On the Speed of the Electric Wave through Submarine Cables, 549 



From the imponderable nature of electricity (considered for a mo- 

 ment as a fluid), from its incompressibility, and other circumstances, 

 the author infers that the electric current commences flowing out at 

 the one end at the very instant that contact is made with the battery 

 at the other end ; but it is a considerable time before it reaches ah 

 appreciable strength ; it then goes on augmenting in strength, ap- 

 proaching to, but never absolutely attaining, its maximum force. 



As the first part of the wave commences to appear instantly, and 

 as the top of the wave would require an indefinitely long period of 

 time to be reached, it will be evident that the part of the wave best 

 suited for investigation is half the maximum, as at that period the 

 changes of its intensity in a given time are more rapid than at any 

 other. 



When attempting to measure by means of a galvanometer the 

 arrival of the wave at half its maximum, the weight and momentum 

 of the magnet of the galvanometer were found to interfere so much 

 that, excepting through very long lengths of cable, nothing approach- 

 ing to an accurate determination of the speed could be obtained. 

 The use of electro-magnets was equally, if not more, objectionable, 

 as they require a very appreciable but uncertain time to be mag- 

 netized. 



The following method, however, of bisecting the electric wave has 

 obviated these difficulties, and admits of the determination of the 

 relative rates of transmission through cables of different lengths with 

 very great accuracy. 



The machine used consisted of an axle carrying two " commuta- 

 tors." This axle was driven by clockwork, governed, in one case, 

 by a fly rotating in mercury, and in the second experiment by 

 means of a fly in the air, together with a friction spring. 



The commutator consists of two wheels, each wheel being in two 

 halves. Upon the broad edge of the wheel rest two springs, one of 

 which is connected with one pole of the battery, and the other with 

 the other pole. The two halves of the wheel were constantly 

 connected, by means of two other springs, the one with the cable 

 wire, and the other with the earth, so that when the wheel was 

 turned round, during one half of the revolution a positive current 

 was flowing through the cable wire, and during the other a negative. 

 The other commutator on the same axle was precisely similar in 

 construction ; but the two springs resting on the edge of the wheel 

 were connected with two wires of a galvanometer, and one half of 

 the wheel was connected with the receiving end of the cable wire 

 tested, and the other half of the commutator was connected with 

 the earth. 



The two commutators were so arranged that when, by the rotation 

 of the wheel, the current of electricity from the battery was reversed, 

 the connexions of the galvanometer were reversed also ; and there- 

 fore, if the speed of the electric wave through the cable were inde- 

 finitely great, the currents would flow through the galvanometer in 

 one direction, no matter how fast the currents in the cable are re- 

 versed. As, however, a given amount of time elapses before these 

 waves reach their maximum at the distant end of the circuit, and 



