TERRESTEIAL MAGNETISM AT KEW OBSERVATORY. 
85 
pulse, with its crest of course in column n. The differences between tlie figures in any 
corresponding pair of columns n — s and n + s are very small, but on the whole there is 
a tendency for the figure in column 7^ + s to be the larger of the two. From column 
w—15 to column n—9, and again from column n + 9 to column n + 15, the sun.spot 
figure is nearly constant. The pulse is obviously of a favourable type for the 
investigation of the presence or absence of a corresponding pulse in any magnetic 
quantity. 
The pulses in the case of all three seasons from the 11 years are fairly symmetrical 
with respect to column n for values of s up to 5 or more ; but for values of s in excess 
of 9 or 10 there is considerable a-symmetry. The figures show a marked tendency to 
fall from column n+11 to column w +15 in the case of equinox, and a similar tendency 
to rise in the case of summer. These phenomena are presumably “accidents,” and 
would disappear if a large number of 11-year periods were combined ; but they merit 
attention, in view of their possible Influence on the magnetic phenomena presently to 
be described. 
Whether the more rounded character of the pulse in winter is wholly accidental is 
more open to doubt. If a natural phenomenon, characteristic of the season, it would 
seem to imply some direct action of the earth on sunspots, and so far as I am aware 
astronomers have not succeeded in establishing any such action. 
The diflerence between the sunspot maximum and minimum years is marked, the 
pulse being much more rounded and less accentuated in the former. This may have 
been a peculiarity of the particular sunspot cycle, but it seems not unlikely to be 
fairly representative of sunspot maximum and minimum. Towards minimum there 
are usually a number of days without visible sunspots, and any finite number, however 
small, is infinite when compared to zero. 
Omitting the last line of Table IV., which represents the exact antithesis of the 
conditions represented by the first eight lines, it will be seen that the sunspot area 
even on the fourth day after the representative day of largest area is still some 30 per 
cent, above the mean for the period. It would thus be quite in harmony with the 
hypothesis of a direct instantaneous action of sunspots upon the earth, if the range 
of the magnetic elements were decidedly above the mean for at least 4 days after the 
day of largest area. But, on this hypothesis, the excess above the average range on 
the fourth day after the day of largest area should be only some 30 per cent, or less 
of the excess on the actual day of largest area. 
If the effect on the absolute magnetic range in individual days followed a similar 
law to that exhibited by the range of the mean diurnal inequality for the year, its 
absolute size would depend on the absolute difference in the sunspot area. For instance, 
if we compare the results in Table IV. for the groups of years of sunspot minimum 
and maximum, the total range of sunspot area in the former case—-being the excess at 
day n over day w-1-14—would be in Greenwich units 
(292-34) X 163/100 = 420, 
