230 MAGNUS ON THE DEVIATION OF PROJECTILES, 



is fixed, which turns in two holes, C and D, of the hoop CDG. 

 AB is moveable in this hoop like the beam of a common balance. 

 The hoop itself is suspended in G from a string, around which 

 the whole apparatus easily revolves. And hence it is possible to 

 give the axis AB any required position. 



If one of the two discs, E or F, be set in rotation, the axis 

 AB remains in the position first given to it, and a somewhat con- 

 siderable force is requisite to change the same. This difficulty 

 is much increased when both discs rotate in the same direction. 

 If the hoop CDG, however, be held firm with the hand, the 

 axis is immediately and with facility moveable ; it is indeed 

 surprising how this moveability changes, according as the hoop 

 is held firm or left free. These phaenomena are exactly similar 

 to those described in the apparatus with the three rings. 



If, however, one of the discs rotate to the right, the other to 

 the left — these directions having reference to an observer at one 

 extremity of the axis AB — then, even if the hoop CDG be left 

 free, the axis AB is easily moveable, at least as long as both 

 discs rotate with equal velocity. 



If they have unequal velocities it is more difficult to move the 

 axis, and the greater the difference of these velocities the greater 

 this difficulty becomes. 



If on a solid of revolution, whose axis is freely moveable 

 in all directions, a force acts during its rotation with great ve- 

 locity around its axis, and the direction of this force pass through 

 the axis but not through the body's centre of gravity, a lateral 

 motion will, as we know, take place ; but its velocity is always 

 very small, not only in comparison to the velocity of rotation, 

 but also to the motion which the same force would have im- 

 parted to the axis had not the body rotated. On this account 

 the axis appears to retain its position unchanged. 



The action of every force passing through the axis, but not 

 through the centre of gravity, causes a motion of the former, 

 which motion, however, is exceedingly small when the body 

 rotates quickly, and on that account the axis is apparently 

 fixed. 



The above-described experiment, with the apparatus shown in 

 fig. 6. Plate III., particularly illustrates the connection between 

 this fixity of the axis and its lateral motion : for it shows that 



