354 



Results of Makerstoxjn Observations, 1844. 



TABLE XXI. — Number of differences in 1000 (without reference to sign), which occur between 

 the limits of successive minutes in each hour of 1844. 



Note on the D'mrnal Variation of the Magnetic Declination. — It is obvious from the previous investiga- 

 tions that the diumal variation is a compound phenomenon ; in its observed form it presents a curve with two 

 maxima and two minima, the principal ma.ximum occurring immediately after noon, the secondary maximum 

 shortly after midnight, and the principal minimum about 10'' p.m. or B*" a.m., at the former if the amount of 

 disturbance for the year be considerable. Wlien the diurnal variation is considered in days nearly free from 

 intermittent disturbance, it is found that the minimum about 10'' p.m. almost wholly disappears, and the mini- 

 mum about 6'' A.M. is increased. Although, however, the lO'' p.m. minimum and the secondary maximum 

 wholly disappear, the diumal curve, even when unafl'ected by disturbance, does not become a regular curve of 

 two branches ; on the contrai-y, there is still evidence of the action of a secondary or superposed cause of varia- 

 tion in the flattening of the curve from about ff' p.m. till .about 2'' a.m., in the sudden check in the rate of the 

 easterly progression about 6'' p.m., and its sudden increase again about 2'' a.m. These epochs, it will be 

 observed, are the times of the principal minimum and maximum for our periceci. We might suppose the diur- 

 nal variation, therefore, when unaffected by intermittent disturbances, due to two operations of the same cause, 

 a direct action proper to one side of a zone producing a regular motion having its greatest westerly limit after 

 noon, and its most easterly limit about C" a.m. ; and an indirect or reflected action upon the opposite side of 

 the same zone producing similar motions simultaneously, but of less magnitude. Such a hypothesis, which has 

 an analogy to that of tides (say in the atmosphere), would satisfy the simplest form of the diurnal curve for 

 the year, and for all the months for which the sun is north of the equator ; another element perhaps requires 

 consideration for the winter months. 



The form of the diumal curve is different at different seasons of the year. (Colonel Sabine has shewn, in 

 his discussion of the St Helena Observations, that the diumal curve has two opposite forms at St Helena, 

 according as the sun is to the north or south of the equator ; the one corresponding to the diurnal motion for 

 the nortliern hemisphere, the other to that for the southern hemisphere. An examination of the Makerstoun 

 Observations will shew an equivalent fact. During the months that the sun is north of the equator the form 

 and range of the diurnal curve have but little variation, but they change considerably in the months for which 

 the sun is south of the equator. If we take the means of the 10 days of least disturbance in the two months 



