255 | 1.464 2*773 || 2:364 1°502
1°508 || 48-5 83-58 97-50 184°65 157°47 || 100-00 || 1 260 | 1.470 2°784 || 2:374 || 1:508
1°509 ... ... 83-87 97-84 185-30 | 185-01 || 100-35 | 1.265 1-476 2-795 2°384 || 1:514
1:510 || 48-7 102 || 84-09 || 98*10 | 185 °79 || 158°43 || 100-62 || 1:270 | 1.481 || 2:805 || 2:392 || 1 °519
1°511 . ... ... I 84-28 || 98-32 | 186°21 | 158-79 || 100-84 || 1:274 || 1 °486 || 2:814 2°400 || 1 °524
1°512 84-46 98°53 | 186-61 159°13 101-06 || 1 °277 || 1 °490 2°822 || 2°406 || 1 °528
1°513 84-63 98-73 || 186'98 || 159'45 || 101 °26 || 1:280 || 1 °494 || 2 829 || 2°413 || 1 °532
1“514 ... ... 84-78 98-90 187-30 || 159-72 | 101 °44 || 1:283 || 1:497 ||2-835 | 2°418 1°535
1°515 . 49-0 103 || 84.92 99-07 | 187-63 || 160-00 || 101-61 || 1:287 || 1 -501 || 2-843 2°424 || 1 °539
1°516 ... ... 85-04 || 99-21 | 187°89 160°22 || 101 °75 || 1 289 || 1:504 || 2°848 || 2:429 || 1 °543
1°517 ... 85-15 99 °34 || 188-14 160°43 101 '89 || 1:292 || 1 °507 || 2:854 2°434 || 1 °546
1°518 ... 85-26 99°46 | 188°37 | 160-63 102-01 || 1:294 || 1 °510 |2-860 2°439 || 1'549
1°519 ... ... I 85-35 | 99.57 | 188'58 160-81 102-12 || 1:296 | 1.512 ||2-864 || 2:442 || 1 °551
1-520 || 49-4 || 104 || 85-44 99-67 | 188-77 | 160-97 || 102°23 || 1 299 || 1:515 || 2:869 || 2:447 || 1:554
Table for the Correction of the observed Specific Gravity of Nitric
devised to allow for its effect when present.
Acid for differences of Temperature between 13° and I7° C.

Specific Correction for Specific Correction for
Gravity. tº Gravity. + 1°.
1*000—1°020 + 0-0001 1°281–1 °310 + 0-0010
1 *021–1°040 0.0002 1°311—1 °350 0°0011
1-041—1-070 0°0003 1 °351–1 °365 O'0012
1*071—1°100 0°0004 1°366–1 “400 0°0013
1*101—1°130 0.0005 1°401–1°435 0 °0014
1*131––1 °161 O'0006 1°436–1°490 0 °0015
1*162–1 °200 0-0007 1°491–1°500 0 °0016
1°201—1°245 0'0008 1°501–1 °520 0-0017
1°246–1°280 0.0009
Lunge and Marchlewski!
have shown that Hirsch's view, that I per cent. Of nitrous acid corre-
sponds to an increase of O,OI in the specific gravity, is untenable. Their
observations gave the following results for the strongest commercially
important acid. -
The following example, dealing with nitric acid containing nitrous
acid, will illustrate the use of these tables.
* Z. angew. Chem., 1892, 5, Io.
A sample of nitric
328 MANUFACTURE OF NITRIC ACID
acid was found to contain 2.93 per cent. N2O4, and to have a sp. gr.
of 1.4994 at 20°. From the table on p. 327, which may be used
without hesitation for temperatures lying somewhat below 13° or
above 17°, it will be seen that the specific gravity becomes, at 15°,
1.4994+(o-oor 6× 5) = 1.5074. According to the table on p. 324, O.OI8o
must be deducted from this weight, for the 2.93 per cent. (or in round
figures 3 per cent.) N2O,. This gives, therefore, a sp. gr. for the pure
acid = 1.5oz.4—ool 80 = 1.4894. For this value the round figure I-490
is taken from the table on p. 326, giving 89.6O HNOs as the percentage
of the sample. If instead of this value, which takes account of the
N.O, present, the value (1.5074) obtained by simply correcting for
temperature had been taken, the percentage of nitric acid found would
have been given as 97.1349,39–97. 31 obtained as the mean of 1.5oz
with 97.13 per cent. HNOs and 1,508 with 97.50 per cent. HNO, ;
that is, instead of the correct value 89.60 per cent, the erroneous value
of 97.31 per cent.
In the application of these high strength nitric acids for nitration
purposes it is customary to regard the whole of the N2O, as inactive,
and the table (p. 324) is calculated from this standpoint. It might
in many cases be more accurate to regard half of the nitrogen peroxide
as active, according to the equation —-
N.O., + H2O = HNO3 + HNO,.
In such case a much smaller deduction should be made, and for
this a special table might advantageously be drawn up.
Lunge and Marchlewski have endeavoured to prepare similar
tables to make allowance for the nitrogen peroxide present in nitric
acid solutions of two lower strengths (sp. gr. I-4509 and I-4018),
which are included in the range of commercial acids. They were not
able, however, to obtain concordant results, apparently owing to the
partial or complete conversion of the nitrogen peroxide to nitrous and
nitric acids; such tables are still less possible for weaker acids (cf.
p. 329).
Estimation of Nitrogen peroxide (Hyponitric acid).-This estima-
tion for the correction to be applied to the specific gravity reading of
high strength nitric acid solutions (cf. p. 324) is carried out as follows:
—The acid is allowed to flow slowly and in small quantities at a time,
from an accurately calibrated burette, into a measured volume of semi-
normal potassium permanganate solution (15.82O g. KMnO, per litre),
warmed to 40°, until the colour disappears. The acid should be allowed
to stand for some time in the burette before the titration is made, until
it has attained the temperature of the room as measured by an
accurate thermometer; this is indicated by the volume remaining
NITRIC ACID 329
constant. The burette should be graduated in ºr c.c., and it should
be possible to read with certainty to O.OI c.c. The number of c.c. of
acid required to decolorise the permanganate solution, multiplied by
the sp. gr. corresponding to the room temperature, gives the weight of
acid used, from which the percentage content in NgC), is calculated in
the manner described under the estimation of nitrous vitriol (p. 342).
Each I c.c. M/2 permanganate solution corresponds to O-O23OO g. N2O, ;
thus with a consumption of n c.c. permanganate, and ºn C.C. of the acid
under examination, the content in No-ºº: g. per c.c. The
estimation of the three nitrogen acids in admixture is described in the
section on Sulphuric Acid (p. 34O).
The Total Acidity is generally determined in practice by the
hydrometer only, notwithstanding the great uncertainty inseparable from
the method in this particular case (cf. p. 324). It may, of course, be deter-
mined by titration, and this should be done in all important cases. This
estimation can be carried out without difficulty in the case of slightly
fuming acids, by either pipetting a quantity of the previously diluted
acid or, preferably, by weighing out a portion of the concentrated acid
in the bulb-tap pipette used for fuming sulphuric acid (Fig. I34, p. 390).
In titrating, attention must be paid to the destructive action of nitrous
acid on methyl orange (cf. p. 66). In the case of strongly fuming red
acids, even the method of weighing in the special pipette fails to give
satisfactory results, owing to the impossibility of completely washing
out the nitrogen oxides, especially nitric oxide, which are continu-
ally evolved from the strong acid. On this account the following
directions given by Lunge and Marchlewski should be adopted :—
Io c.c. of the acid are allowed to drop slowly, preferably from a burette,
into ice-cold water, the solution made up to IOO C.C., and an aliquot part
titrated with a very accurately prepared standard sodium hydroxide
solution.
The percentage of total nitrogen acids may also be determined in
the nitrometer, employing the bulb form (Fig. 45, p. 136), so as to
accommodate the larger volume of nitric oxide evolved. The test is
carried out exactly as described under the analysis of saltpetre, the
quantities taken being I c.c. of the acid, measured in an accurate
pipette, and IO c.c. of concentrated sulphuric acid. This method
is, however, only employed in cases in which a large proportion of
sulphuric acid is already present, as, for example, in nitrating and
spent acids.
Other Tests.- Fixed residue. This consists chiefly of sodium
sulphate with small quantities of oxide of iron, etc., and is estimated by
evaporating 50 c.c. of the acid to dryness, in a place protected from dust,
igniting, and weighing.
330 MANUFACTURE OF NITRIC ACID
Krauch” states that IO g. of pure nitric acid should leave, on
evaporation in a porcelain dish, only a very minute and scarcely weigh-
able residue ; on an average he obtained a residue of from 2 to 3 mg.
from 50 g. of pure concentrated acid.
Sulphuric acid (Krauch): IO g. are evaporated to about I c.c. in a
porcelain dish, the residue dissolved in 30 c.c. of water, the solution
transferred to a beaker, heated, and barium chloride added. No sign of
a precipitate should appear in the case of acid, nitric. pur., even after
standing for some time.
In carrying out this test, it is necessary to bear in mind that
the presence of strong nitric acid retards the precipitation ; the
greater part of the nitric acid must either be removed by evaporation
or nearly neutralised by the addition of chemically pure sodium
carbonate.
For the quantitative determination of sulphuric acid the solution
is rendered nearly neutral by the addition of pure sodium carbonate,
and precipitated by barium chloride as described on p. 274. If the
acid leaves an appreciable residue on evaporation, this must be taken
into account, as it consists mainly of sodium sulphate.
Chlorine. (Krauch): 50 c.c. of distilled water to which a few
drops of silver nitrate solution have been added should show no
change on the addition of 5 to IO c.c. of “pure” nitric acid. This
method of carrying out the test removes any uncertainty which may
arise should the distilled water employed not be absolutely free from
chlorine.
For the quantitative estimation, the acid is neutralised with
chemically pure sodium carbonate and titrated with silver nitrate
solution; a very slightly alkaline reaction is not prejudicial.
Heavy metals and alkaline earths (Krauch): 20 g. are diluted with
water and treated with excess of ammonia, ammonium sulphide, and
ammonium oxalate. In the case of acid. nitric. pur., no darkening or
cloudiness should be formed.
Iron is detected qualitatively by adding potassium thiocyanate to
the previously diluted solution ; quantitatively, by supersaturating the
acid with ammonia, heating for some time, and collecting the precipitated
hydroxide on an ash-free filter. Traces of iron are best determined
colorimetrically by means of potassium thiocyanate according to the
directions given by Lunge” (cf. p. 381). -
Iodine. Krauch states that nitric acid is coloured yellow by the
presence of gºo per cent of iodine, and that it may be identified by
shaking the acid with chloroform (Biltz). A yellow coloration may
1 The 7esting of Chemical Reagents for Purity, trans. by J. Williamson and L. W. Dupré,
I903, p. 185. &
* Z, angew. Chem., 1896, 9, 3.
NITRIC ACID 331
also be due to chlorine compounds, and is generally to be ascribed to
the presence of lower nitrogen oxides. The iodine is not usually
present in the free state, but as iodic acid dissolved in the nitric acid.
The oxygen compounds of iodine, as well as iodine itself, are recognised
by cautiously adding a very dilute solution of sulphurous acid or a few
drops of sulphuretted hydrogen water to the diluted acid and extracting
any iodine so liberated by carbon bisulphide, or testing by the addition
of starch solution. An excess of Sulphurous acid or of sulphuretted
hydrogen destroys the reaction.
An alternative method is to treat the solution with chemically pure
zinc, which reduces the iodic acid, whilst the nitrous acid formed at the
same time liberates iodine from the hydriodic acid produced. The
solution is then well shaken with carbon bisulphide.
According to the Pharmaceutical Commission of the German
“Apotheker-Verein,” iodine and iodic acid are tested for by shaking
the acid, diluted with twice its volume of water, with a small quantity of
chloroform ; the solvent should not be tinged violet even after the addi-
tion of a small piece of zinc to the acid solution.
Beckurts * gives the following as the most delicate test for iodine
in nitric acid :—I c.c. of the acid is boiled to remove lower oxides
of nitrogen, and to oxidise all the iodine to iodic acid, and I c.c.
of well-boiled water is then added, followed by a few drops of a
solution of potassium iodide and starch in air-free water. A
blue coloration indicates the presence of iodine in the original solu-
tion ; a blank test made with the potassium iodide and pure acid is
essential.
ExAMINATION OF NITRATING AND SPENT ACIDS. (Mixtures
OF SULPHURIC ACID, NITRIC ACID, ETC.)
Mixtures of sulphuric and nitric acids are prepared for coal-tar
colour works and for explosives factories for nitrating purposes. On the
other hand, waste acids are produced in these works, containing in
addition to the original components (of which the greater portion of the
nitric acid has naturally been utilised) much nitrous acid and a small
quantity of organic matter. Such residual acids as contain large
amounts of organic matter, and have attained thereby a tarry con-
sistency, as for example the acids from the purification of benzol and
petroleum, will not be considered. They contain no nitric acid, and
may be very considerably purified from the tarry matter by dilution;
the tarry constituents consist, in the main, of pyridine bases, condensed
hydrocarbons, etc.
* Arch. Pharm., 1887, 93. * Fischer's Jahresber., 1886, 305.
332 MANUFACTURE OF NITRIC ACID
According to Guttmann, the average composition of spent acid is as
follows:—

From
From From Nitrobenzene,
Nitroglycerine. Nitrocellulose. PicrºAsia,
65C.
HNO, 10 10 I
H2SO, 70 80 65
2 20 10 34
100 100 100
No notice is taken of lower nitrogen oxides, and of organic matter
in these data.
The following constituents are estimated in nitrating and similar
mixtures:—
I. Total Acidity.—Two to 3 g. are weighed off in the bulb
pipette, allowed to flow cautiously into a large volume of water, and the
total acid estimated by titration with normal sodium hydroxide solution.
Litmus may be used as indicator, but if so, prolonged boiling is neces-
sary; methyl orange may also be employed, notwithstanding the nitrous
acid present, if the titration be carried out in the cold and the operation
conducted as described on p. 66.
2. Lower Nitrogen acids are estimated by allowing the acid to flow
into a measured volume of potassium permanganate solution; the
method is more fully described under the testing of nitrous vitriol (p. 342;
cf. also p. 329). They may be calculated to HNO, or N.O, or to N.O,
(hyponitric acid). The lower nitrogen oxides contained in the strong
nitric acid consist mainly of N2O, ; on mixing with concentrated
sulphuric acid this is converted into equimolecular proportions of
HNOs, and of SO2(OH)(ONO). In the case of N.O., each c.c. of
seminormal permanganate solution used corresponds to O'oz 300 g.
N2O, ; if therefore a represents the number of c.c. of permanganate
taken, y the c.c. of acid required to decolorise this, and s the specific
gravity of this acid, then the N2O, content in g. per litre is equal to
23-OO 4: 2.3OO +
, or in percentage by weight to
3. Total Nitrogen acids are estimated in the nitrometer (p. 132).
The value found, less that obtained in determination (2), gives the
content of nitric acid, and (1)–(3) that of sulphuric acid.
4. Sulphuric acid. Two to 3 g. are weighed in the bulb-pipette
(Fig. I 34, p. 390), transferred to a small porcelain dish, and heated on the
water-bath for a half to one hour, a small quantity of water being
at the same time added to destroy all nitrosyl-sulphuric acid. The heat-
NITRATING AND SPENT ACIDS 333
ing is continued until on shaking the beaker all nitrous smell has dis-
appeared. The removal of the nitric acid is facilitated by occasionally
blowing on to the acid and rotating the dish. The contents are then
washed into a beaker and titrated with normal or N/2 sodium hydroxide
and methyl orange; this gives the sulphuric acid only.
Lunge and Berl’ have shown that this method gives low results, and
they recommend the determination of the sulphuric acid by difference,
from the total acidity less the total nitrogen acids, as given above.
The sulphuric acid may also be determined by precipitation as
barium sulphate, but the above method effects a considerable saving
of time, as a large number of samples can be evaporated simultaneously,
and then only require to be titrated to complete the analysis,
SULPHURIC ACID MANUFACTURE
RAW MATERIALS
These have already been described under—Sulphur (p. 264), Spent
oride (p. 279), Pyrites (p. 272), Zinc blende (p. 289), Mitre (p. 306), and
AVitric acid (p. 323).
CONTROL OF WORKING CONDITIONS
In the manufacture of sulphuric acid by the lead-chamber process,
many factors have to be taken into account, and a certain relationship
must exist between these for the regular and satisfactory working of
the plant. These factors comprise more especially the temperatures in
the various parts of the system, the colour of the gases in the chambers,
the draughts, the strengths of drip- and bottom-acid, of the Gay-Lussac
and Glover acids (also the temperature of the last), the quantity of
nitrogen oxides present in the acids, the composition of the gases, etc.
The rules to be observed in, and the conclusions to be drawn from,
these observations are fully described in Lunge's Sulphuric Acid and
A/ka/.”
This section is restricted to a description of such methods of investi-
gation as are essential to the satisfactory carrying out of the process.
EXAMINATION OF THE GASES
It is necessary to differentiate : Inlet gases, Chamber gases from
various parts of the system, and Exit gases beyond the Gay-Lussac
to Wer.
The examination of the inlet gases (kiln gases from the burning of
pyrites or blende) has been described under the preparation of
* Z. angew. Chem., 1905, 18, 1687. * Third edition, 1903, p. 675 et seq.
334 MANUFACTURE OF SULPHURIC ACID
sulphurous acid (p. 299). The determination deals solely with the
estimation of sulphur dioxide and sulphur trioxide. It is unnecessary
to estimate the oxygen accompanying these gases, since it must always,
in normal working, be in reciprocal relationship to the sulphur oxides
present.
The examination of the gases in the chambers is generally carried
out by observing the colour (more particularly in the hind portion of
the system and at the exit), the temperature as measured by means
of thermometers inserted at different places in the walls of the chambers,
and the pressure under which they stand, for which purpose the
manometer and anemometer described on pp. 165 et seq. may be
employed, though frequently this is only done in a very crude manner
by removing plugs or lutes. -
Generally speaking, a chemical analysis of the chamber gases is not
made, and under ordinary working conditions little would be gained
from such an analysis, owing to the difficulty of obtaining really average.
samples; and further, it is unnecessary, since the process can be
perfectly controlled by the analysis of the inlet and outlet gases.
Should it be desirable in a special and exceptional case to investigate
the chamber gases, the methods described under the investigation of
the exit gases will be found perfectly satisfactory. For more exact
determinations, reference should be made to the methods employed by
Lunge and Naef." .
The exit gases from the Glover tower are in the first place
examined for their content in oxygen, which constitutes one of the most
important factors for the regulation of the process in general, and for
the setting of the damper in the exit flue in particular. In many of the
foreign works this estimation is looked upon as sufficient, but in England,
and more recently in Germany, in consequence of the legal restrictions
imposed, the gases are also tested for total acidity.
I. Oxygen.—The oxygen is estimated by absorption, and measuring
the resulting decrease in volume. The absorption is effected either
by means of an alkaline solution of pyrogallol, which, however, requires
to be frequently renewed, and thus involves much time and expense,
or preferably by thin rods of moist phosphorus, which allows many
hundreds of analyses to be carried out in succession. It must, however
be remembered that phosphorus is not acted upon by oxygen at
temperatures below 16°, consequently when the determination has to
be made in a cold place, the absorption vessel must be slightly warmed
in any convenient manner. Lunge recommends the use of an alkaline
solution of sodium hydrosulphite, as suggested by Franzen (cf. p. 209), in
preference to either of the above reagents for the absorption of oxygen.
Since the water with the absorbents will take up any acid gases present,
* Chem. Ind., 1884, 7, 5.
CONTROL OF WORKING CONDITIONS 335
it is necessary to wash the gas by passing it through a solution of
sodium hydroxide before proceeding with the absorption of the oxygen.
When, as is usually the case, it is only necessary to make several
single tests in the course of the day, the use of a special aspirator
is unnecessary, as the gas burette itself suffices for this purpose by
filling and emptying the burette three or four times in succession from
the special opening in the gas main ; it is then safe to assume that
a fair sample of the gas in the main has been collected, and the analysis
may be proceeded with.
The practice of making a continuous test, in addition to the above,
is to be strongly recommended ; that is, to draw the gas slowly during
the full twenty-four hours into a special vessel, from which the sample
is taken for analysis. For this purpose any suitable aspirator of wood
or metal may be employed, provided the acid vapours are previously
removed from the gases. The apparatus for the estimation of the acids,
to be described later, effects this removal very conveniently. On the
other hand, more simple appliances may be used, so long as they permit
at least IO litres of the gas to be drawn off and measured in the
twenty-four hours, this quantity
being essential if a representative
average sample is required.
A good form of apparatus, devised
by Strype for the investigation of the
exit gases, is described and figured in
Lunge's Sulphuric Acid and Alkali."
The apparatus described by Davis,”
Lovett,” and Pringle * do not possess
any features of special value.
The estimation of oxygen is best
carried out by moist phosphorus in
an Orsat apparatus (p. 198), provided
with two absorption tubes, the first of
which is filled with sodium hydroxide
solution to remove the acid gases,
and the second with phosphorus. |2|=Lº
The manipulation is exactly the # # tº sº
same as in the examination of chim- ===
ney gases (p. 198). - Fº
Lindemann's apparatus as modi-
fied by C. Winkler” (Fig. 126) can also be employed. The measuring
tube A is fitted above with a three-way tap, and has a capacity of
* Vol. i., 3rd edition, p. 739. * /. Soc. Chem. And, 1882, 1, 2Io. -
* Chem. News, 1880, 41, 188. * Ibid., 1883, 2, 53.
* Handbook of Technical Gas Analysis, p. 51.

336 MANUFACTURE OF SULPHURIC ACID
IOO c.c., of which 25 c.c. are contained in the lower cylindrical portion,
which is graduated in tº c.c. B is the absorption vessel filled with thin
rods of phosphorus, and C the levelling bottle. The manipulation is
the same as in the Orsat apparatus.
M. Liebig's apparatus" is arranged for use with alkaline pyrogallol;
the gas is aspirated by means of a rubber bellows into a 50 c.c. pipette,
and forced from this through the absorbing solution into a graduated
measuring tube.
2. Examination for Acids.-If only the sulphur dioxide is to be
determined, a measured volume of the exit gas is drawn through
sodium hydroxide solution, and the resulting solution, after consider-
able dilution with water, poured into chlorine or bromine water. This
solution is then acidified with hydrochloric acid, heated to boiling,
and precipitated with barium chloride. Each I g. BaSO, corresponds
to 93.75 c.c. SOs. Reich's method (p. 3OO) is inapplicable in this
case, owing to the disturbing action exerted by the nitrous gases
present.
For a complete examination the following directions will suffice.
On the one hand the sum of the sulphur acids is determined, and on
the other the sum of the nitrogen acids, irrespective of the degree of
oxidation. The following scheme agrees in the main with that adopted
by the British Alkali Makers' Association in 1878; improvements have,
however, been made in some of the analytical details.
A small volume of gas is drawn continuously from the Gay-Lussac
exits by means of a constant aspirator, at such a rate that at least
24 cubic feet (O-68 cub. m.) are collected in the twenty-four hours. The
volume V drawn off must be measured with sufficient accuracy, either
by graduating the aspirator or by means of a gas meter; the volume
is then corrected by Tables VI. and VII. (Appendix) to o’ and 760 mm.,
and the corrected volume called V’. To make the comparisons more
complete, a record is kept of the number of cubic feet of chamber space
for each pound of sulphur burnt and entering the chambers in the
twenty-four hours (or kilos Sulphur per cubic metre), as calculated
from the weekly average; further, of the distance from the tower, of
the point at which the samples are taken. The gas is drawn through
four absorption bottles, each of IOO C.C. capacity, and giving a column
of liquid of at least 75 mm. in depth. The openings of the inlet-tubes
may not exceed mm. diameter, as measured by a standard wire.
The first three bottles each contain IOO c.c. of normal sodium hydroxide
solution (free from nitrate and equal to 31 g. Na,C) per litre), the fourth
IOO c.c. of distilled water. The gases are examined for—(1) Total
acidity, measured as sulphur trioxide; (2) Sulphur; (3) Nitrogen in
the form of acids, both the latter being measured in grains per cubic
* Post, Chem. Tech. Analyse, 2nd edition, I, p. 7oo.
CONTROL OF WORKING CONDITIONS 337
foot, or g. per cubic metre of the gas (reduced to o' and 760 mm.).
The operation is conducted as follows:—
The contents of the four bottles are combined, the bottles washed
out with a small volume of water, and the total divided into three
equal parts, one portion being kept as a reserve. The first portion
is titrated with normal sulphuric acid (49 g. H2SO, per litre) or with
normal hydrochloric acid, thus affording a measure of the total acidity,
SO2, H2SO, N2Os, and HNO3. The second portion is added gradually
to such quantity of a warm solution of potassium permanganate,
rendered strongly acid by means of pure sulphuric acid, that a slight
excess of permanganate remains; this excess is finally so far neutralised
by a few drops of sulphurous acid solution that only a faint rose tint
persists. All the nitrogen acids are now present as nitric acid, and
there is no excess of sulphurous acid in the solution. The nitric acid is
estimated by means of ferrous sulphate, employing for the purpose a
Solution containing IOO g. crystallised ferrous sulphate, and IOO g. pure
sulphuric acid per litre. To 25 c.c. of this solution contained in a flask
are added a further 20 to 25 c.c. of pure concentrated sulphuric acid,
and when the solution is cold the mixture resulting from the above
permanganate treatment is poured in. The flask is closed by a stopper
furnished with two tubes; the first is connected with a Kipp's apparatus
generating carbon dioxide, the second luted in water. The air in the
apparatus is expelled by the carbon dioxide, and the contents of the
flask heated until the solution, which at the start is dark-coloured owing
to the nitric oxide present, has become of a bright yellow colour. This
may require from a quarter to one hour according to the quantity of
nitric acid present, and the volume of sulphuric acid added. The
ferrous salt which remains unoxidised by the nitric acid is titrated back
by a N/2 permanganate solution equivalent to O.OO4 g. oxygen per I c.c.
(cf. p. 99). Since the strength of the acid ferrous sulphate solution
alters somewhat rapidly, it is necessary to compare it each day with
the permanganate solution by titrating 25 c.c., the quantity taken for
the test, with permanganate. If a c.c. of sulphuric acid are used in the
first titration, y C.C. of permanganate in titrating back the unoxidised
ferrous salt, and 2 c.c. of permanganate are equivalent to 25 cc. of the
ferrous sulphate solution, the figures required are obtained according to
the following equations; the measurement of volume (V°) being of
course made in cubic metres or in cubic feet according to the method of
expression —
(I) Total acidity expressed in-
o. I2O (100 – 3)
—y-
Grains SOs per cubic foot = lºssº-9
=
Grams SOs per cubic metre
338 - MANUFACTURE OF SULPHURIC ACID
(2) Sulphur in—
o.Oo3 (600 – 6x – 2 + y)
V.
_ o. 12346 (600 – 6x – 2 + y)
V.
Grams per cubic metre =
Grains per cubic foot
(3) Witrogen in—
Grams per cubic metre •ºg- )
o. 108o3(2 – y)
Grains per cubic foot V’
In place of the above method it will often suffice to simply deter-
mine the total acidity by titrating with W/IO sodium hydroxide solution
and phenolphthalein, either according to Lunge's method, described on
p. 302, or by employing the ten-bulb tube described below. -
The maximum escape allowed under the Alkali Act (1906) equals
4 grains per cubic foot, equivalent to 9. I5 g. SOs per cubic metre of
residual gases, before admixture with air or smoke; in Germany the
limit is 5 g. where pyrites is burnt, and 8 g. in the case of blende,
all acids being calculated to SOs.
The alkali inspectors have recently adopted a mixture of I vol. N/2
alkali and IO vols. neutral hydrogen peroxide for the absorption of the
acid gases. This overcomes the difficulty caused by secondary reactions
between sulphites and nitrites which is liable to occur' when sodium
hydroxide solution is used alone.
3. Nitric oxide may always be present in the exit gases even after
these have passed through the absorption bottles. Should it be desired
to estimate this gas, a ten-bulb absorption tube (Fig. 127) must be placed
FIG. 127.
between the last flask of the series and the aspirator. The tube is filled
with 30 c.c. of N/2 potassium permanganate solution to which I c.c. of
sulphuric acid of sp. gr. I-25 has been added. After the gas has passed
for twenty-four hours the tube is emptied and washed out. Fifty c.c. of
ferrous sulphate solution (the strength of which according to the pre-
ceeding section is = 22 c.c. permanganate) are added, and the decolorised
solution titrated back with permanganate solution till it again becomes
rose-coloured; the number of c.c. of permanganate required is called
" Carpenter and Linder, J. Soc, Chem, Ind., 1902, 21, 1490.

CONTROL OF WORKING CONDITIONS 339
u. The nitric oxide has thus reduced (30+ u–22) c.c. seminormal
permanganate solution. This corresponds to :
o.OO7(30+ u – 22)
N = P
3V
grams nitrogen per cubic metre of the gas when V' represents the total
volume, as registered by the aspirator; or in grains per cubic foot,
N = O. Io903(30+ u – 22)
3V'
A Winkler's coil, or other suitable apparatus, may be used instead of
the above absorption bulbs.
To test whether all the nitric oxide has been absorbed, Griess'
reagent (a-naphthylamine and sulphanilic acid), as modified by
Lunge and Ilosvay," may be employed, by allowing air to mix with
the gases, leaving the bulb-tube and then examining for higher oxides
of nitrogen. The above reagent will always show a slight reddening
with such gases even where the absorption has been made as complete
as possible, but for most practical purposes such traces may be
neglected.
According to Divers,” a concentrated alkaline solution of sodium or
potassium sulphite forms an excellent medium for the absorption of
nitric oxide in gas analysis.
Von Knorre and Arndt" mix the gas with hydrogen and pass the
mixture very slowly through a Drehschmidt platinum capillary tube
(p. 214) heated to bright redness. If the gas be passed too quickly or the
heating be insufficient, some ammonia always results. The reaction is:
2 NO + 2 H2 + 2 H2O + N,
(4 vols. 4 vols. 2 vols.)
Consequently each volume of nitric oxide corresponds to a contraction
of I+ volumes.
Von Knorre * states that it is more convenient to absorb the nitric
oxide in a mixture of 5 volumes of saturated potassium bichromate
solution and I volume concentrated sulphuric acid than in the
acidified permanganate solution described above, since the bichromate
mixture is perfectly stable at ordinary temperatures, and does not
evolve oxygen when shaken with indifferent gases; further, the nitric
oxide is oxidised quantitatively to nitric acid by this mixture. A
liberation of absorbed nitric oxide to indifferent gases cannot therefore
occur in this case, as it may when ferrous sulphate is employed. This
reagent is consequently well adapted for the removal of nitric oxide
from mixtures of gases, and also for the estimation of larger amounts
of this gas; for the estimation of smaller quantities such as occur in
* Z. angew. Chem., 1890, 3, 568. * Ber., 1899, 32, 2136.
* J. Chem. Soc., I899, 75, 82. * Chem. Ind., 1902, 25, 534.
340 MANUFACTURE OF SULPHURIC ACID
the above cases, absorption in acid permanganate solution and subsequent
titration is to be preferred.
4. Nitrous oxide in chamber or exit gases has scarcely ever been
estimated up to the present. The earlier known method of Winkler,"
which consists in the removal of the nitrogen acids and oxygen from
the gases and conversion of the nitrogen compounds in the residue into
ammonia by passing them together with excess of hydrogen over gently
heated palladium asbestos, and that of Lunge,” which is carried out
by first acting upon the gases with concentrated potassium hydroxide
solution and permanganate, shaking the residue with absolute alcohol,
and decomposing the gases expelled from the latter into nitrogen and
oxygen by means of a glowing palladium wire, do not suffice where the
quantity of nitrous oxide present is so small and is admixed with a
largely preponderating amount of nitrogen. The process worked out
by Pollak” is preferable. The mixture of NO, N,O, and N, from which
all oxygen has previously been removed by means of moist phosphorus, is
burnt either with hydrogen or with carbonic oxide in a Drehschmidt tube.
The contraction, Ve, is noted, oxygen added to the residual gas, and the
mixture again burnt in the capillary. Two-thirds of the resulting
contraction is equal to the volume of hydrogen that remains after
the first combustion, and this subtracted from the volume of hydrogen
originally added gives the volume of hydrogen used, VW. We thus
have :—
(1) x N.O + yNO + 2 N = W
(2) x + 3/2 y = Vc
(3) x + y = Vw
from which :—
3 = 3VW – 2Ve
y = 2 (Vc – Vw.)
2 = V – VW
In the case of combustion with carbon monoxide the contraction,
Ve, and the carbon dioxide formed, Vk, are measured.
This gives:—
(1) xN,O+ yNO+z N = W
(2) 1/2 y = Vc
(3) x + y = Vl.
from which it follows:—
3C = Vk - 2Vc
y = 2Vc
2 = V – Vk
For the proof of these formulae, see Treadwell, loc. cit.
5. Loss of Sulphur.—Lunge “has published a formula which allows
* Industrie-Gase, vol. 2, p. 429. * Ber., 1881, 14, 2188.
* Treadwell, Analytical Chemistry, vol. ii., p. 612.
* Cf. Sulphuric Acid and Alkali, vol. i., p. 749; Ding!. polyt.J., 1877, 226, 634.
CONTROL OF WORKING CONDITIONS 341
the quantity of sulphur burnt, expressed in grams per litre on the exit
gases, to be calculated from the percentage volume of oxygen present in
these. The sulphur loss may be calculated by comparing this value with
the quantity of sulphur acids present in the exit gases. The formula is :
J /*
3 = (20.95 – a)o.OO9637 1.oo367t 7to
where r equals the total sulphur burnt, expressed in grams per litre on
the exit gas, a the percentage volume of oxygen in the exit gas, t its
temperature, and h the pressure.
EXAMINATION OF THE WORKS' ACIDS
The strength of the Drips from the chambers is observed at
frequent intervals during the day and their content in nitrous vitriol
gauged, at least by a rough test. The latter can, of course, be more
exactly estimated by means of permanganate solution. Tests of the
bottom-acid are also made. -
For collecting the “drips” Lunge" recommends the use of small
cylinders of 20 c.c. capacity and correspondingly small hydrometers
having a range of about 20° Tw, in place of the customary larger
cylinder and full-range hydrometers. The smaller cylinders have the
advantage that the acid is renewed about every ten minutes, and thus
show any changes more quickly. -
For a rapid, approximate determination of the Nitrosity (percentage
of nitrogen acids) of the drips, Lunge” recommends the following
method for testing a set of four chambers. The test should be per-
formed at least once, better twice, daily, and for each test a stand
furnished with eight ordinary test tubes, each I 3 cm. high, is employed.
The test tubes are each filled to a height of IO cm. with samples of the
bottom-acid and of the drip-acid respectively, of each of the chambers,
taken in turn. The strength of each sample is taken at the same time by
the hydrometer and written on the stand at the bottom of the respective
tube. Concentrated ferrous sulphate solution, which does not require to
be free from ferric salts, is then cautiously added so as to form a layer
about 1 cm. deep on the acid in each cylinder. In the presence of very
minute traces of nitric acid or of the lower oxides of nitrogen a yellowish
coloured ring results at the junction of the liquids. In the presence of
larger amounts of nitrogen acids the ring is darker, or with still greater
quantities the whole of the ferrous sulphate layer assumes a deep
brown to black colour, and may even begin to froth owing to the heat
evolved driving off the dissolved nitric oxide from the solution. By an
inspection of the various colours in their sequence from chamber to
chamber, coupled with the strength of the acid from each chamber and
* Sulphuric Acid and Alkali, vol. i., p. 507. * /bid., p. 699.
342 MANUFACTURE OF SULPHURIC ACID
the colour of the gases in the latter, a very good indication of the
operation of the chamber process is obtained. The estimation of the
sulphur dioxide in the inlet gases and of the oxygen in the exit gases
complete the necessary testing.
The bottom-acid in the first chamber should show no trace of
nitrous acid. In the middle chambers the test should indicate a small
quantity of nitrous acid in the bottom-acid and a more pronounced
quantity in the drips. The bottom-acid in the last chamber should give
at least a moderate indication, and the drips a very marked reaction.
Should selenium be present the ferrous sulphate test is not so easily
applied; it may, however, be successfully employed after some practice.
For the quantitative estimation of the nitrogen acids in vitriol the
potassium bichromate method was formerly generally employed, but it
has now been replaced by the permanganate method, which gives a much
sharper end-reaction. The details of the permanganate method are
given under “Gay-Lussac acid" (p. 343).
It has been suggested that the daily production of acid might be
deduced by measuring the drips from the chambers. Lunge" has
shown, however, that this view is incorrect, owing to the drip tables act-
ing as contact surfaces. They consequently not only collect the acid
normally produced at the point of the chamber considered, but also the
much larger volume of acid condensed locally either by mechanical
means or owing to a better admixture of, and accompanying reaction
in, the chamber gases impinging against the solid wall.
The yield of chamber acid should be determined directly by measur-
ing the acid dip in the chambers, which is done by means of vessels
communicating with the interior. It is advisable to draw up a table
for each chamber, so as to see at a glance the volume of acid, in cubic
feet or litres, corresponding to any observed dip. If the specific gravity
be determined at the same time by means of the hydrometer, the weight
of acid in the chamber can be calculated.
Examination of the Glover acid. The strength (140° to I5O’Tw., sp.
gr. 1.7 to 1.75) and temperature are determined in case the acid is to be
pumped back at once to the Gay-Lussac tower; also the content in
nitrous compounds, should more than traces of these be present. (This
determination is described in connection with the finished acid, p. 377.)
Other tests are carried out exactly as in the case of the nitrous vitriol
from the Gay-Lussac tower.
Examination of the Gay-Lussac acid (AVitrous vitriol). Generally only
the nitrous acid is determined; this is not present in the vitriol in
the free condition, but is combined to form nitrosyl-Sulphuric acid
SO,(OH)(ONO). Nitric acid only occurs in poor nitrous vitriol; it is
included in the total nitrogen found by the nitrometer test.
* Z. angew. Chem., 1889, 2, 265.
EXAMINATION OF THE WORKS’ ACIDS 343
Nitrous acid is usually estimated by the permanganate method; the
bichromate method is to-day seldom employed (cf. supra). The test is
carried out in the manner proposed by Lunge," who has shown that the
earlier methods employed were inaccurate, owing to the ready loss of
nitric oxide and production of nitric acid that took place before the
reaction with permanganate was effected. A seminormal perman-
ganate solution, prepared and standardised as described (p. 99), is
employed. The essential point to be observed is that the nitrous
vitriol must always be added to the permanganate, and never the
reverse. The test is carried out by allowing the sample to flow from a
glass-tap burette into a measured volume of the M/2 permanganate
solution, diluted with five times its volume of warm water (30° to 40°),
until the colour just vanishes. At the ordinary temperature the reaction
proceeds too slowly, whilst at too high a temperature or in too concen-
trated solution hydrated manganese dioxide separates out ; such separa-
tion gives trouble, but it does not prevent the titration being carried to
a finish, since the dioxide goes into solution again before the completion
of the titration. The quantity of permanganate taken varies according
as the vitriol to be tested contains much or little nitrous acid; I c.c. per-
manganate corresponds to O.OO95O2 g. N2O3. For chamber acid or other
similar acid, 5 c.c. of permanganate solution at most are taken ; for a
strong nitrous vitriol up to 50 c.c. If the volume of permanganate solu-
tion taken be denoted by ar, and the volume of nitrous vitriol required
to decolorise it by y, the weight of N2Os in g. per litre is given by the
formula º* . Should it be desired to express the nitrogen content
otherwise than as N.O.s, the factor 9.502 is substituted by I5.75 for HNO3,
by 29.83 for nitric acid of 66°.5 Tw. at 15° C., by 25-44 for nitric acid of
77° Tw., and by 21.258 for NaNO3.
The following table (p. 344) saves calculation for cases in which 50
c.c. of permanganate solution are employed. The number of c.c. of
nitrous vitriol required are given in column y, the content in g. per litre
in column a, and in column & the percentage by weight, assuming the
acid examined to be of sp. gr. 1.7 (140° Tw.). For other strengths of
acid the percentage by weight is found by dividing the values in column
a by ten times the specific gravity of the acid. -
Riegler” has proposed to estimate the nitrous acid by the gas-volu-
metric method, using hydrogen peroxide according to the reaction :-
HNO, + H2O,- H2O + HNOs.
Grützner 8 estimates the nitrite by determining the chloride produced
by interaction with chloric acid by Volhard's method (p. 123):—
- 3HNO, + HCIO,-3HNO3 + HC1.
1 Ber., 1877, Io, Io'75. * Z, anal. Chem., 1897, 36, 665. * Arch. Pharm., 35, 241.
344
MANUFACTURE OF SULPHURIC ACID
Table for the estimation of Nitrous Acid in Nitrous vitriol.
Expressed as HNO, NaNO3, and as nitric acid of 66°-5 and 77° Tw, at 15° C., in
which 50 c.c. M/2 potassium permanganate is taken for the titration, and the
percentages by weight are based on sulphuric acid of 140° Tw. as unit.


Acid - HNO3. NaNO3. Nitric Acid 66°-5 Tw. || Nitric Acid 77° Tw.
required.
3/. a. .b. (t. b. Q. b. 0. b.
G. per | Percentage G. per | Percentage G. per | Percentage G. per | Percentage
C.C. Litre. by Weight. Litre. by Weight. litre. by Weight. Litre. by Weight.
10 78-8 4°61 106-3 6-22 || 149-1 8-72 127-2 7:44
11 71.6 4°19 96-6 5-65 135°6 7-93 115-6 6-76
12 65°6 3-84 88°6 5*18 124°3 7.27 106-0 6°20
13 60-6 3'54 81-8 4-78 114-7 6'70 97.8 5-72
14 56°2 3°29 75-9 4°44 106°5 6°23 90°8 5°31
I5 52°5 3.07 70-9 4°14 99°4 5°80 84'8 4°96
16 49°2 2°88 66°4 3°88 93-2 5*45 79.5 4-65
17 46 °3 2-71 62°5 3°65 87-7 5*13 74-8 4’37
18 43.7 2°56 59°0 3°45 82-9 4'85 70-7 4°13
19 41°5 2°42 55-9 3-27 78°5 4°59 66-9 3-91
20 39°4 2°30 53-1 3'11 74-6 4 °36 63-6 3.72
21 37-5 2-19 50-6 2.96 71-0 4°15 60-6 3'54
22 35°8 2-09 48°3 2°82 67-8 3-96 57.8 3:38
23 34°2 2°00 46°2 2*70 | 64 °8 3-79 55°3 3°23
24 32°8 1-92 44°3 2’60 62-1 3°62 53-0 3°11
25 31°5 1°84 42°5 2°49 59-7 3°49 50°8 2.97
26 30°3 1.77 40°9 2°39 57-4 3°35 48°9 2-86
27 29-2 1-71 39°4 2°30 55-2 3°25 47-1 2.75
28 28-1 1 *65 38°0 2-22 53-3 3°12 45°4 2-66
29 27.1 1°59 36-7 2-15 51*4 3-01 43°9 2°56
30 26°3 1°54 35°4 2-07 49-7 2-91 42°4 2'48
31 25 °4 1°49 34°3 2-01 48°1 2°81 41*0 2°40
32 24'6 1 *44 33°2 I-94 46"6 2-73 39.7 2-32
33 23.9 1°40 32°2 1°88 45 °2 2*64 38°5 2-25
34 23-2 1°35 31-3 I '84 43-9 2°56 37-4 2*19
35 22°5 1°32 30°4 1-78 42°6 2°49 36°3 2-13
36 21-9 1°28 29°5 1-73 41°4 2°42 35°3 2-07
37 21°3 1 *24 28-7 1°68 40°3 2°36 34°4 2°01
38 20-7 1 °21 28-0 l"64 39°3 2°30 33°5 1°96
39 20-2 1-18 27-3 1°60 38°2 2°23 32°6 I '91
40 19.7 1-15 26°5 1°55 37-3 2°18 31°8 1°86
41 19-2 1-12 25°8 1°51 36 °4 2-13 31°0 1°81
42 18-8 1 * 10 25-3 1°48 35-5 2°08 30°3 I-77
43 18°3 1-07 24-7 1°45 34°6 2-02 29'5 1-73
44 17.9 1*05 24°2 1-42 33-9 I '98 28-9 1.69
45 17-5 1*02 23°6 1°38 33-1 1 - 93 28-2 1 *65
46 17-1 1*00 23-1 1°35 32°4 1 '90 27.6 1-62
47 16-7 0-98 22°6 1°32 31-7 1-85 27-0 1-58
48 16'4 0.96 22*2 1°30 31-1 1-82 26°5 1 *55
49 16-1 0-94 21-7 1-27 30°4 1-78 25-9 1°51
50 15-8 0°921 21-3 I “25 29°8 1-74 25°4 1°49
55 14°3 O'837 19°3 1-13 27-1 I •59 23-1 1°35
60 13-1 0-768 17-7 I”04 24°5 1 *45 21-2 1°24
65 12-1 0.709 16°4 0.96 22-9 I "34 I9'6 1°14
70 11-2 0-658 15-2 0°89 21°3 1-25 I8-2 I •06
75 10°5 0 °614 14-17 0-829 I9°9 1-16 16°96 0-99
80 9-85 O-576 13°3 0.778 18-6 1-09 15°9 0-93
85 9-2 0°542 12°5 0.731 17.5 1-03 14°9 0-87
90 8-7 0-511 11°8 0°692 16-5 0-967 14°1 0°825
95 8-3 0°485 11.2 0-655 15-7 0-91S I 3-4 0.783
100 7-9 0°461 10°6 0-620 14 "9 0-873 12.7 0.744
NITROUS VITRIOL 345
The estimation of the total nitrogen acids present in nitrous vitriol
is generally effected in the nitrometer, the use of which for this purpose
has been already described (p. 134). Should the acid contain, in addition
to the nitrous acid, appreciable quantities of sulphur dioxide, which can
readily be detected by the smell, a little powdered potassium perman-
ganate is placed in the cup of the nitrometer; a large excess of perman-
ganate must not be used, since it interferes considerably with the reaction.
The observed gas volume must be reduced to o' and 760 mm. by
means of the Tables VI. and VII. (Appendix), unless the gas-volumeter
(cf. p. 138) be employed; with the latter instrument it is especially advis-
able to use a separate decomposition vessel.
The nitrogen compounds are calculated from the corrected nitric
oxide volumes by means of the following table. The table takes
account of the various ways of stating results adopted in different
works, and will be found useful in other than sulphuric acid works.
Column a gives the values in milligrams throughout, column b the per-
centage by weight, assuming that I c.c. of acid of 142° Tw. has been
taken for analysis; for other strengths of acid the values under a must
be taken as the basis for calculation (cf. p. 343).
N. NO. N2O3.
C.C.
NO. 0. b. 0. b. 0. e
Ing. Percentage. Ing. Percentage. Img. Percentage.
1 0-6257 0-0366 1°3402 0.0784 1°6975 O'0993
2 1 2514 0-0732 2°6804 0-1568 3°3950 0°1986
3 1-8771 0°1098 4°0206 O-2352 5-0925 0-2979
4 2°5028 0"I 464 5 °3608 0°3136 6'7900 0.3972
5 3-1385 0-1830 6-7010 0°3920 8’4875 0 °4965
6 3-7542. 0-21.96 8 °0412 0°4704 10-1850 O'5958
7 4-3799 0°2562 9-3814 0°5488 11°882.5 O'6951
8 5 °0056 0°2928 10-7216 0-6272 13°5800 0-7944
9 5*6313 0°3294 12-0618 0°7056 15-2775 0-8937
HNO3. Nitric Acid 66°-5 TW. NaNO3.
C.C.
NO. ſº, b. 0. º - b.
IIlg. Percentage. IIlg. Percentage. IIlg. Percentage.
1 2°8144 0 °1646 5 °333 0-312 3-7986. 0-2221
2 5-6288 0°3292 10 *666 0-624 7-5972 0 °4442
3 8 *4432 0°4938 15-999 0 '936 11°3958 0 °6663
4 11:2576 0 °6584 21°332 I •248 15°1944 0-8884
5 14-0720 0 '8230 26*665 1-560 18°9930 1*1105
6 16 °8864 0°9876 31°998 1-872 22°7916 1°3326
7 197008 1-1522 37-331 2°184 26*5902 1°5547
8 22°5152 1°3168 42°664 2-496 30°3888 1°7768
9 25°3296 1°4814 47-997 2-808 34-1874 1°9989




One hundred parts of nitric acid of 66°.5 Tw. correspond to 71.23
parts pure NaNO3, or 74.20 parts of 96 per cent. Chili saltpetre.
346 MANUFACTURE OF SULPHURIC ACID
Ratio of the three nitrogen acids to each other. The following
formulae may be used to determine the relationship between the
N,Os, N.O., and HNO3, present together in sulphuric acid when the
results of the permanganate titration and the nitrometer determination
of total nitrogen in the form of NO are known :-
a = c.c. NO, as found in the nitrometer;
b = c.c. O, calculated from the permanganate titration (I c.c. O =
I-4278 mg., consequently I c.c. N/2 permanganate =OOO4 g. =
2.80I 5 C.C. Oxygen);
a = vol. NO, corresponding to N present as N2Os;
y = vol. NO, 33 35 N2O4;
g = vol. NO, 3 y 33 HNO,
If 46 -a, then —
ar=46 – a ; y = 2 (a – 26) or = a – a.
If 46.<a, then :-
y=46; 3 = a – 46.
That is, if the oxygen is sufficient to allow of all the nitrogen acids being
present as N.O., they are calculated as such ; if there is more than
sufficient oxygen for this the excess is calculated to HNO3, whilst if
there is less oxygen the deficit is calculated to N2O4. In practice, the
ordinary nitrous vitriol contains no N2O, on which account it is better
to calculate the total oxygen deficit (as determined by permanganate)
to N2Os or to nitrosyl-Sulphuric acid, the rest of the nitrogen being
calculated to nitric acid.
FINISHED PRODUCT : SULPHURIC ACID
Sulphuric acid appears in commerce as chamber acid of IO6° to
I 16° Tw.; as 140° Tw. acid from the Glover tower, or after evaporation
in lead pans; as ordinary 168° Tw. acid (93 to 95 per cent.); as extra
concentrated acid (96 to 98 per cent.); as commercial monohydrate, and
as fuming sulphuric acid. The last will be dealt with separately.
Actual sulphuric acid, generally referred to as monohydrate, H2SO,
has a specific gravity of I-853 at O’ and, according to different observers,
of I-8372 to I-8384 at I 5° compared with water at 4°. The German
Standards Commission calculate the specific gravity of actual mono-
hydrate as I-8357 at I 5"/4” by extrapolation; it should, however, be
borne in mind in connection with this extraordinarily low value, that
owing to the remarkable irregularities that occur in the density curve
when such high concentrations are reached extrapolation may be very
misleading. The maximum specific gravity (I-8415) lies between 97
and 98 per cent, acid, and the density of the monohydrate is thus
increased by addition of a little water as well as by addition of sul-
phuric anhydride. The monohydrate begins to boil at 290°, and at
SULPHURIC ACID 347
first evolves some SOs, the boiling point rising to 338°, at which
temperature 98 per cent of H2SO, and 2 per cent of H.O are present."
The strength of sulphuric acid is generally determined by its specific
gravity. The earlier tables in use have been found to be altogether
unreliable, especially at the higher concentrations, and Lunge and Isler”
consequently worked out a new table based on fresh observations, using,
however, the values previously determined by Lunge and Naefº for
acids of over 90 per cent. H2SO4. All uncertainty in this connection
has now been removed by the very exhaustive and painstaking investi-
gations of Domke, Bein, Fischer and others, whose work is embodied
in the reports of the German Standards Commission.* This work,
undertaken in the first place in connection with the establishment of
the normals for percentage hydrometers, has been carried out with such
care that the results may safely be looked upon as exact; the report
contains in addition a detailed description and criticism of all previous
work on the subject.
The following tables are taken from this source. (1) The table for the
estimation of the sp. gr. of pure Sulphuric acid-water mixtures from the
percentage content for I 5°C/4°C.; the fifth decimal place, which was
only determined for purposes of calculation, is omitted, but has been
used to round off the fourth figure. (2) An extract from the tables for
reducing to 15° C. the reading found at some other temperature. The
original memoir contains many other tables of less importance for the
present purpose, and which are not included here.
In these tables the values are given for progressively increasing
percentages of sulphuric acid. The form adopted in the tables prepared
by Lunge, based upon measurements made in conjunction with Isler and
Naef, is much more useful for works and laboratory purposes. These
tables are accordingly included ; they have been recalculated to the
values of the German Standards Commission for all cases in which they
differed by more than O. I per cent. Larger variations than this
scarcely appear below the specific gravity of I-560 but between this
point and I-680 differences rising up to about 3 per cent, occur; the
causes of these divergencies have not been discovered up to the
present. Still greater differences exist at certain of the higher con-
centrations, especially from 96 per cent. onwards, but at these strengths
the determination of the acid content by means of the specific gravity
is quite unreliable, and consequently the readings do not come into
question for practical purposes. Perhaps the most pronounced variation
(o. 54 per cent.) occurs at 99% per cent., and it is interesting to note that
Koechlin and Gerber have shown that at this extremely high concen-
tration acids prepared in different ways may, although of the same
' Cf. Lunge, Sulphuric Acid and Alkali, vol. I, p. 173. * Chem. Ind., 1883, 6, 37.
* Z. angew. Chem., 1890, 3, I31. * Abhandlungen, 1904, vol. 5.
348
MANUFACTURE OF SULPHURIC ACID
chemical composition, possess different physical properties, a view
hinted at by Lunge in 1883.
A greater degree of accuracy than
O. I per cent. is not likely to be attained in practice, and this limit
is more than sufficient for the tables, since the figures given
are for chemically pure acid which is not found in ordinary practical
It may be noted that the German Standards Commission always
worked to temperatures as recorded by the hydrogen scale, which at 15°
work.
gives values O. I lower than those recorded on the mercury scale.
Table for determining the Specific Gravity of Pure Sulphuric
Acid-water Mixtures from the Percentage Content.

# •0 •] •2 | •8 *4 •5 •6 •7 •8 •9
#:
: 5 Density of Sulphuric Acid at +15° C. compared with Water at +4° C.
§ 5. for percentages given in units in the vertical column at the side, and in decimal parts in the
73 horizontal line above.
ÚO
0 || 0 °9991 || 0 °9998 || 1'0005 || 1'0012 1-0019 || 1 0026 1-0033 || 1:0040 || 1-0047 | 1.0054
1 || 1-0061 1-0068 || 1:0079 || 1:0082 1-0088 || 1-0095 | 1.0102 | 1.0109 || 1:01.16 1.0122
2 || 1 0129 || 1:01:36 || 1:01.43 || 1:01.49 || 1-0156 || 1:01.63 || 1:01.70 || 1:01.76 || 1-0183 1:01.90
3 1-019.7 | 1-0203 || 1-0210 || 1:02.17 | 1:02:24 1-0230 1-0237 1-0244 1-0251 | 1.0257
4 || 1 °0264 1-0271 1*0277 1*0284 1*0291 || 1 °0298 || 1:0304 || 1:03.11 || 1:03.18 || 1:0325
5 || 1:0332 || 1 0338 || 1:0345 || 1:0352 || 1:0359 || 1:0366 || 1:0373 || 1-0380 | 1-0386 || 1:0393
6 || 1:0400 || 1:0407 || 1:04.14 || 1:0421 || 1:0428 || 1:0435 | 1-0442 | 1-0449 || 1:04:56 || 1:0462
7 || 1:0469 || 1-0476 | 1-0483 || 1:0490 || 1:0497 || 1:0504 || 1:05.11 || 1:0518 || 1:05.25 || 1:0532
8 || 1-0539 1.0546 || 1:05.54 || 1:0561 | 1.0568 || 1:05.75 || 1:0582 | 1.0589 || 1:05.96 || 1:0603
9 || 1-0610 || 1 °0617 | 1-0624 1-0631 || 1 °0638 || 1:0645 || 1:0653 1-0660 | 1.0667 1*0674
10 || 1°0681 1-0688 || 1-0695 || 1:0702 || 1:07.10 || 1:07.17 | 1.0724 | 1.0731 || 1:0738 1-0745
11 || 1:07:53 | 1.0760 | 1-0767 | 1.0774 || 1:0781 || 1:0789 || 1:0796 | 1.0803 || 1:08.10 | 1.0818
12 || 1-0825 | 1.0832 1*0839 1-0847 | 1.0854 || 1-0861 1-0868 1-0876 | 1.0883 1-0890
13 || 1-0898 || 1-0905 || 1-0912 || 1:0920 | 1-0927 || 1:0934 || 1:0942 || 1:0949 | 1.0956 || 1:0964
14 || 1:0971 || 1:0978 || 1:0986 || 1:0993 || 1:1000 || 1:1008 || 1:1015 1-1023 1-1030 || 1:1038
15 || 1-1045 1°1052 || 1:1060 || 1:1067 || 1:1075 || 1 1082 || 1:1090 1-1097 || 1-1105 || 1 III2
16 || 1-1120 1-1127 1-1135 || 1:1142 1*1150 || 1:1157 | 1-1165 | 1.1172 1-1180 1-1187
17 | 1-1195 || 1:1202 || 1:1210 || 1:1217 || 1:1225 || 1:1233 || 1:1240 || 1:1248 1-1255 1-1263
18 || 1:12.70 | 1-1278 1-1286 || 1:1293 || 1:1301 || 1:1309 || 1:1316 || 1:1324 1-1331 1-1339
19 || 1-1347 || 1-1354 || 1 1362 || 1:1370 || 1:1377 || 1-1385 || 1:1393 || 1:1400 | 1-1408 || 1-1416
20 || 1-1424 1-1431 || 1:1439 || 1:1447 || 1 °1454 || 1:1462 || 1:1470 || 1:1478 - 1-1485 | 1-1493
21 || 1:1501 || 1:1509 || 1:1516 || 1:1524 || 1:1532 || 1:1540 || 1:1548 || 1:1555 1-1563 1-1571
22 1°1579 || 1 1587 || 1 °1594 || 1:1602 || 1:1610 || 1:1618 || 1:1626 1-1634 1-1641 1-1649
23 || 1 1657 1-1665 1-1673 || 1:1681 || 1 1689 || 1 1697 || 1:1705 || 1-1712 1-1720 | 1.1728
24 1-1736 || 1:1744 || 1:1752 || 1:1760 || 1:1768 || 1:1776 || 1:1784 || 1:1792 | 1.1800 1-1808
25 || 1-1816 || 1 1824 || 1:1832 || 1:1840 || 1:1848 || 1 1856 || 1:1864 || 1 1872 1-1880 || 1 1888
26 || 1-1896 || 1:1904 || 1 °1912 1°1920 | 1 °1928 || 1:1936 || 1:1944 || 1:1952 || 1:1960 || 1:1968
27 || 1 1976 || 1:1984 || 1:1992 || 1:2000 || 1:2008 || 1 2016 || 1:2025 || 1:2033 1-2041 | 1.2049
28 || 1:2057 || 1:2065 | 1.2073 || 1:2081 || 1:2089 || 1:2098 || 1:2106 || 1:2114 | 1.2122 | 1.2130
29 || 1:2138 | 1.2146 || 1:2155 || 1:2163 || 1:2171 || 1:2179 || 1:2187 || 1:2196 || 1:2204 || 1:2212
30 || 1-2220 | 1.2228 || 1 °2237 1'22.45 || 1:22:53 || 1:2261 || 1:22.70 || 1:2278 || 1:2286 || 1:2294
31 1-2302 || 1:2311 || 1 °2319 || 1:2327 | 1 °2335 || 1:234.4 I-2352 || 1:2360 | 1-2368 || 1:2377
32 1°2385 || 1 °2393 || 1 *2402 || 1 ‘2410 || 1 °2418 || 1:2426 || 1:2435 | 1:2443 || 1:24:51 || 1:2460
33 1-2468 || 1 °2476 || 1 °2485 1“2493 || 1:2501 || 1 2510 || 1:2518 || 1:2526 1:2535 | 1.2543
34 || 1:2552 || 1:2560 | 1:2568 || 1:2577 || 1:2585 || 1:2594 | 1.2602 || 1:2610 | 1.2619 || 1:2627
SPECIFIC GRAVITY OF SULPHURIC ACID
349
Specific Gravity of Pure Sulphuric Acid-water—Continued.

# •0 •1 •2 •3 *4 •5 •6 .7 •8 •9
#:
à 5 Density of Sulphuric Acid at +15° C. compared with Water at +4° C.
§ 5, for percentages given in units in the vertical column at the side, and in decimal parts in the
Ž horizontal line above.
35 || 1:2636 || 1:2644 | 1.2653 || 1:2661 || 2:26.70 || 1:2678 1-2686 || 1:2695 || 1:2703 || 1:2712
36 1-2720 || 1-2729 || 1:2737 || 1:2746 || 1:2754 || 1:2763 || 1:27.71 1-2780 || 1:2788 || 1:2797
37 1 °2806 || 1 2814 || 1:2823 || 1:2831 || 1:2840 || 1:2848 || 1:2857 || 1:2866 | 1.2874 1°2883
38 1°2891 || 1 °2900 || 1:2909 || 1:2917 | 1.2926 || 1:2935 | 1.2943 || 1:2952 || 1:2961 || 1:2969
39 1-2978 || 1 °2987 || 1 2995 || 1:3004 || 1:3013 || 1:3022 || 1:3030 | 1.3039 || 1:3048 1°3057
40 1°3065 1-3074 || 1:3083 || 1:3092 || 1:3101 || 1:3109 || 1:3118 1°3127 | 1.3136 || 1:3145
41 1°3153 1-3162 1-3171 || 1 °3180 1-3189 || 1.3198 || 1 °3207 || 1 °3215 1°3224 || 1 °3233
42 1°3242 || 1 °3251 || 1:3260 || 1:3269 || 1:3278 || 1:3287 || 1:3296 || 1:3305 || 1:3314 || 1 °3323
43 1°3332 1-3341 || 1 °3350 || 1:3359 || 1 °3368 || 1:3377 | 1.3386 1°3395 || 1 3404 || 1 °3413
44 1°3423 1:34.32 || 1:3441 || 1:3450 | 1.3459 || 1:3468 | 1.3478 || 1:3487 | 1.3496 || 1 °3505
45 I-3514 I-3524 || 1:3533 || 1:3542 | 1.3551 || 1:3561 | 1.3570 | 1.3579 || 1:3589 || 1:3598
46 1°3607 || 1:3617 | 1.3626 || 1-3635 | 1-3645 || 1:3654 || 1:3664 || 1 °3673 || 1 °3682 J "3692
47 1°3701 || 1:371.1 | 1.3720 || 1 °37.30 | 1.3739 || 1:3749 || 1-3758 || 1-3768 || 1:37.77 1-3787
48 1°3796 || 1 3806 || 1:3816 || 1:38.25 | 1-3835 || 1:3844 || 1:3854 || 1 "3864 || 1:3873 || 1:3883
49 1°3893 | 1.3902 || 1:3912 1-3922 || 1 3931 || 1:3941 | 1.3951 || 1:3961 || 1 °3970 1°3980
50 || 1:3990 || 1:4000 || 1:4010 || 1 “4019 || 1:4029 || 1:4039 || 1:4049 1-4059 | 1' 4069 || 1:4079
51 1 “4088 || 1:4098 || 1:4108 || 1:41.18 || 1:4128 || 1:4138 | 1.4148 || 1:41:58 || 1:4168 || 1:41.78
52 1°4188 || 1:41.98 || 4-4208 || 1:4218 || 1:4228 || 1:4238 || 1:4249 || 1:4259 || 1:4269 || 1:4279
53 1-4289 || 1:4299 || 1:4309 || 1:4319 || 1:4330 || 1:4340 || 1:4350 || 1:4360 | 1:43.70 || 1:4381
54 1-4391 || 1:4401 | 1.4411 | 1.4422 || 1:4432 || 1:44.42 || 1:4453 | 1.4463 || 1:4473 || 1:4484
55 || 1:4494 || 1:4504 || 1:4515 || 1:4525 || 1:4535 | 1.4546 || 1:4556 | 1'4567 || 1:4577 | 1'4587
56 1°4598 || 1:4608 || 1:4619 || 1:4629 || 1:4640 || 1:4650 || 1:4661 1'4671 || 1'4682 || 1'4692
57 1°4703 || 1:4714 | I-4724 || 1:4735 | 1.4745 || 1:4756 || 1:4767 || 1:4777 || 1:4788 1°4798
58 1°4809 || 1:48.20 | I-4837 1:4841 1'4852 || 1:4863 1-4873 || 1:4884 || 1:4895 || 1 °4905
59 I'4916 || 1:4927 | 1:4938 || 1 °4949 || 1:4960 || 1:49.70 || 1 4981 || 1:4992 || 1:5003 || 1 ‘5013
60 1°5024 || 1:5035 | 1:5046 1-5057 || 1:5068 || 1:5079 || 1:5090 || 1 °5101 || 1:5112 || 1 °5122
61 1°5133 1-5144 || 1 °5155 1-5166 || 1:5177 || 1-5188 || 1:5199 || 1 °5210 || 1:52:21 || 1 °5232
62 1'5243 || 1:5,254 || 1 °5265 | 1.5276 || 1:5287 || 1:5298 || 1:5309 || 1:5321 || 1:5332 || 1:53:13
63 1°5354 || 1 °5365 || 1:5376 || 1 °5387 || 1:5398 || 1:54.10 || 1:5421 1°54'32 1°5443 | 1.5454
64 1'5465 || 1 °5477 || 1 °5488 || 1 °5499 || 1:5510 || 1:5521 || 1:5533 || 1 °5544 || 1:5555 | 1'5566
65 1:5578 1-5589 1-5600 | 1-5612 | 1.5623 || 1-5634 || 1:5645 || 1:5657 || 1:5668 || 1:56.79
66 1-5691 || 1-5702 || 1:5713 || 1:5725 | 1.5736 || 1-5748 || 1:5759 || 1:57.70 || 1:5782 | 1.5793
67 || 1:58.05 1-5816 || 1:5827 | 1:58.39 1-5850 || 1:5862 || 1:5873 1-5885 1°5896 || 1:5908
68 || 1:5919 1°5931 || 1:5942 1-5954 || 1-5965 || 1:5977 | 1.5989 || 1:6000 || 1:6012 | 1.6023
69 1'6035 | 1-6046 || 1-6058 1-6070 | 1.6081 | 1.6093 I-6104 || 1 °6116 || 1:6128 || 1:6139
70 | 1.6151 1-6163 || 1:6174 || 1-6186 1-6198 || 1:6209 || 1-6221 | 1-6233 || 1:6245 || 1:6256
71 || 1-6268 I-6280 I-6291 || 1:6303 || 1:6315 || 1.6327 | 1:6338 || 1:6350 || 1:6362 || 1:6374
72 1°6385 || 1:6397 || 1:6409 || 1:6421 || 1:6433 || 1:6444 || 1:6456 || 1:6468 1°6480 || 1:6492
73 || 1:6503 || 1 °6515 || 1 °6527 || 1:6539 || 1.6551 || 1:6563 | 1.6574 || 1 °6586 || 1°6598 || 1°6610
74 1:6622 || 1-6634 || 1:6645 | 1.6657 | 1.6669 || 1-6681 1-6693 || 1:6705 | 1.6717 | 1-6728
75 1-6740 | 1.6752 | 1.6764 | 1.6776 || 1:6788 || 1.6799 || 1:6811 | 1.6823 1°6835 | 1-6847
76 1°6858 || 1 °6870 || 1:6882 || 1:6894 || 1:6906 || 1-6917 | I-6929 || 1:6941 || 1-6953 | 1-6965
77 1-6976 || 1-6988 || 1:7000 | 1.7012 || 1-7023 || 1-7035 | 1.7047 | 1-7058 || 1-7070 17082
78 1-7093 || 1:7105 || 1-7117 | 1-7128 1-7140 || 1-7151 1-7163 1-7175 1-7186 || 17198
79 || 1-7209 || 1:7221 || 1:7232 | 1.7244 1-7255 || 1:7267 || 1:7278 || 1:7289 || 1:7301 || I-7312
350
MANUFACTURE OF SULPHURIC ACID
Specific Gravity of Pure Sulphuric Acid-water—Continued.
•0
•l
•2
•3
*4
•5
•6
•7
•8
•9
;
i
Density of Sulphuric Acid at +15° C. compared with Water at +4°C.
for percentages given in units in the vertical column at the side, and in decimal parts in the
horizontal line above.
1°7324
1°7435
1°7544
1°7649
1-7748
1°7841
1-7927
1-8006
1-8077
T “814]
1°8198
1°8248
1°8293
1°8331
1°8363
1-8388
1-8406
1°8414
1 °84.11
1-8393
1-7335
1-7446
1-7555
1-7659
1.7758
1-7850
1°7935
1-8013
I '8084
1°8147
1-8203
I '8253
1-8297
1-8335
1°8366
1-8390
1-8407
1°8415
1-84.10
1°8391
1-7346
1-7457
1 °7566
1-7669
1-7767
1°7859
1°7943
1-8021
1°8090,
1°8153
1-8208
1°8258
1°8301
1-83.38
1-8369
1-8392
1°8408
1°8415
1-8409
1-8388
1-7357
1°7468
1-7576
1-76.79
1.7777
1.7868
1°7951
1-8028
1-8097
1-8158
1°8213
1°8262
1°8305
1°8341
1-8371
1°8394
1°8409
1°8415
1-8408
1-8385
1-7369
1-7479
17587
1-7689
1-7786
1-7876
1-7959
1°8035
I '81 03
1°8164
1-82.19
1-8267
1-8309
I '8345
1-8374
1°8396
1-84.10
1°8415
1-8406
1°8381
1-7380
1°7490
1 7597
1°7699
17796
1°7885
I-7967
1°8042
1°8110
1°8170
1°8224
1-82.71
1°8313
1°8348
1-8376
1-8398
1-84.11
1°8415
1°8405
1-8378
1°7391
1-7501
1-7608
1-7709
1°7805
1°7894
1-7975
1°8049
1°8116
1°8176
1-8229
1-8276
1-8317
1°8351
1-8379
1-8400
1°8412
1°8414
1°8403
1-7402
17512
1.7618
1-7719
1-7814
1-7902
1-7983
1°8056
1°8122
1°8181
1°8234
1°8280
1°8320
I '8354
1°8381
1°8401
1°8413
1°8414
1-8401
1°7413
1°7523
1°7628
1-7729
1-7823
1-7911
1-7991
1°8063
1°8129
I '8187
1°8239
1°8284
1°8324
1-8357
1°8384
1°8403
1°8414
1°8413
1°8398
1-7424
1°7534
1°7639
1-7738
17832
1-7919
1-7998
1-8070
1°8135
1°8192
1°8244
1°8289
l”8328
1°8360
1°8386
1°8404
1°8414
1°8412
1°8396
(1-8374) (1-8370) (18366)|(18362)
(1-8357)
(The bracketed values from 99.6 to 100 per cent. are extrapolated.)
Specific Gravity of Sulphuric Acid Solutions.
Lunge, Isler, and Naef.

