550 Royal Society : — 



as also a given time elapses after the battery has been reversed 

 at the one end before the current is reversed at the other or 

 distant end, it is clear that by gradually augmenting the rate of 

 rotation of the commutator until the wheel is a quarter of a revo- 

 lution in advance of the wave, a point is arrived at when the galva- 

 nometer's connexions are reversed precisely at the moment that the 

 wave reaches its maximum strength, and consequently the wave is bi- 

 sected, one half of it flowing through the galvanometer in one direction, 

 and the other half in the other. At this rate of rotation the galva- 

 nometer falls to zero ; because, the wave being exactly bisected, the 

 one half tends to deflect the needle to the right, and the other to the 

 left, but, owing to the weight of the needle and the rapidity of the 

 reversals, it (the needle) stands nearly steadily at zero. The galva- 

 nometer used consisted of a rather heavy astatic pair of needles 

 suspended by a silk fibre. The wire acting upon the needles was 

 about the twentieth part of an inch in thickness, in order that it 

 should offer no serious resistance to the electric current. Its re- 

 sistance was less than one Varley unit (1 mile copper wire -^ inch in 

 diameter). 



The rate of rotation necessary to obtain the first zero is the point 

 recommended for comparing the relative speeds of the electric waves 

 through submarine cables of different dimensions. 



By augmenting the speed beyond that necessary to produce the 

 first zero, the needle becomes deflected in the opposite direction and 

 gradually approaches a maximum ; that is to say, when the electric 

 wave is half a revolution behind, the currents all flow through the 

 galvanometer in one direction again. This is termed the second 

 maximum (the first maximum being that obtained when the wheel 

 is not rotating at all) ; and by augmenting the speed still more, until 

 the wave is three-quarters of a revolution behind, the wave is again 

 bisected and a second zero is obtained, and so on. 



The great variation of speed necessary to give these and other re- 

 sults was such that the means then at the author's disposal in the 

 first experiments were not sufficiently regular to admit of very ac- 

 curate readings. 



The experiments now communicated were made upon two cables, 

 one containing six conducting wires, a portion of which was laid in 

 the Mediterranean. This cable had been lying exposed to sun and 

 weather in the East India Docks for some years, and the gutta percha 

 had become deteriorated to a considerable extent ; its exact length 

 was not known ; and from these combined causes it could not be used 

 for determining the rate at which the wave travels through given 

 lengths, but it has served to demonstrate that Thomson's "law of the 

 squares" is substantially correct in practice. 



In the experiments made on this cable, the resistance of the galva- 

 nometer was equal to one mile of the cable. The battery power 

 used averaged from 12 to 36 cells of Daniell's battery, each cell 

 offering a resistance of one-sixth of a mile of the cable. 



The first experiment was made upon two wires forming a loop of 

 about 150 miles in length; and when the currents were reversed at 

 the rate of 15*16 per second, the needle came to zero. 



