54 



SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 65 



and the compression is due entirely to the action of the increasing 

 number of groups of eight — which might account for the increasing 

 compression, though, on consideration, it cannot be definitely said 

 that it would. 



There is another reason why atomic volumes should not increase 

 as fast as N, which is not general but applies only to the halogens 

 and the other " negative " elements for which we chose corrected 

 " effective numbers " on account of their tendency to form the group 

 of eight. This tendency to form the group of eight will be dimin- 

 ished as N increases ; 1 hence the effective number of valence mag- 

 netons will be increased. 



In this unavoidably involved set of predictions there is still another 

 factor to be taken into account. As the surface of the atom increases 

 with N, the average force on unit area that is caused by the action 

 of a given number of valence magnetons must diminish — more 

 rapidly at first than for large values of N. This will be true for all 

 except the inert elements (see c, above). Hence, as iV increases, 

 atomic volumes (except for He, Ne, etc.) will increase faster than 

 we have supposed, and the final predicted curve could be got from the 

 curve in the diagram by multiplying the separate values which it 

 represents by a set of factors whose relative magnitudes are roughly 

 indicated by the curved line OB. 



The extremely qualitative and even uncertain nature of this reason- 

 ing is very apparent, but it is the best that can be done under the 

 circumstances. 



If atoms have envelopes, as has been assumed in this section, we 

 would expect the envelopes in a diatomic gas molecule to be dis- 

 tributed somewhat as follows : 



the compressions due to collisions (see T. W. Richards, Journ. Amer. 

 Chem. Soc, 36, 617-634, 1914) affecting the envelopes chiefly. The 

 advantage of this idea is obvious, for if polyatomic molecules were 



1 This effect has not been alluded to as yet, but it is doubtless connected with 

 the fact that as the number of groups of eight in the atom increases, the 

 valence magnetons must be more and more scattered. 



