TBANSACTIONS OF SECTION B. 105 



elements of Group VII. holds good also for the elements of the 

 other natural groups. It was proved, however, that the law, 

 in its present form, cannot be extended to inorganic com- 

 pounds, but must be confined to the compounds of elements 

 with hydrocarbon radicles. The law, then, fully stated, is as 

 follows : — 



If liquid compounds of the elements of the same natural 

 group zvith the sayne hydrocarbon radicle he compared at tem- 

 peratures ativhich they have equal vapour-pressures, the volumes 

 occupied by unit weights are inversely proportional to the 

 molecular weights and directly proportional to the absolute 

 temperatures. 



It is easy to prove as a corollary that such compounds also 

 undergo equal expansions in changing from liquid into vapoitr. 



Further, when the above law is considered in connection 

 with a thermo-dynamical relation discovered a few years ago 

 by Eamsay and Young, it follows that the latent heats of 

 vaporisation of such compounds folloto exactly the same laio of 

 relation to temperature as do the molecular vokimes themselves. 

 This last generalisation is due to Mr. E. F. J. Love, who called 

 my attention to it. 



When different liquid compounds are compared, what de- 



V 



termines the equality or inequality of m ? In the case of very 



nearly-related compounds, as defined in the law, we get 

 equality. That is to say, you can change C.2H5CI into C2H5I 

 or C^HsBr (for example) without materially altering the value 



V 



of ns; but you cannot change either of them into ( 02115)20 



without such alteration; nor can you replace the ethyl by 



V 

 propyl without such alteration. Further, -m has a different 



value in the compound CCI4 (for example) from what it has in 

 the chemically similar compound SnCli. Hence that kind of 

 chemical similarity which we are accustomed to associate 



with isomorphism has to do with equality of m ; but it is not 



the only condition necessary. 



Eeference to Professor Young's research on the physical 

 constants of benzene and its monohalogen derivatives (Trans. 

 Chem. Soc, 1889) — a research which it would be difficult to 

 praise too highly — afforded a clue to the other essential factor. 

 For in the compounds C0H5F, CeHgCl, CeH^Br, and GJIJ. we 



V 

 have a case in point : the value for ^ is constant except for 



the usual very slight variations already mentioned. Now, 

 these compounds, according to Young, have all the same, or 



