212 ESSAYS BIOGRAPHICAL AND CHEMICAL 



pointing out the identity of this wave-length with the 

 principal auroral line ; and a week later, Professor Arthur 

 Schuster, in a letter to Nature, called attention to the 

 same coincidence. 



It therefore appeared probable that the aurora might 

 be produced by electric discharges in the upper atmo- 

 sphere, through a gas in which krypton was present in 

 considerable amount. 



In the meantime Professor Paulsen, of Copenhagen, 

 has been examining the photographed spectra of the 

 aurora collected by the Icelandic and Finland expeditions; 

 and it appears probable that many of the lines seen in the 

 auroral spectrum are identical with those of the common 

 gas nitrogen, seen at the cathode terminal of a vacuum- 

 tube. Professor Paulsen has had the great kindness to 

 send me prints of two photographs, the lines of which are 

 numbered I. and II. in the following table. I have added 

 in a parallel column the corresponding wave-lengths of 

 krypton lines. 



Lines of auroral and of cathode Lines of 



nitrogen spectrum. krypton. Intensity. 



I. II. 



1 Absent. 3140 



2 3160 



3 3370 



4 Absent. 3540 



5 3580 3580 3590 7 



c 3718 10 



6 3710 Absent. < 3719 8 



I 3721 7 



7 3760 3760 3754 5 



8 3800 3800 3805 4 



9 3920 3920 3920 8 



f 3995 6 



10 400 * 00 I 3998 5 





