CHAPTER I. 



INTRODUCTION. 



The free vibration periods of molecular aggregations may be studied 

 by means of their emission or absorption spectra. Emission generally 

 implies a high temperature and uncertainty of the composition of the 

 radiator, due to dissociation; hence it is more advantageous to study a 

 substance at a low temperature by means of its absorption spectrum, 

 since we then have a more definite knowledge of its composition. This 

 is especially true of the highly complex compounds which are so abun- 

 dant in organic chemistry. 



Although generally assumed that in gases the molecules have the 

 advantage of a greater freedom of vibration, it will be show^n in the 

 present paper that, starting with a simple liquid or solid compound, the 

 free vibration of a given molecule or radical is not always seriously 

 modified as the molecular complexity of the compound is increased. 

 In fact, it has been found that there may be two sets of particles (or 

 "ions"), which vibrate side by side, yet still retain the free period 

 that they had when present, alone, in a simpler compound. Further- 

 more, their free vibration periods, manifested as absorption maxima, 

 occasionally seem to repeat themselves harmonically throughout the 

 spectrum investigated. This presence of groups of ions, each group 

 having its own free period of vibration, is in accord with present ideas 

 of absorption and anomalous dispersion. 



Previous investigations of infra-red absorption spectra have never 

 been extended farther than about y i^ for alcohols and to about lOju. for 

 several other compounds. Now, it so happens that with the limited 

 dispersion at our disposal (if that be the true reason), all carbohydrates 

 investigated show a large absorption band between the wave-lengths 

 3.0 /x and 3.5 /i, and then there are, as a general rule, no marked bands 

 until we arrive at 6 jx. Beyond this point, to the limit of the working 

 transparency of rock salt at 15 ^u,, there are numerous large, well-defined 

 bands. These facts were noticed in some preliminary work, when it 

 became evident that, in order to gain a better knowledge of infra-red 

 absorption spectra than obtained for the region up to 7/1, a very 

 extended^ and systematic investigation of the spectrum beyond this 

 point would be necessary, in magnitude not unlike Kayser and Runge's 

 work on the emission spectra of the metals, for the optical region. 



^Subsequently it was found that this was the conclusion of all the preceding inves- 

 tigators in this field, who, however, never made further contributions to the work. 



3 



