188 TROWBRIDGE. HIGH ELECTROMOTIVE FORCE. 
and switchboards could be further reduced. The danger in using the batteries evi- 
dently increases with the number of switches which must be manipulated. This 
danger is very great, for 20,000 cells, even of the small test tube type, can yield for 
a few seconds a current of eight amperes at 40,000 volts. In using the batteries one 
is speedily convinced that a steady current of large amperage at voltages as low as 
200 can be dangerous to life. The amperage is just as important a factor as elec- 
tromotive force in considering this danger. 
o ™*~ —& 
L 
Figure 2, Plate XXV, shows the arrangement of connections and switches on the 
back of each compartment of the battery. There are 48 compartments. In each 
compartment there are 7 shelves, 60 cells to each shelf; thus 420 cells, or 820 volts, 
to a compartment. These cells are connected to the line for charging or discharging 
by the switches S. When these switches are thrown to the right hand of one facing 
the switchboard, the cells of each shelf are connected in series up to switches S 4 S 4 . At 
this point 3J shelves are put in multiple; that is, 210 cells, or 420 volts. The charg- 
ing current of 500 volts thus has to overcome this difference of potential. All of the 
switches S to S 4 S 4 are thrown to the right, both in charging and discharging. To dis- 
charge it is only necessary to throw switch S 5 to the left, thus putting all the cells 
in series. The time necessary to throw the cells from charging to discharging is ten 
minutes. From two to three hours of charging with a voltage of 500 and a current 
of two amperes is generally necessary to bring the cells up to working voltage. As 
in the case of all lead accumulators or storage batteries, a transient or pseudo voltage, 
if the cells are not fully charged, is shown, which quickly disappears in connecting 
the terminals of the cells. 
One of the chief causes of the deterioration of an accumulator is the buckling of the 
plates due to overdischarging. This buckling causes the positive and negative plates 
circuit the cells. This buckling can readily happen unless 
the plates are massive, or held apart rigidly. Massive plates add much to the weight 
of the cells, and when a large number are needed for great electromotive force the 
weight is prohibitive. The plates are 1.2 cm. broad, 14 cm. long, and are approxi- 
mately 1 mm. thick, weighing approximately one-sixteenth of a pound. They are 
roughened deeply by a suitable die. 
At first I adopted the expedient of keeping the plates apart by means of rubber 
bands. Experience, however, has shown that this is 
touch, and thus to short 
is shown that this is a very temporary expedient, for 
the rubber disintegrates, and also becomes covered in time with the active material 
which falls from the plates ; they then serve as conductors to short-circuit the cells. 
In a new arrangement of the cells I have dispensed with the rubber separators and 
