MOLECULAR AGGREGATION 255 



atomic nature, such modes of vibrations must 

 have existed in the body at a lower temperature 

 and at all temperatures so long as its composition 

 remains precisely the same. The coefficient of 

 absorption will, however, be constant, as this does 

 not depend upon the amplitudes of the vibrations. 



Thus a change in absorptive spectrum means a 

 change in the molecular structure of the body. And 

 this is the point it is our endeavour to bring out. 



As in the case of fluorescence so also here with 

 incandescent bodies, flames, and other high tem- 

 perature sources of light, the change of absorption 

 is the test of change of molecular or chemical 

 composition. 



Therefore, a comparison of the curves of in- 

 tensity and absorption will enable us to judge how 

 far the radiation is due to intramolecular and 

 how far to isolated molecular or atomic motions. 



This test, and the hypothesis of which it is but 

 the test, form the basis of our conceptions, as 

 well as of our experiments on the theory of 

 luminosity. 



A body will always absorb the radiation it 

 emits ; but if the simpler process without change 

 of molecular composition takes place, variation in 

 the amount of the absorption for any particular 

 wave-length will not correspond to variations in the 

 intensity of emission for the same wave-length. Thus 

 Balfour Stewart and Kirchoff's law applies, strictly 

 speaking, niore accurately to the phenomenon of 

 luminosity of a chemical nature than other of the 



