the salt constantly change places and enter into fresh combination, one consequence of which would be a gradual passage of the nitric element from the growing silver to the copper plate. This actually takes place, and there is a diminution of the salt at the ends of the silver branches, giving rise to an upward current and a condensation of nitrate of copper against the copper plate, which gives rise to a downward current. These two currents are seen in every reaction of this nature. In the case of silver and copper, however, it has been proved that the crowding of the salt toward the copper plate is more rapid than would follow in the usual "polarization theory." The instrument used in determining this point was a divided cell in which two plates, one of silver and one of copper, connected with a wire, are each immersed in a solution of its own nitrate, contained in each division of the cell, and separated merely by parchment paper. The crystals of silver deposited on the silver plate in this experiment are very brilliant. There are indications of the liquid being put into a special condition by the presence of two metals which touch one another. Zinc alone is incapable of decomposing pure water, but if copper or platinum be deposited on the zinc in such manner that the water can have free access to the junction of the two metals, a decomposition is effected; oxide of zinc is formed and hydrogen gas is evolved. At ordinary temperature the bubbles of gas rise slowly through the liquid, but, if the whole be placed in a flask and heated, pure hydrogen is given off in large quantity. We have also found that iron or lead similarly brought into intimate union with a more electro-negative metal, and well washed, will decompose in pure water.
As might be expected, the action of magnesium on water may be greatly enhanced by this method; and a pretty and instructive experiment may be made by placing a coil of magnesium in pure water at the ordinary temperature, when there will be scarcely any visible effect, and, then adding a solution of sulphate of copper, the magnesium is instantly covered with a growth of the other metals, and at the same time the liquid seems to boil with the rapid evolution of the hydrogen bubbles from the decomposed water.
When, however, the force of the two metals in contact has to traverse a layer of water, the resistenceoffered by the fluid prevents its decomposition. This must also be an important element in the decomposition of a metallic salt dissolved in water—and, in fact, we have found that the addition of some neutral salt, such as nitrate of potassium, increases the action, apparently by diminishing the resistance of the liquid. If, too, we increase the quantity of the dissolved salt, we get more than a proportional increase of deposited metal. Thus, in an experiment made with different strengths of nitrate of silver, the following results were obtained in ten minutes, all the circumstances being the same except the strength of the solution:
