472 MEMOIRS OP THE AMERICAN ACADEMY 



and St. Pierre being 2,300 megohms, the total insulation is 3,880,000 ohms. The 

 insulation of each knot of cable between St. Pierre and Brest being 2,405 megohms, 

 the total insulation is 2,000,000 ohms. The total insulation of the gutta-percha 

 is expressed by 1,300,000 ohms ; and is, therefore, about seventy-two times greater 

 than that of the conductor. If we suppose Duxbury and Brest to be united by a 

 homogeneous cable throughout the entire length, and suppose also that the remote 

 end of the 'able is put to ground, if the insulation is perfect, the tension will diminish 

 regularly from its maximum at the battery end down to zero at the ground end. 

 Under these circumstances the charge of electricity it would hold, with the battery 

 of 36 volts, would be one half of what the same cable would contain when it was dis- 

 connected from the ground ; that is, 23,580 microfarads. In the actual case, the two 

 branches of the cable were very different ; moreover, the two cables were not imme- 

 diately joined, but a Leyden condenser was used at St. Pierre, one surface of which 

 was connected with the branch which extended to Brest, and the other surface with 

 the branch which proceeded to Duxbury. 



The current sent into the • cable would be, according to Ohm's formula, n , . If we 



° K -\-r 



substitute for (E) the electromotive force of the battery used, viz. 36 volts, and for 

 (R) the resistance of the battery (which was about 68 ohms), and for (r) the total re- 

 sistance of the conductor, which was about 8,153 ohms for the St. Pierre and Brest 

 brand*, and about 8,980 ohms for the St. Pierre and Duxbury branch, adding also the 

 resistance of the galvanometer, we should have ¥ f f ¥ (or .003826 of one veber) for 

 the strength of current between Brest and St. Pierre, and t ^%q (or .003517 of one 

 veber) for the strength of current between St. Pierre and Duxbury. When s 



o 



sent from Duxbury to St. Pierre, and then repeated by the condenser between St. Pierre 

 and Brest, 3,517 microvebers flowed into the circuit every second between Duxbury 

 and St. Pierre, and 3,826 microvebers between St. Pierre and Brest. When signals 

 were sent from Brest to St. Pierre and then forwarded by the condenser to Duxbury, 

 the number of microvebers circulating every second in the two branches was less than 

 the numbers just mentioned in the ratio of about 253 to 337, since the battery used 

 at Brest was smaller than that which operated the Duxbury end in this ratio. This 

 maximum strength of the current would not be attained, throughout the circuit, until 

 after connection with the battery had been maintained for an indefinitely long period 

 of time. But ninety per cent of the maximum would be reached in the time allotted 

 to signals of Class IV., and a large fraction of the maximum even in the times assigned 

 to Classes II. and III. For the battery would supply the 18,756 microfarads required 

 to charge fully the longer branch of the cable in five or six seconds, and the 4,824 



