On the Viscosity of Ice. 



495 



deduced from three measurements of the diameter, at the top, in the 

 middle, and at the bottom, differed from the maximum stress, as 

 deduced from the total load and the least area of section. 



The last column but one in Experiment No. 1 gives the minimum 

 temperature the preceding night, measured, not in the box, but by a 

 thermometer freely exposed out of doors. Since the ice specimen was 

 enclosed in the double box its temperature did not sink as low as those 

 given in this column. The numbers here are only to be taken as sig- 

 nifying that the night was or was not specially cold. Unfortunately 

 in this experiment no maximum and minimum temperature observa- 

 tions were taken. In Experiments 2 and 3 the maximum and 

 minimum temperatures were observed, when recorded, by a maximum 

 and a minimum thermometer suspended in the box, and their deter- 

 minations furnished "the range of temperature." 



Experiment 1. 

 Mean original length of specimen not subjected to 



tension, measured at 9.15 on Feb. 7th = 233'48 mm. 



Mean length with load of 5'5 kilos = 234*22 „ 



Giving a sudden increase in length of 0*74 ,, 



Measured just afterwards it had become = 234*58 ,, 



After 13 hours (at night) under same load = 234" 98 „ 



Giving a mean hourly extension of 0*03 ,, 



After 2 hours (temp. —5*0°) the length became = 235'04 ,, 



Giving an hourly extension of - 02 ,, 



The load on the specimen was then increased to 12*5 kilos., and the 

 mean length became 235*12 mm. 



VOL. XLIl. 



