52 MR. DANIELL ON VOLTAIC COMBINATIONS. 



the two balls, of the diameter of 4 inches by 3j nearly, beyond which there was 

 a general diffusion of darker red precipitate (fig. 3.). The surface of this oval was 

 perfectly smooth and compact, and presented no appearance of greater thickness or 

 inequality at the centre than at any other part. The two balls seem to have acted as 

 one oblong piece of generating metal would have done ; and, upon the hypotliesis of 

 a radiant force from each, no increase was perceived at the points upon which the 

 double set of rays might have been supposed to impinge. 



I was desirous to repeat once more the experiments with the opposite sides of the 

 conducting plate, and to connect them more closely with some of the previous series, 

 by measuring the effects upon the calorific galvanometer, and by varying the form of 

 the plate. For this purpose I took a circular plate of copper of the same diameter 

 as the sphere, and immersed it in the copper solution. A zinc ball of 1 J inch dia- 

 meter, placed in a membrane of acid over the centre, gave the following results : — 



First Series. Second Series. 



Ball at top . . . 79 82 



Ball at middle . . 89 92 



Ball at bottom . . 103 105 



After these experiments the copper was found diffused over both sides of the plate, 

 but did not extend to the centre of the under side. 



The upper surface of the plate having been thoroughly covered with lac varnish, 

 the experiment was repeated with the following results : — 



Ball at top 69 



Ball at middle 75 



Ball at bottom 70 



The precipitated copper was here found deposited upon the under side in a ring about 

 two inches in breadth from the edge, and there was very little in the centre of the 

 plate, and none upon the upper surface. 



I finally covered the under surface with varnish, leaving the upper exposed, with 

 the following results : — 



Ball at top 73 



Ball at middle 83 



Ball at bottom 93 



Hence it appears, that the under surface, which by itself is capable of sustaining an 

 action from the ball in the centre of the solution nearly as great as the upper surface, 

 when combined with the latter, adds no more than 10°, or about one-eighth to its 

 efficiency. It appears also that, whereas with the upper surface, the action increases 

 in some inverse ratio of the distance, of the generating from the conducting surface, 

 with the under surface, there is a maximum point, on both sides of which it decreases. 



