"معاملات الجمعية الجغرافية في بومباي. من يناير ١٨٦٣ حتى ديسمبر ١٨٦٤. (تحرير السكرتير.) المجلد XVII." [‎١‎٩‎٧] (٥٢٤/٣٦٨)

MAGNETIC INTENSITY OF THE EARTH.
197
thermometer placed three feet above the ground, whereas it is not im
probable that the variations of temperature of the mass of the shell of air
likely to act influentially (if it acted at all) would have been nearly as
well represented by the thermometer in the box as by that outside.
The approximate value for the temperature co-efficient of the bifilar
magnetometer for 184/ has been obtained by me approximately (1*58
scale divisions, about half of that employed in the Bombay volumes for
1847 and 1848)having corrected the Monthly Means (as obtained by me
from the table of Daily Means, p. 459, 1847) and allowed for an in
crease of reading at Bombay at the rate of 0*8 small scale divisions
monthly, I have found the following quantities, which are placed
alongside of corresponding means for the same year, deduced by me in a
similar manner from the bifilar observations at Hobarton, Van Die-
man’s Island : —
Variations of Bifilar Monthly Means.
1847 Hobarton. Bombay.
sc. div. sc. div.
January 7 29 6T3
February 3*95 3*2G
March 2-12 (P54
April 0-00 (POO
May 3*73 2-10
June 7*51 5*54
July 6-55 510
August 5-/9 4*91
September F02 2-12
October 2 , 19 DOl
November 3 91 4*03
December 4 51 2*92
These quantities (where the unit is one ten thousandth
of the horizontal force at the respective places) agree pretty well, consi
dering that the determination of the unit co-efficient depends at
both places upon the torsion circle, a method which I have found
to be frequently inaccurate, and that there may be errors at both places
which I have no means of correcting.
I have felt more reason for a careful examination of the erroneous
result as to the magnetic effect of the atmosphere, that I believe it in
some degree attributable to conclusions arrived at by mvself.