Skip to main content

Full text of "Quantitative chemical analysis"

See other formats


TEXT-BOOKS OF SCIENCE 



ADAPTED FOR THE USE OF 



ARTISANS AND STUDENTS IN PUBLIC AND SCIENCE SCHOOLS. 



QUANTITATIVE CHEMICAL ANALYSIS. 



* PRINTED BY 

SPOTTISWOODE AND CO., NEW-STREET SQUARE 
LONDON 



QUANTITATIVE 



CHEMICAL ANALYSIS; 



BY 



T. E. THORPE, PH.D., B.Sc.(ViCT.), F.R.S. 

PROFESSOR OF CHEMISTRY IN THE NORMAL SCHOOLS OF SCIENCE, SCIENCE AND 
ART DEPARTMENT, SOUTH KENSINGTON, LONDON. 



NINTH EDITION. 



NEW YORK: 

JOHN WILEY & SONS, 15 ASTOR PLACE. 
1891. 



PREFACE 

TO 

THE FIFTH EDITION. 



THE PRESENT EDITION of this work will be found to 
include a considerable amount of new matter. Many 
valuable hints and suggestions have been received 
from teachers and others both in this country and 
in America. Professor Frankland has kindly looked 
over the section on Water Analysis ; Professor Dittmar 
has fui nished an account of his method for the valua* 
tion of chrome ore ; and Mr. Watson Smith has sent 
the results of his experience of the work in the labora- 
tory of the Owens College, Manchester. To these 
and to other gentlemen who have furnished him with 
additions and corrections, the Author begs to tender 
his grateful acknowledgments. He also desires to 
express his indebtedness to Mr. C. H. Bothamley, 
Assistant-Lecturer on Chemistry in the Yorkshire 
College, for aid in the revision of the book. 

YORKSHIRE COLLEGE, LEEDS : 

237347 



PREFACE. 



THE AIM of this book is to teach the principles of 
Quantitative Chemical Analysis by the aid of ex- 
amples chosen, partly on account of their practical 
utility, and partly as affording illustration of the more 
important quantitative separations. 

It is divided into five distinct parts. The first part 
gives a description of the balance, of the mechanical 
principles involved in its construction, and of the 
manner of using it. It also contains an account of 
the operations generally or most frequently occurring 
in Quantitative Analysis ; such as the process of fil- 
tration, the incineration of filters, and so forth. 

The second part consists of a graduated series of 
examples in simple gravimetric analysis, commencing 
with the analysis of copper sulphate, and ending with 
the estimation of arsenic, antimony and tin. 

The third part treats of volumetric analysis. The 
more important volumetric processes are fully de- 
scribed, and their application is illustrated by ex- 
amples of scientific as well as of technical interest 



viii Preface. 

The fourth part contains an account of the methods, 
gravimetric and volumetric, employed in the analysis 
or valuation of ores, minerals, and of the more im- 
portant industrial products, such % as copper and lead 
ores, iron and manganese ores, limestone, cast and 
wrought iron, soda-ash, bleaching powder, &c. Con- 
siderable space has been allotted to the important 
subject of water- analysis. 

The fifth part treats of the general processes of 
organic analysis. 

The Author's thanks are due to his assistant, Mr. 
Dugald Clerk, for the attention he has bestowed on 
the drawings for the woodcuts. The illustrations in 
the section on * Water-analysis ' are taken, with Mr. 
Sutton's kind permission, from his work on ' Volumetric 
Analysis/ In the account of Frankland and Arm- 
strong's method of determining the amount of organic 
carbon and nitrogen in water, it will be seen how 
much the Author is indebted to Mr. Wm. Thorp's 
excellent description of the process in that manual 
Mr. Crookes has also kindly allowed the use of the 
figures in illustration of Luckow's process for assay- 
ing copper-ores. 



CONTENTS. 



PART I. 

PRINCIPLES OF QUANTITATIVE ANALYSIS. 

PAGE 

The Balance I 

The Weights . 16 

The Operation of Weighing ....... 20 

General Preliminary Operations . . . . . -33 



PART II. 



SIMPLE GRAVIMETRIC ANALYSIS. 

I. Copper Sulphate . . . . . . . 70 

II. Sodium Chloride ....... 79 

III. Pearl-ash 83 

IV. Rochelle Salt 85 

V. Dolomite . 85 

VI. Barium Sulphate . . . . . . -91 

VII. Indirect Estimation of Barium and Calcium . . 92 



Contents. 

PAGE 

VIII. Ferrous Ammonium Sulphate . -. , . ". . -94 
IX. Determination of Nitric Acid .... 96 

X. Potash-alum . . / 98 

XL Glass . . . 99 

XII. Felspar 101 

XIII. Brass, Bronze, Gun-Metal, Bell-Metal . . .103 

XIV. German Silver . ~ % . . , . . . 106 
XV. Britannia Metal 107 

XVI. Type-Metal . . . .108 

XVII. Fusible Metal ... 109 



PART III. 

SIMPLE VOLUMETRIC ANAL YSIS OF SOLIDS AND 
LIQUIDS. 

I. Determination of Chlorine by Standard Silver Solu- 
tions . . . . . . . .124 

II. Indirect Determination of Potassium and Sodium 
by means of Standard Silver Solution and Potas- 
sium Chromate . . . . . . .128 

III. Estimation of Chloric Acid ..... 129 
Ilia. Determination of Chlorine in presence of Sulphites . 129 

ALKALIMETRY. 

IV. Valuation of Soda- Ash .... v 137 
V. Estimation of Alkaline Hydrate in presence of Car- 
bonate ,'. .. . , . .'; ..' . . -138 

VI. Estimation of Sodium Carbonate in presence of Potas- 
sium Carbonate . . -.' T ,-. ,., . . T39 
VII. Determination of Ammonia . , . ',.--,,. 140 



Contents. xi 



ACIDIMETRY. 

PAGE 

VIII. Determination of the Strength of Acetic Acid, Pyro- 

ligneous Acid, Vinegar . . . . . .142 

IX. Determination of combined Carbon Dioxide . . 143 
X. Estimation of Carbonic Acid in Natural Waters . 144 
XI. Estimation of Carbon Dioxide in Artificially Aerated 

Waters 145 

XII. Determination of Combined Acids in Salts . . 146 



ANALYSIS BY OXIDATION AND REDUCTION. 

XIII. Determination of the Strength of the Permanganate 

Solution ........ 148 

XIV. Volumetric Estimation of Calcium by means of Potas- 

sium Permanganate . . . . . -152 
XV. Volumetric Estimation of Lead by Permanganate Solu- 
tion 153 

XVI. Valuation of Manganese Ores by means of Potassium 

Permanganate Solution . . . . . -154 

XVII. Estimation of Potassium Ferrocyanide by Permangan- 
ate Solution . . . . . . .154 

XVIII. Estimation of Potassium Ferricyanide by Permangan- 
ate Solution . . . . . . -154 

XIX. Valuation of Bleaching-powder by Iodine and Sodium 

Thiosulphate Solutions . . . . .158 

XX. Estimation of the Amount of Chlorine in Aqueous 

Solutions of the Gas . . . . . .158 

XXI. Estimation of the Amount of Sulphur Dioxide in 

Aqueous Solutions of the Gas . . . 159 

XXII. Estimation of Sulphuretted Hydrogen in Aqueous So- 
lutions of the Gas . . . . . . -159 

XXIII. Estimation of Hydrocyanic Acid . . 160 

XXIV. Estimation of Antimony in Tartar Emetic . .161 
XXV. Determination of Tin by Iodine Solution . 162 



xii Contents. 



ANALYSES BY MEANS OF IODINE AND SODIUM THIO- 
SULPHATE SOLUTIONS, WITH PREVIOUS DISTILLA- 
TION WITH HYDROCHLORIC ACID. 

PAGE 

XXVI. Analysis of Potassium Bichromate .... 163 

XXVII. Estimation of Arsenious Acid .... 165 

XXVIII. Analysis of Chlorates, Bromates, and lodates . -165 
XXIX. Estimation of Iron by means of Iodine and Thiosul- 

phate Solutions . . . . . . .166 

XXX. Estimation of Nitric Acid by Solutions of Iron, Iodine, 

and Sodium Thiosulphate . . . . .167 

XXXI. Valuation of Manganese Ores by Distillation with 
Hydrochloric Acid, and Titration with Iodine and 
Thiosulphate Solutions . . . . .167 



PART IV. 

GENERAL ANALYSIS, INVOLVING GRAVIMETRIC AND 
VOLUMETRIC PROCESSES. 

I. Nitre . ... .' . ' . v .168 

II. Gunpowder . . . . ''. . . . 17 

III. Limestones. Hydraulic Mortar . . . . 174 

IV. Clays . . . ' f . ' V ' . . ' ' . i . 180 
V. Assay of Manganese Ores (Pyrolusite, Braunite, &c. ) 

Gravimetrical Method of Fresenius and Will . 185 
VI. Volumetric Determination by means of Iron and Po- 
tassium Permanganate Solution . '.' . .189 
VII. Volumetric Determination by means of Oxalic Acid 

and Potassium Permanganate . . . .189 
VIII. Determination of Moisture in Manganese Ores . 190 



Contents. xiii 

PAGE 

IX. Determination of the Amount of Hydrochloric Acid 

required to decompose Manganese Ore . . igi 

X. Bleaching Powder ....... 193 

XI. Black- Ash ; Soda-Ash ; Vat-Waste ... 198 
XII. Estimation of Sulphur in Pyrites, by means of Copper 

Oxide ......... 205 

XIII. Assay of Copper Ores (Mansfeld Process) . . 205 

XIV. Assay of Copper Ores (Luckow's Process) . . . 207 
XV. Assay of Copper Ores by Precipitating the Metal by 

means of Zinc ....... 2 IO 

XVI. Copper Pyrites . . . . . , -211 

XVII. Iron Pyrites 215 

XVIII. Kupfernickelstein ' 215 

XIX. Iron-Ores 215 

XX. Titaniferous Iron-Ore (Ilmenite) .... 225 

XXI. Wrought and Cast Iron and Steel .... 226 

XXII. Iron Slags .' 249 

XXIII. Assay of Zinc- Ores ...... 250 

XXIV. Assay of Tin-Ores ....... 252 

XXV. Separation of Tin from Tungsten . . . -253 

XXVI. Wolfram 2^4 

XXVII. Scheelite 255 

XXVIII. Galena 255 

XXIX. Refined Lead 258 

XXX. White Lead . . ' 267 

XXXI. Chrome Iron-Ore 268 

XXXII. Smaltine : Cobalt-Glance 272 

XXXIII. Fahl-Ore (Tetrahedrite) 276 

XXXIV. Determination of Silver in Solutions .... 279 
XXXV. Assay of Silver in Bullion, Coin, Plate, &c. . .281 

XXXVI. Assay of Gold . 286 

XXXVII. Separation of Gold, Silver, and Copper . . 286 

XXXVIII. Estimation of Mercury 287 

XXXIX. Coal 289 

XL. Examination of Water used for Economic and Techni- 
cal Purposes .... . 291 



xiv Contents. 

PAGE 

XLI. Determination of the Amount and Nature of the Gases 

dissolved in Water .'',.' . " .. ' . . 327 
XLII. Guano . . ~" V . , x . , \ . ' , '' . . 331 

XLTII. Bone-Dust . 337 

XLIV. Superphosphates 337 

XLV. Ashes of Plants ....... 340 



PART V. 



ORGANIC ANALYSIS. 

I. Analysis of Bodies containing Carbon and Hydrogen, 

or Carbon, Hydrogen, and Oxygen . . . 348 
II. Analysis of Organic Substances containing Nitrogen 358 

III. Analysis of Organic Substances containing Chlorine. 

Bromine, and Iodine " . . . , |r 364 

IV. Analysis of Organic Substances containing Sulphur 

and Phosphorus . . \ . . . . 366 



APPENDIX . .-.,,,. , V "' - - 369 



QUANTITATIVE 

CHEMICAL ANALYSIS. 

PART I. 
PRINCIPLES OF QUANTITATIVE ANALYSIS. 

THE BALANCE. GENERAL PRELIMINARY OPERATIONS. 

QUANTITATIVE ANALYSIS is that branch of Practical 
Chemistry which treats of the processes by which we deter- 
mine the relative amounts of the constituents of a body. 
QUALITATIVE ANALYSIS informs us simply of the nature of 
these constituents, and teaches us how they may be sepa- 
rated. The latter form of analysis always precedes the 
former, for, obviously, we must first know what are the 
elements present in a substance before we can proceed to 
estimate their proportions. 

The methods of Quantitative Analysis are subdivided 
under the two heads of Gravimetric Analysis and Volumetric 
Analysis. By means of Gravimetric Analysis we seek to 
weigh the known constituents of a substance either in their 



2 Quantitative Chemical Analysis. 

elementary condition,' or in 'the form of combinations which 
admit of exact weighing, and of which the composition is 
already accurately known. Supposing that we wish to deter- 
mine the composition by weight of a sixpenny piece : quali- 
tative analysis tells us that the coin is made up of silver and 
copper. We may determine the proportion of the two metals 
in the solution of the coin, either by separating them out and 
weighing them in their metallic state, or we may convert the 
silver into silver chloride, and the copper into cupric oxide, 
and weigh the two compounds. Since we know the com- 
position of the silver chloride and cupric oxide, we can 
calculate the amount of silver and copper respectively con- 
tained in them, and in this manner determine the relative 
quantities of the metals present in the coin. In practice, it 
is usually found more convenient to estimate the constitu- 
ents in a body by the aid of combinations of known com- 
position, rather than to attempt to isolate the elements. It 
is evident, therefore, that a correct knowledge of the pro- 
portion in which the several elements are present in these 
fiduciary combinations is of the highest value to the analyti- 
cal chemist ; and, further, that the exact determination of 
the combining weights of the elements becomes to him a 
matter of primary importance. 

But it will be obvious on reflection that we can deter- 
mine the quantitative composition of the sixpenny piece 
without directly weighing either the metals, or their com- 
binations with chlorine and oxygen. We might determine 
the amount of the silver, for example, by ascertaining the 
quantity of hydrochloric acid required to convert it com- 
pletely into silver chloride. We know that if we add hydro- 
chloric acid to solution of silver in nitric acid (silver nitrate) 
we obtain insoluble silver chloride ; and that if we add a 
sufficiency of hydrochloric acid the whole of the silver will 
be thrown out of solution : 

AgNO 3 + HC1 = AgCl + HN0 3 . 



The Balance. 3 

Now, if we know how much hydrochloric acid (HC1) is 
contained in any given volume of the solution which we 
employ to precipitate the silver, and if we have the means 
of recognising the exact point at which the formation of 
silver chloride ceases, we can calculate from the volume of 
acid required the amount of the silver, since, as the equation 
tells us, 36*46 parts of hydrochloric acid are equivalent to 
107*93 parts of silver. This is the fundamental principle o* 
volumetric analysis, a form of quantitative analysis in which 
we seek to estimate the amount of a substance from the 
determinate action of reagents in solutions of known 
strength, the amount of the reacting substance being calcu- 
lated from the volume of liquid used. Many examples 
might be adduced to show the wide applicability of this 
principle of analysis. Let us suppose that we wish to deter- 
mine the amount of sodium hydrate in an aqueous solution 
of this substance. If we add a few drops of litmus tincture 
to the liquid we obtain a blue colouration, which, on the con- 
tinued addition of hydrochloric acid, eventually becomes 
permanently red. The acid combines with the alkali to 
form common salt, which is without action on the colour of 
litmus ; the final change in colour shows us that the whole 
of the sodium hydrate is in combination, and that the acid 
is in very slight excess. If the strength of our hydrochloric 
acid solution is known to us, that is, if we can say how 
many grams of H Cl are contained in 1,000 cubic centimetres 
(for instance) of the liquid, we can calculate, from the 
number of cubic centimetres we require to add to the soda 
solution coloured with litmus before it is permanently red- 
dened, how much sodium hydrate is contained in the alka- 
line liquid originally taken, from the knowledge that 
HCl=NaHO ; i.e. that 36*46 grams of hydrochloric acid are 
equivalent to 40*04 grams of sodium hydrate. 

THE BALANCE. 

The balance affords the only practicable means of mea- 
B 2 



4 Quantitative Chemical Analysis. 

suring the mass of the various forms of matter contained in 
a substance. Practically speaking, this instrument consists 
of an inflexible metallic lever or beam suspended near its 
centre of gravity on a fulcrum or pivot, the masses to be 




compared being also suspended from pivots placed at the 
extremities of the beam, equidistant from, and in the same 
horizontal line with, the central fulcrum. For a complete 
treatment of the mechanical theory of the balance we 



The Balance. 



must refer to special treatises on the subject : in this 
work we shall mainly confine ourselves to the essential 
points in its construction and mode of use, and only touch 
on the mechanical problem in so far as it appears necessary 
to enable the student to understand the conditions of sen- 
sibility and accuracy in the instrument. Fig. i gives a 
representation of a modern chemical balance. The beam 
a a has the shape of an 
acute rhomboid ; this 
form of construction 
combines lightness 
with inflexibility and 
strength : on the pos- 



FlG. 2. 




FIG. 3. 



session of these quali- 
ties in the beam much 
of the sensibility and 
accuracy of the balance 
depends. Through the 
centre of the beam passes a triangular piece of hardened 
steel or agate, termed a knife-edge, the lower edge of which 
turns upon a horizontal plate of polished agate connected 
with the pillar. At the 
end of each arm is a 
similar knife-edge fixed 
in the reverse position, 
and bearing an agate 
plate from which de- 
pend steel hooks to hold 
the wires attached to the 
pans (fig. 2). These 
terminal knife-edges are 
fixed in brass settings, 
and admit of being ad- 
justed so as to bring them into exactly the same plane with 
the centre edge. Their relations to the middle knife-edge 
may be altered by means of the little screws shown in the 




6 Quantitative Chemical Analysis. 

figure. Various other methods of arranging the terminal 
knife-edges and pan-suspensions have been proposed. Fig. 3 
represents a form adopted by continental balance- makers. 

As the efficacy of the instrument depends to a large 
extent on the preservation of the sharpness of the knife- 
edges and the smoothness of the agate planes, it is desirable 
to prevent their contact when the balance is not in use. 
This is effected by means of the frame b b (fig. i), which 
lifts the centre knife-edge about 0*2 millimetre from the 
centre plane : at the extremities of the frame are steel 
points which enter into little hollows in the lower surface of 
the pan-suspensions, and raise them from the terminal 
knife-edges. This frame is attached to a rod descending 
through the pillar, and connected with an eccentric worked 
by a milled-head screw (s) situated on the outside of the 
balance-case : by means of this movement, the rod, and 
with it the frame, can be raised or lowered at pleasure. In 
balances of the highest class there is a second eccentric con- 
nected with a system of bent levers which carry supports for 
the pans ; by means of these supports the pans can be 
steadied whilst the weights are being transferred, or their 
vibrations can be checked preparatory to releasing the 
beam. In some balances the pan-supports are worked by 
an independent screw : in, others they are worked in con- 
junction with the movement which raises or lowers the 
frame. Where all the movements are controlled by a single 
screw these are not made to act quite simultaneously. 
When the balance is to be set in vibration, the first action 
of the screw lowers the pan-supports ; it next brings down 
the centre knife-edge upon the agate plane, and gradually 
allows the pan-suspensions to drop simultaneously upon the 
terminal knife-edges. For the proper performance of these 
movements great nicety of workmanship is needed, for it is 
not only requisite that the beam and pan-suspensions should 
be properly raised when wanted, but it is also necessary 
that the edges and planes should be brought into contact in 



The Balance. j 

a constant position. The movements of the beam are in- 
dicated by a vertical pointer which oscillates before an 
ivory scale fixed to the pillar; this ivory scale is usually 
graduated into 20 parts, and its middle point or zero is 
exactly behind the needle when the beam is horizontal. Any 
inequality in the weight of the arms is compensated by 
means of a small vane fixed on the top of the beam above 
the central knife-edge, which may be turned to the right or 
left as occasion requires. In some balances this compen- 
sation is effected by means of little screws travelling along 
fine threads attached to the ends of the beam (see fig. 3). 
The stability of the beam is regulated by the aid of a 
weight termed the gravity-bob (g) moving along the rod 
attached to the upper edge of the beam over the centre knife- 
edge on which the vane works. 

In order to protect the instrument from the fumes of the 
laboratory, and to prevent air-currents from interfering with 
its action during the operation of weighing, it is enclosed in 
a glass case, the back, front, and sides of which can be 
opened at will. The case is supported on levelling screws, 
by which it can be adjusted to horizontality in accordance 
with the indications of a spirit-level or plumb-line attached 
to the instrument. When an object too bulky to be brought 
within the balance-case has to be weighed, on releasing the 
screws at the base, the pillar and beam can be turned 
through an angle of about 60, so that the ends of the beam 
project beyond the back and front of the case. The proper 
adjustment of the beam on the part of the balance-maker is 
an operation of the greatest nicety. To ascertain if the 
three knife-edges are in the same plane, he first poises the 
beam without weights on the pans, and moves the gravity- 
bob until the vibrations, as indicated by the pointer, become 
very slow he then puts equivalent weights into the pans, 
and again sets the beam vibrating : if its rate of vibration is 
unaltered, the adjustment is perfect. If the beam vibrates 
too quickly or oversets, the gravity-bob is raised or lowered 



8 Quantitative Chemical A nalysis. 

so as to bring the vibrations to the original rate : the 
number of turns required to effect this is noted, and then 
the bob is turned in the contrary way through double the 
number of revolutions, and the slow motion is again pro- 
duced by means of the adjustments at the ends of the 
beam. To determine whether the terminal knife-edges are 
at equal distances from the centre edge, the beam is poised 
with weights, and the pans, together with their suspensions, 
are changed from side to side. If the equilibrium is un- 
disturbed, the edges are properly adjusted ; if, however, one 
side appears heavier than the other, a small piece of bent 
wire termed a rider (see fig. 7) is placed on the lighter side, 
and pushed along the beam until the equilibrium is again 
established : the rider is now pushed along half way towards 
the centre of the beam, and the adjustment made at one 
end. The knife-edges may be known to be parallel by 
hanging little hooks upon them and equipoising the beam : 
on sliding the hooks along the knife-edges, equilibrium 
should be maintained. The student will better appreciate 
the skill required in these adjustments when we treat of 
the circumstances, other than those due to imperfect work- 
manship, which modify the action of the balance. 

We will next briefly state the main conditions upon which 
the stability, sensibility, and accuracy of the instrument 
depend. 

i. The centre of gravity of the balance must be situated 
below the point of suspension, i.e. the centre knife-edge. If 
the balance were suspended at its centre of gravity, it would 
be in the condition of neutral equilibrium, and the beam 
being once disturbed would have no tendency to reassume 
horizontality, but would remain in any position given to it ; 
that is to say, the beam would not oscillate. If, on the 
other hand, the centre of gravity is above the point of sus- 
pension, the beam would be in the state of unstable equi- 
librium, and would tend to overset with the least prepon- 
derating weight. 



The Balance. 9 

2. The terminal points of support (knife-edges for pan- 
suspensions) must be in the same line with the centre 
point of suspension. In other words, an imaginary straight 
line drawn from one terminal edge to the other should 
just touch the lowest point of the centre knife-edge. If 
the centre edge is below the line joining the extreme 
points of suspension, the centre of gravity of the whole 
system will be raised in proportion as the load is increased, 
and at a certain point the centre of gravity will be exactly 
at the point of suspension of the beam, which will then 
cease to oscillate ; a continued increase of weight will 
now raise the centre of gravity above the point of sup- 
port, when the beam will overset. But if the centre edge 
is situated above the line joining the extreme edges, the 
centre of gravity of the whole system is lowered in pro- 
portion to the increase in load ; that is, its sensibility 
will diminish with the increase of weight on the pans. 
When the three edges are in the same plane an increase 
of weight continually tends to bring the centre of gravity 
nearer the point of support, but it can never be made 
to coincide with it : consequently the balance will never 
cease to vibrate. We thus see why it is necessary that 
the beam should be perfectly inflexible. If the weights 
acting on the terminal edges caused them to sink below 
the level of the centre edge, the centre of gravity of the 
whole system would become lowered, and the sensibility 
be lessened. Theoretically, increased weight creates in- 
creased sensibility in the instrument : practically, however, 
this increase in sensibility is more than counterbalanced 
by other effects of increased weight 

Let a be the central knife-edge supporting a beam seen end- 
on (fig. 4), and b and c the terminal points of suspension ; the 
straight line b c touches the point a of the centre knife-edge. 
Fronitf draw the vertical a d\ then the centre of gravity of 
the beam must be somewhere on this line a d : let us suppose 
it to be at e. If equal weights w are suspended from b and 



io Quantitative Chemical Analysts. 




The Balance. 1 1 

<r, we may conceive that each of the weights acts respectively 
at b and c : their common centre of gravity falls at # , and the 
centre of gravity of the whole system of beam and load is 
somewhere between a and e along the vertical a d. Since 
the centre of gravity remains vertically under the point 
of support, the equilibrium is undisturbed. Suppose now 
that an additional weight w is made to act at c y the centre 
of gravity of the combined load is no longer coincident 
with a, but falls somewhere along the line b c in the direction 
of t, say at f\ the centre of gravity of the whole system is 
at some point along the line e /, say at g. The centre of 
gravity is no longer vertically under the point a : accord- 
ingly the beam tends to revolve until this condition obtains 
(fig. 5). The arm a c therefore falls, whilst, of course, the 
arm a b is proportionately raised. The angle made by the 
beam in its new position of equilibrium with the horizontal, 
due to the preponderance a/, is termed the angle of deviation : 
it is equal to the angle g a e. This angle is the measure of 
the sensibility of the instrument. It is evident, moreover, 
that (3) the sensibility of the balance is augmented by bringing 
the centre of gravity as near the point of support as possible, 
whereby g, due to an overplus, becomes proportionately 
raised towards the line b c, and the angle of deviation g a e 
increased in the direct ratio of the change. If e is so far 
raised as to be identical with 0, then on the addition of the 
weight w, g would fall on the line b c ; that is, the angle of 
deviation becomes a right angle, and the beam consequently 
oversets. The distance between the point of support and 
centre of gravity in a sensitive instrument probably does not 
exceed T J F of a millimetre. The arrangement by which 
the alteration in the centre of gravity is effected has already 
been described. There is, however, a limit of approxima- 
tion beyond which, in ordinary work at least, it is practically 
inconvenient to go, and for another reason than that just 
adduced. As the centre of gravity nears the point of sup- 
port, the rapidity of the vibrations of the beam diminishes, 



12 Quantitative Chemical Analysis. 

and ultimately becomes very slow : the operation of weigh- 
ing may thus need a greater expenditure of time than is 
warranted by the degree of accuracy required. 

4. The sensibility of the instrument increases with the 
length of the arms. If ac be made longer, the distance af 
will be proportionately increased, and the point g will be 
further removed from the vertical a d : the line ag will 
therefore form a greater angle with a e that is, the angle 
of deviation will be increased. Here, too, in practice, we 
quickly find the limit to the length of beam. By increasing 
the length of the beam we increase its weight, and by in- 
creasing its weight we diminish its sensibility. 

5. The sensibility is dependent upon the lightness of the 
beam. We may conceive that the weight 2W + w, acts at 
the point f t and that the weight x of the beam acts at e. 
The position of the centre of gravity g t along the line 
e /, is obviously dependent upon the relation of the 
weights acting at e and/: if e x = (2W + w) /, then g 
will be equidistant from e and f, and the smaller x becomes 
in proportion to 2W + w the further will g be removed from 
<?, and therefore the greater will be the angle of deviation. 
With the view of increasing the sensibility by diminishing 
the weight, other substances than brass have been proposed 
as the material of balance-beams, and beams have actually 
been constructed of aluminium, which possesses a specific 
gravity of only 2 '6, less than one-third that of brass. 

6. The sensibility of the balance is also affected by the friction 
between the knife-edges and agate planes. The immediate 
effect of this friction at the terminal knife-edges is prac- 
tically to vary the length of the arms. Let us suppose 
that the impediment to the free motion of the planes of the 
pan-suspensions is at its maximum, or, in other words, that 
the planes of the pans are maintained parallel to the direc- 
tion of the beam during the oscillations of the instrument. 
Such an instrument would be perfectly useless as a balance, 
for as one arm was depressed by the action of a prepon- 
derating weight, the heavier pan would be thrown inwards, 



The Balance. 1 3 

but its tendency to move would be counterbalanced by 
the other pan being thrown correspondingly outwards. 
This variableness, practically speaking, in the length of the 
arms may be perceived in balances of which the edges 
have become blunted by wear. Supposing that the width 
of the knife-edges is x, and the distance between their 
middle points is jy, a glance at the figure (fig. 6) enables us 

FIG. 6. 




to see that the real lengths of the arms must be x y and 
x + y. Therefore two loads possessing the ratios of x -\-y 
and x y will be in apparent equilibrium. It is said that 
some balances will indicate one part in a million ; if such an 
instrument possessed a 20-inch beam, the width of the 
knife-edges cannot exceed ^nnnnro* of an inch (an inappre- 
ciable amount even under a microscope), and the two arms 
must be adjusted to equality within this length. 

When properly adjusted, a balance should satisfy the fol- 
lowing tests : i. When the beam is released the pointer 
should coincide with the zero on the scale, or make slow 
excursions tending to equality of amplitude on either side. 
2. The equilibrium should not be disturbed when the pans are 
removed. 3. If possible, the position of the beam should be 
reversed, so as to cause the arm which points to the right to 
point to the left, and the beam again be made to oscillate : it 



14 Quantitative Chemical A nalysis. 

should vibrate exactly as before, and finally acquire a horizon- 
tal position. Imperfect workmanship in the middle knife-edge 
or in the planes on which it works is immediately indicated 
if the beam now behaves differently. In a good balance the 
centre knife-edge, whether of steel or of agate, is made in 
one piece, and runs through a perforation in the beam : if, 
as is occasionally the case, the knife-edge is made in two 
parts, one being affixed to each side of the beam, it becomes 
almost impossible to bring the edges into exactly the same 
straight line. 4. Inequalities in the lengths of the arms 
may be detected by loading the pans, after the beam has 
been found to be in equilibrium, with counterpoising 
weights, and transferring the weights from one pan to the 
other ; if equilibrium is retained, the lengths of the arms are 
equal. A final test of the efficacy of the balance consists 
in weighing an object several times in succession with the 
greatest exactitude which the instrument admits of: if it is 
trustworthy the greatest differences will not exceed o'2 mil- 
ligram. 

It is very desirable that the student, so soon as he has 
had a little practice in weighing, should make himself tho- 
roughly acquainted with the capabilities of the instrument 
which he employs. An hour's careful observance of its 
behaviour under varying conditions will materially conduce 
to accurate weighing, and save much subsequent time. He 
should, in the first place, determine for himself the degree 
of its sensitiveness by accurately noting the deviation from 
the zero-point, which an overweight of a milligram effects ; 
(i) when the balance is unloaded, and (2) when carrying 
50 grams on each pan ; and (3) when carrying 100 
grams. He should at the same time observe the variation 
in the rates of oscillation as the load increases. As the 
balance approaches equality, after some experience he 
will readily be able to estimate the weight required to 
establish equilibrium, from the extent and rapidity of the os- 
cillations. 



The Balance. 1 5 

The balance-room should have a northerly aspect, and it 
should not be liable to great or sudden variations in tem- 
perature. If possible, the instruments should be so arranged 
within the room, that in weighing the light falls over the 
right shoulder of the operator. The position of the several 
balances should be fixed, and they should be moved as 
seldom as possible, otherwise their adjustment to horizon- 
tality will be continually disturbed, and the shaking of the 
beams and pans will inevitably interfere with the constancy 
of their indications. Frequent attempts at readjustment 
by inexperienced hands will certainly disturb the regular 
action of the instrument. If, on commencing to weigh, the 
balance is found not to be in equilibrium, the beam and 
pans should be lightly brushed with a camel's-hair brush, 
and the horizontality of the beam again tested. It should 
be carefully borne in mind that none of the adjustments 
ought to be disturbed without sufficient reason, and only 
after a proper inspection of the several parts of the instru- 
ment. In a laboratory where the same balance is used by 
several operators, the necessary adjustments should be in- 
variably made by the assistant, for no one can have con- 
fidence in the indications of an instrument which is liable to 
hasty and improper alteration. A balance suffers more from 
imperfect preservation than from proper usage. The fumes 
of the laboratory should be carefully excluded from the 
balance-room, and the student should never neglect to 
securely close the doors of the balance-case when he has 
finished weighing. Careful exclusion of acid fumes will do 
much to prevent the corrosion of the polished knife-edges 
and suspensions : if rusting commences at any one point it 
will rapidly extend over the entire surface of the steel. In 
order to diminish the humidity of the air a small dish con- 
taining dried carbonate of potash or a piece of freshly-burnt 
lime should be kept within the case. A balance in constant 
use will require cleaning about every three or four months. 
The instrument -should be carefully taken to pieces, and the 



1 6 Quantitative Cliemical A nalysis. 

loose parts dusted ; the beam, pans, and suspensions should 
be rubbed with a piece of soft leather, and the movements 
cleaned and oiled. Occasionally the agate planes will re- 
quire repolishing by the maker, as the constant working of 
the knife-edges wears a minute groove in them, easily per- 
ceptible by a lens. Lastly, all delicate instruments should 
be encased in a well-fitting baize or linen bag when not in 
frequent use ; this will greatly tend to keep out dust and 
acid fumes. 

THE WEIGHTS. 

Since the chemist mainly concerns himself with the rela- 
tive weights of the constituents of a substance, it is, for the 
greater number of his operations, a matter of indifference 
what unit he adopts. He needs merely to assure himself 
that its multiples and submultiples actually have the values 
which are assigned to them. Experience shows, however, 
that it is highly desirable that a common unit should be 
adopted by chemists, and that it should be directly con- 
nected with some national standard. Accordingly, we 
employ almost exclusively the metric system, of which the 
gram is the standard weight. The reasons which have con- 
duced to the adoption of this system are (i) that the unit 
is moderately small ; (2) that its multiples and submultiples 
are derived from it by decimal multiplication and division, 
i.e. by the simplest possible system ; and (3) that it bears a 
very simple relation to the measures of capacity and length. 
A set of weights extending from fifty grams to a milligram 
will be found most generally useful. Such a set should con- 
tain pieces of the following denominations : 

grams grams gram gram gram 

50 5 0-5 0-05 0-005 

2O 2 O'2 O'O2 O-OO2 

10 I O'l O'OI 0-001 

IO I O'l O'OI O'OOI 

I O'OOl 

the entire twenty-two pieces making up 101 grams. The 



The 



I V eights. 



larger weights, down to one gram, are kept in receptacles 
lined with velvet to prevent their being scratched \ the smaller 
ones are also kept in separate compartments covered by a 
plate of glass. The box should be furnished with small 
forceps, with which the weights are invariably to be handled, 
as they should never be touched by the fingers (fig. 7). 

FIG. 7. 




Several substances have been proposed as the material for 
weights ; for example, rock-crystal and platinum have actually 
been used by reason of their durability. In general, how- 
ever, only the smaller weights are constructed of platinum, 
the others being made of brass gilded by the electrotype 
process. If the gilding is of moderate thickness, and the 
weights are preserved with due care, they will be found to 
be practically unchanged even after many years' use. 
Probably the most convenient shape for the 
brass weights is that of a short frustum of 
an inverted cone, to the base of which a 
handle is attached (fig. 8). Beneath the 
handle is a small cavity containing the 
minute pieces of foil or wire required to 
adjust the weights. The smaller weights are 
preferably made of pieces of stout platinum foil, one corner 
of each being turned up for holding in the forceps ; the very 



FIG. 8. 




1 8 Quantitative Chemical A nalysis. 

small ones are occasionally made of palladium or aluminium. 
All the pieces should have their values plainly stamped 
upon them, and the compartments of the smaller weights 
should be large enough to admit of their being easily with- 
drawn ; otherwise the foil will soon become bruised, and the 
denomination of the weight rendered indistinct. 

The greatest care should be taken to preserve the weights 
from the action of acid fumes. The lid of the box in which 
they are contained should be tightly fitting, and the box 
itself should be kept in a bag of soft leather. No attempt 
should ever be made to clean the weights by rubbing ; any 
dust may be removed by a camel's-hair brush. If by chance 
the small platinum weights become dimmed or soiled, they 
may be brightened without injury by holding them for a 
moment in the flame of the Bunsen lamp. Brass weights un- 
gilded generally become heavier by corrosion ; when gilded 
they usually become lighter by wear. Since this action 
occurs only at the surface, which increases as the square of 
the diameter, whilst the mass increases as the cube, it 
follows that the smaller weights become more quickly erro- 
neous than the larger. The error thus caused by age is, 
however, exceedingly minute. The slight tarnish is so ex- 
cessively thin that it requires a very delicate instrument to 
detect its influence. 

We strongly recommend the operator to test his set of 
weights ; for however carefully the remainder of his quanti- 
tative work may be done, if he weighs with grossly in- 
accurate weights, his results, if not valueless, will at least be 
inexact. The exact determination of the values of the 
separate pieces in a set of weights in terms of one of them 
selected as a standard is an operation of some skill, and 
requires a considerable expenditure of time. As a rule, the 
weights of the best makers possess greater accuracy than is 
required of them in ordinary quantitative processes ; never- 
theless their examination should never be omitted. The 
readiest method of detecting errors in the values of the 



The Weights. 19 

denominations is to place one of the gram weights on the 
pan of a delicate balance, adjusted to perfect equilibrium, 
and equipoise it with pieces of brass or small shot, and 
finally tinfoil. The weight is then removed and replaced 
by the second gram weight, and the balance caused tq 
oscillate. If the excursions on either side of the zero are 
of equal amplitude, the weights are equivalent ; if not, the 
deviation must be noted. It should not exceed i division 
of the scale from the zero-point. The third gram weight is 
next tried in the same way, and it is then replaced by 
platinum weights of the smaller denominations to make up 
i gram, and the balance again caused to oscillate, every 
deviation from ,the equilibrium being carefully noted. In 
the same way the 2 gram piece is compared with two of the 
single grams, the 5 gram piece with the 2 -f i + i + i gram 
pieces, and each of the 10 gram pieces with the 5 + 2 + 1 + 1 + 1 
gram pieces. The larger pieces also should not show greater 
variations than i division from the zero, since the value of 
i scale division with a heavy load on the pans is almost 
invariably greater than with a diminished load, for the reasons 
already given. Thus in a certain balance tested by the 
author, i scale division, with no load on the pans, was 
equivalent to 0-29 milligram ; i.e. a preponderance of 0-29 
milligram would cause a variation of i division from the 
zero-point. Under varying loads the value of i division 
on this balance was as follows : 

Load on each pan Value of i scale division 

grams milligram 

10 0-32 

25 0-35 

50 0-39 

100 0-51 

The process of testing should be repeated from time to 
time, particularly when the weights begin to be tarnished. 
It is also desirable that the student should be able to 
compare his weights with a standard set. 

c 2 



2O Quantitative Chemical Analysis. 



THE OPERATION OF WEIGHING. 

There are three methods of determining the weight of a 
body by means of the balance : (i) by the process of 
direct weighing, (2) by that of reversal, and (3) by that of 
substitution. As the first is the most expeditious, it is the 
one usually employed, although it is not the most accurate 
method. The substance to be weighed is placed on one 
pan, conveniently that on the left, and the weights are 
placed on the other. The pan originally selected to contain 
the object to be weighed should be used for the same 
purpose subsequently. The effect of any inequality in the 
length of the arms is thus in a great measure obviated. 
Supposing that we wished to determine the amount of copper 
iira piece of the pure metal, and that for the purpose of the 
estimation we wished to weigh out i gram of the metal, and 
that the arms of our balance were of unequal length, the one 
on the left being 99 millimetres, whilst the other on the 
right was 101 millimetres long. We place a i gram weight 
on (say) the /(//-hand pan, and counterpoise it with copper. 
Since two masses acting on a lever are in equilibrium when 
the products of the weights into their distances from the 
fulcrum are equal, it follows that the weight of copper 
equivalent to the i gram weight would be 0-9802 gram, for 
99 x i'o=ioi x 0*9802. The metal is next dissolved and 
the copper precipitated, these operations being so carefully 
performed that nothing is lost. The precipitated copper is 
then weighed, and by mischance it is brought on the left-hand 
pan ; when counterpoised with the weights, it would appear 
to weigh 0-9705 gram, since 99x0-9802 = 101x0-9705: 
accordingly the pure copper would appear to contain only 
97 '05 per cent, of metal. Had we invariably employed the 
same scale for the same purpose, we should have determined 
what we required to know viz., that the copper contained 
100 per cent, of the metal although we might not know 
(which would be perfectly immaterial) what was the true 



The Operation of Weighing. 21 

weight of the metal employed in the analysis. And by 
similar reasoning we see that, in the analysis of a substance 
containing any number of constituents, we ought to use one 
and the same pan as the object-pan ; the results, being 
strictly comparative, are thus independent of the imperfec- 
tion of the instrument, since the apparent weights of all the 
bodies weighed are altered in exactly the same ratio. 

We can, however, obtain the absolute weight of the copper 
taken for analysis, either by the method of reversal or by 
that of substitution. In the method of reversal (known as 
Gauss's) the object (the copper for example) is weighed 
first in one pan, and then in the other. If the weights are 
identical, the true weight of the object is at once given. If 
the weights are unequal, their geometric mean will be the 
true weight; this is found by multiplying the apparent 
weights together and taking their square root. Practically 
the common arithmetic mean of the two weights will be 
sufficiently accurate, unless their difference is considerable. 

In the method of substitution (due to Amiot *), the body 
to be weighed is first accurately counterpoised ; it is then 
removed, and equilibrium again established by placing 
weights in its stead. Obviously the absolute weight of the 
body is expressed by the value of the weights substituted. 
In practice this method of weighing may be facilitated by 
using one of the larger weights, heavier than the body to be 
weighed, as a counterpoise, and adding weights to the object 
until equilibrium is established. The object is then removed, 
and weights substituted until the balance is again in equili- 
brium. The substituted weights express the real weight of 
the object. 

Theoretically these methods are faultless ; practically they 

are subject to at least two minute errors one due to the 

impossibility of maintaining the edges of suspension in a 

perfectly uniform position on the plates whilst the beam is 

' being repeatedly released and arrested, the other due to 

* Known also as Borda's, 



22 Quantitative Chemical Analysis. 

insensibility. As a balance, however sensitive, requires 
some weight to make it turn, a difference equal to this turn- 
ing weight may exist between weights apparently equal. 
The error due to insensibility may be eliminated by weigh- 
ing the object several times in succession, since the balance 
ought to turn as readily one way as the other. 

The weights should be placed on the weight-pan methodi- 
cally, and not taken haphazard from the box. A very little 
experience is sufficient to tell, roughly speaking, the weight 
of an object. Let us suppose that we wished to determine the 
weight of a platinum crucible which we afterwards found to 
be 26*715 grams. We place on the weight-pan the 20 
gram piece and release the beam this proves too little ; we 
add one of the 10 gram pieces this is too much ; we sub- 
stitute the 5 gram piece for the 10 gram the weight is again 
too little; we add the 2 gram piece this is again too much; 
we substitute one of the i gram pieces for the 2 gram piece 
the weight is too little ; we add the 0*5 gram, still too little; 
also the 0*2 gram, still too little, although we observe that 
the rate of the vibration of the beam becomes much slower 
the balance is rapidly approaching equilibrium. It is at 
this point that the skill of the operator comes into play 
an expert weigher, familiar with the indications of his instru- 
ment, can almost intuitively tell, from the extent and rapidity 
of the vibrations, what weight is required to establish equili- 
brium. Until this experience is acquired it will be well to try 
the addition of the remaining weights in the methodical 
manner above indicated. We find that 267 grams is not quite 
sufficient to equipoise the crucible ; we add 0*1 gram too 
much (the pointer swings with increased energy in the 
opposite direction) ; substitute 0-05 for the o-i gram still 
too much; try cro2 the pointer vibrates much more slowly, 
but still indicates that the weight is too great ; we substitute 
o'oi gram for the 0*02 gram the pointer swings with the 
same slowness, but shows an equal deflection to the opposite 
side of the zero ; we add '005 gram the pointer now makes 
excursions of equal amplitude the balance is in equilibrium. 



The Operation of Weighing. 23 

Having assured ourselves that such is the case by reading 
off the position of the pointer at the end of the vibrations 
several times in succession, we finally arrest the motion of 
the beam, and determine the aggregate value of the weights 
employed, (i) from the vacant spaces in the box, and (2) 
from the denominations on the weights themselves. This 
double reading should never be neglected : the one method 
serves to control the other. 

It must be carefully borne in mind that whenever an 
exchange of weight is necessary, the motion of the beam 
must be arrested : under no circumstances must anything be 
removed from the pans when the balance is free to oscillate. 
Neglect of this precaution will quickly ruin the instrument. 

The difficulty of transferring and reading the weights is 
greatly increased when we arrive at such minute fractions as 
the milligram. In order to obviate the inconvenience in- 
separable from the use of a number of small weights, which 
are apt to be lost or erroneously read, Berzelius proposed 
to estimate the last minute fractions (the milligrams and 
tenths of a milligram) by a small movable weight sliding 
along the beam. A piece of brass gilt, aluminium, or 
palladium wire, weighing exactly i centigram, is bent in the 
form represented in A, fig. 7 ; this weight is called a rider , 
and by means of the movable rod (fig. i ) worked from the 
outside of the balance case it can be placed on one arm 
of the beam (that to which the weight-pan is suspended) at 
any required distance from the centre edge. The arm from 
the centre to the terminal edge is divided into ten equal 
parts, and each of these is (generally) subdivided into five 
equal parts. The rider weighs exactly i centigram or 10 
milligrams when placed just above the terminal knife-edge. 
When acting at a point on the arm exactly midway between 
the two edges, it exerts only half this effect in other words, 
it now becomes equivalent to 5 milligrams ; when acting at 
a fourth of the distance from the centre edge, it is equal to 
2-5 milligrams, at three-quarters the distance 7^5 milligrams, 
and so on. The employment of this very simple contrivance 



24 Quantitative Chemical A nalysis. 

is attended with much economy of time and trouble, and 
with greater accuracy, as the rider allows the minutest varia- 
tion to be estimated, and diminishes the chance of error in 
reading the weights. 

The subdivisions of the centigram may also be rapidly 
estimated by means of the arrangement represented in fig. 9. 

FIG. 9. 





To the bottom end of the rod carrying the screw c is fixed 
a small movable pointer d, which moves over a graduated 
arc. The pointer can be pushed along the arc by means 
of the arm / worked by the milled-head screw G, placed on 
the outside of the balance case. This pointer acts as a 
small weight, its value depending on its proximity to the 
plane of the beam. The balance-maker effects the gradua- 
tion of the arc by placing a centigram weight upon one 
pan, and moving the index in the direction of the opposite 
pan until equilibrium is established ; one milligram of the 
weight on the pan is now removed, and the position of the 
index along the arc again noted when equilibrium is re-esta- 
blished. Successive milligrams are removed, and the position 



The Operation of Weighing, 5 

of the pointer required to bring about equilibrium is repeatedly 
determined. The divisions are then subdivided into tenths, 
each representing 0*1 milligram. 

The following general rules may prove serviceable to the 
student in weighing : 

i. Before commencing to weigh, see that the rider hangs 
upon the projecting pin of the sliding rod, and not upon the 



2. Next test the equilibrium of the balance by cautiously 
lowering the supports and setting the beam in oscillation. 
If the balance is not in equilibrium, seek for the cause of 
disturbance, and brush the pans, &c., with a camel's-hair 
brush, and again test the equilibrium before attempting to 
move the vane or alter the terminal screws. 

3. When it is necessary to arrest the motion of the beam, 
or to transfer any of the weights, or to remove the body 
weighed, the supports should be raised the moment that the 
pointer is opposite the zero of the scale. The screw regu- 
lating the eccentric should be gently turned as the pointer 
travels towards the zero, so that immediately they are in 
coincidence, a very slight but rapid turn may arrest the 
beam without any jerking or vibration. If the screw is 
suddenly turned when the pointer is at the end of an excur- 
sion, the beam will be jerked, and its original position of 
equilibrium, as shown on the ivory scale, will inevitably be 
disturbed. Carelessness in arresting the oscillation of the 
beam greatly interferes with the uniform behaviour of the 
instrument, and with inexperienced operators is the most 
frequent cause of disarrangement. 

4. All shaking of the room or table on which the balance 
stands should be carefully avoided. The operator should be 
seated so as to have the ivory scale in direct line of vision, 
but he should also be able to remove the weights, &c., 
readily, and to work the sliding rod carrying the rider. 

5. If the balance is very nearly in equilibrium, it occa- 
sionally happens that it refuses to vibrate immediately after 



26 Quantitative Chemica 

releasing the beam, and the pointer remains stationary. By 
gently wafting the air down upon one of the pans, vibration 
to the required extent may be readily set up. In very 
accurate weighing it is not always desirable to reopen the 
doors of the case for this purpose : vibration may then be 
brought about by gently touching the beam or top of the 
rider with the pin of the sliding rod. 

6. As a general rule, the substance to be weighed should 
never be placed directly upon the object-pan, but should be 
contained either in a crucible or on a watch-glass. Owing 
to its hygroscopic nature paper cannot be used if the exact 
weight of the body is of importance (and when it is not 
ordinary scales should be used instead of the chemical 
balance). 

7. A body should never be weighed when warm. By 
placing a warm object on the pan the indications of the 
instrument are affected to a marked extent. The ascending 
air-current produced acts against the object-pan and beam 
above it, and the body appears to weigh less than it ought 
to weigh. The warm air, after a time, also affects the portion 
of the beam against which it strikes, and by increasing its 
length disturbs the equality in the arms. All substances 
condense upon their surface a certain amount of air and 
moisture, the weight of which depends upon their tempera- 
ture. From this cause also the weight of a body when 
warm is always less than when cold. A silver crucible, 
weighing when cold 38-880 grams, was heated over a lamp, 
and placed whilst hot on the pan ; it now appeared to 
weigh only 38*835 grams : weighed again when cold, it 
regained its original weight. If the crucible had contained 
0-5 gram of a body to be heated, a loss of 0-045 gram, or 
9 per cent., would have apparently occurred, when in fact 
the weight of the body might have been unchanged. 

A platinum crucible when rubbed with a dry cloth and 
immediately weighed always weighs sensibly less than after 
half an hour's exposure to the air of the balance case, owing 
to the condensation of the air upon its surface. It is ad- 
visable, therefore, to allow the crucible, if freshly wiped, to 



The Operation of Weighing. 27 

remain upon the balance pan or in the case some little time 
before being weighed. 

8. Hygroscopic substances must be weighed in well- 
covered crucibles, or in stoppered tubes, or between watch- 
glasses. Liquids must be weighed in covered vessels or in 
stoppered flasks. Expedients required in particular cases 
will be mentioned hereafter. 

Within certain limits, the deviation from the horizontal 
in a balance is proportional to the weight causing it. 
Advantage may be taken of this fact to estimate the last 
fractions of a weight (i.e. the parts of a milligram) with 
great accuracy. When the balance is very nearly in equili- 
brium, it is caused to oscillate, and the position of the 
pointer when at rest determined from successive observa- 
tions of the extreme points of the vibrations. So soon as 
the excursions of the pointer fall within a certain limit, 
their extent commences to decrease at a regular and uniform 
rate. Let # a^ a 3 ..... be the extreme points con- 
secutively reached by the pointer in its oscillations ; then the 
equilibrium of the balance x is 



r = - 



n 



If we know the weight corresponding to a given deviation 
from the zero, the estimation of the minute fraction 
required for exact equilibrium becomes an easy problem. 
We have first to determine the values of one division of the 
graduated scale (i.e. the weight required to make the pointer 
deviate one division) for varying loads. The balance, having 
been adjusted as nearly as possible, is made to oscillate, and 
the extreme positions of the pointer in its excursions observed 
through a telescope. An odd number (conveniently seven) 
of successive readings are made so soon as the pointer reaches 
division 6 on the scale : the arithmetical mean of the half- 
differences between consecutive pairs of observations gives 
the position of rest of the pointer along the graduated 
scale. 



28 Quantitative Chemical A nalysis. 

In an actual determination of the position of rest (x) the 
following readings were made. The deviations of the pointer 
from the zero are marked + when they occur to the left of 
the observer, and when they occur to his right : 

+ 5'5 

- 4-5 + 0-50 
+ 5-0 + 0-25 
4-0 + 0-50 
+ 4-5 + 0-25 

- 3'5 + 0-50 
+ 4 f o + 0-25 

* = ^ = + ' 375 

An overweight of 0-5 milligram is then made to act on the 
beam, the balance again set in oscillation, and successive 
readings again taken : 

Example : +8-5 

- 2-0 + 3-25 
+ 7'5 + 275 

- 1-0+ 3-25 
+ 67 + 2-85 

- 07 + 3-00 

+ 6'2 



6 

The overweight is then removed, and the position of 
equilibrium again determined : the second determination 
usually differs to a slight extent from the original one, owing 
to unavoidable variations in the relative positions of the 
plates and edges. The mean position is therefore taken as 
the true point. In the case cited the second determination 
gave + 0-260 : accordingly the mean point is + 0-3 1 7. 
Then the deviation due to the overweight of 0-5 milligram 
would be 2'975 0*317 = 2-658 divisions; or i division 
of the scale would be equivalent to 0-188 milligram (). An 
overweight of i milligram is next caused to act on the beam, 
and the balance is again made to vibrate. This weight 
ought to produce double the amount of deviation caused by 



Weighing by Vibrations. 29 

the o'5 milligram : if any difference is observed, the mean of 
the two observations is taken as representing the true value 
of S. The determination of 5 must be made with varying 
loads, for, as already explained, the sensibility of a balance 
is seldom constant. In the instrument which gave the fore- 
going readings the sensibility increased with the load : 

Load Value of 8 

grams milligrams 

o- 0-209 

10 0'202 

50 0-188 

It would obviously be absurd to employ such a refined 
method of weighing unless we are assured that the differ- 
ences in the relative values of the weights we use fall within 
the errors of observation. However good a set of weights 
may be, the values of the several pieces are never in exact 
accordance with their denomination : the 50 gram piece, for 
example, is seldom if ever exactly fifty times the weight of 
each of the i gram pieces, nor has each of the 10 gram 
pieces exactly -1th of the weight of the 50 gram piece. The 
method of determining minute weights by vibrations, as 
above described, affords a simple means of comparing the 
pieces in a set of weights, and of estimating their true values. 
A delicate balance, placed in a room as little subject as 
possible to vibrations and changes of temperature, is care- 
fully adjusted, and the value of S determined on it in the 
manner already described (i), with the pans empty; (2), 
with a load of 10 grams; (3), with one of 20 grams; and 
(4), with one of 50 grams.* 

* In these determinations the greatest care is necessary to preserve 
the balance under perfectly uniform conditions. The operation should 
be conducted in a room (best in a cellar) set apart for the purpose. If 
the instrument is exposed to a sudden change of temperature, its equi- 
librium will almost inevitably be disturbed, owing to the unequal expan- 
sion of its arms. A rise of temperature also affects its sensibility (i) by 
increasing the distance between the centre edge and centre of gravity 
and (2) by flexure of the beam. The value of one scale division on a 
delicate balance is invariably greater in summer than in winter. 



3O Quantitative Chemical Analysis. 

A slight mark' is made with a sharp-pointed needle on one 
of the lo-gram pieces, best near its number : similarly one 
of the i gram pieces is marked ', the other is marked ". One 
of the two platinum OT gram pieces and one of the 0*01 
gram pieces have each a second corner turned up j these 
weights are respectively styled o'i' and o f oi' gram. The 
object of these markings, &c., is to enable the weights to be 
again recognised. One of the weights (say the unmarked 
10 gram piece) is considered as normal; this is placed on 
the pan to the right of the operator, and is tared with a piece 
of the same denomination from a similar set. The beam is 
then cautiously released and made to vibrate, the excur- 
sions of the pointer being observed through a telescope 
placed at a convenient distance (10 or 12 feet) from the 
balance. The position of rest x is deduced from the read- 
ings in the manner already described. The 10 gram is then 
replaced by the 10' gram piece, the balance is again caused 
to oscillate, and a second set of readings taken. The adopted 
standard is again placed on the pan, and a third set of obser- 
vations made, and again the 10' gram piece is substituted 
and another set of readings taken. From the mean 
position of rest (x') deduced from the series with the 10' 
gram piece, we determine the direction and extent of its 
difference from the adopted standard, i.e., the unmarked 
10 gram piece. The following mean results of actual read- 
ings will serve to show the degree of uniformity which may 
be expected : 

Tare v. 10 gram. Tare v. 10' gram. 

Observation x Observation x! 

I + 3-09 II + 3'10 

III + 3-09 IV + 3-16 

V + 3-09 VI + 3-11 

vii + 3-09 

Mean + 3*09 Mean + 3-12 

It appears, therefore, that the lo'gram weight is 0*03 of a 
division heavier than the normal weight. Under a load of 
10 grams one scale division on this balance corresponded to 



Weighing by Vibratiohs. 



o*32 milligram ; accordingly the 10' gram piece weighs 
10 + 0-00032 x -03 grams, or lo'ooooi grams when the 10 
gram piece is taken as normal. The set 5 + 2+1 + I'+i" 
gram pieces is in like manner compared with the standard 10 
gram piece. The following mean readings were actually- 
obtained : 

Tare v. standard 10 gram Tare v. 5 + 2 + 1 + 1' + ! 



II + 25 

IV + 2-82 

VI + 2-90 

VIII + 2-82 

Mean + 2*85 



I + 3-08 

III + 3'II 

V + 3-09 

VII + 3-10 

Mean + 3-09 

Hence it appears that the series 5 + 2 + i + i / + i" is 
lighter than the standard by 0-24 of a division, and ac- 
cordingly weighs IO- (0*00032 x 0*24) or 9-99992 grams. 
The 5 gram piece is then repeatedly compared with the 
2 + 1 + 1' + i" series, and the 2 gram piece with the i + 1' 
and i + i" and i' + 1" pieces, and so on, the higher and 
lower denominations being compared in exactly the same 
manner. The results of the comparisons are thrown together 
in a table which should accompany the set of weights : in 
using these the sum of the corrected values of the several 
denominations is taken as the true relative weight of the 
object weighed. This table may conveniently resemble the 
one annexed, which contains the results of a comparison of 
a remarkably good set of weights by Staudinger of Giessen. 
D= Denomination of weight. 
W=True relative value. 



D 


W 


D 


W 


D 


W 


D 


W 


100 


99-99971 


5 


5-OOOO2 


P*5 


0-50002 


0-02 


0-02002 


.So 


49-99971 


2 


I -99997 


O'2 


19997 o-oi 


O'OIOOI 


20 


19-99989 


I 


0-99995 


O'l 


IOOOI 


O'OI 


O'OIOO2 


IO 


lO'OOOOO 


i' 


0-99998 


o-i' 


-09999 !o-oi*i 0-00996 


10' 


10-00001 


I" 


i -ooooo 


O-O5 


05001 ooi* 0-01004 












* Riders 







32 Quantitative Chemical Analysis. 

The determination of the weight of a body with the 
greatest attainable accuracy is a problem of no slight 
difficulty. It not only demands on the part of the operator 
considerable skill and a thorough acquaintance with his in- 
strument, but also the knowledge of certain numerical data, 
some of which indeed can only be approximately known to 
him. Every substance immersed in a fluid is apparently 
diminished in weight by the weight of the fluid displaced \ 
and since all our weighings are made in the fluid which 
everywhere surrounds us, viz., the air, a balance carrying two 
weights in equilibrium simply shows that the weight of the 
body weighed less the weight of the air which it displaces is 
equal to the aggregate values of the weights less the total 
volume of air which they displace. Since every body dis- 
places so much air as is contained in the space it occupies, 
it follows that, in order to determine the true weight of an 
object weighed in air, we must know also : 

1. The volume of the body weighed (v). 

2. The total volume of the weights (v'). 

3. The weight of a given volume of dry air under stand- 
ard conditions (L). 

The weight of the object z>L=the aggregate values of 

the weights 2/L, 
or 

the true weight of the object=the aggregate values of the 
weights + VL Z/L. 

When the volume of the body weighed is equal to that of 
the weights employed, Z^L -Z/L=O: in this case only does 
the balance directly give the true weight of a body. When 
the volume of the body weighed is less than that of the 
weights, the expression (Z^L V'L) is negative : the apparent 
weight of the object is greater than its real weight. On the 
contrary, when the volume of the body weighed is greater 
than that of the weights, the apparent weight is less than the 
real weight, since VL is greater than V'L. 



Displacement of A ir. 3 3 

We can determine v and v' either from the linear dimen- 
sions of the bodies, or more easily, and more accurately, from 
their densities. The value of L requires correcting for varia- 
tion in temperature, pressure, and atmospheric moisture. 
We have therefore to observe the thermometer, barometer, 
and hygrometer at the moment of weighing : v and v' are 
also not invariable but are dependent on the temperature at 
the time of weighing ; we require therefore (when the greatest 
possible accuracy is desired) to correct for their expansion. 

It is seldom necessary to correct the indications of the 
balance to this extent, since it is only in the estimation of 
the combining weights of the elements, and in certain 
physico-chemical determinations that such extreme accuracy 
is needed. In such cases Table XI. in the Appendix will 
be found useful. The scope of this work will not permit of a 
fuller discussion of the precautions and corrections necessary 
in the exact determination of weight. We would refer for 
more complete information to a memoir by Bessel, in the 
' Astronomische Nachrichten,' vol. vii., or to Schumacher's 
paper, 'Ueber dieBerechnung der beiWagungenvorkommen- 
den Reductionen ' (Hamburgh, 1838), in which ^principles 
of the corrections are very fully explained : the physical data, 
however, need revision. In Kuppfer's work, * Travaux de la 
Commission pour fixer les Poids et Mesures de Russie ' (St. 
Petersburg, 1841), is given a full account of the method of 
vibrations; and, lastly, in Prof. W. H. Miller's classical 
memoir, * On the construction of the New Imperial Standard 
Pound' ('Phil. Trans.' Part III. for 1856), the best manner 
of conducting Gauss's method of reversal is described ; the 
tables of correction therein given are based on the most 
accurate data. 

GENERAL PRELIMINARY OPERATIONS. 

Before we commence any quantitative investigation it is 
desirable that we should have a clear conception of its 
object that we should understand the question our inquiry 
is intended to settle. If we steadily bear in mind the 

D 



34 Quantitative Chemical A nalysis. 

reason of our labour we shall be guided in the proper 
selection of the specimen of the substance which we desire 
to analyse. Supposing that we have a mineral, and wish to 
determine its composition with the object of elucidating its 
constitution, we ask ourselves in the outset would an analysis 
of this particular specimen afford a proper solution of the 
question ? We examine it with a lens, or by some other 
appropriate means, to learn if it is free from foreign 
matter ; if it is imbedded in a matrix, we carefully remove 
the adhering gangue ; we then break up the mineral, and 
select the cleanest and apparently purest pieces ; in short, 
we do everything that the nature of the case suggests to 
assure ourselves that we have an individual body to analyse, 
and not a mixture of substances. 

Again let us suppose that we are called upon to examine a 
cargo of ironstone, or other heterogeneous mass, with the view 
of ascertaining its value. We should not content ourselves 
with examining the first lump of the mineral which came 
under our notice, but we should carefully select and mix a 
sufficient quantity from various parts "of the mass, reduce 
the mixture to coarse powder, thoroughly intermix it, and 
then take a portion for analysis. 

In the plan of instruction given in this book, the first 
work of the beginner in quantitative analysis is the determi- 
nation of the constituents of simple and definite compounds 
of which the composition is already established. One of the 
objects of these exercises is to afford him the means of 
gauging his progress in manipulative skill. To this end the 
substances to be analysed must actually contain only those 
constituents they are represented to contain ; in other words, 
they must be pure. If their purity cannot be guaranteed, 
the main object of these exercises is missed : the student is 
not in a position to compare his experimental results with 
the supposed composition of the body, and from the want of 
a sure control he fails to acquire that degree of proper con- 
fidence in his work which every operator ought to possess. 



Mechanical Division. 35 

Many of the operations of quantitative analysis are of 
continual recurrence. They may therefore be most conveni- 
ently described in this introductory part. 

Mechanical Division. In order to render the substance 
we wish to analyse more susceptible to the action of solvents 
or fluxes, it is generally desirable to reduce it to a more or 
less finely-divided condition. This operation is usually con- 
ducted in mortars. The kind of mortar to be employed 
depends upon the hardness of the substance ; in all cases 
the material of the mortar and pestle must be considerably 
harder than the body to be powdered, otherwise the sub- 
stance to be analysed will be inevitably contaminated with 
the material of the mortar. In general, smooth porcelain 
mortars suffice for pounding salts, whilst most minerals 
require to be powdered in mortars made of agate. The 
agate mortar and pestle should be free from cracks or 
crevices : the pestle may be conveniently inserted into a 
wooden handle, which renders it much easier to use. Very 
hard substances should first be broken into small pieces by 
wrapping them in paper, and striking them with a hammer 
upon a smooth surface of iron. The pieces should then 
be reduced to coarse powder in the steel mortar (fig. 10). 
a is a solid block of steel, into the slight 
cavity at the top of which fits the hollow 
cylinder b ; in this cylinder are placed the 
pieces of the substance to be crushed. The 
solid pestle c is then placed in the cylinder, 
and repeatedly struck with a hammer until 
the pieces are sufficiently broken up. Some 
minerals which suffer no change on ignition 
(e.g. quartz containing gold) may be disinte- 
grated by being repeatedly heated and 
thrown into cold water. 

In order to obtain the complete decomposition of many 
minerals and insoluble bodies by the action of fluxes, it is 

D 2 




36 Quantitative Chemical A nalysis. 

necessary to reduce them to the finest possible state of sub- 
division. This was formerly frequently effected by the 
process of elutriation. The substance was triturated with a 
little water in an agate mortar, and the pasty mass thrown 
into a quantity of distilled water. After settling for a minute 
or two the turbid liquid was poured into another vessel, and 
the subsident portion was rinsed back into the mortar and 
again triturated ; this process being repeated until a sufficient 
amount of the suspended substance was obtained. After 
standing for a few hours, to allow the finely-divided matter 
to subside, the supernatant liquid was decanted off, and the 
powder thoroughly dried. This process is now less frequently 
employed in quantitative analysis than formerly, for the 
reason that it is found that very few substances are entirely 
unacted upon by water; even finely-divided felspar and 
granite give up a portion of their constituents, and 
powdered glass loses weight considerably when thus 
treated with water. In the case of mixed substances it very 
generally happens that some portions are more easily 
reduced to powder than others, and that some have a very 
different specific gravity from others ; hence it is always 
doubtful if a complex substance after elutriation has exactly 
its original composition. 

The majority of bodies may be obtained sufficiently 
finely divided by patient pounding and careful sifting. The 
sifting is best effected through fine cambric or muslin. A 
piece of the clean and dry fabric is tied over a beaker, about 
10 centimetres in diameter, and the powder is thrown upon 
the cover, which is then gently tapped with a glass rod. That 
which fails to pass through the cover is again triturated and 
sifted, the process being repeated until the entire mass has 
passed through into the beaker. The powder is again to be 
returned to the mortar in small portions at a time (using not 
more than will cover a sixpenny piece), and ground until 
every trace of grittiness has disappeared, and the substance 
cakes in an impalpable dust round the pestle. 



Desiccation. 



37 



FIG. 



Desiccation. It has already been stated that before we 
can proceed to analyse a substance we must be assured that 
it is free from all unessential constituents. The most 
frequent of these unessential constituents is moisture, by 
which term we also understand the water over and above 
that which may be proper to the constitution of the com- 
pound. This mechanically-held water may be due to the 
method by which the body has been prepared, as in the 
crystallisation of salts, or it may be derived from the atmo- 
sphere, as in the case of certain minerals. The majority of 
substances require to be dried before they can be analysed 
quantitatively. The method by which this can be most pro- 
perly and readily accomplished depends upon the nature of 
the body. If the substance contains water of crystallisation, 
repeated pressure between folds of filter-paper often suffices 
to remove the moisture. Occasionally it will be better to place 
the finely pulverised body in an 
artificially dried atmosphere, over 
some hygroscopic substance, such 
as calcium chloride or strong sul- 
phuric acid. Fig. 1 1 represents 
a convenient form of drying appa- 
ratus or desiccator^ as it is often 
termed. It consists of a glass 
bell-jar with ground and greased 
rim, resting on a plate of ground 
glass ; the dish is partly filled with 
strong sulphuric acid ; the tripod 
may be made of glass rod, and 
the circular plate to support the 
dish or crucible containing the 
substance of thin wood or metal. It is sometimes desirable 
to hasten the desiccation by conducting it under diminished 
pressure ; the apparatus has therefore an arrangement to 
connect it with the air-pump or other instrument for pro- 
curing a vacuum. 




FIG. 12. 



3 8 Quantitative Chemical A nalysis. 

Fig. 1 2 represents a more portable form of desiccator ; 
it is especially convenient for allowing hot crucibles, &c., 

to cool in a dry atmosphere pre- 
paratory to weighing them. The 
lid and lower portion are of glass, 
ground together, their perfect con- 
junction being secured by a slight 
film of grease. A brass rim, fitting 
into the aperture of the lower 
vessel, which contains sulphuric 
acid or calcium chloride, carries 
a triangle to support the crucible, 
&c. 

Substances experiencing no alteration in the neighbour- 
hood of 100 may be more quickly dried in the steam-bath. 

FIG 13. 





Fig. 13 represents a simple and convenient form of this 
apparatus ; it consists of a chamber surrounded on five 



Desiccation. 



39 



sides by an outer case of sheet copper, in which the water 
is placed, and which only communicates with the air at a 
and b. Water continually drops into a to replenish that 
lost by evaporation, and the steam makes its escape through 
b. The atmosphere within the chamber may be renewed 
through the holes c c in the door. 

If the body bears a higher temperature without change, 
it may be heated in the air-bath. Fig. 14 represents a very 

FIG. 14. 





simple form of this apparatus ; it is made of sheet-copper, 
and is heated by the lamp /, the flame of which should be 
surrounded by an earthenware cylinder, indicated by the 
dotted lines in the figure. The substance to be dried is 
placed on the shelf within the chamber, the temperature of 
which is given by the thermometer /. In certain analytical 
operations it is desirable to maintain the bath at a constant 



4O Quantitative Chemical Analysis. 

temperature for a considerable time ; the flame must there- 
fore be kept of a constant size, and be corrected for varia- 
tions in the pressure of the gas. It is very convenient when 
the bath itself can be made to regulate its temperature, so 
that if it becomes over- or under-heated it can momentarily 
cut off or increase the supply of gas. The apparatus shown 
in the figure is fitted with one of the many forms of regu- 
lators which have been described. The U-shaped tube 
contains mercury, into which dips a tube connected with the 
gas supply ; the gas from this tube passes through a narrow 
slit, thence up a glass tube, through the side-tube r, with 
which the caoutchouc tube of the lamp is connected. By 
means of a loosely-fitting screw, the gas- supply tube can be 
raised or depressed within the mercury, so that the length 
of the slit, and therefore the amount of gas passing through 
the apparatus for a given pressure, can be varied by 
surrounding it with more or less mercury. The other end of 
the U-tube is connected with a reservoir of air a, placed in 
the bath. If the screw is so regulated as to maintain a 
given temperature when the reservoir is heated, any increase 
or diminution of this temperature will be accompanied by a 
proportional increase or diminution in the volume of air 
withiri this reservoir, and a corresponding rise or fall in the 
height of the mercury surrounding the orifice through which 
the gas issues to the lamp. By means of this arrangement a 
uniform temperature within the bath (within 2 at 150-! 70) 
may be readily maintained. 

Weighing out the substance. Having obtained it in a fit 
state for analysis, and having fixed upon the scheme of 
separation to be adopted, the student next weighs off a 
certain amount of the substance, and proceeds to treat it in 
accordance with the requirements of his plan, No exact 
general rules can be given as to the amount which will be 
required for the analysis, since so much depends upon the 
nature of the body, and the proportion of its several con- 
stituents. The greater the amount taken, the more accurate, 



Weighing- out the Substance. 41 

c&teris paribus, should be the analysis, since the unavoidable 
errors in precipitating, washing, and weighing do not exercise 
the same degree of influence on the final result, when the 
quantity of the substance is large, as when it is small. On 
the other hand, the smaller the quantity operated upon the 
sooner will the analysis be finished, but at the same time 
the demand upon the manipulative skill of the operator will 
be increased. The object for which the analysis is required 
can alone tell us how far we should sacrifice accuracy to 
time. As the student will glean from the following examples, 
no strictly uniform plan can be given of the manner in 
*vhich substances should be weighed off for analysis; in 
general, however, the body, especially when in the state of 
powder, is weighed out from tubes. The light tube con- 
taining the body, and fitted with a good cork, is accurately 
weighed, the tube is removed from the pan, the cork is with- 
drawn, and the proper quantity of the substance cautiously 
shaken out into a beaker or. crucible, as the case demands ; 
the cork is now replaced, and the tube and its contents are 
again weighed. The difference between the two weighings 
gives the amount of the body taken for ^analysis. 

The further treatment of the substance depends upon the 
nature of the constituent or constituents to be estimated. 
The experience to be gained from the examples which 
follow will suggest the proper methods. It most frequently 
happens that the body is to be brought into a state of 
complete or partial solution, and the constituents separated 
either by evaporation or by precipitation, or by both pro- 
cesses. Thus we can determine the nitre in gunpowder by 
treating that substance with water, whereby the salt is 
dissolved, separating the solution from the undissolved 
portion, evaporating it to dryness, and weighing the residue. 
We can analyse common salt when in solution by precipita- 
ting the chlorine by the addition of silver nitrate, and 
weighing the silver chloride produced : the solution still 
contains the sodium (now as sodium nitrate) ; this, after 



42 Quantitative Chemical Analysis. 

the removal of the excess of the silver by appropriate 
means, can be obtained by evaporation. Precipitation can 
only be resorted to when the precipitate is practically inso- 
luble in the liquid in which it is formed, and when, pos- 
sessing a constant composition, it admits of being freed 
from foreign substances, and of being accurately weighed. 

Evaporation. Liquids are generally concentrated by evapo- 
ration in porcelain basins placed over a lamp, care being 
taken that the solution never enters into actual ebullition, as 
this would occasion loss by portions being projected from 
the dish. Unless the evaporation is conducted in a room set 
apart for the purpose, it will be advisable, indeed actually 
necessary, if many persons work together in the laboratory, 
to protect the liquid from dust. A piece of glass rod bent 
before the lamp in the form of a triangle, and covered with 
a sheet of filter-paper, and supported on a stand over the 
dish, forms an efficient shield (fig. 17, p. 45). In the evapo- 
ration of acid liquids it must not be forgotten that the 
fumes may dissolve out the inorganic constituents of the 
paper (iron, lime, &c.) and the condensed vapour dropping 
back into the dish may contaminate the liquid with those 
substances. In such cases the paper used must be freed 
from these soluble matters by treatment with acid in the 
manner to be hereafter described. 

Liquids containing gas, which is evolved in bubbles on 
the application of heat, are very liable to sustain loss by 
spirting. In such cases the dish should be covered with a 
large watch-glass, and the liquid should only be gently 
heated so long as the evolution of gas continues. When 
it has ceased, the projected portions maybe rinsed from the 
watch-glass back into the dish. The evaporation of such 
liquids may also be conducted with less chance of loss in 
obliquely-placed flasks, which should only be half filled : the 
portions spirted strike against the upper part of the flask, 
and are washed back again into the main body of the liquid 
by the condensing steam. With the flask so tilted the liquid 



Evaporation. 



43 



may even be gently boiled, with a very remote chance of 
anything being projected. 

Occasionally a liquid has to be evaporated to dryness in 
a platinum crucible, in order that the residue may be 
weighed. If the boiling point is much higher than that of 
water, the evaporation is best conducted by heating the 
crucible, placed obliquely, in the manner seen in fig. 15, 
the heat being directed upon the crucible above the level 



FIG. 15. 




FIG. 1 6. 




of the liquid, By placing the lid in the position indicated 
in the figure the evaporation is materially accelerated since 
a current of air is thus caused to play over the surface of the 
liquid. A little piece of wire gauze placed on the top of the 
lamp, enables the smallest flame to be produced without any 
fear of the gas igniting within the tube. This method of 
evaporation from the surface is especially serviceable if the 
liquid contains a precipitate ; by heating the crucible at the 
bottom it is almost impossible in such a case to prevent loss by 



44 Quantitative Chemical A nalysis. 

succussion or bumping. It is also useful when the heated solu- 
tion has a tendency to creep up the sides of the crucible ; the 
liquid evaporates as it ascends, and meets the heated surface, 
and the residue is prevented from passing out over the rim. 
Or the crucible may be placed in a vertical position with the 
lid partially over the side, so that a small flame placed beneath 
the lid heats the extreme end to dull redness. By conduc- 
tion the whole lid becomes hot, and radiates sufficient heat 
to effect a tolerably rapid evaporation of the liquid. 

But as a rule it is safer to conduct the evaporation of 
liquids over the water-bath. Fig. 16 represents one of the 
simplest forms of this apparatus ; it consists of a vessel of 
sheet-copper, about 1 5 centimetres in outside diameter, par- 
tially filled with water, and set over a lamp ; the vessel to be 
heated by the steam is placed on the top. The feath is fur- 
nished with a number of flat rings of various diameters, 
adapted to receive vessels of different sizes. In order to 
guard against the effect of inadvertent evaporation of the water 
in the bath, the apparatus, as represented in the figure, has 
a simple contrivance for turning off the gas when the copper 
basin becomes dry. The lamp is provided with a cock, the 
lever of which is prolonged and weighted with lead : it is kept 
in position by a piece of thin thread passing over the rim of 
the basin, and attached to a hook at the bottom. When 
the basin becomes dry, the thread chars, and breaks, and the 
lever falls and shuts off the gas. 

It is far better, however, so to arrange the bath that 
the water is continually replenished. The apparatus seen in 
fig. 17 is designed with this object. The water flows in 
from the main at a ; by raising or depressing the glass tube 
which slides watertight through a piece of caoutchouc tube 
slipped over the lower portion of a, any required height of 
water may be obtained in a, and accordingly in the bath. 
The overflow runs through />, and may be carried away by 
a piece of attached caoutchouc tube. This bath has also 
a number of flat rings, to suit vessels of various sizes. 

Bunsen has devised an excellent form of water-bath, which, 



Evaporation. 



45 



when once regulated, necessitates no attention on the part 
of the operator in regard to the water supply. It is repre- 
sented in fig. 1 8, p. 46. 

The bath A is made of sheet copper, and is partially 
filled with water, which is heated by the lamp a, the flame 

FIG. 17. 




of which passes into the chimney shown in the figure 1by 
dotted lines. The fresh water enters from the apparatus B. 
This consists of a wide glass cylinder, nearly filled with 
water, in which is a float, the l6wer cylindrical part of 
which contains mercury, whereby it is maintained in a ver- 
tical position, and at a certain height in the water. Through 
the upper opening dips a tube/, connected with the water 
supply ; this tube is fastened to the cylinder, but the depth 



46 Quantitative Chemical Analysis. 

to which it dips into the float can be so regulated that at 
the proper level of water the float rises, and the mercury 
cuts off the entrance of the water. The water-bath is con- 
nected with the cylinder by means of the tube e ; as the 
water evaporates, its level in B sinks, whereby the float falls, 
until the end of the tube /is uncovered by the mercury ; 
water now enters and flows over into B, and the float, again 

FIG. 18. 




rising, shuts off the water. This apparatus is more espe- 
cially adapted to a large laboratory, since any number of 
the water-baths may be connected together, one cylinder 
and float serving to replenish them all, without waste of water. 
It must not be forgotten that the material of the vessel in 
which the evaporation is conducted, is, in general, more or 
less attacked by the liquid ; it has already been stated that 



Evaporation. 47 

even pure water dissolves very appreciable quantities of 
glass. The influence of the matter dissolved from the 
flasks, &c., used in the operations, is too frequently lost sight 
of in quantitative analysis : there is no doubt that the results 
are affected to a greater degree than is usually supposed. 
Experiment has shown, that, in the case of new vessels, the 
amount of glass dissolved by a heated liquid is always greatest 
in the first hour, and gradually diminishes, until it reaches a 
certain amount, after which the quantity passing into solu- 
tion is, within certain limits, proportional to the time oY 
action. The amount dissolved is proportional to the surface 
on which the liquid acts, and is independent of the amount of 
liquid vaporised, so long as it is maintained at the boiling 
temperature ; that is, the mere rapidity of the evaporation is 
without influence. The amount dissolved is in proportion 
to the temperature of the liquid. 400 c.c. of boiling water 
in a glass flask of 600 or 700 c.c. capacity* dissolved in 
the first hour 8^9 milligrams ; in three hours, 14*8 milli- 
grams ; in six hours 22-5 milligrams ; in twelve hours, 32*5 
milligrams ; in twenty-four hours, 53-3 milligrams, and 
in thirty hours, 66*5 milligrams. The same quantity of 
dilute hydrochloric acid (n per cent), boiling in a similar 
flask, dissolved only 4*2 milligrams in the first hour ; 5*1 milli- 
grams in the third ; 7-3 milligrams in the sixth ; 9-4 milligrams 
in the twelfth ; and 1 7 'o milligrams in thirty hours. Dilute 
hydrochloric acid exerts much less action therefore than pure 
water. Nitric acid in like manner exerts comparatively little 
action on glass. 400 c.c. of dilute ammonia (9 per cent), 
dissolved from a precisely similar flask 67 milligrams in 
one hour ; in three hours, 15*5 milligrams; in six hours, 
25*3 milligrams ; in twelve hours, 43*9 milligrams ; in 
twenty-four hours, 84-8 milligrams, and in thirty hours, 99-6 
milligrams. It appears that the extent of action of am- 
monia-water varies very slightly with its strength, A 
solution of ammonium chloride (7 per cent.), dissolved in one 

* Composition of glass, SiO a 73-8, GaO 8-6, Na 2 O 14-0, K 2 O 0-60. 



48 Quantitative Chemical Analysis. 

hour 4-2 milligrams ; in six hours, 7-3 milligrams ; in fifteen 
hours, 9 '6 milligrams; and in thirty hours, 14-6 milligrams. As 
a general rule, liquids possessing an acid reaction, even when 
they contain salts in solution, dissolve less of the glass than 
when they have an alkaline reaction. The comparatively 
small quantity dissolved by the ammonium chloride solution 
is in a great measure due to the fact that this liquid acquires 
an acid reaction on boiling, owing to the dissociation of the sal- 
ammoniac : the liberated hydrochloric acid appears to exert 
a preservative action on the glass. Dilute sulphuric acid, 
however, exerts a marked action, twice as strong indeed as 
that of water. The amount dissolved by alkaline fluids is 
very considerable, even when the quantity of alkali is small ; 
in six hours 400 c.c. of a boiling liquid, containing i per 
cent, sodium carbonate, dissolved 34-8 milligrams ; the 
addition of - 4 V of a per cent, of caustic potash to water 
increases its action threefold. Certain salts, as ammonium 
carbonate, calcium chloride, common salt, potassium chlo- 
ride, nitre, act upon glass to the same extent as water ; 
sulphate and phosphate of sodium act much more energeti- 
cally, the action of the latter salt being six times that of 
water. Direct experiment has shown that the glass in all 
these cases is virtually dissolved ; the liquids do not extract 
any one constituent in preference to others. Very little 
difference is observed in the action of the liquids on glass 
of varying composition, but the true Bohemian glass, rich in 
silica, and poor in soda, is of all the least attacked. Porcelain 
vessels are scarcely acted upon by any heated liquids, with 
the exception of the alkalies, and even in their case the 
action is very much smaller than with glass : therefore 
vessels of the former material should invariably be used in 
evaporations, &c., whenever circumstances permit.* 

The precipitation of substances intended to be collected 
and weighed is usually effected in beakers, on account of 

* Emmerling, Ann. der Chemi und Pharm,, i$o.-2$J. 



Precipitation. 



49 



FIG. 19. 



the facility with which the bodies may be transferred, either 
to the filter or to the vessels in which they are to be 
weighed. In cases where the liquid is strongly alkaline, or 
where it has to be heated for some time, it is better to use 
porcelain basins. The separation of the precipitate from the 
liquid in which it is formed, is accomplished either by decan- 
tation or \>y filtration, or by a combination of these processes. 
In general, where the liquid has to be filtered, it is advisable 
to allow it to stand at rest for some hours after the addition 
of the precipitant, for the reasons : (i) that the complete 
separation of the substance in the insoluble form occurs only 
after some time ; and (2) that in certain cases, if thrown on 
the filter immediately after precipitation, it is apt to pass 
through the pores of the paper. Before proceeding to 
separate the liquid, the operator must invariably assure 
himself that the precipitant is in excess, by adding a few 
drops of its solution, and noting 
if any further turbidity is pro- 
duced. The clarification of the 
liquid on standing allows this 
to be ascertained with greater 
certainty. In cases where the 
precipitate forms only after some 
time, the precipitant must be 
added to a portion of the su- 
pernatant liquid, poured into 
another vessel. 

The separation of precipitates 
by decantation is but seldom 
resorted to, on account of the 
length of time which it occupies, 
and the comparatively large 
amount of water needed for 
thorough washing. If, however, 

the precipitate has a high specific gravity, and is practically 
insoluble in water, as in the case of silver chloride, metallic 

E 




50 Quantitative Chemical Analysis t 

mercury, &c., the process may be advantageously employed. 
The subsidence of the precipitate takes place most readily in a 
vessel of the form seen in fig. 1 9, p. 49 : this should be made 
of glass sufficiently thin to be heated without risk of fracture, 
since warming greatly accelerates the subsidence. The clear 
liquid is conveniently removed by a syphon, the longer limb of 
which can be closed by a pinch-cock, so that when the flask is 
replenished with the washing fluid, the syphon, being always 
filled with liquid, can again be set in action without the 
operator being under the necessity of refilling it. The precipi- 
tate is then transferred by the aid of the wash-bottle to the 
crucible or dish in which it is to be weighed, the fluid used 
in the transference being poured away, as far as practicable, 
and the precipitate dried. The decanted liquid should in- 
variably be set aside, in order to allow any of the insoluble 
substance which may have inadvertently been carried over, to 
subside : if any is detected, it is separated from the liquid in 
the manner described, and either added to the main quantity 
or weighed by itself. 

In the majority of cases, decantation and filtration are 
combined ; that is, the liquid is poured through the filter 
without disturbing the precipitate, and the precipitate, after 
having been agitated with fresh washing water, is allowed to 
subside during the time occupied by the contents of the 
filter in passing through the paper. Filter paper should 
permit of rapid filtration, and yet possess pores sufficiently 
minute to prevent the passage even of the most finely 
divided precipitates ; it should, moreover, be as free as 
possible from inorganic matter. The Swedish filter-paper, 
with the water-mark ' J. H. Munktell/ is generally con- 
sidered to fulfil these conditions in the highest degree. It 
contains about 0-4 to 0-6 per cent, of ash, consisting of 

Silica Alumina Iron Lime Magnesia 

35-16 3-84 45-06 14-09 1-01=99-16.* 

* From an analysis communicated by Mr. Walter Dearden, Owens 
College. 



Filtration. 



FIG. 20. 



The amount and nature of the ash vary, however, with 
different ' makes ' of the paper. 