100 parts by weight of chemically
1 Litre of chemically pure Acid
pure Acid contain per cent.
Sp. Gr. contains kilograms of
at
Degrees
s Degrees
Baumé.
# TWaddell.
(in vacuo.)
Acid
"Of
142° TW.
Acid Acid
Of
142° Tw.
Acid
f
SO3. O
106° Tw.
H2SO4, SO3. H2SO4.
Of
106° Tw.
1°000
1°005
1*010
1*015
1*020
1*025
1*030
1*035
1*040
1*045
1*050
1*055
1*060
1°065
1-070
1-075
0-001
0°008
0-013
0-019
0.025
0.032
0-038
0°044
0-051
0-057
0-063
O'070
0-076
0-082
0°089
0°096
0-001
0°009
0-016
0-023
0-031
0-039
0-046
0-054
0-062
0.071
0.077
0-085
0°093
0-102
0°109
0-117
0-001
0°013
0°020
0-030
0-040
0-049
0-059
0-070
0.079
0'089
0-099
0°109
0°119
0-129
0°140
0°150
0-001
0°015
0-025
0-037
0-050
0-062
0-074
0-087
0-099
0-112
0-124
0° 136
0°149
0°161
0-174
0°188
i
i
I
;
:
:
1
2
9
I
II "24
12°14
13-05
13-96
I
SPECIFIC GRAVITY OF SULPHURIC ACID
351
Specific Gravity of Sulphuric Acid Solutions—Continued.

100 parts by weight of chemically
1 Litre of chemically pure Acid
Sp. Gr. pure Acid contain per cent. contains kilograms of
aft, Degrees Degrees
#. Baumé. Twaddell. SO H.jSO Aft Aft SO HoSO Aja Aft
* º *Aºzº lº O O *Rs . g O O
(in vacuo.) 3 2SV4 142° TW. 106° TW. 3 2SV4 142° TW. 106° TW.
1*080 10°6 16 9°47 II*60 | 14-87 18°56 || 0-103 || 0-125 || 0-161 || 0-201
1*085 11-2 17 10-04 || 12:30 15-76 | 19.68 || 0-109 || 0-133 || 0-171 || 0-213
1°090 II '9 18 10-60 | 12-99 || 16*65 20°78 || 0-116 || 0-142 || 0-181 0.227
1°095 12°4 19 11-16 || 13:67 17°52 21-87 || 0-122 || 0-150 || 0-192 || 0-240
1“100 13-0 20 11-71 || 14-35 | 18-39 22-96 || 0-129 || 0-158 || 0:202 || 0-253
1°105 13-6 21 12°27 | 15'03 || 19°26 24-05 || 0-136 || 0-166 || 0:212 || 0:265
1*110 14-2 22 12'82 15-71 20:13 25°14 || 0-143 || 0-175 0.223 || 0-279
1*115 14°9 23 13'36 16°36 || 20°96 || 26°18 || 0-149 || 0-183 || 0-234 0°292
1*120 15°4 24 13°89 || 17 °01 || 21°80 27-22 || 0-156 || 0-191 || 0:245 || 0°305
1-125 16°0 25 14'42 || 17-66 22-63 || 28-26 || 0-162 || 0-199 || 0-255 || 0-318
1°130 16 °5 26 14-95 | 18-31 || 23°47 | 29-30 || 0-169 || 0.207 || 0-265 || 0-331
1“135 17-1 27 15°48 18°96 || 24°29 || 30-34 || 0-176 || 0-215 || 0:276 || 0-344
1*140 17.7 28 16°01 || 19°61 || 25°13 || 31°38 || 0-183 || 0°223 || 0-287 || 0°358
1°145 18-3 29 16°54 || 20°26 || 25°96 || 32°42 || 0-189 || 0-231 || 0:297 0°371
1°150. 18°8 30 17-07 || 20-91 26'79 || 33°46 || 0-196 || 0-239 || 0.308 || 0-385
1°155 19°3 31 17°59 || 21°55 27-61 34-48 || 0:203 || 0:248 || 0-319 || 0°398
1*160 19°8 32 18°11 22°19 || 28°43 || 35°50 || 0-210 || 0-257 || 0-330 || 0°412
1°165 20-3 33 18-64 22-83 29°25 || 36-53 || 0:217 | 0.266 || 0-341 || 0-426
1-170 20-9 34 19:16 || 23°47 || 30-07 || 37.55 || 0-224 || 0-275 || 0-352 || 0:439
1°175 21°4 35 19°69 || 24*12 || 30°90 || 38°59 || 0°231 || 0°283 || 0-363 0°453
1*180 22-0 36 20-21 | 24-76 || 31°73 || 39-62 || 0-238 || 0:292 || 0-374 || 0:467
1*185 22°5 37 20-73 || 25-40 || 32°55 | 40-64 || 0-246 || 0°301 || 0-386 0-481
1*190 23-0 38 21:26 26-04 || 33°37 || 41-66 || 0-253 || 0-310 || 0-397 || 0:496
1°195 23-5 39 21°78 - 26*68 || 34°19 || 42°69 || 0°260 || 0°319 || 0°409 || 0-511
1°200 . 24°0 40 22-30 || 27:32 || 35'01 || 43-71 || 0-268 || 0.328 || 0-420 || 0-525.
1°205 24°5 41 22°82 27-95 35'83 44-72 || 0-275 || 0:337 || 0:432 || 0-539
I •210 25-0 42 23°33 28°58 36°66 45-73 || 0-282 0-346 0-444 || 0-553
I-215 25 °5 43 23°84 || 29°21 37°45 || 46-74 || 0-290 0-355 || 0°455 || 0-568
1-220 26-0 44 24°36 || 29°84 || 38°23 || 47°74 || 0-297 || 0-364 || 0-466 || 0:583
1.225 26°4 45 24'88 || 30-48 || 39-05 || 48-77 || 0-305 || 0-373 || 0:478 || 0-598
1°230 26-9 46 25°39 || 31°11 || 39°86 || 49°78 || 0-312 || 0°382 || 0:490 || 0-612
1-235 27.4 47 25'88 31.70 | 40-61 || 50-72 || 0-320 || 0-391 || 0-502 || 0-626
1-240 27-9 48 26°35 | 32°28 || 41°37 || 51-65 || 0-327 || 0-400 || 0:513 || 0-640
1°245 28°4 49 26°83 || 32°86 || 42°11 || 52°58 || 0°334 || 0°409 || 0-524 || 0-655
1*250 28:8 50 27°29 || 33°43 || 42°84 || 53-49 || 0-341 || 0°418 || 0-535 | 0°669
1°255 29°3 5]. 27°76 34-00 || 43°57 54-40 || 0.348 || 0-426 || 0-547 || 0-683
1°260 29-7 52 28-22 || 34°57 || 44-30 || 55-31 || 0-356 || 0:435 | 0-558 || 0-697
1°265 30°2 53 28°69 || 35°14 || 45°03 || 56-22 || 0-363 || 0-444 || 0°570 0-711
1-270 30°6 54 29°15 || 35-71 || 45°76 || 57.14 || 0-370 0-454 || 0-582 0-725
1-275 31°1 55 29-62 36°29 || 46°50 58.06 || 0°377 || 0°462 || 0-593 || 0-740
1°280 31°5 56 30-10 || 36-87 || 47°24 || 58-99 || 0-385 || 0-472 || 0-605 || 0-755
1-285 32°0 57 30°57 || 37°45 47-99 || 59-92 || 0-393 || 0°481 || 0-617 || 0-770
1°290 32°4 58 31-04 || 38-03 || 48-73 || 60-85 || 0°400 || 0-490 || 0-629 || 0-785
1°295 32°8 59 31°52 || 38°61 || 49°47 | 61-78 || 0°408 || 0-500 || 0-641 || 0-800
1°300 33°3 60 31°99 || 39°19 || 50°21 || 62.70 || 0°416 || 0-510 || 0-653 0-815
1°305 33.7 61 32°46 || 39°77 | 50-96 || 63-63 || 0-424 || 0-519 || 0-665 || 0:830
1°310 34°2 62 32°94 | 40°35 || 51-71 64°56 || 0°432 || 0-529 || 0-677 || 0-845
1°315 34°6 63 33°41 | 40°93 || 52°45 || 65°45 || 0°439 || 0°538 || 0-689 || 0-860
1°320 35-0 64 33-88 || 41°50 | 53-18 66-40 || 0-447 || 0-548 || 0-702 || 0-876
1°325 35°4 65 34-35 || 42-08 || 53-92 || 67'33 || 0°455 || 0:557 || 0-714 || 0-892
1°330 35°8 66 34-80 || 42-66 54°67 | 68°26 || 0-462 0-567 || 0-727 | 0-908
1°335 36°2 67 35-27 || 43°20 || 55-36 69°12 || 0-471 0.577 || 0-739 || 0-923
1°340 36°6 68 35-71 43-74 56-05 || 69'98 || 0-479 || 0-586 || 0-751 || 0-938
352 MANUFACTURE OF SULPHURIC ACID
Specific Gravity of Sulphuric Acid Solutions—Continued.
100 parts by weight of chemically 1 Litre of chemically pure Acid
Sp. Gr pure Acid contain per cent. contains kilograms of
#. Degrees | Degrees
4° Baumé. Twaddell. SO HoSO *. d * O HoSO A. d Aft
º ** e Q\}\, Jºſie O O SO3. QYOV-J is O O
(in vacuo.) * | **** | 1.42. Tw. 106. Tw. | * * | 142*Tw.|10&Tw.
1°345 37-0 69 36-14 || 44-28 || 56-74 || 70-85 || 0:486 || 0°596 || 0°763 || 0°953
1°350 37-4 70 36-58 44-82 | 57-43 || 71-71 || 0:494 || 0-605 || 0°775 || 0-968
1°355 || 37.8 71 37-02 || 45-35 | 58-11 || 72°56 || 0-502 || 0-614 || 0-787 || 0°983
1°360 38-2 72 37-45 45.88 || 58-79 || 73°41 || 0-509 || 0-624 || 0-800 || 0°998
1°365 38-6 73 37-89 || 46-41 || 59°48 || 74°26 || 0-517 | 0:633 || 0-812 1*014
1:370 39-0 74 38-32 16-94 60°15 75-10 || 0-525 || 0°643 || 0-824 1*029
1°375 39°4 75 38.75 47-47 60-83 || 75-95 || 0-533 || 0°653 || 0°836 1*044
1°380 39 •8 76 39-18 48-00 | 61.51 76-80 || 0:541 || 0°662 || 0-849 || 1 °060
1-385 40 "1 77 39.62 || 48-53 62-19 77-65 || 0:549 || 0-672 0-861 || 1:07.5
1 "390 40°5 7S 40-05 || 49-06 || 62-87 78.50 || 0:557 0-682 || 0-873 || 1:091
1°395 40°8 79 40°48 || 49°59 || 63°55 79°34 || 0-564 || 0-692 || 0-886 || 1:107
1°400 41°2 80 40-91 50°11 || 64°21 80-18 || 0-573 || 0-702 || 0-899 || 1:123
1°405 41-6 81 41 °33 50-63 64 °88 81-01 || 0-581 || 0-711 || 0-912 || 1:138
1°410 42-0 S2 41-76 || 51°15 65°55 81-86 || 0-589 || 0°721 || 0-924 1*154
1° 415 42-3 83 42-17 | 51-66 | 66°21 82-66 || 0-597 || 0-730 || 0-937 || 1:170
1°420 42.7 84 42-57 52-15 | 66-82 83°44 || 0-604 || 0-740 || 0°949 || 1:185
1°425 43-1 85 42-96 || 52-63 67°44 | 84-21 || 0-612 0-750 0-961 || 1:200
1°430 43°4 86 43-36 53-11 | 68-06 | 84-98 || 0-620 || 0-759 || 0-973 || 1:215
1°435 43°8 87 43°75 53.59 | 68-68 || 85-74 || 0-628 || 0-769 || 0-986 || 1:230
1°440 44°l 88 44-14 || 54-07 || 69°29 86-51 || 0-636 || 0-779 || 0-998 || 1:246
1°445 44' 4 89 44'53 54-55 69-90 87-28 || 0:643 || 0-789 || 1:010 || 1:261
1°450 44 °8 90 44-92 || 55.03 || 70°52 | 88-05 || 0-651 || 0-798 || 1:023 || 1:277
1°455 45°1 91 45-31 || 55-50 71-12 || 88-80 || 0:659 || 0-808 || 1:035 | 1:292
1-466 45°4 92 45-69 || 55-97 || 71-72 | 89-55 || 0-667 || 0-817 | 1-047 || 1:307
1°465 45°8 93 46-07 || 56-43 || 72°31 || 90°29 || 0-675 || 0-827 | 1:059 || 1:323
1°470 46*1 94 46-45 56-90 | 72-91 91-04 || 0-683 || 0-837 | 1.072 | 1.338
1-475 46°4 95 46.83 57.37 | 73°51 | 91-79 || 0-691 || 0-846 || 1-084 || 1:354
1°480 46°8 96 47-21 57.83 || 74°10 92-53 || 0-699 || 0-856 || 1:097 || 1:370
1°485 47°l 97 47.57 58-28 || 74°68 93-25 || 0-707 || 0-865 || 1:109 || 1:385
1°490 47-4 98 47-95 || 58-74 || 75-27 | 93-98 || 0-715 || 0-876 || 1:122 || 1:400
1°495 47-8 99 48-34 59-22 || 75-88 || 94-75 || 0:723 || 0:885 || 1:134 || 1:417
1 "500 48°1 100 48-73 || 59-70 76°50 95-52 || 0-731 || 0-896 || 1:147 || 1:433
1°505 48°4 101 49-12 60-18 77°12 96-29 || 0:739 O'906 || 1:160 | 1.449
1°510 48°7 102 49-51 || 60-65 77°72 || 97-04 || 0-748 || 0-916 || 1:174 || 1:465
1°515 49'0 I03 49.89 61-12 || 78-32 97-79 || 0-756 || 0-926 | 1.187 || 1:481
1°520 49°4 104 50-28 || 61.59 || 78-93 98-54 || 0-764 || 0-936 || 1:199 || 1:498
I ‘525 49-7 105 50-66 || 62-06 || 79°52 99-30 || 0-773 || 0-946 || 1:213 1“514
1 °530 50-0 106 51-04 || 62.53 || 80-13 || 100-05 || 0-781 || 0-957 || 1:226 1-531
1°535 50-3 107 51-43 | 63-00 || 80-73 || 100-80 || 0-789 || 0-967 || 1:239 || 1:547
1°540 50°6 108 51-78 || 63-43 | 81-28 || 101-49 || 0-797 || 0-977 || 1:252 || 1:563
1'545 50°9 109 52-12 63.85 81-81 | 102-16 || 0-805 || 0-987 | 1.264 || 1:579
1°550 51-2 110 52.46 64.26 82°34 || 102-82 || 0-813 || 0-996 || 1:276 || 1:593
1555 51°5 111 52.79 64.67 82-87 | 103-47 || 0-821 || 1:006 || 1:289 || 1:609
1°560 51°8 112 53-22 || 65-20 | 83°50 | 104-30 || 0-830 || 1:017 | 1.303 || 1-627
1°565 52°1 113 53.59 || 65-65 | 84-08 || 105-03 || 0-839 || 1:027 | 1.316 || 1:644
1°570 52°4 114 53-95 | 66-09 | 84-64 105-73 || 0-847 1-038 | 1.329 1*660
1-575 52.7 115 54'32 | 66-53 85’21 | 106°42 || 0:856 || 1-048 || 1:343 || 1:677
1-580 53-0 116 54-65 | 66.95 || 85-78 107-10 || 0-864 || 1:058 1°356 || 1:692
1°585 53-3 117 55.03 || 67-40 || 86°34 || 107-85 || 0-872 || 1:068 | 1.369 || 1:709
1°590 53-6 118 55-37 || 67-83 || 86-88 || 108-52 || 0-880 || 1:078 || 1:382 || 1:726
1°595 53-9 119 55-73 | 68°26 || 87°44 || 109-21 || 0-889 || 1:089 | 1.395 || 1:742
1°600 54°l 120 56-09 | 68.70 | 88-00 || 109-92 || 0-897 || 1:099 || 1:409 || 1°759
1-605 54°4 121 56-44 || 69°13 | 88°55 || 110-61 || 0-906 || 1:110 || 1:422 || 1775

969.3 || 6.9L. & g89. I | 9/8. I | 98./7I 30.8LI | 0I-36 || 61.9/, 99T 038. I
689.3 | #9 I. 3 || I39. I | 3/8. I | 70. If I 9/./II | 06. It 80.9/ * º º tº g tº 638. I
Z89.3 || 8 FI.Z | 919. I | 898. I | 31.97I Ig. III || 01. I6 98.7/ #.99 838. I
g/9.2 złI.Z I/9. I | #98. I | 07-97L | 93. ZII | 09. I6 | 69.72 ‘’’ 138. I
999. & Q9I.Z 999. I | 093. I | 00.97I 86.9 II | 93. I6 67.7/ tº g tº 3.99 938. I
199.3 || 8&I.Z | I99. I | 993. I | 09.971 I9.9LI 00. I6 || 63.71 .99 I tº º º 938. I
Ogg. Z || 3&I.Z 999. I | 393. I | 83.9'ſ I | 98.91.I | 08-06 || &I-71, tº º º 3.99 738. I
gfg. & 9L.I.Z Igg. I 873. I | 96.77|| || 01.9LI | 09.06 || 96.8/. tº e g 838. I
g39.3 || 0II.Z | Zł9. I gig. I | #9. WI 78.9 II | 07-06 || 08.84 I.99 33.8. I
839.3 | #01.2 gi/9. I | Ifºg. I | 38.77|| 69.9LI | 03-06 | 89.8/. tº tº $ © tº wº I38. I
339.3 | 660. Z | 699. I | 888. I | 80.77I 88.9 II | 90-06 || Ig.8/. j9I 0.99 038. I
689.3 | p 10.8, 8I9. I | 333. I g9.3?I IZ-7II 9T-68 96.3/ 39T 8. 79 9I8. I
899.3 870. Z | 869. I g08. I | 8%. If I 91.8 II | 08.88 80.3/ 39T 9. P9 OI8. I
039.3 | 930.3 || ISg. I | I63. I | 91.07I 93.3TI 09.18 || 09. II, I9'I #.j9 908. I
30g. & g00.3 ggg. I | 1.13. I | 90.681 38. III || 36.98 || 96.0/. 09 I 3.59 008. I
6/7.3 || 886. I | 67 g. I g93. I | 80.88 I | 89.0LI 08.98 gy.0/. 69 I 0.59 961. I
ggſ. Z g96. I | #8 g. I Z93. I | p I. 18L | 38.60I | 01-98 || 96.69 89T A-89 06/. I
Zgiº. 3 || 1:56. I | 6 Ig. I | 0%. I 91.99 I g0.601 || 0 I-98 || 1.j.69 /9I 9.89 98/. I
Z07.3 836. I | #09. I | 833. I | 03.98 I | 13.80I | 09.78 || 86.89 99T 3.89 08/. I
988.3 || II6. I I67. I | 813. I | 87.78T 39. IOI 30.78 || 09.89 99T 0.89 9//- I
g98.3 | #68. I | 815. I | 103. I I IQ.88I I6.90I I9.88 || LI-89 #9 I 8. 39 0//- I
fºg. & 118. I g9%. I 96 I. I | 08.38T | Ig.90I I0.88 91-19 39T 9. 39 99/. I
Ić9.3 698. I Igy. I g8I. I | 06. IgT | 79.901 || 77.68 || 08.19 39 I 8.39 091. I
303.3 gi/8. I | 685. I g/L. I | 0%. Iglſ 80.90I 00.38 £6.99 Ig|I I. 39 g91. I
#83.3 | 668. I ZZiff. I g0I.. I | 67.08I 39.70I 99. I8 | 89.99 09:I 8. I9 09/.. I
g93.3 | FI8. I | 9 If. I | 99T. I | 6/.6%I g6.801 || 3I. I8 || 33.99 6FT 9. I9 gf/. I
/7Z.Z | 66/. I | #05. I | 9PI. I | 60.6%I | 88.80T | 89.08 || 98.99 8?I jº. I9 071. I
833. & #8/. I | 363. I 98.I.. I | 88.8%I | 38.301 || 73-08 || 09.99 /j/I I. I9 98/. I
60Z. & 69/. I I88. I | 1.3.I.. I 89.1%T | 93.30I | 08.6/ | ?I.99 97I 6.09 08/. I
I6I.Z | #g/. I 693. I | 8T.I.. I | 86.9%I 69. IOI 98.6/ | 84.79 gºl 9.09 93/. I
&1 I. 3 | 681. I | 1.93. I | 80 I. I | 13.9%I 8T. I0I | 66.8/, 87.79 ##I 7.09 031. I
#9T. & ga/. I 958. I | 660. I | 19.9&I 99.00I 87.8/, 10.79 gift I 3.09 9II. I
98.I.Z OI!. I fgg. I | 680. I | 98.7%I 00.00I | 70.8/, 01.89 Z; I 0.09 0Iſ... I
ZII.3, 969. I | 333. I 080. I | 91.7&I 77.66 || 09.1/, 98.89 If I 1.69 g01. I
00I. & I89. I | 3Ig. I | I/O. I | 17.8%I | 68.86 AI.// | 00.89 Of I 9.69 001. I
Z80.3 /99. I IO3. I &90. I | 1/.3%I 38.86 9/.9/ | #9.39 63.I 3.69 969. I
#90.3 || 399. I | 683. I | 890. I | 80.3%I //-/6 || 88.9/, 63.39 89 I 6.89 069. I
970. Z 839. I 8/Z. I gi/O. I | 88. I&I IZ. 16 f6.9/, 86. I9 A3 I /.89 989. I
630.3 g39. I | 89%. I g30. I | 09.0%I 69.96 || 09.9/ | 89. I9 99T #.89 089. I
&IO. & II9. I | 69%. I | 130. I I II.0&I 91.96 || 80.9/, 6%. I9 Q9I Z.89 919. I
g66. I | 869. I | 9%. I | LIO. I | 98.6LI 39.96 || 99.7/ I g6.09 W9I 6. /9 0.19. I
//6. I | #89. I | 093. I 600. I | 11.8TI 80.96 || 73.7/, I9.09 99T 1. A9 999. I
096. I | 019. I | 933. I | 000. I II. 8 [I #9.76 I8.8/, 93.09 33L jº. 19 099. I
fºé. I /gg. I g|IZ. I | 866.0 || 77. III || 30.76 || 07.8/ | 66.69 Ig|I I. 19 999. I
936. I 879. I #03. I | 886.0 || 3/.9TI g?.86 || 96.3/ | 19.69 03 I 6.99 099. I
606. I | 639. I | 36|I. I | g 16.0 || 90.9LI 36.36 99.3/ | 33-69 6&I 9.99 Qj9. I
Z68. I | 919. I Z8I. I 996.0 || 07.9 II | 88.36 ZI.Z/ | 88.89 8%I 8.99 079. I
g/8. I 309. I | a LI. I | 1.96.0 II. WII 88. I6 || 01. I/ | 89.89 /.3L 0.99 989. I
698. I | 687. I 39 I. I | 856.0 || 30. VII 63. I6 || 13. II, 81.89 9&I 8.99 089. I
Zī’8. I | 3/ſ. I IgE. I | OF6.0 || 93.8LI | #1.06 || 98.01 | #8. 19 Q&I 9.9% g39. I
g38. I 39%. I | IFI. I | Ig6.0 | 89.3 II | 03.06 || 37.0/ | 67. 19 f/3.I 3.99. 039. I
OI8. I | 677. I Igſ. I | 836.0 || 00. &II | 99.68 || 00.0/, 91. 19 3&I 0.9% 9I9. I
&61. I g3?. I | 0&I. I | WI6.0 || 08. III | 0I.68 99.69 61.99 3&I /...?9 0I9. I
º *AW *AA *AA
*ºol ºn *OSøH ‘8OS ºo: º *OS3H ‘8OS ‘(onopa wº)
pjoy plow plow prov Toppsall'oumvg | +
Sø9132CI soolgoCI ;
Jo sue13olpi supejudo *quao Jod uſequoopſow aind “IE) “ds
plow 9.1md AIIboguoud Jo 914; I I AIIgopulouq go quâ19A Aq S41ed 00I
£98
'pamuliuod—suolºnioS plov opinudins Jo A41Aero olypods
CIIC)V OIRIſl Hol'IſlS HO ALIAV Hº) OISHIO'HaS

3
5
4
MANUFACTURE OF SULPHURIC ACID
Specific Gravity of Sulphuric Acid Solutions—Continued.

100 parts by weight of chemically 1 Litre of chemically pure Acid
Sp. Gr. pure Acid contain per cent. contains kilograms of
# Degrees | Degrees
Baumé. Twaddell.
o so. Hso. " | * 1 so HøSO Aja - Aft
(in vacuo). 3• * |lºw.lio.ºrw. 3• * |142%rw.|106°rw.
1-831 65-5 e - e. 75-46 92-43 || 118-41 || 147-88 || 1:382 1-692 || 2:169 || 2:708
1°832 e ‘s e * - e. 75-69 92-70 118-73 || 148-32 || 1:386 1-698 || 2:176 2-717
1°833 65°6 tº º º 75°89 || 92-97 || 119-07 || 148-73 || 1:391 || 1 704 || 2:184 2-727
1°834 tº º 0 76-12 93-25 || 119°43 || 149-18 || 1:396 || 1 710 |2-191 || 2:736
1°835 65.7 167 76°38 || 93-56 || 119-84 149-70 || 1:402 || 1-717 || 2:200 || 2:747
1°836 e tº 76°57 || 93-90 | 120-19 || 150-08 || 1:406 || 1:722 || 2-207 || 2°755
1-837 º º 76-90 94-25 | 120-71 || 150-72 | 1.412 1-730 || 2:217 || 2:769
1°838 65°8 tº e Q 77°23 94°60 | 121*22 151°36 || 1:419 || 1 °739 || 2:228 2°782
1°839 - - - 77°55 95-00 | 121-74 152-00 || 1:426 || 1:748 || 2:239 || 2-795
1-840 || 65.9 iás | 73-04 || 35-30 | 122.5i is?-96 | 1.436 || 1759 2.254 || 2:314
1°8405 e - e. 78-33 95'95 || 122°96 || 153°52 || 1:451 || 1:765 2°262 2°825
1-84.10 tº e te • * tº 78-69 || 96°38 123°45 154°20 || 1:448 1-774 || 2:273 || 2:838
I '8415 e tº e e - - 79°47 97°35 | 124-69 155-74 || 1:463 1792 || 2:296 || 2-867
I '84.10 tº tº o * - a 80-16 || 98-20 | 125-84 || 157-12 || 1:476 1-808 || 2:317 || 2:893
1-8405 e tº e * * * 80°43 || 98°52 | 126'18 157-62 || 1:481 || 1 '814 || 2:325 2'903
1°8400 © tº º * - e. 80-59 || 98.72 | 126°44 157-94 || 1:483 || 1:816 || 2:327 | 2'906
1°8395 e is tº • e e 80°63 || 98°77 | 126°50 | 158°00 || 1:484 || 1:817 || 2:328 2'907
1°8390 tº ſº & * * * 80'93 99°12 || 126-99 || 158-60 || 1'488 || 1-823 2-336 2.917
1°8385 tº e Q © tº º 81-08 || 99-31 | 127°35 | 158-90 || 1:490 | 1.826 || 2:339 2.921
A table giving the relationship between the percentage content and
the readings of the hydrometer employed in the United States has been
published by Elliot."
The tables calculated by Richmond* and by Marshall 3 from
Pickering's measurements," and which they regard as very exact, are not
really so, since Pickering only employed a relative and not an absolute
unit for his determinations; the figures consequently require to be
corrected by multiplying by the factor O.9974, as given by the German
Standards Commission.”
Temperature Correction.
Since in practice the specific gravity determinations are only seldom
made at exactly I 5°C., it is necessary to have a means of correcting
readings made at a higher or lower temperature.
Lunge has compiled a table, based on an extended series of observa-
tions, for this purpose. An abridged form of the complete table is given
below."
A very complete table has also been published by the German
Standards Commission.
* J. Soc. Chem. Ind., 1898, 17, 45. * Ibid., 1890, 9, 479. * /bid., 1899, Io, 6.
*J. Chem. Soc., 1890, 57, 63 and 331. * Report, 1904, p. 221.
* The complete table is given in Zhe Alkali Makers' Handbook, Lunge and Hurter, 2nd
edition, pp. IoS-Io8,
SPECIFIC GRAVITY OF SULPHURIC ACID 355
Lunge's table agrees very well with the latter down to a sp. gr.
I. IOO = about 15 per cent. H2SO4. The table is inapplicable for weaker
acids, owing to an error which cannot now be traced ; it is,
however, seldom likely to be required for such weak acids in
practice.
Table showing the Influence of Temperature on the Specific
Gravity of Sulphuric Acid.
a. Specific gravity at I 5"/4°; under t, changes in specific gravity at the temperature t.

t. t. t t t. t. t
C!. 0° 10° 20 30° 40° 50° 60
1°840 + 0°015 + 0°005 — 0°005 — 0°015 — 0°025 — 0°034 — 0°044
1-820 16 5 5 16 26 37 47
1-800 17 5 5 16 27 37 47
1.780 17 5 5 16 27 37 47
1:760 I6 5 5 16 26 36 47
1-740 16 5 5 15 25 35 45
1.720 15 5 5 15 25 35 44
1-700 15 5 5 14 24 33 43
1-680 15 5 5 14 24 33 42
1°660 14 5 5 14 23 32 41
1 *640 14 5 4 14 23 32 40
1-620 14 4 4 14 22 31 40
1-600 14 4 4 13 22 31 39
1-580 14 4 4 13 22 30 39
1-560 13 4 4 13 21 30 38
1°540 13 4 4 13 21 30 38
1°520 13 4 4 13 21 29 37
1°500 13 4 4 12 21 29 37
1°480 13 4 4 12 20 28 36
1 °460 I2 4 4 12 20 28 36
1°440 12 4 4 12 20 28 35
1°420 12 4 4 12 19 27 35
1°400 12 4 4 12 19 27 34
1°380 12 4 4 11 19 27 34
1°360 11 4. 4 11 19 26 34
1°340 11 4 4 11 19 26 33
1 *320 11 3 4 11 18 26 33
1°300 11 3 3 11 18 26 33
1 °280 11 3 3 11 18 25 33
1°260 11 3 3 11 18 25 32
1*240 11 3 3 10 18 24 32
1*220 10 3 3 10 17 24 31
1°200 10 3 3 10 17 23 30
1-180 10 3 3 10 16 23 29
1°160 9 3 3 9 15 22 28
1°140 8 3 3 8 14 20 27
1-120 8 2 2 8 14 19 25
1°100 7 2 2 7 13 18 24
1*080 6 2 2 7 12 17 23
1*060 5 2 2 6 10 16 21
1*040 3 1 1. 5 9 14 20
1*020 2 1 1 4 8 13 18
1*010 2 I I 4 7 12 17
For temperatures below 15° the values given in the column t must,
of course, be subtracted from the observed reading; for temperatures
356 MANUFACTURE OF SULPHURIC ACID
above 15° the corrections must be added to give the value at 15°. A
table has also been calculated by Fuchs" for this correction.
The following table, p. 357, worked out by the Chemische Fabrik,
Griesheim, will be found useful in many cases.
It is to be noted that the values given in all the above tables only
hold good for chemically pure acid. In commercial acids the specific
gravities at the highest concentrations are appreciably higher than the
values given in the tables, but the variations between acids made at
different works are too irregular to permit of a table being prepared for
such acids. The specific gravity is influenced in the case of nitrous vitriol
by the percentage of nitrous acid present, and in the case of chamber
and concentrated acids by the presence of sulphurous acid, lead sulphate,
nitrogen oxides, arsenic, and iron. The quantity of these usually present-
in sulphuric acid is too slight to markedly affect the specific gravity;
nevertheless exceptional cases may occur in which the acid is strongly
contaminated with iron, aluminium, or sodium salts. The iron may
result, for example, from pyrites dust, the aluminium from the packing
of the Glover tower, or from the fireclay frequently employed for
temporarily repairing leaks, the sodium salts from solutions of Chili
saltpetre or Glauber's salts, which sometimes get into the chambers
through carelessness. Nitric acid, which occurs in appreciable quantity
in sulphuric acid recovered from spent, nitrating acids, also causes an
increase in the specific gravity.
The influence of impurities on the specific gravity of sulphuric acid
has been dealt with by Lunge,” also in a recent paper by Marshall.”
and in special detail by the German Standards Commission.” According
to the Commission the Glover acid, as might be expected, is the most
impure, the differences between the actual percentages found by analysis
and those obtained by calculation from the tables varying from O. 5 to
1.87 per cent. “Commercially pure” chamber acid showed differences
of O-O I to O-26 per cent. ; acids of 142° Tw., of (exceptionally) O-O3 to
1.55 per cent. ; acids of 168° Tw. made from chamber acid, differences
of O-24 to O-76 per cent, in One instance only O.O2 per cent. ; whilst
acid free from arsenic differed from the actual percentage by from oog
to o-65 per cent. Acids made by the contact process showed differences
of only O-O2 to O-23 per cent, and acid manufactured from sulphuretted
hydrogen of only O.O2 to O. 18 per cent.
Since the changes in specific gravity with strength are very slight
in the case of high-strength acids, and as it is just in these acids
that the impurities, always present in commercial acids, have the
greatest effect on the specific gravity, which is increased by their
presence, the tables should only be used for guidance in the works
* Z. angew. Chem., 1898, II, 950. *J. Soc. Chem. Ind., 1902, 21, 1503.
* Sulphuric Acid and Alkali, vol. i., p. 189 et seq. * Report, 1904, 5, p. 243 et seq.
Reduction of the Hydrometer Readings for Sulphuric Acid between 65° and 66 B.
(164° and 168° Tw.) to 15° C.
The hydrometer reading is taken from the first vertical column, and the observed temperature from the top line. The reading
which lies vertically below the observed temperature and on the line containing the observed specific gravity gives
the specific gravity at 15°.
65-00 |64'80 64'84 64°88 |64'92 || 64-96 || 65-00 || 65-04 || 65-08 || 65-12 |65-16 || 65-20 65-24 || 65-28 || 65°32 |65-36|| 65-40 || 65°44 || 65°48 || 65-52 65-56 |65-60
65-10|64-90 64'94 || 64'98 |65-02 |65-06 |65-10 |65-14|65-18|65-22 || 65-26 |65-30 || 65.34 || 65.38|65-42|65-46 |65'50|65'54 || 65°58|65°62|65-66 |65-70
65-20165-00 || 65-04 || 65-08 || 65-12 || 65-16 || 65-20 || 65.24 || 65-28 || 65-32 || 65-36 |65-40 || 65-44 |65-48 || 65-52 || 65-56 |65-60 |65:64 || 65-68 || 65-72 || 65-76 || 65-80
65-30 |65-10 || 65°14 || 65-18 || 65-22 || 65-26 65-30 || 65-34 65-38 |65-42|| 65-46 65-50 || 65-54 || 65-58 || 65-62 65-66 || 65’70 |65’74 65°78 || 65'82 |65'86 || 65'90
65°50′ 65°30 65°34 65°38 65-42 65-46 65-50 65-54 65-58 65-62 65-66 65-70 65-74 65-78 65-82 65-86 65-90 65-94 || 65'98 || 66-02 | 66-06 | 66°10
65-60165-40 || 65°44 65-48 || 65-52 65-56 || 65-60 |65-64 || 65-68 || 65-72 65-76 65-80 || 65-84 || 65-88 || 65'92 || 65-96 | 66-00 | 66-04 ||66°08 | 66°12|66°16 | 66°20
65°70|65°50 || 65°54 || 65-58 65°62 65-66 || 65-70 65-74 || 65-78 || 65-82 65-86 65-90 65-94 || 65-98 66:02 66-06 | 66°10 66-14 | 66°18 66-22 | 66°26 | 66°30
65-80 || 65°60 65°64 65-68 65-72 || 65-76 65-80 || 65-84 || 65-88 65'92 || 65-96 || 66-00 | 66-04 || 66-08 ||66-12 | 66-16 | 66-20 | 66-24 66°28 66°32 66°36 | 66°40
65°90 |65’70 || 65’74 || 65°78 || 65-82 || 65-86 || 65:90 65-94 || 65'98 |66'02 | 66-06 ||66-10 |66-14 | 66-18|66-22 | 66°26 | 66-30 | 66°34 | 66°38 |66°42 | 66°46 | 66°50
66-00 || 65°80 || 65°84 || 65'88 || 65'92 || 65°96 | 66-00 | 66-04 || 66-08 || 66°12 | 66°16 | 66°20 | 66°24 | 66-28 || 66°32 | 66°36 | 66-40 | 66°44 | 66°48 || 66°52 66'56 | 66'60
65-40 |65°20' 65:24|65-28 |65°32' 65:36|65-40 |65.44|65-48 |65-52|65-56 |65-60 |65-64|65-68|65-72|65-76 |65-80 |65'84 |65'88 |65'92 |65-96 |66-00 |
§






358
MANUFACTURE OF SULPHURIC ACID
-
for acids of over 90 per cent. strength; for selling purposes, actual
analyses should always be made.
Melting Points of Sulphuric Acid. Knietsch.”
Enlarged by Lunge by the addition of the corresponding percentage of H2SO4.
The melting points given in the table are the temperatures which remain constant
during the period of complete crystallisation ; that is, from the temperature at
which crystals begin to appear in the cooled acid until the stage when the mass



becomes solid after removal from the freezing mixture.
Sulphuric Acid.
Total HoSO. Melting- Total HoSO Melting- Total HoSO Melting-
pºt. pºt. Point. pºt. lº,t. Pogº. º*t. rº. Point.
I 1-22 — 0°6 23 28°17 — 40° 1 80 98-00 + 3-0
2 2°45 — 1-0 |Under 81 99°25 + 7-0
3 3:67 – 1-7 e - tº tº e 6 — 40 °0 81-63 || 100-00 + 10°0
4 4°90 — 2-0 61 74-72 — 40°0 82 tº º º + 8°2
5 6-12 - 2-7 62 75°95 – 20 °0 83 – 0°8
6 7-35 – 3°6 63 77-17 – 11°5 84 – 9°2
7 8-57 — 4°4 64 78’40 — 4°8 85 – 11°0
8 9°80 – 5 °3 65 79-62 — 4°2 86 – 2°2
9 11*02 – 6°0 66 80-85 + 1-2 87 + 13°5
10 12:25 – 6-7 67 82°07 + 8-0 88 + 26-0
11 13°47 — 7-2 68 83°39 + 8-0 89 + 34°2
12 14-70 – 7-9 69 84°52 + 7-0 90 + 34°2
13 15°92 — 8°2 70 85-75 + 4*0 91 +25°8
14 17-15 – 9°0 71 86-97 — 1 °0 92 + 14°2
15 18°37 – 9°3 72 - 88-20 – 7-2 93 + 0°8
16 19°60 – 9 °8 7 89°42 – 16°2 94 + 4*5
17 20 '82 – 11°4 74 90°65 – 25 °0 95 + 14°8
18 22*05 — 13°2 75 91.87 — 34°0 96 +20°3
19 23-27 – 15°2 76* 93°10 – 32-0 97 + 29°2
20 24°50 – 17-1 77* 94°83 – 28-2 98 + 33°8
21 25-72 – 22°5 78* 95°05 – 16*5 99 + 36°0
22 26°95 – 31°0 79 96-77 — 5°2 100 +40°0
* So-called 168° Tw.
Boiling Points of Sulphuric Acid.” \
<!--> * o S. 2-d; in . +5 o S. Uſ? •q.) s +: ** 9 v, -aš º
§ | # | # | ##3 | #; # | # | #3 |##| # # |##3
:3, #3 §§ #5: ...?. #; #3 #5: :3 | #3 #; #53
#r. JD Co f; §P- #5 UD Cºy ñā §2. #. āş §§ flº-
5 1*031 4-2 101 56 1°459 || 45°4 || 133 82 1°758 62°2 218.5
10 1*069 9-2 102 60 1-503 || 48-3 || 141-5 | 84 || 1°773 || 63-0 || 227
15 I-107 || 13-9 103-5 || 62°5 1°530 50-0 || 147 86 || 1-791 || 63-8 || 238°5
20 1-147 | 18-5 105 65 1°557 || 51-6 || 153-5 || 88 1-807 || 64°4 || 251-5
25 1*184 || 22-4 106-5 || 67°5 || 1:585 || 53-3 | 161 90 | 1.818 65-0 || 262-5
30 1*224 || 26°4 I08 70 1 *615 55-0 || 170 91 || 1 °824 || 65°3 || 268
35 I 265 || 30-2 110 72 1 °639 56°3 174°5 || 92 || 1:830 65°45 274-5
40 1°307 || 33-9 114 74 1°661 57°4 180°5 || 93 || 1 '834 65-65 28.1°5
45 1°352 || 37-6 118-3 || 76 1 °688 || 58 °8 || 189 94 || 1°837 65°8 || 288'5
50 I “399 || 41°1 124 73 1°710 || 60-0 || 199 95 || 1 °840 || 65°9 295
53 1°428 43°3 128°5 || 80 1°7'33 61-0 || 207 tº º º º º º e & © e - B
* Aer, 1901, 34, 41oo.
* Lunge, Ber., 1878, II, 370.
IMPURITIES IN SULPHURIC ACID - 359
QUALITATIVE EXAMINATION OF SULPHURIC ACID FOR IMPURITIES
The impurities likely to be present in ordinary commercial sulphuric
acid are:—Sulphates of sodium (less frequently potassium), ammonium,
calcium, aluminium, iron and lead, exceptionally also of zinc and
copper; arsenic, selenium, thallium, titanium, nitrogen oxides, hydro-
chloric acid, sulphurous acid, and hydrofluoric acid.
Acidum sulfuricum purissimum is, according to Krauch, tested for
fixed residue, nitric acid, Selenium, reducing substances, lead, other
metals, arsenic, ammonia, and the halogens.
General examination for gaseous impurities (Warington). Two
kilos of the undiluted acid are thoroughly shaken in a flask, of a
capacity equal to twice the volume of the acid, so as to saturate the
air in the flask with the gases dissolved in the acid. The atmosphere
in the flask is then tested for sulphurous acid by iodised starch paper
and for gaseous oxides of nitrogen by potassium iodide starch
paper. Any blue coloration produced by the latter gases will not be
destroyed by sulphurous acid unless this is present in considerable
excess. Sulphuretted hydrogen reacts like sulphurous acid in this
test.
Sulphurous acid. This gas will decolorise a starch solution rendered
faintly blue by iodine ; or it may be converted into sulphuretted
hydrogen by means of zinc or aluminium, and examined by lead acetate
paper or by an alkaline solution of sodium nitroprusside. The latter is
a very delicate test.
Hydrochloric acid. Two g. are diluted to 30 c.c., and a few drops
of silver nitrate solution added; in the case of acid, sulfuric, puriss,
no turbidity should result. The hydrochloric acid present in ordinary
commercial vitriol arises from the sodium chloride present in the
nitre.
The qualitative examination for traces of Nitrogen Acids is best made
with diphenylamine. The diphenylamine is dissolved in about 100
parts of pure sulphuric acid, or failing this, in an acid freed from
nitrogen compounds by boiling with the addition of a very small
amount of ammonium sulphate, and the solution diluted with about tº
of its volume of water. The solution may be used at once, or may be kept
if desired : it becomes, however, discoloured and less sensitive after some
time. In testing concentrated sulphuric acid for nitrogen acids about
2 C.C. of the acid are placed in a test-tube or conical test-glass, and about
I c.c. of the diphenylamine solution added carefully so that the two
layers only mix slowly; in the case of weaker acids or other solutions of
lower specific gravity the order is reversed, and the heavier diphenyl-
amine solution placed below. The slightest trace of nitrogen acids is
* The Testing of Chemical Reagents for Purity, p. 305.
360 MANUFACTURE OF SULPHURIC ACID
shown by the development of a beautiful blue colour at the junction of
the two solutions."
In the presence of Selenium, which also gives the above reaction with
diphenylamine, nitrogen acids, if present in somewhat large quantity,
may be recognised by the decolorising effect on indigo solution, whilst if
only present in traces they may be detected by the reddening produced
in a solution of brucine sulphate. Selenium is detected by the formation
of a brownish-red precipitate on the addition of a concentrated solution
of ferrous sulphate to the vitriol; this precipitate cannot be confused
with the simple coloration produced by nitric oxide. The detection of
nitrogen acids by means of ferrous sulphate has been described above
(p. 341).
Mitrous acid can be detected by a number of very delicate reactions,
which are not given by nitric acid. These include the blue colora-
tion produced in potassium iodide starch solution, or still better in zinc
iodide starch solution. Its presence is detected with especial ease by
the formation of azo-colours, a reaction first discovered by Griess,
who recommended the use of metaphenylenediamine, which gives rise
to a yellow coloration, and also the still more sensitive reaction with
Sulphanilic acid and a-naphthylamine, which produces a rose coloration.
The latter reagent has the disadvantage that the naphthylamine solution,
even when prepared from a perfectly white salt, becomes dark coloured
after a relatively short time, and as a result loses its sensitiveness.
Further, in very dilute solutions, such as I in IOOO millions, the
reaction takes place so slowly that it is impossible to be quite certain
that the coloration arises from the substance tested and not from
nitrous acid present in the air; the time may be appreciably shortened
by warming the solution, but even in this case fifteen to twenty minutes
may be necessary. -
Ilosvay” found that by using acetic acid instead of sulphuric acid or
hydrochloric acid, the time required for the reaction is very much
decreased, and that at the same time the colour is developed to a much
greater extent. He overcomes the difficulty of the discoloration of the
naphthylamine by adopting the following method of preparing the
solution :-0.5 g. of sulphanilic acid is dissolved in 150 c.c. of dilute
acetic acid and o. 1 g of solid naphthylamine is boiled with 20 c.c. of
water, the colourless solution poured off from the bluish-violet residue,
and 150 c.c. of dilute acetic acid added to this solution. In testing, a few
c.c. of the sulphanilic acid solution are added to about 20 c.c. of the
solution to be examined, the mixture warmed to 70° to 80°, and the
naphthylamine solution then added. In the presence of one part of
nitrous acid in IOOO million parts of solution the red coloration, due to
the azo-colour produced, appears in about a minute; if the nitrous acid
* Cf. Lunge, Z, angew. Chem., 1894, 7, 345. * Bull. Soc. Chim., 1889 [3], 2, 317.
IMPURITIES IN SULPHURIC ACID 361
be present in relatively large amount, about I : IOOO, only a yellow solu-
tion is obtained unless a concentrated solution of naphthylamine be
employed.
Lunge 1 mixes the solutions of sulphanilic acid and naphthylamine
prepared according to Ilosvay's directions, together, and keeps the
mixture in a well-stoppered bottle. It is unnecessary to exclude light
from the mixed solutions, but it is essential to prevent access of air, since
this may possibly contain nitrous acid as an impurity. By thus combin-
ing the two reagents in a single solution, any contamination by nitrous
acid from the atmosphere is at once evidenced by the red colour of the
solution. A solution that has become red may be very quickly rendered
fit for use by shaking with zinc dust and filtering. With the reagents
ready mixed, the reaction may also be hastened by warming the solution
to 70° to 80°. (Cf. infra, under the quantitative application of the
method.) -
Riegler” recommends the use of naphthionic acid or sodium
naphthionate and 8-naphthol, but this has no advantage over Griess'
reagents, and according to Riegler himself is not so sensitive (I : IOO
million). This reagent can be used colorimetrically.
H. Erdmann & employs sulphanilic acid in hydrochloric acid solution,
or, still better, p-amino-benzoic ester, followed by the acid sodium salt of
I-amino-8-naphthol-4-6-disulphonic acid (also known as amino-naph-
thol-K acid, and sold under the name “Water-analysis reagent—
Bagdad”). The coloration produced by this reagent is a bright wine-
red. Mennicket has defended this method against certain adverse criti-
cism, and states that one part of sodium nitrite in 2000 million parts of
water can be detected by its aid.
Brucine in sulphuric acid solution reacts only with nitric acid and
not with nitrous acid when sulphuric acid is present in large excess;
nitrous acid only gives a coloration if the solution contains at least two
parts of water to each part of sulphuric acid. Consequently, to test only
for nitric acid, the solution should contain sulphuric acid equal to 2/3 of its
volume, and the test be carried out by adding I c.c. of a solution of O-2 g.
brucine in IOO c.c. strong sulphuric acid, to 50 c.c. of the solution to be
examined. In the presence of O.OI mg. nitrate-nitrogen, a red colour—
subsequently passing through orange to a golden yellow *—results.
Hydrofluoric acid may be detected by warming the acid in a
platinum dish, covered with a glass plate coated with wax in which
figures have been scratched.
Ammonia. Two g. of the acid are diluted with about 30 c.c. of
water, the solution made alkaline by the addition of a potassium
* Z. angew. Chem., 1889, 2,666. * Z, angew. Chem., 1900, 13, 33.
* Z. anal. Chem., 1896, 35, 677 ; 1897, 36, 306, 377, 665. * /bid., 1900, 13, 235 and 7II.
* Cy Winkler, Z, angew. Chem., 1901, I4, 17o ; and, Lunge, ibid., 1901, 14, 241.
362 MANUFACTURE OF SULPHURIC ACID
hydroxide solution containing 3 to 4 g. pure hydroxide, and ten to fifteen
drops of Nessler's reagent added; no distinctly yellow or brownish-red
coloration should result.
Krauch states that by this method I mg. NHs in IOO g. concen-
trated sulphuric acid will produce a distinct yellow coloration and
turbidity.
Solid impurities. Lead is shown by any turbidity which results on
the addition of five volumes of strong alcohol to one volume of the
acid; simple dilution with water is sufficient if the lead be present
in considerable quantity. The presence of lead may be confirmed by
examining the precipitate in the blowpipe flame, etc.
Iron. The acid is boiled, after the addition of one drop of pure
nitric acid, diluted slightly, and when cold, treated with an excess of
thiocyanate solution. A blank test should always be made with the
nitric acid employed, to make sure that any red coloration produced is
not due to this.
Venable 1 employs a mixture of cobalt nitrate and strong hydro-
chloric acid. Traces of ferric salts change the blue colour of the solu-
tion to green; ferrous salts do not affect it.
Selenium is recognised by the red coloration and subsequent red
precipitate produced by the addition of ferrous sulphate to the solution;
the reaction is hastened by warming. Sulphurous acid is a better re-
agent than ferrous sulphate for this test. The presence of Ooi per cent.
of selenium is shown, after some hours, by either reagent. Codeine may
also be used, but it is less sensitive than the preceding reagents, requir-
ingo. 5 per cent, of selenium to show a reaction.” Selenic acid cannot be
tested for by any of these reagents, but it may be detected by means of
acetylene, which also reacts with selenious acid, and produces a red
coloration in the presence of only O-OOI per cent. of selenium ; a little
hydrochloric acid hastens the separation of the selenium, which is
soluble in hot sulphuric acid, forming a green solution.
DETECTION AND APIPROXIMATE ESTIMATION OF ARSENIC
Examination for this impurity is always important, and it is abso-
lutely necessary in all cases where the acid is sold as pure, or is
intended for use in the preparation of substances such as glucose,
tartaric acid, mineral water, yeast, etc., which either directly or in-
directly may be used as food-stuffs. Sulphuric acid, prepared from
pyrites, unless specially purified, usually contains from O. I to O.2 per
cent, of arsenious oxide, and, in exceptional cases, up to 1 per cent.
and more.
* Z. anal. Chem., 1889, 28, 699.
* Jouve, Chem. Centr, 1901, I., 1389; and, Orlow, ibid., 1901, I, 480.
DETECTION OF ARSENIC 363
Very special attention has been directed to the presence of arsenic
in sulphuric acid, owing to the numerous and in some instances fatal
cases of arsenical poisoning that occurred in the year 1900. It was
found that these cases arose from the consumption of beer, in the brewing
of which glucose containing arsenic had been employed, the arsenic in
the glucose being traced in turn to the sulphuric acid used in its manu-
facture. The investigation into the origin of this poisoning led to the
publication of much important work on the detection and estimation of
arsenic in various materials."
Of the various methods of testing for arsenic, that of Marsh, more
recently and correctly known as the Marsh-Berzelius test, is by far the
most generally employed; the Reinsch and the Gutzeit methods are
also largely used.
As a sequel to the outbreak of arsenical poisoning referred to
above, a joint committee of the Society of Chemical Industry and of
the Society of Public Analysts was appointed in 1901, to investigate
the various methods for the detection and approximate estimation of
arsenic in beer, brewers' materials, food-stuffs, and fuels;” this com-
mittee reported in favour of the Marsh-Berzelius test as the most
reliable and generally applicable method. g
According to Hehner,” ordinary pyrites vitriol contains on the average
O.2 per cent of arsenious acid (As,Os); such acid, after purification, he
considers permissible for use in the preparation of foods and for the other
purposes mentioned above, provided that the arsenic left in does not
exceed O.O.5 mg. As,Cs in IO g. of the acid, that is, one part As,Cs in
2OO,OOO parts of acid. Very frequently the amount of arsenic remain-
ing after purification is less than this limiting value, but no acid
prepared from pyrites, and perhaps not even that from Sicilian brim-
stone, is absolutely free from arsenic; on the other hand, such freedom
might be expected in acid made by the “contact process” when platinum
is employed as the contact material. Acid prepared by the ferric oxide
contact process is not free from arsenic (Conroy). Arsenic present to
the extent of O.OO1 mg. may be detected and estimated without special
difficulty.
A. The Marsh-Berzelius Test.
The actual test, as proposed by Marsh in 1827, consisted in the
formation of a dark stain on a piece of cold porcelain brought into a
burning mixture of hydrogen and arseniuretted hydrogen. For more
exact purposes, this is generally combined with the Berzelius reaction
by heating the arsenical hydrogen gas during its passage through a
' Cf. Report of the Royal Commission on Arsenical Poisoning, 1903; J. Soc. Chem. Ind.
I904, 23, I 59.
* The full Report of the Committee is published in the J. Soc. Chem. Ind., 1902, 21, 94.
* / Soc. Chem. Ind., 1901, 20, 188.
364 MANUFACTURE OF SULPHURIC ACID
glass tube, and so effecting a decomposition of the arseniuretted
hydrogen with the separation of the arsenic in the form of a mirror.
The following are the details and conditions for carrying out the
test as recommended by the joint committee of the Society of Chemical
Industry and the Society of Public Analysts.
The materials required are:—
I. Hydrochloric acid. The purest hydrochloric acid obtainable is very
rarely free from arsenic. To the “pure” acid as purchased for analysis,
diluted with distilled water to a sp. gr. of I. IO, sufficient bromine is
added to colour it strongly yellow (about 5 c.c. per litre); Sulphurous
acid, either gaseous or in aqueous solution, is then added in excess,
and the mixture allowed to stand for at least twelve hours; or hydro-
bromic acid and sulphurous acid may be used. The acid is then
boiled till about one-fifth has evaporated, and the residue can either be
used direct or it may be distilled; the whole of the arsenic is volatilised
with the first portion.
- 2. Sulphuric acid. This is more frequently obtainable arsenic-free
than hydrochloric acid. If not procurable, to about half a litre of
sulphuric acid, “pure for analysis,” a few grams of sodium chloride
are added and the mixture distilled from a non-tubulated glass retort,
the first portion of the distillate, about 50 c.c., being rejected. For
the purpose of the test, one volume of the distilled acid is diluted with
four volumes of water.
3. Mitric acid can, as a rule, be obtained free from arsenic, without
much difficulty, the pure redistilled acid being used. This should be
tested by evaporating 20 c.c. in a porcelain dish, which should then be
washed out with dilute acid, and tested as described below.
The purified acids should be prepared as required, and should not
be stored for any length of time. If this be unavoidable, however,
Jena flasks are to be preferred, since most bottle glass is liable to com-
municate traces of arsenic.
4. Zinc, Arsenic-free zinc is obtainable from chemical dealers.
It should be re-granulated by melting and pouring it from some
height into cold water. A. H. Allen holds it to be essential, both for
a regular evolution of hydrogen and for the formation of uniformly
deposited brown-coloured mirrors, that the zinc should contain a trace
of iron.
5. Calcium hydroxide. Even when made from white marble this
is not always free from arsenic; a selection must, therefore, be made
from various samples. If pure lime is not obtainable, magnesia may
equally well be used, and can more readily be obtained of sufficient
purity.
6. Calcium chloride. This salt often contains arsenic, and before being
used as a drying agent must be freed from the volatilisable part of the
APPROxIMATE ESTIMATION OF ARSENIC 365
impurity by moistening it with strong hydrochloric acid, fusing, and re-
granulating.
The apparatus to be used is shown in Fig. 128.
A bottle or flask, a, holding about 200 c.c. (for frothing materials this
should be preferably wider at the top than at the bottom) is fitted with a
doubly bored cork, or rubber stopper, or with a ground-in glass connection
carrying a tap funnel, b, holding about 50 c.c., and an exit tube, c. The
latter is connected with a drying tube, d, containing first a roll of blotting-
paper, e, soaked in lead acetate solution and dried, or a layer of cotton wool
w - §lº:
!. -
FIG. 128.
prepared in a similar way, then a wad of cotton wool, f, then a layer,
g, of granulated calcium chloride, and finally a thick wad, h, of cotton
wool. To this tube is fitted a hard glass tube, i, drawn out as shown
in the figure to a thinner tube, Æ, and of such external diameter that
at the place where the arsenic mirror is to be expected, the tube just
passes through a No. 13 Birmingham wire gauge (corresponding with
olog2 inch=2.34 mm.). The exact size is not material, but all tubes
used for standards and tests should be as nearly as practicable of the
same diameter. A good Bunsen flame is used to heat the hard glass
tube close to the constriction. About one inch of tube, including the
shoulder, ought to be red hot. A piece of moderately fine copper gauze,

366 MANUFACTURE OF SULPHURIC ACID
about one inch square, wrapped round the portion of the tube to be
heated, assists in ensuring an equal distribution of heat.
Method of testing. About 20 g. of zinc are placed in the bottle, a,
and washed with water to clean the surface, as particles of dust may
contain arsenic; all the parts of the apparatus are connected up, and a
sufficient quantity of acid, prepared as described, is allowed to flow from
the funnel, à, so as to cause a fairly brisk evolution of hydrogen. When
the hydrogen flame, which during the heating of the tube, i, should be
kept at as uniform a height as possible (about a quarter of an inch),
burns with a round, not pointed tip, all air has been removed from the
apparatus. The plug in d prevents the flame from striking back. The
burner is then placed under the hard glass tube as described, and more
acid (IO to 20 c.c. is generally enough) run in as required. With good
materials no trace of a mirror is obtained within half an hour. Great
care must be taken that when additions of acid are made to the zinc,
no air-bubbles are introduced, since in presence of air the arsenic
mirror may become black and unevenly distributed; when the experi-
ment is properly conducted the mirror is brown. Should the blank
experiment not be satisfactory, it must be ascertained, by changing
the materials methodically, whether the fault lies with the acid, zinc, or
other materials, or with the apparatus.
Preparation of standard mirrors. When a satisfactory blank experi-
ment has been obtained, a series of standard mirrors are prepared under
the following conditions:—
A hydrochloric acid solution of arsenious oxide, containing oool mg.
As Os per c.c., is prepared by diluting a stronger solution with distilled
water. Two c.c. of this Solution, equal to O.OO2 mg. of arsenious oxide,
are introduced into the apparatus, a new tube, ié, having been joined to
the drying tube. If the zinc is sensitive, a distinct brown mirror is
obtained after twenty minutes. It is important to note that some “pure”
zinc is, from a cause at present unknown, not sufficiently sensitive; that
is to say, the addition of minute quantities of arsenic produces no
mirror. The portion of the tube, ik, containing the mirror is sealed
off while still filled with hydrogen; in contact with air the mirrors
gradually fade. Mirrors are then similarly made with O.OO4, O.OO6, O.OO8,
and O.OI mg. of arsenious oxide. With a little practice it is easy to
obtain the deposits of arsenic neatly and equally distributed. The
standard mirrors, properly marked, are mounted on a white card or
porcelain slip. It is important to bear in mind that the first stage of
every test must be a blank of at least twenty minutes.
Hydrochloric acid is somewhat more sensitive than sulphuric acid;
that is to say, it gives rather denser mirrors with minute quantities of
arsenic. If for any reason, sulphuric acid is preferred, the set of
standard mirrors must be prepared with this acid.
APPROXIMATE ESTIMATION OF ARSENIC 367
Organic materials, such as beer, yeast, etc., cannot be tested, when
sulphuric acid is used, without destruction of the organic matter, whilst,
as a rule, they can be directly tested with hydrochloric acid. However,
many materials are met with in which it is preferable to destroy the
organic matter.
Procedure zwithout destruction of organic matter. The apparatus is
started and a blank experiment allowed to go on for twenty minutes.
If no trace of a deposit is obtained, IO c.c. of the liquid to be tested
and Io c.c. of hydrochloric acid are put into the funnel, b, and slowly
introduced into the bottle, care being taken to exclude air-bubbles.
Some materials, such as beers, are apt to froth, hence the necessity
for the slow introduction of the sample. If after about ten minutes no
mirror appears, another IO c.c. of the liquid with IO c.c. of hydrochloric
acid are added, and the experiment continued for from fifteen to twenty
minutes, acid being added from time to time as may appear necessary.
The report also deals with the special precautions to be observed in
the examination of malt, hops, sugar, and other brewing materials, as
well as the particulars for cases in which the organic matter present
must be destroyed, as is frequently essential. Fuel, for example, is
incinerated after mixing with lime or magnesia, and the resulting
residue extracted with hydrochloric acid to obtain the “total arsenic.”
If sulphites are present they must be oxidised by bromine, the
excess of the latter being removed by heating.
The committee found that arsenic acid as well as arsenious acid can
be detected and estimated by the procedure described. The quantitative
results are, of course, only approximate, and cannot be relied upon for
quantities of arsenious oxide below O.OO3 mg. As an additional pre-
caution, a second tube should always be substituted for that containing
the mirror, and the experiment continued for a further period of from
fifteen to twenty-five minutes, and any arsenic deposited added to that
previously obtained. As a further check, the tests should always be
made in duplicate.
To prove that the mirror obtained actually consists of arsenic, the
narrow portion of the tube, which should not contain more arsenic than
corresponds to O-OI mg. As,Cs, is cut off, the hydrogen displaced by air,
and the ends of the tube then fused together. The sealed-up tube is
then drawn several times through a Bunsen flame until the mirror has
disappeared; on cooling, minute crystals of arsenious oxide appear.
The sparkling of the crystals can be seen with the naked eye, if the
tube be held in front of a luminous flame, and the crystals may be
readily identified under the microscope.
By carrying out the test in the manner described, it is possible to
detect, when working with 20 g. or 20 c.c. of substance, so low a per-
centage as O-OOOO15 As2Os or one part As2O5 in 7,OOOOOO.
368 MANUFACTURE OF SULPHURIC ACID.
The foregoing description is intended for the detection of extremely
minute quantities of arsenic in food-stuffs and the like for forensic
purposes; fuel is included on account of its use in the “kilning” of
malt.
For the examination of sulphuric acid the method may be simplified
by suitable modifications. For works purposes, for example, in testing
whether, in purifying, all the arsenic has been precipitated by sulphuretted
hydrogen, a much simpler apparatus will suffice, the test being frequently
carried out in a flask fitted with a glass tube drawn out to a point and
bent at right angles, and the examination made by observing whether
the hydrogen flame will produce a stain on cold porcelain. If no stain
is produced, it is assumed that only harmless traces of arsenic remain in
the acid.
A further refinement in the method for the detection of arsenic has
resulted from the work of the special Committee," appointed by the
Board of Inland Revenue, of which Prof. T. E. Thorpe acted as chairman.
In the report two methods of carrying out the Marsh-Berzelius test are
described, namely, the zinc method and the electrolytic method. The
details for the first agree with those of the earlier arsenic committee
(p. 364), except that the apparatus recommended is much smaller and
the quantities of acid and zinc employed correspondingly less, The gas
evolution takes place more slowly and the arsenic is deposited on a
smaller surface of glass, which latter allows of a better comparison with
the standard mirrors. These improvements are considered essential by
the Inland Revenue Committee. Where cost of apparatus need not
be considered and where electrical current of suitable strength is avail-
able, the evolution of hydrogen by electrolysis is recommended. The
design of the apparatus and the manner of carrying out the test, as
worked out in the Government Laboratory, are illustrated and minutely
described in the report. In this method the conditions of experiment
may be made absolutely uniform, and more exact comparisons with the
standard mirrors should, in consequence, be obtained ; further, all
troubles arising from the presence of arsenic in the zinc are
eliminated.
The apparatus, Fig. I29, consists of the following parts. The glass
vessel A, open below, is fitted with a ground-in stopper furnished with a
tap funnel the tube of which projects slightly below the stopper. The
stopper is provided with a bent tube B', to which the calcium chloride
tube C is attached by a ground-glass connection. A strong platinum
wire, a, is also fused into the stopper, and to it a conical-shaped
* Report of the Committee appointed by the Commissioners of Inland Revenue to specify the
ingredients of beers and the materials used in their preparation which are liable to be contaminated
by arsenic, and to prescribe tests, etc.; published by Wyman & Sons, London, 1903. Cf. also,
J. Chem. Soc., 1903, 83, 974. -
APPROXIMATE ESTIMATION OF ARSENIC 369
platinum cathode, aſ, pierced with several holes, is attached. The vessel
A is supported by a porous cell D, the distance between the walls of the
two vessels being from 2 to 3 mm. The porous cell is made of Pukall's
porous porcelain, the walls of which are from I to 1.5 mm. thick; it stands
in the thick-walled, glass anode vessel E. The anode consists of a strip of
platinum foil, 2 cm. broad, coiled loosely round the cell D, and connected
to the circuit by a strong platinum wire. The whole stands in a
cooling-vessel, F, by aid of which the temperature is kept below 50°.
The tube C contains firstly a plug of cotton wool, then for a length
of 5 cm. pure, rather finely granulated, calcium chloride, which must be
renewed after every three or four experiments, followed by a second
plug of cotton wool, and finally a roll of lead acetate paper; this is
FIG. 129.
prepared by soaking filter paper in a cold saturated solution of lead
acetate, and after drying in the air, cutting it into strips I cm. wide and
then rolling it into a coil so as to fit loosely in the tube. A further
small spiral roll of lead paper is also inserted in the exit-tube from C,
which is connected with the constricted glass tube G by means of a small
piece of non-vulcanised rubber, care being taken that the ends of the
glass tubes are in contact. The tube G is made from a piece of Jena
glass tubing of 3-5 mm. internal and 5 mm. external diameter, and
is cleaned by successive treatment with acid, alcohol, and water. It is
then dried and heated in the blowpipe so that a portion of the tube,
about 2 cm. in length and 5 cm. from the end, is softened, when it is
drawn out to a length of 7 to 8 cm. and to a uniform external diameter
of 2 mm. The tube is then cut off near the end of the narrowed

