382 POPULAR LECTURES AND ADDRESSES. 



the suspended watch presented, quite independently 

 of the swinging due to the rolling of the ship. 



When the top points of support are arranged 

 precisely to the critical condition, I find that the 

 \vatch will, of itself, take sometimes one mode of 

 vibration, sometimes the other. But a very slight 

 deviation in either direction from the critical 

 arrangement suffices to do away with this indif- 

 ference, and to insure that, when the watch is 

 steady and left to itself, it will take up either 

 always the gaining or always the losing mode of 

 vibration. But even then it may be compelled to 

 take up cither mode by properly-timed touches 

 with the finger, and it continues vibrating accord- 

 ingly when left to itself. Thus, when the top 

 points of support are adjusted, cither precisely, 

 or somewhat approximately, to the critical con- 

 dition, the watch may be made to go either faster 

 or slower than its proper rate, by applying the 

 hand to cause it to take up either mode of 

 vibration at pleasure, and then leaving it to itself. 

 This last experiment ought not, however, to be 

 pushed too far with a valuable chronometer, as the 



