Prof. Draper on the Production of Light by Heat. 393 



Having thus by repeated experiments ascertained the contin- 

 ued extension of the more refrangible end as the temperature 

 rose, it became necessary to obtain observations for points below 

 1210°, the limit of visibility through the telescope. I therefore 

 carried the prism nearer to the platinum, and looking with the 

 unassisted eye directly through it at the refracted image, I found 

 it could be distinctly seen at a temperature as low as 1095°. 

 Under these circumstances the total length could not be compar- 

 ed by direct measurement with the other observations, and the 

 result given in fig. 2 is from the best judgment I was able to 

 form : the colors were red and greenish-gray. 



The gray rays emitted by platinum just beginning to shine 

 appear to be more intense than the red ; at all events the wires 

 in the field of the telescope are more distinctly seen upon them 

 than upon the other color. I give them the designation of gray, 

 for they appear to approach that tint more closely than any other ; 

 and yet it is to be remarked that they are occupying the position 

 of the yellow and green regions. 



Already we have encountered a fact of considerable impor- 

 tance. The idea, that as the temperature of a body rises it be- 

 gins to emit rays of increasing refrangibility, has obviously to be 

 taken with a certain restriction. Instead of first the red, then 

 the orange, then the yellow, &c. rays, in succession, making 

 their appearance, in which case the spectrum should regularly 

 increase in length as the temperature rises, we here find, at the 

 v ery first moment it is visible to the eye, it yields a spectrum 

 reaching from the fixed line B to nearly F ; that is to say, equal 

 to about two-thirds the whole length of the interference spec- 

 trum, and almost one-half of the prismatic. 



It is to be remarked, that while the more refrangible end un- 

 dergoes a great expansion, the other extremity exhibits a corres- 

 ponding though a less change. As very important theoretical 

 conclusions depend on the proper interpretation of this fact, we 

 ttust not forget that, to a certain extent, it may be an optical 

 deception, arising from the increased brilliancy of the light. 

 While the rays are yet feeble, the extreme terminations may be 

 s <> faint that the eye cannot detect them ; but as the intensity 



rises, 



and 



the spectrum is the consequence. 



It is agreed by optical writers, that to the human eye the 

 yellow is the brightest of the rays. In the prismatic spectrum 

 the true relationship of the colors is not ]>erceived, because the 

 •ess refrangible are crowded together, and the more refrangible 

 unduly spread out. But in the interference spectrum, where the 

 c °lors are arranged side by side in the order of their wave-lengths, 

 the centre is occupied by the most luminous portion of the yel- 



declin 



Second Series, Vol. IV, No. 12— Nov., 1847. 



60 