2 A
370 MANUFACTURE OF SULPHURIC ACID
portion and the latter bent upwards to a height of I cm.; it is
supported in slots in the metal cone H, surrounding the Bunsen flame;
a piece of platinum gauze, e, about 2 cm. square, is wrapped round the
portion of the tube to be heated in the burner. The small burner J
has a circular base 12 mm. high and a tube 6 cm. high, and 5 mm.
internal diameter.
The resistance of the apparatus is 1.4 ohms, and working at 7 volts
potential difference between the pole wires and employing a current of
5 amperes, 40 c.c. of hydrogen are liberated per minute; this gives a
steady flame 2 mm. high. The original description contains directions
for arranging the apparatus so as to permit several tests being carried
out simultaneously. -
The sulphuric acid solution employed in the apparatus is prepared
from pure sulphuric acid, tested especially for freedom from arsenic,
which is diluted in the proportion of 7 vols. of water to I vol. of con-
centrated acid.
To apply the test to sulphuric acid, 5 c.c. of the sample are diluted
with 20 c.c. of water, O. 5 g. potassium metabisulphite added, the
solution boiled to drive off the sulphur dioxide, and allowed to cool.
The object of the addition of the metabisulphite is to reduce any
arsenic acid or arsenate present to the “arsenious” state, as this elec-
trolytic method is only applicable to arsenic in the condition of an
arsenite or of arsenious acid. * .
To carry out a test, the vessels A, B, and E are first thoroughly
washed with water, cold water poured into the cooling vessel F, and the
tube G connected with the drying tube C as described. The wires a
and b are then connected with the current by means of binding screws,
30 c.c. of dilute sulphuric acid poured into the anode vessel E, and 20 c.c.
into the porous cell D through the funnel B ; the current is then switched
on ; after ten minutes the air will have been sufficiently displaced to allow
the hydrogen to be lighted. At the same time the burner J is lighted, and
the flame so adjusted that the platinum gauze e is maintained at a red
heat. If at the end of a further fifteen minutes no brown ring has
appeared in the narrow portion of the tube G, it is safe to assume that the
acid is free from arsenic. If this be the case, 2 c.c. of rectified amyl alcohol
are admitted to D through B, and followed immediately by the acid to
be tested, which has previously been prepared as described above, and
the funnel rinsed with 5 c.c. of water; the funnel tube is thus kept full
of liquid, and so excludes air from the apparatus. The object of the
addition of the amyl alcohol is to prevent frothing. If arsenic be present,
a deposit will be formed, after a few minutes, at a point I to 2 cm.
removed from the heated portion of the tube; after thirty minutes
practically all the arsenic will in most cases have been deposited. The
tap in B is then opened, the Outer end of G immediately held with a
APPROXIMATE ESTIMATION OF ARSENIC 3.71
pair of forceps, and a small pointed flame directed against the tube
between the end and the deposit at a point 3 cm. from the latter. The
tube fuses together immediately and is drawn off. The electric current
is then switched off and the tube G heated and drawn out near its wider
end, care being taken not to heat the arsenic mirror. The piece of tube
about 4 cm. long so obtained and containing the arsenic mirror, is placed
on a piece of white paper and compared with standard mirrors prepared
from very small quantities of arsenic employed in the form of a solution
of pure arsenious acid in hydrochloric acid, and diluted so that I c.c.
contains O.OI mg. As2O3.
In order to obviate the necessity of reducing arsenic acid and
arsenates, Trotman adds a few drops of zinc sulphate solution to the
contents of the inner cell; this causes the hydrogen to be evolved in a
state of supertension which suffices to reduce any “arsenic” arsenic
present. The same object has been more satisfactorily attained by
using cathodes of metals, such as lead and zinc, which have a consider-
ably greater supertension than platinum. Sand and Hackford * recom-
mend the use of lead electrodes, and have described a suitable apparatus
for the electrolytic estimation of minute quantities of arsenic, whilst
W. Thomson” recommends zinc electrodes; in both cases a complete
reduction of “arsenic” arsenic is said to be effected. Chapman and
Law “ have shown that arsenite solutions are completely reduced in
presence of sulphuric acid when cathodes of lead, tin, or of cadmium
are used, but that with other metals, including iron, a large proportion
of the arsenic is retained in the cell; arsenate solutions behave similarly,
but in no case is the whole of the arsenic evolved as hydride.
Kühn and Saeger" have described a modified form of the Marsh
test, which permits of a quantitative estimation of the arsenic. Ackroyd"
states that only the brown and not the blue modification of the arsenic
mirror is suitable for quantitative comparison; the former results in the
case of organic liquids (glucose, beer, etc.), the latter from inorganic
solutions. Only dilute solutions should be employed in testing, and for
exact work fresh standards should be made each time.
Bertrand 7 has proposed the following method for the detection of
rºorg mg. arsenic or less. The acid used is warmed to from 30° to a
maximum of 60°, the zinc added, and all air driven out of the apparatus
by pure carbon dioxide; I to 2 drops of platinic chloride solution and
15 c.c. of dilute sulphuric acid (I : 5) added, followed after an interval of
ten minutes by the solution to be tested. The gas evolved is dried by
passing through a plug of cotton wool previously heated to I2O", and
* J. Soc. Chem. Ind., 1904, 23, 177. * Analyst, 1906, 31, 3.
* /. Chem. Soc., 1904, 85, IoI8. * Ber., 1890, 23, I798.
*J. Soc. Chem. Ind., 1904, 23, 799. * /. Soc. Chem. Ind., 1902, 21, 1900.
7 Bull. Soc. Chim., 1902 [3], 27, 851.
372 MANUFACTURE OF SULPHURIC ACID .
then through a carefully cleaned glass tube of 1 mm. diameter, which is
narrowed by drawing out at a point some Io to 15 cm. behind
the place at which the ring is to be produced. A length of 20
cm. of the tube is heated to incipient redness. In the case of thick-
walled tubes it is advisable to confine the heated space by means of
strips of filter paper which are kept moist. Finally, the tube is sealed
at both ends in a current of hydrogen. According to Treadwell," the
addition of platinic chloride is inadmissible. Pederson” gives similar
details to those of Bertrand.
The zinc sometimes appears to be quite inactive (cf. p. 366), but such
zinc has the peculiar property of becoming active on remelting and
granulating. Allen ascribes this to the trace of iron, which may be taken
up on melting in an iron ladle, and on this account he purposely adds a
trace of ferric sulphate. Large quantities of iron are, however, to be
avoided, since according to Parsons and Stewart” they cause a portion
of the arsenic to be retained in the hydrogen generating flask. The
influence of metallic salts on the sensitiveness of zinc has been especi-
ally studied by Chapman and Law.” Salts of palladium, platinum,
nickel, and of cobalt prevent the reduction of arsenic to its hydride by
zinc, whilst those of cadmium, tin, lead, aluminium, magnesium, and
the alkali metals have no effect. Alloys of most metals with “sensitive”
zinc, similarly inhibit the reduction, with the exception of the alloys of
zinc with tin and with cadmium. If, however, 2 g of cadmium sulphate,
or of stannous chloride, or of lead acetate, be added to one of these
insensitive alloys (with the exception of that with nickel), the whole of
the arsenic is evolved as hydride. Similarly, the addition of 2 g. of
cadmium sulphate renders insensitive zinc capable of reducing arsenic
compounds; also, magnesium is rendered sensitive by the addition of a
cadmium, zinc, or lead salt.
Arsenic-free zinc is readily prepared, according to Hehner,” by melt-
ing about 500 g of zinc with about I g. of metallic sodium ; on stirring, a
black scum is formed on the surface of the metal, which is removed.
The fused metal is then poured into a second crucible, again treated
with sodium, and granulated. If this treatment is carefully carried out,
the zinc obtained is absolutely free from arsenic.
The suggestion to employ aluminium foil and sodium hydroxide in
place of zinc, cannot be recommended for exact work. According to
Hehner," it is possible, when working with aluminium and sodium
hydroxide solution, to fail in obtaining a mirror even when O.2 mg. of
arsenious oxide is present in 25 c.c. of the solution.
Great differences of opinion exist as to the influence of sulphites,
* Analytical Chemistry, vol. i., p. 191. * Analyst, Igoó, 31, 3.
* Chem. Centr., 1903, I., 250. * J. Soc. Chem. Ind., Igo2, 21, 675.
* J. Amer. Chem. Soc., 1902, 24, Ioos. * Chem, News, 1901, 83, 34.
APPROXIMATE ESTIMATION OF ARSENIC 373
which are frequently present in the case of beer. According to some
authorities they interfere with the formation of the arsenic mirror, and
should consequently be oxidised with bromine water before testing (cf.
p. 367), whilst others maintain that they exert no prejudicial effect. The
extremely small amount of sulphur dioxide which may be present in
sulphuric acid need scarcely be considered. The presence of sulphites
only calls for consideration in the examination of beer, to which calcium
bisulphite has been added to check fermentation ; traces of sulphur
dioxide may also arise from Sulphuring casks or hops.
Allen purifies the hydrochloric acid intended for use in the Marsh
test by adding a slight excess of potassium permanganate, and dis-
tilling. The distilled acid is quite free from arsenic, but the first
fraction coming over contains chlorine and must be rejected. Ling and
Rendle” effect the purification by boiling hydrochloric acid of constant
boiling-point, mixed with a small proportion of methyl alcohol, in a
reflux apparatus, under reduced pressure, with bright electrolytic copper,
free from arsenic; a little arsenic-free granulated zinc is added to the
acid. The acid is then distilled over pure copper.
Other contributions concerning the estimation of minute quantities
of arsenic have been made by Lockemann,” Mai and Hurt,” Bishop,”
Bertrand and Vamossy,” and especially in the last report of the
Analysis Committee of the International Congress of Applied
Chemistry." - -
That selenium affects the delicacy of the test, has been shown both
by Dawydow * and by Berry.” In this connection Rosenheim and
Tunnicliffe” have drawn attention to the fact that selenium, similarly
to arsenic, may give rise to symptoms of poisoning." Rosenheim states
that the presence of selenium is not shown by the Marsh test, but
that when present with arsenic it influences the size of the arsenic
mirror, and may even, under certain conditions, entirely prevent its
formation. The Reinsch test may be satisfactorily employed in presence
of selenium, provided silver foil is used instead of copper. Further, the
Gutzeit test is not influenced by the presence of selenium, though
Bettendorf’s test is. Schindelmeiser” has shown that in the Marsh test
selenium, if present, is deposited on the zinc, and that after such
deposition is complete arseniuretted hydrogen is evolved and may then
be detected as usual.
* / Soc. Chem. Ind., 1902, 21, 903. t 7 Report, 1906, pp. 280-318.
* Analyst, 1906, 31, 37. * Chem. Centr., 1895, I., 811.
* Z. angew. Chem., 1905, 18, 416. * /. Soc. Chem. Ind., 1901, 20, 322.
* Z, anal. Chem., 1904, 43, 537. * /bid., 1901, 20, 390.
* /. Amer. Chem., Soc., 1906, 28, 178. * Chem. Mews, 1901, 83, 280.
* Chem. Centr., 19C6, I., 1461. * Chem. Centr., 1902, II., 960.
374 MANUFACTURE OF SULPHURIC ACID
B. Reinsch's Test.
This test depends on the fact that clean copper when immersed
in a hydrochloric acid solution of arsenious acid becomes covered with
a grey coating of copper arsenide As,Cus; the deposit is formed in the
cold if much arsenic is present, but only on warming if the solution be
dilute. Arsenic acid only gives the reaction on warming. Since antimony
also gives a coating of similar appearance, any deposit formed must be
specially examined for arsenic (cf. infra).
The Reinsch test has now been discarded to a considerable extent,
owing to its not being approximately quantitative in character, and
also because it is not reliable when the arsenic is present as arsenic
acid, or when sulphites are present. The test is, however, of consider-
able value in certain cases. According to Allen it is preferable to
other tests for detecting arsenic in beer, etc. To carry out the reaction
he purifies the hydrochloric acid as described above (p. 364), adds
a small quantity of the acid and bromine water to IOO c.c. of
the beer, boils for a few minutes to oxidise any sulphite present,
adds a small quantity of cuprous chloride to reduce the arsenic
to the arsenious condition, immerses I sq. cm. of copper foil in the
solution and boils for half an hour, adding water from time to time
to replace that lost by evaporation. If the copper becomes darkened,
it is dried on the water-bath, cut into strips, and examined for arsenic
by heating one of the strips in a narrow test-tube, and observing
whether a sublimate of the characteristic octahedra or tetrahedra of
arsenious oxide is formed. Larger crystals are obtained by previously
warming the upper portion of the subliming-tube, and the appearance
of the crystals may be made more distinct by filling the tube with
water. A similar procedure, together with a minute description of the
various details involved in carrying out the test, has been published
by a committee appointed by the Brewer's Central Association,
Manchester;” the addition of an oxidising agent to destroy the
sulphite and of a reducing agent to convert arsenate to arsenite, is
omitted.
C. Gutzeit's Test.
This test is based on the action of arseniuretted hydrogen on solid
silver nitrate or mercuric chloride, prepared by allowing a drop of the
respective solution to dry on filter paper, or preferably, according
to Eidenbenz, by dissolving a small crystal in situ on the paper, and
drying. Silver nitrate is first turned yellow, owing to the formation of
silver arsenide, and subsequently black, owing to separation of metallic
silver; mercuric chloride is turned yellow owing to the formation of a
* /. Soc. Chem. /nd., 1901, 20, 281. * /øid., 1901, 20, 646.
DETECTION OF ARSENIC 375
mercurous salt having the composition As (HgCl). The test is
generally performed by placing a small piece of granulated arsenic-free
zinc in a test-tube, adding the material to be tested, and then, provided
the material itself is not acid, a small quantity of dilute sulphuric acid.
A plug of cotton wool is placed in the upper part of the tube and
the tube capped with the previously prepared piece of filter paper.
Sulphuretted hydrogen, phosphoretted hydrogen, and antimoniuretted
hydrogen interfere with the reaction. The test is, however, frequently
employed in the examination of commercial acids and of other sub-
stances; its applications and reliability have been specially studied
by Flükiger, Richardson,” Gotthelf.” Dunstan and Robinson, and by
Goode and Perkin.”
Kirkby's modification" of the Gutzeit test, which is frequently
employed, is based on the observation that hydrogen sulphide
may be completely removed from
the evolved gases by means of a
5 per cent, solution of lead acetate,
without any loss of arsenic. The
apparatus used for carrying out the
test is shown in Fig. 130. The
hydrogen is evolved in the flaska,
and is purified by passing through
five bulbs, the two lower, bà,
being half filled with 5 per cent.
lead acetate solution, and the two
upper, cc, with water. The puri-
fied gas passes into a small funnel, d,
which is capped with the spotted
and dried filter paper.
Tyrer' washes the gas with a 10 per cent. solution of lead acetate
in a two-bulb apparatus of different type to the above. A simple form of
apparatus, but perhaps not so reliable, has been described by Dowzard.”
F. W. Richardson” recommends the Gutzeit test, owing to its
simplicity as compared with the Marsh test, and the possibility of
carrying it out without special precautions. -
Hehner” objects to the test on the ground that it is impossible to
be certain that the stain obtained has been caused by arsenic,
since hydrogen phosphide, etc., will also produce similar stains; apart
from this, he considers the test very satisfactory and delicate. The
FIG. 130.
* Arch. Pharm., 1889, 27, 1. * Ibid., 1901, 20, 281.
*J. Soc. Chem. Ind., 1902, 21, 901. 7 /öld.
* /bid., 1903, 22, 191. * Ibid., 1900, 19, II45.
*Ibid., 1904, 23, 999. * Ibid., 1902, 21, 902; cf. Gotthelf, ibid., 1903, 22, 191.
* Ibid., 1906, 25, 507. 10 Ibid., 1901, 20, 194.

376 MANUFACTURE OF SULPHURIC ACID
method of testing proposed by Bird" overcomes Hehner's objection.
The stain is treated with boiling hydrochloric acid, whereby a stain
produced by phosphoretted hydrogen changes to lemon-yellow, one
due to sulphuretted hydrogen disappears, one due to antimoniuretted
hydrogen changes to pale grey, whilst that produced by arseniuretted
hydrogen changes to brick-red, and may even be recognised when
all the gases mentioned are present together in the test. The arsenic
stain vanishes on treatment with bromine-hydrochloric acid, and the
arsenic may be detected in the resulting solution by means of the
brownish-red coloration produced on addition of stannous chloride.
The application of Gutzeit's test to the quantitative estimation of
arsenic has recently been very fully investigated by Sanger and Black.”
D. Bettendorf's Test.
A few drops of the solution to be tested are added to I c.c. of con-
centrated hydrochloric acid, and , C.C. of a solution of stannous chloride
dissolved in an equal weight of strong hydrochloric acid added to the
mixture. If arsenic be present the solution quickly turns brown, and a
dark precipitate of metallic arsenic gradually separates out. The
change occurs more readily on warming the solution. The reaction
does not take place with aqueous solutions of arsenious acid, but only
with the arsenious chloride which is formed from it in the presence of
strong hydrochloric acid. Hydrogen phosphide and antimoniuretted
hydrogen are not reduced by stannous chloride, and consequently do not
interfere with the test.
According to Messel.” O.OI mg. AssCs in I c.c. sulphuric acid may
be detected by this method.
The “tin foil test,” which consists in the addition of strong hydro-
chloric acid to vitriol, then tin foil, and warming the solution, is
practically identical with the Bettendorf test.
Of other tests that have been proposed, the following may be
mentioned :—Donath * mixes IO to I5 C.C. of the acid with an equal
volume of water, adds a strongly acid solution of Stannous chloride,
heats nearly to boiling-point, and adds gradually a solution of sodium
sulphite. Should arsenic be present, it separates out after some time
in the form of finely divided yellow arsenious sulphide.
Seybel and Wikander" make use of the yellow precipitate Asſa,
produced on addition of potassium iodide. Free chlorine, ferric salts,
nitrous acid, and lead interfere with this reaction.
* /. Soc. Chem. Ind., 1901, 20, 390. * / Soc. Chem. Ind., 1901, 20, 192.
* Ibid., 1907, 26, III.5. * Z. anal. Chem., 1897, 36, 664.
* Chem. Zeit., 1902, 26, 50.
ESTIMATION OF IMPURITIES IN SULPHURIC ACID 377
QUANTITATIVE ESTIMATION OF SULPHURIC ACID AND
OF ITS IMPURITIES
Free Sulphuric acid is estimated almost exclusively by titration
with a standard solution of sodium hydroxide; the gravimetric deter-
mination by precipitation as barium sulphate is much less exact, and
includes, moreover, any combined sulphuric acid which may be present.
On the other hand, the volumetric method includes, as sulphuric acid,
other acids which are present in the free state, but in commercial sulphuric
acid the quantity of such other acids may be neglected as unimportant.
In the case, however, of nitrating and of spent acids, the process
described on p. 332 must be employed.
For the volumetric estimation, 2 to 3 g of the acid are weighed off
in a bulb pipette (Fig. I 34); pipetting Concentrated sulphuric acid is
precluded by reason of its viscosity. The outside of the pipette is care-
fully cleaned, and the acid then allowed to flow into a relatively large
volume of water, and the pipette re-weighed without washing out. The
pipette need not be cleaned and dried for further determinations ; it
suffices to wash it out several times with the fresh acid to be tested.
Working with the quantity of acid given above, it is best to
titrate with normal sodium hydroxide solution. The most satisfactory
indicator is methyl orange, used sparingly (cf. p. 65), at the ordinary
temperature and not in a warm solution. The amount of nitrous acid
present in commercial vitriol does not interfere with the titration, but
when larger quantities are present the test must be made as described
on p. 66.
Sulphurous acid, if present in appreciable quantity, is best estimated
by means of iodine (p. I 16). Should nitrogen acids be present, sulphurous
acid can only exist in traces, and cannot in such cases be determined
quantitatively. t
Nitrous acid (Nitrosyl-sulphuric acid), if present to any considerable
extent, is estimated by potassium permanganate (p. 343). Such small
amounts as cannot be determined with certainty either by per-
manganate or by the nitrometer, may be estimated colorimetrically.
Of the various methods proposed for this purpose, that in which the
modified Griess reagent (a-naphthylamine and sulphanilic acid) is
used, is the best."
Lunge and Lwoff found that a large excess of this reagent is neces-
sary in order to obtain reliable colorimetric results, and that in the
absence of such excess the intensity of the coloration increased at a
much slower rate than corresponds to the percentage of nitrous acid
present. Under suitable conditions and with at least one hundred times
* Ilosvay, Bull. Soc. Chem., 1894, II, 216; Lunge and Lwoff, Z, angew Chem., 1894, 7, 348;
the whole of the literature relating to the subject is given in this paper.
378 MANUFACTURE OF SULPHURIC ACID
the quantity of reagent theoretically necessary, the intensity of the
coloration corresponds to the percentage of nitrous acid. By observing
the details given below for the preparation of fresh reagent, the addition
of I c.c. suffices in all cases to which the colorimetric method is applicable;
the applicability is, of course, restricted to dilute solutions. -
As is well known, the red coloration produced by the action of very
minute traces of nitrous acid only appears after some time, and continues
to grow in intensity for several hours. This might appear to be a draw-
back in the application of the test for colorimetric determinations, but it
is not really so, for if all conditions, more particularly the time allowed
for reaction, are equal, the colour intensity and percentage of nitrous acid
remain comparable. If it be desired, therefore, to compare a solution of
unknown strength with a known standard solution, it is only necessary
to treat both with the reagent at the same time; the comparison may
then be made either at the end of five minutes, or of half an hour, or
after twenty-four hours, and always with the same result; for though the
absolute colour intensity continually increases, the relative values remain
the same. -
In the case of aqueous solutions the comparison may generally be
made directly, or at all events in a quarter of an hour, after the addition
of the reagent. The presence of free mineral acid even in traces retards
the reaction and renders the coloration less intense; a large amount
of such acid entirely stops the reaction. This is due to the fact
that the “coupling” of diazonium compounds is more or less
inhibited by the presence of mineral acids. This difficulty is readily
and completely overcome by adding sodium acetate in sufficient
quantity to take up the whole of the mineral acid. The sodium
acetate employed must, of course, give no coloration with the reagent,
which is not always the case with commercial acetate. A “normal
solution ” for purposes of comparison must be prepared for the test.
Dilute aqueous solutions of sodium nitrite or of nitrous acid are not
suitable, owing to their instability. The nitrous acid is therefore
employed in the form of the perfectly stable nitrosyl-sulphuric acid,
and the “normal solution” is prepared by dissolving O.O493 g. pure
Sodium nitrite (= IO mg. N) in IOO c.c. of water and adding Io c.c. of
the solution to 90 c.c. of pure sulphuric acid; the solution thus contains
I/IOO mg. of nitrite nitrogen per c.c.
In carrying out the test, I c.c. of the “reagent” diluted with about
40 c.c. of water is placed in each of two colorimeter cylinders; to the
one is added 5 g. of solid sodium acetate and I c.c. of the “normal
solution,” to the other, 5 g. sodium acetate and I c.c. of the acid to be
tested, and the whole well mixed so that the nitrous acid may react
with the reagent at the moment of liberation. The colorations
produced may be compared when desired; as a rule, five minutes'
ESTIMATION OF IMPURITIES IN SULPHURIC ACID 379
standing will prove sufficient. It is both unnecessary and disadvan-
tageous to warm the solutions in this case. The solutions are best
mixed by means of glass tubes blown below into a bulb corresponding
to the inside diameter of the cylinders, as in the Nessler test for
ammonia; by moving these stirrers up and down three or four times
efficient mixing is secured.
The “reagent” is prepared by dissolving O. IOO g. pure white
a-naphthylamine in IOO C.C. of water by boiling for one-quarter hour,
adding to this solution 5 c.c. of glacial acetic acid, or the equivalent
quantity of weaker acetic acid, followed by a solution of I g. of
sulphanilic acid in IOO C.C. of water. The mixture must be kept in
well-stoppered bottles. The solution readily discolours, but a faint rose
coloration may be disregarded since it disappears when I c.c. of the
reagent is diluted to the 50 c.c. used in the test; any stronger coloration
may be removed by reducing the solution with zinc dust. One c.c. of
the reagent will, after ten minutes, indicate distinctly the presence of
I/IOOO mg. nitrite nitrogen in IOO c.c. of water.
Nitric acid.—For the quantitative colorimetric determination of
nitric acid alone the brucine reaction may, according to Lunge and
Lwoff, be employed, if the comparisons be made not with the initial
red coloration but with the sulphur-yellow colour, which appears later.
The solutions required are:—(1) a brucine solution prepared by dis-
solving O-2 g. brucine in IOO c.c. of pure concentrated sulphuric acid;
and (2) a normal solution of potassium nitrate containing Tºg mg. nitric
acid nitrogen per c.c., prepared by dissolving O-O722 g. of the pure
salt in IOO C.C. of distilled water and making IO c.c. of this stock solution
up to IOO c.c. with pure concentrated sulphuric acid. These solutions
are best kept in well-stoppered burettes fitted with glass taps; that
intended for the “normal” solution should be graduated in ºr c.c.
The sulphuric acid to be tested may be used directly, provided
it is of not less than 1.7 sp. gr. ; weaker solutions or acids must be
mixed, in measured ratio, with pure concentrated sulphuric acid until
the above strength is approximately attained (cf. p. 361); thus water
alone will require the addition of three times its volume of concentrated
acid.
Narrow 50 c.c. graduated cylinders of pure white glass are employed
in comparing the colours; the graduated portion should be about 24 cm.
in length, and the cylinder should extend a further IO cm. beyond the
last graduation mark, to allow for mixing. The Hehner cylinders, which
are furnished with side taps at a height of about 5 cm. from the bottom,
are very convenient. An actual colorimeter will, of course, give the
most accurate results.
The test is made as follows:—I c.c. of the normal solution and I c.c.
of brucine solution are placed in one of the cylinders and pure con-
380 MANUFACTURE OF SULPHURIC ACID
centrated sulphuric acid added up to the 50 c.c. mark; the mixture is
then transferred to a flask and heated to 70° to 80°. When the colour
has changed to sulphur yellow the solution is cooled and returned to
the cylinder. The solution to be tested is treated similarly, it having
been ascertained by a preliminary test whether the liquid should be used
alone or after admixture with more concentrated sulphuric acid. Part
of the solution is then poured away, or run off through the side tap, from
one or other of the cylinders until the depth of colour is the same in
each ; the result is then obtained in the usual way by calculation from
the measure of normal solution required per given volume of the
unknown solution.
Selenious acid, according to Lunge,' has no effect on brucine.
In the presence of appreciable quantities of iron salts such as may
occur, for example, in certain makes of concentrated sulphuric acid, the
brucine test is less delicate, and the colour changes produced are not
uniform even when equal quantities of nitric acid are present.
Lead.—Concentrated acids are diluted with an equal volume of
water and twice the volume of alcohol, and allowed to stand for some
time. The precipitated lead sulphate is filtered off, washed with
dilute alcohol, dried, and ignited. The precipitate should be separated
from the filter paper as completely as possible; the latter must not be
ignited in a platinum crucible. I g. PbSO, =0.6829 g. Pb.
Iron.—If the quantity present is not too small, titration with per-
manganate may be employed after first reducing the iron to the ferrous
state. This reduction may be carried out in many ways. The most
general method is by means of chemically pure zinc, which should
always be tested for freedom from iron, the reduction being hastened
by warming, and carried out in a flask fitted with a Bunsen valve, or still
better with a Contat bulb (Fig. 39, p. IO6), or, according to Winkler, by
wrapping a zinc rod with platinum wire. The reduction is complete
when a drop of the solution, withdrawn by the aid of a capillary tube,
produces no red coloration with potassium thiocyanate. The solution
is allowed to cool, poured through a funnel provided with a platinum
cone or nearly closed by a glass stopper (not through filter paper), to
keep back the zinc, the flask and undissolved zinc washed with
thoroughly boiled, air-free water, and the solution titrated. Should the
zinc contain iron, a blank experiment must be made on at least 3 g of
the zinc, a weighed quantity of this taken for the reduction, and the
operation continued until all the zinc has been dissolved.
Skrabal” employs about one hundred parts of zinc for each part of iron
present, from which it will be seen that the smallest trace of iron in the
zinc may lead to very appreciable error. Any titanic acid present will also
be reduced by the zinc, and must consequently be taken into account.
* Aer., 1887, 20, 2031. * Z, anal. Chem., 1903, 42, 359.
ESTIMATION OF IMPURITIES IN SULPHURIC ACID 381
For very small quantities of iron, Gintl" uses palladium charged with
hydrogen, and so avoids introducing foreign matter into the solution.
The hydrogenised palladium is prepared by electrolysing sulphuric acid
with palladium as the cathode. Winkler 2 recommends a cylinder of
palladium gauze for this purpose; this is suitable for reducing larger
amounts of ferric salts.
Ebeling * proposes to add a little thiocyanate to the iron solution
itself in order to ascertain when reduction is complete, the heating being
continued until the red coloration has disappeared. Volhard * has
shown that this is unsuitable, partly because the thiocyanate is
destroyed in the process, and partly because traces of iron require a com-
paratively large excess of thiocyanate for their detection.
Other methods of reduction possess no advantages over the use
of zinc for the estimation of small quantities of iron in commercial pro-
ducts, and are consequently seldom employed for this purpose. These
include reduction by sulphuretted hydrogen recommended by Tread-
well,” sulphurous acid, stannous chloride, etc. The case is, of course,
different when large quantities of iron have to be determined, as for
example in ores, when reducing agents other than zinc are frequently
employed.
In the foregoing and analogous cases it is advisable to use AV/2O
potassium permanganate solution, prepared by diluting the seminormal
solution (p. 99); I c.c. will correspond to the O.OO2795 g. Fe. Further,
a considerable quantity of the acid, say 50 c.c., should be employed
for the test, since as a rule the quantity of iron present is only very
small.
For very small amounts of iron the colorimetric ferric thiocyanate
method may be employed. It is scarcely possible to obtain nitric acid
absolutely free from iron for oxidising the iron in the solution to be
tested to the ferric state, but this difficulty may be overcome by employ-
ing as pure an acid as possible, working only with small quantities, and
by making a blank test for comparison. The best results are obtained
under the following conditions worked out by Lunge."
Glass-stoppered 25 c.c. cylinders of perfectly colourless glass, gradu-
ated in ſo c.c. divisions, are employed in the test. The total height of
each cylinder is 17 cm. and the internal diameter 13 mm. ; to facilitate
mixing, the graduated portion does not reach beyond a point 5 cm. below
the stopper. The cylinders should be as similar in size as possible, so
that equal volumes reach to the same height in each. At least three
* Z. angew. Chem., 1902, I5, 398 and 424. * Massanalyse, 3rd edition, p. 95.
* Z, angew. Chem., 1901, I4, 57.I. * /bid., 1901, I4, 609.
* Analytical Chemistry, vol. ii., p. 87. According to Cappadow (Gaz, chim., 31 (ii.), 217) and
Skrabal (Z anal. Chem., 1903, 42, 359), this method is inexact, owing to iron sulphide being
carried down with the precipitated sulphur.
* Z, angew. Chem., 1896, 9, 3.
382 MANUFACTURE OF SULPHURIC ACID
cylinders are required; it is preferable to have four or six. The neces-
sary reagents are: (1) a IO per cent. Solution of potassium thiocyanate;
(2) pure ether; (3) a solution of ammonium iron alum prepared by dis-
solving 8-630 g. iron alum in a litre of water, and diluting I c.c. of this
solution to IOO c.c.; the solution will thus contain O-OIO g. Fe per litre.
This dilute solution soon decomposes, especially in the light, and can
only be kept for a few days even in a dark place. The stronger solution,
containing the 8-630 g. iron alum per litre, may be kept for a consider-
able time if protected from air and light; sometimes, however, it
becomes cloudy even after a short time. The stability is increased by
the addition of a little sulphuric acid, and if, in preparing the solution,
5 c.c. of pure concentrated sulphuric acid are added before filling to the
litre mark, the solution will remain perfectly clear for a long period. The
quantity of sulphuric acid thus introduced into the test scarcely amounts
to 1 mg., and is therefore negligible. (4) Pure nitric acid as above.
Fifty c.c. of the diluted sulphuric acid are taken for the test and first
oxidised by warming with exactly I c.c. of nitric acid. At the same
time a second I c.c. of nitric acid (4) is diluted to 50 c.c. with distilled
water. Should the experiment indicate later that the sulphuric acid
solution should be still further diluted, the nitric acid solution must
also be diluted to the same extent, so that in both cases the amount
of iron introduced with the nitric acid will be the same. Any nitric
acid which produces more than a very faint reddish tinge with the
thiocyanate solution must be rejected.
Exactly 5 c.c. of the prepared solution of the sulphuric acid to be
tested are then placed in one of the stoppered cylinders (A), and 5
c.c. of the diluted nitric acid in a second cylinder (B). To the latter a
suitable quantity, say I c.c. of the iron alum solution, is then added from
a burette, and to A an equal volume of water, so that the solutions in
the two cylinders are brought to the same dilution. Five c.c. of the
thiocyanate solution are next added to each solution. Both mixtures
will redden, but the coloration is frequently of a somewhat dirty
yellowish red, and its intensity bears no relationship to the iron content
of the solution. Finally, IO c.c. of ether are added to each cylinder
and the contents thoroughly agitated. The aqueous solution contains a
double thiocyanate of potassium and iron. This double salt is split
up on shaking with ether, which dissolves only or preferably the iron
thiocyanate, as is evidenced by the rose-red coloration of the ethereal
solution as compared with the yellowish red of the aqueous solution.
Shaking must be continued until the aqueous layer becomes colourless.
The ethereal solution gradually darkens in colour, presumably owing
to a further splitting up of the double thiocyanate; the solutions
should therefore be prepared in rapid sequence, and preferably only
compared after standing for several hours.
ESTIMATION OF IMPURITIES IN SULPHURIC ACID 383
Differences of considerable magnitude are detected immediately, so
that it only becomes necessary to employ, in addition to the cylinder A,
containing the sulphuric acid solution, two other cylinders, B and C, to
which have been added the approximately correct quantities of iron
alum solution ; the final comparison, however, is, as stated, not made
for some hours. The solutions should not be allowed to stand too long,
for instance, over night, for it sometimes happens that after long
standing the ether becomes nearly, or even perfectly, colourless, only a
strongly coloured narrow zone of liquid remaining at the dividing line
between the ether and the water. This only occurs occasionally, and is
to be attributed to traces of impurities, which are, however, present in
such extremely small quantities that they cannot be identified.
The degree of accuracy of the method may readily be brought to
+ O. I c.c. iron alum solution, that is, to + O.OOOOOI g. Fe, on the 5 c.c. of
Solution taken, provided the volume of alum solution used does not
exceed 2 c.c., equivalent to O-OOOOO2 g. Fe. That is, the reading may
be made to 1/2O of the total iron present; this may be looked upon as
satisfactory, where only thousandths or hundredths of a per cent. come
into account. It is not, however, Sufficient for appreciably greater
amounts, and for such the volumetric method must be employed.
The above degree of accuracy may be attained without difficulty by
merely examining the ethereal layer by transmitted light. The com-
parison may be made more accurate by looking through the ethereal
layer obliquely from above, or by holding the cylinders a little above a
white surface, and looking through the full depth of liquid in the
cylinder. This plan is much better than allowing the cylinders to stand
on the white surface. Colorimeters, as ordinarily constructed, are not
suited for this particular determination, no provision being made for
shaking, preventing loss of ether, etc. Seyda” has described a method
differing but little from the above, and intended for use in the examina-
tion of water.
Hydrochloric acid.—Ten c.c. of the acid are boiled in a small flask
and the vapours evolved conducted to a surface of water con-
tained in a second flask. The hydrochloric acid is dissolved by the
water, and may be estimated by titration with alkali or with M/Io silver
nitrate solution.
Arsenic.—About 20 g. of the acid are diluted with water, the
solution filtered from any precipitated lead sulphate, and then treated
with a current of sulphurous anhydride until the solution smells
strongly of the gas, to reduce all arsenic compounds to the arsenious
condition. Prolonged action, and a considerable excess of sulphur
dioxide are necessary for the reduction. The excess of sulphur dioxide
is removed by heating, aided by a current of carbon dioxide, the
* Chem. Zeit., 1898, 22, Io&6.
384 MANUFACTURE OF SULPHURIC ACID
Solution exactly neutralised by addition of sodium carbonate and
bicarbonate, and titrated with M/IO iodine and starch solution till the
colour-change to blue occurs. One c.c. of the iodine solution corre-
sponds to O.OO495 g. As, Os. Iron must first be removed should it be
present in appreciable quantity.
The presence of lead, antimony, Copper, platinum, etc., somewhat
complicates the determination."
According to Böckmann, the reduced solution, freed from excess of
sulphur dioxide as above, is treated for some hours at a moderate
temperature with a current of sulphuretted hydrogen, which precipitates
the arsenic and other metals which yield sulphides insoluble in acid
solution. The precipitate is filtered off and well washed, at first with
water rendered slightly acid with hydrochloric acid, and then with hot
water until the washings leave no residue when evaporated on platinum
foil. The precipitate, which contains arsenic sulphide and antimony
sulphide, is dried and weighed. It is then moistened with cold water
and, when thoroughly damped, extracted with dilute ammonia. The
residue remaining on the filter, consisting of sulphur and antimony
sulphide, is washed with hot water, dried, and weighed as above. The
difference between the two weights gives the content of arsenic sulphide
sufficiently accurately for technical purposes.
Neher” precipitates the arsenic in acid solution by sulphuretted
hydrogen. Hattensaur * dilutes 500 c.c. of sulphuric acid with 500 c.c.
of water, adds 500 c.c. of dilute hydrochloric acid (I : 2) to the solution,
cooling during the addition, and passes sulphuretted hydrogen for three-
quarters to one hour through the continually cooled solution, finally
filtering off the precipitate, which consists of pure As,Ss free from lead.
The filtration may be performed in a platinum or porcelain Gooch crucible
(p. 26); any trace of sulphur is removed by a final washing with hot
alcohol, and the precipitate weighed in the crucible. In the absence of
a Gooch crucible, the precipitate may be dissolved on the filter in 20 c.c.
of dilute ammonia (I : 2), the solution evaporated in a porcelain crucible,
the residue oxidised to arsenic acid, and estimated as magnesium
ammonium arsenate. If the arsenic is weighed as pentasulphide, the
determination can be made in from three to four hours.
Very minute quantities of arsenic are estimated approximately by
comparison of the arsenic mirrors made according to the Marsh-
Berzelius test (cf. p. 363 et seq.).
Atterberg 4 estimates arsenic colorimetrically by boiling with strong
hydrochloric acid, collecting the distillate in water, evaporating this
with nitric acid, and finally reducing by stannous chloride or sodium
hypophosphite. -
* L. M'Cay, J. Amer. Chem. Soc., 1885, 7, 6. * Z. angew. Chem., I896, 9, 130.
* Z, anal. Chem., 1893, 32, 45. * Chem. Zeit., 190I, 25, 264.
FUMING SULPHURIC ACID - 385
FUMING SULPHURIC ACID (ANHYDRIDE, OLEUM)
Fuming sulphuric acid is generally regarded as a solution of sulphur
trioxide (sulphuric anhydride, SOs) in sulphuric acid monohydrate
(H2SO4). As a matter of fact, the chief constituent is generally pyro-
Sulphuric acid (H2S2O.), and the so-called 45 per cent, oleum consists
entirely of this. The other varieties below 45 per cent. are mixtures
of pyrosulphuric acid with monohydrate, and those above 45 per cent.
mixtures of pyrosulphuric acid with anhydride. The strength is always
given as percentage of anhydride, no account being taken of the presence
of pyrosulphuric acid. The latter is looked upon, from the analytical
standpoint, as a mixture of fifty-five parts H2SO, with forty-five parts
SOs.
Properties of fuming sulphuric acid. Pyrosulphuric acid, H2S,O,
that is oleum of 45 per cent. SOs, as well as those acids which contain
in addition to the pyrosulphuric acid only little H2SO, or SOs, that is
acids from slightly below 40 per cent. up to nearly 60 per cent.
SOs, are solid; on the other hand, the varieties ranging from O to
nearly 40 per cent, and from 60 per cent. to 70 per cent. SOs, are oily
liquids. From 70 per cent. onwards the oleum is once more solid,
until it finally passes to pure anhydride.
Melting Points of Oleum. Knietsch."

* | Mating point | *;" | Mating point. | *:::::" Melting-Point.
0 + 10 35 + 26-0 70 + 9-0
5 + 3°5 40 + 33-8 75 + 17-2
10 — 4°8 45 + 34°8 80 +20°0
15 – 11°2 50 - + 28°5 85 + 33°0 §§
20 – 11'0 55 + 18°4 90 + 34-0 (27-7
25 — 0°6 60 - + 0-7 95 +36-0 (26-0
30 +15°2 65 + 0°8 100 +40-0 (17-7
Boiling Points of Oleum. Knietsch.”

, Free, * * * º Barometric Pressure,
sº ºl, º: . Boiling-Point. Hºrtºn
82°3 . 3*64 212 759
83°4 9-63 170 759
86°45 26°23 125 759
89°5 42-84 92 759
93-24 63-20 60 759
99.5 97.2 43 759
* Ber, 1901, 34, 41oo. The figures in brackets are the melting points of the fresh non-
polymerised acids.
* Ibid., 411o,
2 B
386 MANUFACTURE OF SULPHURIC ACID
Specific Gravity of Fuming Sulphuric Acid at 35° C.
Knietsch."
Total SO3 Free SO3 Specific Total SO3. Free SO3 Specific
per cent. per cent. Gravity. per cent. per cent. Gravity.
81-63 0 1°8186 91-18 52 1°9749
81-99 2 1°8270 91°55 54 1°9760
82°36 4 1°8360 91-91 56 1.9772
82-73 6 1°8425 92°28 58 1°9754
83'09 8 1°8498 92°65 60 1-973S
83°46 10 1°8565 93-02 62 1°9709
83-82 12 1°8627 93°38 64 1-96.72
84 °20 14 1°8692 93-75 66 1°9636
84°56 16 1°8756 94°11 68 1°9600
84 °92 18 1°8830 94 °48 70 1°9564
85-30 20 1°8919 94-85 72 1-9502
85-66 22 1°9020 95 °21 74 1°9442
86°03 24 1°9092 95°58 76 1-9379
86°40 26 1°9158 95°95 78 1°9315
86-76 28 1°9220 96°32 80 1°9251
87°14 30 1°9280 96*69 82 1°9183
87°50 32 1°9338 97-05 84 1°9115
87-87 34 1°9405 97.42 86 1°9046
88°24 36 19474 97-78 88 1°8980
88-60 38 19534 98°16 90 1°8888
88-97 & 40 1°95S4 98°53 92 1°8800
89°33 42 1°9612 98-90 94 1-8712
89-70 44 1°9643 99 °26 96 1°8605
90-07 46 1-96.72 99 •63 98 1°8488
90°44 48 1°9702 100'00 100 1-8370
90°81 50 19733 e e tº - - - e tº tº

This table may be used in calculating the quantity of concentrated
sulphuric acid to be added to an oleum to obtain an oil of any desired
lower strength in free SO, Gerster” gives the following formula for
this purpose:— -
ô — a
42 – C
3 = IOO
where a represents the quantity of sulphuric acid to be added to
IOO parts of the oleum, a the total SOs per IOO parts of the acid desired,
6 the total SOs per IOO parts of the original strong oleum, and c the
SOs per IOO parts of the acid used for dilution. The values for a and b
are taken from the table; c is obtained by multiplying the percentage
of H2SO, present in the acid used for dilution by O.816.
Impurities. These may be the same as those present in ordinary
sulphuric acid, but since fuming acid is now made exclusively by the
contact process, they will only occur in inconsiderable quantity. The
examination for these impurities is carried out exactly as in the case
of ordinary sulphuric acid (p. 377 et seq.).
* Ber., 1901, 34, 41or, * Chem, Zeit., 1887, II, 3.
FUMING SULPHURIC ACID 387
Quantitative Analysis of Fuming Sulphuric Acid
The sampling of oleum is a matter of some difficulty. In the case
of liquid oleum, or in that of the partially or completely crystalline
products up to 45 per cent. SOs, the difficulty is not very great, since
the latter acids may be liquefied without danger by warming to 30°
in closed vessels on the sand-bath. In works the soldered leaden
vessels containing the acid are usually stored in a warmed room, so
that the contents are always liquid. It is advisable to remove the
stopper or other seal before warming, and to substitute a watch-glass;
this does not cause any appreciable loss of SOs, and checks the
development of pressure, which might easily lead to an accident on
removing the stopper.
Products rich in SOs cannot be completely liquefied by warming.
A jelly-like residue always remains, but as this possesses the same
composition as the liquefied portion, the sample may be drawn from
the latter.
The portion for analysis is taken from these larger samples.
Solid oleum (pyrosulphuric acid) must be liquefied in the sample
bottle by gently warming, before being withdrawn by the pipette for
analysis; it will then remain liquid sufficiently long to allow of its flow-
ing from the pipette even after weighing. Actual anhydride and oleum
approaching this strength cannot be handled in this manner, as they
fume too much. In such cases Stroof's method may be followed.
This consists in weighing several portions of the anhydride in a glass-
stoppered bottle, and then adding a weighed quantity of accurately
analysed monohydrate sufficient to produce an oleum of about 70
per cent, which is liquid at the ordinary temperature. Solution is
hastened by warming to 30° to 40°, with the stopper placed loosely in
the bottle. The analysis of the mixture is then carried out in the
ordinary way.
The value of an oleum depends chiefly on the content of free
anhydride, SO, ; the results are accordingly calculated in the follow-
ing manner. First, the total acidity is determined by titration and
calculated to SOs; the difference between IOO and the percentage
so obtained is regarded as water, for each 18. I parts of which 80.06
parts SOs are necessary for the formation of H2SO, ; after all water
has been allowed for in this way, any SOs remaining is taken as
free anhydride. It must not, however, be forgotten that other substances
* Cf. Fürstenau, Chem Zeit., 1880, 4, 18; Möller, ibid., 1880, 4, 569; Becker, ibid., 1880,
4, 6oo; Winkler, Chem. Ind., 1880, 3, 194; Clar and Gaier, ibid., 1881, 4, 251 ;
Rosenlecher, Z. anal. Chem., 1898, 37, 209; Setlik, Chem. Zeit., 1889, 13, 1670 ; Rabe, Chem.
Zeit, I90I, I5, 345; Sulphuric Acid and Alkali, I, p. 238.
388 MANUFACTURE OF SULPHURIC ACID
may be present in addition to the water, especially sulphurous acid, SOs.
which, as will subsequently be seen, considerably affects the result,
Solid constituents, in more than traces, are also occasionally present;
these must be estimated and the percentages obtained deducted from
the water value, otherwise the result obtained for the percentage of
free SOs will be too low.
The analysis of fuming sulphuric acid or anhydride is carried out as
follows:—
The oleum is weighed off in tared, thin-walled bulb tubes of about 20
mm. diameter, drawn out at both ends to form long capillaries. Three
to 5 g. of the melted, completely homogeneous oleum, which is sufficient
to nearly half-fill the tube, are drawn in by the following device.
An ordinary narrow-necked flask is fitted with a rubber stopper,
through which passes a well-ground glass tap provided with a piece of
rubber tubing at its free end. A partial vacuum is produced in the
flask by suction (with the mouth), the tap closed, the rubber tube passed
over one of the capillaries of the weighing bulb, and the other
capillary immersed in the oleum ; on opening the tap the desired
quantity of oleum is drawn into the bulb.
The bulb is then cleaned, one of the capillary ends sealed by fusion, and
the whole weighed. Loss of anhydride by evaporation or by absorption
of moisture through the non-sealed, capillary end does not occur to any
appreciable extent during the weighing. In weighing, it is advisable to
support the capillary ends of the bulb-tube on a small platinum crucible
with two notches cut in the rim ; this will prevent damage to the balance
should breakage occur. -
The weighed bulb is then placed in a small Erlenmeyer flask
(Fig. 131), so that the bulb closes the neck and the point of the open
capillary dips rather deeply into the water contained in
the flask. By this means any loss of anhydride by
evaporation is excluded. The point of the upper capil-
lary is then broken, and after all the oleum has run
out the tube is washed by dropping water through the
upper capillary, and finally by sucking in water several
times so as to completely fill the bulb. The solution
is diluted to 500 c.c. and 50 c.c. taken for titration.
The titration is made with N/5 sodium hydroxide
solution (I c.c. = O.OO8 g. SOs), using methyl orange as
indicator (not litmus, cf. p. 391). Sulphurous acid is determined by
titrating a second portion of the solution with iodine solution, and the
quantity found deducted from the total acidity, as found in the titration
with alkali. - s
Clar and Gaier weigh the anhydride (oleum) in a glass bottle
(Fig. I32), 58 mm. high and 17 mm. wide, fitted with a long, ground-
FIG. 181.

FUMING SULPHURIC ACID 389
in glass stopper broadened out to a bulb above the neck of the flask,
and pierced on the top with a small hole which can be closed by a
small glass stopper. The interior of the stopper is filled with glass
wool moistened slightly with water. Two to 3 g of
the melted anhydride or fuming acid are introduced
into the bottle, the stopper quickly inserted, and the
whole weighed. The head of the stopper is then
weighted by wrapping a strong platinum wire round
the neck, and the bottle allowed to slide, mouth down-
wards, into a suitable flask of about 2 litres capacity
placed in an inclined position, and containing about
500 c.c. of water at 50° to 60°. The flask is next placed
in an upright position and covered with a watch-glass.
It is advisable to bind the stopper to the bottle with
thin platinum wire, to prevent its falling out.
With suitable weighting the bottle takes up an
oblique position in the water, the head being directed
downwards, which is the best position for the subsequent
FIG. 133.
operation. At first the warmth of the
water expands the air in the bottle and
causes some acid to escape; later on, water
is drawn in and dilutes the contents of the
bottle without producing too violent a
reaction. The dilution may be hastened by cooling the
flask, but shaking or other violent movement must be
avoided. Finally, the bottle is washed inside and out
with water, the cooled solution made up to I litre, and
100 c.c. titrated with M/5 alkali as above. .
A very convenient apparatus for weighing fuming sul-
phuric acid or melted anhydride is the glass tap-tube
recommended by Winkler (Fig. 133).
The conically narrowed portion ending in a capillary
must be absolutely uniform, and the tap, which may not
be greased, should fit perfectly. The longer portion of
the tube is one-half, or at most two-thirds, filled by suction
with the acid to be examined, the tap closed, and the tube
then inverted so that the acid runs down to the tap. The
end of the tube is carefully cleaned with paper, and the
tube and contents weighed in a horizontal position. No
alteration in weight during weighing need be feared. The tube is then
placed with the point downwards in a beaker containing water and the
acid allowed to enter the water very gradually, the rate of the flow
being best regulated by means of a screw clip. In the case of very
strong acids or of pure anhydride which will remain liquid for a con-


390 MANUFACTURE OF SULPHURIC ACID
siderable time, dilution is effected by allowing the acid to escape on
to a layer of coarsely powdered, crystallised, and perfectly neutral
Glauber's salt. In this way dilution takes place quietly and without
danger, on account of the water of crystallisation present in the salt.
Finally, a few drops of water are allowed to enter the weighing-tube
from above, and after standing for a short time the tube is thoroughly
washed out. The Glauber's salt is dissolved in water, the solution made
up to a known volume, and an aliquot portion titrated. This plan does
not, however, give quite accurate results, as the colour change of the
indicator (methyl orange) is less sharp in presence of the Glauber's salt.
The most convenient form of weighing apparatus, not only for oleum
but also for other liquids, such as fuming acids of all kinds, ammonia,
— . . etc., which must be weighed out of contact with air, is the
Ho! bulb-tap pipette designed by Lunge and Rey (Fig. I 34).
The taps a and c must fit perfectly without grease. To
fill the pipette the tap c is closed, a opened and, whilst
applying suction at d, closed so that a partial vacuum is
obtained in the bulb b. The point e of the tube is then
placed in the acid, and c opened, whereby the acid rises in
the pipette. Care must be taken that it does not reach the
tap c.; all vapours are retained in b. The tap c is then
closed, the end of the tube e cleaned, and after returning
the pipette to the protecting tube f the whole is weighed.
In the case of strongly fuming nitric acid and similar
liquids a drop of the liquid may escape from the point e
during the weighing ; in such cases it is advisable to place
a little water in the tube f, taking care, however, not to
insert the empty pipette before the final weighing, lest its
point become wetted. The pipette is then removed from ſ,
the point e dipped into water, and the contents allowed to
escape gradually by opening the tap c.; a little water is
. admitted to 6 through d and a, and the whole finally washed
after allowing to stand for some time. If only O. 5 to I g. of acid have
been weighed off, the whole of the resulting solution is titrated, the
result obtained being more accurate in this case than by diluting to
a large volume and only titrating a portion. For larger quantities of
acid the solution must be diluted to a measured volume and an aliquot
part titrated.
The strongest oleum (over 70 per cent.) cannot be run into water
directly without undergoing loss. Such acid is weighed in the bulb-
tubes described above, both capillary ends fused, and the bulb placed
in a flask containing a fairly large volume of water. The flask is
then closed with a glass stopper, the bulb broken by shaking the flask,
and after standing for some time the whole is titrated.
FIG. 134.

FUMING SULPHURIC ACID 39i
As already stated (p. 388), a deduction must always be made for
sulphurous acid, which is seldom absent from commercial oleum. The
sulphurous acid is generally estimated by titration with iodine and the
quantity found deducted from the total acidity. Lunge 1 has pointed
out that an error may easily occur in the latter titration, unless due
attention be paid to the indicator employed and the stage at which the
change of colour takes place; great differences exist between the
various indicators in this respect. Thus with phenolphthalein the
change occurs when I mol. SO, has combined with 2 mol. NaOH, that is,
when NagsOs has been formed; with methyl orange, on the other hand,
the change takes place when only I mol. NaOH has been added for
I mol. SO2, that is, as soon as NaHSOs has been produced. Litmus
cannot be employed, since the results obtained are quite indefinite, lying
between the limits given above (cf. p. 70).
It is thus inadmissible to employ litmus in titrating fuming
sulphuric acid, since it will not show what allowance is to be made for
the Sulphurous acid; nor is phenolphthalein a good indicator in this
case, Owing to the unavoidable presence of carbonate in the sodium
hydroxide solution. Ammonia is unsuitable for the titration. Methyl
orange should therefore be used, bearing in mind that I c.c. normal
sodium hydroxide solution (O.O4OO6 g. NaOH) is neutralised by 3 mol.
SOs (O.O4OO3 g. SOs), but by a full molecule of SO, (O-06405 g. SO).
Consequently for each I c.c. N/Io iodine solution required for the
titration of the latter only O-os c.c. normal alkali solution must be
allowed, and not the equivalent ratio o, I c.c. If this point be neglected
very considerable error may result, since everything other than SO,
and SOs is assumed to be water; thus an incorrect allowance for sulphur
dioxide will not only show too little trioxide, but also a corresponding
quantity of water in excess of the real value, and since the latter has to
be credited with its corresponding amount of SO, (= 4,443 times the
quantity of water), the percentage of free trioxide may come out very
much too low.
In an actual case the incorrect allowance for SO2 led to an error
of 84 per cent. in the free SOs. This is, however, an extreme instance.
If the proportion of free trioxide is calculated by means of the table
(p. 394), instead of by multiplying the water by 4.443, due allowance
must similarly be made for the content of sulphur dioxide. Thus,
taking the actual case just referred to, analysis proved the acid
to contain 95.2 I per cent. SOs, 2.43 per cent. SO2, and 2.36 per cent.
H2O. In using the table the SO, must be left out of account, that is,
the SO, and H2O are the only factors to be considered. It will not
do to merely take from the table the value corresponding to a total
percentage of 95.2 I SOs; this would only give 73.95 per cent, free SOs,
* Z, angew. Chem., 1895, 8, 22.
392 MANUFACTURE OF SULPHURIC ACID
an even more incorrect result than the above. The correct procedure is
to add the 2.36 parts H2O to the 95.21 parts SOs, giving in all 97.57
parts of acid and water with a percentage content of 97-58 SOs. Accord-
ing to the table, it is found by interpolation that such acid corresponds
to 86.80 per cent, free SO, and 12.20 per cent. HASO, This 86.80
per cent, must then be recalculated to the original acid, and so allow
97.57 × 86.8O
IOO
SO, as the final value. The result obtained thus involves consider-
ably more calculation than multiplying the percentage of water by its
equivalent of sulphur trioxide; it is only fair to state, however, that this
example does not lend itself particularly well to the table, which is
drawn up for mixtures of anhydride and water only.
In the foregoing instance, as in all methods hitherto published for
determining the strength of oleum, everything other than SO2 and SOs
is regarded as water. It is, however, advisable to estimate the fired
impurities by evaporation, since otherwise their weight multiplied by
4,443 is erroneously deducted from the free SOs,
Rosenlecher 1 has described the method employed at Freiberg which
is used for checking the process in the works. Weighing-tubes of the
dimensions shown in Fig. 135 are made from glass tubing of 5 to 6 mm.
for the contained SO2, giving = 84.69 per cent. free
diameter, a number of which are prepared at one time. The ends of the
capillaries are narrowed to 3 mm., and for strong oleum or anhydride
to 4 mm. The bulbs are filled by suction applied by the mouth
through a capillary rubber tube fixed to the shorter capillary, a test-
tube filled with soda crystals being interposed when necessary. Suction
is applied until the liquid begins to rise in the bulb, and is stopped
before the heavy anhydride vapours reach the shorter limb. The bulb
is then turned so that the limbs point upwards, and the bulk of the
liquid brought into the bulb by gentle tapping. Each bulb is carefully
wiped with filter paper and supported opposite a corresponding number
in a notched cardboard box. In weighing, the bulb is placed on a
platinum crucible (p. 388), or held in a suitable brass wire support.
If the above dimensions of the capillaries are adhered to, there is no
danger of moisture being absorbed or of loss by evaporation, even in
* Z, anal. Chem., 1898, 37, 209.

FUMING SULPHURIC ACID 393
the case of pure anhydride. Care must, however, be taken not to warm
the bulbs in handling, either before or after weighing. To bring the
acid into solution the bulb is placed in a bottle containing the indicator
and 20 to 30 c.c. of water, at the temperature of the room. The
previously wetted glass stopper of the bottle is then inserted, the
bottle held in a horizontal position (the colour of the indicator should
remain unchanged up to this time), the stopper securely fixed, and
the bottle well shaken until the tube has been completely shattered
and all white vapours have disappeared. The solution is then
titrated in the bottle itself. The differences in the results obtained
seldom exceed o. 15 per cent. even with rapid working. (This state-
ment apparently relates to the total acidity and not to the free
anhydride.)
Dobriner and Schranz' dissolve 6 to 8 g. of the oleum contained
in a sealed tube by breaking the tube in a stoppered litre flask con-
taining about 150 c.c. of water, and adding a weighed quantity of
chemically pure dry sodium carbonate, so that only 3 to 4 c.c. of
normal solution are required for titrating back. By this means errors
in the standard solution, and in the burette, etc., are reduced to a
minimum.
The results of the titration are first calculated to percentage of
total SOs (combined and uncombined with water), each I c.c. normal
sodium hydroxide solution corresponding to O.O4OO3 g. SOs ; the ratio
between free SO, and the H2SO, present may then be taken from
the following table drawn up by Knietsch.” Or the general formula,
SOs = S – 4.443 (IOO – S),
may be employed, in which SOs stands for the free sulphur trioxide and
S for the percentage total SOs, as found by titration.
Grünhut” gives the following table (p. 395), which allows the percent-
age to be read directly to the hundredth part. The left-hand portion
shows the amount of trioxide corresponding to the total percentage of
H2SO, as found. For IO6 per cent. H.SO, for example, the point of
intersection of the vertical column Io and the horizontal row 6 gives the
desired percentage 26.657. The right-hand portion contains the values
for the decimal places. The first decimal is found in the vertical
column marked by the asterisk, and the horizontal line lying to the
right is then followed until the column having the second decimal place
for its heading is reached. Thus for O-78 per cent. H2SO the value
is 3-465. The desired result is then obtained by adding the per-
centages found for the whole numbers and for the decimals. Thus, had
the titration given 106.78 per cent. H,SO, the percentage of trioxide
* Z, angew. Chem., 1896, 9, 453. * Aer, I90I, 34, 41 I4.
* Z. anal. Chem., 1899, 38, I67.
394 MANUFACTURE OF SULPHURIC ACID
present in the oleum would be 26.657+ 3-465 = 30. I2 per cent. The
corrections for SO, and fixed residue must, of course, be made as
described (p. 391), before making use of these tables. -
Table for finding the Percentage of Free SOs in Oleum from the
Total SOs as determined by Analysis. Knietsch.

