542 MR JOHN ALLAN BROUN ON THE HORIZONTAL FORCE 
southern hemisphere was least at the September equinox and greatest at the 
March equinox (in 1844-45). 
79. If we subtract from the first group in Table XXIII. the differences due to 
March 29, 1844, and April 13, 1845, and from the last group the differences due 
to January 9, 1845, three days of marked disturbance that we have subtracted 
from the sums in Table XXI., we shall have— 
TABLE XXIV.—Svms From TABLE XXIII., omittinc Taree Days or LARGE 
DISTURBANCE, WITH RATIOS TO HOBARTON. 






Makerstoun. Trevandrum, Singapore. Hobarton. 
PERIOD. 




Ratio. 






























Sums, Sums. Ratio. Sums. Ratio. Sums. Ratio. 
Feb. 5 to May 7,. 982-75 | 1-38 | 246-13 | 1:16 | 207-90 | 0:98 | 212-41 | 1-00 
May 7 ,, Aug. 6, . 960:32 | 1:41 | 248:52 | 1:35 | 186-78 | 1:02 | 183°95 | 1-00 
Aug. 6 ,, Nov. 5, . 322:14 | 1-22) 319-88 | 1-21 || 253°24 | 0:96 | 263260 E06 
Nov. 5 ,, Feb. 5,. 3840:05 1:42',|°3385:29 | 1:38 | 270-80) | 1:12 5249-00 TiiEoG 


The ratios of Hobarton to Makerstoun are more nearly constant here than in (78). 
In all cases we have to remark the nearness to equality in the amount of move- 
ments at Trevandrum and Makerstoun, as well as in those at Hobarton and 
Singapore. 
80. From all the sums we arrive at the following important result :—That 
the amount of variations of daily mean horizontal force does not depend (or 
depends but little) on season in the meteorological sense, the amounts of move- 
ment being least in both hemispheres in the quarter when the sun is in apogee— 
that is, during the northern summer and southern winter. More extensive series _ 
will prove that the amount is greatest also in the equinoctial quarters, though — 
the difference betwixt the amounts for these and the December solstice is not 
considerable. 
81. If we now examine Plate XXVII. again, we shall perceive a series of 
movements having a period of about 26 to 30 days (66, 2d.), which are, how- 
ever, only marked in certain seasons. The months of 1844-45 in which this 
period is seen are January to April and August to December,—perhaps least 
distinctly in December and January ; it is not visible in May, June, or July. ~ 
82. When we attempt to determine the duration of this period from the pro- 
jected results for all the places, we find the time from minimum to minimum % 
approximately as follows:— 

