some Mercury Resistance Standards^ 71 



The mercury with which the tubes were filled in 1885 has re- 

 mained in them since that date. Now recent observations had 

 seemed to indicate that a small change had taken place in some 

 of the platinum-silver standards, and it became desirable to 

 compare them again with M. Benoit's tubes. The method 

 was the same as that of my former paper. The tubes im- 

 mersed in melting ice were compared with the B.A. standards. 

 The temperature of the room was kept very low, from 0°*5 to 

 3° C, and thus the errors caused by the conduction of heat into 

 the tubes through the copper connecting rods were avoided. 

 The connecting pieces used to connect the tubes to the bridge 

 differed from those employed in 1885, being a modification 

 of the platinum cups described in my paper * on the specific 

 resistance of mercury. A hollow platinum cup, about 3*5 cm. 

 long and rather more than 1 cm. in diameter, is secured firmly 

 into an ebonite tube ; the outside of the tubes are turned to 

 fit the ground portion of the glass vessels which form the ends 

 of the mercury tubes, thus taking the place of the stoppers 

 which usually close the tube ; the platinum cups dip into the 

 mercury, the surface of contact being about 12 sq. cm. The 

 cups had previously been platinized ; on the present occasion 

 they were merely cleaned with nitric acid and distilled water. 

 Stout copper rods, well amalgamated, fit tightly into the inside 

 of the cups ; the other ends of these rods are in contact with 

 the bridge. Thus the connexion with the bridge is made 

 through the rods and across the platinum of the cups, avoiding 

 contact between the mercury and copper. 



The resistance of the connexions w 7 as determined, and it was 

 shown that the resistance of the platinum was negligible. 

 This was done by making the rod and cup part of a circuit, 

 the resistance of which was measured. The rod was then 

 removed from the cup, and the latter was removed, so that the 

 part of the copper which had been in the cup was now in direct 

 contact with the mercury into which previously the cup had 

 dipped. No change in the resistance of the circuit was produced 

 by this. To secure this result it was necessary that a consider- 

 able area of the platinum should dip into the mercury ; and 

 this condition was in the experiments always carefully at- 

 tended to. The resistance of these connecting-pieces was thus 

 found to be -00291 B.A. unit. 



The tubes were compared both with the B.A. units and also 

 with the Legal- Ohm standards. In the comparison with the 

 B.A. units, a large amount — about 300 divisions, or '015 B.A. 

 unit — of the bridge-wire was used. Since the resistance of the 



* Phil. Trans. 1888. 



