22 



CLARKE 



the development of the constant ; and the simplest explanation 

 of the fact is the one which I have just given. In the phe- 

 nomena of neutralization we have a reaction between separated 

 ions, and the forces which hold the latter in equilibrium come 

 into play. Hence the ionic combination H — OH and the atomic 

 unions in H — O — H are not necessarily the same, although 

 one fundamental constant seems to govern both operations. 

 The only definite conclusion to be reached at present is, that the 

 heats of formation of the oxygen compounds usually exceed the 

 normal, by differences which are constant for each class. 



Among the organic compounds containing both oxygen and 

 nitrogen, Thomsen gives only six examples, and these are not 

 altogether satisfactory. They respond, approximately, to the 

 formula used for the amines ; but this involves the unproved 

 assumption that the substances, during combustion, dissociate 

 completely ; in other words, that the internal oxygen is ignored. 

 The data are given below, when 



4^ 



12a + 65 + <^m — c —\ 



= constant. 



Mean, 13474 



The questionable character of this mean is evident, although 

 two of the quotients are good. For this class of substances more 

 data are necessary, with a more elaborate discussion than is now 

 practicable. 



Up to this point we have considered only linear molecules, that 

 is, molecules whose atoms are united in such a manner as to 

 form the so-called chains. The ring compounds, which are 

 mainly of the aromatic series, remain to be examined. In the 

 first group of substances n represents the number of atoms in 



