222 [Dec. 12, 



the liquids, consisting of currents of the electrolyte which flow from 

 the positive to the negative electrode, and of the mercury which flow 

 in an opposite direction ; these I shall leave out of consideration, as 

 they are only secondary results. 



12. Nearly all the experiments described in this paper were made 

 with the solution of cyanide of potassium and mercury already 

 described (3). 



Influence of the Electrodes. 



13. The crispations and sounds never (or very rarely, and under 

 conditions which I have not recently met with) occur at the anode ; 

 nor had the mass, size, form, or position of surface, distance, che- 

 mical composition, or physical condition of the anode any effect upon 

 the direction of the rows of elevations, nor any direct effect (their 

 indirect effects will be described hereafter) upon the sounds. The 

 elevations and sounds occurred with anodes of all sizes and shapes, 

 and in all positions ; also with anodes of platinum wire and sheet, 

 A ring cathode of mercury within a ring anode of mercury gave the 

 sounds equally well. By using a straight and narrow strip of mer- 

 cury as anode inside a narrow ring cathode, no difference in the size 

 of the crispations at different parts of the ring occurred. 



14. With regard to the influence of the cathode, the sounds 

 occurred as well with a very shallow layer of mercury as with a deep 

 mass. Sounds were obtained with nearly as great facility in the central 

 pool (if not too large) by making that the cathode, as in the annular 

 portion. The direction of the rows of elevations was dependent upon 

 the form of the boundary outline of the negative mercury surface ; a 

 rectangular surface gave rectangular rows, and a circular surface gave 

 circular rows, and in each case the rows were parallel to the boundary 

 of the mercury surface : a small shallow ring of gutta percha held 

 upon the surface of a large rectangular cathode of mercury caused 

 the rows of crispations to be circular inside the ring, whilst the rows 

 outside remained rectangular. A convenient apparatus, though a 

 temporary one, for examining the direction of the rows of crispations 

 was made by joining together rectangular pieces of thin sheet-glass 

 by sealing-wax in the form shown in fig. 6. The vessel is 3 inches 

 long, 2 inches wide, and 1 inch deep; its lower part is divided into 

 three equal portions by two strips of glass, A and B, about th of an 

 inch wide, so as to form spaces Q D, and E for three pools of mer- 

 cury ; a vertical slide of glass, F, rests by its lower edge upon the 



