5 
case, sulphide of sodium paper is used as a secondary guide. 
The experimental results gave the mean error, in colorless 
liquids, as 0.0026 grs. of oxide of lead ; while in dark liquids, 
the average error equalled 0.0042 grs. of oxide of lead. 
The second section treated of baryta. This is effected in 
like manner, but by the aid of neutral chromate of potash. 
The mean error was 0.0024 grs. of baryta. 
The third section discussed the estimation of combined 
sulphuric acid, by the double use of chloride of barium or 
other salt of baryta and chromate of potash. An excess of 
the former is added, and the surplus ascertained by fractional 
addition of the latter salt. Mean error 0.0022 grs. sulphuric 
acid. A salt of lead may be similarly used. 
The fourth section was devoted to the consideration of 
sulphuretted hydrogen. The estimation of this compound in 
ores, waters, &c. the author finds may be made with great 
nicety by use of solutions of lead and bichromate of potash. 
A similar plan may be adopted with regard to carbonic acid, 
and this was detailed under the fifth heading. Collective 
allusion was also made to a numerous series of compounds, 
which give rise to insoluble bases with oxides of barium 
and lead. Such bodies may often be estimated advantage- 
ously by application of the general method given. 
By modifying the above processes, metrical separations of 
sulphuric acid and sulphurous acid, of sulphuric acid and 
hydrochloric acid, of sulphuric acid and sulphurous acid from 
carbonic acid, and the latter from sulphuretted hydrogen, 
can be accomplished. 
The concluding section embraced the use of lead or baryta 
solutions for the determination of chromic acid. The process 
is applicable to the oxide of chromium, which must be 
converted into chromic acid. 
By so doing, oxide of chromium may be metrically 
separated from alumina and other oxides, and also from 
chromic acid. 
