INFLUENCE OF SUSPENSION ON WATCH. 373 



ment wheel would be that represented by Figure 2, 

 and the clock would go faster than if its frame 

 were perfectly fixed. In Case II., the vibrations 

 maintained by the escapement would be those re- 

 presented by Fig. 3, and the clock would go some- 

 what slower than its proper rate. Case I. could 

 never occur in practice, but may be experimentally 

 illustrated by hanging the works of a clock on a 

 light stiff frame, movable round a horizontal axis. 

 Case II., Fig. 3, with CP' much shorter in propor- 

 tion to P' P than shown in the diagram, represents 

 the actual circumstances of an ordinary pendulum 

 clock, which, owing to want of perfect rigidity of 

 the frame, must experience a little of the influence 

 of the pendulum in the manner there illustrated, 

 causing the rate of the clock to be somewhat 

 slower than it would be if the support of the pen- 

 dulum were absolutely fixed. The clock cases of 

 the best astronomical clocks do not seem well 

 adapted to give the steadiness necessary for good 

 results ; and it is wonderful that their performances 

 are not worse than they are found to be. The 

 pendulum ought to be hung from a massive stone 



