302 The Rev. T. R. Robinson's Experimental Researches on the 



electrodes of the contact theory). As I have formerly sho\vn, it varies by heat. 

 I assigned 004986 as the change for 1° Fahr., but this value was obtained by 

 dividing the current, and without means of correcting the rheostat for tempe- 

 rature. I have since obtained by better methods — 



e at 60° = 62-229, change for V = 0-06735. 



It is not affected by the quantity of sulphiu-ic acid mixed with the water to in- 

 crease its conducting power, being almost identical whether this be ^ or ^ of 

 the electrolyte. Nor is it (within very wide limits) by the size of the electrodes ; 

 being the same when they oppose surfaces of 19 square inches (the size of the 

 platinum in the battery), of 3, or of 0-75, the intensity of the battery being 

 given. 



But there is a change, real or apparent, depending on that intensity. The 

 value above given was obtained with two Groves' ; with three it is 69-137 at 

 60°, and with four 75-052. It is my present belief that this seeming increase 

 is caused by two things: by the internal resistance of the cells decreasing in 

 consequence of being heated by the current, and by the rheostat wire being 

 hotter within than at its surface. The thermometer immersed in the water 

 gives merely the latter temperature, and therefore the resistance correction is 

 too small.* This, however, I hope soon to be able to determine. 



After this long preface (which I hope will not be useless to any one who 

 may engage in these or similar researches), I proceed to state in the following 



* Taking the equation 



m tan (b = F = , 



and introducing a resistance p, which produces the deflection 0*, 



^_ m(p + dR) 



cot 0' - cot 0' 



dR being any change of the cells' resistance. Introduce now the voltameter, and a similar equa- 

 tion gives E-e. Now if the wire be hotter than we reckon, we use a value of p less than the 

 truth, E - e is therefore too little, as we compute it; but E, as separately determined, is also too 

 little, nay, even more so, because the current is stronger when the voltameter is not in circuit. 

 Therefore e will be too great. To obtain access to the truth, it will be necessary, first, to deter- 

 mine the law of the cell's resistance as connected with its temperature ; and secondly, to measure 

 the wire's temperature not by an immersed thermometer, but by its own expansion. 



