124 



On Ice as cm Electrolcjte. 





Temperature, in 



Swings, first and 



Time. 



degrees Centi- 



last of a series of 





grade. 



four. 



h m 

 9 20 



-§•5 



rei 



t61-5 



9 25 



-8-8 



r57 



9 30 



-77 



J 58 

 t60-5 



9 37 



-35 



r57 



162-5 



9 44 



-2-5 



roo 



161 



9 55 







r47-5 

 152-5 



In the above experiments the ice condenser always remained 

 short-circuited while the temperature was being raised between 

 each set of experiments. The results obtained appear to show 

 that the specific inductive capacity of ice does not materially 

 alter from — 9°*5 C. to — 2°'5 C. ; the apparent change 

 near the melting-point is very probably due to the increased 

 conductivity of the ice allowing much of the charge to be lost 

 before the discharge through the galvanometer. Our preceding 



experiments have shown that the conductivity of < ^^^ v 



increases regularl}' without discontinuity from — 10° C. to +10° 

 C, although altering very rapidly at 0°C., where the change of 

 state occurs. The specific inductive capacity, on the other 

 hand, appears to change very little while the dielectric (ice) 

 is solid, and perhaps also changes very little while the dielec- 

 tric (water) is liquid. It must therefore probably undergo a 

 great change at the melting-point, since we have shown that 

 the specific inductive capacity of water at 8°*7 C. is about 

 2240 times that of ice at — 13°*5 0. The specific inductive 

 capacity seems therefore to be a stress and strain phenomenon, 

 and to be intimately connected with the rigidity of the body ; 

 whereas the connexion between conductivity and rigidity 

 seems to be less marked. This, we think, bears out all recent 

 theories in molecular physics. The complete iuA^estigation of 

 the connexion between the conductivity and specific inductive 



capacity of -j ^.^^^^ I at different temperatures will form the 



subject of a second paper. 

 March 22, 1877. 