The paper should be cut into niters of various sizes by the 
aid of circular patterns made of tin-plate or sheet- zinc ; 
these may with advantage have the radii 3, 4, 5, 6, and 8 
centimetres. The filters possessing these radii are respectively 
designated as Nos. 3, 4, 5, 6, and 8. The filters should be 
treated with dilute hydrochloric acid (which extracts nearly 
the whole of the inorganic matter, with the exception of 
the silica), and afterwards be thoroughly 
washed with water and dried. This 
treatment with acid may be conveni- 
ently made in the apparatus represented 
in fig. 20. The ready-cut filters are 
placed in the vessel, and covered with 
dilute hydrochloric acid (i part acid to 
20 of water), which is allowed to act 
for a few hours. On opening the pinch - 
cock at the bottom, the acid liquid flows 
away the filters are then to be repeat- 
edly washed with water until every trace 
of acid has disappeared, after which they 
may be dried in the water-bath. 

The operator must now determine the amount of ash left 
on burning the prepared filters, since this requires to be sub- 
tracted from the final weighing of the separated substance. 
A light porcelain crucible is heated over the lamp, placed in 
the desiccator and weighed when perfectly cold. The cru- 




FlG. 21. 



FIG. 22. 




cible is placed on a smooth glazed sheet of paper, fig. 21; 
one of the filters (No. 5, for example) is repeatedly folded 



E 2 



52 Quantitative Chemical Analysis. 

over in plaits ,of about i centimetre in breadth, and tightly 
rolled between the finger and thumb from one end of the 
folded length to the other until it has the form seen in fig. 
22. About half the length of a piece of platinum wire 40 
centimetres long is wrapped round the rolled-up filter, which 
is now lighted at the lamp and held over the crucible. The 
flame quickly disappears, and the paper becomes reduced 
to a mass of glowing carbon. As soon as the last spark has 
died out but not till then the flame is made to play on the 
ash held over the crucible, to complete the combustion of 
the carbon. The ash should now be white or have at most 
a reddish-gray tinge, without the least trace of blackness. 
Care should be taken not to heat the ash more strongly than 
is necessary to burn the carbon, or it may fuse to the plati- 
num wire. This more readily happens with filters which 
have not been treated with acid, and which, therefore contain 
comparatively large quantities of lime, iron, &c. The ash 
is now shaken out of its platinum cage into the crucible, and 
by tapping the wire against the rim of the crucible, any 
adhering traces are readily detached. This process is to be 
repeated with five additional filters ; the crucible is again 
placed in the desiccator, and when cold re-weighed. The 
difference between the two weighings divided by 6, gives the 
amount of ash left by a No. 5 filter. Call this amount a : it 
is easy from this to calculate the ash left by each size of 
filter. A No. 5 filter has a radius of 5 centimetres ; its super- 
ficial area is =r 2 7T or 5x5x3-14=78-5 sq. centimetres. 
Required, for example, the weight of ash of No. 8 filter (or): 
8 x 8 x 3-14 area=2oo'9 centimetres ; and 78-5 : 200*9 * 
a \ x. It is convenient to prepare a large number of such 
filters at a time, and to calculate and arrange in a little table 
for use the amount of ash left by the different sizes. 

Glass funnels should always be employed in quantitative 
analysis: the sides should be inclined at an angle of 60, and 
should be free from irregularities or bulgings; the stem 
should not be too short or too wide, and the end should be 



Filtration. 53 

cut obliquely. The size of the funnel to be employed of 
course depends upon the size of the filter required ; the filter 
must never project beyond the funnel ; it should be within 
one or two centimetres from the edge. The size of the 
filter in its turn depends upon the bulk of the precipitate to 
be filtered : the precipitate should not occupy more than 
half the capacity of the filter, or the process of washing will 
be very tedious. 

The filter paper should be carefully folded, and properly 
placed in the funnel, moistened with hot water (unless cir- 
cumstances forbid this), and pressed with the finger so as to 
cause it to fit closely to the funnel ; for the better it fits, the 
more rapidly will it filter, and the less will be the danger of 
rupture on washing. The funnel is placed in a convenient 
stand, so that the edge of the stem touches the side of the 
vessel intended to receive the filtrate. By allowing the 
liquid to flow down the side, all splashing and consequent 
risk of loss of the filtrate is avoided. The liquid to be 
filtered should never be poured directly into the funnel, but 
down a thin glass rod, the stream being so directed as to fall 
against the side of the filter ; if poured into the apex, loss by 
splashing will inevitably ensue. The rim of the vessel con- 
taining the liquid to be filtered should be slightly greased 
with lard (free from salt) ; this prevents the chance of the 
liquid running down the outside of the vessel. Whilst not 
in use, the rod is placed in the vessel containing the preci- 
pitate, or, if this ought not to be disturbed, in a little flask or 
beaker, which is afterwards rinsed out into the filter so soon 
as the whole of the precipitate has been transferred. It is 
advisable to cover the various vessels with glass plates 
during the progress of the filtration, to prevent dust falling 
into them. The plate covering the beaker in which the 
filtrate is received must of course have a small hole at the 
side to admit the stem of the funnel ; this may be readily 
snipped out by a pair of pliers, or by a key, the wards of 
which allow of the insertion of the glass. 

It frequently happens that small particles of the precipitate 



54 



Quantitative Chemical Analysis. 



firmly attach themselves to the sides of the vessel and can- 
not be rinsed out on to the filter. To remove them, the end 
of the glass rod should be covered with a short piece of thin 
unvulcanised caoutchouc tubing (about i centimetre long): by 
rubbing this against the sides of the vessel, the last traces of the 
precipitate may, generally, be readily detached. Or, instead 
of the rod coated with india-rubber, a feather may be used, the 
plumules of which have been torn away to within 2 centi- 
metres of the end ; those remaining are to be cut parallel to 
the quill and within -5 centimetre of it. In transferring 
precipitates from a basin, the little finger may be often ad- 
vantageously used to rub away any of the substance from the 
sides. In all cases it must not be forgotten that the rubbing 
instrument, after use, must in its turn be carefully rinsed. 
If the substance cannot be detached by mechanical means, 
it must be re-dissolved and again precipitated. 

The form of wash-bottle best adapted for use in quantita- 
tive analysis is seen in fig. 23 ; as the nozzle is moveable, 



FIG. 23. 



FIG. 24. 





the jet may be directed to any required spot. Fig. 24 shows 
another kind of wash-bottle with moveable nozzle : it is 
especially convenient for washing down the precipitate from 
an inverted beaker held over the funnel. The orifice of the 
nozzle should not be too large, or the amount of water re- 



Filtration. 5 5 

quired to bring a precipitate on to the filter becomes 
unnecessarily great. In washing a precipitate on the filter, 
the stream should be directed round the edge of the paper, 
and care should be taken that the force of the jet is not so 
great as to rupture the paper. Carelessness in directing the 
jet will inevitably cause portions of the precipitate to be pro- 
jected out of the funnel. The operator should also guard 
against the formation of channels in the mass of the precipi- 
tate, through which the water tends to flow without coming 
into contact with the bulk of the substance. He should never 
refill the filter with liquid until the previous quantity has 
passed through : neglect of this rule not only retards the 
process of washing ; but often occasions a turbid filtrate. As 
a rule hot water should be employed in washing ; its use 
accelerates the process greatly ; the few cases in which it is 
objectionable will be mentioned hereafter. In order that the 
heated wash-bottle may be conveniently handled, a coil of 
thick string, or some other badly conducting material, may 
be wrapped round its neck. 

Occasionally the washing is conducted with other liquids 
than water. Thus in the estimation of potash and ammonia 
the double platinum salts are washed with alcohol, and in 
the determination of magnesia and phosphoric acid the pre- 
cipitate is washed with dilute ammonia water. One separate 
wash-bottle should be employed for all the special liquids j 
as it will be comparatively seldom used, it may have only 
half the ordinary capacity. By attaching a small piece of 
caoutchouc tubing to the end of the shorter tube of this 
wash-bottle beneath the cork, cutting a slit through the 
caoutchouc to within a centimetre from the end, and stopping 
it nearly up to the slit with a small piece of glass rod, a simple 
valve is formed which effectually prevents the escape of 
the vapour of these special washing fluids some of which, 
like sulphuretted hydrogen and ammonia, are very irri- 
tating. The valve only opens by inward pressure and closes 
immediately when this pressure is withdrawn. 

We cannot too strongly impress upon the beginner the 



56 Quantitative Chemical Analysis. 

necessity of conscientiously performing the operation of 
washing ; imperfect or careless washing is a very frequent 
source of error in quantitative analysis. He should 
invariably ascertain, before he discontinues the opera- 
tion, that the liquid passing through no longer contains 
any of those substances which it is the object of the washing 
to remove : thus in the determination of chlorine as silver 
chloride, he should test the filtered wash-water by adding to 
a portion of it, collected apart in a test-tube, a drop of dilute 
hydrochloric acid ; if the silver chloride has been washed 
free from the excess of silver nitrate, not the faintest turbidity 
will be produced. 

In certain cases, however, such methods of testing the 
perfection of the washing are inapplicable. It is obvious 
that if we wash twice with a given quantity of water, 
we reduce the impurity more than if we wash once with 
double the quantity. For, let the original impurity be 
i gram, and let us add 10 grams of wash-water and filter off 
10 grams : there will then remain T ' T th of the original 
impurity. At the second washing there will remain y T th of 
that, or T ^ T of the original. If we had added the 20 
grams at once, the impurity would have been only reduced 
to ^ T . It is evident that for the same amount of wash- 
water we shall get the best result by using small quantities 
at a time, and washing many times. 

The following table gives an approximation to the smallest 
volume of wash-water, and minimum number of washings 
required, to reduce the precipitate to a given state of purity. 
It is obtained by regarding the apparent volume of the pre- 
cipitate at the bottom of the beaker or on the filter as 
consisting wholly of a solution of impure matter, which it is 
required to reduce to a certain degree of purity, by successive 
dilutions with a constant volume of water. 

Let v be the volume of the precipitate at the bottom of 
the beaker or on the filter, regarded as above, v the amount 
of wash-water used at each washing, n the number of 



Filtration. 



57 



washings. Also let * be the fraction of the original amount 
of impurity which remains after n washings, then 



Further, if w be the total volume of water employed in the 
11 washings, w=;z v, and (i) becomes 



If we make n infinite, a well known algebraical theorem gives 
W = v log a ...... (3) 

and this value of W is the smallest volume of water "by the 

use of which the impurity can be reduced to - of its original 
amount. 



! 


i 


i 


i 


100,000 


50,000 


* 0,000 


10,000 


I. 


II. 


III. 


I. 


II. 


III. 


I. 


II. 


III. 


I. 


II. 


III. 


v 






v 






v 






v 








n 


w 




n 


w 




n 


W 




n 


w 








v 












v 






0-5 2S-4 


14-2 


0-5 


2 6- 7 


i3'3 


o*5 


24-4 


I2'2 


o-s 


227 


if 4 


i 16-6 


16-6 




I 5 -6 


15-6 




143 


143 


i 


i.V3 


I3-3 


2 ID'S 


2 I'D 


2 


9- 


197 


2 


9-0 


18-0 


2 


8-4 


16-8 


3 


8-3 


24-9 


3 


7-8 


23-4 


3 


7-1 


21-4 


3 


6-6 


19-9 


4 7'i 


28-6 


4 


67 


26-9 


4 


6-1 


24-6 


4 


57 


22-9 


5 1 6-4 


32-1 


5 


6-0 


30-2 




5'8 


27-6 




ST." 


257 


6 


5 '9 


35 '5 


6 


5-6 


33'4 


6 


S'i 


30'S 


6 


47 


28-4 


7 


svs 


38-8 


7 


S'2 


36-4 


7 


4-8 


33'3 


7 


4'4 


31-0 


8 


5-2 


42-0 


8 


4'9 


39'4 


8 


4'5 


36-1 


8 


4'2 


33 '5 


9 


5*9 


45 * 


9 


47 : 42-3 


9 


4*3 


,^7 


9 


4-0 


36*0 


10 


4'8 


48-0 


10 


4*5 


45-i 


10 


4'i 


4I-3 


10 


3'8 


38-4 


ii 


4-6 


51*0 


ii 


4 '4 


47'9 


ii 


4-0 


4.V8 


ii 


37 


40-8 


12 


4'5 


5.V9 


12 


4-2 


50-6 


12 


3'9 


46-3 


12 


3-6 


43'i 


13 


4'4 


56-4 


13 4'I 


53'3 


J 3 


3-8 


48-8 


13 


3 '5 


45 '4 


H 


4-2 


59'4 


14 4-0 ! 55-8 


14 


37 


Si-i 


14 


3'4 


47'5 


15 


4-2 


62-3 


15 


3'9 SB'S 


15 


3'6 


.S3 "6 


IS 


3'3 


49 -8 


16 


4'i 


65-0 


16 


3-8 


6i'i 


16 


3'5 


56-0 


16 


3'3 


53 >0 


17 


4-0 


67-8 


17 


37 


63-6 


17 


3 "4 


58-3 


17 


3'2 


S4'2 


18 


3'9 


70-4 


18 


37 


66-1 


18 


V4 


6o-s 


18 


3-1 


56'3 


19 


3-8 


74'3 


19 


3-6 


68-6 


19 


3'3 


62-8 


19 


3'i 


58-4 



58 Quantitative Chemical Analysis. 

By taking the logarithm of formula (i) we obtain 
log a 

- 



the logarithms being common logarithms, and this formula 
enables us to find the least number of washings requisite to 

bring down the impurity to a fraction of its original 

& 

amount, by the use of a quantity v at each washing. 
The foregoing table has been calculated from it. The top- 

most line of the heading shows the fraction i. 

a 

By employing for each treatment the same volume of wash- 
water, and approximately determining the relative volumes 
of the precipitate, and of the washing liquid, used each time, 
we may obtain from the table on the preceding page, calculated 
from the foregoing formula, the minimum number of treat- 
ments required to reduce the impurity in the precipitate to 
unnnro *7mro *WOT or TOTTO a ^ its weight. Column I. gives 
the relation of the volume of the precipitate to the volume of 
the washing-fluid employed for each treatment. Column II., 
the minimum number of treatments necessary for the par- 
ticular extent of washing desired, and Column III., the total 
volume of wash-water which will be obtained. (See p. 57.) 

Let us suppose that we have a precipitate occupying a volume 
at the bottom of the beaker of thirty cubic centimetres, and 
that the amount of liquid we employ to wash it each time 

is fifteen cubic centimetres, then - is of course 0*5, and 

v 

if we wished to remove the impurities to the yffjwth part, 
we learn from Column II. of the table that we must treat it at 
least 27 ^267) times with this amount of water (viz., 15 c.c.) 
that is to say, the minimum amount of wash- water needed will 
be 399 cubic centimetres. In cases of simple decantation from 
beakers, the volume occupied by the precipitate, as compared 
with the fluid above it, may be very easily determined by 
laying a strip of paper along the side of the beaker, marking 
off the height of the precipitate, and level of the liquid, and 



Filtration. 



finding the number of times the length corresponding to the 
height of the precipitate may be folded into the length corre- 
sponding to the depth of the supernatant liquid. In applying 
this table to niters, the capacity c of these must be calculated ; 
it is given by the formula 






where r is the radius of the filter-paper. 

The following table shows the capacity in cubic centi- 
metres of various filters, placed in a funnel whose oppo- 
site sides form an angle of 60. 



No. 3 
4 


6 - i c.c. 
I4-5 


No. 5 
6 


28-3 c.c. | No. 7 

49-0 ;,, '8 

i 


77-8 c.c. 
116-1 ,, 



When the whole of the precipitate has been brought on to 
the filter, the unoccupied volume of the latter is determined 
by filling it with water from a burette. If w be the amount 

of water required to fill it, then gives the entry - in 

Col. I. of the table on p. 57, whence we obtain the minimum 
number of washings required.* 

The rapidity with which a liquid filters is proportional to 
the difference of pressure exerted on its upper and lower 
surfaces. By the ordinary method of filtration this differ- 
ence seldom exceeds six millimetres of mercury. By increas- 
ing the difference we accelerate one of the most tedious of 
the operations of quantitative analysis. The following 
arrangement effects this acceleration to the desired extent. 
A glass funnel is chosen of about 8 centimetres in depth, 
the sides of which are free from irregularities or bulgings, 
and subtend an angle of 60. The stem should be long and 

* Bunsen, Ann. der Chem. u. Pharm., vol. cxlviii. p. 269. In the 
absence of exact knowledge respecting the nature of precipitates, whether 
pervious or impervious to liquids, and in what degree, or whether 
different liquids have different powers of adhering to or penetrating 
precipitates, we must regard the above process as an attempt only to 
place the operation of washing upon a quantitative basis. 



6o 



Quantitative C/icmical Analysis. 



narrow, and the end should be cut obliquely. A small 
cone, i to i J centimetre in depth, of thin platinum foil or 
gauze, and having exactly the angle of the funnel, is dropped 
into the apex, and over it is fitted the filter, with all the prer 
cautions described on p. 53. The following is the readiest 
method of obtaining the platinum cone of the desired 
shape. A circular piece of writing paper, 10 or 12 centi- 
metres in diameter, is folded like a filter, and placed in the 
funnel so as to fit accurately to its sides, especially near its 
apex. It is kept in position by a few drops of sealing wax, 
and is saturated with oil by means of a feather, care being 
taken that no drops of the oil remain at the point of the 
paper cone. A thin cream of plaster of Paris is then poured 
into the paper mould, and a small handle is inserted into it 
just before the mass becomes solid. In a few hours the 
plaster cast will be dry enough to be removed from the 
funnel, together with the oiled paper. The outside of the 
paper is now thoroughly oiled, and inserted into a small 
crucible, or similarly shaped vessel, of 4 or 5 centimetres in 
height, filled with cream of plaster of Paris. As soon as the 
outer mould is dry, the plaster cone is removed, and the paper 
rubbed off it. In this manner a solid cone, fitting into a hollow 
cone, is obtained, both of which possess exactly the angle of 
FIG. 25. inclination of the fun- 

nel (fig. 25). Apiece 
of platinum foil, of 
such thickness that 
i square centimetre 
weighs 0-15 gram, is 
cut into the shape and 
size represented in 

fig. 26 : it is divided by a pair of scissors 
along the line a b, as far as the centre a. 
The foil is then held in the flame of the lamp for a few 
minutes, to render it pliable, and placed against the plaster 
cone, so that the point a is at the end of the cone ; the side 
a b d is folded against the cone, and over this is folded the 



FIG. 26. 





Filtration. 



61 



remainder, a b c, so that the foil also becomes a cone, the 
sides of which have exactly the same inclination as those of 
the plaster cast, and also of the funnel. The shape of the 
platinum funnel may be completed by dropping it into the 



FIG. 



FIG. 27. 




hollow mould, and pressing it down by means of the plaster 
cone : this shape of course may, at any time, be again given to it 
by simple pressure between the two cones. The platinum cone 
should allow no light to pass through its apex : when pro- 
perly made it will support a filter filled with liquid, under a 
pressure of an atmosphere, without the paper breaking : the 



62 Quantitative Chemical Analysis. 

small space between the folds of the foil is quite sufficient to 
allow of the passage of a rapid stream of water from the filter. 
The stem of the funnel is pushed through a caoutchouc 
cork, pierced with two holes, and fitting into a thick glass 
flask (A, fig. 27); the second hole carries a piece of glass 
tube ending immediately under the cork and leading to the 
instrument which creates the difference in pressure. This 
is also seen in fig. 27. A brass tube, a a, about i 
metre in length and 8 millimetres in internal diameter, has 
its upper end cut obliquely in the manner seen in fig. 2 7 A. 
At about 5 centimetres from the end is a side tube c of 
equal diameter and 5 centimetres long, into which is screwed 
a short piece of tube d\ the ends of this tube */are fitted 
with narrow brass tubes, e and/, 4 millimetres in diameter 
and 2 centimetres long. Over f is pushed a piece of thick 
caoutchouc tube 4 centimetres in length. This tube must 
be made of good caoutchouc : it should be about 6 milli- 
metres in external diameter, and its bore should not exceed 
2 millimetres in width. Before introducing it into <r, a piece 
of wood, somewhat wider than its bore, is pushed into it, 
and the caoutchouc is cut through to the wood by a smart 
blow on the head of a chisel, 2 centimetres broad, placed 
against the tube at 15 millimetres from the end. The wood 
is then withdrawn, and the end of the caoutchouc tube is 
stopped air-tight by a short length (i centimetre) of glass 
rod, held firmly in position by binding wire. The thick 
caoutchouc tube so cut, forms a very efficient valve, which, 
on the application of pressure from within, opens, but closes 
immediately by outward pressure. The tube being of con- 
siderable thickness in the walls, is rigid, and does not collapse 
even under a pressure of an extra atmosphere. The upper 
end of the tube a a is connected by means of a short piece 
of elastic caoutchouc tubing with the water-supply; this 
tube should be bound round with calico to within 5 or 6 
centimetres of the end near the brass tube, since it will be 
subject to considerable inward pressure. On allowing a 
sufficient amount of water to flow through, it commences 
to pulsate as the india-rubber valve intermittently opens and 



Filtration. 63 

shuts. Rapid suction is thus set up, and the instrument 
exhausts a closed vessel in a comparatively short time to 
within the pressure due to the tension of aqueous vapour 
corresponding to the temperature of the water flowing down 
the tube, plus the tension required to open the caoutchouc 
valve. The degree of exhaustion is determined from the 
height of the mercury in the manometer ;;z, which is con- 
nected with the tube d by means of a piece of strong caout- 
chouc tubing. The entire apparatus is fixed upon a board, 
which may have a foot if it is desired to move it from place 
to place in the laboratory ; or it may be fixed in a position 
where the water can most conveniently flow away.* By 
connecting the tube h with the flask holding the funnel (or 
with an intermediate vessel to which several flasks are 
attached) we diminish the pressure to which the under sur- 
face of the liquid to be filtered is exposed, so that the filtrate 
is driven with greatly increased rapidity through the pores of 
the paper ; the filter itself is prevented from being pushed 
through into the stem by the closely-fitting little platinum 
cone which supports it. 

The diminution of pressure may also be readily brought 
about by the aid of the little apparatus seen in fig. 276, which 
is specially applicable to water Fig. 27, B. 

under high pressure. The 2- 

apparatus is attached by a 



stout piece of caoutchouc tub- ,u d /, ; 
ing to the water tap; the '* 




water flowing in in the di- 
rection indicated by the ar- 
rows. When forced through 
the narrow internal tube b 
into the sharply bent fall- 
tube c c, a partial vacuum is 
created in the bulb-shaped portion, d, and hence within the 

* For an explanation of the principle of this apparatus, see a paper 
by Mendelejeff, Kirpitschoff, and Schmidt, Ann. der Chem. u. Pharm., 
January 1873. See also Jagn, Ann. der Chem. u. Pharm., Feb. 1873. 



64 Quantitative Chemical Analysis. 

filter-flask or other vessel with which it is connected by 
the T-tube e, which may also be put in connection, if ne- 
cessary, with .a manometer. In the tube / is a small caout- 
chouc valve, similar to that described in the apparatus shown 
in fig. 2 7 A, to prevent the possible reflux of the water. 

To use this apparatus for filtering, the liquid resting over 
a precipitate is cautiously poured on to the filter fitted to the 
funnel, with the precautions detailed on p. 53, the action of ,the 
pump is set up, and as the liquid flows through into the flask 
fresh portions are added until the whole has been decanted. 
The precipitate is then transferred in the ordinary manner and 
washed by the addition of water from an open-mouthed vessel, 
and not by a jet from the wash-bottle. The fluid in which 
the precipitate was originally formed, together with that 
necessary to transfer the precipitate to the filter, should be 
allowed to flow away completely before the process of wash- 
ing is commenced. Immediately after the precipitate is 
drained, but before any channels commence to form in it, the 
filter is to be filled up with water, poured cautiously down 
the side of the funnel. When this wash-water has drained 
away, the suction is continued until the precipitate is seen 
to shrink, when the filter is again filled up over the edge and 
to within i centimetre of the brim of the funnel. This process 
is to be twice repeated, after which the precipitate may be 
drained almost dry by continuing the action of the pump 
for a few minutes. This method of filtration and washing 
is exceedingly rapid as compared with the old plan, and 
requires very little wash-water by reason of the compression 
which the precipitate suffers. Thus a precipitate of chromium 
sesquioxide weighing about 0-24 gram, required i hour 48 
minutes, and 1050 cubic centimetres of water to wash it to 
within 3~ff^oir by tne ld method, whilst with the new plan 
the same amount of sesquioxide required only from 12 to 14 
minutes, and from 39 to 41 cubic centimetres of water. 
(Bunsen.) 

A further advantage attending the use of the suction 
apparatus arises from the condition of the precipitates after 



Drying and Igniting Precipitates. 65 

filtration. The chromium sesquioxide, for example, is left so 
dry, that without further desiccation, the precipitate, wrapped 
in the filter, may be placed in the crucible over the lamp, 
and, after cautiously charring the paper, maybe ignited without 
any apprehension of loss by proj ection. Many other precipi- 
tates which experience no alteration when ignited in contact 
with paper, such as ferric oxide, alumina, &c., may be treated 
in the same way. The paper being nearly dry may also be 
readily detached from the funnel ; when opened out on a 
flat surface, the coherent precipitate may easily be removed 
by means of a thick platinum wire, so as scarcely to leave a 
trace upon the filter. This ready method of removing the 
precipitate is of great value when we have occasion to treat 
it with a solvent or flux. 

This suction apparatus may be used for a variety of pur- 
poses in addition to that of filtration ; it may be employed 
as an aspirator in quantitative operations, since the volume 
of air passing through the tube can be regulated with the 
utmost nicety by the aid of the screw- clamp ; it may also be 
applied to the evacuation of desiccators or vessels in which the 
concentration of liquids in vacuo is conducted, to freeing 
crystals from mother-liquors, &c. 

Drying the precipitate. -In the majority of cases it is neces- 
sary to dry the precipitate thoroughly before it can be further 
treated with the view of determining its weight. The water 
in the stem of the funnel is removed by filter-paper, and the 
mouth of the funnel closed, to protect the precipitate from 
dust, by placing a moistened filter over it ; on drying, the 
paper adheres to the rim and makes a very efficient cover. 
The funnel is then placed in the steam-bath represented in 
fig. 13 (p. 38), and kept there until the paper and precipitate 
are completely dried. This method of drying the precipitate 
is preferable to that of supporting the funnel directly over 
the lamp, for in addition to the risk of cracking the stem, 
the latter method has the further disadvantage of causing 
the precipitate, by reason of the manner of heating, to adhere 



66 Quantitative Cliemical Analysis. 

to the paper. When dried in the steam-bath, the precipi- 
tate, in contracting, detaches itself from the filter ; so much 
so, that many curdy or gelatinous precipitates like silver 
chloride, or ferric or chromic oxides, may be almost com- 
pletely shaken out of the funnel into the crucible in which 
they are to be weighed. This ready separation of the dried 
precipitate from the paper materially conduces to accuracy 
in determining its weight. 

Igniting and weighing the precipitate. Since the precipitate 
requires to be weighed in a perfectly dry state, it is in 
general necessary to remove it from the filter and to ignite it. 
A porcelain or platinum crucible is heated, allowed to cool 
in the desiccator, and weighed. It is then placed on the 
sheet of black glazed paper, together with the platinum wire 
and feather (Fig. 21, p. 51). The filter is removed from the 
funnel, opened out, and the detached fragments of the pre- 
cipitate allowed to fall into the crucible. The portions of 
the precipitate adhering to the filter are loosened by rubbing 
its sides together, care being taken that its surface is not 
thereby destroyed, otherwise filaments of paper are apt to 
contaminate the precipitate ; these may either escape burn- 
ing in the subsequent ignition, or if burnt, may alter the 
composition of the precipitate. The detached precipitate 
is then added to the main quantity already in the crucible. 
Care must be taken to remove as much of the precipitate 
as possible from the filter ; but however carefully the opera- 
tion may be performed, a considerable amount of the sub- 
stance invariably remains, either on the surface of the paper 
or contained within its pores. When the substance suffers 
no change by ignition with carbonaceous matter, it may be 
recovered by burning the paper, and adding the ash to the 
crucible. The known weight of the filter-ash is then sub- 
tracted from the increase in the weight of the crucible. 
The filter is burnt in the manner described on p. 51. The 
paper should be so folded that the soiled half of the filter is 



Weighing Precipitates. 67 

in the centre : there is thus less chance of loss from projection 
or from the precipitate fusing to the heated platinum wire. 
The ash is shaken into the crucible, which is then ignited, at 
first gently, and with the lid on ; afterwards more strongly, and 
with the lid removed. The degree and duration of the heat 
depend, of course, on the nature and amount of the pre- 
cipitate : as a general rule from five to ten minutes at a low 
red heat will be sufficient. The crucible is then placed in 
the desiccator, and weighed when cold. It must be heated 
a second time and again weighed, to ascertain t*hat its 
weight is constant. 

Some precipitates suffer change when ignited in contact 
with carbonaceous matter, or become altered in composition 
at the high temperature necessary to burn the carbon com- 
pletely. Thus silver chloride becomes reduced to metallic 
silver in contact with carbon, and calcium carbonate is 
converted into caustic lime at a red heat. In weighing 
silver chloride, for example, the precipitate is detached as 
far as practicable from the filter, and the crucible in which 
it is placed is gently heated until the chloride fuses, and 
when cold it is weighed. The paper is now folded in the 
usual way, the soiled portion being in the centre, and it is 
burned in the manner described, and the ash added to the 
fused silver chloride. The crucible is again weighed : its 
increase of weight gives the amount of the filter-ash, together 
with the quantity of metallic silver which has been reduced 
from the state of silver chloride by contact with ignited car- 
bonaceous matter. Since 108 parts of silver correspond to 
143-5 of silver chloride, the amount of silver chloride in the 
pores of the paper can be readily calculated from this 
reduced silver : it is of course added to the weight of the 
main quantity of the chloride. The cases in which it is 
necessary to weigh the filter-ash separately will be mentioned 
as they occur. 

Whenever practicable, a platinum crucible should be em- 
ployed on account of the readiness with which it may be 

F 2 




'68 Quantitative Chemical Analysis. 

heated to redness. Indeed, in some cases, its use is almost 
indispensable, as in the conversion of calcium oxalate into 
carbonate, and of magnesium-ammonium-phosphate into 
magnesium pyrophosphate. Platinum vessels, however, 
cannot be used for the ignition of silver chloride or bromide 
FlG 28 or of lead chloride. Many oxides, 

^ sulphides, and phosphides cannot 
be heated in contact with pla- 
tinum without injury to the metal. 
After prolonged ignition over a 
lamp, especially if the reducing 
portion of the flame be permitted 
to impinge upon it, the lower por- 
tion of the crucible loses its lustre 
and appears to be superficially corroded. This appearance 
is said to be due to the formation of a carbide of platinum. 
Red-hot platinum crucibles should never be touched with 
brass tongs or placed in brass rings, as black stains are thus 
formed on the metal. They are best heated on pipe-clay 
triangles or on thin platinum triangles supported on a triangle 
of stout iron wire, fig. 28. Clean iron tongs will be found 
more generally convenient than brass tongs. Platinum 
crucibles may be cleaned by scouring with moist sea-sand, 
which does not scratch the metal ; stains which cannot thus 
be removed are got rid of by heating with acid potassium 
sulphate, or borax, the crucible being afterwards thoroughly 
washed with hot water and scoured with sea-sand. 

Collection of precipitates on weighed filters. Occasionally 
we have to deal with a precipitate which cannot be ignited to 
ensure the expulsion of moisture before being weighed. The 
precipitate must be weighed therefore on the filter on which 
it is collected. Accordingly the weight of the filter itself 
must be known. The paper is folded in the usual manner, 
placed in a stoppered tube, or between well-fitting watch- 
glasses pressed together by a clip, and heated in the 
steam-chamber for an hour or so. The stoppered tube, 



Weighing Precipitates. 69 

or watch-glasses, together with the filter-paper, are allowed 
to cool in the desiccator, and weighed when cold. The 
filter is then fitted into the funnel, and the precipitate is 
brought on to it, the tube or the watch-glasses being mean- 
while left in the balance-case. The paper and precipitate 
are first dried in the funnel, the filter is then detached from 
the glass, and placed in the tube or between the watch- 
glasses, heated for some hours in the bath, and repeatedly 
weighed until the weight is constant. 

Another plan is to weigh two filters of equal size (A and 
B) against each other, and mark the difference in weight on 
B. The precipitate is collected on A, the filtrate and wash- 
ings being allowed to pass through B ; both are dried and 
weighed against each other, and the original difference in 
weight allowed for. 



70 Quantitative Chemical Analysis. 

Q 

PART II. 
SIMPLE GRAVIMETRIC ANALYSIS. 

I. COPPER SULPHATE. CuSO 4 + 5H 2 O. 

Preparation.^ order to obtain this salt in a fit state for 
analysis, it is necessary to purify it by recrystallisation. The 
blue vitriol of commerce not unfrequently contains ferrous 
sulphate, which cannot be removed even by repeated crystal- 
lisation, as the two sulphates tend to crystallise together. 
By heating the solution with a few drops of nitric acid, the 
ferrous salt is oxidised to ferric sulphate, and on concentrat- 
ing the liquid, crystals of pure copper sulphate are easily 
obtained. Two hundred grams of clean, well-developed 
crystals of the commercial salt are* dissolved in about a 
quarter of a litre of hot water, a few drops of nitric acid are 
added, the solution is filtered, if necessary, and boiled for 
half-an-hour ; on cooling the liquid deposits crystals of the 
pure salt. After standing for a few hours the solution is 
poured off, and the mother-liquor is drained as far as 
possible from the crystals. The crystalline mass is broken 
up by means of a glass rod, and dried by pressure between 
folds of filter-paper. It is advisable not to use too great a 
degree of force in pressing the salt, as the sulphate may 
thus become mixed with filaments of filter-paper, which 
interfere with the accuracy of the analytical operations. 
When the greater portion of the moisture has been removed 
by repeated pressure between filter-paper, the salt is wrapped 
in a fresh sheet of dry paper, and the folds are placed 
under a heavy weight for an hour or two. Whilst the salt is 
drying, the apparatus required for its analysis is got ready. 
Two small thin test-tubes, to hold about 6 or 8 grams of the 
salt, are cleaned and dried, and fitted with good, clean, soft 



Copper Sulphate. 7 1 

corks. A couple of beakers of 250 cubic centimetres capacity, 
and two large watch-glasses to cover their mouths, a filter- 
flask fitted with its funnel, and two thin glass rods, all per. 
fectly clean and dry, are also needed. 

The complete analysis of copper sulphate necessitates 
the determination (i) of the water of combination \ (2) of 
the copper ; and (3) of the sulphuric acid. 

i. Determination of the water of combination. Copper sul- 
phate gives up the whole of its combined water on heating ; 
4 molecules being readily expelled at 100-110, and the 
remaining molecule at about 200. The determination of 
the amount of water expelled at different temperatures 
may be made by means of the air-bath (fig. 14, p. 39), or 
in the apparatus represented in fig. 29. The test-tube 
(a) contains the tube and salt to be dried; it is about 
8 centimetres long, and 2 centimetres wide ; into it is 
placed the narrower and shorter tube, containing the weighed 
amount of salt : the tube a is closed with a cork pierced 
with two holes, into which are fitted narrow glass tubes 
bent at right angles : one of these tubes passes nearly to the 
bottom of the test-tube. The narrow tubes are connected 
by means of caoutchouc tubing with the small flasks c and d y 
containing strong sulphuric acid : the tube e of the flask d 
is in connection with the filter-pump, by means of which a 
current of air, dried by aspiration through the acid in the 
flask c, is drawn over the salt. The test-tube dips into a 
small quantity of oil contained in a beaker of 400 cubic 
centimetres capacity. The tube is held by means of a clamp 
attached to a retort- stand. The oil is heated by a small 
flame, and the temperature is ascertained by a thermometer 
placed near to the tube. 

By the time this piece of apparatus has been fitted up, 
and the beakers, &c., are cleaned, the copper sulphate 
under the weight will be dry. One of the test-tubes which has 
been fitted with a cork is nearly filled with the dried salt ; any 



72 Quantitative Chemical Analysis. 

adhering powder is wiped from the upper portion and edge 
of the tube, and the cork is replaced. The remainder of 
the copper sulphate is set aside in a stoppered bottle ; it will 
be useful for subsequent analytical operations. 

Weigh the other test-tube and cork, and introduce about 

FIG. 29. 




i -5 2 grams of the copper salt, taking care not to soil the 
edge of the tube ; replace the cork and weigh again. The 
increase of weight gives the amount of salt taken for the deter- 
mination. Take out the cork, and leave it in the balance 
case. Drop the little tube containing the salt into the wider 
test-tube of the drying apparatus, insert the cork and bent 
tubes,heat the oil-bath to ioo-no,with frequent stirring, and 



Copper Sulphate. 73 

aspirate a gentle current of air through the sulphuric acid. 
In about an hour the greater portion of the water will have 
been expelled: the little tube is withdrawn from the wider one 
by means of the forceps, allowed to cool in the desiccator, 
and carried to the balance-room. When cold, the cork in 
the balance-case is inserted into the tube, and the loss of 
water which the sulphate has suffered is determined by a 
second weighing. The cork is once more withdrawn, left in 
the balance-case, and the tube again heated in the oil- 
bath to 1 00-110, a current of dry air being swept over it 
for about half-an-hour, after which it is again weighed when 
cold, in order to determine if it has parted with an addi- 
tional quantity of water. If the second weighing is within 
0*00 1 o gram of the first, the loss may be set down as 
constant; but if the weighings differ by more than this 
amount, the tube and salt must be again heated for half-an- 
hour, and weighed a third time, the process being repeated 
until a constant weight is obtained. 

The temperature of the oil-bath is next raised to 200, 
and maintained at this point for about an hour, a 
current of dry air being passed uninterruptedly through the 
apparatus. After cooling, the loss of weight experienced by 
the salt (which is now nearly white) is again determined, 
the tube is once more replaced in the bath, again heated, 
and again weighed, the operation being repeated until no 
further loss of water is perceptible. 

Arrange the results of the several weighings in the follow- 
ing manner in your note-book, The numbers here given 
are the results of an actual determination. 

ANALYSIS OF COPPER SULPHATE. 
(Date.) i. Determination of Water. 

Tube + cork + salt 6 '3400 grams. 
Tube + cork 4 -8905 , , 

Weight of salt taken i -4495 



74 Quantitative Chemical Analysis. 

Weight of tube + cork + salt after drying : 

After i hour at 105 -112 5'934O grams 

After 30 min. additional : no 5 -9210 ,, 

io7-no 5-9180 

io7-ii2 5-9176 ,, 

Water expelled at 1 10 0-4224 ,, 
or 29*14 per cent. 

After drying for i hour at I9O-2OO 5-8250 ,, 
After further drying for half an hour at 200 5*8183 ,, 

205 5-8176 

Additional loss of water O'looo ,, 
or 6 '9 1 per cent. 

2. Determination of the Copper. Whilst the salt is drying 
in the oil- or air-bath, proceed with the estimation of the 
copper. This is effected by precipitating it as cupric oxide, 
by the addition of caustic soda solution. 

CuSO 4 + 2 NaHO = CuO + Na 2 SO 4 + H 2 O. 

The corked tube containing the copper sulphate is care- 
fully weighed, and about a gram of the salt is shaken out 
into one of the clean and dry beakers. On replacing the 
cork, and again weighing the tube, its loss of weight gives 
the exact amount taken for analysis. Care must, of course, 
be taken that all the copper sulphate removed from the 
tube finds its way into the beaker. The salt is then dis- 
solved in about 50 cubic centimetres of hot water : the 
solution should be perfectly clear, and free from suspended 
matter. It is boiled, the mouth of the beaker being mean- 
while closed by one of the large watch-glasses, in order to 
prevent the projection of any of the solution on ebullition. 
The lamp is drawn aside, and a clear solution of caustic soda 
is poured into the liquid, drop by drop, down the side of the 
beaker, the liquid meanwhile being kept in agitation. A preci- 
pitate is at once formed ; it is at first of a bluish green colour, 
but it rapidly darkens as it falls through the hot liquid, and 
ultimately becomes nearly black. These changes of colour 
are due to the passage of the copper oxide from the 
hydrated to the anhydrous condition. The precipitate is 



Copper Sulphate. 75 

allowed to settle, when, if sufficient soda has been added, 
the liquid will be colourless. Ascertain that the alkali 
is in excess by testing the solution with a slip of red- 
dened litmus-paper. Fold a No. 5 filter, drop it into the 
platinum cone, moisten thoroughly with hot water, and fit it 
carefully to the funnel in the manner already described 
(see p. 53). Next slightly grease the edge of the beaker, and 
by means of the glass rod decant the clear liquid on to the 
filter (taking care not to disturb the precipitate of copper 
oxide) and set the pump in operation. When the whole of 
the liquid has been decanted on to the filter, pour about 
30 or 40 cubic centimetres of hot water over the precipitated 
oxide, boil for a few minutes with the glass cover on, allow 
to settle, and again pour the supernatant liquid on the filter. 
Repeat the washing by decantation and then carefully rinse 
every particle of the precipitate with hot water on to the filter. 
It may happen that a small quantity of copper oxide obsti- 
nately adheres to the beaker and cannot be removed by 
washing. Pour about 2 cubic centimetres of hot water into 
the beaker, and a couple of drops of nitric acid : by the aid 
of the glass rod, the dilute acid solution may be made to 
dissolve the adhering oxide. This is reprecipitated by the 
addition of a few drops of soda solution, and thrown on to the 
filter. Now fill up the filter five times with hot water, taking 
care to allow the whole of the wash-water to run through 
before a fresh addition is made. If these instructions are 
followed the precipitate will be thoroughly washed. The 
funnel is withdrawn from the flask, its mouth is covered with 
paper to protect the copper oxide from dust, and the whole 
is placed in the drying chamber. Whilst the precipitate is 
drying, clean, dry and heat a small porcelain crucible and 
lid (No. i size), place them together in the desiccator, and 
when cold, carefully weigh them. When the copper oxide 
is dry, it is detached as far as possible from the filter, and 
transferred to the weighed crucible. The funnel is cleaned, 
if necessary, by rubbing it with the edge of the paper ; 



76 Quantitative Chemical Analysis. 

the filter is burned, and the ash allowed to fall into 
the crucible. One drop of nitric acid is allowed to moisten 
the oxide and filter-ash, and the crucible is gently warmed 
until the mass is dry, when the heat is raised until 
the bottom of the crucible is red hot. It is allowed to 
cool slightly, and whilst still warm, is transferred, together 
with its lid, to the desiccator, and when quite cold, again 
weighed. The crucible and lid are once more heated, 
and again weighed on cooling ; care should be taken that 
the reducing gases from the flame do not find their way into 
the crucible. The second weighing ought not to differ more 
than 0-5 milligram from the first weight. If the difference 
is greater, the operation must be repeated until a constant 
weight is obtained. The increase in weight of the crucible 
and lid gives the amount of copper oxide contained in the 
quantity of salt taken for analysis plus the ash of No. 5 filter. 
The details of an actual determination will show how the 
results ought to appear in the note-book : 

ANALYSIS OF COPPER SULPHATE. 
2. Determination of Copper by precipitation with Caustic Soda. 

(1) Weight of tube + salt . . 10-6052 

(2) _9J8o5_ 

Salt taken . . I '0247 

Weight of crucible + lid + CuO + ash (i) 8 1 530 

(2) 8-1527 

Crucible + lid . . . 7-8240 

0-3287 

Less filter-ash No. 5 0*0023 
0-3264 

Calculation* ^'i x 0-3264 x IOO log 63-1 j_-8ooo 
79-1x1-0247 log -3264 1-5137 

log IOO 2-OOOO 
33137 

log 79-1 i -8982 

1-4155 

' .; log I -O247 'OIO5 

1-4050=25-41 Cu. 



Copper Sulphate. 77 

3. Determination of the Sulphuric acid. This is effected 
by adding barium chloride to the solution of the copper 
salt and weighing the precipitated barium sulphate : 

CuS0 4 + BaCl 2 = BaSO 4 + CuCl 2 

About one gram of the copper sulphate is weighed out into 
the second beaker, and is dissolved in 30-40 cubic centi- 
metres of water, a few drops of hydrochloric acid are added, 
and the solution is heated to the boiling point. The lamp 
is now drawn aside and solution of barium chloride is 
added drop by drop. In order to determine whether an 
excess of the precipitant has been added, the barium sul- 
phate is allowed to subside, and when the liquid is sufficiently 
clear, a drop of the barium chloride solution is poured down 
the side of the beaker. If an increased turbidity ensues, the 
liquid is again heated and a further quantity of barium 
chloride added : the precipitate is once more allowed to 
settle, and the liquid again tested by the cautious addition of 
barium chloride. When you have assured yourself that the 
precipitation is complete, cover the beaker and set it aside 
in a warm place for a few hours. If you attempt to filter 
the turbid liquid immediately, the finely divided precipitate 
will inevitably pass through the pores of the paper, but on 
standing, especially after precipitation from a hot solution 
slightly acidified with hydrochloric acid, the barium sulphate 
becomes denser and more granular. When the precipitate 
has completely settled add one more drop of barium chloride 
to again assure yourself that the precipitation is complete, 
and proceed to collect the barium sulphate. Fit a No. 5 
filter carefully into the funnel, grease the edge of the beaker, 
pour the clear liquid, by the help of the glass rod, on to the 
filter, and cautiously set the pump working. When the whole 
of the liquid has passed through, pour 40-50 cubic centi- 
metres of hot water over the barium sulphate and boil the 
liquid for a few minutes ; the precipitate is allowed to settle 
and the liquid (which will now be slightly turbid) poured on 



78 Quantitative Chemical Analysis. 

to the filter. It is well to stop the flow of water in the 
pump until the first filter-full of liquid has passed through, 
otherwise there is danger of the precipitate making its way 
through the filter. When the whole of the liquid on the 
filter has passed through, fill up the funnel again with the 
wash-water, cautiously set the pump in operation, rinse the 
precipitate on to the filter with hot water, remove any barium 
sulphate adhering to the beaker by means of a feather, and 
wash five or six times with hot water. Take care to allow 
the whole of the liquid to pass through before a fresh 
quantity is poured in, or the filtrate may become turbid. 
Dry the precipitate and transfer it from the filter to a 
weighed porcelain crucible, burn the filter and add the 
ash to the crucible. Place the crucible on the triangle 
and cautiously heat it with the lid on for 2 or 3 minutes, 
increase the heat and keep the bottom red hot for 5 or 
10 minutes, allow to cool in the desiccator, and when cold, 
weigh : repeat the heating and again weigh. The two 
weighings ought to agree. The results should thus appear 
in the note-book : 

ANALYSIS OF COPPER SULPHATE. 

3. Determination of Sulphuric Acid by precipitation as Barium 
Sulphate. 

Weight of tube + salt (before) 9-5803 
(after) 8-6115 



0-9688 



(1) Crucible + lid + BaSO 4 + ash 87317 

(2) ,, ,, 8-7315 

Crucible + lid 7-8242 



0-9073 
Less ash No. 5 -0023 

0-9050 



Copper Sulphate. 79 

Calculation: ?6_>^o -9050x100 log 96 1-9823 

233-1x0-9688 log -905 1-9566 

log 100 2-0000 



log 233-1 2-3676 

J-57'3 

log -9688 1-9863 

SO 4 = i -5850 = 38 -46 p. ct 
The complete analysis is therefore as follows : 

Calculated Found 

Cu 63-1 25-33 25-41 

so 4 96-0 38-54 38-46 

4H 2 O expelled at ioo-i 10 72-0 28-90 29-14 

H 2 O ,, 200 18-0 7-23 6-91 

249-1 loo-oo 99*92 



II. SODIUM CHLORIDE (Na Cl). * 

Preparation. Common salt rarely contains more than 98 
per cent, of sodium chloride, its principal impurities being 
calcium sulphate and magnesium chloride. These sub- 
stances cannot be readily removed by recrystallisation, but 
by adding hydrochloric acid to a strong solution of 
the salt, pure chloride of sodium is precipitated, and 
the magnesium chloride and calcium sulphate remain dis- 
solved. About 70 grams of salt are dissolved in a quarter 
of a litre of hot water, the solution is filtered and saturated 
with hydrochloric acid gas. The apparatus represented in 
fig. 30 may be conveniently used for this purpose. The 
flask a contains the salt and sulphuric acid, and the evolved 
hydrochloric acid gas, after passing through a small quantity 
of the acid solution contained in the bottle b, is led into the 
filtered brine. The exit-tube of the apparatus is replaced by 
a small funnel dipping into the solution of salt : this method 
of delivering the gas into the liquid prevents the possibility 
of the precipitated sodium chloride interfering with the pas- 



So Quantitative Chemical Analysis. 

sage of the gas by closing the outlet. The salt begins to 
separate out almost immediately, and in an hour or so the 
process may be interrupted. The liquid is poured from the 
precipitated salt, which is washed once or twice with pure 
strong hydrochloric acid solution, allowed to drain, and 
heated gently in a porcelain basin. The moisture cannot 

FIG. 30. 




be removed by filter-paper, as the strong acid would cause 
the contamination of the salt with the iron, &c., contained 
in the ash. The mass should be heated gently in a porcelain 
crucible until all the acid is expelled, powdered roughly 
while still warm, and a portion introduced into a small dry 
tube fitted with a good cork. The remainder of the salt is 
placed in a stoppered bottle : it will prove useful in sub- 
sequent operations. 



Sodium Chloride. 81 

i. Determination of the Chlorine. This is effected by pre- 
cipitation as silver chloride, by means of silver nitrate 
solution. 

NaCl + AgNO 3 = AgCl + NaNO 3 . 

About 0*5 gram of the salt is weighed out into a beaker of 80 
cubic centimetres capacity and dissolved in 30-40 cubic centi- 
metres of water ; a few drops of pure nitric acid are added, 
together with a solution of silver nitrate. If sufficient silver 
solution has been added, the chloride separates out as a dense 
curdy precipitate. When you have satisfied yourself that all the 
chlorine is precipitated, heat the liquid to near the boiling 
point, stirring it occasionally by means of a glass rod, and 
allow the precipitate to settle by placing the beaker, protected 
from the dust, in a warm place for a few hours. Be careful 
to protect the silver chloride as much as possible from the 
light. Pour the clear liquid on to a filter, wash twice by 
decantation with hot water, carefully rinse the precipitate on 
to the filter, wash it 5 or 6 times with hot water and dry 
it in the steam bath. Transfer the dried chloride, detached 
as completely as possible from the filter, to a weighed 
porcelain crucible and heat very gradually, increasing the 
temperature until the chloride begins to fuse at the edges of 
the mass; allow to cool in the desiccator and weigh. If the 
chloride has been carefully protected from light it will have 
at most but a slight violet tinge, and the fused portion will 
have the appearance of horn. The filter-paper is folded in 
the manner described on p. 67, and burnt, the ash being 
allowed to fall on to the chloride. The crucible and its 
contents are again weighed : the increase in weight gives 
the amount of metallic silver originally adhering as silver 
chloride to the filter, together with the ash of the paper. 
The known weight of ash in the filter subtracted from the 
total increase gives the amount of reduced silver ; this is 
calculated to silver chloride, and the amount added to the 
main quantity. An actual example may make this clearer : 

G 



82 Quantitative Chemical Analysis. 

Sodium Chloride taken 0*4065 grams 

Crucible + AgCl 8-9710 

+ Ag + ash 8-9813 

Crucible 7-9860 

8-9813- 8-9710 = 0-0103 
Less ash 0023 

Ag=oo8o = AgCl o-o 1 06 
8-9710-7-98600-9850 
Total AgCl = 0-9956 
= 6o . 6o cent 



143 '5 * 0*4065 

The fused silver chloride may readily be detached from 
the crucible by placing over it a small piece of zinc and 
adding a few drops of dilute sulphuric acid. The semi- 
reduced mass, together with the silver chloride, precipitated 
by adding a few drops of hydrochloric acid to the nitrate 
should be put into a bottle labelled ' silver residues.' When 
these residues have sufficiently accumulated they are to 
be worked up as directed in the Appendix. 

Determination of the Sodium. The salt is converted into 
sodium sulphate by the action of strong sulphuric acid. 
Clean, ignite, and weigh a platinum crucible and lid, intro- 
duce into it 0-5 to i gram of salt, and again weigh. The 
increase of weight gives the amount of sodium chloride taken. 
Place the crucible on a triangle in the slanting position repre- 
sented in fig. 15, and add drop by drop pure strong sulphuric 
acid. Do not heat the crucible for ten or fifteen minutes, or 
until the reaction has moderated. There is no danger of any 
of the substance being lost by projection, if care is taken 
to place the crucible as directed, and to add the acid cau- 
tiously. Now heat the crucible gently from the top, placing 
the lid as indicated in the figure, and allow the flame to 
approach the bottom of the crucible very gradually. The 
operation must be done slowly, and with constant watching, 



Pearl-ash. 83 

or a portion of the sulphate may be lost by spirting. In a 
few minutes the whole of the hydrochloric acid will have 
been expelled, and dense fumes of sulphuric acid will be 
evolved. As these diminish, the heat is gradually raised, 
until the crucible attains a full red heat. Maintain it at this 
temperature for fifteen or twenty minutes, put on the lid, 
allow to cool, and weigh. Again heat to redness for five 
minutes, and weigh a second time. The operation is to be 
repeated, until a constant weight is obtained. The fused 
mass should be quite white. 

III. PEARL-ASH. 

Good pearl-ash of commerce contains from $3 to 96 
per cent, of potassium carbonate, the rest consisting of 
water, alkaline sulphates, chlorides, silica, &c. To deter- 
mine its value it is merely necessary to estimate the per- 
centage amount of potassium carbonate and water. The 
quantity required to make up 106 is taken as the measure of 
the impurities. 

Determination of the Moisture. Weigh out from 3 to 4 
grams of the coarsely-powdered ash from a tube into a small 
weighed porcelain crucible provided with a lid. Gently 
heat the ash for half-an-hour over a small gas flame, and 
weigh when cold ; again heat, and again weigh. It must 
not be forgotten that potassium carbonate is highly hygro- 
scopic ; the weighings must, therefore, be made as expedi- 
tiously as possible. 

Determination of the Potassium. The carbonate is con- 
verted into the double chloride of platinum and potassium 
(PtCl 4 2KCl). Weigh out about 2 grams of the carefully 
sampled carbonate and dissolve in 30-40 c.c. of water. Filter, 
if necessary, into a 250 c.c. flask, wash the filter thoroughly 
from every trace of alkali, fill the flask up to the mark with 
distilled water, and shake vigorously. Transfer two lots of 
50 c.c. each to porcelain basins, cover the basins with 
glass plates, and add dilute solution of hydrochloric acid 

G 2 



84 Quantitative Chemical Analysis. 

in slight excess. Heat the basins on the water- bath, and 
when the expulsion of the carbonic acid is finished, rinse 
the covers with a few drops of hot water, and add solu- 
tion of pure platinum tetrachloride. It is necessary to 
add the solution of platinum in considerable excess ; about 
i gram of the metal is required for 0*5 gram of the car- 
bonate. Evaporate the solutions on the water-bath until 
they become pasty ; remove the dishes, and, when nearly cold, 
pour over the crystals about 25 cubic centimetres of rectified 
methylated spirit. Do not attempt to break up the highly 
crystalline scales of the double salt The action of the 
alcohol in dissolving out the excess of platinum tetrachloride 
may be facilitated by imparting a gentle rotatory motion to 
the contents of the dish.* Cover, and allow to stand for 
five or ten minutes ; pour the clear liquid on to a No. 2 
filter (which ought to be first washed with hot water, and 
then with alcohol), and drain the liquid as far as possible 
from the precipitate : again add about 10 cubic centimetres 
of spirit to the precipitate, and shake ; allow to stand five 
minutes, and pour the clear liquid through the filter. 
Repeat the digestion with spirit a third and fourth time : the 
solution will now be nearly colourless. Transfer the precipi- 
tate to a weighed crucible, by the aid of a glass rod, and a 
stream of alcohol from a small wash-bottle ; pour the alcohol 
through the filter, and wash the paper carefully with alcohol 
until the filtrate is absolutely colourless. Dry the double 
chloride at 70, heat to 100 in the steam -chamber, and 
weigh the precipitate. If the above instructions have been 
properly followed, scarcely a stain will be left on the paper. 
The filter is burnt, the ash added, and the whole again 
weighed : the increase in the weight of the crucible after 
subtracting the ash, may without sensible error be considered 
as platinum : it is calculated to K 2 Ptd6, and the amount 

* Water would dissolve the double chloride : 100 parts of water 
dissolve about I part of the salt at the ordinary temperature ; whereas 
the same quantity of the salt requires 26,400 parts of alcohol of 80 per 
cent, and 42, 6co parts of absolute alcohol. 



Dolomite. 85 

added to the main quantity. When calculated to potassium 
carbonate, the result of the two determinations ought not 
to differ by more than o - 2 per cent. 

IV. ROCHELLE SALT. C 4 H 4 KNaO 6 . 4H 2 O. 
(Separation of Potassium and Sodium.} 

Preparation. 40 grams of cream of tartar, and 30 grams 
of crystallised sodium carbonate, are added successively in 
small portions at a time, to 150 cubic centimetres of boiling 
water. The liquid must be tested to ascertain that it is 
alkaline ; it is filtered, if necessary, and concentrated by 
evaporation. On cooling, the solution deposits fine large 
crystals of the potassium-sodium-tartrate. 

Analysis. Powder a few of the crystals, press between 
filter-paper, and weigh off from a tube about 1*5 gram 
into a platinum dish or crucible. Heat gently, so as to 
melt the salt in its water of crystallisation, and gradually 
increase the flame until the mass is perfectly dry. Ignite at 
a low red heat in the draught-chamber for some time ; allow 
the charred mass to cool, and digest it repeatedly with hot 
water; filter, acidulate with pure hydrochloric acid, and 
evaporate to dryness in a weighed platinum dish, with all 
the precautions necessary to avoid loss by spirting. As soon 
as the residue is perfectly dry, heat it very gently for five 
minutes over the lamp, transfer to the desiccator, and when 
cold weigh the mixed chlorides. Dissolve in a small 
quantity of water, transfer to a porcelain dish, and separate 
the potassium with platinum tetrachloride, as directed in the 
foregoing example. Enough platinum chloride must be 
added to convert both the alkaline chlorides into the 
double salts of platinum. The sodium-platinum-chloride is 
readily soluble in alcohol, especially if previously moistened 
with water. 

V. DOLOMITE. 

This substance is essentially a double carbonate of 
lime and magnesia. No definite relation, however, exists 



86 



Quantitative Chemical Analysis. 



between the amounts of the two carbonates, as the calcium 
and magnesium replace each other in all proportions. Occa- 
sionally the mineral occurs associated with the isomorphous 
carbonates of iron and manganese. 

The portion employed for analysis is finely powdered, dried 
in the steam-bath, and introduced into a small corked tube. 

Determination of the Carbonic Acid. The mineral is 
decomposed by dilute hydrochloric acid, and the carbon 

FIG. 31. 




dioxide, freed from moisture, is absorbed by soda-lime. 
The determination is most accurately effected by means of 
the apparatus represented in fig. 31. The flask A has a 
capacity of about 150 cubic centimetres : it is fitted with a 
caoutchouc cork, containing two holes, into one of which is 
inserted the little bulb-tube a, containing a few drops of 
strong sulphuric acid. This serves to regulate the rapidity 
of the decomposition, by indicating the speed with which 
the gas travels through the apparatus. The tube , which 



Dolomite. 87 

may be 10 centimetres high, is filled to a depth of 6 centi- 
metres with pumice saturated with solution of copper sul- 
phate, and strongly heated until all the water has been 
expelled : its use is to absorb any vapour of hydrochloric 
acid which may pass over from the flask. The remainder 
of the tube is filled with coarsely-powdered calcium chlo- 
ride. In the tube c the carbonic acid is absorbed ; seven- 
eighths of it are filled with soda-lime ; and as this sub- 
stance, in combining with carbonic acid, becomes heated, 
and parts with a little water, the remaining one-eighth of 
the tube is filled with calcium chloride : c is also con- 
nected with a small unweighed tube containing calcium 
chloride, in order to prevent the possibility of the weighed 
tube absorbing atmospheric moisture. Caoutchouc corks are 
used to close the tubes ; if, in their absence, ordinary corks 
are employed, they must be cut, and covered with sealing- 
wax. The tubes may be suspended as in the figure, or by 
wires from a glass rod running through a cork held in a 
clamp on the retort stand. When not in use the tubes are 
closed by stoppers of glass rod. In the other hole of the 
caoutchouc stopper of A is fitted a bulb-tube, passing down 
nearly to the bottom of the flask, drawn out and turned 
up in the manner represented in the figure ; this can be 
closed by means of a small screw clamp : d contains soda- 
lime : it is attached to the caoutchouc tubing on the bulb- 
tube at the close of the experiment. 

Weigh out about 1-5 gram of the carbonate into A, and 
add about i o cubic centimetres of water. Weigh c without 
the stoppers and tubing, and put the several parts of the 
apparatus together, setting the flask A upon a piece of wire 
gauze on a tripod. Nearly fill the bulb-tube with hydro- 
chloric acid solution, diluted with its own volume of water. 
Open the clamp, and allow enough acid to pass into the 
flask to set up the evolution of the carbonic acid, and as the 
action diminishes add more acid, until the tube is empty. 
Close the clamp, and connect the exit-end of the apparatus 
with an inverted wash-bottle, of about 600 cubic centimetres 



88 Quantitative Chemical Analysis. 

capacity, and filled with water, the jet of which you have 
transferred to the short, obtusely-bent tube, and temporarily 
closed by a piece of caoutchouc tubing and clamp; heat 
A slowly to boiling, boil for about ,one minute, cautiously 
open the clamps, remove the lamp, and aspirate a slow 
current of air through the apparatus, regulating the speed by 
means of the screw. Detach c, allow it to cool, and weigh it; 
the increase of .weight gives the amount of carbonic acid in 
the mineral. This method of estimating carbon dioxide is 
very accurate, and is generally applicable ; it is expeditious, 
and has the advantage of being direct. The tubes may be 
used a great number of times without their contents being 
changed, if they are well stoppered when not in use. 
A. very simple and accurate method of determining carbon 
dioxide in salts and minerals consists in heating the sub- 
stance with fused and powdered potassium bichromate in a 
short combustion-tube (about 25 cm. long), passing the 
evolved gas through a tube containing calcium chloride, and 
absorbing it in a weighed soda-lime tube. Or a weighed 
portion of the carbonate is heated with about four times 
the quantity of fused borax in a platinum crucible to dull 
redness, and the carbon dioxide determined from the loss 
of weight. 

Determination of the Silica, Iron (and Manganese), Lime, 
and Magnesia. Decant the solution from the flask A into a 
porcelain basin, rinse the flask with a few cubic centimetres 
of hot water, adding the washings to the main quantity of the 
liquid, and evaporate the whole to complete dryness on the 
water-bath, in order to render the small quantity of silica 
insoluble. Moisten the dried residue with a few drops of 
strong hydrochloric acid, cover the basin with a glass plate, 
allow to stand for a few minutes, add hot water, and filter 
the solution through a No. 3 filter ; wash thoroughly, drain 
the paper by the action of the pump, fold the filter, and 
without further drying throw it into a weighed platinum 
crucible : cautiously heat until the paper is dry, and incine- 



Dolomite. 89 

rate. The increase in weight of the crucible, minus the 
filter-ash, gives the amount of silica. 

To the filtrate from the silica, add a little bromine-water, 
then ammonium chloride and ammonia in slight excess ; 
heat gently for some time and filter. Wash the precipitate 
once or twice, re-dissolve it in hydrochloric acid by heating, 
add one more drop of bromine, again precipitate with ammo- 
nia, and filter. Dry and weigh the oxide of iron (and man- 
ganese*): the precipitate may be ignited without complete 
drying. The second precipitation effects the removal of small 
quantities of lime and magnesia precipitated with the ferric 
oxide. 

Mix the ammoniacal filtrates, and add ammonium oxalate 
in quantity sufficient to precipitate the lime and to convert 
the magnesia into oxalate. Presence of excess of ammonium 
oxalate prevents the slight solubility of calcium oxalate in 
chloride of ammonium solution. Allow the liquid to stand 
for ten or twelve hours. Decant the clear liquid on to a filter 
and wash once or twice by decantation, taking care to disturb 
the precipitate as little as possible. Dissolve the calcium 
oxalate in a small quantity of hydrochloric acid, heat to 
boiling, add a few drops of ammonium oxalate and a slight 
excess of ammonia. This double precipitation of the calcium 
oxalate effects the separation of a small quantity of co-pre- 
cipitated magnesia. Filter off the oxalate, wash thoroughly 
with hot water, and dry. Transfer the dried precipitate to a 
weighed platinum crucible, and if its weight does not exceed 
i gram, proceed to convert it into caustic lime. Burn the 
filter, and add the ash to the crucible. Heat the crucible 
gently, with the lid on, and gradually increase the flame 
until the bottom is red hot. Now expose the crucible to a full 
red heat over the blow-pipe for fifteen minutes, occasionally 
removing the lid for a few seconds ; allow to cool in the 
desiccator, and weigh. By this treatment the oxalate is con- 
verted, first into carbonate and then into caustic lime. The 

* For a. method of determining the manganese, usually present in 
small quantity only, in limestones and dolomites, see Part IV. 



90 Quantitative Chemical Analysis. 

heating must be repeated until the weight is perfectly con* 
stant ; that is, until the carbonate is wholly converted into 
lime. If the quantity of the oxalate exceeds i gram, its 
conversion into oxide is accomplished with difficulty and 
requires prolonged heating. In this case it is better to trans- 
form the oxalate simply into carbonate by heating gently over 
a small flame scarcely sufficient to make the bottom of the 
crucible appear red hot by diffuse daylight. The conversion 
into carbonate is rendered visible by a slight change of colour 
which creeps over the heated oxalate. After heating for ten 
minutes, weigh, and repeat the operation with the lid on 
until the weight is constant. Weigh the filter-ash and the 
small quantity of adhering lime on the lid. Moisten the 
carbonate with a few drops of water and test it with a slip of 
reddened litmus paper ; it ought not to show the slightest 
trace of alkalinity. If the paper becomes blue, the crucible 
has been overheated ; in that case transfer the ash of the 
filter to the crucible, add a few drops of ammonium carbonate 
solution, evaporate to dryness on the water-bath, heat very 
gently for a few minutes and weigh. The ammonium car- 
bonate reconverts the caustic lime into carbonate. The 
nitrate from the oxalate of lime is poured into a porcelain 
basin, concentrated considerably, and rendered strongly acid 
by nitric acid. About 3 grams of the acid are employed for 
each gram of the sal- ammoniac supposed to be present. The 
dish is covered with an inverted funnel and gently heated, 
when a rapid effervescence sets in, owing to the decompo- 
sition of the ammonium chloride. The liquid is evaporated 
to dryness, when, if sufficient nitric acid has been added, all 
the sal-ammoniac will have been expelled.* The funnel is 
rinsed, and the saline mass in the dish dissolved in water, 
ammonium chloride added in small quantity, the solution 
rendered strongly alkaline by ammonia, filtered if necessary, 
and mixed with solution of sodium phosphate. The liquid 

* This method of removing the excess of sal-ammoniac is preferable 
to that of evaporating to dryness in a platinum basin and igniting. By 
the latter plan there is danger of loss from the tendency of the salt to 
creep over the side of the dish during the evaporation. 



Barium Sulphate. 91 

is well agitated by shaking the beaker, covered with a glass 
plate, and set aside for twenty-four hours. The clear liquid 
is poured on to a filter and the precipitate washed by decan- 
tation in the beaker, and afterwards on the filter by dilute 
ammonia-water (i part strong ammonia and 5 of water), 
until the filtrate acidulated with pure nitric acid gives only 
a slight opalescence with silver nitrate. Pure water dissolves 
the precipitate to a slight extent (i part in 15,300 of water); 
in dilute ammonia it is much less soluble (i part in 45,000). 
The presence of a large excess of ammonium chloride in- 
creases its solubility; hence the necessity of expelling the 
greater portion of this salt before precipitating the magnesia. 
The dried phosphate is detached as completely as possible 
from the filter, transferred to a weighed platinum crucible and 
very gradually heated (at first with the lid on) for fifteen 
minutes. The temperature is now raised until the crucible 
is red hot, when the lamp is withdrawn and the filter-ash 
added. Take care that the filter is burnt as completely 
as possible, or the crucible will be corroded by the action of 
the carbon on the phosphate. The crucible is then strongly 
ignited for a quarter of an hour, allowed to cool, and weighed. 
The residue (magnesium pyrophosphate Mg 2 P 2 O 7 ) should 
be white, or, at most, have a very slight tinge of grey.* 

VI. BARIUM SULPHATE (Ba SO 4 ). 

The pure substance is prepared by adding a dilute solu- 
tion of barium chloride to an excess of hot and moderately 
diluted sulphuric acid. The precipitate is washed once or 
twice with hot water, dried, and ignited. The method of 
analysis is founded upon the decomposition of the sulphate 
by prolonged fusion with an alkaline carbonate. 

About i gram of the ignited precipitate is weighed out 

* In any analysis involving the separation of several substances it is 
advisable to preserve the weighed precipitates in small corked tubes or 
between watch-glasses until the analysis is finished. Questions re- 
garding the purity or identity of the substances separated frequently 
arise, which, of course, cannot be answered if the bodies are thrown 
away. 



92 Quantitative Chemical Analysis. 

into a platinum crucible and mixed with 3 parts of a dry 
mixture, in equivalent proportions, of pure potassium and 
sodium carbonates. This mixture may be conveniently 
obtained in a state of purity by igniting Rochelle salt 
(C 4 H 4 KNaO6 + 4H 2 O) in a platinum basin, extracting the 
charred residue with hot water, filtering and evaporating to dry- 
ness. The mixture of barium sulphate and alkaline carbonates 
is fused at a bright red heat for thirty or forty minutes, allowed 
to cool, the mass extracted with water containing a few drops 
of ammonia and ammonium carbonate, and filtered.* The 
filtrate contains the sulphuric acid in union with the alkalies; 
the residue consists of barium carbonate. It is washed with 
water containing a few drops of ammonia and ammonium 
carbonate, dried, and 'weighed. It is dissolved in the 
crucible in a few drops of hydrochloric acid, and a slight 
excess of sulphuric acid added; the mixture is cautiously 
evaporated to dryness, and the residue ignited and weighed. 
Its amount ought to be equal to that originally taken. 

The filtrate containing the alkaline sulphate is acidulated 
with hydrochloric acid, heated to boiling, barium chloride 
added, and the precipitate washed, dried, ignited, and 
weighed. Its weight ought to equal that of the barium sul- 
phate analysed. 

The same process is applicable to the analysis of strontium 
and calcium sulphates. 

VII. INDIRECT ESTIMATION OF BARIUM AND CALCIUM. 

It is frequently possible to determine the amount of a 
substance A, by combining it with a second body B, so as to 
form a definite compound A B. By estimating the quantity 
of B in the combination, we can readily calculate the amount 
of A which must be present. Thus we can determine the 
amount of Ag in a solution by estimating the amount of Cl 
required to precipitate it completely. In like manner we 

* Barium carbonate dissolves in 14,000 parts of cold water and 
15,500 of boiling water. It is ten times less soluble in water con- 
taining a slight quantity of ammonia and ammonium carbonate. 



Indirect A nalysis. 93 

could determine the amount of Ba and Ca from the quantity 
of CO 2 respectively contained in their carbonates. Such 
indirect determinations are based on the law of constant 
proportion, which states \h&\. the same substance always consists 
of the same elements united in the same proportion. 

But it will be obvious on a little reflection that we can 
determine the amount of barium and calcium in a mixture of 
their carbonates, by estimating the amount of carbon dioxide 
contained in a known weight of the mixed compounds. The 
possibility of such an estimation is based upon the wide 
difference which exists between the combining weight of 
barium (137*2) and that of calcium (40-0), both of which 
substances combine with 44 of CO 2 . Supposing that we had 
found that 2 grams of the mixed carbonates had evolved 0*67 
gram of carbon dioxide. Then, if the whole of the carbon 
dioxide were combined with the barium, the amount of the 
barium carbonate would be 3*002 grams. 

Eq. C0 3 Eq. BaCO 3 . CO 2 found. 

44 : 197-2 :: 0-67 : = 3-002. 

But the weight of the mixed carbonates taken was only 2 
grams. The deficiency (3-002 2 -o)= i '002, is proportional 
to the amount of calcium carbonate present. This amount 
is thus found : The difference between the atomic weights of 
BaCO 3 and CaCO 3 is to the atomic weight of CaCO 3 , as the 
difference found is to the quantity of calcium carbonate contained 
in the mixture. 

BaCO 3 - CaCO 3 = 197-2 - 100 = 97-2 

BaCO 3 - CaCO 3 . CaCO 3 . 

97*2 : 100 :: 1*002 : = 1-032. 
Accordingly the composition of the mixture is 

Calcium Carbonate 1*032 
Barium Carbonate 0*968 
2-000 
or expressed centesimally, 

Calcium Carbonate 51-6 
Barium Carbonate 48*4 

lOO'O 



94 Quantitative Chemical Analysis. 

Similarly we might determine the proportion of the bases 
present in the mixture by estimating the weight of sulphuric 
acid necessary to form the two sulphates. The method of 
calculation is, mutatis mutandis, precisely similar to that 
above given. 

Weigh out into a platinum crucible about equal weights 
of pure and recently ignited barium and calcium sulphates 
(o'5 gram of each is a convenient quantity to take). Mix 
with 3 or 4 pts. of the mixture of sodium and potassium 
carbonates, and proceed exactly as described in the preceding 
example. The weighed barium and calcium carbonates are 
then decomposed in the apparatus described in p. 84, and 
the amount of carbon dioxide determined with great care. 
From the weight of CO 2 obtained, the proportion of the two 
bases is calculated in the manner above described. 

As a control, determine the amount of sulphuric acid in 
the nitrate, after acidulation with hydrochloric acid, by pre- 
cipitation with barium chloride, according to the method 
described on p- 75, and again calculate the proportion of the 
two bases. This exercise will afford a good trial of the 
manipulative skill of the operator. 

VIII. FERROUS AMMONIUM SULPHATE. 
Fe(NH 4 ) 2 2 SO 4 + 6H 2 O. 

To prepare this salt, 27-8 grams of recrystallised ferrous 
sulphate and 13*2 grams of pure ammonium sulphate are 
separately dissolved in the least possible quantity of water 
at a temperature of about 40. The solutions are mixed, a 
few drops of sulphuric acid are added, and the mixture is 
stirred constantly until cold. The greater portion of the 
salt separates out in a finely-divided state : if the solution is 
now set aside for a few hours a further quantity of the double 
salt crystallises out. Pour off the mother-liquor, allow the 
crystalline powder to drain, and dry it thoroughly between 
filter-paper. 

Determination of the Ammonia. Weigh out about i gram 



Ferrous Ammonium Sulphate. 



95 



of the salt into the retort (fig. 32), the neck of which is 
contracted at a, and the upper portion filled with fragments 
of broken glass. The tube b is filled with strong soda-lye, 
which can be delivered little by little on opening the clamp. 
The flask c is fitted with a caoutchouc cork and bent tube <?, 
on which is blown a bulb. The short wide tube d is filled 
with fragments of glass, previously well washed with water j 
through this tube hydrochloric acid is poured. The tube 
e is so arranged that it just dips beneath the surface of 

FIG. 32. 




the liquid in the flask. The retort containing the weighed 
quantity of ammoniacal salt, dissolved in a small quantity of 
water, is fixed on a clamp, the tube e inserted into its neck, 
a small quantity of soda solution allowed to flow into the 
retort, and the liquid heated to boiling. Care must be 
tiken to prevent the liquid, if it shows any tendency to 
froth, from passing over into the flask. The broken glass 
in the neck of the retort tends to prevent such a mishap. 
The liquid should be boiled for fifteen or twenty minutes 



96 Quantitative Chemical Analysis. 

after the caustic soda solution has been added. When the 
evolution of ammonia is finished, the tube e is disconnected, 
the powdered glass in d washed with distilled water, the 
tube e drawn up from the liquid and washed with distilled 
water. The ammoniacal solution is poured into a porcelain 
basin, an excess of platinum tetrachloride is added, and the 
whole is evaporated just to dryness on the water-bath. The 
double chloride is washed with strong alcohol, and is trans- 
ferred to the weighed platinum crucible in the manner de- 
scribed on p. 82. The salt is dried at 70 or 80, and heated 
to 1 00 for ten or fifteen minutes, after which its weight will be 
constant. The little filter is dried and ignited on the lid of 
the crucible and weighed separately. The weight of the 
residual platinum is calculated to that of double salt, and the 
amount added to the main quantity. By way of control, the 
double salt may be gently heated so as to expel the greater 
portion of the ammonium chloride ; the crucible is then 
raised to a full red heat, and the metallic platinum weighed. 
Determining as platinum, however, is not more accurate than 
weighing the double salt, owing to the readiness with which 
finely-divided particles of the metal are carried away in the 
vapour of the escaping ammonium chloride. 

IX. DETERMINATION OF NITRIC ACID. 

Dr. Gladstone and Mr. Tribe have found that a thin 
plate of zinc coated with copper (formed by placing the 
former metal in a solution of copper sulphate for a few 
minutes) decomposes water, particularly on warming, with 
the formation of zinc hydrate and the evolution of hydrogen. 

The hydrogen so eliminated is capable of reducing nitric 
acid in combination to the state of ammonia : 

NO 3 K + 4H 2 = NH 3 + HKO + 2H 2 O. 

This reaction constitutes the basis of a method of deter- 
mining nitric acid in nitrates. 

About 25-30 grams of thin sheet zinc are placed in a flask 



Nitric Acid. 97 

of about 200 c.c. capacity, and covered with a moderately- 
concentrated and slightly-warmed solution of copper sul- 
phate. In about ten minutes a thick spongy coating of 
copper will be deposited on the zinc : the liquid is poured 
off the metals, which are now well washed with cold water, 
and covered with about 40 or 50 c.c. of pure water. Weigh 
out about 0-5 gram of pure nitre into the flask, which is then 

FIG. 33. 




placed in a sand-bath and connected with the condensing 
arrangement shown in fig. 33. The receiver and U tube 
contain a few cubic centimetres of dilute hydrochloric acid. 
The liquid is gradually heated to boiling and distilled for 
about an hour. The distillate is poured from the receiver 
and evaporated to dryness in a porcelain basin over the 
water-bath, with excess of platinum tetrachloride, and the 
double chloride is treated exactly as in the foregoing example. 



98 Quantitative Chemical Analysis. 



X. . POTASH-ALUM. A1 2 (SO 4 ) 3 .K 2 SO 4 .24H 2 O. 

The salt is purified by recrystallisation, powdered, dried 
between blotting-paper, and placed in a well-corked tube. 

Determination of the water. About i gram of the 
double salt is heated to 120 in the apparatus represented in 
fig. 29, until it ceases to lose weight. The loss is equiva- 
lent to 10 atoms of water. The temperature is now raised 
to 200, and the heat maintained at this point until the 
weight is once more constant. The salt should now be 
perfectly anhydrous. 

Determination of the Alumina and Potassium Sulphate. 
Weigh out about i gram of the crystalline salt into a porce- 
lain basin, dissolve in hot water, and add a moderate 
quantity of ammonium chloride solution, together with a 
slight excess of ammonia. Heat the liquid to boiling and 
maintain it in gentle ebullition for some time, keeping the 
basin covered with a sufficiently large watch-glass to avoid 
loss by projection. Rinse the watch-glass into the basin, 
and pour the clear supernatant liquid on to the filter, wash 
the precipitate once or twice by decantation and transfer it 
also to the filter ; wash three or four times with boiling water, 
and, after the whole of the liquid has passed through, keep 
up the action of the pump for ten minutes. Remove the 
filter containing the precipitate from the funnel, and, without 
further drying, place it in a weighed platinum crucible ; heat 
gently for a few minutes to char the paper, and gradually in- 
crease the flame until the crucible is red hot. Keep it at this 
temperature for ten or fifteen minutes, occasionally removing 
the lid, and then ignite it strongly over the blow-pipe for 
five or ten minutes ; place the crucible in the desiccator, and 
weigh when cold. 

Ignition over the blast-lamp expels the last traces of water, 
together with the minute quantity of sulphuric acid which 



Glass. 99 

is precipitated with the alumina from a solution containing 
sulphates. 

The filtrate from the precipitate of alumina contains the 
potassium sulphate : it is evaporated to dryness in a weighed 
platinum basin, gently heated to expel ammoniacal salts, and 
moistened with a few drops of pure sulphuric acid^ in order 
to transform the potassium chloride, which invariably forms 
when potassium sulphate is heated with ammonium chloride, 
back into sulphate. The mass is again heated, with all the 
precautions detailed on p. 82, in order to expel the excess 
of sulphuric acid, and when cold, the potassium sulphate 
is weighed. 

XI. GLASS 

consists of a mixture of the alkaline silicate with cer- 
tain insoluble silicates, generally of calcium, lead, iron, 
aluminium, magnesium, or manganese. The best window 
glass has approximately the composition, Na 2 OQxO.6SiO 2 . 
In flint glass the lime is replaced by oxide of lead. The 
pale green variety used for chemical apparatus is mainly 
made up of silicates of lime and potash, mixed with smaller 
quantities of iron and alumina. 

In order to analyse it, the glass is reduced to the finest 
possible state of division, and fused in a platinum crucible 
with four times its weight of a mixture of equal parts of 
sodium and potassium carbonates. When cold, the crucible 
is placed in a porcelain basin, and the mass boiled out with 
water, hydrochloric acid is added in excess, and the whole 
is evaporated to complete dryness over the water-bath. The 
dried mass is then moistened with strong hydrochloric acid, 
hot water is added, and the silica is filtered off, repeatedly 
washed with hot water, dried and weighed. The solution con- 
tains the lead, iron, alumina, manganese, lime, and magnesia. 
The alkalies cannot, of course, be determined in this por- 
tion, as they are mixed with the salts required to decompose 

II 2 



IOO Quantitative Chemical Analysis. 

the glass. Pass sulphuretted hydrogen through the filtrate, 
to precipitate the lead ; filter, dry it, and convert it into 
sulphate by treatment with strong nitric acid. Add a few 
drops of bromine to the filtrate, and heat gently ; add 
ammonia, and filter off the iron, alumina, and manganese. 
The lime and magnesia are separated as in No. V. 



FIG. 34. FIG. 35. 




Determination of the Alkalies. About 1*5 gram of the 
finely-powdered glass is weighed out into a platinum crucible, 
and intimately mixed with 9 grams of calcium carbonate, and 
i *5 gram of ammonium chloride, and heated to bright red- 
ness for an hour, in a small table furnace (figs. 34, 35). The 
platinum crucible should be protected from the direct action 
of the fire by being placed in a small clay crucible, with a 



Glass. IOI 

little calcined magnesia at the bottom. When' cold the 
contents of the crucible are boiled with water in a silver or 
platinum dish, filtered, the filtrate acidified with hydrochloric 
acid, and evaporated to dryness to render the silica in- 
soluble. The residue is treated with hot water and filtered ; 
to the filtrate, ammonia, ammonium carbonate, and a few 
drops of ammonium oxalate are added to throw down the 
lime. The liquid is boiled, to render the precipitate dense 
and granular. It is filtered off ; the liquid is evaporated to 
a small bulk in a porcelain basin, pure nitric acid is added 
in quantity, and the whole is evaporated to dryness to de- 
stroy the ammonium chloride. The saline residue is dis- 
solved in a little water, and filtered if not quite clear, and 
again evaporated to dryness with a small quantity of strong 
hydrochloric acid, whereby the nitric acid is expelled. If 
the quantity of the mixed alkalies is considerable, this 
treatment with hydrochloric acid must be repeated once or 
twice before the nitric acid is completely dissipated. The 
alkaline chlorides are again dissolved in a little water, and 
evaporated to dryness in a weighed platinum dish, heated 
gently, and weighed. The potassium chloride is then sepa- 
rated by platinum tetrachloride in the manner described in 
No. IV. p. 83. Its amount, subtracted from the sum of the 
chlorides, gives the sodium chloride. 



XII. FELSPAR (Orthoclase, Albite}. 

The group of the felspars may be regarded as silicates of 
alumina united, in varying proportions, with silicates of the 
alkalies and alkaline earths. The varieties, orthoclase and 
albite, differ from one another in crystalline form and in 
chemical composition. Orthoclase crystallises in forms 
belonging to the monoclinic system, and the alkali it contains 
is chiefly potash, whereas albite is triclinic, and its alkali 
consists mainly of soda. 



102 t . jQiiQ_ntita.tive Chemical A nalysis. 

The following analyses serve to show this characteristic 
difference in composition : 

Orthoclase Albite 

Silica 6476 67-62 



Alumina 

Ferric oxide 

Lime 

Magnesia 

Potash 

Soda. 

Loss on ignition 



17-60 16-59 

0-50 2-30 

0-65 0-85 

0-30 1-46 

I4'i8 0-51 

175 10-24 
0-65 

100-39 99-57 



The methods employed in the analysis of these minerals 
are identical with those described in No. XL The only 
point which needs special mention is the separation of the 
iron and alumina. 

The solution containing the iron, alumina, lime, magnesia, 
and alkalies, from which the silica has been removed by 
evaporation, is mixed with a little nitric acid, boiled for some 
time, and a slight excess of ammonia added, whereby, on 
boiling, the iron and alumina are precipitated. Care must be 
taken not to employ a very large excess of ammonia, other- 
wise the precipitation, even after protracted boiling, 'will not 
be complete. The mixed oxides are washed -thoroughly with 
hot water, dried as far as possible by the action of the pump, 
and transferred to a platinum dish ; the small portions remain- 
ing on the filter are dissolved in hot hydrochloric acid, the 
solution being allowed to drop into the dish. When the whole 
of the acid solution has passed through, the dish is removed 
from under the funnel, a beaker put in its place, and the 
filter thoroughly washed with hot water, the washings being 
collected in the beaker. The precipitate in the dish is now 
dissolved in the least possible quantity of hydrochloric acid, 
an excess of a concentrated solution of pure caustic potash 
added, the liquid heated to boiling, and a lump of the pure 
hydrate added, in quantity sufficient to dissolve the alumina, 
and the boiling is continued for a few minutes. The contents 
of the dish are now washed into the beaker, diluted with a little 



Brass, Bronze, &c. 103 

water, and filtered, the ferric oxide being repeatedly washed 
with hot water, dried, and weighed. The alkaline solution 
containing the alumina is acidified with hydrochloric acid, 
a few crystals of potassium chlorate are added to destroy any 
organic matter present, which would tend to retain a small 
portion of the alumina in solution ; the liquid is concentrated 
in a porcelain basin, and ammonia is added in slight excess, 
and the liquid is boiled until a piece of turmeric paper held 
in the steam is no longer turned brown. The precipitate is 
filtered, dried, ignited over the blow-pipe in a platinum 
crucible, at first very gently, and with the lid on, and then 
for 5 or 10 minutes to bright redness. 

