SOME PHYSICAL PROPERTIES OF ICE 



621 



the rate of i . 76 gm. per minute was observed before it was used 

 for the series of experiments on torsion by constant force in which 

 the effect of elastic fatigue, as previously stated, was observed in 

 pronounced form. The other two curves were obtained after that, 

 and hence it is possible that the flatness of the first curve com- 

 pared to the curve for .88 gm. per minute, which should not be 

 the case, may be due to fatigue. 



o I 234567 



Torsion in degrees 



Fig. 8. — Torsion curves, by increasing force, of the second bar used for torsion 

 by constant force. 



The curves in Figure 7 were obtained after the observations for 

 the torsion by constant force. We have seen before that the test 

 piece was already subjected to fatigue in that case, and consequently 

 the present curves will be affected in the same way. As a very 

 rough approximation, therefore, we may compare the curves for 

 the rates of .88 gm. and 44.0 gm. per minute in Figure 8, assum- 

 ing the effect of fatigue to be the same after it has abruptly appeared 

 and also neglecting the effect of temperature, which did not differ 

 very much. The former curves are so much steeper that this 



