202 



hours only during which time any auroral phenomena would 

 be visible there was, generally speaking, an increase in 

 auroral intensity for several hours after that immediately 

 succeeding twilight. The hour of maximum frequency is 

 remote from the daylight hours. With regard to the detec- 

 tion of a variation in the quantity of auroral phenomena 

 noted month to month, making due allowance for obscuring 

 causes, such as daylight and blizzards, no great difference was 

 observed. It was remarkable, however, that whereas during 

 April, May, and August some of the most brilliant were seen, 

 aurorae were never entirely absent for a day during June and 

 July, though the majority of the displays were much calmer 

 and localized than during the preceding or succeeding months. 



3. Auroral Storm Periods. 



It is generally the case that when brilliant aurorae are 

 observed notable auroral phenomena are witnessed on several 

 preceding or succeeding days. At Cape Royds this was 

 notably the case. The most strongly-defined displays occurred 

 during periods of two, three, four, or five consecutive days 

 of remarkable auroral activity. As examples, may be men- 

 tioned the bright aurorae of the 1st, 2nd, 3rd, 4th r and 5th 

 of June, the 19th, 20th, 21st, and 22nd of August, the 18th. 

 19th, 20th, and 21st June, the 22nd, 23rd, and 24th of May, 

 and the 20th, 21st, and 22nd of April. 



Furthermore, not only was there a bunching of the days 

 on which the more notable auroral effects were observed, but 

 on each day at a corresponding time manifestations peculiar 

 to the period transpired. For example, during the after- 

 noons of the 2nd, 3rd, 4th, and 5th of June bands ran out 

 across the zenith in a nearly E. and W. direction; at other 

 times such bands were quite a rare occurrence, so that their 

 repeated appearance at this time must have special signi- 

 ficance. On the three successive nights of the 19th, 20th, and 

 21st of June striking circular and elliptical aurorae took place 

 in the eastern sky. Again, on the 19th, 21st, and 22nd of 

 August, auroral curtains formed in sweeping curves, whilst 

 at other times straight or but slightly-bowed bands and 

 curtains were the general rule at Cape Royds. 



Although undoubtedly there are to be expected several 

 agents co-operating to produce auroral phenomena, it is 

 undoubtedly owing to a single factor that such periods of 

 auroral maxima are determined. Assuming that aurorae are 

 directly produced by the radiant energy proceeding from a 

 sun-spot, such a succession of auroral manifestations would 

 be expected during the few days that the portion of the sun's 

 surface containing the particular spot is turned towards the 



