MI;, s. \v. .1. SMITH ox TIII-; NATIIM: I>F KLKITKOTAPILLABY PHENOMENA. 85 



I nit the surface tension increased when the contact broke, showing that the number 

 given convspoiidfil to ;i point on tin- drs-nitlm^ bnmch of the KC1 curve. The 

 horizontal distance between the maximum and 30*5 on the KC1 curve is approxi- 

 mately the same as in the former case. It corresponds to a potential difference of 

 about "24 volt. Both results are somewhat uncertain, partly owing to the difficulty 

 of determining the exact points of maximum surface tension on the capillary curves; 

 but as the effects observed are so considerable, and as they are further corroborated 

 by the results described below, there can be little doubt concerning their significance. 

 The results can be further tested by experiments similar to PASCHEN'S by measuring 

 the KM.F.s of cells of the type 



Hg 



KI KC1 



I- 



in which the vertical arrow signifies the dropping electrode. The following approxi- 

 mate results were obtained ; 



Hg 



Hg 



Hi: 



Hg 



KC1 







KC1 





-KCl 



KI 



KI 



,, "I 



= -522 volt. 



= 784 



= '383 



= '127 



The first and second results give '262 volt and the third and fourth give '256 volt 

 as the difference between the potentials 



n 



and [fKCl 



JH 



The difference between the chemical effects during the dropping of mercury into 

 * I have found, since this was written, that G. MEYER ('Wied. Ann., 1 vol. 56, 1895) measured the 



F..M.F. of a cell of the type nKCl 



Y 

 above :iru in accord with this result. 



iuul found it to be '284 volt. The observation* tfetcrilwd 



