conformity with the Dynamical Theory of Heat and Gases. 249 



librio of pressure and temperature (and conforming to the pri- 

 mary laws of Mariotte, and of Dalton and Gay-Lussac), there 

 exists the same number of separate parts. On the other hand, 

 these parts cannot be in all cases the atoms that enter into 

 chemical union, or what is understood by the term "com- 

 bining element." The researches of Dumas and Mitscher- 

 lich have brought out this fact very obviously in the case of 

 sulphur, phosphorus, arsenic, and selenium. In sulphur the 

 ultimate element of the vapour or gaseous molecule must con- 

 sist of six chemical atoms up to the temperature of 860° C, at 

 which, according to MM. Deville and Troost's researches, the 

 vapour-density is reduced to one-third, showing that the gaseous 

 molecule had then split into three, each of which was made up 

 of two chemical atoms. In selenium, the vapour molecule must 

 consist of four chemical atoms ; and at 1040° C. there was found 

 such alteration of vapour-density as led to the anticipation that 

 from about 1200° to 1400° the ultimate vapour molecule would 

 contain only two chemical atoms (MM. Deville and Troost, 

 I860). 



If we were to view the ultimate moving part of all the pri- 

 mary gases as consisting of at least two chemical atoms, the 

 same symbols would express not only the combining proportions 

 of all the bodies entering into union, but the actual composition 

 of a volume of the compound, and we should get rid at once of 

 all the ambiguity and difficulties in the arithmetic of combining 

 proportions. The absolute necessity of doing so may, I think, 

 be easily demonstrated from the examples chosen by Mr. Odling 

 in his lecture. 



The argument of the lecture is to show from the phenomena 

 of substitution that the hydrogen of the water molecule is expe- 

 rimentally divisible into two equal parts; for it is from such 

 phenomena that the chief, if not only chemical proof is to be 

 found. To quote Mr. Odling's words, " It is evident, for instance, 

 that the molecule of marsh-gas must contain four proportions of 

 hydrogen, because we are able to replace one-fourth,-^ wo -fourths, 

 three-fourths, and four-fourths of its hydrogen by four succes- 

 sive substitutions to produce a series of bodies differing from one 

 another by a regular gradation of properties." 



In marsh-gas the ratio of Hy, the quantity of hydrogen by 

 weight, to the carbon Qx, is ^. If the vapour-density of carbon 

 were 3 to hydrogen unity, the numerator added to the denomi- 

 nator of the fraction £, making 4, would represent the vapour- 

 density of marsh-gas, and be represented by HC (#=y=l), in 

 which the letters represent the vapour-densities of each of the 

 constituents respectively. But the vapour- density of marsh-gas 

 is 8; and assuming the vapour-density of carbon to be 12, we 



