Prof. Draper on the Production of Light by Heat. 395 



that a ray falling near the edge, and one falling near the back, 

 after dispersion, will paint their several spectra on the screen ; 

 the colors of the one not coinciding with, but overlapping the 

 colors of the other. In such a spectrum there must undoubtedly 

 be a general commixture of the rays ; but may we not fairly in- 

 quire whether, if an elementary prism were used, the same facts 

 would hold good ; or, if the anterior face of the prism were cov- 

 ered by a screen, so as to expose a narrow fissure parallel to the 

 axis of the instrument, would there be found in the spectrum it 

 gave every color in every part, as in Sir David Brewster's original 

 experiment ? M. Melloni has shown how this very considera- 

 tion complicates the phenomena of radiant heat ; arid it would 

 seem a very plausible suggestion that the effect here pointed out 

 must occur in an analogous manner for the phenomena of light. 

 I proceed now to the third branch of the inquiry, — to exam- 

 ine the relation between the temperatures of self-luminous bodies 

 and the intensity of the light they emit, premising it with the 

 following considerations. 



The close analogy which is traced between the phenomena of 

 light and radiant heat lends countenance to the supposition, that 

 the law which regulates the escape of caloric from a body will 

 also determine its rate of emission of light. Sir Isaac Newton 

 supposed that whilst the temperature of a body rose in arith- 

 metical progression, the amount of heat escaping from it increas- 

 ed in a geometrical progression. The fallacy of this was subse- 

 quently shown by Martin, Erxleben, and Delaroche ; and finally 

 Dulong and Petit gave the true law, " when a body cools in 

 vacuo, surrounded by a medium whose temperature is constant, 

 the velocity of cooling for excess of temperature in arithmetical 

 Progression increases as the terms of a geometrical progression, 

 diminished by a constant quantity." The introduction of this 

 constant depends on the operation of the theory of exchanges of 

 heat; for a body, when cooling under the circumstances here 

 given, is simultaneously receiving back a constant amount of 

 heat from the medium of constant temperature. 



Whilst Newton's law represents the rate of cooling of bodies, 

 and therefore the quantities of heat they emit, when the range 

 °f temperature is limited, and the law of Dulong and Petit holds 

 t0 a wider extent, there are in our inquiry certain circumstances 

 to be taken into account not contemplated by those philosophers. 

 Dulong and Petit throughout their memoir regard radiant heat 

 as a homogeneous agent, and look upon the theory of exchanges, 

 x vhich is indeed their starting point and guide, as a very simple 

 a ffair. But the progress of this department of knowledge since 

 their times has shown, that precisely the same modifications as 

 a ^ found in the colors of light, occur also for heat; a fact con- 

 veniently designated by the phrase " ideal coloration of heat , 



