392 DISTRIBUTION OF HEAT IN THE SPECTRUM. [MEMOIR XXVIII. 



should cease in the green or blue or in any middle ray. 

 On the contrary, calorific effects ought to be traceable 

 throughout the entire length of the spectrum. These 

 views on the transference of motion from the ether to the 

 particles of ponderable bodies, and conversely, I endeavor- 

 ed to explain in detail in a memoir on Phosphorescence 

 inserted in the Philosophical Magazine, February, 1851, 

 p. 98, etc. (Memoir VIII.). I had previously indicated 

 them in the same journal, February, 1847 (Memoir V.). 

 A given series of waves of red light impinging upon an 

 extinguishing surface will produce a definite amount of 

 heat, and similar series of violet waves should produce 

 the same amount ; for though an undulation of the lat- 

 ter may have only half the length of one of the former, 

 and therefore only half its vis viva, yet, in consequence 

 of the equal velocity of waves of every color, the impacts, 

 or impulses, of the violet series will be twice as frequent 

 as those of the red. The same principle applies to any 

 intermediate color, and hence it follows that every color 

 ought to have the same heating power. 



Description of the Apparatus employed. 



The optical arrangement I have employed for carrying 

 the foregoing suggestions into practice is represented by 

 Fig. 87, and in a horizontal section by Fig. 88. 



A ray of sunlight reflected by a Silberman's heliostat 

 comes into a dark room through a slit, a, one millimeter 

 wide. It then passes through a prism, b. On the front 

 face of this prism is a black paper screen, c c, having a 

 rectangular opening, just sufficient to permit the light of 

 the slit to pass. After refraction the dispersed rays fall 

 as a spectrum on a concave metallic mirror, dd, nine inch- 

 es in aperture and eleven in focus for parallel rays. I 

 have sometimes used one of speculum metal, but more 

 frequently one silvered on its front face. In front of this 



