1891.] 



On the Plasticity of an Ice Crystal. 



331 



expression for PV if it had been normal to the curve P'Q' at 

 P' and QV ha,d been drawn perpendicular to it, and PP' is normal to 

 the curve P'Q' at P'. 



Exp. 8. This was an experiment on a bar composed of three 

 crystals designed to investigate the action at the interfaces of crys- 

 tals. The bar bent a good deal, but nearly the whole bend occurred 

 in the middle of one of the crystals. I had cut nicks in the sides of 

 the bar to test for migration of the interfaces within the ice, but 

 found none. It appears, in fact, that the interfaces do not in any 

 way assist the plasticity, but hinder it by fettering the sliding of the 

 layers in the separate crystals. 



Exp. 9. Out of some thick ice formed on the surface of the 

 water in a foot-bath I cut a bar which was all one crystal. When 

 the bar was in position the optic axis was horizontal, and inclined at 

 about 60 to the length of the bar. 



Breadth, 13 mm. ; depth, 1T7 mm. ; distance between supports, 

 38 ram. 



A weight of 1'29 kilos, was applied for 16| hours from 4.50 P.M. 

 on January 29 to 9.15 A.M. on January 30, during which time the 

 maximum temperature was 5 C., the minimum 12'7 C., and the 

 mean about 8'6 G. The depression of the middle was 1*9 mm. A 

 little consideration will show that by the theory of the sliding layers, 

 the upper and lower surfaces of the bar should be bent in such a 



inner as to still contain straight lines perpendicular to the optic axis, 

 some such deformation was observed, but it was not very definite. I 

 noticed that the numerous bubbles which were originally parallel to 

 the axis were still parallel to the upper and lower surfaces in their 

 neighbourhood. 



I now set up an arrangement for obtaining more accurate measure- 

 ments of the rate of bending. A large square aperture in an iron 

 plate was bridged by a curved iron bar rigidly attached to the plate. 

 The bar of ice was placed across the aperture. A loop attached to 

 the curved bar supported a wire lever, of which the long arm served 

 as a pointer on a scale, and the short arm carried a stirrup which 

 embraced the ice. When the bar bent the stirrup was depressed 

 md the pointer raised about twenty- eight times as much. 



This part of the apparatus was placed in a cigar box, at one end of 

 which the pointer projected through a slit, while there was a hole in 

 the bottom to allow the string, to which the weight was attached, to 

 pass through. The Six thermometer was on a level with the ice, 

 and could be read by gently lifting the lid without disturbing any- 

 thing. The mirror and scale with which the position of the pointer 

 was read were fastened to the box. 



The only part of the stirrup that touched the ice was the flat piece 

 of tin at the bottom. This was slightly roughed and made flat, so 



