Often the experimenter is in need of a good dry battery of a certain size or shape for some particular work, where the ordinary standard sized cell is either too large or not of the right shape for the same work. An inexpensive dry cell that will produce results, and can be made by anybody capable of handling a soldering iron, has long been the desire of every experimenter. The cell as described herein has been used for many purposes and with marvelous results by the writer, and has been used in other shapes where economy in space was desired, such, for instance, as in wireless telegraphy.

To make a cell of the standard size, a strip of zinc of medium thickness, 8 1/4 inches long and 6 inches wide, is necessary. The zinc is rolled into a cylinder 6 inches long and 8 inches in circumference, thus leaving a quarter of an inch which is to be tightly soldered. A zinc cap is next soldered on one end of the cylinder. Any solder showing on the inside is to he well shellacked. Do not shellac any zinc surface, as that will interfere with the action of the battery. Line the inside of the cylinder with a thin layer of blotting paper. For the positive pole of the battery the carbon from a wornout cell is the best if procurable, but if not, a bundle of are light carbons with the copper surface well filed off is the next best. The chemicals for producing the action that generates the electricity are. 1/4 pound of sal ammoniac, 1/4 pound of chloride of zinc (paste form), 1/4 pound oxide of zinc, and 1/4 pound plaster of Paris. These salts should be thoroughly mixed with a mortar, and packed tightly in the cylinder about the carbon, which is a half inch from the bottom of the cell. This paste will fill the cylinder to within half an inch of the top. the rest being filled with paraffine. A copper terminal soldered to the zinc and a heavy copper wire scraped clean and bound about the protruding end of the carbon, form the negative and positive connections for the cell.

This cell can be easily formed in various other shapes to suit the experimenter's fancy. The cell herein described, if constructed according to specifications, will produce very satisfactory results, especially in ignition work or for wireless telegraphy. It will register about I 2/5 volts and between 10 and 15 amperes.