588 
CHEMISTRY: G. N. LEWIS 
equilibrium is shifted in favor of the quinoid type in polar solvents, and 
in a very polar solvent such as SO2 the molecules may dissociate into 
positive and negative ions of which the former (and probably the latter 
also) is largely in the quinoid form. 
The recognition of the existence of the quinoid form is indispensable 
to an understanding of numerous reactions which characterize the hexa- 
arylethanes, but since we are here concerned with that part of the sub- 
stance which remains in the benzoid form we may proceed at once to 
the next conclusion. 
II. In nonpolar solvents such as benzene the substituted ethanes 
exist chiefly in the benzoid form and are to a greater or less extent disso- 
ciated into the triarylmethyls. The undissociated part is colorless. 
The product of dissociation, though colored, has also the benzoid form. 
The fact of dissociation has been abundantly demonstrated. The 
mere fact that the equilibrium between colorless and colored forms 
changes in favor of the latter with increasing dilution^ shows that the 
change from colorless to colored is attended by an increase in the num- 
ber of molecules. Measurements of the degree of dissociation, whether 
through molecular weight determinations or by colorimetric methods, 
show that while the parent substance, hexaphenylethane, is at ordinary 
temperatures only to a few per cent dissociated, the dissociation is nearly 
complete in the case of some other aryl derivatives. Schmidlin^ has 
shown that the process of dissociation is not instantaneous but requires 
a short though appreciable time, and hence he was able to show that it 
is the colored substance which reacts rapidly with oxygen or iodine, 
thus exhibiting the characteristics of a free radical. Indeed it might be 
possible under favorable circumstances to obtain the degree of dissocia- 
tion in a new way by rapid titration of the colored substance against 
atmospheric oxygen. 
Gomberg has hesitated to adopt the conclusion that the colored tri- 
arylmethyls retain the benzoid form, on the ground that there is no 
analogy to support the assumption that a colored substance could be 
formed by the mere dissociation of a colorless substance such as the 
substituted ethane. It is true there is no analogy in organic chemistry 
except among these very compounds of trivalent carbon, and the cor- 
responding compounds of bivalent and quadrivalent nitrogen which 
have been investigated with great perspicuity by Wieland and his col- 
laborators. If, however, we examine the whole field of chemistry, we 
see that triphenyknethyl belongs to a class of substances which are 
almost invariably colored and which are typified by NO2 which, like 
