A number of these elements are placed side by side, with non-conducting frames intervening, so as to form chambers through which the hydrogen gas is passed along one side of the element and air along the other. This peculiar construction allows us to get a very large amount of duty from a very small amount of platinum. One of the batteries before you, consisting of seven elements, with a total effective surface of half a square meter, contains 21/2 grammes of platinum leaf and 7 grammes of platinum black, a total of 91/2 grammes of platinum, and produces a current of 2 amperes and 5 volts, or 10 watts, when the outer resistance is properly adjusted. This current is equal to nearly 50 per cent. of the total energy obtainable from the hydrogen absorbed in the battery. In order to maintain a constant current, we have from time to time (say once an hour) to interchange the gases, so as to counteract the disturbing influence produced by the transport of the sulphuric acid gas from one side of the diaphragm to the other. This operation can easily be performed automatically by a commutator worked by a clock. The water produced in the battery by the oxidation of the hydrogen is carried off by the inert gas mixed with the hydrogen, and by the air, of which we use a certain excess for this purpose. This is important, as if the platinum black becomes wet, it loses its absorbing power for the gases almost completely and stops the work of the battery. To avoid this was in fact the great difficulty in designing a powerful gas battery, and all previous constructions which employed the electrolyte as a mobile liquid failed in consequence. The results obtained by our battery are practically the same whether pure oxygen and hydrogen or air and gases containing 25 per cent. of hydrogen are used; but we found that the latter gases must be practically free from carbonic oxide and hydrocarbons, which both interfere very much with the absorbing power of the platinum black. We had thus to find a cheap method of eliminating these two gases from the producer gas, and converting them at the same time into their equivalent of hydrogen. The processes hitherto known for this purpose, viz., passing a mixture of such gases with steam over lime (which I mentioned some time ago) or over oxide of iron or manganese, require high temperatures, which render them expensive, and the latter do not effect the reaction to a sufficient extent for our purpose. We have succeeded in attaining our object at a temperature below that at which the gases leave my producers, viz., at 350 deg. C. to 450 deg. C., by passing the producer gases, still containing a considerable excess of steam, over metallic nickel or cobalt. These metals have the extraordinary property of decomposing almost completely, even at the low temperature named, carbonic oxide into carbon and carbonic acid and hydrocarbons into carbon and hydrogen.