and Nature of the Cosmic Electric Rays. 85 



ordinary colour. At the same time other parallel drapery- 

 shaped arcs appeared which, however, were greenish-yellow. 



At other times the aurorse seem to be painted red all over. 

 On April 17, 1917, the author observed at Christiania 

 another interesting case of red-coloured aurorse, which on 

 this occasion appeared on the northern sky in various forms. 

 Sometimes it had the form of a diffuse arc, sometimes it was 

 more like the pulsating form, and all the while long rays 

 passing nearly to the zenith were formed. While the forms 

 changed the position was kept fairly unaltered, indicating 

 that the change of form was probably not accompanied with 

 any essential change of the properties of the cosmic rays. 

 Usually the aurora had the ordinary colour, but then all of 

 a sudden a large part of it became dark red all over, and the 

 change of colour was not accompanied with any essential 

 change of form. It appeared as if the aurora, with the 

 shape it had taken, had suddenly been painted red. There 

 was in this case no difference in colour between the lower 

 and the upper part of the auroral streamers. The same 

 phenomenon was also observed by Professor St^rmer from 

 another place near Christiania. 



A similar phenomenon was observed by Krogness at 

 Troms0 on November 29, 1918. An auroral display of 

 ordinary colour suddenly — with its rays, streamers, and 

 arcs — turned dark red, and this situation lasted for a com- 

 paratively long time — about half an hour. This is the only 

 instance of this kind that has been observed since the 

 Observatory was erected. 



With regard to the explanation of these curious colour 

 phenomena, a number of possibilities have been mentioned 

 by the author * in some previous papers. If the colour 

 change were of such a nature that it could be regarded as a 

 function of the height, it would be most natural to suppose 

 that it was due to a change in the composition of the atmo- 

 sphere. An aurora like that observed on October 11, 1912, 

 with a red bottom edge, would naturally suggest an ex- 

 planation of the sort that the red colour appears where the 

 electric rays were penetrating down below a certain height 

 of the atmosphere. If so, the bottom edge of the red 

 aurora should be nearer the ground than that of an ordinary 

 greenish-yellow aurora. There are, however, as yet no exact 

 height-measurements of this type of aurora. If, however, — 

 as indicated hj the auroral displays on April 17, 1917 and 

 • November 11, 1918, — the red colour appears all along the 



* L. Vegard, Jahrbuch tl. Bad. u. Electronic, xiv. \\ 468 (1917). 



