LAEGE DOUBLE-CURRENT VOLTA-FARADIG APPARATUS. 265 



induction tliat is always suflScient ; but the induction of my ap- 

 paratus, is not, so to speak, saturated by these three elements ; for 

 I have obtained from it a much greater force, with the fifteen or 

 twenty elements of bisulphate of lead that I can adapt to it at my 

 own house. 



Although M. Marie-Davy has advised the use only of the proto- 

 sulphate of mercury, I have substituted the bisulphate in order to 

 obtain greater power. I have only employed the proto-sulphate, 

 as I have said, for the formation of a battery composed of from 

 thirty to a hundred elements, for the application of continuous 

 currents. 



Manufacture of the carbon plates. — -After numerous experiments 

 with regard to the best kinds of coal and. coke, to the proportions 

 in which they should be mixed, to the manner of setting the 

 mixture in the moulds, to the form of the latter, and to the amount 

 of burning required, I succeeded in manufacturing carbon plates 

 that were hard and free from fissures. The manufacture of these 

 plates being, however, both difficult and costly, I sought to replace 

 them by carbon sawn out of gas-coke. 



The coke deposited on the interior of the retorts used for making 

 gas is formed, as is known, from the bituminous matter called 

 petroleum. This petroleum becomes carbonized, and deposited, 

 layer upon layer, on the inner surfaces of the retorts in which the 

 coal is distilled. It forms at length a thick crust, which it is 

 necessary to remove from time to time. Unfortunately, the carbon 

 plates formed from gas-coke, which are excellent when immersed 

 in nitric acid, and which, being very pure and free from earthy 

 matter, seem then to increase power, do not succeed for the system 

 that- 1 have adopted in my instruments. In a few minutes (two or 

 three) the current furnished by them diminishes by more than 

 three-fourths, when the circuit is closed. We then observe the 

 singular phenomenon — which is produced also in piles of zinc and 

 copper — that the current recovers its intensity when the circuit has 

 been for a short time interrupted. I attribute this rapid weaken- 

 ing of the current to the great density of the gas-coke, the inter- 

 stices of which become obstructed by the salts that are formed by 

 the chemical action of the pile, and that are rapidly dissolved 

 when the circuit is left open. The carbon plates that I have had 

 manufactured, without being too soft, which would interfere with 

 their durability, are yet hard enough. It is not until after some 

 time (five or six months of frequent use) that the salts resulting 

 from the interior chemical action of the pile are deposited in the 

 interstices in sufficient quantity to diminish the power of the ap- 

 paratus by resisting the passage of the current. It is important 



