THE TRANSATLANTIC LONGITUDE. 89 



otlier cable, and the signal thus exhibited upon the Valencia galvanometer. Ob- 

 viously, when the ground-connection was made with the zinc of the Valencia battery, 

 this disturbing action would be the greatest for those signals of which the tension 

 would thus be for a moment partially neutralized ; namely, for the positive signals. 



Valencia. Newfoundland. 



Cable of 1866 ■ 



Cable of 1865 . 



No other explanation than this has suggested itself; and though, as already stated, 

 this scarcely appears adequate, it would yet derive some color from the absence 

 of any analogous differences for the two classes of signals in the first experiment 

 of the same day, in which the ground-connection was made to the middle of each 

 battery. On the other hand, a similar though inferior difierence does exhibit itself 

 in the third experiment, where no earth-contact was made; and it seems safer to 

 assume some additional and yet undiscovered mistake in the arrangement of the 

 connections, and therefore to discard the observations of November 10 altogether, 

 than to attempt to draw any inferences from them. These wouM contradict the 

 experiments upon other days, when the connections were managed somewhat more 

 effectively, although not without mistakes in the Newfoundland batteries on both 

 the 6th and 9th of November. 



On November 16th all the arrangements seem to have been correctly made, each 

 battery consisting of 4 cells, and the earth-connections at both stations being made 

 with the zincodes in the second experiment, and with the middle of the batteries in 

 the third. 



In the former case, all the positive signals found earth at the other extremity of 

 their respective cables without affecting the second cable at all, and therefore with- 

 out manifesting themselves upon the galvanometer at the distant station ; while 

 negative signals, which differed from the positive ones only by the interchange of 

 the cables used for the respective electrodes, were of course received and recorded. 

 Thus we have on this occasion only the " negative" signals; i. e., those in which the 

 platinodes went to the cable of 1866 at Valencia, and to that of 1865 at New- 

 foundland. 



In the latter case the effect of the arrangement would be to substitute two cir- 

 cuits (each consisting of one cable with two cells at each extremity and earth-con- 

 nections), for the one circuit, formed by a cable with four cells at the signal-giving 

 station and with earth-connections ; were it not that a very small portion of each of the 

 two first-named circuits is common to the two, being formed by the piece of metal 

 which unites the short-circuit of the local battery with the connected or " looped" 

 cables. This will be readily seen from the diagram, which represents a positive 

 signal from Valencia. Both sets of signals from Valencia were received at New- 



12 Septeipber, 1869. 



