CHEMISTRY: G. N. LEWIS 
589 
triphenylmethyl, associates at low temperature or high concentration 
to form the colorless double molecule. They are the substances which 
possess what I have called, in a recent paper,^ odd molecules. 
I will review briefly the significance of this term. If we assume that 
every neutral atom possesses in its outer shell a number of electrons 
corresponding to the ordinal number of the element's group in the peri- 
odic system, for example, H, 1; C, 4; N, 5; 0, 6; CI, 7, then in almost 
all compounds there is an even number of such electrons. Those few 
substances which possess an odd or unpaired electron are said to have 
odd molecules. Thus sodium in the state of vapor has 1, monatomic 
iodine has 7, NO2 has 17, CIO2 has 19, (C6H5)3C has 91, (C6H5)2N has 
63, and (C6H5)2NO has 69. Since an unpaired electron is always held 
in the molecule by weak constraints, it is usually capable of oscillating 
with so low a characteristic frequency as to absorb visible light. This 
absorption, moreover, involving only a single electron, might be expected 
to be of the simplest possible character. 
In fact it has been shown by Wieland^ that several substances of this 
type, with odd molecules, have very characteristic absorption spectra 
possessing ordinarily a single absorption band. Thus the spectrum not 
only of (C6H5)2NO but also of the triarylmethyls is like that of NO2 
and by no means resembles that of such quinoid substances as tri- 
phenylmethyl ion. There seems therefore every reason to believe that 
triphenylmethyl is colored not because it assumes the quinoid form but 
because it possesses an odd electron which, as in every other known 
substance of this class except NO, has a sufficiently low natural fre- 
quency to absorb visible light. 
But whether or not this point be admitted, the fact that equilibrium 
is rapidly established between the ethanes and their corresponding 
methyls shows conclusively the non-existence of the first type of steric 
hindrance. Instead of a retardation due to the large groups we find a 
reaction velocity which, among organic compounds, is exceptionally 
high. 
Therefore in the discussion of this type of reaction the term, steric 
hindrance, tacitly came to be employed in the second sense, implying a 
crowding out of large groups by other large groups, and this view 
seemed to be favored by the discovery by Schlenk that the degree of 
dissociation of the hexarylethanes ordinarily increases with increasing 
complexity of the substituted aryl groups. 
In order to obtain a comprehensive view let us consider four sub- 
stances which in the light of the periodic system must in certain re- 
