570 MR. F. E. SMITH, MR. T. MATHER, AND DR. T. M. LOWRY 



through the electrolyte in the syphon at high temperatures. The heat produced by 

 the passage of such a current raised the temperature of the liquid sufficiently to 

 vaporise a portion of it and so break the circuit at the bend of the tube. 



Our results are not in agreement with those of most other observers. Lord 

 RAYLEIGH* found a higher deposit at 50 C. than at 15 C., and a higher deposit at 

 ] 5 C. than at 4 C. The temperature coefficient was therefore positive and averaged 

 about O'OOl per cent, per I 8 C. for the range 4 C. to 50 C. LEDUC found a negative 

 coefficient. RICHARDS, COLLINS, and HEIMROD obtained at 60 C. and also at C. a 

 larger deposit than at 20 C. They state, however, that the apparent gain at C. 

 was undoubtedly due to the difficulty in washing the deposited silver. The tempera- 

 ture coefficient obtained from their results is about O'OOl 7 per cent, per 1 C. for the 

 range 20 C. to 60 C. 



MERRILL, t who experimented with fused silver nitrate and solutions at normal 

 temperatures, concluded that temperature has no effect on the mass of the deposit. 



Liquid Inchisions in the Deposit. 



The appearance of the deposits in the platinum bowls varied with the solutions 

 used, but with pure solutions they were generally of very open texture. The 

 deposits were usually dried in an electric oven at 160 C., but in eight cases we 

 reheated at 240 C. without observing any loss in weight, and on three other occasions 

 some deposits in silver bowls were heated to over 400 C. In no case was there a 

 diminution in weight, but in two of the latter experiments a gain of 3 parts in 100,000 

 was recorded. This was possibly due to the formation of silver sulphide. 



Lord RAYLEIGH and Mrs. SIDGWICK* sometimes found no loss on a second heating, 

 but more often a slight decrease. RICHARDS^ found a mean loss of 18 parts in 100.000 

 when the deposits were reheated over an alcohol flame to constant weight. The most 

 extensive observations have been made, however, by VAN DIJK, who after washing 

 and drying at 150 C. reheated in an electric furnace to 500 C., and in some cases to 

 600 C. No loss in weight was observed. VAN DIJK used smaller bowls and 

 crucibles than RICHARDS, but the amounts of silver deposited by him are comparable 

 with those deposited by RICHARDS ; it is apparent, therefore, that the texture of the 

 deposits must have been different or that something in addition to silver was 

 deposited in RICHARDS' experiments. 



Size of Kathodes. 



In general the area of the kathode surface in our experiments was 200 sq. centims., 

 but in observation 58d the bowl L of 100 sq. centims. kathode surface was used, and 



* RAYLEIGH and SIDGWICK, 'Phil. Trans.,' 175, p. 411, 1884. 



t MERRILL, 'Phys. Rev.,' 10, p. 170, 1900. 



J RICHARDS, COLLINS and HKIMROD, 'Proc. Am. Ac.,' 35, p. 145, 1899. 



VAN DIJK, 'Ann. der Phys.,' 19, p. 266, 1906. 



