Ice as an Electrolyte. 123 



obtained at the end of the first two minutes of insulation, to 

 155, obtained at the end of the hist two minutes of insulation. 

 Exactly the same experiment being repeated ten times with 

 ice at a constant temperature of about —17° C, the deflec- 

 tion 212, obtained at the end of the first two minutes of insu- 

 lation, was only reduced to 203 at the end of the last two 

 minutes of insulation. It would therefore appear from this 

 set of experiments, that the smaller amount of polarization 

 produced by the high resistance of the ice only allowing a 

 small current to flow during the short-circuiting of the zinc 

 plate and copper box was of rather more importance than the 

 solidity of the ice retarding dissipation of polarization. 



VI. Further experiments on the Specific Inductive Capacity of 

 Ice. — The zinc disk G H (fig. 1) was now removed and the ori- 

 ginal copper disk replaced. With a charging electromotive 

 force as small as 0*174 volt, we found that, in the case of ice, 

 it was always necessary to carefully correct the swing for the 

 opposing permanent electromotive force existing between the 

 copper disk and the copper box, whereas with water at about 

 + 10° C, when using the same charging electromotive force, 

 it w-as not necessary to make any such correction. Rough 

 measurements gave this opposing electromotive force at 

 — 12°*4 C. as 0'017 volt, and showed that as the temperature 

 rose there seemed to be a regular decrease : at 0° C. it was only 

 0*003 volt, and at higher temperature it was immeasurably 

 small by the rough method we employed. To avoid this cor- 

 rection being of so much importance, we used an electromo- 

 tive force of 0*87 volt to determine the capacity of ice at dif- 

 ferent temperatures. We did not find that the time of char- 

 ging or the time of short-circuiting aflected our results nearly 

 as much as in the case of water ; in fact, the time of short- 

 circuiting our ice condenser between successive tests of its ca- 

 pacity hardly affected the results at all ; but during a series 

 of observations, the time of short-circuiting between every 

 two being fifteen seconds, and the time of charging being 

 gradually increased from five seconds up to thirty, the swings 

 increased from 56 to 70 divisions of the scale. We therefore, 

 as on the former occasion, always charged for ten seconds 

 and short-circuited for fifteen. Four or five consecutive ob- 

 servations of this kind always showed a slight increase from 

 the first to the last, the temperature being kept quite constant. 

 Several series at difierent temperatures are shown in the fol- 

 lowing Table; — 



