TKOWHRIDGE. — HIGH ELECTROMOTIVE FORCE. ]S7 
are placed side by side with an air space between them, and are kept apart by lass or 
porcelain tubes. 
It is necessary that the cells of each tray should be connected with lead wire, and 
wires and the switches, together with the eop- 
between these lead 
per conductors of the 
far removed from the cells. "\\ 
the battery was first installed, these connecting had wires were led through glass 
tubes in the wood partitions separating the trays. It was soon found, however, that 
the sulphuric acid crept along the lead connecting wires, a distance of eight to ten 
inches from the cells, covering the glass conduits and soaking into the wood parti- 
tions, thus short-circuiting the cells. Accordingly the gla-s tubes were x moved and 
the lead wires were led through airspaces. Thi disposition remedied the difficulty 
My experience with various insulating material has taught me that dry air is the best 
insulator that one can obtain at moderate expense; and next to it is kiln-dried wood 
soaked in parafhne. The latter, however, must not he exposed to acid fumes. 
The scheme of electrical switches for this experimental plant has been the tubject 
of much thought. It was desirable that a range of voltage extending from the con- 
tents of one of the unit trays — 60 cells, approximately 120 volts — through all the 
multiples of this unit up to 40,000 volts, could be obtained. At first the cells were 
charged in multiple with a current of 60 volts, the cells being arranged in multiple, 
20 cells in each derived circuit, the charging current through each cell being one- 
thirtieth of an ampere. 
The switchboards are placed on the back of each of the unit cabinets, there being 
eight of these cabinets. The switches are three point switches (Figure 2, Plate XX V). 
When they are thrown to the right, the batteries are in multiple for charging; when 
thrown to the left, the cells are in series. An entirely similar switchboard serves to 
throw each cabinet into multiple or tandem. 
It is evident that with a charging current of only 60 volts a large number of 
switches are necessary. The following arrangement, suggested to me by Dr. II. W. 
Blackwell, who has studied the disposition of the cells for use in his work on electric 
double refraction, has greatly simplified the arrangement of switches, and lessened the 
work of throwing the cells from multiple to series: 
A 
f 500 volts is now used. The cells in each cabinet 
d in multiple, 60 in each shelf, 3 
420 volts. This change merely 
equires throwing a one point switch on the switchboard, the cells remaining in 1 
,nd being divided at a middle point, giving 210 cells in each deri\ d branch. ^ 
uitable transformer and a mercury rectifier it is evident that the number of sv 
