THE TRANSATLANTIC LONGITUDE. 99 



The question as to the route by which signals are transmitted, when part of the 

 circuit is formed by the earth, is thus disposed of; and the position maintained in 

 the memoir above cited seems entirely corroborated, although it loses its theoretical 

 significance. Prof. Kuhn, in his learned and valuable Handbuch der Elektricitdts- 

 Ze/ire,^ while doing the fullest justice to the former investigation in other respects, 

 takes exception to the propriety of my inferences regarding this question, but 

 careful reconsideration has failed to convince me of any flaw in the argument, such 

 as it is, notwithstanding my distrust of any reasoning from which so eminent a 

 physicist would dissent. 



Our experiments with the cables are inadequate for any decided deductions 

 regarding the relative velocity when the earth forms a part of the circuit, but it 

 may be well to examine for a moment what they appear to indicate. 



The transmission-time for the several signals in our exchanges of November 10 

 and 16 may be approximately determined by a method different from those which 

 we have thus far employed. Since the experiments occupied but a comparatively 

 short time on each of these days, we may suppose the clock-errors to have remained 

 constant during each series. Then, from those experiments in which no earth-con- 

 nection was made, we may deduce the constant difference of the two clock-times ; 

 and a comparison of this quantity with the difference of clock-times as deducible 

 from any set of signals will afford a near approximation to the actual time of their 

 transmission. 



Thus we have from II. 3 and IV. 3, supposing the speed the same in each direc- 

 tion — 



Date. 



Signals. 



Diff. ( 



)f records. 



Error of noting. 



True interval. 



Diff. 



of clocks. 



November 10. 



A^'alencia, 



2'' 62' 



"^ 20'. 790 



-|-0=.331 



21M21 



2'> 52° 



2P.382 





Newfoundland, 





21.917 



—0.275 



21.642 







November 16. 



Valencia, 





21.184 



-fO.331 



21.515 



2 52 



21.753 





Newfoundland, 





22.236 



—0.275 



21.991 







and adopting these values of the difference of clocks, we obtain as the translnission- 

 times — 



Experiment. Signals. 

 I. 1. Valencia, 



Nevffoundland, 

 I. 2. Valencia 



Newfoundland, 

 I. 3. Valencia, 



Newfoundland, 



IL 1. Valencia, 0.181 



Newfoundland, 0.308 



IL 2. Valencia, 0.279 



Newfoundland, 0.288 



II. 3. Valencia, 



Pos. & neg. 



Make-circuit. 



Break-circuit. 



0^202 











0'.633 



0'.343 



0.271 









0.636 



0.383 



0.301 







.... 



0.434 



0.308 



, .260 

 Newfoundland, 



J 0.2. 



' Allgemeine Encyklopadie der Physik. Bd. XX, p. 494, Leipzig, 1866. 



