ANALYSIS OF THE GALVANIC PILE, \Q$ 



tube, the chemical effects went on with the same rapidity: 

 and the increase of the quantity of the fluid; by connecting 

 the pile with the prime conductor of the machine, which 

 made the pile emit spontaneous flashes, did not sensibly 

 increase these chemical effects. Not having then to contend 

 with the idea that negative and positive states were proper- 

 ties of any substance, I concluded only from the above 

 experiments, that our strongest means of depriving the pile 

 of electric fluid left always enough of the latter for the aV- 

 culation produced by the property of the former, so small 

 is the quantity thus set in motion ; and that the increase of 

 its absolute quantity did not sensibly increase that which 

 circulates in the pile. 



First dissection of the pile. 



This second kind of pile is represented at fig. 2, No» 2,: Exp. 4. 



the tripods, in this dissection, are placed between the twl» ^Vet clotn ill 



r r contact with 



metals, thus separating groups in which these metals have each metal, & 



between them wet cloth. the m< * tals se- 



parated. 



Exp. 4. The experiments made on the continuous pile 

 having been detailed, I shall only assemble under this head 

 the comparative effects of the two piles. 



1. This new mode of experiments haviug been made im- 

 mediately after the former, the free pile (I mean without 

 connexion either between its extremities or with the ground) 

 had the same electric indications. 



2. No shock was perceived. 



3. When the glass tubes were applied, the same chemical 

 effects were produced in them, at the same wires; they only 

 began later, and were a little smaller. 



4. The condenser applied to the three points a, c, b, 

 manifested, in the same circumstances, the same residua of 

 electric signs. 



At first it might be thought, that this dissection indicate? 

 the efficient groups for both electric and chemical effects; 

 but the following experiment will show, that the efficient 

 groups for these two functions are different. 



S-econd 



