Miuichester Memoirs, Vol. liv. {\g\o), No. H, 7 



and E. von Meyer'" in 1807. x-\. Wurtz, putting the date 

 even earlier, says "the truth of the fixity of chemical 

 proportion was definitely established in the year 1806.""^ 

 These writers do not adduce any evidence in support 

 of their statements. They seem to consider their case so 

 probable that proof is unnecessary. On the contrary their 

 case is not even probable. It depends on a fatal under- 

 estimate of the influence of BerthoUet. He occupied a 

 commanding position in the world of science, so that his 

 ideas could not fail to receive consideration in full. Not 

 only were his main ideas of the highest intrinsic value but 

 his teaching on the very subject of constant proportion, 

 in the light of the knowledge which was then available, 

 was extremely plausible. There are tw^o reasons for this. 

 First, chemists in the XVII Ith century had concentrated 

 their attention on the outstanding compound of each pair 

 of elements, and on this insufficient basis the doctrine of 

 constant composition had been founded. BerthoUet 

 raised a new problem by studying the relation between 

 the different compounds of the same elements. It has 

 already been suggested that his teaching did not so much 

 contradict Lavoisier's and VVenzel's and Richter's as go 

 beyond it. While holding in general that affinity tends 

 to unite substances in all proportions, he pointed out that 

 this tendency could be limited by physical factors such as 

 cohesion and insolubility and elasticity, in which case the 

 compounds would be produced on which the supposition 

 of fixed proportion had been based. 



In the second place the wretched state of chemical 

 analysis only too easily afforded data in support of 

 variable proportion. BerthoUet's theory suited the exis- 

 tence of discordant analyses of the same substance by 



''■> "Hist, of Chem.," Eng. trans., p. 194, 1906. 



-" '• The Atomic Tiieory," Eng. trans., p. 9, 1880. 