SO3 SO3 SO3 SO3 SO3 SO3
Total. Free Total. Free. Total. Free. Total. | Free. Total. | Free. Total. | Free.
81 °63 0-0 84-7 16-7 || 87°8 || 33-6 || 90-9 || 50°5 || 94-0 || 67-3 || 97-0 || 83-7
81-7 0 °4 84-8 17.2 87-9 34°1 91°0 51-0 || 94°1 || 67.9 || 97.1 84 °2
81-8 0-9 84 °9 17-8 88-0 || 34-7 || 91-1 51-6 || 94-2 | 68°4 || 97.2 84-8
81-9 1-5 85-0 18°3 88-1 35-2 || 91°2 52°1 || 94-3 || 69-0 || 97.3 85-3
82-0 2°0 85-1 18-9 || 88°2 || 35°8 || 91°3 52°6 || 94°4 || 69°5 || 97°4 85.8
82°1 2-6 85.2 19°4 88°3 36°3 || 91°4 53-2 || 94-5 70°l 97-5 86°4
82°2 3-1 85-3 20-0 88°4 36°8 91-5 53-7 || 94-6 || 70-6 || 97.6 86-9
82°3 3-6 85*4 20-5 88°5 37°4 || 91-6 54'3 || 94-7 || 71.2 || 97.7 87°5
82°4 4°2 85-5 21-0 88°6 37.9 || 91-7 54.8 || 94-8 71-7 || 97.8 88-0
82°5 4-7 85-6 21-6 || 88-7 || 38°5 || 91-8 55-4 || 94-9 || 722 || 97.9 || 88-6
82.6 5 °3 85-7 22*2 || 88-8 || 39-0 || 91-9 || 55-9 || 95-0 || 72°8 || 98.0 | 89.1
82.7 5:8 85-8 22*7 || 88-9 39-6 || 92-0 || 56°4 || 95.1 | 73-3 || 98-1 | 89.7
82-8 6°4 85-9 23-2 89°0 40°1 92°1 57-0 || 95-2 || 73°9 || 98-2 90°2
82.9 6-9 86-0 23.8 89-1 40°6 92°2 57.5 || 95-3 || 74°4 || 98-3 90.7
83-0 7-5 86°1. 24°3 89-2 41°2 92-3 || 58-1 || 95-4 || 75-0 || 98-4 91.3
83-1 8-0 86°2 24°9 89-3 || 41°7 92°4 || 58-6 || 95 °5 75°5 || 98-5 91.8
83-2 8-5 86°3 25°4 89 °4 42-3 || 92-5 59.2 || 95-6 || 76°1 || 98-6 92°4
83°3 9-1 || 86.4 26-0 89°5 42°8 92°6 59-7 || 95-7 || 76-6 || 98.7 92.9
83°4 9-6 86°5 26°5 89°6 43°4 || 92-7 60-3 || 95-8 77°1 98-8 93°5
83°5 10-2 86°6 27-0 89.7 43°9 92°8 60°8 || 95-9 || 77-7 || 98-9 94'0
83-6 10-7 86-7 27.6 89 °8. 44°5 92°9 61°3 || 96-0 || 78°3 || 99-0 94 "6
83-7 11°3 86.8 28-1 89-9 45-0 || 93°0 61.9 || 96-1 || 78°8 || 99-1 95°1
83-8 11°8 86-9 28-7 || 90-0 || 45°6 || 93°1 62°4 || 96-2 || 79°3 || 99-2 95-6
83-9 12-3 87-0 29-2 90°1 46"I 93-2 63-0 || 96-3 || 79°9 || 99-3 96-2
84-0 12-9 87-1 29.8 90°2 46-6 || 93-3 63°5 || 96°4 || 80°4 || 99°4 96-7
1 84.1 13°4 87.2 30°3 90°3 47°2 || 93-4 64°1 || 96°5 || 81-0 || 99-5 97.3
84 °2 14-0 87-3 30-9 90°4 47.7 93°5 64°6 || 96*6 81°5 || 99.6 97.8
84 °3 14°5 87°4 31°4 90°5 48°3 93-6 65°2 96-7 || 82-0 || 99-7 98’4
84°4 15:1 87-5 31-9 90-6 || 48°8 || 93-7 || 65-7 || 96-8 || 82-6 || 99.8 98-9
84°5 15-6 87-6 32°5 || 90-7 || 49-4 || 93-8 | 66-2 || 96-9 || 83-1 || 99-9 || 99.5
84°6 16-2 87.7 33-0 90°8 49°9 || 93-9 66-8 s e tº - © tº e e º to º 0
Setlik 1 has described a simple and rapid, method for the estimation
of fuming sulphuric acid, which depends on the known property of oleum
to fume in the air until all anhydride present has been converted to
monohydrate. The analysis is carried out as follows. At least 50 g., or
preferably IOO g., to simplify the calculation, of the sample are weighed
off on an ordinary balance to within O. I g., into a long-necked flask
of from 130 to 200 c.c. capacity, and distilled water dropped in very
slowly from a tap burette graduated in ºr c.c. and provided with a fine
exit-tube.
* Chem. Zeit., 1889, 13, 1670.
FUMING SULPHURIC ACID 395
Table for finding the Percentage of SOs in Oleum from the
Total Acidity calculated as H2SO4. Grünhut.
Whole Numbers. - Hundredth Parts.
#:



1O 11 12 O 1 2 3 4. 5 6 7 8 9
0 || 44'428 88-857
4°443 || 48°871 | 93-300
8-886 53°314 97.743
13°329 57-757 ...
17-771 |62-200
22*214|66-643
26°657 || 71.085
31°100 || 75°528
35°543 || 79°971
39°986 84°414
O 0-044 || 0-089 || 0-133 || 0°178 || 0°222 |.0°267 || 0°311 || 0°355 || 0-400
0°444 || 0°489 || 0-533 || 0-578 || 0-622 || 0-666 || 0-711 || 0-755 || 0-800 0-844
0°889 || 0-933 || 0-977 1-092 || 1-066 || 1-111 || 1-155 || 1:200 || 1:244 || 1:288
1°333| 1.377 || 1:422 || 1:466 || 1:511 | 1-555 1-599 || 1:644 || 1:688 | 1.733
1°777|1-822|1-866|1-910 | 1.955 | 1999|2-044 || 2:088 |2-133 |2-177
2°221 2°266 || 2:310|| 2:355 |2-399 |2-444 || 2:488 2°532 |2-577 || 2:621
2°666 |2-710 |2-755 |2-799 ||2-843 |2-888 |2-932 |2-977 ||3-021 || 3:066
3° 110 || 3' 154 || 3:199 || 3:243 || 3:288 || 3:332 || 3:377 3°421 || 3:465 || 3:510
3°554 || 3:599 || 3:643 || 3:688 || 3-7.32 || 3-776 || 3-821 3-865 |3-909 || 3'954
3°999 || 4-043|4-087 || 4-132 |4°176 || 4-221 || 4-265 || 4-310 || 4:354 |4-398
As the reaction is very violent, means must be provided for keeping
the flask well-cooled. At the start the drops of water falling into
the oleum fizz violently and give rise to dense fumes; as the titra-
tion proceeds, the reaction becomes less violent and the fumes less
dense. Towards the end of the titration the shaking must be continued
sufficiently long, after the addition of each drop, to allow the sulphuric
acid fumes to be completely absorbed; this is necessary to permit of
proper observation. The reaction is finished when no further fumes form
on the surface of the acid and a drop of water falling in the centre dis-
solves quietly. -
The calculation of the analysis is very simple: 9 c.c. H2O saturate
40 g. SOs, or IOO C.C. H2O correspond to 444 g. SOs. If, for example,
IOO C.C. of oleum have been taken, and 6 c.c. of water required, the
sample contains 26-64 per cent, free SOs. A table such as the following
may readily be prepared to give the content in free SOs directly from
the titration.

c.c. Water required º .C. W ir *
... ." Free so, [...Y.;"| Free so so)
0-1 0°444 1-1 4°882
0-2 0-888 1-2 5°332
0-3 I •333 1-3 5-777
0°4 - 1-776 1°4 6°221
O'5 2-222 1°5 6*666
0-6 2°664 1-6 7-104
0-7 3-110 1-7 7-544
0-8 3°551 1-8 7°996
0-9 3°996 1-9 8°440
1-0 4°444 2-0 8°888
It is advisable to have an approximate knowledge of the strength of
the oleum before making the analysis, since an oleum containing over
396 MANUFACTURE OF SULPHURIC ACID
35 per cent. SOs does not lend itself to direct titration, owing to the
violence of the reaction. In such cases the sample should be reduced to
from 30 to 35 per cent. oleum by addition of sulphuric acid monohydrate.
The monohydrate used for this purpose must, of course, have been care-
fully prepared, and should be stored in well-stoppered bottles. Setlik
recommends the method for works use, and claims that with a little
practice the test is quickly carried out, and is as accurate as titration
with alkali, especially if the acid be contaminated with sulphurous
acid, arsenious acid, sulphate of iron, or other impurity.
Rabe's method" is somewhat similar in principle, but sulphuric acid
of known strength (say 95 per cent.) is employed for titrating in place
of the water. This substitution only complicates the method and
renders the results less exact.”
Both methods are intended for works use only, and not for the
analysis of acids intended for sale.
LITERATURE
LUNGE, G.-The Manufacture of Sulphuric Acid and Alkali, vol. i., 3rd edition, 1903;
vols. ii. and iii., 2nd edition, 1895-96.
LUNGE, G., and E. BERL.-The Technical Chemist's Handbook, 1908.
KRAUCH, G-7%e Testing of Chemical Reagents for Purity, trans. from 3rd German
edition, by J. A. Williamson and L. W. Dupré, 1903.
* Chem. Zeit, 1901, 25, 345.
* Cf. Lunge, Sulphuric Acid and Alkali, vol. i., p. 246.
THE MANU FACTURE OF SALT CARE AND OF
HYDROCH LORIC ACID
By Professor G. LUNGE ; translated by JAMES T. CONRoy, B.Sc., Ph.D.
A. BRINE AND SALTWORKS
BRINE is regularly tested by the hydrometer both at the spot where
it is obtained (bore-holes, etc.), and in the works where it is employed
(saltworks, ammonia-Soda works, etc.). The temperature at which the
readings are made should of course be taken into account, and the
accuracy of the hydrometer ascertained. The method of checking
hydrometers, together with a description of the errors usually found in
these instruments, more particularly the cheaper kinds, have already
been described (cf. p. 156 et seq.).
A saturated solution of sodium chloride boils at Io9'.4, and contains
at this temperature between 29.4 and 29.5 per cent of the salt.
The relationship between the specific gravity of salt solutions and
the percentage content of salt, at 15° C., is given in the following
table:— -

Specific Gravity. P“gººse Specific Gravity. | **:::::ge Specific Gravity. Pegºsº
1:00725 1 1-07335 10 1°14315 19
1*01450 2 1-08097 11 1-15107 20
1*02174 3 1*08859 12 1°15931 21
1*02899 4 1-09622 13 1-16755 22
1*03624 5 1*10384 14 1°17580 23
1*04366 6 1*11146 15 1*18404 24
1*05108 7 1*11938 16 1-19228 25
1*05851 8 1-12730 17 1-20098 26
1*06593 9 1*13523 18 1°20433 26°395
H. C. Hahn has more recently determined the specific gravity of
solutions of sodium chloride, reducing the values to water at 4°, and
observing all possible precautions. Owing to the very special refinements
* /. Amer. Chem. Soc., 1898, 20,621 ; Chem, 6entr, 1898, II., 699.
397
398 MANUFACTURE OF SALTCAKE AND HYDROCHLORIC ACID
used in the determinations, they are hardly applicable to technical
conditions of work.
ANALYSIS OF BRINES AND MOTHER LIQUORS
The following determinations are made :-
1. Specific gravity, by means of a densineter or a hydrometer; in
the latter case the hydrometer readings are reduced to sp. gr. by means
of the table given on p. 16O.
2. Total Chlorine (eapressed as sodium chloride). Ten c.c. of the brine
are diluted to IOOO c.c.; of this IO c.c. are titrated with silver nitrate
solution (cf. p. 401). *
3. Sulphuric acid. Fifty c.c. of the brine are rendered acid by the
addition of a few drops of hydrochloric acid, diluted with an ap-
proximately equal volume of water and, when hot, treated gradually
with a hot solution of barium chloride. The settled precipitate is
repeatedly washed by decantation with hot water, acidulated with
hydrochloric acid, and finally well washed on the filter (cf. p. 274).
4. Owides of Iron, Aluminium, Calcium, and Magnesium. Two hundred
and fifty c.c. of the brine are warmed with a little nitric acid, ammonia
added in excess, and the whole heated for some time; the precipitate
is filtered, dissolved in hydrochloric acid, reprecipitated, and ferric oxide
and alumina determined in the usual manner. The calcium and
magnesium are determined in the filtrate as described on p. 405. The
calculation of the analytical results is made in exactly the same manner
as in the analysis of salt (p. 402).
5. Bicarbonates of ferrous Iron, Calcium, and Magnesium. Five
hundred c.c. of the brine are boiled fairly vigorously in a tall beaker or
Erlenmeyer flask, fresh water added, and the boiling and evaporation
repeated several times. Finally, hot water is added to the highly con-
centrated solution to dissolve any separated salt, the solution filtered, the
insoluble residue washed with hot water, dissolved in hydrochloric acid,
and precipitated with ammonia solution, free from carbonate. The
precipitation is repeated if necessary. The precipitate of ferric
hydroxide is filtered off, dissolved whilst still moist in dilute sulphuric
acid (1 : 4), and the undiluted reduced solution titrated with per-
manganate solution, and the volume of permanganate necessary to
produce a faint rose coloration in the solution alone deducted from
the reading obtained. Calcium and magnesium are determined in
the filtrate as described on page 405 et seq. The bicarbonates found
are entered in the analysis as carbonates, and the quantity so found
must be deducted from the total calcium, magnesium, and ferric
oxide obtained under 4, before calculating the rest of the calcium and
magnesium to Sulphates or chlorides. Calcium bicarbonate is, as a rule,
SODIUM CHLORIDE (ROCK SALT) 399
present to an appreciable extent in brine (from O2 to O.5 g. and
more per litre); on the other hand, magnesium bicarbonate appears
in much smaller quantity, and its estimation may, as is usually the
case in works testing, be neglected.
6. Water. The percentage of water is arrived at by deducting the
sum of the weights of the solid constituents present in a litre, from the
weight of the litre of brine.
The results are calculated on the litre of brine or mother liquor.
B. SODIUM CHLORIDE (ROCK SALT)
Sodium chloride melts, according to Carnelly, at 772°; according
to v. Meyer and Riddle,” at 815°.4.
The commercial products may contain, in addition to moisture and
water of hydration combined with calcium sulphate, etc., the following
impurities: chlorides of calcium, magnesium, and potassium ; sulphates
of calcium, sodium, magnesium, and potassium ; magnesium carbonate,
organic substances (bitumen, mineral oil), gaseous hydrocarbons, and
clay. According to the nature of the clay content, which seldom
exceeds o. I per cent, the salt is coloured usually a bluish green, but
sometimes yellow, brown, reddish brown, or greenish (Schwarzenberg).
Other impurities which occur less frequently and only in traces, are:
potassium bromide, potassium iodide, and lithium chloride, all of which,
in consequence of their relatively greater solubility, accumulate in the
mother liquors. Thus Krauch found in the concentrated mother
liquors of brine from Werl 3:3754 g. KBr, O.OI37 g. KI, and 8.9833 g.
LiCl, per litre. Magnesium borate is found admixed with rock salt at
Stassfurt. Occasionally the percentage of soluble foreign salts present
is so great that the rock salt is rendered useless for food purposes. -
An appreciable percentage of magnesium chloride (O.2 per cent, is
sufficient) gives a sharp saline flavour to the salt and renders it hygro-
scopic. Distinct amounts of calcium chloride also make the salt
hygroscopic. .
The following methods of analysis apply to ordinary salt, and to
pure sodium chloride intended for analytical purposes, respectively.”
I. ORDINARY SALT
Sampling. For the method of taking an average sample, cf. p. 13.
Qualitative analysis. The salt is, from time to time, examined
* J. Chem. Soc., 1879, 35, 280. * Ber., 1893, 26, 2447.
* A description of the analysis of “Denatured Salt” is included in the German edition, vol. i.,
p. 409.
400 MANUFACTURE OF SALTCAKE AND HYDROCHLORIC ACID
qualitatively for potassium, alkali bromides, and iodides, and, when it is
intended for culinary purposes, also for metallic salts (lead, copper, tin)
which may have been taken up from the vessels used in the purification
process. These metals are detected in the usual manner. A simple
method of examining for the other impurities mentioned consists in
extracting a considerable quantity of the salt with water, using an
insufficient quantity to effect complete solution, evaporating the filtered
extract to one-third of its bulk, again filtering, and dividing the filtrate
into two portions. To the first portion platinic chloride is added and the
mixture thoroughly shaken; a lemon yellow precipitate indicates the
presence of potassium chloride. The second portion is treated with
chlorine water, drop by drop, and shaken with chloroform before each
fresh addition; any iodine present is liberated first, the bromine subse-
quently, and each may be recognised in turn by the colour imparted to
the chloroform. (Cf. also p. 404 for Krauch's method of testing for iodine.)
Quantitative analysis. For works’ purposes a shortened method of
analysis is usually adopted. The constituents generally estimated are
moisture, total chlorine, expressed as sodium chloride, sulphuric acid,
calculated to calcium sulphate, and matter insoluble in water. These
figures are amplified periodically by the estimation of calcium,
magnesium, ferric oxide, and of the percentage insoluble in hydro-
chloric acid whereby the percentage of sand and clay is obtained, and
also, by difference, the approximate amounts of calcium and magnesium
carbonates present (cf. infra). The quantities of magnesium chloride
and of sodium sulphate, or calcium chloride, present may be arrived
at by the following simple methods, which will be found sufficiently
accurate for technical purposes.
I. Water.—(a) Moisture. Five g. of the salt are heated in a
platinum crucible, well covered to prevent loss by decrepitation, at first
very gradually and finally for some minutes at a low red heat. The
loss in weight gives the total water present, that is, the moisture proper
plus that chemically combined; the amount of the latter is usually very
small. This method has the disadvantage that, no matter what care be
taken, a small loss of salt always occurs owing to the unavoidable
decrepitation. By employing as new a platinum crucible as possible,
with smooth walls and a well-fitting lid, and placing the stand, Bunsen
burner, etc., on a sheet of black glazed paper, the loss, which may reach
I per cent. or over, is very much diminished; but the method lacks
certainty and cannot be carried out quickly or without direct attention.
In works where dozens of water estimations, including those of damp
salt, which may contain up to 15 per cent. of water, have to be made,
the inaccuracy of the above method is very pronounced, especially in
cases where the salt contains over 7 or 8 per cent of water, as the decrepi-
tation is then greater.
SODIUM CHLORIDE (ROCK SALT) 401
Böckmann consequently recommends the following method, which,
whilst perfectly accurate, permits of a large number of water determina-
tions being made simultaneously and without special attention during
the drying process.
A small, perfectly dry Erlenmeyer flask, about 14 to 15 cm. high,
and of about 250 c.c. capacity, is weighed together with a dry funnel
inserted in the neck. About 5 g. of salt are then introduced into the
flask, this quantity being just sufficient to form a thin layer on the
bottom, and the flask, funnel, and salt accurately weighed. The object
of the funnel is to allow the dried salt to be cooled in the open
instead of in a desiccator, and to prevent loss of salt by decrepita-
tion during the expulsion of the combined water described under ö.
The flask is then placed on a previously warmed, portable sand-bath,
40 cm. by 20 cm., and large enough to carry eight flasks, the tempera-
ture of the sand being about 140° to 150°. The funnels are removed,
and placed, each near its corresponding flask, on the bench. The water
soon begins to come off, condensing at the start on the upper portions
of the flasks; at the end of three to four hours all the water present as
moisture is driven off from the salt without any crackling sound. The
funnels are replaced in the flasks, which are then allowed to cool on a
thick glass or marble slab, and weighed.
(b) Chemically combined Water. After the determination of the
moisture the flasks are carefully heated on a wire gauze, or on a metal
(aluminium) plate, over the free flame, with the funnels in position, to
prevent loss by decrepitation. The water held in chemical combination
by calcium sulphate, etc., is thus driven off; the evolution of this small
quantity of water is accompanied by violent crackling. Since the
amount of combined water present is very frequently under O. I per
cent, and seldom more than 4 to $ per cent., its determination can
generally be omitted in technical analyses.
2. Total Chlorine (expressed as Sodium chloride).-Teng. of the
finely ground average sample are dissolved in lukewarm water, and
after cooling, made up to 500 c.c. at I 5° in a graduated flask. Ten c.c.
of this solution, corresponding to O-2 g. salt, are diluted with IOO c.c.
of water, IO drops of a Io per cent, potassium chromate solution added,
and the whole titrated with silver nitrate solution (p. 123). A deduc-
tion of O2 c.c. silver nitrate solution is made from the observed reading,
since this quantity is necessary to produce the red colour with the
chromate. Each I c.c. N/Io silver nitrate solution equals O.OO5850 g.
NaCl. If 5.850 g. of salt are dissolved to 500 c.c., and 25 c.c. of this
solution taken for the titration, the percentage of sodium chloride in the
salt is obtained on multiplying by 2 the number of c.c. taken.
3. Sulphuric acid.—Ten g. of the sample are dissolved in lukewarm
water and a little hydrochloric acid added; should much earthy matter
2 C
402 MANUFACTURE OF SALTCAKE AND HYDROCHLORIC ACID
be present the salt should be digested for some time with hydrochloric
acid, to ensure complete solution of the calcium sulphate present. The
cooled solution is then made up to 500 c.c. and passed through a dry,
pleated filter. Two hundred and fifty c.c. of the filtrate are heated to
boiling and precipitated by the addition of hot barium chloride solution
in slight excess. The barium sulphate is first washed by decantation
with water acidulated with hydrochloric acid, and finally on the filter
(ºf p. 274).
4. Matter insoluble in water and acids, Calcium and
Magnesium.—Fifty g. of the finely ground average sample are dis-
solved in warm water and filtered through a filter paper previously tared
against one of equal weight (p. 28). The insoluble matter is carefully
washed from the filter into a small glass mortar and ground with sufficient
water to dissolve all the gypsum present. The water is decanted and the
operation repeated a few times. The insoluble matter is again collected
on the filter and dried at Ioo”; the weight obtained represents the
matter insoluble in water (clay, sand, oxide of iron, calcium carbonate,
etc.). This residue is then treated on the filter with warm dilute hydro-
chloric acid, the solution obtained precipitated by the addition of
ammonia, the precipitated ferric hydroxide dissolved in dilute sulphuric
acid (I : 4), and the iron estimated, after reduction by zinc, by titration
with permanganate solution. The direct titration of the hydrochloric
acid extract, after reduction, in presence of manganese sulphate (2O c.c.
of a 20 per cent. Solution), yields less accurate results owing to the lack
of sharpness of the end-reaction. The residue insoluble in hydrochloric
acid is well washed with water, dried at IOO", and weighed to give the
percentage of sand and clay. The difference between the sum of the
weights of the sand, clay, and iron oxide, and the weight of the total
matter insoluble in water, may usually be regarded as representing the
carbonates of calcium and magnesium present.
The calcium and magnesium are determined in the filtrate from the
insoluble matter in the usual manner after addition of ammonium
chloride and ammonia (cf. p. 405).
Calculation of Results. All magnesium found is calculated to
magnesium chloride and its equivalent in sodium chloride deducted
from the “total" sodium chloride, to give the actual content of the
latter. Should more sulphuric acid be found than corresponds to the
soluble calcium, such excess is calculated to Sodium sulphate. In the
reverse case the excess of calcium is calculated to calcium chloride, and
a corresponding deduction made in arriving at the percentage of sodium
chloride.
J. and S. Wiernik 1 recommend a direct determination of the
magnesium chloride present in sodium chloride and in brine, and they
* Z, angew. Chem., 1893, 6, 43.
SODIUM CHLORIDE (ROCK SALT) 403
further state that wholly incorrect results may be obtained by the
usual method of calculation. They extract the dried salt with absolute
alcohol, which dissolves out magnesium chloride only, and, after evapo-
rating off the alcohol, estimate either the magnesium as pyrophosphate
or the chlorine by titration. Both methods should give the same result
when expressed as magnesium chloride. The total calcium, magnesium,
chlorine, and sulphuric acid are estimated in the original solution in the
usual manner. The sulphuric acid found is first calculated on the
calcium, then on the excess of magnesium that may be present above
that found as magnesium chloride by extraction with alcohol; any
sulphuric acid still remaining is regarded as existing as sodium
sulphate. Chlorine found as magnesium chloride is deducted from the
figure obtained for “total chlorine,” and the difference stated in terms
of sodium chloride.
All such calculations are quite illusory in the case of dilute solutions
where the salts are either almost completely ionised or in part dis-
sociated hydrolytically; in the case of solid salts and concentrated
solutions such as the above, however, the greater part of the salts are
actually present as such.
2. PURE SODIUM CHILORIDE FOR ANALYTICAL PURPOSES
According to Krauch, pure sodium chloride appears under the
following names: Natrium chloratum chem. pur., matrium chloratum
puriss., natrium chloratum puriss. exsicc., natrium chloratum puriss.
fus. The natr. chlorat, chem. pur. is absolutely pure; the various
varieties of natr. chlorat. “puriss.” are generally sufficiently pure
for analytical purposes, but not always so, a point to be borne in mind
in standardising silver nitrate solution. These preparations usually
contain traces of sulphates of calcium or magnesium. Kubel' found
magnesium chloride and ammonium chloride in commercial natr.
chlorat. pur.
Krauch's method of examining natr. chlorat, chem. pur. is as
follows:—
Complete Solubility and freedom from Sulphate.—Three g. of salt
and 20 c.c. of water should yield a clear, neutral solution, which, when
diluted to 80 c.c., heated to boiling, and treated with barium chloride,
should show no signs of a precipitate even after standing for several
hours.
Examination for Alkaline Earths and heavy Metals.-Three g. of
the salt are dissolved in 50 c.c. of water, the solution heated to boiling,
and ammonium oxalate, sodium carbonate, and ammonium sulphide
added. No turbidity should result.
* Arch. Pharm., 1888, 226,440.
404 MANUFACTURE OF SALTCAKE AND HYDROCHLORIC ACID
Iodine.—Twenty c.c. of the aqueous solution (I : 2O) treated with one
drop of ferric chloride solution and a little starch solution should show
no blue coloration.
Potassium.—The concentrated solution of the salt should give no
precipitate on addition of platinic chloride even after long standing.
C. SULPHURIC ACID
The examination of sulphuric is conducted as previously described
(p. 359 et seq.). The acid employed in the manufacture of Saltcake is
either non-purified chamber acid, Glover acid, or acid concentrated
in pans placed over the pyrites burners. The acid employed in the
preparation of saltcake intended for use in the manufacture of the better
qualities of glass should be as far as possible free from iron. The
estimation of iron is described above (p. 380 et seq.).
D. SALTCAKE
The saltcake drawn from the furnaces is spread out in long ridges
on a stone-paved floor to cool. Samples are taken morning and
evening, corresponding to the output of each furnace, from various places
in the several rows, and an average of the lots so taken ground in a
suitable mill.
The judging of the sample by its appearance is described later
under the Leblanc soda process (p. 422).
For works’ purposes, estimations I and 2 given below are sufficient;
for purposes of sale or purchase, the other tests given should be made.
I. Free Acid.—Twenty g. Of the saltcake are dissolved in water and
made up to 250 c.c.; 50 c.c. of this solution are titrated with normal
sodium hydroxide solution, using litmus or methyl orange as indicator.
Each I c.c. alkali corresponds to I per cent. SOs. All acidity, whether
arising from HCl, NaHSO, or from iron and aluminium salts of acid
reaction, is calculated to SOs. Should large quantities of iron and
aluminium salts be present, and it be desired to avoid the influence of
these in this determination, no special indicator is added to the solution,
and the end-reaction is judged by the first appearance of a permanent
flocculent precipitate, which occurs as soon as the free acid and the acid
sulphate have been neutralised by the alkali.
2. Sodium chloride.—Fifty c.c. of the solution prepared under I are
exactly neutralised by the addition of the necessary volume of normal
sodium hydroxide solution, and the neutral solution titrated with
M/IO silver nitrate solution after the addition of a little potassium
SALTCAKE 405
chromate solution (cf. p. 123). Each I c.c. silver nitrate solution, after
deducting o-2 c.c. from the total as allowance for the end-reaction,
corresponds to O. 146 per cent. NaCl. Or a solution containing
2.9054 g. AgNO, per litre may be employed, each I c.c. of which
corresponds to O.OOI g. NaCl, or, with the above amount of saltcake, to
o.O25 per cent. NaCl. This weaker silver nitrate solution should
always be used in the case of high-grade Saltcake.
3. Iron.—Ten g. of the sulphate are dissolved in water, the iron
reduced by addition of zinc and sulphuric acid and titrated with
permanganate solution (p. 380). Very small amounts are estimated
colorimetrically (p. 381). According to Ost, sulphate manufactured in
lead pans contains from O,OO9 to O.O29 per cent. Fe; that manufactured
in iron pans O.O62 to O. I 30 per cent. Fe.
4. Insoluble Matter when present is estimated in the usual way.
5. Calcium.—Ten g. of the saltcake are dissolved in water, a little
hydrochloric acid being added if necessary, and the solution treated with
ammonium chloride and ammonia, the latter in slight excess. The
annmoniacal solution is heated to boiling and the calcium precipitated
by the addition of a boiling solution of ammonium oxalate in consider-
able excess. The solution is allowed to stand for twelve hours, and the
precipitate then washed with boiling water several times by decantation
and finally on the filter, a small quantity of ammonium oxalate being
added to the wash-water. The filter paper and precipitate are dried
and burnt off in a platinum crucible (this may be done whilst the
precipitate is still slightly moist), after which the whole is strongly
ignited for twenty minutes over a Müncke or Tecluburner. Should
the precipitate be large in amount, which is not likely to be the case, it
must be heated in the blowpipe flame until the weight becomes
constant. The precipitate is regarded as CaO ; any ferric oxide present
must, of course, be deducted. The desiccator used must be protected, to
prevent access of moisture and carbonic dioxide (p. 29).
Should any considerable quantity of magnesium be present—which is
never likely to occur, however, in saltcake—the washed calcium oxalate
precipitate must be transferred to a beaker, dissolved in a little warm
hydrochloric acid, and reprecipitated by the addition of 2 to 3 c.c. of
ammonium oxalate followed by ammonia in slight excess. The precipi-
tate is allowed to settle for twelve hours and the reprecipitated calcium
oxalate treated as above. The two filtrates are then combined for the
magnesium determination.
Instead of converting the calcium oxalate to oxide and weighing, it
has been recommended to dissolve the precipitate in dilute sulphuric
acid and to titrate with permanganate solution, calculating the calcium
from the oxalic acid thus found. Of course, in this case it is inadmis-
* Z, angew. Chem., 1896, 9, 9.
406 MANUFACTURE OF SALTCAKE AND HYDROCHLORIC ACID
sible to add ammonium oxalate to the water employed for washing; on
the other hand, such omission may lead to a loss which is by no means
negligible. Walland" overcomes this difficulty by standardising the
permanganate solution against calcium oxalate prepared from pure
calcium oxide obtained by ignition of the carbonate, and treated in
exactly the same way as the precipitate obtained in the actual analysis,
so that the errors are the same in the two cases.
Such a method always has disadvantages, and is only to be
recommended when, as in technical work, a large number of analyses
have to be made simultaneously; when this is not the case it offers no
saving of time as compared with the ordinary method by ignition.
6. Magnesium is estimated by addition of ammonium phosphate to
the filtrate from operation 5. For very exact work this addition should
be made in neutral solution and as far as possible in the absence of
ammonium salts, the necessary ammonia being added subsequently.
The filtrate from 5 should therefore be evaporated to dryness and, after
gentle ignition to drive off the ammonium salts, extracted with a
small quantity of hydrochloric acid. The solution is filtered from any
separated carbonaceous matter, exactly neutralised with ammonia,
heated to boiling, and sodium phosphate solution added, drop by drop,
to the boiling solution until no further precipitation results. The
mixture is allowed to cool, ammonia Solution equal to one-third of its
volume added, and the whole allowed to stand for three hours, or should
only a small amount of magnesium be present, for twelve hours, before
filtering. The precipitate is washed with 2% per cent, ammonia solution,
dried, separated as completely as possible from the filter paper, the
latter burnt separately in a spiral of platinum wire, and the ash and
precipitate then ignited, first at a low heat and finally strongly. The
precipitate is thus converted to pyrophosphate, Mg,F,C), one part of
which corresponds to O-3624 g. MgO.
In the analysis of saltcake it is usual to precipitate the magnesium by
direct addition of sodium phosphate to the filtrate from 5 without
removing the ammonium oxalate, and observing the other precautions
given; in this case the precipitate should be allowed to stand twenty-
four hours before filtering.
This method of separating calcium and magnesium is essentially in
accord with the conditions given by Treadwell.” An alternative method,
differing somewhat from the above, has been described by T. W.
Richards.” The magnesium salts must not be present in the solution in
greater concentration than corresponds to ºr normal. Ten equivalents of
ammonium chloride are added to the solution, followed by Oxalic acid
in sufficient quantity to combine with the whole of the calcium. It is
* Chem. Zeit., 1903, 27, 922. * Analytical Chemistry, vol. ii., p. 70.
* Z, amorg. Chem., 1901, 28, 88.
SALTCAKE 407
advantageous to reduce the dissociation of the oxalic acid by the
previous addition of three or four equivalents of hydrochloric acid. A
drop of methyl orange is added, the solution boiled, and very dilute
ammonia added gradually at suitable intervals and with continued
stirring; the addition of the ammonia solution should occupy half an
hour. When the solution has been made neutral, a large excess of
ammonium oxalate is added, and the whole allowed to stand for four
hours. The precipitated calcium oxalate is thoroughly washed with
water containing ammonium oxalate. The filtrate contains the whole
of the magnesium with the exception of O. I to O.2 per cent., which
causes a corresponding error in the calcium, the whole of which is con-
tained in the precipitate.
Heraeus' states that platinum crucibles are easily injured by the
ignition of magnesium ammonium phosphate, both in the case of
pure platinum and of platinum-iridium. The effect takes place, how-
ever, in a very irregular manner, and is to be attributed to the
liberation of phosphorus, which occurs at 900°, arising partly from the
presence of carbon or reducing gases, and also partly from the hydrogen
liberated by the decomposition of ammonia at the high temperature.
The action due to ammonia is stronger if free ammonium phosphate be
present in the precipitate, and may very easily occur if Gooch crucibles
are used in which old precipitates have been allowed to remain.
The action is most marked when the heating is carried out rapidly in
a covered crucible. Up to the present it has not been possible to define
the conditions under which the crucibles are least attacked.
A rapid, and for many purposes sufficiently exact, method of deter-
mining calcium and magnesium consists in adding sodium carbonate
to the filtrate from the silica, ferric oxide, and alumina, evaporating to
dryness, igniting to remove the ammonium salts, taking up with sodium
carbonate solution, extracting with hot water and weighing the calcium
and magnesium carbonates remaining in the dish after drying at 200°;
or by weighing as CaO + MgO after ignition. The quantity of each
base present may be determined indirectly from the difference in weight
between the CaCO3 + MgCOs and the CaO + MgO ; it is, however, better
for this purpose to convert both into the sulphates. Christomanos” has
applied this method to the analysis of a large number of samples of
magnesite. It must, however, be borne in mind that this, like all
other indirect methods, may lead to somewhat large errors.
7. Aluminium.—A considerable quantity of ammonium chloride or of
ammonium nitrate is added to the solution of the saltcake, the mixture
heated nearly to boiling in a porcelain or platinum dish and ammonia
added, but not in too great excess, as the removal of an excess by
boiling off is not only unnecessary but harmful; the precipitate is then
* Z. angew. Chem., 1902, I5, 917. * Z, anal. Chem., 1903, 42, 606.
408 MANUFACTURE OF SALTCAKE AND HYDROCHLORIC ACID
washed three times by decantation with hot water to which one drop of
ammonia has been added, washed out on to the filter paper, thoroughly
washed with hot water, and dried on the pump (cf. p. 273). Any small
quantities adhering to the sides of the dish are wiped off with a piece
of filter paper. The filter paper and precipitate may be ignited, whilst
still moist, in a platinum crucible, and finally ignited strongly, pre-
ferably over the blowpipe. The weight of ferric oxide, as found under
3, must be deducted from the weight obtained.
The ammonia employed must be tested by barium chloride for
freedom from carbonate, which, if present, would cause precipitation
of the calcium ; if necessary, it should be purified by distillation over
lime.
8. Sodium sulphate.—One g. of the sample is dissolved in water,
the calcium, together with the iron, precipitated as under 5, the solution
filtered, and the filtrate, after the addition of a few drops of pure
sulphuric acid, evaporated to dryness. The dried residue is ignited,
first alone, and again after the addition of a small piece of ammonium
carbonate, and weighed. From the weight obtained there must be
deducted (a) the sulphate equivalent of the sodium chloride found under
2 (I-OOO g. NaCl = I-215 g. NagsO, or I c.c. N/IO silver nitrate solution
= O.OOI77 g. NassO4), and (b) the magnesium sulphate equivalent of the
magnesium found under 6 (I-OOO g. MgO = 2.9836 g. MgSO4). The residue
represents the Na2SO, actually present in the original I g. of saltcake.
Isbert and Venator” proceed as follows:–About 2 g of the saltcake
are dissolved in the least possible volume of hot water; ferric oxide,
alumina, calcium carbonate, and magnesium carbonate precipitated by
the addition of ammonia and ammonium carbonate, the precipitate
dissolved in hydrochloric acid and reprecipitated, the precipitate well
washed with hot water, and the total filtrate collected in a platinum
dish. The filtrate—which is about IOO c.c. in volume and contains, in
addition to the sulphate, the free acid and sodium chloride—is treated
with ammonium sulphate or sulphuric acid, and evaporated on the
water-bath to convert the chloride into sulphate. The residue is gently
ignited, to volatilise all ammonium salts, and weighed. From the
weight of NagsO, so obtained the weight of sulphate corresponding to
the chloride present in the original sample must be deducted.
Koninck's and Grossmann's methods are described in Lunge's
Sulphuric Acid and Alkali.” The latter is, according to the author
himself, subject to a constant, inexplicable error of 1.3 per cent, and
consequently cannot be recommended for technical work.
According to Fenton,” the total sodium may be estimated in the
following manner. An excess of the potassium salt of dihydroxy-
* Z. angew. Chem., 1889, 2, 66. * Vol. ii., pp. 91 and 92.
* /. Chem. Soc., 1898, 73, 167.
HYDROCHLORIC ACID 409
tartaric acid is added to a concentrated neutral Solution of the sodium
salt to be examined, and the mixture maintained at O’ for half an hour.
The precipitated sodium salt of dihydroxytartaric acid is filtered off,
washed with ice-cold water, dissolved in excess of dilute sulphuric acid,
and titrated with permanganate solution, which readily oxidises the
dihydroxytartaric acid. Magnesium salts do not interfere with the
reaction, but ammonium salts must be excluded.
A. HYDROCHLORIC ACID
The daily control of the process in the works is confined to measuring
the strength of the acid flowing from the condensers, jars, etc., and to
determining the completeness attained in the absorption. The strength
of the acid is measured by means of the hydrometer, and in some
works tests are only taken once in the day. It is, however, preferable
to allow the acid, as it leaves the plant, to flow through a glass cylinder
in which the hydrometer floats; in this way the strength of the acid can
be seen at any moment without drawing a sample.
It is especially important to check the amount of acid which escapes
condensation and so passes into the atmosphere. Should the air be
damp a practised eye can, to a certain extent, judge the degree of
absorption by the appearance of the escaping vapours. Whilst white
fumes are continually evolved from the open pipes of the pan con-
densers, even when condensation is complete, these consist entirely of
steam, and may be distinguished from hydrochloric acid fume by the
ease with which they are dissipated in the air. Hydrochloric acid
vapour, on the other hand, forms dense white fumes, which, in a moist
atmosphere, spread to form a heavy cloud over a considerable area,
and persist for a considerable time. Very frequently the fumes only
become apparent when the gases come into contact with the outer air.
According to the Alkali Act (1906), 95 per cent. of all the hydro-
chloric acid produced in a works must be condensed, and no gases are
permitted to escape into the atmosphere which contain more than
# grain HCl per cubic foot (= O-457 g. per cubic metre), and the total
acidity of all the gases present must not exceed the equivalent of
4 grains SO3 per cubic foot (=9. I5 g. per cubic metre). In these
regulations the volume of the gases is reduced to 60° F (15°. 5 C.) and
3O inches mercury (almost exactly 760 mm.).
Examination of the Exit Gases.—The examination of the chimney
gases is carried out in a Fletcher bellows, which serves both as aspirator
and absorbing vessel. They are constructed to draw ºr of a cubic foot
of gas at one aspiration, but they should in all cases be standardised
by filling with air at the normal working capacity and measuring the
410 MANUFACTURE OF SALTCAKE AND HYDROCHLORIC ACID
volume aspirated by expelling it into an inverted graduated vessel
filled with water, correcting the volume obtained for temperature and
pressure. In examining chimney or other gases the bellows are con-
nected with the chimney by means of a porcelain, glass, or platinum
tube, of I2 mm. diameter, which extends some considerable distance
into the chimney. Both bellows and tube are first washed out with
distilled water, 200 to 300 c.c. of distilled water then introduced, and the
necessary number of aspirations made. The contents of the bellows
are well shaken after each aspiration to allow all the acids present to
be dissolved by the water. When the operation is complete, a little
water is forced into the connecting tube and allowed to flow back into
the bellows to wash out any acid that may have condensed in the tube.
The liquid in the bellows is then washed into a porcelain dish, and if
necessary filtered from soot. Any sulphurous acid present is oxidised
by potassium permanganate, excess of the latter removed by a trace of
ferrous sulphate, the solution neutralised by pure sodium carbonate, a
little potassium chromate added, and the whole titrated with W/Io or
M/IOO silver nitrate solution. Each I c.c. M/IO silver nitrate solution =
O.OO3646 g. HCl. - -
In the Alkali Inspector's Report for 1898 (No. 35) an addition of
hydrogen peroxide, free from chlorine, to the water put into the
bellows was recommended. This addition effects the immediate oxida-
tion of any sulphurous acid present in the gases. The total acidity is
determined by titration with sodium carbonate solution, and subse-
quently the chloride as above, by means of silver nitrate. Certain
difficulties are liable to occur when working in this manner. Thus
discoloration may arise owing to incomplete oxidation of organic
matter by the hydrogen peroxide, and this may under certain conditions
lead to the reduction of the chromate; further, it is difficult to obtain
hydrogen peroxide free from chlorine. For these reasons the process
has been modified, and is now carried out as follows. The total acidity
is determined as before with sodium carbonate solution and methyl
orange, a few drops of potassium permanganate solution being added
in cases where the solution is very dark owing to the presence of sooty
matter. The neutralised solution is treated with O.5 g. of calcium or
magnesium carbonate, followed by 5 to IO drops of a 5 per cent, ferrous
sulphate solution, the mixture stirred for a minute and then decanted or
filtered. The chloride is then estimated in the filtrate in the usual
manner. The addition of ferrous sulphate gets rid of the organic
matter which is carried down with the ferrous carbonate precipitate, and
so gives a neutral solution in which the hydrogen peroxide will not
exert any reducing action on the chromate; also, it precipitates the
arsenic and copper which are found in testing the gases from copper-
works.
HYDROCHLORIC ACID 411
The potassium permanganate, added to oxidise the organic matter,
must be employed with caution, since the manganese sulphate produced
may reduce the chromate, with the production of a green-coloured
solution; this will, however, not occur if the solutions are neutralised
as described.
Should any of the difficulties above referred to be met with, it is
best to oxidise with nitric acid and estimate the chloride by Volhard's
method (p. I23).
A continuous test may of course be made, as in the case of the
exit gases of the sulphuric acid process (p. 335), by employing a large
aspirator and selecting a suitable type of absorption apparatus (cf.
p. 302); the flasks described in the Alkali Inspector's Annual Report
for 1898 are specially suitable for this purpose.
EXAMINATION OF THE GASES IN THE HARGREAVES’ PROCESS
In this process" the burner gases from pyrites or blende are drawn,
together with steam, through hot salt, packed in iron cylinders, whereby
the sodium chloride is converted into sodium sulphate, the sulphur
dioxide being gradually absorbed and hydrochloric acid liberated. The
progress of the reaction is followed by withdrawing and testing samples
of the gases passing through the connecting pipes between adjacent
cylinders. The tests are made in the following manner:— -
(a) Total acidity, best estimated by Lunge's method (p. 299).
(b) Sulphur dioxide, by Reich's method (p. 3OO).
(c) Hydrochloric acid. The test is made in the sample taken for
test a by titration with silver nitrate, either by Mohr's method (p. 123),
or by Volhard's method (p. 123).
The content of sulphur trioxide is obtained by deducting b-i-c
from a.
For continuous work test, a or ò may be omitted; either of these in
combination with test c is sufficient.
PROPERTIES OF HYDROCHLORIC ACID
The strength of the acid is usually determined by the hydrometer.
The use of the Twaddell scale possesses the advantage that for acids of
ordinary strengths the number of the degrees Tw. and the percentage
of acid in the acid are represented by practically the same figure. The
following table, drawn up by Lunge and Marchlewski,” gives for the
various specific gravities of pure hydrochloric acid the corresponding
degrees on the Twaddell scale, the percentage content in HCl, the
* Cf. Lunge, Sulphuric Acid and Alkali, vol. ii., p. 238.
* Z. angew. Chem., 1891, 4, 135.
412 MANUFACTURE OF SALTCAKE AND HYDROCHLORIC ACID
percentage content in acid of 28°.5 Tw. and of 30°4 Tw., and the weight
of HCl per litre of the acid expressed in grams and per cubic foot
expressed in pounds.
Specific Gravity of Hydrochloric Acid Solutions at 15° C., com-
(Lunge and Marchlewski.)
pared with Water at 4°, and reduced to vacuum.

100 parts by Yº. ºpond to parts
Specific y weignt O 1 litre e
Degrees Grºßg at * - contains 1 ...”
Twaddell. TAF Oršº, ofº, #. lbs. HC1.
(in vacuo). HC1. #. #; &
= 28°5 TW. =80°-4 TW.
0 1°000 0-16 0-57 0.53 1-6 0-10
I 1°005 1-15 4'08 3-84 I2 O-75
2 1*010 2-14 7-60 7.14 22 1:37
3 1-015 3-12 11°80 10°41 32 I '99
4 1*020 4°13 14.67 13°79 42 2°62
5 1-025 5-15 18°30 17-19 53 3:30
6 1*030 6-15 21°85 20-53 64 3-99
7 1*035 7-15 25*40 23.87 74 4°61
8 I "040 8-16 28'99 27-24 85 5°30
9 1-045 9-16 32°55 30°58 96 5'98
10 1*050 10-17 36°14 33-95 107 6-67
11 1*055 11-18 39.73 37-33 118 7-35
12 1 °060 12-19 43°32 40°70 129 8'04
13 1°065 13-19 46-87 44'04 141 8-79
14 1-070 14-17 50-35 47-31 152. 9°48
15 1-075 15-16 53.87 50-62 163 10°16
16 1 °080 16:15 57-39 53-92 174 10°85
17 1-085 17-13 60-87 57-19 186 11°59
18 1°090 18°11 64°35 60°47 197 12°28
I9 1.095 19-06 67-73 63-64 209 13°03
20 1°100 20-01 71°11 66°81 220 13-71
21 1°105 20-97 74°52 70-01 232 14°46
22 I "I LO 21-92 77-89 73-19 243 15°15
23 1*115 22-86 81°23 76°32 255 15'90
24 1*120 23.82 84°64 79-53 267 I6*65
25 1-125 24-78 88-06 82°74 278 17-33
26 1-130 25-75 91°50 85-97 291 18°14
27 I-135 26-70 94'88 89-15 303 18°89
28 1°140 27-66 98.29 92-35 315 I9 °64
29 1°145 28-61 101-67 95°52 328 20°45
30 1°150 29'57 105-08 98-73 340 21*20
31 1-155 30°55 108°58 102°00 353 22°01
32 1-160 31°52 112-01 105 °24 366 22°82
33 1.165 32°49 115*46 108°48 379 23.63
34 1-170 33°46 118-91 111-71 392 24°44
35 1-175 34°42 122-32 114-92 404 25°19
36 1*180 35-39 125-76 118-16 418 26°06
37 1-185 36-31 129°03 121:23 430 26°81
38 1°190 37-23 132°30 I24°30 443 27.62
39 1°195 38°16 135°61 127°41 456 28°43
40 1-200 39-11 138-98 130°58 469 29-24
The correct specific gravity at I 5°C. may be calculated from readings
made between 13° and 17° (and also for temperatures slightly below or
IMPURITIES IN HYDROCHLORIC ACID 413
above this range) by the aid of the accompanying short table. If the
observed reading is made below I 5° the values given in the table must
be deducted for each 1” below the I5°; for observations made above 15°
the corresponding values must be added.
Spec. Grav. I.OOO—I-O4O : + O.OOO2
y) I.O4I—I.O85: O-OOO3
jj I.O86—I. I 20: O-OOO4
3) I. I 2 I-I-I 55 : O-OOO5
}} I. I 56—I-2OO : O-OOO6
Kremers' has published a table giving the variations in the specific
gravity of hydrochloric acid over a temperature range from of to IOO’
(normal temperature 19° 5); the most recent table is that of Fuchs.”
DETECTION OF IMPURITIES”
1. Residue.* Ten g. should leave not more than a very minute and
scarcely weighable residue on evaporation.
According to Krauch, the preparation of absolutely chemically
pure hydrochloric acid is attended by considerable difficulties; he
invariably found, on evaporating 50 g. Of the acid in a porcelain dish, a
residue of about I mg., probably consisting of calcium oxide derived
from the porcelain vessel or from the sulphuric acid.
2. Sulphuric acid.* (a) Five g. are diluted with 50 c.c. of water, and
barium chloride added; no sign of a precipitate should appear after
twelve hours' standing,
(b) Five hundred g. are slowly evaporated on the water-bath till
only about I c.c. remains and the Sulphuric acid estimated in this
residue, an ash-free filter paper being employed. The weight of barium
sulphate found should not exceed 12 mg., corresponding to 1 mg. H2SO,
per IOO g. hydrochloric acid.
Krauch states that scarcely a single sample of the various makes to
be found on the market appears to be absolutely free from sulphuric
acid when tested as under & ; many of the samples examined gave the
sulphuric acid reaction with test a, that is, without evaporating off the
hydrochloric acid, showing, therefore, a greater degree of impurity than
should be accepted.
In commercial work the evaporation of the acid may be omitted and
much time saved by nearly neutralising the acid with pure sodium
carbonate (not with ammonia) before precipitating by addition of
barium chloride. Each one part by weight BaSO,-O-3429 parts SOs.
* Pogg. A mm., 1859, IoS, II5. * Z. angew. Chem., 1898, II, 753.
* The tests marked with an asterisk are due to Krauch and apply only to acid. hydrochloric.
purum. Conc.
414 MANUFACTURE OF SALTCAKE AND HYDROCHLORIC ACID
A rapid method for the approximate estimation of the sulphuric
acid present, well suited to works purposes, has been proposed by
Lunge. The method is based on the degree of turbidity produced by
addition of barium chloride, and is best carried out in the following
modified form proposed by Rürup." Glass tubes having a lower part 6
mm. diameter, sealed below, and an upper part I 5 mm. diameter,
closed with a rubber stopper, are used for the estimation; the cylin-
drical, narrower portion is 250 mm. in length, and is graduated in the
following manner. Acids containing varying proportions of sulphuric
acid, from O-4, O-6, etc., to 3-O per cent. SOs, are prepared, Io C.C. of
each heated to boiling, transferred to the tube, nearly neutralised
with strong ammonia, precipitated by addition of hot barium chloride
solution, and allowed to settle. A mark is then made on the tube corre-
sponding to the level of the settled precipitate and the corresponding
percentage of acid etched at the side. To carry out a test, Io c.c. of the
acid are nearly neutralised with ammonia, heated to boiling, transferred
to the tube, 5 c.c. of a saturated solution of barium chloride added, the
rubber stopper inserted, and the tube well shaken. Five ‘minutes are
allowed for settling before reading off the volume of the precipitate.
The method is said to be accurate to within O-O5 per cent.
3. Arsenic. The detection and estimation of arsenic in hydrochloric
acid is of importance not only in chemico-legal investigations but also
in numerous analytical and technical applications. The methods of
estimating arsenic in sulphuric acid apply also to hydrochloric acid (cf.
p. 362 et seq.).
(a) According to Krauch, hydrochloric acid which will satisfy the
following test may be looked upon as sufficiently pure for most
analytical work. Ten g. of the acid are diluted with Io c.c. of water
and the mixture carefully covered in a test tube with a layer of 5 c.c. of
freshly prepared sulphuretted hydrogen water; no coloration or yellow
ring should appear at the junction of the two solutions even after the
lapse of an hour. The test should be applied to both cold and hot
solutions so as to detect arsenic acid; O-OOOOO5 g. As in I g. acid can be
detected, that is, ºr mg. As in the quantity taken for the test.
(b) According to the German Pharmacopoeia, III., I c.c. of hydro-
chloric acid is treated with 3 c.c. of Stannous chloride solution, prepared
by rubbing to a cream five parts of crystallised stannous chloride and one
part of hydrochloric acid, Saturating with dry hydrochloric acid gas, and
filtering. No coloration should appear after the mixed solutions have
stood for one hour.
(c) Hager's Kramato-method” is extremely sensitive, and quickly and
simply carried out. One c.c. of the hydrochloric acid is diluted with
2 c.c. of water and a little ammonium oxalate added. A drop of the
* Chem. Zeit., 1894, 18, 225. * Pharm. Centr., 1884, 265.
IMPURITIES IN HYDROCHLORIC ACID 415
solution is then evaporated on a strip of brass (previously rubbed clean
with sand and water and dried), the heating being so regulated that no
ammonium salts are volatilised. If arsenic be present, a stain is produced
varying from grey through red to black, according to the quantity of
arsenic in the acid.
(d) A delicate method for the detection of arsenic has been described
by Schlickum." If a minute crystal (O.OI to O.O2 g) of sodium sulphite
is added to a solution of O-3 to O-4 g. stannous chloride in 3 to 4 g.
of hydrochloric acid, both sulphur dioxide and sulphuretted hydrogen are
evolved, the latter being due to the reducing action of the stannous
chloride on the sulphurous acid. If hydrochloric acid containing
arsenic be carefully introduced so as to form a layer on the top of the
Solution, a yellow ring of arsenious sulphide immediately forms at
the zone of contact. The reaction is given by ºr mg. of arsenious acid.
The ring gradually extends in an upward direction, and in the presence
of # mg. of arsenious acid the whole of the acid layer assumes a yellow
colour in the course of a few minutes. The reaction proceeds more
slowly when the arsenic is present in the arsenic condition. The
success of this test depends on the use of a minimum quantity of
Sodium sulphite and of strong hydrochloric acid, so as to prevent
the separation of antimony sulphide.
The two following tests are especially adapted for confirming the
complete absence of arsenic in hydrochloric acid, more particularly in
chemico-legal inquiries.
(e) Otto's method.” Several litres of the acid are treated with a few
crystals of potassium chlorate, and sufficient water added to reduce
the specific gravity to a maximum of I. IO4. The solution is evaporated
on the water-bath in a porcelain dish, the residue taken up with water
and examined by the Marsh-Berzelius test.
(f) Gutzeit's test (p. 374) is extremely delicate and allows the detec-
tion of Tºro mg. As2O3, but it involves the careful observation of the
conditions referred to in connection with sulphuric acid, since sulphu-
retted hydrogen, hydrogen phosphide, and antimoniuretted hydrogen
affect both silver nitrate and mercuric chloride papers in a similar
manner to arseniuretted hydrogen.
Quantitative determination of Arsenic. The arsenic present in
ordinary, non-purified, commercial hydrochloric acid, may be determined
quantitatively either by reducing 20 g. of the acid with sulphurous acid,
driving off the excess of the latter, neutralising with sodium carbonate,
and titrating with iodine or by precipitating with sulphuretted hydrogen,
and weighing the sulphide obtained after extraction with carbon bi-
sulphide. Neither method is free from objection, and in the latter
case it is difficult to be certain that the arsenic is actually present
* Analyst, 1886, 11, 19. * Ausmittelung der Gifte, 1884, p. 146.
416 MANUFACTURE OF SALTCAKE AND HYDROCHLORIC ACID
as pure As,Sa. Kretzschmar" therefore recommends the following
method:—
The highly diluted acid is nearly neutralised by addition of sodium
carbonate, ammonia and yellow ammonium sulphide then added,
followed by chemically pure hydrochloric acid in excess and a strong
current of sulphuretted hydrogen passed for two hours through the
solution, which is kept hot on the water-bath. By this treatment the
precipitation, which otherwise requires from fifteen to twenty-four hours,
is complete in the time stated. The precipitate of arsenious sulphide
is washed, dissolved in potassium hydroxide, aided by chlorine,
or preferably by bromine, and the arsenic finally precipitated in slightly
acid solution by the addition of ammonia and magnesia mixture and
weighed as magnesium pyroarsenate. This method may also lead to
erroneous results; should the ignition be too strong, arsenic may be
distilled off, whilst if too gentle, the conversion to the pyroarsenate is
likely to be incomplete.”
The following abnorinally high percentages of arsenic found by
Buchner in certain non-purified acids, viz., 2-4, 3. I, IO-4, O-7, 4:7, 5-7,
9.7 g. As per IOO kilos crude acid have been stated * to be quite mis-
leading, at all events for well-regulated works.
4. Iron.* Five g. diluted to 25 c.c. should yield no coloration on the
addition of a few drops of potassium thiocyanate solution.
For detecting traces of iron in strong acids, Venable * recommends
making use of the blue solution obtained on treating cobaltous nitrate
with strong hydrochloric acid. Traces of ferrous salts change the blue
colour of such a solution to green; ferric salts are without effect.
For quantitative work the iron is reduced with zinc, the solution
diluted with a large volume of distilled water, a 20 per cent. Solution
of manganese sulphate, free from iron, added, and the whole titrated
with AW/2O potassium permanganate solution (I-582 g. per litre), of
which I c.c. = O.OO2795 g. Fe. A blank experiment on an equal
volume of the distilled water, as used above, is made at the same time
and the volume of permanganate solution necessary to produce a faint
permanent rose tint deducted. Traces of iron are determined colori-
metrically as described above (p. 381).
5. Sulphurous acid. According to Krauch, the addition of a few c.c.
of the diluted acid should not destroy the faint blue colour of water to
which a little iodine and starch solution have been added. In the absence
of chlorine and ferric chloride, sulphurous acid may also be detected by
means of sulphuretted hydrogen (white cloudiness due to separated
sulphur) or by stannous chloride (brown precipitate of stannous sulphide).
For quantitative work, titration with permanganate or iodine solution
* Chem. Zeit., 1891, 15, 269. * Chem. Zeit., 1891, 15, 43.
* Cf. Blattner and Brasseur, Chem. Zeit., 1904, 28, 2II. * Z. anal. Chem., 1889, 28, 699.
IMPURITIES IN HYDROCHLORIC ACID 417
is adopted. It is, however, safer to estimate the total sulphuric acid
present after oxidation by the above reagents, or by hydrogen peroxide,
and to deduct from this the sulphuric acid originally present as found
under 2.
6. Simultaneous examination for Sulphurous and Arsenious acids. If
on adding iodine solution to the acid the iodine is decolorised, at least
one of these impurities is present.
Should this be the case, Hilger' recommends the addition of more
iodine solution till an excess is present; the acid is then transferred to
a test tube and a few pieces of zinc added. The test tube is loosely
stoppered by a cork carrying a piece of silver nitrate paper; arsenic, if
present, produces a darkening of the paper, owing to the arseniuretted
hydrogen evolved. Should no darkening occur, the original acid is
tested for sulphurous acid by first precipitating the sulphuric acid
by means of barium chloride and adding iodine solution to the filtrate
till a coloration is produced ; should sulphurous acid be present in
the original hydrochloric acid, a further precipitation of barium sulphate
will take place.
7. Chlorine. No blue coloration should result when I c.c. of the acid,
diluted with water, is added to 5 c.c. of very dilute freshly prepared
starch solution to which a few drops of potassium iodide solution and
a few drops of dilute sulphuric acid have been added. By carrying out
the test in this manner any blue coloration due to the presence of iodate
in the potassium iodide would appear before the addition of the
hydrochloric acid to be tested. An alternative method of testing is to
shake the acid in a closed flask with a perfectly clean, bright piece of
thin copper foil, first displacing the air in the flask by means of carbon
dioxide; should chlorine be present, part of the copper goes into solution
and may be detected by potassium ferrocyanide, etc. For ordinary
purposes it is sufficient to hold a piece of iodised starch paper in the
vapours given off from the warmed acid ; an immediate blue coloration
shows the presence of free chlorine.
8. Selenium is detected by the Reinsch test and gives a similar stain
on copper to that produced by arsenic; on heating the copper foil
in a dry test tube a sublimate is obtained which dissolves in sulphuric
acid with the formation of a brownish-green coloration.”
According to Reidemeister,” reddish-brown deposits of selenium are
sometimes found in roaster acid but never in pan acid.
9. Estimation of Hydrochloric acid. Ten c.c. of the acid, the specific
gravity of which has been previously determined, are measured from an
accurate pipette, diluted to 200 c.c. with water, and IO c.c. of this solution
taken for the test; or about I g. of the acid is accurately weighed off in
* Jahresber. f. chem. Zech., 1875, p. 445. * Drinkwater, ibid., 1884, p. 348.
* Lunge, Sulphuric Acid and Alkali, vol. ii., p. 98.
2 D
418 MANUFACTURE OF SALTCAKE AND HYDROCHLORIC ACID
the bulb-tap pipette (Fig. 134, p. 390), allowed to flow into water, and the
whole of the solution so obtained taken for titration. The diluted solution
is treated with sodium carbonate, free from chloride, until the reaction is
neutral or very faintly alkaline. This may be done rapidly and without
appreciable loss of acid by spotting, if sodium carbonate solution cor-
responding to the acid content, as determined from the specific gravity
by the aid of the table given above (p. 412), be added from a burette. A
little neutral potassium chromate solution is then added and the solution
titrated with M/IO silver nitrate solution until a distinct faint rose
coloration, which remains permanent on stirring, is produced (cf. p. 123).
From the volume required, the usual deduction of the O.2 c.c. necessary
to produce the coloration must be made. The percentage of HCl in the
acid is obtained by multiplying the number of the remaining c.c. by
72.92 and dividing by the specific gravity of the acid.
The titration may also be carried out by Volhard's method (p. 123).
Should metallic chlorides be present, the above method will obviously
lead to incorrect results; appreciable quantities of such chlorides occur,
however, but seldom. In such cases the total acidity is determined as
described for sulphuric acid (p. 413), the sulphuric acid estimated as
in 2, and deducted from the total acidity. This method may be
employed, of course, even in the absence of metallic chlorides.
LITERATURE
LUNGE, G.-The Manufacture of Sulphuric Acid and Alkali, vol. ii., 2nd edition,
I895.
KRAUCH, G.-The Testing of Chemical Reagents for Purity, translated from the 3rd
German edition by J. A. Williamson and L. W. Dupré, 1903.
THE MANU FACTURE OF SODIUM CARBONATE
By Professor G. LUNGE ; translated by JAMES T. CONRoy, B.Sc., Ph.D.
I. THE LEBLANC SOIDA PROCESS
RAW MATERIALS
I. SALTCAKE
THE analytical examination of saltcake has already been described
(p. 404 et seq.). In the case of Saltcake intended for alkali manufacture,
additional important conclusions are to be drawn from the outward
appearance and behaviour of the sulphate. According to Lunge,' salt-
cake intended for the revolver process must satisfy the following
requirements. It should be very porous and in the state of fine
powder, or at least “spongy”; any lumps present must be quite easily
crushed with a spade, and then fall to fine powder. Hard lumps
nearly always contain a core of raw common salt, which is at once
recognised by its colour and texture; on crushing, its coarsely crystal-
line form and grey colour present a strong contrast to the fine-grained
yellowish or white sulphate. Since saltcake containing salt fluxes
more easily than in the pure state, pieces which are entirely fluxed,
even if inwardly white, are also open to suspicion. Hard, lumpy,
or fluxed saltcake, even if containing 97 per cent. of sodium sulphate,
will not yield a good black ash. Sufficiently pure saltcake may be
deteriorated by calcining so far that the ferric sulphate, which is always
present, has been converted into red ferric oxide; such saltcake
(termed “foxy”) never makes good soda. Good saltcake accordingly
ought to show a little, but not above I-5 or at most 2 per cent. of free
acid (as SOs); in this case it will not, as a rule, contain more than
o. 5 or at most I per cent. of sodium chloride. Quite fresh saltcake, just
as it comes from the roaster, does not yield such good ash as that
which has been lying for some time; even outwardly the difference
is noticeable. Fresh saltcake always shows more or less lumps and
other irregularities in texture, whilst that which has been lying for some
* Sulphuric Acid and Alkali, vol. ii., p. 481.
419
420 THE MANUFACTURE OF SODIUM CARBONATE
time in a large heap, looks quite regular and of even grain. In all
probability, in such large heaps, which retain their heat for a long time,
a further reaction takes place between the undecomposed sodium
chloride and the excess of acid, which improves the quality of the
Saltcake.
2. CALCIUM CARBONATE
This is employed in the form of limestone, chalk, or dried lime-mud,
obtained either from the Chance process (see under “Soda Residues")
or from the causticising pans where the lime process is used.
So far as the revolver process is concerned, the harmful constituents
likely to occur in limestone are, magnesia (dolomitic limestones are
quite unsuitable), clay, sand, and iron. The last three impurities
combine with the soda to form double silicates insoluble in water, and
thus lead to loss of alkali. Many limestones are coloured blue and
even black, owing to the presence of organic matter of a bituminous
character; this impurity is, however, perfectly harmless.
The analysis is generally confined to the determination of moisture,
matter insoluble in hydrochloric acid, calcium and, where necessary,
magnesium. Full details are given under the manufacture of “Bleaching
powder” (p. 481).
3. MIXING SLACK
As a rule, coal is employed for the black ash mixture; occasionally,
however, lignite, wood, charcoal, coke, etc., are used. In taking the
samples for analysis, the rules given on pp. 9 and 24I must be
rigorously observed. The following determinations are necessary.
1. Moisture.—To prevent loss of moisture during grinding, the sample
is only broken down to the size of beans, and as rapidly as possible.
One hundred to 200 g. are then heated for two hours at a temperature
not exceeding I IO’, the quantity of air entering the drying oven being
kept as low as possible. If this be neglected, too great a liberation of
the volatile “bituminous” constituents may occur, and finally an
increase in weight due to oxidation may arise. It is advisable to
employ an air-bath, the upper opening of which is nearly closed, or
else a toluene vapour bath. It is still more satisfactory to heat in a
current of dry carbon dioxide.
2. Coke, that is the non-volatile portion of the coal or fixed carbon.
One g. of the finely ground coal is quickly heated in a platinum
crucible, covered with a tightly fitting lid; the crucible should be
at least 30 mm. deep. An ordinary Bunsen burner giving a flame at
least 18 cm. high is employed, and the heating is continued so long
as any appreciable quantity of combustible gas continues to escape
between the crucible and the lid; this should only take a few
MIXING SLACK 421
minutes. The crucible is then allowed to cool, and weighed. The
platinum crucible should be supported in a thin wire triangle, and
its bottorn should not be more than 3 cm. above the top of the
burner. With too small a flame, or too thick a triangle, the yield
of coke comes out too high. The results are calculated on the
ash-free coal or coke, to facilitate comparisons. A good fuel should
yield 60 to 70 per cent. Of coke.
3. Ash.--The estimation of the ash is very simple in the case of
lignite or turf. Coke requires a very high temperature for incineration.
For the determination of ash in caking coals the sample must be very
finely ground and the heating effected very gradually, so that the
volatile constituents may escape without the powder caking to form
a coke.
Should only an occasional determination of the ash be necessary,
I to 3 g of the finely powdered fuel are heated in a platinum crucible
supported in a hole cut in an obliquely supported piece of asbestos
card (Fig. 136). By this device the air serving
for oxidation is kept apart from the flame
gases and the incineration is correspondingly
accelerated. Working in this way, the opera-
tion is completed in two hours, whereas other-
wise it might still be incomplete after eight or
ten hours. The use of the blowpipe is not to
be recommended, since it is liable to lead to a A. ºº:
loss of ash mechanically. FIG. 136.
Where frequent tests have to be made it is
preferable to carry out the incineration in a muffle, employing a platinum
or porcelain dish ; or combustion may be still more rapidly effected if
the fuel, contained in a platinum boat, be heated in a porcelain tube in
a current of oxygen. In the latter case small pieces of the coal or coke
are taken, since with a fine powder the oxygen does not come sufficiently
into contact with the lower portions.
In case a new supply of coal is to be examined, a determination of
the quantity of the ash only will not be sufficient; the percentages of
silica, alumina, and ferric oxide present should also be estimated accord-
ing to the methods used for the analysis of silicatés (cf. p. 580).
4. Sulphur.—AEschka's method. From O. 5 to I g. of the finely powdered
coal is well mixed in a platinum crucible, by the aid of a glass rod, with
one and a half times its weight of an intimate admixture of two parts of
well-burnt magnesia and one part of anhydrous sodium carbonate, and
the uncovered crucible heated in an oblique position, so that only the
lower half is raised to redness. This is best done as shown in Fig. 136.
The combustion is hastened by frequently stirring the mixture with a
platinum wire. The colour, which is originally grey, changes during