Instead of treating the mixed oxides of iron and alumina 
with caustic potash, they may be washed, dried, ignited 
over the blow-pipe, and weighed together. The mixture, or 
an aliquot portion of it, is then brought into a porcelain 
boat, and strongly heated in a porcelain tube, in a current 
of dry hydrogen for an hour. The boat must be allowed to 
cool in the current of the hydrogen before it is withdrawn. 
The loss of weight which it suffers gives the amount of 
oxygen combined with the iron ; each milligram of loss is 
equivalent to 3*339 milligrams of ferric oxide. With proper 
care this method is very accurate. By way of control 
(and this is more particularly necessary when the amount 
of oxide of iron, compared with that of the alumina, is very 
small), you may treat the weighed mixture with highly 
dilute nitric acid (i part of acid to 30 of water). The dis- 
solved iron is re-precipitated, after filtering, by means of 
ammonia, dried, and weighed. The residual alumina is 
also dried and weighed. 

XIII. BRASS, BRONZE, GUN-METAL, BELL-METAL. 
(Separation of Tin, Copper, Lead, Iron, and Zinc.) 

Weigh out about 5 grams of the finely-divided alloy (a 
portion of a penny, for example) into a flask, and dissolve in 



IO4 Quantitative Chemical Anal} 



'StS. 



25 c.c. strong nitric acid and 15 c.c. water at a gentle heat. 
Place a small funnel in the neck of the flask to prevent loss 
by spirting. When the substance is dissolved, add about 
three^ times the bulk of water, and digest the precipitate on 
the water bath with occasional shaking for about an hour. 
Allow the precipitate to settle, decant the clear liquid, and 
repeat the digestion on the water bath for about an hour 
with dilute nitric acid (i : 6). The oxide of tin is thus 
obtained free from admixed metals ; it is filtered off, 
washed, dried, and weighed in a porcelain crucible. If the 
quantity is at all considerable, it requires to be ignited over 
the blow-pipe before it is rendered completely anhydrous. 
The filtrate is evaporated nearly to dryness with strong hydro- 
chloric acid, to expel the greater portion of the nitric acid ; 
re-dissolved in hot water, and the solution precipitated by 
sulphuretted hydrogen. The clear liquid (which should smell 
strongly of the gas) is filtered off, and the precipitate washed 
once or twice by decantation with hydrochloric acid of 
sp. gr. i -05, through which a stream of sulphuretted hydrogen 
is led, and afterwards with water containing sulphuretted 
hydrogen. The mixed sulphides are drained thoroughly, 
and transferred to a small porcelain basin, and digested with 
nitric acid and about 10 cubic centimetres of dilute sul- 
phuric acid. The solution is evaporated nearly to dryness, 
a small quantity of water is added to dissolve the copper 
sulphate, and the lead sulphate is filtered off without delay 
through as small a filter as possible, and washed with 
water acidulated with sulphuric acid, the filtrate being re- 
ceived in a litre flask. When the copper has been washed 
away, the lead sulphate is washed with dilute alcohol to 
remove the last traces of acid, otherwise the filter-paper 
would blacken on drying and fall to pieces.* The sulphate 
is weighed in a porcelain crucible ; care must be taken- to 
remove the precipitate as completely as possible from the 

* The alcoholic washings are not to be mixed with the filtrate con- 
taining the copper. 



Brass, Bronze, &c. 105 

filter before incinerating. The ash may be moistened with 
one drop of dilute nitric acid, heated gently, a drop of 
sulphuric acid added, and the contents of the crucible care- 
fully dried and ignited. The filtrate in the litre flask con- 
taining the copper is diluted to the mark, and the liquid 
thoroughly mixed by shaking; 100 cubic centimetres are 
withdrawn, and the copper determined, as in No. I. Part. II. 
p. 72, by precipitation with soda. In a second portion of the 
solution determine the amount of metal by precipitation with 
metallic zinc. Transfer 800 c.c. to a porcelain basin, add 
an excess of pure sulphuric acid, and evaporate to dryness to 
expel nitric acid. Dissolve the copper sulphate in a small 
quantity of water, decant the solution into a weighed platinum 
dish, and place in it a piece of pure zinc (about i or 2 grams 
will be sufficient) ; add a few drops of hydrochloric acid, and 
cover the dish with a watch-glass. In about an hour the 
whole of the copper will be precipitated, partly as a coherent 
film on the dish, and partly in red, spongy masses. A drop 
or two of the supernatant fluid should be tested with sulphu- 
retted hydrogen water ; it should, of course, remain colourless. 
Assure yourself that the whole of the zinc is dissolved, press 
the spongy masses of copper together, decant the colourless 
liquid, and repeatedly wash the metal with boiling water 
until the washings give no opalescence when tested with 
silver nitrate or barium chloride. Allow the water to drain 
away, and cover the copper with a small quantity of strong 
alcohol. Pour this away, and dry the copper in the steam- 
bath. The precipitation of the copper may also be effected 
in a porcelain or glass dish; in this case the process 
requires longer time, owing to the absence of the galvanic 
action between the platinum and zinc. A weighed piece of 
platinum-foil placed in the dish, and of course weighed with 
it at the termination of the experiment, accelerates the 
operation. 

The filtrates from the sulphides of lead and copper con- 
tain the iron and zinc ; they are concentrated to a small 



io6 Quantitative Chemical Analysis. 

bulk (about 70 cubic centimetres), filtered into a small flask, 
with a drop or two of nitric acid to oxidise the iron, heated 
for a few minutes, allowed to cool, and mixed with a small 
quantity of freshly precipitated barium carbonate suspended 
in water. The liquid should not contain too much free acid ; 
if a large excess is present, it must be removed by adding 
sodium carbonate before mixing with the barium carbonate. 
The flask is closed and occasionally shaken. After standing 
a few hours the iron is all precipitated. The liquid is 
filtered, ammonium chloride added, and the zinc precipitated 
by sulphuretted hydrogen. The zinc sulphide is filtered off, 
washed, re-dissolved in nitric acid, the solution boiled, and 
the zinc re-precipitated as carbonate by the addition of 
sodium carbonate. The zinc carbonate is filtered, washed, 
dried, and ignited in a porcelain crucible, and weighed as 
oxide. 

The barium carbonate precipitate mixed with the iron is 
dissolved in hydrochloric acid, ammonium chloride is added, 
together with a slight excess of ammonia, and the liquid 
heated and filtered. The washed precipitate is re-dissolved 
in a few drops of hydrochloric acid, and the iron again pre- 
cipitated by ammonia free from carbonate, washed, dried, 
ignited, and weighed as ferric oxide. 



XIV. GERMAN SILVER. 
(Separation of Copper , Zinc, and Nickel. ) 

Weigh out about i "5 gram of the finely-powdered alloy, 
and dissolve in nitric acid at a gentle heat, with the precau- 
tions mentioned in No. XIII. Evaporate the excess of acid, 
and separate any oxide of tin which may be ' formed. Pre- 
cipitate the copper by sulphuretted hydrogen in hot solution : 
re-dissolve the copper sulphide, and again precipitate with 
sulpuretted hydrogen to separate the small quantity of 



German Silver. 107 

co-precipitated zinc. The sulphide is then treated as in 
No. XIII. , and the copper weighed as oxide. 

To the filtrate containing the zinc and nickel is added a 
solution of sodium carbonate until a slight permanent pre- 
cipitate is formed, which is then re-dissolved by the cautious 
addition of a few drops of hydrochloric acid. Pass sul- 
phuretted hydrogen through the liquid, and to ensure the 
complete precipitation of the zinc add a few drops of a very 
dilute solution of sodium acetate, and again treat with sul- 
phuretted hydrogen. Allow to stand for twelve hours, and 
filter off the zinc sulphide ; wash it with sulphuretted hy- 
drogen water, dissolve in hydrochloric acid, and precipitate 
with sodium carbonate, filter, wash, dry, and ignite and weigh 
as zinc oxide. Boil the filtrate containing the nickel after 
the addition of a few drops of hydrochloric acid, and pre- 
cipitate with caustic soda (best in a porcelain basin), filter 
off the nickel hydrate, wash, dry, and ignite and weigh as 
nickel oxide. 

XV. BRITANNIA METAL. 
(Separation of Tin and Antimony?) 

About i *5 gram of the alloy, as finely divided as possible, 
is oxidised in a porcelain basin with strong pure nitric acid, 
and evaporated to perfect dryness. The dried mass is then 
washed into a silver basin, again evaporated to dryness, and 
fused with an excess of sodium hydrate (about eight times 
the bulk). It is treated with a small quantity of water, and 
the liquid mixed with about one-third of its volume of strong 
alcohol. The stannate of soda, together with the excess 
of hydrate, is thus separated from the sodium antimoniate. 
The liquid is allowed to stand for six hours, filtered, and the 
precipitate washed first with weak spirit, and afterwards with 
strong alcohol. The antimoniate is dried, transferred to a 
porcelain crucible, and fused with potassium cyanide. 
Metallic antimony is thus obtained, which can be washed 



IO8 Quantitative Chemical Analysis. 

from adhering salts, dried, and weighed. The nitrate con- 
taining the tin is boiled to expel the alcohol, diluted if 
necessary, acidulated with dilute sulphuric acid, and precipi- 
tated by sulphuretted hydrogen. The tin sulphide is 
filtered off, washed, dried, and transferred to a weighed 
porcelain crucible, and cautiously roasted to oxide, and 
weighed. 

XVI. TYPE-METAL. 
(Separation of Lead, Antimony, and Tin.) 

The alloy in fine powder is treated with nitric acid, and 
tartaric acid is added to the solution. The lead dissolves 
completely, together with the greater part of the antimony. 
The precipitated oxides are filtered off, and separated, as 
in No. XV. The solution containing the lead is evaporated 
to dryness with dilute sulphuric acid, and the lead sulphate 
separated as in No. XIII. The antimony and traces of lead 
in the filtrate are precipitated by sulphuretted hydrogen, 
the sulphide filtered off, washed into a flask, and mixed with 
an excess of yellow sodium or potassium sulphide. The flask 
should be closed with a good cork, and the solution kept 
at a gentle heat. Pour the clear liquid through a filter, and 
repeat the digestion with the alkaline sulphide twice. The 
residue consists of lead sulphide, which may also contain 
copper sulphide : these are separated as in No. XIII. Add 
hydrochloric acid to the alkaline filtrate, until the solution 
is distinctly acid ; allow the liquid to stand, and filter off the 
re-precipitated antimony sulphide. This is dried, transferred 
to a weighed porcelain crucible, moistened with strong nitric 
acid, and treated with ten times its weight of fuming nitric 
acid. The acid boiling at 86 must be employed for this 
purpose : nitric acid of sp. gr. i -42 is not able to effect the 
complete oxidation of the sulphur, as its boiling point is 
about 1 6 higher than the fusing point of sulphur ; by heating 
with this acid the separated sulphur fuses, and forms little 



Fusible Metal. 109 

globules which resist oxidation. The white mass in the 
crucible consists of antimonic acid and sulphuric acid : by 
ignition it is converted into antimony tetroxide Sb 2 O 4 . If 
the amount of sulphur mixed with the precipitated antimony 
sulphide is considerable, it is advisable, before proceeding 
to oxidise with nitric acid, to remove the greater portion of 
it by treatment with carbon bisulphide. 



XVII. FUSIBLE METAL. 

(Separation of Bismuth, Lead, and Tin, with traces of Copper ; 
Iron, and Zinc.) 

The alloy in the state of powder is oxidised with nitric 
acid, and the mass repeatedly digested (three or four times) 
with an excess of ammonia and yellow ammonium sulphide. 

The mixture should be kept in a closed flask, and the 
solution gently warmed. In presence of copper, potassium 
sulphide must be used, as the sulphide of that metal is 
slightly soluble in ammonium sulphide. The tin is dissolved ; 
the bismuth and lead, together with the traces of copper, 
iron, and zinc, remain undissolved. The liquid is filtered, 
and the tin precipitated as sulphide by hydrochloric acid : 
it is filtered off, washed, dried, and roasted in a weighed 
porcelain crucible to the state of oxide. The sulphides of 
bismuth and lead, together with the small quantities of 
copper, iron, and zinc, are dissolved in nitric acid, evapo- 
rated nearly to dryness, water added, and the solution, with- 
out filtering, again treated with sulphuretted hydrogen. The 
lead and bismuth and traces of copper are thus once more 
precipitated as sulphides ; the zinc and iron remain in solu- 
tion. The sulphides are next dissolved in nitric acid, 
sulphuric acid is added, the solution is evaporated to dryness, 
and the lead separated as sulphate. Nearly neutralise the 
solution by the cautious addition of ammonia, add a clear 
solution of common salt, and a large quantity of water. 



IIO Quantitative Chemical Analysis. 

Allow it to stand twenty-four hours, and test the supernatant 
liquid by adding a few drops of water : it ought to remain 
perfectly clear. The bismuth is thus completely precipitated 
as basic chloride (BiCIO, or BiCl 3 . Bi 2 O 3 ). It is filtered 
off, washed with cold water, dried, and fused in a capacious 
porcelain crucible with five times its weight of potassium 
cyanide. The fused mass is treated with water, when the 
metallic bismuth is left behind. The grains of the reduced 
metal are washed with water, and afterwards with spirits of 
wine, dried, and weighed. Separate the copper in the 
filtrate from the precipitated basic chloride of bismuth by 
sulphuretted hydrogen ; redissolve the copper sulphide, and 
precipitate as oxide by sodium hydrate. The iron and zinc 
are separated as in No. XIII. 



PART III. 

SIMPLE VOLUMETRIC ANALYSIS OF SOLIDS 
AND LIQUIDS. 

WE have already indicated the principle of this mode of 
analysis, in showing how it is possible to determine the 
amount of silver in a liquid by the aid of a solution of 
hydrochloric acid, or of sodium chloride, of known strength ; 
and we have also shown how we can ascertain the amount 
of alkali in a solution of sodium or potassium hydrate by 
the use of litmus tincture, and of an acid solution of known 
chemical power. The following examples will serve to 
render the principles of volumetric analysis still clearer. 

If we dissolve a small piece of iron-wire in dilute sul- 
phuric acid, we obtain a solution of ferrous sulphate which 
is almost colourless, or which at most possesses a faint 
green tinge. If we add to the solution some substance 



Volumetric A nalysis. Ill 

which readily parts with its oxygen, the colour will change, 
the greenish tinge will give place to yellow, the ferrous 
salt becoming oxidised to the state of ferric oxide. 

2FeO + O = Fe 2 O 3 . 

A few grams . of potassium permanganate (KMnO 4 ) dis- 
solved in water give a deeply-coloured purple solution. 
Potassium permanganate, when in solution, very readily 
parts with its oxygen ; if we add a few drops of the 
liquid to the solution of ferrous sulphate, containing free 
sulphuric acid, we notice that the colour of the permanga- 
nate is instantly discharged. If, however, we continue to 
add successive quantities of the permanganate we arrive at 
a point when its colour is persistent. Let us consider what 
is the nature of this reaction. The potassium permanganate, 
in presence of free sulphuric acid, is decomposed j perman- 
ganic acid is liberated, and potassium sulphate is formed. 
The permanganic acid, however, in presence of the ferrous 
sulphate and free sulphuric acid, readily parts with its 
oxygen, converting the ferrous salt into the state of ferric 
sulphate, and is itself reduced to the state of manganese 
sulphate, which in solution is colourless. So long as any 
ferrous sulphate remains in solution, this decolourising action 
will continue ; immediately, however, that the whole is con- 
verted into ferric sulphate, the red colour of the perman- 
ganic acid will remain unchanged. 

This reaction may be represented by the equation : 

ioFeSO 4 + 8SO 4 H 2 + 2 KMnO 4 = 5Fe 2 (SO 4 ) 3 
+ K 2 SO 4 + 2MnSO 4 + 8H 2 O. 

If we know the strength of the solution of permanganate 
that is to say, if we determine the number of cubic centi- 
metres we require to add to a solution containing a known 
weight of iron as ferrous sulphate, before the solution is per- 
manently coloured we can employ this solution of perman- 
ganate to determine the amount of iron in any given solu- 



112 Quantitative Chemical A nalysis, 

tion. Let us suppose that we required to add 50 cubic 
centimetres of permanganate solution to 0-5 gram of iron, 
dissolved in dilute sulphuric acid, before the colour was per- 
sistent; then each cubic centimetre of the permanganate 
would be equivalent to o'oi gram of iron. If now we 
added the permanganate to a solution containing iron, say 
from an iron- ore, and found that we needed 25 cubic centi- 
metres before the colour was permanent, we should say 
that the amount of iron in solution was 0-25 gram. 

Instead of potassium permanganate, we may employ 
potassium bichromate as an oxidising agent. The reaction 
which occurs with this reagent may be thus represented : 

6FeSO 4 + K 2 Cr 2 7 + 7SO 4 H 2 = 3 Fe 2 (SO 4 ) 3 
+ Cr 2 (S0 4 ) 3 + K 2 S0 4 + 7 H 2 0. 

This equation tells us that 294*4 parts of potassium 
bichromate will convert 6 eq. or 336 parts of iron from the 
state of ferrous to that of ferric oxide. Potassium bichro- 
mate possesses a bright orange-red colour in solution, and 
if the products of its reaction on ferrous sulphate were 
colourless, we might continue to add the solution of bichro- 
mate until its colour was permanent, when we should know 
that, as the chromic acid was no longer decomposed, the 
whole of the ferrous salt was changed to the state of 
ferric salt. Unfortunately, however, the chromic sulphate 
Cr 2 (SO 4 ) 3 which is produced has a deep green colour in 
solution which entirely masks the tint of the bichromate. 
We are accordingly obliged to have recourse to some other 
method than the persistency of the orange colour, to enable 
us to know when the whole of the ferrous oxide is converted 
into ferric oxide. Ferrous salts give a deep blue precipitate 
or colouration with a solution of ferricyanide of potassium ; 
ferric salts produce no such colouration. If then we sprinkle 
a few drops of the ferricyanide solution on a white surface, 
and from time to time take out a drop of the solution of 
iron which is undergoing oxidation, the gradual diminution 



Volumetric Analysis. 113 

in the intensity of the blue colour will inform us of the 
progress of the reaction, and its cessation will tell us when 
the oxidation is complete. 

A consideration of these cases will enable us to lay down 
the conditions required in a volumetric process. In the first 
place, the reaction which constitutes the basis of the method 
must be constant, even under a diversity of circumstances. 
If, for example, it is modified by the concentration of the 
fluids, or the amount of free acid present, or if precipitates 
are formed during the reaction of variable composition, or 
if the presence of the air seriously affects the process, the 
reaction cannot, except in very special cases, afford the 
basis of a trustworthy method. A volumetric process 
further necessitates that we possess accurate means of deter- 
mining the completion of the reaction. Thus the cessation 
of a precipitate in the case of the silver-salt, and standard 
solution of sodium chloride, denotes that the whole of the 
silver is precipitated. The change of the litmus tincture 
from blue to red indicates that the alkali is neutralised. 
The persistency in the colour of the permanganate solution 
tells us that the whole of the iron is in the state of ferric 
salt. In the case of the bichromate, we learn the same fact, 
from the non-formation of a blue colour with potassium 
ferricyanide. A final reaction must be sensitive, rapid, and 
decisive in its changes ; if it requires considerable time, or a 
large expenditure of the testing fluid, or if it involves the 
passage through a series of closely-related tints or changes 
of colour, it cannot well serve to indicate the termination of 
the intended decomposition. 

In order to carry out a volumetric process, we require : 

1. A solution of the reagent of known chemical strength : 
this we call a standard solution. 

2. The means of accurately determining the completion of 
the reaction. 

3. Accurate measuring vessels (pipettes, litre-flasks, &c.), 
and a graduated instrument termed a burette, for pouring 

i 



114 



Quantitative CJiemical A nalysis. 



determinate quantities of the standard solution into the 
liquid on which it is to act. 

The amount of apparatus specially required for volumetric 
analysis is not very extensive. In addition to a few beakers, 
flasks, porcelain basins, glass stirrers, &c., the student must 



FIG. 36. 




provide himself with a set of measuring flasks, pipettes, 
and burettes. 

The most convenient series of measuring flasks is the 
following : 

(i) 1,000 CD.C.J (2) 500 c.c. ; (3) 300C.C.; (4) 2500.0.; and 



Graduation of Vessels. 115 

(5) looc.c. They should be fitted with well ground glass 
stoppers, and the graduation mark of each should be near 
the middle of the neck. The space between the mark and 
stopper allows the fluid to be more readily mixed by agitation. 
The flasks should be sufficiently thin to be heated without 
risk of fracture. Fig. 36 d represents a convenient form of 
litre-flask. 

The following is the most convenient series of pipettes : 
(i) 100 c.c. ; (2) 50 c.c. ; (3) 25 c.c. ; (4) 10 c.c. ; and (5) 
5 c.c. Several i c.c. pipettes will be also needed ; these 
are readily made from glass tubing. The pipettes have the 
form seen in fig. 36 bb. 

The measuring flasks and pipettes are generally sold with 
the graduating marks, their denomination, and the temperature 
at which the graduation is effected, etched upon them. But 
before employing them the operator must never neglect to 
verify their capacities. It must be borne in mind that 
pipettes are to be graduated to deliver their contents ; 
measuring flasks, however, should be graduated both to 
contain and to deliver. A 50 c.c. pipette, accordingly, needs 
to hold more than 50 c.c. of liquid ; it must hold this 
quantity //w that amount which adheres to the glass when 
the liquid is allowed to flow out. We frequently use the 
measuring flasks to dilute liquids to determinate volumes, 
from which we afterwards withdraw aliquot portions by 
means of the pipettes ; occasionally, however, it is necessary 
to transfer a determinate volume of fluid from the flask ; it 
is desirable, therefore, that the same flask should have a 
double graduation one to contain, the other to deliver. 

1,000 c.c. of distilled water at 4 C. weigh 1,000 grams. 
If, therefore, we place the litre-flask, perfectly clean and 
dry, on one pan of a balance capable of turning with 0-05 
gram when carrying 2 kilos, and tare it, placing 1,000 
grams on the weight-pan, and pouring in water of 4 C. until 
the equilibrium is established; the level of the water 
will indicate to us the proper position of the graduating 

I 2 



Quantitative Chemical A nalysis. 



line. The flask contains 1,000 c.c. of liquid when filled up to 
that line. If we now pour out the water, allow the flask to 
drain for a few seconds, remove the 1,000 grams from the 
weight-pan, and re-adjust the tare of the flask, replace the 
1,000 grams, and again fill up the flask with water at 4, 
until the equilibrium is again established, the level of the 
water will now indicate to us the position of what we may 
call the delivery-mark. The flask filled up to this mark and 
emptied, delivers 1,000 c.c. But a very superficial observance 
of the surface of the liquid in the neck of the flask shows us 
that it is not perfectly horizontal. Unless, therefore, we in- 
variably make some determinate point of the curve to coin- 
cide with the graduating line, our measurements will not be 
uniform. It will be found most convenient to take the 
lowest point of the curve or meniscus as the fixed point. In 
verifying or correcting the graduation of the flask, the true 
mark is scratched with a diamond so as to coincide with the 
lowest point of the curve of water in the neck ; and when it 
is desired that the flask shall be filled with 1,000 c.c., the 
liquid is to be poured in until the lowest portion of its sur- 
face exactly reaches this position. 

The distilled water in a laboratory has very seldom a 
temperature of 4, but as we know from experiment the rate 
at which the liquid expands, it is easy to calculate what would 
be the weight of 1,000 c.c. at any given temperature. This 
weight may be obtained from the following table. 

The weight of 1,000 c.c. of water of t C., when determined 
by means of brass weights in air of o C., and of a tension 
07 6m., is equal to 1,0003: grams.* 



t 


o 


I 


2 


3 


4 


5 


6 


7 


8 


9 


X 


1-25 


1-20 


I'lS 


1-13 


I'I2 


I'I2 


1-14 


1-16 


I '21 


1-27 



* Watts's 'Dictionary of Chemistry, ' vol. i. p. 256. 



Graduation of Pipettes. 



117 



t 


IO 


II 


12 


13 


1.4 


15 


16 


17 


18 


19 

2'55 


X 


1-34 


i'43 


I '52 


I-6 3 


I- 7 6 


I-8 9 


2-04 


2-20 


2-37 


t 


20 


21 


22 


23 


2 4 


25 


26 


27 


28 


29 


X 


274 


2'95 


3-17 


3'39 


3^3 


3-88 


4'i3 


4'39 


4-67 


4 '94 



The student is now in possession of all the data required 
to graduate his measuring vessels. He should fill a large 
beaker with distilled water, place it in the balance-room, and 
ascertain its temperature. Let us suppose that it is 15, and 
that he requires to graduate his litre-flask to contain 1,000 
c.c. On reference to the table, we see that the value of x 
corresponding to 15 is 1*89; accordingly, the weight of 
water necessary to be poured into the flask is 1,000 1*89= 
998-11 grams. In graduating the 250 c.c. flask, he would 
of course take one-fourth of this amount, viz., 249-53 grams. 
He places 998-1, or 249-5 grams, as the case may be, on the 
weight-pan, in addition to the tare of the flask, and fills up 
the flask with water until it is exactly equipoised. He then 
marks with a diamond on the neck of the flask the position 
of the lowest point of the meniscus. He now repeats the 
observation in the manner already described in order to obtain 
the graduation for delivery. 

He next proceeds to re-graduate his pipettes. The light 
frame A B (fig. 37), made of stout brass wire, carries two clips 
of thin sheet brass closed by sliding collars ; through the 
lower clip is inserted the upper end of the pipette to be 
graduated : this is connected by caoutchouc tubing with the 
glass stopcock c, to which a short length of thermometer 
tube can be attached, as shown in the figure. To begin with, 
the thermometer tube is removed, and a piece of wider glass 
tubing placed in the caoutchouc tubing, and, the stopcock 
being opened, the pipette is filled by suction with distilled 



118 



Quantitative Chemical Analysis. 



FIG. 37. 



water a centimetre or so above the mark. The object of 
the glass tube is to prevent the caoutchouc being moistened 
by the lips. The end of the pipette which has been dipped 
beneath the surface of the water is dried by a cloth, the stop- 
cock is reopened, and the water is allowed to flow out again 
by its own weight into the beaker. 
As soon as the flow of water has 
ceased, the pipette is held vertically 
for three or four seconds to allow 
the liquid adhering to the glass to 
flow down into the stem; the end 
is then caused to touch the surface 
of the water. Of the various me- 
thods of delivering pipettes, this is 
most accurate. The glass tube is 
withdrawn from the caoutchouc, and 
the short length of thermometer 
tubing, the end of which is drawn 
out before the lamp so as to make 
the bore of very small diameter, is 
placed in its stead. The whole is 
then suspended from the arm of a 
balance turning with 0*05 gram, in 
the manner represented in fig. 37, 
and accurately counterpoised. The 
pipette is removed from the balance, 
the thermometer-tube is withdrawn, 
and the wide glass tube reinserted ; 
the cock is opened, and the water 
is again drawn into the pipette 
one or two centimetres above the 

mark already etched upon the stem. The end of the pipette 
is again wiped with a dry cloth, the glass tube is replaced by 
the thermometer tubing, and the pipette is again suspended 
from the balance arm. The temperature of the water is ob- 
served j suppose it to be 15, and that the pipette is to deliver 
50 c.c., we find from the table that the weight of water possess- 




Graduation of Pipettes. 1 19 

ing this volume is ? ^- = 49 -90 grams. This weight 

20 

is accordingly placed on the weight-pan, in addition to the 
tare, and the balance is caused to oscillate. In all proba- 
bility the pipette and its contents will be too heavy ; the 
cock is now opened, and one or two drops of water are 
allowed to flow out into a beaker placed below. On 
account of the slowness with which the air finds it way 
through the narrow bore of the thermometer-tube, the 
number of drops may be regulated with great nicety. 
Successive drops are thus allowed to flow out until the 
balance is in equilibrium. The lowest part of the meniscus 
is then marked on the stem. The pipette will now deliver 
50 c.c., if emptied in the manner described. The determi- 
nation should be repeated ; if made with proper care, the 
level obtained in the second experiment will be identical 
with that found in the first. The capacities of the remain- 
ing pipettes are verified in the same manner. 

The Burette. This instrument serves to deliver definite 
volumes of the standard solutions. Various forms of the 
burette have been devised, but the most convenient modifi- 
cations are those of Gay-Lussac and Mohr. Gay-Lussac's 
burette is seen in fig. 36 c. It consists of a tube about 
30 centimetres long, and 1-4 to 1-8 centimetres wide, sealed 
at one end, and furnished with a narrow side-tube, starting 
near the bottom, and running close to the side, to within 
about 2 centimetres from the open end, where it is bent 
slightly in the manner seen in the figure. These burettes 
are usually made in two sizes one to hold 25 c.c., and 
graduated in ^ c.c. ; the other to hold 50 c.c., and gradu- 
ated in i c.c. They are graduated 'for delivery.' The 
correctness of the graduation should be tested previous to 
use, by filling the burette with distilled water, and emptying 
it through the side-tube, until the bottom of the meniscus 
is coincident with the lowest division. The temperature of 
the water is then ascertained, the burette is tared, and the 



120 Quantitative Chemical Analysis. 

weight of the water supposed to be required to fill it is 
placed on the weight-pan, and the instrument is filled up 

FIG. 39. 



FIG. 38. 





with the distilled water 
until the balance is in equi- 
librium ; if the meniscus is 
now coincident with the 
zero point, the instrument 
is correctly graduated. Dif- 
ferences of less than 0-05 
c.c. may generally be neg- 
lected. So long as its bore 
is uniform, and the divisions 
are of an invariable width, 
the instrument need not be 

discarded, even if the lowest point of the meniscus is not 
coincident with the zero. Let us suppose that on a 50 c.c. 
burette, on which 250 divisions=5o c.c., the lowest point of 



Graduation of Burettes. 



121 



FIG. 40. 



the meniscus was coincident with the division correspond- 
ing to 0-8 c.c., then obviously 2504, or 246 divisions, are 
equivalent to 50 c.c., and i di vision =2% c.c., or 0-203 c - c - 
Accordingly, the indications of the burette must be multi- 
plied by i 'oi 6, to give the correct number of cubic centimetres 
delivered. Thus, if in an analysis we had delivered ap- 
parently 25 c.c. from such a burette, we should in reality 
have delivered 25 x roi6 = 25-4 c.c. of liquid. 

In using the burette, the edge of the side-tube should 
be greased slightly ; this prevents the possibility of liquid 
adhering to the outside of the tube when the burette 
is replaced vertically in its support. With a little practice it 
is easy to deliver the liquid in a stream or in drops ; when 
the burette is brought to the vertical, to be read off, it is 
necessary to wait for a 
few seconds before 
making the observa- 
tion, in order that the 
liquid may attain a con- 
stant level. 

Mohr's burette is 
seen in fig. 38. It is 
simply a divided tube, 
contracted at its lower 
end, and fitted with a 
short length of caout- 
chouc tubing into which 
is inserted a glass jet. 
The sides of the caout- 
chouc tube can be 
pressed together by 
means of the spring 
clamp. This form of 
burette is not so gene- 
rally applicable as that 
of Gay-Lussac, since the caoutchouc is acted upon by 




122 Quantitative Chemical A nalysis. 

several of the substances employed in standard solutions. In 
the more modern form of the burette, a glass stop-cock is sub- 
stituted for the india-rubber and clamp. This modification 
(fig. 360) leaves nothing to be desired. It is especially con 
venient where a great number of analyses of the same kind 
have to be made, as in metallurgical laboratories, chemical 
works, &c. In such cases the burette may be conveniently 
arranged as shown in fig. 39. The bottle A contains the 
standard solution ; on opening the clamp a, the liquid fills 
the burette gradually, and without the formation of air- 
bubbles. Fig. 40 shows another method of connecting the 
burette with the reservoir of the standard solution. The 
liquid is driven into the burette by simply blowing through 
the caoutchouc tube a. The graduation of Mohr's burette 
may be verified by filling the instrument with water, and 
allowing successive quantities of, say, 10 c.c., to flow out 
into a weighed beaker. If the 10 c.c. weigh 9-98 grams, the 
burette is correctly graduated. Of course- due care must be 
taken to allow the liquid adhering to the sides" of the tube 
to flow down before the level is read off. 

The correct reading off of the burette may be facilitated 
by the use of a little device recommended by Mohr. A 
broad strip of black paper is pasted on a white card or 
sheet of white paper, and this is held behind the burette, 
so that the edge of the black paper is about 2 mm. 
below the dark zone of the liquid. The lower edge of the 
liquid is thus sharply defined, and may be read off with 
certainty. A little caoutchouc band, slipped round the tube, 
and through the card, renders the arrangement more con- 
venient. 

In reading off the Gay-Lussac burette, the level of the 
liquid should be brought to the direct line of vision. This 
may conveniently be determined by pasting a narrow strip 
of black paper upon the side of the room, ten or twelve feet 
from the operator, and on a level with his eye. The burette 
is held 'perpendicularly between the thumb and first finger. 



Graduation of Burettes. 



123 



FIG. 41. 



in such position that the black strip appears immediately 

behind the level of the liquid. Greater certainty in reading 

off may be attained by the use of Erdmann's float (fig. 

41). It is simply an elongated glass bulb, 

somewhat smaller in diameter than the burette, 

containing a small quantity of mercury. The 

upper end is drawn out, sealed, and bent into 

a little hook, by which the bulb can be lifted 

in and out of the burette by the aid of a bent 

wire. Round the bulb runs a line a, etched 

by means of hydrofluoric acid, or scratched 

by a diamond. The coincidence of this line 

with the division of the burette is taken as 

the reading. The float should move easily 

within the burette, and so that the line is 

always parallel with the divisions ; by its 

means the volume of the liquid delivered may 

be determined to within 0*005 c.c. 

In certain cases the quantity of the liquid 
delivered is determined by weight. The 
solution is contained in the little weighed 
flask seen in fig. 42. The required amount 
is poured through the delivery-tube, 
which should be slightly greased at the 
edge. By weighing the apparatus before 
and after delivery, the amount of liquid 
employed is at once determined. A 
method of making a simple form of 
this apparatus is described under the 
section ' Ash Analysis.' 



FIG. 42. 




We now proceed to the experimental study of certain 
volumetric processes. We shall describe here a few typical 
processes to enable the student to familiarise himself with 
this mode of estimation. Other methods will be given in 
Part IV. 



124 Quantitative Chemical Analysis. 

I. DETERMINATION OF CHLORINE BY STANDARD SILVER- 
SOLUTIONS. 

If we add silver nitrate to a solution of a chloride, say of 
common salt, we obtain a white precipitate of silver chloride; 
if we continue to add the silver solution, the formation of 
this substance goes on until the whole of the chlorine is 
precipitated. If we add silver nitrate to a solution of 
potassium chromate, we obtain a dark-red precipitate of 
silver chromate. If now we mix the alkaline chloride and 
chromate together, and cautiously add, little by little, the 
silver nitrate solution, we notice that the chloride is first de- 
composed ; and white silver chloride continues to be formed 
so long as any chlorine remains in solution. It is only after 
the whole of the chlorine is precipitated that we observe the 
formation of the dark-red silver chromate. This principle 
constitutes the basis of an accurate volumetric process. To 
carry it out we require a standard solution of pure silver 
nitrate free from excess of acid, and a solution of potassium 
chromate. 

Preparation of Pure Silver. Chemically-pure silver is 
frequently needed in volumetric analysis. We not only re- 
quire it in the present process : we shall have occasion to 
use it in determining the strength of the hydrochloric acid 
solution employed in alkalimetry. It is therefore desirable 
that the student should prepare at one time all that he will 
need in this and subsequent operations. 

About 50 grams of standard silver (composed of 12-3 parts 
of silver and i part of copper) are dissolved in dilute nitric 
acid in a thin porcelain basin at a gentle heat ; the solution 
is evaporated to dryness, and the residue heated to fusion. 
The cooled mass is then dissolved in ammoniacal water, 
allowed to stand for a short time, and filtered into a large 
flask. The filtrate is diluted to 2\ litres. 50 c.c. are with- 
drawn, heated nearly to boiling, and mixed with a solution 
of neutral ammonium sulphite (prepared by neutralising 



Pure Silver. 125 

ammonia with sulphur dioxide gas) added drop by drop, 
until the liquid is decolourised. The ammonium sulphite solu- 
tion is then mixed in the proportion demanded by this trial 
with the 2450 c.c. of liquid in the flask, which is then closed 
air-tight. In about 48 hours, nearly a third part of the 
silver is deposited as a crystalline powder, and the remainder 
is thrown down on heating the liquid to 60 or 70 for a short 
time. The liquid is now completely decolourised, unless it 
contains nickel or cobalt, which are not infrequent impuri- 
ties in standard silver, when it will be light-green or pink. 
The precipitated silver is washed with distilled water, and 
digested with strong ammonia, again washed and dried. It 
is then fused with about 3 grams of ignited and powdered 
borax, previously mixed with a little sodium nitrate, in an 
unglazed porcelain crucible, and the button of metal is 
washed with hot water, and rubbed, if necessary, with a little 
sea-sand. It should then be rolled out into foil, sufficiently 
thin to be readily cut with a pair of scissors. 

Preparation of the Standard Solution of Silver. 10794 
grams of the foil are weighed out, placed in a porcelain basin 
provided with a glass cover, and dissolved in dilute pure 
nitric acid on the water-bath. When the whole of the metal 
is dissolved, the under surface of the glass is rinsed into the 
dish, and its contents are evaporated to complete dryness on 
the water-bath, and gently heated over the lamp until the salt 
fuses. The dry and neutral silver nitrate is dissolved in pure 
water, and the solution carefully poured into the litre-flask, 
the dish being repeatedly washed out with fresh portions of 
distilled water. The flask is now filled up to the containing- 
mark with distilled water, the stopper is inserted, and the 
flask well agitated. The liquid constitutes a deci-normal solu- 
tion of silver nitrate : i c.c. = 0-010794 gram silver; it is 
therefore equivalent to 0-003546 gram of chlorine, or 
003646 gram of hydrochloric acid, or -00585 gram of sodium 
chloride. The solution should be poured into a perfectly 



126 Quantitative Chemical Analysis. 

clean and dry bottle, provided with a well-fitting stopper : it 
should be labelled ' Deci-normal Silver Solution.' NOTE. 
By a normal solution is to be understood a solution con- 
taining i eq. of the substance, in grams, dissolved in 1,000 
c.c. of liquid. Thus a normal solution of silver would con- 
tain 107 '94 grams of the metal in i litre of the solution. A 
normal solution of hydrochloric acid would contain 36-46 
grams of HC1 in 1,000 c.c. A deci-normal solution contains 
one-tenth of an equivalent; a centi-normal the one-hundredth 
part of an equivalent, in grams, per litre. 

Preparation of Potassium Chr ornate Solution. The com- 
FIG. 43. mercial salt is recrystallised until it 

is free from chlorine : the solution 
acidified with nitric acid should not 
give the least turbidity on the addi- 
tion of a drop of silver nitrate solu- 
tion. Its solution should be kept 
in a little bottle A, through the 
cork of which runs a narrow tube 
with a mark d scratched upon it. 
This allows of a constant quantity 
of the solution to be withdrawn 
from the bottle. (Fig. 43.) 

The Process. Aquantity of pure sodium chloride (see p. 79), 
is powdered, and gently heated, and whilst warm introduced 
into a small tube, fitted with a good cork. About i gram of 
the chloride is accurately weighed out into the \ litre flask, 
and dissolved in distilled water ; the flask is filled up to the 
containing-TC&fo, and the solution well agitated. The burette 
(either Gay-Lussac's or Mohr's may be used) is rinsed out 
with a little of the standard silver solution (which is thrown 
into the * silver residue ' bottle) and filled up to the zero with 
the silver solution. 50 c.c. of the solution of sodium chloride 
are withdrawn from the flask, and run into a porcelain basin, 




Estimation of Chlorine. 127 

and mixed with a measure of the chromate solution ; the 
silver solution is added, drop by drop, from the burette, 
until the red colour of the silver chromate is permanent 
Each drop of the silver solution forms a red spot in the 
yellow liquid, which quickly disappears, so long as any chlo- 
ride remains in solution ; immediately all the chlorine 
is precipitated, the red colour of the chromate of silver is 
unaltered. The process is now at an end. The volume of 
the silver solution employed is read off, corrected, if neces- 
sary, for the error of the graduation (see p. 121), and o-i 
c.c. subtracted, this expenditure of silver being required to 
render the final reaction evident The analysis should 
be repeated on a second portion of 50 c.c. of solution. 
An actual example will render the method of calculation 
clear. 1*0850 gram of pure salt was dissolved in 250 c.c. 
of distilled warter. 50 c.c. of this solution required in experi- 
ment L, 37-1 c.c.; in experiment II., 37-2 c.c. ; in experiment 
III., 37 -2 c.c. of silver solution. Mean 37-17 c.c. Subtract o-i 
for final reaction. 37*07 x -003546 = 0-1314 gram of 
chlorine. 50 c.c. of liquid contain 0-217 gram of salt 

Accordingly, the salt contains ^=60*56 per cent 

chlorine. Theory requires 60 60 per cent. 

The remainder of the solution of the sodium chloride 
should be poured into a clean and dry stoppered bottle, and 
its strength marked on a label attached to the bottle. It 
will be useful in cases where, in determining chlorine by 
this method, we imagine that we have added an excess of 
silver solution. We have only to add a definite volume, 
say i c.c. of the solution, to the turbid liquid, and after the 
last trace of silver chromate has disappeared, again add the 
silver solution until the final point is exactly obtained. In 
the case above cited, 50 c.c. of salt solution equal 37*07 c.c. 
of silver solution ; accordingly i c.c. of salt = 0-7 c.c. of 
silver. 0-7 + 0-1 (for final reaction) or 0-8 c.c. subtracted 
from the total amount of silver solution employed, gives the 
exact amount used in the analysis. 



128 Quantitative Chemical Analysis. 

II. INDIRECT DETERMINATION OF POTASSIUM AND SODIUM 
BY MEANS OF STANDARD SILVER SOLUTION AND POTASSIUM 
CHROMATE. 

The deci-normal silver solution may be used for a variety 
of estimations in which the amount of chlorine present may 
be taken as a measure of the other constituents. We shall 
have occasion to mention several of the applications of this 
solution in the General Part (Part IV.). We have already 
described the method of estimating potassium and sodium 
by gravimetric analysis : as an example of the above-men- 
tioned applications of the solution of standard silver, we 
proceed to show how these alkalies, when together, may be 
estimated by volumetric analysis. 

From 3 to 4 grams of pure Rochelle salt (C 4 H 4 KNaO6. 
4.H 2 O) are gently heated in a platinum basin nntil the water 
of crystallisation is expelled. The temperature is then in- 
creased until the mass is completely carbonised j the heat 
should not exceed low redness, or a loss of alkali will be 
incurred. The alkaline carbonates in the charred mass are 
then dissolved in a small quantity of hot water, filtered, and 
the charcoal repeatedly washed with successive quantities of 
water. A slight excess of pure hydrochloric acid is added 
to the filtrate contained in a weighed platinum basin, and 
covered with a watch-glass, and as soon as the evolution of 
gas ceases, the under surface of the watch-glass is rinsed 
into the basin, and the liquid is evaporated to complete dry- 
ness, and heated in the air-bath to 180. The alkaline 
chlorides are weighed and dissolved in a small quantity of 
water, the solution poured into a ^-litre flask, and diluted to 
the containing-vxsdL. 50 c.c. are then withdrawn and titrated 
with silver solution and potassium chromate in the manner 
described. From the weight of the chlorides, and of the 
chlorine they contain, we can readily calculate the propor- 
tion of the alkalies in the mixture. 

Let x stand for the potassium, and y for the sodium, s for 



A Ikalimetry. 1 29 

the weight of the mixed chlorides, and A for that of the 
chlorine found. 

[(S-A). 1-54] -- A 

"* ~~~ 7 

0-63 

y - A - [ (S A) 0-91] 

- a -? -^-a-s 

III. ESTIMATION OF CHLORIC ACID. 

Weigh out about 0*5 gram of dry potassium chlorate into 
a small beaker in which you have previously placed about 
20 grams of thin sheet- zinc covered with spongy copper, 
in the manner described on p. 96. Add about 25 c.c. of 
water, cover the beaker with a watch-glass and boil the 
liquid gently for about an hour. Add water to the beaker, 
filter the liquid* into a porcelain basin, and wash the zinc and 
copper in the beaker repeatedly with hot water. By the 
action of the nascent hydrogen, the alkaline chlorate is 
reduced to chloride. The nitrate should be quite neutral. 
Determine the amount of chlorine in the liquid by standard 
silver and potassium chromate solutions, i c.c. of the solu- 
tion is equivalent to 0-01226 gram of potassium chlorate. 

Example. 0-2492 gram of potassium chlorate, treated in 
the manner described, required 20*2 c.c. of deci-normal silver 
solution. 20-2 x -01226=0-2477 gram potassium chlorate. 

IIlA. DETERMINATION OF CHLORINE IN PRESENCE OF 
SULPHITES. 

Add to the solution of the salts a very slight excess of a 
solution of potassium permanganate free from chlorine ; 
neutralise the liquid with pure soda, and then add the 
potassium chromate and standard silver nitrate solution in 
the usual manner. This preliminary oxidation of the sul- 
phurous acid is necessary, otherwise the chromate solution 
would be reduced. 



130 Quantitative Chemical Analysis. 

ALKALIMETRY. 

Preparation of Normal Solution of Hydrochloric Add. i c.c. 
=0-03646 gram HC1. Messrs. Roscoe and Dittmar have 
shown that if a solution of hydrochloric acid containing 20-2 
per cent. HC1 be boiled under the ordinary pressure of the 
atmosphere, the acid and water distil over in the proportion 
in which they are contained in the boiling liquid. If we take 
a solution of the acid having approximately this composition 
and boil it in a retort until about half of it has distilled over, 
we may be sure that the residue contains about 20*2 per cent. 
of acid. This principle affords the basis of a method of 
preparing a standard solution of hydrochloric acid. 

We commence by ascertaining the specific gravity of a 
strong solution of hydrochloric acid by means of the hydro- 
meter (see Appendix), and we then add water to it until its 
specific gravity is reduced to i *i. A solution of this strength 
contains about 20*2 per cent, of HC1. The amount of water 
x which we require to add to a measured quantity of strong 
hydrochloric acid A, of specific gravity a, to reduce it to the 
specific gravity b (in this case i'i), is found from the formula 

. A (a - b) 
~b-T 

Let us suppose that the specific gravity of our acid is 1*16: 
in order to bring its specific gravity down to i'i we shall 

require to add to every 100 c.c. of acid I0 '* - "" * *' = 

ri i 

60 c.c. of water. 500 c.c. of strong acid are mixed with the 
quantity of water required to reduce its specific gravity to i -i, 
and the mixture is brought into a retort connected with a good 
condensing arrangement, and boiled until nearly one-half the 
amount has distilled over. The ebullition may be rendered 
more regular by throwing a few scraps of clean platinum foil 
into the liquid. The residue contains about 20-24 per cent, 
of HC1. 1 80 -8 grams of such acid, when diluted to a litre, 
furnish a solution which is approximately normal. 



Standard Hydrochloric Acid. 131 

Since this acid is of frequent application, the student 
should prepare from 2 to 3 litres of its solution. To deter- 
mine its exact strength, 50 c.c. of the acid solution are run 
into the -J-litre flask, and diluted to the contatning-raaxk after 
shaking, i c.c. approximately equals 0*003646 gram HC1. 
Call this solution A; 25 c.c. of A are further diluted to 250 
c.c. i c.c. = '0003646 gram HC1. Call this solution B. 

Weigh out exactly 1*0794 gram of pure silver into a 
bottle of about 300 c.c. capacity, provided with a well-fitting 
stopper, and dissolve the metal in pure dilute nitric acid. 
The solution should be heated on the water-bath, and the 
fumes of the oxides of nitrogen should be blown out of the 
bottle from time to time. When the silver is completely 
dissolved, and the liquid on agitation gives no trace of red 
fumes, the bottle is removed from the bath and allowed to 
cool. A 100 c.c. pipette is rinsed with a small quantity of 
solution A, which is allowed to flow away. The pipette is 
filled to the mark with the solution A, and emptied into the 
silver solution. The stopper is inserted, and the solution is 
briskly agitated for some time until the silver chloride settles 
out completely and leaves the liquid almost clear. If the 
hydrochloric acid is of exact strength, that is, if it is strictly 
normal, and if 1*0794 gram of pure silver has been 
accurately weighed out, we ought of course to have neither 
silver nor chlorine in excess in solution. In all probability 
we shall have one or other of the bodies in excess. To 
determine which of the two remains in solution, we add i c.c. 
of deci-normal silver solution and note whether a further 
turbidity ensues. If the liquid remains clear, the silver is 
in all probability already in excess : if it becomes turbid it is 
a sign that the chlorine is present in excess. In the latter 
case the solution is again vigorously agitated until the 
liquid is once more clear. We will assume, by way of 
example, that the addition of the i c.c. of deci-normal silver 
solution produced a turbidity. Another i c.c. of silver solu- 
tion is added to the liquid, and we again observe whether a 
turbidity is caused. Let us suppose that the liquid now 



132 Quantitative Chemical Analysis. 

remains clear : it is evident that an excess of silver is in 
solution. The total amount of silver in the bottle is there- 
fore 1*0794 + ('010.794 x 2) = 1*100988 gram. If we 
determine the amount in solution, we can at once tell how 
much is precipitated as chloride, and accordingly calculate 
the amount of HClin the 100 c.c. of diluted acid. It is 
evident that we have at least 0*010794 gram of silver in 
excess, since the addition of the last i c.c. of deci-normal 
silver solution failed to produce a turbidity. We now add 
i c.c. of solution A, and shake the liquid vigorously until it 
is clear. In all, we have added 101 c.c. of solution A, and 
1*100988 gram silver. We now add i c.c. of solution B, 
and note whether a turbidity is produced ; if so, we again 
shake the liquid vigorously until it is clear, and add a 
second i c.c. : if the liquid is still rendered turbid, we again 
shake briskly, and add a third i c.c., and so on until the 
addition of i c.c. of solution B no longer produces any 
change. This last c.c. is not counted, since it shows that 
the HC1 is once more in slight excess, and we shall be nearer 
the truth if we assume that only half of the preceding c.c. is 
necessary for precipitation. In working with the centi- 
normal solution (solution B) it is necessary that the liquid 
above the precipitated silver chloride be perfectly clear, and 
that we wait for a few seconds (say \ minute) before we con- 
clude that no further turbidity is caused by the addition of 
i c.c. of solution. Let us suppose that we found it necessary 
to add 5 c.c. of solution B before the liquid remained clear : 
3*5 c.c. are therefore necessary to precipitate the silver in 
solution after the addition of the i c.c. of solution A. If 
the directions given have been properly followed, we may 
assume without sensible error that i c.c. of solution B con- 
tains 0*000365 gram HC1 : this is equal to '0010794 gram Ag., 
and '0010794x3*5 = 0-003778 gram Ag. 101 c.c. of solu- 
tion A are equivalent to 1-100988 '003778 = 1*09721 
gram Ag., and accordingly 

107-94 : 36-45 :: 1*09721 : x 

x = 0-370514. 



Standard Sulphuric A cid. 133 

101 c.c. of solution A contain 0-370514 gram HC1. Ac- 
cordingly the 500 c.c. would contain 1*83423 gram HC1. 
But the 500 c.c. of A are equivalent to 50 c.c. of the original 
acid. Accordingly i c.c. of the original acid contains 
036684 gram HC1, instead of "03645 gram, the quantity 
required to constitute the normal acid. Instead of diluting 
the acid to bring it to the exact strength, it is better to ex- 
press the difference by a small factor: in this case 3 i-= 

03645 

1-0064. The acid is accordingly labelled ' Standard Hydro- 
chloric Acid, i c.c. = 0*03645 x 1*0064 gram HCl? 
i c.c. of the acid is equivalent to 0*10794 x 1*0064 gram 
Ag., or 0-04004 x 1-0064 gram NaHO. 

At least two determinations should be made of the 
strength of the acid before it is used, and the mean result 
-should be taken as indicating the correct value. 

Preparation of Normal Sulphuric Acid Solution. In 
certain processes the use of standard hydrochloric acid is 
inadmissible ; in such cases we may generally employ a 
normal solution of sulphuric acid. This may be prepared 
by diluting about 60 c.c. of concentrated and pure sulphuric 
acid with five or six times its volume of water, allowing the 
mixture to cool, and making it up to 2 litres. If the sul- 
phuric acid used was concentrated, the solution will now 
contain rather more than 49 grams H 2 SO 4 per litre. To 
determine its exact strength, weigh out about 2 grams of 
recently heated pure sodium carbonate into a weighed 
platinum basin, dissolve it in a small quantity of water, 
cover the solution with a watch-glass, draw the watch-glass 
aside and add 25 c.c. of the acid. Place the liquid on a water- 
bath, and as soon as the evolution of gas has ceased, remove 
the cover, rinse its under surface into the dish, and evaporate 
the liquid to complete dryness. Heat to 180 in the air-bath 
until the weight is constant. The calculation is very simple. 
SO 4 =96 has displaced CO 3 = 60. The increase in weight of 
the dish is proportional to the amount of sulphuric acid 



134 Quantitative Chemical Analysis. 

employed so long of course as there is excess of sodium 
carbonate present. The amount of sulphuric acid x in the 
25 c.c. is thus found : 

The difference between the equivalent of SO 4 and CO 3 , 
yiz. 36, is to the equivalent of SO 4 H 2 , viz. 98, as the differ- 
ence between the first and second weighing of the platinum 
basin is to the sulphuric acid present in the 25 c.c. 

Example. Weighed out 17210 gram of pure dry sodium 
carbonate, added 25 c.c. of the acid, evaporated to com- 
plete dryness, and heated in the air-bath. The difference 
between the weight of the dish + carbonate, and dish -f- 
mixed sulphate and carbonate was 0*465 gram : then 

36 I 98 \\ 0*465 : x 

,.. , 1*266 
x = 1*266 and = 0*05064. 

2 5 

i c.c. of the acid accordingly contained 0*05064 gram 
H 2 S0 4 . 

This method of determining the exact strength of the acid 
solution is quite as accurate, and certainly more convenient, 
than precipitating the sulphuric acid as barium sulphate. 
By way of control the acid solution was treated with barium 
chloride and the precipitate washed, dried, and weighed. 
10 c.c. of the solution gave 1*2043 gram BaSO 4 as the 
mean of three concordant experiments. This is equal to 
0*5063 gram of sulphuric acid. The determination by 
sodium carbonate gave 0*5064 gram. The determination 
of the strength of the acids employed in volumetric analysis 
may, in many cases, be accurately and expeditiously made 
by means of metallic sodium. About 0*5 gram, of clean 
freshly-cut sodium is placed in a short wide-mouthed test-tube, 
previously weighed, together with its well-fitting cork ; the 
tube containing the sodium is corked and again weighed ; the 
metal is thrown upon about 15 c.c. of cold water contained 
in a porcelain basin, which is then quickly covered with a 
large watch-glass. When the action is at an end, the glass 
is raised, a few drops of litmus solution (vide infra) added, 



Standard Soda Solution. 135 

and the acid is run in from a burette until the red tint is 
permanent. 

The determination of the strength should be repeated once 
or twice, and the mean result taken as expressing the true 
value of the acid. If the solution is approximately normal 
it is better not to dilute it, but to calculate the factor required 
to bring its strength to the normal value. Thus the factor of 
the above-mentioned solution would be -f^ = 1*0334. 
It would be labelled therefore ' Normal Sulphuric Acid, i 
c.c.='Q^gram x 1-0334 SO 4 H 2 .' 

If the acid is much above the normal value it will be 
more convenient to dilute it so as to make it as nearly as 
possible of the proper strength. Thus, supposing that we 
had found that i c.c. contained 0-055 gram H 2 SO 4 , 1000 c.c. 
would contain 55 grams. Consequently, according to the 
proportion 

49 : 1000 M 55 : x. x 1123. 

we must add 123 c.c. of water to every 1000 c.c. of the acid 
solution. This may be best eifected by rilling the litre flask 
up to the wntazmng-maxk with the acid solution, and emptying 
it into the dry and clean bottle in which it is to be preserved. 
Now pour into the litre flask 123 c.c. of water (by means of 
the 100 c.c. pipette and the burette), shake the liquid about 
in the flask and pour it into the bottle, and shake the 
mixture. Again pour about half the acid liquid back into 
the litre flask, shake and transfer it once more to the bottle. 

Preparation of Normal Caustic Soda Solution. Dissolve 
from 42 to 45 grams of sodium hydrate in 800 c.c. of water, 
and titrate the solution by normal acid and litmus tincture. 
The alkaline solution is then diluted until it possesses the 
normal strength. 

The caustic soda solution may also be obtained by dis- 
solving about 150 grams of pure dry sodium carbonate in 3 
litres of water, boiling the solution in a clean iron vessel, and 
adding, little by little, 80 grams of freshly-burnt lime made 



136 



Qtiantitative Chemical Analysis. 



into a cream with water.* The mixture must be boiled until 
a small quantity of the clear solution no longer effervesces 
on the addition of an acid in excess. The iron vessel is then 
closely covered, and after standing 12 or 14 hours the clear 
alkaline solution is drawn off by the aid of a syphon. 

Preparation of Litmus Solution. 5 or 6 grams of coarsely- 
powdered litmus are digested with about 200 c.c. of distilled 
water for a few hours. The clear solution is decanted from 
the sediment, and very dilute nitric acid added, drop by drop, 
until the colour is changed to violet. The solution must be 
neither red nor blue, but between the two in colour ; when 
properly neutralised less than ^ c.c. of the standard acid 
should distinctly redden the solution of i c.c. in 100 c.c. of 
water ; on the other hand the same amount of standard alkali 
should render the colour decidedly blue. The solution should 
be kept in a wide-mouthed bottle, the cork of which is so cut 
FlG 44 that the air has ready access 

to the interior of the bottle, 
otherwise the liquid quickly 
loses its colour. Through the 
cork is fitted a short tube on 
which is a mark ; this tube 
serves to deliver a determinate 
volume of the litmus solution. 
Determination of the Strength 
of the Caustic Soda Solution. 
25 c.c. of the standard sul- 
phuric acid solution are poured 
into a porcelain basin, mixed 
with a measure of litmus solu- 
tion, and the alkaline liquid is 
added drop by drop from a 
burette until the colour is just 
turned to blue. Repeat the 

* If this quantity of pure dry sodium carbonate is not at hand, 250 
grams of the bicarbonate are heated to dull redness in a platinum basin, 
in small portions at a time, for ten or fifteen minutes, to expel the 
carbon dioxide. The salt is then treated as above. 




Soda- Ash. 137 

determination, take the mean of the two observations, and 
dilute the alkaline solution until it corresponds volume for 
volume with the standard acid. Thus, supposing you have 
found that 25 c.c. of acid required 22 c.c. of soda for neu- 
tralisation, you will require to add 3 c.c. of water to every 
22 c.c. of lye, or each litre of alkaline liquid will require the 
addition of 136 c.c. of water. 

The diluted liquid should be poured into a large bottle 
fitted with a syphon and wide tube as shown in Fig. 44. The 
wide tube is filled with soda-lime in small pieces to prevent 
the entrance of carbon dioxide. A thin layer of refined 
petroleum or paraffin oil poured on the surface of the liquid 
greatly tends to the preservation of its strength. The exact 
strength of the diluted liquid should then be determined by 
neutralising varying qualities, say 25 c.c., 30 c.c., and 50 c.c., 
of standard acid in the manner above described. The mean 
result of the observations should be taken as the true value. 

IV. VALUATION OF SODA-ASH. 

Soda-ash is a crude sodium carbonate; its value depends on 
the amount of available sodium carbonate which it contains. 
Its impurities, in addition to moisture, mainly consist of 
sodium hydrate, sulphate, chloride, silicate, and aluminate. 
It also not unfrequently contains sodium sulphide, sulphite, 
and thiosulphate. 

Weigh out about 10 grams of the powdered sample into a 
weighed platinum crucible, and heat gently for twenty or 
thirty minutes over the lamp ; place the crucible in the 
dessicator, and weigh when cold. The loss of weight gives 
the amount of moisture contained in the sample. Transfer 
the weighed salt to a beaker, wash out the crucible, and 
dissolve the salt in a small quantity of water, filter (if neces- 
sary) into the J-litre flask, wash the filter thoroughly, and 
dilute to the containing-mark, and shake. Take out 50 
c.c. of the solution, corresponding to i gram of soda-ash, 
pour the liquid into a flask, and add 25 c.c. of standard 
sulphuric acid, and boil the solution for some time until the 



138 Quantitative Chemical Analysis. 

carbonic acid is expelled. Add a measure of litmus solu- 
tion, and standard soda-solutin from a burette, drop by 
drop, until the blue colour of the litmus is restored. 

Example. 10-025 grams of soda-ash were heated, dis- 
solved in water, and the clear solution made up to 500 c.c. 
50 c.c. (corresponding to 1*0025 gram soda-ash) were trans- 
ferred to a flask, 25 c.c. of standard sulphuric acid 
(i c.c. = '049 gram SO 4 H 2 x 1*0204) were added, the liquid 
was boiled, mixed with a measure of litmus solution, and 
standard soda added, drop by drop, until the blue colour 
was restored. 10 c.c. of soda solution=9*8 c.c. of standard 
sulphuric acid. 9*2 c.c. of the alkaline liquid were needed. 

10 \ 9*2 \\ 9*8 I 9*0 

Accordingly 259*0=16*0 c.c. of standard acid have been 
used to decompose the 1*0025 g 1 " 3 - 111 of soda-ash. But 
i c.c. of acid contains 0*049x1*0204 gram of sulphuric 
acid ; this corresponds to 0*053 x 1*0204 of sodium carbo- 
nate. The amount of sodium carbonate in the 1*0025 
gram of original soda-ash is therefore -053 x 1*0204 x 16*0 
=0*865 ram > r m IO parts 1*0025 Io 0*865 
=86 '3 per cent. 

The value of pearl-ash may be determined in exactly the 
same manner. It must not be forgotten that dried potas- 
sium carbonate is very hygroscopic ; due expedition must, 
therefore, be employed in weighing this body. The crucible 
should be closely covered during the operation. The 
\ equivalent of potassium carbonate is 69*1 ; accordingly, 
the factor -0691 is used instead of -053 in the calculation. 

V. ESTIMATION OF ALKALINE HYDRATE IN PRESENCE OF 
CARBONATE. 

The crude carbonates of soda and potash not unfrequently 
contain notable quantities of hydroxide. The alkaline lyes 
used by paper, soap, and starch manufacturers also consist 
of mixtures of carbonated and caustic alkalies. 

To estimate the proportion of the two constituents, a deft- 



Sodium and Potassium Carbonates. 139 

nite quantity of the salt or solution, say 15 grams, or 50 c.c., 
is dissolved in water, and diluted to 250 c.c. The total 
amount of alkali is then determined, say in 50 e.c., in the 
manner described on p. 137, viz., by standard acid, litmus, 
and soda solutions. Take out 100 c.c. of the alkaline solu- 
tion, pour it into a 300 c.c. flask, dilute with a little water, 
heat, and add solution of barium chloride so long as a pre- 
cipitate is formed. The reactions which occur in the case 
of sodium carbonate and hydrate are : 

2 NaHO + BaCl 2 = 2NaCl + BaH 2 O 2 , 
and Na 2 CO 3 + BaCl 2 = 2NaCl + BaCO 3 . 

Fill up the flask to the containing-mark, shake, and allow 
the precipitate to settle. Withdraw 100 c.c. of the clear 
liquid, pour it into a porcelain basin, add an excess, but in 
measured quantity, of standard hydrochloric acid, a measure 
of litmus solution, and then determine the excess of acid 
added by means of the soda-solution. Multiply the amount 
of hydroxide found by 7-5 ; this gives the amount of caustic 
alkali in the weight of the substance originally taken. 

VI. ESTIMATION OF SODIUM CARBONATE IN PRESENCE OF 
POTASSIUM CARBONATE. 

Sodium carbonate, on account of its cheapness, is some- 
times employed to adulterate pearl-ash. The quantity of the 
admixed sodium salt may be estimated in the following 
manner. About 5 grams of the mixture are gently heated 
in a weighed platinum crucible for fifteen or twenty minutes; 
the loss on weighing gives the amount of moisture present. 
The dried mass is dissolved in a small quantity of water, 
and filtered, if necessary ; acetic acid is added in slight 
excess, the liquid is heated to expel carbonic acid, and 
mixed with a dilute solution of lead acetate so long as a 
precipitate of lead sulphate is formed. The liquid is 
filtered, and the excess of lead removed by a stream of 
sulphuretted hydrogen; the solution is again filtered into a 
250 c.c. flask, and diluted to the containing-mark. 50 c.c. 



140 



Quantitative Chemical Analysis. 



of the liquid are then evaporated to dryness, with about 10 
c.c. of dilute hydrochloric acid (IT sp. gr.) in a weighed pla- 
tinum dish. The residual chlorides are dried and weighed ; 
the relative proportion of the potash and soda is then deter- 
mined by means of standard silver and potassium chromate, 
in the manner described on p. 128. 

VII. DETERMINATION OF AMMONIA. 

The quantity of free ammonia in solution may be deter- 
mined, as in the case of caustic soda or potash, by means of 
standard acid and litmus solutions. A definite quantity of 
the solution, say 10 c.c., is transferred to a small tared flask, 
and weighed : its absolute weight and specific gravity are 
thus determined in a single operation. If the 10 c.c. weighed 
9*0 grams, its specific gravity would of course be 0-9000, 
water being i. The weighed quantity of the ammonia is 
then diluted with 6 or 8 times its bulk of water and titrated 
directly in the ordinary manner by the standard acid. 

The operation of taking the specific gravity and weighing 
out the ammonia solution may be rendered more accurate, 

especially if the solution is 
very strong, by the aid of the 
little apparatus seen in Fig. 45. 
In the hole of the caoutchouc 
ball a is inserted a brass tube 
running through and fixed into 
the plate b. The end of the tube 
is closed, its upper edge is 
filed through, and over it is 
slipped a piece of tightly-fitting 
caoutchouc tube, in which, 
immediately over the orifice in 
the brass tube, is a hole pierced 
by a pin. Through the hole is 
inserted the end of the appara- 
tus cc t which is further supported by the holder. The 
caoutchouc ball maybe compressed by the plate d, moveable 



FIG. 45 . 




Ammonia. 141 

along the rods, by the aid of the milled-head screw s, On 
compressing the ball, the air makes its escape through the 
end e of the apparatus, and if, after the expulsion of the air, 
the end ^be placed beneath the surface of a liquid, the liquid 
will be driven into the apparatus in proportion as the screw 
is reversed. If the apparatus has the arrangement seen in the 
figure, it is evident that it may be withdrawn from the caout- 
chouc tube without any of the liquid flowing out. If the 
capacity of the bulb up to a certain mark, say at m, be accu- 
rately estimated, by determining the weight of water it 
contains up to that mark, we can readily determine the 
specific gravity of any liquid introduced into it by simply 
weighing the apparatus filled with the liquid. By reinserting 
the end into the hole in the caoutchouc tube and compressing 
the ball we can deliver any required quantity of the liquid. 
On again weighing the apparatus, its loss of weight im- 
mediately gives us the amount of liquid delivered. The 
apparatus may also be used for weighing out and determining 
the specific gravities of strong acids, fuming liquids, &c. 

Ammonia in combination may be determined by expelling 
it by means of caustic soda or lime, collecting the evolved 
ammonia in an excess of standard acid, and determining the 
excess of acid by soda- solution. The ammoniacal compound 
(say ammonium chloride) is weighed out into the retort a, 
Fig. 32, and the ammonia collected in excess of standard 
acid j the amount of the residual acid is then determined by 
soda-solution. 

The ammonia contained in many commercial salts, in 
ammoniacal gas-liquor, &c., may be determined by this 
method. In estimating the ammonia in guano, magnesia 
must be employed instead of lime or soda, otherwise the 
nitrogenous organic matter present will be partially decom- 
posed with evolution of ammonia. 



142 Quantitative Chemical Analysis. 

ACIDIMETRY. 

The principles involved in the estimation of the strength 
of acid solutions are identical with those we have indicated 
under Alkalimetry. The value of the strong acids, such as 
hydrochloric, nitric, and sulphuric acids, is frequently deduced 
from their specific gravities, and comprehensive tables have 
been calculated showing the percentage amount of the 
various acids in solutions of different densities (see Appen- 
dix). Occasionally it is necessary to control the indications 
of the hydrometer, or specific-gravity bottle, by titrating the 
acid solution. The apparatus described on p. 140 (Fig. 45) 
may be conveniently employed to determine both the 
specific gravity of the liquid and the weight taken for an- 
alysis. The determination of the strength of nitric, hydro- 
chloric, and sulphuric acids by caustic soda and litmus solu- 
tion presents no difficulties : the method will be evident 
from the foregoing descriptions under Alkalimetry. 

VIII. DETERMINATION OF THE STRENGTH OF ACETIC, 
ACID, PYROLIGNEOUS ACID, VINEGAR. 

The estimation of the strength of this acid in its various 
forms cannot be made with very great accuracy by direct 
titration with caustic soda solution, since sodium acetate 
possesses a feeble alkaline reaction, which interferes with the 
correct determination of the final point The method most 
generally applicable consists in adding to a known quantity 
of the acid, a weighed quantity (in excess) of finely-powdered 
marble, heating the liquid to boiling, filtering, washing the 
residual calcium carbonate with hot water, dissolving it in a 
slight excess of normal hydrochloric acid, and titrating with 
caustic soda and litmus solution. This method is par- 
ticularly useful in testing brown pyroligneous acid or highly 
coloured vinegars. 

[Note. Instead of litmus solution, a dilute tincture of 
cochineal may be employed. It may be prepared by digesting 
2 or 3 grams of powdered cochineal in 200 c.c. of a mixture 



Carbon Dioxide. 



143 



of i part of alcohol and 4 of water. It forms an orange 
solution which is turned violet by alkalies : the colour is 
almost unaffected by carbonic acid.] 

IX. DETERMINATION OF COMBINED CARBON DIOXIDE. 

The amount of carbon dioxide in soluble carbonates may 
be readily determined by decomposing them with a solution 
of calcium chloride, throwing the precipitated calcium car- 
bonate on to a filter, washing thoroughly with hot water, 
dissolving in an excess of standard hydrochloric acid, and 
determining the excess of acid by standard soda solution in 
the usual manner. 

The acid carbonates (bicarbonates) require the addition of 
ammonia, with the calcium chloride. 

The carbon dioxide in insoluble carbonates, as in cala- 
mine, ferrous carbonate, white lead, mortar, cements, &c., 
is determined by expelling the gas by the action of hydro- 
chloric acid, absorbing it by 
ammonia, and precipitating by 
the addition of calcium chlo- 
ride. The calcium carbonate is 
further treated in the manner 
above described. 

The decomposition may 
conveniently be effected in the 
apparatus seen in fig. 46. 
The flask A, of about 150 c.c. 
capacity, contains the weighed 
quantity of carbonate, together 
with about 10 c.c. of water: it 
is fitted with a caoutchouc 
cork, in which are inserted the 
bent tube a and the pipette- 
shaped tube , filled with 
moderately-concentrated hy- 
drochloric acid. The flow of 



FIG. 46. 




144 Quantitative Chemical Analysis. 

the acid into the flask may be easily regulated by the clip. 
The flask B contains TO or 15 c. c. of ammonia-water: the 
tube must not dip into the liquid. The wider tube c is 
partially filled with broken glass moistened with ammonia- 
water. Care should be taken that the ammonia is free 
from carbonic acid: its purity may be tested by adding to it 
a few drops of calcium chloride solution : if free from car- 
bonate it will remain perfectly clear. 

To make a determination, warm the flask B, so as to fill it 
with an atmosphere of ammonia, and then cautiously allow 
a few drops of hydrochloric acid to fall on to the weighed 
quantity of carbonate in A. As soon as the whole of the 
carbonate is decomposed, heat the liquid in A to boiling, so 
as to expel the last trace of carbon dioxide, and keep it 
boiling for a few minutes. Wash the bent tube a and also 
the glass in t, add calcium chloride solution to the ammo- 
niacal liquid, boil for some time, filter, wash thoroughly, and 
titrate the calcium carbonate in the manner already described. 
The operation of filtering and washing should be done as 
expeditiously as possible, since the ammoniacal liquid 
absorbs atmospheric carbon dioxide. 

X. ESTIMATION OF CARBONIC ACID IN NATURAL WATERS. 

The amount of carbonic acid in spring, river, or mineral 
water may be accurately estimated in the following manner. 
100 c.c. of the water to be examined are transferred to a dry 
flask, together with 3 c.c. of a strong neutral solution of 
calcium chloride, and 2 c.c. of a saturated solution of 
ammonium chloride. 50 c.c. of lime-water, the strength of 
which is accurately known, are then added, the flask is 
closed by a caoutchouc cork, and its contents, amounting to 
155 c.c., agitated. In about twelve hours the whole of the 
calcium carbonate will have separated out, and the liquid 
will be perfectly clear. 50 c.c. of the clear liquid are with- 
drawn, and the residual amount of lime determined by 
deci-normal hydrochloric acid. 



Carbon Dioxide. 145 

A solution of oxalic acid may be conveniently substituted 
for the hydrochloric acid ; it should be made of such a 
strength that i c.c. is equivalent to i milligram of carbon 
dioxide. This solution may be obtained by dissolving 
2-8636 grams of pure dry crystallised oxalic acid in water, 
and diluting to i litre. The lime-water should be of such 
strength that 25 c.c. are equal to 23 or 24 c.c. of acid. The 
final point of the reaction may be determined by the aid of 
a drop of tincture of pure rosolic acid, which gives a 
splendid red colour in presence of the alkaline earth, which 
disappears on neutralising with an acid. Instead of rosolic 
acid, turmeric paper may be used. Swedish paper, in strips, 
is immersed in tincture of turmeric, and dried. A drop of 
the liquid brought upon the paper gives a reddish-brown 
stain, so long as the least trace of the alkaline earth remains 
in the free state. 

Example. 100 c.c. of the water of the Irish Channel 
were mixed with 3 c.c. of calcium chloride, and 2 c.c. of 
ammonium chloride, together with 50 c.c. of lime-water. 
50 c.c. of lime-water =46 '4 c.c. of oxalic acid ; 

of which i c.c.=i milligram CO 2 . 
After standing fifteen hours 

50 c.c. of solution required 13*3 c.c. of standard acid for 
neutralisation. 

Therefore 

4 6-4-(i3'3 x 3-1) = 5-2. 
100 c.c. of the sea-water contain 5-2 milligrams carbon dioxide. 

XL ESTIMATION OF CARBON DIOXIDE IN ARTIFICIALLY 
AERATED WATERS. 

The determination of the total quantity of carbon dioxide 
in artificial mineral waters, seltzer, soda-water, &c. may 
be readily effected in the following manner. A narrow brass 
cork-borer is pierced with two or three small holes, about 
4 or 5 centimetres from the edge. A piece of caoutchouc 
tubing is slipped over the upper end of the borer, and the 

L 



146 



Quantitative Chemical Analysis. 



FIG. 47 



other end is connected with the bent tube a of the flask B 3 
fig. 47. The flask contains about 25 or 30 c.c. of mode- 
rately strong ammonia ; the end of the glass tube should dip 
beneath the surface of the liquid. The sides and edge of 
the brass tube are rubbed with a little paraffin, and it is then 
screwed through the cork of the bottle of the aerated water 

by holding the tube stationary, 
and turning the bottle round, 
until the holes make their ap- 
pearance below the cork. The 
gas is immediately liberated j it 
makes its escape through the 
holes in the borer, and is absorbed 
by the ammonia-water. The 
greater portion of the residual 
gas is expelled by shaking the 
water, and gently heating it by 
surrounding it with warm water. 
As soon as no more gas is 

evolved, the cork of the soda-water bottle is withdrawn, 
and the liquid added to that contained in the flask, the 
tubes are washed, calcium chloride is added, the liquid 
is boiled for some time, and the amount of calcium carbo- 
nate determined in the ordinary manner. The quantity of 
combined carbon dioxide in the water may be estimated by 
evaporating a second portion to dryness, gently igniting the 
residue, and titrating with standard acid and soda. 

XII. DETERMINATION OF COMBINED ACIDS IN SALTS. 

If we add caustic soda to a boiling solution of copper 
sulphate, cupric oxide is precipitated, and sodium sulphate 
remains in solution : 




CuS0 



H 2 O. 



4 2 NaHO = CuO + Na 2 SO 4 

.If, therefore, we mix a measured quantity (in excess) of 
standard soda solution with a solution of copper sulphate, 



Oxidation and Reduction. 147 

and boil, the amount of residual alkali indicates the quantity 
of the acid contained in the salt. 

Weigh out about 2 grams of copper sulphate into a 
250 c.c. flask, dissolve in 100 c.c. of water, boil, and add 
excess of standard soda solution, allow to cool, and dilute 
with water to the containing-mark, cork the flask, and allow 
the precipitate to settle. Take out an aliquot portion of the 
clear supernatant liquid, and determine the amount of resi- 
dual alkali in solution. The method of calculating the result 
needs no explanation. 

Many other substances, precipitable by sodium hydroxide 
or sodium carbonate, admit of estimation by this method. 
Thus we may determine the amount of acid present in 

Silver salts, with caustic soda. 
Mercury salts, with caustic soda. 
Bismuth salts, with sodium carbonate. 
Lead, nickel, cobalt, zinc, aluminium, manganese, alkaline 
earths, and magnesia, with sodium carbonate. 

In certain cases the acid, as in copper sulphate, or mer- 
curic chloride, may be liberated by treating a boiling solution 
of the salt with sulphuretted hydrogen, filtering, and deter- 
mining the amount of the free acid by caustic soda and 
litmus solutions. 



ANALYSIS BY OXIDATION AND REDUCTION. 

We have already explained the main principle of this 
special form of volumetric analysis ; we have shown, for 
example, how the amount of iron in solution may be esti- 
mated by determining the quantity of oxygen required to 
convert it from the state of ferrous to that of ferric oxide. 
A large number of other substances may be estimated by 
the aid of a solution of a substance which, like potassium per- 
manganate, readily parts with a portion of its oxygen. In 

L 2 



148 Quantitative Chemical Analysis. 

all these cases the amount of oxygen given up is taken as an 
index of the quantity of the substance to be determined, 

Of the many oxidising agents which are known, the most 
generally applicable are (i) potassium permanganate (per- 
manganic acid), and (2) iodine. 



A, ESTIMATIONS BY MEANS OF POTASSIUM 
PERMANGANATE. 

Preparation of Potassium Permanganate Solution. About 
5 grams of pure crystallised potassium permanganate are dis- 
solved in a small quantity of water, and the solution is 
diluted to i litre. It must be contained in a glass-stoppered 
bottle, which should be kept in a cool, dark place, when not 
in use. The solution thus preserved may be kept for a long 
time without experiencing much alteration. 

XIII. DETERMINATION OF THE STRENGTH OF THE 
PERMANGANATE SOLUTION. 

It is absolutely necessary to determine the power of the 
permanganate solution by direct experiment before using it; 
we cannot calculate its strength, i.e. the amount of oxygen 
that it is capable of furnishing, from the weight of the salt 
dissolved, on account of its instability. The most accurate 
method of estimating the strength of the solution consists in 
determining the amount required to transform a known 
weight of iron from the condition of ferrous oxide to that of 
ferric oxide. 

About i gram of fine iron-wire (piano wire), perfectly free 
from rust, is accurately weighed out into a ^-litre flask, and 
dissolved in about 100 c.c. of dilute sulphuric acid (i pt. of 
acid to 6 of water). A small quantity of sodium carbonate 
is thrown into the liquid at the same time in order that the 
air within the flask may be displaced by carbon dioxide. 
The flask is fitted with a cork and bent tube furnished with 



Use of Potassium Permanganate. 149 

a clip ; the end of the tube dips beneath the surface of about 
25 c.c. of water contained in a small flask (fig. 48). 
Whilst the iron is dissolving, the clip is kept open by 
slipping it over the glass tube. The solution of the iron 
may be accelerated by a gentle heat ; the liquid is gradually 
caused to boil, and maintained in brisk ebullition for a 
minute or two, so as to expel the mixture of carbon dioxide 
and hydrogen, and the caoutchouc tube is immediately 
closed and the lamp removed. In a minute or so the clip 
is again opened, when the water from the little flask is driven 
over into the solution of iron : in proportion as it passes 
over, boiling water is poured into the small flask until the 

FIG. 48 




larger one is nearly filled. The caoutchouc tube is once 
more closed, the flask and its contents allowed to cool 
perfectly, and the volume of the liquid made up to the con- 
taining-mark, the stopper of the flask is inserted, and the 
liquid thoroughly mixed by shaking. Whilst the solution is 
cooling, fill a Gay-Lussac's or a Mohr's burette fitted 



15 Quantitative CJiemical Analysis. 

with a glass stop-cock (the permanganate solution gradually 
attacks caoutchouc), previously rinsed out with a little 
of the permanganate solution, read off the level of the per- 
manganate solution, take out 50 c.c. of the iron solution, 
and pour it into about 200 c.c. of water contained in a 
beaker, standing on a sheet of white paper. Add the per- 
manganate drop by drop to the liquid, with constant stirring, 
until the pink colour of the solution is permanent. The 
permanganate is at first decomposed with great rapidity : as 
the iron becomes oxidised the colour disappears more slowly; 
the rapidity of the change indicates the progress of the 
oxidation. The operation of standardising should be - 
repeated once or twice on successive portions of 50 c.c. of 
iron solution, and the mean of the observations taken as 
representing the true value of the permanganate solution. 

ioFeSO 4 + 2 KMnO 4 + 8SO 4 H 2 
= 5Fe 2 (SO 4 ) 3 + 2MnSO 4 + K 2 SO 4 + 8H 2 O. 

Let us suppose that we have weighed out n gram of 
wire and dissolved it in 250 c.c. of liquid : 50 c.c. of the 
solution would be equivalent to 0-2200 gram of iron. The 
iron we have taken, however, is not chemically pure : we 
may assume without sensible error that its impurities amount 
to 0-4 per cent. ; accordingly the amount of iron in the 50 
c.c. is 

i : 0-996 :: 0-2200 : x. $ = 0-2192. 

Let us further suppose that we have required 20 c.c. of 
permanganate solution, as the mean of the experiments, be- 
fore we obtained a permanent coloration with the 50 c.c. of 
iron solution : then 20 c.c. permanganate oxidise 0*2192 gram 
iron from protoxide to peroxide, or 100 c.c. permanganate 
are equivalent to i "096 gram iron. 

The strength of the permanganate solution may also 
be determined by means of pure ferrous sulphate, precipitated 
from its aqueous solution by means of alcohol. The ferrous 
sulphate so prepared keeps unchanged for years. Or, in- 



Use of Potassium Permanganate. 151 

stead of this salt, the double sulphate of iron and ammonium 
FeSO 4 (NH 4 ) 2 SO 4 + 6H 2 O may be employed. It contains 
exactly one-seventh of its weight of iron : 07 gram of salt is 
equivalent to 0*1 gram of iron. 

The strength of the solution may also be estimated by 
means of oxalic acid. If potassium permanganate solution 
is dropped into a solution of oxalic acid acidulated with sul- 
phuric acid, the oxalic acid is completely decomposed into 
carbon dioxide and water : 

5C 2 H 2 4 + 2KMn0 4 + 3H 2 SO 4 
K 2 SO 4 + 2MnSO 4 + ioCO 2 + 8H 2 O. 

The oxalic acid solution requires to be gently heated (to 
about 60) before the reaction commences : at this tempera- 
ture the permanganate is rapidly decomposed so long as any 
oxalic acid remains in the solution. From the foregoing 
equations it is evident that 112 parts of iron (in the state 
of protoxide) and 126 parts of crystallised oxalic acid 
(C 2 H 2 O 4 + 2H 2 O) require exactly the same amount of 
oxygen, viz. 16 parts, for complete oxidation. The amount 
of available oxygen in i c.c. of the permanganate solution 
may therefore be readily calculated. 

Of the several substances which may be used for titrating 
the permanganate solution, metallic iron is on the whole to be 
preferred. All the methods are equally accurate with careful 
manipulation, but fewer sources of error attend the use of 
metallic iron. Crystallised oxalic acid is not readily obtained 
quite pure, it is liable to part with a portion of its water of 
crystallisation, and its solution, especially under the influence 
of light, is apt to decompose. Ammonium oxalate is prefer- 
able to oxalic acid : it may be readily purified by recrystal- 
lisation, and keeps perfectly well : its composition is C 2 O 4 
(NH 4 ) 2 + H 2 O. 142-08 parts of the crystallised salt are 
equivalent to 1 1 2 parts of iron. 

Whichever method be adopted it is absolutely necessary 
that the solution to be titrated should contain free sulphuric 



152 Quantitative Chemical A nalysis. 

acid. If there is a deficiency of free acid, the solution 
becomes brown, and eventually a precipitate is formed. It 
is not a matter of indifference which acid is employed for 
acidulation : nitric acid cannot well be used under any 
circumstances, and hydrochloric acid is liable to be decom- 
posed, and chlorine eliminated, in accordance with the 
equation : 

I4HC1 4- Mn 2 O 7 = 2MnCl 2 + 5C1 2 + 7H 2 O. 

Whenever, therefore, the use of sulphuric acid is inadmis- 
sible, it is better to employ potassium bichromate as an 
oxidising agent (see Analysis of Iron Ores). 

XIV. VOLUMETRIC ESTIMATION OF CALCIUM BY MEANS 
OF POTASSIUM PERMANGANATE. 

Oxalic acid or ammonium oxalate solution added to cal- 
cium chloride or any soluble salt of lime, in presence ol 
ammonium chloride and free ammonia, gives rise to a 
precipitate of calcium oxalate. Calcium oxalate digested 
with dilute sulphuric acid is decomposed, calcium sulphate 
is formed, and oxalic acid passes into solution. We already 
know that it is possible to determine the strength of a per- 
manganate solution by means of an oxalic acid solution of 
known strength : conversely we can determine an unknown 
quantity of oxalic acid by means of a solution of perman- 
ganate, the strength of which is accurately known to us. 
Upon these considerations is based a method of determining 
lime volumetrically. 

Weigh out about 2 grams of marble into a 250 c.c. flask, 
dissolve it in dilute hydrochloric acid, heat to boiling, add 
dilute ammonia in slight excess, and solution of ammonium 
oxalate so long as a precipitate is formed. Allow the pre- 
cipitate to settle, pour the supernatant fluid through a small 
filter, and thoroughly wash the calcium oxalate by decanta- 
tion with hot water. Pour the turbid liquid through the 
filter, but do not bring more of the precipitate on the filter 



Use of Potassium Permanganate. 153 

than you can help. Wash the filter thoroughly, place the 
funnel in the neck of the J-litre flask, and pour into it a 
quantity of dilute sulphuric acid, previously heated. Again 
wash the filter until the filtrate which should of course drop 
into the flask is no longer acid. Add a further quantity of 
dilute sulphuric acid to the flask, dilute with water, and heat 
gently. Allow the liquid to cool, fill up the flask to the con- 
taining-mark, agitate, and quickly transfer 100 c.c. of the 
turbid liquid to a beaker, heat to about 60, and add the 
permanganate solution until the pink colour is permanent. 
Again shake the liquid in the flask, and repeat the deter- 
mination on a second quantity of 100 c.c. The results 
should agree, i eq. of oxalic acid is equal to i eq. of calcium. 
The details of the calculation need no explanation. 

