342 



On the Plasticity of an Ice Crystal. 

 FIG. 10. 



[Mar. 12 



Let be the depression of the middle of the bar, \ the angle either 



half makes with the horizontal, 

 small, 



and 



We have s = -v. 

 2* 



When x > 





IVfdt 



(3). 



This gives the coefficient of plasticity in terms of the unsupported, 

 length of the bar, the weight per unit area of cross section, and the? 

 observed rate of depression. We have employed equation (2)j 

 which is strictly applicable only when the bar is straight and hoi 

 zontal. But, in the cases to which we have to apply these results, \ ^ 

 so small that the error is negligible. It was hardly worth whil 

 calculating the numerical value of p, especially as it has been shoi 

 to depend on the temperature, on the value of U nearly, and 

 on the previous history of the bar. But the above investigation 

 assist any one in estimating, as far as can be done from my expei 

 ments, the rate of distortion of an ice crystal in any given case. 



In several cases in the experiments, after a heavy weight was 

 removed, a slight gradual unbending of the bar took place. At first 

 I thought this a mere consequence of the irregular elastic strains ou 

 the bar, the parts most severely strained gradually bending back the 

 rest. But the magnitude of the recovery seems, on closer examination, 

 to put this explanation out of the question, and I have now little 

 doubt that it is a true molecular effect. 



In Exp. 12, after a stress of 1*69 kilos, per sq. cm. had been 

 removed, the middle of the bar rose O'Ol 04 cm. in four hours. Accord- 

 ing to an experiment by Moseley (' Phil. Trans.,' 1871), Young'* 

 modulus for ice is 92,700 kilos, per sq. cm. Hence, if we neglect 

 the effect of the plastic strains in one bar of ice, the elastic depres- 

 sion under 2'5 kilos, should have been 0'00138 cm., less than oue- 

 seventh of the recovery observed. The permanent or plastic strains 

 in Moseley 's bar are considerable, so that the deduced value of 

 Young's modulus may be too great. Bevan, also by flexure of baw 



