186 TENTH ANNUAL REPORT OP 



ring on tlae hook, taking care to keep the axis of the inner ring hor- 

 izontal. I now spin the spheroid as rapidly as possible, and then 

 Fig. 9.^ whirl the outer ring the same way till the cords are 



twisted so far as to tend strongly to untwist. Letting 

 go the ring, it commences whirling by the force of 

 torsion ; but suddenly the axis of the spheroid throws 

 itself into an oblique position, and instantly arrests 

 the motion of the ring, while the spheroid, with the 

 inner ring, slowly turns itself over.* As soon as it 

 is inverted, the cords untwist, and twist up in the op- 

 posite direction, the spheroid all the while maintaining 

 its own rotation the same way. When they begin the 

 second time to untwist, the spheroid authoratatively 

 interposes, and takes time to turn over quite leisurely^ 

 and get itself ready to whirl in the same direction 

 also. And thus will it operate a number of times be- 

 fore running down. 



This experiment does not need a separate explanation ; it is, in fact, 

 a repetition of the one in which I carried the frame round its vertical 

 axis. But it becomes more striking, for the reasons that the force is 

 more secretly applied by the cord than by the hands ; that it is ap- 

 plied uniformly as well as gently ; and tliat it is repeated as often as 

 the cord is twisted up. A short and thick rope of parallel pords is 

 purposely used, that the inversion may be repeated several times be- 

 fore the spheroid loses its velocity. You will observe, that the outer 

 ring does not move at all by the torsion of the cord, while the axis is 

 reversing itself; that force is wholly expended on the spheroid, com- 

 bining with its own rotation, to produce the inversion of its axis. 



To prepare the instrument for the second experiment, I replace it 

 in the frame, take the inner ring with the spheroid fromthe^ outer 

 ring, and attach to it, at one end of the spheroid's axis, this stiff rod 

 of brass, about six inches long. One end of the rod terminates in a 

 strong fork, which is slipped tightly upon the ring, and confined bv 

 pins. The other end is connected by a hook and swivel, with a wire 

 two feet long. I next remove the cord used in the preceding experi- 

 ment, and hang up, in its place, the wire with the spheroid attached 

 in the manner just described. Having put the spheroid into swift 

 revolution, I lift it up on one side by the ring, till the rod and axis 

 make a right-angle with the wire. Dropping it now from tliis posi- 

 tion, it does not fall, as one would expect, and hang beneath the 

 wire, nor does it even descend in the least, but commences a horizontal 

 revolution about the wire. The spheroid itself revolves vertically, but 

 the system horizontally. And the whole, weighing fifteen pounds, 

 and having its centre of gravity more than a foot from the support, 

 presents the magical appearance of being held up without force. If 

 I elevate it higher, at an acute angle with the wire, it will sustain 



•Figure 9. The arrows T show the direction of torhion. The particles A are moved up- 

 ward: by the rotation of the spheroid, and korizmtallji to the left by torsion— tlies-e forces 

 bcins indicated by the hrokcn-shafl arrows. The douhle-shaft arrow shows the direction of 

 the resultant, wliich corresponds to an elevation of the pole N, and a depression of iS, 



