128 STUDIES IN THE DIFFRACTION SPECTRUM. [MEMOIR VII. 



ferent phenomena of chemical changes occurring through 

 the agency of light become at once allied to a multitude 

 of other optical results. If it were necessary, by a very 

 simple arithmetical process we could determine the num- 

 ber of vibrations executed by a ray bringing about a 

 given decomposition in billionths of a second. The 

 fixed lines used in this way enable us at once to divide 

 the diffraction spectrum into any number of parts, and, 

 by comparing wave-lengths and the velocity of light, to 

 indicate effects either in space or in time." 



The diffraction spectrum, as w r e have seen, differs 

 strikingly from the prismatic in the arrangement of its 

 colored spaces. In the latter, the less refrangible parts 

 are compressed more and more in proportion as their 

 refrangibility is less. Now there is reason to believe 

 that in the former the colored spaces are equally warm, 

 though so feeble is the calorific effect that all attempts 

 at the direct measurement of the heat have proved 

 unsatisfactory. I first made such attempts with very 

 delicate thermo-electric apparatus, but could not obtain 

 sufficiently striking results. Admitting, however, that 

 every ray, irrespective of its color, in the act of extinc- 

 tion will generate the same amount of heat, it necessarily 

 follows that in the prismatic spectrum the heat should 

 appear to increase steadily from the more to the less 

 refrangible end, because in it the compression of the 

 colored spaces is becoming greater and greater, and this 

 is what is actually observed. 



These considerations respecting the distribution of 

 heat in the spectrum lead naturally to the examination 

 of a much more comprehensive problem indeed, one of 

 the most important problems that science presents viz., 

 the constitution of the sunbeam. 



Until the time of Newton it was universally admitted 



