176 Mr Wilberforce, On some experiments on the [Oct. 27, 



time of the commutator, the resistances of the bridge, and the 

 capacity of the condenser. 



The arrangement may be represented by a slight modification 

 of the ordinary Wheatstone bridge diagram. The commutator 

 consists of the spring PN which is fixed at P, and which strikes R 

 and S alternately, making (n) double vibrations a second. The 

 vibrations are regulated and maintained by a current rendered 

 intermittent by passing through a fork interrupter. 



If we neglect all the coefficients of self- and mutual-induction 

 of the resistance coils, and the resistances DR, SB, the relation 

 between G, the capacity, n, and the resistances has been investi- 

 gated by Thomson, and is 



nC=^. 

 cd 



The experiments were all made at the Cavendish Laboratory 

 during July and August, 1884. 



The condenser used was divided into four parts whose reputed 

 capacities were *1, - 2, *3, '4, of a microfarad respectively. 



The galvanometer used was a reflecting one with a resistance 

 of 11,000 B.A. units. A vertical slit was placed in front of the 

 lamp used so that a narrow band of light with very well defined 

 edges was reflected on to the scale. The readings were taken for 

 the left-hand edge and were correct to one scale division, the 

 divisions being a millimetre apart. 



The fork-interrupters used had frequencies (JV 32 and iV 64 ) of 

 about 32 and 64 double vibrations per second respectively. They 

 will be referred to as the "32" and "64" forks. A battery of 

 three pint Groves was used for driving them, while in the bridge 

 three Leclanche' cells were used. 



The resistances (a) and (d) were coils of 10 and 1000 B.A. units, 

 while (c) was a resistance box by means of which we could 

 proceed by single units up to 1000. 



Every precaution was taken to make the resistances of con- 

 nection extremely small. The key used for opening and closing 

 the galvanometer circuit consisted of two wires whose ends dipped 

 into four mercury cups. It was arranged that when the galvano- 

 meter was not in circuit the ends of its coil should be connected. 



Various precautions instantly suggest themselves, which must 

 be taken in order that: 



(i) the conditions assumed in the investigation of our 

 formula should be the conditions of the experiment ; 



(ii) the calculated value of G should be correct within as 

 small a fraction of itself as possible ; 



