52 



As the punctum of the observation is at the commencement of the 

 break, and that may occur in the midst of the little breaks between 

 the seconds, and thus involve a small fraction of error, I have in^ 

 vented a key which will cause a measured break, say |- second long, 

 when either end of the break will be available, and if one is inde- 

 finite, the other will be measurable by scale and dividers. Bain, 

 Steinheil, and Wheatstone, have invented telegraphic clocks ; but, so 

 far as I can learn, their object has been to make one clock drive 

 others at the same rate on various points of the circuit.* The me- 

 chanical action interferes, too, with the pendulum. The invention 

 of a clock, having nothing to disturb or obstruct the pendulum, or 

 otherwise change its rate, and which shall register currently its time, 

 and enter truly the observations in their exact place, in reference to 

 time, is, so far as I know, new. The plan is certainly original 

 with me, whether it be new or not. Atmospheric disturbance of the 

 electrical current may, for a while, interfere with the action of my 

 clock, but when that ceases, the clock will always strike in correctly 

 and will recommence to register its minutes and seconds at their pro- 

 per moment, according to the motions of the parent instrument. Not 

 so with Bain's, which moves other clocks, for if the electricity be in- 

 terrupted, the secondary clocks fall behind, and remain so. 



Seconds ; 



Seconds, and minute zero: 

 Five minute signal : 



Minute zero. 



Five minute signal. 



Hour signal ; 



Tliree Seconds. 



Minute zero. 

 Hour sienal. 



Four seconds. Hour commences. 



Observations : Obs. Obs. 



T have communicated my invention, step by step, to Mr. Bache, 

 who has written me a letter approbating the affair, though he says 

 he does not know whether I am anticipated by any one. The use of 

 this machine in determining longitude is obvious ; for the interval on 

 the fillet, between the breaks for a star transit at one meridian and 



* It is easy to make a pendulum clock move another at a distance, by elec- 

 tro-magnetism, so long as the current passes well, but any atmospheric dis- 

 turbance will put it out of time. 



