Some evidence can, however, be offered on the subject of the tendency to repetition 

 of conditions at the same hour on successive days, this result being largely possible 

 on account of the relatively favourable weather conditions at this station. As an 

 example, we may take the observations at 2 hours between the 26th May and the 

 31st July, at which hour conditions were favourable for observation on 45 occasions. 

 Thirty-two times out of this possible 45 aurora was seen, a " nil " result being recorded 

 on the remaining 13 occasions. During this period of 67 days, the following sequences 

 of aurora on successive days were observed : 



On 1 occasion, 6 days in succession. 

 l ^ 



55 55 " 55 5) 



55 ** 55 ^ 55 55 



9 -7 



}} " 55 55 55 



A " nil " result was recorded during one period of four days, and on two occasions 

 for two successive days. 



If the occurrence of an aurora was subject only to the laws of probability, we 

 would estimate that the chance of the occurrence of an aurora on any one day was 

 32/45, and the chance of the occurrence of an aurora on six successive days was (32/45)", 

 while the chance of the occurrence of four successive " nils " would be (13/45) 4 . 

 Considering further that, during the period of 67 days, a sequence of six was only 

 possible on four occasions owing to unfavourable weather conditions, it seems very 

 clear that there must be a strong tendency for the same conditions to recur at the 

 same time on successive days. 



Not only is this the case, but a scrutiny of the Cape Adare log shows most 

 definitely that there is a marked tendency for the aurora to recur in similar form at 

 the same time on successive days. So pronounced is this tendency that it would be 

 possible in the majority of cases to assign an approximate time for the observation, 

 from the sketch alone. As at Cape Evans, the tendency is for the aurora to originate 

 in a glow low down on the horizon (at a distance), the aurora gradually approaching 

 the station and sometimes passing overhead. 



A very fine example of the recurrence of aurora in the same form at the same 

 hour on successive days is furnished by the observations and sketches at 8 hours 

 between June 7th and June 15th. During this period, an aurora of very distinctive 

 form was often seen at 8 hours (never at other times), which took the form of an arch 

 running from the horizon in the south-eastern quadrant up to the zenith. At the 

 zenith the arch then broke into three or four short branches in the north-western 

 quadrant. A description of the form of the aurora during this period is given below, 

 the observations being made at 8 hours precisely : 



At 8 hours on the 7th, the straight portion of the arch ran from E.S.E. to 

 the zenith, thence branching into three short portions in the N.W. quadrant. 



22 