422 THE MANUFACTURE OF SODIUM CARBONATE
ignition to yellow, red, or brown; the time required for the ignition
seldom exceeds an hour. Hot water is then poured on the ignited
mixture and bromine water added until the solution assumes a pale
yellow colour, after which it is boiled, filtered, and the insoluble residue
thoroughly washed with hot water. The aqueous extract is acidified with
hydrochloric acid, boiled until the excess of bromine has been expelled
and the solution has become colourless, and the sulphate precipitated
by addition of barium chloride as described above (p. 274). Allow-
ance must, of course, be made for any sulphate present in the magnesia
and sodium carbonate mixture. Should the coal gas be rich in sulphur,
it is advisable to work with a spirit lamp ; Lunge, however, finds that
the use of the asbestos card (Fig. 136) suffices to keep the products
of combustion away from the contents of the crucible.
Aſundes/agen's method differs from Eschka's in the use of potassium
carbonate instead of sodium carbonate. Other published methods
are either too troublesome or else too uncertain." Norvicki”
employs a Rose's crucible in carrying out the Eschka method, and so
renders it possible to conduct a current of oxygen through the mixture
during the operation. The application of sodium peroxide for the
determination has already been described (p. 246).
5. Nitrogen is estimated by igniting the fuel with soda-lime, and
absorbing the products formed in standard sulphuric acid as in the
ordinary method of organic analysis; Kjeldahl's method (p. 247) is prefer-
able. This estimation is much more important in the case of mixing
slack than in that of furnace fuels.
CONTROL OF WORKING CONDITIONS
I. BLACK ASH
The outward appearance of the “ball” is a more important guide
than the chemical analysis, more especially as the latter is vitiated by
the impossibility of drawing a truly average sample; nevertheless
analysis should of course not be omitted.
A properly made ball” is easily detached from the bogie. In the
places exposed to the air whilst at a red heat, especially at the surface
projecting from the bogie, it is of a liver-brown colour, at the other por-
tions of the surface a blackish brown. On breaking, a good ball presents
a slate-grey colour and a honeycombed, almost pumice-like structure;
it should be as homogeneous as possible, with only a few particles of coal
here and there. There should be no black streaks (from coal) or white
* For a comparison and examination of these methods, cf. Heath, J. Amer. Chem. Soc.,
1898, 20, 630.
* Stahl u, Eisen, 1903, 23, II4I. * Cf. Lunge, Sulphuric Acid and Alkali, vol. ii., p. 552.
BLACK ASH 423
streaks (from chalk); these indicate bad working. The presence of
many dispersed particles of coal or of chalk point to an excess of this
constituent in the mixture. A pink or purple shade of colour is less
satisfactory. The inside of a ball ought to be all of one shade, except
close to the edge, where it is always a little darker.
Dark coloured balls result when the furnacing is insufficient or when
the mixing has been bad. Red (burnt) balls contain much sodium
sulphide, to the presence of which the colour is due.
The black ash is tested daily for its percentage of free and total
lime, and for sulphide, sulphate, and carbonate of sodium; the deter-
mination of the free lime is important, since it aids the lixiviation
by reason of the disruptive action brought about by hydration, and
unless a certain quantity be present in the ball the lixiviation process
proceeds very slowly and may be very incomplete.
Sulphide, Sulphate, and Carbonate of Sodium. In addition to the above
the sodium hydroxide and chloride present are also estimated more or
less frequently. The sodium found as NagcOa, Nags, and NaOH is
calculated to NagSO, and the value obtained added to that of the
Na,SO, present as such, giving the total sodium expressed as Na2SO4.
By comparing this value with the figure found for “total lime,” it is
possible to ascertain from the analysis whether the correct
proportion of saltcake and lime has been used in the mixing.
The estimation of the various constituents is carried out
as follows." Fifty g. of the average sample are rapidly, but
thoroughly, ground in a mortar (in works it is often possible
to employ some mechanical contrivance for this purpose), and
introduced into a 500 c.c. flask. Lukewarm, distilled water,
freed from carbonic acid by previous boiling, is then poured
on the mass, which is thoroughly shaken at once, and subse- ( \
quently at intervals during two hours. It is important that
the mixture should be thoroughly shaken at the start, other-
wise the mass cakes to a solid block on the bottom of the
flask and cannot then be broken up.
I. Free Lime.-At the end of the two hours the flask is
filled up to the 500 c.c. mark and two portions, each of 75 C.C., |
of the thoroughly mixed contents are taken for the estimation \/
of free and total lime. For this purpose it is advisable to use
a pipette, the outlet of which terminates abruptly, as shown
in Fig. I37, instead of an ordinary pipette, the long narrow outlet tube
of which is very apt to get blocked by the solid matter present.
The frothy material on the outside of the pipette is washed off by
means of a wash-bottle, the contents emptied into a beaker and the
FIG. 137.
* Cº. Lunge, Sulphuric Acid and Alkali, vol. ii., p. 557; Z, angew. Chem., 1890, 3, 570,
424 THE MANUFACTURE OF SODIUM CARBONATE
pipette washed out with water. An excess of barium chloride solution
and a drop of phenolphthalein solution are added, and the mixture
titrated with M/5 hydrochloric acid until the red coloration just
vanishes (cf. p. 72). Each I c.c. of the acid = O.OO56I g. CaO. Pro-
vided the shaking is efficient, concordant results are obtained.
2. Total Lime.—This determination is carried out by first converting
the lime into calcium chloride and then into calcium carbonate, in
neutral solution, by addition of an excess of M/5 sodium carbonate
solution, the excess of this reagent being titrated back by M/5 acid.
Five c.c. of the thoroughly shaken solution, as used under I, are treated
in a small Erlenmeyer flask with a few c.c. of strong hydrochloric acid
and then heated to boiling until all gas has been expelled. A drop of
methyl orange is added to the cooled solution, which is then exactly
neutralised with sodium carbonate solution; 30 to 40 c.c. of N/5 sodium
carbonate are next added and the solution again heated to boiling. By
this treatment the whole of the calcium is precipitated as calcium carbon-
ate, together with a certain quantity of oxide of iron, alumina, and
magnesia. The quantity of the three latter compounds is, however, small
and for the present purpose may be neglected. The mixture is trans-
ferred to a 200 c.c. flask, the flask filled to the mark with water, and the
excess of sodium carbonate determined by titrating IOO c.c. of the
filtered solution with N/5 hydrochloric acid, using methyl orange as
indicator.
If the number of c.c. of acid required, = n and 30 c.c. of sodium car-
bonate solution, have been taken, the total calcium, expressed as CaO,
is = (30 – 2n) × O.OO561, or, expressed as CaCOs, = (30–2n) × O.OIOOI.
The results obtained by these two tests are not to be regarded as
absolute, owing to the impossibility of obtaining a really average sample;
they serve, however, as a guide, provided that they are made in the
thoroughly mixed turbid solution. The same reservation holds for
all the results obtained for black ash, and on this account it is
important to take the physical appearance of the ball into consideration
(cf. p. 422).
After the portions for the above tests have been taken, the 500 c.c.
flask is well stoppered and allowed to stand until the solution has
... become perfectly clear, after which portions are withdrawn for the
following determinations.
3. Total available Alkali.—Ten C.C., - I g. of black ash, are titrated
cold with hydrochloric and methyl orange (p. 61). This gives the
total alkalinity, and is a measure of the Na,CO3, NaOH, and Na,S.
The quantity of Na,CO., present is found by deducting the values found
under 4 and 5 and is equal to O-O5305 g. for each I C.C. of normal acid.
Any error due to small amounts of alumina and silica may be neglected.
4. Caustic Soda is estimated by adding an excess of barium chloride
BLACK ASH 425
solution (IO c.c. of a IO per cent. Solution of BaCl,2H,O will more than
suffice) to 20 c.c. of the liquor contained in a IOO c.c. flask, filling to
the mark with boiling water, shaking and corking; the precipitate
settles well, the solution becoming clear after a few minutes' standing.
Fifty c.c. of the clear solution are pipetted off and when cold titrated
with hydrochloric acid and methyl orange. The solution should not
be filtered, because filter paper absorbs appreciable quantities of barium
salts. A simpler and more exact method is to titrate the solution
(IO c.c.) in presence of the precipitate, employing litmus or, still better,
phenolphthalein as indicator, the colour change taking place when all
alkali hydroxide is neutralised (cf. p. 72). Each I c.c. of acid corresponds
to O.O40O3.g. NaOH in I g., the weight of black ash taken. In this method
any sodium sulphide present is estimated as hydroxide; silica is present
in such extremely minute quantity that it does not interfere with the
determination, though it may do so in the case of the finished caustic
Soda (cf. p. 466).
5. Sodium sulphide.—Ten c.c. of the solution are diluted to about
2OO c.c. with boiled air-free water, acidified with acetic acid and titrated
quickly with iodine solution, employing starch solution as indicator. If
M/IO iodine solution be used (12.697 g. I per litre), each I c.c. = O.OO3908
g. Na2S; if the solution be made up to contain 3:249 g. I per litre, then
each I c.c. =O-Ooi g. Nags. When employing the AW/IO solution the
number of c.c. required divided by IO gives the acid equivalent to be
deducted from test 4. The more exact process, due to Lestelle, described
under the analysis of the finished soda ash, is unnecessary in testing
black ash. A detailed investigation has been made by Marchlewski'
on the estimation of sulphide-sulphur. e
No account need be taken of other sulphur compounds, with the
exception of sulphate; their separation is dealt with under Soda Mother
Liquors (p. 429).
6. Sodium chloride.—Ten c.c. of the solution are neutralised as
exactly as possible with nitric acid, most conveniently by adding the
same number of c.c. of normal nitric acid (63.02 g. HNO3 per litre) as
were required of hydrochloric acid in test 3. The solution is then
boiled until all sulphuretted hydrogen has been driven off, filtered from
separated sulphur and, after the addition of a little neutral potassium
chromate solution, titrated with silver nitrate solution. Or nitric acid
of any convenient strength may be added in excess and the solution
rendered slightly alkaline by addition of sodium carbonate or
bicarbonate after all the sulphuretted hydrogen has been expelled.
Volhard's method, using ammonium thiocyanate as indicator (p. 123)
can also be employed, in which case it is not necessary to neutralise the
excess of nitric acid. Each I c.c. of W/IO silver nitrate solution corre-
* Z, anal, Chem., 1893, 32,405.
426 THE MANUFACTURE OF SODIUM CARBONATE
sponds to O.OO585 g. NaCl; a solution containing 2.9054 g. AgNOs
per litre corresponds per I c.c. to O.OOI g. NaCl. -
7. Sodium sulphate.—Twenty c.c. of the solution are acidified with
hydrochloric acid, in not too great excess, heated to boiling, and a hot
solution of barium chloride added. Should the bulk of precipitated
barium sulphate be small, it may be at once transferred to and
washed on the filter with hot water and ignited in a platinum
crucible whilst still wet. Each one part BaSO, corresponds to O-60885
parts Na2SO4.
By taking advantage of the methods of weighing and taring
described on p. 20, the gravimetric method of estimation is quite as
rapid as the volumetric process (p. 277), and is at the same time
appreciably more accurate.
8. Carbonating test.—An average sample of the full charge, taken
by mixing together equal quantities of the solutions from each ball
tested, is carbonated by a current of carbon dioxide, and filtered. The
sum of the Na,CO3, Na,SO, and NaCl is then determined by evaporat-
ing the filtrate and weighing the dry residue.
2. VAT LIQUOR
The liquor should be as far as possible of a bright yellow colour, and
should not be tinged brown or green. The strength should be from
54° to 57° Tw. (sp. gr. 1.26 to I-28), measured warm. Since a considerable
separation of crystals takes place on cooling, the liquor must be
examined whilst still warm, and if necessary kept at 40°. It is best,
and saves much time, to withdraw small portions (2 to 5 c.c.) in accurate
pipettes from the non-diluted liquor for the following tests.
Specific gravity.—The specific gravity is taken by the hydrometer,
and, as explained above, necessarily in the warm solution. Lungel has
shown that if the temperature of the liquor be taken at the same time
a very close approximation to the amount of solid material in the
liquor may be obtained since the percentage corresponds almost exactly
with the percentage of pure sodium carbonate present in a sodium
carbonate solution of the same specific gravity (cf. Tables, pp. 448 and
449. -
The chemical examination of the vat liquor includes the following
determinations:—
I. Sodium carbonate.—Two c.c. are titrated with normal hydro-
chloric acid. If methyl orange be employed as indicator, the solution is
cooled by addition of cold water. The number of c.c. found under 2,
together with ºn of the number of c.c. found under 3, must be deducted
from the number used in this test.
* Chem. Ind., 1881, 4, 376.
VAT LIQUOR 427
2. Caustic Soda.-Two or 5 c.c. are measured off and treated as
described on p. 424.
3. Sodium sulphide.—This is determined in 2 c.c. of the liquor as
described on p. 425. Errors arising from the presence of other sulphur
compounds may be neglected.
4. Sodium sulphate.—Two C.C. are examined, as on p. 426.
5. Total Sulphur.—Five c.c. of the liquor are oxidised by addition
of strong bleaching powder solution and hydrochloric acid in excess,
taking care that the solution smells strongly of chlorine. The solution
is filtered and precipitated by barium chloride.
6. Sodium chloride.—Two or 5 c.c. are neutralised as described
on p. 425, and titrated.
7. Sodium ferrocyanide.—This may be estimated by de Haën's
permanganate method, viz., precipitation as Prussian blue, decomposi-
tion of this on the filter by sodium hydroxide solution, and titration of
the re-formed sodium ferrocyanide by permanganate. The following
modification of Hurter's copper sulphate method is, however, both better
and quicker. In its original form 1 the method had the disadvantage
that the expulsion of the chlorine from the excess of bleach used was a
very tedious operation, during which decomposition of the ferrocyanide
produced was liable to occur. Lunge and Schäppi * have overcome this
difficulty by adding only so much bleach solution as is actually necessary
for the oxidation. The improved method is carried out as follows.
Twenty c.c. of the liquor, or a larger volume in the case of a low
cyanide content, are acidified with hydrochloric acid and a strong
solution of bleaching powder added from a burette to the well-agitated
liquor. A drop of the mixture is withdrawn from time to time and
added to a drop of ferric chloride, free from ferrous chloride, on a white
plate. The oxidation of ferrocyanide to ferricyanide is complete when
the ferric chloride test gives only a brown solution free from Berlin
blue. A drop of bleaching powder in excess does not matter, but if too
much has been taken or too large a volume of the solution withdrawn
for spotting, a fresh portion of the sample is taken and nearly the full
quantity of bleach solution required run in at once, only a few tests on
the plate being then necessary.
M/IO copper solution (containing 3, 180 g. Cu or 12,488 g. crystallised
copper sulphate per litre) is then added from a burette to the oxidised
liquor, causing precipitation of yellow copper ferricyanide Cusſes(CN)12.
From time to time the solution is tested by mixing a drop of the turbid
liquid with a drop of dilute ferrous sulphate solution on a porcelain
plate. The addition of the copper solution is continued so long as the
test gives the blue colour, due to the interaction of the ferrous sul-
1 Chem. News, I879, 39, 25; Lunge, Sulphuric Acid and Alkali, Vol. ii., p. 626.
* Chem. Ind., 1881, 4, 370.
428 THE MANUFACTURE OF SODIUM CARBONATE
phate and sodium ferricyanide, and until the colour produced on the
plate has a distinct reddish tinge. When this stage has been reached
all the ferricyanide has been converted into the copper compound; the
red colour is due to the reduction of the yellow ferricyanide of copper to
the red ferrocyanide by the ferrous sulphate. The reaction must be
taken as finished as soon as the first marked reddening occurs, even
though this disappears after a short time. Each I c.c. of the copper solu-
tion should correspond to O'oroi 3 g. Na,Fe(CN)4; but later work" has
shown that such is not the case, and that the factor to be used is O.OI 23 g.
Na,Fe(CN), per I c.c. copper solution. It is, however, advisable to
check the copper solution by standardising it directly with pure potas-
sium ferrocyanide.
Zulkowsky” estimates the ferrocyanide by adding the liquor to a
boiling solution of zinc sulphate, acidified with sulphuric acid, until a
blue colour is produced at the line of contact when a drop of the solution
and a drop of ferric chloride are placed a little distance apart on filter
paper. The precipitate has the composition:—
KıFe(CN)6.3Zn,Fe(CN)6. 12H2O.
Zaloziecki” adds zinc carbonate to the solution to be tested, passes
carbon dioxide through the hot solution, and titrates a portion of the
filtrate with normal acid and methyl orange. The ferrocyanide present
in the liquor is calculated from the quantity of sodium carbonate formed
according to the equation:—
2Na,Fe(CN), +42ncOs= 22ngFe(CN)6+ 4 Na,COs
Should the liquor, as in the case of vat liquor, be alkaline before the
addition of the zinc carbonate, the acid corresponding to such alkalinity
must be deducted from that used after the above treatment.
Hawliczek 4 estimates the total cyanogen by heating the black ash,
placed in a wrought-iron tube, in a current of hydrogen at a red heat.
By this treatment the cyanogen is said to be converted quantitatively
into ammonia, which is collected in normal acid.
The thiocyanate may be determined approximately by Hurter's
method, which consists in first precipitating the ferrocyanide by addition
of zinc chloride to the acidified liquor, filtering, and then comparing the
red coloration, produced on addition of ferric chloride to the filtrate, with
a series of standard solutions containing known quantities of thio-
cyanate and ferric chloride.
8. Silica, Alumina, and Ferric oxide. Parnell's method.”—One
hundred c.c. of the liquor are treated with an excess of hydrochloric acid,
a considerable quantity of ammonium chloride solution added, followed
* Chem. Ind., 1882, 5, 79. * Z. angew. Chem., 1890, 3, 2IO and 3OI.
* Ding!. polyt.J., 1883, 249, 168. * /. Soc. Chem. Ind., 1889, 8, 353.
* Chem. Ind., 1880, 3, 242.
CARBONATED LIQUORS - 429
by excess of ammonia, and the solution boiled until all smell of the
latter has disappeared. The precipitate settles well, and may be filtered
and washed without difficulty. On washing with hot water it acquires
a deep blue colour (owing to formation of Prussian blue?). The mixture
of SiO, Al,Os, and Fe2O3 is ignited and weighed.
9. Carbonating Test.—A considerable volume of the vat liquor is
carbonated by passing a current of carbon dioxide through it, and filtered.
The filtrate is evaporated to dryness and the residue tested for alkalinity,
sodium sulphate, and sodium chloride.
An example of the method of calculating the results and of the
practical conclusions to be drawn from the data obtained, is given in
Lunge's Sulphuric Acid and A/kali, vol. ii., p. 628.
3. CARBONATED LIQUORS
These are examined in the same way as vat liquor. The contained
bicarbonate is estimated in addition. The estimation of carbonic acid is
performed most accurately and quickly by the Lunge and Marchlewski
method as modified by Lunge and Rittener” (p. 153), which is equally
applicable for very large and for very small amounts of carbonic acid.
Failing the apparatus necessary for this method, approximate results,
sufficiently accurate for all practical purposes, may be obtained as follows.
The carbonated liquor is titrated cold, using phenolphthalein as
indicator with M/5 hydrochloric acid until the red colour disappears,
observing the precautions given on p. 73, especially cooling to nearly
o”. The other conditions, high concentration and a large percentage
of sodium chloride, are present at the start and are brought about during
the titration respectively. Methyl orange is then added and the
titration continued until the change to pink occurs. If a c.c. of N/5
hydrochloric acid have been taken for the first titration, and b c.c. for
the second, then the number of c.c. corresponding to the constituents of
the liquor are: 5– a for the bicarbonate, 2a for the alkali present as
Na,COs, and a-i-b for the total alkali. The ratio of bicarbonate to
carbonate can thus be readily calculated.
Other methods for determining bicarbonate are described under
“Bicarbonate” (p. 467).
The mother liquors resulting from the manufacture of soda crystals
are examined in the same way as carbonated liquors.
4. SODA MOTHER LIQUORS
Appreciable quantities of sodium sulphite and thiosulphate may
occur in these liquors, in addition to sodium sulphide, especially when
they are derived from non-carbonated liquors.
* Z. angew. Chem., 1906, 19, 1849.
430 THE MANUFACTURE OF SODIUM CARBONATE
The estimation of sulphide-sulphur is generally carried out by
expelling the sulphuretted hydrogen with acid and absorbing it by a
suitable reagent. The detailed communications of Marchlewski” and
of Jannasch” give full information on this method.
The determination is best made in a flask provided with a tap
funnel reaching nearly to the bottom, and with an exit-tube connected
with one or two ten-bulb tubes (Fig. 127), filled with ammoniacal
hydrogen peroxide, either free from sulphuric acid or in which the
quantity of Sulphuric acid present is known. The material is introduced
into the flask; in the case of a solid it is covered with well-boiled
water, and after the air has been expelled from the apparatus by a
rapid current of hydrogen, hydrochloric acid, diluted with an equal
volume of well-boiled water, is slowly admitted through the funnel.
The solution is finally raised to gentle boiling, and hydrogen passed
through for fifteen minutes. The contents of the receiving tubes
are then heated to boiling to complete the oxidation, acidified with
hydrochloric acid, and the sulphur precipitated as barium sulphate.
Sodium hydroxide solution, free from sulphate, may be employed
instead of the ammoniacal hydrogen peroxide, and the absorbed sulphur
gases converted to sulphate by treating the solution with hydrochloric
acid and bromine water and boiling until the excess of bromine has
been expelled. The hydrogen employed must be washed by an alkaline
solution of lead acetate and then with water. - .
Instead of the above absorbents AV/IO iodine solution may be em-
ployed, the first bulb-tube being filled with this solution, and the second
with an equal volume of N/Io sodium thiosulphate solution, to catch
any iodine carried forward. The two solutions are combined at the end
of the experiment, and the excess of thiosulphate determined by
titration. The sulphuretted hydrogen corresponding to the iodine
taken up is calculated from the equation —
H.S +I, + 2 HI+.S.
The presence of carbonate in the liquors does not interfere in
any of these methods, but sulphites and thiosulphates will lead to error,
owing to the liberation of sulphur dioxide.
The following method” is well adapted to the rapid estimation of
sulphide, sulphate, sulphite, and thiosulphate, when present together in
solution. -
I. The sulphate originally present is determined in one portion of
the liquor. The air in the flask used for precipitation is first displaced
by carbon oxide, to prevent oxidation of the lower sulphur compounds
present, the solution heated to boiling, acidified with hydrochloric acid,
and the sulphate precipitated by addition of barium chloride.
* Z, anal. Chem., 1893, 32, 403. * Z, amorg, Chem., 1896, 12, 124, 134, I58.
* Grossmann, Z. anal. Chem., 1889, 28, 79.
soda MOTHER LIQUORS 431
2. A second portion is acidified with acetic acid, diluted with air-
free water, and titrated with iodine, using starch as indicator. This
gives the Na,S+ NagSOs–H Na2S2O3:
3. A third portion is treated with cadmium carbonate to remove the
sulphide, filtered, the filtrate acidified with acetic acid and titrated with
iodine. This gives the NassOs and Nags,Cs present. The difference
between the readings obtained under 2 and 3 gives the Na,S.
4. A fourth portion is oxidised by bromine water or bleaching-
powder solution (p. 427), and the total sulphate estimated. From the
value so obtained, the sulphate found under I and the sulphate
corresponding to the sulphide present are deducted. If B represents
the remaining sulphate, expressed in g. NassO, and derived from the
sulphite and thiosulphate, and A the total sulphite and thiosulphate
found in test 3, the Nagsg0s present = (O-74.1784 B-O-4I4698 A) g.
and the Na,SOs = (0.661417 A–O,295775) g. -
Kalmann and Spüller" give the following process based on the
behaviour of barium sulphite and barium thiosulphate towards water,
the former being practically insoluble, whilst the latter is soluble in a
large volume of water. - -
(a) The total alkalinity is determined in a measured volume of the
solution by titration with normal acid, using methyl orange as indicator.
The acid required corresponds to the sum of the sodium carbonate,
sodium sulphide, sodium hydroxide, and half of the sodium sulphite
present, since NagsOs reacts alkaline and NaHSOs neutral towards
methyl orange (p. 65).
(b) An equal volume of the liquor is acidified with dilute acetic acid
and titrated with M/Io iodine after the addition of starch solution.
The iodine used corresponds to the sum of the sodium sulphide, sodium
sulphite, and sodium thiosulphate present.
(c) A portion of the solution, double the volume of that employed in
each of the tests I and 2, is treated with a solution of an alkaline
zincate, to precipitate the sulphide, and made up to a definite volume.
One-half of this is then filtered off, acidified with acetic acid, and titrated
with N/Io iodine solution and starch. The iodine used corresponds to
the sodium sulphite and sodium thiosulphate.
(d) Excess of barium chloride is added to a larger portion of the
liquor—three or four times that used in test I—the solution made up to
a definite volume with well-boiled water, the precipitate allowed to
settle, and the solution then filtered.
(a) One-third or one-quarter of this solution is titrated with normal
acid, the volume required corresponding to the sodium hydroxide and
sulphide present.
(8) A second one-third, or one-quarter, is acidified with acetic acid
* Ding!. polyt.J., 1887, 264, 456.
4.32 THE MANUFACTURE OF SODIUM CARBONATE
and titrated with N/Io iodine solution. This gives the sulphide and
thiosulphate present.
The calculation is as follows:—
b — d? = A c.c. M/Io iodine solution, corresponding to the NagSOs
b – c = B }} }} 3) Na,S
d6–(5 – c) = C , }} } % Nags,Cs
da – I/Io B = D » normal acid, ) ) NaOH
I – (da – 1/20 A) = E a } • } } NagGOs
The following method for the estimation of sulphite and thio-
sulphate, when present together, is due to Kalmann." In the reaction
NagSOs--Ig-H H2O = Na,SO,--2HI, acid is produced; in the reaction
2Na,S,Os--Is= Na,S,Os--2NaI, the solution remains neutral. The test
is therefore carried out by allowing the solution to be examined to flow
from a burette into a measured volume of iodine solution until this is
just decolorised, then adding methyl orange and titrating the acid
formed in the iodine titration by N/Io sodium hydroxide solution.
The iodine equivalent of the sodium hydroxide used gives the sulphite,
and the difference between this and the total iodine taken corresponds
to the thiosulphate. This method of testing is, of course, only appli-
cable in the absence of carbonate or after any carbonate present has been
exactly neutralised, which is not always easily accomplished.
Dobriner and Schranz” state that only NagS and NaSH, or Nags and
NaOH, and not all three substances, can exist together in solution, and
that this should be taken into account in calculating and stating the
results of analysis. gº
Autenrieth and Windaus * separate sulphite and thiosulphate by
addition of strontium nitrate, which precipitates the sulphite but leaves
the thiosulphate in solution. According to experiments carried out
in Lunge's laboratory by Bruhns, the method gives useful if not quite
accurate results, and it may therefore be recommended, more especially
as it affords a direct separation of the two substances.
Feld 4 describes the following process for estimating sulphur in all
stages of oxidation, the method being more particularly intended for
liquors containing compounds of the alkali earths. Free sulphur, which
may be dissolved in the thiosulphate or polysulphide present in the
liquors, is extracted by carbon bisulphide and separated from this solution
by distillation. The sulphuretted hydrogen is then driven off by
distilling the liquor with magnesium chloride in a current of carbon
dioxide (which makes the reaction quantitative) and collected in M/Io
iodine solution. For this test three absorption vessels are employed;
two filled with iodine solution, and the third with M/IO thiosulphate
* Ber., 1887, 20, 568. * Z. anal. Chem., 1898, 37, 291.
* Z, angew. Chem., 1896, 9, 455. * Chem. Ind., 1898, 21, 372.
SODA MOTHER LIQUORS 433
solution to retain any iodine carried forward by the gas. The
sulphuretted hydrogen liberated may arise from monosulphide,
polysulphide, or sulphydrate. Should polysulphide be present, sulphur
separates out during the distillation with magnesium chloride, and is
subsequently extracted from the residue by means of carbon
bisulphide; should sulphite also be present, a portion of the poly-
sulphide sulphur may combine with this to form thiosulphate. The
residual liquor from the distillation is oxidised by the addition of iodine
in excess, which liberates sulphur from any sulphide of iron present,
according to the equation :-
2FeS +3H2O+6I = Fe,0s + 2S4-6H.I.
The liberated sulphur is extracted with carbon bisulphide as before.
The thiosulphate, or rather the resulting tetrathionate, is decom-
posed by distilling the residue from the previous operation, or
a fresh portion oxidised by iodine, with aluminium and hydrochloric
acid, and collecting the sulphuretted hydrogen evolved in iodine
solution, as above. The conversion of the tetrathionate to sulphuretted
hydrogen is quantitative, but additional sulphuretted hydrogen will, of
course, be produced from any other polythionates present. Sulphurous
acid is estimated by treatment with excess of mercuric chloride, which
decomposes all the sulphur compounds with the exception of sulphite,
followed by distillation with hydrochloric acid, and collecting the
liberated Sulphur dioxide in iodine solution. º
Browning and Howe” dissolve O. I g. of the material in Io or more
c.c. of water, add potassium or sodium hydroxide or ammonia solution
until the reaction is faintly but distinctly alkaline, then excess of zinc
acetate, and filter. The precipitate is examined for sulphide by addition
of acid. Acetic acid is added in slight excess to the filtrate, followed
by barium chloride, and the solution again filtered through a double
filter paper. The filtrate is treated with iodine solution until a distinct
yellow colour is produced, and this is removed by the addition of a little
stannous chloride and hydrochloric acid. Any precipitation at this
stage indicates the presence of sulphite. The filtrate is then treated
with bromine water in slight excess, and the excess removed with
stannous chloride as above; any precipitate found arises from the
presence of thiosulphate in the original liquor.
The following method has been worked out by Lunge and J. H.
Smith.” The sulphate is determined by displacing the air in the pre-
cipitating flask by carbon dioxide, to prevent oxidation, heating the
solution, acidifying with hydrochloric acid, and precipitating with barium
chloride. A second portion is diluted with air-free water acidified with
acetic acid, and titrated with N/Io iodine solution. A third portion,
* Chem. News, 1898, 78, 213. * Chem. Ind., 1883, 6,301.
2 E
434 THE MANUFACTURE OF SODIUM CARBONATE
four times as large, is treated with zinc acetate or cadmium carbonate
to remove the sulphide, made up to a definite volume, allowed to settle,
and a quarter of the solution taken for each of the following determina-
tions: (1) iodine equivalent= M ; (2) treatment with permanganate
solution of strength W (determined according to the equation:
3Na2S2O,--8KMnO,--H,O = 3Na2SO,--3K2SO,--8MnO,--2KOH) in
considerable excess, and without acidifying. This treatment is carried
out by allowing the liquor to flow into the permanganate solution,
then adding an acid solution of ferrous sulphate, and finally titrating
back the excess of this with permanganate. If the net quantity of
permanganate used be called N, the thiosulphate sulphur S, and the
sulphite sulphur s, then :- -
S = 1/7 (8WN – o.oob.4 M)
s = 2WN – 2S.
The quantity of sulphide present is found by deducting the value M
from the figure for the original iodine titration.
This process has been further tested by Lunge and Segaller" and
found to give exactly the same results as are obtained by the method
proposed by Richardson and Aykroyd,” who had, through incorrect
manipulation, stated that the Lunge-Smith method was inexact.
Richardson and Aykroyd estimate the sulphate by precipitating cold
with barium chloride after addition of tartaric acid; a second portion
is treated with iodine, and the acid produced by interaction with the
sulphite, titrated with methyl orange as indicator, exactly as described
by Kalmann (p. 422). The sulphide-sulphur is estimated in the ordinary
Iſla 1111C1.
Dupré and Korn” propose to decompose sodium sulphite by boiling
with sodium acetate and potassium chlorate in acetic acid solution;
the thiosulphate remains unaltered under this treatment. Lunge was
unable to obtain satisfactory results by this method.
A very complete summary of the numerous earlier investigations on
the action of potassium permanganate on thiosulphate, supplemented by
further experiments, has been made by Dobbin,” from which he has drawn
the following conclusions. When neutral solutions of thiosulphate and
permanganate react in the cold, there results a dark brown, flocculent
precipitate of variable composition, the solution remaining neutral. The
quantity of permanganate necessary to produce a permanent coloration is
less than that indicated by theory, assuming the thiosulphate to be
Oxidised to sulphate, and the permanganate to be reduced to manganese
dioxide. The brown precipitate always contains manganese in lower
stages of oxidation than MnO2, in varying quantity, together with traces
* J. Soc. Chem. Ind., 1900, 19, 22.I. * Z, angew, Chem., 1902, I5, 225.
* Ibid., 1896, 15, 171. *J. Soc, Chem, Ind., 1901, 20, 212.
VAT WASTE 435
of sulphur compounds. The solution separated from this precipitate
always contains tetrathionate in addition to Sulphate, and complete
oxidation to sulphate cannot be effected even after prolonged boiling ;
other sulphur compounds do not appear to be present.
In view of these results, it would appear that the accuracy of all the
earlier methods for the separation of the various sulphur compounds
must be called into question.
Dobbin confirms the statement, frequently lost sight of, that barium
sulphate is not altogether insoluble in the presence of thiosulphate, and
that when the latter is present the sulphate determination is inexact.
The total sulphur and total oxidisable sulphur in Soda mother
liquors are also always estimated (cf. p. 436).
5. VAT WASTE
Vat or tank waste must always be examined as a check on the
lixiviation process, some idea as to the completeness of which can be
gauged from the external appearance of the waste.
Thoroughly extracted alkali waste may be recognised as such by its
appearance, and analysis is usually in accord with the conclusions
formed. It forms a uniform mass, neither slimy nor too coarse, of
blue-grey to black-grey colour, and contains very few pieces as large
as peas, the greater proportion being of small size. Even the larger
pieces should readily yield to pressure. The presence of coarse hard
pieces of the size of, or larger than, hazel nuts in the waste indicates
poor lixiviation, and points to a considerable loss of soda.
Whatever the appearance of the waste, a chemical examination is
essential. If it be not intended to recover the sulphur, an estimation of
the available alkali is usually considered sufficient; a test of the total
alkali is made in addition from time to time. If the sulphur is to be
recovered, the further determinations of oxidisable and total sulphur
must always be made, although different methods of testing may be
adopted according to the nature of the process in use.
A. The non-oxidised A/kal, Waste.
One or two samples are taken each day from the heaps of fresh
waste, and placed in large, wide-necked, glass-stoppered bottles, the
samples being as far as possible representative of the bulk. In the
early days of the industry the samples were exposed to the atmosphere
until dry, and all estimations were made on the air-dried waste. Work-
ing in this way very variable results were obtained, according to the
extent of oxidation undergone by the waste. Now the waste is always
examined in the moist condition and the results calculated on this basis.
The percentage of water may be taken in round figures as 40; it is,
436 THE MANUFACTURE OF SODIUM CARBONATE
however, estimated directly when necessary. The constituents usually
estimated are available and total soda, oxidisable and total sulphur.
I. Available Soda.—The earlier work of Lunge" is now out of date,
and Watson's 4 modification of Lunge's method is generally employed.
Twenty g. of the waste are well stirred with from 150 to 200 c.c. of
warm water and allowed to stand for one hour; at the end of this time the
clear liquor is decanted off and treated for five minutes with a current
of carbon dioxide. On passing the gas the solution becomes at first
cloudy, but it clears later on as bicarbonate is formed and sulphuretted
hydrogen evolved. This affords an indication that the whole of the calcium
has been converted into bicarbonate. The solution is then evaporated
to half its bulk or less, filtered from calcium carbonate, and the filtrate
titrated with normal acid, using methyl orange as indicator. The filtrate
still contains calcium compounds, but only in the form of sulphate or
other neutral salt, which exerts no appreciable influence. Watson found
by this method only O.O25 per cent. of soluble soda as an average of a
Series of analyses extending over a year. -
2. Total Soda (including insoluble sodium compounds; Lunge's
method).-17.71 g of waste are heated in a porcelain or iron dish with
sulphuric acid of sp. gr. 1.5 until the waste is thoroughly disintegrated
and the mixture has been transformed into a stiff paste, when it is
evaporated and all free sulphuric acid driven off by heating. Hot water
is added to the residue, which is broken up with a wooden spatula and
transferred to a 250 c.c. cylinder. Pure milk of lime (prepared from
ordinary slaked lime by pouring off the first alkaline aqueous extract) is
added to neutralise the free acid and precipitate any magnesia present,
the cylinder filled to the 250 c.c. mark and the liquid allowed to settle.
Fifty c.c. of the clear solution are pipetted off, treated with IO c.c. of
Saturated barium hydroxide solution, filtered through a dry filter paper,
and 50 c.c. of the filtrate treated with carbon dioxide till all the baryta
is precipitated and, after filtration, titrated with normal acid. Each
I c.c. of normal acid corresponds to I per cent. Na,C) when the above
quantities are taken, this factor including allowance for the volume of
the waste.
3. Total Sulphur.—Two g. of the waste are treated with excess of
strong bleaching powder solution and hydrochloric acid, to convert all
Sulphur to sulphate, the mixture filtered, and the sulphate determined
by addition of barium chloride to the filtrate. Care must be taken that
the solution smells strongly of chlorine after the oxidation.
4. Oxidisable Sulphur.—This is arrived at by difference, by estimat-
ing the sulphur originally present as sulphate in the waste and deducting
the value so obtained from the total sulphur obtained under test 3. Two
* Chem. Ind., 1881, 4, 372; Z. angew. Chem., 1890, 3, 571.
*J. Soc, Chem. Ind., 1890, 9, 1107.
SULPHUR RECOVERY .. 437
g. of the waste are boiled with hydrochloric acid, filtered, the insoluble
matter washed with dilute hydrochloric acid, the filtrate nearly neutral-
ised by addition of pure sodium carbonate and precipitated by barium
chloride. - -
B. The Chance-Claus Sulphur Recovery Process."
I. Estimation of Sulphur present as Sulphide.—The apparatus
employed consists of a flask fitted with a tap-funnel and gas-exit tube,
the latter being connected with an absorption vessel, such as that
shown in Fig. I 27 (p. 338), charged with a solution of alkali hydroxide
and preferably connected with an aspirator. Two g. of the waste and
a small quantity of water are introduced into the flask, and hydrochloric
acid, diluted with an equal volume of water, gradually run in from the
funnel until decomposition is complete. The solution is then boiled to
drive off the whole of the gas, much water being carried forward and
condensed in the absorption vessel during the operation. When about
one-third of the water has been evaporated and the two-thirds remaining
in the flask are boiling hot, the funnel tap is opened, the apparatus
allowed to cool, and the contents of the absorption vessel transferred to
a 500 c.c. flask and made up to this volume. An aliquot part of the
solution is taken, diluted considerably with well-boiled water, neutralised
with acetic acid, and titrated with M/IO iodine solution, each I c.c. of
which =OOOI6 g. S. (Cf. also p. 430 for estimation of sulphide-sulphur.)
2. Sulphur present as Sulphide in Carbonated mud.—Six g. are
taken for the analysis, which is otherwise carried out exactly as under
test I.
3. Sulphide-sulphur and Carbonic acid in vat waste.—This estima-
tion is but seldom made. The apparatus required consists of a small
flask fitted with a tap-funnel and connected with a U-tube filled with
Sodium sulphate, to absorb any hydrochloric acid that passes over, and
a sufficiently large number of calcium chloride tubes to thoroughly dry
the gas. Beyond the drying tubes two potash bulbs are provided and
these are in turn connected with a weighed calcium chloride tube.
Two g. of the waste together with some water are introduced into the
flask and a current of nitrogen then passed through the apparatus.
(The nitrogen is best prepared from lime-kiln gas by first washing this
by bubbling through sodium hydroxide Solution, then passing it through
a red-hot tube filled with copper turnings, and finally washing with alkali
hydroxide and baryta water.) The waste is decomposed by hydro-
chloric acid, the solution boiled, and all sulphuretted hydrogen and carbon
dioxide carried forward to the absorption tubes by the current of
nitrogen, which is continued for a considerable time. The gain in
* Cf. Lunge, Z, angew. Chem., 1890, 3, 573. For a description of the process, see Lunge's
Sulphuric Acid and Alkali, vol. ii., p. 867.
4.38 THE MANUFACTURE OF SODIUM CARBONATE
weight of the potash bulbs and final calcium chloride tube gives the
weight of CO, and H2S contained in the waste. The H2S is estimated
by treating the potassium hydroxide solution as described above under
test I, and the CO, obtained by taking the difference between the two
determinations.
4. Sulphur present as Sulphide in solutions of calcium sulphide
or sodium sulphide.—Ten c.c. are diluted to 250 c.c., an aliquot
part of this diluted with a large volume of air-free water, acidified
with acetic acid, and titrated as under I. Should thiosulphate be present,
it is estimated as under 5 and deducted from the total. Should poly-
sulphide be present, only the sulphur liberated as sulphuretted hydrogen
and not that precipitated by addition of acid is shown by this method.
5. Soda, Lime, and Thiosulphate in sulphur liquors.-The total
alkalinity (CaO + Na,C) is determined by titrating 5 c.c. of the liquor
with hydrochloric acid, using methyl orange as indicator. A second
portion of 50 c.c. is treated with carbon dioxide to drive off all the
sulphuretted hydrogen (tested by lead paper), the solution boiled to
decompose the calcium bicarbonate, diluted to 50 c.c., allowed to settle,
and 50 c.c. of the clear solution titrated as above, the acid required
representing the Na,C) present. The CaO is obtained as the difference
of the two titrations. *
A further portion of the carbonated liquor is titrated for thiosulphate
by M/IO iodine solution, I c.c. of which =O.OO6412 g. S as Na,S,Os.
The separation of the various sulphur compounds has been described
above (p. 43O).
6. Lime-kiln gases.—The carbon dioxide is estimated in any form
of gas burette or in an Orsat apparatus (Fig. 76, p. 199); the latter
allows the oxygen to be determined at the same time.
7. Gas from the Gas-holder.—(a) The sulphuretted hydrogen and
carbon dioxide together are estimated as in test 6.
(b) The sulphuretted hydrogen is determined in a wide-necked flask
of known capacity, say about 500 c.c. The flask is fitted with a double
bored rubber stopper through which pass two glass tubes, the one reach-
ing nearly to the bottom of the flask, and the second ending just below
the stopper. Both tubes are provided with taps outside the flask. The
gas to be examined is passed through the flask until all air has been
expelled; 20 or 25 c.c. of normal sodium hydroxide solution are then
admitted through one of the taps, and after thorough shaking the liquor
is poured off, the flask well washed out, and the liquor and washings
made up to a known volume. An aliquot part of the solution is then
diluted with a large volume of air-free water, acidified with acetic acid
and titrated with iodine. It is advisable to use an iodine solution con-
taining I I-46 g. I per litre, since I c.c. of such a solution corresponds to
I c.c. H2S measured at O’ and 760 mm. pressure.
SULPHUR RECOVERY 439
Lunge and Rittener's method (cf. pp. I 53 and 514) can also be
applied to the analysis of mixtures of carbon dioxide and sulphuretted
hydrogen. The two gases are absorbed together by sodium hydroxide,
and an aliquot part of the sample then transferred to a second burette
and absorbed with N/Io iodine solution, the excess of which is titrated
back after the absorption. gº
8. Exit Gases from the Claus kiln.—These gases contain small
amounts of sulphuretted hydrogen and sulphur dioxide, both of which
compounds by interaction with iodine solution give rise to two molecules
of hydriodic acid for each atom of sulphur present; but whilst the
sulphuretted hydrogen does not cause any further increase in the acidity,
the sulphur dioxide gives rise to an equivalent of sulphuric acid. The
sum of the H.S +SO, is consequently measured by estimating the
quantity of iodine converted into hydriodic acid, and the sulphur dioxide
by the determination of the acidity remaining after the hydriodic acid
thus formed has been neutralised. Since, however, the passage of a
large volume of gas through iodine solution volatilises a portion of the
iodine, it is necessary to insert a flask containing sodium hydroxide or,
preferably, sodium thiosulphate solution. One or more litres of the gas
are aspirated through 50 c.c. of N/IO iodine solution contained in a bulb-
tube (Fig. 127, p. 338), followed by a similar tube filled with 50 c.c. of
M/Io sodium thiosulphate solution. When the operation is finished, the
contents of the two tubes are emptied into a beaker and titrated with
M/IO iodine solution and starch; the number of c.c. required (= m)
multiplied by O.OO1603 gives the total sulphur present as SO2 + H2S.
The blue coloration is then removed by addition of a drop of thio-
sulphate solution, methyl orange added, and the solution titrated with
M/IO sodium hydroxide. If the number of c.c. necessary to neutralise
the solution be called m, then (m – n)x OO1603 gives the quantity of
sulphur present as SO3.
II. THE AMMON IA-SODA PROCESS
RAW MATERIALS
I. Rock-salt and Brine (cf. p. 397 et seq.).
2. Gas liquor, Ammonium sulphate, and other Ammonium salts
are treated of in Vol. II., under the section “Ammonia,” and are examined
as there described.
3. Limestone (cf. p. 481).
4. Quicklime is analysed as described under Bleaching powder,
p. 483.
. Coal
# 3.) (cf. p. 24.I et seq., and p. 42O).
440 THE MANUFACTURE OF SODIUM CARBONATE
CONTROL OF WORKING CONDITIONS
I. Ammoniacal Brine.
(a) Sodium chloride. The solution is acidified by nitric acid, and the
sodium chloride estimated by Volhard's method (p. 123), or by titra-
tion in the neutral or faintly alkaline solution as described on pp. 123
and 40I.
(5) Free and combined Ammonia. Ten c.c. of the solution are diluted
to about IOO c.c. with water and boiled in a distilling flask until all free
ammonia and ammonium carbonate have been driven off. The evolved
gases are collected in a measured volume of normal hydrochloric acid,
the excess of which is determined by titration. When this has been
done sodium hydroxide solution is added to the residue in the distilling
flask and the distillation continued until all the combined ammonia has
been liberated and absorbed in normal acid as above.
2. Bicarbonate vessels (Carbonators). — Free and combined
ammonia are determined as under I.
3. Mother liquors.
(a) Free and combined Ammonia, as above.
(b) Unchanged Sodium chloride. Ten c.c. are evaporated in a platinum
dish, the residue ignited to drive off all ammonium chloride, and weighed.
4. Bicarbonate.
(a) Total Alkalinity (cf. p. 61 et seq.).
(b) Carbonic acid (cf. p. 429 et seq.).
(c) Moisture, determined by ignition, the carbonic acid found
under (b), and the corresponding water from bicarbonate being deducted.
5. Ammonia distillation.
(a) Free and combined Ammonia, in the mother liquor, as above, under
I (b).
b Milk of Lime, as under Bleaching powder, p. 484.
(c) Excess of Lime in the stills. One hundred c.c. are boiled until all the
ammonia has been driven off, ammonium sulphate is then added and
the boiling continued. The ammonia so liberated corresponds to the
excess of lime present; it is absorbed in normal acid, and titrated.
6. Lime-kiln gas.-Estimation of carbon dioxide (cf. p. 438).
7. Analysis of the Finished product, as described under “Finished
Soda Products,” p. 446.
III. MANUFACTURE OF CAUSTIC SODA
The methods described are restricted to those relating to the manu-
facture of caustic soda by the lime process; the methods involved in
the Löwig process (causticising by ferric oxide), follow naturally from
MANUFACTURE OF CAUSTIC SODA 44.1
the former. The analysis of the liquors obtained in the electrolytic
processes are referred to on p. 442.
A. CAUSTICISED LIQUOR
This liquor is examined in the same way as the vat liquors of the
Leblanc process (p. 426 et seq.); as a rule, only the specific gravity,
total alkalinity, sodium carbonate, and sulphur compounds are estimated.
A table showing the percentage of sodium hydroxide corresponding to
various specific gravities is given under “Caustic Soda,” p. 464.
B. FISHED SALTS
The fished salts from the strong liquors consist essentially of mono-
hydrated sodium carbonate and anhydrous sulphate; those fished from
weaker liquors contain the same two salts in the hydrated condition.
For analysis, 50 g. of the salts are dissolved in water and made up
to IOOO C.C.
I. Total Alkalinity is determined in 20 c.c. of the solution, exactly
as in the case of caustic soda.
2. Sodium chloride.—Twenty c.c. are treated with nitric acid in
excess, boiled to decompose the sulphur compounds, and filtered if neces-
sary; after neutralising the, excess of nitric acid with sodium carbonate,
the solution is titrated with silver nitrate solution as described on
pp. 123 and 40I.
3. Sodium sulphate.—Twenty c.c. are acidified with hydrochloric
acid, and hot barium chloride solution added to the boiling solution.
4. Sodium sulphate from oxidisable sulphur compounds (sodium
sulphite and thiosulphate).—Twenty c.c. of the solution are treated
with bleaching powder solution in excess, followed by hydrochloric acid,
until the solution acquires an acid reaction and smells distinctly of
chlorine (cf. p. 427). Barium chloride is then added and the precipitated
barium sulphate collected and weighed. From the weight so obtained
the weight of barium sulphate obtained in test 3 must be deducted.
C. CAUSTIC “BOTTOMS’’
Caustic bottoms are examined for insoluble matter, total alkalinity,
and sodium carbonate.
I. Insoluble matter.—Twenty g. of the sample are dissolved in
water, and the solution filtered into a 500 c.c. flask. The insoluble
matter is well washed, burnt whilst still moist in a platinum crucible,
ignited, and weighed.
2. Total Alkalinity.—Fifty c.c. of the above solution are titrated
442 THE MANUFACTURE OF SODIUM CARBONATE
hot with normal hydrochloric acid, using phenolphthalein as indicator.
Methyl orange is not suitable in this case, owing to the presence
of alumina, the quantity of which may amount to from 2 to 3 per cent."
3. Sodium carbonate is estimated as described under Caustic Soda
(p. 466).
D. LIME MUD
The lime mud is examined for sodium hydroxide, and sodium
carbonate, free lime, and calcium carbonate.
I. Total Soda.-Ammonium carbonate is added to the mud and
the mixture evaporated to dryness to decompose all insoluble sodium
compounds. The operation is repeated, the whole of the ammonium
carbonate driven off, the residue washed with hot water, filtered, and the
filtrate titrated for alkalinity. The result is best expressed in terms of
Na,C) (O.O3 IO5 g. per I c.c. normal acid), although the sodium found
may originally have been present, partly as hydroxide and partly
as carbonate.
2. Caustic Lime.—The mud is titrated with normal hydrochloric
acid, employing phenolphthalein as indicator, as described on pp. 72
and 423. From the result obtained, the value found under test I, so far
as this represents sodium hydroxide, must be deducted; no appreciable
error will be introduced if the amount to be deducted is assumed as
equal to half the value found under I.
3. Calcium carbonate.—The total calcium is determined by titrating
with acid and methyl orange or litmus and the caustic lime found
under test 2 deducted from the total.
IV. ELECTROLYTIC ALKALI LIQUORS
It is scarcely necessary to state that the methods of analysis are the
same, so far as the works tests are concerned, for both potash and soda
solutions. In the case of potassium compounds a determination of the
potassium in the finished product should, of course, be made; for this
purpose the methods described in the section “Potassium Salts” (p. 520)
are applicable. For the sake of simplicity the expressions sodium car-
bonate, caustic soda, sodium chloride, etc., are to be taken to include the
corresponding potassium compounds, in cases where potassium com-
pounds form the raw materials.
The essential products found in the liquors produced in the elec-
trolysis of sodium chloride are:—sodium hydroxide, sodium carbonate,
sodium chloride, sodium hypochlorite, and sodium chlorate.
These are the same as those found in “Eau de Javel,” the com-
* Lunge, Z, angew. Chem., 1890, 3, 300.
CRUDE FUSED SODA 443
mercial bleaching solution in which sodium is the basis; the proportions
are, however, quite different in the two cases. The same methods of
analysis are, however, applicable in both instances; these are given in
the section dealing with Bleaching compounds (p. 509).
V. CRUDE FUSED SODA OF THE CELLULOSE
INDUSTRY
Under this name" is understood the product which results when the
waste alkaline liquors obtained in the manufacture of cellulose from
wood or straw by the “sulphite process” are, after addition of sodium
sulphate, evaporated to dryness and calcined. The product consists
essentially of carbonate, hydroxide, silicate, sulphide, sulphite and
sulphate of sodium, but contains, in addition to these, small amounts of
Sodium chloride, iron oxide, lime, magnesia, and alumina. Since these
latter substances are without influence on the process, the amount
present is seldom estimated. A correct knowledge of the composition
of the crude fused soda (recovered soda) is essential to the proper
carrying out of the boiling operation, and indicates which constituents
are to be removed and which to be added for successful working. In
the early days of the process this recovered soda was examined in the
same way as Leblanc black ash. Viewed from a purely qualitative
standpoint the two materials correspond fairly well, though, from a
quantitative aspect, the ratio of the various constituents differs con-
siderably in the two cases. As is well known, the solution obtained on
lixiviating black ash consists in the main of sodium carbonate together
with a fair proportion of sodium hydroxide, a little sulphate, very little
sulphide and still less sulphite, silicate, and aluminate; the cyanogen
compounds, etc., may be left out of account in this connection. The
crude fused soda, on the other hand, contains very large quantities of
sulphide and also marked quantities of silicate, especially in cases where
straw has been treated.
The methods employed for the estimation of the extremely small
percentages of sulphide and silicate present in Leblanc liquors, although
sufficiently exact for such purposes, are not directly applicable to the
analysis of the wood-pulp product and, as Lunge and Lohöfer” have
shown, may easily lead to wholly unreliable results. A consideration of
special importance is the fact that the separation of carbonate and
hydroxide by means of barium chloride is quite impossible in the presence
of silicate; thus from a solution of sodium metasilicate only a portion of
the silica (50-60 per cent.) is precipitated, even in the presence of a
* Cf. Z. angew. Chem., 1901, I4, IIo2 ; /. Soc. Chem. Ind., 1901, 20, 123.I.
* Z, angew. Chem., 1901, 14, II25; J. Soc. Chem. Ind., 1902, 21, 70.
444 THE MANUFACTURE OF SODIUM CARBONATE
very large excess of the reagent. As the result of their investigations,
Lunge and Lohöfer recommend the following method for determining
the chief constituents of crude fused sodium carbonate.
Fifty g. of the average sample are dissolved by continued shak-
ing with 500 c.c. of water free from air and carbon dioxide. Solution
is effected in a litre flask at a temperature of 45°, and the solution is
diluted to IOOO c.c. before taking portions for analysis.
I. Insoluble matter is estimated in the usual manner, as described
under black ash.
2. Alkalinity.—Twenty c.c. = I g. substance are titrated with normal
hydrochloric acid and phenolphthalein till the solution becomes
colourless, when methyl orange is added and the titration continued
until the change to red takes place. The number of c.c. required for
each of these stages is noted. The titration must be carried out in the
cold solution, and the best results are obtained if the temperature is
not much higher than O’C. (cf. pp. 73 and 429). The solution, which is
Originally colourless, assumes a very faint yellow tinge when the
colour change with phenolphthalein is reached ; on further addition of
acid the solution becomes strongly yellow and milky sulphur is
deposited. The colour change with methyl orange is nevertheless
quite distinct.
3. Sulphide and Sulphite.—Twenty c.c. = I g. substance are diluted
to about 200 c.c. with air-free water, acidified with acetic acid and
rapidly titrated with M/IO iodine solution, using starch paste as
indicator. The sulphide present is found by deducting from the value
so obtained the equivalent of the sulphite found under 4.
As there is a risk of loss of sulphuretted hydrogen when the titration
is carried out in an acidified solution, Conroy recommends adding the
sulphide liquor to a measured volume of an acidified N/Io iodine
solution, either until the latter is decolorised, or to use an excess of
iodine solution and titrate back with sodium thiosulphate.
4. Sulphite.—The sulphide present is precipitated by adding a
solution of sodium zincate (prepared by adding sodium hydroxide
solution to a solution of zinc acetate until the precipitate first formed
goes into solution) to IOO c.c. of the solution, the whole made up to
250 c.c. and filtered through a dry filter paper. Fifty c.c. of the filtrate
(= 1 g. substance) are acidified with acetic acid and titrated as above
with N/Io iodine solution and starch, the iodine required corresponding
to the sulphite present.
5. Silicate.—Twenty c.c. of the solution are treated with excess of
hydrochloric acid, evaporated to dryness, and the silica estimated gravi-
metrically in the usual way. One part SiO2 = 2.028 parts NassiO3. It
is best to add the hydrochloric acid in an atmosphere of carbon dioxide,
* Private communication.
CRUDE FUSED SODA 445
excluding air as far as possible, so that the filtrate from the silica may
be available for the estimation of sulphate. These precautions are
necessary to prevent oxidation of the sulphuretted hydrogen and
sulphur dioxide.
6. Sulphate is estimated as barium sulphate by the addition of
barium chloride to the acidified filtrate from the silica determination.
This method is only exact provided all oxidation has been excluded
during the addition of the hydrochloric acid in estimation 5. Further,
incorrect results are obtained when relatively large quantities of thio-
sulphate are present, but this point has little bearing in the analysis of
wood-pulp soda, as thiosulphate is not likely to be present in the fresh
material.
The calculation of the quantities of the various constituents present is
very tedious, and occupies considerable time, if the results of the titra-
tions are each calculated directly to parts by weight of the individual
constituents. It is much easier to arrive at the results by comparing the
equivalents of the normal solutions employed, as will be evident from
the following example of an actual analysis:—
I. Insoluble matter.
(a) IO.Oo39 g. substance gave I.O836 g. residue.
(b) IO.OOOO g. 39 I.O805 g. ,
Weight of Ignited Residue.
(a) I.OOOO g., thus giving O.OS36 g. carbonaceous matter.
() o,9904 g., , , O.O90I g. ) )
2. Alkalinity in c.c. A/5 HCl.
With phenolphthalein, 49.39; 49.33; mean, 49.36 c.c.
, methyl orange, 76.74; 76.74 ; mean, 76.74 c.c.
3. Nags + NagsOs.
40.23 c.c. AW/IO I \
40.27 c.c. N/Io I J Mean
4. Na2SOs.
O.40 c.c. M/Io I
o.40 c.c. M/Io I
5. Nasio,
o.O699 g. SiO,
O.O7OI g. SiO2
6. Na;SO4.
O.O536 g. BaSO,
o.O534 g. BaSO,
, 40.25 c.c. AW/IO I.
} Mean, o.40 c.c. M/Io I.
} Mean, 0.0700 g. SiO2.
} Mean, 0.0535 g. BaSO4.
* Qſ. Richardson and Aykroyd, J. Soc. Chem. Ind., 1896, 15, 171 ; Annual Report on Alkali, etc.,
Works, 1899, p. 47; also, Dobbin, J. Soc. Chem. Ind., 1891, Io, 218.
446 THE MANUFACTURE OF SODIUM CARBONATE
Calculation.
I c.c. M/5 HCl corresponds to O.OIo6 g. Na,COs and to o.oo&o g. NaOH.
I c.c. M/10 I ) ) o.Oo39 g. Na2S ,, O.OO63 g. Na2SOs.
I g. SiO2 }} 2.028 g. NagsiCs.
I g. Na2SiOs }} 81.63 c.c. M/5 HC1.
I g. BaSO4 }) O.6089 g. NagsO4.
The 76.74 c.c. M/5 HCl with methyl orange correspond to the :—
Na2CO3 + NaOH + NagsiO3 + Nags + 1/2 NagsOs present.
The 49.36 c.c. M/5 HCl with phenolphthalein correspond to the –
NaOH + Na,SiO3 + 1/2 Na,S+ 1/2 Na,COs present.
[76.74 – O.I.O (for 1/2 NagsO3)]–49.36 = 27.28 c.c. A/5 HCl.
2 x 27.28 = 54.56 c.c. M/5 HCl = Na,S+ NagcCs.
40.25 c.c. A/IO I = Na,S+ NagSOs.
40.25 – O.40 = 39.85 c.c. M/IO I = Na2S.
O.40 c.c. AV/IO I = Na2SOs.
54.56-º- 34.64 c.c. Ms HC-Na,CO,
49.36 – 27.28 = 22.08 c.c. M/5 HCl = NaOH + NagSiO3:
I g. substance, therefore, contains:—
NagcCs • = 34.64 x O.OIO6 = 0.3672 g.
NagsiOs == o.O7OO x 2.028 = O.I.42O g.
= II.59 c.c. A/5 HCl)
NaOH = (22.08 – II.59) × 0.008 = 0.0839 g.
Na,S F: 39.95 × O.OO39 = O. I554 g.
NagSOs =: o.4o x O.OO63 = O.OO25 g.
NagsO, Fº o.O535 × 0.6089 = O.O326 g.
Insoluble matter = Residue + Io = O.I.O8I g.
(of which the carbonaceous matter = o.oo&6 g.)
VI. FINISHED PRODUCTS OF THE SODA INDUSTRY
The various products of the manufacture are: soda ash, Soda crystals,
crystal carbonate, caustic soda, and bicarbonate, together with an inter-
mediate product known as caustic ash; the method of analysis of the
last does not call for special comment.
The individual products vary somewhat in their properties accord-
ing to the method of manufacture followed in their preparation. Thus,
for example, a non-carbonated ash made by the Leblanc process may
contain both caustic soda and sodium sulphide whilst the presence of
these compounds is practically impossible in the product of the ammonia-
soda process. The chief impurity of Leblanc soda, according to per-
centage, is sodium sulphate, whilst in ammonia-soda the chief impurity
is chloride. Commercial bicarbonate prepared by the ammonia-Soda
SODA ASH -- 447
process may contain a small percentage of ammonia, whilst in bicarbon-
ate prepared from soda crystals, ammonia is never present. Generally
speaking, however, the same methods are adopted for the examination
of any individual product, irrespective of its origin. Any special char-
acteristics of importance dependent on the process of manufacture will
be referred to in the description of the methods of analysis employed.
Special attention should be paid to the methods of sampling given
On p. I3.
A. SODA ASH
This product consists essentially of sodium carbonate, but it may
contain as impurity, small quantities of other sodium salts together with
alumina, ferric oxide, water, etc.
Chemically pure sodium carbonate" contains 58:53 per cent. Na2O,
and 41.47 per cent. CO2, and has a specific gravity of 2.5. Various
values have been found for the melting-point by different observers ;
Carnelly gives 814°, le Chatelier 8 Io’, Victor Meyer 849°; the earlier
determination of IO98°, due to Victor Meyer and Riddle, has been shown
to be incorrect. On fusion, a small quantity of sodium oxide is formed,
and at a yellow heat the loss of carbon dioxide may rise to 1; per cent.
This loss on heating is inconsiderable at temperatures below the melting-
point and may be altogether obviated by heating in a current of
carbon dioxide. No loss occurs below a temperature of 300° (cf. p. 85).
The first of the following tables gives the percentage content corre-
sponding to the various specific gravities of solutions of pure sodium
I. Specific Gravity of Solutions of Sodium Carbonate
at 15° C. = 60° F.
Percentage by Weight. Kilos per 1 cubic metre.
Specific
Gravity.
Na2CO3. NagCO3 10H2O. Na2CO3. N a2CO3 10H2O.
1°007 0.67 I '807 6'8 18 °2
1*014 1°33 3'587 13°5 36°4
1*022 2°09 5-637 21 “4 57-6
1*029 2-76 7'444 28°4 76-6
1 *036 3°43 9°251 35°5 95'8
1*045 4°29 II*570 44-8 120°9
1*052 4°94 l3°323 52-0 140°2
1*060 5-71 15°400 60°5 163°2
1*067 6-37 17-180 68-0 I83°3
1*075 7-12 I9 °203 76.5 206°4
1*083 7-88 21°252 85 °3 230°2
1°091 8-62 23°248 94'0 253-6
1*100 9°43 25 °432 103-7 279-8
1*108 10-19 27-482 112-9 304°5
1*116 10°95 29°532 122-2 329-6
I lºš II '81 31°851 132°9 358°3
1 *134 12°61 34°009 143°0 385-7
1°142 13°16 35*493 150°3 405 °3
1 *152 14°24 38°405 164°1 442°4

* For further details, cf. Lunge, Sulphuric Acid and Alkali, vol. ii., p. 43.
448 THE MANUFACTURE OF SODIUM CARBONATE
carbonate, together with the corresponding percentage of decahydrated
sodium carbonate, and the weight of sodium carbonate, anhydrous and
hydrated, contained in I cubic metre of solution at I 5°C. The data
(at 15°) are calculated by Lunge from Gerlach's determinations. The
second table gives corresponding data for degrees Twaddell, the per-
centage content of Na2O and of Na,COs, and the content of these per
cubic foot of solution. The third table (at 30°) is based on Lunge's
own determinations, and deals with the stronger liquors which frequently
occur in practice and which can only exist above the ordinary atmos-
pheric temperature.
2. Specific Gravity of Solutions of Sodium Carbonate in degrees
Twaddell at 15 C. = 60 F.

Percentage by Weight. Lbs. per 1 cubic foot of solution.
TWaddell.
Na2O. Na2CO3. Na2O. Na2CO3.
1 0.28 0-47 0-172 0°294
2 0-56 0-95 0°350 0 °598
3 0°84 1°42 0°525 0-898
4 I •II 1°90 0.707 1-209
5 1°39 2°38 0°889 1°521
6 1-67 2.85 1-070 1°830
7 1-95 3-33 1.257 2°149
8 2-22 3-80 1°441 2°464
9 2°50 4°28 1-631 2°788
10 2-78 4-76 1°852 3-116
II. 3'06 5°23 . 2°012 3°440
I2 3°34 5-71 2°206 3-772
I3 3-61 6-17 2°396 4'097
14 3-88 6*64 2°591 4°430
I5 4°16 7-10 2.783 4-759
16 4°42 7.57 2°981 5'0.98
17 4*70 8'04 3°181 5*439
18 4-97 8°51 3°382 5°783
I9 5°24 8-97 3°582 6-125
20 5°52 9°43 3°783 6°468
21 5-79 9-90 3°989 6'821
22 6°06 10°37 4°197 7.177
23 6°33 10-83 4°403 7.529
24 6-61 11:30 4°615 7-891
25 6°88 11.76 4°825 8'249
26 7-15 12°23 5'040 8-617
27 7-42 12-70 5°256 8.988
28 7-70 13-16 5*465 9°354
29 7.97 13-63 5°691 9-713
30 8°24 14°09 5°908 10'103
According to Lunge, these tables not only give the percentage of
Na2COs in solutions of pure sodium carbonate, but they give with
* Chem. Ind., 1881, 4, 376.
SODA ASH 449
almost equal accuracy the percentage of solid matter, that is, sodium
carbonate plus impurities, in ordinary vat liquor (cf. p. 426).
3. Specific Gravity of Concentrated Solutions of Sodium Carbonate
at 30° C. = 86° F.

Specific Degrees 100-lbs. contain lbs. 1 litre contains grms.
Grayity Twaddell.
at 30°. Na2CO3. Na2CO3, 10 aq. Na2CO3. Na2CO3, 10 aq.
1°310 62 28-13 75-91 368-5 994 °5
1°300 60 27 °30 73-67 35.4 °9 957-4
1°290 58 26°46 71°40 341 °3 92.1 °0
1:280 56 25-62 69°11 327-9 884-7
1:270 54 24-78 66'86 314-7 849.2
1 260 52 23.93 64°59 301 °5 813-2
1°250 50 23:08 62°15 288°5 778-5
1*240 48 22*21 59°94 275-4 743-0
1°230 46 21°33 57-55 262°3 707-8
1*220 44 20-47 55°29 249-7 673-8
1°210 42 19°61 52°91 237-3 640°3
1°200 40 18°76 50-62 225-1 607-4
1°190 38 17-90 48°31 214°0 577.5
1*180 36 17:04 45°97 201*1 542°6
1-170 34 16-18 43°38 189°3 510-9
1°160 32 15°32 41°34 177.7 479°5
1°150 30 14-47 39°04 164°4 449°0
T -140 28 13-62 36-75 155-3 419'0

A table showing the influence of temperature from O’ to IOO’ on the
specific gravity of sodium carbonate solutions has been published by
Lunge." Liebig” has made use of Lunge's figures in calculating the
mean values given in the table below.
4. Influence of Temperature on the Specific Gravity of Sodium
Carbonate Solutions.
(Approximate mean values for + 1° C.)

