414 PROCEEDINGS OF THE AMERICAN ACADEMY. 



Regarding the various straight lines as practically parallel within the 

 range of our experiments, and giving to A its average value, we may ex- 

 press all our results approximately by the equation E= .0160 log 10 /+ B, 

 where B varies between the extreme values 0.014 and 0.069, E being 

 given in volts and /in millionths of a milliampere. 



On account of the temporal changes in B it was difficult to find how 

 this quantity changed with the temperature, for the values obtained on 

 different days at different temperatures differed hardly more from one 

 another than those obtained on different days at the same temperature. 

 Attempts to solve this problem by suddenly changing the temperature 

 back and forth failed because we found that after a change of temperature 

 several hours at least were necessary for the establishment of constant 

 conditions. The effect of temperature on the polarization cannot be 

 large, and on the whole our experiments indicate, if anything, that at 

 a given potential the current is smaller the higher the temperature. 



This result appears at first sight opposed to the general law that the 

 velocity of jmysico-chemical reactions increases rapidly with the tempera- 

 ture, but it must be borne in mind that this law holds only for given 

 concentrations of the reacting substances, while, when the potential is 

 kept constant and the temperature raised, the concentrations of the 

 reacting substances may diminish very considerably. 



We have not yet mentioned an interesting phenomenon which ap- 

 peared when the potential was carried much higher than 0.7 volt. At 

 these high potentials the current would frequently begin to fluctuate, and 

 at the same type the hydrogen, instead of making a single bubble at the 

 top of the mercury meniscus, would form a number of small bubbles 

 which appeared not only at the surface of the mercury but also between 

 the mercury and the tube, sometimes as much as a centimeter below the 

 surface. This is net such a break in the polarization curve as we have 

 mentioned above, which would indicate the sudden formation of some 

 new product of the electrolysis, for it is not associated with a sudden rise 

 in the current, but on the contrary usually brought the current down to 

 an abnormally low value. 



Probably this phenomenon is due to changes in the surface tension 

 of the mercury produced by the high polarization, but it might be ex- 

 plained by the formation of an unstable amalgam of hydrogen, and this 

 hypothesis should not be discarded until further experiments have been 

 made. 



It will be seen that the experiments which we have described in this 

 paper have presented a number of new problems rather than solved the 



