THE TRANSATLANTIC LONGITUDE. 

 Nov. 10. Signals FROM Newfoundland; Battery at Valencia. 



95 



Exper't. 



Earth con- 

 nection. 



5' interval. 

 Makes. Breaks. 



lO'i interval. 

 Makes. Breaks. 



Excess of time for make-circuits. 

 5» 10' 1 Diflf. 



I. 1 . . 

 I. 2 . . 

 I. 3 . . 



Middle 



Zinc 



None 



22". 293 22' 000 

 22.260 22.056 

 22.101 21.919 



22=360 

 22.063 



2P.915 

 21.910 



+ 0=293 

 4-0.204 

 + 0.128 



+ 0".385 

 + 0.163 



+ 0M81 

 + 0.025 



Nov. 16. Signals FROM "Valencia ; Battery at Newfoundland. 



III. 1 . . 



in. 2 . . 



III. 3 . . 



Middle 



Zinc 



None 



20.920 

 21.045 

 21.061 



20.196 

 20.816 

 21 030 



21.010 

 20.911 

 21.120 



20.935 

 20.895 

 21.030 



—0.124 

 —0.169 

 —0.031 



—0.015 

 —0.082 

 —0.090 



+ 0.049 

 + 0.081 

 + 0.053 



It is thus manifest that in general a longer time was required for the transmis- 

 sion of signals after an interval of ten seconds, than after an interval of five seconds. 

 In those cases where no earth-connection existed, and the signals were alternately 

 positive and negative, the cable was meanwhile assuming its electrical equilibrium, 

 so that a positive signal was transmitted more rapidly through the conductor when 

 it was affected with a larger amount of negative electricity, and a negative signal 

 more rapidly through a conductor containing more positive electricity. This affords 

 new testimony to the erroneous character of the supposition that the conductor 

 must be charged through any portion of its length, in order to transmit a signal 

 beyond this portion. 



As showing the continued existence of currents (doubtless engaged in establishing 

 equilibrium) during the intervals between the signals, it may be of interest to men- 

 tion that on one occasion when the two cables had been joined at Heart's Content 

 without battery, and while the Valencia battery had been temporarily disconnected, 

 signals from Newfoundland were distinctly received. They were weak, and the 

 deflections of the needle were scarcely one-fifth as large as usual, yet they were 

 none the less distinct, and a complete set of signals, ten in number, at proper inter- 

 vals and preceded by a " rattle," was recognized at Valencia. No other record of 

 them was made, than the fact of their transmission by alternation of the make-cir- 

 cuit and break-circuit signals, although no battery had been connected with the 

 cable for several minutes. 



On the 16th of November I made a series of experiments at Valencia, for the 

 purpose of ascertaining the effect of changes in the electromotive force upon the 

 speed of the signals, and whether these signals could, by the interpolation of any 

 resistance between them and the galvanometer, be made to traverse the double 

 length of the cable before reaching the galvanometer at the same station. 



The results of these experiments may be very briefly stated, after mentioning some 

 details regarding the signal-key or commutator. The construction of this key was 

 such, that very little time was lost in pressing down either button, the interval being 

 as nearly as I could estimate, about one-seventieth of a second, or approximately 

 O'.OIS. All signals by which currents were sent were given in this way, but the 

 break-circuit signals were given by removing the thumb from the button, which was 



