observations was even more seriously restricted by the unfavourable meteorological 

 conditions. The sky remained overcast or a blizzard raged, often for days at a time, 

 particularly during the winter of 1912. 



This statement in regard to the low intensity of aurora 1 observed at Cape Evans 

 appears to be borne out by the large proportion of " clear '' hours on which no aurora 

 could be seen, neglecting, that is to say, hours when meteorological or twilight 

 conditions precluded the possibility of auroral observations. 



Though auroral intensity was estimated in the majority of cases, it appears there 

 was a tendency on the part of certain observers to avoid any mention of intensity in 

 logging the other features. Neglecting such cases, there remain 463 observations 

 between April 24th and September 30th, 1911, for which some indication is given in 

 regard to auroral intensity. These 463 observations are divided up as follows : 



Very bright and bright . . . . . . 46 cases. 



Moderate and fairly bright . . . . . . 169 ., 



Faint 220 



Very faint 28 



The above groups of auroral intensity are, of course, relative, simply expressing 

 the fact that, with reference to the mean brightness of all auroraB observed, this 

 brightness was far exceeded in 46 cases, while the brightness was considered to be far 

 less in 28 cases. 



Corresponding to the same period (April 24th to September 30th, 1911), the 

 number of observations of " Clear, no aurora " at exact hours was 605, on all occasions 

 when twilight, moon and weather conditions did not preclude the possibility of 

 recording a positive result. These two numbers, 605 and 463, are therefore not 

 strictly comparable.* 



Strictly comparable numbers are, however, available for the period May 13th 

 to July 31st, inclusive, counting only observations at exact hours between 4 p.m. 

 and 8 a.m. For this period, there were 656 occasions on which conditions for 

 observation were favourable, while aurorse were recorded on only 236 occasions, 

 or 36 per cent, of the whole. 



The figures given above, showing the number of recorded observations of 

 auroraB of various estimated brightnesses, are of more than passing interest, indicating 

 as they do that the aurora at Cape Evans in 1911 tended to appear of definite brightness, 

 the number of very faint and very bright aurora? being small in comparison with the 

 number of aurora? of mean brightness. It would seem, therefore, that the frequent 

 occurrence of " no aurora " is probably a real effect, corresponding to a total absence 

 of aurora, rather than to occasions when the aurora was present, but of such low 

 intensity as to be invisible to the eye. If this were not the case, one would, I think, 

 expect to find a much higher proportion of aurora? in the " very faint " class. 



* A record of " Clear, no aurora," was only logged at the exact hour, positive results being occasionally 

 recorded also at other times. 



5 A. 3 



