98 
DK C. CHKEE: SOME PHENOMENA OF SUNSPOTS AND OF 
largely influenced by something essentially local. The conclusion which I ultimately 
(jxpressed'^' was that “ whilst in the opinion of the writer further investigation is 
required to justify the flnal acceptance of any of Mr. MAUNDEif S views, his paper is a 
most important one.” 
Several other criticisms of Mr. Maunder’s work were published at the time. One 
which influenced me to some extent in undertaking the investigation now to be 
described, was made by Prof H. H*. TuRNERf on behalf of Prof. Schuster, in the 
discussion of Mr. Maunder’s paper before the Royal Astronomical Society. Prof. 
Schuster apparently considered the data to raise a presumption, rather than aftord a 
demonstration, of the existence of a period, and after applying his periodogram 
methods he seemed to think there was more to be said for a 13’64 than a 27'28-day 
period. As a flnal summary of his views it is stated {loc. cit., p. 84) :—■ 
“ We have in fact a choice between two interpretations— 
“ 1. Magnetic storms are apt to occur at times which, starting from a certain point, 
are multiples of 13'64 days. During some years the odd multiples and during other 
years the even intervals are principally concerned. 
“ 2. Magnetic storms often recur after several successive intervals which are equal 
to some lapse of time sufficiently near 27'28 days to fall within the limits of rotation 
of sunspot zones.” 
§ 21. If we look at Tables V. and VII. we see that the interval between the crests of 
the principal and secondary pulses is somewhere about 15 days, but the crests are so 
rounded that the interval might well be a day or more short of this. The fact reminded 
me of Prof Schuster’s remark about a 13'64-day period; it also suggested that we 
might have to do with the half of Dr. Schmidt’s period. It thus appeared desirable 
to ascertain deflnitely whether the magnetic data employed in the previous investiga¬ 
tions did or did not show a period. This was investigated in the following way :— 
The 5 days of largest H range in each month were selected, and the magnetic 
character figures put down in successive columns for each of these and the 35 subsequent 
days. The columns were numbered n to n-l-35, the column n including the 
representative days of largest H range. A year or two’s data sufficed to show that 
something was to emerge, so the enquiry was applied to the wliole 11 years, and the 
investigation was extended so as to include 5 days (columns n —5 to w—l) preceding 
the selected days in column n. 
The method may appear of somewhat a liybrid character. Naturally one would 
have preferred to base the enquiry entirely on the H ranges, or else entirely on the 
character figures. The olqection to the former course was that a range is a 3 or 4 
figure result. The research entailed dealing with 60 x 41 x 11, or over 27,000 entries, 
and having for these character figures instead of ranges meant a great economy of 
* Lor., cit., p. 14. 
t ‘ 01)servcatory,’Jiin. 1.3, 1905, p. 80, 
