OF THE EARTH’S MAGNETISM. 541 
October 5 to November 2; while the most marked minimum at all the places 
occurs near the summer solstice. ; 
75. It should be noted that the differences in the above table include the vari- 
ation due to the annual period and secular change. If we put the differences into 
four groups, each having an equinox or solstice as the middle point, and subtract 
the difference between the first and last daily mean of each group, from the sum 
for that group (in order to eliminate approximately the effect of secular change), 
we shall obtain the following quantities :— 
TABLE XXIII.—Sums or DIFFERENCES, WITH EFFECT OF SECULAR CHANGE ELIMINATED 
FOR EACH SEASON, 1844-45. (1:00=0:00014 X.) 












Makerstoun. Trevandrum. Singapore. Hobarton. 
PERIOD. 
Sum. Ratio. Sum. Ratio. Sum. Ratio. Sum. Ratio. 
Feb. 5 to May 365°92 1:40 | 260-69 1:05 | 223:04 | 1:19 | 233-47 | 1-27 
Aug. 6 ,, Nov 322°14 | 1°24 | 819-88 | 1:26 | 253°24 | 1:36 | 263-60 | 1-48 
Nov. 5 ,, Feb. 379:06 | 1:46 | 348°88 | 1-40 | 281:22 | 1°51 | 250-48 | 1:36 




dons 
May 7, Aug. 6,. .| 260:32 | 1:00 | 248-52 | 1:00 | 186-78 | 1-00 | 183-95 | 1-00 
_5, 
5, 

76. The number of years’ observations here included is too few for certain 
conclusions on small differences, but there can be little doubt these numbers 
show that, in both hemispheres, the variation of daily mean horizontal force is less 
in the quarter about the June solstice (when the sun is in apogee) than in any 
other quarter. 
77. The ratios of the sums for the summer solstice to those for the winter 
solstice differ little for all the four places; and if we add the sums for Maker- 
stoun to those for Trevandrum (which do not differ greatly), and those for 
Hobarton to the sums for Singapore (which also are nearly equal to each 
other), we shall find— 
Makerstoun plus Trevandrum, June solstice : December solstice = 1-000: 1-43] 
Hobarton plus Singapore, June solstice : December solstice = 1-000 : 1-434 
78. If we compare the amounts in each quarter at Hobarton with those at 
Makerstoun, we obtain the following results :-— 
March equinox, Hobarton : Makerston=1 : 1:57; 
June solstice, 58 if; =A 
September equinox, ,, me a EF 
December solstice, _,, i —=11 5 1dsile 
or the excess of the movements in the northern hemisphere over those in the 
