I06 INFRA-RED ABSORPTION SPECTRA. 



that the small spectrometer was used, and that in general it was not 

 practicable to make the examinations in the manner just described. A 

 shift of 0.02 fi was easily observed in the bands of xylene and mesitylene 

 at 3.4 fi. 



An interesting fact to be noticed in this connection is that all the prom- 

 inent lines found in the two oils just mentioned are present in all the 

 petroleum oils studied, as well as in many other compounds, like myricyl 

 alcohol, piperidine, etc. For a larger dispersion the transparent region 

 at 4 to 6 /i. remains so for some oils, while in others numerous small 

 bands were found. 



The difference between the spectra of the oils (aliphatic series) and 

 the benzene spectrum (carbocyclic series) has been noticed under the 

 question of structure. The benzene spectrum, as well as that of its 

 methyl derivatives, is banded, " channeled," i. e., the lines occur in 

 groups, just as Pauer^ found for the ultra-violet. He found the bands 

 of the benzene spectrum, which extend from 0.267 /* to 0.235 i^- con- 

 densed and shifted toward the visible spectrum for toluene, the xylenes, 

 aniline, etc., and considered it due to increase in molecular weight. If 

 we consider the center of gravity of the benzene bands at about 0.245 i"» 

 and that of the methyl derivatives at about 0.267 /x, this shift amounts 

 to 0.02 /x, while for aniline it is about 0.05 /x. 



As a whole, there is no evidence of a real shifting of the maximum, 

 with increase in molecular weight, if we except the xylenes and the 

 gases mentioned, for the region at 3.1 to 3.5 fi, which is not resolved for 

 most compounds. The condition is similar to that in Kayser and Runge's 

 work for the emission spectrum of the elements, in which they observed 

 that, for the alkali metals, the violet lines shifted toward the longer 

 wave-lengths, with increase in atomic weight, but they could not estab- 

 lish^ this relation for all the elements. 



EI^FECT OF TEMPERATURE. 



There is little to be said on this subject, for no effect due to rise of 

 temperature of about 20° has been observed. As already mentioned 

 the absorption cell was between the spectrometer slit and the Nernst 

 heater, with the double sheet-iron-asbestus shutter intervening. It was 

 necessary to have the heater close to avoid loss of radiation. Conse- 

 quently, in the course of a series of observations, lasting four hours, the 

 cell would unavoidably become warmer, due principally to the raising 

 of the shutter in making observations. This was first noticed in thymol, 



^Pauer : Ann. der Physik, 61, p. 363, 1897. 

 ^Kayser's Spectroscopy, II, p. 591. 



