Ice as an Electrolyte. 121 



very great^ increase in conductivity occurred at about 1° C. ; 

 it is possible, however, that the 'apparently corresponding 

 points of maximum temperature and maximum conductivity 

 shown in the curves (fig. 2) may not have really exactly cor- 

 responded, although changes in the temperature always ac- 

 companied changes in the current. 



The curves R 8, T V do not appear to be logarithmic. Tlie 

 coordinates of the points from w^hich they have been drawn 

 are given in the following Table : — 



Temperature, Eesistance per 



in degrees cubic centimetre, 



Centigrade. in megohms. 



-12-4 2240 



- 6-2 1023 



- 5-02 948-6 



- 3-5 642-8 



- 3-0 569-3 



- 2-46 484-4 



- 1-5 387-6 



- 0-2 284-0 



+ 0-75 118-8 



about + 2-2 24-8 



+ 4-0 9-1 



+ 7-75 0-54 



+ 11*02 0-34 



V. The copper disk GH (fig. 1) was now removed and 

 replaced by a disk of zinc of exactly the same size. The 

 pieces of glass used to separate the zinc disk from the bottom 

 of the copper box were the same as those previously employed 

 being 0-324 centim. thick. Distilled water was poured in so 

 as to cover the zinc plate, and the cover was cemented on 

 the box as before. In the following experiments, K J (fig. 1) 

 was a strip of zinc cut out of the same sheet as the disk Gc H 

 and bent up : the strip K J and the disk G H were therefore 

 continuous without joint. The point J w^as joined to one of 

 the electrodes of a Thomson's quadrant electrometer, and the 

 copper box D C to the other electrode. Both with water and 

 with ice the maximum electromotive force obtained was one 

 volt about. 



In the first sets of experiments. A, B, C, the zinc disk and 

 copper box were alternately joined together and insulated from 

 one another, the result of the short-circuiting being, of course 

 to^ diminish the electromotive force by polarization. While 

 this was being done the temperature was gradually low^ered. 



