THE INTEKIOR OP THE EARTH. 249 



kind of rotation. The spinner imparts to the top a motion of trans- 

 hition as well as a rotation, and if we wish to study the rotation we 

 must arrest the translation in some way. This we can do by letting 

 the top fall upon a hard surface, in which the iron peg soon wears a 

 minute hole for itself, and the effect is to stop the translation of the 

 top without modifying seriously the rotation. Then we can see that, 

 while the top is turning very rapidly around an axis, this axis is 

 itself rotating in a comparatively leisurely way. Just the same thing 

 is occurring with the earth — the point (or pole) at which the axis of 

 the daily rotation pierces the surface of the earth is continually in 

 motion. If we could take to the neighborhood of the pole a modern 

 instrument and if we could observe there at leisure and in comfort, 

 we should have no particular difficulty in finding the position of the 

 pole within a meter. But if we should repeat these observations a 

 few months later, we should find that the pole had wandered away to 

 some distance. To be sure, this distance would not be great, and all 

 the wanderings of the pole that have thus far been observed could 

 be plotted to true scale on the floor of a room not much larger than 

 the one we are in. Of course, if the pole is moving, so, too, is the 

 earth's Equator ; and thus the latitudes of all points on the earth are 

 varying. Such wanderings as these need not disturb the peace of 

 mind of those gentlemen who like to discover the Arctic or the Ant- 

 arctic Pole. Under the circumstances that the polar explorer must 

 work and with the meager instruments he can transport, he is glad 

 to determine his latitude within half a mile of the truth. 



We must understand that it is only in our time, and only after the 

 lapse of many years since Euler published his memoir, that latitude 

 variations have actually been observed. There w^as nothing in 

 Euler's theory to indicate how large a variation to look for, since 

 this is a matter that depends upon the whole complex of " initial 

 conditions," of which our knowledge is the very vaguest. But this 

 theory does tell us what the period of variation should be, since 

 this depends upon the shape of the earth and the distribution of the 

 material within it, and precisely the information that is here needed 

 is afforded by a study of precession. Applying this information, 

 Euler was able to say that the period of the latitude variation should 

 be 10 months. Bessel at Konigsberg, in 1842, later Peters at Pul- 

 kova, Nyren also at Pulkova, Downing at Greenwich, and Newcomb 

 at Washington, all searched their observations for evidence of a 

 latitude variation having a period of 10 months, but all in vain. 

 Astronomers concluded that if latitude variations existed at all, their 

 extent was too small to be detected by instruments of the precision 

 that had then been attained. 



Toward the end of the nineteenth century vague whisperings that 

 this conclusion might be ineorrect seem to have been in the air. 



