460 



respect to a certain line or equator on the earth's surface, which can- 

 not yet be precisely traced for want of sufficiently numerous stations 

 (but which seems to approach to the line of least intensity, and is 

 very far from coinciding with the geographical equator), and in the 

 next, and for its other influential cause, to the fact of the sun's 

 having north or south declination ; so that the whole diurnal change 

 in any one of the elements, and at any station, is made up of two 

 portions, one of which retains the same sign, and a constant co- 

 efficient all the year round ; the other changes sign, and varies in 

 the value of its coefficient with the annual movement of the sun 

 from one side of the equator to the other. 



That, consequently, for a station on the magnetic equator (so de- 

 fined), the mean amount of diurnal change is nil, when taken over 

 the whole year, but that on any particular day in the year it has a 

 determinate magnitude, which passes through an annual periodicity, 

 with opposite characters in opposite seasons. And that for a station 

 in middle latitudes the mean diurnal fluctuation is not nil, but such 

 as during every part of the year to exhibit an easterly deviation in 

 the morning hours, and a westerly in the evening hours, for stations 

 north of the magnetic equator, and vice versa for those south of it ; 

 but that the amount of this deviation, or the amplitude of the diurnal 

 fluctuation, varies with the seasons, being exaggerated or partially 

 counteracted by the alternate conspiring and opposing influence of 

 the sun's declination during the summer and winter seasons. 



As regards the irregular disturbances, though arbitrary and capri- 

 cious in extent and in the moments when they may be expected, in- 

 dividually, they nevertheless obey, with great fidelity, the law of 

 averages when grouped in masses, and treated separately from those 

 of the former class. So handled they are found to conform in their 

 average effect, at each of the twenty-four hours of the day, and on 

 each day of the year, to the very same rules as regards the sun's 

 daily and annual movement, with one remarkable point of difference, 

 viz. that their hours of maxima and minima are not identical with 

 those of the regular class, but that each particular station has, in 

 this respect, its own peculiar hours, analogous to what is called the 

 " establishment" of a port in the theory of the tides. And that in 

 consequence, the superposition of these two systems of diurnal fluc- 

 tuation gives rise to a series of compound variations analogous to the 



