﻿Temperature on the Spectra of Incandescent Gases, 203 



the effect being the same as an increase of pressure would have 

 produced with weaker discharges'*. 



9. In our investigations hitherto, only two parts of the spec- 

 trum have been compared. But when the object is to deter- 

 mine generally the changes in the contrast with which a bright 

 line stands out from the darker ground, the alterations in the 

 brightness of the ground on both sides of the line must be taken 

 account of. Here two cases may occur : either the values of 

 the absorptive power on the two sides are equal, so that its curve 

 declines symmetrically from the maximum ; or this symmetry 

 does not exist. It is obvious that in the latter case (which, 

 being evidently the most general, will be the most frequent) the 

 widening produced in the lines by alterations of density must 

 be quicker on that side of the line on which the absorptive 

 power is the greater and hence its curve declines less steeply. 

 Although, as above remarked, the maximum of brightness of a 

 line cannot be displaced by alterations of density alone, yet the 

 effect of the asymmetry mentioned will in general be that the 

 middle of the line widened by increase of the pressure will no 

 longer coincide with that of the line previously to the widening, 

 so that, on the supposition made, the following theorem may be 

 advanced : — 



The middle of a line widened by increase of the pressure is dis- 

 placed toward that side of the spectrum in which the greater values 

 of the absorptive power of the incandescent gas are situated. 



Hence we must not at once refer the displacement of the cen- 

 tres of unequally wide lines of the same substance in different 

 sources of light to an alteration of the refrangibility by the ap- 

 proximation or removal of the luminous body. In general, such 

 a cause must not be assumed as certainly existing until the dis- 

 placement of several lines of the same substance, in a qualita- 

 tively and quantitatively accordant manner, has been shown. 



10. The widening of lines through increase of the vapour- 

 density of the incandescent gas can be very simply shown in the 

 lines of sodium, if we, by pushing a grain of chloride of sodium 



* Compare Pogg. Ann. vol. cxxxvii. p. 344 et seqq. Wiillner here 

 compares the spectrum of hydrogen at a pressure of 300 millims. in a 

 Geissler's tube, as seen when produced by discharges from a Leyden jar, 

 with that produced at much higher pressure by discharges from the simple 

 induction-apparatus. His words are : — 



"With further increase of pressure H/3 and Hy expand continually more, 

 so that soon they appear as only maxima of brightness on a continuously 

 luminous background which is growing brighter; at the same time H a be- 

 comes gradually less sharp, indistinct at its edges, and widened, so thai when 

 the pressure of the gas is 300 millims. the spectrum produced by using the 

 jar has about the same appearance as when, without the jar, the pressure of 

 the gas amounted to nearly 3 atmospheres. The brightness of the spectrum 

 is also about the same." 



