There is, therefore, a real tendency for days of strong disturbance to be immediately 

 preceded and followed by days of strong disturbance. This analysis is not, however, 

 free from the disturbing influence of the moonlight in preventing auroree from being 

 visible. 



Notwithstanding this tendency for disturbed auroral conditions to continue for 

 a whole day and even for successive days, it is not at all infrequently found that, 

 even on a fairly disturbed day, " no aurora " is logged at one or two of the hours of 

 observation. This fact might, of course, be inferred from the preliminary statement 

 in this section regarding short period variations in brightness ; the remarkable fact 

 being that the negative results recorded at these hours are very often duplicated at 

 the same hours on the following day or even days. 



An example of this tendency to the recurrence of " no aurora " at the same hour 

 on successive days is given in Table 2. 



TABLE 2. Recurrence of Auroral Displays at the same hours on Successive Days, 



Cape Evans, 1911. 



Hours on which aurora was observed are marked 0. 



Hours on which aurora could not be observed on account of weather conditions are marked X. 



Hours on which weather conditions wen- favourable but no aurora was observed are not marked. 



