152 AN EMPIRICAL STUDY OF GYRATING BODIES. 



wire an inch long or so in the small hole at the end of 

 the axle, and then to place over it a small loose ring 

 with eight or ten inches of string attached. By pulling 

 backward or forward, the same paradoxical movements 

 will be produced that were produced with the stick. 



ANOTHER PARADOX. 



If we wish to make our top gyrate more slowly we 

 must strike the axle from behind ; in other words, at- 

 tempt to push it along faster. It will show its contrari- 

 ness by going slower. To make it go faster we strike it 

 in front, as if we meant to stop it. 



The philosophy of this is equally simple. The gyra- 

 tion, as has been shown, increases with the load. The 

 nearer the top is to a vertical, the less is that part of the 

 load which tends to tilt it. This part, for distinction, 

 may be called the effective load. 1 I have shown that 

 striking the top from behind causes it to rise, and the 

 nearer the axis is to the vertical the less effective the 

 load becomes ; and, in consequence of that, the gyration 

 or precessional movement grows less rapid. 



TRACINGS MADE BY THE TOP. 



The paths described by the "points" are very curious 

 and beautiful. They may be preserved by using a large 

 pane of smoked glass for the points to rest on, and after- 

 wards, flowing the traces with demar varnish. 



No description can do justice to the exquisite beauty 

 of these paths. They are spiral, large or small, accord- 

 ing to the size of the " points" and the obliquity of the 

 axis. 



The path is a remarkably regular spiral, generally 

 largest at the beginning, but very gradually diminishing 

 to the end. In a few cases, the spirals remained of uni- 

 form size. These were very small — possibly one-eighth 



] 1 1 varies as the cosine of the obliquity. 



136 



