THE EARTH A MAGNET. 19 



careful comparison of Beaufoy's observations, that the 

 needle reached the limit of its western digression (at 

 Greenwich) in March 1819, at which time the declina- 

 tion was very nearly 25. In Paris, on the contrary, 

 the needle had reached its greatest western digression 

 (about 22^) in 1814. It is rather singular that 

 although at Paris the retrograde motion thus presented 

 itself five years earlier than in London, the needle 

 pointed due north at Paris six years later than in 

 London, viz., in 1663. Perhaps the greater amplitude 

 of the needle's London digression may explain this 

 peculiarity. 



' It was already sufficiently difficult,' says Arago, 

 * to imagine what could be the kind of change in the 

 constitution of the globe which could act during one 

 hundred and fifty-three years in gradually transferring 

 the direction of the magnetic needle from due north to 

 23 west of north. We see that it is now necessary 

 to explain, moreover, how it has happened that this 

 gradual change has ceased, and has given place to a 

 return towards the preceding' state of the globe.' * How 

 is it,' he pertinently asks, ' that the directive action of 

 the globe, which clearly must result from the action of 

 molecules of which the globe is composed, can be thus 

 variable, while the number, position, and temperature 

 of these molecules, and, as far as we know, all their 

 other physical properties, remain constant ? ' 



But we have considered only a single region of the 

 earth's surface. Arago's opinion will seem still juster 

 when we examine the change which has taken place 



c 2 



