ON THE DETERMINATION OF TRANSATLANTIC 



1 



faces separated by a thickness of only & of an inch of glass. Wln-n a cur. nfl is fully 

 established and the wire is charged, an immersed cable-line will conduct as well as an 



connection is first made with the battery, or first broken the 



■line. But when the 



charge in one case, and the discharge in the other, will travel more slowly in the cable, 

 line than in the air-line; in other words, the time of the variable state of tin m- 



* 



ductor will be prolonged. 1 



In the cable between St. Pierre and Brest, the inner surface of 1 1 vt gutta-percha 

 (which represents approximately the surface of copper to be charged) amounts to 

 about 700,000 square feet, and the outer surface to about 2,oou,000 square feet. In 

 the cable between Duxbury and St. Pierre, the inner surface contains about 100,000 

 square feet, and the outer surface about 300,000 square feet. Both cables united cor- 

 respond to a Leyden arrangement, in which one surface has about 800,00 -qua re feet 

 and the other about 2,300,000 square feet. If the dielectric were glass instead of 

 gutta-percha, the equivalent thickness of the glass would be, according to Thump son's 



formula already given, T multiplied by .043 of an inch for the St. Pierre and Brest 



k 



k' 



cable, and T multiplied by .084 of an inch for the Duxbury and St. Pierre cable ; U 

 and k representing respectively the inductive capacities of glass and guttarpcrch. 

 Experiment 2 shows that the ratio % is about Jf The electrostatical rapacity of 



the St. Pierre and Brest cable has been already given as equal to .404 of a micro- 

 farad for each nautical mile. Its total capacity is about 1,042 microfarads The 

 electrostatical capacity of the Duxbury and St. Pierre cable is .358 of a microfarad 

 for each nautical mile, and the total electrostatical capacity is about 288 microfarads. 

 The electrostatical capacity of both cables is about 1310 microfarads. 



This total electrostatical capacity of both cables expresses the whole amount of 

 electricity they contain when one end is united to a battery having an electromot.ve 

 force of one volt, and the other end is put to air; that is, it disconnect.,! with .he 

 ground. As the forty cells of Minolta's battery used at Duxbury were eqmval, ,,, to 

 about 33.7 of Daniell's cells, its electromotive force would be equal to about 36 volts, 

 that of a simple Daniell's cell being taken as 1.079 volts. If one end of the un> ed 

 cables is connected to this battery, and the other end is insulated ,„ the an so ha 

 the difference of potential between the inner and outer surfaces of * 8°»££ 

 envelope is equal to 36 volts, the whole amount of e.ectnc, y require ^charge e 

 cable is 47,160 microfarads. The insulation of eaeh knot of cable between Duxbury 



1 See also experiments of W. Siemens, Ann. de Ch. et de Ph., XXIX. 394. 

 * Clark and Sabine's Electrical Tables and Formula?, p. 67. 



