444 



MEMOIRS OF THE AMERICAN ACADEMY 



one or two minutes at a time. II. The other kind of signals used may be called hand- 

 signals or key-signals} On December 15, 23, and 31, and on January 3, six sets, of ten 

 breaks each, were sent to Cambridge and received from Cambridge. A few failed of 

 being properly recorded on the chronograph-sheets. On January 14, 22, 24, 26, and 

 28 and on February 9 and 10, the Cambridge and Duxbury line was connected, by 

 means of a relay-magnet, with the cable key and clock and chronograph circuit at 

 Duxbury, and when the cable signals belonging to Classes II., Ill, or IV. were received 

 from Brest or sent to Brest, they were recorded on the Duxbury and Cambridge 

 chronograph-sheets. These are called cable-key signals, to distinguish them from hand- 

 signals made with a telegraph key at arbitrary times. 



After the programme already described had been successfully carried out, with no 

 essential variation, by the officers of the Coast Survey in charge respectively of the 

 three stations, the materials were all placed in my hands by the Superintendent, Pro- 

 fessor Benjamin Peirce, in order that I might deduce from them the differences of 

 longitude between Cambridge, Duxbury, and Brest. The computations have been 



made under my direction, and have been carefully re-examined by me. Those designed 

 to ascertain the clock and instrumental corrections at Cambridge were made by Mr. 

 Henry Gannett. Those necessary to obtain the clock and instrumental corrections at 

 Duxbury and Brest, and also those intended to give a precise determination of the 

 differences of longitude, corrected for the Personal Equation in observing transits and 

 noting cable signals, were made by Mr. Lucius Brown. I shall now describe the 

 methods of computation which have been followed. 



The plan which has been adopted for the precise determination of the local time at 

 each of the three stations is essentially the same as that upon which F. G. W. Struve 

 proceeded in working up the results of his two chronometric expeditions to ascertain 

 the differences of longitude between Pulkowa and Altona, and between Altona and 

 Greenwich, and which has become familiar in the Coast Survey Service by the labors 

 of the late Mr. Sears C. Walker and Dr. B. A. Gould. I have not thought that obser- 

 vations made with small portable Transit>instruments, such as those which are used at 

 the Coast Survey stations, would be accurate enough to justify the added labor of 

 assigning different weights to the stars observed for time, according to their declina- 



after the manner indicated by Struve and J. A. C. Oudemans 



2 



1 These are sometimes called arbitrary signals, to distinguish them from clock signals. 

 1 Dissertatio Astronomica Inaugurate. 



