﻿Temperature on the Spectra of Incandescent Gases. 199 



trura with the brightness of the homologous place in the charcoal- 

 spectrum, we can, neglecting the absorption in cosmical space, 

 and taking due account of that in our atmosphere, ascertain in 

 the manner indicated a lower limit for the temperature of the 

 nebula, as soon as we succeed in determining the temperature of 

 the galvanically incandescent charcoal. 



7. The dependence of the position of the lines of a disconti- 

 nuous spectrum on the temperature and quality of the incan- 

 descent gas, above recognized theoretically as admissible and 

 probable, is very remarkable, and appears to me quite adequate 

 to explain the remarkable phenomenon, discovered by Pliicker, 

 of the so-called spectra of different orders of one and the same 

 body ; for the value of the expression for the ratio of brightness 

 of two adjacent places in the spectrum, 



Excr = [1-(1-Aa)*]J\ 



E V [i_(i_ AX| )*]J x ; 

 is dependent, when o- is constant, only on the values of the ab- 

 sorptive powers A A and Aw since ~ for this case may always be 



taken as =1. But these values may have, for the same wave- 

 length and continuous alteration of the temperature, similar 

 maxima and minima to those which they in fact possess for the 

 same temperature and continuous alteration of the wave-length, 

 whereby they produce the phenomenon of discontinuous spectra. 

 The simplicity and continuity attributed to KirchhofFs function 

 J refers only to the ratio between the magnitudes E A and A Kf 

 not to the magnitudes themselves. While that function is 

 the same for all bodies, E A and A A (as functions of the tempera- 

 ture and wave-length) are directly dependent on the particular 

 condition and character of the body. We thus see that the ex- 



Exo- 

 pression for ^ — , with alteration of the temperature, may solely 



through alteration of the values of Ax and Ax, assume different 

 values which are greater or less than unity. From this it follows 

 that the ratio of brightness of two adjacent places in the spectrum 

 may be reversed by alterations of temper ature, and a minimum ap- 

 pear in the place of a former maximum. 



Hence, in relation to the changes of intensity of adjacent parts of 

 a spectrum, there is an essential difference between the effects of 

 temperature and those of pressure : the ratio of intensity may be 

 reversed by changes of temperature ; its reversal is not possible 

 by changes of pressure. By increase of the latter a difference 

 of intensity can only be made to vanish, it cannot be reversed. 

 When, therefore, inversions of this kind are observed in different 

 spectra of one and the same substance, this is the result of dif- 



