142 



THE ELECTRICAL RESISTANCE AND 



tube in the direction of its length,, although the tube was 

 enclosed in three separate vessels, in a refrigerator and closely 

 packed with felt. The junction of the thermo-couple was 

 enclosed in a capillary-tube, (see Fig. 7), so lihat it could be 

 placed inside the "potential-terminal-tube/' c l (see Fig. 5). 

 A reading of the galvanometer was taken when the end of the 

 capillary tube reached the end -of the "potential-terminal-tube." 

 The capillary tube was then drawn up about 3 centimeters or 

 so and another galvanometer reading was made. The average 

 of these two readings wag taken as the average temperature 

 of the ice between the potential terminals. 



When the water was frozen, cracks were observed through 

 the ice in the "IT" tube, and it was found impossible to obtain 

 ice at low temperatures, by freezing in the tube, without the 

 cracks 



These cracks may have a considerable effect on the 

 resistance of the ice. So the accuracy of the values of the 

 specific resistance as given in this paper, is limited by this 

 uncertainty. 



F 



t- 



I?' 



1; 



7 



PIG. 8. 



