594 



PROCEEDINGS OF THE AMERICAN ACADEMY. 



tains a similar and nearly equal capacity Cj and the variable resist- 

 ance R. Ci and C2 are rolled paraffin paper condensers of about 22 

 microfarads each, composed of 10 units of 2.2, mf. connected in 

 parallel by No. 11 copper wires soldered at each point and also to 

 the bridge connections. The entire wiring of the bridge was non- 

 inductive. 



The cell was maintained at constant temperature in a thermostat 

 and was connected by soldered No. 11 wires terminating in a heavy 

 mercury-metal switch-block so arranged that the cell coald be placed 



Figure 2. Bridge connections for measuring low electrolytic resistances. 



in circuit, or cut out, without change in the block resistance and with- 

 out short-circuiting the cell. The ^-inch wires of the block switch 

 were kept continually under mercury whichever way the switch was 

 turned, thus keeping contacts bright when the cell was out. 



The variable resistance R consisted of two parts: One of two parallel 

 wires shunted by a heavy sliding block, and. the other, in series with 

 this, was a vertical U-tube of glass containing mercury, as shown in 

 Figure 3. The connecting wires dip into the mercury in the two arms 

 and the cross-section of one arm was decreased or increased by lower- 

 ing or raising the glass rod. This smoothly variable resistance proved 

 to be of the greatest aid in obtaining the close settings desired. 



In parallel with R was a box Wheatstones bridge and galvanometer. 

 R could be thrown over into the A^-arm of this bridge by means of a 

 mercury switch, but as the box bridge had a resistance of about 

 1000 ohms it was found unnecessary to cut it out while making 

 settings on the cell, and during the later measurements the switch was 



