92 ME. GEORGE C. SIMPSON ON THE 



light which retains its form and colour for a comparatively long time, and colours 

 which flash out and disappear immediately. A very interesting fact struck me with 

 regard to the latter class of colour. It is generally known that an aurora arch is 

 often composed of a series of spear-like shafts of light arranged perpendicularly to the 

 direction of the arch, and which appear to be in constant motion. A number of these 

 spears will suddenly become brilliant and the lower ends shoot out of the arch into 

 the black sky below. The brilliancy will then run along the arch like a wave of 

 light, lighting up all the spears as it goes along. I noticed that the " front" of such 

 a wave of brilliancy and the points of the spears when shooting out were bright red, 

 but as soon as the motion stopped the colour disappeared, while the more violent the 

 motion the purer and brighter the red. It appeared as if some physical process 

 accompanied the passage of the aurora beam through the air and gave out a red 

 light. For example, if the air had to be ionized before the discharge could pass 

 through, then the process of ionization produced red light. If the motion was 

 particularly violent, the production of red light would be followed by a production of 

 brilliant green light, so that if a bright wave passed along an arch two waves of 

 colour would appear to travel along, first a wave of red light, closely followed by a 

 green wave, the two travelling so closely together as to appear one wave having a 

 two-coloured crest. Similarly spears shooting out with a great velocity would appear 

 to have red and green tips. 



The question of the relation of clouds to auroras has been very often raised. Three 

 of my observations bear on this point. 



On the evening of October 11, 1903. after a fairly active display, the aurora 

 disappeared ; but its place was taken by a system of narrow bands of cirrus clouds 

 stretching right across the sky, which, being illuminated by the bright moon, had all 

 the appearances of the aurora. That they did not form part of the aurora could only 

 be decided at first owing to no line appearing in the spectroscope when pointed at 

 them ; but later there could be no doubt, as they partly obscured the moon. 



On October 26 a very similar phenomenon again appeared ; that which at first was 

 taken to be aurora later turned out to be cloud. 



On December 13 the most brilliant aurora display of my stay took place. The 

 whole display reached a climax at 9.45, when a most brilliantly coloured corona shot 

 out from the zenith. While this final brilliant display was taking place the sky 

 suddenly became thinly overcast, and the aurora was only visible later as bright 

 patches through the clouds. 



It has long been a matter of controversy as to whether the aurora ever extends 

 into the lower regions of the atmosphere. Several observers positively affirm that 

 they have seen it quite close to the ground. This may be due to an optical illusion ; 

 one evening I was, for a considerable time, in doubt as to whether the aurora was 

 really under the clouds or not. All over the sky were detached clouds, the clouds 

 and spaces between them being of about the same size and shape. Right across the 





