1878.] 



Contact Theory of Voltaic Action. 



199 



force of contact tends to produce a current of electricity from N to M 

 across the junction. 



Sir W. Thomson's method of measuring the difference of potential 

 between two metals in contact, which consisted in hanging a needle 

 symmetrically oyer two similar plates of different metals, and observ- 

 ing its deflexions when it was charged, first positively and then nega- 

 tively, requires, in order to be accurate, all the delicacy of the arrange- 

 ment and even more complicated contrivances than those employed in 

 the quadrant-electrometer. It will be easily seen that a series of 

 experiments for determining the difference of potentials between pairs of 

 metals, pairs of liquids, and between liquids and metals, could only with 

 the very greatest difficulty be carried out by a method which virtually 

 requires the taking to pieces and refitting of a quadrant-electrometer 

 for every experiment. The authors' attention was, therefore, directed 

 to devising a method of investigation in which a delicate quadrant- 

 electrometer would be employed to measure the difference of potentials 

 between two similar metal plates themselves acted on inductively by 

 parallel and equidistant plane surfaces of the two substances in contact, 

 the difference of potentials between which it was desired to measure. 

 Thus let A and B be two insulated brass plates connected with the 

 electrodes of the electrometer. Let G under A and D under B be the 

 surfaces whose difference of potential is to be measured. A and B are 

 both put to earth and then insulated. G and B are made to change 

 places with one another, G now being under B and B under A. The 

 deflection of the electrometer needle will now give a measure of the 

 difference of potential between G andD.* 



The apparatus, by means of which our idea has been carried out, 

 has, by a process of natural selection, assumed the form shown in the 

 accompanying drawing. (Plate 7.) The circular brass plates A and 

 B, 20*4 centims. in diameter, are supported and kept insulated by 

 means of artificially dried glass stems. The leaden cups /, containing 

 pumice stone moistened with sulphuric acid, slide in H by a bayonet- 

 joint, and may be lowered by the handles I so as to rest on the brass 

 plates when the instrument is not in use ; thus preventing access to 

 H of the external air. 



The permanent adjustment for coincidence of the plane surfaces of 

 the brass plates is effected by means of three wood-screws in each of 

 the brass caps K; temporary adjustment for horizontality of the brass 

 plates is effected by means of a spirit level and three levelling screws 



* Suppose C and D to be very large, and that the protecting covering of the 

 apparatus is far away from the insulated plates. Let the coefficients of induction of 

 the two sets of quadrants when there is a difference of potential d between them be 

 s + dm and s — dm. We know from the position and shape of the needle that this is 

 approximately true. Let a be the area of each of the insulated brass plates A and 

 JB; t the distance from A to C or from B to D. Let a be the potential of the sur- 



