Magnetic Declination. 349 



When the diiferences are combined for periods of three months, we find the following approximate epochs 

 of maximum and minimum : — 



Nov., Dec, Jan. ; minimum, T"" a.m., Makerstouu mean time; maximum, 8^ p.m. 



C Value nearly con- 

 Feb.., Mar., April; between S*" a.m. and 4'' p.m. ; ll*" p.m. < stant between 5'' 



[ P.M. and l** A.M. 



May, June, July ; between o*" p.m. and G*" p.m ; 2'' a.m. 



Aug., Sept., Oct. ; l* p.m. ; Si" p.m. 



These indicate that the diurnal law of variation of disturbances varied in 1844, with season; this may 

 be due, however, to the intermittent character of the phenomenon, and the differences may disappear in a 

 greater number of observations. 



Annual Variation of the number of Positive Differences. — The following are the numbers of differences in 

 100 which are positive in each month of 1844 : — 



In order to obtain more certain mean values for each epoch, the mean for each three months is given 

 below as the mean for the middle month : — 



From which it appears, that the number of excursions of the north end of the declination magnet, to the west 

 of the monthly mean position in 1844, was a maximum in January and about .Tuly, and a minimum in April 

 and October ; or a ma.'cimum after the solstices, and a minimum after the equino.xes. The reverse, of course, 

 holds for the number of negative deviations. 



Diurnal Variation of the number of Positive Differences.— The numbers of positive differences in 100 for 

 each hour of Makerstoun mean time for 1844, are as foUow : — 



1211 lhi.M. 2l> 3h 4l> 5h ek ^h sh gh i|)h jjli ijtltp.M. 21> S^ *>• 51' 61' 71i 81' 9'' 101' n,, 

 58-6 58-3 55-7 53-2 50-0 44-0 36-0 39-2 41-3 38-9 37-9 41-7 S7-6 42-7 43-0 46-2 47-8 48-7 59-2 67-5 70-7 66-6 G7-8 66-2 



Whence the number of excursions to the west of the monthly mean position for each hour is a minimum about 

 O*" or 10*' A.M., and a maximum about S*" p.m. The reverse is the case for the number of excursions to the east 

 of the mean position, and the numbers of positive and negative differences are equal about 4'' a.m. and 5^ p.m. 

 This result is evidently related to that already obtained for the mean effect of disturbance. 



The previous conclusions have reference to the differences of the observations from their arithmetical 

 mean, wliich, it has been shewn, is in all probability not the normal mean ; we are therefore still ignorant of 

 the actual sums of the positive and of the negative excursions at different epochs. These differences might 

 be determined approximately from the mean effects of disturbance, and the total sums of diiferences already 

 found ; the error would be due solely to the observations occurring between the arithmetical mean and the 

 normal mean. The positive and negative sums of difference have, however, been determined, by employing the 

 normal mean (as deduced from the five days of least disturbance in each month) instead of the arithmetical 

 mean ; the differences thus obtained may be considered as due to disturbances of all classes, but chiefly to 

 intermittent disturbances. For this reason, the following results are distinguished from the former results, in 

 terming them means of disturbances, instead of means of differences. 



Annual Variation of the Mean Disturbance. — The following are the average disturbances for a single 

 observation of magnetic declination, in each month of 1 844 : — 



This gives the same law as that already found from the sums of differences ; the values are but slightly 

 different, being on the whole less for the disturbances or differences from the normal mean. 



The mean (hsturbance of a single observation in 1844, ......= 1''70 



MAG. and met. obs. 1844. 4 T 



