136 ABSTRACTS FOR THE MAGNETIC DECLINATION. 
TABLE I.—The Monthly Means of the Declinometer Readings in Westerly Declination, for the Hours 20, 23, 
2, and 5. 
J Mean of 
Month. 204. 7 . 4 poten atets 
1841. <2 25°. 
July 29.49 
August 35-88 
September 32-06 
October 30-22 
November 31-82 
1842. 
January 
February 
March 
April 
August 
September 
October 
November 
December 
The mean of the Declinations at 20" and 5" have been given as an approximate mean for the month. The 
difference of the Declinations at 20" and 2? is given as the approximate diurnal range. The latter must be very 
near the truth, as 20" and 2" are nearly the hours of least and greatest Westerly Declination throughout the year. 
It seems very probable, that the Declination in the months of July and August 1841, and of April 1842, are 
considerably affected by the torsion of the suspension thread. While this would destroy the value of the mean for 
the month, the diurnal range may be considered as unaffected. 
In order to shew the regular changes well, it is advisable to eliminate, as far as possible, those days of irre- 
gularity or disturbance in which the positions of the magnet differ considerably from the mean. The following 
rule was adopted for the determination of the days of irregularity which should be rejected. 
If a, b, and ¢, be the approximate diurnal ranges for three successive months, those days in the middle month 
a+2b+e 
3 
For the first and last months, the sum 6+ ¢ or a+ has been used. (See Table XXIII.) 
This rule was formed chiefly for the two components of force, for which it is much better suited than for the 
Declination. 
Table II. has been formed after rejecting the days of disturbance thus determined. 
are rejected in which the diurnal range exceeds 