This method is especially convenient when a number of 
estimations of lime have to be made in succession. It may 
be accelerated by adding an excess of ammonium oxalate 
of known strength to the solution of the lime salt, 
diluting to a definite bulk, allowing the precipitate to 
settle, withdrawing an aliquot portion of the clear liquid, 
acidulating with sulphuric acid, heating to 60, and adding 
the permanganate until the coloration shows that the 
reaction is finished. We of course know the amount of 
ammonium oxalate we have used originally; the titration 
tells us the excess remaining in solution ; the difference 
expresses that combined in the insoluble lime salt. Since 
i eq. oxalic acid is equal to i eq. of calcium, we can readily 
calculate the quantity of the alkaline earth from the amount 
of permanganate solution used. 

XV. VOLUMETRIC ESTIMATION OF LEAD BY PERMAN- 
GANATE SOLUTION. 

In certain cases lead may be estimated by the foregoing 
methods, but on account of the slight solubility of the lead 
oxalate in water containing ammonia, the results are not 
quite so accurate as in the case of lime. 



1 54 Quantitative Chemical A nalysis. 

XVI. VALUATION OF MANGANESE ORES BY MEANS OF 

POTASSIUM PERMANGANATE SOLUTION. 

(See Part IV.) 

XVII. ESTIMATION OF POTASSIUM FERROCYANIDE BY 

PERMANGANATE SOLUTION. 

Potassium permanganate solution added to a solution of 
potassium ferrocyanide, acidulated with sulphuric acid, 
converts that salt into potassium ferricyanide. If the solu- 
tion is sufficiently dilute there is no difficulty in perceiving 
the termination of the reaction. It is advisable to titrate the 
permanganate solution (which should contain about i gram 
KMnO 4 per 1000 c.c.) with a solution of potassium ferro- 
cyanide of known strength; 5 grams of the recrystallised 
salt (K 4 FeCy6 + 3H 2 O) dissolved in 500 c.c. of water forms 
a convenient solution. A measured quantity of the solution, 
say 25 c.c., is placed in a porcelain basin and diluted with 
about ten times its bulk of water, together with a quantity 
of pure sulphuric, acid, and the potassium permanganate is 
added from a burette, with constant stirring, until the pure 
yellow colour of the solution changes to a reddish-yellow 
tint. To determine the amount of pure ferrocyanide in a 
sample of the article, weigh off about 3 grams, dissolve in 
water, and make up the solution to 250 c.c. 25 c.c. of the 
liquid are transferred to a porcelain basin, diluted with 
water, and treated in the manner directed. 

XVIII. ESTIMATION OF POTASSIUM FERRICYANIDE BY 

PERMANGANATE SOLUTION. 

This substance may be analysed by reducing it to the 
state of ferrocyanide, and determining the amount of the 
reduced salt in the manner already described. The weighed 
quantity of the ferricyanide is rendered strongly alkaline with 
caustic potash solution, heated to boiling, and mixed with a 
concentrated solution of ferrous sulphate, added little by 



Use of Iodine, &c. 155 

little until the colour of the precipitate is black, owing to 
the formation of triferric tetroxide. The alkaline solution 
is diluted, filtered, and the filtrate made up to 300 c.c. ; 50 
or 100 c.c. are then transferred to a porcelain basin, 
strongly acidified with sulphuric acid, and titrated with 
potassium permanganate solution. 



B. ANALYSES BY MEANS OF IODINE AND SODIUM 
THIOSULPHATE (HYPOSULPHITE) SOLUTIONS. 

When iodine is brought into contact with sodium thio- 
sulphate (hyposulphite) the following reaction occurs 

2Na 2 S 2 O 3 + I 2 = Na 2 S 4 O 6 + 2NaI 
iodine and sodium thiosulphate forming sodium tetrathionate 
and sodium iodide. If now we mix with the thiosulphate 
solution a small quantity of starch, and add the iodine so- 
lution drop by drop, the sensitive blue colour of the iodide of 
starch will continue to be destroyed as fast as it is produced, 
so long as the foregoing reaction occurs. Immediately that 
the whole of the thiosulphate has been converted into 
tetrathionate, the least excess of iodine will act upon the 
starch, and the blue colour will be permanent. If, therefore, 
we have a solution of thiosulphate of known strength we can 
readily estimate the amount of iodine in a solution containing 
an unknown quantity of that element 

A solution of iodine, in presence of substances which 
readily take up oxygen, decomposes water, forming hydriodic 
acid with its hydrogen and giving up the oxygen to the 
oxidisable substance. Thus in the case of arsenious acid 

As 2 O 3 + 2l 2 + 2H 2 O = 4HI + As 2 O 5 . 
In the case of sulphur dioxide 

SO 2 + I 2 + 2H 2 O = SO 4 H 2 + 2HI. 

These reactions afford the basis of an exact and generally 
applicable volumetric process. 



1 56 Quantitative Chemical A nalysis. 

The method requires 

(1) A solution of iodine of known strength. 

(2) A solution of sodium thiosulphate of known strength. 

(3) A solution of starch. 

i. Preparation of the Iodine Solution. This may most 
conveniently be of deci-normal strength : i c.c. should con- 
tain therefore 0-012685 gram iodine. About 13 grams of 
pure iodine are weighed out into a litre flask, together with 
about 1 8 grams of pure potassium iodide, the whole is dis- 
solved in about 300 c.c. of perfectly cold water, and diluted 
so as to be of deci-normal strength. Thus supposing that we 
had weighed out exactly 13 grams of iodine, we should 
require to dilute the liquid to 1024 c.c., since 

i2'685 : 1000 :: 13 \ x x = 1024. 

The pure iodine may be obtained by intimately mixing re- 
sublimed iodine with about one-fourth of its weight of 
powdered potassium iodide, heating the mixture in a large 
porcelain crucible placed on an iron plate, and surmounted 
by a precisely similar crucible. The resublimed crystals are 
loosened from the sides of the crucible and placed over 
strong oil of vitriol within a bell-jar for a few hours, in order 
to deprive them of the last traces of hygroscopic moisture. 
The dried iodine is then quickly transferred to a clean dry 
test-tube, into which a second test-tube is fitted, in the 
manner seen in fig. 49. The stoppered 

FIG. 49. FIG. 50. , , . - iv 

tube represented in fig. 50 may also be 
employed. The tubes containing the 
iodine are accurately weighed, a portion of 
the substance is transferred to the litre- 
flask containing the potassium iodide, dis- 
solved in water, and the tubes are again 
weighed. The loss of weight gives the 
amount of iodine taken : care should be 
taken that the amount is not less than 
12-685 grams. Any powder adhering to the sides of the 



Use of Iodine, &c. 157 

flask is washed down into the liquid ; when the whole of the 
iodine is dissolved the solution is diluted to the proper 
degree. The liquid must not be heated ; if the weighing 
out and the dilution have been carefully conducted the 
solution will be strictly deci-normal. The solution is most 
conveniently preserved in small stoppered bottles of about 
200 c.c. capacity, which should be filled to the neck and 
kept in a cool, dark place. 

2. Preparation of Deci-normal Solution of Sodium Thiosul- 
phate. i litre contains 24*8 grams of the salt. About 25 
grams of the recrystallised salt, previously powdered and 
dried between filter-paper, are accurately weighed out into 
the litre flask, dissolved in water, and the solution diluted so 
that i c.c. = 0-0248 gram of thiosulphate. The solution 
should also be kept in the dark : when exposed to light, it 
slowly decomposes, with the precipitation of sulphur. 
Accordingly a fresh solution of the salt should be prepared 
from time to time. 

3. Preparation of Starch Solution. i gram of pure wheaten 
starch is mixed with a small quantity of water and rubbed 
to a thin cream in a mortar. The paste is poured into 150 
c.c. of boiling water, the liquid is allowed to stand, and the 
clear solution decanted from the sediment. The solution 
should be prepared before the beginning of each series of 
experiments, since it decomposes after a time. By adding 
about 10 c.c. of glycerine to it, or saturating it with common 
salt, the solution keeps better : it is so readily prepared, 
however, that it is preferable to make a fresh solution when 
wanted. 

If any doubt should exist as to the exact strength of the 
thiosulphate solution, it may be readily standardised by the 
aid of a deci-normal solution of potassium bichromate. 
Potassium bichromate solution added to potassium iodide, 
in presence of free hydrochloric acid liberates iodine : 
K 2 Cr 2 O 7 + 6KI + I4HC1 = 3X2 + 8KC1 + Cr 2 Cl 6 +7H 2 O. 



158 Quantitative Chemical Analysis. 

294-3 parts of potassium bichromate liberate 761-1 parts of 
iodine : accordingly i c.c. of deci-normal solution of 
bichromate liberates 0-012685 gram of iodine. 

25 c.c. of deci-normal bichromate solution made by dis- 
solving 4-907 grams of potassium bichromate in a litre of 
water are placed in a flask and mixed with about 10 c.c. of 
solution of pure potassium iodide (i of salt to 10 of water), 
together with about 5 c.c. of pure hydrochloric acid, and 200 
c.c. of water. The standard thiosulphate solution is then 
added drop by drop until the iodine has nearly disappeared : 
a few drops of starch solution are added, and the addition of 
the thiosulphate continued until the last trace of the blue 
colour of the iodide of starch vanishes. Since the 25 c.c. of 
bichromate solution liberate 25 x 0-012685 = 0-3171 gram 
iodine, we can readily calculate the amount of iodine 
equivalent to i c.c. of the sodium thiosulphate solution. 

XIX. VALUATION OF BLEACHING-POWDER BY IODINE 
AND SODIUM THIOSULPHATE SOLUTIONS. 

(See Part IV.) 

XX. ESTIMATION OF THE AMOUNT OF CHLORINE IN 

AQUEOUS SOLUTIONS OF THE GAS. 

Prepare a dilute solution of chlorine in water, measure off 
a definite quantity of the liquid, and transfer it to a solu- 
tion of potassium iodide. Determine the amount of the 
liberated iodine by solution of sodium thiosulphate, adding 
the latter liquid until the iodine has nearly disappeared ; add 
2 or 3 c.c. of starch solution, and continue the addition of 
the thiosulphate until the blue colour just vanishes. 

Add a second portion of the chlorine water to the solu- 
tion of potassium iodide, and a measured quantity (in 
excess) of sodium thiosulphate, and determine the amount 
of the residual thiosulphate by means of starch and stan- 
dard iodine solution, adding the latter until the blue colour 
of the solution is persistent. The results of the two expe- 
riments should agree. 



Use of Iodine, &c. 159 

XXI. ESTIMATION OF THE AMOUNT OF SULPHUR DIOXIDE 
IN AQUEOUS SOLUTIONS OF THE GAS. 

Prepare a very dilute solution of sulphur dioxide (by 
adding 10 c.c. of a saturated solution of the gas to a litre of 
water), transfer a definite quantity of the liquid to a known 
amount (in excess) of standard iodine solution, and deter- 
mine the amount of residual iodine by starch and sodium 
thiosulphate solution. 

To a second measured portion of the solution of the gas, 
add 2 or 3 c.c. of starch solution, and the standard solution 
of iodine, until the blue colour of the iodide of starch is 
persistent. The reaction between sulphur dioxide and 
iodine may be thus represented : 

SQ 2 + I 2 + 2H 2 = S0 4 H 2 + 2 HI. 

If, however, the solution is concentrated, the sulphuric 
acid reacts upon the hydriodic acid, and sulphur dioxide and 
free iodine are again formed 

SO 4 H 2 + 2HI = SO 2 + I 2 + 2H 2 O. 

So long as the solution contains only about 0*05 per cent, 
of sulphur dioxide, the first reaction alone takes place. 

XXII. ESTIMATION OF SULPHURETTED HYDROGEN IN 
AQUEOUS SOLUTIONS OF THE GAS. 

When sulphuretted hydrogen is brought into contact with 
free iodine, the following reaction ensues : 

H 2 S + I 2 = 2HI + S. 

This reaction, however, is liable to be modified by the 
concentration of the solution ; experiment has shown that 
it can only be depended upon when the solution contains 
not more than 0^04 per cent, of the gas. 

Prepare a dilute solution of sulphuretted hydrogen, 
measure off a definite quantity, add starch liquor and solu- 
tion of iodine until the colour is persistent. 



160 Qitantitative Chemical Analysis. 

Place in a flask about the same quantity of iodine solu- 
tion which you have consumed in the foregoing experiment, 
and then add a measured quantity (in excess) of the solution 
of sulphuretted hydrogen. Determine the excess of sul- 
phuretted hydrogen by starch and standard iodine solutions. 

The amount of sulphuretted hydrogen in mineral waters 
may be readily determined by this method. For such esti- 
mations it will be more convenient to dilute the iodine solu- 
tion to ten times its bulk ; 100 c.c. are transferred to the 
litre flask, and the flask is filled up to the containing-mark. 

Measure off 250 c.c. of the mineral water, transfer to a 
beaker, add i or 2 c.c. of starch liquor and centi-normal 
solution of iodine, until the blue colour is persistent. Let 
us suppose we have used 20 c.c. of centi-normal solution of 
iodine ; this amount of the iodine solution is placed in 
a beaker, and mixed with 250 c.c. of the mineral water ; 
i or 2 c.c. of starch liquor are added, and then, cautiously, 
centi-normal solution of iodine, drop by drop, until the blue 
colour of the iodide of starch remains. It will be found in 
general that i or 2 c.c. more of iodine solution are required 
in the second experiment than in the first. Now pour 
250 c.c. of distilled water into the beaker, add the same 
bulk of starch liquor used in the last experiment, and, drop 
by drop, the solution of iodine, until the blue colour is pro- 
perly defined. This experiment gives the amount of the 
iodine solution required to produce the final reaction ; subtract 
this quantity from the amount of iodine solution required in 
the second experiment. The remainder shows the amount 
of iodine solution equivalent to the sulphuretted hydrogen 
present ; this amount may readily, be calculated from the 
above equation. 

XXIII. ESTIMATION OF HYDROCYANIC ACID. 

When potassium cyanide is mixed with solution of iodine, 
iodide of potassium and iodide of cyanogen are formed : 

KCy + I 2 = KI + Cyl. 



Use of Iodine, &c. 1 6 1 

Two eq. or 2537 parts of iodine correspond to i eq. or 
65*17 of potassium cyanide. This principle affords the 
basis of an exact method for determining the value of 
potassium cyanide, or solutions of prussic acid. 

In the case of potassium cyanide, weigh out about 
2 grams of the salt into a J-litre flask, dissolve in water, 
dilute to the mark, shake, and transfer 50 c.c. to a beaker 
containing about 200 c.c. of water, add 100 c.c. of a satu- 
rated solution of carbonic acid gas in water (to convert 
the alkaline carbonates, always present in the commercial 
article, to acid carbonates), and add solution of iodine, until 
the liquid possesses a slight and permanent yellow tinge. 
The cyanide must be free from alkaline sulphide. 

In the case of free hydrocyanic acid, add a very slight 
excess of caustic soda, and then solution of carbonic acid, 
and proceed in the manner above described. The specific 
gravity of the ordinary preparations of hydrocyanic acid is so 
little removed from that of water, that it is more prudent to 
determine the amount taken for analysis by weighing, rather 
than by measuring the solution in a pipette, by aspiration in 
the ordinary manner. 

XXIV. ESTIMATION OF ANTIMONY IN TARTAR EMETIC, 
AND OF ARSENIC AND ARSENIOUS ACIDS IN COM- 
MERCIAL ARSENIATES. 

These methods depend upon the conversion of antimony 
or arsenic trioxide, in an alkaline solution, into pentoxide, by 
solution of iodine : 

Sb 2 O 3 + 2l 2 + 2H 2 O == Sb 2 O 5 + 4HI 
As 2 O 3 + 2l 2 + 2H 2 O = As 2 O 5 + 4HI 

Weigh out about 2 grams of the tartar emetic, dissolve in 
water, and dilute to 250 c.c. Transfer 20 c.c. of the solu- 
tion to a beaker, add the same amount of a saturated 
solution of sodium bicarbonate, and 2 c.c. of starch liquor. 
Now add the iodine solution until the blue colour is 

M 



1 62 Quantitative Chemical Analysis. 

persistent for about 5 minutes. The fluid acquires a reddish 
tint just before the reaction is completed : the blue coloura- 
tion, even after the reaction is finished, fades after a time, say 
in 15 minutes. 

In the case of commercial arseniates weigh out about 3 
grams of the substance into a -J-litre flask, and dissolve in 
about 150 c.c. of warm water, add a small quantity of 
sodium acetate and acetic acid, and boil the solution for 15 
minutes to decompose any nitrites : when cold make up to 
500 c.c. 

To determine the amount of the arsenite withdraw 50 c.c. 
of the solution, add 25 cb.c. of a saturated solution of pure 
NaHCO 3 , starch paste, and decinormal iodine solution 
(i c.c. = "00495 g r & m As 2 O 3 ). 

Now pass sulphur dioxide gas into the solution in the 
flask to reduce the pentoxide 

As 2 O 5 + 2SO 2 + 2H 2 O = As 2 O 3 + 2H 2 SO 4 , 

boil to expel the excess of sulphur dioxide, and when cold 
make up to 500 c.c. Withdraw 50 c.c. of the solution, add 
25 c.c. of the sodium bicarbonate solution, starch paste and 
iodine solution as before (i c.c. =0-00575 gram As 2 O 5 ). 

XXV. DETERMINATION OF TIN BY IODINE SOLUTION. 

The metal is dissolved in hydrochloric acid, and mixed 
with a solution of Rochelle salt, and a concentrated solution 
of sodium bicarbonate is then added until the liquid is no 
longer acid. Add about i c.c. of starch solution and deci- 
Rorinal solution of iodine, until the blue colour is persistent. 
2537 parts of iodine are equivalent to 118 of tin. The 
solution of the metal is best effected in a stream of carbon 
dioxide : the addition of a few scraps of platinum-foil ac- 
celerates the process. 

This method is of course applicable to the valuation of 
c Tin crystals.' 



Use of Iodine, &c. 1 63 

ANALYSES BY MEANS OF IODINE AND SODIUM 
THIOSULPHATE SOLUTIONS, WITH PREVIOUS 
DISTILLATION WITH HYDROCHLORIC ACID. 

A number of substances containing oxygen, when heated 
with hydrochloric acid, are decomposed in such manner that 
free chlorine is evolved in amount bearing some simple ratio 
to the quantity of oxygen present. Thus when manganese 
dioxide is heated with hydrochloric acid, the following 
reaction occurs : 

Mn0 2 + 4 HC1 = MnCl 2 + C1 2 + 2H 2 O 
70-92 parts of chlorine are therefore equivalent to 86-04 of 
manganese dioxide. If the chlorine be led into a solution 
of potassium iodide, iodine will be liberated in exact propor- 
tion to the chlorine evolved : the amount of iodine liberated 
is a measure therefore of the amount of real manganese 
dioxide in the sample analysed. 

Similarly when potassium bichromate is heated with 
hydrochloric acid, chlorine is evolved : 

K 2 Cr 2 7 + i 4 HCl = Cr 2 Cl 6 + ;H 2 O + 2KC1 + 3C1 2 . 
212-76 parts of chlorine are equivalent therefore to 294-3 
parts of potassium bichromate. As we have seen from the 
method of titration given on p. 158, No. XX. the disengaged 
chlorine,. in presence of excess of potassium iodide, liberates 
an equivalent quantity of iodine, which can be estimated by 
means of starch and sodium thiosulphate solutions. 

XXVI. ANALYSIS OF POTASSIUM BICHROMATE. 

Into the little bulb a, which has a capacity of about 60 c.c. 
(fig. 50 A), accurately weigh out about 0*3 or 0-4 gram of pure 
fused potassium bichromate, add about 25 c.c. of pure fuming 
hydrochloric acid, and connect the bulb with the bent tube 
b by means of a tightly-fitting caoutchouc tube, which has 
been previously boiled in caustic soda solution to remove 
any adhering sulphur. The bent tube is connected with a two- 
bulb U-tube by means of a caoutchouc cork, which should 

M 2 



164 



Quantitative Chemical A nalysis. 



also have been previously cleansed by caustic soda solution. 
In 'certain cases it is advisable to have a second U-tube, 
which is connected with the first by corks and a bent tube. 
Both the tubes are placed in a beaker and are surrounded 
by cold water: they each contain about 25 c.c. of strong 
solution of potassium iodide. Gently heat the bulb con- 
taining the bichromate and acid ; chlorine is readily evolved 

FIG. SOA. 




and decomposes the potassium iodide in the U-tubes: 
after two or three minutes' heating, the whole of the chlorine 
will have been eliminated. Heat the liquid to boiling, so as 
to drive over the last traces of the gas ; remove the lamp, 
and allow the whole to stand for 5 or 10 minutes to effect 
the complete absorption of the chlorine ; empty the contents 
of the U-tubes, when quite cold, into a beaker (the solution 
in the second tube need not be added unless it contains 
liberated iodine), dilute with water, and titrate with . starch 
and thiosulphate solution. 



Use of Iodine, &c. 165 

XXVII. ESTIMATION OF ARSENIOUS ACID BY THE AID 

OF THE FOREGOING REACTION. 

If we mix the weighed quantity of bichromate with finely- 
powdered arsenious acid (not in excess) only a portion of 
the chlorine demanded by the equation is evolved; the 
deficit has served to bring about the oxidation of the arsenic 
trioxide to pentoxide : 

As 2 O 3 + 4C1 + 2H 2 O = As 2 O 5 -f 4HC1. 
Consequently every 4 eq. of missing chlorine, i.e. less than 
that which would be obtained by distilling the weighed 
amount of bichromate alone with hydrochloric acid, represent 
i eq. of arsenic trioxide. The details of the method are 
identical with those of the foregoing example. 

XXVIII. ANALYSIS OF CHLORATES, BROMATES, AND 
IODATES. 

i eq. of potassium chlorate, heated with a large excess of 
hydrochloric acid, evolves an amount of chlorine partly 
free and partly in combination with oxygen sufficient to 
liberate 6 eq. of iodine. 761*1 parts of iodine are therefore 
equivalent to 122-56 of potassium chlorate. The weighed 
quantity of the chlorate (about 0*2 gram is sufficient) is 
placed in the distillation flask and heated with excess of 
hydrochloric acid. The remainder of the operation is con- 
ducted in the manner described. Bromates and iodates are 
best analysed by the method of digestion instead of by that 
of distillation. A strong bottle of about 120-150 c.c. is 
fitted with an accurately-ground stopper: the weighed 
amount of the bromate is placed in the bottle, the requisite 
amount of a saturated solution of potassium iodide and 
hydrochloric acid is added, and the stopper is firmly fastened 
down by binding wire. The bottle is then placed on the 
water-bath ; when the decomposition is complete it is allowed 
to cool, and its contents are diluted with water ; the solution 
is emptied into a beaker, and the titration proceeded with 
in the usual manner. 



1 66 Quantitative Chemical A nalysis. 

In the case of iodates and bromates, only 4 eq. of iodine 
are liberated for each eq. of the acid. 

The method of digestion may be frequently substituted 
for that of distillation. It is of course absolutely necessary 
that the stopper of the bottle fits perfectly: to test it, it 
should be tied down in the empty bottle, which is then to 
be immersed in hot water ; if any air-bubbles make their 
escape between the stopper and the neck, the bottle is 
useless for this purpose. The stopper, if nearly tight, may be 
re-ground with a little fine emery and water. In every case 
it must be carefully tested before use. 

XXIX. ESTIMATION OF IRON BY MEANS OF IODINE AND 
THIOSULPHATE SOLUTIONS. 

When ferric chloride is added to a warm solution of 
potassium iodide the following reaction ensues : 

Fe 2 Cl 6 + 2KI = 2FeCl 2 + 2 KC1 + I 2 . 
126*85 parts of iodine correspond to 56 parts of iron. 

Dissolve 5*02 grams of clean piano- wire (corresponding 
to 5 grams of pure iron) in dilute hydrochloric acid, in a 
^-litre flask; add a few crystals of potassium chlorate, to 
convert the iron into ferric chloride, boil the solution for 
some time, to expel the excess of chlorine, and when cold 
dilute the solution to the containing-mark. 10 c.c. of the 
solution correspond to o'i gram of iron. Transfer 20 c.c. 
of the iron solution to the stoppered bottle, cautiously add 
dilute caustic soda solution until a slight precipitate of oxide 
of iron remains, and then i c.c. of hydrochloric acid 
(sp. gr. i'i). Now add about 4 grams of potassium iodide 
to the clear dark-yellow solution, insert the stopper, and 
fasten it down by means of binding wire. Heat the bottle 
over the water-bath, for ten or fifteen minutes, to about 60, 
allow it to cool completely, open it, and add sodium thio- 
sulphate from a burette, until the solution is nearly de- 
colourised, add i c.c. of starch liquor, and continue the 
addition of the thiosulphate until the blue colouration just 
disappears. 



Use of Iodine, &c. 1 67 

This method is particularly useful for the determination of 
small quantities of iron. The solution must, of course, be 
fully oxidised, and contain no other substance which can 
eliminate iodine. If strongly acid, it must be partially 
neutralised by soda solution, in the manner described. 

XXX. ESTIMATION OF NITRIC ACID BY SOLUTIONS OF 
IRON, IODINE, AND SODIUM THIOSULPHATE. 

Free nitric acid, added to a solution of ferrous chloride, 
converts the iron into ferric chloride. Ferric chloride, as we 
have seen in the foregoing process, may be estimated by 
means of iodide of potassium and sodium thiosulphate. 

Weigh out about 07 gram of iron- wire, and dissolve it in 
a small quantity of hydrochloric acid, in a flask in a 
current of carbon dioxide, in order to prevent the least 
chance of oxidation. Weigh out about 2 grams of nitre 
into a ^-litre flask, dissolve in boiling water, and dilute 
with boiled water to the containing-mark. Transfer 25 c.c. 
of the solution, corresponding to 0-2 gram of nitre, to the 
iron solution, and quickly re-insert the cork ; gently heat the 
liquid, at length to boiling, to expel the nitric oxide. Main- 
tain a rapid . current of carbon dioxide throughout the 
operation. As soon as the solution is of a pure yellow 
colour, allow it to cool in a current of carbon dioxide, add 
a sufficiency of potassium iodide, allow the solution to stand 
for a short time, and determine the amount of liberated 
iodine by starch and thiosulphate solution, exactly as de- 
scribed in the preceding method. The quantity of thiosul- 
phate used, multiplied by 0*0021, gives the amount of nitric 
acid present. 

XXX. VALUATION OF MANGANESE ORES BY DISTILLATION 
WITH HYDROCHLORIC ACID, AND TITRATION WITH IODINE 
AND THIOSULPHATE SOLUTIONS. 

(See Part IV.) 



i68 



Quantitative Chemical A nalysis. 



PART IV. 

GENERAL ANALYSIS, INVOLVING GRAVIMETRIC 
AND VOLUMETRIC PROCESSES. 

I. NITRE. 

THE crude nitre of commerce invariably contains alkaline 
chlorides and sulphates, together with more or less insoluble 
matter and moisture. Samples of nitre are occasionally met 
with containing sodium nitrate, arising either from imperfect 
decomposition of the Chili saltpetre, from which the nitre 
was prepared, or from wilful adulteration. Such nitre is 
highly hygroscopic, and requires to be purified before it can 
be used for the manufacture of gunpowder. 

The presence of sodium nitrate, or excess 01 common 
salt in the nitre, may be readily detected by means of the 
spectroscope. The quantity of admixed sodium nitrate 
may be approximately determined by ascertaining the 
amount of water taken up from a perfectly moist atmo- 
sphere. Pure nitre placed over the surface of water for 
a fortnight remains comparatively dry ; sodium nitrate, 
under the same circumstances, absorbs one-fourth of its 
weight of water. The following table shows the amount of 
water taken up by 100 grams of the mixed nitrates : 



Percentage of Sodium Ni-"l 
trate . . . / 


0*5 


I 


3 


5 


10 


Amount of Water (in grams) "1 
absorbed in 14 days . J 


2'5 


4 


10 


12 


19 



Determination of Moisture. About 20 grams of the 
sample are gently heated in a weighed platinum crucible, 



Nitre. 169 

until the salt commences to fuse. When cold, the crucible 
is again weighed. The loss gives the amount of moisture. 

Determination of Insoluble Matter. The contents of the 
crucible are washed out into a porcelain basin with hot 
water, dissolved, filtered, and the insoluble matter dried and 
weighed. The filtrate is received into a ^-litre flask, allowed 
to cool, and diluted to the containing- mark. 

Determination of Chlorine. 100 c.c. of the liquid, corre- 
sponding to 4 grams of nitre, are transferred to a porcelain 
basin, and the chlorine estimated by means of standard 
silver nitrate and potassium chromate solutions. If the 
amount of chlorine is very small, it is advisable to use centi- 
normal silver solution. 

Determination of Sulphuric Acid. 250 c.c. of the liquid 
are transferred to a beaker, heated to boiling, and acidu- 
lated with a small quantity of hydrochloric acid. Barium 
chloride solution is added, and the liquid is set aside for a 
time, to allow the precipitate to subside perfectly. As 
barium sulphate is slightly soluble in solutions of the 
alkaline nitrates, the precipitation is not quite complete. 
The clear supernatant liquid is poured through the filter, 
and the precipitate is washed several times by decantation 
with boiling water, taking care to allow it to settle as com- 
pletely as possible, before pouring the liquid on to the 
filter. The barium sulphate still retains co-precipitated 
nitrate. This may be best removed by a solution of copper 
acetate. Crystallised copper acetate is dissolved in a small 
quantity of hot water, containing acetic acid ; one drop of 
sulphuric acid is added, and then one drop of barium chlo- 
ride solution; the mixture is boiled and filtered. About 
5 or 10 c.c. of the saturated solution, according to the amount 
of the precipitate, is added to the barium sulphate, together 
with a small quantity of water and a few drops of acetic 



170 Quantitative Chemical Analysis. 

acid. The liquid is heated to boiling, and maintained in 
ebullition for ten or fifteen minutes. The amount of acetic 
acid added should be sufficient to prevent the precipitation 
of any basic salt of copper. Pour the liquid through the 
filter, transfer the precipitate, and wash it thoroughly with 
hot water. Dry, ignite, and weigh it. This procedure is 
recommended to be followed in all precipitations of sul- 
phuric acid by barium salts, in presence of considerable 
quantities of alkaline nitrates or chlorides. 

Determination of Nitric Add. Fuse a few grams of the 
nitre at the lowest possible temperature, and pour out the 
liquid mass into a warm porcelain dish, powder it quickly, 
and transfer it to a tube. Place 2 or 3 grams of powdered 
quartz in a platinum crucible, heat to redness, and weigh 
accurately after cooling. Add about 0-5 gram of the nitre, 
and again weigh. Mix the nitre and silica by the aid of a 
thin glass rod, taking care, of course, that nothing adheres 
to the rod, and heat the crucible gradually to a low red heat, 
keeping it at this temperature for twenty or thirty minutes, 
transfer to the desiccator, and weigh when cold. The loss 
of weight gives the amount of nitric acid. Care must be 
taken to regulate the temperature properly, or the sulphates 
and chlorides present (particularly the latter) may partially 
volatilise. Potassium bichromate and borax may be substi- 
tuted for the powdered quartz ; the latter substance, how- 
ever, is preferable. 

The nitric acid may also be determined by the method 
described on p. 96. 

II. GUNPOWDER. 

Gunpowder is an intimate mixture of sulphur, nitre, and 
charcoal. It invariably contains also a small quantity of 
moisture. In the following scheme of analysis all these con- 
stituents are determined in a single portion of the sample. 

A light glass tube (a, fig. 51), about 10 centimetres long 



Gunpowder. 



171 



and i centimetre in diameter, is drawn out near the end by 
the aid of the blow-pipe. The contracted portion should 
measure about 5 centimetres long, and possess an internal 
diameter of 0*2 centimetre. At the point where the tube is 
narrowed, place a plug of recently-ignited asbestos, from i -5 
to 2 centimetres long. Accurately weigh the tube, place in 
it about 3 grams of the triturated powder, and again 
weigh : the increase of course gives the amount taken for 

FIG. 51. 




analysis. By the aid of the filter-pump aspirate a gentle cur- 
rent of air, dried by sulphuric acid, through the tube for 10 
or 1 2 hours, and again weigh. The loss indicates the amount 
of moisture. 

The tube is next fitted into a light flask of about 25 c.c. 
capacity, provided with a side-tube in the manner seen in 
fig. 51. The flask is accurately weighed and connected with 
a tube of thin glass 25 to 30 centimetres long. Portions of 



172 Quantitative Chemical Analysis. 

about 3 c.c. of carbon bisulphide (free from moisture, and 
rectified by agitation with mercury, and redistillation) are 
poured over the powder. The filtrate running into the 
flask should be perfectly clear. As soon as the flask is 
about half filled, it is heated by hot water at about 70. 
The carbon bisulphide distils over, and is collected in a 
dry test-tube surrounded by cold water. The distillate is 
again poured on to the powder, and again distilled, the 
operation being repeated 6 or 8 times. After the last distil- 
lation the residual sulphur is gently heated, and a current of 
dry air is drawn through the entire apparatus by attaching 
the side-tube to the filter-pump. The flask is re-weighed : 
the increase in weight gives the amount of the sulphur in the 
powder which it is possible to extract by means of bisulphide 
of carbon. 

The tube containing the residual charcoal and nitre, 
together with the minute quantity of sulphur still left in the 
exhausted powder, is heated to 100, and a current of air, 
dried by sulphuric acid, is drawn over it. The tube is again 
weighed : its decrease in weight will be slightly greater than 
the increase in weight of the flask containing the sulphur. 
The difference is the amount of moisture which the powder 
would give up if dried at 100. This slight difference is to 
be added to the quantity of moisture already determined. 

About i gram of the exhausted powder is shaken out into 
a porcelain basin, and the tube is re-weighed. The loss of 
weight shows the amount transferred to the basin. This 
powder is then gently heated with nitric acid (perfectly free 
from sulphuric acid) and a few crystals of potassium 
chlorate. The liquid is evaporated to dryness with hydro- 
chloric acid, dissolved in water, filtered if necessary, and 
the sulphuric acid precipitated in the ordinary manner by 
barium chloride. The barium sulphate is dried, ignited, and 
weighed, and the amount of sulphur equivalent to it is added 
to the main quantity extracted by the sulphide of carbon. 

The tube containing the remainder of the exhausted pow- 



Gunpowder. 



173 



FIG. 52. 




der is now treated with water to extract the nitre. It 
is fitted into the bell- 
jar (fig. 52), standing on 
a plate of ground glass. 
Within the jar and under- 
neath the tube is a 
weighed platinum dish. 
The side tube is con- 
nected with the filter- 
pump. A few cubic 
centimetres of cold water 
are poured on to the pow- 
der, and the pump is set 
in operation so as to cause 
the liquid to fall, drop by 
drop, into the basin. To 
avoid loss by splashing, 
the basin should be as near to the edge of the tube as 
possible. Successive small quantities of water of a gradually 
increasing temperature are now poured over the powder, 
water as hot as possible being used for the last washings. 
50 c.c. of water are amply sufficient to extract all the nitre 
from the residue of the powder, if care be taken not to use 
too much water for each washing. The use of large quantities 
of water for washing is not advisable, since appreciable 
quantities of organic matter are thereby liable to be dissolved 
out of the charcoal. The solution of the nitre is evaporated 
to dryness in the dish, dried at 120, and weighed: the 
weight is of course calculated upon the entire quantity of 
powder. The asbestos-plug is now detached from the tube 
by the aid of a platinum wire, and the tube and its contents 
are again dried at 100, and weighed. It will be generally 
found, if the process has been properly carried out, that the 
weight of the charcoal is slightly greater (generally from i to 
2 milligrams) than the amount calculated from the quantity 
of nitre found : the difference is due to the fact that the pure 



174 Quantitative Chemical Analysis. 

charcoal retains water more tenaciously, even after drying 
for some time at 100, than when mixed with nitre. 

It is sometimes necessary to determine the amount of car- 
bon, hydrogen, and oxygen in the residual charcoal : this 
may readily be effected by mixing it, together with the 
asbestos, with lead chromate, and burning it in a stream of 
oxygen, and collecting the carbon dioxide and water in the 
manner described under organic analysis. In calculating 
the amount of hydrogen, due regard must be taken of the 
water retained by the charcoal after drying at 100. 

III. LIMESTONES. HYDRAULIC MORTAR. 

Limestone is essentially calcium carbonate, containing 
more or less magnesium carbonate, ferrous and manganous 
carbonates or oxides, alumina, silica, and alkalies. Many 
limestones also contain variable quantities of clay, sand, and 
organic matter, together with chlorine, fluorine, phosphoric 
and sulphuric acids, and iron pyrites. 

The method given on p. 85, which includes the determi- 
nation of the essential constituents, will generally suffice for 
the examination of limestone for technical purposes, but 
occasionally it is required to estimate the substances which 
are present in smaller proportion. 

Powder about 100 grams of the mineral, mix uniformly, 
and dry at 100. Weigh out about 2 grams into the flask A 
(fig. 31), and determine the amount of carbon dioxide in the 
manner directed on p. 86". Rinse the solution into a porce- 
lain basin, evaporate to complete dryness, moisten with a 
few drops of hydrochloric acid, dissolve in hot water, filter 
through a weighed filter, wash the residue, and dry it at 100. 
It may consist of sand, clay, and separated silicic acid, and 
organic matter. The proportion of these several substances 
will be estimated hereafter. 

Add a few drops of bromine-water to the filtrate, and then 
ammonium chloride and a slight excess of ammonia, cover 
the beaker, and heat it gently for some time. The precipi- 



Limestones, &c. 175 

tare contains the oxides of iron, manganese, and aluminium, 
together with the phosphoric acid ; it is thrown on to a filter, 
washed once or twice, re-dissolved in a small quantity of 
hydrochloric acid, again mixed with bromine-water, and re- 
precipitated with ammonia, and again filtered. The second 
precipitation effects the removal of the small quantities of 
lime and magnesia which are invariably thrown down with 
the oxides on the first precipitation. The precipitate is well 
washed, dried, and weighed. 

The lime and magnesia in the mixed nitrate are separated 
as directed on p. 88. The lime may be estimated volume- 
trically, as described on p. 152, or it may be weighed as 
carbonate or oxide. 

Determination of the Constituents present in small quantity. 
Dissolve about 50 grams of the mineral in dilute hydro- 
chloric acid in a porcelain basin, heat the solution gently to 
expel carbon dioxide, and filter through a weighed filter into 
a litre flask, wash the residue thoroughly, dry, and weigh it. 
Dilute the filtrate up to the containing-mark. 

Analysis of the Insoluble Residue. (a) Weigh out about 
one-fourth of the insoluble matter into a platinum basin, and 
boil it with strong solution of pure sodium carbonate. Filter, 
and determine the silicic acid in solution by acidulation with 
hydrochloric acid, and evaporation to dryness in a platinum 
basin. 

(b) Weigh out another portion into a platinum crucible 
and fuse with pure sodium and potassium carbonates, extract 
with hot water, acidulate with hydrochloric acid, evaporate 
to dryness, and separate the silica in the usual manner. 
Deduct the amount of the silica soluble in solutions of 
alkaline carbonate (a). 

(c) A third portion of the residue is weighed out into a 
platinum boat and heated in a current of oxygen in a com- 
bustion-tube partly filled with copper oxide, in order to 



176 Quantitative Chemical Analysis. 

determine the amount of organic matter (humus). Accord- 
ing to Petzholdt humus contains 58 per cent, of carbon. 

(d) Iron pyrites is not an infrequent constituent of lime- 
stones : it is found in the insoluble residue, after treatment 
with dilute hydrochloric acid. To determine its amount the 
remainder of the insoluble residue is heated with nitric acid 
and potassium chlorate, and the proportion of the pyrites 
is calculated from the quantity of sulphuric acid obtained. 

Analysis of the Solution. Transfer 500 c.c. of the liquid 
to a porcelain basin, evaporate to complete dryness, and 
heat the saline mass until fumes of hydrochloric acid are no 
longer visible. Moisten with strong hydrochloric acid, add 
hot water, and filter the solution, wash the residue, and 
weigh it in a platinum crucible. It consists mainly of silica, 
but may contain sulphates of strontium and barium. Call 
it Pp. I. 

To the filtrate add a few drops of nitric acid, boil, add 
ammonia, and again boil until the fluid no longer smells of 
ammonia, filter, wash the precipitate once or twice, dissolve 
it in hydrochloric acid, and precipitate again with ammonia. 
Call it Pp. II. : it contains ferric oxide, alumina, and phosphoric 
acid. The mixed filtrates are received in a flask (which they 
should nearly fill), and mixed with ammonium sulphide. 
The flask is closed and set aside for 24 hours in a warm 
place. The precipitate consists of manganese sulphide. 
Filter it off and wash it with water containing ammonium 
sulphide. Call it Pp. III. 

Treatment of the Precipitates /., //., and ///.Pp. I. 
Moisten the weighed precipitate with pure hydrofluoric acid 
and a drop or two of sulphuric acid, and evaporate to dry- 
ness. Repeat this operation, and fuse any residue with 
sodium carbonate, digest with hot water, filter, dissolve the 
washed precipitate in hydrochloric acid, and add a drop or 
two of sulphuric acid to the solution. Filter into a small 



Limestones, &c. 177 

flask (filtrate a), wash the precipitate, and digest it on the filter 
for 1 2 or 1 5 hours with solution of ammonium carbonate, the 
tube of the funnel being meanwhile closed by a rod during the 
digestion. Open the tube, wash the precipitate with water, and 
treat it with dilute hydrochloric acid (filtrate b\ again wash 
with water, and weigh the residual barium sulphate. Mix a 
and b, add ammonia and ammonium carbonate ; if a precipi- 
tate forms on standing it consists of strontium carbonate ; it is 
filtered, washed with ammonia water, dried, and weighed. 

Pp. II. Dissolve in hydrochloric acid in a small flask, add 
pure tartaric acid, ammonia, and ammonium sulphide. Close 
the flask, and after standing a few hours, wash the iron sul- 
phide with water containing a few drops of ammonium 
sulphide. Dissolve in hydrochloric acid, add a crystal or 
two of potassium chlorate, boil, and precipitate with ammonia, 
and weigh the ferric oxide. 

To the yellow-coloured filtrate, containing the alumina 
and phosphoric acid, add a small quantity of pure sodium 
carbonate and nitre, evaporate to dryness, and ignite until 
the residue is free from carbon. Digest with water once or 
twice, pour off the solution into a beaker, and treat the 
residue with warm hydrochloric acid, add the solution to 
that contained in the beaker, filter if necessary, and mix the 
filtrate with ammonia. Filter off the precipitate and weigh 
it. Mix the filtrate with a few drops of magnesia mixture : 
if a precipitate is again formed, it consists of ammonium 
magnesium phosphate : it is washed with ammonia water 
and weighed as magnesium pyrophosphate. In that case 
the first precipitate has the composition A1 2 P 2 O8 : if no 
precipitate form, it is probably a mixture of alumina and 
aluminium phosphate. It is redissolved in a small quantity 
of hydrochloric acid and mixed with molybdic acid solution, 
and treated as directed under Iron Ores Estimation of 
Phosphorus. 

Pp. III. consists almost entirely of manganese sulphide. 
It is treated with moderately dilute acetic acid, and the solu- 

N 



178 Quantitative Chemical Analysis. 

tion is filtered if necessary ; if any insoluble matter remains, 
it is tested for the metals of Group III. The filtrate is heated 
to boiling, nearly neutralised with caustic soda, and mixed 
with bromine water, and the manganese dioxide filtered, 
washed, dried, ignited, and weighed as trimanganic tetroxide.* 

Determination of the Alkalies.^ Transfer 300 c.c. of the 
solution to a flask, add bromine water, heat gently, and mix 
with ammonia and ammonium carbonate. Allow the 
liquid to stand for some hours, filter, wash, evaporate the 
filtrate to dryness in a platinum dish, and ignite to remove 
the ammonia salts. Dissolve in water, boil with a little milk 
of lime, filter, wash, remove the excess of lime by ammonium 
carbonate and oxalate, filter, wash, and evaporate the filtrate 
to dryness ; again ignite, dissolve in a few drops of water, 
filter once more if necessary, acidulate with hydrochloric 
acid, and evaporate the solution of the mixed alkaline 
chlorides to dryness in a weighed platinum dish. The 
potassium and sodium are then separated by means of plati- 
num tetrachloride, as directed in No. IV. p. 85. 

Determination of Sulphuric Acid. To the remainder of 
the original solution, add one or two drops of barium 
chloride, and allow the liquid to stand for some time. If 
any precipitate of barium sulphate is formed, filter it off, 
wash, dry, and weigh it. 

* Chatard determines the small quantity of manganese present in dolo- 
mites, limestones, &c., by dissolving the mineral in dilute nitric acid, 
boiling the solution with a small quantity of lead peroxide, filtering 
through an asbestos filter, and volumetrically determining the amount of 
permanganic acid in the liquid by a dilute solution of ammonium 
oxalate of known strength. 

t Alkalies when present in limestones or dolomites may be readily 
detected by strongly heating the mineral in a platinum crucible, boiling 
with a little water, filtering, acidulating with hydrochloric acid, 
adding ammonia and ammonium carbonate, filtering, evaporating the 
filtrate to dryness, and examining the residue by means of the specto- 
scope. The ammonium carbonate precipitate may also be treated with 
hydrochloric acid, evaporated to dryness, and examined for barium and 
strontium in the same manner. (ENGELBACH.) 



Limestones, &c. 179 

Determination of Chlorine. Chlorides are occasionally 
present in dolomites and limestones : the amount of chlorine 
in them may be determined by dissolving a quantity of the 
mineral in dilute nitric acid, filtering, and adding a few drops 
of silver nitrate to the filtrate, and treating the silver chloride 
as directed in No. II. p. 81, 

Determination of Fluorine. A large quantity of the 
mineral is dissolved in acetic acid, the solution is evaporated 
to dryness, and heated to expel the excess of acetic acid. 
The mass is repeatedly treated with water, the residue is 
weighed, and a portion is tested for fluorine. If this substance 
is found in estimable quantity, fuse the remainder of the 
insoluble residue with sodium carbonate, boil with water, 
filter, and wash with boiling water and solution of ammonium 
carbonate. Heat the filtrate, which contains all the fluorine, 
with an additional quantity of ammonium carbonate, and after 
some time filter off the precipitated silica and alumina. Boil 
the filtrate until the ammonium carbonate is completely ex- 
pelled, and add a few drops of calcium chloride solution ; if 
any precipitate of calcium fluoride forms, filter it off, wash 
thoroughly with hot water, dry, ignite, and weigh. To ensure 
the absence of calcium carbonate in the weighed precipitate, 
digest it with dilute acetic acid, allow the precipitate to 
subside, pass the liquid through a small filter, wash by 
decantation, dry the precipitate and filter, burn the latter, 
add the ash, and again weigh. 

Determination of Water retained in the Mineral after heating 
to 100. This is effected in the apparatus seen in fig. 53. 
The weighed quantity of the mineral (about 3 grams) is 
heated in the bulb, made of difficultly -fusible glass, in a 
stream of aspirated air, dried by passing through the first 
calcium-chloride tube ; the water evolved is absorbed in the 
second weighed tube, which also contains calcium chloride. 
The increase in the weight of this tube at the termination of 

N 2 



i8o 



Quantitative Chemical Analysis. 



the experiment gives the amount of moisture present in the 
mineral. The little flask contains strong sulphuric acid ; it 
serves to indicate the rate of the current of air, and prevents 
the possibility of moisture diffusing into the weighed calcium- 
chloride tube. 

CEMENT-STONE, as the material used in the manufacture 
of hydraulic mortar is termed, is an impure limestone con- 
taining a considerable quantity of ferrous carbonate, alumina, 

\ 

FIG. 53 




and silica. Hydraulic mortar owes its property of harden- 
ing under water to the gradual formation of hydrated sili- 
cates of lime and alumina, which are very dense and insolu- 
ble in water. 

IV. CLAYS. 

Clay is a hydrated aluminium silicate, derived from the 
decomposition of felspar. The purer varieties are perfectly 
white, and contain but small quantities of lime, magnesia, and 



Clay. 181 

oxide of iron. The red colour of the common brick-clay 
after burning is due to the ferric oxide which it contains. 
The different clays used in the arts may be classed under 
the heads of slate clay, common clay, fire-clay, plastic clay, 
and kaolin. These varieties have essentially the same quali- 
tative composition ; their different properties are mainly 
due to the relative amounts of the admixed substances. 
Pure clay is nearly infusible ; but if mixed with a sufficient 
amount of iron and lime, it may be more or less readily 
melted, especially if free silica be present. The ease with 
which it may be fused depends upon the proportion of these 
admixtures. The following analyses will serve to indicate 
the composition of ordinary fire-clays : 

No. i, BEST STOURBRIDGE CLAY. No. 2, INFERIOR FIRE-CLAY, 
FAULTY GLASS POT. No. 3, BRICK CLAY OF AVERAGE QUALITY. 

T. 2. 3. 

Silica . 73*82 69-91 49 -44 



Alumina 
Ferrous Oxide 
Lime 
Magnesia 
Alkalies . 
Water 



15-88 17-44 34-26 

2'95 Fe 2 3 2-89 Fe.Oa 774 

trace 3 -08 I -48 

trace 4-47 5-14 

0-90 2-21 

6-45 1-94 



100-00 100-00 



The silica in clay exists partly as sand, partly as hydrate, 
and partly in combination with bases ; the amount of free 
hydrated silica seldom exceeds i per cent. The sand may 
vary from 15 to 60 per cent. 

The porcelain earth, or kaolin of the Chinese, is almost 
pure hydrated silicate of aluminium, containing undecom- 
posed felspar and free silica. It has been formed from 
orthoclase by the gradual abstraction of the whole of the 
alkalies, and of about f of the silica. When freed from felspar 
and admixed silica, its average composition is Al 2 O 3 2SiO 2 + 
2H 2 O, but varieties of kaolin are frequently found in which 
the relation of alumina to silica is very different. 

For many of its applications, it is important to know the 



1 82 Quantitative Chemical Analysis. 

relative proportions of coarse and fine sand present in the 
sample, in addition to the true clay. The chemical exami- 
nation, is, therefore, usually preceded by a mechanical 
analysis. 

A. Mechanical Analysis. 

This may be effected with sufficient accuracy by elutria- 
tion. The powdered clay is agitated in a stream of water ; 
the coarse particles subside, the finer particles are carried 
away by the current, and are received in a large beaker, or 
other suitable vessel, where they are allowed to settle. Of 
the several forms of apparatus proposed for this purpose, the 
simplest is that devised by Schulze. To a tall and narrow 
glass (one of the old forms of champagne glass does very 
well) about 20 centimetres deep, and 7 centimetres in dia- 
meter at the mouth, is fastened a brass rim, about 2 centi- 
metres broad, carrying a short tube inclined downwards. 
A slow stream of water is allowed to pass into the glass, 
through a funnel-tube about 40 centimetres long, and 
7 millimetres in diameter ; the bulb of the funnel should be 
about 5 centimetres in diameter, and its end should be 
drawn out until the opening is only \\ millimetre in 
diameter. Triturate 30 or 40 grams of the air-dried clay in 
a mortar, transfer it to a porcelain dish, and boil it with 
about 70 c.c. of water for thirty minutes. Repeatedly crush 
the sedimentary matter with a pestle, and agitate the liquid 
so as to disintegrate the clay completely. Allow to cool, 
and transfer the contents of the dish to the narrow glass ; 
suspend the tube-funnel within the glass, so that its end is 
about 2 or 3 millimetres from the bottom, and so regulate 
the stream of water that the funnel is kept constantly half- 
filled with water. The finer particles are stirred up, and are 
carried away in the current of water, through the lateral 
tube, into the beaker placed to receive it ; the coarse sand 
remains in the elutriating glass. The flow of the water is 
arrested when it runs away nearly clear, and the liquid still 



Clay. 183 

in the glass is decanted into the beaker. The coarse sand 
is washed out into a porcelain crucible, dried, ignited, and 
weighed. 

In about six hours the liquid in the beaker will be nearly 
clear ; the whole of the fine sand will certainly have been 
deposited by that time. The supernatant liquid is poured 
away : the deposit is rinsed back into the narrow glass, and 
the process of elutriation is repeated. The flow of the water is 
so regulated that its level in the funnel-tube is about 3 centi- 
metres higher than that in the glass ; in about four hours the 
whole of the clay proper will have been carried away through 
the discharge-pipe. The residual fine sand is then rinsed into 
a weighed porcelain crucible, dried, ignited, and weighed. 

The water in the air-dried sample is determined by igni- 
ting a second portion for some time ; the amount of the clay 
proDer is determined by difference. 

B. Chemical Analysis. 

The air-dried clay is powdered as finely as possible, a 
portion is weighed out into a large porcelain crucible, and 
heated in the steam-chamber for several days. The moisture 
is calculated from the loss. 

About 2 grams of the dried powder are then heated in a 
platinum dish with excess of moderately-concentrated sul- 
phuric acid for about 8 or 10 hours, and the mass is 
evaporated to dryness to expel the acid. When cold, the 
residue is boiled with water, the solution is filtered, and the 
insoluble matter, consisting of the silica originally existing in 
union with the bases in the clay, and also of the small 
quantity of free silica, together with the sand, is washed, 
dried, and weighed. It is then repeatedly boiled with solu- 
tion of sodium carbonate in the platinum dish, the liquid 
filtered, and the residual sand washed with hot water, next 
with water slightly acidified with hydrochloric acid, and 
finally with pure water. It is then dried and weighed. Its 
weight, subtracted from the total weight of the residue left 



1 84 Quantitative Chemical Analysis. 

after treatment with sulphuric acid, gives the amount of true 
silicic acid. 

The quantity of uncombined silicic acid in the clay may 
be determined by boiling a weighed quantity of the sample 
dried at 100 with a strong solution of sodium carbonate, 
filtering the liquid, and evaporating to dryness with excess 
of hydrochloric acid. The separated silicic acid is washed, 
dried, and weighed. Titanium dioxide is not an infrequent 
constituent of clays. It may be detected in the residue after 
treatment with sulphuric acid : an aliquot portion is heated 
with hydrofluoric and sulphuric acids, whereby the greater 
portion of the silica is volatilised : the residue is fused with 
acid sulphate of potassium. Dissolve in cold water, filter if 
necessary, and boil. The titanium dioxide isreprecipitated: 
it is to be washed, dried, and weighed. To the filtrate 
obtained after treatment with sulphuric acid, and containing 
the bases, add a slight excess of lead nitrate solution, and 
allow the turbid liquid to stand for a few hours. Filter, 
wash the lead sulphate, adding the washings to the filtrate, 
and pass sulphuretted hydrogen through the liquid to remove 
the excess of lead, again filter, and evaporate the solution to 
dryness. The alumina, iron, lime, magnesia, and alkalies 
are obtained as nitrates. Heat gradually to about 250 until 
no more fumes of nitric acid are evolved. The residue 
consists of alumina and ferric oxide, calcium, magnesium, 
and alkaline nitrates. Moisten the residue with a concen- 
trated solution of ammonium nitrate, and heat on the water- 
bath until no further evolution of ammonia is perceived. 
Add hot water, filter, wash the alumina and ferric oxide, 
dry, ignite, and weigh. Transfer the weighed oxides to a 
porcelain boat, and heat in a stream of dry hydrogen or coal- 
gas for half an hour. Treat the mixture of alumina and 
reduced iron with very dilute nitric acid (i pt. acid to 40 of 
water), warm, filter, and precipitate the ferric oxide by 
ammonia, wash, dry, and weigh it. The mixture may also 
be heated with dilute sulphuric acid, filtered, reduced with 



Manganese Ores. 185 

a small piece of zinc, and titrated with a weak solution 
of potassium permanganate. Deduct the weight of ferric 
oxide from that of the mixed oxides : the difference gives 
the quantity of alumina. 

To the nitrate containing the nitrates add ammonium 
oxalate and weigh the precipitate as caustic lime. Evaporate 
the nitrate to dryness, ignite to expel the ammonium salts, 
add excess of oxalic acid (sufficient to convert all the bases 
present, considered as potash, into quadroxalates), treat with 
a small quantity of water, and again evaporate to dryness. 
Ignite gently : the magnesium oxalate is converted into 
magnesia, and the alkaline oxalates into carbonates. Treat 
repeatedly with small quantities of water, filter, dry, and weigh 
the magnesia. 

Add a few drops of hydrochloric acid to the filtrate, 
evaporate to dryness, ignite gently, and weigh the alkaline 
chlorides. The potash and soda may be separated by platinum 
tetrachloride (see p. 84), or their proportion may be deter- 
mined by a dilute standard silver solution in the manner 
described on p t 127. 

The magnesia may also be precipitated by sodium phos- 
phate, and the alkalies determined in a separate portion by 
ignition with calcium carbonate and ammonium chloride 
(see Glass, p. 99). 

ASSAY OF MANGANESE ORES (PYROLUSITE, 
BRAUNITE, &c.). 

Of the many methods which have been proposed for the 
valuation of these substances, those of Bunsen and of 
Fresenius and Will are on the whole the most accurate 
and convenient. As the value of the oxide depends upon 
the amount of chlorine which it yields on heating with 
hydrochloric acid, the method of Bunsen is perhaps the 
more generally applicable, since it directly determines the 
amount of chlorine thus evolved. It has the advantage too 
of being rapidly carried out, and is therefore well adapted 



1 86 Quantitative Chemical Analysis. 

to the requirements of manufacturing establishments. When 
manganese dioxide is brought into contact with hydrochloric 
acid, the following reaction occurs : 

MnO 2 + 4HC1 = MnCl 2 + C1 2 + 2H 2 O. 

70*92 parts of chlorine or 2537 parts of iodine are equiva- 
lent to 86*04 parts of manganese dioxide (Mn = 54*04, 
Schneider). If the chlorine be led into a solution of 
potassium iodide, an equivalent quantity of iodine is libe- 
rated ; this may be determined by the method given on p. 164. 

Weigh out 0*5 gram of the finely-powdered ore into the 
little bulb a (fig. 50), add hydrochloric acid, and drive over the 
chlorine into the U-tube, in which you have previously 
placed 25 cubic centimetres of the strong solution of potas- 
sium iodide. 

The remainder of the operation is conducted as described 
on p. 163. 

V. Gravimetrical Method of Fresemus and Will. 

This process depends upon the action of manganese 
dioxide on oxalic acid, in presence of sulphuric acid ; when 
these substances are brought together, the manganese dioxide 
parts with an atom of its oxygen to the oxalic acid, which is 
thereby completely converted into carbon dioxide and water, 
and the manganese protoxide combines with sulphuric acid 
to form manganous sulphate ; thus : 

Mn0 2 + C 2 2 (OH) 2 + H 2 S0 4 = MnSO 4 + 2CO 2 + 2 H 2 O. 

88 parts of carbon dioxide evolved, correspond therefore to 
86*04 parts of manganese dioxide. 

The decomposition may be effected in the apparatus 
represented in fig. 31. From 2 to 4 grams of the finely- 
powdered ore, according to its supposed richness, are 
weighed out into the flask A, and covered with dilute sul- 
phuric acid. A strong solution of oxalic acid is then allowed 
to flow from the bulb-tube, and the evolved carbonic acid 



Manganese Ores. 187 

collected in the weighed soda-lime tube c. It is advisable to 
place a weighed potash-apparatus (Geissler's form is the 
most convenient) before the U-tube. 

Certain ores of manganese contain considerable quantities 
of earthy carbonates ; on treatment with an acid their car- 
bonic acid is of course liberated, and thus tends to in- 
crease the apparent value of the sample. By means of 
the apparatus above mentioned, this amount of carbonic 
acid may be readily determined ; it is only necessary to 
re-weigh the absorption tubes before the addition of the 
oxalic acid. Of course, the precaution must be taken to 
aspirate air through the appa- 
ratus before disconnecting 
the several parts. 

The original apparatus de- 
vised by Fresenius and Will 
is represented in fig. 54. It 
consists of two flasks, a and , 
connected together by a thin 
glass tube d^ one end of which 
terminates just below the 
cork of a, the other ends a 
few millimetres above the 
bottom of b. The thin tube c passes down to the bottom 
of a -, the short tube e ends just below the cork of b. The 
two flasks are of about 100 cubic centimetres capacity ; they 
are made of thin glass, so as to weigh as little as possible. 
Weigh out into a from 3 to 5 grams of the powdered 
sample, add from 5 to 6 grams of sodium oxalate, and half 
fill b with strong sulphuric acid. Weigh the entire apparatus. 
Place a short piece of caoutchouc tubing over c 9 and close 
it by a glass rod, slip a longer piece of caoutchouc tubing 
over <?, and aspirate two or three bubbles of air from a 
through the sulphuric acid. On discontinuing the suction, 
the acid rises in the tube ; if the column remains stationary, 
the apparatus is air- tight. Now aspirate more air from a 9 




1 88 Quantitative Chemical Analysis. 

so as to cause a small quantity of sulphuric acid to pass 
over from b. Carbonic acid is immediately evolved, and is 
dried by passing through the sulphuric acid in b. When the 
evolution of gas begins to slacken, draw over in the same 
manner fresh quantities of the acid. The complete decom- 
position of the ore requires from five to ten minutes ; the ter- 
mination of the reaction is indicated by the appearance of 
the residue in #, which will no longer be black, and also from 
the non-evolution of carbonic acid when a fresh quantity of 
sulphuric acid flows over into a. Remove the glass stopper 
from c, and aspirate a slow current of air, dried by sulphuric 
acid, through the apparatus for five minutes. When the 
flask is cold, remove the caoutchouc tubing, re-weigh, again 
aspirate dry air through it, and again weigh. The two 
weighings ought to agree. The loss of weight indicates the 
amount of carbon dioxide evolved by the action of the man- 
ganese dioxide upon the oxalic acid. 

In the case of ores containing carbonates, treat the 
weighed portion in a with a little water, add two or three 
drops of dilute sulphuric acid, and heat on the water-bath. 
In about ten or fifteen minutes test the liquid with a slip of 
blue litmus paper ; if it is not acid, add a few more drops 
of dilute sulphuric acid, and continue the heating. When 
the carbonates are completely decomposed, neutralise with 
caustic soda, free from carbonic acid, allow the liquid to 
cool, add the sodium oxalate, connect the flasks together, 
and proceed as above. 

With proper care this method gives very concordant 
FIG. S4A. results ; two analyses should not differ by 
more than 0-25 per cent. 

The great weight of the apparatus, and the 
surface it exposes, are its chief disadvantages. 
To obviate these sources of error, various 
modified forms of it have been devised ; one 
of the most convenient of these is seen in 
fig- 54A. 




Manganese Ores. 189 

VI. Volumetric Determination by means of Iron and 
Potassium Permanganate Solution. 

About i '5 to 2 grams of iron wire, perfectly free from 
rust, are accurately weighed out, and dissolved in 100 c.c. of 
dilute sulphuric acid (i of acid to 4 of water), in the appa- 
ratus represented in fig. 48, p. 149. About the same 
weight of the finely-powdered ore to be tested is added to 
the solution of iron, and the liquid is heated gently, until 
the whole of the manganese is dissolved. The solution is 
boiled, the water allowed to recede (see p. 148), and the 
contents of the flask made up to 250 c.c. When cold the 
amount of residual ferrous sulphate is determined by per- 
manganate solution : 

2FeSO 4 + MnO 2 + 2H 2 SO 4 = Fe 2 (SO 4 ) 3 + 
MnSO 4 + 2H 2 O. 

112 parts of iron correspond to 86 '04 of manganese dioxide. 
Example. Weighed out 1*562 grams of iron wire. 
1-562 x -996 = 1*556 pure iron. Dissolved in the 
requisite quantity of dilute sulphuric acid, and added 1*285 
gram of manganese dioxide. When cold, the solution 
required 20-2 c.c. of permanganate solution; i c.c. perman- 
ganate = 0*01096 Fe. Accordingly, 20*2 x 0*01096 = 
0*2215 Fe. J '556 0*2215 = z '3345 gram of iron has 
been oxidised by the manganese dioxide. 

112 \ 86*04 I" 1 '3345 x ' x r ' O2 5 
and 1*285 I<02 5 ' I0 79'8o. 

The ore, therefore, contained 79*80 per cent, of manga- 
nese dioxide. 

VII. Volumetric Determination by means of Oxalic Acid 
and Potassium Permanganate. 

Since the amount of oxalic acid in solution can be readily 
determined by means of potassium permanganate (see p. 151), 
the reaction between the manganic oxide, oxalic acid, and 



IQO Quantitative Chemical Analysis. 

sulphuric acid may be made the basis of a volumetric 
method. It is merely necessary to heat the finely-powdered 
ore with an excess of oxalic acid, and on the completion of the 
decomposition to determine the amount remaining in the 
liquid by means of standard potassium permanganate solution. 
About 2 grams of the ore are weighed out into a flask, and 
gently heated with 50 cubic centimetres of normal oxalic acid 
solution, and 5 or 6 cubic centimetres of strong sulphuric acid : 
when the decomposition is at an end (which may be known 
by the absence of any black grains in the sediment), the 
solution is filtered into a J-litre flask, the sediment and filter 
washed, and the liquid diluted to 250 cubic centimetres. 
After shaking, 100 cubic centimetres of the liquid are intro- 
duced into a beaker, and the solution titrated by potassium 
permanganate (see p. 151). The determination is repeated 
with a second portion of 100 cubic centimetres. The mean of 
the two results multiplied by 2-5 gives the amount of residual 
oxalic acid, and this, subtracted from the amount originally 
taken, shows the quantity of acid decomposed by the man- 
ganese. From the equation 

MnO 2 + C 2 O 2 (OH) 2 = 2COi + MnO + H 2 O 

it is seen that 90 parts of oxalic acid correspond to 86-04 of 
manganese dioxide. 

Many samples of manganese ore contain more or less 
ferrous oxide, which becomes oxidised at the expense of a 
portion of the chlorine evolved on treating the mixture with 
hydrochloric acid. The method of Fresenius and Will, and 
the volumetric modification above described, are therefore apt 
to assign too high a value to the manganese, the quality of 
which, as we have already remarked, depends solely upon 
the quantity of available chlorine which it can liberate. 

VIII. Determination of Moisture in Manganese Ores. 

All manganese ores contain variable amounts of moisture, 
the exact determination of which is a point of some 



Manganese Ores. IQI 

importance. By repeated experiments it has been found that 
a temperature of 120 maintained for about an hour and 
a half is sufficient to expel the whole of the hygro- 
scopic moisture, without eliminating any of the true water of 
hydration. A few grams of the powdered oxide are intro- 
duced into the weighed tube, fig. 29, and heated to 120 
for about an hour and a half. The loss of weight gives the 
amount of moisture. If it is preferred to dry the sample at 
1 00 the heat must be maintained for at least six hours 
before a constant weight will be obtained. 

The dried and finely-powdered oxide is exceedingly hy- 
groscopic, and if the sample is dried before being analysed 
there is great risk that in the operations of weighing the 
portion will take up fresh quantities of moisture. It is 
better to keep the powdered sample undried, in a well- 
corked test-tube, and to weigh out portions for the deter- 
mination of the oxide and water at the same time, and to 
conduct the two operations simultaneously. The percentage 
amount of oxide in the sample when dried may be after- 
wards readily calculated. 

IX. Determination of the Amount of Hydrochloric Add 
required to decompose Manganese Ore. 

Two samples of manganese ore may show on analysis the 
same amount of available dioxide that is, may liberate the 
same amount of free chlorine and yet require very different 
amounts of hydrochloric acid to effect their complete de- 
composition. The elimination of a given quantity of chlorine 
by means of hausmannite, for example, requires the expendi- 
ture of twice the amount of hydrochloric acid needed to 
yield the same quantity from the binoxide : 

Mn 3 O 4 + 8HC1 = 3MnCl 2 + C1 2 + 4H 2 O. 
Mn0 2 + 4 HC1 = MnCl 2 + C1 2 + 2 H 2 O. 

Moreover, since most ores of manganese contain carbonates, 
and gangue, decomposable by hydrochloric acid, it is always 



192 Quantitative Chemical Analysis. 

necessary to employ a larger quantity of acid than corresponds 
to the amount of available binoxide found. Accordingly it 
is often required to determine exactly the quantity of acid 
needed to effect the complete decomposition of the ore. 

Dissolve a quantity of recrystallised copper sulphate in 
warm water, and carefully add solution of ammonia, witVi 
constant stirring, until the bluish-green precipitate is very 
nearly dissolved. If the exact point is overstepped, add a 
little more solution of copper sulphate until the precipitate 
just reappears. Filter the solution, and determine its value 
by withdrawing 10 cubic centimetres, and adding standard 
sulphuric acid until the liquid remains permanently turbid. 
Now weigh off about i gram of the powdered manganese 
ore into a small flask fitted with a cork, in which is fixed 
an obtusely -bent tube about 3 feet in length, and add to it 
10 cubic centimetres of moderately-concentrated hydrochloric 
acid (sp. gr. ri), and heat gently ; the bent tube is to 
be so arranged that the condensed water flows back 
into the flask. Whilst the liquid is heating, measure off a 
second quantity of 10 cubic centimetres of the acid, run it 
into a beaker, and add the standardised ammonio-copper 
solution, with .constant stirring, until the liquid is just 
rendered turbid. Heat the contents of the flask more 
strongly for a few minutes to expel the chlorine; let the 
flask cool, add a quantity of cold water to it, throw the solu- 
tion on to a filter, wash, and again titrate with the copper 
solution. The difference expresses the amount of hydro- 
chloric acid required to decompose the ore. As the amount 
of chlorine evolved is known from a determination made by 
one of the preceding methods, the quantity remaining in com- 
bination as manganous chloride, &c., is readily calculated. 
This method, although not absolutely exact, affords results of 
sufficient accuracy for technical purposes. The ordinary acidi- 
metric methods are here inapplicable, since the manganous 
chloride possesses an acid reaction which interferes with the 
process. This solution (known as Kieffer's) is frequently of 



Bleaching Powder. 193 

service in testing liquors containing free acid or salts 
which redden litmus ; in determining, for example, the 
amount of free acid in liquids from galvanic batteries, 
&c. It may be also used in determining the strength of 
vinegars, as the brown colour of the solution in no way 
interferes with the completion of the reaction. It is neces- 
sary from time to time to redetermine the strength of the 
solution by standard acid, as it experiences slight alteration 
on keeping. 

X. BLEACHING POWDER. 

Bleaching powder or chloride of lime is formed by the 
action of chlorine upon calcium hydrate. The relation of 
its constituents in the freshly-prepared substance is repre- 
sented by the formula Ca 3 H6O6.Cl 4 . When allowed to stand 
in contact with air and light, chloride of lime suffers 
decomposition, and, after treatment with water, the calcium 
chloride is found to have increased in quantity, whilst the 
hypochlorite has suffered a corresponding diminution. 
When exposed to moist air containing carbonic acid, 
bleaching powder is decomposed, hypochlorous acid is 
evolved, and calcium carbonate formed. When, there- 
fore, chloride of lime is used as a disinfectant, the active 
agent in ordinary circumstances is hypochlorous acid, and 
not free chlorine, as formerly supposed. At a moderate 
temperature (50) dry chloride of lime is converted into cal- 
cium chlorate, and the mass becomes pasty from the 
separation of water : 

3 Ca 3 H 6 6 Cl 4 = sCaC! 2 + Ca(ClO 3 ) 2 + 3 CaH 2 O 2 + 6H 2 O. 

This change goes on at a diminished rate even in direct 
sunlight. Chloride of lime is decomposed by water, calcium 
hydrate separates out, and calcium chloride and hypo- 
chlorite pass into solution : 

Ca 3 H 6 6 Cl 4 = CaH 2 2 + CaCl 2 + Ca(ClO) 2 + 2 H 2 O. 
It is highly probable that the hypochlorite thus formed 



194 Quantitative Chemical Analysis. 

is only produced by the action of the water, and does not 
exist pre-formed in the bleaching powder. 

Since the value of the commercial article depends entirely 
upon the amount of hypochlorous acid which it can produce, 
and since the circumstances of heat, moisture, air, and light 
exercise such an important influence upon the proper pro- 
duction and stability of the bleaching powder, it is evident 
that, as manufactured and stored, it must -vary very consider- 
ably in quality. The most concentrated preparation which 
can be obtained by saturating calcium hydrate with chlorine, 
contains about 38-5 per cent, of available chlorine, but the 
great bulk of the substance found in commerce rarely contains 
more than from 32 to 37 per cent, of which i or 2 per cent, 
is without bleaching power, being present in the form of cal- 
cium chlorate. In badly-made bleaching powder the amount 
of chlorate present is occasionally equal to 8 or 10 per cent, 
of available chlorine nearly one-fourth of the amount 
which ought to be contained in the product. Many 
methods have been proposed to estimate the available 
chlorine present in bleaching powder, the majority being 
based on the oxidising effect of the hypochlorites, but a 
great number are inaccurate, in that they do not take cogni- 
sance of the presence of this admixed chlorate, which, under 
the circumstances of the valuation-processes, reacts like 
chlorine, although it has no bleaching effect. 

The best and most convenient chlorimetrical methods 
hitherto proposed are those of Penot and Bunsen. 

Penofs Method. This process is based upon the conversion 
of an alkaline arsenite, by the chloride of lime solution, into 
an arseniate : 

As 2 O 3 + Ca(ClO) 2 = As 2 O 5 + CaCl 2 . 

The final point of the reaction is determined by means of 
potassium iodide and starch ; so long as any hypochlorite 
remains undecomposed, a drop of the solution brought into 
contact with potassium iodide and starch renders that mixture 



Bleaching Powder. 195 

blue. This mixture of iodide and starch is conveniently 
employed in the form of test-papers. 3 grams of arrowroot, 
potato, or wheat starch are rubbed into a thin cream with 
50 or 60 cubic centimetres of warm water. Pour the mixture 
into about 200 cubic centimetres of water, and heat the 
liquid, with constant stirring, until it boils : now add i gram 
of potassium iodide and i gram of pure carbonate of soda 
dissolved in a little water, and dilute the mixture to 500 cubic 
centimetres. Moisten a number of strips of Swedish filter- 
paper, or other unsized paper of good quality, with the 
solution, and when dry, preserve them in a wide-mouthed 
stoppered bottle. To prepare the arsenious acid solution, 
powder a quantity of the purest sublimed arsenious acid (free 
from arsenic sulphide), and weigh off exactly 4-95 grams into 
a litre flask, add about 25 grams of recrystallised sodium car- 
bonate (free from sodium sulphide, sulphite, or thiosulphate) 
and 200 cubic centimetres of water. Boil the solution gently, 
and shake it continually until the arsenious acid is dissolved: 
when the solution is cold dilute it exactly to one litre. This 
constitutes a deci-normal solution of arsenious acid : the 
equivalent of As 2 O 3 is 1 98. i eq. can take up 2 atoms of oxygen 
to form As 2 O 5 , or is equivalent to 4 of Cl. Since it is diffi- 
cult to weigh out exactly the required quantity of arsenious 
acid, it is preferable to take a round number, about 5 grams, 
and dilute proportionally. 

Example. 5*013 grams were weighed out into the litre 
flask, 25 grams of sodium carbonate and 200 cubic centi- 
metres of water added : after complete solution and cooling 
the liquid was diluted to i litre, and 127 cubic centimetres 
of water were added by means of a burette ; since 

4-95 : 1000 :: 5*013 : 10127. 

The solution in the flask is well shaken, and decanted off 
into a number of small well-stoppered bottles : this precaution 
diminishes the liability of the solution to change on exposure 
to the air. If the solution is perfectly free from sodium 

o 2 



196 Quantitative Chemical Analysis. 

thiosulphate, or sulphite, sodiuin or arsenic sulphides, there 
is far less chance of it suffering alteration.* 

The sample of bleaching powder to be tested is well 
mixed, and about 10 grams are weighed out into a porcelain 
mortar ; 50 or 60 cubic centimetres of water are added, and 
the mixture is rubbed to a thin cream ; it is allowed to settle 
for a few minutes, and the supernatant liquid (which is still 
turbid) poured into a litre flask. The sediment in the mortar 
is triturated with fresh water, and the operation is repeated 
until the whole of the chloride of lime has been brought 
into the litre flask. Fill up to the mark, and shake. Have 
the burette ready filled to the zero mark, withdraw 50 cubic 
centimetres of the turbid solution, run it into a beaker, and 
add the arsenious acid solution, with constant stirring, until 
a- drop from the beaker, taken out on a glass rod, and brought 
into contact with a strip of the iodised paper moistened 
with water on a white plate, no longer gives a blue stain. 
There is no difficulty in hitting the final point ; the gradually 
increasing faintness in the blue colour of the drops indicates 
with great accuracy the progress of the reaction. In making 
a second determination, care must be taken to shake the 
contents of the litre flask before withdrawing the solution ; 
if this precaution be neglected, the second determination 
will give a much lower result a difference of 2 or 3 cubic 
centimetres being not unfrequently obtained in testing the 
clear and the turbid liquids. 

Example. 10*99 grams of bleaching powder were treated 
as directed, and diluted to i litre. 50 cubic centimetres of 

* When a great number of chlorimetrical estimations have to be 
made it will be found convenient to modify the above method in the 
following manner : The weighed quantity of arsenious acid is dissolved 
by a gentle heat in 10 or 15 c.c. of glycerine, and diluted with water to 
I litre. The weighed sample of bleaching powder is treated with 
water as directed, and a portion of the turbid solution poured into a 
burette. 25 c.c. of the standard arsenious acid solution are delivered 
into a flask, mixed with I cubic centimetre of indigo solution, and the 
bleaching powder solution added, with constant shaking, until the blue 
colour is discharged. 



Bleaching Powder. 197 

the turbid solution required 47*3 cubic centimetres of the 
arsenious acid solution to complete the reaction. Since 
i cubic centimetre of this solution is equivalent to 0*003546 
of chlorine, this would correspond to 47-3 x 0-003546 = 
0^1677 gram in the 50 cubic centimetres of solution. But the 
50 cubic centimetres contain 0-5495 gram of the bleaching 



powder ; hence the substance contains ___ I0 = 

Q'5495 
30-52 per cent, of chlorine. 

The amount of calcium chloride present in a sample of 
bleaching powder may be determined by first estimating the 
hypochlorite in the manner above described, and then 
adding to a second portion of 50 cubic centimetres a slight 
excess of ammonia, and warming. The hypochlorite is thus 
converted into the chloride, with the formation of water and 
nitrogen : 

3Ca(C10) 2 + 4NH 3 = 3 CaCl 2 + 6H 2 O 4- 4 N. 

The chlorine is then determined in the solution, after 
boiling, and cautiously neutralising with nitric acid, by 
means of standard silver solution. In normal bleaching 
powder the amount of available chlorine will be equivalent 
to that existing as calcium chloride. To determine the 
amount of chlorate present, a third portion is heated with 
ammonia, strongly acidified with pure sulphuric acid, and 
digested with metallic zinc. In a few hours the nascent 
hydrogen will have completely reduced the chloric acid to 
the state of hydrochloric acid, and on again precipitating 
the chlorine, the increased amount over the second determi- 
nation shows the quantity existing as chlorate. 

Another method of estimating the total chlorine present 
in bleaching powder, is to boil the turbid solution with a 
solution of ferrous sulphate and potash, whereby the hypo- 
chlorous and chloric acids are reduced to hydrochloric acid : 

HC10 + HC10 3 + 8FeO = 2 HC1 



198 Quantitative Chemical Analysis. 

The solution is filtered, acidified with nitric acid, and the 
chlorine precipitated with silver nitrate. 

In some parts of the Continent, particularly in France, it 
is customary to represent the amount of available chlorine, 
not in percentages, but in chlorimetrical degrees, represent- 
ing the number of litres of chlorine at o, and 760 milli- 
metres which i kilo, of the sample should yield. Thus, if a 
sample is reported to be of 100, it means that i kilo, of it 
would yield 100 litres of chlorine, measured at the standard 
temperature and pressure. Since a litre of chlorine weighs 
3-177 grams, this sample would contain in a kilo. 3177 
grams, or 3177 per cent. Conversely it is easy to see that 
3177 per cent, would be equal to 100, since 3177 per 

cent, is equal to 3177 per mille, and 3 I 1_7 = i o. 

Bunseifs Method (Modified.} Withdraw 20 cubic centi- 
metres of the turbid solution, made in the manner above 
described, and place it in a beaker, add about 15 cubic 
centimetres of potassium iodide solution, acidify with hydro- 
chloric acid, and determine the iodine liberated according 
to the method given on p. 157. The amount of available 
chlorine in the sample is thus measured by the quantity of 
iodine which it can set free. 

XL BLACK-ASH ; SODA- ASH ; VAT- WASTE. 

Black-ash is the product obtained by heating the mixture 
of sodium sulphate (salt-cake), calcium carbonate, and small 
coal or slack in a reverberatory furnace, in the manufacture 
of soda by Leblanc's process. It consists essentially of a 
mixture of carbonate and caustic soda with sulphide and 
carbonate of calcium. In addition it contains small quantities 
of sodium sulphite, thiosulphate (hyposulphite), sulphide, and 
undecomposed sulphate and chloride, together with alumina, 
ferrous sulphide, sand, and unburnt carbon; to the last-named 
substance is mainly due the characteristic colour of the product. 

On lixiviating the fused mass with tepid water, the greater 



lack- Ask, &c. 199 

portion of the sodium compounds pass into solution, and on 
evaporating the clear liquid crude soda-ash is obtained. The 
insoluble matter remaining in the lixiviating tanks, and con- 
sisting mainly of calcium sulphide and carbonate, is termed 
tank- or vat-waste. 

The difference in composition of these various products is 
well seen in the following analyses : 

I. BLACK-ASHGERMAN MAKE (ANALYSED BY FRESENIUS). 
Sodium carbonate . . 31 -982 
Sodium hydrate . . 6-104 
Sodium silicate . . . 1-019 
Soluble in water Sodium aluminate . . i -080 
Sodium sulphide . . 0-133 
Sodium sulphite . . 0-216 
Sodium chloride . . 0*288 

40-822 

C Calcium sulphide * . 
Calcium carbonate 
Lime 

I Ferrous sulphide 
Insoluble in water ( Silica 

I Alumina . 
Soda 
Carbon 
IjSand ... 

59-026 

99-848 

II. REFINED SODA-ASH^-GLASGOW (ANALYSED BY BROWN). 

Sodium carbonate . . . . . .80-92 

Sodium hydrate . . . . . .3-92 

Sodium silicate . . . . . . .1-32 

Sodium aluminate . . . . . I "Oi 

Sodium sulphate . . . . . 7 '43 

Sodium sulphite . . . . . .I'll 

Sodium thiosulphate ...... trace 

Sodium sulphide . . . . . .0*23 

Sodium chloride . . . . . .3-14 

Insoluble matter . . . . . .0-77 

99-85 

* The combination of lime and sulphur in the portion insoluble in 
water has been rearranged in accordance with the views now generally 
held as to the manner in which these bodies are united in the vat-waste. 
In the original analyses the sulphur (excluding that present in the ferrous 
sulphide) was calculated to calcium oxysulphide, 3CaS.CaO. 




2OO Quantitative Chemical Analysis. 

The following scheme gives the method for the complete 
analysis of black-ash. 

About 30 grams of the finely-powdered ash are digested 
with water at a temperature of about 45 in a flask of 500 
c.c. capacity. The solution should be hastened as far as 
possible by repeated shaking ; the insoluble matter is then 
allowed to subside, and in a few hours the clear supernatant 
liquid is poured through a folded filter into a J-litre flask. 
The residue, which should be kept as far as practicable 
in the flask in which it was originally placed, is quickly 
washed with cold water, and the washings added to the 
main bulk of the filtrate. Discontinue the washing when 
the filtrate commences to be turbid, and fill up the 1-litre 
flask to the containing-mark. Close and shake it. If the 
knowledge of the nature of the soluble matter is not im- 
mediately wanted, it is better to proceed at. once with the 
examination of the insoluble portion, since this is apt to 
suffer alteration on standing. 

A. Analysis of Insoluble Matter. 

Wash the precipitate from the filter back again into the 
flask in which the main portion of the residue is contained, 
and without delay attach the flask to the rest of the apparatus 
represented in fig. 55, the several parts of which should be 
already weighed, as indicated below, and put together. A is 
the flask containing the insoluble matter. It is fitted with a 
caoutchouc cork, through which is passed a 100 c.c. pipette 
filled with hydrochloric acid of sp. gr. i *i. To the upper end 
of the pipette is attached a piece of caoutchouc tubing which 
can be closed by a screw clamp. The other end of the tube 
is connected with a soda-lime tube, s ; A is connected with 
the flask B, of 300 c.c. capacity, and containing a cold 
saturated solution of copper acetate free from sulphuric acid. 
Both flasks stand on an iron plate which can be heated by a 
lamp ; B is connected with the two-bulbed U-tube <:, also 
containing copper acetate solution, standing in an empty 



Black- Ash, &c. 



201 



beaker, and fitted with bent tubes in such a manner that 
the gas traversing the apparatus passes twice through the 



FIG. 55. 




same quantity of liquid (fig. 56). The two U-tubes d and 
e are filled with calcium chloride to dry the gas : f is a 
potash-apparatus (of the form known as F 

Geissler's) ; it is filled to the extent indi- 
cated in the figure with solution of potash 
of sp.gr. 1-27 (containing about 30 per 
cent, of potassium hydrate), and is ac- 
curately weighed : g and h are U-tubes 
filled with soda-lime and calcium chloride : 
g only is weighed, its object is to absorb 
the last traces of carbon dioxide ; h is 
placed merely to prevent the absorption 
of atmospheric carbon dioxide and moisture by g : the 
caoutchouc tube at the end leads to the filter-pump or 
other aspirating arrangement. As soon as the apparatus is 
put together, fill the pipette with hydrochloric acid, fit the 




2O2 Quantitative Chemical A natysis. 

cork into A and open the clamp, so as to cause the acid to 
enter the flask. The insoluble matter in A is immediately 
decomposed, t and sulphuretted hydrogen and carbon dioxide 
are simultaneously evolved ; the former is absorbed in B 
and c ; the latter, after being dried by passing through d and 
^, is absorbed by/ and g. When the evolution of the gas 
slackens, add more acid, until the decomposition is complete. 
Heat the iron plate until the contents of A and B are in gentle 
ebullition, pour hot water into the beaker, open the clamp, 
and aspirate a slow current of air (about 5 litres) through the 
apparatus. 

The increase in the weight of/ and g gives the amount of 
carbon dioxide. The copper sulphide is thrown on to a filter, 
and, without washing, the precipitate, together with the filter, 
is transferred to a flask and treated with hydrochloric acid and 
potassium chlorate. The liquid is filtered into a litre flask, 
the filter washed, and the filtrate diluted to the containing- 
mark. Withdraw two portions of 100 c.c., and determine in 
each the sulphuric acid by means of barium chloride. 233-2 
parts of barium sulphate are equivalent to 32 parts of sulphur. 

Pour the contents of the flask A on to a weighed filter, 
receiving the filtrate in a J-litre flask, wash the insoluble 
matter, consisting of sand and carbon, dry at 100 and 
weigh. Ignite the weighed mixture to burn off the carbon, 
and weigh the residual sand. The difference between the 
two weighings gives the carbon. 

Make up the filtrate to 500 c.c., and transfer 200 c.c. to 
a porcelain basin, add a small quantity of nitric acid, and 
evaporate to dryness on the water-bath. Treat the separated 
silica in the usual manner and weigh it. Precipitate the 
iron and alumina from the filtrate by ammonia, weigh 
them together, fuse the mixture with a little acid potassium 
sulphate, and determine the iron volumetrically by potas- 
sium permanganate. The alumina is found by difference. 
Determine the lime, magnesia, and alkalies by the method 
given on p. 184. 



Black- Ash, &c. 203 

B. Analysis of the Soluble Portion. 

1. Withdraw 50 c.c., and determine the total alkali pre- 
sent by standard acid, litmus, and soda, in the manner 
directed on p. 139. Calculate as sodium carbonate. 

