Mr AIRY ON A NEW CONSTRUCTION OF THE GOING-FUSEE. 219 



a force in the axis of the barrel, which, by a proper application of bell- 

 cranks, could easily be effected by a weight. But the construction pro- 

 duced a trifling friction in the ordinary going of the clock, and was not 

 elegant, and I therefore abandoned it. 



Finally, I had the good fortune to imagine a construction entirely dif- 

 ferent, with which, in all respects, I am fully satisfied. It is based upon 

 the following principle. If a lever abc, fig. 1, is used to produce pressure 

 at the point c, b being its fulcrum, and a the point at which the force is 

 applied : then the same effect may be produced on c, by making a the 

 fulcrum, provided that at b we apply a force exactly equal and opposite to 

 the pressure which the fulcrum at b sustained when the force was applied 

 at a. 



To apply this to the first wheel of a clock. Suppose (for simplicity of 

 present consideration and of future construction) the axes of the first wheel 

 and second wheel to be in the same horizontal plane. Let a fig. 2. be the 

 point of the barrel from which the weight depends : b its center, c the 

 point at which the toothed wheel, connected with the barrel, acts upon the 

 pinion of the next wheel. Then, during the ordinary action of the clock, 

 abc may be considered as a lever, of which b is the fulcrum, sustaining a 

 downwards pressure, a the point of application of the force, and c the point 

 on which it is to produce an effect. Suppose, in the operation of winding 

 up, the force acting at a to be removed. Then, by the theorem which I 

 have lately mentioned, the same effect may still be produced on c; pro- 

 vided that we can so arrange our mechanism that a shall, during the 

 operation of winding up, become the fulcrum ; and that a force shall act 

 upwards at b, exactly equal to the downwards pressure which b sustained 

 during the clock's ordinary motion, the point b being not fixed (as before) 

 but moveable. The mechanism necessary for this purpose is extremely 

 simple. 



Instead of supposing the pivot b of the barrel to turn in a hole in the 

 clock-plate, let it turn in a hole in the arm of a frame, fig. 3, of which one 

 side is bad, and which is itself a lever whose fulcrum projected in a is the 

 line joining two pins turning in holes of the clock-plate, corresponding 



E E2 



