LECTURE XL] HISTORY OF CHEMISTRY. 2OI 



be avoided in the cases of the other elements. The different 

 vapour densities which were found in the case of sulphur 14 can 

 only be explained by the assumption that, at low temperatures, 

 the molecule consists of three times as many atoms as at a very 

 high temperature. 



It is certainly not without interest, that, in 1857, Clausius 

 was led, by the mechanical theory of heat, to the divisibility of 

 the physical molecule. 15 Since, according to this theory, the 

 kinetic energy of translatory motion in equal volumes of two 

 gases at the same pressure, is proportional to the absolute tem- 

 perature, Clausius concluded that the kinetic energy of trans- 

 latory motion of the single molecules of all gases at the same 

 temperature is the same. This assumes the fulfilment of Avo- 

 gadro's hypothesis. 



The number of facts drawn from different branches of natural 

 science which tell in favour of this hypothesis might be multi- 

 plied still further : but I confine myself to the statement of those 

 that I have mentioned, and pass on to chemical arguments 

 which, after all, were the only ones that eventually led to the re- 

 cognition of the hypothesis. Amongst these, the experiments 

 which were now carried out and led to the conception of the 

 chemical molecule, distinctly take the first place. I will not 

 assert that this conception was not already extant ; but it now 

 appeared in a much more definite form. This latter assertion 

 will certainly be justified when I collect here the facts and 

 hypotheses which existed and exercised an influence upon the 

 determination of molecular weights by chemical methods, prior 

 to Williamson. 



The atomic theory gave the first clue to these magnitudes. 

 The formula of every compound required to be expressible by 

 means of multiples, by whole numbers, of the atomic weights ; 

 but this had no obligatory consequences so long as the atomic 

 weights had not been determined with certainty, because, in 

 case of necessity, even the atomic weight of a constituent could 



14 See Dumas : Ann. Chim. [2] 50, 170, and Deville and Troost, 

 Comptes Rendus. 49, 239. 15 Pogg. Ann. 100, 353. 