For Temperatures from For Specific Gravities
0° to 30°. 30° to 40°. 40° to 50°. 50° to 70°. 70° to 100°. From TO
0.0002 0-0004 0-0004 0-0005 0-0005 1°010 1*050
0-0003 0°0004 0 °0004 0-0006 0-0005 I'060 1-070 o
0-0004 0-0004 0-0004 0.0006 0°0006 1*080 1:110 Ug
0-0004 0-0004 0-0005 0.0006 0-0006 1*120 1°170 ſº
0-0004 0°0004 0.0006 0-0007 0-0007 I “ISO 1-200 | *
0-0005 0°0004 0-0005 0-0007 0-0007 I*210 1*240
0-0005 0-0005 0-0007 0-0007 1°241 1.252 &
0-0005 0-0005 0-0006 0'0008 1°263 1°285 t;
* Alkali Makers' Handbook, p. 136. * Post, Chem, tech, Analyse, 2nd edition, vol. i., p. 795.
#
2 F
450 THE MANUFACTURE OF SODIUM CARBONATE
A very thorough investigation of the specific gravities of solutions
of sodium carbonate and of sodium hydroxide has been published by
R. Wegscheider and H. Walter;" the results differ but slightly from
the figures given above up to the third decimal place.
The chemical analysis of commercial soda ash is generally confined
to the determination of the available alkali or alkalimetrical degree;
a complete analysis is, however, made from time to time as a check
upon the process.
Estimation of the available Alkali of Soda ash.—The method
generally used is as follows: 26.50 g. of soda ash are weighed off in a
small beaker and dissolved by boiling with water in a larger vessel.
The solution, together with the small amount of insoluble matter, is
transferred to a 500 c.c. flask, made up to this volume, well shaken and
filtered, if necessary, through a pleated filter paper covered by a watch-
glass. Fifty c.c. of the filtrate are then titrated with normal hydro-
chloric acid and methyl orange. For an ash containing 98 per cent.
Na,CO3, 49 c.c. of normal acid (I c.c. = 2 per cent.) will be required for
the titration.
In the method adopted by the German alkali manufacturers, the
ash is always ignited before determining the percentage of alkali and
the results calculated on the ignited material; this gives the only
satisfactory results. 2.6525 g. are taken for the analysis, dissolved in
water, and the solution titrated directly without filtering. Each I c.c.
of normal acid corresponds to 2 per cent. NagcCs.
- If the readings are made to ſo C.C. in a 50 c.c. burette gradu-
ated in ºr c.c., as may very readily be done in the manner described
on p. 49, the error in the reading will not exceed oos per cent.
Na2COs.
A normal solution of hydrochloric acid is employed, containing
36.46 g. HCl per litre, standardised both by means of chemically
pure sodium carbonate and by silver nitrate (cf. p. 83). Litmus
solution or phenolphthalein may be used as indicator, but in each
case prolonged boiling is necessary; it is much more convenient and
accurate to titrate cold, with methyl orange as indicator (cf. pp. 60
and 65).
It will be noted that in the method generally adopted here, the
alkalinity is determined in the solution only, whilst in the German
method the alkalinity due to calcium carbonate, magnesium
carbonate, ferric oxide, etc., in the insoluble matter is included in the
total. This does not, however, cause any appreciable difference, at all
events in the case of ammonia-soda, in which the total insoluble matter
(inclusive of sand, carbonaceous matter, and other substances not
affecting the alkalinity) does not exceed 4 per cent.
* Monatsh, Igoš, 26, 685; 1906, 27, 13.
VALUATION OF SODA ASH 451
As the whole of the weighed substance is titrated directly in the
German method, any error due to relative inaccuracy in the content
of the pipette and measuring flask is avoided." * -
Various methods are in vogue of stating the results of the titration,
i.e. the strength of the ash. In the scale proposed by Gay-Lussac, the
strength is expressed in terms of available soda, under which all the
substances present which react with normal acid, such as carbonate,
hydroxide, silicate, and aluminate, are included. Chemically pure
sodium carbonate contains 58:53 per cent. Na2O, and would thus corre-
spond to 58:53 degrees Gay-Lussac. These degrees are generally
referred to as “English’ or “Newcastle degrees,” which, however, give
the commercial strength always higher than the real value, as the
former is based on the assumption that the chemical equivalent of
Na,CO, is 54, instead of the true value, 53.05. That is, the com-
mercial percentage of Na2O is higher than the actual percentage in
the ratio of 31.4 to 31. A scale known as the “Liverpool test” was
formerly in use in Lancashire; it was an altogether erroneous method
of statement, and is hardly ever employed now.”
In Germany, the “degrees” indicate percentage of Na,COs, which
is rational as applied to Sodium carbonate itself; but the scale is also
applied to all other sodium compounds, such as caustic soda, which act
upon the test acid, and the strength is accordingly quoted in terms of a
substance which occurs only as an impurity in the caustic alkali. The
scale is, however, in general use commercially.
In France and Belgium, Soda compounds of all kinds (also potash,
baryta, etc.) are quoted on a basis similar to Gay-Lussac degrees, that
is, according to their titrimetric value, leaving out, however, all reference
to the particular alkali, whether sodium hydroxide, sodium carbonate, or
potassium carbonate. The scale employed is that of Descroizilles, and
the degrees indicate the quantity of sulphuric acid monohydrate (H2SO.)
neutralised by IOO parts of the alkali examined. Since IO parts of
chemically pure sodium carbonate are equivalent to 9:245 parts H2SO,
the standard acid is prepared in such a way that 92.45 half cubic centi-
metres (so-called “divisions”) will exactly neutralise 5 g. of pure sodium
carbonate; or, in other words, the standard acid contains exactly IOO g.
H,SO, to the litre. The “Descroizilles sulphuric acid” is prepared by
adding about 3150 c.c. of concentrated sulphuric acid to 50 litres of
water, and standardising, as described on p. 86.
The following table (p. 452) shows the relationship between the
English, German, and French degrees, and is applicable to sodium
hydroxide and all soda products. &
* Cº. Lunge, Sulphuric Acid and Alkali, vol. ii., p. Too; also under “Insoluble Matter " in
this section (p. 455).
* Ibid., p. IoS, and vol. iii., p. 819.
452
THE MANUFACTURE OF SODIUM CARBONATE
English, German, and French Commercial Alkalimetric
Degrees.
O 3. £º O wº 3.S. C § 32.
; : 3 || 5 § 3 §§ | g : # | # 33 || 3 is I gå #| ### | #9 §§
##|###| #| # |###|###| #| #|##| ###| # #
Zºš | ###| | #2 | ##| || 2.É | ###| | #3 | ## £35 | ###| | #2. 3 #
#3"|#"| #" | ##| ||73 |fia" | #" | ##|Héº fiz. 5 £5
º § £& 3- 3 || RE I 3 § pr; ºr
0-5 || 0-51 || 0-85 || 0-79 26-5 26-85 || 45-31 || 41°88 || 52.5 53-19 | 89-76 | 82°98
1 || 1:01 | 1.81 || 1:58 || 27 | 27°35 | 46-17 || 42°67 || 53 || 53.70 || 90-61 | 83°77
1-5 || 1:52 2-56 || 2:37 || 27-5 27-86 || 47-02 || 43°46 || 53°5 54-20 | 91°47 | 84-56
2 || 2:03 || 3:42 || 3-16 || 28 28°36 || 47-88 || 44-25 || 54 || 54-71 || 92°32 || 85 °35
2.5 || 2:54 || 4-27 || 3-95 || 28-5 28-87 || 48-73 || 45-04 || 54-5 55-22 || 93-18 86°14
3 || 3-04 || 5-13 || 4-74 || 29 || 29-38 || 49°59 || 45 ‘83 || 55 55°72 || 94-03 || 86°93
3-5 || 3:55 5'98 || 5-53 || 29.5 29-89 || 50-44 || 46-62 || 55°5 56°23 || 94'89 || 87-72
4 || 4-05 || 6-84 || 6-32 || 30 || 30-39 || 51°29 47°42 || 56 56-74 95°74 || 88°52
4-5 4-56 || 7-69 || 7-11 || 30-5 || 30-90 52-14 || 48-21 || 56-5 57-24 96.60 | 89-31
5 || 5-06 || 8-55 || 7-90 || 31 || 31°41 53-00 || 49-00 || 57 || 57-75 97°45 90°10
5-5 5-57 9-40 || 8-69 || 31.5 || 31°91 || 53-85 49-79 || 57-5 58°26 || 98-31 || 90-89
6 || 6-08 || 10-26 9-48 || 32 || 32°42 || 54-71 50°88 || 58 58-76 || 99°16 || 91°68
6-5 || 6-59 || 11-11 || 10-27 || 32.5 || 32-92 || 55-56 || 51°37 || 58-5 59-27 | 100'02 92°47
7 || 7-09 || 11.97 || 11-06 || 33 33°43 || 56'42 52°16 || 59 59-77 | 100'87 | 93-26
7.5 || 7-60 | 12.82 | 11.85 || 33-5 || 33-94 57-27 | 52-95 || 59.5 | 60-28 || 101°73 || 94-05
8 || 8-10 || 13.68 || 12°64 || 34 || 34°44 || 58'13 || 53°74 || 60 | 60°79 || 102°58 94-84
8-5 | 8-61 || 14-53 || 13°43 || 34.5 || 34-95 || 58'98 || 54°53 || 60°5 || 61°30 || 103°44 || 95-63
9 9-12 || 15-39 || 14-22 || 35 | 35-46 || 59-84 55°32 || 61 | 61-80 || 104-30 | 96:42
9°5 9-63 | 16-24 15:01 || 35.5 35-96 || 60-69 56°11 || 61-5 62-31 || 105-15 97.21
10 || 10-13 || 17-10 || 15°81 || 36 || 36-47 || 61-55 56°90 || 62 62°82 106°01 || 98-00
10-5 | 10-64 || 17-95 16-60 || 36-5 || 36'98 || 62-40 || 57-69 || 62°5 63-32 || 106'86 || 98-79
11 || 11-14 | 18-81 || 17-39 || 37 || 37-48 || 63-26 58°48 || 63 63-83 || 107-72 99-58
11.5 11-65 | 19.66 | 18'18 || 37.5 || 37-98 || 64°11 59-27 | 63-5 64’33 || 108°57 100-37
12 | 12-17 | 20-52 | 18-97 || 38 || 38°50 | 64-97 | 60-06 || 64 || 64-84 || 109°43 || 101°16
12.5 | 12.68 21:37 || 19-76 || 38°5 || 39-00 || 65-82 60-85 || 64-5 || 65°35 | 110°28 || 101-95
13 || 13-17 | 22-23 || 20°55 || 39 || 39°51 | 66-68 || 61-64 || 65 65'85 111-14 || 102-74
13.5 | 13-68 || 23-08 || 21-34 || 39-5 40-02 || 67-53 || 62°43 || 65°5 | 66°36 || 111-99 || 103°53
14 || 14-18 23-94 | 22°13 || 40 | 40-52 | 68°39 || 63-22 | 66 | 66-87 112'85 | 104-32
14.5 || 14-69 24-79 |22'92 || 40-5 || 41.03 || 69°24 || 64-01 || 66°5 || 67-37 113°70 105-11
15 15-19 || 25-65 23°71 || 41 || 41-54 || 70-10 || 64-81 || 67 || 67°88 || 114-56 || 105-90
15-5 15.70 || 26-50 24°50 || 41.5 || 42-04 || 70'95 || 65-60 || 67°5 | 68°39 || 115°41 || 106-69
16 16-21 || 27-36 || 25°29 || 42 42.55 71°81 | 66°39 || 68 || 68-89 || 116-27 | 107-48
16-5 | 16-73 28-21 |26-08 || 42.5 43-06 || 72-66 67-18 || 68°5 | 69-40 || 117-12 | 108-27
17 | 17-22 || 29-07 || 26'87 || 43 || 43-57 | 73°52 || 67-97 || 69 || 69-91 || 117-98 || 109-06
17.5 || 17-73 29-92 || 27°66 || 43.5 44-07 || 74°37 68-76 || 69°5 || 70°41 118-83 || 109-85
18 18-23 || 30-78 || 28°45 || 44 44-58 || 75°23 69°55 || 70 || 70-92 || 119-69 || 110-64
18-5 | 18.74 || 31-63 29-24 || 44°5 || 45-08 || 76-08 || 70°34 || 70°5 || 71°43 | 120°53 || 111-43
19 || 19-25 || 32°49 || 30-02 || 45 || 45-59 || 76-94 || 71°13 || 71 || 71-93 | 121°39 || 112:23
19.5 | 19-76 || 33-34 || 30-82 || 45-5 |46-10 || 77-80 || 71-92 || 71°5 | 72°44 || 122-24 || 113.02
20 |20-26 34-20 || 31-61 || 46 || 46-60 | 78-66 | 72-71 || 72 | 72-95 || 123°10 || 113-81
20-5 20-77 || 35-05 || 32-40 || 46-5 || 47-11 || 79-51 | 73°50 || 72°5 | 73°45 | 123-95 || 114-60
21 | 21-27 | 35-91 || 33-19 || 47 || 47.62 | 80°37 || 74°29 || 73 || 73-96 || 124-81 | 115-39
21-5 21-78 || 36-76 || 33-98 || 47.5 || 48-12 || 81-22 || 75-08 || 73°5 || 74°47 | 125-66 || 116-18
22 22:29 || 37-62 || 34°77 || 48 || 48.63 || 82-07 || 75-87 || 74 || 74-97 | 126-52 | 116-97
22°5 22-80 || 38-47 || 35°56 || 48-5 || 49-14 || 82-93 || 76*66 || 74°5 || 75-48 || 127°37 | 117.76
23 || 23-30 || 39-33 || 36°35 || 49 || 49-64 || 83-78 || 77°45 || 75 || 75-99 || 128-23 118-55
23°5 || 23-81 | 40-18 || 37°14 || 49°5 || 50-15 | 84-64 || 78°24 || 75°5 || 76°49 | 129-08 || 119-34
24 || 24-31 || 41-04 || 37-93 || 50 | 50-66 | 85-48 || 79°03 || 76 || 77-00 | 129-94 | 120-13
24°5 24'82 || 41-89 || 38°72 || 50°5 || 51-16 || 86°34 || 79-82 || 76°5 || 77°51 || 130-79 | 120-92
25 25°32 || 42-75 || 39°51 || 51 || 51-67 || 87-19 || 80-61 || 77 78-01 || 131-65 | 121-71
25.5 25.83 || 43-60 | 40-30 || 51.5 52-18 88-05 || 81-40 || 77.5 78-52 | 132.50 | 122:50
26 || 26°34 || 44°46 || 41°09 || 52 || 52°68 88-90 | 82°19
EXAMINATION OF SODA ASH 453
Böckmann has drawn attention to the importance of always stating
the results of the determination of the available alkali of soda ash on
the dry material, so as to avoid disputes that may arise from any
absorption of moisture that may have taken place during storage.
When exposed to the air, soda ash takes up moisture fairly rapidly, and
may absorb up to 10 per cent, whilst the loss on ignition of a soda ash
properly handled in drawing from the calcining furnace and in packing, is
always less than O. I per cent. For this reason, it is certainly desirable
that the results should be expressed on the dry substance as well as on
the sample as received.
GENERAL EXAMINATION OF SODA ASH
Under this heading the following estimations are included, in
addition to the alkalinity determination: specific gravity, clearness of
the aqueous solution, degree of fineness, matter insoluble in water,
oxide of iron, and proportion of sodium chloride and sodium sulphate.
I. Specific gravity or Density (Böckmann).—The term specific
gravity, or more correctly density, when applied to soda ash, does not
mean the actual specific gravity as determined in the pyknometer, where
the intervening space between the various particles of the ash is filled
with some liquid, such as benzene, which does not exert a solvent
effect. Such a determination is of no practical value, as under these
conditions all samples of soda ash, no matter what the method of
manufacture, would give approximately equal values, since the impurities,
which are only present in small amounts, and do not differ greatly in
specific gravity from the soda ash itself (in round figures, Na,COs, 2.5 ;
Na2SO, 2.6; NaCl, 2.1) have but little influence.
It is, however, necessary for many purposes to be able to express
in figures the “bulkiness” of the ash, a property depending on the
method of manufacture and grinding adopted. The expression refers
only to the cubic weight, that is, the weight of soda ash which can be
closely packed into a vessel of definite capacity.
Böckmann recommends for this determination a strong cylindrical
glass vessel, prepared by cutting off the upper part of a specimen bottle
and grinding the edges of the lower part until the capacity is as nearly
as possible IOO c.c." It is not essential that the capacity should be
IOO c.c., but the employment of a vessel of this size saves a considerable
amount of calculation. Any value lying between 99.5 c.c. and IOO.5 c.c.
may be regarded as representing IOO c.c. for this determination. If, for
example, the weight of ash filling such a vessel be IO2 g., then the “cubic
weight” will be I.O3, I-O2, I-OI respectively, according as the capacity
* Special bottles for “cubic weight” determination of this capacity are made by Desaga,
Heidelberg.
454 THE MANUFACTURE OF SODIUM CARBONATE
is taken at 99.5, IOOO, or 100.5; that is, a margin of £0.5 c.c. is allowed.
The capacity of the vessel to within O. I c.c. and its weight to within
O. I g. are determined once for all, and both values are etched on the
glass. Any similar vessel may of course be used, or even a small strong
beaker (without lip) of suitable size, though this is rather apt to get broken.
The ground and dry ash to be examined is filled into the vessel in
about six separate portions, the glass being tapped on the table
after each addition in order to pack the soda as closely as possible
Finally, the ash standing above the top of the vessel is removed by
sliding a small glass plate over the rim, after which the vessel and soda
are weighed on a balance turning to O. I g. The density obtained is
calculated to the second decimal place. The results agree to within
two units in the second decimal place; it is thus unnecessary to weigh
to nearer than O. I g.
Concordant results can only be obtained by this method by filling
in the ash in small quantities at a time and tapping the glass after each
addition, until the particles are thoroughly packed together so that a
finger-nail does not make an impression in the mass.
If the ash is filled into the vessel without pressure and tapping, the
results are quite different to those obtained by the above method, but
they are concordant amongst themselves and are consequently equally
applicable in practice. This alternative method is more suited as an
empirical test to be used in the packing-room. The vessel may con-
veniently consist of a box made of well-seasoned wood and having the
internal dimensions 40 × 2.5 × 20 cm., corresponding to a capacity of 20
litres, or similar suitable dimensions to give a capacity of I cubic foot.
The soda ash is filled in with a clean wooden spade, care being taken
to disturb the box as little as possible in introducing the material. The
excess of ash is carefully removed by drawing a straight piece of wood
over the top of the box, which is then weighed. The weight of the ash
× 50 gives the weight of ash per cubic metre.
Another plan for determining the density is to use a metal cylinder
about 8.5 cm. high and 3 cm. diameter, filed down so as to hold 62.5 c.c.
of water. The weight in grams of ash that fills a vessel of this capacity
is also equivalent to pounds per cubic foot, as I g. per 62.5 c.c. = 1 lb.
per cubic foot.
Soda ash may be divided into three grades, according to the
density—“light” (density O.8 to I.O.), “medium” (about 1.0 to 1.25),
and “heavy” (about 1.25 to 1.50). .
2. The Clearness of the Aqueous solution (Böckmann).—The
examination of the solution for turbidity is made by dissolving 25 g.
of the ash in 500 c.c. of warm water in a beaker, and comparing the
solution, after cooling, with a standard solution prepared in a similar
Iſlan11621.
EXAMINATION OF SODA ASH 455
The solutions obtained from ammonia-ash are relatively very
clear, on account of the great purity of the product; this point is of
advantage in colour and similar works. It will, of course, be under-
stood that a commercial product like Soda ash always contains traces of
impurities and that a perfectly clear solution is not to be expected. .
3. The degree of Fineness (Böckmann).-The degree of fineness
attained in the grinding is only determined in the case of a very heavy
ash containing a corresponding proportion of coarser particles, and
only then, as a rule, when there is reason to suppose that the grind-
ing has been insufficient. For the determination, IOOO-1 500 g. of
the soda are sieved through a sieve of 35 cm. diameter and 2 mm.
mesh, and the coarser particles remaining on the sieve weighed. The
amount of the residue should not exceed 5 per cent. of the sample
taken.
4. Insoluble matter (Böckmann)—Fifty g. of the sample (or IOO g.
in the case of a very pure sample) are weighed off in a large beaker on
a balance turning to O. I g. and sufficient water added to dissolve the ash.;
the contents of the beaker are agitated gently and continuously to
prevent the formation of lumps, which would retard the solution. The
whole is then allowed to settle for from 4 to 3 hour in a warm place.
As a rule, the insoluble matter settles well, and when this is the case
the greater bulk of the solution can be syphoned off or decanted; the
remainder is then filtered through a tared filter paper and the insoluble
residue well washed with hot water. The two filter papers and in-
soluble matter are dried at IOO’ and weighed.
To estimate the iron present in the insoluble matter, the dried and
weighed filter is moistened with water and the oxide of iron dissolved
in warm hydrochloric acid. The acid filtrate is treated with ammonia,
and the precipitated ferric hydroxide filtered off and dissolved in dilute
sulphuric acid (I : 4). The solution is reduced by zinc and then titrated
with permanganate. The titration may be made in the hydrochloric
acid solution after reducing, diluting largely with water, and adding
manganese sulphate solution, but the end-reaction is less sharp in this
C3 SC. -
5. Sodium chloride.—Two g of ammonia-ash, or 5 g. of Leblanc
ash, are dissolved in water and nitric acid added until the solution is
perfectly neutral or very faintly alkaline to sensitive litmus paper.
Potassium chromate is then added and the solution titrated with silver
nitrate solution, as described on p. 123; or the determination may be
made in acid solution by Volhard's method (p. 123).
6. Sodium sulphate.—Five or ten g. of the sample are dissolved in
hydrochloric acid and the sulphate precipitated by the addition of a hot
solution of barium chloride to the boiling solution.
456 THE MANUFACTURE OF SODIUM CARBONATE
THE COMPLETE ANALYSIS OF SODA ASH
This includes, in addition to the determinations already described,
the estimation of the various constituents of the insoluble matter (sand,
carbonaceous matter, alumina, calcium carbonate, magnesium carbonate,
and ferric oxide), and the estimation of bicarbonate, hydroxide, sulphide,
sulphite, silicate, and aluminate of sodium. The five last-named impuri-
ties are only found in Leblanc soda. Such gross impurities as sulphide of
iron, ferrocyanides, etc., which were formerly met with in Leblanc ash,
need hardly be looked for to-day. In addition to the above, it is
necessary to determine the moisture, should any opportunity for its
absorption have occurred between the times of manufacture and
analysis. This may be done by drying in a desiccator over concen-
trated sulphuric acid, or with greater certainty, by drying for half an
hour at 300°, or by gentle ignition (cf. pp. 83 and 447).
Before carrying out the various determinations, it is advisable to
ascertain whether sodium hydroxide, sodium sulphide, and sodium
sulphite are present or not. For this purpose, a portion of the solution
is shaken with an excess of barium chloride in absence of air, and the
filtrate tested for alkalinity by means of sensitive litmus paper. A
second portion of the alkaline solution is tested for sulphide by
means of sodium nitroprusside or lead acetate paper, and a third
portion examined for sulphite by testing whether the solution after
acidifying with acetic acid and adding starch paste will decolorise a
dilute solution of iodine.
For the analysis, IOO g. of the ash are weighed into a large
beaker, agitated with warm water until solution is complete, the
solution allowed to settle for about half an hour in a warm place,
filtered through a tared filter paper into a litre flask, the insoluble
matter thoroughly washed and the filtrate and washings made up to
IOOO C.C. -
I. Sodium chloride.—Twenty c.c. (= 2 g. ash) in the case of
ammonia-ash, or 50 c.c. (= 5 g. ash) in the case of Leblanc ash, are
neutralised with dilute nitric acid or dilute sulphuric acid, and titrated
with silver nitrate solution (p. 401).
2. Sodium sulphate.—Fifty C.C. (= 5 g. ash) in the case of Leblanc
soda, or IOO c.c. (= IO g. ash) in the case of ammonia-soda, are
rendered slightly acid by addition of hydrochloric acid and the sulphate
precipitated in the hot solution by addition of barium chloride.
3. Sodium bicarbonate.—This can only occur in ammonia-soda,
and is present only in traces. It is estimated as described on p. 468.
At least 5 g. of the sample should be taken and dissolved without
agitation in cold water.
4. Sodium hydroxide.—This can only occur in incompletely carbon-
ANALYSIS OF SODA ASH 457
ated Leblanc soda or in caustic ash. One hundred c.c. of the solution
(= Io g. ash) are taken and analysed as described on p. 424.
5. Sodium sulphide.—This is estimated by Lestelle's method," by
titrating 50 c.c. of the solution (= 5 g. ash) with an ammoniacal
solution of silver nitrate containing I 3-8 I g. Ag per litre and corre-
sponding per I c.c. to O.OO5 g. Na;S. The solution is heated to boiling,
ammonia added, and the silver solution run in drop by drop from a burette,
graduated in ºn c.c., until no further precipitation of the dark-coloured
silver sulphide results. To get the end-point more accurately it is
advisable to filter the solution towards the end of the titration, repeat-
ing the filtration as often as may be necessary. Working with the
above quantity of solution, each I c.c. of silver nitrate solution used
corresponds to O. I per cent. NagS.
The silver solution is prepared by dissolving 13.81 g. pure silver in
pure nitric acid, adding 250 c.c. of ammonia Solution, and diluting to
IOOO C.C.
6. Sodium sulphite.—Fifty c.c. of the solution (= 5 g. ash) are
acidified with acetic acid, and after the addition of starch solution,
titrated with M/Io iodine solution till the blue colour appears. Each I
c.c. N/Io iodine solution equals O.OO63O8 g., or O. 1262 per cent. NassOs.
Or the iodine solution containing 3.249 g. I per litre, described under
black ash (p. 425), may be employed, each C.C. of which corresponds to
o.oO1613 g. or O.O323 per cent. NagsOs. A deduction must, if neces-
sary, be made for the percentage of sulphide found under test 5, in
the ratio of 1.3 c.c. N/IO iodine solution, or 5 c.c. of the weaker solution,
for each I c.c. of the silver solution used.
7. Sodium silicate and Sodium aluminate.—One hundred c.c. of
the solution (= Io g. ash) are acidified in a porcelain dish of about I
litre capacity by the gradual addition of hydrochloric acid. The
solution is then evaporated to dryness on the water-bath and finally
dried at a higher temperature. On extraction with water, the silica
remains behind, and the alumina goes into solution ; each constituent
is then estimated in the usual way.
The determination of silica and alumina is seldom required, but when
it is necessary it is better to weigh off Io g. of the ash and treat it
separately in a porcelain or platinum dish. This prevents any error
caused by the alkaline Solution attacking the glass vessels used as
above. The silicate present in soda ash is too small in amount to
interfere with the separation of the carbonate and hydroxide (cf. p. 425).
8. Insoluble matter (Böckmann).—The filter paper through which
the solution of the IOO g. of ash has been filtered is moistened with
water and the ferric oxide, alumina, and calcium and magnesium
carbonates dissolved by addition of hydrochloric acid. The residue
* Comptes rend., 1862, 55, 739.
458 THE MANUFACTURE OF SODIUM CARBONATE
on the filter is well washed with hot water and the filter paper again
dried and weighed, against the same paper as tare as before. This
gives the sand and carbonaceous matter. The quantity of sand
present is estimated by burning the filter paper and insoluble residue in
a platinum crucible and deducting from the weight so obtained the
ash found on burning the filter paper employed as tare; the difference
of the above weighings gives the percentage of sand. .
The filtrate containing the portion of the residue soluble in hydro-
chloric acid is then submitted to the so-called “complete” commercial
analysis, which consists in estimating the iron oxide volumetrically as
described on p. 38o, and taking the rest of the insoluble matter as
calcium carbonate ; that is, the content in alumina and magnesium
carbonate is not taken separately into account, but is looked upon as
part of the calcium carbonate. This is sufficiently accurate for
technical purposes. -
In the method of estimating the available alkali by direct titration
of the soda solution, without previous filtration (p. 450), the following
substances, if present, are determined in the titration along with the
Sodium carbonate: calcium and magnesium carbonates, the Sesqui-
oxides in the portion insoluble in water, sodium hydroxide, sodium
bicarbonate, sodium sulphide, sodium sulphite, sodium silicate, and
sodium aluminate. If the titration is made in the filtered solution,
any alkalinity due to basic substances in the insoluble matter is excluded.
In any case, however, the influence of these impurities, both soluble
and insoluble, plays but a small part in the titration of an average ash
whether manufactured by the ammonia or Leblanc process, provided
the product is not a poor second grade soda. For example, supposing
an ash gives on titration a total alkalinity, including that of the insoluble
matter, equivalent to 98.4 per cent. Na,COs and the insoluble matter
equals O-33 per cent, a deduction of O-3 is made from the total and the
percentage of sodium carbonate present put down at 98. I. A 98° soda
ash must at least consume acid equivalent to 98 per cent. Na,COs when
the filtered solution is titrated, and the acid used may unhesitatingly be
calculated to NagcCs for commercial purposes, since the soluble im-
purities are only present in traces and for all the important technical
applications of soda ash, with the exception of the manufacture of soda
crystals, the impurities will react in the same manner as their equivalent
of sodium carbonate.
Should it be necessary, in an exceptional case, to ascertain the real
quantity of NagcCs present as accurately as possible, the sodium
carbonate equivalent to the impurities determined as above, is de-
ducted from the total available alkali. The direct estimation of sodium
carbonate, that is, of the carbonic acid content, is best made by the
gas-volumetric method of Lunge and Marchlewski (p. I49).
ANALYSIS OF SODA ASH 459
A good soda ash should not contain more than O-4 per cent. of
matter insoluble in water, nor more than about O. I per cent. insoluble
in hydrochloric acid, whilst the ferric oxide should not exceed O.O2 per
cent. The impurities present in ammonia-ash frequently fall consider-
ably below these limits.
The amount of sulphate present in ammonia-ash will not exceed
O. I per cent, unless it has been specially added, whilst in the best
Leblanc soda ash it may reach from to I per cent, and in ash of low
strength may even exceed 8 per cent.
Sodium chloride is present to the extent of from 3 to 2% per cent. in
ammonia-ash according as it is 98° or 96°/98° ash. Good Leblanc ash
contains only from 4 to per cent. -
9. Physical and Qualitative Tests. – Various physical and
qualitative tests may be advantageously applied to the lower grades
of ash made by the Leblanc process."
A good quality of second grade ash should be white in colour even
when hot, not yellow or reddish, though it cannot be expected to be as
white as “refined alkali.” On the other hand, a yellowish product is
often found to be better carbonated than a pure white product; this is
the case when oxide of iron is present to any considerable extent.
Frequently the ash possesses a bluish tinge, which arises from
ultramarine or sodium manganate, either formed in the black ash or,
occasionally, purposely added. A grey colour indicates poor carbonat-
ing and calcining; such ash will as a rule be found to contain much
caustic alkali, together with incompletely oxidised sulphur compounds.
Good ash should only show a few dark or red specks after grinding.
The proportion of sodium hyroxide should in no case exceed 2 per
cent. except in the case of “caustic ash"; if the ash be intended for
the manufacture of soda crystals, a maximum of I per cent. Na,C) is
frequently stipulated, but it is difficult to maintain this exactly, especi-
ally when, as in the Tyne works, all the mother liquors are worked
up and carbonated, not by gas, but with sawdust. It is, however,
readily done when Mactear's mechanical carbonator is employed.
The ash should contain at most only very slight traces of sulphide.
Even in the case of a moderately calcined and carbonated ash it
should be impossible to detect sulphide by means of lead paper, and
one drop of iodine should be sufficient to produce a blue coloration
when added to a solution containing starch and I g. of the soda ash.
An alkaline solution containing lead (sodium plumbate) acts better
than the ordinary lead paper. Ash manufactured by the Leblanc
process always contains minute quantities of sulphur compounds of the
lower degrees of oxidation, but these can only be detected when large
quantities of the ash, e.g., 50 g. are worked upon. A second quality ash
* Cf. Lunge, Sulphuric Acid and Alkali, vol. ii., p. 684.
460 THE MANUFACTURE OF SODIUM CARBONATE
may be regarded as satisfactory in this respect when the oxidisable
sulphur compounds present do not exceed O. I per cent. of sulphur;
for most purposes twice or three times this amount does no harm.
Sodium thiosulphate cannot occur in calcined soda ash, since it
would be decomposed in the early stages of calcination. Sodium sul-
phite, on the other hand, is nearly always present, though in very
small amount, in the poorer qualities of ash. Its presence may be
detected by iodine solution or other reagents (cf. p. 359).
A good second grade ash should not contain more than from I to Ił
per cent. of insoluble matter; 13 per cent. should be looked upon as a
maximum. The insoluble matter consists chiefly of calcium carbonate
accompanied by alumina and silica; ferric oxide is only present in
traces, except in the case of an ash of reddish-yellow colour and very
bad appearance.
The moisture in freshly made ash ranges from 4 to $ per cent, and
when the ash is well packed it should not rise much over I per cent.
even after some time. When the moisture reaches 2 per cent, the ash
becomes lumpy and discoloured; as already stated, it may absorb up
to IO per cent. of moisture if exposed to a damp atmosphere.
The two neutral salts, sodium chloride and sodium sulphate, which
are always present in soda ash, are practically without effect, prejudicial
or otherwise, in the quantities normally contained in ordinary ash; this
does not, of course, hold when they are purposely added to reduce the
strength.
EXAMINATION OF CHEMICALLY PURE SODIUM CARBONATE
According to Krauch, the natrium carb. puriss, generally employed
for analytical work contains minute traces of iron, sodium chloride, and
sodium sulphate. On this account only the purest and completely
anhydrous product, natrium carb. sicc. chem. pur. should be employed
for standardising solutions and other similar work. Kissling” found
Merck's natrium carbonic, sicc. pulv. chem. pur. (for analysis), prepared
from natr. carbonic chem. pur. cryst., to consist only of sodium oxide,
carbon dioxide, and water; it lost O.63 per cent. CO, and 14.76 per
cent. (corresponding to I molecule) H2O at 150°, and therefore con-
tained a little bicarbonate. According to Krauch, ordinary chemi-
cally pure anhydrous carbonate contains from 2 to 3 per cent. of
Water.
The various impurities are tested for as follows.”
* Silica. From 5 to Io g. are evaporated with an excess of
dilute hydrochloric acid, the residue dried for some time at IOO" and
- * The Testing of Chemical Reagents for Purity, p. 281.
* Chem. Zeit., 1890, 14, 136. * The tests marked * are taken from Krauch, loc. cit.
PURE SODIUM CARBONATE 461
then dissolved in a little hydrochloric acid and about 150 c.c. of water;
the resulting solution should be clear and free from any flocculent
precipitate of silica.
* Sulphuric acid. Five g. are dissolved in 75 c.c. of water, the
solution made slightly acid by addition of hydrochloric acid, heated to
boiling, and barium chloride added; no separation of barium sulphate
should be apparent after twelve hours' standing.
* Sodium chloride. The slightly acid solution of 2 g. in 20 c.c. of
water and dilute nitric acid should remain unaltered on addition of
silver nitrate.
* Arsenic. Five g. of granulated arsenic-free zinc are placed in a
Marsh generating flask of 200 c.c. capacity, dilute sulphuric acid (I : 3)
added and the reagents tested in the usual way for freedom from
arsenic. Fifteen g, of the sodium carbonate are then dissolved in a
small volume of water, the solution rendered acid by addition of pure
dilute sulphuric acid added to the generating flask, and a steady slow
evolution of gas maintained for half an hour; no indication of arsenic
should occur during this time.
Soluble organic Iron compounds. Twenty-five g. of the carbonate
are dissolved in IOO c.c. of lukewarm water, the solution filtered
through a pleated filter paper and the filtrate treated with half its
volume of sulphuretted hydrogen water. The solution should remain
absolutely colourless and show no greenish or dark coloration after
standing for an hour. A blank test should be previously made with
the distilled water employed.
Krauch (loc. cit.) makes a general examination for heavy metals in a
similar manner: 2O g. of the carbonate are dissolved in 60 c.c. of water,
the solution made acid by hydrochloric acid, and sulphuretted hydrogen
water added ; no coloration or turbidity should result. The solution
should remain equally clear and free from green coloration after
addition of ammonia and ammonium sulphide.
* Phosphoric acid. Twenty g. of the carbonate are dissolved in
60 c.c. of water, and nitric acid added in considerable excess, followed
by a solution of ammonium molybdate. No sign of a precipitate should
appear after the solution has been allowed to stand for two hours at a
moderate temperature.
Ammonium salts. According to Krauch (loc. cit.), several grams of the
carbonate are heated in a test tube containing a piece of moistened
turmeric paper in the upper part of the tube, which is loosely closed.
One-fiftieth of a per cent of ammonia may be detected in this manner,
whilst not less than I per cent. of ammonium salts can be recognised by
smell without the paper. Detection by smell may be made much sharper
by dissolving IO g. of the carbonate in 500 c.c. of water, and distilling the
solution in the apparatus described on p. 31 I, under Ulsch's method of
462 THE MANUFACTURE OF SODIUM CARBONATE
nitrate estimation, and smelling the gases evolved at the exit-tube from
time to time.
* Sodium thiosulphate. The aqueous solution is acidified with
acetic acid (I : 50), and tested with silver nitrate. If, at the end of
several minutes, the solution only shows a white opalescence (chlorine),
it may be taken that neither thiosulphate nor arsenic is present in
appreciable quantity. A reddish or yellowish turbidity indicates arsenic,
whilst a brown or black turbidity indicates thiosulphate (cf. p. 468).
* Potassium. The yellow flame coloration produced by the
carbonate should not show any, or only a transitory red tinge when
viewed through cobalt glass or an indigo prism. According to Krauch,
even fractions of a per cent. of potassium salts are sufficient to produce
a persistent red coloration in the flame.
* Sodium hydroxide. Traces of this impurity are best recognised
qualitatively by Dobbin's reagent, ammoniacal potassium mercury
iodide. Kissling 1 gives the following method for its preparation. A
solution of 5 g. of potassium iodide is treated with a solution of mercuric
chloride until the precipitate formed just ceases to redissolve. The
solution is filtered, and I g. of ammonium chloride added to the filtrate,
which is then cautiously treated with sufficient dilute sodium hydroxide
solution until a slight permanent precipitate is formed. The solution
is finally filtered and the filtrate diluted to I litre.
To apply the test, a small quantity of the carbonate solution is
placed on a watch-glass and the reagent added ; the least trace of
hydroxide gives a yellow coloration.
B. SODA CRYSTALS
Ordinary soda crystals consist essentially of decahydrated sodium
carbonate, Na,COs. IoH,O, which, in the chemically pure state, contains
2I-695 per cent. Nago, 15.371 per cent. CO2 (together, therefore, 37.07
per cent. Na,CO3), and 62.93 per cent. H.O. A description of the
properties of this salt and of other hydrated compounds of sodium
carbonate containing less than Io molecules of water of crystallisation
is given in Lunge's Sulphuric Acid and Alkali.” The most important
of these other salts is known as “Crystal Carbonate,” which contains
82.0 per cent. of Na,CO3, 17 per cent. of H2O, and I per cent. of other
salts (practically Na,CO3.H2O), but the quantity made is small when
compared with that of the ordinary decahydrated soda crystals.
Commercial soda crystals seldom contain the theoretical quantity
of water; either a small excess, which should not exceed I per cent,
is present, due to the drying only being carried out at ordinary
temperature, or there is too little water, owing to weathering, from
* Chem. Zeit. Rep., 1898, 14, 136. * Vol. ii., pp. 43 and 7II.
SODA CRYSTALS 463
prolonged exposure to the atmosphere. Weathered soda crystals lose
considerably in appearance and this sometimes gives rise to complaint,
although the change is entirely in the buyer's favour.
The commercial product is never free from sodium chloride and
sodium sulphate. The percentage of sodium chloride should not
appreciably exceed O.5 per cent, whilst the percentage of sodium
sulphate, the presence of which is essential for the production of hard
crystals," is seldom less than I and may rise to 2. A higher percentage
of sulphate is to be regarded as inadmissible, especially as it is fre-
quently added as an adulterant. Commercial soda crystals should test
at least 34 per cent. NagcCs by titration, allowing for all impurities;
generally 35 per cent. will be found, and in weathered samples the
percentage will naturally be still higher. Manufacturers would gladly
have a margin extending down to 32 per cent, but this is not always
granted.
The estimation of the available alkali and of the impurities is
carried out exactly as described under soda ash.
Grosser impurities, such as insoluble matter, iron, etc., were fairly
frequent in the yellow crystals common in the early days of manufacture,
but they are seldom found in the fine, colourless, transparent crystals
which are now supplied. A yellow coloration may be due in part to
organic matter. - -
“Chemically pure crystallised sodium carbonate” (natrium carb.
cryst. chem. pur.) is examined as described on p. 46O.
C. CAUSTIC SODA
Chemically pure sodium hydroxide can only be obtained by treating
metallic sodium with water and is not prepared on the manufacturing
scale. Its properties are described in Lunge's Sulphuric Acid and
Alkali,” where Honigmann's table of the boiling points of solutions of
sodium hydroxide is also given.
The specific gravity of aqueous solutions of sodium hydroxide has
been determined by various observers. Tünnermann's determinations
(1827) were made use of by Schiff (1858) for calculating a formula
connecting density with concentration and this formula was employed
by Gerlach (1867) for the preparation of his table of specific gravities
which was in general use until twenty years ago. This table was
revised by Hager (1883), and an independent series of determinations
were published subsequently by Pickering * (1894). Bousfield and
Lowry “ have recently shown that all these determinations are unreliable,
owing to the difficulty of preparing solutions of known concentration ;
* Cf. Lunge's Sulphuric Acid and Alkali, vol. ii., p. 704. * Vol. ii., p. 81.
* Phil. Mag., 1894, 37, 359; J. Chem. Soc., 1895, 67, 545. * Phil. Trans., 1905, 204, 253.
464 THE MANUFACTURE OF SODIUM CARBONATE
in every case the solutions appear to have been standardised by
titration, a method which they have shown to be inaccurate to the
extent of from O-2 to O.3 per cent. Hager's table is in their opinion less
accurate than that of Gerlach ; they regard Pickering's relative values
for the specific gravity at different concentrations as substantially
accurate, but the absolute values as unreliable, to the extent of O.OOI
at 50 per cent. sodium hydroxide, and roughly to a proportionate
extent at the lower concentrations.
Bousfield and Lowry' have accordingly made a fresh series of
determinations of the specific gravity of sodium hydroxide solutions,
which is given in the following table; the determinations were made
at 18°C., but are calculated by means of a series of temperature
coefficients to 15° also, in the original paper. The solutions were
prepared from weighed quantities of metallic sodium and the densities
determined by means of the pyknometer; the values may be regarded
as accurate to within o.o.1 per cent. of sodium hydroxide, and involve
a possible error of o-OOOI in the sp. gr. of a 50 per cent. Solution.
Specific Gravity of Solutions of Sodium Hydroxide
at 18° C. = 64.4 F.

Fººt. Specific Gravity. £3. Specific Gravity.
0 0-99866 26 1°2860
I 1*01003 27 1°2968
2 1-02127 28 1°3076
3 1*03241 29 1 °3184
4 1*04349 30 1°3290
5 1*05.454 31 1°3396
6 1°06'559 32 I •3502
7 1-07664 33 1°3605
8 1-08769 34 1°3708
9 1.09872 35 1°3811
10 1-10977 36 1 °3913
11 1*12082 37 1°4014
12 1°13188 38 1°4115
13 1°14294 39 1°4215
14 1*15400 40 1°4314
15 1°16505 41 1°4411
16 1°17610 42 1°4508
17 1*1871.4 43 1°4604
18 1-19817 44. 1°4699
19 1°20920 45 1°4794
20 1 22022 46 1°4890
21 1°23121 47 1°4985
22 1-24220 48 1°5080
23 1-25317 49 1°5174
24 1°26412 50 1°5268
25 1.27506
A detailed table, giving the variation of specific gravity with changes
of temperature from O’ to IOO at intervals of IO’, is also given by
* Zoc, cit.
CAUSTIC SODA 465
Bousfield and Lowry." A similar table up to temperatures of 50°, but
based on the earlier determinations of specific gravity, is given in the
Alkali Makers' Handbook (p. 143).” -
It is important to bear in mind that solutions of sodium hydroxide
prepared from commercial products show greater variations in the
degree of purity than is the case with solutions of sodium carbonate or
of acids.
The ordinary rules for sampling, described under “General Methods
of Technical Analysis” (p. 7), are not applicable to caustic soda, which
comes on to the market in solid blocks packed in iron drums. The
block filling the drum is not perfectly uniform throughout ; the portions
most nearly representing the average are those next to the bottom and
sides of the drum, as these portions have solidified fairly rapidly
on packing; the central portions which have solidified more slowly
are less regular in their composition, as the impurities, more especially
chloride and sulphate, have time to accumulate in the fluid core. In
sampling, pieces should therefore be taken from as many different parts
as possible. In the works the sampling is best done while the material
is still liquid ; the most satisfactory procedure is to draw three samples
from each pot as it is emptied, viz., from the top, the middle, and the
bottom respectively, and to pour these, one after the other, on to a plate
and to take the middle portion of the whole for analysis. The three lots
can subsequently be readily separated, as the surfaces of the layers
solidify between the pouring of each sample.
The samples readily take up moisture and carbon dioxide, on the
surface, even when preserved in well-stoppered bottles, which is shown
by the formation of a bright white crust; this crust should be removed
by scraping before the samples are weighed for analysis.
According to Böckmann, the sample should be roughly powdered
as rapidly as possible, to prevent absorption of moisture; if single
pieces be taken instead of the more homogeneous, coarse powder,
differences up to I per cent. may result with different portions of the
same product. Such powdering is not, however, to be recommended
since Some absorption of moisture and of carbon dioxide cannot be
avoided and may give rise to errors appreciably exceeding I per cent.
The chemical analysis of caustic soda is generally confined to the
determination of the total alkalinity and of the caustic soda, or more
correctly the available soda inclusive of sodium silicate and sodium
aluminate. Sodium chloride, sodium sulphate, and water may also be
determined. Other impurities are only present in small amount and
are consequently only estimated in exceptional cases; the methods
described under black ash and soda ash are then employed.
* Zoc. cit., p. 279. -
* CAE also, Wegscheider and Walter, Monalsh., 1905, 26, 685; 1906, 27, 13.
2 G
466 THE MANUFACTURE OF SODIUM CARBONATE
Analysis of Caustic Soda-Fifty g. of the sample, freed from
crust by scraping, as described above, are dissolved in water, the
solution made up to IOOO c.c., and portions of this solution with-
drawn for the following tests.
1. The Total Alkalinity is estimated by titrating 50 c.c. of the solution
= 2.5 g. substance with normal acid, using methyl orange as indicator.
The result is variously calculated according to the scale employed (see
p. 451). In England the results are either expressed in percentage of
“real Na,C)” or in percentage of “commercial Na2O,” the latter figure
being arrived at by multiplying the percentage of real Na2O by 31.4
and dividing by 31. In Germany the alkalinity is expressed as a per-
centage of sodium carbonate, and in France in degrees Descroizilles.
2. The Sodium Hydroxide actually present may be estimated in
various ways. The only other alkaline substance to be taken into
account is sodium carbonate; the quantities of aluminate and silicate
are extremely minute, except in the case of caustic bottoms (p. 441).
The most exact method of estimating the carbonate is to determine
the carbon dioxide evolved on treatment with a strong acid. This may
be done either gravimetrically by the Fresenius-Classen method, or more
rapidly and accurately by the method of Lunge and Marchlewski
(p. I49). -
The method depending upon titration with phenolphthalein and .
hydrochloric acid, after the addition of an excess of barium chloride,
(pp. 73 and 424), is almost equally exact. This method gives the
caustic soda directly. -
For routine testing in the works the method described on p. 73 is
to be recommended. It is not quite so accurate as the above, but it
has the advantage of rapidity and serves as a check on the total alkalinity
(test I). Fifty c.c. of the solution, preferably cooled nearly to o', are
titrated first with hydrochloric acid and phenolphthalein until the red
coloration just disappears, which occurs when all the sodium hydroxide
has been neutralised and the carbonate present has been converted to
bicarbonate; methyl orange is then added to the solution and the
titration continued until the yellow colour changes to pink. If n c.c.
of acid are required for the first part of the titration, and m additional
c.c. for the second, 2 me corresponds to the Na,CO3 present and n – m to
the NaOH.
3. The determination of Chloride, Sulphate, and other impurities is
carried out as described under Soda Ash, p. 456. -
4. Water. Caustic soda may contain up to .30 per cent of water
when it reaches a broker or analyst. This is more especially the case
when samples have been sent in badly sealed boxes, etc.
In such instances the water cannot be accurately determined by
direct heating in a porcelain crucible owing to unavoidable mechanical
SODIUM BICARBONATE 467
loss due to small particles being carried away by the escaping steam;
on the other hand, if caustic soda be heated in a drying oven to 140°
a gain in weight may very easily occur, due to absorption of carbon
dioxide; such a gain almost invariably occurs when the water content
is low.
The moisture determination is therefore best made as described
under Salt (p. 401). About 5 g. of the caustic soda are heated to I 50°
on the sand-bath for three or four hours in an Erlenmeyer flask of the
prescribed dimensions. The funnel must be kept in position the whole
time (in this the method differs from that employed in the case of salt),
to prevent absorption of carbon dioxide. The flask and funnel are
placed on a marble slab, after the heating is completed, allowed to cool
in the air, and finally weighed.
Caustic Salts are analysed in the same way as caustic soda.
D. SODIUM BICARBONATE
Sodium bicarbonate, NaHCOs, contains 36.94 per cent. Nago, 52.34
per cent CO, and 10.72 per cent. H.O. In the pure condition, that is,
when fully saturated with carbon dioxide, it reacts alkaline towards litmus
and, according to earlier accepted views, neutral to phenolphthalein.
Küster" states that in dilute solution it reddens phenolphthalein in
consequence of hydrolysis and Siemens” has found that such reddening
is produced by even the purest bicarbonate. The solubility and other
properties are described in Lunge's Sulphuric Acid and Alkali.”
On exposure to the air the powder loses carbon dioxide, even at the
ordinary temperature, and much more rapidly at somewhat higher
temperatures. The aqueous solution, in a similar manner rapidly, parts
with some of its carbon dioxide and is then found to contain appreciable
quantities of the normal carbonate.
Commercial bicarbonate is very seldom absolutely free from normal
carbonate, but the quantity of the latter should only be very small.
Bicarbonate should dissolve in water to a perfectly clear solution and
chloride and sulphate should only be present in traces. Formerly
bicarbonate was frequently found to contain sodium thiosulphate, due to
its having been manufactured from Leblanc soda liquors by the Deacon-
Hurter process. Bicarbonate manufactured by the ammonia-soda
process may contain a little ammonia as carbonate or chloride, but at
the present time the high percentage (2.6 per cent. ammonium car-
bonate) found by Lehmann” is not likely to occur; a much smaller
percentage than this would make itself evident by the odour.
Bicarbonate of soda is very largely used in medicine and in bakeries,
* Z, anorg. Chem., 1896, 13, I27. * Vol. ii., p. 54.
* Fischer's Jahresöer, 1897, 455. * Chem. Ind., 1887, Io, 88.
468 THE MANUFACTURE OF SODIUM CARBONATE
and it should consequently not contain metallic impurities in sufficient
quantity to be detectable by sulphuretted hydrogen or by ammonium
sulphide. The detection of small quantities of arsenic is carried out
as described in p. 362.
Qualitative analysis of Sodium bicarbonate.—This is directed
in the first place to the examination for chloride and sulphate, which
impurities should only be present in very small quantity; this applies
also to ammonia. Thiosulphate is detected, according to Mylius," by
adding dilute sulphuric acid and zinc, and examining for sulphuretted
hydrogen by means of lead paper. The methods of Salzer” and of Lüttge”
are less exact. Salzer adds a drop of iodine solution to the cold saturated
solution of the bicarbonate; if thiosulphate is present, the iodine is
decolorised. Lüttge treats the bicarbonate with excess of hydrochloric
acid and then adds barium nitrate. According to Musset, very small
amounts of thiosulphate can be detected by the grey coloration due to
mercuric sulphide, which is formed when 5 g. of the bicarbonate are
ground with O. I g. of calomel and two drops of water.
Thiocyanates are detected, according to Utescher,” by agitating
a large quantity of the bicarbonate with water and testing the solution
obtained with ferric chloride in hydrochloric acid solution.
Several methods for detecting the presence of the normal carbonate
in bicarbonate are known. The methods depending upon the red-
dening of phenolphthalein are, according to Küster's observations,
(cf. p. 467) inaccurate. The method usually employed is the reaction
with mercuric chloride. A solution of mercuric chloride in two parts
of water is added to a solution of the bicarbonate in fifteen parts of
water, when a white cloudiness, which gradually becomes brown after
several minutes, is produced. This method is, however, by no means
reliable, and the same may be said of all other qualitative tests for the
normal carbonate. The presence of the normal carbonate can only be
proved with certainty by a quantitative examination.
The detection of normal carbonate by means of magnesium sulphate
is absolutely untrustworthy. It is better, according to Leys,” to
employ a saturated solution of calcium sulphate which, in the presence
of sodium carbonate, gives a precipitate of amorphous calcium carbonate,
which is easily distinguishable from the crystalline calcium sulphate.
Kubli 7 recommends the use of quinine hydrochloride, which is not
precipitated by bicarbonate containing less than 2 per cent. of normal
carbonate.
Quantitative analysis of Sodium bicarbonate consists in de-
1 Fischer's Jahresher., 1886, 282. * Z, angew, Chem., 1890, 3, 3II.
* Chem. Ind., 1887, Io, 27. * Apoth.-Zeit., 1888, 610.
* Chem. Zeit. Æeb., 1889, 13, 305. * Chem. Centr., 1898, I., 752.
7 Abid., 1898, II., 416.
SODIUM BICARBONATE - 469
termining the percentage of available alkali and of carbon dioxide.
As a matter of fact, the latter determination alone is sufficient.
The indirect method of estimating total carbon dioxide by driving
it off and noting the loss of weight in one of the known form of apparatus,
is certainly not sufficiently exact. Trustworthy results can only be
obtained by estimating the carbon dioxide directly, either gravi-
metrically by absorption with soda lime, etc., or gas-volumetrically by
the method of Lunge and Marchlewski as modified by Lunge and
Rittener (p. 153).
The method described on p. 429 may also be employed, but since
this necessitates bringing the bicarbonate into solution, care must be
taken to avoid loss of carbon dioxide during this operation. The
estimation is carried out as follows:—
5.o g, are weighed off in a small beaker and dissolved in about
IOO c.c. of water which has been previously thoroughly boiled and
again cooled to 15° to 20°. Solution is effected in a large beaker
of 900 to IOOO C.C. capacity, agitation of the Solution being avoided
as far as possible; the operation may be assisted by cautiously
breaking up any lumps of bicarbonate that remain on the bottom of
the beaker with a glass rod, care being taken not to stir up the
solution at all violently. Unless these precautions be observed,
loss of carbon dioxide may easily occur and the results obtained
will not be reliable. The water employed should not be below
15° nor above 20° in temperature; in the former case, solution is
difficult, and in the latter, loss of carbon dioxide is likely to result.
Ten g. of pure sodium chloride are then added to the solution,
which is cooled to nearly O’, and titrated with M/I hydrochloric acid,
using phenolphthalein as indicator, the titration being carried to the
point where the red colour just vanishes (= a c.c. acid). The delivery
end of the burette should dip into the solution. Methyl orange is next
added, and the titration continued until the colour changes to yellow
(= & c.c. acid). 2a c.c. correspond to the alkali present as Na,COs, and
b– a to that present as NaHCO3 (cf. p. 429).
Sundström's method," which consists in adding sodium hydroxide to
the bicarbonate solution until the latter has been converted into
carbonate, this point being determined by the brown coloration pro-
duced by spotting with silver nitrate, has been examined by Lunge”
and found to be reliable. It is, however, less convenient than the
method described above, and is in no way more accurate.
Estimation of the total Carbon dioxide. This is a more reliable
check than the above on the composition of bicarbonate; the quantity
should at least reach 50 per cent. (Theory, 52.34 per cent.). The
best plan is to determine the carbon dioxide present as bicarbonate
* Private communication to Prof. Lunge. * Z. angew. Chem., 1897, Io, 169.
470 THE MANUFACTURE OF SODIUM CARBONATE
directly. A simple, rapid and exact method for this determination has
been described by Lunge,” which consists in heating the sample to a
certain definite temperature and estimating the carbon dioxide driven
off; the method is only applicable to solid bicarbonate and not to
solutions. -
The apparatus shown in Fig. I 38 is employed in carrying out the
test. It is connected with a bulb-nitrometer (p. 136), or preferably,
with a gas-volumeter (p. 138), or with any other suitable apparatus
for measuring or weighing the carbon dioxide evolved.
A glass tube, 65 mm. long and 6 mm. internal diameter, is connected
at one end with a wider tube, provided with a tightly fitting ground-in
stopper, a, and at the other end with a capillary tube, d, 60 mm. long.
The stopper a is continued in the form of a glass rod, Ö, 30 mm. long,
which, whilst not ground into the tube, fits it fairly tightly. This
leaves a free space, c, 35 mm. long and 6 mm. diameter, which is
Ž" S
- - - - Aº Aº - - - - - - - - - -º §
^ N
*— + 30– - - + * * * * .357/27727- - 2 º sºme mºst sº sº º s 6/2772772– ams smºs, as sº mº ºm *
Cºğığā #####
*% Žº.º.º. º. ºf ºf 4° N o!
62 s
ls — — — —-30––––- º
|
N
FIG. 138.
separated from the capillary by means of a small plug of asbestos or
glass-wool. This free space will hold about O.850 g. of powdered
bicarbonate, a quantity which, provided the quality be good, will give
rather more than I IO c.c. of carbon dioxide, measured at O’ and
760 mm. This volume can be readily measured in the universal gas-
volumeter, which is graduated from I to 30 c.c. and from IOO to 140 c.c.,
the intervening space being taken up by the bulb. Should the free
space in the decomposition tube be too small or too large for the
measuring apparatus employed, it may be altered by making the glass
rod b shorter or longer, as may be necessary.
The stopper extension & allows the space c to be well heated without
risk of cracking the ground-in portion a, and so permits this part to be
made perfectly tight by using a little vaseline or other grease.
The heating is effected in the air-bath e, which consists of an iron
crucible bored in two corresponding positions on opposite sides and
covered with a piece of asbestos card, f, through which the thermometer
g is placed. A sheet of asbestos card, h, is fixed over the capillary d:
* Z. angew. Chem., 1897, Io, 522.

SODIUM BICARBONATE 471
this card extends below the flame used to heat the crucible e and thus
protects the gas-volumeter from the heat; an interval of at least ten
minutes must be allowed after each operation before the volume of gas
collected is adjusted and read off.
To carry out the determination, the tube is first weighed empty,
then filled with bicarbonate and again weighed. Care must be taken
that none of the bicarbonate remains in the portion fitted with the
stopper ab; any particles adhering to the walls in this portion of the tube
must be removed by a rubber-covered rod or other suitable contrivance
before the stopper, which should be covered with a little vaseline, is
inserted. The tube is then placed in the air-bath, as shown in Fig. 138,
so that the whole of the bicarbonate gets heated, and the free end of
the capillary d connected with the side capillary tube of the gas-
volumeter. The small capillary space between these two tubes is
evacuated two or three times by lowering the levelling tube of the
volumeter until a vacuum exists in the measuring tube, shutting off
from the capillary and expelling the air from the measuring tube by
raising the levelling tube. Three such evacuations may be carried out
in the space of a minute. The levelling tube is then lowered and the air-
bath heated by a flame of moderate size to between 260° and 270°, an
operation requiring on the average about seven minutes. The heating
is continued for a further three minutes, after which the tap connecting
the capillary with the measuring tube is closed, the apparatus allowed
to stand for about ten minutes to attain the atmospheric temperature
and the levelling and reduction tubes adjusted in the usual manner to
allow the volume of gas evolved to be read off directly as dry gas
corrected to o' and 760 mm. Since the gas is always collected moist
in this determination the reduction tube must be set accordingly; the
condensation of water is very small and exerts no influence on the
reading. The percentage of carbon dioxide may be calculated directly
from the volume read off, since the free capillary space is so small
that no appreciable quantity of air remains after evacuating three times
as described above. Lunge, however, recommends transferring the gas,
after measurement, to an Orsat apparatus and proving that it is com-
pletely absorbable by alkali hydroxide, which he finds is always the
case. One c.c. dry CO2 at O’ and 760 mm. pressure weighs I-9766 mg.,
and corresponds to 7.5524 mg. NaHCOs. To arrive at the actual
percentage of bicarbonate in the sample examined, the figure 7.5524 is
multiplied by the number of c.c. of gas liberated and divided by the
weight of substance taken. If it be desired to also determine the
amount of normal carbonate present, a second portion of the sample
may be titrated with normal acid and methyl orange, or the total carbon
dioxide liberated by decomposing with hydrochloric acid may be
determined (p. 149). In the latter case, twice the volume of carbon
472 THE MANUFACTURE OF SODIUM CARBONATE
dioxide found in the bicarbonate estimation is deducted from the
total carbon dioxide found, and the remainder calculated to normal
carbonate. -
The residue in the decomposition tube used in the determination
of the bicarbonate can be washed out and titrated ; this saves a
weighing, but it is scarcely as accurate as the titration of a separate
portion, owing to the risk of losing some of the material. This estima-
tion of normal carbonate is, as a rule, unnecessary.
THE EXAMINATION OF COMMERCIAL LIQUID CARBON DIOXIDE
The following particulars are taken from a very complete investiga-
tion of the subject by Lange." In determining the value of commercial
liquefied carbonic acid it is, as a rule, sufficient to find out what pro-
portion of gases not absorbable by potassium hydroxide solution is
contained in the liquid. It is generally unnecessary to determine the
volume of air present in the gas space of the cylinder. This and the
determination of the carbon dioxide present in the gas space need
only be done in very exact work, when a series of samples otherwise
similar have to be examined. In determining the percentage of air in
samples of the gas drawn from the cylinder, it will generally suffice to
estimate the carbon dioxide in the gas space before and after the
taking of the sample. The percentage of air present in the sample is
then taken as the mean of the two determinations, provided that the
carbon dioxide last examined is not perfectly air-free.
Any suitable form of apparatus for gas analysis may be employed
for the determination of the contained air. Lange makes use of the
modified Winkler gas-burette shown in Fig. 139. -
The tube A has a capacity of IOO c.c. and is narrowed at the top
to a tube of 5 c.c. capacity, graduated in ºr c.c. It is connected at its
lower end with the tube B by means of a piece of rubber tubing; a
run-off tap is not provided. The bent tube e is connected by a length
of rubber tubing with a glass tube, which dips into about 250
c.c. of the solution contained in the bottle D, clamped to the stand. To
charge the apparatus, potassium hydroxide solution of sp. gr. I-297 is
introduced into B in sufficient quantity to more than half fill the tubes
B and A ; the stopper carrying the bent tube e and the attached
rubber tubing is then fixed in position in B, and the alkali solution
driven out of the tube A by blowing through a rubber tube attached
above the tap 5. The tap a is closed as soon as the level of the liquid
in A has sunk below the tap. The tube B is then filled up with potas-
sium hydroxide solution and fixed at such a height on the stand that
on opening the tap a the solution exactly fills the bore of the latter.
* Chem. Ind., 1900, 23, 530.
LIQUID CARBON DIOXIDE - 473
The apparatus is then ready for use. The tap a is turned through 90°
and the gas to be examined introduced ; on closing the tap b and
opening a, the absorbent solution flows from D into B and A. After
the absorption is completed the tap b is opened, when the solution flows
back again into D until the level of the liquid reaches the tap a, where
it remains, leaving the apparatus ready for a further test. More than
4OO analyses can be made with a single charge of solution.
The examination of the
carbon dioxide is carried out
as follows. An adapting piece
is screwed, using some luting
material, into the outlet pipe
of the cylinder, which is placed
in a vertical position, and a
piece of rubber tubing attached
to the adapter. The cylinder
valve is then opened and care-
fully regulated, so as to give a
steady and not too rapid cur-
rent of the gas when the rubber
tube is connected with the tapa,
which is so turned that the gas
enters the tube A, driving out
the contained air through the
tap 6. After a minute the
tube A will be filled with the
gas; the operation may, how-
ever, be continued until needle-
shaped crystals of potassium
bicarbonate are formed in the
upper narrow portion of A.
The tap & is then closed, and
after the gas in A has been
brought to atmospheric pres-
sure by removing the rubber
tubing, the tap a is turned through 90° to bring A and B into com-
munication. The potassium hydroxide solution immediately rises in
A ; the absorption may be hastened by gradually turning A into a
nearly horizontal position. Towards the end of the absorption the tube
is tilted to and fro and then clamped. The bottle D is then raised
until the levels of the liquid in A and D are the same, and the volume
of the gas read. When many determinations have to be made, the
adjustment with D may be dispensed with and the volume read directly
by the aid of a suitable correction table. The difference between two
FIG. I.39.

