102 Mr. R. T. Glazebrook on a Method of 



Another set of observations was then taken to test the 

 effect of varying the battery-power driving the commutator, 

 thus altering the force with which the spring P N is attracted 

 to G, three Grove cells being used in place of two. The 

 values of d found in two separate observations were 905*5 

 and 905. 



Observations were made to test the effect of varying 

 the period of the driving-fork. Forks of frequencies of ap- 

 proximately 16, 32, 64, and 128 to the second were taken, 

 and the pitch of the lowest of these determined absolutely by 

 comparison with the clock in the method described by Lord 

 Rayleigh (< Nature/ xvii. p. 12; Phil. Trans. 1883, Pt.L). This 

 fork and the 32 fork were then set going on independent cir- 

 cuits, and a pointer attached to the 32 fork was watched over 

 the top of a plate of thin metal which had been all along fixed 

 on to the 16 fork, the eye being placed so that this pointer was 

 visible for only one position of the 16 fork in each vibration. 

 In this way the beats between the octave of the 16 fork and 

 the 32 fork were visible, and could easily be counted. 



Thus the period of the 32. fork was found, and then in a similar 

 manner those of the 64 and 128 forks. By loading with wax 

 one of the two forks under comparison, it was easy to find 

 which of the two was gaining. 



Each of these forks was used in turn to drive the commu- 

 tator, and the value of the capacity calculated from the 

 observations. The values of the resistances were as follows: — 



a was a standard coil of 10 B.A. units, d a resistance of 

 1000 B.A. units, taken from a box by Messrs. Elliott, Bros., 

 and c a variable resistance from another box by the same firm. 



g, the galvanometer-resistance, was about 11,000 B.A. 

 units ; and b, the battery-resistance, from 5 to 6 B.A. units. 

 The galvanometer-resistance was higher than necessary, but 

 it was for other reasons the most convenient instrument for 

 the purpose. The value of c varied from about 1800 to about 

 240 B.A. units. 



It will be found that, with these values, the equation giving 

 the capacity may be written 



nC = 



cd 



I 1+ «K)} 



[The approximate equation given in Maxwell is 