2. Transfer 100 c.c. to a J-litre flask, mix with barium 
chloride solution so long as a precipitate forms, fill up the 
flask to the containing-mark, shake, and close it. Determine 
the sodium hydrate in an aliquot portion of the clear liquid 
by standard acid and litmus. 

3. Transfer 50 c.c. to a large beaker, dilute with 200 c.c. 
of water, add acetic acid until the liquid is very nearly 
neutral, and determine the joint amount of the sulphide and 
sulphite of sodium by means of starch paste and standard 
iodine solution. 

4. Transfer 100 c.c. to a i-litre flask,' and add zinc 
sulphate solution made strongly alkaline by potash, until a 
considerable precipitate is formed. This contains the whole 
of the sulphur present as sodium sulphide ; the rest of the 
sulphur remains in the solution. Dilute to the containing- 
mark, shake, allow to settle, and determine the sulphur still 
in solution in an aliquot portion of the clear liquid by means 
of starch and standard iodine solution, after acidifying with 
acetic acid in the manner previously directed. From the 
amount of iodine used, the sulphur dioxide is readily cal- 
culated: 2537 parts of iodine are equivalent to 126*15 of 
sodium sulphite. The difference between this and the pre- 
vious determination of the total amount of iodine required 
gives the quantity of sodium sulphide : 2537 parts of iodine 
correspond to 78*15 of sodium monosulphide. 

5. Evaporate 100 c.c. to dryness in a thin porcelain 
basin with a small quantity of pure nitre, and gently fuse 
the residue. The sodium sulphide and sulphite are thereby 
oxidised to sulphate. Digest the mass with hot water, filter 
the solution into a 250 c.c. flask, wash the insoluble matter, 
and fill up the flask to the mark, and shake. Withdraw 100 



204 Quantitative Chemical Analysis. 

c.c., and determine the sulphuric acid, after acidulating with 
hydrochloric acid, as barium sulphate. Subtract the amount 
corresponding to the sodium sulphide and sulphite ; the 
remainder is calculated to sodium sulphate.* In another 100 
c.c. determine the chlorine by standard silver. 

6. Transfer IOQC.C. to a porcelain dish, acidify with hydro- 
chloric acid, and separate the silica in the usual way. In the 
filtrate determine the alumina by precipitation with am- 
monia. 

Arrangement of the Results. A. Insoluble Portion. Cal- 
culate the iron to ferrous sulphide, and the remainder of the 
sulphur to calcium sulphide, CaS. The rest of the calcium 
is to be set down as lime, CaO. The silica, alumina, and soda 
are set down uncombined, as we have no knowledge of their 
state in the insoluble portion. The sand and carbon are, of 
course, directly determined. 

B. Soluble Portion. Combine the silica and alumina 
with soda, to form sodium silicate and aluminate, Na z SiO 3 , 
and Na 2 Al 2 O 4 ; calculate the amount of sodium carbonate 
equivalent to these compounds, together with that corre- 
sponding to the hydrate and sulphide, and subtract the joint 
amount from the result obtained by determining the total 
alkali with standard acid and soda. The remainder gives 
the real sodium carbonate present in the solution. 

The foregoing scheme of analysis is, of course, equally 
applicable to the analysis of soda-ash and vat-waste. Vat- 

* It is sometimes requisite for the purposes of the manufacturer merely 
to determine the quantity of sodium sulphide in the liquor. This may 
be readily effected by the following process, due to Lestelle. The solu- 
tion is mixed with ammonia, heated to boiling, and titrated with a weak 
standard solution of silver, made by dissolving 4*3475 grams of pure 
silver nitrate in water in a litre flask, adding excess of ammonia, and 
diluting to 1,000 c.c. I c.c. is equivalent to I milligram of sodium sul- 
phide. When the sulphur is nearly all precipitated, the liquid is 
filtered, and the addition of the silver solution continued until, after again 
filtering, only the faintest turbidity is produced. The method is expedi- 
tious, and after a little practice gives accurate results. 



Copper Ores. 205 

waste is treated exactly like the insoluble portion of black- 
ash ; soda-ash like the soluble portion. 

XII. ESTIMATION OF SULPHUR IN PYRITES, BY MEANS 

OF COPPER OXIDE. 

(Particularly applicable to roasted Pyrites.) 

A rapid and sufficiently accurate method for technical 
purposes consists in heating from 5 to 10 grams of the ore, 
according to its richness in sulphur (if more than 10 per 
cent, of sulphur is present, 5 grams, if less than 10 per cent, 
10 grams are taken), intimately mixed with 5 grams of pure 
sodium carbonate, and about 50 grams of dry copper oxide, 
in a porcelain crucible, to a low red heat, for about a quarter 
of an hour, with frequent stirring. The sulphur is oxidised to 
sulphuric acid, and combines with the alkali. When cold 
the mass is treated with water, and the amount of sodium 
carbonate remaining determined by titration with normal 
acid. 

XIII. ASSAY OF COPPER ORES (MANSFELD PROCESS). 

About 5 grms. of the finely-powdered ore are weighed out 
into a flask, and mixed with 40 c.c. of moderately-concen- 
trated hydrochloric acid (sp. gr. n6). 6 c.c. of dilute 
nitric acid (made by mixing equal bulks of water and pure 
acid of sp. gr. 1-2) are added, and the flask is gently heated 
for 30 minutes on a sand-bath, after which it is boiled for 
15 minutes. The whole of the copper is now in solution : 
the extraction is complete, even in the case of very rich 
ores, provided sufficient attention has been paid to the 
powdering. The solution is filtered into a large beaker, into 
which a rod of zinc, weighing about 50 grms., and surrounded 
with a piece of thick platinum foil, has been previously 
placed. It is necessary that the zinc employed should be as 
free as possible from lead. The precipitation of the metallic 
copper commences immediately, and is generally complete 
in about half an hour. The rod of zinc is withdrawn, and 



2o6 Quantitative Chemical A nalysis. 

the precipitated copper repeatedly washed by decantation. 
If the amount of the copper does not exceed 6 per cent, 
(which may be approximately known from the bulk of the 
reduced metal), it is dissolved in 8 c.c. of the dilute nitric 
acid, prepared as above. The beaker is gently warmed, 
and the amount of copper in the liquid titrated by solution 
of potassium cyanide, after previous addition of 10 c.c. of 
ammonia solution, prepared by diluting i vol. of ammonia- 
water (sp. gr. o'93) with 2 vols. of water. When the 
amount of copper in the ore exceeds 6 per cent, the metal 
is dissolved in 16 c.c. of the nitric acid solution, and the 
liquid is washed, into a 100 c.c. flask, diluted to the con- 
taining-mark, shaken, 50 c.c. withdrawn, mixed with- 10 c.c. 
of the dilute ammonia, and titrated with potassium cyanide. 
The experiment may be repeated with the second portion of 
50 c.c. 

When a solution of potassium cyanide is mixed with an 
ammoniacal solution of copper sulphate or nitrate, the azure 
blue colour gradually disappears with the formation of 
copper-ammonium-cyanide, free ammonium cyanide, ammo- 
nium formate, and urea. The reaction is only constant so 
long as the amount of free and combined ammonia present 
is invariable. 

The strength of the solution of the potassium cyanide is 
thus tested : Exactly 5 grams of chemically pure copper, pre- 
pared by the electrotype process, are weighed out into a 
litre flask, and dissolved at a gentle heat in 266-6 c.c. of the 
dilute nitric acid (made in the manner above described). 
On cooling, the solution is diluted to the containing-mark. 
30 c.c. of this solution, containing 0-15 grm. of metallic 
copper, are placed in a beaker and mixed with 10 c.c. of 
the dilute ammonia liquid, and the solution of potassium 
cyanide is added from a burette, with constant stirring, until 
the blue colour of the liquid just disappears. The strength 
of the cyanide should be so arranged that i c.c. of the solu- 
tion is equivalent to 5 milligrams of copper. The titration 



Copper Ores. 207 

of the solution of the sample of ore is made in exactly the 
same manner. If exactly 5 grams have been taken, and the 
cyanide is of the above strength, each cubic centimetre of 
the solution required for decolourisation is equivalent to o'i 
per cent, of copper. The number of cubic centimetres 
needed, divided by 10, gives the percentage of copper at 
once. 

This method is very expeditious, and if due care be exer- 
cised, it is very accurate. It must be borne in mind that it 
is strictly comparative, and the titrations must be made 
therefore in a uniform manner. The presence of a 'very- 
small quantity of lead exercises no influence on the results," 
but the action of zinc is more injurious. Care must be taken 
therefore to wash the precipitated copper thoroughly before 
dissolving it in the dilute nitric acid. The solutions must 
be quite cold before titration, since less potassium cyanide 
is needed to decolourise a solution when warm than when 
cold. Thus while 30 c.c. of copper solution, containing 0*15 
grm. copper, and 10 c.c. normal ammonia solution, required 
at the ordinary temperature exactly 30 c.c. of potassium 
cyanide solution, the same quantities at about 45 required 
only 28-9 c.c. (STEINBECK). 

The solution of potassium cyanide requires to be titrated 
from time to time, since its strength is not invariable. 

XIV. ASSAY OF COPPER ORES (LUCKOW'S PROCESS).* 

This process, which is now largely used in many German 
establishments, depends upon the fact that copper is pre- 
cipitated in the metallic state from acid solutions by a weak 
galvanic current. The operations required by this method 
may be best described under the following heads : i. Roast- 
ing the ore ; 2. Solution of the roasted ore ; 3. Precipitation 
of the copper ; 4. Weighing the copper. 

i. Roasting the Ore. This operation is only necessary 
when the ores are bituminous. Weigh out about 2 grams of 

* Zeitsch. fur Anal. Chemie, Fresenius, 1869, p. i. 



208 Quantitative Chemical Analysis. 

the finely-powdered sample into a thin porcelain crucible, 
and heat it over a Bunsen flame for 10 minutes, occasionally 
stirring it with a thick platinum wire, so as to expose fresh 
surfaces to the oxidising action of the air. The bituminous 
matter and the greater portion of the sulphur will be expelled 
at the expiration of this time. 

2. Solution of the Ore. The roasted powder is transferred 
to a small flat-bottomed beaker, about 5 centimetres in 
height and 3 centimetres wide, and treated with 6 c.c. of 
nitric acid of sp. gr. 1*2, 4 c.c. of a dilute sulphuric acid 
(prepared by mixing equal volumes of the strong acid and 
water), and 25 drops of hydrochloric acid. The addition of 
the sulphuric acid increases the oxidising action of the nitric 
acid, and converts any lime which may be present into the 
difficultly-soluble calcium sulphate. The liquid is evaporated 
to complete dryness on a sand-bath, the beaker being 
meanwhile covered with a funnel, the stem of which has been 
cut off. This operation requires about an hour. The 
addition of the hydrochloric acid facilitates the evaporation 
and decreases the tendency of the liquid to spirt. When 
dry, break up the mass with a glass rod. 

3. Precipitation of the Copper. Wash the cover inside 
and out with dilute nitric acid (i vol. of acid of sp. gr. 1-2 
diluted with 6 vols. of water), and also the sides of the 
beaker, until it is about half-filled with the dilute acid. Add 
a few drops of a strong solution of tartaric acid (which is best 
preserved in a beaker simply covered with a piece of paper) 
to the liquid, and place the spiral, represented in fig. 56^7, 
within the beaker. This spiral consists of a piece of 
platinum wire about 18 centimetres long, and i millimetre 
thick ; two-thirds of it are so bent in circles that the straight 
portion of the wire projects as if it were the axis of the spiral. 
The outer convolution is so large that it just touches the 
sides of the beaker : the vertical portion of the wire is there- 
fore exactly in its centre. If the evaporation has been 
carefully attended to, the acid liquid remains quite clear : 



Copper Ores. 



209 



should it be turbid, add i c.c. of a concentrated solution of 
barium nitrate, and agitate the mixture by the aid of the 
platinum spiral. A piece of stout platinum foil, to which a 
thick platinum wire has previously been attached, is then bent 
into a cylinder of 3 centimetres long and 18 millimetres in 
diameter (fig. 56 &); this is accurately weighed, and supported 
in the beaker about i millimetre from the spirals : the ver- 
tical portion of the spiral becomes therefore the axis of this 



FIG. 56 a. 



FIG. 560. 



FIG. 56 c. 




cylinder. The wire supporting the foil is fixed by means of a 
screw, a, to the arm a b of the stand (fig. 56 c\ the other screw, 
, holds the wire leading from the zinc pole of a constant 
battery. A small screw clamp, c, is fastened to the end of 
the platinum spiral, and connects the arrangement with the 
other pole of the battery. Immediately the circuit is closed, 
copper commences to be deposited on the platinum foil, and 
bubbles of oxygen are given off from the spiral : in about 8 
hours the whole of the metal will certainly be precipitated, 

p 



2io Quantitative Chemical Analysis. 

even from rich ores and by a weak current. A stream of 
water is run into the beaker so as to displace the acid liquid, 
which is allowed to flow over the sides. The platinum 
cylinder is then withdrawn and disconnected from the stand, 
washed with hot water, and then with a few drops of alcohol. 
It is heated in the steam-bath, and weighed when cold. Its 
increase in weight gives the quantity of metallic copper 
present in the sample. As the process of precipitation 
needs no superintendence, it may be allowed to go on 
during the night : if the operation be commenced in the 
evening, the reduced copper will be ready for washing 
and weighing in the morning. The copper should show its 
characteristic red colour, and be free from any saline deposit : 
the absence of this deposit is evidence of the perfect re- 
moval of copper from the liquid. 

XV. ASSAY OF COPPER ORES BY PRECIPITATING THE 

METAL BY MEANS OF ZINC. 

(See p. 105.) 

About 5 grams of the finely-powdered ore are gently 
heated with strong aqua regia in a deep porcelain crucible, 
covered with a watch-glass. The glass is rinsed with water, 
and the liquid is evaporated to complete dryness. The 
dried residue is heated, to expel the unoxidised sulphur ; if 
the ore contains much pyrites, it will be necessary to treat it 
again with strong nitric acid (B.P. 86},and evaporate a second 
time to dryness, and roast. 

Treat the dried mass with hot water to extract the copper 
(now present as sulphate), filter into a weighed platinum 
dish, wash the insoluble residue, adding the washings to the 
solution already in the dish, and precipitate the copper by 
means of zinc, in the manner directed on p. 105. To prove 
the accuracy of this method, i gram of pure metallic copper 
was mixed with 0*5 gram of the following substances, either 
in the metallic state, or as salts, viz. : Gold, silver, platinum, 
tin, lead, iron, zinc, nickel,. 'cobalt, bismuth, arsenic, ura- 



Copper Pyrites. 2 1 1 

mum, mercury, molybdenum, antimony, sulphur, silica, and 
calcium phosphate. The mixture was treated in the manner 
described, and the copper precipitated by zinc ; the amount 
of reduced copper was 0-996 gram. In more than twenty 
determinations of copper in various combinations, the 
average amount of the metal obtained by this method was 
997 per cent, of the actual quantity present. (MoHR.) 

XVI. COPPER PYRITES. 

This mineral constitutes the most abundant ore of copper : 
it consists essentially of sulphur, iron, and copper, and 
when pure contains 34*8 per cent, of copper. It is almost 
invariably mixed, however, with more or less antimony, 
arsenic, bismuth, lead, manganese, zinc, nickel, and cobalt 
in addition to considerable quantities of silicious substances 
and gangue. When a complete quantitative analysis of the 
ore is to be made, it must always be preceded by a careful 
qualitative examination. 

Pulverise about 20 grams of the mineral in an agate 
mortar, and dry at 100. 

Determination of the Sulphur. About i gram of the ore 
is weighed out into a porcelain dish, a few crystals of potas- 
sium chlorate are added, together with about 50 cubic centi- 
metres of pure nitric acid (sp. gr. 1*35), and the mixture is 
heated on the water-bath. To prevent loss by spirting, the 
dish should be covered with a large watch-glass or funnel. 
Fig. 57. As the evolution of chlorine diminishes, add oc- 
casionally a crystal of potassium chlorate. In about an hour 
the whole of the sulphur will be oxidised ; the cover is rinsed 
with hot water, and the liquid -in the dish concentrated to a 
small bulk, a small quantity of strong hydrochloric acid is 
added, and the solution is evaporated to perfect dryness, in 
order to render the silica insoluble. Moisten the dried re- 
sidue with strong hydrochloric acid, add hot water, filter, and 
wash the insoluble portion by decantation. This consists 

P 2 



? 1 2 



Quantitative Chemical A nalysis. 



FIG 



mainly of silica and imdecomposed gangue ; it may contain, 
however, a small quantity of lead sulphate, left undissolved 
by the hydrochloric acid. The last traces of the sulphate 
may be removed by heating the mass 
with a solution of ammonium acetate 
(made by mixing solutions of am- 
monia and acetic acid), and adding 
the liquid to the main quantity of the 
filtrate. Add to the solution a few 
crystals of tartaric acid, to prevent the 
precipitation of traces of iron, heat 
the liquid to boiling, and add excess 
of barium chloride. Allow the liquid 
to stand ; filter, wash the precipitate 
by decantation with hot water, and 
digest it with a solution of ammonium 
acetate, which removes any traces of 
barium nitrate, which the precipitate 
is very apt to retain. The barium 
sulphate is then washed on to the filter, dried, and weighed. 

Determination of the Copper, Iron, &>c. 2 or 3 grams of 
the powdered ore are oxidised with fuming nitric acid in a 
dish, a few cubic centimetres of strong sulphuric acid are 
added, and the liquid evaporated to dryness. The residue is 
dissolved in hydrochloric acid, water added, the liquid 
allowed to stand, and then filtered into a flask. The insolu- 
ble portion consists mainly of silica and gangue, but it may 
still retain a small quantity of lead as sulphate ; this is to be re- 
moved by repeatedly boiling the residue in the dish with dilute 
hydrochloric acid, and passing the liquid through the filter. 

The residue is then transferred to the filter, washed with 
hot water, dried, and weighed. The filtrate is warmed to 
about 70, and a current of sulphuretted hydrogen passed 
through it. The copper, lead, bismuth, tin, arsenic, and 
antimony are precipitated. Allow the liquor to stand, pour 




Copper Pyrites. 213 

the supernatant liquid through a filter, and wash by de- 
cantation with water containing a little sulphuretted hydro- 
gen. The small portion of the sulphide adhering to the 
filter is washed back into the flask, and the precipitate is 
gently heated with a moderately concentrated solution of 
potassium sulphide ; after a short digestion (thirty or forty 
minutes) add water (otherwise a little copper may remain 
in solution), and again filter. The sulphides of arsenic and 
antimony, mixed with sulphur, obtained by adding hydro- 
chloric acid to the filtrate, are filtered off, and oxidised by red 
fuming nitric acid (B.P. 86), the solution concentrated, an 
excess of sodium carbonate added, and the whole evapo- 
rated to complete dryness, and fused in a silver dish. The 
fused mass is further treated as in No. XV. Part II. 

The sulphide of copper containing the small quantities of 
lead and bismuth, is dried and transferred to a small porce- 
lain basin, and dissolved in nitric acid. The solution is 
evaporated to a small bulk, chloride of ammonium added to 
dissolve the bismuth, and then dilute sulphuric acid. Allow 
the precipitate to settle completely, pour off the clear liquid 
through a filter, quickly wash twice or three times with a 
little water containing a drop or two of sulphuric acid, rinse 
the lead. sulphate on to the filter by means of alcohol, and 
wash the filter paper thoroughly with alcohol. Do not mix 
the alcoholic washings with the main quantity of the filtrate. 
The bismuth is best precipitated as carbonate. Nearly 
neutralise the filtrate containing the copper, and add excess 
of ammonium carbonate, gently warm, filter, dissolve the 
precipitate, which consists partly of bismuth carbonate, 
partly of basic bismuth sulphate mixed with copper, in a 
few drops of nitric acid, and again add ammonium carbo- 
nate, which reprecipitates the bismuth as pure carbonate free 
from copper, and heat gently for some time ; wash, dry, and 
ignite the precipitate, and weigh it as bismuth trioxide, 
taking care to detach the dried carbonate as completely as 
possible from the paper before incineration. Boil the solu- 



214 Quantitative Chemical A nalysis. 

tion containing the copper, add caustic soda, and continue 
the boiling until the solution is free from ammonia, filter, 
wash, dry, and ignite. 

The filtrate from the original precipitate by sulphuretted 
hydrogen contains the iron, zinc, nickel, cobalt, and man- 
ganese. It is concentrated slightly, mixed with nitric acid, 
and boiled until the iron is peroxidised, allowed to cool, 
mixed with a large quantity of strong solution of am- 
monium chloride, and then, drop by drop, with ammonium 
carbonate, until the fluid is just turbid (it must not show the 
least trace of distinct precipitate). Heat to boiling, and 
maintain the fluid in ebullition until the carbonic acid has 
been expelled. Allow the precipitate to settle, add one 
drop of ammonia to the clear liquid, and, if no precipitation 
ensues, a few cubic centimetres of ammonia : filter, and wash 
the precipitate with water containing a little ammonium 
chloride. Dry, ignite, and weigh the ferric oxide. The 
filtrate contains the manganese, nickel, cobalt, and zinc. 
Concentrate to a small bulk, and then add a strong solution 
of sodium acetate and acetic acid until the fluid is distinctly 
acid, heat to boiling, and whilst hot pass a rapid current of 
sulphuretted hydrogen into the liquid. The zinc, nickel, and 
cobalt are precipitated ; the manganese remains in solution. 
The precipitate is thrown upon a filter and washed with 
water containing sulphuretted hydrogen. 

The manganese in the filtrate is determined by boiling 
the liquid to expel the sulphuretted hydrogen, adding caustic 
soda until it is nearly neutral, and then a few drops of 
bromine, and wanning until the manganese separates out as 
binoxide. This is filtered, well washed, and converted into 
protosesquioxide (Mn 3 O 4 ) by ignition. The filtrate will 
contain any lime and magnesia derived from the gangue ; 
these earths may be separated in the usual manner by am- 
monium oxalate and sodium phosphate. 

The mixed sulphides of zinc, nickel, and cobalt on the 
filter are dried, transferred to a small beaker, the filter burned, 



Iron Pyrites, &c. 215 

and the ash added, and the whole dissolved in dilute aqua 
regia ; solution of caustic potash is added until the solution 
is slightly alkaline, then acetic acid until it is distinctly acid, 
and lastly a strong solution of potassium nitrite. Allow the 
solution to stand for at least 24 hours, take out a small por- 
tion of the clear liquid, and add to it a few more drops of 
potassium nitrite ; if after the lapse of a couple of hours 
no further precipitation ensues, the separation of the cobalt 
is complete. Pass both portions of the solution through 
a small filter, wash and dry the yellow crystalline potassium- 
cobalt nitrite, transfer it to a small crucible, incinerate the 
filter, add the ash, moisten the whole with strong sulphuric 
acid, expel the excess by heat, and weigh the residue ; it has 
the composition 2CoSO 4 + 3K 2 SO 4 , and contains 18*0 per 
cent, of cobalt oxide. 

The filtrate containing the zinc and nickel is boiled, car- 
bonate of soda added, and the metals separated as in 
No. XIV. Part II.* 

XVII. IRON PYRITES 
may be analysed by the methods adopted for copper pyrites. 

XVIII. c KUPFERNICKELSTEIN ' 

is a mixture of sulphides of copper and nickel, containing 
arsenic, iron, cobalt, lead, &c. It is obtained as an inter- 
mediate product in the preparation of copper-nickel and 
German silver. It may be analysed by the processes described 
under copper pyrites. 

XIX. IRON-ORES. 

The ores of iron more commonly used for the extraction 
of the metal are the magnetic oxide, red and brown haematite, 
specular ore, spathic ore, and clay iron-stone. The following 
analyses show the characteristic features of these varieties. 

* If the copper alone is to be determined in the ore, the above pro- 
cess is recommended whenever an accurate estimation is required. 



216 



Quantitative Chemical Analysis. 





i 


2 


3 


4 


5 





Ferric oxide 


70-23 


94-23 


90-05 


2'7S 


2-72 


0-40 


Ferrous oxide . 


29-65 







48-I2 


40-77 


45-86 


Manganous oxide 




0-23 


0-88 


0-83 




o-q6 


Alumina . 




0-63 


0-14 


I "63 


5-86 


Lime 




0-05 


0-06 


i-75 


0-90 ! 1-37 


Magnesia . 




trace 


O'2O 


2-29 


0-72 1-85 


Silica 




4-90 


0-92 


1-62 


lo-io 10-88 


Carbonic acid . 









39-92 


26-41 


31 -02 


Phosphoric acid. 




trace 


0-09 


'54 




O'2I 


Sulphuric acid 
Iron Pyrites 




0-09 
0-03 


> traces 


0-22 





trace 
o-io 


Water . ] 




0-56 


9-22 


"45 


I'OO 


i -08 


Organic matter 










0-39 


17-38 


0-90 




99-88 


100-72 


100-76 


100-51 


100-00 


100-29 



1. Magnetic ore. Dannemora. 

2. Red Haematite. Ulverstone. 

3. Brown Haematite. Dean Forest. 



4. Spathic ore. Westphalia. 

5. Blackband ore. Scotland. 

6. Clay iron-stone. Dudley. 



General Method for the Complete Analysis of Iron Ores. 
The ore is carefully sampled, and an average portion of it is 
reduced to a moderately fine powder, dried under the desic- 
cator or at 1 00, according to circumstances, and kept in a 
well- corked tube or bottle. 

Determination of the Moisture. In ores containing car- 
bonic acid, and organic matter, the water can only be 
estimated by direct weighing. 2 or 3 grams of the powdered 
ore are introduced into the bulb-tube (fig. 53), and ignited 
in a slow current of air dried by the chloride of calcium tube: 
the water is condensed in the weighed tube containing 
calcium chloride. The rate at which the air passes through 
the apparatus is seen in the small flask containing sulphuric 
acid. After an hour's gentle ignition the tube is re-weighed ; 
its increase in weight shows the amount of moisture present 
in the sample. 

Determination of the Carbonic Add. This is best effected 
by means of the apparatus represented in fig. 31, p. 86. 



Iron Ores. 217 

Weigh out i or 2 grams of the ore into the flask A, and pro- 
ceed as directed (see No. V. Part II.). 

Determination of the Silica, Iron, Manganese, Sulphur, 
Phosphorus, Alumina, Lime, Magnesia, &>c. Introduce 
about 8 or 10 grams of the finely-powdered ore into a porcelain 
basin, and gently heat with concentrated hydrochloric acid, 
mixed with a little nitric acid, until the mineral is completely 
decomposed. Some varieties of ore for example, blackband 
iron-stone contain such an amount of organic matter that it 
is difficult to determine when they are completely dissolved. 
As a rule not more than 30 or 40 minutes will be required for 
complete decomposition. Some specimens of magnetic iron- 
stone and micaceous haematites dissolve with great slowness 
even in concentrated hydrochloric acid. In such cases it is 
better to heat the weighed portion of the finely-divided ore in 
a current of hydrogen or coal gas until water ceases to be 
evolved : the reduced iron will then readily dissolve in the 
acid. Fusion with acid sulphate of potassium also effects the 
decomposition of such ores. The solution is evaporated to 
dryness, and the residue drenched with strong hydrochloric 
acid, the liquid warmed, diluted with hot water, allowed to 
settle, and filtered into a -J-litre flask. The residue in the dish 
is again heated with a small quantity of hydrochloric acid, 
diluted with water, allowed to settle, and the clear liquid 
poured through the filter. This process is repeated until the 
silicious residue appears quite white and free from iron. It is 
then thrown on to the filter and washed thoroughly with hot 
water, dried, ignited, and weighed. It consists of gangue and 
separated silica. The amount of silica may be determined by 
boiling the weighed residue with a solution of sodium carbo- 
nate in a platinum dish, filtering and determining the weight 
of the gangue remaining : the difference gives the quantity of 
silica. It is sometimes required to determine the nature of the 
gangue : this is accomplished by the methods given in Nos. 
XI. & XII. Part II. Titanic acid is not an unfrequent 



2 1 8 Quantitative Chemical A nalysis. 

constituent of iron-ores : when present it will be found partly 
in the silica separated by evaporation to dryness, partly in the 
hydrochloric acid solution. In order to determine its amount 
a large quantity of the ore is decomposed by hydrochloric 
acid, in the manner above described, evaporated to dryness, 
drenched with hydrochloric acid, and the residue filtered off 
and washed. The silicious matter is transferred to a plati- 
num dish and repeatedly treated, in a * draught-place ' or in the 
open air, with hydrofluoric acid, and a little sulphuric acid;* 
the residue is fused with potassium-hydrogen sulphate, dis- 
solved in a little cold water, and filtered, if necessary. Am- 
monia is added in slight excess, the solution boiled for some 
time, filtered, and the precipitated titanic acid washed, dried, 
ignited, and weighed. To the hydrochloric acid solution a 
few cubic centimetres of nitric acid are added, and the liquid 
is boiled, ammonia added, and the liquid again boiled. The 
precipitated oxide of iron and alumina carries down the rest 
of the titanic acid : these are filtered off, washed, dried, and 
transferred to a platinum dish, and fused with potassium-hy- 
drogen sulphate. The mass is dissolved in a large quantity 
of cold water, neutralised with sodium carbonate, and solution 
of sodium thiosulphate added in slight excess. When the 
solution becomes nearly colourless, boil until all sulphur' 
dioxide is expelled, filter, wash the precipitate with hot 
water, dry, and ignite it gently in a porcelain crucible. The 
residue contains all the titanic acid mixed with alumina. It 
is treated with strong sulphuric acid, filtered if necessary, 
diluted and boiled for some time, and the separated titanic 
acid filtered off and weighed. 

The filtrate from the silica, &c,, separated from the 10 
grams of iron, is diluted to 500 cubic centimetres and well 
mixed by shaking. 

Determination of the Sulphur. Take out 100 cubic centi- 

* If the sulphuric acid be omitted, a portion of the titanium will be 
lost by volatilisation as fluoride. 



Iron Ores. 219 

metres of the solution and evaporate nearly to dryness to 
expel the greater portion of the free acid, dilute with about 
200 cubic centimetres of water, boil, and add one or two 
drops of barium chloride solution. After standing about 
24 hours the barium sulphate is filtered off and weighed. 

Determination of the Phosphoric and Arsenic Acids. To 
TOO cubic centimetres of the solution, add a few cubic centi- 
metres of a clear solution of molybdate of ammonium in 
nitric acid. [This solution is prepared by dissolving 10 
grams of powdered ammonium molybdate in 40 c.c. of dilute 
ammonia (sp. gr. 0*96), and mixing the solution with i6oc.c. 
of dilute nitric acid (120 cubic centimetres of strong acid to 
40 cubic centimetres of water). The mixture is heated to 
about 40 for some hours, and the clear liquid drawn off.] 
The mixture of iron salt and molybdate is kept in a warm 
place (not above 40) for 24 hours, filtered, and washed, the 
precipitate treated with ammonia on the filter, and magnesia- 
mixture added to precipitate the dissolved phosphoric acid. 
After standing a few hours the magnesium ammonium phos- 
phate is filtered, and treated in the usual manner. Many iron- 
ores contain notable quantities of arsenic, which is precipi- 
tated from the solution on adding molybdic acid. If quali- 
tative analysis has shown the presence of arsenic, it must be 
removed before precipitating the phosphoric acid by transmit- 
ting a current of sulphuretted hydrogen through the 100 cubic 
centimetres of iron solution, filtering, heating the filtrate, if 
turbid, with a little nitric acid, and then adding the molybdic 
acid solution. The arsenic in the precipitate does not 
represent the entire amount in the ore, since a consider- 
able portion is lost on the evaporation of the solution to 
dryness in order to render the silica insoluble. If it be 
desired to determine the amount of arsenic actually present, 
a larger quantity of the ore must be heated with aqua regia, 
filtered, and treated with sulphuretted hydrogen, and the 
arsenic determined in the precipitate (see No, XVIII, 



22O Quantitative Chemical Analysis. 

Part II.) Any black residue left after oxidation may con- 
sist of copper oxide. 

Determination of the Manganese, Alumina, Lime and Mag- 
nesia, Potash and Soda. 100 cubic centimetres of the solution 
are boiled with a little nitric acid, acid carbonate of ammonia 
added until the fluid is nearly neutral, and then to the dear 
red liquid, ammonium acetate in excess : boil for some time, 
until the precipitate settles on removing the lamp. Filter 
into a flask, wash with water containing a little ammonium 
acetate, dry, ignite, and weigh. The precipitate consists 
of ferric oxide, alumina, and phosphoric acid, and contains 
traces of silicic acid left in solution. The weighed substance 
is fused with acid-sulphate of potassium, the fused mass 
treated with hot water, and the silica filtered off and weighed. 
The amount of alumina is determined by subtracting the 
total quantity of ferric oxide, phosphoric acid, and silica from 
the original weight of the ignited precipitate. If it is re- 
quired to determine the alumina directly, add tartaric acid 
to the solution, then ammonium chloride, and ammonium 
sulphide. The precipitate after standing a few hours is 
filtered off, and washed with water containing ammonium 
sulphide. Add sodium carbonate and nitre to the filtrate, 
evaporate to dryness, ignite, dissolve in dilute hydrochloric 
acid, filter if necessary, and precipitate the alumina by the 
addition of ammonium chloride and ammonia in slight 
excess, boil until the ammonia is expelled, and wash 
thoroughly with hot water. From the weight of this precipi- 
tate the amount of phosphoric acid is to be subtracted, 
since it is precipitated in the process together with the 
alumina. 

The filtrate from the basic acetates contains the manganese, 
alkalies, and alkaline earths. A few drops of bromine are 
added, the solution is heated to 40 or 50, and the flask 
tightly corked. In a few hours the whole of the man- 
ganese separates out as binoxide : it is filtered off, dried, 



Iron Ores. 221 

ignited, and weighed as Mn 3 O 4 . Concentrate the filtrate, 
add ammonia and ammonium oxalate, and convert the pre- 
cipitate into lime by ignition. The filtrate contains the 
magnesia and alkalies mixed with a large quantity of 
ammonium salts. It is evaporated to dryness, ignited to 
expel the ammoniacal salts, treated with a small quantity of 
water, about i gram of oxalic acid added, and the solution 
again evaporated to dryness and ignited. In this process 
the alkalies are left as carbonates, and the magnesium chloride 
is converted into magnesia. The residue is treated with a little 
water, and the magnesia filtered off, dried, and weighed. The 
carbonates in solution are converted into chlorides, weighed, 
and separated as in No. IV. Part II. If great accuracy is 
required, the nitrate containing the excess of platinum is 
reduced by means of hydrogen, filtered from the precipitated 
metal, and the small quantity of magnesia usually present in 
solution precipitated by the addition of ammonia and a drop 
or two of sodium phosphate. (See Analyses of Plant-ashes.) 

Determination of the Iron by Solution of Potassium Bichro- 
mate. When a solution of potassium bichromate is added 
to a liquid containing a ferrous salt and free hydrochloric 
acid, the iron is converted into a ferric salt in accordance 
with the reaction : 

6FeCl 2 + K 2 Cr 2 7 + i 4 HCl = 3Fe 2 Cl 6 + 2KC1 + Cr 2 Cl 6 

+ 7H 2 0. 

The final point of the reaction that is, the point at which the 
whole of the iron is converted into ferric chloride is ascer- 
tained by bringing a drop of the solution in contact with 
potassium ferricyanide, when no blue colouration will be 
produced. So long as the faintest trace of ferrous chloride 
remains, a drop of the liquid will colour the potassium ferri- 
cyanide. It is evident from the equation that i eq. or 
294-42 parts of the bichromate will convert 6 eq. or 336 
parts of iron to the state of a ferric salt. By dissolving 
4-907 grams of pure dry potassium bichromate in a litre of 



222 Quantitative Chemical Analysis. 

water, a solution is obtained of which i cubic centimetre is 
equivalent to "0056 gram of iron. 

The potassium bichromate is purified by recrystallisation, 
dried between blotting paper, fused, and roughly powdered. 
About 5 grams of the salt are then weighed out into a litre 
flask, and diluted with so much water that each cubic centi- 
metre contains 0*004907 gram of the bichromate. Suppos- 

FIG. 58. 




ing that 5*073 grams have been weighed out, this would 
require 1033-8 cubic centimetres, since 

4-907 : 1000 :: 5-073 : 1033-8. 

The litre flask containing the salt is rilled up to the con- 
taining-mark, and the 33-8 cubic centimetres added from a 
burette. If there is not sufficient space between the mark 



Iron-Ores. 



223 



and stopper of the flask, the 33*8 cubic centimetres are poured 
into the dry bottle in which the solution is to be preserved, 
and the 1,000 cubic centimetres of bichromate solution added; 
the liquid is well shaken, the litre flask refilled with it, and 
the liquid poured back : if this process is repeated two or 
three times the solution will be of uniform strength. It is 
advisable, however, to test the strength of the solution by 
direct experiment. For this purpose about 0*2 gram of fine 
pianoforte wire is weighed out with the greatest care, and dis- 
solved in a few cubic centimetres of pure hydrochloric acid, 
in the flask a represented in fig, 58. During the solution a 
slow current of carbonic acid is passed through the apparatus; 
the exit tube e is furnished with a little valve, made by 
cutting a short slit in a small piece of caoutchouc tubing, 
slipping the one end over the tube and stopping the other 
by means of a glass rod.* This little valve opens by inward 
pressure only ; as soon as the pressure is applied outwardly, 
the sides of the slit are pressed together and effectually 
prevent the entrance of air. In this manner any chance of 
the ferrous solution within the flask oxidising is prevented. 
Whilst the iron is dissolving, make a solution of potassium 
ferricyanide by dissolving a minute portion of the salt in 
about 15 cubic centimetres of water in a test-glass. The 
solution of the ferricyanide should be very dilute, other- 
wise it gives a reddish precipitate towards the completion of 
the assay. Spread a few small drops over a plate or porce- 
lain slab by means of a glass rod, and fill up the burette with 
the solution of bichromate. When all the iron is dissolved, 
the solution is boiled for a minute, and mixed with a quantity 
of cold water, and the bichromate is added to it with 
constant stirring. The solution in the flask changes rapidly 

* It is convenient, when a number of such estimations have to be 
made, to determine the weight of a certain length of the piano-wire, and 
to cut off a portion .when needed equivalent to 0-2 gram. The amount 
taken must of course be controlled by weighing the wire. A milli- 
metre scale fastened on to the balance-table will be found very useful 
for this and similar purposes. 



224 Quantitative CJiemical Analysis. 

in colour and becomes dark green. A brown colour in- 
dicates that a deficiency of hydrochloric acid is present. 
From time to time a drop of the solution is brought from 
the flask in contact with the ferricyanide upon the slab. 
When the intensity of the blue produced begins to 
diminish, the bichromate solution must be more slowly 
added. The mixed drops soon acquire a greenish tint, and 
when the last trace of this colour disappears the reaction is 
finished. A slight correction requires to be made on the 
quantity of iron taken, on account of the impurity present in 
it. If pianoforte wire be used, it may be assumed, without 
sensible error, that it contains 997 per cent, of iron.* 

Eocample. A solution of bichromate was made in accord- 
ance with the directions given: 5*073 grams of the re- 
crystallised and fused salt being dissolved in 1033*8 cubic 
centimetres of water. 0*2097 gram of piano- wire was 
then dissolved in hydrochloric acid. This solution required 
38*0 cubic centimetres of bichromate solution before the 
blue colour disappeared : 38*0 cubic centimetres bi- 
chromate are therefore equal to (0-2097 x '997) = '2091 
gram of iron, or i cubic centimetre of bichromate is equiva- 
lent to 0*0055 gram of iron. 

25 cubic centimetres of the solution of the iron ore to be 
tested are measured off into the flask, a few cubic centi- 
metres of hydrochloric acid added, and two or three small 
pieces of pure zinc. The flask is connected with the carbonic 
acid apparatus, and the evolution of the hydrogen promoted 
by a gentle heat. When the solution has become nearly 
colourless, or possesses only a faint tinge of green, a minute 
drop is withdrawn and tested with potassium thiocyanate, 
water is added, and the bichromate solution is poured in 
until the iron is completely converted into ferric chloride. 

* The bichromate solution may also be standardised by means of 
ferrous sulphate precipitated by alcohol (p. 150) : the salt thus purified 
may be preserved in a stoppered bottle without experiencing the least 
change. 



Titaniferous Iron-Ore. 225 

The experiment is repeated with a second portion of 25 
cubic centimetres of the original solution. 

In ores containing a mixture of ferrous and ferric oxides, 
as magnetic or spathose iron-stones, the amount of the 
former oxide is determined by dissolving from i to 3 grams 
of the sample, according to its supposed richness, in hydro- 
chloric acid, in a stream of carbonic acid, adding water, and 
titrating with bichromate solution according to the method 
just given. 

XX. TITANIFEROUS IRON-ORE (ILMENITE). 

Ilmenite is a naturally occurring ferrous titanate (FeTiO 3 ), 
but it is seldom met with quite pure ; it usually contains 
ferric oxide, manganous oxide, magnesia, alumina, silica, &c. 
Many iron-sands contain considerable quantities of ilmenite, 
associated with magnetic oxide of iron. As such ores are 
occasionally used in the manufacture of iron, it may be 
desirable to give a method for their analysis. 

The weighed portion of the mineral, in a state of fine 
powder, is fused with about eight times its weight of acid 
sodium sulphate, and on cooling the fused mass is digested 
with cold water. The insoluble matter is filtered off, washed, 
dried, and weighed : it is usually free from titanic acid. It 
should, however, be treated by the method given in No. XIX., 
p. 218. Dilute the filtrate considerably, add a little nitric 
acid to it, and boil for some time to precipitate the titanic 
acid. Filter the precipitate, wash, dry, and ignite it, with the 
addition of a fragment of ammonium carbonate, to ensure the 
volatilisation of sulphuric acid, which the precipitate is apt 
to carry down with it. Titanic acid is slightly hygroscopic : 
it must be weighed, therefore, as expeditiously as possible. 
The iron, magnesia, lime, and alumina remain in solution, and 
may be separated by the methods given in No. XIX., p. 217. 

Titaniferous ores may also be decomposed by heating 
with sulphuric or hydrochloric acid, under pressure. A rapid 
and accurate method of estimating iron and titanium when 
present in solution together, after fusing the ore with acid 

Q 



226 Quantitative Chemical Analysis. 

sodium sulphate, and treating the fused mass with water, 
is founded on the behaviour of these substances, when 
reduced to their lowest state of oxidation, towards a solu- 
tion of potassium permanganate. The solution thus ob- 
tained is diluted to a determinate amount, an aliquot 
portion withdrawn, and reduced by zinc and sulphuric acid 
in the apparatus represented in fig. 58 (p. 222). The iron 
and titanic acid are thus brought to the state of ferrous and 
titanous oxides, and their joint amount is determined by 
addition of standard permanganate solution until the rose- 
colour of the latter solution is permanent. A second aliquot 
portion is brought into the flask, and an apparatus for the 
disengagement of sulphuretted hydrogen or sulphur dioxide 
is substituted for the carbonic acid flask. The iron is thus 
reduced to the ferrous state, the titanic acid remains un- 
changed. The solution is boiled to expel the excess of the 
reducing agent, whereby a portion of the titanic acid is 
precipitated, filtered rapidly, and the solution is again titrated 
with permanganate. The amount of ferrous oxide is thus ob- 
tained : the difference between the titrations gives the amount 
of permanganate corresponding to the titanic acid present. 
The weight of the latter substance is found from the equation 

Ti 2 O 3 + O = 2TiO 2 . 

1 6 parts of oxygen are therefore equivalent to 164 parts of 
titanic oxide. 

This process is also applicable to the analysis of rutile 
(native titanic oxide), sphene or titanite (a silico-titanate of 
calcium, CaSiO 3 .CaTiO 3 ), &c. 

XXI. WROUGHT AND CAST IRON, AND STEEL. 
Cast iron, in addition to chemically-combined carbon and 
graphite, contains variable quantities of silicium, phosphorus, 
sulphur, manganese, and copper ; and in very much smaller 
quantities, aluminium, chromium, titanium, zinc, nickel, cobalt, 
arsenic, antimony, tin, vanadium, magnesium, potassium, 
lithium, sodium, and nitrogen. The greater number of these 



Wrought and Cast Iron, &c. 



227 



substances are present in such very minute quantities, that 
their exact quantitative determination is a matter of great 
difficulty. It is very rarely necessary to attempt their esti- 
mation, since their presence in such small amount probably 
exercises little or no influence on the quality of the iron. 
Cast iron occurs in two leading varieties, viz. as grey and as 
white cast iron. The difference between the varieties is 
determined by the state of the carbon present in them. In 
grey cast iron the greater amount of the carbon is in the form 
of graphite, interspersed throughout the mass in a state of 
mechanical mixture ; in the white variety the carbon is 
chemically combined with the iron. Intermediate varieties, 
in which grey and white cast iron are mixed together in 
varying proportions, are classed as mottled cast iron. 

The following analyses of grey and white cast iron obtained 
from the same ore (spathic, containing about 42 per cent. 
iron) clearly show this characteristic difference : 

Grey. White. 

Iron 

Combined Carbon 
Graphite. 
Silicon . 
Sulphur . 
Phosphorus 
Manganese 
Xickel and Cobalt 

Wrought iron approaches more nearly to the character of 
pure iron : it contains much less carbon than cast iron, melts 
at a higher temperature, and is heavier. It also contains 
much less silicon, but the amounts of sulphur and phosphorus 
are as variable as in cast iron. The subjoined analysis is of 
Swedish iron of excellent quality : 



90-584 


93-I83 





4' 100 


2795 





4-414 


0-230 


0-039 


0-030 


0-099 


0-073 


1-837 


2-370 


traces 


0-014 


99768 


loo-ooo 



Iron . 
Carbon . 
Silicon . 
Sulphur . 
Phosphorus 



99*863 
o-o54 

O-O28 

o-o55 

traces 

loo-ooo 



Q 2 



228 Quantitative Chemical Analysis. 

Steel is intermediate in properties and purity between 
cast iron and wrought iron : it differs, however, from these 
varieties by its remarkable and valuable property of becom- 
ing hardened by heating and sudden cooling. In this state 
it is extremely brittle, and is without the characteristic fibre 
of malleable iron. Its tenacity, however, is much greater 
than that of wrought iron. Its average specific gravity is 
intermediate between that of cast iron and wrought iron. 
The following is an analysis of steel of medium quality : 



Carbon . 

Silicon . 

Sulphur . 

Phosphorus 

Manganese 

Iron (by difference) . 



0-501 
0-106 
0-002 
0-096 
0-144 



100-000 



The action of acids upon these varieties of iron is remark- 
able. When heated gently with strong hydrochloric acid 
white cast iron is completely dissolved, whereas grey cast 
iron leaves a residue of graphite. The combined carbon 
present in both varieties combines with the nascent hydrogen 
evolved by the action of the acid, giving rise to hydrocarbons 
of the C n H 2n series, C 2 H 4 . . . C6H 12 , which commu- 
nicate a peculiar odour to the issuing gas. 

The diluted acid at ordinary temperatures attacks cast 
iron but slowly : when dissolved by the aid of heat, white 
cast iron deposits a portion of its combined carbon ; this is 
soluble in potash, and on ignition leaves a black residue, 
containing silica. The residue from grey cast iron, in addition 
to graphite, also contains this carbonaceous matter together 
with a black magnetic substance containing iron. When 
dried this residue occasionally takes fire in contact with 
oxygen, and is converted into a mixture of ferric oxide and 
silica. 

It is said that the action of acids upon steel varies with its 
hardness : soft steel is far more readily dissolved than hardened 



Wrought and Cast Iron, &c. 229 

steel. Concentrated hydrochloric acid dissolves soft steel 
completely, but when diluted it leaves a larger amount of the 
black magnetic substance above mentioned than is yielded 
by malleable iron. It is worthy of note that diluted nitric 
acid completely dissolves this carbonaceous matter, forming 
a yellowish-brown coloured liquid, the intensity of the colour 
being proportional to the amount of combined carbon present. 
It is said that steel may be distinguished from cast and 
wrought iron by the action of hydrochloric acid of sp.gr. 1*134. 
With steel the acid occasions a rapid evolution of gas, which 
suddenly ceases in a short time (in about 20 seconds), 
whereas with cast or wrought iron the disengagement is con- 
tinuous. 

The complete analysis of iron, in addition to the estima- 
tion of the metal, necessitates the determination of the 
carbon, existing both in a state of combination and as 
graphite, of silicon, sulphur, manganese (zinc, cobalt, 
alumina, titanic acid), phosphorus, nitrogen, and admixed 
slag. The iron, however, is usually determined by dif- 
ference. 

Determination of the Total Carbon. Among the many 
accurate methods which have been proposed for the estima- 
tion of the total carbon contained in iron, those of Wohler, 
Weyl, and Ullgren are distinguished by reason of the ease 
and expedition with which they may be carried out. In all 
these processes the carbon is ultimately weighed as carbon 
dioxide. 

(a) Wohler 's Method. By burning the iron in a stream of 
oxygen. The iron must be previously reduced to the finest 
possible state of division, by filing with a hard file, and 
powdering in a large agate mortar. If the metal is very 
hard it is broken on a clean anvil, stamped to powder in the 
steel mortar, and passed through a fine sieve. From 3 to 6 
grams of the metal (according to its supposed richness in 
carbon) are weighed out into a platinum boat, and brought 



Wrought and Cast Iron, &c. 231 

into a short piece of combustion tubing drawn out at one 
end to a narrow tube (fig. 59). The other end is closed 
with a caoutchouc cork, and is connected with a gasometer 
filled with oxygen. To the narrowed end of the combustion 
tube is attached a chloride of calcium tube, connected with 
a weighed U-tube filled -J with soda-lime, and -J- with calcium 
chloride (fig. 60). A is a gasometer containing oxygen ; by 
means of the cock s the amount of the issuing gas can be 
easily regulated. The rate of its passage can be seen in b, which 
contains solution of caustic potash, destined to absorb any 
traces of carbonic acid and chlorine which may be present 
in the oxygen. The cylinder c is partially filled with soda- 
lime with the same object : the upper half and the two U- 
tubes contain calcium chloride, by which the gas is thoroughly 
dried. The combustion tube containing the weighed amount 
of iron rests in a gas furnace j d is a chloride of calcium 
tube : it is placed merely as a precaution against moisture 
passing into the weighed U-tube, and need not be weighed. 
The little U-tube /con tains one or two drops of concentrated 
sulphuric acid, sufficient to fill the bend ; its object is to 
prevent the possibility of the calcium chloride in the weighed 
tube absorbing atmospheric moisture. The process of com- 
bustion needs no particular attention ; if the heat is sufficiently 
powerful and the iron finely-divided, the whole of the carbon 
is converted into carbon dioxide. When the process is 
finished, b is detached from the gasometer, and a slow current 
of air is aspirated through the apparatus to expel the oxygen 
before the U-tube is again weighed. 

(b) Weyl's Method* In pulverising the iron for analysis, 
especially if very hard, there is considerable risk of mixing 
the sample with iron from the file, &c., employed. Weyl's 
method, by dispensing with the necessity of reducing the 
iron to powder, obviates this inconvenience. A piece of the 

* This method appears to have been described so far back as 1857 
by Binks, in a paper read before the Society of Arts. 



232 Quantitative Chemical Analysis. 

iron weighing from 10 to 15 grams is suspended by a pair of 
pincettes provided with platinum points in a beaker con- 
taining dilute hydrochloric acid. Care must be taken that the 
platinum points in contact with the iron are not moistened 
with the acid, or its solvent action will be impeded 
from the separation of carbon between the points and the 
metal. Connect the upper portion of the pincettes with the 
wire of a positive pole of a single element of Bunsen's battery. 
To the wire of the negative pole is attached a slip of plati- 
num foil, which is also immersed in the liquid. By regulating 
the distance between the foil and the metal, the strength of 
the current may be so modified that not a trace of ferric 
chloride is produced. The due regulation of the intensity 
is of the utmost importance, for if it is too strong the iron 
becomes passive, and chlorine is evolved from its surface, 
which brings about the oxidation of any separated carbon. 
This formation of chlorine is immediately rendered evident 
by the yellowish colour of the concentrated solution of iron as 
it falls away from the metal. When the operation succeeds, 
hydrogen only appears at the platinum foil, no gas being 
evolved from the positive pole (the iron). In 12 or 
15 hours the whole of the iron immersed will be dissolved : 
the separated carbon retains the shape of the metal. The 
undissolved portion is detached from the spongy mass of 
carbon, dried, and weighed. Its weight subtracted from that 
of the iron originally taken gives the amount employed for 
the carbon determination.* The carbon is washed slightly, 
and thrown into a short piece of combustion tube containing 
a plug of ignited asbestos. This acts as a filter and retains 
the carbon : care must of course be taken that the plug is 
sufficiently compact to prevent any particles of carbon 
passing through with the filtrate. The tube is gently heated 

* Instead of suspending the iron by the pincettes as above described, 
it may be broken into several pieces, and supported on a small platinum 
tray pierced with a number of holes. In this way the whole of the iron 
may be dissolved. 



Wrought and Cast Iron, &c. 233 

and a current of air aspirated through it. When perfectly 
freed from moisture, the plug together with the carbonaceous 
residue, is drawn by means of a bent wire into the middle of 
the tube, and is mixed with a small quantity of copper oxide 
by the aid of the wire. The combustion tube is heated in a 
gas furnace, and the carbon dioxide collected in a weighed 
soda-lime tube as described above. To ensure perfect com- 
bustion it is advisable to heat the carbon and copper oxide 
in a stream of oxygen : for this purpose the apparatus repre- 
sented in fig. 60 may be employed. Instead of burning it, 
the separated carbon may be treated as in the next method. 

(c) Ullgren's Method. It is necessary for this method that 
the sample should be in a state of coarse powder. Grey cast 
iron is preferably taken in the form of borings : white cast 
iron should be coarsely pondered. Dissolve 10 grams of 
copper sulphate in about 50 cubic centimetres of water, and 
into this solution, contained in a small beaker, weigh out about 
2 grams of the iron. Heat gently and with constant stirring 
until the iron is completely dissolved ; allow the solid portions 
to settle and pour away as much as possible of the clear 
fluid. Rinse the solid particles with any adhering copper 
solution into the small flask A (fig. 31), which should have a 
capacity of about 150 cubic centimetres. Care must be 
taken not to employ too large a quantity of wash water in 
rinsing the reduced copper and carbon into the flask : the 
total fluid in the flask should not exceed 25 cubic centimetres. 
Add to the flask 40 cubic centimetres of concentrated sul- 
phuric acid : if more than 25 cubic centimetres of wash 
water have been used, proportionally more acid must be em- 
ployed. Allow the mixture to cool, add about 8 grams 
of chromic acid, and connect the flask with the system of 
tubes represented in fig. 31. Heat the liquid gradually, and 
regulate the flame so as to maintain a regular evolution of 
gas : as it slackens, increase the heat, until white fumes make 
their appearance in the body of the flask. Remove the lamp 



2 34 Quantitative Chemical Analysis. 

and aspirate a slow current of air through the apparatus (3 
or 4 litres). Weigh the soda-lime tube and again aspirate 
air, in order to determine if the increase of weight due to the 
absorption of the carbon dioxide is constant. The prin- 
ciple of the method is evident. In contact with the sul- 
phuric and chromic acids the carbon separated on dissolving 
the iron is converted into carbon dioxide. 

Note on the preparation of Chromic Acid. 300 grams of coarsely- 
powdered commercial potassium bichromate are warmed with 500 cubic 
centimetres of water and 420 cubic centimetres of sulphuric acid. When 
the salt is dissolved the solution is allowed to stand ten or twelve hours, 
when the acid potassium sulphate separates out. The mother liquor is 
decanted, and the crystals allowed to drain for an hour. The solution 
is heated to 80 or 90, and gradually mixed with 1 50 cubic centimetres 
of sulphuric acid, and then with the same quantity of water, when the 
precipitated chromic acid will be redissolved. The solution is concen- 
trated in a porcelain basin until small spicular crystals appear on the 
surface of the liquid : after standing a few hours an abundant crop of 
chromic acid crystals will be obtained. The mother liquor on further 
evaporation will yield a fresh quantity of crystals. The chromic acid 
thus obtained may be drained by means of the filter pump, and dried 
on a porous tile placed beneath a bell jar : it is very hygroscopic and 
must be kept in a well-stoppered bottle. The crystals are not quite 
pure, as they contain small quantities of potash and sulphuric acid, but 
the presence of these substances in no way interferes with their employ- 
ment in the above method. 

Instead of absorbing the liberated carbonic acid by means 
of soda-lime, Ullgren prefers to use potash-pumice. This is 
prepared as follows : i part of caustic potash is dissolved in 
3 or 4 parts of water, the solution is heated to 100 in an 
iron pot, and a quantity of pumice, in pieces somewhat less 
than the size of a pea, is added until the mass becomes nearly 
dry. Whilst still hot it is transferred to a wide-mouthed 
stoppered bottle, and briskly shaken until, on cooling, the 
small pieces no longer adhere to each other. 

Determination of the Graphite. The above methods, it 
will be observed, determine merely the total amount of carbon, 
and give us no information respecting the proportion present 
as graphite and as combined carbon. In order to determine 



Wrought and Cast Iron, &c. 235 

the graphite, 3 to 5 grams of the sample are dissolved in 
moderately-concentrated hydrochloric acid at a gentle heat 
(comp. p. 228). The combined carbon combines with the 
nascent hydrogen, and is evolved together with the excess of 
this gas : the graphite remains undissolved. Filter the 
residue through asbestos, contained in a short piece of com- 
bustion tube (fig. 59, p. 230), wash it with hot water, then 
with potash, alcohol, and a little ether. Dry it, mix it with 
a little copper oxide, and burn it in a stream of oxygen as 
directed on p. 233 (see fig. 60). On deducting the weight of 
the graphite thus obtained from the total amount of the car- 
bon, the difference gives the quantity of combined carbon. 
This process is the most uniformly accurate, but for technical 
purposes it may be thus simplified. Treat the weighed sample 
of iron with dilute hydrochloric acid, and when the metal is 
nearly dissolved, add a large quantity of strong hydrochloric 
acid. The insoluble matter is collected on a weighed filter, 
washed with hot water, with dilute hydrochloric acid to re- 
move iron, and with potash to remove silica. In washing with 
the alkali there is occasionally a brisk evolution of hydrogen, 
owing to the oxidation of the silicon to silicic acid. The 
insoluble matter is lastly washed with alcohol and ether to 
remove any traces of adhering hydrocarbons, dried at 120 
and weighed. The filter is thrown into a small platinum 
crucible, and incinerated : the weight of the residue, less that 
of the filter-ash, gives the amount of silica (and titanic acid) 
mixed with the graphite. 

Estimation of Combined Carbon in Steel and Wrought Iron. 

Eggertz's Method. In the case of metallic iron containing 
but a small proportion of combined carbon, this method is 
readily applicable. When such iron is dissolved in nitric 
acid, the solution becomes coloured more or less brown in 
proportion to the amount of carbon present (comp. p. 229). 
By comparing the depth of this coloration with that of a 
standard tint, equivalent to a known quantity of carbon, the 



236 Quantitative Chemical Analysis. 

determination of the combined carbon in a sample of steel 
or wrought iron may be made in a very short time, and with 
great accuracy. The process is therefore especially appli- 
cable in steel works, where such determinations are of frequent 
occurrence. It is thus conducted. Two thin test-tubes, 
made of the same glass, are divided into 0-5 cubic centimetre 
by means of a pipette, the graduation being scratched on the 
side by a diamond. A piece of steel weighing not less than 
100 grams, supposed to contain from 07 to ro per cent, of 
carbon, is finely powdered, and the amount of carbon it con- 
tains determined by one of the methods above described. 
As this sample is to serve as the standard of comparison, 
the amount of carbon which it contains must be estimated 
with the utmost possible accuracy : it is advisable therefore 
to make several determinations, and to take the mean of the 
results. Supposing, for the sake of illustration, that the 
sample contains 075 per cent, carbon. One decigram of 
the steel or wrought iron to be tested, and one decigram of 
the standard steel (containing 075 per cent carbon), are 
weighed out with the greatest accuracy on small tared watch- 
glasses. The samples are brought into thin dry test-tubes, 
and covered with about 2 cubic centimetres of dilute nitric 
acid (sp. gr. 1*2), perfectly free from chlorine. In a few 
minutes the greater portion of the metals will be dissolved. 
The tubes are placed in a beaker containing water at 80, 
which is maintained at this temperature until all action on 
the metals is at an end. In the case of steels this ceases in 
about two hours. Allow the tubes to cool, and decant the 
clear supernatant liquid from the undissolved matter into the 
graduated test-tubes ; the solution of the standard steel is 
poured into the one tube, that of the iron into the other. 
To each portion of residue add two or three drops of nitric 
acid, and heat gently over the lamp : if no further evolution 
of gas occurs, the insoluble matter consists merely of silica or 
graphite. Allow these solutions to cool, and add them to 
the contents of the graduated tubes. Dilute the solution 



Wrought and Cast Iron, &c. 237 

of the standard steel until it exactly measures 7-5 cubic 
centimetres : each cubic centimetre is equivalent therefore 
to o'oooi gram of carbon. Now add water, drop by drop, 
to the liquid contained in the other tube, agitate the 
mixture, and compare the tint with that of the liquid in 
the standard tube. If the tints are equal read off the 
volume of the liquid : each cubic centimetre represents one 
tenth per cent of combined carbon. Thus, supposing 
that on diluting the liquid to be tested to 4-5 cubic centi- 
metres it gave a depth of colour equivalent to that of the 
standard solution, then the sample contains 0-45 per cent, 
carbon. It must be remembered that this method is 
strictly comparative : to ensure accuracy the circumstances 
of temperature, time, amount of acid used, &c., must be as 
nearly as possible identical. The normal solution must 
be prepared afresh for each series of comparisons, as it 
gradually becomes paler on keeping, especially if exposed 
to light 

When such determinations of carbon in steel or wrought 
iron are of frequent occurrence, as in iron works, it is more 
convenient to prepare a number of tubes containing the 
brown solution, corresponding to different percentages of 
carbon. The coloured solutions may be readily obtained by 
digesting roasted coffee in dilute spirit : if kept in the dark 
when not in use they maintain their intensity of colour un- 
impaired for a long time. Fig. 6 1 represents a wooden frame 
containing the tubes : these are about f of an inch in dia- 
meter, and about 3^- to 4 inches long. The tubes after the 
introduction of the properly-diluted solutions are hermetically 
sealed. The liquid in the tube placed in the second hole 
in the rack is made to correspond to the tint of a solution 
of i gram of steel containing 0*02 per cent of combined 
carbon in 15 cubic centimetres of nitric acid of sp. gr. 
i -20. The solution in the tube in the fourth hole cor- 
responds to that of the same quantity of iron containing 0*04 
per cent carbon; and so on in regular succession, each tube 



238 Quantitative Chemical Analysis. 

"increasing in value by 0*02 per cent. The last tube is equiva- 
lent therefore to 0-3 per cent. The process is thus conducted ; 



FIG. 61. 




i gram of the iron or steel to be tested, in the state of fine 
powder, is weighed out into a large test-tube, and digested at 
70 or 80 with 10 cubic centimetres of the dilute nitric acid, 
free from chlorine, for about half an hour. The solution is 
quickly cooled by immersing the tube in cold water, and is 
filtered, without the residue being disturbed, through a small 
dry filter into a test-tube of the same size and made of the 
same glass as those containing the standard solutions. The 
insoluble matter is then treated with 5 cubic centimetres of 
the nitric acid, heated gently, and the solution added to the 
main portion. The entire solution is mixed by shaking, and 
its colour compared : the holes in the stand allow the tube 
to be placed side by side with the standard solution : the 
number affixed to the tube with which it corresponds in 
colour indicates the percentage amount of carbon in the 
sample. If the steel or wrought iron contains more than 0-3 
per cent of carbon, 0*5 gram is taken, or the solution is 
diluted with an equal volume of water, shaken, and half of 
it poured away. The comparison is assisted by placing a 
white screen of paper behind the tubes. In this manner a 
carbon determination may be made to within -01 per cent 

Estimation of the Sulphur. When iron containing sulphur 
is treated with hydrochloric acid, the whole of the sulphur is 
evolved as sulphuretted hydrogen. By passing the sulphu- 
retted hydrogen into a solution of bromine in hydrochloric 
acid, it is completely absorbed and converted into sulphuric 
acid ; by converting the sulphuric acid into barium sulphate, 



Wrought and Cast Iron, &c. 



239 



the amount of sulphur may be readily determined. The 
apparatus required for the purpose is seen in fig. 62. The 
flask of 150 cubic centimetres capacity is fitted with a 
caoutchouc cork, which has previously been boiled in dilute 
caustic soda to remove the sulphur contained in it : the 
cork is pierced with two holes, into one of which fits a 
straight piece of tubing, curved at its lower extremity : near 



FIG. 62. 




the upper end is a bulb fille^ with dilute hydrochloric acid r 
the tube is closed by means of a small piece of clean caout- 
chouc tubing and a clamp. The second hole of the cork 
contains a tube bent at right angles, to which is adapted a 
small U-tube, containing the solution of bromine. It will 
be seen by the arrangement of the glass tubes that the gas 
passes twice through the liquid contained in the U-tube. 



240 Quantitative Chemical A nalysis. 

Into the flask weigh out 10 or 15 grams of the finely divided 
iron, fill the bulb with moderately diluted hydrochloric acid, 
insert the cork, open the clamp, and by aspiration at the 
exit-tube of the absorption apparatus bring the acid into the 
flask. Heat the flask gently, and introduce fresh acid from 
time to time until the iron is completely dissolved. Con- 
nect the U-tube with an aspirator, and gently draw a current 
of air through the apparatus to remove the last traces of 
sulphuretted hydrogen in the flask. Transfer the liquid 
from the U-tube to a beaker, boil to expel excess of bromine, 
and add barium chloride. 

it is always advisable to test the solution of ferrous chloride 
for sulphuric acid by concentrating it, and adding one or two 
drops of barium chloride : if any precipitate is thus obtained it 
is to be filtered off, washed, dried, and weighed. The insoluble 
residue should also be tested for sulphur by fusing it with 
nitre and sodium carbonate, dissolving in water, and testing 
the acidified solution with barium chloride. Usually, how- 
ever, the residue will be found to be quite free from this 
substance. A very convenient method of converting 
the sulphur in iron to the state of sulphuric acid consists 
in absorbing the sulphuretted hydrogen in a dilute solution 
of potash or soda free from sulphate. Decant the alkaline 
solution into a beaker, add a few drops of bromine free 
from sulphuric acid, heat gently, acidulate with hydro- 
chloric acid, boil, add a few drops of barium chloride, and 
after standing about twenty-four hours filter off the precipi- 
tated barium sulphate. 

The sulphuretted hydrogen may also be absorbed by a 
dilute solution of a cadmium salt : the cadmium sulphide 
possesses the advantage that it can be dried at 100 without 
alteration. 

When a number of estimations of sulphur in iron have to be 
made, it is convenient to employ a volumetric method to 
determine the sulphuretted hydrogen evolved. Both the 
following plans will be found to give concordant results. 
The U-tube (fig. 62) is filled with a solution of caustic soda 



Wrought and Cast Iron, &c. 241 

( i of soda to 5 of water) to absorb the sulphuretted hydrogen. 
The operation of dissolving the iron is made exactly as in the 
foregoing methods. When the action is at an end, pour the 
contents of the U-tube into a large beaker, dilute with about 
200 cubic centimetres of boiled water, acidify with hydro- 
chloric acid, add a little starch paste, and add standard 
iodine solution until the solution is turned blue. 

Or wash the alkaline solution into a small beaker 
containing water, to which a measured quantity, say 10 
cubic centimetres, of deci-normal arsenious acid solution 
is added. Add hydrochloric acid to distinct acid reaction, 
set aside for a few hours in a warm place, dilute to a de- 
terminate quantity, say 100 cubic centimetres, filter through 
a dry filter, withdraw 25 cubic centimetres of the filtrate, 
neutralise with sodium bicarbonate, add a small quantity of 
starch liquor, and a dilute solution of iodine from a burette 
until the blue colour is permanent. The determination is 
repeated with a second portion of 25 cubic centimetres. 
This method is based upon the following reaction between 
the arsenious acid solution and the sulphuretted hydrogen : 



As 2 3 + sH 2 S = As 2 S 3 + 3H 2 O. 

i cubic centimetre of deci-normal arsenious acid solution = 
0*0024 gram sulphur. With proper care this method affords 
very accurate results. 

Determination of Nitrogen. From the experiments of 
Schafhautl, Despretz, Boussingault, Fremy, and others, it 
appears that nitrogen is an invariable constituent of cast and 
wrought iron, and steel ; some of these authorities are of 
opinion that when present in estimable proportion it exerts 
a marked influence upon certain of the physical properties 
of the metal. The iron is said to be rendered white and 
brittle and less liable to change in the air or water (Des- 
pretz ; Buff). According to Frdmy the nitrogen exists in 
the iron in two conditions, since when the metal is dissolved 



242 Quantitative Chemical Analysis. 

in hydrochloric acid, a portion is converted into ammonia, 
whilst another portion remains in combination with the car- 
bonaceous residue. On the other hand, Erdmann, Stahl- 
schmidt, Stuart, and Baker assert that the quantity of nitro- 
gen usually present in iron is too minute to have the least 
influence upon the metal. The following methods of deter- 
mining the nitrogen present in iron are due' to Ullgren.* 

Determination of the Nitrogen which forms Ammonia on 
dissolving the Iron in Hydrochloric Acid. About 2 grams of 
the finely-divided iron are treated, in a flask provided with a 
bent tube, with a solution of 10 grams crystallised copper 
sulphate, and 6 grams fused sodium chloride. When the iron 
is dissolved, add excess of boiled milk of lime, boil for some 
time, and determine the evolved ammonia as in No. VI 1 1. p. 94. 

The quantity of the ammonia in the distillate, if very 
minute, may be determined by Nessler's method (see Water 
Analysis). 

Determination of the Nitrogen present in the Carbonaceous 
Residue. By combustion with mercuric sulphate, and 
measurement of the evolved nitrogen. The apparatus em- 
ployed for the purpose is seen in fig. 63. A is a piece of 
combustion tubing about 30 centimetres long ; it is filled as 
far as b with 1 2 grams of dry magnesite or bicarbonate of 
soda. From b to c is placed the mixture of about cri gram 
of the carbonaceous residue (dried at 120-130), with from 3 
to 4 grams of the mercuric sulphate. The mixture is made 
in a small glazed porcelain or agate mortar, which is rinsed 
with a fresh portion of the sulphate after the introduction of 
the mixture. From c to d is a layer of coarsely-powdered 
pumice, previously mixed with mercuric sulphatef and a little 
water, and dried : its object is to prevent the possible evolu- 
tion of carbon monoxide. The remainder of the tube is 

* Ann. d. Chem. u. Pharm. cxxiv. 70, cxxv. 40. 

f Preparation of the mercuric sulphate. This salt is obtained by 
gradually adding 4 parts of mercury to 5 parts of strong boiling sulphuric 
acid, and heating the mixture until sulphur dioxide ceases to be evolved, 
and the whole is converted into a dry saline mass. 



Wrought and Cast Iron, &c. 243 

filled with fragments of pumice soaked in a concentrated solu- 
tion of potassium bichromate, and allowed to drain: its object 
is to absorb the sulphur dioxide which is disengaged. The 
various mixtures are separated by plugs of recently-ignited 



FIG. 63. 



d\ 




asbestos. The tube is fitted with a caoutchouc cork and 
bent tube. The vessel B is designed to collect and measure 
the evolved nitrogen. The narrow portion holds about 20 
cubic centimetres, and is graduated into y^ths of a cubic 
centimetre : the bulb holds about 40 cubic centimetres, 
and the lower portion from 20 to 30 cubic centimetres. Fill 
the tube with mercury and invert it in the trough. By means 
of a pipette pass up potash solution (i part potassium hydrate 
and 2 of water) until the bulb is nearly filled (to within 10 
cubic centimetres), and then add 15 cubic centimetres of a 
clear and saturated solution of tannic acid. Now gently 
heat the magnesite or bicarbonate of soda at the extreme 
end of the tube, and gradually heat the tube until about 
half the substance has been decomposed : the air within the 
apparatus is thus expelled. Bring the end of the delivery- 
tube under B, heat from b to c very gently to remove any 
moisture present, then heat from c to d gradually ; and when 



244 Quantitative Chemical Analysis. 

this portion of the tube is nearly red hot, heat the part 
b c to a strong red heat. Heat from b to d until no more gas is 
evolved, and sweep out the gases within the tube by heating 
the undecomposed portion of magnesite. Close the end of 
B with the thumb, and transfer the tube to a vessel of water : 
the mercury and potash will be replaced by water ; adjust 
the levels of the liquid within and without the tube, and read 
off the volume of nitrogen, making the necessary corrections 
for tension of aqueous vapour, temperature, and pressure, 
and calculate the weight of the nitrogen found. 

Determination of the Silicon, Iron, Manganese, Cobalt, 
Nickel, Zinc, Alumina, Titanic Acid, Alkaline Earths, 
and Alkalies* Weigh out about 10 grams of the finely- 
divided iron into a porcelain basin, cover it with a large 
watch-glass, and dissolve it in moderately dilute hydrochloric 
acid, add a few drops of sulphuric acid, and evaporate to 
dryness on the water.-bath, heating until the mass no longer 
smells of hydrochloric acid. Moisten the dried salt with 
hydrochloric acid, heat on the water-bath, add water, filter into 
a capacious porcelain basin, wash and dry the residue. Set it 
aside and label it ' Pp. I.' Heat the filtrate with nitric acid, 
boil, add water until the liquid measures at least 1,500 cubic 
centimetres, and gradually add ammonium carbonate in dilute 
solution until the fluid just loses its transparency (it must not, 
however, show any sign of distinct precipitate). Heat to 
boiling, and maintain the liquid in ebullition until the car- 
bonic acid is expelled. If the solution is not too concen- 
trated the ferric hydrate separates rapidly. Add now a few 
drops of ammonia, filter, and wash the precipitate with water 
containing a little ammonium chloride. The only condition 
necessary to ensure success is the proper dilution of the 
liquid : for the quantity of iron taken it should not measure 
less than \\ litre. Dry the washed precipitate, set it aside, 
and label it < Pp. II.' 

* Compare Lippert, Zeits. f. anal. Chemie, ii. 39. 



Wrought and Cast Iron, &c. 245 

Add ammonia in slight excess to the nitrate from ' Pp. II.,' 
and boil until the free ammonia is nearly expelled, filter, 
and, without washing, redissolve in hydrochloric acid, and 
again precipitate with ammonia. Filter, wash and dry the 
precipitate : call it 'Pp. III.' 

Mix the two filtrates, acidify with hydrochloric acid, con- 
centrate in a porcelain basin, transfer to a small flask, add 
ammonia and ammonium sulphide, cork the flask, and let 
the liquid stand in a warm place for at least twenty- four hours. 
Decant the clear fluid through a filter, and wash it with 
water containing a little ammonium sulphide ; allow it to 
drain as far as possible, spread the filter on a glass plate, and 
wash off the precipitate into a small flask ; add acetic acid, 
and cork the flask. Label the flask ' Pp. IV.' 

Transfer the filtrate to a porcelain basin, evaporate, add 
nitric acid, and heat until the ammonium chloride is de- 
composed. Evaporate to dryness, dissolve in a little 
water, filter if necessary, add ammonium oxalate, and, after 
standing twenty-four hours, filter off the calcium oxalate, 
wash it and convert it into lime by ignition. Evaporate the 
filtrate to dryness in a platinum basin, ignite at first gently 
and then to a red heat. Treat with water, and filter off the 
magnesia and weigh it. The solution of the alkalies is acidi- 
fied with hydrochloric acid, evaporated to dryness, the mixed 
chlorides weighed, and separated as in No. IV. Part. II. 

The residue marked * Pp. I. ' contains the graphite, silica, 
a portion of the phosphorus as phosphide of iron, titanic 
acid, barium sulphate. Fuse it with sodium carbonate 
and a little nitre, soften the mass with water, add hydro- 
chloric acid in excess, evaporate to dryness and separate 
the silica. Weigh it, and treat it in a platinum basin 
with a moderately-concentrated solution of sodium car- 
bonate : if it dissolves completely, it is pure ; if a weighable 
amount remains, determine its quantity and test it for 
titanic acid and barium sulphate. To the filtrate from the 
silica, add ammonia, filter off any precipitate which forms, 



246 Quantitative Chemical Analysis. 

redissolve it in hydrochloric acid, and again add ammonia. 
Filter and dry, and add the precipitate which forms to 
' Pp. III.' To the nitrate add ammonium sulphide, and 
add any precipitate which separates out to * Pp. IV.' 
Test the nitrate for the alkaline earths in the manner 
above directed, and if their amount admits of determination, 
weigh them. 

'Pp. Il.'and 'Pp. III.' contain the ferric oxide and alumina, 
and the small quantity of titanic acid which may have 
passed into the hydrochloric acid solution. The mixed pre- 
cipitates are placed in several porcelain or platinum boats, 
and strongly ignited in a glass tube in a current of dry 
hydrogen until the formation of water ceases. Allow the 
reduced metal to cool in the current of the gas, and treat 
the mixture with dilute nitric acid (i part of acid to 30 of 
water), until the iron is dissolved : filter the liquid into a 
litre flask, dilute to 1,000 cubic centimetres, and determine 
the iron by precipitating an aliquot portion by means of 
ammonia. This method is better than that of determining 
the iron in a fresh portion of the sample, unless the sample 
is perfectly homogeneous. The residue left after treatment 
with dilute nitric acid is fused with acid- potassium sulphate, 
digested with water, and any residual silica filtered off and 
weighed. The filtrate is boiled for some time, and if any 
titanic acid separates out, it is filtered off and weighed. Add 
ammonia and boil : dissolve the precipitate in hydrochloric 
acid, transfer the liquid to a small test-tube, add a little 
tartaric acid, then ammonia, and ammonium sulphide. 
Allow the liquid to stand until any iron sulphide which forms 
has completely settled, filter, redissolve in hydrochloric acid, 
boil with a few drops of nitric acid, add ammonia, and filter 
off the ferric hydrate ; dry, ignite, and weigh. To the yellow 
filtrate containing the alumina, add a little pure sodium 
carbonate and nitre, evaporate to dryness, and ignite until 
the mass is completely white. Rinse the residue into a 
beaker, dissolve in hydrochloric acid, and add excess of 
ammonia. Filter off the precipitate, wash, dry, and weigh it. 



Wrought and Cast Iron, &c. 247 

Mix the filtrate with a few drops of magnesium sulphate 
solution : if a precipitate of magnesium-ammonium phosphate 
forms after standing, calculate the weighed precipitate as 
A1PO 4 . If no precipitate is formed, the amount, of phosphoric 
acid, determined as under, is to be subtracted from the 
weighed precipitate. The remainder gives the amount of 
alumina. This precipitate will also contain any chromium 
present in the iron. Its amount is in general exceedingly 
minute. In case the quantity happens to be more consider- 
able than usual, fuse the mixed oxides with a little sodium 
carbonate and nitre, dissolve in water, add a small quantity 
of ammonium nitrate, evaporate nearly to dryness, add a little 
water, filter, reduce the alkaline chromate with sulphurous 
acid, and precipitate the chromic oxide with ammonia, boil 
for some time, filter, dry, and weigh. The difference 
between the weight of the oxide thus obtained and that of 
the original precipitate gives the alumina. 

* Pp. IV.' consists principally of manganese sulphide : 
it may also contain zinc, copper, nickel, and cobalt. By 
digestion with acetic acid the greater portion of the man- 
ganese sulphide will have been dissolved : filter, spread the 
filter containing the residue on a glass plate, and wash the 
precipitate into a small beaker containing sulphuretted hy- 
drogen water acidulated with hydrochloric acid. Allow the 
liquid to stand for a short time : the zinc and the remainder 
of the manganese pass into solution; the copper (which is 
estimated as under), cobalt, and nickel remain undissolved. 
Filter, evaporate the acid solution to a few cubic centimetres, 
add excess of soda solution, boil, and filter off any man- 
ganese precipitated, redissolve it in hydrochloric acid, and 
add the liquid to the acetic acid solution. Pass sulphuretted 
hydrogen through the alkaline filtrate and treat the zinc 
sulphide as in No. XIII. Part II. p. 106, 

The filter containing the nickel, cobalt, and copper is dried 
and incinerated, dissolved in a few drops of aqua regia, 
treated with sulphuretted hydrogen, filtered if necessary, 
and the nickel and cobalt separated as in No. XVI. Part IV. 



248 Quantitative Chemical Analysis. 

To the solution containing the manganese, add sodium 
carbonate in slight excess, boil, filter, wash the manganous 
carbonate, and ignite it until the weight is constant. It con- 
sists of mangano-manganic oxide. Care must be taken to 
remove the precipitate as completely as possible from the 
filter before it is incinerated. 

Determination of the Phosphorus, Copper, Arsenic, and 
Antimony. Weigh out about 10 grams of the iron in a state 
of fine powder into a capacious flask, place a small funnel in 
the neck, and pour strong nitric acid in small quantities at 
a time over the metal. Warm the liquid, and when all 
visible action is at an end, heat the residue with a fresh por- 
tion of nitric acid. Mix the solutions, add hydrochloric acid 
and evaporate to complete dryness : again add hydrochloric 
acid and evaporate a second time to dryness. Dissolve in 
water, and saturate the liquid with sulphuretted hydrogen. 
After the precipitate has completely settled, filter into a 
large flask, dry the precipitate, and digest it with carbon 
disulphide to remove the sulphur. Separate the copper, 
arsenic, and antimony in the black residue which remains, 
according to No. XVI. Part IV. 

Pass a current of carbonic acid through the filtrate con- 
tained in the flask to dissipate the dissolved sulphuretted 
hydrogen, add a small quantity of ferric chloride solution, 
nearly neutralise with sodium carbonate, add a little barium 
carbonate, and cork the flask ; after standing about twenty- 
four hours, the precipitate will contain the whole of the 
phosphoric acid ; filter it off, wash, dissolve in hydrochloric 
acid, remove the barium by the addition of sulphuric acid, 
filter, concentrate the filtrate, add molybdic acid solution, 
and proceed as in No. XIX. Part IV. p. 218. 

The residue unattacked by aqua regia not unfrequently 
contains phosphide of iron. Fuse it, therefore, with sodium 
carbonate and nitre, extract with water, and test the solution 
for phosphoric acid. 



Iron Slags. 



249 



Determination of Admixed Sla% in Cast Iron. 

By treating a piece of the metal by Weyl's method (p. 231), 
with very dilute hydrochloric acid, any admixed slag is not de- 
composed, and remains with the graphite, &c. The residue is 
collected on a filter, washed, and ignited in a platinum crucible 
until the carbon is consumed. The ignited mass is boiled 
with solution of sodium carbonate to dissolve the free silica, 
the insoluble matter ignited first in a stream of hydrogen, and 
then in dry chlorine (free from air), and treated with dilute 
hydrochloric acid, and again with boiling sodium carbonate 
solution. Filter off the insoluble matter, dry it, and weigh 
it as slag. 

The following analysis by Fresenius of ' Spiegel-eisen,' 
produced from the spathic ore of Stahlberg, near Miisen, 
was made by the foregoing method : 

. 0*014 

0-997 



Iron . 


. 82-860 


"Magnesium 


. 0-045 


Nitrogen 


Manganese 


. 10-707 


Calcium . 


. 0-091 


Silicon 


Nickel . 


. o'oi6 


Potassium 


. 0-063 


Carbon 


Cobalt . 


trace 


Arsenic . 


. 0-007 


Slag. 


Copper . 


. 0-066 


Antimony 


. 0-004 




Aluminium 


. 0-077 


Phosphorus 


0-059 




Titanium . 


. o'oo6 


Sulphur . 


. 0-014 





0-665 



100-014 



XXII. IRON SLAGS. 

The slag produced in the manufacture of cast iron may be 
regarded as a double silicate of lime and alumina, in which a 
portion of the lime is replaced by ferrous and manganous 
oxides, magnesia, and alkalies. Phosphoric acid and calcium 
sulphide are also usually present. The average composition 
of the blast-furnace slag is represented by the following 
analysis : 



Silica . . 41*85 
Alumina . . 1473 
Ferrous oxide . 2-63 
Manganous oxide 1-24 



Lime 

Magnesia . 
Potash . 
Calcium . 



30-99 
4-76 
1-90 



Sulphur . 
Phosphoric acid 



0-92 
0-15 



100-32 



Occasionally, however, either from carelessness or from 
defective working of the furnaces, the amount of iron in the 
slag is considerably augmented. Some slags are completely 



2 5 Quantitative Chemical Analysis. 

decomposed by treatment with hydrochloric acid : others 
are only partially acted upon. All of them yield more or 
less sulphuretted hydrogen on heating with hydrochloric 
acid. The amount thus evolved may be determined by one 
of the methods described on p. 239. 

The methods of analysis described in Nos. XI. and XII. 
Part II. are generally applicable to slags which resist the 
action of hydrochloric acid. 

A very convenient method of decomposing slags is to 
heat them in a finely-divided state in a sealed tube for two 
hours for 200 in an air bath, with a mixture of three parts 
by weight of strong sulphuric acid and one of water, or with 
hydrochloric acid containing 25 per cent. HC1. About one 
gram of the powder is introduced into a strong tube of 
Bohemian glass rounded at one end ; the other end is thick- 
ened before the blow-pipe and drawn out. Add the acid, 
and carefully seal the tube. When the action is at an end, 
open the tube after cooling, and rinse out its contents into a 
porcelain dish, and evaporate to dryness in the ordinary 
way to render the silica insoluble. The remainder of the 
analysis is conducted after Nos. XI. and XII. Part II. This 
process is often applicable to the analysis of natural silicates : 
the determination of mixtures of ferric and ferrous oxides 
may also be conveniently made by this method. 

Tap-cinder may be analysed by the above method, or 
according to the processes given in Nos. XX. XXI. Part IV. 

XXIII. ASSAY OF ZINC-ORES. 

From 0-5 gram to i gram (according to its supposed rich- 
ness) of the finely-powdered ore is dissolved in aqua regia, 
the solution is evaporated to dryness, the residue heated 
with water, filtered, and mixed with ammonium carbonate 
and ammonia solutions. The liquid is gently heated on a 
sand-bath for half an hour, filtered, and the insoluble matter 
washed with ammonium acetate solution. The filtrate con- 
taining the zinc is mixed with an excess of sodium or am- 
monium sulphide, and after standing for a short time the 



Zinc Ores. 251 

cipitate is washed first by decantation and afterwards on the 
filter with warm water containing ammonia, until the filtrate 
no longer discolours an alkaline solution of lead acetate. 
The filter is pierced, and the zinc sulphide is carefully 
washed down into a 500 c.c. flask containing an excess of 
ferric chloride solution and some free hydrochloric acid. 
The flask is well closed and set aside in a warm place for a 
short time. The mixture should occasionally be agitated to 
accelerate the reaction. The zinc sulphide in contact with 
the ferric chloride and free hydrochloric acid is converted 
into zinc chloride, sulphur is separated, and the iron is re- 
duced to the state of ferrous chloride. 

ZnS + Fe 2 Cl 6 = ZnCl 2 + 2FeCl 2 + S. 