474 THE MANUFACTURE OF SODIUM CARBONATE
successive tests should not exceed o.os per cent. by volume. By
graduating the upper part of the tube in ºr C.C., it is possible to
approximate to Tºp C.C. in the readings, and so obtain very accurate
results.
To draw a sample from the liquid contents of the cylinder, it should
be laid on its side, on a suitable support, so that the outlet tube of the
valve points upwards. As a rule, it will be found possible to obtain a
satisfactory and not too violent current of gas by slowly and cautiously
opening the valve; small particles of solid carbonic acid will always be
found to issue with the gas. In some cases the setting of the valve
is very troublesome and the current of gas is so violent and inter-
mittent that a very slight turning of the valve may cause the rubber
tubing to burst. When this is the case a reducing valve may be
inserted and the gas allowed to issue for some considerable time to
thoroughly replace the air in the reducing valve before the sample is
taken.
The calculation is carried out as follows:—
If I kilo of carbon dioxide corresponds, at the temperature at which
the analysis is carried out, to A litres, then the calculated G1 kilos
with # 1 volume per cent. CO, occupy a volume of G1.A.O.OIAE, ; in all,
G kilos or GA litres are present. The volume per cent. of CO2 present
in all, is thus given by the equation:—
GA : IOO = (G1.A. o.o.1#1 + G2. A. O.OLAE2): x
G1% + Gaft,
or x = ---G--
or, more conveniently for calculation, the total quantity of air present
in the carbonic acid is given by the equation —
G1(IOO – Ál)+G's (IOO – #3)
y = G
That is, the carbon dioxide present in the cylinder in the liquid and
gaseous states respectively is first calculated to weight; each weight is
then multiplied by the corresponding percentage of air in the liquid
and gaseous portions and the products added together. The true
average air content of the total carbon dioxide, in volume percentage, is
then found by dividing the figure obtained above by the net weight of
the contents of the cylinder.
The methods for the examination of liquid carbon dioxide have
been more recently investigated by Thiele and Deckert, who have
pointed out the importance of determining the gases other than carbon
dioxide, and the difficulty of doing so if the carbon dioxide is de-
* Z, angew. Chem. I907, 20, 737.
LIQUID CARBON DIOXIDE 475
termined by a gas-volumetric process, owing to the small volume of
the residue. They accordingly recommend a method in which the
carbon dioxide is absorbed over potassium hydroxide in a special
measuring tube, fitted into a flask, which is weighed before and after
the absorption. A considerable volume of the unabsorbed gases is
thus obtained, which is transferred to a gas-volumeter for measurement
and subsequently analysed if desired; the increase in weight of the
flask gives the content of carbon dioxide.
LITERATURE
LUNGE, G.-The Manufacture of Sulphuric Acid and Alkali, vol. ii., 2nd edition,
1895.
KRAUCH, A.—The Testing of Chemical Reagents for Purity, translated from the
3rd German edition by J. A. Williamson and L. W. Dupré, 1903.
THE CHLORINE INDUSTRY
By Professor G. LUNGE ; translated by JAMES T. CONRoy, B.Sc., Ph.D.
RAW MATERIALS
THE raw materials employed in the Weldon process and for the
preparation of chlorine on a small scale, are hydrochloric acid, the
examination of which has already been described (p. 409), and native
manganese peroxide; the latter is necessary in the Weldon process to
make good the loss of manganese that occurs in the operations
involved. In the Deacon process hydrochloric acid is the only raw
material; in the various electrolytic processes the raw material is either
potassium chloride or sodium chloride.
In the Weldon process for regenerating the manganese dioxide,
lime, or milk of lime, is required. The lime employed is very fre-
quently prepared by burning limestone in the works where it is used.
Certain special properties are essential in lime intended for this
purpose. Slaked lime is also required in the manufacture of bleaching
powder and milk of lime in the manufacture of potassium chlorate.
The method of examining the potassium chloride employed in the latter
industry is given in the section on “Potassium Salts” (p. 533). This
section is restricted to the methods employed for the examination of
manganese ore, quick lime, slaked lime and milk of lime.
I. MANGANESE ORE
Native manganese ore consists of manganese dioxide admixed with
larger or smaller quantities of impurities. A description of the various
minerals in which it is present, the localities in which they occur, etc.,
will be found in Lunge's Sulphuric Acid and A/ka/.1
The technical examination of manganese ore intended for use in
Weldon stills, is confined to the estimation of moisture, active oxygen,
carbon dioxide, and to the quantity of hydrochloric acid necessary for
decomposition.
1 Vol. iii., p. 268.
476 , P
MANGANESE ORE 477
I. Moisture.—According to R. Fresenius, manganese ore only parts
with the whole of its hygroscopic moisture at 120°; at higher tempera-
tures a portion of the combined water is also driven off. The method
of drying recommended by Fresenius based on this consideration, is,
however, not followed in works, the practice being to dry the ore at Ioo”
in the following manner.
A fair quantity of the very finely ground ore is weighed off, and
spread out in a thin layer on a large watch-glass, and heated directly on
a briskly boiling water-bath, or at IOO in a drying-oven, until the
weight is constant. Drying should be complete after four hours; if it
be continued for six hours, it is safe to assume that constant weight has
been attained.
2. Active Oxygen is always expressed as percentage by weight of
manganese dioxide. This includes all oxygen combined to form oxides
higher than manganous oxide, MnO ; or in other words, all oxygen
capable of reacting with hydrochloric acid to liberate chlorine. One
part active oxygen = 5:438 parts MnO2.
Of the many methods of estimation which have been proposed,
only a few are actually employed. Gay-Lussac's method, which con-
sisted in measuring the volume of oxygen liberated on boiling the ore
with concentrated sulphuric acid, has gone completely out of use. A
method formerly much used is that of Berthier and Thompson, as
modified by Fresenius and Will; it depends on the oxidation of Oxalic
acid according to the equation :-
MnO2+H,SO, +C2H2O, = MnSO, + 2 H2O + 2CO3.
The ore is warmed with strong sulphuric acid and Oxalic acid, and
the carbon dioxide liberated thoroughly dried before leaving the
apparatus; the weight of real MnO, (or its equivalent) in the sample is
then calculated from the loss in weight arising from the liberated carbon
dioxide. Any carbon dioxide due to the presence of carbonate in the
ore must, of course, be allowed for. The method is subject to the error,
common to all determinations in which a small loss of weight is
obtained as the difference between two weighings of a relatively heavy
piece of apparatus, and which is now known to be much greater than
was previously supposed. In addition there are other sources of error,
especially those arising from the difficulty in dissolving many of the
ores. The proposal to weigh the carbon dioxide directly by absorbing
it with soda lime only partially overcomes these objections.
Better results are obtained by applying the reaction volumetrically,
by starting, for example, with a known weight of oxalic acid and
titrating back the excess with permanganate solution; this modification”
has not, however, found any extended application in works practice.
* Cf. Lunge, Sulphuric Acid and Alkali, vol. iii., p. 274.
478 THE CHLORINE INDUSTRY
The same may be said of Bunsen's method, which consists in boiling
the ore with strong hydrochloric acid, absorbing the evolved chlorine in
potassium iodide solution, and titrating the liberated iodine with sodium
thiosulphate. In principle this method ought to be the best, since it
most nearly corresponds to the technical application of the ore. Correct
results can, however, only be obtained by the very careful observation
of various precautions and by carrying out blank tests. Consequently,
the method is but seldom employed, notwithstanding the fact that the
labour involved has been greatly reduced by the introduction of con-
venient apparatus."
The method most generally adopted is the “Ferrous Sulphate”
method of Levol and Poggiale. Carried out in the following manner
described by Lunge, the process is looked upon as a standard method
in the heavy chemical trade, as it is convenient and exact, and gives
concordant results in the hands of different analysts.
1.0875 g. of the very finely ground ore thoroughly dried at IOO’
are treated in a round-bottomed flask fitted with a Bunsen valve, or
still better with a Contat bulb (Fig. 39, p. 106), with 75 cc. of acid ferrous
sulphate solution which is added from a pipette in three portions
of 25 c.c. at a time. The ferrous sulphate solution is prepared by
dissolving IOO g. of pure ferrous sulphate and IOO C.C. of pure Con-
centrated sulphuric acid in I litre of water; it should be standardised
daily when in use with N/2 permanganate solution, using the 25 c.c.
pipette employed in the tests (cf. p. 103). After the addition of the
ferrous sulphate solution, the flask is heated until the whole of the
manganese dioxide has gone into solution and the insoluble residue is
no longer dark coloured. The Bunsen valve must, of course, remain
gas-tight during the subsequent cooling. Occasionally the external
pressure is sufficient to break the flask when a Bunsen valve is em-
ployed; for this reason the Contat bulb is to be preferred, since it
allows a solution of sodium bicarbonate to enter the flask during the
cooling, and the carbon dioxide liberated effectually keeps out atmos-
pheric oxygen whilst maintaining the normal pressure. When the
flask and contents are quite cold the solution is diluted with from IOO
to 200 c.c. of water, and titrated with permanganate until the faint
rose-tint ceases to disappear instantaneously on stirring the solution,
and remains permanent for at least half a minute; subsequent de-
coloration may be disregarded. The quantity of permanganate
solution required is deducted from that corresponding to the 75 c.c.
of ferrous sulphate solution; each I c.c. of the remainder corresponds
to O.O2175 g. MnO, or for the weight of ore taken above, to 2 per
cent. MnO2.
Another method, characterised by extreme rapidity, and which is
* Gſ, Lunge, Sulphuric Acid and Alkali, vol. iii., p. 274.
MANGANESE ORE 479
very useful as a check on the above, is Lunge's' gas-volumetric method,
based on the interaction of peroxide of hydrogen and manganese
dioxide according to the equation —
MnO, + H2O, + H2SO, a MnSO, + 2 H2O + Og:
The value of the peroxide is determined by the volume of oxygen
evolved, which, as will be seen from the above equation, is equal to twice
the quantity of “active oxygen’ present in the manganese dioxide, the
hydrogen peroxide being of course present in excess. Each I C.C. of
oxygen evolved at O’ and 760 mm, corresponds to O.OO3885 g. MnO,
If O. 1943 g. of the manganese ore be taken for analysis, each I c.c. of
oxygen evolved = 2 per cent. MnO, ; and if O-3885 g. ore be taken, each
I c.c. of oxygen = I per cent. MnO2. The former quantity is suitable
for small nitrometers, the latter for the larger apparatus.
The operation is carried out in the nitrometer with the auxiliary
flask as described on p. 137, the volume of gas obtained being reduced
to the volume of dry gas at O’ and 760 mm. It is, however, more con-
venient to employ the “Gas-volumeter” (p. 138), whereby the corrected
volume is obtained without calculation.
The following points should be observed when carrying out this test.
The ore must be ground until extremely fine, so as to allow of complete
decomposition and the quantity weighed off must be placed in the outer
compartment of the decomposition flask, care being taken that none gets
into the inner compartment; a few c.c. of dilute sulphuric acid are first
added to decompose any carbonate present in the sample. Hydrogen
peroxide is then placed in the inner compartment, an excess over the
amount necessary to decompose the ore being added and the bottle
tightly closed by the rubber stopper connected with the nitrometer tap.
The pressure is equalised by opening the tap to the air, so that the
mercury comes to the zero point of the nitrometer. In placing the
stopper in the decomposition flask, and also in the subsequent shaking,
the flask should be held by the neck only, so as to avoid warming by
the hand as far as possible: a more satisfactory plan is to allow the
bottle to stand for at least ten minutes in water, at the room tempera-
ture, both before and after the decomposition. After the mercury has
been brought to the zero mark the flask is tilted, so that the hydrogen
peroxide flows from the inner compartment on to the ore, and the
bottle is then shaken. Shaking should only be continued for two
minutes, since after this period there is a risk of the insoluble matter
acting upon the hydrogen peroxide with spontaneous evolution of
oxygen. Shaking for a longer period is to be avoided, even in cases
where decomposition is incomplete owing to insufficient grinding. In-
* Aer, 1885, 18, 1872; 2. angew. Chem., 1890, 3, 8 and 75; cf. also, Lunge, Sulphuric Acid
and Alkali, vol. iii., p. 278.
480 THE CHLORINE INDUSTRY
complete decomposition is evidenced by the presence of dark material
in the otherwise light coloured residue of silicate; longer agitation does
not aid the decomposition, but only promotes the spontaneous decom-
position of the peroxide. The level of the mercury should be adjusted
and the volume of gas read as soon as the initial temperature has been
attained.
The evolved gas may be regarded as fully saturated with aqueous
vapour, and the reduction should be made by the aid of the “moist
reduction tube” (p. 140); should the apparatus be provided with a
“dry reduction tube,” the correction must be carried out as described
On p. I4I.
A. Baumann" employs an azotometer for the decomposition (p. 125),
and to avoid calculating the gas volume to O’ and 760 mm., weighs
off a quantity of ore corresponding to the temperature and barometric
pressure; this is facilitated by means of a table giving the quantity to
be weighed off under varying conditions. The use of such a table
does not allow of any variation in temperature between the time of
weighing off the ore and the time of making the analysis, and since
a difference of 1° C. means a rise or fall of O.4 per cent. in the result,
this point is of considerable importance.”
A. Baumann “has also devised a volumetric method based on the
same reaction. The manganese ore is allowed to react with an excess
of hydrogen peroxide, previously standardised by permanganate solu-
tion, and the excess of peroxide is titrated back. The hydrogen per-
Oxide solution employed is prepared by diluting the commercial
peroxide with dilute sulphuric acid (I : IO) until the mixture will
decompose approximately its own volume of permanganate solution.
For the analysis O-4 to I.O. g. Of the very finely ground manganese ore is
weighed off, and treated in a tall beaker or in a flask with exactly 50 c.c.
of the hydrogen peroxide solution. The mixture is shaken at frequent
intervals during half an hour, after which time any undecomposed
hydrogen peroxide is titrated with permanganate solution. Should there
be between five to ten samples for analysis, the first will be ready for titra-
tion when the measured volume of peroxide solution has been added to
the last sample, and the various lots can be titrated in the order in
which they were weighed off. In some cases exact titration is rendered
difficult owing to the ore forming a markedly turbid brown coloured
solution. This difficulty may be overcome by carrying out the decom-
position in a IOO c.c. graduated flask, making the solution up to IOO c.c.
after half an hour, when decomposition is complete, filtering through a
double filter paper, and taking 50 c.c. of the filtrate for titration. The
number of c.c. required must be doubled for calculating the results on
the quantity weighed off.
* Z. angew. Chem., 1890, 3, 78. * Cf. Lunge, ibid., 136. * Ibid., 72.
MANGANESE ORE 481
MacLachlan' is of opinion that the estimation of the active oxygen
in hydrogen peroxide by means of an acid solution of potassium per-
manganate is altogether unreliable. Lunge maintains that this view
only applies to the volumetric determination, and accordingly Baumann's
method is to be regarded as inexact until further investigation proves
the contrary. Lunge has fully confirmed the accuracy of the gas-
volumetric method by a large number of tests.
3. Carbon dioxide. — This estimation is important, since carbon
dioxide constitutes a very harmful impurity in chlorine intended for the
manufacture of bleaching powder. Traces of carbonate may be detected
by stirring the powdered ore on a watch-glass with water until all
adhering air-bubbles have been removed, then adding a little dilute
hydrochloric acid, and viewing the surface of the solution from the side.
The liberated carbon dioxide comes off in the form of small, briskly
evolved gas-bubbles, and cannot possibly be confused with air. The
quantitative determination may either be carried out gravimetrically by
liberating the carbon dioxide with dilute sulphuric or nitric acid, and
absorbing in soda lime, or still better and more rapidly, by the gas-
volumetric method described on p. 149. A good ore should not contain
more than I per cent, of carbon dioxide.
4. The Hydrochloric acid necessary for decomposition.—One g. of
the ore is dissolved in Io c.c. of strong works acid of known strength,
the operation being carried out in a flask provided with a reflux con-
denser, and solution assisted by heating. The solution is then allowed
to cool and normal sodium hydroxide added until the separated reddish-
brown, flocculent precipitate of ferric hydroxide no longer dissolves on
shaking. The volume of sodium hydroxide solution required is calcu-
lated to the strength of the acid employed for the decomposition, and
the figure thus obtained for the excess of acid deducted from Io, the
number of the c.c. of acid taken in the first instance.
Débourdeaux” recommends determining the chlorine value and the
requisite acid in a single operation. The ore is treated with a mixture
of oxalic and sulphuric acids, and after solution the excess of the
former is determined by titration with permanganate, and the acid used
by titrating with ammonia and fluorescein. This method is not likely
to effect any saving of time.
II. LIMESTONE
A description of the characteristics of the most suitable limestone
for the manufacture of bleaching powder is given in Lunge's Sulphuric
Acid and Alkali.” The properties required in lime intended for the
* Chem. Soc. Proc., 1903, 19, 216. * Comptes rend., 1904, 138, 88.
* Vol. iii., p. 440.
2 H
482 - THE CHLORINE INDUSTRY
Weldon process are somewhat similar; much depends in both cases on
the absence of magnesia. The technical examination is carried out as
follows:—
I. Insoluble matter. — One g. of the limestone is treated with
hydrochloric acid, and the insoluble residue thoroughly washed, dried,
and ignited. Should appreciable quantities of organic matter be present,
the insoluble matter is first weighed on a tared filter paper, after drying
at 100°, and then ignited; the difference between the two weighings
gives the organic matter.
2. Lime.—One g. of the limestone is dissolved in 25 c.c. of normal
hydrochloric acid and the excess of acid determined by titration with
normal sodium hydroxide solution. The number of c.c. required is
deducted from 25 ; the remainder multiplied by 2.806 gives the percent-
age content of CaO, or multiplied by 5-OO6, the percentage content of
CaCOs. This method of calculating the result gives any MgO present
in terms of CaO but this is quite permissible in limestones intended for
use in alkali and bleaching powder works, since they should contain
only very little magnesium oxide.
For the same reason, it is equally permissible to calculate the lime
content from a determination of the carbon dioxide.
The latter is very frequently determined by loss. Very many forms
of apparatus for this purpose are in use." The most rapid and accurate
method is the gas-volumetric method described on p. 149. If O-4497 g.
of the sample be taken, each I c.c. of gas obtained, measured at O’ and
760 mm., corresponds to I per cent. CaCOs.
Should the limestone contain appreciable quantities of sesquioxides,
these must first be removed by precipitation with ammonia (free from
carbonate), and the calcium then precipitated as oxalate. In exact
work the precipitates should be dissolved and reprecipitated (cf. p. 405).
According to Passon,” this separation may be obviated by adding
phenolphthalein to the acid solution of the limestone, and then sufficient
ammonia to produce a distinct red coloration after precipitation is
complete. A Io per cent, solution of citric acid is then added until the
colour disappears and the precipitate goes into solution, after which a
further Io c.c. of citric acid solution is added, the whole diluted with water,
and the calcium precipitated by the addition of ammonium oxalate to
the boiling solution. Iron, aluminium, magnesium, and phosphoric acid
do not interfere when this method of precipitation is adopted.
3. Magnesia is, as a rule, only determined in limestone intended
for use in the manganese recovery process, or in the manufacture of
bleaching powder. Two g. of the sample are dissolved in hydrochloric
* Cº. Clowes and Coleman, Quantitative Analysis, 6th edition, 1903, p. 128; Treadwell,
Analytical Chemistry, vol. ii., p. 292.
* Z. angew. Chem., 1898, II, 776.
LIME - - 483
acid, the calcium precipitated by the addition of ammonia and
ammonium oxalate, and the magnesium determined in the filtrate by
precipitation with sodium phosphate (cf. p. 4O6).
4. Iron is usually only determined in limestone intended for the
manufacture of bleaching powder. Two g. are dissolved in hydrochloric
acid, the solution reduced by metallic zinc, diluted, and titrated with
permanganate after the addition of manganese sulphate solution free
from iron (cf. p. 380).
III. LIME
A. QUICK LIME
I. Free Lime—One hundred g. of the average sample are weighed
off, carefully slaked, and the resulting cream transferred to a 500 c.c.
flask, which is then filled to the mark. One hundred c.c. of the
thorougly shaken contents are then drawn off with a pipette and diluted
to 500 c.c. in a second flask, and 25 c.c. (= I g. quick lime) of the latter
well-mixed solution taken for analysis. A small quantity of an alcoholic
solution of phenolphthalein is first added and the solution then titrated
with normal hydrochloric acid until the red coloration disappears. The
colour change takes place when all the free CaO has been neutralised
and before the CaCOs has been attacked. Each I c.c. of normal acid=
o:O2805 g. CaO. The titration must be effected gradually, and with
thorough mixing, otherwise accurate results cannot be obtained
(cf. p. 72).
C. Stiepel has devised a “lime-calorimeter” for determining the free
lime by measuring the heat evolved on slaking. The calculation is
based on a liberation of 515OO calories in the combination CaO + H2O.
As may be imagined, the results obtained with such an apparatus will
not be very exact, and experience has shown that at times they are
quite unreliable. -
Maynard" estimates free caustic lime in admixture with other sub-
stances by extracting with pure glycerine at a temperature of 40°.
The extraction is continued for five days in a thermostat, and the
mixture frequently shaken. The whole is finally filtered at 60° and the
lime determined in an aliquot part of the filtrate.
2. Carbon dioxide.—The CaO and CaCOs are titrated together by
dissolving the lime in a measured volume of normal hydrochloric acid,
and titrating back the excess of acid with normal sodium hydroxide
solution. The CaCO3 present can then be calculated by deducting the
value found for the CaO under test I. In very exact work the carbon
dioxide should be determined directly as described on p. 149.
* Bull. Soc. Chim. 1902 [3], 27, 851 ; Chem. Mews, 1903, 87, Io9.
484 THE CHLORINE INDUSTRY
B. SLAKED LIME
I. Water.—About I g. of the substance is weighed off from a
stoppered weighing bottle, and gradually heated to a red heat in a
platinum crucible; it is then allowed to cool in the desiccator and
weighed. The loss in weight=water--carbonic acid.
2. Carbon dioxide is estimated as described above (A, No. 2).
3. Content of milk of Lime in caustic Lime.—Blattner's specific
gravity method. In the case of a thin milk, the reading must be made
quickly, and before the lime has time to settle out. In the case of
thicker milks, the hydrometer is placed gently in the solution, and the
containing cylinder slowly rotated, so as to cause gentle shaking, until
the hydrometer ceases to sink in the liquor. The cylinder employed
should not be too narrow. The following table, reduced from that
given by Blattner, gives the relationship between lime content and
specific gravity at I 5° C.

Degrees Grams Lbs. CaO Degrees Grams Lbs. CaO
Twaddell. CaO per litre. per cubic foot. Twaddell. CaO per litre. per cubic foot.
2 11-7 - 0-7 28 177 11 "I
4 24 °4 1°5 30 I90 II '9
6 37. I 2°3 32 203 12.7
8 49°8 3-1 34 216 13°5
10 62°5 3-9 36 229 14 °3
12 75-2 4-7 38 242 15-1
14 87-9 5°5 40 255 15-9
I6 100 6°3 42 268 16-7
18 113 7.1 44 281 17.6
20 126 7.9 46 294 18°4
22 138 8-7 48 307 19-2
24 152 9°5 50 321 20-0
26 I64 10°3 tº g tº gº tº e gº º tº
CONTROL OF WORKING CONDITIONS
I. MANUFACTURE OF CHLORINE BY MEANS OF NATIVE
- MANGANESE ORE
In this process analysis is confined to the determination of the
free acid in the manganese liquors coming from the chlorine stills. The
quantity of free acid present is determined by simple titration with
normal sodium hydroxide solution, the end-point being decided by the
appearance of a permanent, flocculent precipitate of ferric hydroxide. The
residual acid in a well-finished liquor may be less than 5 per cent. in cases
where the stills are heated indirectly; as a rule, the free acid rises to 6 per
cent. and over, especially when the heating is done with open steam.
* Dingl, polyt.J., 1883, 250, 464.
THE WELDON PROCESS 485
II. THE WELDON PROCESS
The finished liquors are examined in the following manner according
to Lunge's modification of Weldon's method.
The liquors from the stills are tested for free acid as described
above; with recovered manganese, such free acid should not exceed I
per cent. Occasionally the liquors are also analysed for their manganese
content, the method described under Weldon mud being employed (cf.
infra).
The neutralised and settled liquors are tested for neutrality by
means of litmus, and the degree of freedom from suspended matter
judged by appearance; the percentage of calcium chloride present is
determined by precipitation with oxalic acid in acetic acid solution."
The precipitated mud should further be repeatedly tested during
oxidation, as a check on the blowing operation. The method of testing
the mud is described below.”
The calcium chloride liquors which flow away from the settled mud
are examined for freedom from mud carried forward mechanically.
The mud obtained from the settling-tanks or filter-presses must be
examined for both soluble and insoluble manganese.
Hydrochloric acid used in the Weldon process should be as free as
possible from sulphuric acid.
The Examination of Weldon Mud.
I. Manganese dioxide. — For this determination a standard acid
solution of ferrous sulphate is required. It is prepared by dissolving
IOO g. of crystallised ferrous sulphate and IOO c.c. of concentrated pure
sulphuric acid in I litre of water (cf. p. 103), and is standardised by
diluting 25 c.c. with IOO to 200 c.c. of cold water and titrating with
M/2 potassium permanganate solution, until the rose coloration remains
permanent for at least half a minute. The ferrous sulphate solution
gradually changes in value, and must therefore be checked from day
to day.
The mud must be thoroughly shaken before testing (mere stirring is
not sufficient), and IO c.c. drawn off with a pipette immediately and
before any settling has occurred. The outside of the pipette is freed
from mud by washing, and the IO c.c. of mud allowed to flow into 25
c.c. of the ferrous sulphate solution contained in a beaker, any particles
adhering to the inner walls of the pipette being washed into the beaker
by the aid of a wash-bottle. The mixture is then shaken until all the
mud has dissolved, IOO c.c. of water added, and the solution titrated
with permanganate. If the 25 c.c. of acid ferrous sulphate solution
required a c.c. of permanganate, and the IO c.c. of the mud and ferrous
* Qſ Lunge, Sulphuric Acid and Alkali, vol. iii., p. 350. * Cf. ibid., p. 353.
486 THE CHLORINE INDUSTRY
sulphate y c.c. of permanganate, then 2. I75 (r-y) gives the number of
grams of MnO, present in I litre of the mud, and 2. 175(4-y)xO.O624
the number of pounds per cubic foot.
2. Total Manganese.—Ten c.c. of the mud are measured off, observ-
ing the precautions described under test I, and boiled with strong hydro-
chloric acid until chlorine ceases to be evolved. The excess of acid is
then neutralised by addition of powdered marble or precipitated calcium
carbonate, a strong filtered solution of bleaching powder added, the
whole boiled for a few minutes until the colour becomes deep red, taking
care that the mixture still smells of bleaching powder, when alcohol is
added drop by drop until the red colour is destroyed. The whole of the
manganese is then present as peroxide, which is filtered off and washed.
The filtrate should be tested for freedom from manganese by adding
bleaching powder solution, when no brown coloration should result.
Washing is continued until the washings cease to give a reaction with
iodised starch paper. The filter paper and precipitate are then thrown
into 25 c.c. of standard ferrous sulphate solution, and dissolved by
stirring as above. Should any of the manganese dioxide remain
undissolved, a further 25 c.c. of the ferrous sulphate solution is added.
The solution is diluted with IOO c.c. of water, and titrated with per-
manganate, the results being calculated as under test I, and expressed
was MnG, per litre or per cubic fgº
3. The “Base,” that is, tº noxides in the mud which neutralise
hydrochloric acid but do not liberate chlorine. It may contain lime,
magnesia, ferrous Oxide, and manganous Oxide.
Tenºc.c. of the well-shaken mud are added from a pipette to 25 c.c.,
or 50sp, in the case of a very high base, of normal oxalic acid solution
(63 g. st, oxalic acid per litre), previously diluted to about IOO c.c.
and warmed to a temperature of 60° to 80°, and the mixture well
stirred until the precipitate ceases to appear yellow and becomes pure
white. This should not take long, if the operation is carried out at the
temperature specified. The solution is diluted to 202 c.c. (2 c.c. repre-
senting the volume occupied by the precipitate) in a suitably marked
2OO c.c. flask, filtered through a dry pleated filter paper, and IOO c.c.
titrated with normal sodium hydroxide solution. Tincture of litmus or
phenolphthalein must be used as the indicator, since methyl orange can-
not be employed with Oxalic acid (cf. p. 65). The oxalic acid taken up
by the mud serves (I) to reduce the MnO, to MnO with liberation of
carbon dioxide; (2) to combine with and neutralise the manganous
Oxide so formed ; (3) to Saturate the monoxides, including manganous
oxide originally present, that is, the “base.” The quantities taken up
under I and 2 are equal, and the two together have the same value as
the MnO, as determined in test I, and so equal ar—y, since the oxalic
acid solution is normal, whilst the permanganate solution is only semi-
THE DEACON PROCESS 487
normal. If 2 c.c. of normal sodium hydroxide are used in titrating
back the excess of oxalic acid, the substances under 3 correspond
to the oxalic acid originally taken, 25 (or 50) c.c. less a -y and 23;
therefore this value w = 25 (or 50) — (t+28)+y. By the term “base "
is understood the ratio between the value for the substances under 3,
expressed by w, and the value for I, expressed by #2 (the sodium
hydroxide solution being normal and the permanganate semi-normal);
this ratio therefore equals #, When 25 c.c. oxalic acid solution are
used, it becomes:— -
50 – 24 - 43+29 (*#) — 2
4–3) 4 – y
º º IOO — 4 3
or, when 50 c.c. Oxalic acid are taken = ( *:::) — 2.
III. THE DEACON PROCESS
In this process the ratio between hydrochloric acid gas and air must
be determined both in the gases leaving the Saltcake pot, and also in
the gases leaving the Deacon decomposer. -
In the case of the gases from the saltcake pot it suffices to deterºis'
the volume percentage of hydrochloric acid gas present, as the residual
gas is essentially air. This is effected by aspirating the gaseºhrough
a measured volume of sodium hydroxide solution of known strength,
coloured with litmus or methyl orange until the indicator changes
colour; the volume of non-absorbed gas (air) is ascertained by measur-
ing the water which has run from the aspirator. Since the ratio of the
volume of hydrochloric acid gas to the sodium hydroxide solution used
is constant, and the volume of the air is given by the water collected,
the ratio between the acid and the air is easily calculated. The principle
of the method and the apparatus employed are the same as those
described under Lunge's modification of Reich's method for the analysis
of kiln gases (p. 299). -
The gases leaving the decomposer are tested, as a rule, for free
chlorine and for unchanged hydrochloric acid, and occasionally also for
water vapour and for carbon dioxide.
I. Ratio between the free Chlorine and the unchanged Hydro
chloric acid gas.—Decomposition value. Simple absorption of the gases
in sodium hydroxide solution followed by titration for available and total
chlorine is not applicable in this case, since it is impossible to avoid the
formation of chlorate." This difficulty has been overcome by the follow-
ing method, devised in the works of Messrs Gaskell, Deacon, & Co.
* Cf. C. Winkler, Industrie Gase, vol. ii., p. 318.
488 THE CHLORINE INDUSTRY
Five litres of the gas leaving the decomposer are aspirated through
250 c.c. of sodium hydroxide solution of sp. gr. I.O75, divided between
two or three bottles, to absorb both the hydrochloric acid gas and the
chlorine. The apparatus should be fixed as near to the decomposer as
possible, and the time during which the gas sample is taken should
correspond with the period of working a charge through the Saltcake
pot. The contents of the several absorption bottles are united, diluted
to 500 c.c., and tested as follows:—
I. One hundred c.c. of the solution are heated to boiling with 25 c.c.
of the acid ferrous sulphate solution, prepared and standardised as
described in p. 478, in a flask provided with a Contat bulb (Fig. 39, p. IO6).
When cold, the contents of the flask are diluted with 200 c.c. of water,
and titrated with N/2 permanganate solution. Let y represent the
number of c.c. so required and a the number of c.c. of permanganate
necessary to oxidise the 25 c.c. of ferrous sulphate solution.
2. A small volume of sulphur dioxide solution is added to Io C.C. of the
above alkaline solution and the mixture acidified with dilute sulphuric
acid, so that there is a distinct smell of sulphur dioxide. The acidified
Solution is heated to boiling, allowed to cool, a few drops of perman-
ganate solution added, if necessary, to oxidise any remaining sulphurous
acid, after which the solution is neutralised with pure sodium carbonate
diluted with water and titrated with M/Io silver nitrate solution, employ-
ing neutral potassium chromate as indicator. Let & stand for the
number of c.c. silver nitrate solution required. The percentage decom-
position of the hydrochloric acid gas is then given by the expression
a; - y
*g-2. and the volume of air per I volume HCl by tºrr.
Should some other volume of gas, n, be aspirated instead of the 5 litres,
the constant, 43-53, must be altered to z #624. assuming that the
50 × O.OO3645
remaining operations are carried out exactly as above, and that I litre
HCl at 50° and 760 mm. pressure weighs 1.624 gram.
Younger" passes the gas through a solution of arsenious acid and
employs an aspirator which gives the weight of chlorine in unit volume
of the gases directly. The percentage of hydrochloric acid in the gas is
determined by titrating the resulting solution with silver nitrate. The
absorption is effected in a cylinder containing IOO c.c. of an aqueous
Solution of arsenious acid, each I c.c. of which corresponds to O. 15432
grains (=O.OI g. chlorine). The charged cylinder is connected with the
bottle B (Fig. I4O), containing about I g. potassium iodide dissolved in
water ; the arsenious acid solution is coloured blue by a very small
quantity of indigo-carmine.
* / Soc. Chem. Ind., 1889, 8, 88.
THE DEACON PROCESS - 489
A cubic foot (o-o283 cub. m.) in capacity of the aspirator C is divided
into any convenient number of parts; for example, I 12. One side of
the gauge glass D is calibrated to give the grains of chlorine per cubic
foot of gas, and the corresponding volume of gas aspirated is given
on the other side of the same lines. The liberation of iodine from the
potassium iodide solution and the simultaneous bleaching of the indigo-
carmine indicate the completion of the test, and when this occurs the
aspiration is interrupted and the readings taken. Thus, for example, if
the mark E represents I cubic foot and the water stands at this level on
the completion of the test, the readings indicate that I cubic foot of the
gas contains 15:432 grains of chlorine. Should
the water-level stand at the cubic foot mark,
the presence of 30.864 grains chlorine per cubic
foot of gas is indicated by the scale, and so on.
The hydrochloric acid is determined by
titrating IO c.c. of the arsenious acid solution
with N/Io silver nitrate solution. Should the
gases be free from hydrochloric acid the silver
nitrate necessary for titration will be 28.2 c.c.,
this volume corresponding to the hydrochloric
acid derived from the chlorine absorbed. Any
volume over and above this 28.2 c.c. corresponds
to the hydrochloric acid present as such in the
gases.
The iodine liberated in the potassium iodide
bottle generally indicates about O'2 to O-3
grains of chlorine. The potassium iodide bottle
is not necessary as an indicator for the com-
pletion of the reaction; its main function is to
show that the aspiration is not carried out too
rapidly, and that no chlorine passes unabsorbed through the arsenious
acid solution.
In a subsequent communication * Younger gives the following
particulars for the analysis of Deacon gases: A measured volume of the
gas is drawn first through weighed bottles containing concentrated
sulphuric acid, the gain in weight giving the moisture present; secondly,
through a tube containing a solution of arsenious acid, to arrive at the
chlorine and hydrochloric acid content, as above; and finally, through an
apparatus in which the oxygen may be absorbed and measured.
The following example, showing the methods of calculating the
results, is taken from Winkler,” the data for the gases being corrected
to the real litre weights. The free chlorine was determined by an
arsenious acid solution, of which I c.c. =O.OO25 g. Cl, and the total
* / Soc. Chem. Ind., 1890, 9, 159. * /ndustrie Gase, vol. ii., p. 317.

490 THE CHLORINE INDUSTRY
chlorine by a silver nitrate solution, of which also I c.c. =OOO25 g. Cl.
The following volumes of these solutions were consumed per litre of
unabsorbed gas. Fifty c.c. arsenious acid solution = 50 × .OO25 = O. 125 g.
free chlorine, and IOO c.c. silver nitrate solution= IOO × O.OO25 = O-25 g.
total chlorine. The extent of the decomposition is given by the
equation — -
O.25O : O.I.25 = IOO : x,
3 = SO,
or a 50 per cent. decomposition had been effected. These two deter-
minations thus give the decomposition value and the ratio between the
chlorine and hydrochloric acid present in the gas.
Should it be desired to ascertain the volume percentage of the two
gases leaving the Deacon decomposer, it is necessary to calculate to c.c.
the ascertained weight of free chlorine, and of hydrochloric acid. One
c.c. Cl–O.OO3219 g, consequently the O. I25 g. Cl found correspond to
4O2 C.C. Cl, which is the volume present per I litre of non-absorbable gas.
To calculate the volume of the hydrochloric acid present, since its
weight corresponds to half that of the total chlorine in the gases, it is,
O.25O
= O.I 25 g. Cl- o. I 285 g. HCl.
One c.c. HCl weighs O.OOI64I g.; therefore:—
O. I285 g. = 78.3 c.c. HCl
per litre of non-absorbable gas. The gas leaving the decomposer thus
possesses the following composition :-
IOOO C.C. Oxygen and nitrogen,
78.3 a hydrochloric acid gas,
40.2 m chlorine ;
or, expressed in percentages by volume:—
89.40 vol. per cent. Oxygen and nitrogen,
7-OO , 3) hydrochloric acid gas,
3-6O , yy chlorine.
2. The percentage by volume of Hydrochloric acid gas in the
gases before entering the decomposer is arrived at from the total
chlorine as follows.
The total chlorine found, namely O-250 g., corresponds to O-257 g. or
I 56.6 c.c. HCl. Consequently the gases entering the decomposer
correspond to :— ©
IOOO.O C.C. alr,
I 56.6 c.c. HCl,
which gives 13.60 volume per cent. HCl."
It is usual to simplify the calculations as far as possible, and such
* This leaves out of account the oxygen which has interacted with the HCl, and assumes that
there is no leakage into the decomposer (Conroy). -
THE DEACON PROCESS 491
simplification does not affect the result to any great extent. For
example, the weight of I c.c. chlorine may be rounded off to O-Oo32 g.,
and that of I c.c. HCl gas to O.OOI6, thus saving trouble in calculat-
ing the inactive chlorine to hydrochloric acid; that is, the weight of I c.c.
HCl is taken to be one-half that of I c.c. of chlorine. If these simplifica-
tions be introduced in working out the above example, the following
values are obtained :— -
For the Gases leaving the Decomposer.
O.I 25 g. active chlorine
.OO32
o. 125 g. inactive chlorine
º .oO16
= 39.06 c.c. = 3.49 vol. per cent. (instead of 3.60)
-
78.12 c.c. = 6.99 vol. per cent. (instead of 7.004)
For the Gases entering the Decomposer.
o.2 sºme = I56.25 C.C. = I3.51 vol. per cent. HCl
(instead of
13.60).
It is unnecessary to make any corrections for temperature and
pressure, since the errors from this source are throughout essentially of
the same magnitude, and the results obtained are therefore sufficiently
comparable.
3. Carbon dioxide.—The estimation of this gas is important in the
examination of Deacon gases, since the presence of any appreciable
amount in the chlorine renders the production of high-strength bleach
impossible.
Hasenclever's method" of estimation consists in passing a measured
volume of the gas (previously freed from hydrochloric acid by bubbling
through a wash-bottle containing water) through an ammoniacal solution
of barium chloride. After the absorption is complete the solution is
heated, the barium carbonate filtered off and thoroughly washed with
boiled water. The washed precipitate is then either ignited and weighed,
or dissolved in hydrochloric acid and converted into barium sulphate.
The weight of carbon dioxide in the gas is calculated from the weight
of precipitate obtained (I g. Baso, = O. 1885 g. CO3), and compared
with the volume of gas, say 20 litres, collected in the aspirator. -
According to Sieber,” this method is only applicable to dilute gases
containing not more than IO per cent. of chlorine, a condition always
satisfied by Deacon gases. In the case of more concentrated gases, such
as electrolytic chlorine, which may contain carbon dioxide arising from
the carbon electrodes used, the method is unsuitable, owing to the solu-
bility of barium carbonate in barium chloride. The object of boiling
the solution before filtering off the barium carbonate is to destroy any
carbamate present in the liquors. It is also possible to determine the
* Cf. Winkler, loc. cit., p. 368. * Chem. Zeit., 1895, 19, 1963.
492 - THE CHLORINE INDUSTRY
carbon dioxide by absorbing the gases with sodium hydroxide solution,
destroying the sodium hypochlorite formed by addition of sodium
arsenite, and then liberating the carbon dioxide by means of Sulphuric
acid. This method is, however, inconvenient, since it necessitates a
previous determination of the carbon dioxide in the various reagents
employed; it is also uncertain, when only small amounts of carbon
dioxide are present. A better plan is to absorb the gases in Sodium
hydroxide solution of known carbonate content, to decompose the
sodium hypochlorite by boiling with cobalt oxide, and then to decompose
with sulphuric acid, passing the liberated carbon dioxide through a
solution of potassium iodide previously saturated with carbon dioxide
and air, to retain any traces of chlorine carried forward. The purified
gas is then collected in a cooled potash bulb, and weighed.
In this connection Lunge" remarks that the gas must, of course, be
dried beforehand, and that the process, in addition to being very Com-
plicated, is quite unreliable for the determination
ſº of small quantities of carbon dioxide, especially
| # on account of the necessity of working with a
solution of potassium iodide saturated as above.
It is both simpler and more accurate to absorb
the gas in sodium hydroxide solution, to decom-
pose the hypochlorite completely by boiling with
ammonia, and then to estimate the carbon dioxide
in the ordinary way either by liberating the gas
(p. 149) or by the barium chloride method (p. 72).
# == Treadwell” makes use of the burette shown in
f # Fig. 141, for estimating the carbon dioxide in
º' electrolytic chlorine. The burette is of the Bunté
type, and is provided with a levelling tube. The
capacity of the burette, from tap to tap, must be
accurately determined, the most suitable capacity
b
**
º
º
i
B
AW
(I,
ſº º --- being IOO c.c. The burette B should be thoroughly
º dried before making a determination, and the
chlorine gas to be examined should, after drying
---- by passing through a calcium chloride tube, be
- 41. allowed to flow through the burette for from five
to ten minutes before closing the taps. At the
end of this time the lower three-way tap a is closed, then the upper
two-way tap 6, and the temperature and barometric height noted.
The lower tap is then connected by the rubber tube to N, which is
filled with a 5 per cent. solution of sodium hydroxide; a is then turned
so as to connect N with the outside air, as shown in the figure.
As soon as the liquid flows through the tap, the latter is closed,
* Aſandbuch der Sodaindustrie, vol. iii., p. 698. * Analytical Chemistry, vol. ii., p. 614.


THE DEACON PROCESS - 493
N raised, and the tap a turned so as to allow a little of the solution to
enter B. The tap a is then closed, and the burette suitably inclined so
that its walls are thoroughly wetted with the solution; this promotes
the rapid absorption of the chlorine and carbon dioxide. The opening
and closing of the tap a and the movement of the burette is repeated
several times until absorption is complete. When all the chlorine and
carbon dioxide have been absorbed, the tubes B and N are adjusted to
the same level, and the volume of unabsorbed gas (oxygen, nitrogen,
carbon monoxide) read off. As a rule, the residue only amounts to from
o. 5 to I per cent.
To estimate the chlorine, the funnel on the top of the burette is first
washed out with water, in order to remove any small quantity of
chlorine that may remain from the introduction of the sample, the tap
a closed, and the solution in N run out into a beaker through the side
connection of a ; N and the rubber tubing are then disconnected and
washed out. The contents of the burette are transferred to the same
beaker and the burette well washed with water, added through the
funnel. The absorbed chlorine is titrated with arsenious acid. One
hundred c.c. M/IO sodium arsenite solution are added (this quantity
will always be found sufficient), then phenolphthalein, followed by hydro-
chloric acid till the red colour first disappears, and the excess of
arsenite solution determined by titration with M/IO iodine solution,
starch solution being used as indicator. The number of c.c. of chlorine in
the gas, measured dry at O’ and 760 mm., are obtained by multiplying
the number of c.c. of arsenite solution consumed by I. IoI 5. The
value so obtained must be recalculated, to give the volume at the
temperature and pressure at which the gas was measured. The carbon
dioxide is obtained as the difference between the chlorine and the total
gas absorbed.
This method is subject to the error that appreciable quantities of
chlorate are formed in the absorption of the chlorine, and the results are
always from O-7 to O.77 per cent, too low." Treadwell and Christie.”
have accordingly modified the process by first absorbing the chlorine
with an alkaline arsenite solution free from carbonate, and then the
carbon dioxide by alkali hydroxide. The arsenite solution is made by
dissolving 4.95 g. of arsenious oxide in dilute potassium hydroxide
solution, adding phenolphthalein, exactly neutralising with sulphuric
acid, and diluting to I litre. One hundred c.c. of this M/IO solution are
introduced into the burette through the tap a, for the absorption, and
subsequently IO c.c. of a I : 2 potassium hydroxide solution through the
funnel above b. After taking the readings, the solutions and washings
are collected, phenolphthalein added, and the whole neutralised with
hydrochloric acid; 60 c.c. of sodium bicarbonate solution (40 g. per litre)
* Z. angew. Chem., 1905, 18, 1930. 2 /øia.
494 THE CHLORINE INDUSTRY
are then introduced, and the excess of arsenite solution titrated with
M/Io iodine solution and starch. One c.c. M/IO solution=OOO3545 g.
Cl, or 1. Ior 5 c.c. N.T.P. The carbon dioxide is determined by differ-
ence, as above.
Correct results may also be obtained if two samples of the gas are
taken at the same time in a couple of Bunté burettes arranged in series,
the gas in the first being treated with sodium hydroxide solution as
above, and that in the second with a solution of potassium iodide, and
the chlorine estimated by titrating the liberated iodine with standard
arsenite or thiosulphate solution. Each I c.c. M/IO solution required
corresponds to O.OO3545 g., or I. IOI 5 c.c. dry chlorine measured at O’ and
760 mm. The volume under these conditions is calculated to the
volume under the conditions of analysis by means of the formula
76O(273+?)
(5–f)273
perature tº ; or the Tables VI, VII, and VIII. (Appendix) may be used.
The carbon dioxide is arrived at as the difference between the volumes
absorbed in the two burettes. It can also be determined directly in a
single burette by introducing exactly 45 c.c. N/5 sodium hydroxide
solution the carbonate in which has been determined, as described below
(p. 514). These 45 c.c. are placed in the funnel of the burette after the
sample has been collected, and as much as possible allowed to enter the
burette; the whole is then shaken. From 5 to Io c.c. of a neutral 3 per
cent, solution of hydrogen peroxide are added to decompose the sodium
hypochlorite, the burette again shaken, the contents run into a 200 c.c.
flask, the burette well washed with water free from carbon dioxide, and
the solution made up to 200 c.c. Fifty c.c. of this solution are taken
for the determination of the carbon dioxide which may be carried out
very accurately by the Lunge and Marchlewski method (p. 149), or suffi-
ciently accurately, for most purposes, by adding sodium chloride and
phenolphthalein and titrating with M/5 hydrochloric acid at a tempera-
ture slightly above o’."
Treadwell and Christie.” have modified this method as follows:–
The chlorine is absorbed by a 5 per cent, solution of potassium iodide,
then the carbon dioxide by IO c.c. of potassium hydroxide solution, and
the total volume of absorbed gas determined. The excess of hydroxide
converts the liberated iodine into iodide and iodate. The liquid from
the apparatus is then run into a beaker containing IO c.c. of strong
hydrochloric acid, when the iodine is again liberated, and is titrated
with thiosulphate. The results are accurate, but the method presents
no advantage over the arsenite method as described above.
Lunge and Rittener” absorb the chlorine in a Bunté burette with .
where fstands for the tension of aqueous vapour at the tem-
* Cf. Offerhaus, Z. angew. Chem., 1903, 16, Io93. * Ibid., 1905, 18, 1930.
* Z, angew. Chem., 1906, 19, 1853.
MANUFACTURE of BLEACHING Powder 495
an W/Io solution of sodium arsenite and then the carbon dioxide with
sodium hydroxide ; the unchanged arsenite is titrated back and the
carbon dioxide obtained by difference from the volume of total
absorbable gases found. The details of the method are given on p. 514
The above methods are of considerable importance in the analysis of
electrolytic chlorine. Nourisson examines such chlorine in an Orsat
apparatus by first absorbing the chlorine by stannous chloride, then the
carbon dioxide by a solution of sodium hydroxide, and finally the
oxygen by metallic copper and ammonia solution. Schloetter” absorbs
the chlorine by hydrazine Sulphate, each 2 volumes of chlorine absorbed
liberating 1 volume of nitrogen, and subsequently the carbon dioxide by
sodium hydroxide solution.
4. Moisture in the Deacon gases may be determined when neces-
sary by inserting a U-tube filled with pumice moistened with concen-
trated sulphuric acid, or a bulb apparatus containing sulphuric acid,
between the gas main and the absorbing bottle for the chlorine and
hydrochloric acid, and determining the gain in weight. It is essential to
drive all chlorine and hydrochloric acid gas out of the absorbing tube
before weighing, by passing a current of dry air through the U-tube for
a sufficient time; also it is advisable to insert a second weighed sulphuric
acid tube behind the drying apparatus during the operation.
IV. THE MANUFACTURE OF BLEACHING POWDER
The method of examining the lime used in filling the chambers has
already been described (p. 484). The main point to be attended to in
the manufacturing operation is the strength of the bleach, which must be
brought to a specified standard. This is controlled by analysis; the
various methods available are described under “Finished products”
(p. 499). It is, of course, essential that the sample drawn from the
chamber should be representative of the bulk.
The chambers must not be opened until it has been ascertained that
this may be done with safety to the workmen, and without giving rise to
undue nuisance.” Before the chamber is opened, the chlorine present
in the gas space must not exceed 5 grains per cubic foot (= I I-5 g. per
cubic metre); this limit was agreed to by manufacturers about twenty
years ago, but the present practice is to restrict the limit to 2% grains
per cubic foot. *
The determination of the chlorine present in the chamber atmos-
phere may be carried out in an Orsat apparatus, the simplified form
described by Fleming-Stark, Fig. 142, being especially suitable.
The burette a is filled with water and connected by means of a
* Chem. Zeit., 1904, 28, Ioy. & Cº. Lunge, Sulphuric Acid and Alkali, vol. iii., p. 458.
* Z. angew. Chem., 1904, 17, 30I. * J. Soc. Chem. Ind., 1885, 4, 3II.
496 THE CHLORINE INDUSTRY
rubber tube with the reservoir à, a tap c being inserted between the two.
This tap is provided with two passages at right angles to each other, the
one of small and the second of large diameter. This arrangement
allows of a strong flow of water when filling the burette, and of a
diminished flow when the gas is forced into the absorbing solutions.
The four tubes d are filled
with an aqueous solution of
potassium iodide, and each
may be connected through
a corresponding glass tap
with the burette. Each
absorption tube is pro-
vided with a double-bored
stopper, and a tube reach-
ing almost to the bottom
of the absorption vessel
passes through one of these
openings. This tube is
narrowed at its lower end,
*º-- *=º ºk. So as to break up the gas-
FIG. 142. - bubbles entering the ab-
Sorbing solution, and at its
upper end is connected through the tube e with the bleaching powder
chamber. The tube passing through the second opening in the stopper
is cut off just below the stopper, and serves to connect the absorbing
vessels with the burette. A small wash-bottle containing potassium
iodide solution and starch is inserted between the absorption tubes and
the burette. The two-way tap g, provided between the wash-bottle and
the burette, permits the air to escape during the filling of the burette
without passing through the wash-bottle.
In using the apparatus, 387.7 c.c. of gas, as measured in a, are drawn
through one of the absorption tubes d'; the solution in the wash-bottle
affords an absolutely safe indication of the completeness of the chlorine
absorption. The contents of d are then washed into a beaker, and
titrated with M/IO sodium arsenite solution. The grains of chlorine per
cubic foot are obtained by multiplying by 2 the number of c.c. of arsenite
solution so required.
The Government inspectors make use of the simple apparatus shown
in Fig. 143* A is an ordinary rubber pressure ball of about 100 c.c.
capacity, provided with a small hole, B, in the mouthpiece. The end of
the mouthpiece passes through one of the two holes in the cork C, a
glass tube, D, bent at right angles passing through the second. This
latter tube reaches nearly to the bottom of the cylinder E, and is
* Lunge, Sulphuric Acid and Alkali, vol. iii., p. 458.

MANUFACTURE OF POTASSIUM CHLORATE 497
narrowed down at the lower end, so that only a very fine needle can be
inserted. The cylinder E is filled with the solution described below, and
the outer end of D is inserted in an opening in the
bleach chamber at a height of 2 feet above the floor.
To take a sample of the chamber gas, A is com-
pressed, the hole B being covered by the finger, and
the pressure then released ; by the expansion of the
bulb the gas in the chamber is drawn through the
tube D and the solution in E. The operation is
repeated until the solution in E becomes coloured
through separation of iodine, and the number of
aspirations required to cause this is noted. Each
delivery of the bulb corresponds to 4 ozs. (about
IOO C.C.), or g-Ho of a cubic foot. The solution em-
ployed is prepared by dissolving O-3485 g. of arsenious
acid in Sodium carbonate, neutralising with sulphuric
acid, adding 25 g. of potassium iodide, 5 g. of precipi-
tated calcium carbonate, 6 to IO drops of ammonia,
and diluting the whole to I litre. Twenty-six c.c. of
this solution are employed for each test, and a little ==E=::=::::::::::::::::::
starch solution is added as indicator. Under these FIG. 143.
conditions five deliveries of the bulb will produce a
coloration when the gas contains 5 grains of chlorine per cubic foot, ten
deliveries when the chlorine content is only 2% grains per cubic foot, and
SO O11.
V. THE MANUFACTURE OF POTASSIUM CHILORATE
In the old process of manufacture, in which a warm solution of milk
of lime is treated with chlorine, the methods of examining the lime and
milk of lime are those already described (p. 484). The resulting liquor
must be examined for its content of chlorate, since this is the basis on
which the potassium chloride necessary for conversion into chlorate is
calculated, and further for the percentage of calcium chloride present, as
a check on the working of the process.
Determination of the Chlorate.—Two c.c. of the liquor are measured
off with an accurate pipette and boiled, after the addition of a little hot
water and a drop of alcohol, to remove any dissolved free chlorine. A
flask fitted with a Bunsen valve or Contat bulb (Fig. 39, p. IO6) is used
for the test, but the valve is not inserted during the above preliminary
heating. The complete removal of chlorine is indicated by the disap-
pearance of all smell of chlorine and of the red colour. The flask and
contents are allowed to cool, 25 c.c. of the acid ferrous sulphate solution
(corresponding to a c.c. N/2 permanganate solution) described on p. 485,
added, the whole boiled for ten minutes, and the valve inserted. When

2 I
498 THE CHLORINE INDUSTRY
cold, the solution is titrated with N/2 permanganate solution, & c.c., say,
being necessary to produce the coloration. The chlorate contained in
I litre of the liquor will then be 5.105 (a – 6) g. calculated as KClO3, and
the potassium chloride theoretically necessary to convert the calcium
into the potassium salt will be 3. 106 (a –b) g per litre.”
The chlorate remaining in the mother liquors is similarly determined.
The chlorate may also be determined by Bunsen's method, namely,
boiling with strong hydrochloric acid and absorbing the liberated
chlorine in a solution of potassium iodide; the method is, however, no
more advantageous in this case than in that of manganese ore (p. 478).
Rasenack” describes a method which consists in precipitating the
chloride present by addition of silver nitrate, reducing the chlorate in
the filtrate by means of zinc, and titrating the resulting chloride. This
method is much more troublesome than the above, and the same may be
said of all other reduction methods depending on the estimation of the
chloride produced. An iodometric method has been proposed by de
Koninck and Nihoul.”
Other methods for the determination of chlorates are described
On p. 5 II.
Calcium chloride.—The chloride (calcium chloride) in the liquors
is estimated as follows:—I c.c. of the liquor is treated as above, to
destroy free chlorine, until the red colour disappears, a little neutral
potassium chromate added, and the solution titrated with M/IO silver
nitrate solution, as described on p. 123. Each I c.c. of the silver nitrate
solution corresponds to chlorine equivalent to 7.46 g. KCl per litre.
Should it be necessary to determine the free Chlorine and Hypo-
ch/orite present in the liquor, a measured volume of the latter is
allowed to flow into an excess of potassium iodide solution and the
iodine liberated determined by titration with thiosulphate. It is not
permissible to add sodium carbonate to fix the chlorine before adding
the potassium iodide, since, owing to the formation of iodic acid and
to the oxidation of the thiosulphate, too much iodine is then required.*
The “bleaching chlorine” may, however, according to Pontius, be
titrated directly with potassium iodide solution after addition of sodium
bicarbonate (cf. p. 503).
The titration may also be carried out by Penot's method, described
under bleaching powder, by addition of sodium arsenite solution, until a
drop of the titrated liquor ceases to produce a blue colour on potassium
iodide starch paper; or the arsenite solution may be added in excess
and titrated back with iodine solution.
* Cf. also, Rosenbaum, Z. angew. Chem., 1893, 6, 80.
* Dammer, Lexikon der Verſälschungen, p. 423.
* Z. angew. Chem., 1890, 3, 77.
* Friedheim, Z, anorg, Chem., 1893, 4, 145.
BLEACHING POWDER 499
FINISHED PRODUCTS
I. BLEACHING POWDER
Owing to the unstable nature of this substance, very special attention
must be given to the collection and preservation of the samples.
Exposure to air and to daylight is very prejudicial ; exposure to Sun-
light still more so. The samples are drawn from the separate casks by
a suitable auger (p. 13), and placed one after the other in a wide-
mouthed bottle, which must be closed immediately after each portion
has been introduced. The mixing and filling into the separate sample
bottles, together with the sealing up of the latter, is performed as
rapidly as possible in the manner prescribed on p. II. The samples
must be kept in a cool, dark place prior to testing, and the tests should
be made without delay.
The technical analysis of bleaching powder is confined to the deter-
mination of the “available” or “bleaching” chlorine, that is, to the com-
pound CaCl(OCl) which splits up into Ca(OCl), and CaCl, on solution
in water, and consequently shows the reactions of a hypochlorite. The
quantity of available chlorine present is, in England and America,
always expressed as percentage by weight on the bleaching powder, and
the same method is usually followed in Germany and most other
countries. In France, however, and to a certain extent in other
countries, the strength is quoted in terms of Gay-Lussac degrees, which
indicate the number of litres of chlorine gas, reduced to O’ and 760 mm.,
capable of being evolved from IOOO g. of bleaching powder. The
following table shows the relationship between the English and French
degrees.


French | Per cent. | French | Per cent. French | Per cent. Fronch | Per cent. | French Per cent.
Degrees. Chlorine. Degrees. Chlorine. Degrees. Chlorine. Degrees. Chlorine. I Degrees. Chlorine.
63 20*02 77 24-47 91 28 -92 105 33°36 119 37°81
64 20 °34 78 24-79 92 29-23 106 33°68 120 38°13
65 20-65 79 25:10 93 29 °55 107 34°00 121 38°45
66 20-97 80 25°42 94 29.87 108 34°32 122 38-77
67 21°29 81 25 °74 95 30-19 109 34°64 123 39°08
68 21°61 82 26'06 96 30°51 110 34°95 124 39°40
69 21°91 83 26'37 97 30 "S2 111 35-27 125 39°72
70 22-24 84 26'69 98 31-14 112 35 °59 126 40’04
71 22°56 85 27:01 99 31 °46 113 35 °9 | 127 40°36
72 22°88 86 27-33 100 31°78 114 36°22 128 40°67
73 23°20 87 27-65 101 32-09 115 36°54 tº e ºl tº e tº
74 23 °51 88 27-96 102 32°41 116 36 °86
75 23°83 S9 28°28 I03 32-73 117 37 -18
76 24-15 90 28-60 104 33-05 118 37-50
The following table, prepared by Lunge and Bachofen, gives the
* Z, angew. Chem., 1893, 6, 326.
500 THE CHLORINE INDUSTRY
available chlorine corresponding to various specific gravities of bleaching
powder solutions.

spectºriº || 3: spºravity || 3: spºils | . .
e Grms. per litre. & Grms. per litre. & Grms. per litre.
1 *1155 71-79 1 "0800 49°96 1*0350 20*44
1*1150 71°50 1°0750 45-70 1*0300 17:36
1*1105 68°40 1-0700 42°31 1*0250 14-47
1 *II.00 68°00 1*0650 39°10 1*0200 11 “41
1*1060 65°33 1*0600 35°81 1-0150 8’48
I "1050 64°50 I*0550 32°68 1°0100 5-58
1*1000 61°50 1*0500 29'60 1*0050 2.71
1°0950 58°40 1*0450 26*62 1°0025 1°40
1°0900 55-18 1*0400 23-75 1°0000 Trace
1*0850 52.27 tº e e & © tº © & © • * *
Very many methods" have been devised for the estimation of the
available chlorine in bleaching powder, of which the following are the
most important. 4.
Gay-Lussac's Arsenious acid method.—This, the oldest method,
was introduced by Gay-Lussac in 1835, and is still in general use in
France. It consists in treating the bleach with a hydrochloric acid
Solution of arsenious acid, prepared by dissolving 4,409 g, of arsenious acid
in hydrochloric acid, and diluting with water to I litre. Ten c.c. of this
solution are rendered blue by the addition of a drop of indigo solution,
and titrated with a solution of bleaching powder until the blue colour
disappears. The bleach solution is prepared by rubbing IO g. of the
bleaching powder with water and diluting to IOOO c.c. Since O.O4409 g.
AS,C), corresponds to IO c.c. of chlorine gas measured at O’ and
760 mm., it is only necessary to divide the number of c.c. of the bleach
solution required to decolorise the indigo solution into IOOO, to obtain
the number of litres of chlorine gas corresponding to I kilo of bleach.
The method is, however, very inexact, and the results obtained vary
considerably, according to the degree of dilution and the excess of acid
present. Chlorine and arsenious acid can exist side by side in dilute
solutions, and consequently the bleaching of the indigo is no indication
that all arsenious oxide has been oxidised. Further, the indigo is
always partially decolorised where the bleach solution enters the
arsenious acid solution, so that the coloration gradually becomes weaker
as the titration proceeds, thus causing the end-reaction to be very
indistinct. -
On account of this gradual decolorisation, Denigés” replaces the indigo
sulphate by potassium bromide, the end of the titration being indicated
by the appearance of a distinct yellow colour. This indicator is also,
* Cf. Lunge, Sulphuric Acid and Alkali, vol. iii., p. 429.
* /. Pharm. Chim, I891 [5], 23, IoI.
BLEACHING POWDER 501
according to Denigés, very suitable for titrating commercial hypochlorite
solutions (Eau de Javel) when these are coloured pink through presence
of permanganate. In this case the small amount of permanganate
present is first decomposed by the arsenious acid solution. The
estimation is carried out in the usual manner by adding a few drops of
concentrated potassium bromide solution to IO c.c. of the bleach solu-
tion, and titrating the arsenious acid with this solution until a permanent
faint yellow coloration is obtained.
Graham's Ferrous sulphate method.—This method, formerly in
general use in England and Germany, is no more accurate than the
above and is, moreover, distinctly inconvenient to carry out. It depends
on the oxidation of ferrous salts by bleaching powder in acid solution.
3.9 g, of crystallised ferrous sulphate, FeSO, 7H2O, which require for
complete oxidation O. 5 g. of chlorine, are dissolved in 50 c.c. of water
acidified with sulphuric acid, and titrated with a mixture of 5 g. of
bleaching powder and IOO c.c. of water until a drop of the solution
ceases to give the ferrous iron reaction when added to a drop of
potassium ferricyanide solution. Ferrous ammonium sulphate, which
has been strongly recommended by Mohr in place of ferrous sulphate, is
quite unsuitable for this test, as a portion of the chlorine is used up in
decomposing the ammonia (Biltz).
All methods of bleaching powder analysis must, according to general
commercial practice, permit of the determination not only of the
chlorine in the solution, but also of that present in the insoluble matter,
since the latter too exerts a bleaching action. The insoluble matter
must be uniformly distributed through the solution so that the whole
forms a uniform thin cream. Hence all methods similar to the above,
in which the bleaching powder cream is run in from a burette, are neces-
sarily inaccurate, since the distribution of insoluble matter can never be
uniform under such conditions. This source of error is avoided in the
following methods.
The sample is prepared for analysis by triturating it with water to a
thin cream. The grinding should be extremely intimate, and neither too
much nor too little water should be used in the operation. The triturated
cream should not contain so little water that the bleaching powder
separates out again in lumps while being washed into the litre flask; if
this does occur, more water must be added to the cream first prepared,
and the whole again triturated for a short time in the mortar. When
the operation has been properly performed, the cream will readily mix
on shaking with the further water added in the litre flask. -
Bunsen's Iodometric method depends on the separation of an
amount of iodine equivalent to that of the available chlorine, when a
bleach solution and potassium iodide are treated with hydrochloric acid.
The iodine liberated remains dissolved in the excess of potassium iodide
502 -- THE CHLORINE INDUSTRY
present, and is titrated with M/IO sodium thiosulphate solution (p. 117).
If I g. of bleaching powder be taken for the titration, the percentage of
available chlorine contained in the bleach is found by multiplying the
number of the c.c. of thiosulphate required by O-355. The titration
should be performed rapidly to prevent loss of chlorine, and any con-
siderable excess of hydrochloric acid must be avoided. It is best to
prepare the cream with I g. of bleach and about IOO C.C. of water, and to
add to this 2 to 3 g of potassium iodide and about IO drops of hydro-
chloric acid, and to stir once quietly with a glass rod so as to distribute
the separated iodine evenly throughout the Solution. The thiosulphate
is then added rapidly and without stirring until the colour of the solu-
tion has been reduced to a faint yellow, starch solution added until the
colour becomes deep blue, and the titration continued slowly and drop
by drop to completion.
When carefully carried out, the method gives accurate results but
great care is necessary, and in no case are the results better than those
obtained by Penot's method. The method is consequently but seldom
employed in the works, more especially as the consumption of potassium
iodide renders it expensive. Wagner's proposal to utilise the decolor-
ised solutions as a solvent for the iodine is not practicable, partly owing
to the dilution and partly for other reasons.
A very disturbing effect that frequently occurs in this method, is a
rapid reappearance of the blue colour; in the case of impure solutions
this often makes it impossible to determine the end-point of the reaction.
R. Schultz” has found that such cases are of very common occurrence in
determining the excess of bleaching powder in disinfected waste liquors,
and that the effect is to be ascribed to the presence of oxidising sub-
stances such as ferric oxide and calcium chlorate. According to Schultz,
very satisfactory results may be obtained, if instead of hydrochloric acid,
acetic acid be added to the mixed waste liquor, together with potassium
iodide solution and the resulting solution titrated with thiosulphate.
In the case of strongly coloured liquors, the end-point is judged by
comparing the colour of the titrated solution with that of the original,
diluted to an equal bulk.
Bunsen's Distillation method is even less acceptable than the
above for technical work. It consists in decomposing the bleaching
powder with hydrochloric acid, in a flask, driving off the chlorine by
boiling the mixture and passing the evolved gases into a solution of
potassium iodide contained in an inverted retort, the liberated iodine
being determined by titration. Any chlorate present in the bleach will
of course give rise to chlorine by this method. Fogh * has devised an
apparatus entirely of glass with ground-in parts for the determination.
* Z. angew. Chem., 1903, 16, 833.
* CAE. Foerster and Jörre, J. prakt. Chem., 1899, 59, 58.
BLEACHING POWDER 503
It has frequently been shown that the method possesses sources of
error," and in no case does the simplicity of the operation compensate
for the errors introduced. It is worthy of note that Bunsen 2 himself
never once used the distillation method in carrying out his classical
work on hypochlorites. -
Pontius' Potassium iodide method.”—This consists in titrating
bleaching powder with potassium iodide solution after previous addition
of sodium bicarbonate. 7.09 g, of bleach are dissolved in water in the
usual manner and the cream diluted to IOOO c.c.; 50 c.c. of the solution,
corresponding to O-3545 g. of bleaching powder, are measured off, and
about 3 g of solid sodium bicarbonate added to and dissolved in the
solution. One to two c.c. of starch solution are then added and the solu-
tion immediately titrated (before the liberated hypochlorous acid can
react with the starch Solution) with M/IO potassium iodide solution until
the coloration, which is at first reddish brown, and then blue, no longer
disappears and the colour becomes a bright and permanent blue. The
reaction consists in the formation of potassium iodate from the mixture
of hypochlorous acid, sodium bicarbonate, and potassium iodide :-
3Ca(OCl, (=6C)+6NaHCO,4 KI-KIO,4-3CaCO,4-6NaC14-3CO,4-3H,0.
This process is also suited to the examination of bleaching solutions,
the hypochlorous acid in this case being liberated by addition of boric
acid. The method is especially applicable to the approximate deter-
mination of Substances of quite unknown strength, and Lunge considers
it more convenient for this purpose than Penot's method; the latter is,
however, more exact in all cases.
Penot's Arsenious acid method.—This method, which finds the
greatest acceptance in practice, is, whilst free from all sources of error,
both simple and easy of manipulation.* The following details for carry-
ing it out are recommended by Lunge.
An alkaline solution of N/IO sodium arsenite is employed in place of
the acid solution used in the Gay-Lussac method. The preparation of
this solution has been described on p. 122. In alkaline solution the
oxidation of arsenious to arsenic acid, which requires 4 atoms of chlorine,
per gram molecule, takes place smoothly and without loss of chlorine,
the end-point of the titration being shown with extreme sharpness by
the blue coloration produced on iodised starch paper. An excess of
the arsenite solution may be added, if preferred, and such excess
titrated back with W/Io iodine solution. This latter method is, how-
ever, seldom employed, since it necessitates the use of two standard
solutions and is in no way more accurate.
In carrying out the test, 7.09 g, of the thoroughly mixed bleach
* Cf. Winkler, Z. angew. Chem., 1903, 16, 33. * Chem. Zeit., 1904, 28, 54.
* Annalem, 1856, 86, 265. *J. prakt. Chem., 1896, 54, 59.
504 THE CHLORINE INDUSTRY
samples are ground in a porcelain mortar with a little water to a uniform
thin cream, diluted by addition of more water, and the whole washed
into a litre flask and made up to IOOO c.c. The lip of the mortar should
be greased on its under side with a little fat or vaseline, to facilitate the
transference of the liquid from the mortar to the flask. Fifty c.c. of the
thoroughly mixed solution (=O-3545 g. bleaching powder) are pipetted
off and transferred to a beaker, where they are thoroughly stirred, whilst
the M/Io arsenite solution is run in until nearly the full quantity
expected to be required has been added. A drop of the solution is
then placed on filter paper which has been moistened with a little
potassium iodide starch solution. According to the depth of the blue
colour obtained (in presence of larger quantities of chlorine a brown
stain is produced), more or less additional arsenite solution is added and
the spotting repeated until the test paper remains white, or until only a
quite inappreciable coloration is produced. Each I c.c. of the arsenite
required represents I per cent. of available chlorine in the sample.
Suitable iodised starch paper for this test is prepared as follows:–
One g. of starch is boiled with IOO c.c. of water, the solution filtered and
O. I g. of potassium iodide added to the filtrate. The filter paper is then
soaked in this solution and dried on a porcelain or similar plate at 40°
to 50°. For the spotting test the paper should be moistened, as the
reaction is much cleaner and more delicate with moist than with dry
paper. Three or four pieces of the paper are spread on a glass plate and
any excess of water allowed to flow off; they are then ready for spotting.
The disappearance of the blue colour may be very sharply determined
if the glass plate is examined by transmitted light towards the end of
the titration.
Lunge's Gas-volumetric method.—Penot's method is so simple,
convenient, and accurate, that no other method is really needed. It is,
however, desirable to have an independent method, which may, when
required, be used as a check. Lunge's gas-volumetric method serves
this purpose. The hypochlorite is decomposed with hydrogen peroxide
in a nitrometer or gas volumeter, and the volume of the oxygen
liberated measured. The reaction takes place according to the follow-
ing equation —
CaOCl3+H,O2=CaCl2 + H2O + O.
That is, the volume of active chlorine present is exactly equal to the
volume of the oxygen liberated." The operation is carried out in a
nitrometer provided with a decomposition bottle, as in the analysis of
manganese ore (cf. pp. I 37 and 479). The addition of acid, as recom-
mended by Vanino, is of doubtful value.”
* Cº. Lunge, Chem. Ind., 1885, 8, 168; Ber, 1886, 19, 868; 2. angew. Chem, 1890, 3, 8;
Baumann and Vanino, Z. angew. Chem., 1890, 3,80, 186, 509.
* Cf. Z. angew. Chem., 1890, 3, 136.
BLEACHING POWDER 505
The bleaching powder solution is advantageously prepared so that 25
c.c., the volume taken for the reaction, contain I g. of bleaching powder.
This will be the case if 20 g. of the bleach are triturated with water as
described above, and the resultant cream diluted with water to 500 c.c.;
working with 25 c.c. of such a solution, each I c.c. of oxygen evolved,
reduced to O’ and 760 mm., equals O-3 I66 per cent. Of chlorine. The
test should be carried out with a bulb nitrometer graduated to 140 c.c.
Should it be desired to work with a smaller nitrometer, say of 3O to 50
c.c. capacity, only 5 c.c. of the above solution should be taken, in which
case each I c.c. of oxygen evolved will correspond to 5 × O-3166= I-583 per
cent. of chlorine. A still more convenient plan is to dissolve 7,915 g.
of bleach in 250 c.c. of water, and to take IO C.C. of the milky solution
for the test; the percentage of available chlorine is then equal to the
number of c.c. of oxygen liberated. A 50 c.c. instrument can be used
under these conditions.
The hydrogen peroxide is introduced into the inner compartment of
the decomposition bottle, and rendered just alkaline by the addition of
a few drops of sodium hydroxide solution to prevent the evolution of
carbon dioxide; the bleaching powder solution is then placed in the
outer compartment. The shaking should only be continued for from
one to two minutes after mixing the solutions, and the volume of gas
evolved should be read off immediately and before any spontaneous
evolution of oxygen occurs. The addition of a considerable excess of
sodium hydroxide to the peroxide must be avoided, but it is safe to add
the hydroxide until the solution is distinctly alkaline and a flocculent
precipitate appears. This addition should, of course, be made to each
portion of peroxide immediately before the test. The hydrogen per-
oxide solution should not be too concentrated ; I c.c. treated with
excess of bleaching powder solution in the nitrometer should not give
more than 7 c.c. total oxygen. If it is too strong, a little water may be
added directly to the solution in the decomposing flask. A small excess
of water is immaterial, and it also makes no difference whether the
hydrogen peroxide is employed in considerable or only in small excess.
Baumann" has shown that in carrying out the gas-volumetric
method with hydrogen peroxide it is very easy to obtain too high a
result when the hydrogen peroxide solution has been kept in a
stoppered bottle and has not been shaken before use; under these
circumstances the thorough shaking necessary in the analysis may give
rise to the liberation of appreciable quantities of oxygen (from o 5 to 2
mg. oxygen from IO c.c. of hydrogen peroxide). This oxygen results
from the continuous, gradual decomposition of the peroxide and remains
dissolved in the peroxide solution, owing to the somewhat higher
pressure exerted in the stoppered bottle.
* Z. angew. Chem., 1891, 4, 450.
506 THE CHLORINE INDUSTRY
The fear that the solid particles present in the milky bleach solution
might exert a catalytic influence and thereby lead to high results, has
proved to be groundless. The oxygen can be collected over either
mercury or water.
According to Lunge, the gas-volumetric method of estimating bleach
always gives higher results than are obtained by Penot's method. As
the result of a very extended series of tests, he states that such excess
of available chlorine amounts on the average to O. 15, or at most oz
per cent. The cause of this difference has not yet been definitely
elucidated.
Vanino's Hydrogen peroxide method.” – This is a volumetric
method for the analysis of bleaching powder based on the above reaction.
|
*::siſ
| i
º | | ... "
|
FIG. 144.
The bleaching powder is decomposed by adding an excess of hydrogen
Peroxide, standardised by means of permanganate, and the undecom-
posed peroxide is titrated back. The method is more complicated than
that of Penot, and offers no special advantages.
* Z, angew. Chem., 1890, 3, 83.

