162 Dr. C. Chree : Auroral and 



11-year cycles, and the sun-spot data for the 33 years of many 

 and the 33 years of few sun-spots from these eleven cycles 

 are those given in the last two lines of Tables I. and II. 

 For the purposes of the present enquiry, these sun-spot data 

 answer sufficiently closely to either the Scandinavian or the 

 Lovering auroral data. 



The last two lines of Table V. give auroral data for New 

 York State, derived from 3 x 3 years of largest sun-spot 

 frequency and 2x3 years of least sun-spot frequency 

 included between 1828 and 1849. 



Each of the three pairs of comparative figures in Table V. 

 shows a decided rise in the spring maximum in the years of 

 few as compared to the years of many sun-spots. 



The average monthly frequency from May to August is 

 also less in the years of few sun-spots, very decidedly so in 

 the case of Lovering's data ; in the New York data the 

 depression of the midwinter minimum is the more decided. 

 On the whole, the phenomena resemble those already de- 

 scribed, i. e. the annual variation is accentuated in the years 

 of few sun-spots. The differences, however, between the 

 selected years of many and the selected years of few sun- 

 spots are less conspicuous than those between the 33-year 

 periods ; and such differences as exist may not unreasonably 

 be partly ascribed to the differences in the annual variations 

 of the sun-spot figures in the last two lines of Tables I. & II. 

 This explanation cannot, at the same time, go very far, in 

 view of the fact that there is no marked depression in the 

 auroral frequency in September in the groups of 30 years of 

 few sun-spots. 



The differences between the average number of auroras in 

 the year from the first two groups of contrasted years in 

 Table Y. are much less than the corresponding differences 

 in Table III., notwithstanding that the sun-spot differences 

 are larger in the case of Table Y. The ratio of the mean 

 frequency from the period or group of years of many sun- 

 spots to the corresponding frequency from the period or 

 group of years of few sun-spots takes the following approxi- 

 mate values in the several cases (p. 163). 



Obviously the conclusions one would draw as to the extent 

 of the influence of sun-spot on auroral frequency would vary 

 immensely according to one's method of attacking the 

 problem. 



§ 11. Table YI. deals with the same Scandinavian and 

 New York data as the previous tables, but instead of selecting 

 groups of 3 years of S max. and S min., it selected groups of 



