20 PROCEEDINGS OF THE AMERICAN ACADEMY. 



matter by the analysis of the probable causes of P reveals that this 

 quantity may be dependent upon several ultimate factors, — such as the 

 changing solution pressure of the chlorine ions due to changes in tem- 

 perature and the already detected formation of mercuric chloride, the 

 changing solubility of mercurous chloride with the temperature and the 

 presence of other chlorine ions, as well as the possibly varying "heat of 

 formation " of mercurous chloride. Moreover, a modifying circumstance, 

 entirely disconnected with the electrode itself, lies in the possible genera- 

 tion of potential from the contact of the warm solution with the cold 

 solution. That these influences are subordinate ones, except in the cases 

 of hydrochloric acid and ammonia, is indicated by the above tables ; but 

 further explanation must carry one too far into the region of hypothesis, 

 and had better be postponed until a more satisfactory experimental basis 

 is established. 



The results of this paper may be summed up as follows : — 

 (a) The temperature coefficient of the potential of the calomel elec- 

 trode increases with the dilution of the electrolyte. 



(h) The kation of the electrolyte influences the result by affecting the 

 degree of dissociation of the chloride in solution. 



(c) Both of these effects may be approximately computed by a simple 

 logarithmic formula based upon Nernst's hypothesis. 



(d) The accuracy of the results is, however, affected by at least one 

 important modifying influence, the "catalytic" decomposition of mercu- 

 rous chloride into mercuric chloride and mercury. 



(e) This side reaction is responsible for the slight inconstancy of the 

 normal calomel electrode. 



(y) The " decinormal electrode " is much more uniform in its be- 

 havior than the normal, and hence for some purposes might be a more 

 useful means of measuring potential differences. 



(g) Hydrochloric acid and ammonic chloride are anomalous in their 

 behavior. 



This research was begun in the Laboratory of Physical Chemistry at 

 Leipsic, upon the suggestion of Professor Ostwald. It was continued by 

 his kind consent at the Chemical Laboratory of Harvard College, and I 

 hope that some of the problems left unsettled may be solved by future 

 work here. Especially is it my intention to apply the differential equa- 

 tions to other cases, such as an electrode of zinc in a zincic salt. 



Cambridge, Mass., May, 1897. 



