For purposes of classification, auroras were described under the following 

 headings : 



(1) Glows, including cloud-like patches of indeterminate form. In the great 



majority of cases, glows were seen on, or very close to, the horizon. 



(2) Streamers. These appeared in the form of rays or cones, the apices being 



directed away from the horizon. 



(3) Arclies or bands. This form of the aurora usually appeared straight when 



passing through the zenith and gently curved at other times, the curvature 

 being always in the same sense. The aspect of these arches suggested that 

 they were, in point of fact, straight bands in the upper atmosphere, the 

 gently curved appearance being due to perspective. 



(4) Curtains. This designation was applied to aurorse, particularly when of curved 



shape, the radius of curvature varying both in sign and magnitude. 



(5) Corona. This term was applied to the aurora when in the form of discontinuous 



streamers at a high altitude, all streamers being directed towards a single 

 point near the zenith. This form of aurora was only seen on two or three 

 occasions at Cape Evans. 



In a number of cases, streamers were associated both with arches at low altitudes, 

 and with curtain forms. It seems probable, as Vegard suggests, that the streamer is 

 the primitive form of auroral display, the aspect of other types being assumed 

 according to the angle from which the streamers are viewed and the spatial distribution 

 of the streamers. 



2. INTENSITY OF AURORA AT CAPE EVANS. 



Mention has already been made of the fact that aurora was rarely seen during 

 the brighter portions of the day, even in the depth of winter. There can be no 

 doubt, in fact, that the presence of very little twilight was sufficient to obscure all 

 but the brightest aurorse seen at Cape Evans. The same holds true for auroral 

 observations when the moon was above the horizon, and aurorse were very rarely 

 seen close to a bright moon. This result is possibly due to the fact that the year was 

 close to the sunspot minimum and unfavourable for auroral displays, but the result 

 was felt to be disappointing in view of the restricted opportunities for observation. 

 That the position of the station was not favourable for auroral observations appears 

 to be substantiated by the fact that the colours (red or green) usually observed with 

 bright aurorse were seldom seen, mention being made of colour on only 14 occasions 

 during the year. On the occasions when it was estimated that aurora could have 

 been readily seen, the phenomenon was actually recorded only about one hour in 

 three ; on clear days during the dark hours, therefore, the observer was almost twice 

 as likely to log " no aurora " as to record a positive result. 



Apart from this restriction in the number of observations, due possibly to an 

 unfavourable year and certainly to the unfavourable situation, the number of possible 