The solution (which should have a yellow colour, indica- 
ting an excess of ferric chloride, and be free from any smell 
of sulphuretted hydrogen) is diluted to the containing-mark, 
well shaken, and an aliquot portion of the liquid titrated 
with acid-chromate of potassium solution. If the liquid is 
quite cold and dilute, the free sulphur exercises no influence 
upon the result. 

The alkaline solution of zinc may also be titrated with a 
solution of sodium sulphide of known strength. This method 
is largely used in many zinc works on the Continent. 
Saturate a solution of soda with sulphuretted hydrogen, and 
add a second quantity of the alkaline liquid until the smell 
of the gas is no longer perceptible. Dissolve 10 grams of 
pure zinc in dilute sulphuric acid (taking care not to use a 
very great excess of acid), and dilute the solution to i litre. 
Also dissolve a few crystals of sodium tartrate in water, add 
a small quantity of caustic soda and lead acetate, and heat 
the mixture until the liquid is clear. 

Transfer 25 c.c. of the standard zinc solution to a beaker ; 
add a mixture of ammonium carbonate and ammonia suffi- 
cient to redissolve the precipitate first formed, and spread a 
few drops of the alkaline lead solution on a piece of filter- 
paper placed on a porcelain slab or plate, and run in the 
solution of sodium sulphide until a drop of the liquid with- 



252 Quantitative Chemical Analysis. 

drawn by a glass rod and brought into contact with the lead 
acetate forms a black ring at the point of contact. If neces- 
sary, the solution of the sodium sulphide is diluted to a con- 
venient strength for titration, and its exact value again de- 
termined in the above manner on quantities of 25 c.c. or 50 
c.c. of the standard zinc solution. 

The amount of zinc in the ammoniacal liquid obtained 
by treating the ore is then determined in exactly the same 
manner by the sodium sulphide solution. The strength of 
the solution of sodium sulphide must be re-determined before 
each series of experiments, as it experiences alteration by 
exposure to the air. 

Many zinc-blendes contain notable quantities of copper, 
which by combining with the sodium sulphide increases the 
amount of zinc apparently present. The safest plan in such 
a case is to remove the copper by sulphuretted hydrogen in 
an acid solution ; filter, evaporate the filtrate with nitric acid, 
dilute, add ammonia, and proceed as directed. 

In the case of ores which contain alumina and manganese, 
zinc is more accurately estimated by a standard solution of 
potassium ferrocyanide. The solution of the ore prepared 
as above is acidified with hydrochloric acid, and the ferro- 
cyanide solution (previously standardised by means of pure 
zinc) is added until a drop withdrawn from the liquid gives a 
brown colour with solution of uranium nitrate placed on a 
porcelain slab. (Comp. pp. 333, 334.) 

XXIV. ASSAY OF TIN-ORES. 

The amount of tin in its ores may be easily estimated by 
fusing with potassium cyanide : the process, although not 
absolutely correct especially in presence of lead or copper 
yields results of sufficient accuracy for technical purposes. 
About 6 or 8 grams of the finely-powdered ore is placed in 
a smooth porcelain mortar, and intimately mixed with five 
times its weight of commercial potassium cyanide. The 
mixture is projected into a small clay crucible, in the bottom 
of. which a quantity of powdered potassium cyanide has been 



Tin-Ores. 253 

previously placed, sufficient to form a layer of i or 2 centi- 
metres in depth, and the mortar is rinsed with a fresh quantity 
of cyanide, which is poured on the top of the mixture. It 
is advisable by way of control to prepare two such crucibles, 
and to take the higher result as the true one. The crucibles 
are heated to a moderate red heat, and the cyanide is kept in 
fusion for 10 or 15 minutes : they are removed from the fire 
and gently tapped, to promote the formation of a single 
button of the reduced metal. After cooling they are broken, 
and the buttons of metal are extracted and weighed after 
the adhering flux has been removed. The saline mass 
should be triturated with water in order to be certain that 
the reduction has been effectual and that the whole of the 
metal has been collected in one piece. The silica in the ore 
unites with the alkaline carbonate always contained in the 
commercial cyanide. If the ore contains any considerable 
portion of lead or copper, these must first be removed by 
digestion with strong hydrochloric acid. 

XXV. SEPARATION OF TIN FROM TUNGSTEN. 
Commercial stannates of soda are frequently mixed with 
sodium tungstates. The best method of determining the 
proportion of the two acids is to fuse the mixture with 
potassium cyanide, when the tin only is reduced to the 
metallic state. About 2 grams of the powdered salt is 
mixed with four times its weight of fused and powdered 
potassium cyanide, and the mixture heated in a porcelain 
crucible. By treating the fused mass with water the alkaline 
tungstate dissolves, together with the excess of the cyanide : 
the reduced metal is washed, and converted into oxide by 
treatment with nitric acid. The filtrate is heated with nitric 
acid to decompose the potassium cyanide, evaporated nearly 
to dryness, the residue dissolved in alkali, acidified with 
nitric acid, and the tungstic oxide precipitated by means of 
mercurous nitrate. The mercurous tungstate is washed with 
water containing a little mercurous nitrate., dried, and con- 
verted into tungstic oxide by ignition in contact with air. 



254 Quantitative Chemical Analysis. 

XXVI. WOLFRAM. 

This mineral is a tungstate of iron and manganese 
(FeMn") WO 4 , in which the proportion of the two metals is 
variable. It occurs associated with tin -ore, tungstate of cal- 
cium, galena, &c., in Cornwall, Cumberland, in the Hebrides, 
France, Bohemia, and North America. 

The finely-divided mineral is heated with aqua regia until 
it is completely decomposed. The solution is evaporated 
to complete dryness over the water-bath, water added, and 
the manganese and ferric chlorides filtered off. The residual 
tungstic acid is washed with alcohol, dissolved in ammonia, 
the solution filtered from any residue, evaporated to dryness 
in a capacious porcelain crucible, gently heated to expel 
ammonia, and ignited in contact with air. The tungstic 
oxide is then weighed : it should have a pure yellow colour, 
free from any greenish tint. 

Test the residue for niobic acid"" by heating it in the 
Bunsen flame with microcosmic salt : the oxide of niobium 
forms a colourless bead in the outer flame, but a violet- 
coloured bead inclining to blue in the inner flame if the fused 
salt is saturated with the oxide. By adding a trace of ferrous 
sulphate the colour is changed to blood-red. As very similar 
reactions are afforded by tungstic oxide, care must be taken 
to ensure that the whole of this substance is removed by 
ammonia before the test is made. Another portion may be 
heated with a bead of sodium carbonate, by which it is dis- 
solved, forming whilst hot a transparent mass, which becomes 
turbid on cooling ; if whilst still hot it is moistened with a 
drop of tin chloride, and heated in the lower reducing flame, 
it gives a grey mass, which dissolves in hydrochloric acid, 
producing a light amethystine tint. 

The alcoholic filtrate is evaporated to dryness, the residue 
dissolved in water, and the iron and manganese separated 
as in No. XIX. Part IV. p. 220. 

* Columbite, a niobate of iron and manganese, not unfrequently oc- 
curs associated with wolfram- 



Scheelite : Galena. 255 

XXVIL SCHEELITE (CaWO 4 ). 

This mineral is readily decomposed by strong nitric acid. 
The solution is evaporated nearly to dryness, alcohol added, 
and the liquid filtered. Tungstic oxide is left undissolved : 
calcium nitrate goes into solution. The oxide is treated in 
the manner above described : the solution is evaporated 
nearly to dryness, water added, and the lime estimated in 
the usual manner. 

XXVIII. GALENA. 

This substance is essentially lead sulphide, but it is almost 
invariably mixed with more or less iron, copper, silver, anti- 
mony, and zinc, and silicious matter (gangue). 

Determination of the Sulphur. The ore is reduced to the 
finest powder and dried at 100. About i gram of the sub- 
stance is weighed out into a large porcelain crucible, gently 
heated with potash solution (free from sulphate) for an hour, 
and a slow current of chlorine conducted into the liquid. 
The galena by this treatment is decomposed ; the sulphur is 
oxidised to sulphuric acid, which combines with the potash, 
and the lead is converted into binoxide. The liquid is fil- 
tered, acidified with hydrochloric acid, and the sulphur 
precipitated as barium sulphate. 

Determination of the Lead, Iron, Zinc, &>c. About 1*5 to 
2 grams of the ore are oxidised with red fuming nitric acid 
(B.P. 86) in a flask, in the mouth of which is placed 
a small funnel. The sulphur is thus completely oxidised, and 
the galena converted into lead sulphate. Evaporate off the 
excess of acid, and add about 5 cubic centimetres of 
moderately-strong sulphuric acid, and evaporate nearly to 
dryness. Add about 20 cubic centimetres of water, filter, 
wash the residue with water containing sulphuric acid, and 
remove the sulphuric acid by washing with alcohol, otherwise 
the paper will fall to pieces on being dried. Do not mix 
the alcoholic washings with the acid filtrate. The operation 



2 56 Quantitative Chemical Analysis. 

of washing must be done without delay, and with as little 
water as possible, otherwise a perceptible amount of lead sul- 
phate will be dissolved. The residue in the filter is dried 
and ignited in a weighed porcelain crucible ; care must be 
taken to remove as much of the lead sulphate as possible 
from the paper before incineration. The ash may be 
moistened with one drop of nitric acid and then one drop of 
sulphuric acid, and the whole carefully dried, ignited, and 
weighed. The substance consists of lead sulphate mixed 
with sand, silica (gangue). When weighed, it is carefully 
transferred to a small beaker and heated with hydrochloric 
acid, which dissolves the lead sulphate, leaving the silicious 
matter unchanged. Allow the liquid to become clear by 
standing, and pour it through a small filter, again digest with 
hydrochloric acid and again filter, repeating this treatment 
three or four times until the filtrate is no longer blackened 
by sulphuretted hydrogen water. Wash the residue on to 
the filter with hot water, dry and weigh, and subtract the 
weight, minus the filter ash, from the original weighing : the 
difference gives the amount of lead sulphate. 

Pass sulphuretted hydrogen through the filtrate, and deter- 
mine the copper and antimony as in No. XVI. Part IV. 
Iron and zinc are precipitated from the filtrate, after treat- 
ment with sulphuretted hydrogen, by means of ammonium 
sulphide : they may be separated as in No. XVI. Part IV. 

Determination of Silver in Galena. Galena rarely con- 
tains as much as 0*5 per cent, of silver, but this metal can be 
profitably extracted, even when it does not exceed one- 
twentieth of this amount in the ore. The exact determina- 
tion of the small quantities of silver almost universally present 
in galena becomes, therefore, a matter of importance. When 
an argentiferous galena is smelted, the whole of the silver is 
found in the reduced metal. About 50 or 60 grams of the 
finely-divided galena are mixed with twice their weight of 
sodium carbonate and 20 grams of nitre, and placed in a clay 



Silver in Galena. 257 

crucible. A layer of well-dried common salt (about 8 or 10 
millimetres deep) is placed over the mixture, and the crucible 
is heated to bright redness. It is allowed to cool, broken, 
and the button of reduced lead extracted. This is flattened 
on an anvil and freed from slag, &c., by rubbing and washing 
with water. The following equation represents the reaction : 

4Na 2 CO 3 + yPbS = 4?b + 3(PbSNa 2 S) + Na 2 SO 4 + 4CO 2 . 

The object of the nitre is to decompose the double sulphide 
of lead and sodium : the lead is separated, and the sodium 
sulphide oxidised to sulphate. The button of lead (which 
should weigh from 35 to 40 grams, if the operation has been 
properly conducted) is slowly dissolved in pure dilute nitric 
acid until about 5 or 10 grams of the metal remain : this is 
withdrawn from the solution. It contains the whole of the 
silver. It is dissolved in dilute nitric acid, and the solution 
is diluted with a large quantity of water. A few drops of 
highly-dilute hydrochloric acid are added, and the liquid is 
allowed to stand until the silver chloride has completely 
settled : this is filtered off, washed repeatedly with hot 
water, and weighed. (Compare also No. XXXIV. p. 279. 

The object of only partially dissolving the button is to 
avoid the presence of an undue amount of lead in solution, 
since the nitrate of this metal dissolves silver chloride to a 
perceptible extent. The results obtained by this method, if 
properly conducted, are more exact than those given by 
cupellation. 

A very ready method of assaying galena for technical pur- 
poses consists in decomposing the ore by means of zinc and 
hydrochloric acid. About 2 grams of the finely-powdered 
sulphide are weighed out into a tall beaker and covered with 
a piece of pure zinc about an inch in diameter and a quarter 
of an inch thick.* Pour into the beaker about 120 cubic 
centimetres of dilute hydrochloric acid ( i part acid to 4 of 

* Obtained by dropping the molten metal upon a smooth surface of 
wood or metal. 

S 



258 Quantitative Chemical Analysis. 

water), cover the beaker with a watch-glass, and gently heat 
(to 40 or 50) for 15 or 20 minutes, occasionally stirring 
the liquid. When the evolution of sulphuretted hydrogen 
ceases, and the liquid becomes clear, the decomposition is 
complete. The supernatant liquid is poured on to a filter, 
in which a small piece of zinc is placed, and the zinc and 
lead in the beaker washed with hot water by decantation 
until the filtrate has no longer an acid reaction. The lead 
is transferred to a weighed porcelain crucible, any portions 
adhering to the zinc being rubbed off by a glass rod. The 
small particles on the filter are washed into a porcelain basin 
and added to the crucible. The water in the crucible is 
poured away and the lead gently dried in a current of coal 
gas to prevent it from oxidising. If gangue is mixed with 
the reduced lead, its amount may be determined by dis- 
solving the metal in dilute nitric acid, and washing, drying, 
and weighing the insoluble residue. 

XXIX. REFINED LEAD. 

Recent improvements in refining, and the introduction 
of such improved methods of desilverisation as Pattinson's 
crystallisation process or Parkes' zinc process, have so far 
perfected the process of manufacturing softened lead that 
this article seldom contains less than 99-9 per cent, of the pure 
metal. Pure as such lead may appear, the presence in it of 
minute traces of iron, copper, &c., yet exercises a very im- 
portant influence in its application to the manufacture of 
vitriol chambers, evaporating pans, and to the preparation of 
white lead, flint glass, &c. 

The following substances are found in refined lead : silver, 
copper, bismuth, cadmium, zinc, iron, nickel, and antimony. 
Cobalt, arsenic, and manganese are seldom present in esti- 
mable quantity. These metals are derived partly from the 
ores and partly from the employment of Parkes' process, the 
impurities being introduced in the zinc used. 



Refined Lead. 259 

The Method of Analysis* The lead to be analysed is 
cut up into large pieces, and the surface of each piece is 
scraped with a clean bright knife until it is perfectly bright 
and free from any apparent impurity. Weigh off two por- 
tions of 200 grams each of the lead into flasks of 1,500 
cubic centimetres capacity, and add to one portion about 500 
cubic centimetres of pure nitric acid of sp. gr. i -2, and so 
much water that no lead nitrate separates out. The action 
of the acid may be promoted by a gentle heat ; care 
must be taken not to employ a greater excess of nitric acid 
than that given. The solution is allowed to stand from 12 to 
24 hours. Since 200 grams lead give 310 grams of nitrate, 
and i part of lead nitrate requires about 2 parts of water 
for solution, there is no possibility of lead nitrate separating 
out from the solution if it measures about i litre. If a crys- 
tallisation occurs, it is a sign that too great an amount of 
nitric acid has been used, lead nitrate being far more in- 
soluble in dilute nitric acid than in water. 

To the other 200 grams add nitric acid of the above 
strength in small portions at a time, always keeping the 
metal in excess, and heat the liquid until only about 5 or 10 
grams of lead remain undissolved, and the solution com- 
mences to turn yellow in consequence of the formation of 
lead nitrite. In the residual metal the whole of the silver 
is concentrated. It is withdrawn from the liquid, dissolved 
in nitric acid, the solution diluted with a large quantity of 
water, and a few drops of a solution of lead chloride added, 
or i cubic centimetre of hydrochloric acid of 1*12 sp. gr., 
previously diluted with 50 cubic centimetres of water. This 
quantity of acid is more than sufficient to precipitate all the 
silver without throwing any lead chloride out of solution. 
Set the beaker aside for two or three days. Draw off the 
clear liquid by means of a syphon, and collect the precipitate 
on a small filter, wash thoroughly with boiling water, dry it, 

* Fresenius, Zeits. fur anal. Chem. vol. viii. 1869. 
S 2 



260 Quantitative Chemical Analysis. 

and incinerate in a small weighed porcelain crucible. If the 
amount of silver chloride is so considerable that there is a 
possibility of its being incompletely reduced by the combus- 
tion of the filter-paper, the residue must be heated for a few 
minutes in a stream of hydrogen before weighing. The 
amount left, after subtracting the filter ash, gives the quan- 
tity of silver in the 200 grams of lead. The refined metal 
seldom contains more than 0-0015 per cent, of silver. 
(Mean of 12 specimens, 0-0013 P er cent) 

The solution of the other portion of 200 grams is used for the 
estimation of the remaining impurities. As a rule it remains 
perfectly clear even after standing. Occasionally, however, in 
the case of lead rich in antimony, a more or less considerable 
precipitate forms after a time. The precipitate is filtered off 
and set aside ; the mode of examining it will be given here- 
after. Bring the clear solution or filtrate into a 2-litre flask, 
add 115 grams (about 62 or 63 cubic centimetres) pure and 
concentrated sulphuric acid, shake, allow to cool, and fill up 
to the mark ; again shake the liquid so as to mix it thoroughly, 
and allow the precipitate to settle. The amount of sulphuric 
acid to be added is so calculated that about 10 or 12 grams 
are in excess. As soon as the lead sulphate has completely 
settled, the clear liquid is drawn off by means of a syphon 
previously filled with the liquid. In this way rather more 
than 1,750 cubic centimetres may be drawn off. Accurately 
measure off 1,750 cubic centimetres of the solution, and 
evaporate it in a porcelain basin, in a draught-place free 
from dust, until sulphuric acid fumes make their appearance, 
indicating that the nitric acid has been expelled. Allow the 
liquid to cool, add about 60 cubic centimetres of water, and 
filter off the small quantity of lead sulphate which separates 
out, and wash the precipitate with a little water. 

The slight precipitate of lead sulphate frequently retains 
small quantities of antimony. It is therefore dissolved in 
hydrochloric acid, and the solution diluted with sulphuretted 
hydrogen water, the liquid warmed, and sulphuretted hydro- 



Refined Lead. 261 

gen passed through it. The precipitate is allowed to sub- 
side, filtered, washed, the filter-paper spread out in a 
porcelain basin, and heated with a solution of pure am- 
monium or potassium sulphide, to which a small quantity of 
pure sulphur has been added. Filter the solution, wash, 
acidify the filtrate with hydrochloric acid, and allow the 
precipitate to settle at a gentle heat. 

The solution filtered from the lead sulphate, and diluted 
to 200 cubic centimetres, is heated to about 70, and treated 
with sulphuretted hydrogen, allowed to stand 12 hours at 
a gentle heat, and filtered through a small filter. The 
washed precipitate is heated with potassium sulphide solu- 
tion (containing sulphur) in the manner above described. 
The filtrate is acidified with hydrochloric acid and allowed 
to stand until the precipitate has completely subsided. 

The filter and residue insoluble in potassium sulphide are 
heated in the dish nearly to boiling with dilute nitric acid 
(i part acid 1*2 sp. gr. and 2 of water). When the pre- 
cipitate is dissolved, filter, wash the paper slightly, dry, and 
incinerate, and add the ashes to the nitric acid solution ; 
add 2 cubic centimetres of dilute sulphuric acid, and evapo- 
rate until the nitric acid is expelled j dilute with a little 
water, and filter off the small quantity of lead sulphate which 
separates out. Nearly neutralise the filtrate with pure caustic 
potash, add sodium carbonate, and a small quantity of 
potassium cyanide solution free from potassium sulphide, 
and heat gently. If a precipitate is formed, it is filtered off, 
washed, and dissolved in dilute nitric acid, and the bismuth 
precipitated by ammonium carbonate, and weighed as oxide. 
To the filtrate a little more potassium cyanide is added, 
together with a few drops of potassium sulphide. The 
precipitate which ensues contains the cadmium and silver. 
It is filtered off, dissolved in dilute nitric acid, the silver 
precipitated by hydrochloric acid, the filtrate evaporated 
nearly to dryness, and a few drops of sodium carbonate 
solution added. The cadmium precipitate is filtered off, 



262 Quantitative Chemical Analysis. 

dried, ignited, and weighed as oxide. The reduction and 
volatilisation of the cadmium may be prevented by moisten- 
ing the filter with ammonium nitrate. The nitrate from the 
mixed silver and cadmium sulphides is mixed with a small 
quantity of sulphuric and nitric acids, and evaporated nearly 
to dryness ; a few drops of hydrochloric acid are added, and 
the solution heated until the last traces of hydrocyanic acid 
have disappeared ; the solution is filtered, if necessary, and 
the copper precipitated and weighed as sulphide. 

When cadmium is absent, the separation of the copper and 
bismuth may be effected by means of ammonia and am- 
monium carbonate. In this case it must not be forgotten 
that the silver is to be removed by the addition of hydro- 
chloric acid before the copper is precipitated as sulphide. 

The precipitates obtained by acidifying the sulphide of 
ammonium solution are filtered off, washed, dried, and 
repeatedly treated with carbon disulphide to remove the 
sulphur. The little filter and the residue are then together 
warmed with a few drops of red fuming nitric acid, the 
porcelain basin being covered with a watch-glass j the solu- 
tion is heated to expel the excess of nitric acid, sodium car- 
bonate added in slight excess, and then a small quantity of 
sodium nitrate. The solution is evaporated to dryness, and 
carefully heated until the residue melts and the mass be- 
comes white. When cold the fused mass is transferred to 
a small mortar and carefully broken up by rubbing it in a 
little cold water. The solution is filtered and the residue is 
washed with water containing alcohol. The sodium anti- 
moniate is dissolved in hydrochloric acid, to which a little 
tartaric acid has been added, and the solution treated with 
sulphuretted hydrogen, and set aside for a few hours. 

The soluble portion of the fused mass which contains all 
the arsenic, together with a little antimony, is evaporated to 
expel the alcohol, an excess of sulphuric acid is added, and 
the solution again evaporated to expel the nitric acid, water 
added, the liquid heated to 70 C, and treated with sulphu- 



Refined Lead. 263 

retted hydrogen. When the precipitate has settled, it is 
filtered through a small filter and washed with water. It is 
then treated on the filter with a cold concentrated solution 
of ammonium carbonate, the filtrate being repeatedly poured 
back on to the filter, to obviate the use of a large excess of 
ammonium carbonate. The arsenic sulphide is dissolved : 
the antimony sulphide mixed with a little sulphur remains un- 
dissolved. This residue is dissolved in a little strong hydro- 
chloric acid diluted, and treated with sulphuretted hydrogen, 
and the precipitated sulphide added to the main quantity 
obtained from the sodium antimoniate. The antimony sul- 
phide is best filtered through a small tube in the bottom of 
which a little asbestos has been placed. The tube containing 
the asbestos is gently heated over the direct flame and 
weighed. When the whole of the precipitate has been trans- 
ferred to the little tube, it is warmed to expel the greater 
portion of the water, and then gently heated in a stream of 
dried carbon dioxide until the antimony sulphide becomes 
black. The tube is allowed to cool in a current of the gas, 
the carbon dioxide displaced by atmospheric air, and the tube 
and antimony sulphide again weighed. 

The solution of arsenic sulphide and ammonium carbonate 
is acidified with hydrochloric acid, and the turbid solution 
treated with a little sulphuretted hydrogen water, filtered, 
and the arsenic sulphide filtered through a weighed tube 
containing asbestos, and heated in the manner prescribed in 
the case of the antimony sulphide. 

The filtrate and washings from the original precipitate by 
sulphuretted hydrogen are concentrated, poured into a flask, 
rendered alkaline by ammonia, and mixed with ammonium 
sulphide. The flask, which must be at least half full of 
liquid, is well corked and allowed to stand 24 hours. When 
the slight precipitate has subsided the liquid is filtered, the 
filtrate acidified with acetic acid, and evaporated at a gentle 
heat to facilitate the separation of a small quantity of nickel 
sulphide. This, mixed with sulphur, is filtered off, slightly 
washed, and dried and incinerated. 



264 Quantitative Chemical A nalysis. 

The precipitate formed by the ammonium sulphide is 
treated on the filter with a mixture of i part hydrochloric 
acid (sp. gr. 1-12) and 6 parts sulphuretted hydrogen 
water, the filtrate being repeatedly poured back on to the 
filter. The sulphides of iron and zinc are dissolved : the 
nickel and cobalt sulphides remain on the filter. The filter 
is dried, incinerated, and mixed with the ash of the filter 
containing the nickel sulphide derived from the sulphide of 
ammonium solution. The mixed ashes are treated with 
aqua regia, the solution concentrated, a little water added, 
filtered, rendered alkaline by ammonia, a few drops of am- 
monium carbonate added, filtered into a platinum basin and 
treated with a few drops of potash solution until no more 
ammonia is evolved. Filter off the slight flocculent pre- 
cipitate, wash, dry, ignite, and weigh, and test the precipitate 
(which generally consists mainly of nickel oxide) for traces 
of cobalt with the blowpipe. The solution containing the 
iron and zinc, to which a few drops of nitric acid are 
added, is concentrated by evaporation, and rendered alka- 
line by ammonia, the ferric oxide filtered off, again dissolved 
in a few drops of hydrochloric acid, and again precipitated 
by ammonia, filtered, washed, dried, and weighed. By way 
of control the weighed precipitate may be fused with potas- 
sium-hydrogen sulphate, dissolved in water, and reduced 
with zinc, and the iron estimated volumetrically by a dilute 
permanganate solution. 

The filtrate from the ferric hydrate is mixed with a little 
ammonium sulphide, and allowed to stand at least for 24 
hours at a gentle heat. If anything separates out it is 
filtered off, washed, and digested on the filter with dilute 
acetic acid in order to separate any admixed manganese 
sulphide. The residue on the filter is dried, and weighed as 
zinc sulphide. The acetic acid solution is concentrated to a 
few cubic centimetres, and mixed with excess of caustic 
potash to precipitate the manganese. 

Before we can proceed to express the results centesimally 



Refined Lead. 265 

it is necessary to determine the quantity of lead correspond- 
ing to the 1,750 cubic centimetres of solution taken for 
analysis. This can only be estimated when we know the 
volume occupied by the lead sulphate obtained from the 200 
grams of metal, when suspended in water. By repeated 
experiments it has been found that it occupies the space 
of 44*99 grams, or in round numbers, 45 grams of wafer 
at 1 6 C. The 2-litre flask, when filled to the mark, 
holds therefore 1,955 cubic centimetres solution, and 45 
cubic centimetres lead sulphate. But the 1,955 cubic cen- 
timetres of solution were equivalent to 200 grams lead ; 
therefore the 1,750 cubic centimetres would be equal to 
179-03 grams, or in round numbers, 179 grams of the original 
lead. 

The solution of the lead, when containing unusually large 
quantities of antimony, not unfrequently forms, on standing, 
a white precipitate of antimony oxide and antimoniate of anti- 
mony, which occasionally retains arsenic. This precipitate is 
filtered off, washed, and the arsenic and antimony separated 
as above. In calculating the result, it must not be forgotten 
that this precipitate is obtained from 200 grams of the 
metal, whilst the remaining portion in solution is assumed 
to be derived from 179 grams of lead. 

The determination of the minute quantity of sulphur con- 
tained in lead may be effected by heating the metal in 
chlorine gas, when the sulphur is converted into the chloride, 
which, when led into water, is decomposed, with the forma- 
tion of sulphuric acid : this may be precipitated in the 
usual way, and weighed as barium sulphate. The best 
method of carrying out this process is to heat about 100 
grams of the lead in the form of a thick rod about i 
centimetre in diameter in a combustion-tube about i metre 
long. In the middle the tube is narrowed, and the end is 
drawn out and bent downwards, and dips into a small three- 
bulbed U-tube containing water. The lead is placed in the 
anterior portion of the tube ; the other serves to collect the 



266 Quantitative Chemical Analysis. 

lead chloride, which flows over the little bridge as fast as it 
is produced, leaving the metal exposed to the further action 
of the gas. The combustion-tube is connected with a small 
tube containing ignited fragments of charcoal, over which 
the chlorine passes before it comes in contact with the 
heated lead. The charcoal must be kept at a red heat 
throughout the operation : it serves to free the chlorine 
from any trace of accompanying oxygen, which might oxidise 
the lead sulphide to sulphate. Vulcanised stoppers, on 
account of the sulphur they contain, cannot be used to 
connect together the several pieces of the apparatus. Or- 
dinary corks must therefore be employed. When the 
entire apparatus is filled with chlorine the lead is gradually 
melted, care being taken to place the combustion-tube in 
such a position that the metal does not come in contact with 
the cork it should flow against the bridge, but not above it. 
When the heat is properly regulated the lead burns slowly 
to lead chloride, which flows over and collects in the empty 
portion of the tube. When care is taken to regulate the 
stream of chlorine and not to overheat the chloride, but 
little of this substance passes over into the U-tube. The 
solution is washed out of the condensing tube into a small 
beaker, heated, and the sulphur precipitated by the addition 
of a few drops of barium chloride.* The following analysis 
of three specimens of soft lead, executed by the above 
methods, will give some idea of the nature and amount of 
its impurities : the results are represented centesimally : 

0-00385 
00190 

00553 
02639 
00129 



* Bannow and Kramer, Ber. Deutschen Chem. Gesells.' July 1871. 



Silver . 




0-0020 


O '0006 2 


Copper 




00228 


OOO3I 


Cadmium 




trace 


oooio 


Bismuth 




00040 


oooio 


Antimony 




00173 


00186 


Iron . 




00035 


00012 


Zinc . 




00014 


00023 


Sulphur 




00076 


00008 



White Lead. 267 



XXX. WHITE LEAD. 

This substance is a compound of lead carbonate and 
hydrate in variable proportions. In general the relation 
between the hydrate and carbonate may be represented by 
the formula 2PbCO 3 + PbH 2 O 2 , although specimens of the 
composition 3PbCO 3 + PbH 2 O 2 and 5PbCO 3 + PbH 2 O 2 
are occasionally obtained (Mulder, Phillips). As found in 
commerce it is frequently mixed with barium sulphate (heavy 
spar), barium carbonate (witherite), calcium carbonate, and 
zinc oxide. These bodies cannot always be regarded as 
adulterants ; the heavy spar, for example, serves to protect 
the lead from the rapid action of sulphuretted hydrogen, 
and unless present in large excess does not interfere with 
the opacity or body of the pigment. But these substances 
are not unfrequently added in undue quantity ; and perfectly 
pure white lead is now comparatively rare as an article of 
commerce. 

Determination of the Carbon Dioxide. From i to 2 grams 
of the substance are weighed out into the flask A, fig. 31, and 
decomposed by moderately- dilute nitric acid. The operation 
is carried out exactly in the manner described on p. 86. 

On the termination of the experiment the liquid in the 
flask A is filtered if necessary, and the residue washed, dried, 
and weighed : it may consist of the sulphates of barium or cal- 
cium. The weighed residue is boiled in a platinum or porcelain 
basin with solution of pure sodium carbonate for an hour or 
so, care of course being taken to replenish the dish with water 
from time to time. Any calcium sulphate present is com- 
pletely decomposed, and, on pouring the liquid through a 
small filter, the sulphate in solution may be detected by the 
addition of barium chloride. The calcium carbonate formed 
may be dissolved out by dilute hydrochloric acid, and the lime 
precipitated by means of ammonium oxalate. The washed 
precipitate is rendered caustic by ignition and weighed. 



268 Quantitative Chemical Analysis. 

The residual barium sulphate may also be weighed, by way 
of control. 

The nitric acid nitrate containing the lead, &c., is evapo- 
rated nearly to dryness to expel the excess of the acid, 
diluted with water, and the liquid saturated with sulphuretted 
hydrogen. The lead is completely separated ; it is filtered 
off, and converted into sulphate by oxidation with nitric 
acid. The weighed precipitate should then be heated with 
a dilute solution of sodium thiosulphate, which dissolves the 
lead sulphate and leaves unattacked any barium sulphate 
which may be present. To the filtrate from the lead sulphide 
add ammonia and ammonium sulphide ; wash the precipi- 
tated zinc sulphide, dissolve it in dilute hydrochloric acid, and 
re-precipitate as carbonate, and convert into oxide by ignition. 
The filtrate from the zinc sulphide contains the lime and 
baryta : these are separated as in No. VII. Part II. If baryta 
is found no calcium sulphate can be present. If it is desired 
to determine the water directly, this can be effected by means 
of the apparatus shown in fig. 53. 

Arrangement of the Results. Calculate the baryta found 
in the last filtrate to barium carbonate, and the lime to 
sulphate or carbonate according to circumstances. The 
residual amount of carbon dioxide is converted into lead 
carbonate, and the remainder of the lead to lead oxide. 
The water, zinc oxide, and barium sulphate are set down as 
such in the statement of the analysis. 

Zinc white may be also analysed by the foregoing method. 
In addition to the adulterants mentioned above, kaolin is not 
unfrequently met with : this is left undissolved on treating 
the pigment with dilute acid. 

XXXI. CHROME IRON-ORE. 

This mineral occurs massive in various parts of the world, 
particularly in Norway, Siberia, Asia Minor, Silesia, and 
North America ; it consists essentially of a compound of 



Chrome Iron-Ore. 269 

ferrous oxide and chromium sesquioxide. It belongs to the 
spinelle group of minerals, and is isomorphous with magnetic 
oxide of iron. Its formula, FeOCr 2 O 3 , requires 677 per 
cent, of chromium sesquioxide. Usually, however, the chro- 
mium is replaced by aluminium, and the iron by magnesium 
to a considerable extent, and the ore is of very good quality 
when it contains 50 per cent, of chromium sesquioxide. 

Determination of the Chromium Sesquioxide. Grind a few 
grams of the carefully- sampled mineral in an agate mortar, 
and pass the powder through a fine muslin sieve. When 
you have collected about two grams of the fine dust, again 
grind it in portions of a decigram at a time in the agate mortar 
until every feeling of grittiness has disappeared, and the ore 
cakes in an impalpable powder round the pestle. Too much 
care cannot be given to the grinding : the success of the 
analysis entirely depends upon the ore being in the finest 
possible state of division. Weigh out about 0-5 gram of 
the powder into a platinum crucible of about 50 cubic centi- 
metres capacity, and place over it 12 times its weight of 
recently-fused potassium-hydrogen sulphate, and gently heat 
the crucible so as merely to melt the sulphate. Keep it 
melted at a gentle heat for 15 or 20 minutes, and gradually 
increase the temperature until the bottom of the crucible 
becomes red hot. Care must be taken that the fused mass 
does, not rise above half way up the crucible. In a few 
minutes the mixture will fuse quietly and dense fumes of 
sulphuric acid will be evolved; the heat is now gradually 
increased until the crucible is at a bright red heat ; in about 
half an hour remove the lamp, and add about 6 parts of 
powdered sodium carbonate, again fuse the mixture, and, 
keeping the temperature for an hour at a red heat, add little 
by little the same quantity of nitre. The temperatui 
crucible is now increased and kept at a full red JJH&; for 20 
minutes; it is allowed to cool, transferred to a porcelain basin, 
and the mass boiled out with water. The solution is filtered, 



270 Quantitative Chemical Analysis. 

and the residue washed with hot water until the filtrate comes 
through colourless. It is not necessary to transfer the whole 
of the insoluble matter to the filter. Quickly dry the filter 
and its contents, detach the ferric oxide and return it to the 
porcelain basin, burn the filter and add the ash, and digest 
the whole at a gentle heat with moderately-concentrated 
hydrochloric acid. If the fusion has been properly con- 
ducted the residue will dissolve ; any black insoluble matter 
left undissolved denotes that the grinding has not been done 
with sufficient care. This insoluble portion must be collected 
on a small filter, dried, and the filter, &c., folded up, thrown 
into the platinum crucible, ignited, and the residue again 
fused with potassium bisulphate, sodium carbonate, and nitre, 
and the fused mass again boiled out with water, filtered, and 
the filtrate added to the main solution. A few grams of 
ammonium nitrate are added to the total filtrate, and the 
liquid is evaporated nearly to dryness, water is added, and 
the solution is filtered from the alumina, silica, &c. The fil- 
trate, which should be received in a porcelain basin,* is then 
made strongly acid with sulphurous acid, and boiled until 
. the gas is nearly expelled, a slight excess of ammonia is 
added, and the solution is again boiled until it becomes 
colourless. Pour the liquid on to a filter, wash the precipi- 
tate by decantation with hot water, and by means of a feather 
transfer it to the filter, and wash carefully with hot water. 
After the sixth washing allow the filter and precipitate to 
drain thoroughly by keeping up the action of the pump for 
about ten minutes, remove the filter, fold it, and, without 
further drying, transfer it to a weighed platinum crucible 
and cautiously heat with the lid on. Gradually increase the 
heat, and as soon as the paper is charred, remove the lid, 
placing it at the edge of the crucible (see fig. 15), and ignite 




tiling with ammonia is conducted in glass vessels there is 
great probabfll^ that the precipitated chromic oxide will be contaminated 
with silica. Trie error from this cause may amount to 0-5 per cent. 
(Compare Souchay, Fres., 'Zeits.' iv. 66.) 



Chrome Iron Ore. 271 

strongly for 10 or 15 minutes. If the precipitate has been 
drained sufficiently by the action of the pump, there is no 
danger of the oxide being projected from the crucible on 
ignition. On treating the weighed precipitate with a few 
drops of water the liquid ought to remain colourless : a yellow 
colour indicates that the oxide has been imperfectly washed 
from alkaline salts. 

Complete analysis of chrome iron ore (Dittmar's process). 
Fuse together a mixture of equal weights of pure borax glass 
and sodium potassium carbonate in a platinum basin, and 
break up the fused mass when cold. 10 grams of this 
mixture are placed in a platinum crucible of about 50 c.c. 
capacity, fused, and allowed to cool. About i gram of the 
finely powdered ore is added, and the mass again fused, at 
first with the lid on : the crucible is now inclined, and the 
contents are stirred with a stout platinum wire. In about 
15 minutes the whole of the ore should be dissolved. The 
lid is now placed in the position seen in fig. 15, the fused 
mass being occasionally stirred to promote its oxidation : 
this is usually complete in about half an hour. Allow to 
cool, and add about 5 grams of pure potassium carbonate, 
again fuse, and pour out the melted mass into a platinum 
basin, which should be quickly covered with a clock glass 
as the solidifying substance decrepitates. The portion ad- 
hering to the crucible is dissolved off by hot water in a 
porcelain basin, and the solidified mass in the dish is added 
to the solution. Add a few drops of alcohol to reduce any 
manganate which may be present, and digest on a water- 
bath until the whole is disintegrated and the alcohol ex- 
pelled. Filter and wash the precipitate until the filtrate is 
colourless, and dissolve the precipitate in hot dilute sul- 
phuric acid to ascertain if any ore is left undecomposed. 
If any be found it must be treated again with a small quan- 
tity of the fusion mixture, and the mass dissolved out with 
water as before. The filtrate is concentrated to about 
200 c.c., and divided by weighing or measuring into two 



2 7 2 Quantitative Chemical A nalysis. 

approximately equal portions. To the one portion is added 
a quantity of dilute sulphuric acid and a known weight (in 
excess) of pure ferrous sulphate, and the amount of un- 
oxidised iron determined by standard potassium bichromate 
solution as directed on p. 221. 

Cr 2 O 3 : 6Fe = 152-2 : 336 = -4529 : i. 

If only the amount of chromium is required, the determina- 
tion may be repeated on the second portion. To determine 
the amount of silica and alumina, add ammonium nitrate to 
the solution, evaporate to dryness, add water, and filter off 
the alumina and silica. 

The sulphuric acid solution is treated with sulphuretted 
hydrogen, filtered, and the filtrate concentrated to a small 
bulk, and poured into an excess of strong and fresh potash 
solution contained in a platinum basin, filtered, and the 
alumina and small quantity of silica precipitated by boiling 
with ammonium chloride : the precipitate is added to that 
obtained as above, and the substances are separated as in 
an ordinary silicate analysis. The separation of iron,, lime, 
and magnesia is effected as described on p. 87. 

XXXII. SMALTINE: COBALT-GLANCE. 

Smaltine or tin-white cobalt is an arsenide of cobalt : 
cobalt-glance is essentially a sulpharseniate of cobalt. Both 
minerals frequently contain, in addition to cobalt, arsenic, 
and sulphur, variable quantities of nickel, iron, lead, bismuth, 
copper and gangue. 

Determination of 'Sulphur. See Copper Pyrites, p. 211. 

Determination of Silica and the Metals. Treat 2 grams of 
the finely-divided substance with strong nitric acid, and 
evaporate to dryness with sulphuric acid (hydrochloric acid 
is inadmissible, since a small quantity of the arsenic would 
be volatilised on heating). Moisten the dried mass with 
sulphuric acid and add water. On standing, the supernatant 
liquid ought to become quite clear : a turbidity indicates the 



Smaltine : Cobalt-Glance. 273 

presence of basic salts. Filter the liquid into a flask, wash 
the precipitate slightly with water containing sulphuric acid, 
remove the acid from the paper by alcohol, dry, and weigh ; 
the insoluble matter consists of silica, and calcium and lead 
sulphates. Transfer the weighed precipitate to a small beaker 
and boil with dilute nitric acid : the sulphates are thus dis- 
solved. Throw the residual silica on a filter, wash, dry, and 
weigh. The nitrate containing the lead and calcium is evapo- 
rated to dry ness with hydrochloric acid, water added, the 
solution boiled, and the lead precipitated with sulphuretted 
hydrogen and converted into sulphate by treatment with 
strong nitric acid. The lead sulphate is weighed : its weight, 
plus that of the silica, subtracted from the original weight of 
the precipitate (SiO 2 + PbSO 4 + CaSO 4 ), gives the cal- 
cium sulphate. 

To the filtrate from the insoluble residue (SiO 2 .PbSO 4 . 
CaSO 4 ) a strong solution of sulphurous acid is added in 
small quantities at a time, and the liquid boiled after each 
addition. The liquid is maintained at a temperature of 
60 or 70 by surrounding it with warm water, and the 
metals precipitated by sulphuretted hydrogen. Allow the 
solution to stand for some time, so that the greater portion 
of the sulphuretted hydrogen may be removed by diffusion, 
and filter the liquid. Wash and drain the filter thoroughly 
by the action of the pump, spread it on a glass plate, and 
remove the arsenic sulphide by means of a thick platinum 
wire, transferring it to a large porcelain crucible. Fold the 
filter and replace it in the funnel, and dissolve the small 
quantity of adhering sulphide by a few drops of strong 
potash solution, and allow the liquid to fall into the porcelain 
crucible. A little more potash is added to the crucible until 
the arsenic sulphide is dissolved. A stream of chlorine is 
then passed into the solution until it is perfectly colourless : 
it is then gently heated for some time, and filtered from the 
small quantities of the oxides of copper and bismuth. Dis- 
solve the residue in nitric acid, add an excess of ammonium 
carbonate, heat gently for some time\ filter, redissolve in 

T 



2 74 Quantitative Chemical Analysis. 

nitric acid, and again precipitate the bismuth by the addition 
of ammonium carbonate in excess, dry, and weigh the 
bismuth as trioxide. Mix the two filtrates and precipitate 
the copper by means of caustic potash. 

The filtrate containing the arseniate of potassium is acidi- 
fied with hydrochloric acid, and ammonia, and * magnesium 
mixture' added. The magnesium-ammonium arseniate may 
be obtained highly crystalline (in which state it washes 
better), by first adding the magnesia mixture to the acid 
liquid, and then a large quantity of ammonia about one- 
fourth of the bulk of the entire liquid. After standing 24 
hours, pour the liquid on to a weighed filter, wash the pre- 
cipitate with ammonia water until the washings acidified 
with nkric acid give only a slight opalescence with silver 
nitrate. The magnesium-ammonium arseniate is dried in 
the air-bath at 120, and ignited at a low red heat in a 
weighed porcelain crucible. The ignited residue has the 
composition Mg 2 As 2 O 7 , corresponding ta magnesium pyro- 
phosphate. 

To the rose-coloured filtrate containing the cobalt, nickel, 
and iron, add nitric acid, and boil, and then solid sodium 
carbonate, until the liquid is nearly neutralised, and becomes 
slightly turbid, owing to the separation of ferric hydrate : 
complete the precipitation by adding a solution of succinate 
of soda or ammonia, wash and dry the precipitate, ignite it, 
and weigh it as Fe 2 O 3 . 

The nickel and cobalt are co-precipitated by adding 
potash to the boiling solution ; the precipitate is filtered off 
and washed,, drained as far as practicable by the pump, the 
filter spread on a plate, and the precipitate, detached as far 
as possible by means of a thick platinum wire, transferred to 
a small porcelain dish : the filter is re-folded and dropped 
back again into the funnel. Treat the precipitate in the dish 
with dilute hydrocyanic acid, and then with potash solution, 
and again with hydrocyanic acid, and warm on the water- 
bath, until no further solution occurs. A minute quantity 
of substance frequently remains undissolved : it consists of 



Smaltine : Cobalt- Glance. 



275 



paracyanogen, and retains a small quantity of the mixed 
oxides. This is added to the small quantity remaining on 
the filter, which is now washed, dried, ignited, and weighed. 
The proportion of the two metals contained in it is calculated 
from the results of the after-separation. The solution of 
nickel and cobalt is boiled to expel the excess of acid : it is 
reddish-yellow, and consists now of cobalticyanide of potas- 
sium and double cyanide of nickel and potassium. Add to 
the hot solution precipitated (yellow) mercuric oxide, and 

FIG 64. 




again boil for some time. All the nickel is precipitated, 
partly as cyanide, partly as sesquioxide, the mercury com- 
bining with the liberated cyanogen. Wash the precipitate 
thoroughly, dry it, and heat it for some time in a covered 
porcelain crucible to expel the excess of mercuric oxide. 
The filtrate is nearly neutralised with nitric acid, and a neutral 
solution of mercurous nitrate is added in excess, when 
cobalticyanide of mercury is precipitated ; this is washed, 
dried, and ignited, in an open porcelain crucible, until the 
weight is constant : it has the composition Co 3 O 4 . 



276 Quantitative Chemical Analysis. 

By way of control it may be reduced to the metallic state, 
by heating in a stream of dry hydrogen. When the decom- 
position is apparently finished, allow the metal to cool in a 
current of the gas and weigh : again heat in hydrogen and 
again weigh, repeating the operation until the weight is con- 
stant. The apparatus employed for the reduction is seen 
in fig. 64. The crucible lid is pierced with a hole and pro- 
vided with a porcelain tube. 



XXXIII. FAHL-ORE (TETRAHEDRITE) 

Consists of a mixture of sulphantimonites and sulpharsenites 
of copper, silver, iron, zinc, and mercury. Every specimen 
of the oreY however, does not contain all these substances : in 
some the^silver, zinc, and mercury are entirely absent ; mothers 
the whole ofMhe arsenic is replaced by antimony. The com- 
position of the mineral maybe represented by the general 

formula ^ 3 : N "' 4 \ S 7 , in which M denotes Cu and Ag, 

oD2 I ASa-* 

and N" denotes Fe, Zn, and Hg. 

Determination of the Sulphur. See Copper Pyrites, p. 209. 

Determination of the Metals. The decomposition of the 
finely-powdered mineral is best accomplished by heating it 
in a stream of dry chlorine. The apparatus required for 
this purpose is seen in fig. 64A. The substance is placed in 
the bulb-tube a-, the second bulb serves to collect the 
greater portion of the sublimate, and thus prevents the 
narrower portion from being stopped up by the chlorides. 
The mineral is best weighed out from a tube of a diameter 
sufficiently narrow to allow of its being introduced into the 
bulb-tube. The bulb-tube is connected with the U-tube b, 
which is filled to the extent of ^ of the two upper bulbs with 
a mixture of hydrochloric and tartaric acids. The flask A 
contains the chlorine mixture (i part of salt, i of manganese 
dioxide, i of water, and 2 of sulphuric acid : the sulphuric acid 



Fdhl-Ore. 



277 



and water are previously mixed, and allowed to cool) ; the 
small flask c contains strong sulphuric acid, and the U-tube d 
sulphuric acid and pumice. Allow this portion of the ap- 
paratus to become rilled with chlorine before joining it to 
the bulb-tube, and do not heat the mineral until the gas is 
apparently without further action upon it. Indeed as soon 
as the chlorine comes in contact with the fahl-ore immediate 
signs of decomposition ensue, and the bulb becomes very 
hot. When it cools, gently heat it in order to drive off the 

FIG. 64 A. 




volatile products into the second bulb. Care must be taken 
to send only a gentle stream of chlorine through the appara- 
tus, otherwise there is danger of a portion of the sublimate 
passing through the liquid unabsorbed. As soon as reddish 
vapours of ferric chloride make their appearance, discontinue 
the heating, and allow the apparatus to cool. Divide the 
bulb-tube a by means of a file between the two bulbs, and 
allow the portion containing the sublimate to remain in a 
damp place for 24 hours, in order that it may take up 



278 Quantitative Chemical Analysis. 

moisture from the atmosphere : this prevents the evolution 
of the great heat and consequent possibility of loss by vola- 
tilisation, when the chlorides are subsequently treated with 
water. In the meantime proceed with the analysis of the 
fixed residue contained in the first bulb ; this contains the 
silver, copper, zinc, and a portion of the iron. It is placed 
in a beaker and digested with dilute hydrochloric acid until 
everything is dissolved, with the exception of the silver chlo- 
ride. This is filtered off, washed with hot water, dried, and 
weighed. The copper in the filtrate is precipitated by sul- 
phuretted hydrogen, re-dissolved in nitric acid, and weighed 
as oxide. The solution containing the zinc and iron is 
boiled with nitric acid for a short time, and the metals 
separated as in No. XIII. Part II. 

The solution of hydrochloric and tartaric acids is poured 
into a beaker, the tube rinsed out, and the bulb containing 
the sublimate added. This solution contains the arsenic, anti- 
mony, mercury, and the remainder of the iron ; if it is cloudy, 
from the separation of a little antimony, warm \\. gently, some- 
times the turbidity is due to the separation of sulphur ; in that 
case the liquid should be filtered. The solution is heated 
to about 60, and a current of sulphuretted hydrogen is 
passed through it for some time until the fluid smells strongly 
of the gas. It is allowed to stand for 12 or 15 hours, filtered, 
and washed with sulphuretted hydrogen water. The re- 
mainder of the iron is found in the filtrate ; it is precipitated 
by ammonium sulphide, washed with sulphuretted hydrogen 
water, re-dissolved in hydrochloric acid, the solution boiled 
with a little nitric acid or a few crystals of potassium chlo- 
rate, and the iron re-precipitated as ferric oxide by ammonia, 
washed, dried, and weighed. 

The mixed sulphides of antimony, arsenic, and mercury 
are treated with potassium sulphide until the residue of 
mercuric sulphide is quite black. This is filtered off through 
a weighed filter, washed with water containing potassium 
sulphide, then .with pure water, two or three times with 
alcohol, and finally with bisulphide of carbon, until a drop 



Silver : Pisani's Method. 279 

of the filtrate evaporated on a watch-glass leaves no residue. 
The filter is once more dried at 100 and weighed. 

The sulphide of potassium solution is transferred to a 
porcelain crucible and saturated with chlorine gas. The 
solution is heated on a water-bath, and mixed with a con- 
siderable excess of hydrochloric acid and evaporated to 
about half its bulk ; an equal volume of hydrochloric acid 
is again added to the solution, again evaporated to one-half. 
A freshly prepared solution of sulphuretted hydrogen is next 
added (in the proportion of TOO c.c. of solution to each 'i 
gram of antimonic acid supposed to be present), when anti- 
mony pentasulphide separates out. The excess of sulphu- 
retted hydrogen is quickly expelled by blowing a stream of 
air through the liquid, and the precipitate is brought upon a 
weighed filter, drained by means of the pump, and washed 
six or eight times with alcohol, then with carbon bisulphide, 
and again with alcohol. The antimony pentasulphide is 
dried at 110, and weighed. 

The aqueous filtrate is mixed with a few drops of chlorine 
water, heated on the water- bath, and treated with a stream 
of sulphuretted hydrogen, and after standing for 24 hours in 
a warm place the precipitated arsenic pentasulphide is 
transferred to a weighed filter, and washed with alcohol and 
carbon bisulphide, as in the case of the antimony sulphide, 
and treated at 110 until its weight is constant. 

Bournonite(2PbSCuS.Sb 2 S 3 ); Boulangerite (3PbS.Sb 2 S 3 ); 
red silver ore, 3Ag 2 SAs 2 (Sb 2 )S3, and nickelspeiss may also 
be analysed by decomposition in a stream of chlorine. 

XXXIV. DETERMINATION OF SILVER IN SOLUTIONS. 

An excess of a standard solution of sodium chloride is 
mixed with a determinate volume of the silver solution to be 
tested, a few drops of potassium chromate solution are added, 
and the excess of the chlorine in solution is determined by 
adding a solution of silver of known strength until the orange 
colour of the silver chromate is persistent (compare No. I. 



280 Quantitative Chemical Analysis. 

Part III.). This method is especially applicable to the deter- 
mination of the strength of the silver solutions employed in 
photography. 

Pisanfs Method (particularly applicable to the Estimation of 
Silver in Alloys and Ores). 

If a solution of the blue compound of iodine and starch 
be added to a neutral liquid containing silver nitrate, the 
colour is rapidly destroyed, the iodine combining with the 
silver to form silver iodide (and iodate ?). Immediately the 
silver is completely precipitated the iodised starch solution 
colours the liquid permanently blue, and thus marks the 
completion of the process. 

This method is, of course, only accurate in the absence of 
substances, other than silver, which effect decomposition of 
the blue solution. Tin, arsenic, and antimony, mercury, iron, 
and manganese protoxides and gold must accordingly be 
absent. Copper and lead do not influence the reaction. 

Weigh out about 2 grams of iodine and 15 grams of pure 
starch into a porcelain mortar, add 6 or 8 drops of water, 
and mix intimately ; transfer the mass to a flask, and heat it, 
well closed, on the water-bath for an hour. The violet-blue 
colour of the mixture will now have changed to dark greyish- 
blue. Dissolve in water and dilute considerably. 

To ascertain the value of the deep bluish-black solution, 
transfer 10 c.c. of a neutral solution of silver nitrate, contain- 
ing i gram of silver per litre, to a beaker, add a small quan- 
tity of pure precipitated calcium carbonate, and run in the 
solution of iodised starch, with constant stirring, until the 
yellow colour of the liquid (due to the silver iodide formed) 
changes to greenish - blue. The 10 c.c. of silver solution 
should require about 50 c.c. of the iodised starch solution. 
The object of the calcium carbonate is to neutralise the 
nitric acid liberated: it also renders the completion of the 
reaction more distinctly visible. 

The minute quantities of silver contained in lead ores and 



Silver Assay. 281 

in refined lead may be readily estimated by this method. 
The nitric acid solution, prepared as directed on p. 259, 
is mixed with sulphuric acid to remove the lead, the 
liquid is filtered, mixed with calcium carbonate in excess, 
and again filtered. A small additional quantity of the cal- 
cium carbonate is added to the filtrate, which is titrated with 
the standardised iodine solution in the manner above de- 
scribed. 

A determinate volume of the standard iodine solution, 
p. 156, mixed with clear starch liquor, and diluted to centi- 
normal strength, may be employed with equally good results. 
(Field.) 

XXXV. ASSAY OF SILVER IN BULLION, COIN, PLATE, &c. 

The method about to be described is that generally prac- 
tised in the European mints : it was originally devised by 
Gay-Lussac, and has been rigidly investigated by Mulder, 
Probably no quantitative process is susceptible of such a 
high degree of accuracy as the estimation of silver by the 
* humid method,' as the process of Gay-Lussac is generally 
termed, in contradistinction to the older process of cupella- 
tion. 

We have already indicated the leading features of this 
method in describing the process for exactly estimating the 
strength of a standard hydrochloric acid solution (p. 131). 
This very simple process is, however, complicated by the 
following circumstance : If we add i eq. of silver nitrate to 
i eq. of sodium chloride, both dissolved in water, we should 
expect that all the silver would be precipitated, and that we 
should obtain no subsequent turbidity by the further addi- 
tion either of salt or of silver solution. But in reality we 
find that the addition of either of the solutions produces a 
turbidity. This remarkable fact probably depends upon 
the solvent action of the sodium nitrate, and upon the ex- 
istence of a certain degree of equilibrium between the silver 
nitrate and common salt, which is destroyed, with the im- 



282 Quantitative Chemical Analysis. 

mediate formation of silver chloride, by the addition of either 
of the bodies. 

It thus happens that if we add a decimal solution of salt 
(vide infra), drop by drop, to a solution of silver until no 
further turbidity is produced, and then add decimal solution 
of silver to the liquid, we again notice the formation of a 
slight precipitate. If we continue to add the decimal silver 
solution until the turbidity ceases, and once more add 
decimal salt solution, we shall again observe the formation 
of another precipitate. If we determine the number of drops 
required to pass from one limit to the other, we observe that 
the same number of each is needed. Suppose that we had 
added in the first place the salt solution until the exact point 
was reached at which no further turbidity was produced, and 
that we required to add 20 drops of the silver solution before 
the precipitate again ceased to form, we should find that it 
would be necessary to add 20 drops of the salt solution be- 
fore this point of the non-formation of a turbidity was again 
reached. If we add exactly half this number of drops, viz., 
10, we reach what Mulder terms the critical point, that is, the 
point at which both salt and silver produce equal pre- 
cipitates. 

We have, therefore, three different methods of determining 
the completion of the reaction : a. We may add the salt 
solution until the turbidity just ceases ; /3. We may stop at 
the neutral point ; or y. We may go back with silver solution, 
and continue the addition until no further turbidity is pro- 
duced. Whichever method we adopt in standardising the 
solution of salt we must afterwards invariably employ. Thus 
we must not at one time end with salt and at another end 
with silver. Mulder has shown that the error by such a 
procedure amounts to i milligram per gram of silver : by 
employing first a and then /3 the difference amounts to 0-5 
milligram at 16 C. Mulder has also shown that the degree 
of error varies slightly with the temperature and dilution of 
the liquids. 



Silver Assay. 283 

On the whole it is most convenient to adopt the first plan 
i.e. to continue the addition of the salt until no further pre- 
cipitate is formed. We require for this method : 

1. Solution of Sodium Chloride. Dissolve 5*4145 grams of 
salt in distilled water, and dilute to i litre. The temperature 
of the solution should be 16 when measured. 100 c.c. 
of this solution == i gram of silver. Call it * Salt Solution 
No. i.' 

2. Decimal Solution of Sodium Chloride. Transfer 50 c.c. 
of the above solution to a J-litre flask, and dilute to the con- 
taining-mark with water at 16. Call it ' Salt Solution No. 2.' 

3. Decimal Silver Nitrate Solution. Weigh out exactly 
o'5 gram of pure silver (see p. 124) into a J-litre flask, dis- 
solve in 3 c.c. of pure nitric acid, and dilute with water at 
1 6 to the containing-mark. i c.c. contains i milligram of 
silver. 

4. Test Bottles. The bottles specially made for the method 
should be procured. They are of white glass, about 250 c,c. 
in capacity, and are fitted with accurately- ground stoppers, 
the lower portion of which is pointed. On touching the 
side of the bottle with the point of the moistened stopper, the 
adhering liquid is readily detached. The bottles are placed 
in well-fitting cases of cardboard or vulcanite, and when in 
use are wrapped in a black cloth in order to protect the silver 
chloride from the light. 

We commence the process by determining the exact value 
of the salt solution. Weigh out with the greatest possible 
accuracy from i-ooi to 1-003 gram of pure silver, place it in 
a test bottle, add 5 c.c. of pure nitric acid (sp. gr. 1-2), and 
heat the bottle (placed obliquely) on the water-bath until the 
silver is dissolved. From time to time blow out the nitrous 
fumes from the bottle, occasionally shaking the liquid to 
promote their expulsion. When solution is effected, allow 
the bottle to cool for a short time and place it in water at a 
temperature of about 16. Remove it in about half an hour, 
wipe it, and place it in its case. Transfer 100 c.c. of salt 



284 Quantitative Chemical Analysis. 

solution No. i, measured with the greatest care, to the bottle. 
Moisten the glass stopper with distilled water, insert it firmly 
in the neck, cover the bottle with the black cloth, and shake 
the whole violently for a minute or two, or until the silver 
chloride settles completely, leaving the fluid perfectly clear. 
Take out the stopper, rub it on the bottle to remove the 
adherent silver chloride, replace it, and shake down the 
silver chloride on the sides of the glass by giving the liquid 
a rotatory motion. As soon as the chloride is deposited, 
again open the bottle, incline it, and allow \ c.c. of decimal 
salt solution (Sol. No. 2) to flow in against the lower part of 
the neck of the bottle. The salt solution should be added 
from a Mohr's burette, graduated into OT c.c., and fitted 
with a glass stopcock. After the addition of the i c.c. of salt 
solution, raise the bottle from its case, and note the degree 
of turbidity, insert the stopper, and shake until the liquid 
is again quite clear. Repeat the agitation after each addition 
of the salt solution, and, as the turbidity decreases, add the 
solution in very small quantities : towards the end only two 
drops should be added at a time. At this point read off the 
burette, and continue the addition of the salt solution by two 
drops at a time, reading off the burette after each addition, 
and agitating the liquid until no further precipitate is pro^ 
duced. When the last two drops fail to produce a turbidity 
the process is at an end. The previous reading that is, the 
one before the addition of the last two drops is taken as the 
correct one. 

If by mischance the exact point of the non- formation of a 
turbidity has been overstepped, add 2 c.c. of decimal silver 
solution, shake, and continue the addition of the salt solution 
until the proper point is again reached. 

To take the most complicated case : let us suppose that 
we have weighed off 1*0023 gram of pure silver, added 
100 c.c. of salt solution No. i, and 4*2 c.c. of the decimal salt 
solution. We have reason to believe that we have over- 
stepped the proper point, and we therefore add 2 c.c. of 



Silver Assay. 285 

decimal silver solution, and again 1*8 c.c. of decimal salt, 
when turbidity ceases. 

Amount of silver taken = 1-0023 + -0020 = 1-0043 gram. 

This required 100 c.c. of No. i salt solution + 6 c.c. of 
No. 2 salt solution, or altogether 100-6 c.c. of No. i 
solution for complete precipitation. Calculate the quantity 
required for i gram of silver : 

1-0043 " 1*000 100*6 = 100-17. 

We thus find that 100*17 c.c. of the salt solution No. i ex- 
actly precipitates i gram of pure silver. It is desirable from 
time to time to repeat the determination of the strength of 
the salt solution. 

For the actual assay, weigh out about i -085 gram of standard 
silver ( 1 2*3 of silver to i of copper) in the test bottle, dissolve 
in 5 or 6 c.c. of nitric acid, add 100 c.c. of the salt solution 
No. i, and proceed with the addition of the decimal salt 
solution as directed. Let us assume that we had weighed 
out 1*085 g ram > an d tnat it was necessary to add 6 c.c. of 
decimal silver solution before turbidity ceased : this, with- 
out sensible error, we may assume to be equal to 0*6 of 
solution No. i. We thus find that 100*6 c.c. of the strong 
solution were needed to precipitate all the silver in the 1*085 
gram of the alloy. But 100*17 c - c - of the solution were 
equal to i gram of pure silver. Therefore 

100*17 : roo'6 :: i = 1*0043; 

and accordingly 1,000 parts of the alloy would contain 
1*085 io II 1*0043 = 925*6 parts of pure silver. 

In the case of alloys of which the composition is not 
approximately known it is necessary to determine it by a 
preliminary trial before the regular assay is made. Weigh 
off from 0*5 to i gram of the alloy according to its richness 
in silver, dissolve in 4 or 5 c.c. of nitric acid in the usual 
manner, and add sodium chloride solution No. i from a 



286 Quantitative Chemical Analysis. 

burette, until no further precipitate is formed. Then calculate 
the amount of the alloy which will contain 1*002 gram of 
silver, and proceed with the assay in the manner directed. 

In the case of alloys containing sulphur and gold, digest 
with the least possible quantity of nitric acid, blow out the 
nitrous fumes from the bottle, add strong sulphuric acid, and 
boil until the gold separates completely, and proceed with 
the assay in the usual way. 

XXXVI. ASSAY OF GOLD. 

For a description of the most accurate methods of assaying 
alloys of gold by cupellation, we would refer the student to 
Professor Jevons' excellent article on the subject in Watts' 
' Dictionary of Chemistry,' vol. ii. p. 932. Longmans : 1869. 

XXXVII. SEPARATION OF GOLD, SILVER, AND COPPER. 

I. When the amount of silver in the alloy does not exceed 
15 per cent, the whole of the gold and copper maybe dissolved 
out by means of nitro-hydrochloric acid, the silver remaining 
as silver chloride. The solution of the finely-divided alloy 
after complete decomposition is evaporated nearly to dryness 
to expel the greater portion of the free acid, water is added, 
the solution filtered, and the silver chloride is washed, dried, 
and weighed. To the solution oxalic acid is added, whereby 
the gold is completely reduced to the metallic state. After 
standing for about 48 hours, the liquid is filtered, and the 
gold washed, dried, and weighed. The copper in solution 
is precipitated as sulphide by means of sulphuretted hydro- 
gen : it may either be weighed as such, after covering it with 
sulphur and heating in a stream of sulphuretted hydrogen, 
or be converted into oxide by re-solution and precipitation 
with sodium hydrate. 

II. If, on the other hand, the amount of gold in the alloy 
does not exceed 15 per cent., the whole of the silver is dis- 
solved by prolonged boiling with moderately-concentrated 



Mercury. 287 

nitric acid.* The liquid is evaporated nearly to dryness, 
water added, the silver precipitated by hydrochloric acid, 
and the copper in the filtrate by sulphuretted hydrogen. 
The gold after weighing is dissolved in cold nitro-hydro- 
chloric acid, to ascertain that it is perfectly free from silver. 
If any silver chloride is found it must of course be filtered 
off and weighed. 

III. By heating alloys of gold, silver, and copper with 
strong sulphuric acid, the metals may be separated, whatever 
may be their proportion. The finely-divided alloy is heated 
with the concentrated acid until no more sulphur dioxide is 
evolved, and the acid begins to volatilise. Water is then 
added and the liquid boiled, and the silver and copper 
separated as above. It is advisable to treat the gold again 
with the acid before finally weighing it : if any additional 
silver and copper are obtained their weight must of course 
be added to the main quantities of these metals. 

XXXVIII. ESTIMATION OF MERCURY. 
In ores and compounds containing mercury the amount 
of the metal may be readily determined by heating the sub- 
stance with quicklime. The process is conducted in the 
apparatus represented in fig. 65. Into a piece of combus- 
tion tube, about 50 centimetres long and rounded at the 
end, introduce a layer, 5 centimetres in length, of powdered 
magnesite (a). Weigh out about 5 grams of the substance 
into a glazed porcelain mortar and mix it intimately with 
quicklime. Introduce the mixture into the tube (a to b\ 
and rinse out the mortar with a fresh portion of quicklime 
(b to c) ; fill up the rest of the tube to within a few centi- 
metres of the end with powdered quicklime (c to d) t and draw 
out the tube before the blow-pipe in the manner represented 
in the cut. Gently tap the tube so as to make a channel for 

* The nitric acid must not be too strong, otherwise more than traces 
of the gold would be dissolved by the nitrous acid formed. 



288 Quantitative Chemical Analysis. 

the gas, and place it in the combustion furnace, and immerse 
the drawn-out end just beneath the surface of water contained 
in a small flask. Heat the tube from d to <r, at first gently, 
and then to bright redness, and gradually heat the portion of 



FIG. 65. 
t> c 



the tube containing the mercurial compound to redness. 
When the substance is completely decomposed, heat the 
magnesite to expel the traces of mercurial vapour within the 
tube. Whilst the tube is still red hot, cut off the tube at *, 
and wash any adhering mercury into the flask. Agitate the 
mercury beneath the surface of the water so as to bring it 
together into one globule, and in about half an hour decant the 
clear water ; pour the mercury on to a small weighed watch- 
glass, remove the water as far as practicable by filter-paper, 
and before re-weighing it place the watch-glass and metal 
for an hour or two beneath a bell-jar containing strong sul- 
phuric acid. The only mercurial compound which thus 
resists complete decomposition is the iodide ; it may be 
readily analysed, however, by substituting copper filings for 
the lime. Cinnabar may be readily analysed by heating it 
with nitric acid of sp. gr. i -4 in a closed tube for a couple 
of hours : the sulphur is converted into sulphuric acid, which 
may be determined as barium sulphate. Silica, heavy spar, 
&c., remain undissolved. The mercury is precipitated as 
calomel by phosphorous acid as described below. 

In the case of very poor ores of cinnabar containing large 



Coal. 289 

quantities of bituminous matter, such as those of Austria, 
the foregoing process may be judiciously modified by first 
extracting the organic matter from the powdered ore by means 
of benzol, thoroughly drying the residue, and treating it with 
a solution of barium sulphide (containing about 50 grams of 
barium to the litre). The mercuric sulphide is precipitated 
from the solution by means of hydrochloric acid, filtered, 
dried, and digested with carbon bisulphide to remove ad- 
mixed sulphur. The impure mercuric sulphide is then 
dried, re-dissolved in aqua regia, the liquid concentrated, 
and mixed with solution of phosphorous acid, allowed to 
stand about 12 hours in the cold, and the precipitated calo- 
mel collected on a weighed filter, washed with hot water, 
and dried at 100. 

Mercury may be readily determined by electrolysis. The 
solution of the metal is poured into a weighed platinum dish 
and slightly acidified with sulphuric acid, and the dish is 
connected with the zinc pole of a battery of six bichromate 
cells, the carbon end being attached to a piece of platinum 
foil which dips into the solution. Mercurous chloride 
gradually separates, and this is finally reduced to the metal, 
and in about an hour all the mercury is precipitated. The 
metal is washed with water, alcohol, and ether, and dried 
over oil of vitriol and weighed. 

[Note on the Preparation of Phosphorous Acid. Small pieces of 
phosphorus are introduced into a saturated solution of copper sulphate 
contained in a corked flask ; metallic copper is first reduced, which is 
ultimately converted into black copper phosphide, and the acid solution 
consists of a mixture of phosphorous and sulphuric acids. The latter 
may be removed by the cautious addition of baryta water. The pre- 
cipitate is allowed to settle, and the clear solution of phosphorous acid 
decanted and preserved in a well-stoppered bottle, as it oxidises on 
exposure to air. It is a very powerful reducing agent, and may be use- 
fully applied in a variety of cases.] 

XXXIX. COAL. 

In the proximate analysis of coal we require to determine 
the amount of moisture, volatile matter, coke, ash, and 

u 



290 Quantitative Chemical A nalysis. 

sulphur. If the actual amount of carbon, hydrogen, and 
nitrogen is needed, recourse must be had to elementary 
organic analysis. 

Determination of the Moisture. About 2 grams of the 
finely-powdered coal are weighed out and dried between 
watch-glasses at 105-110 for an hour, and the loss is set 
down as moisture. With the quantity of coal taken the loss 
of weight appears to be greatest at the end of this time : on 
further heating, it actually increases in weight, an effect due 
probably to the oxidation of the finely- divided pyrites. 

Determination of the Volatile Matter. About 2 grams of 
the powdered undried coal are heated for four minutes over 
a Bunsen flame, and then immediately, without cooling, for 
the same length of time over the gas blow-pipe flame. The 
loss is set down as volatile matter + moisture. The residue 
gives the coke + ash. 

Determination of the Ash. From 3 to 5 grams of the 
finely- divided coal are heated in a small platinum dish over 
a Bunsen lamp. Usually the incineration proceeds with 
rapidity : if it is found necessary to increase the draught of 
air over the heated mass, the arrangement described in the 
section on Analysis of Ashes of Plants may be employed. 

Determination of the Stilphur. The sulphur in coal exists 
in two modifications as pyrites and as calcium sulphate. 
The sulphur contained in the pyrites alone influences the 
economical application of the fuel. 

The total amount of sulphur may be determined by heat- 
ing about 2 grams of the powdered coal with four times its 
weight of pure sodium carbonate in a platinum dish, or the 
coal may be heated in a current of oxygen and the gases 
passed through a solution of hydrochloric acid and bromine, 
the sulphuric acid being determined by precipitation with 
barium chloride. The sulphur existing as calcium sulphate 
may be determined by boiling 5 grams of the finely-powdered 
coal with a solution containing about the same weight of pure 
sodium carbonate, whereby the calcium sulphate is decom- 



Water Analysis. 291 

posed, calcium carbonate and sodium sulphate being formed. 
Filter the solution, acidify with hydrochloric acid, and add 
barium chloride. The difference between the total amount 
of sulphur and that found after boiling with sodium car- 
bonate represents the amount as pyrites. The same process 
is of course applicable to the determination of the iron 
sulphide and gypsum in coke. 

Determination of Specific Gravity. It is occasionally 
desirable to ascertain the weight of a cubic foot of the fuel, 
or the number of cubic feet corresponding to a ton. This is 
easily calculated when the specific gravity of the coal is 
known. The specific gravity is readily determined by weigh- 
ing the coal in air ; and in water by suspending it from the 
arm of the balance by a hair or thin wire. The piece taken 
should not be too small, and care should be taken that no 
air bubbles adhere to it during the weighing. It is desirable, 
too, that the coal be soaked sufficiently : this is easily effected 
by immersing the lump, after attaching the hair or wire to 
it, in water, in the flask of the filter-pump, and exhausting 
the air within the apparatus as far as practicable. The 
weight of a cubic foot of the coal in pounds is found from 
the expression : 

log. sp. gr. + i'79588=log. wt. of cb. foot. 
The number of cubic feet in a ton 

==1 '55437 -log- sp. gr.=log. cb. ft. 

XL. EXAMINATION OF WATER USED FOR ECONOMIC 
AND TECHNICAL PURPOSES. 

i. Collection of the Sample. The water to be analysed 
should be collected in stoppered glass bottles those known 
as ' Winchester Quarts,' which hold about 2 J litres, may be 
conveniently employed. As a rule two of these bottles will 
contain sufficient water for an ordinary examination. If, 
however, an exhaustive analysis is required, double or even 
treble this amount may be necessary. Care must be taken 

u 2 



292 Quantitative Chemical Analysis. 

that the bottles, and the vessels employed to fill them, are 
quite clean, and a due amount of judgment should be used to 
obtain a representative sample. In collecting the water from 
a river or tank the bottles themselves should be immersed be- 
low the surface, and rinsed once or twice with the water. In 
taking the water from a pump or pipe a considerable quan- 
tity should be allowed to flow away before the sample is 
collected. The bottles should be filled up nearly to the neck 
and the stopper tied down with string : no luting or sealing- 
wax should be used. 

Glass bottles are preferable to stoneware jars, for the 
reason that earthenware is not readily cleaned ; moreover it is 
liable to affect the hardness of the water, as the clay not un- 
frequently contains notable quantities of calcium sulphate. 

2. Preliminary Observations. Fill a tall narrow cylinder 
of white glass with the water to be examined, and compare 
its colour with that of distilled water contained in a similar 
cylinder. Heat a portion to about 30 in a wide test-tube, 
shake, and note if the water possesses any peculiar odour or 
taste. 

In the outset the analyst must decide whether the water for 
analysis is to be filtered or not. His decision will depend upon 
the manner in which the water is used by the consumer. If it 
be considered necessary to filter the sample, care must be 
taken that the paper employed is free from ammonia. It 
should be steeped in distilled water for some time before 
use, dried and folded, and heated in a weighed tube 
for some hours at 120. It is placed in the desiccator 
and weighed when perfectly cold. It is now properly fitted 
into the funnel, and the filter-flask is replaced by a clean 
* Winchester Quart,' in which the filtrate is directly received. 
The quantity of the water to be filtered is measured ; as soon 
as the whole has passed through, the funnel is removed from 
the bottle and washed with distilled water (the washings 
must not be allowed to mix with the filtered water), again 



Water Analysis. 293 

dried at 120 for some hours in the stoppered tube, and again 
weighed. The increase in the weight gives the quantity of 
total suspended matter in the known volume of the water. 
Incinerate the paper in a small platinum crucible, treat the 
residue with a few drops of solution of ammonium carbonate, 
dry and weigh ; the quantity in excess of that contained in the 
filter gives the amount of suspended inorganic matter in the 
water. 

If it is considered unnecessary to filter the water, care 
should be taken to shake the bottle before withdrawing 
portions for analysis. 

3. Estimation of the Ammonia. It is desirable to proceed 
at once with the determination of this constituent, since it is 
the most liable to change. The method of estimation is 
based upon the fact that an alkaline solution of mercuric 
iodide added to a liquid containing ammonia produces a 
brown colouration, due to the formation of the iodide of 
tetramercurammonium. This test, known as Nessler's, is 
capable of detecting i part of ammonia in 20,000,000 parts 
of water. 

Preparation of the Nessler Test. Dissolve 35 grms. of 
potassium iodide in 120 c.c. of water, transfer 5 c.c. of the 
solution to a clean beaker, and add, little by little, a cold con- 
centrated solution of mercuric chloride to the remainder 
until the mercuric iodide ceases to be re-dissolved on stirring. 
Add the 5 c.c. of the potassium iodide to re-dissolve the 
remaining mercuric iodide, and cautiously continue the ad- 
dition of the corrosive sublimate solution until a very slight 
precipitate only remains. Now add an aqueous solution of 
potash, prepared by dissolving 100 grams of 'stick 5 potash 
in 200 c.c. of water, and dilute the mixture to 500 c.c. The 
liquid should be allowed to stand for a short time, and a 
portion decanted into a small bottle for use. As the small 
bottle becomes empty it is replenished from the other. In 
addition we require : 



294 Quantitative Chemical Analysis. 

(a) A Standard Solution of Ammonium Chloride. Dis- 
solve 07867 grm. of pure ammonium chloride in a litre of dis- 
tilled water. Pour it into a clean stoppered bottle, and label 
it c Ammonium Chloride Solution No. i.' Withdraw 100 
c.c. of this solution and dilute it also to i litre. Call it 
'Ammonium Chloride Solution No. 2.' i c.c. of the latter 
solution contains -025 of a milligram of ammonia. The solu- 
tion should be delivered from a Mohr's burette, fitted with 
glass stopcock, and graduated to tenths of a cubic centi- 
metre. 

(/3) A small Pipette to deliver about i c.c. of the Nessler 
Test. This may readily be made from a short piece of glass 
tube. Also several cylinders, marked A, B, c, D, &c., 
about 20 cm. in height and of 60 c.c. capacity. To graduate 
them, transfer 50 c.c. of water to each, and mark the level of 
the liquid on the glass. Also two or three pieces of thin glass 
tube, about 30 cm. in length, and 3 mm. in external diameter, 
the ends of which should be blown into bulbs of such 
diameter that they will readily pass into the cylinders ; the 
other ends are sealed. 

(y) Distilled Water free from Ammonia. The distilled 
water of the laboratory must be tested for ammonia. Rinse 
one of the cylinders with the water and fill it up nearly 
to the top ; add i c.c. of the clear Nessler solution, and 
agitate with the bulb-tube (/3) by drawing it up and down a 
few times within the cylinder. If, after standing for five 
minutes, the water remains perfectly uncoloured, it may be 
considered free from ammonia. If it shows a yellow or brown 
tint, re-distil it after addition of about i gram of pure sodium 
carbonate ; collect the distillate in a Winchester quart, as 
soon as 50 c.c. received in one of the cylinders gives no 
reaction for ammonia on testing with the Nessler solution. 
If ordinary water is used, the distillation must not be carried 
to dryness, and the water remaining in the retort or boiler 
must be thrown away before a fresh quantity is distilled. 



Water Analysis. 295 

TJee Process. Transfer 50 c.c, of the natural water to be 
tested to one of the glass cylinders standing on a sheet of 
white paper, add i c.c. of the Nessler solution and agitate 
with the bulb-tube. Run 50 c.c. of distilled water into a 
second cylinder, add *2 c.c. of Ammonium Chloride Solution 
No. 2, mix thoroughly, and compare the tints in the two 
cylinders. If they are about equal in intensity, take half a 
litre for the estimation : if the coloration in the natural 
water is the more intense, take a proportionately less quantity. 
This testing is simply preliminary : its object is to afford an 
idea of the proper quantity to take for the actual estimation. 
Observe whether the natural water becomes turbid after the 
addition of the Nessler test : a decided precipitate is due to 
lime -or magnesia salts, and indicates hardness. 500 c.c. of 
the water, or a less quantity if the preliminary testing has 
shown that ammonia is present in considerable amount, are 
placed in a capacious retort, and connected with a Liebig's 
condenser, which should be freed from ammonia by previously 
blowing steam through it for a few minutes. If less than 500 
c.c. of water have been taken, the liquid in the retort should be 
made up to this volume by the addition of pure distilled 
water before the distillation is commenced. Add about i 
gram of recently-heated and pure sodium carbonate, note if 
much precipitate is formed, and distil rapidly over the direct 
flame. Collect 50 c.c. of the distillate in one of the cylinders, 
A ; when filled replace it by a second cylinder, B. When 
the second cylinder is full, remove the lamp ; add i c.c. of 
Nessler's solution to the second 50 c.c. of the distillate (i.e. B), 
agitate, and place it on a sheet of white paper. Now fill up 
a third cylinder, z, with pure distilled water to within a few 
centimetres of the level of that in B, add as much standard 
Ammonium Chloride Solution No. 2 as you think will pro- 
duce the same depth of colour as in B, and afterwards i c.c. 
of Nessler's solution : add a little distilled water if necessary, 
so as to make the two levels in the tubes coincident. 
Agitate and compare the tints. If the colour of the liquid in 



296 Quantitative Chemical Analysis. 

the two cylinders, after standing about 5 minutes, is equal, 
we at once know the amount of ammonia contained in B : 
it is equal to that contained in the volume of standard am- 
monium chloride solution added to z. If the intensity in z is 
not equal to that in B, pour away the contents of the former 
cylinder, rinse it, fill it with a second portion of distilled 
water, add more or less ammonium chloride solution, as the 
case may be, and i c.c. of the Nessler test.* 

If the quantity of ammonia in B does not exceed -01 of a 
milligram (equal to 0-4 c.c. of the standard ammonium 
chloride solution), the distillation may be discontinued : if 
the amount is greater than this, the boiling must be renewed, 
and successive portions of 50 c.c. of the distillate tested until 
the above limit is reached. If the quantity of ammonia in B 
does not exceed that corresponding to 0-8 c.c. of the 
standard solution of ammonium chloride, the amount in A 
may be determined in the manner directed ; if the quantity is 
greater than this, 25 c.c., or less if need be, of the solution must 
be transferred to another cylinder, diluted to 50 c.c. with 
pure distilled water and tested as above. A colour produced 
by more than 4 c.c. of the ammonium chloride solution can- 
not be conveniently compared, since the liquid is apt to 
become turbid. In the case of waters known to contain 
much ammonia, as in sewage, distil over 100 c.c. at once 
into a larger cylinder, withdraw an aliquot portion, dilute to 
50 c.c., and titrate in the manner directed. 

The determination of the ammonia in water used for 
drinking is of great importance, since an undue proportion 
of this substance denotes contamination with sewage. 
Sewage may contain from 2 to 10 parts of ammonia in 
100,000 parts of liquid : river- waters maybe said to contain 
on the average about o-oi part, although this amount is 



* The addition of more ammonium chloride solution after the Nessler 
test has been mixed with the liquid would cause a turbidity, which pre- 
vents accurate comparison, 



Water Analysis. 297 

subject to great variation. Bad well-waters sometimes con- 
tain as much as 0-5 to i part in 100,000 parts. 

Estimation of ' Albuminoid Ammonia? Messrs. Wanklyn 
and Chapman have found that many nitrogenous organic 
substances yield either the whole or a definite portion of 
their nitrogen in the form of ammonia when boiled with an 
alkaline solution of potassium permanganate. Hippuric 
acid parts with all its nitrogen as ammonia when thus treated : 
whereas albumen gives up only 10 per cent, of ammonia, 
uric acid 7 per cent, and creatine 12-6 per cent. Since it is 
highly probable that the azotised organic matter contained 
in water is of an albuminoid nature, its quantity may be 
approximately estimated by determining the quantity of 
ammonia yielded by boiling the water with an alkaline solu- 
tion of potassium permanganate ; according to Messrs. 
Wanklyn and Chapman, 'the disintegrating animal refuse' 
in the water 'would be pretty fairly measured by ten times the 
albuminoid ammonia which it yields.' * 

The liquid remaining in the retort after distillation with 
sodium carbonate (Estimation of Ammonia) is mixed with 50 
c.c. of a solution obtained by dissolving 8 grams of potassium 
permanganate and 200 grams of potassium hydrate in i litre 
of distilled water. The mixture should be boiled for some 
time previous to use and preserved in a well-stoppered 
bottle. After the addition of the permanganate solution, 
heat the retort over the naked flame, and distil successive 
portions of 50 c.c., and determine the quantity of ammonia 
present in them with the Nessler solution. As soon as the 
distillate contains less than '01 milligram of ammonia, the 
process may be considered at an end. Add together the 
several quantities of ammonia obtained. The succussive 
ebullitions occasionally noticed in bad water may be dimin- 
ished by throwing a number of recently-ignited pieces of 
pumice into the liquid. 

* Wanklyn and Chapman, ' Water Analysis,' p. 68. 



298 Quantitative Chemical Analysis. 

Estimation of Organic Carbon and Nitrogen (Frankland 
and Armstrongs Process). 

Drs. Frankland and Armstrong have proposed to estimate 
the carbon and nitrogen contained in the organic matter 
present in water by direct combustion. The water is 
evaporated to dryness, and the residue is mixed with cupric 
oxide and burnt as in the elementary analysis of an organic 
compound. The resultant gas is collected over mercury, 
and the proportion of carbon dioxide and nitrogen deter- 
mined by gasometric analysis. The process occupies con- 
siderable time, but if the evaporation of the water be 
commenced as soon as the ammonia-determination has been 
made, the residue will be ready for the combustion (which 
occupies about an hour) by the time that the hardness, total 
soluble matter, nitrates, &c., have been estimated. 

i. Evaporation of the Water. The quantity of the water 
needed for analysis will depend upon its quality.. If less than 
0*05 part of ammonia in 100,000 parts of water has been 
found, a litre should be taken ; if more than 0-05, but less 
than 0-2, half a litre will suffice; if more than 0-2, and less 
than i -o, a quarter of a litre should be used. Of sewage, which 
is much richer in organic carbon and nitrogen, 100 c.c., or 
even less, may be taken. 

