274 On various phcenomena of Rotating Bodies. 



that the iucipieut rotation round the vertical axis be not checked, 

 otherwise the weighted side of the apparatus falls immediately. 



When this side falls as low as it can, that is, until one of the 

 protuberances x or // strikes against the piece pu, the rotation 

 round the vertical axis ceases, because the rotation round the 

 horizontal axis is checked. If, however, the weighted side be 

 raised witli the hand, the rotation commences again round the 

 vertical axis. In this case a peculiar motion generally sets in. 

 While the ajjparatus rotates round the vertical axis, the weighted 

 side moves up and down like a pendidum. 



If instead of lifting this side the rod tu be carried round with 

 the hand, and the entire apparatus thus turned round its vertical 

 axis with a somewhat greater velocity than it would have turned 

 of itself if the motion round the horizontal axis had not been 

 checked, then the weighted side rises of itself, and the rotation 

 round the vertical axis continues of itself, provided that the discs 

 AB and CD rotate with sufficient velocity. 



When the two discs AB and CD are impressed with equal 

 velocities but in opposite directions, then the apparatus, although 

 its motions are perfectly unhindered, is easily moveable, and the 

 smallest surplus weight at vi or n causes the overweighted side 

 to sink immediately. If one disc only rotated, the entire appa- 

 ratus would turn round the vertical axis vw in an opposite direc- 

 tion to that of the motion M'hcn both the discs rotate. Hence it 

 is solicited in two opposite directions round this axis, and as the 

 rotating forces are equal, they neutralize each other. In this 

 way the two rotations round the vertical axis are checked. Hence 

 the same consequence follows as if only a single disc rotated, or 

 as if both moved in the same direction, the rod tu being held 

 fast by the hand. 



The whole of these experiments may be made without the use 

 of weights. When the screw e is loosened, the rod mn may be 

 moved back and forward in the collar os, so that the apparatus 

 may assume the position given to it in fig 1. The weight of 

 the disc CD acts then at the end of a longer arm than that of 

 the other, AB, and in this way the same effect is obtained as if 

 weights had been placed at n. The greater the difference of the 

 two arms of the lever, the more quickly does the apparatus 

 rotate. 



By moving mn in the collar os, it is possible to make another 

 experiment. If the distances of the rotating discs from the 

 vertical axis ow be unequal, and the two discs be set in rotation 

 with as nearly the same velocity as possible in opposite directions, 

 then the apparatus is perfectly mobile. In like manner, when 

 the shorter arm is so weighted as to establish an equilibrium, 

 then the slightest weight, which while the discs are at rest is 



