APPENDIX. 



G83 



supposition be correct ; but at present it is out of our power to 

 assign any reason whatever, 



Such was the state of the analysis of organic bodies when M. 

 Dumas, by an admirable series of experiments, demonstrated that 

 the atomic weight of carbon, as admitted by the continental che- 

 mists, from the experiments of Berzelius, namely, 0-76438, is 

 about 2 per cent, too high ; and that the true number is O75.* 

 To determine the exact composition of carbonic acid, MM. Du- 

 mas and Stas placed diamonds successively in a porcelain tube, 

 which was heated to redness, and a current of oxygen gas passed 

 through it till the diamond was converted into carbonic acid. 

 The oxygen passed previously through a tube filled with frag- 

 ments of pumice imbibed with caustic potash, and a second tube 

 filled with fragments of caustic potash, to deprive the oxygen 

 gas of every trace of carbonic acid with which it might happen 

 to be mixed. It then passed through a tube filled with frag- 

 ments of pumice impregnated with sulphuric acid, in order to 

 deprive it of any water which it might contain. Beyond the por- 

 celain tube was luted to it a long tube filled with oxide of cop- 

 per, which was kept in a state of ignition during the process, and 

 through which the surplus oxygen and the carbonic acid formed 

 all passed. To this was luted a tube bent like U, and filled with 

 fragments of pumice soaked in sulphuric acid, to imbibe any 

 water that might be formed during the combustion ; then a Lie- 

 big's tube containing caustic potash, then two tubes in U filled 

 with pumice impregnated with caustic potash, and lastly, a tube 

 in U filled with fragments of pumice impregnated with sul- 

 phuric acid, and finally, another tube containing potash in pow- 

 der. The whole apparatus is represented in the figure below. 



It was ascertained by preliminary trials that, in this appa- 



* Ann. de Chim. et de Phys. (third series,) i. 5. These experiments have 

 been repeated arid confirmed by Erdmann and Marchaiit. Ibid. iii. p. 500. 



