﻿Temperature on the Spectra of Incandescent Gases. 197 



belonging to a definite wave-length \, with the density cr of the 

 incandescent gas, is the following : — 



E te =[i-(i-A x n^. 



Accordingly E\<r vanishes for cr=0, and for cr=oo reaches the 



maximum value -r^« 



For a given value of <? and a given temperature, E\ ff for a 

 certain value of X will be an absolute maximum ; or, in other 

 words, among the various bright lines of a discontinuous spec- 

 trum, one will be the brightest, since in a given spectrum both 



E x 

 A A and -r~ vary as functions ofX. 



x 

 Having regard, now, to the fact that, so soon as the value of 



Ex ff sinks below a certain limit (given by that of the sensibility 



of our eye), the place in question of the spectrum vanishes to our 



perception, from these considerations results the following 



theorem : — 



If with the temperature constant the density of an incandescent 

 gas is constantly diminished, the number of the lines of its spectrum 

 must also be diminished and finally, in general, the whole spectrum 

 be reduced to a single line, the situation of which depends on the 

 temperature and quality of the gas. 



I believe that this theorem may be regarded as confirmed by 

 the observations published during the past year, by Eranklaud 

 and Lockyer, in the Proceedings of the Royal Society, No. 112. 

 The passage in question is as follows : — 



" Under certain conditions of temperature and pressure, the 

 very complicated spectrum of hydrogen is reduced in our instru- 

 ment to one line in the green corresponding to E in the solar 

 spectrum. 



"The equally complicated spectrum of nitrogen is similarly 

 reducible to one bright line in the green, with traces of other, 

 more refrangible, faint lines." 



Yet these observations do not permit us at once to draw con- 

 clusions as to the temperature of those heavenly bodies which, 

 like many of the nebulae, present the remarkable phenomenon of 

 very simple spectra ; the preceding considerations show that such 

 conclusions would be inadmissible, since, at any temperature, 

 sufficient rarefaction of the incandescent gas may reduce its spectrum 

 to a single line, the situation of which, in the case of one and the same 

 substance, is dependent on the temperature only. 

 J Having regard to the above-demonstrated principle of the 

 equivalence of the density and the thickness of the radiating layer, 

 it may even be maintained that the values of the temperature 



