THE TRANSATLANTIC LONGITUDE. 101 



the opportunity of adding some few facts to those heretofore established seemed 

 worth improving, although obtained with no special apparatus, and entirely col- 

 lateral and subordinate to the astronomical purposes of the expedition. And further- 

 more, the question has an especial interest for me, as having been among the first 

 to demonstrate and measure nearly twenty years ago the transmission-time of the 

 galvanic signals, which had previously been assumed to be instantaneous. The 

 duration of our signal-currents was intended to be uniformly one-quarter of a 

 second, but depended upon the skill and care of the observer, no automatic signal- 

 giver having been employed. Every electrician knows how greatly the strength 

 of the current is augmented by an increase of its duration from 0\2 to 0\3; yet the 

 duration of the signals varied frequently through a larger range than this. StiU 

 the actual length of each signal is recorded upon the chronograph-register, and its 

 average did not vary much from the prescribed duration of 0'.25. 



It appears manifest that not an electrical charge or discharge, but simply an 

 electrical disturbance, is requisite for transmitting a signal; that an inductive 

 impulse, sufficient to deflect the galvanometers employed, was transmitted through 

 one cable, having at each end a condenser with 10 cells, in somewhat less than the 

 third of a second, five seconds after the transmission of an impulse of the opposite 

 sort ; that with a circuit formed by the two cables, a smaller electromotive force 

 sufiiced to transmit the signals with yet greater rapidity ; that the signals travelled 

 more rapidly through a cable which had not recovered its electrical equilibrium 

 after a current of the opposite character ; and that the speed of the signals is modified 

 by the earth-connections, more readily than by changes^ in the battery-power. And 

 the very marked differences, found in the rates of transmission, between signals 

 given by completing an interrupted circuit and those given by interrupting a closed 

 circuit, may perhaps lead to investigations which will afford an explanation. 



' Jenkin (Phil. Trans. CLII, 982) arrived at the conclusion that the electromotive force of the 

 battery has no appreciable effect on the velocity with which the current is transmitted. But he 

 would doubtless consider that some qual ideations to the general statement should be taken for granted. 



