OF THE EARTH’S MAGNETISM. 537 
or for the last term, putting the coefficient in jo999: we have :— 
DOGG OS), i wk nl) fae Ee: Ley ClO 
This coefficient agrees nearly with that for Toronto, the Cape, Hobarton, and 
Makerstoun, but the epochs of the maxima and minima differ twenty-three days 
from those for the last two places. 
61. We may place the results for the second term thus :— 
Epochs of 
Maximaand Minima 
PERC after the days of Solstices 
and Equinoxes. 
Bifilar observations, Makerstoun (1844-48), é , 298... 3 . 64 
se Hobarton (1844-48), Y : 254 . : 2/16 
Absolute intensity, Toronto (1845-52), : : 312. ; 23 
Bs The Cape (1847-49), ; : 2:84 . : . 15 
3 ses Hobarton (1846-50), 4 ‘ PD) 6 : 5 PY 
The agreement of these results shows that the assumption (46) was well founded ; 
and the results, from the observations of absolute horizontal intensity, confirm 
the annual law derived from the bifilar observations. 
62. In the previous investigations, we have considered the annual movement 
as represented by twelve monthly means, corresponding to the middle day of each 
month ; it must be obvious that we in this way fail to perceive all the minor 
variations of the monthly mean. In order to compare these variations at the 
four places, Makerstoun, Hobarton, Singapore, and Trevandrum, I have obtained 
the four weekly means corresponding to every day in the years 1844 and 1845, 
for these places. These four weekly means are projected in Plates XXV. and 
XXVI. 
63. In comparing these variations, we observe that even the minute changes 
are followed similarly at each place, but that they are combined at certain periods 
of the year with a greater continuous increase or diminution of force at one place 
than at another. This similarity of movement depends on the similarity of the 
variations of the daily means, which we shall consider immediately. 
64. If only one daily mean be much less at one place than at the others, there 
will be a sudden drop of the curve for four weeks, with as sudden a rise there- 
after as we see in Plate XX V., March 14th to L5th, and April 11th to 12th, 1844, 
at Makerstoun, due to the fall, March 29, 1844 (see Plate X XVII. of daily means) ; 
if, however, the force should increase suddenly and permanently at any epoch, 
the curve will show an increase continuously for the four weeks following this 
sudden change. Some such cause, real or accidental, produces the difference be- 
tween the curve for Makerstoun and for the other places in January 1845 (see 
Plate XX VI.) The further consideration of these variations I shall leave till 
another occasion. 
