APPENDIX. 675 



this process does not answer when the body analyzed contains 

 azote in the form of nitric acid, not even when mixed with six 

 times its weight of sugar. Indeed, the late experiments of M. 

 Reiset have shown that the process of Varrentrapp and Will is 

 not susceptible of absolute accuracy. When bodies destitute of 

 azote, as sugar and stearin, are heated in a combustion tube 

 with a mixture of lime and hydrate of soda, a certain portion of 

 ammonia always makes its appearance, derived from the azote of 

 the common air contained in the tube. This azote first unites 

 with carbon, and forms cyanogen, and the cyanogen is ultimately 

 converted into ammonia. Sugar treated in this way gave 1'03 per 

 cent, and stearin 0'92 per cent of ammonia. The error from this 

 source in the eighteen analyses made by Varrentrapp and Will, 

 namely, of urea, uric acid, taurin, oxamide, caffein, asparagin, 

 melamin, hippuric acid, amygdalin, narcotin, piperin, brucin, 

 harmalin, fibrin, albumen, and casein, protein and oil of mustard 

 was exceedingly small. But if Manzini's analysis of cinchovina 

 be correct, that it contains 7 '18 per cent, of azote, the error, 

 when the azote is determined by mixing it with sugar, and col- 

 lecting the ammonia formed, is so great that the azote is increas- 

 ed from 7-18 to 11*95, or 4-77 per cent.* 



Such are the methods of determining the carbon, hydrogen, 

 and azote contained in organic bodies. These being added to- 

 gether, and the sum subtracted from the weight of the organic 

 body subjected to analysis, the remainder must represent the 

 weight of oxygen which the body contains ; but which, from the 

 way in which the carbon and hydrogen are obtained, united to 

 oxygen, cannot be evolved in a separate state. 



Having obtained the weight of each constituent in 100 parts - 

 of the organic body subjected to analysis, the next step is to de- 

 termine the number of atoms of each ingredient contained in an 

 integrant particle of the organic body. To determine this we 

 must know the atomic weight of the body under examination. 

 Now this body may be an acid, or a base, or a volatile neutral 

 body, or a fixed volatile body. 



1. To determine the atomic weight of an acid we must, in the 

 first place, ascertain how much water it contains when in crys- 

 tals, how much of this water can be driven off by the highest tem- 

 perature which it can bear without decomposition, and how 



* Ann. de Cbim. et de Phys. (3d series), v. 469. 



