104 HISTORY OF CHEMISTRY. [LECTURE VI. 



gases (at the same temperature and pressure) there is the same 

 number of (physical) atoms, but that these are still divisible by 

 , chemical means. "We call atoms, the groups of chemical 

 molecules that exist isolated in the gases. The atoms of the 

 elementary gases always contain a certain number of molecules 

 which is unknown to us." 49 The ratio of the densities of the 

 gases gives Dumas the ratio of their atomic weights. In fixing 

 the atomic weights of the solid elements he makes use of the 

 law of Dulong and Petit, which he regards, accordingly, as hold- 

 ing for groups of chemically smallest particles molecules, as 

 we should now say. Further, he employs for the same pur- 

 pose, the relative densities of volatile compounds, making 

 assumptions, from analogy, as to the volume relations of the 

 unknown elementary gases contained in them. Thus he finds 

 the atomic weight of sulphur from the density of sulphuretted 

 hydrogen, which he assumes to be constituted like water and 

 to consist of 2 volumes of hydrogen and i of sulphur vapour ; 

 and that of phosphorus from phosphuretted hydrogen, which 

 he supposes to be constituted like ammonia. His determina- 

 tion of the atomic weight of carbon is noteworthy. He deduces 

 it from the relative densities of ethylene and of marsh gas. He 

 assumes in the latter (as Gay-Lussac had also done previously) 

 2 volumes of hydrogen for i of carbon vapour, and in the 

 former, equal volumes of the two. He thus finds the atomic 

 weight of carbon to be one half of what Berzelius had estimated 

 it to be, that is 6, if that of hydrogen is assumed equal to i. 

 In general, however, the values which he assigns to the atomic 

 weights of the better known elements are the same as those of 

 Berzelius. Mercury, silicon, etc, form exceptions. Dumas 

 does not state the weights of the chemically smallest particles. 

 Berzelius contested the principles of the system just con- 

 sidered, although they were so closely related to his own. 50 

 He thinks that it is absurd to assume fractions of atoms, and 

 says it was formerly the custom to abandon hypotheses as soon 



49 Dumas, Traite. I, 41. 50 Berzelius, Jahresbericht 1828, 80. 



