228 Mr. J. N. Lockyer. On the Classification [Jan. 10 r 



(1.) That 16 of the aurora lines nearly coincide with air lines, 

 8 with the positive-pole spectrum of nitrogen, 4 with the negative-pole 

 spectrum of nitrogen, and 3 with lines of hydrogen. 



(2.) That the aurora spectrum greatly resembles that of lightning, 

 and regards it as consisting of several superposed spectra. The 

 variable character of the spectrum is accounted for by the absence 

 sometimes of one, sometimes of another, of these elementary spectra. 



(3.) The brightness of the aurora, according to M. Gyllenskiold,. 

 does not depend upon the energy of the electrical discharge which 

 produces it, but upon some cause with which we are not acquainted. 



Note. It is not out of place to suggest that the brightness of the 

 aurora may depend upon the varying quantities of meteoric dust in 

 the atmosphere at different times. 



(4.) Two kinds of aurora are distinguished, viz., red ones and 

 yellow ones. In the former, the positive-pole spectrum of nitrogen 

 is predominant, while in the latter the negative-pole spectrum is pre- 

 dominant. Laboratory experiments have shown that the positive- 

 pole spectrum of nitrogen is given by dense moist air, while the nega- 

 tive-pole spectrum is given by rarefied dry air; and Gyllenskiold 

 suggests that yellow auroras are formed in the higher parts of the 

 atmosphere, and the red ones in the lower layers. 



(5.) That the observations bear out Angstrom's suggestion that 

 some of the bands belong to the negative-pole spectrum of nitrogen. 

 He says : " Nos observations confirment done 1'opinion d'Angstrom, 

 que les bandes faiblement lumineuses de 1'aurore boreale appartiennent 

 au spectre du pole negatif ; auxquelles les bandes et les lignes de 

 1'azote se joignent dans certains cas." He observes that the cha- 

 racteristic line of the aurora appears in company with the negative- 

 pole spectrum, and says it is probable that some of the more refran- 

 gible bands of the positive-pole spectrum also appear at the same 

 time. Both the positive and negative-pole spectra are very rich in 

 violet and ultra-violet rays, and Gyllenskiold's observations support 

 Angstrom's view, that the characteristic line is due to the fluorescence 

 of oxygen produced by the violet light of the negative pole. 



This fluorescence, however, cannot be reproduced in experiments 

 with Geissler tubes, and M. Gyllenskiold concludes that the origin of 

 the characteristic line still remains unexplained, but he suggests that 

 its origin may eventually be discovered by investigation of the 

 fluorescent spectra of various chemical substances. 



The characteristic aurora line therefore remains unexplained by 

 M. Gyllenskiold. As regards the remaining lines, he states that 

 sixteen nearly coincide with air lines, but it is important to note that 

 these are not the sixteen strongest air lines. Some of the lines fall 

 near to bands in the positive-pole spectrum of nitrogen, as Gyllenskiold 

 points out, but equally strong- or stronger bands are not seen in the 



