Mr. J. Prideaux on the Theory of Voltaic Action. 253 



and we know how greatly approximation increases the acti- 

 vity of voltaic plates. 



29. If this transference be the real conducting agency, it 

 should follow, that when the poles are disconnected, as the 

 electricity accumulates, and counterbalances those attractions 

 and repulsions between the plates, the action between them 

 should become gradually weaker, and at last cease, but be 

 renewed on restoring the connexion. And such we know to 

 be the fact. 



Also, when the plates, and consequently the intermediate 

 imperfectly conducting strata of liquid, are multiplied, resist- 

 ance to the passage of electricity is increased. It must there- 

 fore accumulate, in degree, on each pair of plates ; and the 

 atom of hydrogen from the zinc of pair A should be unable 

 to discharge itself into the copper of pair B, unless its charge 

 be high enough to overcome the resistance forward, and vice 

 versa. Hence the electricity at the poles of a numerous vol- 

 taic battery, though greater in tension, should be less in quan- 

 tity of current than at those of a battery equal in surface, but 

 in fewer divisions. 



And this also may be seen to be true by the following ex- 

 periment. 



30. For comparisons of this kind I employ a trough of 

 wood, twenty-four inches long, seven wide, and four deep, 

 divided by transverse partitions into six cells, and well lined 

 with cement. Each of these cells contains a small calorimotor, 

 nearly on Dr. Hare's plan, composed of 10 zinc and 11 copper 

 plates, each three inches square, the zinc and copper working 

 into separate mercury boxes (3, &c.) on the top of each calori- 

 motor. 



By passing connecting wires along 13, 23, 33, 43, &c, and 

 1 c, 2 c, 3 c, 4 c 9 &c, the whole six sets become a single pair ; 

 but connecting them 13, 2 c; 23, 3 c; 33, 4 c, &c, they become 

 as many pairs as there are sets. And it is easy to understand 

 how, by arranging the connexions, they become two or three 

 pairs. It is this convenient divisibility which led to the pre- 

 ference of the number 6. 



The following Table shows the deflections, in proportion to 

 the manner in which the calorimotors were divided. 



Connexion 



Experiments 



into 1 pair 60° 58° 58° 59 c 



2 pair 49 48 48 49 



3 pair 42 42 42 43 



6 pair 30 30 .'30 30 



