732 
Proceedings of the Royal Society 
serious. I will state at the end of the paper at what resistance I 
usually stop charging. The zinc should have a smooth surface, with 
no edges to deposit on. It deposits best on a horizontal plate, as 
might he expected, and I have some reason to believe deposits better 
for having light excluded. This point would require careful investi- 
gation. It is very difficult to insulate any part of the zinc plate. 
The zinc appears in hunches, having come through invisible holes. 
The only effective insulation I have found is sheet gutta-percha. 
The form of cell I most frequently use is a tray cell. Thin 
sheet copper is folded at the corners on a wooden mould into a 
shallow tray, with sides sloping outwards. The bottom of each 
tray has stamped in it a depression about inch deep, across which 
is soldered copper gauze. The sides of the tray are carefully 
insulated with gutta-percha, and the solder varnished or plated with 
copper. The shallow depression is filled with cuprous chloride 
paste, through the holes in the wire gauze, and is then covered with 
parchment paper. Each tray is kept apart from the tray beneath 
by gutta-percha supports, or by a piece of old fishing-net. The 
cuprous chloride is reduced by means of zinc plates inserted into 
the trays, and the battery is ready for use. The only objection to 
the tra3 r s is the gradual formation of bubbles of hydrogen, when a 
charged cell is left standing, from the slight local action on the 
zinc plate. These bubbles cannot escape very readily, but it seems 
sufficient to slightly incline the cells. The electromotive force of 
this cell I have already stated. The plates may be brought very 
close together, as very little water is sufficient to dissolve all the 
chloride of zinc required. The resistance is consequently very 
low, a primary cell of 1 square inch surface having a resistance of 
1*2 ohm. 
At the end of this paper I give a summary of one experiment 
with a primary cell, and of two experiments with a secondary cell. 
The storage of this cell for a given surface is very small. This is 
to be expected, from what I have already stated about the limited 
storage of secondary cells. With so small an EMF as mine, it is 
not possible to store much on a given surface. The materials of 
my cell are so light, however, that I can store about 30,000 ft. lbs. 
for every lb. weight of metal in the cell. 
In the primary cells the current is usually pretty steady until 
