1879.] Aqueous Solutions ivith Mercurial Electrodes. 473 



it. KKKKisa wooden frame for supporting the apparatus, and L 

 is a weight to keep it steady. This kind of apparatus was used in all 

 the experiments, and acted very satisfactorily. I devised one of 

 different construction for a similar purpose many years ago. (See 

 " Phil. Mag." June, 1859.) 



Liquid Thermo -Pair with Mercury Electrodes. 



The mercury employed was prepared by distilliug the ordinary 

 metal twice, then shaking it strongly in succession with dilute nitric 

 acid, water, solution of cyanide of potassium, distilled water, and 

 filtering it. Distilled water was used in making all the solutions. 



In using the apparatus, the mercury was first poured into the 

 basins ; the tubes E and F placed, taking particular care that the 

 mercurial column in F and its contact with the pool were complete. 

 Two very shallow india-rubber bungs were next inserted in the lower 

 ends of the bent tube C, the tube then put in position, and filled with, 

 the liquid. The solution was now put into the basins ; the bung D 

 inserted tightly, the shallow ones then removed, and the bungs in- 

 serted in A and B. The completeness of the circuit was next tested 

 by attaching a galvanometer to the instrument, and, if necessary i & 



2 k 2 



