522 
DR. WALTER NOEL HARTLEY ON THE ABSORPTION SPECTRA 
molecule ; if it were, then it would be impossible to distinguish between the three 
isomeric xylenes, or to identify them by a simple inspection of their absorption spectra, 
or even to distinguish them from ethylbenzene. 
It may be conceived to be due in part to an irregularity in the motions of the 
molecules, such as would arise from encounters between unsymmetrically shaped 
bodies ; or if of symmetrical form, such as vortex rings or spheres, the rings or 
spheres being not throughout of uniform density ; in other words, the distribution 
of mass within the molecule being different for each isomer. 
The same effect upon the spectrum as that due to a want of uniformity in the 
distribution of the mass of the parts within the molecule could be brought about by 
periodic contractions and expansions occurring within it. 
It is conceivable that it might be caused by a transference of a part of the intra¬ 
molecular energy to the kinetic energy of the molecule, the intra-molecular vibrations 
being actuated by the radiant energy of the spark. 
A striking feature in all the benzene homologues examined is the existence in their 
spectra of certain bands which remain unaffected by temperature and pressure, and 
these bands are common to all the substances, whether they be isomeric substances or 
not. They are the bands which are subject to displacement when the molecules are 
o 
weighted, and their displacement is uniform and measurable in Angstrom Units. 
It is necessary to refer to the constitution of the solution spectrum of benzene as 
photographed by Messrs. Baly and Collie, because they have ascribed to benzene 
exactly seven, and no more than seven, solution bands (‘ Chem. Soc. Trans.,’ 1905, 
vol. 87, p. 1332). 
The head of a band appears at wave-length 2656, and this they consider to be the 
second of the solution bands. Every solution band must have its origin in a group 
of vapour bands , or it may be derived from a single vapour band. A study of the 
vapour spectrum under varying conditions of temperature and pressure shows that 
the second solution band , which is narrow and of very feeble intensity at ordinary 
temperatures, is not represented by any corresponding band or group in the vapour 
spectrum at the same temperature and with the same quantity of substance which 
gives rise to six groups of bands. 
At a temperature of 45° C. and barometer 759'5 mm. the vapour spectrum has two 
feeble narrow bands at wave-lengths 2636 and 2633. Friederichs and also Grebe 
measured a vapour band at X 2633, which apparently corresponds to the solution band 
measured by Baly and Collie at 2656, but it is narrow and feeble, and there is no 
strong band or group near with which it might be confounded. The same solution 
band has been photographed (Hartley and Dobbie, ‘ Chem. Soc. Trans.,’ 1898, vol. 73, 
p. 695), but it was found to be too feeble to be measured, and also it was different in 
character from the six principal bands. Moreover, its position was so close to the first 
band at the least refrangible end that it might easily have been mistaken for an 
extension of it caused by dilution or by decreasing the quantity of substance. 
