LEEDS CHEMISTS’ ASSOCIATION. 
421 
by adding the reagent till no further precipitation occurs; secondly, by using some other 
substance which shall show when excess of reagent has been added, and thirdly, by adding 
reagent until a precipitate is commenced to be produced or the liquid becomes turbid. 
The mutual estimations of Cl by means of Ag, and of Ag by means of Cl, are the only 
ones that can be accurately made in the first way, as in few or none of the other cases 
where insoluble precipitates result on the addition of any reagent, do these precipitates 
settle down so readily and completely as to leave the liquid clear enough to see the exact 
point when precipitation ceases on addition of more reagent. The estimation of Ag by 
Cl w r as one of the earliest, as it is now one of the most important, of the applications of 
volumetric analysis. It is a point of great value to be able readily to ascertain the 
amount of real silver contained in the various alloys, solutions, and ores of this important 
and costly metal, and the old process by cupellation, as it is called, occupies considerable 
time. The volumetric assay of silver solutions is effected by means of a standard solu¬ 
tion of NaCl, which is added to the silver with constant and strong agitation to cause 
the precipitated AgCl to clot together and leave the solution clear, so that the point 
may be seen when a drop of the salt solution ceases to produce any further turbidity. 
It is well in this analysis to work with dilute solutions containing only one-tenth of 
the ordinary standard quantity, i.e. an equivalent in grains in 1000 grain-measures. 
The analysis of chlorides by means of standard silver solutions is of course the 
same in principle as the above, only reversed, the silver being added to the chloride. 
It has, however, a considerable advantage over its companion process in that we can 
make use of another substance to show the end of the reaction, thus bringing it 
under the second division of analysis by precipitation. This is KOCrG 3 which forms 
with silver a blood-red chromate of silver, which, however, is not produced as long as 
any chloride is present in the solution, chlorine having a greater affinity for Ag than 
Cr0 3 has. If a little chromate of potash be added to a chloride, and then the silver, 
the precipitate remains white until the chloride is all decomposed, when the appearance 
of the red colour at once indicates the end of the operation. So great a convenience is 
this, that in estimating silver, it is often well to add at once a known excess of salt and 
then estimate the amount of this excess by standard silver and chromate of potash. 
This facility for backward or residual, as well as for indirect analysis, is one of the great 
recommendations of the volumetric method. Thus if in any analysis the mark be over¬ 
stepped by accident or carelessness, and too much of the reagent added, the operation is 
not spoiled, as the amount of this excess can generally be as readily ascertained as that of 
the original solution. The estimation of sulphuric acid, when in combination as sul¬ 
phates, by precipitation by chloride of barium is another instance of the use of an indi¬ 
cator contained in the solution. If the precipitation alone were relied on, the difficulty 
caused by the non-subsidence of the precipitate would be fatal; but if a known excess 
of BaCl be at once added to the hot solution, this excess may be tolerably readily 
ascertained by means of standard chromate of potash solution. This precipitates the 
baryta as a yellow powder, but the delicacy of the indication is owing to the great 
colouring power this chromate of potash has for water, a very small quantity indeed 
communicating a decided yellow colour to the solution. As soon, therefore, as the BaO 
is all precipitated the liquid becomes yellow, and a very slight subsidence of the pre¬ 
cipitate is sufficient for this to be observed. 
The estimation of phosphoric acid or phosphates by nitrate of uranium is another 
instance of the use of an external indicator of the end of the reaction, but in this case it 
is not contained in the Solution itself, but is brought in contact with it from time to 
time. 
All soluble phosphates are precipitated by nitrate of uranium in the presence of NH 3 
(under certain conditions which I need not here specify). The precipitate produced is 
one that settles very slowly, so that the end could not be known by the reagent pro¬ 
ducing no further precipitation. When, however, nitrate of uranium and ferrocyanide 
of potassium are brought together, an intense brown colour results. The uranium is 
therefore added to the phosphate till a small drop of the solution mixed with a drop of 
ferrocyanide on a white plate shows the commencement of this brown colour, and the 
uranium thus is shown to be in excess. 
The third kind of analysis by precipitation, namely, that in which the end of the 
reaction is shown by the commencement of formation of a precipitate in a previously 
clear liquid, includes only two substances (if ^ve except the use of ammonio-sulphate of 
