Refractivity and Atomic Interaction. 95 



with the following examples, in which such terms do play an 

 essential part. 



Benzene. 3 Here we have, according to the usual ring- 

 shaped formula, three double bonds between the carbon 

 atoms. Thus, 



6H 



(Ha). 



6-55 2 



(Ry). 



6'73 2 



60 



14'47 8 



14'79 6 



3 "= 



5*05 8 



5'67 9 



-*■' calc. 



= 26-09 



27-21 



-^' obs. 



= 26-0 



27-1 



This again is a satisfactory agreement from the point of 

 view of the above Table. However, the mere presence, in 

 the account, of three double-bond terms |= (which gives 

 an item nearly equal to that due to five hydrogen atoms) is 

 a manifest confession of non-additivitj, i.e. of the presence 

 of considerable atomic interaction. Still, in cases like this 

 the chemist does not speak yet of " exaltation " proper. 

 Nor does he call the triple-bond term |= an exaltation, the 

 latter name being reserved for such cases in which there 

 are extra's even to these sanctioned and fixed extra-terms. 

 The discovery of |= is due to J. W. Briihl, who has done 

 much valuable work in this domain *. Notice that |= is as 

 much as |- of the C-refr activity itself, and, what deserves a 

 peculiar emphasis, the dispersion of [=, from H a to H^, is 

 even about six times stronger than that of the C itself. 

 This is equally true of |=. Yet both of these are looked 

 upon as normal extra-terms, since they are of commonest 

 occurrence. The name of " exaltation," or, its negative, 

 " depression," is used when these ordinary extra-terms are 

 insufficient to maintain the additive law, e. g. when owing, 

 say, to the proximity or not of two multiple bonds, there is 

 still an outstanding disagreement from that law. But from 

 the physical point of view these extra-terms, primary or 

 secondary, being so many manifestations of non-additivity, 

 will appropriately in all cases be called exaltations when 

 positive, and depressions when negative, with respect to 

 certain standards. In doing so, however, we must not 

 deprive the double- and the triple-bond terms of their 



* The reader may consult Briihl's account in " The Development of 

 Spectro-Chemistry," Proc. Roy. Inst, xviii. 1 (1906), p. 122. 