Transfer the measured quantity of the water to a large 
flask, add to it 20 c.c. of a saturated solution of washed 
sulphurous acid, and, if it does not exceed 250 c.c., boil the 
mixture for a few seconds to expel the carbon dioxide pre- 
sent. Transfer the water, little by little, to a hemispherical 
glass dish, 10 centimetres in diameter, and shaped somewhat 
like a finger-bowl : during the evaporation the glass dish 
should be supported in a copper basin provided with a pro- 
jecting flange and resting on the water-bath, and over it 
should be placed a glass shade, about 1 2 in. high (such as 
is used for covering statuettes). The steam condenses in 
the inside of the shade and flows down into the copper dish, 
filling the space between the two dishes. The excess of 



Water Analysis. 299 

water flows out by a small lip on the edge of the copper dish, 
and is led off by a piece of tape. The destruction of the 
nitrates and nitrites by the sulphur dioxide may be greatly 
accelerated by the addition of 2 or 3 drops of ferrous chloride 
solution (prepared by dissolving well-washed ferrous hydrate 
precipitated from ferrous sulphate by pure soda solution in 
the minimum quantity of pure hydrochloric acid) to the first 
dishful of the water \ if the water is free from carbonates it 
will be necessary also to add i or 2 c.c. of a solution of 
sodium bisulphite in order to combine with the sulphuric acid 
formed, which if free would decompose the organic matter on 
concentration. If, however, the water in the glass dish or 
flask ceases at any time during the progress of the evapora- 
tion to smell of sulphur dioxide, more of the solution should 
be added. If the water is found to contain much nitric 
acid it may be necessary to digest the residue with ,a dilute 
solution of sulphur dioxide, and again evaporate to dry- 
ness, to ensure the complete elimination of the inorganic 
nitrogen. 

2. Combustion of the Residue. Introduce a small quantity 
of cupric oxide in fine powder (made by oxidising the metal 
in air*) into the dish, and mix it thoroughly with the residue 
by the aid of a small steel spatula : this should be very 
elastic, so that by accommodating itself to the curvature of 
the dish the dried residue may be completely detached from 
the glass. Fill about 3 centimetres of the carefully-cleaned 
combustion tube (which should be about 40 centimetres in 
length and i centimetre in internal diameter, sealed at 
one end and rounded like a test-tube) with pure copper 
oxide, and transfer the whole of the mixture in the glass dish 
to the tube, rinsing the dish with small successive portions of 

* The copper oxide may be prepared by cutting copper wire or sheet 
into small pieces, washing the metal, and heating it in a muffle. The oxide 
obtained by strongly heating the copper nitrate cannot be well employed, 
as it is very difficult to expel the last traces of nitrogen from it. The 
cupric oxide remaining in the tube after the combustion (with the excep- 
tion of that with which the substance was mixed) may be used again 
after it has been re-heated in a current of air. 



30O Quantitative Chemical Analysis. 



FIG. 66.' 







7 




FIG. 67. 





Water A nalysis. 30 1 

pure cupric oxide in fine powder. Add copper oxide to the 
tube until it is a little more than half-filled, insert a cylinder 
of metallic copper, about 8 centimetres in length, made by 
wrapping fine copper gauze round a piece of thick copper 
wire,* and then add 2 centimetres of granular copper 
oxide in order to oxidise any carbon monoxide which 
might be formed on burning. The end of the tube is 
softened in the flame of the blow-pipe and drawn out to 
form a tube about 150 millimetres long and 4 millimetres in 
diameter. Bend the tube at right angles, fuse the edges in 
the flame, place it in the combustion furnace and attach it 
to the Sprengel pump. 

Fig. 66 shows the arrangement of this apparatus as applied 
to the purpose of exhausting the combustion tube, a is a 
glass funnel maintained full of mercury, and connected by 
means of a short piece of caoutchouc tube, on which is a 
screw clamp, />, with a long narrow tube which passes nearly to 
the bottom of a wider tube, d, 90 centimetres in length and 
about i centimetre in internal diameter ; the upper end of 
d is connected with a glass funnel in the manner represented 
in the figure, d is connected with the tube/^ by a piece 
of strong caoutchouc tube covered with tape and provided 
with a screw clamp. The tube fg is about 6 millimetres in 
diameter and 600 millimetres in length; it is attached to a 
tube,gVz k, about 1,500 millimetres long and 6 millimetres in 
external diameter, but with a bore of only i millimetre. The 
portion g h is about 20 centimetres long ; the portion h k is 
about 130 centimetres. To give them stability the tubes are 
fastened together by caoutchouc and copper wire. At the 
upper portion of the bend is a tube, h /, about 1 2 centimetres 
long and 5 millimetres in diameter. The combustion tube 

* The cylinder must be previously heated in the lamp, so as to 
oxidise it superficially : it is then placed in a tube and heated in a 
current of hydrogen, in order to reduce the oxide formed. The 
cylinder must be allowed to cool in the gas before it is withdrawn from 
the tube. 



302 Quantitative Chemical Analysis. 

o is connected with the tube h I by the tube / m n, 
of the same diameter as the tube h k. The tube / m n is 
connected with the tube h I and with the combustion tube 
by short lengths of well-fitting caoutchouc tube ; the joint 
at / is bound round with copper wire, and is surrounded 
with glycerine, contained in the wider tube supported by a 
cork on h I; the joint at n is in like manner surrounded by 
a wider tube filled with water. On the tube lmn'\<$> a small 
bulb, which is immersed in cold water during the combus- 
tion ; its object is to receive the greater portion of the water 
formed on burning the residue. The tube h k is re-curved at 
k, where it ends in the mercury-trough /. The trough 
p (shown in plan at B, fig. 67) is cut out of a solid 
piece of mahogany. It is 20 centimetres long, 15-5 centi- 
metres wide, and 10 centimetres deep, outside measurement. 
The edge rr is 13 millimetres wide, and the shelf s is 65 
millimetres wide, 174 millimetres long, and 50 millimetres 
deep from the top of the trough. The channel / is 25 milli- 
metres wide, and 75 millimetres deep ; at one end of it is a 
circular well, w, 42 millimetres in diameter, and 90 milli- 
metres deep. The recesses u u receive the re-curved ends 
of the Sprengel pumps : the object of having two recesses is 
to allow of two experiments being made simultaneously : 
each recess is 40 millimetres long, 25 millimetres wide, and 
75 millimetres deep. The tubes destined to receive the gases 
are supported against the iron wires v v. The trough stands 
upon four short legs, and has a side tube and clamp, ^, to 
draw off the mercury to the level of the shelf s when 
necessary. 

When everything is arranged, heat the fore part of the 
tube containing the metallic copper and unmixed copper 
oxide, and allow a gentle stream of mercury to flow from 
the funnel a : on reaching h the metal passes down the tube 
h k in detached portions, each carrying before it a small 
quantity of air from the combustion tube. Care must be 
taken so to control the flow of mercury that it does not rise 
into the tube / m n. The bulb on the tube I m n is sur- 



Water Analysis. 

FIG. 6a 



303 




304 Quantitative Chemical Analysis. 

rounded by hot water during the exhaustion in order to 
expel any moisture which may remain in it from a previous 
experiment. If the fall is properly regulated the exhaustion 
will be complete in about ten minutes, when the mercury 
will be heard to fall with a sharp clicking sound. The 
action of the pump is now arrested ; a small tube filled with 
mercury is inverted over the end, k, of the tube, the 
hot water in x is replaced by cold water, and the rest of the 
tube is gradually heated to redness. In about an hour the 
combustion will be terminated. The pump is again set in 
operation and the gases are transferred to the tube. 

Measurement and Analysis of the Gases. The gases pro- 
duced in the combustion consist of carbon dioxide, nitrogen, 
nitrogen dioxide, and occasionally, if the operation has been 
conducted too rapidly, sulphur dioxide and carbon mon- 
oxide. Their measurement and analysis may be conveni- 
ently made in the apparatus seen in fig. 68, which is essen- 
tially that devised by Frankland for the separation of gases 
incidental to water- analysis, a c d is the measuring tube : 
the portion a is about 370 millimetres long and 18 milli- 
metres wide, c is 40 millimetres long and 7 millimetres 
wide, and d is 175 millimetres long and 2-5 millimetres 
in diameter. To the upper end of d is attached a tube 
with capillary bore, bent at right angles and provided 
with a stopcock, /. The measuring tube is graduated from 
below upwards at intervals of 10 millimetres, the zero being 
about 100 millimetres from the lower end. The upper por- 
tion of d is divided into millimetres. Attached to the tube 
and stopcock/, is a steel cap, shown on a larger scale at B, 
fig. 68. The lower portion of a is drawn out until it is only 
about 5 millimetres wide : the tube b, which is about i -2 
metre long and 6 millimetres in internal diameter, is also 
narrowed at the lower end. Both a and b pass through the 

* In newer forms of the apparatus Frankland has dispensed with the 
steel caps : the tube from the laboratory vessel being fitted by a cap of 
un vulcanised caoutchouc into a cup- shaped vessel attached to the ca- 
pillary tube of the measuring apparatus. 



Water Analysis. 



305 



caoutchouc stopper o, which is fitted into the glass cylinder 
n n, which is filled with cold water with the view of giving a 
definite temperature to the enclosed gas : this temperature 
is ascertained by a thermometer, e, suspended by a hook 
from the edge of the cylinder. Uniformity in the tempera- 
ture of the mass of the water may be ensured by agitating it 
with an iron wire, the end of which is bent in the form of a 
ring. The tube b is graduated into millimetres, the zero 
being about 10 millimetres above the stopper o, and on a 
level with that of a c d. The rubes a and b are supported 
by the wooden clamp p (seen in end elevation and plan at 
B and c); the clamp is drawn together by two screws, the 
tubes being covered with caoutchouc where they fit into the 
holes to protect them from breakage. The clamp is sup- 
ported by an upright piece of wood (seen in B) which is 
screwed into the base : a and b are connected by tubes 
of caoutchouc, covered with tape and bound with wire, to 
the tube ^, which is also connected with the long caout- 
chouc tube leading to the glass reservoir /. This tube, 
which should have an internal diameter of not more than 2 
millimetres, passes through the steel clamp r, the lower por- 
tion of which is fixed into /. The reservoir t is suspended 
by a cord passing over pulleys, in the arm of the iron rod s. 
On releasing the loop on the cord from the hook v, the re- 
servoir sinks from about 10 centimetres above the level of 
the stopcock / to the level of the bottom of the clamp /. 
In the jar >, termed the laboratory vessel, the gases are 

FIG. 69. 

2> & E fl ' 



subjected to the action of absorbents. It is 100 millimetres 
high and 40 millimetres in internal diameter, and is fur- 

x 



306 Quantitative Chemical Analysis. 

nished with a capillary tube, glass stopcock, and steel cap, 
g h, exactly like fg. The mercury trough /, seen in plan 
in D and in section in E in fig. 69, is cut out of a solid piece of 
mahogany : it is 265 millimetres long, 80 millimetres broad, 
and 90 millimetres deep, outside measurement. The rim 
a a is 8 millimetres broad and 15 millimetres deep. The 
channel b is 230 millimetres long, 26 millimetres broad, 
and 65 millimetres deep. In the larger excavation at the 
end of the channel is placed the laboratory vessel : it is 45 
millimetres in diameter and 20 millimetres in depth below 
the top of b. The small cavities c c are to receive the 
capsule employed to transfer the tube containing the gases 
from the trough of the Sprengel pump. The trough / rests 
on a telescope-table, and its height is so adjusted that when 
the laboratory vessel is placed in the cavity the faces of the 
steel caps are in exact coincidence. (Fig. 68.) 

Before using the instrument the ' corrections for capillarity ' 
must be determined. When the mercury in the tube b and 
in the measuring tube a c d is freely exposed to atmospheric 
pressure, it will be noticed that the levels of the metal in the 
two tubes are not coincident ; the level in b is slightly higher 
than in a ; on the other hand the level in c and d will be 
found to be higher than that in b. The difference in each 
portion should be determined by taking several observations : 
the correction will also include the error arising from dif- 
ference of level in the zeros of the graduations of b and a c d. 
The determination of the levels should be made by the aid 
of a telescope sliding on a vertical rod. 

To determine the capacity of the measuring tube at each 
graduation, first fill the entire tube with mercury, so that the 
metal drips from the cap g. Close the stopcock/, and slip 
a piece of caoutchouc tube over the cap : attach the other 
end of the tube to a funnel filled with distilled water ; lower 
the reservoir /, and open the clamp r and the stopcock/. 
As the mercury flows into the reservoir, water is drawn through 



Water Analysis. 307 

the capillary tube. As soon as it is below the zero on a, close 
/, remove the caoutchouc tube from the cap and slightly grease 
it, to allow water to pass through it without adhering. Raise 
the reservoir, open/, and expel the water until the upper por- 
tion of the mercury meniscus is coincident with the zero of the 
graduation. Now allow the water to flow out into a small tared 
flask until the level of the mercury is coincident with the 
next graduation, controlling the influx of the mercury by 
the clamp r. Read off the temperature of the water in the 
cylinder , weigh the water in the flask, and calculate its 
volume from Table II. in the Appendix. Repeat the deter- 
mination between successive graduations on the whole length 
of the tube in exactly the same manner. 

A table, showing the capacity of the tube at various points, 
is then constructed, the intermediate graduations being ob- 
tained by interpolation : as the calculations are much facili- 
tated by the use of logarithms, it will be found more con- 
venient to set down in this table the logarithms of the capa- 
cities in place of the natural numbers. 

To use the apparatus, grease the stopcocks/ and h and 
the faces of the caps g and g' with a little resin cerate mixed 
with oil. Fill a c d with mercury and close /. Place the 
laboratory vessel in its cavity, and suck out the air as far as 
practicable by the aid of a caoutchouc tube, which is re- 
moved as soon as the jar is filled. Any remaining air may be 
drawn away by aspirating at g'. Close h, and fasten the 
faces of the caps tightly together by the aid of the clamp A. 
Of course the entire apparatus must be quite free from air, 
FlG- 70 . and on opening the stopcocks the 

mercury should flow freely through 
the capillary tubes. To determine 
if the several joints of the apparatus 
are air-tight, close h, and lower the 
reservoir, until it is on a level with^. 
Since the stopcocks and joinings are 
thus .subjected to a pressure of nearly half an atmo* 

X 2 




308 Quantitative Chemical Analysis. 

sphere, any imperfection which may cause leakage will be 
readily detected. After the trial replace t in its original 
position. 

The gas to be analysed is decanted into the laboratory 
vessel and treated with one or two drops of strong potassium 
bichromate solution, to ascertain if it is free from sulphur 
dioxide. If this gas is absent the colour of the solution will 
be unaltered; if present a portion of the chromic trioxide 
will be reduced, and the liquid will become green. If any 
change is observed, pass up a few more drops of the solution, 
to complete the absorption of the gas. Open the stopcocks 
and lower the reservoir, and transfer the gas to the measuring 
tube ; close h so soon as the liquid in the laboratory vessel 
is within 10 mm. from the stopcock. The quantity of gas 
remaining in the capillary tube is too minute to affect the 
experiment. The apex of the mercury meniscus in a c d (as 
seen through the telescope) is made to coincide with the 
nearest division on the tube by allowing mercury to flow in 
from the reservoir /. Read off the levels of the mercury in 
tubes b and a c d ; note the temperature of the water in n, 
together with the height of the barometer. 

Pass a few drops of strong potash solution by means of a 
pipette into the laboratory vessel, and return the gas to it. 
The absorption of the carbon dioxide will be complete in 
about five minutes. The gas now consists of nitrogen mixed 
with a small quantity of nitric oxide ; it is again brought into 
the measuring tube, and its volume is ascertained in the 
same manner as before. If the volume of the gas is very 
small it is possible that it may already contain oxygen ; if so, 
any nitric oxide which might have been formed will have been 
converted into nitrogen tetroxide, which will have been ab- 
sorbed by the potash solution. The volume of gas abstracted 
in this case is, however, too small to affect the result. To 
ascertain if oxygen is present, pass up a small quantity of a 
cold saturated aqueous solution of pyrogallic acid into the 
jar, and by gently shaking the stand of the trough, throw the 




Water Analysis. 309 

liquid up against the sides of the jar in order to promote 
the absorption. As soon as the liquid runs down from the 

glass without the forma- 

FlG. 71. 

tion of a dark red stain, 
the absorption of the 
oxygen is complete. If 
oxygen be absent, it will 
be necessary to intro- 
duce a few bubbles of 
that gas in order to ox- 
idise the nitrogen dioxide which may be mixed with the 
nitrogen. The addition of the oxygen may be conveniently 
made from the pipette shown in fig. 71. The bulbs #and b 
are about 5 cm. in diameter ; the neck joining them is 
narrowed, so that mercury flows through it but slowly. To 
use the instrument fill the bulb b and the tubes d and c 
with mercury ; introduce the tube ^into a small tube contain- 
ing oxygen, standing over the mercury trough, and gently 
aspirate by the aid of the caoutchouc tube e. A few bubbles 
are readily drawn over into b, and the gas is confined by the 
mercury in d and c \ on introducing the limb d beneath the 
edge of the laboratory jar and gently blowing through ^, the 
oxygen may be transferred to the gas under examination. 
Allow the mixed gases to remain in contact with the potash 
solution for a few minutes. When the nitrogen tetroxide and 
excess of oxygen have been absorbed, transfer the gas back 
again to the measuring tube and determine the volume of 
the residual nitrogen. 

The three reduced volume-readings ist, of the total 
gas (A) ; 2nd, of the nitrogen and nitrogen dioxide (B) ; and 
3rd, of the nitrogen (c) furnish all the data for obtaining the 
total volume of nitrogen and carbon dioxide in the gaseous 
mixture. 

A B = VOl. Of CO 2 . 

B + c = vol. of N. 



Quantitative Chemical Analysis. 



From the corrected volumes the weights of the carbon 
and nitrogen are readily calculated. 

The calculation, as Dr. Frankland has pointed out, may 
be simplified by considering the original gaseous mixture as 
nitrogen, so far as volume-weight is concerned. If A be the 
weight of the total gas, B its weight after treatment with 
potash, and c after absorption by pyrogallate, the weight 
of carbon will be f (A B), and the weight of nitrogen 

? + c? since the weights of carbon and nitrogen in equal 

volumes of carbon dioxide and nitrogen, measured under 
the same conditions, are as 6 : 14, and the weights of 
nitrogen in equal volumes of that gas and of nitrogen dioxide 
are as 2 \ i. By using the annexed logarithmic table for the 
reduction of cubic centimetres of nitrogen to grams for each 
tenth of a degree centigrade, the calculation becomes the 
work of a few moments only. 

Table for the reduction of Cubic Centimetres of Nitrogen to Grams. 



Log 



0-0012562 
(i + 0-00367/) 760 



for each tenth of a degree from o to 30 C. 



t 


o'o 


0*1 


O'2 


'3 


0-4 


o'S 


0-6 


0*7 


0-8 


0-9 





6-21824 


808 


793 


777 


761 


745 


729 


713 


697 


681 


I 


66 5 


649 


633 


617 


601 


586 


570 


554 


538 


522 


2 


507 


491 


475 


429 


443 


427 


412 


396 


380 


364 


3 


349 


333 


3i8 


302 


286 


270 


255 


239 


223 


208 


4 


192 


177 


161 


H5 


130 


114 


098 


083 


06 7 


051 


5 


035 


020 


004 


*989 


*973 


*957 


*942 


*926 


*9 


*895 


6 


6-20879 


864 


848 


833 


817 


801 


786 


770 


755 


739 


7 


723 


708 


692 


676 


661 


645 


629 


614 


598 




8 


567 


552 


536 


52i 


505 


490 


474 


459 


443 


428 


9 


413 


397 


382 


3^6 


35 i 


335 


320 


34 


289 


274 


10 


259 


244 


228 


213 


198 


182 


167 


151 


136 


121 


ii 


106 


090 


075 


060 


045 


029 


014 


*999 


* 9 8 4 


* 9 6 9 


12 


6-19953 


938 


923 


907 


892 


877 


862 


846 


831 


816 


13 


800 


785 


770 


755 


740 


724 


709 


694 


679 


664 


H 


648 


633 


618 


603 


588 


573 


558 


543 


528 


513 


15 


497 


482 


467 


452 


437 


422 


407 


392 


377 


362 



Water A nalysis. 311 

Table for the reduction of Cubic Centimetres of Nitrogen to Grams cont. 



t 


o'o 


O'l 


0'2 


0-3 


'4 


o'S 


o'6 


0-7 


0.8 


o- 9 


16 


346 


331 


3^ 


301 


286 


271 


256 


241 


226 


211 


17 


196 


181 


166 


J 57 


136 


121 


106 


091 


076 


06 1 


18 


046 


031 


016 


001 


*986 


* 97 I 


*956 


*94i 


*926 


*9ii 


19 


6*18897 


882 


867 


852 


837 


822 


807 


792 


777 


762 


20 


748 


733 


718 


703 


688 


673 


659 


644 


629 


614 


21 


600 


58S 


570 


555 


540 


526 


5ii 


496 


481 


466 


22 


452 


437 


422 


408 


393 


378 


363 


349 


334 


319 


23 

24 


305 

158 


290 
143 


% 


261 
114 


246 
099 


084 


216 
070 


202 

55 


187 
041 


172 
026 


25 


OI2 


*997 


*982 


* 9 68 


*953 


* 93 8 


* 9 24 


*909 


*8 95 


*88o 


26 


6-I7866 


851 


837 


822 


808 


793 


779 


764 


750 


735 


27 


721 


706 


692 


677 


663 


648 


634 


619 


605 


590 


28 


576 


561 


547 


532 


518 


503 


489 


475 


460 


446 


2 9 


432 


417 


403 


388 


374 


360 


345 


331 


316 


302 



An example of the mode of calculation will serve to render 
the process more intelligible. Let us assume that we have 
made the following readings : 

A. B. c. 



Volume of gas 
Temperature 



(After treatment 
with KHO) 

5 -oo c. c. 0-40 c. c. 0-40 c. c. 

15 I5-I I5-2 



Height of mercury in a, c, d . 300 
i, ,i b . . 

Difference .... 
Add tension of aqueous vapour 
(Table III. Appendix) 



Deduct for capillarity . 

Deduct from height of bar 
Pressure on dry gas 



mm. mm. 
300 480 
200 350 

loo 130 
127 12-8 


mm. 
480 
330 

150 

12-9 

162-9 


112-7 142-8 
Add for f 
capillarity \- 


111-7 J 45'3 

760-0 760-0 
in-7 145-3 


165-4 

760-0 
165-4 



6483 



614-7 594-6 



312 Quantitative Chemical Analysis. 

Log. of vol. of gas .."*,." 0-69897 1-60206 1-60206 

5 " 9482 



Pressure on dry gas . . 2-81178 2-78866 2-77422 

Log. of weight calc. as N. . 37O572 4'58554 4*57095 

Weight calculated as N. -0050783 -00038507 '00037235 

Weight of carbon = 3( -0050783 --0003851) = O<O02O1I4 



Weight of nitrogen = . O0037 8 7I 

Sometimes, especially when the amount of carbon is 
large, small quantities of carbon monoxide may be formed, 
and may escape complete oxidation by the copper oxide 
placed in the anterior portion of the tube. This gas remains 
mixed with the nitrogen after absorption with potassium 
pyrogallate solution. Its amount may be determined when 
the whole of the gas is transferred to the measuring tube, in 
the last determination of volume, by removing the labora- 
tory vessel, washing it, refilling it with mercury, and again 
attaching it to the face of the cap. A few drops of solution 
of cuprous chloride are then introduced into the vessel, and 
the gas allowed to act upon it. In about five minutes the 
absorption of the carbon monoxide will be complete ; the 
residual nitrogen may then be returned to the measuring 
tube, and its volume determined. If any carbon monoxide 
is found, its weight as nitrogen is calculated in the manner 
described, and added to that corresponding to the carbon 
dioxide before multiplying by fj its weight must also be 
deducted from that corresponding to the volume after treat- 
ment with potash. 

Since the accuracy of this method of combustion depends 
upon the perfection of the vacuum obtained by the Sprengel 
pump, and is liable to be affected to some slight degree by 
nitrogen retained in the copper oxide, absorption of ammo- 
nia during the evaporation, &c., it is advisable that the 



Water A nalysis. 313 

experimenter should perform several blank determinations, 
to ascertain the accumulated effect of these errors. This 
should be done by evaporating to dryness in the manner 
described a litre of pure distilled water, with the usual 
quantities of sulphurous acid and ferrous chloride solutions, 
together with about 0*1 gram of recently ignited sodium 
chloride, to afford a residue w r hich can be transferred to the 
tube. The residue is then to be burnt, and the gases analysed 
as directed : the amounts of carbon and nitrogen thus found 
are to be deducted from the quantities obtained in the sub- 
sequent analyses of water-residues. The corrections should 
amount to '00006 gram of carbon, and '00005 f nitrogen 
per litre of water. The amount of nitrogen existing as NH 3 
must be subtracted from the quantity of N thus found ; the 
remainder may be set down as organic nitrogen. 

The estimation of the organic carbon and nitrogen in 
water is of great importance in determining the degree of 
organic contamination which it has experienced. Good 
drinking water should not contain more than 0*2 part of 
carbon, and 0*02 part of nitrogen per 100,000 parts of water. 
Sewage usually contains about four parts of carbon, and two 
parts of nitrogen. The ratio of carbon to nitrogen is of 
especial importance ; the lower the ratio the more objection- 
able is the organic matter. The ratio in water for domestic 
supply may vary from five to twelve ; sewage varies from 
one to three ; polluted river- water from three to five. (For 
further details consult Frankland's ' Water Analysis,' Van 
Voorst, 1880.) 

Dittmarr's and Robinson's process. In this process the 
organic carbon is determined by weighing as carbon dioxide 
as in an ordinary combustion, whilst the organic nitrogen is 
converted into ammonia by ignition with soda or soda-baryta, 
as in the method given on p. 334. 

Determination of Organic Carbon. Evaporate i litre of 
the water, after treatment with i c.c. of a saturated solution 



3 1 4 Quantitative Chemical A nalysis. 

of sulphurous acid, as described on p. 298, in a glass dish 
under the bell jar, and transfer the dried residue by the 
aid of a spatula to a platinum boat, which is then to 
be introduced into a short combustion tube, one end of 
which has been previously drawn out, and into which is 
placed (i) a spiral of silver wire gauze to reduce any nitrogen 
oxides which may be formed, and (2) a layer of granular 
copper oxide. To the drawn-out end of the tube is attached 
a small V-shaped tube, containing a solution of chromic 
acid in 60 per cent, sulphuric acid, to which is adapted 
a short tube filled with calcium chloride. The carbon 
dioxide is absorbed in a light tube containing soda-lime 
and calcium chloride, previously weighed. The platinum 
boat containing the residue having been placed in the 
combustion tube, the posterior end is connected by a 
cork and bent tube with a gas-holder containing air freed 
from carbon dioxidt. The oxide of copper and silver are 
first heated, and then the platinum boat, a stream of the 
purified air being meanwhile sent through the apparatus. 
The carbon dioxide freed from water and sulphur dioxide by 
passing through the chromic acid solution is absorbed by 
the soda-lime. 

Determination of Organic Nitrogen. Place half a litre 
of the water in a flask connected with a condenser, and 
distil, collecting the distillate in the manner directed on 
p. 295, so as to determine the ammonia. The distillation 
is to be conducted until only about 30 c.c. are left in the 
flask. Add solutions of sulphurous acid and ferrous chloride 
to destroy nitrates, and if the solid residue is known to be 
small, add too a small quantity of potassium sulphate. 
Complete the evaporation under the bell jar, and when dry 
transfer the residue to a large silver or copper boat, moisten 
with a drop of water, and cover with 2 or 3 grams of a fused 
mixture of equal weights of baryta and soda. Introduce the 
boat into a short combustion tube connected at one end 
with a U-tube containing a known volume of water acidu- 



Water A nalysis. 315 

lated with a few drops of pure hydrochloric acid, and 
to the other end adapt an arrangement for sending a 
current of hydrogen through the apparatus. Heat the 
tube to redness; the organic nitrogen is converted into 
ammonia, which is absorbed by the dilute hydrochloric 
acid : its amount is estimated by the Nessler test, as de- 
scribed on p. 295, on one-tenth of the solution diluted to 
50 c.c. 

Blank experiments should be made in both determina- 
tions, and the necessary corrections introduced. 

Estimation of Total Soluble Matter. Ignite and weigh a 
platinum dish, place it on a glass ring on the water-bath, and 
fill it with the water to be examined, previously measured in 
a 250 c.c. flask. As the liquid evaporates, add successive 
portions from the flask ; rinse the vessel when empty with a 
small quantity of distilled water, and pour the washings into 
the dish. When the water has entirely evaporated, heat the 
residue to 100, for an hour, or until its weight is constant. 
The increase in the weight of the dish gives the amount of 
soluble matter contained in the ^-litre of water. 

Estimation of Nitrates and Nitrites* This may be 
effected in the residue obtained from the preceding deter- 
mination, by the action of precipitated copper and zinc. 
The apparatus seen in fig. 72 serves for the decomposition. 
A is a flask of 100 c.c. capacity, fitted with a caoutchouc 
stopper, containing (i) the tube funnel b, provided with a 
glass stopcock, and (2) the bent glass tube c, which is con- 
nected, by means of a well-fitting stopper, with one of the 
cylinders, <?, used for the ammonia estimation (p. 292). This 
stopper also carries the bent tube d, which is about 5 milli- 
metres in internal diameter, and is partially filled with frag- 
ments of well-washed glass. Place 3 or 4 grams of very 

* Jour. Chem. Soc. June 1873. 



3 6 



Quantitative Chemical Analysis. 



FIG. 72. 



thin sheet-zinc in small pieces in A, and cover it with a 
tolerably concentrated solution of copper sulphate. Allow 
the solution to act upon the zinc for ten or fifteen minutes, 
pour off the supernatant liquid, and fill up the flask several 
times with cold distilled water to wash the precipitated cop- 
per. After the last washing, remove the water as far as 
practicable. Add about 25 c.c. of distilled water to the 

residue obtained in the 
determination of the 
total soluble matter, to- 
gether with a piece of 
recently ignited lime, 
about the size of a 
hemp-seed, and boil 
the liquid (to destroy 
any urea which may be 
present), until 4 or 5 c.c. 
only remain. Transfer 
the liquid to the flask A, 
rinse the dish with dis- 
tilled water, so as to 
make up the volume in 
A to about 15 c.c. or 
20 c.c. Fit in the cork 
of the flask, add one 
drop of dilute hydro- 
chloric acid (free from 
ammonium chloride) to 
the cylinder *, together with 2 or 3 c.c. of distilled water, and 
also moisten the glass in d with two drops of the acid. Fit 
the cork into <?, and place the tube in a beaker of cold water, 
as represented in fig. 72. Heat the liquid in A to boiling, 
and distil it over into e ; when A is nearly empty, fill up the 
funnel with hot water, turn the stopcock, allow the water to 
flow into the flask, and continue the ebullition until e 
contains about 40 c.c. of liquid. All the nitrates will be 




Water Analysis. 317 

reduced, and the ammonia will be expelled. Raise the 
retort stand so as simultaneously to remove the tube from 
the water in the beaker, and the flask A from over the lamp. 
Wash the fragments of glass in d\ the water is readily 
drawn over into e by the contraction of the air in A on 
cooling. Fill up the tube e to the mark, and agitate the 
liquid by the aid of the bulb-stirrer. Transfer 5 c.c. to 
a second tube, dilute with distilled water, add i c.c. of 
Nessler's solution, and agitate. If the degree of colouration 
is measurable, determine the quantity of ammonia required 
to produce it in the manner described on p. 295 ; if the tint 
is too dark for comparison, take a smaller quantity; if too 
light (as it will be unless the water is very bad), take a 
larger quantity, say 10 c.c. or 20 c.c. of the distillate, in 
accordance with the indications of the preliminary trial. 
The following determinations made on known quantities of 
nitre may serve to show the degree of accuracy of which 
this method is capable : 

TAKEN. FOUND. TAKEN. FOUND. TAKEN. FOUND. 
mgms. mgms. mgms. mgms. mgms. mgms. 

1-67 I 1 ' 68 2-50. ,( 2 ' 42 3-34- 3-22 

' \i-72 \2'49 4'J6 . . 4-01 

If the operator possesses the gasometric apparatus shown 
on p. 303, the amount of nitrogen in the water existing 
as nitrates and nitrites may be readily and accurately 
estimated by determining the volume of nitric oxide 
evolved on agitating the concentrated water, acidulated 
with strong sulphuric acid, with metallic mercury. This 
process, which is an adaptation by Frankland of Crum's 
method for the refraction of nitre, is conducted as fol- 
lows : The residue from 500 c.c. of the water is dissolved 
in a small quantity of hot water, the chlorine precipitated 
by addition of a slight excess of silver sulphate, the 
liquid filtered and concentrated to a bulk not exceed- 
ing 2 c.c., and is then transferred to the apparatus seen 
in fig. 7 2 A. This consists of a stout tube about 20 c.m. 




318 Quantitative Chemical Analysis. 

long and about 1*5 c.m. internal diameter, fitted with a stop- 
cock. The tube is filled with 
mercury to the stopcock, and in- 
verted in a basin, also containing 
mercury. The concentrated fil- 
trate is then brought into the 
little cup together with the wash- 
ings, and is allowed to enter 
the tube by cautiously turning the 
stopcock. About i^ times the 
volume of the aqueous solution 
is then poured into the cup, and 
thence allowed to pass into the 
tube, which is then firmly closed 
by the moistened thumb held 

obliquely, and vigorously shaken for about five minutes. 
On the completion of the reaction the gas, nitric oxide, is 
transferred to the measuring apparatus, and its volume de- 
termined. If 500 c.c. of water have been used, the volume 
of the nitric oxide denotes the volume of nitrogen, as nitrates 
and nitrites, in 1000 c.c., since nitric oxide contains half its 
volume of nitrogen. If considerable quantities of nitrites 
are present, they must be oxidised to nitrates by adding a 
very dilute solution of potassium permanganate to the acidu- 
lated water until the pink colour is permanent. Sodium 
carbonate is then added to alkaline reaction, and the water 
is evaporated and treated in the manner above described. 

Estimation of Nitrites. A solution of meta-phenylene- 
diamine (C6H 8 N 2 ) in dilute sulphuric acid gives a dark red 
or reddish violet colouration with small quantities of nitrous 
acid, which may be made the basis of a method for the de- 
termination of the nitrites present in natural waters. The 
method requires 

(i) A Solution of Metaphenylenediamine. Dissolve 5 
grams of the base in i litre of water containing 
a slight excess of sulphuric acid. 



Water A nalysis. 319 

(2) Dilute Sulphuric Add\ i part of acid to 2 of distilled 

water. 

(3) Standard Solution of pure Sodium Nitrite. Dissolve 

0*406 gram of silver nitrite in boiling distilled 
water, and add a hot solution of pure common salt 
so long as silver chloride is precipitated. Dilute 
to i litre, allow the precipitate to settle, and make 
up each 100 c.c. of the clear solution to i litre, 
i c.c. of this solution is equivalent to o-oi mgrm. 
of N 2 O 3 . The solutions must be kept in well- 
stoppered bottles, which should be quite full. 

(4) Four narrow cylinders of colourless glass : these 

should be of such size that 100 c.c. of water rise 
to a height of about 18 c. : the level of the water 
should be marked or etched on the glass. Also 
several graduated pipettes or burettes. 

Fill one of the cylinders to the mark with the water to be 
tested, and mix with i c.c. of the dilute sulphuric acid and 
i c.c. of the solution of meta-phenylenediamine, and stir 
with the bulb tube. If a red colour appears immediately 
the amount of the nitrite is probably too great for accurate 
comparison : in that case take 50 c.c., or 25, or even 10 c.c. 
of the water, make up to 100 c.c., and repeat the trial. The 
colour should appear only at the expiration of 70 or 80 
seconds. 

Into the other three cylinders place measured quantities 
from 0*3 to 2-5 c.c. of the standard nitrite solution, dilute to 
the mark with distilled water, and add to each i c.c. of the 
sulphuric acid and meta-phenylene diamine solutions, and 
compare the tints formed with that in the first cylinder. 
This comparison give a first approximation. The trials are 
repeated on a fresh sample of the water and with the quan- 
tities of standard nitrite now deemed necessary to produce 
a similar tint. Lastly, a final series of comparisons, starting 
simultaneously, should be made, and the colours compared 
after the lapse of 20 minutes. 



320 Quantitative Chemical Analysis. 

Estimation of Chlorine. By Standard Silver Nitrate and 
Potassium Chr ornate Solutions. Dissolve 2-3944 grm. of 
pure dry silver nitrate in distilled water, and dilute to i litre ; 
i c.c. of this solution is equivalent to 0-5 milligram of chlo- 
rine. The potassium chromate solution should be strong 
and neutral. Transfer 50 c.c. of the water to be tested 
to a porcelain basin, colour with 2 drops of the potassium 
chromate solution, and add the silver solution, drop by 
drop, until the permanent red colour of the silver chromate 
makes its appearance, If 50 c.c. have been taken, the 
number of c.c. of the silver solution employed gives the 
amount of chlorine in parts per 100,000. 

In many highly-coloured waters the final point of the 
reaction is not distinctly visible. In such a case add a small 
quantity of lime-water (free from chlorides) to the measured 
portion of the water, pass washed carbonic acid through the 
liquid, boil, and filter. The colouring matter will in general 
be carried down by the precipitated chalk : the amount of 
chlorine in the filtrate may then be determined as directed 
above. It will be sometimes necessary, however, to destroy 
the organic matter by heat : the measured portion of the 
water, after addition of a small quantity of lime-water, must 
be evaporated to dryness and the residue gently heated. 
Treat the saline matter with hot water, filter, and determine 
the chlorine in the usual manner. 

The quantity of chlorine contained in the water will often 
afford another indication of its purity. Very pure waters, as 
a rule, contain comparatively small quantities of chlorine, 
less than i part in 100,000 ; when contaminated with 
sewage (which contains on the average about 10 parts in 
100,000), the quantity is largely increased. Of course in 
judging of the character of a drinking-water from the amount 
of chlorine it contains, due regard must be had to the nature 
of the strata through which it percolates. Water originating 
from springs in the neighbourhood of the sea, especially if 
the district be sandy, may contain considerable amounts of 
chlorine, and yet be free from sewage-matter. 



Water Analysis. ' 321 

Estimation of ' Hardness! Waters are familiarly spoken 
of as * hard ' and * soft ' : these terms have reference to the 
action of soap upon the water. A ' hard ' water necessitates 
the expenditure of much soap before it will give a lather ; 
this expenditure is caused by the action of the lime and mag- 
nesia salts, which decompose the soap, or stearate of soda, 
forming insoluble stearates of lime and magnesia. These 
substances constitute the pellicle or scum which forms upon 
the surface of a hard water after treatment with soap. Before 
a ' hard ' water can give a lather useful for detergent action, 
sufficient soap must be used to convert all the lime and 
magnesia salts present into stearates. The estimation of 
the hardness or soap-destroying power of the water becomes 
therefore an important element in determining its value for 
economic purposes. 

The soap-destroying power of the water is measured 
directly : a solution of soap of known strength being made to 
act upon a definite volume of the water until a permanent 
(and detergent) lather is obtained. 

I. Preparation of the Strong Soap Solution. Pound, in 
small quantities at a time, in a mortar, 3 parts of lead plaster 
and i part of dry potassium carbonate. Mix thoroughly, and 
add a small quantity of methylated spirit, and triturate until a 
thin creamy mixture is obtained. After standing for some 
hours, pour the clear solution through a filter, and exhaust 
the residue repeatedly with fresh portions of spirit. If the 
solution remains clear on standing, proceed to determine its 
exact strength by the aid of a standard solution of calcium 
chloride. 

II. Preparation of Standard Calcium Chloride Solution. 
Weigh out into a porcelain or platinum dish exactly *2 gram 
of finely-powdered marble, cover the dish with a large watch- 
glass, and dissolve the marble in dilute hydrochloric acid. 
Heat the basin on the water-bath, and when the expulsion of 

Y 



322 Quantitative Chemical Analysis. 

the carbon dioxide is at an end, rinse the under- surface of 
the watch-glass into the basin and evaporate to complete 
dryness. Add a small quantity of water, and again evaporate 
to ensure the complete removal of the excess of the hydro- 
chloric acid. Dissolve in water, and dilute to 1,000 c.c. 

III. Dilution of the Strong Soap Solution. Transfer 50 c.c. 
of the standard calcium chloride solution to a bottle of 25oc.c. 
capacity, provided with a well-fitting stopper, fill up a burette 
with the soap solution and add it to the water in the bottle 
in quantities of a few drops at a time. After each addition 
of the soap solution insert the stopper, and shake briskly. 
The process is finished when a uniform lather is obtained 
which is permanent for at least 3 minutes, and which may be 
re-formed by again shaking the liquid. Read off the burette 
and dilute the soap solution with a mixture of 2 vols. of 
methylated spirit and i vol. of water, until about 12 c.c. of 
the diluted mixture are equivalent to 50 c.c. of the standard 
calcium chloride solution; allow it to stand for 24 hours, 
filter it if necessary, again determine its strength, and dilute 
it with the mixture of alcohol and water, until exactly 14*25 
c.c. are required to produce a permanent lather with 50 c.c. 
of the standard calcium chloride solution. 

The Process. Transfer 50 c.c. of the water under examina- 
tion to the 250 c.c. bottle, shake vigorously, and suck out 
the air from within the bottle by the aid of a glass tube, to 
remove the carbon dioxide expelled on agitating. Fill up 
the burette with the soap solution, and add it, i c.c. at a 
time, to the water; after each addition insert the stopper, and 
shake vigorously. As soon as a froth begins to form, add 
the solution of soap in smaller quantities until a uniform per- 
manent lather is obtained. If more than 16 c.c. of the 
soap solution are required the operation must be repeated on 
a smaller quantity of the water. Transfer 25 c.c., or less if 
it is very hard, of the water to the bottle, and add sufficient 
distilled water to make up the volume to 50 c.c., and again 



Water A nalysis. 



323 



add the soap solution, in small quantities at a time, until the 
lather is obtained. Multiply the volume in c.c. of soap 
solution used by the number expressing the fraction of 50 c.c. 
taken : thus, if 25 have been taken, multiply by 2 : if 10, 
multiply by 5. The weight of calcium carbonate in 
100,000 parts of water, corresponding to the number ofc.c. 
required for 50 c.c. of the water, is given in the following 
table : Column I. gives volume of soap solution : Column 
II. the corresponding amount of calcium carbonate per 
100,000 parts. 

Table of Hardness. 



i 


II. 


L 


II. 


I. 


II. 


I. 


II. 


I 


II. 


I. 


II. 


c.c. 

07 


oo 


c.c. 

3-3 


3-64 


c.c. 

5 "9 


7-29 


C.C. 

8'5 


11-05 


C.C. 

II'I 


15-00 


c.c. 

137 


19-13 


8 


16 


4 


77 


6-0 


43 


6 


20 


2 


16 


8 


29 


9 


32 


5 


90 


*i 


'57 


7 


'35 


'3 


32 


9 


44 


I'O 


48 


6 


4-03 


2 


71 


8 


50 


4 


48 


14-0 


60 


I 


63 


7 


16 


'3 


86 


'9 


65 


5 


63 


i 


76 


'2 


79 


8 


29 


'4 


8-00 


9-0 


80 


6 


79 


'2 


92 


'3 


'95 


'9 


43 


'5 


'14 




95 


7 


'95 


'3 


20-08 


'4 


i-ii 


4'0 


'57 


6 


29 


2 


12-11 


8 


16-11 


'4 


24 


I 


27 
43 


i 

'2 


71 

86 


8 


'43 

'57 


"3 
'4 


26 
41 


'9 

I2'O 


27 
'43 


| 


40 

56 


7 


56 


'3 


5 -oo 


'9 


71 


5 


56 


'I 


'59 


7 


71 


8 


69 


'4 


14 


7-0 


86 


6 


71 


2 


75 


8 


87 


9 


82 


5 


29 


i 


9-00 


7 


86 


3 


90 


'9 


21-03 


2'0 


'95 


6 


43 


2 


14 


8 


13-01 


'4 


17-06 


15-0 


19 


I 


2-08 


7 


57 


'3 


29 


'9 


16 


15 


22 




'35 


'2 


21 


8 


7i 


'4 


'43 


10 -o 


3i 


6 


38 


'2 


'Si 


'3 


'34 


9 


86 


5 


'57 


I 


46 


7 


'54 


'3 


68 


"4 


'47 


5' 


6-00 


6 


71 


2 


6 1 


8 


70 


'4 


85 


'5 


60 


i 


14 


7 


86 


'3 


76 


'9 


86 


'5 


22-02 


6 


73 


2 


29 


8 


lO'OO 


4 


91 


13-0 


18-02 


6 


18 


7 


86 


'3 


'43 


'9 


15 


5 


14-06 


i 


17 


7 


35 


8 


99 


4 


57 


8-0 


30 


6 


21 


2 


33 


8 


52 


'9 


3-12 


'5 


71 


i 


45 


7 


'37 


'3 


*49 


'9 


69 


3-0 


2 5 


6 


86 


2 


60 


8 


52 


'4 


65 


16-0 


86 


i 


38 


7 


7-00 


'3 


75 


9 


68 


5 


81 






2 


'Si 


8 


14 


'4 -90 


II'O 


84 


6 


97 







In waters rich in magnesian-salts the lather acquires a 
characteristic curdy appearance, easily recognised after a 



Y2 



324 Quantitative Chemical Analysis. 

little experience. To familiarise himself "with the difference 
in the lathers occasioned by calcareous and magnesian waters, 
the student should make a dilute solution of magnesium sul- 
phate in strength equal to that of the standard calcium chloride 
solution, and compare the lathers obtained by adding a slight 
excess of soap solution to equal volumes of the two liquids. 
If a trial has shown the presence of magnesia salts in large 
proportion, the experiment should be repeated, with the 
water so far diluted with distilled water, that 50 c.c. of the mix- 
ture require only about 7 c.c. of the standard soap solution. 

It is well known that the hardness of water is occasionally 
diminished by boiling : such a water contains magnesium and 
calcium carbonates, dissolved in free carbonic acid. On boil- 
ing the water the free carbonic acid is expelled and the car- 
bonates are almost entirely precipitated, no diminution of 
the hardness will occur on boiling, unless it exceeds three parts 
per 100,000, since the carbonates are dissolved to that extent 
by water free from carbonic acid. It sometimes happens 
that the hardness, even when considerable, is not lessened 
by boiling : in this case it is due to calcium and magnesium 
sulphates or chlorides : such a water is termed permanently 
hard : a water which owes its soap- destroying power to car- 
bonates is said to be temporarily hard. It is generally desira- 
ble to distinguish between temporary and permanent hardness 
in the analysis. For this purpose transfer 200 c.c. of the water 
to a flask, and heat to boiling. After half an hour's gentle 
ebullition, remove the lamp, allow the water to cool slightly, 
filter it through a small filter, make its volume up to 200 c.c., 
and again determine the hardness on 50 c.c. of the filtrate. 

The number of c.c. of soap solution employed shows the 
permanent hardness, and this subtracted from that of the 
unboiled water gives the temporary hardness. Sometimes, 
although rarely, the hardness of the water is in part due to 
free hydrochloric or sulphuric acids ; it is advisable, there- 
fore, to ascertain its reaction before determining its soap- 
destroying power. 



Water Analysis. 325 

The hardness of water is of importance in determining its 
value for manufacturing purposes. Hard waters are a source 
of much annoyance to the manufacturer. The hard crust or 
1 cake ' which forms in steam boilers consists of sulphate of 
lime and carbonates of calcium and magnesium, often mixed 
with co-precipitated organic matter. 

Detection of Lead and Copper. Concentrate a litre of the 
water to about 50 c.c., and filter it, if necessary, into one 
of the cylinders used for the estimation of the ammonia : 
add two or three drops of acetic acid, and 2 c.c. of a 
freshly-prepared and saturated solution of sulphuretted hy- 
drogen. If a brown colouration is produced fill a second 
cylinder with distilled water, acidulate with two or three 
drops of acetic acid, mix with 2 c.c. of the sulphuretted hy- 
drogen solution, and add a standard solution of lead con- 
taining T V milligram per c.c. (obtained by dissolving 0-1831 
gram of crystallised lead acetate in a litre of distilled water) 
until the colouration is equal in both tubes. Copper may be 
accurately estimated by means of a solution of potassium 
ferrocyanide. The reddish-brown tint thus produced is com- 
pared with that formed under similar circumstances in dis- 
tilled water containing a known quantity of copper. Iron 
may be determined by a similar process. (Carnelley, Chem. 
News, XXXII. 308.) 

To determine if the water has any action on lead fill two 
beakers with the sample ; into one place a bright strip of 
the metal, and into the other a strip which has been tarnished 
by previous exposure to water, and leave them in contact 
with the sample for 24 hours. The strips are then removed 
and the water tested in the manner above described. Many 
waters which rapidly attack a clean surface of lead have no 
action on the tarnished metal : other waters, especially when 
containing nitrates and nitrites, act on lead whether bright 
or tarnished. 

The estimation of the amounts (i) of ammonia, (2) of 
organic carbon and nitrogen (or, if preferred, of the * albu- 



3 26 Quantitative Chemical A nalysis. 

minoid ammonia '), (3) of nitrogen as nitrates and nitrites, 
(4) of chlorine, (5) of total soluble and suspended matter, 
(6) of the hardness, and (7) of the presence or absence of 
lead, afford the principal data in determining the value of a 
sample of water for domestic supply. 

Occasionally it is necessary to ascertain more particularly 
the nature of the inorganic matter in solution : the brewer, 
for example, frequently wishes to know the actual amount of* 
the calcium and magnesium sulphates and carbonates present. 
The estimation of the various inorganic constituents may be 
readily made by methods already described at length. 

Estimation of Silica. Evaporate not less than a litre of 
the water to dryness (best in a platinum dish), after acidifying 
with a few drops of hydrochloric acid. Dry the saline 
residue thoroughly, moisten with hydrochloric acid, dilute 
with hot water, and filter off the separated silica. 

Estimation of Iron. Add two drops of nitric acid to the 
filtrate from the silica, boil and add ammonium chloride and 
a slight excess of ammonia, allow the precipitate to settle, 
pour the supernatant liquid through a small filter, redissolve 
the precipitate in the least possible quantity of hydrochloric 
acid, and again add ammonia. Transfer the precipitate to 
the filter, wash with hot water, dry, and weigh the ferric oxide. 

Estimation of Lime. Add excess of ammonium oxalate to 
the filtrate from the preceding estimation, filter the precipi- 
tated calcium oxalate, and, after washing and drying, ignite 
it strongly and weigh as caustic lime. 

Estimation of Magnesia. Concentrate the filtrate from 
the calcium oxalate, add sodium phosphate and ammonia, 
and treat the magnesium-ammonium phosphate in the usual 
manner. 

Estimation of Sulphuric Acid. Acidify a litre of the water 
with a few drops of hydrochloric acid, concentrate to 80 or 
loo c.c. and add excess of barium chloride solution. Filter 
off the barium sulphate and weigh it 



Gases in Water. 3 2 7 

Estimation of Phosphoric Acid. Many samples of water 
rich in lime and magnesia salts contain estimable quantities 
of this acid. To determine its amount acidify a litre of the 
water with nitric acid, concentrate to 50 c.c., and add solu- 
tion of molybdic acid (p. 218). After standing for 24 hours 
treat the yellow precipitate as directed on p. 219. 

Estimation of Sodium and Potassium. Add a few drops 
of barium chloride to a litre of the water, to precipitate the 
sulphuric acid, and boil with pure milk of lime to throw 
down the magnesia, iron, and phosphoric acid. Filter, con- 
centrate the filtrate, add ammonia, ammonium carbonate, 
and a few drops of ammonium oxalate ; again filter, and 
evaporate the filtrate to dryness, ignite to expel ammoniacal 
salts : treat the residue with a small quantity of water, filter, 
if necessary, acidify with hydrochloric acid, and evaporate to 
dryness in a weighed platinum dish. The alkalies may then 
be separated by platinum tetrachloride, or their proportion 
may be determined by dilute standard silver solution and 
potassium chromate. 

XLI. DETERMINATION OF THE AMOUNT AND NATURE OF 

THE GASES DISSOLVED IN WATER. 

Pure natural water (rain water) when thoroughly aerated 
contains about 2073 c.c. of gases per litre, composed of 

Nitrogen .... 13*08 c.c. 
Oxygen. .... 6*37 
Carbon dioxide . . . 1*28 

2073 

The ratio of the oxygen to the nitrogen is as i : 2-05. 

If the water becomes contaminated with putrescent 
organic substances the quantity of oxygen rapidly diminishes. 
It is abstracted from the water in oxidising the carbon, 
hydrogen, and nitrogen of the organic matter. By deter- 
mining the ratio of the oxygen to the nitrogen in the gases 
which it contains, we may often ascertain whether such 
putrefactive changes are in actual operation in the water. 



328 



Quantitative Chemical A nalysis. 



Several methods have been proposed for expelling the 
gases contained in water : * one of the simplest of these is 
due to Reichardtf Fig. 73 represents the apparatus re- 
quired for this method. A is an ordinary flask of from 800 to 
1,000 c. c. Its capacity must be accurately known. The 
narrow cylindrical vessel B serves as a gasholder. It is 
fitted with a caoutchouc stopper pierced with three holes, 
through one of which passes the bent tube a, the longer 
limb of which ends in B at about one-third of its height 

FIG. 73. 




from the bottom : the other end is fitted into the caoutchouc 
stopper of A. The second hole in the stopper of B contains 
the tube b, which passes nearly to the bottom of the bottle : 
b is connected by means of a caoutchouc tube, which can be 
closed by a clamp, with the glass tube c, which runs nearly 
to the bottom of c. The third hole of the stopper of B 
contains the tube d, the end of which must be on a level 
with the under surface of the cork : d is connected with the 

* See Bunsen's Gasometry, translated by Roscoe ; Miller's Inorganic 
Chemistry, p. 187; M'Leod, Chem. Soc. Jour. 

f Fresenius, Zeits. fur Anal. Chemie, vol. xi. p. 271. 



Gases in Water. 329 

narrow delivery tube ^, terminating beneath the surface of 
the mercury in the trough, in which is supported the ab- 
sorption tube T, destined to receive the gases from the 
water, and which is therefore filled with mercury before the 
commencement of the experiment. 

The flask A is completely filled with the water to be 
examined ; B and c are also partially filled with recently 
well-boiled and still warm distilled water, and b and ,t, d and 
e, are connected together ; both the clamps being open. By 
simply blowing through/ the bottle B and all the tubes are 
completely filled with the warm water : the clamps of b and 
d are now successively closed, and the tube a is inserted 
into the flask A, care of course being taken that no air- 
bubbles lodge beneath the surface of the cork. Open the 
clamp of b and c, and gradually boil the water in A. The 
gas expelled on ebullition collects in B : as soon as the 
whole is eliminated, which will require at least an hour, and 
the water in B is nearly boiling, open the clamp of d and <?, 
cautiously blow through f, so as to expel the water con- 
tained in d and <?, which is allowed to flow out into the trough, 
and displace the gases from B into the measuring tube by 
again blowing through f. On removing the lamp from beneath 
A, the water, on receding, should completely fill the flask. If 
it does not, continue the ebullition and collect the small 
portion of the remaining gas. 

Allow the moist gas in the tube T to acquire the temper- 
ature of the air (obtained from the thermometer /). Read off 
the level of the mercury within and without the tube, 
together with the temperature and barometric pressure at 
the time of observation, and calculate the volume of the 
gas when dry at o and i metre pressure. Introduce a ball 
of potash attached to a platinum wire into the tube, and in 
six or eight hours again read off the levels of the mercury. 
Quickly withdraw the piece of potash, moisten it slightly 
with water, and reintroduce it into the tube, and at the 
expiration of another hour again determine the levels of the 



3 3 Quart titative Chemical A na lysis. 

mercury, the temperature, and atmospheric pressure. If 
the second and third readings are identical the absorption 
of the carbon dioxide is complete. Reduce the volume to 
o and i metre pressure : it is not necessary to deduct the 
tension of aqueous vapour corresponding to the tempera- 
ture observed, since the gas may be assumed to be dry 
after having been in contact with the potash. Bring into 
the tube a papier-mache bullet, soaked with a solution of 
potassium pyrogallate, and from time to time read off the 
level of the mercury within the tube. As soon as the con- 
traction appears to be finished withdraw the papier-mache' 
bullet, and accurately determine the position of the levels 
of the mercury within and without the tube, together with 
the temperature and pressure. Reduce the volume of the 
residual gas (nitrogen) to oC and i metre pressure. The 
volume absorbed by the alkaline pyrogallate represents 
the amount of oxygen. 

The potash balls for the absorption of carbon dioxide may 
be readily made by inserting the end of a platinum wire 
into a little notch filed in the face of a bullet-mould, on the 
opposite side to the orifice through which in bullet-making 
the fused lead is poured. A small quantity of potash is 
melted in a silver dish and poured through the hole until 
the mould is completely filled. When quite cold the potash 
ball may be readily detached from the metal. The short 
projecting piece formed by the hole through which the 
potash has been poured should be cut away and the ball 
preserved in a stoppered bottle until used. 

The papier-mache' balls are made in a similar way : filter- 
paper, converted into pulp by maceration in water, is forced 
through the hole into the mould until it is quite filled, and 
the mould is placed in the steam chamber and heated until 
the paper is dry. 

Of course if the operator possesses the apparatus described 
on p. 303 et sey.y the accurate analysis of the gas by the use 
of liquid reagents becomes the work of a few minutes only. 



Manures. 331 

MANURES. 

XLII. GUANO. 

THIS substance is the excrement of sea-birds, more or less 
altered by exposure to the weather. It contains ammonia 
in combination with uric, oxalic, carbonic, and phosphoric 
acids, phosphates and sulphates of lime, magnesia, and 
alkalies, and more or less organic matter, water, sand, &c. 
It is very variable in composition, and is often largely 
adulterated. Its fertilising power mainly depends upon the 
ammonia and phosphoric acid which it contains. 

Mix the sample carefully, and transfer about 50 grams to 
a stoppered bottle ; the several portions used in the analysis 
are to be taken from this quantity. 

i. Determination of the Moisture. 

Weigh out about 5 or 6 grams of the guano into the tube, 
fig. 29, p. 70, heat the oil-bath to 120, aspirate a slow 
current of dry air through the apparatus, and repeatedly 
weigh the tube until the weight is constant. The flask 
contains 5 c.c. of normal acid, diluted with water, to absorb 
the ammonia, which volatilises with the steam when guano is 
heated ; the quantity of residual acid may be determined 
with litmus and a dilute soda solution in the usual way. 
The loss in the weight of the tube, minus the amount of 
ammonia retained in the flask, gives the quantity of moisture. 

2. Determination of Fixed Inorganic Matter. 

About four-fifths of the dried guano is transferred from the 
tube to a weighed platinum dish. Re-cork the tube securely ; 
the remaining portion of the dried guano will be used for the 
estimation of the nitrogen. Gently ignite the portion in the 
dish. The ash should be nearly white ; if it is of a reddish 
colour, adulteration with sand or clay may be suspected. 
The quantity of the ash in the better class of guanos does 



332 Quantitative Chemical Analysis. 

not exceed 35 per cent. The loss of weight gives the organic 
matter, together with the ammonia and combined water. 

3. Determination of the Insoluble Matter, Sand, &<r. 

Boil the weighed portion of the ash with dilute nitric acid 
for 15 minutes.* If the substance effervesces strongly the 
guano has in all probability been adulterated with calcium 
carbonate. Heat on the water-bath for some time, add water, 
filter into a 300 c.c. flask, wash the residue, dry, and weigh it. 

4. Determination of the Phosphoric Acid, Lime, Magnesia, 
Sulphuric Acid, and Alkalies. 

Make up the filtrate to the containing-mark, and shake 
the liqui'd. 

(a) Phosphoric Acid, Lime, and Magnesia. Transfer 
100 c.c. of the solution to a -J-litre flask, add excess of 
ammonia, and then acetic acid to acid reaction. Dilute the 
liquid to the mark, and shake. 

a. Phosphoric Acid. By Standard Uranium Solution. 
When a solution of acetate or nitrate of uranium is added 
to a solution of phosphoric acid, containing ammoniacal 
salts and free acetic acid, a light greenish-yellow precipitate 
of the double uranium-ammonium phosphate is produced. 
This precipitate is insoluble in water, and in acetic acid ; 
the stronger acids, however, dissolve it. 

A solution of the acetate or nitrate of uranium gives a 
reddish-brown colour with potassium ferrocyanide. This 
colour is not produced so long as any phosphoric acid is in 
solution. These reactions form the basis of an accurate 
volumetric method for the estimation of phosphoric acid. 

Preparation of the Standard Solution of Uranium. 
Dissolve about 35 grams of well-crystallised uranium 
nitrate or acetate (the former, however, is preferable) in 
900 c.c. of water. The solution, mixed with sodium 
acetate, must be standardised by a solution of pure sodium 

* In burning off the organic matter the phosphates are partially 
converted into pyrophosphates. By boiling with dilute nitric acid the 
pyvophosphates are reconverted into orthophosphates. 



Manures. 333 

phosphate of known strength. Dissolve 100 grams of 
sodium acetate in 900 c.c. of water, and dilute with strong 
acetic acid to i litre. Dissolve 10-085 grams of pure and 
non-effloresced crystals of sodium phosphate, previously 
dried by pressure between filter-paper, in a litre of water. 

Transfer 50 c.c. of the sodium phosphate solution, corre- 
sponding to o-i gram P 2 O 5 , to a beaker, add 5 c.c. of the 
sodium acetate solution, and heat the mixture on the water- 
bath. Remove it when its temperature is about 80, and 
quickly run in 10 or 12 c.c. of the uranium solution, with 
constant stirring. Now add the uranium solution more 
cautiously, in quantities of 0^5 c.c. at a time, and test the 
mixture after each addition. For this purpose bring a few 
drops of the turbid, but nearly colourless, liquid, on to a 
porcelain slab, and add to it a small drop of potassium ferro- 
cyanide solution; if the least excess of uranium salt be 
present, the mixed drops will acquire a reddish-brown 
colour. If this colour is not at once perceived, continue the 
addition of the uranium solution until it just appears. 
Replace the beaker on the water-bath, and in a few minutes 
again transfer a few drops to the slab, and test a second time 
with the ferrocyanide. If the colouration is still distinctly visi- 
ble, the process is finished. The solution of the uranium salt 
must now be diluted, until 20 c.c. are exactly required for the 
50 c.c. of sodium phosphate solution : 20 c.c. thus become 
equivalent to 0*1 gram of P 2 O 5 , or i c.c. = 5 milligrams 
P 2 O 5 . The exact strength should be again determined by 
several trials after the dilution, has been effected. 

To apply this process to the determination of the phos- 
phoric acid contained in the solution of the guano, transfer 
50 c.c. of the acetic acid solution (a) of the phosphate to a 
beaker, heat on the water-bath, and proceed exactly as 
described. Repeat the determination on a second portion 
of the liquid. 

/3. Lime. Transfer 100 c.c. of the liquid (a) to a beaker, 
heat, and add excess of ammonium/ oxalate ; the calcium 



334 Quantitative Chemical Analysis. 

oxalate is filtered off after standing, and weighed as carbo- 
nate, or it is treated with sulphuric acid, and titrated with 
potassium permanganate. 

y. Magnesia. The filtrate from the lime precipitate is 
mixed with ammonia and sodium phosphate, and the mag- 
nesium-ammonium phosphate weighed as pyrophosphate. 

(b} Determination of Sulphuric Acid and Alkalies. 
Transfer the remainder (200 c.c.) of the liquid in 4 to a 
beaker, heat, add barium chloride, and filter off any barium 
sulphate which may form. To the filtrate add milk of lime 
or baryta water, boil, filter, and precipitate the excess of the 
alkaline earth with ammonia and ammonium carbonate, filter, 
evaporate to dryness, and ignite, to expel the ammoniacal 
salts. The residue is treated with a small quantity of water, 
filtered again, if necessary, and the filtrate evaporated to 
dryness in a weighed platinum dish. The proportion of the 
alkalies in the washed chlorides may then be determined by 
a dilute standard solution of silver nitrate and potassium 
chromate. 

5. Determination of the Nitrogen. 

a. Nitrogen existing as Ammonia. From i to 5 grams 
of the guano, according to its supposed richness in ammo- 
nia, as determined by the amount evolved in drying, are 
weighed out into the retort, fig. 32, and boiled with magnesia 
for some time. The ammonia is gradually expelled, and 
may be collected in the flask c, containing a known quan- 
tity of standard sulphuric or hydrochloric acid, diluted with 
water. The quantity of the residual free acid is then to be 
estimated by standard soda and litmus solution. The use 
of caustic soda or lime in expelling the ammonia is in- 
admissible, since these substances would convert a portion 
of the nitrogenous organic matter into ammonia. 

ft. Nitrogen existing as Azotised Organic Matter, Uric 
Acid, &>c. By Ignition with Soda-lime. Many organic sub- 
stances containing nitrogen, not in the form of nitroxides. 



Manures. 335 

when heated with a caustic alkali, give up the whole of 
their nitrogen in the form of ammonia. This reaction con- 
stitutes the principle of a convenient method for estimating 
the amount of organic nitrogen existing in manures. 
The following articles are required for this method : 

(a) Combustion-tube. This should have the form seen in 
fig. 74; it is about 40 centimetres long, and from 10 to 12 
millimetres in internal diameter. 

(b) Soda-lime. Heat a sufficient quantity of the coarsely 
powdered substance in a porcelain basin, just before it is 
wanted, and allow it to cool. A mixture of equal weights of 
dry slaked lime and dehydrated sodium carbonate may be 
used instead of soda- lime. 

(c) Oxalic Acid, This should be well dried in the water- 
bath so as to expel all its water of crystallisation. 

(d) Asbestos. Ignite a small quantity in the gas-flame 
before use. 

(e) A bulbed U-tube, fitted with caoutchouc stopper and 
bent tube. On the end of the bent tube is a cork, which 
fits tightly into the combustion-tube. 

1 Introduce a layer, about 3 centimetres in length, of the 
dried oxalic acid, mixed with a small quantity of soda-lime, 
into the posterior end of the tube, and afterwards an 
equal bulk of soda-lime. Weigh out from the tube the 
remainder of the dried guano obtained in i, into a dry 
porcelain mortar, and mix it with soda-lime. Bring the 
mixture without loss of time (since it is apt to part with a 
small quantity of ammonia) into the tube, and rinse out the 
mortar with a fresh portion of soda-lime. The substance 
should be mixed with a sufficient amount of soda-lime to 
occupy about 20 centimetres of the length of the tube. 
Fill up the tube to within 5 centimetres of its length with 
soda-lime, and insert a loosely-fitting plug of the re- 
cently-ignited asbestos. Fit in the cork of the U-tube, 
transfer 10 c.c. of standard acid to the U-tube, and add 
sufficient water to fill the bulbs to the extent indicated in 
the figure. Gently tap the combustion-tube on the table 



3 3^ Quantitative Chemical A nalysis. 

so as to make a passage for the evolved gases. Place 
the combustion- tube in the furnace, and gradually heat it 
along its entire length, beginning at the end nearest the 
U-tube. The heat must be sufficient to cause a steady 
evolution of gas ; towards the end it should be increased, to 
break up any cyanides which may have been formed. Do 
not heat the extreme end of the tube where the oxalic acid 
is situated, until the evolution of the gas has almost 
finished. When the combustion is at end, and the evolution 
of gas has totally ceased, cautiously heat the oxalic acid ; 
this occasions a brisk current of carbon dioxide, which 
sweeps out all the ammonia remaining in the tube. Remove 
the U-tube when the evolution of the gas has nearly finished, 

FIG. 74. 



add a few drops of litmus solution, and dilute caustic 
soda solution from a burette, until the free acid is nearly 
neutralised. Transfer the liquid to a beaker, wash out the 
bulbs, and complete the addition of the soda solution. By 
operating in this manner there is less chance of loss arising 
from incomplete transference of the acid liquid, and less 
washing water is afterwards required. It will be found 
most convenient to use a soda solution, of which about 3 c.c. 
are equivalent to i c.c. of normal acid. If it is preferred to 
determine the nitrogen by weight (and this should be done 
if much empyreumatic matter be present in the liquid of the 
bulbs), rinse the bulbs into a beaker, and pour the solution 
through a moistened filter. Evaporate the filtrate to dryness 
with excess of .platinum tetrachloride, transfer the washed 
double salt to a weighed porcelain crucible, in the manner 
directed on p. 84, and dry it slowly; heat the crucible 



Manures. 



337 



to bright redness, and weigh the residual platinum. 194-4 
parts of platinum are equivalent to 28 of nitrogen. 
The amount of the nitrogen cannot be calculated from 
the weight of the double salt, since it is apt to contain 
considerable quantities of compounds of platinum with 
organic bases. These bases, however, contain the same pro- 
portion of nitrogen and platinum as the ammonium-platinum 
chloride. By determining the amount of platinum left on 
ignition, the proportion of the nitrogen is therefore readily 
calculated. 

Deduct the amount of nitrogen corresponding to the 
ammonia found in o : the difference shows the quantity of 
organic nitrogen. 

XLIII. BONE- DUST. 

1. Moisture. Dry a weighed portion at 120-130 in the 
air-bath. 

2. Carbonic Acid. Determine this constituent by means 
of the apparatus represented in fig. 31, p. 86. 

3. Fixed Constituents. See p. 331. 

4. Insoluble Matter and Sand, &c. See p. 332. 

5. Soluble Matters after Treatment with Hydrochloric Acid. 
See p. 332. 

6. Fat. Treat about 6 or Sgrams of the sample with boiling 
ether in an apparatus adjusted for distillation per ascensum, 
and dry the insoluble matter at 120-130 in the air-bath. 
From the loss of weight deduct the moisture found in i ; the 
remainder gives the quantity of fatty matter. 

7. Gelatigenous Matter. Add together the amounts of the 
several constituents : the difference required to make up 
100 may be set down as gelatigenous substance. 

XLIV- SUPERPHOSPHATES. 

The phosphoric acid existing in the majority of naturally 
occurring phosphates is not very readily dissolved by water, 
and is therefore not in the form in which it can be rapidly 

z 



338 Quantitative Chemical Analysis. 

assimilated by plants. The manure manufacturer converts 
a portion of the phosphoric acid into the soluble modifica- 
tion by treating the phosphorite, bone-dust, spent bone- 
black, &c., with sulphuric acid. In this operation the in- 
soluble tricalcium phosphate (Ca 3 P 2 O8) is converted into 
the soluble monocalcium phosphate, CaH 4 P 2 O 8 , calcium 
sulphate being simultaneously produced. The pasty mass 
which runs out of the apparatus in which the mixture of 
phosphate and acid is made, gradually becomes dry on 
standing, partly from the evaporation, and partly from the 
assimilation of the water. To increase the fertilising power 
of the material, or to satisfy the tastes of the consumers, the 
manufacturer frequently adds various substances, such as dried 
or liquid blood to the material before or after treatment with 
the acid. Superphosphate therefore consists essentially of 
monocalcium phosphate, CaH 4 P 2 O8 (so-called soluble phos- 
phate), mixed with tricalcium phosphate, Ca 3 P 2 O8 (insoluble 
phosphate), calcium sulphate, oxides of iron, alumina, mag- 
nesia, and alkalies, and more or less organic matter and 
moisture. 

Sample the mixture carefully, and transfer a portion to a 
stoppered bottle, from which the quantities employed for the 
several estimations are to be taken. 

1. Water. Dry a weighed portion of the sample at 170 
in the air-bath until it ceases to lose weight. The loss 
gives the quantity of moisture and water existing in com- 
bination with the calcium sulphate. 

2. Weigh out 10 grams of the undried superphosphate into a 
mortar, add a small quantity of cold water, and triturate with 
the aid of the pestle ; allow the suspended matter to settle 
for a few minutes, and pour the liquid through a filter into a 
500 c.c. flask. Repeat the extraction with cold water in the 
same manner several times in succession, and finally wash 
the residue with hot water, transferring the washings to the 
filter. Dilute the filtrate to the mark and shake the liquid. 
Weigh the insoluble portion. 



Manures. 339 

I. Examination of the Filtrate. 

(a) Estimation of the Ferric Phosphate and Soluble Calcium 
Phosphate. Transfer 200 c.c. of the liquid to a platinum 
dish, and evaporate, adding an excess of sodium carbonate 
and a little nitre so soon as the whole of the liquid has been 
brought into the dish. When the mass is dry, ignite gently, 
mix with a little water, and rinse the contents of the dish 
into a beaker, add excess of hydrochloric acid, and heat 
until the liquid is clear. Mix with excess of ammonia, 
acidulate with acetic acid, and filter off and weigh the ferric 
phosphate, receiving the filtrate in a 250 c.c. flask. Dilute to 
the mark and shake. Withdraw successive portions of 50 c.c., 
and determine the phosphoric acid by uranium solution. 
Transfer 100 c.c. of the liquid to a beaker, and determine 
the lime-sxA magnesia, as directed on p. 333. 

(b) Estimation of Organic Matter and Alkalies. Evaporate 
100 c.c. in a platinum dish, adding milk of lime until the 
liquid is distinctly alkaline. Dry the residue at 180 and 
weigh. Ignite the dried mass and again weigh ; the difference 
gives the quantity of organic matter. Boil the weighed residue 
with lime water, then with pure water ; filter, and add barium 
chloride to precipitate the sulphuric acid ; mix with am- 
monia, ammonium carbonate, and oxalate ; filter, evaporate 
to dryness with hydrochloric acid, ignite the residue, treat 
with water, filter, and weigh the alkaline chlorides. 

(c) Determination of the Sulphuric Acid. Heat 100 c.c. of 
the liquid in a beaker, acidulate with a few drops of hydro- 
chloric acid, and precipitate with barium chloride. 

II. Examination of the Insoluble Portion. 

(a) Determination of the Carbon. Ignite gently in a pla- 
tinum dish ; the loss of weight gives the quantity of organic 
matter and charcoal. 

(b} Determination of Sand, Clay, &c. Boil the ignited 
portion repeatedly with dilute hydrochloric acid, filter into 

z 2 



34O Quantitative Chemical Analysis. 

a | -litre flask, and wash with hot water. The insoluble 
residue consists of sand and clay. Dilute the filtrate to the 
mark and shake. 

(c) Determination of Phosphoric Acid, Iron, Lime, and 
Magnesia. Transfer 100 c.c. c/f the above solution to a beaker 
and proceed exactly as in I. (a). 

(d) Determination of Sulphuric Acid. In 100 c.c., by 
barium chloride in the usual manner. 

(e) Determination of Total Nitrogen. In from i to 2 grams 
of the original substance by ignition with soda-lime. (See 

P. 335-) 

(y~) Determination of Ammonia. Superphosphates are 

occasionally mixed with ammoniacal salts. The amount of 
this ammonia is determined as in No. VIII. Part II. 

The results of the analysis should be arranged according 
to the subjoined form : 

/ Phosphoric acid* 

Soluble constituents ^ esia 

\ Ferric oxide 

{Phosphoric acidf 
Lime 
Magnesia 
Ferric oxide 
Alumina 

Total calcium sulphate, 

,, organic matter and charcoal, \ 
,, sand and clay, 
,, moisture, 

* Equal to per cent, soluble phosphate. 

f Equal to per cent, insoluble phosphate. 

I Containing per cent, of nitrogen, equal to per cent, ammonia. 



XLV. ASHES OF PLANTS. 

The substances generally present in estimable quantity in 
the ashes of plants are silica, phosphoric, sulphuric, and car- 
bonic acids, chlorine, potash, soda, lime, magnesia, iron, and 
manganese. In much smaller quantity are sometimes found 



Ashes of Plants. 34 1 

alumina, lithia, strontia, baryta, rubidia, copper, fluorine, 
iodine and bromine, cyanides and cyanates, boracic acid, 
sulphides, &c. 

Certain of these substances are, however, never present 
in plants: thus the cyanides and cyanates are formed by 
the mutual action of the carbon and nitrogen in the plant at 
the high temperature of the incineration : in the case of 
ashes rich in alkalies and alkaline earths they may have 
been also formed by the action of the nitrogen in the air 
during the burning. Probably too, all the sulphur found in 
the ash did not originally exist as sulphuric acid : not 
unfrequently sulphur exists in the unoxidised state in a 
plant, and in combination with carbon, hydrogen, and 
nitrogen, forming peculiar organic acids. In presence of 
the bases the sulphur becomes converted into sulphuric 
acid during the incineration: sometimes a portion of the 
sulphur escapes oxidation, or when oxidised is again reduced 
by the admixed charcoal, giving rise to the sulphides. The 
main quantity of the carbonic acid present in the ash is 
derived from the destruction of organic acids combined with 
the alkalies. 

In order to obtain the ash in a proper scate for analysis, 
the portions of the plant to be incinerated must be freed 
as far as possible from adhering soil, &c., by brushing or 
rubbing. In the case of small seeds the best plan is to 
treat them in a beaker with a small quantity of water, stir 
them with a glass rod for a minute or so, and throw them on 
a sieve, the meshes of which are sufficiently coarse to allow 
the sand to pass through, whilst retaining the seeds. Repeat 
this operation several times, but take care not to allow the 
seeds to remain too long in contact with the water, or por- 
tions of the soluble salts will be dissolved out. Place the 
seeds in a cloth and rub them between its folds, and dry 
them on a water-bath. The substance to be incinerated is 
weighed, and placed in a shallow porcelain basin fitting into 
a muffle, which is to be gradually heated to low redness. 



342 



Quantitative Chemical Analysis. 



FIG. 75. 



Great care must be taken duly to regulate the heat : if it is 
too high, the process of incineration will be retarded : the 
salts will fuse, and enclose the carbonaceous matter, thus 
protecting it from the action of the air. Moreover, at a high 
temperature, chloride of sodium would volatilise, and a part 
of the phosphorus would be lost. It seldom facilitates the 
operation to stir the heated mass, as its porosity and loose- 
ness of aggregation are thereby destroyed. The supply of air 
must be adequate, but not excessive, otherwise particles of 
the ash are apt to be carried away in the draught. In the 
ash of vegetables, the amount of alkali 
is frequently so considerable that it is 
almost impossible to obtain the mass 
quite white at a temperature sufficiently 
low to prevent it fusing. In this case 
it is best to char the body in a Hessian 
crucible at a low red heat (scarcely 
visible in daylight), extract the soluble 
portion with water, and complete the 
incineration of the residue in a muffle. 
In all cases the ash must be weighed, 
properly mixed in a smooth porcelain 
crucible, and preserved in a well-stop- 
pered bottle. 

The ash of organic substances may 
in general be readily obtained free from 
carbonaceous matter by the simple ar- 
rangement seen in fig. 75. The mass 
is charred in a porcelain dish at a low 
red heat, and as soon as the evolu- 
tion of empyreumatic matter ceases, the neck of a large 
retort is supported over the dish by means of a clamp, and 
the heating is continued until the mass is white. The increased 
current of air playing over the heated mass facilitates the 
combustion of the carbon. This method is liable, however, 
to increase the amount of sulphates present in the ash, owing 





Ashes of Plants. 343 

to the action of the sulphuric acid derived from the coal 
gas : in cases where great accuracy is required the Bunsen 
lamp must be replaced by a spirit lamp. 

From 7 to 10 grams of the well-mixed and finely-powdered 
ash are placed in a glass cylinder of about 300 cubic centi- 
metres capacity, provided with a well-fitting stopper. About 
25 cubic centimetres of distilled water are then added, and 
carbon dioxide is passed into the cylinder. The delivery 
tube of the apparatus (which must not dip into the liquid) 
is occasionally withdrawn, the stopper inserted, and the 
liquid shaken to promote the absorption of the gas. When 
the caustic bases are completely neutralised and the solution 
saturated (which is evidenced by the cessation of the partial 
vacuum, and also by the bubbles passing upwards between 
the bottle and its stopper when the latter is cautiously lifted 
after the liquid has been shaken), the contents of the cylinder 
are washed into a porcelain dish,* evaporated to complete 
dryness, again heated with a small quantity of water to 
dissolve the alkaline salts, and after standing a short time 
FIG 7 6 filtered through a weighed filter. The 

filtrate is again evaporated to dryness, 
the saline residue treated with a small 
quantity of water, and the calcium sulphate 
which separates out filtered off through a 
weighed filter. The filtrate is received in a 
small weighed flask of 150 cubic centimetres 
capacity, provided with a side tubulus (fig. 76). This is 
easily made by directing the flame of the blowpipe upon the 
side of the flask until the glass is softened, when on touching 
the softened part by a thick platinum wire it will adhere, and a 
portion of the glass may be drawn out in the form of a 
narrow tube. The wire is detached from the tubulus by 
scratching the latter with a cutting diamond. Care must be 

* If calcium carbonate crystallises on the side of the cylinder it may 
be removed by adding a little water, saturating it with carbonic acid, 
and dissolving the thin crust by vigorously shaking the liquid. 




344 Quantitative Chemical Analysis. 

taken in filtering the liquid containing the soluble portion 
of the ash, into- this flask that the end of the funnel does not 
dip into the liquid ; the funnel must be maintained in such 
a position that the drops in falling into the flask are not 
splashed against its upper sides. The filtrate is diluted to 
about 60 cubic centimetres, and well mixed by shaking. The 
edge of the tubulus is slightly greased, and the flask and solu- 
tion weighed. The liquid is then divided in to six portions, con- 
tained in little beakers, to serve for the determination of the 
sulphuric acid, alkalies, chlorine, phosphoric and carbonic 
acids, the sixth portion being reserved in case of accident. 
The object of the tubulus is to allow of the liquid being 
poured from the flask into the beakers : the amount taken 
for each determination is indicated by the loss of weight 
suffered by the flask and solution. 

The carbonic acid is determined volumetrically by deci-nor- 
mal sulphuric acid and litmus solutions ; the sulphuric acid 
and chlorine by precipitation as barium sulphate and silver 
chloride. The portion for the phosphoric acid determina- 
tion is acidified with hydrochloric acid solution, boiled to 
expel carbonic acid, allowed to cool, ammonia added, 
together with a few drops of magnesia-mixture, and the mag- 
nesium-ammonium phosphate weighed as pyrophosphate. In 
order to determine the amount of the alkalies, the solution is 
boiled with a slight excess of baryta water (best in a platinum 
or silver dish) ; the sulphuric, carbonic, and phosphoric acids, 
together with the greater portion of the magnesia dissolved, 
are thus separated : the excess of baryta is removed by 
ammonia and ammonium carbonate. The filtrate is evapo- 
rated to dryness in a platinum dish, gently heated, re-dissolved 
in a few drops of water, filtered if necessary, a few drops of 
hydrochloric acid added, and the liquid evaporated to dry- 
ness, heated, and the mixed alkaline chlorides weighed. 
The potassium chloride is then separated by platinum tetra- 
chloride,.or the relative amount of the two chlorides deter- 
mined by standard silver. In cases where the amount of the 



Ashes of Plants* 



345 



FIG. 77. 




soluble portion of the ash is comparatively large, more than 
traces of magnesia will remain in solution with the alkaline 
salts. This portion of the magnesia is found in the nitrate from 
the double chloride of potassium and platinum : its amount 
may be estimated by evaporating the alcoholic solution to 
dryness, re-dissolving in water, and 
transferring the liquid to a small flask 
provided with a tightly-fitting cork, 
furnished with two tubes, as in fig. 
77. Hydrogen is led through the 
tube a, and the exit tube , within 
the flask, is sufficiently long to reach 
just above the surface of the liquid, 
so as to ensure the thorough expul- 
sion of the air. When the flask is 
completely filled with hydrogen, the 
ends of the tubes are closed by 

stoppers, and the flask is placed in direct sunlight, when the 
platinum is quickly reduced to the metallic state, and the 
solution becomes colourless. The process of reduction 
may, if necessary, be facilitated by heating the solution on a 
water-bath before the transmission of the gas. If the capa- 
city of the flask is small, it will be requisite to refill it once or 
twice with hydrogen to ensure the complete reduction of the 
platinum ; it is then desirable to displace the remaining gas 
by a rapid current of carbonic acid, otherwise an explosion 
might occur, particularly if the contents of the flask are 
warm, owing to the surface action of the finely-divided 
platinum on a mixture of air and hydrogen. The colourless 
solution is then filtered from the reduced metal, and, after 
concentration, the magnesia precipitated by sodium phos- 
phate and ammonia. This method is recommended to be 
used in all accurate separations of the alkalies from magnesia : 
it is moreover a rapid and easy mode of recovering the excess 
of platinum used in the determination of potassium or am- 
monium salts. 



346 Quantitative Chemical Analysis. 

In the insoluble portion are contained lime, magnesia, 
ferric oxide (alumina), silica, phosphoric, sulphuric, and 
carbonic acids. This is dried at 100 and weighed. It is 
detached as far as possible from the filter, and the latter in- 
cinerated. The ash from the filter-paper is allowed to fall 
into a porcelain basin, and treated with water saturated 
with carbonic acid, evaporated to perfect dry ness on the water- 
bath, and mixed with the main quantity of the insoluble por- 
tion in a smooth porcelain mortar. The carbonic acid is 
determined in about i to 2 grams of the substance according 
to the method given in No. V. Part II. ; the solution in the 
flask serves for the determination of the silica, sand, charcoal, 
and sulphuric acid. The phosphoric acid, iron (alumina), man- 
ganese, lime, and magnesia are determined in about 2 grams 
of the remainder of the insoluble matter. The weighed 
portion is dissolved in nitric acid, and after separation of the 
silica in the usual manner, the solution is again evaporated 
nearly to dryness in a porcelain basin, and dilute nitric acid 
added until the bases are completely dissolved, strong fuming 
nitric acid (saturated with the lower oxides of nitrogen) 
added until calcium nitrate begins to separate : a few 
more drops of dilute nitric acid are now added to destroy 
the slight turbidity. The nitric acid solution of the sub- 
stances is thus in the highest possible state of concentration. 
It is covered with a large watch-glass, gently warmed, and 
about 2 grams of tin-foil added in small portions at a time. 
The tin is rapidly oxidised, and the supernatant liquid 
becomes perfectly clear. The preliminary heating of the 
solution is absolutely necessary, since in the cold the metal 
is apt to become passive, when it resists the action of the 
acid. Care must be taken to keep the nitric acid in sufficient 
excess, in order to prevent the formation of hydrated mon- 
oxide, which renders the solution inconveniently turbid. 
When all action is at an end, and the tin fully oxidised, the 
contents of the dish are evaporated nearly to dryness, water 
is added, and the solution filtered The precipitate contains 



Ashes of Plants. 347 

all the phosphoric acid ; the bases are found in the filtrate. 
The precipitate, detached as far as possible from the filter, is 
digested in the smallest possible quantity of highly- concen- 
trated potash solution j on the addition of water the solution 
will become perfectly clear, provided no great excess of the 
alkali has been used. The small amount of the precipitate 
still adhering to the filter is also dissolved in a few drops of 
potash solution, and added to the main portion of the liquid. 
The mixture is then saturated with sulphuretted hydrogen, 
acetic or sulphuric acid added in very slight excess, and the 
precipitated tin sulphide separated by the filter-pump. The 
filtrate is concentrated to a small bulk, filtered from the 
slight amount of tin sulphide, which often separates on 
evaporation, and the phosphoric acid precipitated by mag- 
nesia-mixture and ammonia. The filtrate from the insoluble 
tin phosphate is treated with sulphuretted hydrogen to 
remove the lead with which the foil is frequently mixed, 
filtered, evaporated to a small bulk, boiled, ammonia added 
in slight excess, and the iron and alumina filtered off : they 
are separated as in No. XII. Part II. The filtrate from the 
precipitate by ammonia contains the manganese, lime, and 
magnesia. These are separated as in No. XIX. Part IV., 
p. 220. 



Quantitative Chemical Analysis. 



PART V. 
ORGANIC ANALYSIS. 

