338 



Results of Makeestoun Observations, 1844. 



TABLE VI. — Diurnal Variations of Westerly Declination for each Month in 1844. 



The approximate times of the secondary maximum ( + ) and ma.ximum ( — ) are as follow ; the hours in 

 parentheses are times of inflexions, which do not attain the character of maxim a or minima. 



Jan. Feb. March. April. 



+ (3''A.M.)? 2" 3"? 2"? 



-(S'-A.M.)? 41" 'liy\ 7^-" A.M. 



May. June. July. 



3" A.M. (12'>P.M.)? (I'-A.M.)? 

 12'' P.M. (6''P.M.)^ (7" A.M.)? 



Aug. Sept. Oct. Nov. Dec. 



12"? P.M. 121'] P.M. 6''A.M. 2" (3''A.M.)? 



9"? P.M. 4"? P.M. S^-A-M. S" (7''A.M.)1 



The irregularities in the means for a single month, are too considerable to render approximations to the 

 epochs of the secondai-y points of much value ; on the whole, however, it may be stated, that in the 8 months 

 which have the principal minimum before midnight, a secondary minimum occurred from 5*" to S" a.m., or 

 within a few hours of the eppch of the principal minimum of the remaining 4 months ; and in these the 

 secondary minimum perhaps occun-ed between B^and 12" p.m. The secondary maximum (or inflexion) occurred 

 between midnight and 3 a.m. in each month excepting October. 



Strictly speaking, only one maximum and one minimum are marked distinctly in the solstitial months ; 

 but although the secondai-y points are not absolutely exhibited, there are points of inflexion, in the projected 

 means, near the corresponding times ; there are periods when the velocity of change is zero, or has a minimum 

 value, though it does not change sign. These seem equally results of the same physical cause, and are 

 probably only less distinct from the superposition of other motions. 



It is evident that a single year's observations, combined in the usual way, are insufficient for the 

 determination of the true law of diurnal variation in each month of the year. Tlus insufiiciency is chiefly due 

 to the eflect of intermittent disturbances, which destroy the regularity of the continuous diurnal disturbance. 

 From results which follow, with reference to intermittent disturbances, it will be seen not only that these are 

 in excess at particular hours, but that they affect a particular sign, the mean disturbance deduced from any 

 number of observations at a given hour being a positive or negative quantity. It is certain, therefore, that no 

 accumulation of observations, combined in the usual way, can exhibit the law of the simple diurnal variation, 

 and it becomes a matter of importance to endeavour to deduce this law at once from the series we possess. 



A considerable acquaintance with the motions of the magnets, and a careful examination of the 

 observations, have equally led me to separate the diurnal motions into three classes — 1st, On particular 



