OF THE EARTH’S MAGNETISM. 559 
have been obtained by interpolation from the projected four weekly means; and 
the series before and after February 1844 were connected by a mean for February 
obtained from the projected four weekly means, reference being made to the 
Cape curve, which the St Helena curve resembled at this epoch, and to the 
facts that there was a diminution of force till January, whilst after March the 
force increased at St Helena. 
120. Hoparton,* 1842, April—The last monthly adjustment was made 
April 1842. The monthly means for January, February, and March 1842 are 
interpolated from those at Makerstoun. 
121. 1843, June.—No observations were made during the first half of the 
month on account of repairs in the Observatory. The correction of the mean of 
the last thirteen days to the whole month was found as follows :— 
By Makerstoun bifilar mean (June) minus (June 16 to 80)= — 0°86. 
The correction including four other stations, Trevandrum, Singapore, Cape, and 
St Helena, was found =—1:52; but the similarity of movements at Hobarton 
and Makerstoun gives the result from the latter alone much the greater weight. 
— 0:86 was employed as the correction. 
122. 1843, July.— After the adjustment at this time, the bifilar readings in- 
creased with much rapidity, showing some cause of error similar to that at 
Toronto after February 1843. A careful examination of the means has induced 
me to reject all the monthly means from July 1843 to March 1844, both inclu- 
sive; for the purposes of comparison, the Makerstoun monthly mean changes 
have been substituted for these months in Table VI. The daily means also 
for January, February, and to March 14, 1844, have been corrected, upon the 
assumption of a regular accidental change, by —0:15 (1=0 00014) counting back- 
wards from March 15 to January 1. (See Table XXXIIT.) 
123. 1846, March 31.—The arms of the torsion circle were turned through a 
small angle (1° 27’), to diminish the scale readings, which had increased consider- 
ably since the previous adjustment. By comparative readings with a small 
bifilar before and after moving the arms of the torsion circle, the alteration 
caused in the scale readings was found —75°5.; The first observation noted after 
turning the arms was 45-0, at 1846, April 1° 0° Gott. (temp. 54°-0), which, 
according to the comparison with the small bifilar, may be considered equivalent 
to about 120-5 before turning the arms; hence, to convert the series before turn- 
ing the circle-arms into ten-thousandths of X, we have the formula 
(N—120-5) 0-000229 ; 
* Tat. 42° 52”5 S., Long. 95 49™ 50s H. Height above the sea, 105 feet. Hobarton Observa- 
tions, vols. i., ii., and iii., printed under the superintendence of Lieutenant-Colonel E, Saninz. The 
temperature coefficient employed for 1842-43 was g’=1- 50 sc. div., instead of the more accurate 
value 1-52 (see Trans. Royal Soc, Ed., vol. xxii. p. 481). 
} Hobarton Observations, vol. i., Adjustments, &c., p. xliii. 
VOL. XXII. PART III. 7B 
