LECTURE XV.] HISTORY OF CHEMISTRY. 299 



of the atomic weights, just as the hypothesis of Avogadro was 

 in that of the molecular weights. Even at this time the carrying 

 out of these principles was still confronted by great difficulties. 

 It is true that Deville and Troost 6 showed, at this date, that 

 the vapour density of sulphur, at about 1000, was only one- 

 third of the number previously found by Dumas and Mitscher- 

 lich (compare p. 105) at lower temperatures; so that it was 

 possible to adopt the molecular formula S 2 for sulphur. The 

 anomalies previously observed in the cases of mercury, phos- 

 phorus, and arsenic, remained, however; but these were not 

 an obstacle to Cannizzaro. The chemical relations had to 

 stand aside in order to procure acceptance of the principle. 

 He assumed that only one-fourth of a molecule of phosphorus 

 and of arsenic, respectively, was contained in two volumes of 

 phosphuretted and of arseniuretted hydrogen ; whilst half a 

 molecule of nitrogen is present in two volumes of ammonia, 

 and a molecule of mercury in two volumes of mercuric chloride. 

 Consequently the divisibility of the molecule is different, 

 according to Cannizzaro, even in chemically analogous sub- 

 stances. Even although this appeared to be a bold view, still 

 no decisive reasons could be established against it. 



The assumption of differences of constitution amongst the 

 elementary molecules, although striking at first, seemed after a t/ 

 time to be fully justified. Why should not a state of matters 

 be met with in the case of the elements similar to that observed 

 amongst compound substances, the molecules of which are 

 known to present the greatest variety with respect to the num- 

 bers of their atoms ? Cannizzaro very aptly compares the 

 elements with the hydrocarbons the molecules of hydrogen, 

 oxygen, etc., with the so-called alcohol radicals, methyl, ethyl, 

 etc., and the molecules of mercury, zinc, and cadmium, with 

 the olefines, a view which may also be extended to the derivatives 

 of both classes of substances : 7 



6 Comptes Rendus. 49, 239 ; Annalen. 113, 42. 7 Wurtz, Le9ons de 

 philosophic chimique, 172. 



