TRANSACTIONS OF SECTION A. 289 



because the batteries were found next day not to be in good order. In taking the 

 mean I reject them, and find 



Cap. A = 03306 1 „, 

 „ B = 0-323 } mfds - 



To determine the capacity by loss of charge the connections were arranged as 

 below. 



Fig. 1. 



B is the battery, s a Varley's 'slide resistance,' dividing the potential into 

 10,000 parts. 



a, a resistance in pt. a g wire = 200,000 ohms. 



b, a variable resistance. 



x, a selenium resistance from 200 megs, to 270 megs. 



r, a pt. a g wire resistance = 1,000,000 ohms. 



The resistances a and r were lent me by Mr. H. A. Taylor, who had made 

 them with great accuracy, and in conjunction with whom a few of the observations 

 now given were made. 



The selenium resistances were made by Mr. Bassett ; the selenium bars were 

 sealed in glass tubes with platinum terminals. The tubes inserted in holes bored 

 in a plate of ebonite and filled in with paraffine wax, ensuring a very high degree 

 of insulation. 



O is the condenser, K a Lambert's discharging key, E a Thomson's quadrant 

 electrometer, and G a sensitive galvanometer. 



The method of observation was as follows : 



a. /and g were pressed down and b adjusted, so that x:r = o:b. 



£. At a piven instant g is raised and the slides gradually moved to the 

 expected position of the spot after the condenser had discharged itself for the 

 proper time. 



y. At a given instant / is raised, the slide reading, and position of the spot 

 noted as soon as possible, and also after given intervals. 

 8. Observation a repeated. 

 1879. n 



