378 GIBBS AND PRATT. 
the colored modification than in either of the para compounds 
previously described. 
Additional proof of the quinoid structure of the salts of o- and 
p-nitrophenol is found in the work of Baly, Edwards, and 
Stewart ® which shows that the band heading at 300 pp is 
present in p-nitroanisole and in its incipiency in o-nitroanisole, 
and that neither of these curves is altered in any way by the 
addition of alkalies. The fixation of the labile hydrogen atom 
in these cases, as in many others which could be cited, precludes 
the change to the quinoid form. 
For purposes of comparison we have photographed phenol with 
various concentrations of alkali, and the small shift in the 
position of the band, about 20 yw», is of an entirely different 
character from that noted in the nitrophenols and p-nitroso- 
phenol. The change in the band heading at 270 pp» on the 
addition of alkalies is not manifested by an increase in the 
persistence, but by a broadening of the band and a progressive 
movement toward the longer wave lengths. The extent of this 
movement depends upon the concentration of the alkali, but is not 
very great in the presence of a large excess of sodium ethoxide. 
Our plates check with the corresponding curves charted by Baly 
and Ewbank® and it does not seem necessary to plot the inter- 
mediate curves obtained with fractional equivalents of sodium 
ethoxide. 
Baly, Tuck, and Marsden *° have noted that all the nitro com- 
pounds containing the hydroxyl or amino groups in the ortho 
position give evidence of two bands, while the para isomerides 
show only one absorption band. They state: 
The reason for this is not clear, but it seems to be characteristic of 
nearly all ortho disubstituted benzenes with two powerful groups of 
different types. 
It seems to us that the generally accepted idea, that two 
groups para to each other are in closer relation than when in 
the ortho position, will, in some cases at least, have to be modified. 
Isorropesis appears to be more easily excited between the groups 
in the ortho than in the para positions. It is probable that at 
one stage of the vibration of the benzene ring the two groups 
para to each other may be in close proximity from which position 
they recede to a maximum distance in accordance with the expla- 
*‘Tbid., 518. 
*Tbid. (1905), 87, 1351. 
*Tbid., 590. 
