LECTURE VI.] HISTORY OF CHEMISTRY. [05 



as they led to absurdity. Dumas stands entirely isolated in his 

 views, but, in spite of this, he would probably have adhered to 

 them, had he not himself discovered facts which caused him to 

 doubt the accuracy of Avogadro's hypothesis. 



Dumas was not only an ingenious thinker, but he was also 

 an excellent experimenter ; and, since he had chosen the den- 

 sities of gases and vapours as the basis of his atomic theory, he 

 thought it necessary to increase our knowledge respecting these 

 densities. He succeeds in elaborating a method for carrying 

 out determinations of this kind at high temperatures, and em- 

 ploys it for ascertaining the relative densities of the vapours of 

 iodine, phosphorus, sulphur, mercury, etc. 51 His results, from 

 which he anticipated confirmation of his views, lead him to 

 abandon them. He finds the density of phosphorus vapour to 

 be twice as great, and that of sulphur vapour to be three times 

 as great as he had previously assumed, whilst that of mercury 

 vapour is only one half of what he had supposed. In view of 

 these facts he begins to doubt ; in fact he declares that even 

 the simple gases do not contain, in the same volume, the same 

 number of chemical atoms. According to him, the assumption 

 may still be made that there are the same number of molecular 

 or atomic groups present in equal volumes of all gases ; but 

 that this is only a hypothesis, which cannot be of any service. 52 

 Dumas is obliged to admit that Gay-Lussac's law, when applied 

 in the manner he had done to the determination of atomic 

 weights, furnishes erroneous results. Hence he believes that it 

 cannot be employed for this purpose, and he now abandons 

 Avogadro's hypothesis. 



Berzelius, too, can no longer maintain the identity of 

 volume and atom in the cases of the elementary gases, and has 

 to confine his proposition to the incondensable elastic fluids. 53 

 It must be admitted that the law, when so stated, was not 

 capable of any extended application, and was more than 

 insufficient for the determination of the atomic weights of the 



51 Ann. Chim. [2] 33, 337; 44, 288; 49, 210; 50, 170. 52 Le9ons. 

 268 and 270. 53 See p. 92. 



