442 MEMOIRS OF THE AMERICAN ACADEMY. 



the batteries. With this qualification, this class of signals is useful for the determi- 

 nation of longitude and of the velocity with which the charge travels. On the 

 other hand, signals which belong to Classes I, III., and IV. are employed for an exact 

 measurement of the velocity with which the discharge of the battery is propagated 

 through the cable. A preliminary trial, made by Mr. Varley on December 19, re- 

 vealed the fact that two seconds was insufficient to charge the cable so as to give 

 unequivocal signals, and that the signals of five seconds' duration were better. Ac- 

 cordingly, signals of Class I. were not used at all, and those of Class III. were employed 

 more frequently than those of Class II. The signals of Class IV. were seldom used. 

 Moreover, the original plan laid out beforehand by Mr. Dean was modified so as to make 

 Class II. to comprise six sets of signals, of ten each, and Class III. to comprise six sets 

 ol ngnals, of six each, to be sent in each direction. A good night's work would, there- 

 fore, include one hundred and ninety-two signals. 



The following data in regard to the Atlantic cable which connects Duxbury with 

 Brest are extracted from the day-book of Mr. Goodfellow. They were communicated 

 to him by R. T. Brown, Esq., Superintendent of the Duxbury Station of the French 

 Transatlantic Cable Company : 



Length of cable from Duxbury to St. Pierre, 749 nautical miles. 



Length of cable from St. Pierre to Brest, 2580 " " 



Mean cable resistance per knot of 6087 feet, at 75° Fahr., between Duxbury and St. 



Pierre, 



11.99 ohms, 

 between St. Pierre and Brest, . 3.16 " 



Mean gutta-percha resistance per knot, at 75° Fahr., between St. Pierre and Brest, . 2405 megohms. 



between Duxbury and St. Pierre, 2300 " 

 (when laid and without regard to temperature and pressure.) 

 Electrostatic capacity per knot between Duxbury and St. Pierre, .... 0.358 farads. 



bet ween St. Pierre and Brest, 0.404 " . 



Weight of the conductor per knot between Duxbury and St. Pierre, ... 107 lbs. 





Wei 



between St. Pierre and Brest, .... 400 " 



150 " 



between St. Pierre and Brest, .... 400 " 



0.282 inch. 



Diameter of the cores between Duxbury and St. Pierre, 



between St. Pierre and Brest, .. . , . # , 0.468 " 

 Specific conducting power compared with pure copper, between Duxbury and St. Pierre, 92.9 per cent 



p . between St. Pierre and Brest, . 94.3 " " 



Reliance of the galvanometer used in the observations for longitude (temp, of 56 



xahr.), # # 



gal 



1188 ohms. 

 39 inches. 



4 



