on the Analysis of the Carbonates. 357' 



utes, and ought never to be omitted, when great accuracy is in 

 view. As proving its importance, we may state that in the great 

 number of experiments, which we have made within the last few 

 months, by the method above described, we have found the 

 amount of absorbed carbonic acid, to be rarely less than one twen- 

 tieth, and sometimes as much as one fifteenth of a cubic inch ; 

 varying thus from one twentieth of a per cent, to one fifth of a per 

 cent, of the whole weight of thai substance contained in the carbo- 

 nate employed. 



That the carbonic acid thus united with the liquid, cannot be 

 expelled by Rose's method, is apparent from the fact that its re- 

 moval can only be effected by an actual boiling of the liquid, and 

 this if attempted in the flask, would lead to far more serious er- 

 rors, than that proposed to be corrected. In proof of the latter 

 statement, we would refer to the following experiments. 



1st. Having prepared a solution with carbonate of lime, and 

 the usual charge of dilute hydrochloric acid, and boiled it to ex- 

 pel the dissolved carbonic acid, we introduced it into a small bot- 

 tle furnished with an ample drying tube, the junctions being all 

 secured air tight. After careful counterpoising at 64°, we heated 

 it gradually over a small lamp, until it began briskly to boil. On 

 withdrawing it from the lamp, the chloride of calcium was found 

 to have been moistened by the condensed steam, for about half 

 the length of the tube. The original temperature restored, the 

 instrument was placed in the scale. It had lost five tenths of a 

 grain. 



2d. Supposing that this loss might be due to the escape of hy- 

 drochloric acid, we made a similar trial with sulphuric acid, and 

 found the reduction of weight to be about six tenths of a grain. 



3d. Still further to assure ourselves that the hydrochloric 

 acid had not escaped in the former experiment, we renewed the 

 charge, and while heating the liquid, passed the vapor and air, as 

 they escaped from the drying tube, through a solution of nitrate 

 of silver in a test glass. No impression was made upon the test 

 solution, up to the period at which the former experiment was 

 discontinued. But as soon as the whole length of the drying 

 tube was moistened by condensed vapor, the escape of hydro- 

 chloric acid was indicated by dense curds of the precipitated chlo- 

 ride. A like trial with the sulphuric solution gave, even earlier, 

 the same result, the sulphuric acid carried over with the steam, 



