OR MOON AT THE EARTH'S SURFACE. .'ill 



magnet throughout the night are comparatively small, and do not 

 correspond, with change of sign only (as required by the foregoing 

 law), to those which take place at the homonymous hours of the day. 

 The phenomena of the lunar-diurnal variation are even more op- 

 posed to the foregoing law, the variation having two maxima and 

 two minima of nearly equal magnitude in the twenty-four lunar 

 hours, and its values at homonymous hours having for the most part 

 the same sign. Hence the phenomena of the diurnal variation are 

 not caused by the direct magnetic action of the sun and moon. 



It is true that if we proceed another step in the approximation, 

 nnd include in the values of the disturbing forces the terms con- 



b c 



turning the first powers of -, -, the former will produce in the re- 

 sulting values of AS, AJ7, and AF, terms containing sin 20, cos 20, 

 and giving rise therefore to a semidiurnal inequality. But the co- 

 efficient , by which these terms are multiplied, amounts in the 



case of the sun to -g-gVjr on ly> while in that of the moon it is about 

 yV ; and the magnitude of the semidiurnal inequality should bear 

 to that of the diurnal the ratios designated by these small fractions. 

 The facts are altogether opposed to this result. The coefficient of 

 the solar-diurnal inequality of the declination at Dublin, in the 

 mean of the entire year, is 3'- 52, while that of the semidiurnal is 

 2'-13, nearly two-thirds of the former. In the case of the lunar- 

 diurnal variation, the semidiurnal inequality exceeds the diurnal. 



