Decomposition of Water by the Galvanic Battery. 93 



the deflection was a little less than that which was produced by 

 the current from 3 cells when it passed through one decom- 

 posing cell and the coil. The difference arose from the imper- 

 fect insulation of the cast-iron cells. But when the current was 

 sent from the same 12 cells through 4 decomposing cells, and 

 then through the helix of an electro-magnet b)'^ which a small 

 magnetic machine was driven, the speed of the machine was 

 considerably greater than when the current from 3 cells passed 

 through one decomposing cell, and then through the coil of the 

 electro-magnet. The wire coiled on the electro-magnet was 

 about 50 feet long and onc-cighth of an inch thick. The coil of 

 the galvanometer is about 7 feet long and three-eighths of an 

 inch in diameter. The resistance in the latter was insensible 

 compared with the resistance in the former. 



The ratio which the acting surface of each electrode should 

 bear to that of the zinc in each circle may be found by putting 

 a pair of large plates into a glazed vessel, and connecting them 

 with the opposite ends of a battery of 4 cast-iron cells in each 

 of which the zinc plate is small, so that the current will pass 

 through the coil of a galvanometer. If a solution of carbonate 

 of soda be then gradually poured into the glazed vessel until the 

 needle ceases to recede from the magnetic meridian, that is, 

 until no more electricity is transmitted through the fluid, it will 

 be found that the acting surface of each of the electrodes, which, 

 is covered by the fluid, is about once and a half as great as the 

 acting surface of the zinc plate in each circle. By pouring the 

 fluid to any height whatever into the glazed vessel so as to 

 increase the acting surface of the electrodes, no increase will be 

 produced in the deflection of the needle. Hence there is a limit 

 to the conducting power of fluids for electricity of low intensity. 

 A solution of carbonate of soda (no matter how thick and short 

 the column of fluid may be) will not, when interposed between 

 the opposite ends of a nitric acid battei-y of 4 cells, conduct 

 more than about one-half of the electricity which will be con- 

 ducted by a short, thick wire connected with the opposite ends 

 of the same battery. In investigating the ratio which the surface 

 of the electrodes should bear to the surface of each zinc plate of 

 the battery, I used solutions of carbonate of soda of different 

 degrees of strength, and found that the conducting power of a 

 solution, which contains an ounce in each quart of watei', is very 

 little inferior to that of the strongest solution, whilst it foams 

 far less. If a neutral salt could be found, which, when mixed 

 with carbonate of soda, would jircvent its foaming, and would 

 not be decomposed by the voltaic current, sheet-iron plates 

 coated with an alloy of lead and tin would be in every way 

 preferable to platina electrodes. I have tried nitrate of potash. 



