﻿M. F. Zollner on the Spectrum of the Aurora borealis. 125 



a Geissler's tube filled with atmospheric air, and with a thickness 

 of the radiant layer of 1 millim. a sufficient amount of light is 

 developed for the observation of the atmospheric spectrum. But 

 if now this thickness of the luminous layer be compared with the 

 thicknesses present in the aurora borealis, the latter must evi- 

 dently be taken as incomparably greater, and, especially at spots 

 far from the zenith, may be estimated at miles*. But if we even 

 assume for the thickness of such a layer only 1 kilometre, this, 

 at the same temperature as that in the Geissler's tube in the 

 present example, could have only the millionth part of the den- 

 sity of the air contained in the tube, and so only 0'00005 mil- 

 lim. pressure at 0° C, if the spectrum of the incandescent gas 

 in the tube were to correspond perfectly with that of the aurora 

 borealis f. Here, then, the question arises, whether it is allow- 

 able to assume such low rates of pressure in those regions of our 

 atmosphere in which the aurora borealis is developed. 



As regards the height the accounts of various aurorse boreales 

 are very divergent. Thus, not to mention earlier accounts, 

 HansteenJ finds for the aurora of January 7, 1831, the height of 

 26 [German] geographical miles, by combining the measure- 

 ments of the apparent height of the dark segment at Berlin and 

 at Christiansand in Norway. On the contrary, the observations 

 of Farquh arson § make it probable that auroras boreales in ge- 

 neral have a much less altitude, and sometimes descend quite to 

 the region of the clouds. 



The latter view is corroborated by many more recent observa- 

 tions, especially in the polar regions, made by Parry, Wrangel, 

 Franklin, Hood, Richardson, and others; so that the height of 

 from 10 to 20 geographical miles, assumed in the following cal- 

 culations as that of the lower parts of the aurora borealis, may 

 be presumed to be at all events rather too great than too small. 

 If, assuming a uniform temperature everywhere of 0° and the 

 normal condition of the barometer at the surface of the earth, we 

 calculate the atmospheric pressure at those heights, we shall find 



* This may offer a very simple explanation of the fact that, in general, 

 with increasing distance from the zenith the brightness of the aurora 

 borealis increases, as far as the boundary of the dark segment. 



t In all these considerations, it is obvious that the apparent magni- 

 tude of the luminous layer is always assumed to be sufficient to cause 

 the variation of distance of the latter to have no effect on the brightness 

 of the spectrum. In like manner the influence of absorption in the 

 deeper and non-luminous lavers of the atmosphere is neglected. 



% Pogg. Ann. vol. xxii. (1831). 



§ " On a definite Arrangement and Order of the Appearance and Pro- 

 gress of Aurora borealis, and its Height above the surface of the earth," 

 Phil. Trans. 1829; and "On a remarkable Appearance of the Aurora 

 borealis below the Clouds," ibid. 1842. 



