GENERAL DISCUSSION OF THE SPECTRA. IO7 



melting point 43°, which in the course of examination as a soHd film 

 became a liquid. It is quite crystalline in a solid film, hence more 

 opaque, due to scattering- of energy, so that it became more transparent 

 on melting. This is the only compound found to behave thus. 



It was an easy matter to repeat the observations on several sharp 

 absorption bands, after finishing the exploration of the whole spectrum, 

 to see whether the transparency remained the same as in the beginning, 

 and in no case was there a change observable, excepting for the solid 

 thymol just mentioned. 



This heating of the cell was applied in examining several compounds 

 with low melting points, e. g., phenol and menthol in their liquid state. 

 It was only necessary to melt them before placing them upon the spec- 

 trometer arm, then to keep the shutter raised longer than usual. This 

 made possible the examination of several compounds which as solids 

 would have been too opaque because of their scattering effect. 



THE EFFECT OF CERTAIN CHARACTERISTIC GROUPS OF ATOMS. 



Having shown that the infra-red absorption spectra depend upon the 

 internal structure of the molecule, and that their maxima are not influ- 

 enced by molecular weight (of the molecule as a whole, e. g., the petro- 

 leum distillates), the next step is to determine, if possible, what groups 

 of atoms, or ions, have the power of absorbing heat waves. This is of 

 considerable importance, since many recorded phenomena have been 

 credited to "the resonance^ of the OH ion in the molecule." 



Aschkinass (loc. cit.) found the absorption bands of water at the 

 wave-lengths 1.5 1 /x, 3.06 /n, and 6.1 /it. Although he says little about 

 the sequence of the maxima, subsequent writers have laid considerable 

 stress upon it, as showing harmonics, i. e., electromagnetic resonance. 

 Marx,^ in finding the dielectric constants of water for electrical waves, 

 finds a double harmonic relation for the electrical region. Ransohoff 

 (loc. cit.) studied six alcohols, and found the bands harmonic at 1.71 

 (3.0) and 3.43 fi. Although the alcohols were " chemically pure," that 

 is a diflferent question from the one of having them "water free," which 

 he does not consider, and the 3.0 /x and 6.06 fi bands may be due to water. 

 The higher alcohols, like glycerin, even if they could be freed from 

 water, are so hydroscopic that they are difficult to investigate. As 

 already mentioned, in the present work alcohols were avoided because 



•Marx: " Potential Fall und Dissociation in Flammengasen," Drude's Ann., 2, p. 795, 

 1900 ; also on Electromagnetic Resonance, Wied. Ann., 66, p. 600, 1898. After a 

 lengthy discussion he concluded that although it is a plausible assumption it has not 

 been proven that electrolytic dissociation in a flame depends upon the effect of the 

 electromagnetic resonance of the OH ion upon the infra-red radiation. 