I. ANALYSIS OF BODIES CONTAINING CARBON AND HY- 
DROGEN, OR CARBON, HYDROGEN, AND OXYGEN. 

ORGANIC substances containing hydrogen, when heated 
with cupric oxide, are converted into carbon dioxide and 
water. By absorbing the products of the combustion in a 
suitably-arranged apparatus, and weighing them, we can 
readily calculate the amount of carbon and hydrogen in 
the substance analysed, from the knowledge that 44 parts 
of carbon dioxide contain 12 parts of carbon, and that 
1 8 parts of water contain 2 parts of hydrogen. If the sum 
of the amounts of carbon and hydrogen is equal to the 
weight of the body taken, the substance contains only these 
elements ; if the body contains oxygen in addition, the differ- 
ence indicates the amount of this constituent. 

Fig. 78 represents the apparatus in which the combustion 
may be conveniently made. The substance is burnt with 
cupric oxide by the aid of a current of oxygen or air. The 
gas-furnace is of the form known as Erlenmeyer's j it consists 
of 24 Bunsen-burners, each provided with a separate stop- 
cock worked by a little lever. The width of the air-passages 
in the burners may be regulated by a short piece of move- 
able tube, so that the amount of air passing into the tube 
may be altered at will. This arrangement serves to prevent 
the flame passing down to the burner at the bottom when 
the gas-current is feeble. The tubes end in a horizontal pipe, 
which is connected with the gas-supply by wide caoutchouc 
tubes. The flames strike against a semi-circular trough of 
well-baked fire-clay, resting on small clay supports ; in this 
trough is placed the combustion-tube. The side plates a, a 



Organic Analysis. 



349 




350 



Quantitative Chemical Analysis. 



FIG. 79. 




(fig. 79) are of clay ; they are moveable, and are supported 
upon a ledge running the entire length of the furnace. 

It will be seen from their peculiar shape 
(fig. 79) that the flames, after diverging 
from beneath the trough, strike against the 
sides ; the heat is thus reverberated, and 
the tube is uniformly and regularly heated. 
By the aid of the clamping screws a slight 
inclination may be given to the ledge on 
which the plates rest, or it may be raised 
or lowered above the burners. 

The following articles are needed to 
make a combustion by means of this appa- 
ratus : 

1. A Piece of Combustion-tube. This 
should be about 4 or 5 centimetres longer 

than the furnace ; and it should be about 2 millimetres 
thick in the glass, and about 12 or 14 millimetres in internal 
diameter. The sharp edges of the tube should be fused in 
the blowpipe flame, so that two caoutchouc stoppers, pierced 
with holes, may be introduced without being cut or torn. 

2. A Calcium Chloride Tube. This serves to 
absorb the water produced : it may conveniently 
be arranged as in fig. 80. It is furnished with 
two bulbs, a and b : in the small neck between 
the bulbs is fused a piece of thin glass tube 
projecting into the bulb a. By carefully regu- 
lating the heat, the greater portion of the water 
produced in the combustion condenses in a : 
if its quantity is not too considerable, it remains 
in this bulb when the tube is held perpendicularly 
with the bulbs uppermost. After having been 
weighed the water may readily be emptied out 
into a little capsule, and its purity tested by its 
taste, smell, action on litmus-paper, &c. A cal- 
cium chloride tube so arranged may be used for a great 



FIG. 80. 



Organic A nalysis. 3 5 1 

number of observations without replenishing, provided that 
on the conclusion of the experiment the bulb be emptied 
and the tube dried by the aid of a narrow roll of filter- 
paper. To fill the calcium chloride tube, place a loose 
plug of cotton-wool within the wide tube, close the end 
with the finger, and suck out the air at the narrow end. 
On suddenly removing the finger, the loose plug is driven 
into the larger bulb : repeat this operation until the long 
fibres of the wool are within the neck between a and b : 
these fibres tend to prevent the formation of drops in the 
narrow tube, and thus to promote the regularity of the 
passage of the gas through the potash bulbs. Fill the 
larger bulb with coarse fragments of spongy calcium chlo- 
ride, gently tapping the tube so as to shake the pieces 
together, and then add smaller pieces (not powder) until 
the tube is nearly filled ; insert a plug of cotton-wool and 
close the tube with a good, softened, tightly-fitting cork, 
through which passes a tube about 4 centimetres long and 
of the same diameter as the tube of the potash bulbs. Fuse 
the sharp edges of the tube before inserting it into the cork. 
After fitting the cork into the calcium chloride tube, cut 
the protruding portion with a sharp knife in the manner seen 
in fig. 80, and neatly cover the surface with sealing wax. 
Take care that the wax is uniformly melted and is in a co- 
herent piece, otherwise portions are apt to be detached in 
FIG. 81. handling the tube between the opera- 

tions of weighing ; the experiment may 
thus be nullified or rendered inexact. 
The ends should then be closed by 
short pieces of caoutchouc tube stopped 
with glass rod. 

3. The Potash Bulbs. This appa- 
ratus serves to absorb the carbon di- 
oxide. The form represented in fig. 81 
is that originally devised by Liebig (by whom, indeed, the 
method of organic analysis by combustion with cupric oxide 




352 Quantitative Chemical A nalysis. 

was first worked out). It is filled to the extent indicated by 
the dotted line in the figure, with strong potash solution, 
prepared by dissolving 3 parts of potash free from carbonate 
in 2 parts of water. The bulbs are readily filled with this 
liquid, contained in a porcelain dish, by dipping the end of 
the tube connected with the larger bulb beneath the surface 
of the potash solution and gently aspirating at the other tube 
until the required amount has been introduced. Carefully 
dry the tube, inside and out, with paper, and close the 
apparatus by short caoutchouc tubes fitted with glass rod. 
Twist a piece of platinum wire round the tubes where 
they touch, in the manner seen in the figure : this serves 
to suspend the bulbs from the hook of the balance-pan. 

The tube connected with the smaller bulb is adapted to a 
short and light drying tube, c (fig. 78), about 5 centimetres 
long, filled with soda-lime contained between loose plugs of 
cotton-wool, as in the calcium chloride tube. The cork is 
to be trimmed and covered with sealing wax in the man- 
ner already described. FIG. 82. FIG. 83. 
This apparatus serves 
to retain any carbon 
dioxide which may 
escape absorption in 
the potash bulbs : it 
is therefore weighed 
with the bulbs. 

Wipe the potash 
bulbs and the calcium chloride tube with a soft clean cloth 
and place them in the balance-case. 

Many other forms of the potash apparatus have been de- 
scribed. Fig. 82 represents a modification due to Geissler: 
it will be seen that the gas passes thrice through the 
potash solution. The apparatus requires no support and is 
readily filled and emptied. Fig. 83 shows a very simple 
form of potash bulbs, originally devised by Mitscherlich, and 
modified by De Koninck. This piece of apparatus is 




Organic Analysis. 353 

admirably adapted for washing or drying gases. Carbon 
dioxide may also be absorbed by soda-lime, as we have fre- 
quently had occasion to observe. This method of absorption 
is especially convenient if the carbon dioxide is mixed with 
comparatively large quantities of other gases. In such a 
case the potash apparatus is replaced by a U-tube rilled with 
soda-lime and calcium chloride, as described on p. 87. 

4. A Platinum Boat, to contain the substance to be 
analysed. This should be of such size as to pass readily 
into the tube. It may conveniently be 70 mm. long and 
8 mm. deep. 

5. Cupric Oxide. Strongly heat some clean copper scales 
in a muffle, and when they are sufficiently cool, transfer them 
to a porcelain basin and heat them with nitric acid (sp. gr. 
1-2). Evaporate the pasty mass to dryness on a sand-bath, 
pound it up and heat it strongly in a covered Hessian 
crucible. Break the crucible, carefully remove any pieces of 
clay, and coarsely powder the fused cupric oxide, pass the 
powder through a sieve of wire gauze, to separate the fine 
portions. The cupric oxide to be used in the analysis should 
be in little pieces about the size of hemp- seed. The finer por- 
tion should be preserved in a stoppered bottle : it is useful 
for the determination of nitrogen, as described hereafter. 

6. Copper Gauze and Wire. Roll two pieces of fine wire 
gauze, about 2 centimetres broad, into plugs of a size just 
sufficient to pass easily, but with a little friction, into the 
combustion-tube. Heat them in the Bunsen flame to re- 
move any adhering greasy matter, and when cold push one 
of them down about 25 centimetres into the combustion- 
tube, and fill up the tube from the other end with the 
coarsely-powdered cupric oxide, occasionally tapping it so 
as to shake the pieces as closely together as possible. When 
the tube is filled to within 6 centimetres of the end, insert 
the second plug of metallic copper. The layer of copper 
oxide should be about 54 centimetres in length : there is no 
necessity to leave a channel above it for the gases, since from 

A A 



354 Quantitative Chemical Analysis. 

the coarseness of the powder there is ample room for their 
escape. Over the end of the tube is placed a small circular 
disc of copper (fig. 78), readily moveable along the tube. 
This serves to protect the caoutchouc stopper and to shield 
the little bulb a of the calcium chloride tube from the heat : 
by moving it backwards or forwards along the tube, as occa- 
sion requires, the condensation within the tube of the water 
produced in the combustion may be entirely prevented. 

Cut another piece of the copper gauze, about 10 centi- 
metres broad, and of the same length as you have found 
suitable for the plugs, and roll it round a piece of stout 
copper wire, about 12 centimetres long; the one end of the 
wire should be bent sharply upon itself so as to hold the 
copper gauze firmly near one corner : the gauze is then 
turned over the wire along its entire length and wrapped 
round so as to form a cylinder, which easily passes into the 
combustion-tube. The other end of the wire should be 
bent so as to form a little ring of less diameter than the tube. 
In the combustion this long cylinder of gauze is placed 
behind the platinum boat containing the substance to be 
analysed. The vacant space in the tube, that is, the portion 
before the first copper plug, should be sufficiently large to 
hold the platinum boat and copper cylinder, and still leave 
room for the insertion of the cork. 

7. An Apparatus for drying and removing Carbon Dioxide 
from the Air and from Oxygen. This may be arranged as in 
fig. 78 : the lower neck of the cylinder is partially closed by a 
few fragments of glass, and the cylinder is half filled with 
soda-lime in coarse fragments : over this is placed a layer of 
cotton-wool, and the remainder is filled with calcium chloride 
in loose spongy pieces. The cylinder is closed by a caout- 
chouc stopper carrying a bent tube and leading to the two large 
U-tubes (a) and (a), also filled with calcium chloride. Before 
entering the cylinder, the air or oxygen traverses the wash- 
bottle b t containing strong solution of caustic potash. This 
removes the greater portion of any accompanying carbon 



Organic A nalysis. 355 

dioxide, and also serves to indicate the speed with which 
the gas passes into the combustion-tube. 

8. A Bell-jar fitted with a Cork and Calcium Chloride Tube. 
The bell-jar stands in a vessel containing water. By con< 
necting it with the potash apparatus in the manner seen in 
fig. 78 the pressure within the combustion-tube is decreased 
in proportion to the height of the column of water within the 
bell-jar above the level of that in the trough. 

The Process. When everything is arranged, gently heat the 
combustion-tube, having previously removed the platinum 
boat, and pass a slow current of dry air over the copper 
oxide to expel any hygroscopic moisture. Whilst the oxide 
is being heated, weigh the potash bulbs and calcium chloride 
tube without their caoutchouc stoppers', when you have de- 
termined their weight, replace the stoppers. In about 10 or 
15 minutes turn down the flames beneath the tube and allow 
the oxide to cool in the current of dry air. We will assume, 
by way of example, that you are about to analyse pure cane- 
sugar. This should be previously powdered and dried in 
the steam-bath. Heat the platinum boat to redness and 
allow it to cool in the desiccator. Weigh it and transfer 
about 0*4 grm. of the sugar to the boat, and again weigh : 
the increase in the weight of the boat shows the amount 
taken for analysis. The sugar may be accurately weighed 
in this manner, as it is not hygroscopic. Stop the air- 
current, and adapt the weighed calcium chloride tube to the 
combustion-tube by means of the caoutchouc stopper, and 
connect the potash bulbs, by a piece of well-fitting caoutchouc 
tubing, with the calcium chloride tube in the manner seen in 
fig. 78. There is no necessity to bind the caoutchouc to 
the glass tubes, since the reduced pressure within the appa- 
ratus, caused by the column of water in the bell-jar, effectu- 
ally prevents leakage. Partially fill the bell-jar with water 
by aspirating at the end of the tube, turn the stopcock 
so as to prevent the entrance of the air, and connect 
the caoutchouc tubing with the end of the soda-lime 

A A 2 



35^ Quantitative Chemical Analysis. 

tube of the potash apparatus (fig. 78). Remove the stopper 
at the further end of the combustion-tube, withdraw the 
cylinder of copper gauze (which will now be superficially 
oxidised), insert the platinum boat containing the weighed 
amount of the sugar, and replace the cylinder, pushing the 
boat nearly to the plug of copper gauze. Again -fit in the 
cork and connect the caoutchouc tube of the wash-bottle (b) 
with the gasometer containing the oxygen. Now cautiously 
open the stopcock of the bell-jar and incline the potash 
bulbs in the manner seen in the figure : the smaller bulb 
should be about half filled with the potash solution. Light 
the first 6 or 8 burners (beginning at the end nearest the 
calcium chloride tube), and gradually heat the tube. As it 
becomes red-hot, light successive burners until it is at a dull 
red heat. Now light the last two or three burners at the 
other end of the tube, immediately under the gauze cylinder, 
so as to heat it gently, and turn on a slow stream of oxygen 
(about a bubble every two seconds suffices at the commence- 
ment of the process). Continue to ignite successive burners 
so as to heat fresh portions of the copper oxide : when the 
tube is at a dull red heat to within 4 or 5 cm. of the 
platinum boat, turn on the gas in one of the burners imme- 
diately underneath the boat, and gently heat it. The sugar 
will quickly melt, become brown, and give off vapours. 
Carefully observe the movements of the liquid in the potash 
bulbs, and regulate the heat so as to preserve a uniform 
passage of gas into the bulbs. 

As soon as the sugar in the boat appears to be completely 
charred, and the amount of carbon dioxide passing into the 
bulbs becomes small, increase the heat beneath the boat (by 
this time the whole of the burners should be lighted), and 
send a slightly brisker current of the dry oxygen (about one 
bubble per second) through the apparatus. The carbo- 
naceous matter within the boat gradually burns : as soon as 
it has disappeared gradually diminish the flames underneath 
the gauze cylinder and platinum boat, turn on a little more 



Organ ic A nalysis. 357 

oxygen, and when the gas appears to pass unabsorbed through 
the potash bulbs gradually lower the flames along the entire 
length of the tube. Close the caoutchouc tube of the 
wash-bottle, and transfer it to the gasometer of air, and send 
a current of air through the tube to displace the oxygen. 
In a few minutes disconnect the potash bulbs and calcium 
chloride tube (taking care to hold the latter so that the 
water condensed in the smaller bulb does not flow out), fit 
in their respective stoppers, wipe them, re-weigh them (of 
course without the stoppers). Allow the combustion-tube to 
cool gradually: if care be taken to anneal it properly it 
will serve a great number of times without rearrangement. 
The heat need not be so high as to distort the tube : the 
great majority of carbonaceous substances, especially if they 
contain oxygen, burn with comparative ease in contact with 
copper oxide and free oxygen. The apparatus is ready for 
a second combustion : if the analysis of the sugar has to be 
repeated it is of course not necessary to wait until the copper 
oxide and tube are completely cold. 

In the analysis of volatile liquids the substance is weighed 
out in little bulbs of the shape seen in fig. 84. These are 

made from tube, obtained by drawing 
______ out a piece of wide glass tubing before 

the blowpipe until it is about 5 milli- 
metres in external diameter. The tube to contain the liquid 
should be about 30 millimetres in length in the wider portion : 
the narrow portion should be short enough to allow the 
tube to rest in the platinum boat. The tube is weighed and 
passed once or twice through the Bunsen flame, and whilst 
still hot the open end is plunged beneath the surface of the 
liquid to be analysed. As the tube cools the liquid is driven 
into it to replace the air expelled on warming. Withdraw 
the tube from the liquid and cause the small portion within 
the bulb to boil briskly so as to drive out all the air, and 
again plunge the end of the tube into the liquid. The bulb 
will now be almost completely filled. Except in the case of 



358 Quantitative Chemical Analysis. 

very volatile liquids it is unnecessary to seal the end of the 
capillary tube : the bulb containing the liquid may be accu- 
rately weighed and transferred to the combustion-tube without 
any appreciable loss from evaporation. The process of 
combustion does not differ in any essential particulars from 
that already described. It is advisable, however, to expel 
the liquid from the bulb before the copper oxide is heated 
in its vicinity. As soon as the copper oxide is red hot to 
within 15 centimetres of the end of the platinum boat, 
remove one of the heated clay plates, and place it immedi- 
ately over the boat : if the liquid is moderately volatile it 
will be readily and gradually expelled. If not, the bulb must 
be heated by a very small flame. The combustion of volatile 
liquids demands great care and attention ; the operation must 
not be hurried, or portions will escape unburnt. 

II. ANALYSIS OF ORGANIC SUBSTANCES CONTAINING 
NITROGEN. 

The determination of the several elements contained 
in an azotised organic compound cannot be very con- 
veniently made in a single operation. It is usually pre- 
ferred to estimate the carbon and hydrogen in one portion, 
and to determine the nitrogen in a second quantity. Nitro- 
genous organic substances when burnt with copper oxide, 
particularly if free oxygen be present, are apt to evolve 
nitroxygen compounds, which condense in the calcium 
chloride tube and potash bulbs, and vitiate the results of the 
carbon and hydrogen determinations. By passing the mixed 
products of combustion over heated metallic copper the 
nitroxygen compounds are decomposed; the oxygen combines 
with the copper, and the nitrogen passes unabsorbed through 
the apparatus. In the combustion of organic substances 
containing nitrogen it is necessary, therefore, to introduce a 
cylinder of copper gauze, about 12 centimetres long, rolled 
on a stout copper wire, exactly like that placed in the 
posterior part of the tube behind the platinum boat. This 



Determination of Nitrogen. 359 

is to be kept at a bright red heat during the operation : the 
carbon and hydrogen may then be determined accurately in 
the ordinary way ; or a length of from four to six inches of 
a mixture of potassium dichromate and manganic oxide is 
placed in the fore part of the tube and kept at a temperature of 
about 250. This mixture readily absorbs nitroxygen fumes. 

Determination of Nitrogen by Volume. Maxwell Simpson's 
Method. This process is applicable to all nitrogenous bodies, 
inorganic and organic. The substance is burnt by a mixture 
of cupric and mercuric oxides in a tube from which the air 
has previously been expelled by a current of carbon dioxide : 
the nitrogen and carbon dioxide, together with the excess of 
free oxygen, are passed over strongly-heated metallic copper, 
which retains the latter gas : the remaining gases are col- 
lected in an apparatus standing over mercury and partially 
filled with strong solution of caustic potash, which absorbs 
the carbon dioxide : the residual nitrogen is transferred to a 
measuring tube standing over mercury, and its volume is 
accurately determined. From the known weight of a litre 
of nitrogen the weight of the gas is readily calculated. 

A piece of strong combustion-tube about 80 centimetres 
long is sealed and rounded at one end like a test-tube. A 
mixture of 1 2 grams of manganous carbonate or magnesite 
dried at 100, and 2 grams of precipitated mercuric oxide are 
introduced into the tube. Insert a plug of recently-ignited 
asbestos, pushing it down to within 2 centimetres from the 
mixture, and afterwards add about i gram of the mercuric 
oxide. Weigh out about 0*6 gram of the nitrogenous substance 
to be analysed into a glazed porcelain mortar, and mix it with 
about 45 times its weight of a previously-prepared mixture 
of 4 parts of finely-powdered and recently-ignited cupric 
oxide and 5 parts of the dried mercuric oxide. Transfer the 
mixture to the tube without loss, and rinse the mortar with 
a fresh portion of the two oxides, adding the rinsings to the 
tube. Push down a second and thick plug of asbestos to 



360 



Quantitative Chemical Analysis. 



FIG. 85. 



within about 30 centimetres from the first, and then a layer, 
about 9 centimetres in length, of pure cupric oxide ; next a 
third asbestos plug, and lastly a layer not less than 20 centi- 
metres long of metallic copper, prepared by reducing gran- 
ular cupric oxide in a stream of carbon monoxide. Draw 
out the end of the combustion-tube before the blowpipe and 
connect it with the bent delivery tube #, which dips be- 
neath the surface of 
the mercury in the 
trough (fig. 85). Be- 
fore placing the tube 
in the furnace tap it 
gently on the table 
to shake down the 
several layers, in or- 
der to leave a chan- 
nel for the escaping 
gases. 

The vessel b has a 
capacity of about 200 
c.c. It is provided 
with a glass stopcock 
and bent delivery tube 
as represented in fig. 85. To ascertain if the stopcock is per- 
fectly air-tight, fill the apparatus completely with mercury 
and place it on its foot : any leakage will immediately reveal 
itself by the mercury flowing out of the tubulus. If the stop - 
cock is found to be tight replace about 20 c.c. of the 
mercury with a strong solution of caustic potash, and place 
the vessel in the mercurial trough with its tubulus beneath 
the surface of the metal. 

Heat a portion, say the posterior half of the manganous 
carbonate or magnesite, and drive out the air within the 
tube by a brisk current of carbon dioxide. At the same 
,time commence to heat the portion of the tube occupied by 
the metallic copper and pure cupric oxide. As soon as the 
escaping gas is free from air (which is readily ascertained by 




Determination of Nitrogen. 



361 



allowing a quantity to pass into a test-tube filled with potash 
solution, when no bubble should be left), and the anterior 
portion of the tube is well heated, insert the end of the 
delivery tube through the tubulus of the vessel, and gra- 
dually heat successive portions of the tube occupied by the 
mixture of nitrogenous substance, cupric and mercuric oxides, 
beginning with the part nearest to the pure cupric oxide- 
As soon as no further evolution of gas is observed, and the 
whole length of the tube (with the exception of the part 
occupied by the undecomposed magnesite or manganous 
carbonate) is at a bright red heat, heat the remainder, so as 
to cause a rapid evolution of carbon dioxide, by which the 



FIG. 86. 




nitrogen still existing in the tube is expelled. Withdraw the 
delivery tube from the tubulus, and allow the gas in b to re- 
main over the caustic potash solution for about an hour to ab 
sorb the last traces of the carbon dioxide. The pure nitrogen 
has now to be transferred to a measuring tube in order that its 
volume may be determined. The bent tube c, fig. 86, which 
is contracted at d, is fitted by the aid of a caoutchouc 
stopper into the tubulus, underneath the surface of the mer- 
cury in the trough. To prevent the possibility of any air 



362 



Quantitative Chemical Analysis. 



FIG. 87. 



adhering to the stopper and so finding its way into the 
nitrogen, it is advisable to moisten the stopper with a solu- 
tion of corrosive sublimate before inserting it into the 
tubulus. Fill up the bent tube with mercury and remove b 
from the trough. Place a drop of water in the measuring tube, 
fill it with mercury, and invert it beneath the surface of the 
metal in the trough. Place the end of the delivery tube e 
beneath the measuring tube, cautiously turn the stopcock, 
and allow the gas to escape from b. When the level of the 
mercury in c approaches the contracted portion, close the 
stopcock, refill the tube with the metal, 
and reopen the stopcock, and so gra- 
dually transfer the nitrogen into the 
measuring tube, closing the stopcock as 
soon as the potash solution touches it. 
Of course the delivery tube is thus left 
filled with nitrogen, but as an identical 
volume of air (viz. that which originally 
filled it) has been transferred to the 
measuring tube, no error is committed. 
Read off the volume of the moist gas 
and correct it for pressure, tension of 
aqueous vapour, and temperature, and 
calculate the weight of the nitrogen : a 
litre of nitrogen under the standard 
conditions of temperature and pressure 
weighs i -2 5 5 gram. 

The above method of measuring the 
volume of the nitrogen may be much 
simplified by the use of the apparatus 
devised by Hugo Schiff,* and repre- 
sented in fig. 87. The burette A, which is fitted with a glass 
stopcock, c, contains about 120 c.c. down to the side tube a, 
and stands in a wooden foot, which may be rendered more 




Fresenius, Zeitschrift fiir anal. Chemie, p. 430. 1868 



Determination of Nitrogen. 363 

stable by being weighted with lead. At about 2 centimetres 
beneath the side tube a, is a second tubulus, ^, inclined 
upwards in the manner seen in the figure. Through this tube 
is poured mercury to a height of 2 or 3 millimetres above 
the lower opening. The vessel B, holding from 150 to 170 
c.c., is supported by a metallic ring attached to the clamp e : 
and may thus be readily placed at any desired height along 
the burette : B is connected by a strong caoutchouc tubing, 
previously soaked in melted paraffin, with the side tube a. 
B is filled with a strong solution of potassium hydrate Of sp. 
gr. 1-5, prepared by dissolving potash in an equal weight of 
water : its neck is closed by a cork, in which a narrow 
opening is cut. On closing the tubulus b with a cork, and 
on opening the stopcock and raising B, the potash solution 
flows over into the burette and completely fills it. The 
stopcock is now closed and the vessel B is lowered nearly 
to the foot of the burette : the stopper may then be with- 
drawn from b without the mercury being forced out. 

The delivery tube of the combustion-tube is then pushed 
through b as soon as all the air has been expelled. The 
volume of the nitrogen is then directly measured, the vessel 
B being raised until the levels of the potash solution in 
both pieces of the apparatus are coincident. The nitrogen 
may without sensible error be assumed to be dry: the 
amount of moisture present in it is probably never more 
than o'oo7 of its volume. This additive quantity serves 
in some measure to compensate for the deficit in the 
amount of nitrogen obtained, due to the impossibility of 
entirely preventing the formation of nitroxygen compounds 
in the process of combustion. 

Estimation of Nitrogen as Ammonia. By Burning with 
Soda-lime. This process, which is applicable to all nitro- 
genous substances excepting the so-called nitro-compounds, 
e.g. nitre-benzol, amyl nitrate, &c., has already been described 
on P- 334- 



364 Quantitative Chemical Analysis. 

III. ANALYSIS OF ORGANIC SUBSTANCES CONTAINING 
CHLORINE, BROMINE, AND IODINE. 

When an organic compound containing a halogen, chlorine, 
for example, is burnt with cupric oxide, cuprous chloride is 
formed, which, being volatile, is carried forward in the 
stream of gas and condenses in the calcium chloride tube, 
and thus renders the determination of the hydrogen in- 
exact. If the gases within the tube contain free oxygen 
the cuprous chloride is more or less decomposed, cupric 
oxide being formed, and chlorine eliminated. This is 
retained partly by the calcium chloride, partly by the 
potash solution. By inserting a cylinder of copper gauze 
in the anterior portion of the tube, the chlorine may be 
arrested so long as the amount of oxygen is not sufficient to 
oxidise the copper. By mixing the cupric oxide with a 
small quantity of lead oxide the chlorine may be entirely 
retained. 

The determination of the carbon and hydrogen in com- 
pounds containing chlorine is best effected by heating with 
lead chromate. This substance is readily made by mixing 
potassium chromate and lead nitrate or acetate solutions, 
thoroughly washing the dense yellow precipitate, drying it, 
heating it to redness in a covered clay crucible, and coarsely 
powdering it. The combustion is made in the manner 
already described in the case of copper oxide. 

Determination of the Halogen. A narrow piece of com- 
bustion-tube about 40 centimetres long is sealed and rounded 
at the end like a test-tube. A small quantity of coarsely- 
powdered and recently-burnt lime is introduced into it, so as 
to occupy a length of 4 centimetres. The compound to be 
analysed, if solid, is weighed out into the combustion-tube, 
and mixed with a quan- FlG 88 

tity of the lime in mo- 
derately fine powder, 
by the aid of a brass wire bent in the manner seen in fig. 88. 



Determination of the Halogen. 365 

By twisting this wire among the fragments the substance 
and the lime are uniformly mixed. The wire is rinsed 
from any adhering powder by a further quantity of lime, and 
the tube is filled with the coarsely-powdered lime to within 
about 3 or 4 centimetres from the open end, placed in the 
furnace and closed by a cork carrying a short piece of bent 
tube, which dips beneath the surface of water contained in 
a small beaker. This serves to maintain a slight pressure 
within the tube and tends to prevent the escape of any of 
the halogen. Commence the operation by heating the an- 
terior portion of the tube, and gradually approach the part 
containing the substance as the lime becomes red-hot. 
Having lighted all the burners beneath it, continue to heat 
the tube until the cessation of gas bubbling through the 
water tells you that the process is finished. When the tube 
is cold, empty the loose fragments of the lime into about 
150 c.c. of water, and half fill the tube with water to dissolve 
any fused substance adhering to the glass. Acidify the liquid 
with moderately dilute nitric acid free from chlorine : an 
excess of nitric acid is readily indicated by the change in 
the colour of the suspended carbonaceous matter. Immedi- 
ately all the lime is dissolved the precipitate becomes quite 
black. The liquid is filtered and treated with silver nitrate 
solution, and the precipitated silver salt washed, dried, and 
weighed. Of course the quantity of the chlorine may be 
estimated volumetrically by standard silver solution and 
potassium chromate if care be taken to neutralise the excess 
of nitric acid by well- washed precipitated calcium carbonate, 
or by the addition of sodium carbonate solution. 

Liquids containing chlorine, &c., are weighed out in bulbs, 
as described on p. 354 : after the introduction of a layer of 
lime, about 4 centimetres long, the bulb is allowed to slide 
down the tube, which is then immediately filled up with lime. 
When about half the length of the tube has been heated, ex- 
pel the liquid from the Ivlb by gently heating the tube where 



366 Quantitative Chemical Analysis. 

it is situated, and conduct the remainder of the operation as 
described. 

Many organic substances containing a halogen may be 
very conveniently analysed by digesting them with water and 
sodium amalgam. The liquid poured off the residual mer- 
cury is acidified with nitric acid and the chlorine determined 
in the usual manner. 

Certain organic iodides are decomposed by heating them 
with an alcoholic solution of silver nitrate : the silver iodide 
thus formed may be filtered off, dried, and weighed. 

IV. ANALYSIS OF ORGANIC SUBSTANCES CONTAINING 
SULPHUR AND PHOSPHORUS. 

The combustion of organic bodies containing sulphur is 
most accurately made with lead chromate: the only pre- 
caution needed is to maintain the anterior portion of the 
tube, to the extent of 15 or 20 centimetres, at a very low red 
heat only. Under these circumstances no sulphur dioxide 
passes into the absorption apparatus, 

Determination of Sulphur. Solid substances containing 
sulphur may be decomposed by fusion with potassium hydrate 
and pure nitre. Place a quantity of potassium hydrate in a 
silver dish, mix it with about J of its weight of nitre and fuse 
the mixture. Allow it to cool and add to it the weighed 
quantity of the sulphur compound. Heat gently, and stir 
continually with a silver spatula, adding little by little a small 
quantity of nitre if the carbon appears to be but slowly con- 
sumed. When the mass is cold, dissolve it in water, acidify 
with hydrochloric acid, boil, and add barium chloride. 
Treat the precipitated barium sulphate in the manner de- 
scribed on p. 169. 

Solid compounds of sulphur may also be analysed by 
digesting them with strong potash solution contained in a 
large porcelain crucible, and passing a stream of chlorine into 



Sulphur and Pliosphorus. 367 

the liquid until the substance is completely decomposed. 
Acidify, heat gently to expel excess of chlorine, filter, and 
add barium chloride. 

Camus' Method. Applicable to the Estimation of Sulphur 
and Phosphorus in Solid and Liquid Substances. The com- 
pound is oxidised by the action of nitric acid of sp. gr. 
1-2. From 0-2 gram to 0-4 gram of the substance is 
weighed out in a thin glass-bulb, care being taken 
that but little air is enclosed within the bulb. The 
sealed bulb is brought into a tube of hard glass of about 
10 or 12 millimetres in internal diameter, sealed and rounded 
at one end like a test-tube, together with from 20 to 60 times 
its weight of nitric acid of sp. gr. 1-2. The tube must not 
be more than half filled with the liquid. It is now softened 
in the blowpipe flame at a few centimetres from the open 
end, and the fused glass allowed to thicken, and it is then 
drawn out into a thick-walled capillary tube. The tube is 
supported in the clamp of a retort stand, and the nitric acid 
caused to boil, so as to expel the air contained within the 
tube : when the acid vapours are freely evolved, the lamp is 
removed, and the capillary opening is closed by the blowpipe 
flame. Allow the liquid to become nearly cold, wrap the 
tube in a thick towel (for safety), and break the bulb by 
shaking it smartly against the ends of the tube. Heat the 
tube to 120-150 for some hours in the air-bath. Allow the 
bath to cool before withdrawing the tube, wrap it in the 
towel, and cautiously warm the point, so as to expel the liquid 
which collects in the capillary tube. Soften the end in the 
blowpipe flame : the enclosed gases will force their way 
through the fused glass. Examine the tube carefully, and if 
you have reason to believe that the oxidation is incomplete, 
re-seal the tube and heat it to 180 for an hour. Allow it to 
cool and open it with the same precautions as before. If no 
more gas escapes the process is finished. Cut off the end of 



368 Quantitative Chemical A nalysis. 

the tube, rinse its contents into a beaker, dilute with water, 
and, in the case of sulphur, add barium chloride. In the 
case of phosphorus, add ammonia, ammonium chloride, and 
magnesia- mixture, and convert the precipitate into magnesium 
pyrophosphate. If sulphur and phosphorus are together 
present, precipitate the sulphuric acid with barium chloride, 
remove the excess of baryta by sulphuric acid, concentrate 
the filtrate by evaporation, and determine the phosphoric 
acid as magnesium pyrophosphate. 



APPENDIX. 



TABLE I. 
Symbols and Atomic Weights of the Elements. 



Element 


Symbol 


Atomic weight 


Observer 


Aluminium 


Al 


27-02 


Mallet 


Antimony 


Sb 


II9-6 


Schneider ; Cooke 


Arsenic . 


As 


74*9 


Kessler 


Barium . 


Ba 


136-84 


Marignac 


Bismuth . 


Bi 


207-5 


Dumas 


Boron 


B 


10-9 


Berzelius 


Bromine . 


Br 


79-76 


Stas 


Cadmium 


Cd 


1117 


Lenssen 


C cesium . 


Cs 


132-7 


Johnson and Allen ; Bunsen 


Calcium . 


Ca 


39-90 


Erdmann and Marchand 


Carbon . 


C 


11-97 


Dumas and Stas ; Liebig 


Cerium . 


Ce 


138-24 


Rammelsberg 


Chlorine . 


Cl 


35'37 


Stas 


Chromium 


Cr 


52-08 


Siewert 


Cobalt . 


Co 


58-6 


Russell 


Copper . 


Cu 


63-12 


Millon and Commaille 


Didymium 


D 


142-44 


Hermann 


Erbium . 


E 


168-9 


Bahr and Bunsen 


Fluorine . 


F 


18-96 


Luca ; Louyet 


Gallium . 


Ga 


69-8 


Lecoq de Boisbaudran 


Glucinum 


Gl 


9-30 


Awdejew ; Klatzo 


Gold 


Au 


196-85 


Thorpe and Laurie 


Hydrogen 


H 


i 


Dulong and Berzelius 


Indium . 


In 


H3'4 


Winkler; Bunsen 


Iodine 


I 


126-54 


Stas 


Indium . 


Ir 


192-5 


Seubert 








f 



B "3 



370 



Appendix. 
TABLE I. continued. 



Element 


Symbol 


Atomic weight 


Observer 


Iron 


Fe 


55'9 


Dumas 


Lanthanum 


La 


I39-33 


Hermann 


Lead . 


Pb 


2O6 '40 


Stas 


Lithium . . . 


Li 


7'00 


Stas 


Magnesium 


Mg 


23-94 


Dumas 


Manganese 


Mn 


54'8 


Dewar and Scott 


Mercury . 
Molybdenum . 
Nickel . ." 


Hg 
Mo 

Ni 


199-8 

95-9 
58-6 


Erdmann and Marchand 
Dumas ; Debray 
Russell 


Niobium . 


Nb 


93'7 


Marignac 


Nitrogen . 


N 


14-01 


Stas 


Osmium . ' . 


Os 


190-8 


Seubert 


Oxygen . . 





15-96 


Nilson 


Palladium J 


Pd 


106-2 


Berzelius 


Phosphorus 


P 


30-96 


Schrotter 


Platinum. ' . 


Pt 


194-38 


Seubert 


Potassium v . 


K 


39-04 


Stas 


Rhodium 


Rh 


104-1 


Berzelius 


Rubidium v . 


Rb 


85-2 


Bunsen ; Piccard 


Ruthenium 


Ru 


I03-5 


Berzelius 


Scandium 


Sc 


44-0 


Nilson 


Selenium . 


Se 


78-9 


Dumas 


Silver 


Ag 


107-67 


Stas 


Silicon . 


Si ' 


28-33 


Thorpe and Young 


Sodium . 


Na 


22-99 


Stas 


Strontium 


Sr 


8734 


Marignac 


Sulphur . 


S 


31-996 


Stas 


Tantalum 


Ta 


182-00 


Marignac 


Tellurium . . 


Te 


125-0 


Brauner 


Thallium 


Tl 


203-50 


Crookes 


Thorium . 


Th 


231-44 


Delafontaine 


Tin. . . 


Sn 


117-4 


Dumas 


Titanium 


Ti 


48-0 


Thorpe 


Tungsten 


W 


183-6 


Schneider ; Dumas; Roscoe 


Uranium . 


U 


239-8 


Ebelmen 


Vanadium 


V 


51-0 


Roscoe 


Ytterbium 


Yt 


173-0 


Nilson 


Yttrium . 


Y 


88-9 


Bahr and Bunsen 


Zinc 


Zn 


647 


Axel Erdmann 


Zirconium 


Zr 


90-4 


Marignac ; Bailey 



Appendix. 



37* 



TABLE II. 
Volume and Density of Water at different Temperatures. 

(Mean results of the observations of Kopp, Pierre, Despretz, Hagen, 
Matthiessen, Weidner, Kremers, and Rossetti.) 



Temp. 


Sp. gr. of Water 

(at o = i) 


Vol. of Water 
(at o = i) 


Sp. gr. of Water 
(at 4 = i) 


Volume of Water 
(at 4 =i) 


O 


I -000000 


I -OOOOOO 


999871 


000129 


I 


i -00005 7 


o '999943 


999928 


000072 


2 


1-000098 


999902 


999969. 


OOOO3I 


3 


I -0001 20 


999880 


999991 


OOOOO9 


4 


1-000129 


999871 


I -000000 


oooooo 


5 


1-000119 


9 9988l 


0-999990 


ooooio 


6 


I -000099 


999901 


999970 


000030 


7 


I -000062 


999938 


'999933 


000067 


8 


1-000015 


999985 


999886 


1-000114 


9 


0-999953 


I -000047 


999824 


1-000176 


10 


999876 


I-OOOI24 


'999747 


I -000253 


ii 


999784 


I -OOO2I6 


999655 


I -000345 


12 


999678 


I -000322 


'999549 


1-000451 


13 


'999559 


I-OOO44I 


999430 


1-000570 


H 


999429 


I-000572 


999299 


I -000701 


15 


999289 


I-0007I2 


999160 


I -000841 


16 


999131 


I -000870 


999002 


I -000999 


17 


998970 


I-OOIO3I 


998841 


I -001 1 60 


18 


998782 


I-OOI2I9 


998654 


1-001348 


19 


998588 


1-001413 


998460 


1-001542 


20 


998388 


I-OOI6I5 


998259 


1-001744 


21 


998176 


I -001828 


998047 


1-001957 


22 


'997953 


I -002049 


997826 


1-002177 


23 


997730 


I -002276 


997601 


I -002405 


2 4 


'997495 


I -0025 1 1 


997367 


1-002641 


25 


997249 


I -002759 


997120 


I -002888 


26 


996994 


I -003014 


996866 


I -003144 


27 


996732 


I -003278 


996603 


I -003408 


28 


996460 


1 -003553 


996331 


I -003682 


2 9 


996179 


1-003835 


996051 


I -003965 


30 


995894 


I-004I23 


995765 


I -004253 


35 


99431 


I-00572 


99418 


I -00586 


40 


99248 


1-00757 


99235 


I -00770 


5o 


98833 


I -OIlSl 


98820 


1-01195 


60 


98351 


I-OI677 


98338 


1-01691 


70 


97807 


I -O2243 


'97794 


I -02256 


80 


97206 


I -02874 


97194 


I -02887 


90 


96568 


1-03554 


96556 


I -03567 


100 


95878 


I -04300 


95865 


1-04312 



B B 2 



372 



Appendix. 



TABLE 
Tension of Aqueous Vapour in 



i 































Mm. 


o 


Mm. 


o 


Mm. 


o 


Mm. 


o 


Mm. 





Mm. 





Mm. 


o'o 


4-600 


2-5 


5 '49i 


S'o 


6 '534 


7 '5 


7'75i 


10 '0 


9'i6 S 


12-5 


10 '804 


15 '0 


12-699 


'I 


633 


6 


'530 


'i 


580 


6 


804 


'i 


'227 


6 


8 7 5 


i 


781 


'2 


667 


'7 


i 


*2 


625 


'7 


'857 


*2 


288 


'7 


'947 


'2 


864 


'3 


700 


8 


608 


'3 


671 


8 


910 


'3 


'350 


8 


1 1 '019 


'3 


'947 


'4 


'733 


'9 


647 


'4 


717 


'9 


964 


'4 


412 


'9 


090 


4 


13-029 


"5 


767 


3' 


687 


'5 


'763 


8'o 


8-017 


'5 


'474 


13 '0 


162 


5 


"112 


6 


801 


'i 


'727 


6 


'810 


'i 


072 


6 


'537 


i 


'235 


6 


-197 


'7 


836 


*2 


767 


'7 


857 


'2 


126 


'7 


601 


"2 


'309 


'7 


28l 


8 


871 


'3 


807 


*8 


904 


'3 


181 


'8 


665 


'3 


'383 


8 


3 66 


'9 


'90S 


'4 


848 


'9 


'951 


'4 


236 


'9 


"728 


'4 


456 


'9 


451 


I'O 


940 


'5 


889 


6'o 


998 


"5 


291 


I I'O 


'792 


'5 


'530 


i6"o 


536 


'i 


'975 


6 


'930 


i 


7-047 


6 


'347 


'I 


857 


6 


'605 


'i 


623 


'2 


S'on 


'7 


'972 


'2 


'095 


'7 


404 


'2 


'923 


'7 


68 1 


'2 


'710 


'3 


'047 


8 


6*014 


'3 


'144 


8 


461 


'3 


'989 


'8 


'757 


'3 


'797 


4 

'5 


'082 
118 


y 
4-0 


'055 
097 


'4 

'5 


'242 


9 
9'o 


'574 


4 
'5 


10-054 

'120 


y 
14-0 


832 
908 


'4 

*5 


885 
'972 


'6 


i55 


I 


140 


*6 


292 


'I 


'632 


6 


l8 7 


'I 


986 


6 


14*062 


'7 


'191 


'2 


183 


*7 


'342 


"2 


"690 


'7 


255 


'2 


12-064 


"7 


151 


8 


228 


3 


226 


*8 


'392 


'3 


748 


8 


322 


3 


142 


8 


-2 4 I 


'9 


'265 


'4 


270 


'9 


442 


"4 


807 


'9 


389 


'4 


'220 


'9 


'331 


2'0 


'302 


'5 


'3i3 


7'o 


'492 


'5 


865 


I2'0 


'457 


"5 


298 


17-0 


'421 


'I 


"340 


6 


'357 


'I 


'544 


6 


925 


'I 


526 


6 


378 


'i 


'5*3 


'2 


378 


'7 


407 


*2 


'595 


'7 


'985 


'2 


'596 


'7 


458 


'2 


605 


'3 


416 


8 


'445 


3 


'647 


8 


9 '45 


'3 


'665 


'8 


538 


'3 


697 


'4 


'454 


'9 


490 


'4 


699 


'9 


105 


"4 


'734 


'9 


'619 


'4 


790 



Appendix. 



373 



III. 

Millimetres of Mercury from o to 34*9 C. 






Mm. 





Mm. 


o 


Mm. 





Mm. 


o 


Mm. 





Mm.. 





Mm. 


i? '5 


14-882 


20 '0 


17 '391 


22-5 


20-265 


25 'o 


23 '55o 


27 '5 


27-294 


30 "o 


3i'548 


32 '5 


36 '37 


6 


'977 


'I 


500 
'608 


6 


389 


'i 


692 
"874. 


6 


'455 


'I 


'729 


6 


576 
7 g.j 


'7 
8 


[5 OJS 

167 


'3 


717 


7 

8 


5*4 
'639 


'3 


34 
976 


7 
8 


778 


3 


32*094 


8 


73 
991 


'9 


262 


'4 


826 


'9 


763 


'4 


24-119 


'9 


'939 


'4 


278 


'9 


37-200 


i8'o 


'357 


'5 


'935 


23 'o 


888 


'5 


261 


28-0 


28'IOI 


'5 


'463 


33 'o 


410 


'i 


'454 


6 


18-047 


'i 


21 'Ol6 


6 


406 


'i 


267 


6 


'650 


"i 


621 


"2 


'552 


"7 


i59 


- 2 


144 


'7 


'552 


"2 


'433 


'7 


'837 


'2 


'832 


'3 


'650 


8 


271 


'3 


272 


8 


697 


'3 


'599 


8 


33-026 


'3 


38 '045 


'4 


'747 


'9 


383 


'4 


400 


'9 


'842 


'4 


'765 


'9 


'215 


'4 


'258 


'5 


845 


21 '0 


'495 


'5 


528 


26*0 


988 


'5 


'93 1 


31 'o 


'405 


'5 


'473 


6 

'7 


'945 


'I 

*2 


"610 


6 


659 


'i 


25'i38 


6 


2g'ioi 


'* 


596 
787 


6 
7 


689 
'006 


/ 

8 


10 045 

145 


'3 


839 


8 


921 


"3 


'438 


8 


441 


'3 


y / 
980 


7 

8 


yoo 
39'i24 


'9 


'246 


'4 


'954 


'9 


22-058 


'4 


588 


'9 


612 


'4 


34'i74 


'9 


'344 


19-0 


346 


'5 


19-069 


24-0 


184 


'5 


738 


29-0 


782 


'5 


'368 


34 'o 


565 


i 


449 


6 


187 


'i 


319 


6 


891 


'i 


956 


6 


'564 


"i 


786 


'2 


552 


'7 


'305 


'2 


'453 


'7 


26-045 


'2 


30-131 


"7 


7 6r 


'2 


40-007 


'3 


655 


8 


'423 


'3 


588 


8 


198 


'3 


'305 


8 


'959 


'3 


'230 


'4 


758 


*9 


'S4i 


'4 


'723 


'9 


'351 


'4 


'479 


'9 


35'i59 


'4 


'455 


'5 


861 


22 '0 


'659 


'5 


858 


27-0 


'SOS 


'5 


654 


32 'o 


'359 


'5 


680 


6 


967 


'I 


780 


6 


996 


'i 


663 


6 


833 


'i 


'559 


6 


907 


'7 


!7'073 


*2 


901 


'7 


23 <:I 35 


'2 


'820 


'7 


31 'on 


'2 


760 


'7 


4i'i35 


8 


179 


'3 


2O "O2 2 


8 


'273 


'3 


978 


8 


190 


'3 


962 


8 


'364 


'9 


285 


'4 


' r 43 


'9 


411 


'4 


27-136 


'9 


369 


'4 

1 


36-165 


'9 


'595 



374 



Appendix. 



TABLE IV. 

BAUME'S HYDROMETER. 
Table for Liquids heavier than Water. 



Degrees 
Baume 


Sp.gr. 


OB. 


Sp. gr. 


R 


Sp. gr. 





I -000 


26 


206 


52 


1-520 


I 


1-007 


27 


216 


53 


i -535 


2 


I-OI3 


2 


226 


54 


1-551 


3 


1-020 


29 


-236 


55 


1-567 


4 


1-027 


30 


246 


56 




5 


1-034 


31 


256 


57 


i -600 


6 


I-04I 


32 


267 


58 


1-617 


7 


1-048 


33 


277 


59 


1-634 


8 


1-056 


34 


288 


60 


1-652 


9 


1-063 


35 


299 


61 


1-670 


10 


I-O7O 


36 


310 


62 


1-689 


ii 


I-078 


37 


322 


63 


1-708 


12 


I -086 


38 


333 


64 


1-727 


13 


094 


39 


'345 


65 


1-747 


14 


101 


40 


'357 


66 


1-767 


15 


109 




-369 


67 


1-788 


16 


118 


42 


382 


68 


1-809 


17 


126 


43 


395 


69 


1-831 


18 


134 


44 


407 


70 


i -854 


19 


143 


45 


421 




1-877 


20 


152 


46 


'434 


72 


1-900 


21 


160 


47 


448 


73 


1-924 


22 


169 


48 


462 


74 


1-949 


23 


178 


49 


476 


75 


1-974 


24 


188 


50 


490 


76 


2-OOO 


25 


. '197 


51 


505 







Appendix. 



375 



TABLE IV. continued. 
Table for Liquids lighter than Water. 



B. 


Sp. gr. 


B. 


Sp. gr. 


B. 


Sp. gr. 


IO 


I -000 


27 


0-896 


44 


0-811 


II 


0-993 


28 


0-890 


45 


0-807 


12 


0-986 


29 


0-885 


46 


0-802 


13 


0-980 


3 


0-880 


47 


0-798 


H 


0-973 


3i 


0-874 


48 


0-794 


15 


0-967 


32 


0-869 


49 


0-789 


16 


0-960 


33 


0-864 


50 


0-785 


17 


o-954 


34 


0-859 


5i 


0-781 


18 


0-948 


35 


0-854 


5 2 


0-777 


19 


0-942 


36 


0-849 


53 


0-773 


20 


0-936 


37 


0-844 


54 


0-768 


21 


0-930 


38 


0-839 


55 


0-764 


22 


0-924 


39 


0-834 


56 


0-760 


23 


0-918 


40 


0-830 


57 


0757 


24 


0-913 


4i 


0-825 


58 


o-753 


25 


0-907 


42 


0-820 


59 


0-749 


26 


0-901 


43 


0-816 


60 


0745 



TWADDELL'S HYDROMETER. 

To convert degrees Twaddell into specific gravity (water 
multiply the number by 5, and add 1,000 to the product. 



i.ooo) : 



To reduce specific gravity (water = 1,000) to Twaddell : deduct 
1,000 and divide the remainder by 5. 



376 



Appendix. 



TABLE V. 

Showing the Percentages of real Sulphuric Acid 

corresponding to various Specific Gravities of Aqueoits Sul 
phuric Acid. 

^ineau j Otto. Temp. 15. 



Specific 
gravity 


Per 

cent. 


Specific 
gravity 


Per 

cent. 


Specific 
gravity 


Per 

cent. 


Specific 
gravity 


Per 

cent. 


I "8426 


100 


1-675 


75 


398 


50 


I-182 


25 


1-842 


99 


1-663 


74 


3886 


49 


I-I74 


24 


I '8406 


98 


I-6 5 I 


73 


379 


48 


I6 7 


23 


1-840 


97 


1-639 


72 


370 


47 


159 


22 


1-8384 


96 


1-627 


7i 


361 


46 


1516 


21 


1-8376 


95 


1-615 


70 


35i 


45 


144 


2O 


I-83S6 


94 


1-604 


69 


342 


44 


I 3 6 


19 


1-834 


93 


I-592 


68 


'333 


43 


129 


18 


1-831 


92 


1-580 


67 


324 


42 


121 


17 


1-827 


9i 


I-568 


66 


315 


41 


1136 


16 


1-822 


90 


i'557 


65 


306 


40 


106 


15 


1-816 


89 . 


1-545 


64 


2976 


39 


098 


14 


1-809 


88 


1-534 


63 


289 


38 


091 


13 


1-802 


87 


1-523 


62 


281 


37 


1-083 


12 


i -794 


86 


1-512 


61 


272 


36 


1-0756 


II 


i -786 


85 


1-501 


60 


264 


35 


1-068 


10 


1-777 


84 


1-490 


59 


256 


34 


1-061 


9 


1-767 


83 


1-480 


58 


2476 


33 


i -0536 


8 


1-756 


82 


1-469 


57 


239 


32 


i -0464 


7 


1-745 


81 


1-4586 


56 


231 


3i 


1-039 


6 


1-734 


80 


1-448 


55 


223 


30 


1-032 


5 


1-722 


79 


I-438 


54 


215 


29 


i -0256 


4 


1-710 


78 


1-428 


53 


2066 


28 


1-019 


3 


1-698 


77 


1-418 


5 2 


198 


27 


1-013 


2 


1-686 


76 


1-408 


5i 


190 


26 


1-0064 


I 



Appendix. 



377 



TABLE VI. 

Giving the Percentage Amount of Hydrochloric Add contained 
in Aqueous Solutions of the Gas of various Specific Gravities. 

Ure. Temp. 15. 



Specific 


HC1 


Specific 


HC1 


Specific 


HC1 


Specific 


HC1 


gravity 


per cent. 


gravity 


percent. 


gravity 


per cent. 


gravity 


per cent. 


2OOO 


40777 


1515 


30-582 


I -1000 


20-388 


I -0497 


10-194 


1982 


40-369 


1494 


30-174 


I -0980 


19-980 


I -0477 


9-786 


1964 


39-961 


1473 


29-767 


I -0960 


19-572 


1-0457 


9379 


1946 


39-554 


1452 


29359 


I -0939 


19-165 


I -0437 


8-971 


1928 


39-146 


1431 


28-951 


I -0919 


18-757 


I-04I7 


8-563 


T9IO 


38-738 


1410 


28-544 


I -0899 


I8-349 


I -0397 


8-155 


I8 93 


38-33 


1389 


28-136 


I -0879 


17-941 


I -0377 


7-747 


1875 


37-923 


1369 


2 7 -728 


I -0859 


I7-534 


1-0357 


7340 


1857 


37-5I6 


1349 


2732I 


I -0838 


17-126 


I -0337 


6-932 


1846 


37-108 


1328 


26-913 


I -0818 


16-718 


I-03I8 


6-524 


1822 


36-700 


1308 


26-505 


I -0798 


I6-3IO 


I -0298 


6-II6 


1802 


36-292 


1287 


26-098 


1-0778 


15-902 


I -0279 


5709 


1782 


35-884 


1267 


25-690 


I -0758 


I5-494 


I -0259 


5-301 


1762 


35-476 


1247 


25-282 


I -0738 


15-087 


I -0239 


4^93 


1741 


35-068 


1226 


24^74 


1-0718 


14-679 


I -0220 


4-486 


1721 


34-660 


I2O6 


24-466 


0697 


14-271 


1-0200 


4-078 


I7OI 


34-25 2 


Il85 


24-058 


0677 


13-863 


I'OlSo 


3^70 


1681 


33-845 


1164 


23-650 


0657 


13-456 


I 'Ol6o 


3-262 


1661 


33-437 


1143 


23-242 


0637 


13-049 


I -0140 


2-854 


1641 


33-029 


U23 


22-834 


0617 


12-641 


I -01 20 


2-447 


1620 


32-62I 


IIO2 


22 -426 


I -0597 


12-233 


I'OIOO 


2-039 


1599 


32-2I3 


1082 


22-OI9 


1-0577 


II-825 


I'OOSO 


I '631 


1-1578 


31-805 


1061 


2I-6II 


1-0557 


11-418 


I -OO6O 


IT24 


I-I557 


3I-398 


1041 


21-203 


1-0537 


II'OIO 


I-0040 


0-816 


IT536 


30-990 


IO2O 


20-796 


1-0517 


10-602 


I -0020 


0-408 



373 



Appendix. 
TABLE VII. 



Showing the Percentage Amount of Nitric Acid (HNO 3 ) 
contained in Aqueous Solutions of various Specific Gravities. 

(Kolb, Ann. Ch. Phys. [4] 136). 

The numbers marked * are the results of direct observations ; the 
others are obtained by interpolation. 



HNO 3 

per cent. 


Specific gravity 


Contrac- 
tion 


HNO 3 

per cent. 


Specific gravity 


Contrac- 
tion 


Ato 


At 15 


Ato 


At 15 


lOO'OO 


J'559 


I-530 


O'OOOO 


68-00 


1-435 


1-414 


0-0784 


99-84* 


1-559* 


1-530* 


O-OOO4 


67-00 


I-430 


I -410 


0-0796 


9972* 


1-558* 


1-530* 


O'OOIO 


66-00 


I-425 


I-405 


0-0806 


99-52* 


1*557* 


1-529* 


O-OOI4 


65-07* 


I -420* 


1-400* 


0-0818 


97-89* 


i-55i* 


1-523* 


0-0065 


64-00 


I'4I5 


1-395 


0-0830 


97-OO 


I-548 


1-520 


0-0090 


63-59 


I-4I3 


1-393 


0-0833 


96-00 


1-544 


1-516 


0-0120 


62-00 


1-404 


1-386 


0-0846 


95-27* 


1-542* 


1-514* 


0-OI42 


61-21* 


I -400* 


1-381* 


0-0850 


94-00 


1-537 


1-509 


0-OI82 


60-00 


1-393 


1-374 i 0-0854 


93-01* 


i-533* 


I -506* 


0-0208 


59-59* 


1-391* 


1-372* 


0-0855 


92-00 


1-529 


I-503 


0-0242 


58-88 


1-387 


1-368 


0-086 1 


91-00 


1-526 


1-499 


O-O272 


58-00 


1-382 


1-363 


0-0864 


90-00 


1-522 


1-495 


O-O3OI 


57-oo 


1-376 


I-358 


0-0868 


89-56* 


1-521* 


I -494* 


0-0315 


56-10* 


i-37i* 


1-353* 


0-0870 


88-00 


i-5i4 


1-488 


0-0354 


55-oo 


I-365 


1-346 


0-0874 


87-45* 


1-513* 


I -486* 


0-0369 


54-oo 


1-359 


i-34i 


0-0875 


86-17* 


1-507* 


1-482 


0-0404 


53-8i 


I-358 


1-339 


0-0875 


85-00 


1-503 


1-478 


0-0433 


53-oo 


1-353 


1-335 


0-0875 


84-00 


1-499 


1-474 


0-0459 


52-33* 


i -349* 


i-33i* 


0-0875 


83-00 


1-495 


1-470 


0-0485 


50-99* 


1-341* 


1-323* i 0-0872 


82-00 


1-492 


1-467 


0-0508 


49-97 


1-334 


1-317 j 0-0867 


80-96* 


1-488* 


I-463* 


0-053I 


49-00 


1-328 


1-312 


0-0862 


80-00 


1-484 


1-460 


0-0556 


48-00 


1-321 


1-304 


0-0856 


79-00 


1-481 


I-456 


0*0580 


47-18* 


i-3i5* 


1-298* 


0-0850 


77'66 


1-476 


I-45 1 


0-0610 


46-64 


1-312 


1-295 


0-0848 


76-00 


1-469 


1-445 


0-0643 


45-00 


1-300 


1-284 


0-0835 


75-00 


1-465 


1-442 


0-0666 


43-53* 


1-291* 


i -274* 


0-0820 


74-01* 


i -462* 


1-438* 


0-0688 


42-00 


1-280 


1-264 


0-0808 


73-00 


1-457 


1-435 


0-0708 


41-00 


1-274 


1-257 


0-0796 


72-39* 


1-455* 


1-432* 


0-0722 


40-00 


1-267 


1-251 


0-0786 


71-24* 


i -450* 


i -429* 


0-0740 


39-00 


1-260 


1-244 


0-0755 


69-96 


1-444 


1-423 


0-0760 


37-95* 


1-253* 


1-237* 


0-0762 


69-20* 


1-441 


1-419* 


0-0771 


36-00 


1-240 


1-225 


0-0740 



Appendix. 
TABLE VII. continued. 



379 





Specific gravity 






Specific gravity 




HNO 3 




Contrac- 


HNO 3 




Contrac- 








tion 








tion 


per cent. 


Ato 


At 15 




per cent. 


Ato 


At 15 




35' 


234 


I-2I8 


0-0729 


2O-OO 


I-I32 


I -120 


0*0483 


33-86* 


226* 


I-2II* 


0-0718 


17-47* 


1-115* 


I-I05* 


0-0422 


32-00 


214 


198 


0-0692 


15-00 


1-099 


1-089 


0-0336 


31-00 


207 


192 


0-0678 


I3-00 


1-085 


1-077 


0-0316 


30-00 


200 


I8 5 


0-0664 


II-4I* 


1-075 


I -067* 


0*0296 


29-00 


194 


179 


0-0650 


7-22* 


1-050 


I -045* 


0-0206 


28-00* 


I8 7 * 


I 7 2* 


0-0635 


4-OO 


1-026 


1-022 


0-OII2 


27-00 


180 


166 


0-0616 


2-00 


1-013 


I -010 


0-0055 


2571* 


1-171* 


157* 


0-0593 


o-oo 


I -000 


0-999 


o-oooo 


23-00 


I-I53 


138 


0-0520 











TABLE VIII. 

Showing the Percentage Amount of Caustic Potash 

Aqueous Solutions of various Specific Gravities. 

Tiinnermann, N. Tr. xviii., 2, 5. Temp. 15. 



n 



Sp. gr. 


Per cent. 


Sp. gr. 


Per cent. 


3300 


28-290 


1437 


I4-I45 


3131 


27-158 


1308 


13-013 


2966 


26-027 


Il82 


1 1 -882 


2805 


24-895 


1059 


IO-750 


2648 


23764 


93 8 


9-619 


2493 


22-632 


0819 


8-487 


2342 


21-500 


0703 


7-355 


2268 


20-935 


I -0589 


6-224 


2122 


19-803 


I -0478 


5-002 


1979 


18-671 


I -0369 


3-961 


l8 39 


I7-540 


I -0260 


2-829 


1702 


16-408 


I-OI53 


1-697 


1568 


15-277 


I -0050 


0-5658 



3^0 



Appendix. 



TABLE IX. 

Showing Percentage Amount of Soda (Na 2 O) in Aqueous 

Solutions of various Specific Gravities. 

Tiinnermann. 



Sp. gr. 


Per cent. 


Sp.gr. 


Per cent. 


Sp.gr. 


Per cent. 


Sp. gr. 


Per cent. 


4285 


30-220 


3198 


22-363 


1-2392 


I5'IIO 


1-1042 


7-253 


4193 


29-616 


3H3 


21-894 


2280 


14-500 


I -0948 


6-648 


4101 


29-011 


3125 


21-758 


2178 


13-901 


I-0855 


6-044 


4011 


28-407 


3053 


21-154 


-2058 


13-297 


I -0764 


5-440 


3923 


27-802 


2982 


20-550 


1948 


12-692 


1-0675 


4-835 


3836 


27-200 


2912 


I9-945 


1841 


12-088 


I -0587 


4-231 


3751 


26-594 


2843 


I9-34I 


1734 


1 1 -484 


I -0500 


3-626 


3668 


25-989 


2775 


18-730 


1630 


10-879 


1-0414 


3-022 


35 86 


25385 


2708 


18-132 


1528 


10-275 


I -0330 


2-418 


3505 


24-780 


2642 


I7-528 


1428 


9-670 


I -0246 


I-8I3 


3426 


24-176 


2578 


16-923 


1330 


9-066 


1-0163 


1-209 


'3349 


23-572 


2515 


16-379 


1233 


8-462 


I -008 1 


0-604 


3273 


22-967 


2453 


I5-7I4 


1137 


7-857 


1-0040 


0-302 



TABLE X. 

Showing the Percentage Amount of Ammonia in Aqueous 

Solutions of the Gas of various Specific Gravities. 

Carius. Temp. 14. 



Sp. gravity 


NH 3 

per cent. 


Sp. gravity 


NH 3 

percent. 


Sp. gravity 


NH S 
percent. 


0-8844 


36 


0-9133 


24 


0-9520 


12 


0-8864 


35 


0-9162 


23 


0-955 6 


II 


0-8885 


34 


0-9191 


22 


0-9593 


10 


0-8907 


33 


0-9221 


21 


0-9631 


9 


0-8929 


3 2 


0-9251 


2O 


0-9670 


8 


0-8953 


3i 


0-9283 


'9 


0-9709 


7 


0-8976 


30 


0-9314 


18 


0-9749 


6 


0-9001 


29 


0-9347 


17 


0-9790 


5 


0-9026 


28 


0-9380 


16 


0-9031 


4 


0-9052 


27 


0-9414 


15 


0-9873 


3 


0-9078 


26 


0-9449 


'4 


0-9915 


2 


0-9106 


25 


0-9484 


13 


0'9959 


I 



Appendix. 



TABLE XI. 

Reduction of Weighings in Air to a Vacuum (G. F. Becker , 
Liebig's Annalen, 1 9 5 , p. 2 2 2 ). 



Brass weights 
for substances whose sp. gr. 
is between 


Correction per 
gram, error less 
than 3*0 mgrm. 


Platinum weights 
for substances whose sp. gr. 
is between 


27738 and 11-064 


O-OOOO67 


51-766 and 13-568 


1 1 -064 


6-904 


O'OOOOOO 


13-568 


7-807 


6-904 


5-019 


+ 0-000067 


7-807 


5-480 


5-019 


3-943 


0-000133 


5-480 


4*222 


3-943 


3-247 


0-000200 


4-222 


3'433 


3-247 


2759 


O-OOO267 


3-433 


2-893 


2759 


2-399 


O-OOO333 


2-893 


2-500 


2-399 


2*122 


0-000400 


2-500 


2-201 


2'122 


I-903 


0-000467 


2-201 




965 


903 


1724 


0-000533 




965 




776 


724 


1-576 O-OOO6OO 




776 




619 


576 


I-452 


O-OOO667 




619 




488 


452 


1-377 


0-000733 




488 




377 


'377 


1-254 


0-000800 




'377 




281 


254 


I-J74 . 


0-000867 




28l 




197 


174 


1-103 


O-OOO933 




I 9 7 




124 


103 


1-041 


o-ooiooo 




124 




059 


1-041 


0-985 


0-001067 




059 


1-002 




0-001133 




002 


0-950 




O-OOI2OO 





PREPARATION OF PURE PLATINUM TETRA- 
CHLORIDE. 

Scrap platinum, which may contain iridium, osmium, &c., is dissolved 
in aqua regia, and the solution is evaporated to dryness ; the residue is 
dissolved in moderately concentrated hydrochloric acid, and again 
evaporated to dryness. The dried chloride is once more dissolved in 
hot water containing free hydrochloric acid, and the solution mixed 
with a large excess of soda-ley. It is again boiled for some time, and a 
few drops of alcohol are added in order to destroy any sodium hypo- 
chlorite which may be formed ; the precipitate is redissolved in hydro- 
chloric acid ; the liquid is filtered, if necessary, and mixed with a hot 
ana saturated solution of ammonium chloride so long as a precipitate 
forms. This process of separating platinum from its congeners, with 
which in the commercial variety of the metal it is almost invariably 



382 Appendix. 

mixed, is based upon the different behaviour of sodium hydrate solution 
towards the higher chlorides of the associated metals. Platinic chloride 
is very slightly, if at all, reduced to platinous chloride on boiling with 
soda solution, whereas the other chlorides are all reduced, with the pro- 
duction of sodium chloride and hypochlorite, and these reduced chlorides 
are not precipitated in union with ammonium chloride. 

The ammonium-platinum chloride, which is of a bright yellow colour 
and free from orange or red crystals, is washed by decantation, dried, 
and gently heated in a platinum crucible, or it may be placed in a piece 
of hard glass tubing and decomposed in a current of coal gas or dry 
hydrogen. The reduced metal should be weighed, dissolved in aqua 
regia, the solution evaporated to dryness with excess of hydrochloric acid 
to expel the last traces of nitric acid, and the residue dissolved at a 
gentle heat in a definite volume of dilute hydrochloric acid. The operator 
in this manner obtains an idea of the strength of the solution. 

Pure platinum chloride may be recovered from the precipitates with 
the alkaline chlorides which are obtained in analytical work by boiling 
them with a solution of sodium carbonate and alcohol : washing the 
precipitated platinum with hot water by decantation and finally with 
hot hydrochloric acid. The spongy metal is then dissolved in aqua 
regia (5HC1 : iHNO 3 ), the solution filtered, and evaporated to dryness 
with hydrochloric acid as above. 



TREATMENT OF SILVER RESIDUES.' 

The mixed silver salts, associated with metallic silver, which accumu- 
late in the course of analytical work, may be conveniently reduced, 
after washing and drying, by heating to fusion with a mixture of sodium 
and potassium carbonates in an earthen or unglazed porcelain crucible. 
The button of metallic silver is washed with boiling water, and dis- 
solved in nitric acid, arid the solution of the nitrate evaporated to 
dryness. 



INDEX OF SEPARATIONS. 



ALUMINA from calcium, iron, magnesia, 
potash, and soda, 101 

from iron in presence of phosphoric 
acid, 177, 220, 246 

from iron, 101, 102, 184, 246, 248 

from copper and bismuth, 213 

from lead and tin, 108 

from tin, 107, 108, 109 

from bismuth, copper, and lead, 213 
BARIUM from calcium, 92 

from strontium, 178 

BISMUTH from antimony, arsenic, copper, 
and lead, 109, 213, 272 

from silver and cadmium, 261 
CALCIUM from magnesium, 88, 185 

from magnesium and alkalies, 99, 180, 
185 

from phosphoric acid, 347 

COBALT from arsenic, antimony, bismuth, 
iron, zinc, &c. 214 

from nickel, 214, 275 

from zinc, 214, 244 

COPPER from antimony, arsenic, and tin, 

212 

from bismuth and lead, 109, 212, 272 

from lead, tin, zinc, and iron, 103 

from zinc and nickel, 106 
GOLD from silver and copper, 286 
IRON from alumina, 101, 102, 184, 246 

in presence of phosphoric acid, 

177, 217, 247 

from manganese, 177, 214, 247 

from nickel and cobalt, 214, 247 

from chromium, 247, 268 

from zinc, 105, 214, 244 

LEAD from arsenic and antimony, 108 



LEAD from bismuth, copper, tin, iron, and 
zinc, 103, 108, 213 

from copper, antimony, iron, zinc, &c. 
25 8 

from silver, 256 

MAGNESIUM from alkalies, 102, 180, 185 

from calcium, 88, 185 

from calcium in presence of phosphoric 
acid, 347 

MANGANESE from alumina, 178, 220, 247 

from iron, 177, 214, 247 

from nickel and cobalt, 214, 247 

from zinc, 214, 247 

MERCURY from silver, copper, arsenic, 

antimony, iron, and zinc, 278 
NICKEL from antimony, arsenic, copper, 

lead, bismuth, &c. 214 

from cobalt, 215, 275 

from iron and manganese, 214, 247 

from zinc, 106, 214, 247 
POTASSIUM from magnesium, 99, 174, 185 

from sodium, 85, 128 

SILVER from copper and gold, 286 

from lead, 256 

from bismuth and cadmium, 261 
SODIUM from magnesium, 99, 174, 185 

from potassium, 85, 128 

from bismuth and lead, 109 

from copper, lead, iron, and zinc, 103 

from tungsten, 253 

TUNGSTEN from iron and manganese, 254 

from tin, 253 

ZINC from cobalt, 215, 244 

from iron, 105, 214, 246 

from manganese, 214, 247 

from nickel, 107, 214, 249 



GENERAL INDEX. 



ACETIC acid, determination of^trength 
of, 142 

Acids in combination, volumetric deter- 
mination, 146 
Acidimetry, 142 
Air bath, 39 

Air in water, analysis of, 327 
Albite. analysis of, 102 
Alkalies, determination of, in glass, too 

limestone, 178 

manures, 332 

plant ash, 340 
Alkalimetry, 130 

Alkaline liquids, action on glass, 47 
Alumina, determination of, 98 
Ammonia, determination of, in gas liquor, 
141 

guano, 141 

by volumetric analysis, 140 

by gravimetric analysis, 94 

in water, 293 

manures, 340 

(albuminoid) in water, 297 
Ammonium chloride solution, action on 

glass, 47 

Analysis (indirect), principles of, 92 
Antimony, estimation of, 107, 108 
volumetric determination of, in tartar- 
emetic, 161 

Arsenic, determination of, in cobalt glance, 
274 

fahl ores, 278 

iron ores, 219 

Arsenious acid, estimation of, by iodine, 

165 

solution, preparation of, 194 
Ashes of plants, analysis of, 340 
Atomic weights and symbols of elements, 

369 



BALANCE, description of, 3 
adjustment of, 7 

conditions of stability, &c. 8 

tests of efficacy, 13 

preservation of, 15 



Balance-room, 15 

Barium and calcium, separation of, 92 

determination of, in limestone, 177 

and strontium sulphates, separation of, 
177 

sulphate, analysis of, 91 
Baume's hydrometer tables, 374 
Bell metal, analysis of, 103 
Bismuth, estimation of, 109, 213 
Black ash, analysis of, 198 

composition of, 199 

Bleaching powder, composition of, 193 

valuation of, 194 
Bone dust, analysis of, 337 
Boulangerite, analysis of. 279 
Bournonite, analysis of, 279 
Brass, analysis of, 103 
Braunite, analysis of, 185 
Britannia metal, analysis of, 107 
Bromates, analysis of, 165 
Bromine, determination of, in organic com- 
pounds, 364 
Bronze, analysis of, 103 
Bullion (silver), assay of, 281 
Burette, graduation of, 119 

Gay Lussac's, 119 

Mohr's, 121 

reading of, 122 



CADMIUM, determination of, as sul- 
phide, 240 

Calcium chlorate, determination of, .in 
bleaching powder, 197 

volumetric determination of, by potas- 
sium permanganate, 152 

Carbon, determination of, in iron and steel, 
229 

Eggertz' method, 235 

Ullgren's method, 233 

Weyl's method, 231 

Wohler's method, 229 

dioxide, gravimetric determination of, 
86 

by volumetric analysis, 143 

in aerated waters, 145 



C C 



3 86 



General Index. 



Carbon dioxide, gravimetric determination 

of, in natural waters, 144 
Carius' method for determination of sulphur 

and phosphorus, 367 
Cement stone, analysis of, 182 
Chlorates, analysis of, 165 

in bleaching powder, 197 
Chlorides, determination of, in limestone, 

179 

Chlorimetrical degrees, 198 
Chlorine, gravimetric determination of, 81 

volumetric determination of, 124 

by iodine solution, 158 

determination of, in organic compounds, 
364 

bleaching powder, 194 
Chloric acid, determination of, 129 
Chrome iron stone, occurrence of, 268 

valuation of, 269 

Chromic acid, preparation of, 234 

Cinnabar, analysis of, 288 

Clay, mechanical and chemical analysis of, 

Coal, valuation of, 289 

specific gravity of, 291 

Cobalt, determination of, as metal, 272 
-as tricobalt tetroxide, 272 

glance, analysis of, 273 

Cochineal, use of, in volumetric analysis, 

142 

Coins, silver, assay of, 281 
Copper, gravimetric determination of, 74 
as metal, 105, 212 

detection of, in waters, 325 

ores, assay of, by Luckow's process, 207 

Steinbeck's process, 207 

oxide, for organic analysis, preparation 
of, 353 

pyrites, analysis of, 211 

separation from gold and silver, 286 

sulphate, analysis of, 71 
purification of, 70 

Crucibles, to remove silver chloride from, 82 

pVECANTATION, 49 
JL-J Desiccation, 39 
Desiccator, 37, 38 
Dolomite, analysis of, 85 



ELUTRIATION, 36 
Erdmann's float, 123 
Erlenmeyer's furnace, description of, 349 
Evaporation, 42 
of liquids containing gas, 42 
of high boiling point, 43 

FAHL ore, analysis of, 276 
Fat, determination of, in bone dust, 
337 

Felspar, analysis of, 102 
Ferrous, ammonium sulphate, preparation 

and analysis of, 94 
Filter ash, determination of, 51 



Filters, capacities of, 59 

incineration of, 51 

paper, analysis of ash of, 50 

preparation of, 51 

pump, 6 1 

weighed, use of, 68 

Filtration, precautions to be observed in, 53 
Fire clay, composition and analysis of, 182 
Fluorine, estimation of, 179 
Funnels, choice of, 53 
Fusible metal, analysis of, 109 



GALENA, analysis of, 255 
determination of silver in, 256 

assay of, 257 

Gases, from natural water, analysis of, 327 

obtained in water analysis, examination 
of, 304 

Gay Lussac's burette, 119 

method of silver assay, 281 
Geissler's potash bulbs, 352 
Gelatigenous matter, determination of, in 

bone dust, 337 

German silver, analysis of, 106 
Glass, action of solutions upon, 47 

analysis of, 99 
Gold assay, 286 

Gold, silver, and copper, separation of, 286 
Graphite, derermination of, in iron, 234 
Gravimetric analysis, definition of, i 
Guano, analysis of, 331 
Gun metal, analysis of, 103 
Gunpowder, analysis of, 171 



HARDNESS of water, estimation of, 
321 . 

Hydrochloric acid, estimation of, by gravi- 
metric analysis, 81 

estimation of, by volumetric analysis, 
126 

determination of amount of, required 
to decompose manganese ore, 191 

normal, preparation of, for volumetric 

analysis, 130 

strength of aqueous solutions of, 377 
Hydrocyanic acid, estimation of, by iodine 

solution, 160 



TNCINERATION of filters, 51 
L of plants for analysis, 341 
Tlmenite, analysis of, 225 
lodates, analysis of, 165 
Iodine, pure, preparation of, 156 

determination of, in organic compound*, 
364 

by sodium amalgam, 366 

by alcoholic silver nitrate, 366 

and sodium thiosulphate, reaction be- 
tween, 155 

solution, preparation of, 156 

use of, in volumetric analysis, 155 
Iron, action of acids upon, 228 



General Index. 



387 



Iron, determination of, by potassium bi- 
chromate, 221 

permanganate, 148 

iodine and sodium thiosul- 
phate, 1 66 

cast, composition of, 227 
varieties of, 227 

ores, analysis of, 215 
varieties of, 216 

pyrites, analysis of, 215 

Swedish, analysis of, 227 

determination of slag in, 249 

wrought, analysis of, 227 

nitrogen in, determination of, 242 

by Ullgren's method, 242 

slags, analysis of, 249 

estimation of, in natural water, 326 



KAOLIN, composition of, 180 
Kieffer's solution, 192 
Kupfernickelstein, analysis of, 215 



LEAD, estimation of, in galena, 255 
refined, analysis of, 258 

in water, 325 

volumetric determination of, 153 
Liebig's potash bulbs, 357 

Lime, estimation of, in dolomite, 88 
felspar, 101 

glass, 99 

manures, 332 

by volumetric analysis, 152 

in water, 326 
Limestone, analysis of, 174 
Liquids, organic analysis of, 348 
Litmus solutions, preparation of, 136 
Litre flasks, graduation of, 115 



MAGNESIA, estimation of, in dolo- 
mite, 85 

felspar, joi 

glass, 99 

manures, 332 

water, 326 

separation of, from lime, 88 
Manganese, determination of, in limestones, 

88, 178 

iron, 247 
ores, 220 

ores, valuation of, by Fresenius' and 
Will's method, 186 

iodine solution, 185 

iron and potassium perman- 
ganate solution, 189 

oxalic acid and potassium per? 
manganate solution, 189 

determination of moisture in, 190 
Manures, analysis of, 331 

Mechanical analysis of clay soils, &c. 182 

division, 35 

Mercury, estimation of, 287 
Mercuric sulphate, preparation of, 242 



Mortars, choice of, 35 
steel and agate, 35 



"VTESSLER'S solution, preparation, 293 
_L\ Nickel, determination of, 106 

separation from copper and zinc, 106 

cobalt, 274 

Nickelspeiss, analysis of, 279 
Niobic acid, detection of, 254 
Nitre, analysis of, 168 

estimation of, in gunpowder, 170 
Nitric acid, action on glass, 47 

estimation of, as ammonia, 96 

by iodine solution, 167 

in nitre, 170 

in water, 315 ^ 

specific gravity ami strength of 

aqueous solution, 378 
Nitrogen, in iron, determination of, 242 

in manures, 334 

determination of, by soda-lime process, 
334. 

in organic analysis, 358 

by SchifF's apparatus, 362 

by Simpson's process, 359 

in water residues, 299 



OLEFINES, evolution of, during solu- 
tion of iron, 228 

Organic analysis, determination of carbon 
and hydrogen, 348 

chlorine, bromine, and iodine, 364 
nitrogen, 358 

phosphorus, 366 

sulphur, 366 

Organic matter, determination of, in lime- 
stone, 176 

in water, by Frankland's process, 
298 

Orthoclase, analysis of, 101 
Oxalic acid, determination of, by perman- 
ganic acid, 151 



PEARL-ASH, analysis of, 82, 139 
'Phosphorus, in iron ores, determina- 
tion of, 217 

iron, determination of, 249 

determination of, in organic compounds 
366 

Phosphoric acid, separation from iron and 
alumina, 346 

in plant ash, determination of, 340 

water, determination of, 327 

manures, determination of, 328 

by uranium solution, 328 
Phosphorous acid, preparation of, 289 
Pipettes, graduation of, 117 

Pisani's method for determination of silver, 
280 

Platinum crucibles, precautions to be ob- 
served in using, 67 

-r- vessels, how to clean, 67, 68 

recovery of, from residues, 345 



CC 2 



388 



General Index. 



Platinum tetrachloride, preparation of, 381 
Potassium bichromate, valuation of, 163 

use of, in volumetric analysis, 112 

and sodium thiosulphaie, reaction 

between, 157 

carbonate, valuation of, 83, 139 

estimation of, in presence of sodium 

carbonate, 139 

chromate solution, preparation of, for 
volumetrical analysis, 126 

cyanide, valuation of, 160 

preparation of standard solution of, 
206 

ferricyanide, estimation of, 154 

ferrocyanide, estimation of, 154 

gravimetric determination of, 83 

hydroxide, determination of, in presence 
of carbonic acid, 138 

iodide paper, preparation of, 195 

permanganate solution, determination of 
strength of, 148 

use of, in volumetric analysis, in, 
148 

and sodium, indirect determination of, 
82, 136 

Potash alum, analysis of, 98 

pumice, preparation of, 234 

solution, strength of, 379 
Precipitates, drying, 65 

igniting, 64, 65 
Precipitating flasks, 50 
Pyrites, copper, analysis of, 212 

iron, analysis of, 215 

in limestone, determination of, 176 
Pyroligneous acid, determination of, 142 
Pyrolusite, analysis of, 185 



"P EICHARDTS method of expelling 

J\. gases from water, 328 

Rider, use of, in weighing, 23 

Rochelle salt, analysis of, 85 

Rosolic acid, use of, in volumetric analysis, 

MS 
Rutile, analysis of, 226 



SAMPLING, method of, 34 
Scheelite, analysis of, 255 
SchifTs apparatus for measuring nitrogen 

in organic analysis, 36^ 
Sifting, method of, 36 
Silica, estimation of, in clay, 181 

glass, 99 

felspar, 102 

water, 326 

plant ash, 346 

in limestone, 175 

* iron ores, 217 

*> in copper ores, 211 

*- iron, 245 

' galena, 256 

smaltine and cobalt glance, 272 

Silver, assay, 281 

determination of, by Pisani's method, 
280 



Silver, separation from gold and copper, 286 

ore, red, analysis of, 279 

preparation of pure, 124 

in galena, determination of, 956 

in solutions, determination of, 279 

residues, treatment of, 382 

solutions, standard, preparation of, 125 
Slag, determination of, in iron, 249 
Smaltine, analysis of, 272 

Soap solution, preparation of, for water 

analysis, 321 

Soda ash, valuation of, 137 
analysis of, 199 

solution, preparation of, for volumetric 
analysis, 135 

strength of aqueous solutions of, 380 
Sodium carbonate, estimation of, in pre- 
sence of potassium carbonate, 1 39 

and potassium, indirect estimation of, 
by volumetric analysis, 128 

gravimetric determination of, 82 

chloride, purification of, 79 
analysis of, 79 

separation from potassium, 85 

1 hydroxide, determination of, in presence 
of carbonate, 38 

salts, estimation of, in water, 327 

thiosulphate, preparation of solution of, 
158 

sulphide, determination of, by Lestelle's 
method, 203 

Soluble matter, determination of, in water, 
315 

Sphene, analysis of, 226 

Spiegeleisen, composition of, 249 

Sprengel pump, use of, in water analysis, 
301 

Starch solution, preparation of, for volu- 
metric analysis, 157 

Steam bath, 38 

Steel, analysis of, 228 

composition of, 229 

action of.acids upon, 229 
Sulphates, determination of, in water, 326 

I Sulphur, determination of, in iron, 238 

volumetrically, 240 

coal, 290 

copper pyrites, 211 

iron ores, 217 

galena, 255 

roasted pyrites, 205 

organic compounds, 366 

by Carius' method, 367 

refined lead, 265 

Sulphuretted hydrogen, estimation of, by 

iodine solution, 159 

in mineral waters, 160 

Sulphur dioxide, estimation of, by iodine 

solution, 159 
Sulphuric acid, action on glass, 48 

gravimetric estimation of, 77 

volumetric estimation of, 135 

separation of, in presence of alkalies, 

169 

determination of, in limestone, 178 

in manures, 332 



General Index. 



Sulphuric acid solution, normal, prepara- 
tion of, 133 

strength of aqueous solution, 376 

Superphosphates, analysis of, 337 



n^ABLE furnace, 100 

J. Tap cinder, analysis of, 250 
Tartar emetic, amount of antimony in, 161 
Tetrahedrite, analysis of, 276 

composition of, 276 
Tin-crystals, valuation of, 162 

determination of, by iodine solution, 162 

ore, assay of, 252 

separation of, 108, 109 

and tungsten, separation of, 253 
Titaniferous iron ores, analysis of, 225 
Titanium, determination of, in clays, 184 

iron ores, 218, 226 
Titanite, analysis of, 226 
Triangle, 68 

Tungsten and tin, separation of, 253 
Twaddell's hydrometer, 375 
Type metal, analysis of, 108 



T TRANIUM solution, preparation of, for 
v_J volumetric analysis, 332 



VAT waste, analysis of, 200 
Vinegar, valuation of, by Kieffer's 
solution, 192 

Volumetric analysis, definition of, 2 
principles of, no 

WASH bottle, 54 
Water, action of, on glass, 47 
bath, 44 



Wash bath, Bunsen's, 46 

collection of, for analysis, 291 

determination of albuminoid ammonia 
in, 297 

ammonia in, 293 

chlorine in, 320 

ni rates and nitrites, 315 

silica in, 326 

iron, 326 

lime, 326 

magnesia, 326 

sulphuric acid, 326 

phosphoric acid, 327 

lead and copper, 325 

sodium and potassium, 327 

total soluble matter, 315 

hardness, 321 

suspended matter, 293 

in combination, determination of, 71 

estimation of sulphuretted hydrogen in, 
159 

in minerals, estimation of, 183 

in manures, estimation of, 331 
Weighing, correction for displacement of 

air in, 32, 381 

by vibrations, 27 

operation of, 20, 21 

by substitution, 21 

precautions in, 25 
Weights, description, 16 

method of testing, 18 

tarnished, to clean, 18 

White lead, composition and analysis of, 
267 

adulterations of, 267 

Wolfram, occurrence and analysis of, 254 



'INC ores, assay of, 250 



PRINTED BY 

SPOTTISWOODE AND CO., NEW-STREET SQUARE 
LONDON 



UNIVERSITY OF CALIFORNIA LIBRARY 
BERKELEY 

Return to desk from which borrowed. 
This book is DUE on the last date stamped below. 



4Dec'50Bt 

1953L 






LD 21-100m-9,'47(A5702sl6)476 




THE UNIVERSITY OF CALIFORNIA LIBRARY