The drying of the substance was a great inconvenience in the Voltaic pile, and the inventor proposed, as a remedy for this, to station the metallic plates at a greater distance from one another, and to fill up the cells or intervals between them with a saline solution. Mr. Cruickshank, an English chemist at Woolwich, improved this construction.—His _trough_ as it is called, is made thus.— Get a wooden trough, made of hard baked mahogany, about thirty inches long, and four or five wide and deep.—On the inside let there be cut in the sides and bottom, and at equal distances from one another, as many grooves, as the number of plates required to be put into the trough;—the grooves of a size to admit the plates. The plates are to be cemented[18] separately to each of the grooves, so that no fluid can pass from one cell to another. In this instrument the plates are constructed by soldering a plate of zinc to one of copper. The zinc, or which is the same thing, the spelter of the shops, should be melted in a vessel which exposes but a small surface to the action of the air, otherwise it would absorb oxygen so rapidly as to be converted into the flowers of zinc.—The melted metal should be poured as soon as possible into a mould of the proper size, made of stone or brass.—It is not necessary that the plates of copper should be more than one tenth of the thickness of those of zinc. The two plates are commonly soldered, not through their whole extent, but about one fourth of an inch from the edge; so that at the edge their union may be complete. Care must be taken that all the plates be cemented to the trough in the same direction; so as to have the copper side of every plate opposite to the zinc side of the next. The liquid employed to fill up the cells between the plates, is formed by diluting muriatic acid with water, in the proportion of one ounce of the former to a pint of the latter. When the trough is not in use, it should be emptied of this solution (which may be preserved for subsequent experiments, unless saturated with the metals) and then rinsed clean with fresh water. This construction is preferable to the Voltaic pile, for experiments in which it is necessary to have the Galvanic action for a length of time. But for occasional experiments the pile is more convenient; as the trough, if suffered to remain long without the fluid, is apt to crack and separate the cement from the plates, which renders it necessary to cement them again.