BI,EACHING POWDER 507
Vanino" recommends the following apparatus (Fig. 144) for the rapid
approximate determination of available chlorine. It is very convenient,
and is adapted for use by workmen in bleaching works, etc. It is also
generally suitable for other tests in which an evolved gas is measured
when rapid approximate results are wanted.
The bleaching powder is weighed off on a hand balance, rubbed to a
cream with water, and introduced into the generating flask A ; the
solution of hydrogen peroxide is placed in c. The outlet tube, p,
of the gas-holder is best drawn out to a point, to reduce the tendency to
form air-bubbles. The gas-holder is filled with water, and before each
test the tube p is lowered so as to allow a few drops of water to
escape; this tube must remain filled with water, both at the beginning
and at the end of each experiment. A measuring cylinder is placed
under the outlet of the tube p. By opening the pinchcock at 6, the
hydrogen peroxide solution flows into the bleach solution and the
oxygen evolved displaces its own volume of water, which is collected in
the measuring cylinder. A certain time is allowed to elapse after each
test before the volume of water in the cylinder is read off. The results
are calculated from Vanino's table (p. 508), the figures in which are for
the moist gas.
Weight of I c.c. Chlorine in milligrams (Panino),
for barometric pressures from 700 to 770 mm. and for temperatures
from Io’ to 25° C. :—
| Value o
f (b – w)3.16696 |
760(1 +o.Oo366t)
A deduction of 1 mm. must be made from the barometric reading at
temperatures from Io’ to 12°, of 2 mm. at I 3° to 14°, and of 3 mm.
at 20° to 25°, so as to reduce the reading of the barometer to o'.
If, for example, I. Io g. of bleaching powder are taken for the test
and the volume of water collected, corresponding to the oxygen evolved,
measures 140 c.c. at IO and 720 mm. pressure, the calculation is
made as follows. From the table the factor corresponding to Io’ and
720 mm. is 2.858; this multiplied by O. I40 the volume, expressed in
litres, of the water collected gives O-4OO as the weight in g. of the
chlorine contained in the I. Io g. of bleaching powder, which, therefore,
tests 36.36 per cent.
* Z. angew. Chem., 1890, 3, 509.
* This table is based on the value 3.16696 g. as representing the weight of I litre of chlorine,
whereas the correct figure according to the most recent determinations is 3.219. Consequently,
the whole of the values given in Vanino's table are 1.6 per cent, too low. Vanino has more
recently (Z. anal. Chem., 1902, 41, 539), extended the table to 30° and for the odd values for
the barometric pressure, within the above limits.
#

0//
891.
098. &
Zº8. &
988. &
/Z8. &
6I8. &
&I8.6
#08. Z
961. &
68/. Z
I8/..?.
f//.3,
99/.3
89/.3,
I9/. &
8; 1. &
98/.3
83/.3
03/.3,
ZIA. &
90/.3
A69.3
069. Z
389.3
f/9. &
A99. &
699. &
IQ9.3,
#79. 3,
989.3,
639. &
I39. &
9I9.3,
909. &
869. 3,
O69. Z
889. &
998.3
/98. &
678. 3,
Žiž8.7,
f88. &
/38.6
6 IS. Z.
II8. &
f08. &
96/.3
88/.3
08/.3
8//-3
Q9/.3
89/.6
09/. &
37/.3
iſ 8/.3,
A31-3
6IA. &
&I/.3
#01. &
969. &
689. &
I89.3,
3.19.3
Q99.3,
899.3
099.3
379.8%
Q89. &
A39.3,
6T9. &
ZI9. Z,
f/09.3
/69.6
618. &
3/8. &
fºS. &
998.3
678. &
If,8. &
998.3
938.3,
8T8. &
OI8. Z,
308. &
g6/.3,
A8/.3
08/.3,
3//- 6
j9/.3.
99/. &
67/.3
If/. &
88/.3
93/.3
8IZ. &
0I/.3
801. &
969. &
A89.3
619.3,
Z/9.3
799.3
999. &
679.3
Ij/9.3
889. &
939.3,
8I9.3
0I9. &
f68.3,
988. 2,
8/8.3,
I/8.3
398. &
'998. &
878. &
078. 3,
388. &
#&S. &
AI8. &
60S. Z.
I08. &
f{}/.3
98/.3
8//-6
0//-3,
89/.3.
99/.3
17/.3
07/.. 3
38/.3
jø/.3
9I/.3,
60/. &
I0/.3
869.7%
989. 3,
8/9.3
0.19.3
399.6,
999. &
/79.3
689.3,
Z39.8%
j69. 3,
806. &
006. 3,
868. J.
Q88. &
A/8.6
698. &
398.3
798. &
978. &
838. 3,
I98. &
838.3,
GIS. &
808. &
008. &
361.6
j/8/.3
A/1-6
69/.3,
I9/.3
79/.3
97/.3
88/.3
08/. &
83/.. 3
9IA. &
A.0/-6
669. &
T69. &
789. &
9/9.3
899.3
099.3
899.3
gi/9. &
/89.3
Z36.3
fT6.3
A.06.3
668.3,
I68. &
j'88. &
9/8. &
898.3
098. &
398.3
gf8. &
/88. &
638. 3,
I38. 3.
#I8. &
908. &
86/.3
O6/.3
38/.3
9//-3
A9/-3
69/.3
Ig/. &
fº Z.Z.
98/.3
83/.3
03/.6
8TA. &
90/.3
/69. &
689. &
389.3
iſ 19.3
999. &
899. &
I99. &
I96.3
gift 6.3
986. &
A 36-3
6 I6. 3,
II6. &
#06. Z
968. &
888. &
088.6
3.18.3
fºg S.Z.
A.Q8.3,
678.3,
lf’8. &
838.3,
Q38. &
ZI8. &
608. Z,
308.3,
76.1. &
98/.3
6//-6
T//-3
39/.3.
99/.3
Af/.3
68/.3
38/.3
j6/.3
9I/.6
801.3
00/.3
369.3
fºg. &
A/9.6
Q96. &
Z96.3
676. &
If 6.3,
386. &
936. &
MI6. &
606. 3,
I06. &
#68. Z
988. 2,
818.6
018.3
398.8%
998. &
/78.3
688. &
I88. &
838.3
g|I8.6
/08. &
66/.3,
36/.3
j/8/.3
9//-3,
89/.3
09/.3
89/.3,
gy/. &
/8/.3
63/.6
I3/.3
8I/- (,
90/.3.
869. &
069.8%
816. &
0/6.3
396.3
j/96.2,
/76.3
686.6
I96.3
836.3
g|I6. &
/06.3
668.3
I68.3
jºb.8. &
9/8. &
898. &
098.3
398.3
fj/8. &
988. &
8&S. &
T38. &
9I8. &
908. &
161-6
68/.3
I8A. &
8//.3,
99/.3
89/.3,
09/.3
37/.3
j9/.3
93/.3
8IA. &
OI/.3
3.0/.3
I66. &
786. K.
9/6.3
896. &
096.2,
396.3
##6. Z
936. &
636.6
T36. &
8I6.3
906. &
/68.3
688. &
I88.3,
3.18.6
Q98.3
A98.3
678. &
Tj9.3
j/98.3
9%8. &
8I8.3,
OI8.3
308. &
#6/.3
98/.3
8//-6
0//-3
891.6
99/.3
Aſ A. &
68/.3,
I84.3
83/.3
9I/.3
Q00.8
A,66. &
686. 3,
I86.3
3/6.6
Q96.3
A.g6.3.
096.3
376.3
f'86. &
936. &
8I6.3
OT6. &
306.2,
#68.3
988. &
8/8.6
048. 3,
398. &
j98.3
/78.3
688. &
I38.3,
838. 3,
QI8. &
/08. 3,
66/.3
I6/.3
88/. &
9//-3
A9/.3
09/.3
39/.6
#7/.3
98/.3
83/.3,
6 Iſ). 3
TIO.8
300.8
Q66.3,
/86. &
6/6.6
I/6. &
996. &
996.3
Af6. &
696.3
I86. Z,
836.3
QI6.3,
A.06.3
668. &
I68. &
888. &
9/8. &
898. K.
098.3
39S. &
#78. &
998.3
838.3,
038. &
&I8.3,
j/08. &
96/.3
88/.3
08/. &
3//-3
#91. &
99/.3
87/.3
07/.3
880.8
fº,0.8
9IO.9
800.8
000.9
366.8%
f'86.3
9/6.2,
S96.2,
096.3
396.2,
776. &
986. Z,
836.3
036.3
&T6.3
#06.3
968. &
888.7%
088. &
3/8.6
j98.3
998. &
878. &
078. 3,
388. &
768. &
9I8. Z,
808. &
I08. &
86/.3
98/.3
A//-6
69/.3
I9/.3
391.6
970.8
180.8
630.8
I&0.8
9IO.9
900.8
/66.3
686. &
I86. &
8/6.6
996. &
/96.3
676. Z,
I?6. &
886. Z,
936. &
MI6.3
606.3
I06. &
868. &
Q88. &
118.6
698. &
I98.3
898. &
Q78. &
/88.6
638. &
I38. &
8T8.3,
Q08. &
A61. &
68/.3
I8/.3
8//-6
991.6
690.8
IQ0.8
370.9
Q80.8
930.8
8I0.8
0I0.8
Z00. 9
#66.3,
986. &
8/6.6
0/6.7%
396.3
j/96.3
976. 3,
886. &
096.3
336. &
7I6.3
906. &
868. &
068. &
Z88. &
f/8. &
998. 3,
898.3
098. &
Žiž8.3
#88. Z
9%8. &
AI8. &
608. &
I08. Z.
86/.3
98/.3
111.6
011,
89/.
99/.
fº/
Z9/,
09/
89/.
99/
#9/
39/
09/
87/.
97/.
##/.
ZW/
Of /
88/.
98/.
78/.
38/.
08/.
83/
93/.
#3/.
3.3/
03/
8I./
9I/
#I/
&I/
0I/
80/
90/
j/0/.
30/.
00/
#
i
|
o96
of 3
o66
o66
oló
606
o6.I
oSI
ell
o9'I
cQ I
of I
c3 I
cGI
ol. I
•0i
5
i
ſ
BLEACHING SOLUTIONS 509
Robert and Roncali i determine hypochlorites by measuring the
nitrogen evolved by decomposition with hydrazine sulphate.
The carbon dioxide present in bleaching powder is occasionally
determined. This may be done by decomposing the bleach with hydro-
chloric acid and passing the liberated chlorine and carbon dioxide into
an ammoniacal solution of calcium chloride (one part CaCl, .6H2O, six
parts water, and ten parts aqueous ammonia, sp. gr. O'96, mixed and
allowed to stand for Some time in a similar manner to that described
on p. 430 for liberating the sulphuretted hydrogen from sulphides). The
chlorine reacts with the ammonia according to the equation:—
8NH3+3Cl2=6NH,Cl + Nº.
The reaction may be hastened by boiling the solution. The carbon
dioxide is converted into calcium carbonate, which is filtered off and
estimated in the usual manner.
The method devised by Lunge and Rittener (p. 514) is both more
convenient and more accurate.
II. BLEACHING SOLUTIONS AND ELECTROLYTIC LIQUORS
Bleaching solutions consist essentially of mixtures of hypochlorites
and chlorides, and in many cases also contain free hypochlorous acid.
The base may be calcium, potassium, sodium (in Eau de Javel and
“chloros”), or magnesium, zinc, etc. Whether such solutions be pre-
pared by the double decomposition of bleaching powder with various
salts, or by passing chlorine into milk of lime, sodium carbonate, etc., or
by the electrolysis of chloride solutions, the constituents are hypochlorite,
chloride, free hypochlorous acid, and chlorate, with bases such as the
alkalis or calcium ; the solutions may also contain carbonate in addition
to caustic alkali. -
The same constituents are present in the liquors which result from
the electrolytic manufacture of caustic alkali or of chlorate, but in this
case the ratio between the various substances is very different from that
existing in bleaching solutions. The same methods of analysis are,
however, applicable in both instances.
James and Richey” have given a useful summary of the various
methods of analysis adopted in electrolytic works.
Available Chlorine.—In the case of bleach solutions the determina-
tion of the available chlorine is in most instances sufficient; this is
carried out as described under bleaching powder, p. 499. Such available
chlorine may be due either to hypochlorites or to the free hypochlorous
acid. Since the bleaching value is very different according to whether
the hypochlorous acid is free or combined, a point on which the excep-
tionally great bleaching value of many liquors, especially electrolytic
* Chem. Centr., 1904, I., 1294. * J. Amer. Chem. Soc., 1902, 24, 469.
510 THE CHLORINE INDUSTRY
liquors, depends, it may be necessary to determine the quantity of free
hypochlorous acid accompanying the hypochlorite salts, in solution.
Insoluble substances or material remaining undissolved, such as zinc
oxide, calcium hydroxide, etc., may be taken to be without action on
hypochlorous acid. Thus (should the liquor be found to contain free
base in solution) free hypochlorous acid may be regarded as absent; the
latter will only occur when all the basic material present has been
Saturated by acid radicals, such as hypochlorous acid, chloric acid, hydro-
chloric acid, carbonic acid, sulphuric acid, etc. The various salts will,
of course, be more or less hydrolytically or electrolytically dissociated
according to the degree of concentration of the solution, but such dis-
sociation is without influence from the analytical standpoint.
Free Hypochlorous acid.—From the above considerations conclu-
sions can be drawn regarding the presence of free hypochlorous acid by
examining the solution, after filtering off from insoluble basic substances,
in the first place for total bases, and in the second place for chlorine
present as chloride (cf. infra), carbonic acid, sulphuric acid, and if neces-
sary for such other acids as may be present, and further for “available
chlorine,” that is, for chlorine present as hypochlorite. If calculation
should show that the quantity of base, the Na,C), for instance, in the
case of a soda solution, is insufficient to neutralise the whole of the acids
found to form sodium chloride, sulphate, carbonate, and hypochlorite, it
may be concluded that free hypochlorous acid is present.
A much simpler method depends upon the fact that when a salt of
hypochlorous acid reacts with potassium iodide, two molecules of the
base are liberated, whilst in the case of free hypochlorous acid, only one
molecule of the base is set free :—
NaOCl 4-2KI+ H2O = NaCl-F 2 KOH + I.
HOC14-2KI = KCl + KOH +I,.
Consequently, if the liberated iodine be removed by titrating the
solution with thiosulphate, the two products can then be distinguished
by titrating with standard acid. Any alkali carbonate present in the
original solution must, of course, be allowed for (cf. p. 516).
The separation of free hypochlorous acid by distillation, with or with-
out the addition of other acids, is too uncertain to be of value.
Lunge" has described a simple method for distinguishing between
free chlorine and hypochlorous acid, based on their reaction with a
neutral Solution of potassium iodide ; hypochlorous acid forms an alkali
hydroxide, whilst in the case of chlorine the solution remains neutral, as
shown in the following equations:—
1. ClOH + 2 KI-KCl + KOH + I.
2. Cl2 + 2 KI = 2KCl + Ig:
* Chem. Ind., 1881, 4, 293.
BLEACHING SOLUTIONS 511
In the first case the liberated iodine reacts with the potassium
hydroxide to produce various compounds, especially potassium iodate ;
this may, however, be prevented by adding, at the start, the requisite
quantity of hydrochloric acid to convert the free alkali into potassium
chloride. If the quantity of acid added be known, it is then only neces-
sary to titrate with standard alkali. The reactions in acid solution
a.1 C : —
Ia. CIOH + 2KI+ HC1 = 2KC14-I, + H2O.
2a. Cl, + 2KI+ HC1 = 2KCl4. I, +HCI.
That is, in the case of hypochlorous acid one molecule of hydrochloric
acid is neutralised for each molecule of hypochlorous acid reacting with
potassium iodide, whilst in the case of chlorine the hydrochloric acid
added remains unchanged. If then, the solution be first titrated with
M/IO thiosulphate solution, and then with M/IO sodium hydroxide solu-
tion, in Ia the volume of thiosulphate solution used will be equal to
twice the difference between the number of c.c. of W/Io alkali necessary
to neutralise the full quantity of acid originally added and that now
required to neutralise the excess of free acid. In 2a, no such difference
will occur. For intermediate cases it is easy to calculate the ratio
between the hypochlorous acid and chlorine, since for each I c.c. M/IO
hydrochloric acid neutralised, 2 c.c. AV/IO thiosulphate must be allotted
to hypochlorite.
This method is difficult of application in the presence of free alkali
or alkaline carbonate, but in such cases free chlorine is hardly likely to
be present.
Chlorates may be estimated by first determining the available
chlorine according to one or other of the methods described on p. 500,
and then determining this and chlorate-chlorine together, by boiling
with ferrous sulphate and titrating back the excess of the latter with
permanganate, as described on p. 497. The difference between the two
figures so obtained gives the chlorine present as chlorate. The chlorine
present as hypochlorite and chlorate may also be determined iodometri-
cally, by boiling with strong hydrochloric acid and collecting the evolved
gas in potassium iodide solution (cf. p. 5O2).
According to Winteler," the latter method is very inexact, but results
accurate to within O. I to O.2 per cent. are obtainable by the ferrous
sulphate method, provided the solution containing the hypochlorite and
chlorate be added to the measured excess of ferrous sulphate solution
and the mixture be allowed to stand for several minutes before heating,
air of course being excluded.
The direct method due to Fresenius” is preferable to the above, but
it is somewhat more troublesome. The solution to be tested is treated
with an excess of neutral lead acetate solution; a precipitate is formed
* Z, angew. Chem., 1903, 16, 33. * Ibid., 1895, 8, 501.
512 THE CHLORINE INDUSTRY
which gradually turns brown with liberation of chlorine, and which con-
tains a quantity of lead peroxide corresponding to the hypochlorite
present:—
Ca(OCl), + PbCl, - PbO2 + CaCl2 + Clg.
The solution is allowed to stand for eight to ten hours, with frequent
agitation, until all smell of chlorine has disappeared. The chlorine
liberated reacts with the lead acetate solution to form lead chloride, lead
peroxide, and free acetic acid. The precipitate, consisting of the peroxide
and chloride of lead, is filtered off and washed, the filtrate and washings
concentrated by evaporation, and the contained lead and calcium pre-
cipitated by addition of sodium carbonate in slight excess; the filtrate,
which then contains the chlorate as the sodium salt, is evaporated
and its chlorate content determined (p 497.).
For-liquors which contain a large amount of hypochlorite together
with chlorate, as is the case with electrolytic and other bleaching
solutions, Ditz and Knöpflmacher"
recommended estimating the chlorate
iodometrically by decomposition at
the ordinary temperature with strong
hydrochloric acid and potassium brom-
ide. The process has recently been
modified by Ditz” as follows; the
method is also suitable for the exami-
nation of chlorate liquors. The appar-
atus employed is shown in Fig. I45.
The bottle a has a capacity of about
I 500 c.c., and may be closed either by
the stopper 6, or by the headpiece c, as
shown in the figure. The latter is
provided with a dropping funnel, d, and
with an absorption tube, e.
The volume of chlorate solution
used for a determination should be such
as to contain approximately O. I g. Of
chlorate, calculated as KCIO, Should it be necessary to estimate the
chlorate in a mixture containing both chlorate and hypochlorite, the
volume taken for analysis should be so regulated that the total con-
sumption of W/Io sodium thiosulphate solution will be between 40 to
50 c.c. The measured volume of chlorate liquor, together with IO C.C.
of a Io per cent. potassium bromide solution, are introduced into the
bottle, and after the absorption tube has been filled to two-thirds of
its height with a 5 per cent. solution of potassium iodide, concentrated
hydrochloric acid is added from the dropping funnel d. If the volume
* Z, angew. Chem., 1899, I2, II95. * Chem. Zeil., I90I, 25, 727.
FIG. 145.

BLEACHING SOLUTIONS 513
of the chlorate solution does not exceed 25 c.c., 50 c.c. of acid are
added, and proportionately more if more than 25 c.c. are required for
the analysis. The liberated bromine is absorbed by the potassium
iodide in e ; to prevent the solution being sucked back into a, the
stopper on the exit-tube of e is inserted. After the whole has been
allowed to stand for five minutes, from 500 to 6oo c.c. of distilled water
are introduced through the dropping funnel, followed at a moderate rate
by about 20 c.c. of the 5 per cent. potassium iodide solution. If more than
50 c.c. of hydrochloric acid have been used for the decomposition, pro-
portionately more water must be added. The mixture is well agitated,
the contents of the absorption tube blown over into a, the tube together
with the headpiece thoroughly washed, the latter replaced by the glass
stopper, and the liberated iodine determined by titration with M/Io
thiosulphate solution. If I c.c. of the thiosulphate solution equals a g.
iodine and the volume required for the titration of the separated iodine
be b c.c., the weight of chlorate present is given by the equation :-
I22.60
76.1.8
g KClOs = a x 5 x = o. I609 a × 5,
and the quantity of potassium chloride equivalent to this by the
equation:—
g KCl = a x b x #. = O.O979 a × 5.
When examining mixtures containing both hypochlorite and chlorate,
the quantity of hypochlorite present must be previously determined and
allowed for.
Kolb and Davidson also carry out the determination at the
ordinary temperature. They dissolve about oo8 g. of the chlorate in
Io c.c. of water, free from air, in a 500 c.c. flask, heat the solution on
the water-bath, and allow it to cool in a current of carbon dioxide ;
2 g. of potassium iodide and 50 c.c. of air-free hydrochloric acid of sp.
gr. I. I2 are then added, the whole allowed to stand for an hour,
protected from the light, diluted with boiled water saturated with
carbon dioxide to 300 c.c., and the liberated iodine finally titrated with
Sodium thiosulphate.
M. Scholtz” reduces the chlorate to chloride with nitric acid and
sodium nitrite, adds a measured volume of W/Io silver nitrate solution,
and titrates back the excess with ammonium thiocyanate, using ferric
ammonium alum as indicator.
Hendrixon * reduces the chlorate with metallic iron, and titrates
the resulting chloride, after oxidising the iron with nitric acid.
* Z. angew. Chem., 1904, 17, 1883; 1905, 18, Ioa'7, 1693. Cf. also, Ditz, ibid., 1905, 18, 1518.
* Arch. Pharm., 1905, 243, 353 ; J. Soc, Chem. Ind., 1905, 24, 904.
* A mer. Chem. J., 1904, 32, 242.
2 K
514 THE CHLORINE INDUSTRY
Chloride-chlorine.—The chlorine present as chloride is most con-
veniently estimated in the following manner. The available chlorine is
first determined by Penot's method (p. 503), whereby it is completely
converted to chloride, whilst the arsenite is oxidised to sodium arsenate.
The latter compound is an exceptionally good indicator for the silver
nitrate titration ; in fact, it is so much superior to potassium chromate
that it is unnecessary to make any deduction for the excess of silver
nitrate necessary to produce the coloration. It is therefore only
necessary to nearly neutralise the free alkali by cautious addition of
nitric acid (a slight excess of alkali is not harmful, excess of free acid is)
and then to titrate with silver nitrate solution, with good agitation, as
described on p. 123, until the precipitate becomes red owing to the
formation of a small quantity of silver arsenate. If preferred, the
chloride may, of course, be determined gravimetrically. On deducting
the chlorine corresponding to the hypochlorite found, from the total
chlorine as determined above, the quantity of chlorine present as chloride
in the original solution is obtained.
Carbonic acid.—In estimating carbonic acid, the details given in
describing the analysis of Deacon gases (p. 491) and of bleaching powder
(p. 509) apply. Naturally, appreciable quantities of carbonic acid cannot
exist in bleach solutions containing calcium oxide or magnesia; they
may, however, be present in potash or soda bleaching solutions and in
electrolytic potash and soda liquors. This carbonic acid may be
estimated by first converting the hypochlorite to chloride by boiling
with ammonia solution free from carbonate, allowing the boiled solution
to cool under exclusion of carbon dioxide, and then determining the
total alkalinity by titrating a portion of the solution with methyl orange
as indicator. The alkali hydroxide is estimated in a second portion of
the solution after addition of barium chloride, as described on pp. 72 and
424. The difference between the two titrations gives the amount of
carbonate present. The carbonic acid may also be determined directly,
after decomposing the hypochlorite in the manner described above, by
liberating it with a strong acid and estimating it either gravimetrically or
gas-volumetrically (p. 149). Blattner's method (cf. infra) can also be used.
The estimation of carbon dioxide in presence of chlorine can also
be carried out by the simplified Lunge-Marchlewski method devised
by Lunge and Rittener" (cf. p. I 53).
The apparatus employed is shown in Fig. 146. The decomposition
flask B has a capacity of about 30 c.c.; it is fitted with a tap-funnel
C and capillary exit-tube D, which is cut off flush with the bottom
of the rubber stopper in B. D is connected with the exit-tube of the
Bunté burette A, as shown ; the ungraduated portion of the burette
must be calibrated.
' Z, angew. Chem., 1906, 19, 1849.
BLEACHING SOLUTIONS 515
To carry out a determination, sufficient substance or solution is
introduced into B to evolve not more than 60 c.c. of gas; D is then
connected with the burette, the taps E and F opened, and the whole
evacuated by attaching an ordinary pump to F. This tap is then
closed and connected with the pressure-bottle G, charged with a
saturated brine solution, which is used as the confining liquid in place
of mercury. A little of the brine solution is allowed to enter the
capillary portion of the bottom of the nitrometer, as a check on any
possible leakage, and the tap F again closed.
\–/ =l to
C/71
F
Fig. 146.
:
|
The substance in B is then decomposed by adding hydrochloric
acid, drop by drop, from the tap-funnel C; when the evolution of gas
slackens the contents of B are heated, and the complete evolution of
the liberated gases facilitated by the addition of from 2 to 3 c.c. of
hydrogen peroxide; on again heating, the evolved oxygen drives
out all the dissolved gas. The burette is then disconnected, allowed
to stand for about twenty minutes and the volume of gas read, after
adjusting the pressure by means of the bottle G. After taking the
reading, a measured excess of an M/Io sodium arsenite solution is
added, through the burette funnel, to absorb the chlorine, and any

516 THE CHLORINE INDUSTRY
portion adhering to the funnel washed in with a little water; this
solution floats on the heavier solution of brine. The pressure-bottle
G is then detached and the carbon dioxide absorbed by sodium
hydroxide solution introduced through E ; as all the chlorine has
been previously removed, no chlorate can be formed during this
absorption. When the absorption is complete the residual volume
of gas is read, which, deducted from the original reading, gives the
volume of carbon dioxide and chlorine, after correcting each reading for
temperature and pressure; the vapour pressure of the brine solution
may be taken as 80 per cent. of that of pure water.
To determine the chlorine, the contents of the burette are washed
out into a beaker, acidulated with hydrochloric acid, an excess of
sodium bicarbonate added, and the unchanged arsenite titrated with
M/Io iodine solution. The volume of carbon dioxide collected is then
calculated by difference.
Bases.—As regards the bases, should only one be present, as is
usually the case, the total quantity may be estimated by conversion to
sulphate by evaporating with sulphuric acid. Calcium and magnesium
may also be determined as oxalate and pyrophosphate respectively,
after first decomposing the hypochlorite by boiling with a strong
acid.
Alkali hydroxide and carbonate.—Blattner 1 describes the following
neat method for determining these constituents in Eau de Javel.
It depends upon the fact that phenolphthalein preserves its red
colour in hypochlorite solution so long as alkali hydroxide is present, but
that the colour is destroyed by free chlorine as soon as all the hydroxide
has been removed, and does not return on the further addition of free
alkali. Ten c.c. of the bleach liquor are diluted with 150 c.c. of
previously boiled and distilled water, a few drops of a I per cent, alcoholic
solution of phenolphthalein added, and the solution then titrated with
normal acid, continuous agitation being maintained during the addition
of the acid. The whole of the alkali hydroxide is neutralised when the
coloration produced by the addition of a fresh drop of phenolphthalein
solution disappears on shaking the solution for five seconds. Each I
c.c. of normal acid used corresponds to O-O3 IO5 g. Na,C) or O.O4OO6 g.
NaOH. The total alkalinity, NaOH + Na,COs, is determined in a second
portion of the liquor by boiling with ammonia until the hypochlorite
has been decomposed and all unchanged ammonia driven off, and then
titrating with acid in the usual manner. The above methods are
especially suitable for the examination of caustic potash or caustic soda
solutions produced electrolytically, which usually contain hypochlorite.
The following is a very simple method of determining the free
alkali. A small quantity of chemically pure, neutral hydrogen peroxide
* Lunge, Sulphuric Acid and Alkali, vol. iii., p. 439.
TOTASSIUM CHLORATE 517
is added to the solution to decompose the hypochlorite, according
to the equation NaOCl.--H2O, - NaCl-HO,-- H.O, and the sodium
hydroxide and carbonate present are then titrated in the usual manner
(cf. pp. 72 and 424). -
H. v. Huber' has described a method for the examination of
electrolytic liquors which contain chromate. Phenolphthalein is, in this
case, unsuitable as an indicator for the free alkali present; methyl
Orange, however, gives a perfectly sharp colour change at the point
where potassium chromate is converted into the bichromate. If, there-
fore, the chromate content is known, it is only necessary to deduct from
the total acid required for titration the quantity necessary to convert
the chromate into bichromate. If the quantity of chromate present is
not known, the chromate is precipitated by addition of barium chloride,
and the solution then titrated with acid. Where the quantity of
chromate present is small, it is not necessary to filter the solution before
titrating. Any hypochlorite or free hypochlorous acid in the liquor
must first be removed; this may be done without affecting the alkalinity,
by addition of neutral sodium sulphite or of sodium thiosulphate. If
Sodium sulphite be used, an excess must be avoided, since with methyl
Orange the colour change only occurs when such excess has been con-
verted to the acid sulphite which involves the consumption of the
corresponding quantity of acid (cf. p. 65).
III. POTASSIUM CHLORATE
The potassium chlorate of commerce is generally almost chemically
pure, and contains only a very small trace of chloride, the quantity of
which should not exceed o.o.5 per cent. For this reason it is necessary
to take a somewhat large quantity of chlorate, say 50 g., for the
determination of the chloride. The chlorate is dissolved in distilled
water, and the chloride precipitated by addition of silver nitrate solu-
tion. The distilled water employed must be perfectly free from
chlorides. The chlorate should dissolve without residue, and the solu-
tion should not be coloured by organic matter. Ammonium sulphide
should not produce any coloration, indicating the absence of iron,
manganese, or lead. Garnier” found arsenic in potassium chlorate.
Potassium nitrate only occurs in potassium chlorate as an
adulterant. According to the German Pharmacopoeia, its presence is
detected by fusing the chlorate and examining the melt for alkalinity;
authorities are, however, agreed that this method is misleading, since even
pure potassium chlorate gives an alkaline reaction after fusion. The
following reactions are more satisfactory.”
* Zeit. Flectrochem., 1901, 7, 396. * Fischer's Jahresher, 1885, p. 260.
* Krauch, The 7esting of Chemical Reagents for Purity, p. 227.
518 THE CHLORINE INDUSTRY
One g. of the salt is heated with 5 c.c. of sodium hydroxide solution;
should ammonia be evolved, indicating the presence of ammonium salts,
the solution is boiled until all ammonia has been expelled. The
Solution is then cooled, O-5 g. of zinc filings and iron powder added, and
the whole again heated. Any ammonia then liberated is due to the
presence of nitrate. According to Scholvien, the potassium chlorate is
heated until the residual potassium chloride has also been brought to
fusion. The melt is dissolved in water, and the solution treated with
dilute sulphuric acid followed by zinc iodide-starch solution. The
coloration produced should not be sufficient to render the solution
opaque. A faint blue coloration is permissible, since with o oſ per cent.
of nitrate the colour produced is so intense that the solution is quite
opaque.
A quantitative determination of the chloric acid in commercial
chlorate is scarcely ever undertaken, owing to the small amount of
impurity present; it may, however, be carried out as described on p. 497.
This determination would be necessary for the indirect estimation of
Sodium and potassium chlorate in a mixture of the two salts; so far,
such mixtures do not, however, appear to occur in practice.
IV. LIQUID CHLORINE
The following physical properties of liquid chlorine are given by
A. Lange.”
Specific gravity at O’, I-469; I 5°, I-426; 30°, 1.381.
Vapour pressure at O’, 3.7; 15°, 5-8; 30°, 8.7 atmos.
One kilo corresponds to a volume of 316 litres measured at o' and
76O mm. pressure.
Critical temperature, I46°.
Critical pressure, 93-5 atmos.
Boiling-point at 760 mm., -33°.6.
Melting-point of the solidified gas, - IO2°.
Conditions of carriage on the German railways:—The capacity of
the containing vessel must be at the rate of O.9 litre for each I kilo
of chlorine; the cylinders must be officially tested up to a pressure
of 50 atmospheres; the pressure test must be repeated annually.
* Apoth. Zeit., 1887, 408.
* 7aschenbuch. d. Berliner Bezirksvereines Deutscher Chemiker, 1898-99, p. 82.
*
LIQUID CHLORINE
519
Pressure and Specific Gravity of Liquid Chlorine.
R. Knietsch."
tºº. PreSSure. Specific Gravity. *:::::::::::
--88 37:5 mm. Hg
85 45°0 , , e tº e
80 62°5 , , 1 *6602 |\
75 88-0 , , 1 *6490
70 118 9 9 1°6382
65 159 3 y 1-6273
60 210 9 * , 1-6167
55 275 9 9 1-6055 >0-001409
50 350 9 3 1°5945
45 445 9 3 1°5830
40 560 9 3 1-5720
35 705 9 3 1°5589 -
33-6 760 9 1:5575 J
30 1°20 Atm 1°5485 |\
25 1°50 , , 1°5358
20 1°84 , , 1°5230
15 2°23 , , 1°5100 -0-001793
10 2°63 , , 1°4965
5 3°14 , 1°4830 .
0 ; : 1°4690
+ 5 4°25 , , 1°4548 |\ a.
10 ; : # }o 00.1978
15 5°75 , , 1°4273 º
20 ; : 1°4118 }o 002030
25 *63 , , 1°3984 •ſyſ).9)
30 §§ 9 9 1°3815 }o 002190
35 9°95 , , 1°3683 tº
40 #}} . ; }0.00200
50 14*70 , , 1°3170 ©
; 18°60 , , 1 °2830 }0 002690
0 23:00 , , 1-2430 { }
80 28°40 , , 1°2000 }o 003460
90 34°50 , , s e e
100 41°70 ,,
110 50-80 ,
120 60°40 , ,
130 71°60 , , © tº gº
146 93°50 , Critical point.
* Annalem, 1890, 259, Ioo.
LITERATURE
LUNGE, G.-The Manufacture of Sulphuric Acid and Alkali, vol. iii., 2nd edition,
1896.
LUNGE, G., and BERL, E.-The Technical Chemists’ AHandbook, 1908.
END OF PART I.
PRINTED BY
{) LIVER AND BOY D
EDINBURGH
STANDARD TEXT Books
PUBLISHED BY
D. VAN NOSTRAND COMPANY,
NEW YORK.
ABBOTT, A. W. The Electrical Transmission of Energy. A Manual
for the Design of Electrical Circuits. Fourth Edition, entirely rewritten
and enlarged. Fully illustrated. 8vo, cloth. . . . . . . . . . . . . . . . . net, $5.00
ASHE, S. W., and KEILEY, J. D. Electric Railways, Theoretically
and Practically Treated. Vol. I, Rolling Stock. 12mo, cloth. 290 pp.,
172 illustrations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . net, $2.50
Electric Railways. Vol. II. The Sub-Station and the Dis-
tributing System. 12mo, cloth. Illustrated. . . . . . . . . . . . . . . . . In press.
ATKINSON, Prof. A. A. (Ohio University.) Electrical and Magnetic Cal-
culations, for the use of Electrical Engineers and Artisans, Teachers,
Students, and all others interested in the Theory and Application of Elec-
tricity and Magnetism. Second edition, revised. 8vo, cloth. Illus-
trated. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . net, $1.50
ATKINSON, PHILIP. The Elements of Electric Lighting, includin
Electric Generation, Measurement, Storage, and Distribution. Tent
edition. Illustrated. 12mo, cloth... . . . . . . . . . . . . . . . . . . . . . . . . . . . . $1.50
The Elements of Dynamic Electricity and Magnetism. Third
edition. 120 illustrations. 12mo, cloth. . . . . . . . . . . . . . . . . . . . . . . . . $2.00
Power Transmitted by Electricity, and its Application by the
Electric Motor, including Electric Railway Construction. Third edition,
fully revised, new matter added. Illustrated. 12mo, cloth. . . . . $2.00
AUCHINCLOSS, W. S. Link and Valve Motions Simplified. Illus-
trated with 29 wood-cuts, 20 lithographic plates, together with a Travel
Scale, and numerous useful tables. Fourteenth edition, revised. 8vo,
cloth. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . $2.00
BEAUMONT, ROBERT. Color in Woven Design. With 32 colored
plates and numerous original illustrations. Thick 12mo. . . . . . . . . . . . . $7.50
BERNTHSEN, A. A Text-Book of Organic Chemistry. Translated
by George McGowan, Ph.D. Fourth English edition. Revised and ex-
tended by the author and translator. Illustrated. 12mo, cloth. . . . $2.50
BIGGS, C. H. W. First Principles of Electricity and Magnetism.
A book for beginners in practical work, containing a good deal of useful
information not usually to be found in similar books. With numerous
tables and 343 diagrams and figures. 12mo, cloth, illustrated. . . . . . . $2.00
2 STANDARD TEXT BOOKS.
BOWIE, AUG. J., Jr., M.E. A Practical Treatise on Hydraulic
Mining in California. With Descriptions of the Use and Construction of
Ditches, Flumes, Wrought-iron Pipes and Dams; Flow of Water on Heav
Grades, and its Applicability, under High Pressure, to Mining. Nint
edition. Quarto, cloth. Illustrated. . . . . . . . . . . . . . . . . . . . . . . . . . . . . $5.00
BOWSER, Prof. E. A. An Elementary Treatise on Analytic Geometry.
Embracing Plane Geometry, and an Introduction to Geometry of Three
Dimensions. 12mo, cloth. Twenty-first edition. . . . . . . . . . . . . . . . . $1.75
An Elementary Treatise on the Differential and Integral Calculus.
With numerous examples. Twenty-first edition, enlarged by 640 addi-
tional examples. 12mo, cloth. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . $2.25
- An Elementary Treatise on Analytic Mechanics. With numerous
examples. 12mo, cloth. Eighteenth edition. . . . . . . . . . . . . . . . . . . . $3.00
An Elementary Treatise on Hydro-Mechanics. With numerous
examples. 12mo, cloth. Fifth edition..... . . . . . . . . . . . . . . . . . . . . . $2.50
A Treatise on Roofs and Bridges. With numerous Exercises. Espe-
cially adapted for school use. 12mo, cloth. Illustrated. . . . . . . . . net, $2.25
CATHCART, Prof. WM. L. Machine Design. Part I. Fastenings.
8vo, cloth. Illustrated. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . net, $3.00
CHRISTIE, W. WALLACE. Chimney Design and Theory. A book
for Engineers and Architects, with numerous half-tone illustrations and
plates of famous chimneys. Second edition, revised. 8vo, cloth. . $3.00
Boiler Waters. Scale, Corrosion, Foaming. . . . . . . . . . . . . . . . In press.
CROCKER, F. B. Electric Lighting. A Practical Exposition of the
Art, for Use of Engineers, Students, and others interested in the Installa-
tion or Operation of Electrical Plants. New edition, thoroughly revised
and fewritten. Vol. I. The Generating Plant. 8vo, cloth. Illus-
trated. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . $3.00
Vol. II. Distributing Systems and Lamps. Third edition. 8vo,
cloth. Illustrated. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . $3.00
— and WHEELER, S. S. The Practical Management of Dyna-
mos and Motors. New edition, thoroughly revised and rewritten,
with a special chapter by H. A. Foster. 12mo, cloth. Illustrated.
In press.
DORR, B. F. The Surveyor’s Guide and Pocket Table-Book. Fifth
edition. Thoroughly revised and greatly extended. With a second ap-
pendix up to date. 16mo, morocco, flaps. . . . . . . . . . . . . . . . . . . . . . . . . $2.00
DRAPER, C. H. Heat, and the Principles of Thermo-Dynamics.
With many illustrations and numerical examples. 12mo, cloth. . . . . . $1.50
ECCLES, Dr. R. G. Food Preservatives; their Advantages and
Proper Use. With an Introduction by E. W. Duckwall, M.S. 8vo, .
202 pp., cloth. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . $1.00
& 4 Paper. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .5
STANDARD TEXT BOOKS. 3
«»
ELIOT, Prof. C. W., and STORER, Prof. F. H. A. Compendious
Manual of Qualitative Chemical Analysis. Revised with the co-operation
of the authors, by Prof. William R. Nichols. Illustrated. Twenty-first
sºon, newly revised by Prof. W. B. Lindsay and F. H. Storer. 12mo,
cloth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . $1.50
EVERETT, J. D. Elementary Text-Book of Physics. Illustrated.
Seventh edition. 12mo, cloth. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . $1.40
FOSTER, H. A. Electrical Engineers’ Pocket-Book. With the
Collaboration of Eminent Specialists. A handbook of useful data for
Electricians and Electrical Engineers. With innumerable tables, diagrams,
and figures. Third edition, revised. Pocket size, full leather, 1000 #.
5.00
FOX, WM., and THOMAS, C. W., M.E. ...A Practical Course in Mechan-
ical Drawing. With plates. Second edition, revised. 12mo, cloth. $1.25
GILLMORE, Gen. Q. A. Practical Treatise on the Construction of
Roads, Streets, and Pavements. Tenth edition. With 70 illustrations.
12mo, cloth. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . $2.00
GOODEVE, T. M. A Text-Book on the Steam-Engine. With a
Supplement on Gas-Engines. Twelfth edition, enlarged. 143 illustra-
tions. 12mo, cloth. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . $2.00
HAEDER, HERMAN, C. E. A. Handbook on the Steam-Engine.
With especial reference to small and medium sized engines. Third
English edition, re-edited by the author from the second German edi-
tion, and translated with considerable additions and alterations by H. H.
P. Powles. 12mo, cloth. Nearly 1100 illustrations. . . . . . . . . . . . . . . . $3.00
HALL, WM. S., Prof. Elements of the Differential and Integral
Calculus. Sixth edition, revised. 8vo, cloth. Illustrated. .net, $2.25
Descriptive Geometry; With Numerous Problems and Practical
Applications. Comprising an 8vo volume of text and a 4to Atlas of illus-
trated problems. Two vols, cloth. . . . . . . . . . . . . . . . . . . . . . . . . . . . net, $3.50
HALSEY, F. A. Slide-Valve Gears: an Explanation of the Action and
Construction of Plain and Cut-off Slide-Valves. Illustrated. Eighth
edition, revised and enlarged. 12mo cloth. . . . . . . . . . . . . . . . . . . . $1.50
HANCOCK, HERBERT. Text-Book of Mechanics and Hydrostatics.
With over 500 diagrams. 8vo, cloth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . $1.75
HECK, Prof. R. C. H. The Steam-Engine. Vol. I. The Thermo-
dynamics and the Mechanics of the Engine. 8vo, cloth, 391 pp. Illustrated.
net, $3.50
Vol. II. Form, Construction, and Working of the Engine. The
Steam-Turbine. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . In press.
HERRMANN, GUSTAV. The Graphical Statics of Mechanism. A
Guide for the Use of Machinists, Architects, and Engineers; and also a
Text-Book for Technical Schools. Translated and annotated by A. P.
Smith, M.E. 7 folding plates. Fourth edition. 12mo, cloth. . . . . $2.00
HOPKINS, Prof. N. MONROE. Experimental Electrochemistry.
Theoretically and Experimentally Treated. 300 pp., 8vo. "...º.o
Inet, ºbj.
4 STANDARD TEXT BOOKS.
JAMIESON, ANDREW, C. E. A Text-Book on Steam and Steam-
Engines. Specially arranged for the Use of Science and Art, City and
Guilds of London Institute, and other Engineering Students. Thirteenth
edition, revised. Illustrated. 12mo, cloth. . . . . . . . . . . . . . . . . . . . . . $3.00
Elementary Manual on Steam, and the Steam-Engine, ... Spe-
cially arranged for the Use of First-Year Science and Art, City and Guilds
of London Institute, and other Elementary Engineering Students. Third
edition. 12mo, cloth. . . . . . . . . p > , , , , , , , , , , , , , , , , , , , , , , , , , , ~ * * * * $1.50
JANNETTAZ, EDWARD. A Guide to the Determination of Rocks:
being an Introduction to Lithology. Translated from the French by G.
W. Plympton, Professor of Physical Science at Brooklyn Polytechnic
Institute. 12mo, cloth. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . $1.50
JOHNSTON, Prof. J. F. W., and CAMERON, Sir CHARLES. Elements
: Agricultural Chemistry and Geology. Seventeenth edition. º
cloth. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . - - - - - - - - - - - - - º
RAPP, GISBERT, C. E. Electric Transmission of Energy, and its
Transformation, Subdivision, and Distribution. A practical handbook.
Fourth edition revised. 12mo, cloth. . . . . . . . . . . . . . . . . . . . . . . . . . . $3.50
REMP, JAMES FURMAN, A.B., E.M. A. Handbook of Rocks; for
use without the microscope. With a glossary of the names of rocks and
tº: lithological terms. Third edition, revised. 8vo, cloth. º
trated . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.50
RLEIN, J. F. Design of a High-speed Steam-engine. With notes
diagrams, formulas, and tables. Second edition, revised and en-
larged. 8vo, cloth. Illustrated. 257 pages. . . . . . . . . . . . . . . . . net, $5.00
LASSAR-COHN, Dr. An Introduction to Modern Scientific Chem-
istry, in the form of popular lectures suited to University Extension Students
and general readers. Translated from the author's corrected proofs for
the second German edition, by M. M. Pattison Muir, M.A. 12mo, cloth.
Illustrated. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . $2.00
LODGE, OLIVER J. Elementary Mechanics, including Hydrostatics
and Pneumatics. Revised edition. 12mo, cloth. . . . . . . . . . . . . . . . . $1.50
LUQUER, LEA. McILWAINE, Ph.D. Minerals in Rock Sections.
The Practical Method of Identifying Minerals in Rock Sections with the
Microscope. Especially arranged for Students in Technical and Scientific
Schools. New edition, revised. 8vo, cloth. Illustrated......net, $1.50
MILLER, E. H. Quantitative Analysis for Mining Engineers. 8vo,
cloth. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . net, $1.50
MINIFIE, WM. Mechanical Drawing. A Text-Book of Geometrical
Drawing for the use of Mechanics and Schools, in which the Definitions and
Rules of Geometry are familiarly explained; the Practical Problems are ar-
ranged from the most simple to the more complex, and in their description
technicalities are avoided as much as possible. With illustrations for Draw-
ing Plans, Sections, and Elevations of Railways and Machinery: an Introduc-
tion to Isometrical Drawing, and an Essay on Linear Perspective and
Shadows. Illustrated with over 200 diagrams engraved on steel. Tenth
thousand. With an appendix on the Theory and Application of Colors.
8vo, cloth. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . $4.00
STANDARD TEXT BOOKS. 5
MINIFIE, WM. Geometrical Drawing. Abridged from the Octavo
Edition, for the use of schools. Illustrated with 48 steel plates.
Ninth edition. 12mo, cloth. . . . . . . . . . . . . . . . . . . . . . ; : - - - - - - - - - - - $2.00
MOSES, ALFRED J., and PARSONS, C. L. Elements of Mineralogy,
Crystallography, and Blow-Pipe . Analysis, from a Practical Standpoint.
336 illustrations. New and enlarged edition. 8vo, cloth. . . . . . . $2.50
NASMITH, JOSEPH. The Student's Cotton Spinning. Third edition,
revised and enlarged. 8vo, cloth. Illustrated. . . . . . . . . . . . . . . . . . . . $3.00
NUGENT, E. Treatise on Optics; or, Light and Sight theoretically
and practically treated, with the application to Fine Art and Industrial
Pursuits. With 103 illustrations. 12mo, cloth. . . . . . . . . . . . . . . . . . . . $1.50
OLSEN, Prof. J. C. Text-Book of Quantitative Chemical Analysis
by Gravimetric, Electrolytic, Volumetric, and Gasometric Methods. With
seventy-two Laboratory Exercises giving the analysis of Pure Salts, Alloys,
Minerals, and Technical Products. 8vo, cloth. Illustrated. 513 pages.
net, $4.00
OUDIN, MAURICE A. Standard Polyphase Apparatus and Systems.
With many diagrams and figures. Third edition, thoroughly revised.
Fully illustrated. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . $3.00
PALAZ, A., Sc.D. A Treatise on Industrial Photometry, with special
application to Electric Lighting. Authorized translation from the French
by George W. Patterson, Jr., 8vo, cloth. Illustrated. . . . . . . . . . . . . . $4.00
PARSHALL. H. F., and HobART, H. M. Armature windings of
Electric, Machines. With 140 full-page plates, 65 tables, and 165 pages
of descriptive letter-press. 4to, cloth. . . . . . . . . . . . . . . . . . . . . . . . . . . . . $7.50
PATTON, HORACE B. Lecture Notes on Crystallography. Revised
edition, largely rewritten. 8vo, 47 pages. . . . . . . . . . . . . . . . . . . . . . $1.25
PAULDING, CHAS. P. Practical Laws and Data on Condensation
of Steam in Covered and Bare Pipes. 8vo, cloth. Illustrated. 102 pages.
net, $2.00
The Transmission of Heat through Cold-Storage Insulation.
Formulas, Principles, and Data relating to Insulation of every kind.
Manual for Refrigerating Engineers. 12mo, cloth. 41 pages. Illustrated.
net, $1.00
PERRINE, F. A. C., A.M., D.Sc. Conductors for Electrical Dis-
tribution; Their Manufacture and Materials, the Calculation of the Cir-
cuits, Pole Line Construction, Underground Working and other Uses.
With diagrams and engravings. 8vo, cloth. . . . . . . . . . . . . . . . . . . . net, $3.50
PERRY, JOHN. Applied Mechanics. A Treatise for the Use of
Students who have time to work experimental, numerical, and graphical
exercises illustrating the subject, 650 pages. 8vo, cloth. . . . . . . net, $2.50
PLATTNER. Manual of Qualitative and Quantitative Analysis with
the Blow-Pipe. From the last German edition, revised and enlarged,
by Prof. Th. Richter, of the Royal Saxon Mining Academy. Translated
by Prof. H. B Cornwall, assisted by John H. Caswell. Illustrated with
78 woodcuts. Eighth edition, revised. 463 pages. 8vo, cloth. net, $4.00
6 STANDARD TEXT BOOKS.
POPE, F. L., Modern Practice of the Electric Telegraph. A Tech-
nical Handbook for Electricians, Managers, and Operators. "Seventeenth
edition, rewritten and enlarged, and fully illustrated, 8vo, cloth.31.50
PRELINI, CHARLES. Tunneling. A Practical Treatise containing
149 Working Drawings and Figures. With additions by Charles S. Hill,
C.E., Associate Editor “Engineering News.” Second edition, revised.
8vo, cloth. Illustrated. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . $3.00
Earth and Rock Excavation. ... A Manual for Engineers, Contractors,
and Engineering Students. 8vo, cloth. Illustrated. 350 pp....net, $3.00
PRESCOTT, Prof. A. B. Organic Analysis. A Manual of the
Descriptive and Analytical Chemistry of Certain Carbon Compounds in
Common Use; a Guide in the Qualitative and Quantitative Analysis of
Organic Materials in Commercial and Pharmaceutical Assays, in the Esti-
mation of Impurities under Authorized Standards, and in Forensic Exami-
nations for Poisons, with Directions for Elementary Organic Analysis,
Fifth edition. 8vo, cloth. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . $5.00
Outlines of Proximate Organic Analysis, for the Identification,
Separation, and Quantitative Determination of the more commonly occur-
ring Organic Compounds. Fourth edition. 12mo, cloth... . . . . . . . $1.75
First Book in Qualitative Chemistry. Twelfth edition. 12mo,
cloth. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . net, $1.50
and OTIS COE JOHNSON. Qualitative Chemical Analysis. A
Guide in the Practical Study of Chemistry and in the Work of Analysis.
Fifth fully revised edition, entirely rewritten. With Descriptive
Chemistry extended throughout. 8vo, cloth. . . . . . . . . . . . . . . . . . . net, $3.50
RANKINE, W. J. MACQUORN, C.E., LL.D., F.R.S., Machinery and
Mill-work. Comprising the Geometry, Motions, Work, Strength, Con-
struction, and Objects of Machines, etc. Illustrated with nearly 300
* Seventh edition. Thoroughly revised by W. J. Millar. sº
cloth. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5.00
The Steam-Engine and Other Prime Movers. With diagram of
the Mechanical Properties of Steam. With folding plates, numerous
tables and illustrations. Fifteenth edition. Thoroughly revised by W. J.
Millar. 8vo, cloth. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . $5.00
IJseful Rules and Tables for Engineers and Others. With
appendix, tables, tests, and formulae for the use of Electrical Engineers.
Comprising Submarine Electrical Engineering, Electric Lighting, and
Transmission of Power. By Andrew Jamieson, C.E., F.R.S.E. Seventh
edition, thoroughly revised by W. J. Millar. 8vo, cloth. . . . . . . . . . $4.00
A Mechanical Text-Book. º Prof. Macquorn Rankine and E. E.
º C.E. With numerous illustrations. Fourth edition. sº
cloth. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.50
Applied Mechanics. Comprising the Principles of Statics and Cine-
matics, and Theory of Structures, Mechanics, and Machines. . With numerous
diagrams. Seventeenth edition. Thoroughly revised by W. J. Millar.
8vo, cloth. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . $5.00
Civil Engineering. Comprising Engineering, Surveys, Earthwork,
Foundations, Masonry, Carpentry, Metal-Work, Roads, Railways, Canals,
Rivers, Water-Works, #. etc. With numerous tables and illus-
trations. Twenty-first edition. Thoroughly revised by W. J. Millar.
8vo, cloth. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . $6.50
STANDARD TEXT BOOKS. 7
RATEAU, A. Experimental Researches on the Flow of Steam
Through Nozzles and Orifices, to which is added a note on The Flow of
Hot Water. Authorized translation by H. Boyd Brydon. 12mo, cloth.
Illustrated. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . net, $1.50
RAYMOND, E. B. Alternating Current Engineering Practically
Treated. 8vo, cloth. Illustrated. 232 pages. Second edition, revised.
net, $2.50
REINHARDT, CHAS. W. Lettering for Draughtsmen, Engineers and
Students. A Practical System of Free-hand Lettering for Working Draw-
ings. New and revised edition. Eighteenth thousand. Oblong boards.
$1.00
RICE, Prof. J. M., and JOHNSON, Prof. W. W. On a New Method
of Obtaining the Differential of Functions, with especial reference to the
Newtonian Conception of Rates of Velocities. 12mo, paper. . . . . . . . . . $0.50
RIPPER, WILLIAM. A Course of Instruction in Machine Drawing
and Design for Technical Schools and Engineer Students. With 52 plates
and numerous explanatory engravings, 4to, cloth. . . . . . . . . . . . . . . . . $6.00
SEVER, Prof. G. F. Electrical Engineering Experiments and Tests
on Direct-Current Machinery. With diagrams and figures. 8vo, pam-
phlet. Illustrated. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . net, $1.00
and TOWNSEND, F. Laboratory and Factory Tests in Elec-
trical Engineering. 8vo, cloth. Illustrated. 236 pages. . . . . . . . net, $2.50
SHELDON, Prof. S., Ph.D., and HOBART, MASON, B.S. Dynamo
Electric Machinery; its Construction, Design, and Operation. Direct-
Current Machines. Fifth edition, revised. 8vo, cloth. Illustrated. $
- net, $2.50
Alternating Current Machines. Being the second volume of the
authors’ “Dynamo Electric Machinery; its Construction, Design, and
Operation.” With many diagrams and figures. (Binding uniform with
volume I.) Fourth edition. 8vo, cloth. Illustrated. . . . . . . . net, $2.50
SHIELDS, J. E. Notes on Engineering Construction. Embracing
Discussions of the Principles involved, and Descriptions of the Material
employed in Tunneling, Bridging, Canal and Road Building, etc. 12mo,
cloth. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . $1.50
SNELL, ALBION T. Electric Motive Power: The Transmission and
Distribution of Electric Power by Continuous and Alternate Currents.
With a section on the Applications of Electricity to Mining Work. 8vo,
cloth. Illustrated. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
STAHL, A. W., and WQODS, A. T. Elementary Mechanism. A Text-
Book for Students of Mechanical Engineering. Eleventh edition, en-
larged. 12mo, cloth. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . $2.00
STALEY, CADY, and PIERSON, GEO. S....The Separate System of
Sewerage; its Theory and Construction. With maps, plates, and illus-
trations. Third edition, revised and enlarged. 8vo, cloth . . . . . $3.00
8 ASTANDARD TEXT BOOKS.
STODOLA, Dr. A. The Steam-Turbine. With an appendix on Gas
Turbines and the future of Heat Engines. ... Authorized Translation from
the Second Enlarged and Revised German edition by Dr. Louis C. Loewen-
stein. 8vo, cloth. Illustrated. 434 pages. . . . . . . . . . . . . . . . . . . . net, $4.50
SWOOPE, C. WALTON. Practical Lessons in Electricity. Princi-
les, Experiments, and Arithmetical Problems. An Elementary Text-
ook. With numerous tables, formulae, and two large instruction plates.
Sixth edition. 8vo, cloth. Illustrated. . . . . . . . . . . . . . . . . . . . . net, $2.00
THURSO, JOHN W. Modern Turbine Practice and Water-Power
Plants. 8vo, 244 pages. Illustrated. . . . . . . . . . . . . . . . . . . . . . . . . net, $4.00
TOWNSEND, F. Short Course in Alternating Current Testing. 8vo,
pamphlet. 32 pages.. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . net, $0.75

URQUHART, J. W. Dynamo Construction. A practical handbook
for the use of Engineer-Constructors and Electricians in charge, embracing
Framework Building. Field Magnet and Armature Winding and Group-
ing, Compounding, etc., with examples of leading English, American,
and Continental Dynamos and Motors; with numerous illustrations.
12mo, cloth. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . $3.00
WALLING, B. T. Lieut.-Com., U.S.N., and MARTIN, JULIUS.
Electrical Installations of the United States Navy. 8vo, cloth. In press.
WEISBACH, JULIUS. A Manual of Theoretical Mechanics. Ninth
American edition. Translated from the fourth augmented and improved
German edition, with an Introduction to the Calculus by Eckley B. Coxe,
à. Mining Engineer. 1100 pages, and 902 woodcut illustrations. 8vo,
cloth. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . $6.00
and HERRMANN, G. Mechanics of Air Machinery. Author-
ized translation with an appendix on American practice by Prof. A.
Trowbridge. 8vo, cloth 206 pages. Illustrated... . . . . . . . . . . . . . . net, $3.75
WESTON, EDMUND B. Tables Showing Loss of Head Due to
Friction of Water in Pipes. Second edition. 12mo, cloth. . . . . . . . . $1.50
WILSON, GEO. Inorganic Chemistry, with New Notation. Revised
and enlarged by H. G. Madan. New edition. 12mo, cloth... . . . $2.00
WRIGHT, Prof. T. W. Elements of Mechanics, including Kinematics,
º and Statics. Third edition, revised and enlarged. 8vo,
Clot s tº º e º º º • Q & Q w e º 9 0 w w tº a c e º 0 t e º 'º C G tº 0 & 0 & 0 e º e o 'º e s s as a • a a a e s a e º a s a $2.50
G-4 cº-
UNI i
1
TY OF MICHIGAN
s
|||||||||||||||
390.15078655381

±±±），
• • • • • • •
C
####
№
∞ √
}，№ſae
№ț¢;&#8-
|- s - ö.-
~·· • • • •
-» № ！！… "
-|·-~•• • • • • • •
*： < ≤ ≤
, ， ， ， º
，·
ſae e º º
© • • • • • • • •
Œ œ • e© ®[…]
·EŒ œ
№.Dae
ſ ºſ.
Œ œ · @
; ： ： .w №
º.a. s. š. ：-）.*', '... --
· e· · · · · · * *·∞•··
·| `ſ -a * ， ， !ae … º* * , ， º „
&.*,，，，-，-， º-º--ººsserwºesº！--~~~~<！-- -*-~ ~~~~? :º ： *****
|-·D
A s v. … + º ：∞ √ √， , * · *ſº----■ ■ ■ ■ ■ ■∞
a ſ s（， ………………… （a → • • • ►e.…… …………… • • • • ? '...'，
|－， ， ， ， º：w º·
ae, * * · * *- -
●ſºn • •º ſae º
• • • • • • • №ſſae-
º ſº ſae ;Țſ.، ،
(）§.ae ：... e º■ ■ ■…
º ^，----·ſ-2,±,±,±,±,±，±），
^ „ “, “ . . . • • • • • • • • • • • • • •º º , .
Ōº∞ √·
·-►●∞w:
„“∞ √≠ √º.º.º.º.º.º.º.…º.º.）"'.
, ’.v · · · · · * *A*, . . .º. ， ， , ， , º_° ） '~：, . • • • • •·
ºs ：- --- e- → º.º:º:º:º rºsºs.--~~~~**************************<！，-,-,-，-- ·+8.
）}º ：tº . 'R■|-ſ.----
• • • •º ·• • •·：··• • • •
*******eſſe~~~~*;**^*^®^^^***
∞ √|- …·e ） » »•^-●
·* ,-
-№© ： * · * *
-ſae• • • • • • • • • • ► ſa a ： » e o
∞ √
prºiz<！-- ***
№ º·• • • ? *
Œ œ(）
∞ √
ſae
∞ → • ，
، ، ، ، ، ،！--. .
! ！! ） ■ ■ ■ ■ ■ Cr + - * *• • • • • ►►*** .% ºſ
w º
∞ √●, ， ， , ，
„… №*******º, mae<!--º--º--~~~~
，--->-º a º. º. v.
ſ ºſ */， ，
º … ！！ !！
· · · · · ** ''; ',
ae ：） {• • • • * , †
·º·：·º·：·º·：·*********-:ºrºrº
º º← • • •،：
œ • • •
W
•， ！…
،* * * · · · · · · · -* （ .. . • • • • • • • •
£
|-■ ■ *-
• º.º… • • • •
·， * *