318 ILLINOIS STATE ACADEMY OF SCIENCE 



To facilitate the changing of connections a multiple 

 switch was constructed (Fig. 2) which accommodates 5 

 trays, the number assembled on each shelf of the cup- 

 board. The terminals of each tra^^ are brought to a pair 

 of mercury cups, as indicated in Fig. 2. The positive 

 terminals are connected to the front roAV of cups, and 

 the negative to the back row. The leads from the charg- 

 ing source are connected to the two end cups as shown. 

 Two types of connectors are used which consist of copper 

 or brass legs attached to a block of hard rubber. The 

 legs are arranged to set in mercury cups. The one type 

 has all of the legs on the front side connected together, 

 and likewise oil the back so that when it is inserted the 

 trays are all connected in parallel, — positive all together 

 on the front and negative on the back. The other type 

 has the legs connected across diagonally so that when 

 the connector is in place the positive terminal of one tray 

 is connected to the negative terminal of the next and so 

 on. One multiple switch is provided for each shelf of five 

 trays. In order to connect the different shelves together 

 in series diagonal cross wires are used. These wires 

 show in Fig. 3. 



The particular* installation which is the subject of this 

 paper, consists of 1500 cells, or 75 individual trays as- 

 sembled in three tiers of 25 trays each. The total po- 

 tential when all are in series and fully charged is 3000 

 volts. In order to reduce any leakage to a minimum 

 value, the five trays on each shelf are supported on an 

 individual paraffined shelf which in turn is supported by 

 four porcelain insulators. In this way each 200 volts is 

 thoroughly insulated from the cabinet as a whole. An 

 idea of this method of assembling may be obtained from 

 Fig. 3. 



In building such a battery, it is, of course, not essential 

 that tlie cells be grouped as described. Almost any num- 

 ber of individual cells may be incorporated in one tray, 

 but it seems that for maximum convenience twenty cells 

 to a tray is about the right number, for if more are added 

 the tray becomes large and heavy. 



