464 



MEMOIRS OF TOE AMERICAN ACADEMY. 



The mean delay of Mr. Dean in recording cable-signals, as computed from three 



hundred and eighty signals, is 8 .334 ± .010. 



The mean delay of Mr. Goodfellow in recording cable-signals, as computed from 

 three hundred and ninety-seven signals, is 9 .298 ± .007. 



TRANSMISSION TIME BETWEEN DUXBURY AND BREST. 



1 



January 



Feb. 



Mean. 





§ 



8 



9 



10 



14 



17 



1.063 

 1.167 



22 



24 



26 



28 



9 



Ci ass EL 



Class III. 



l.in 



1.171 



1.127 



1 . l no 



1.044 



1.129 

 1.056 



1.085 

 1.114 



1.154 



• 



1.164 



1.116 



1.308 



1.090 

 1.157 



1.081 

 1.174 



s 



1.217 



1.110 

 ±.008 



1.149 

 ±.015 



Mr \ v. 



Lin 



1.150 



1.072 



1.093 



1.200 



1.115 



1.159 



1.212 



1.123 



1.178 



1.217 



1.132 

 ±.009 



The time lost by Mr. D 



o 



n 



Brest was S .334, and by Mr. Good 



fellow in noting signals at Duxbury was 8 .298. If 



correct the transmission time 



for the personal and instrumental delay in recording cable-signals, we have 



x 



Correction 



1U32 ± .009 



| (0.334 + 0.298) = .316 ± .006 



Corrected value of x= a .816 i .011. 



The mean transmission time between Duxbury and Cambridge, as obtained from 

 the Arbitrary Hand-signals, is 8 .010 ; from the Cable Key-signals at Duxbury and the 



Mean Time Clock-signals at Cambridge, is S .009 ; and from the Sidereal Clock-signals 

 it both places, is 8 .037. 



When arbitrary hand-signals were sent from Duxbury to Cambridge, arbitrary hand- 

 signals were returned from Cambridge to Duxbury. When such signals are sent, the 

 key breaks only the main line at both stations, and the current acts through a relay- 

 magnet at both stations. Therefore, by combining the results of these signals when 

 sent in both directions, the transmission time is correctly eliminated. When clock- 

 signals from the Mean Time Clock were sent from Cambridge, it only broke the main 

 line as an arbitrary hand-signal would have broken it, and therefore such signals are 

 properly combined with the signals sent from Duxbury to Cambridge by the 



cable 



key 



When sidereal clock-signals are sent in both dire 



may be 



pared, so as to eliminate the transmission time and deduce a correct longitude 



Inas 



much as, however, in this case, the sidereal clock breaks directly the main line, without 



