﻿126 M. F. Zollner on the Spectrum of the Aurora borealis. 



for it at the height of 10 geographical miles 0*078 millim., and 

 at the height of 20 miles 0-00001 millim. 



To a layer 1 metre thick of the incandescent air of the aurora 

 borealis, then, at the height of 10 miles, a pressure of 78 mil- 

 lims. in the Geissler's tube would be equivalent, and at the 

 height of 20 miles a pressure of 001 millim., in order to pro- 

 duce at the same temperature a spectrum just as bright as that 

 of the aurora borealis. 



But since, as before remarked, the thickness of the luminous 

 layers in the aurora is most probably to be estimated as amount- 

 ing to, not metres, but kilometres, we should obtain so high a 

 value for the pressure in the Geissler's tube equivalent to the 

 thickness even at the height of 10 miles (namely, 78 metres of 

 mercury for a layer 1 kilometre thick), that the quantity of elec- 

 tricity produced by even the largest induction-apparatus would 

 not possess sufficient tension to overcome the resistance of the 

 air compressed to such a degree. And even if the tension were 

 sufficient, the temperature produced by the discharge would be 

 so high that the spectrum could only be brilliant and continuous, 

 and hence not comparable with that of the aurora borealis. 



From this it is evident that the quantity of incandescent gas 

 particles in a Geissler's tube, in comparison with that in the 

 aurora borealis, is probably extraordinarily minute. But since, 

 notwithstanding this, the spectrum of a gas rendered incandes- 

 cent in such a tube by electricity must at least possess the bright- 

 ness of the aurora-spectrum in order, even with the best appa- 

 ratus as regards intensity of light, to afford a spectroscopic 

 analysis, it follows that the emissive power of the incandescent 

 gas particles in the tube must be immensely greater than that of 

 those in the aurora borealis. Such a difference of emissive 

 power, however, can only be effected by difference of tempe- 

 rature. 



If, then, the light developed in the aurora borealis arises from 

 incandescent gas particles of our atmosphere, the temperature at 

 which this takes place must be very much lower than that necessary 

 to render the same gases incandescent in Geissler's tubes. 



Still, according to Kirchhoff's theorem, the temperature may 

 not be lower than that of a perfectly black body in a state of 

 incandescence, the continuous spectrum of which, in the places 

 corresponding to the aurora-spectrum, is of equal brightness 

 with this. 



From these considerations it results that all the gas-spectra, 

 of various orders, which we can artificially produce can, generally 

 speaking, only belong to high temperatures ; for the relatively 

 great brightness of so small a quantity of the incandescent par- 

 ticles shows that there must be a great emission of light from 



