1891.] On the PlasticAtij of an Ice Crystal 325 



glass. There were two centres of colour encircled by irregular rings, 

 and these remained much the same when the two faces through which 

 the light passed were rubbed quite flat and the other crystals cut 

 away. There could be no doubt that this crystal had suffered some- 

 thing more than mere elastic distortion. 



The next experiment was very instructive. A thin slip of ice, 

 being a single crystal, was subjected to bending stress as before, and 

 left for several hours. It apparently bent very quickly, for after a 

 few hours it was found crescent shaped, and luckily unbroken, lying 

 at the bottom of the box. The optic axis was bent, and, though its 

 change of direction was rapid where the bend was sharp, there 

 appeared to be no break in continuity. On the other hand, the long 

 narrow bubbles, which were originally no doubt parallel to each 

 other and perpendicular to the slip, were still parallel to each other 

 throughout. In fact, as I noted at the time, the crystal behaved as 

 if it consisted of an infinite number of indefinitely thin sheets of 

 paper, normal to the optic axis, attached to each other by some viscous 

 substance which allowed one to slide over the next with great diffi- 

 culty. This comparison proved to be the key to the whole question 

 of the plasticity of a crystal of ice. 



Further experiment showed that if a bar of ice consisting of a 



single crystal with the axis perpendicular to two of the side faces was 



subjected to bending stress, it would bend freely in the plane of the 



axis either at or below the freezing point, but not at all in a plane 



perpendicular to it. In the bent crystal the optic axis in any part 



was normal to the bent faces in that part. But any series of lines 



drawn in the substance of the ice which were originally parallel to the 



iptic axis and to each other remained parallel to each other, though 



oot, of course, to the optic axis. This was evidenced by the position 



)f long narrow bubbles which frequently form at right angles to the 



)lanes of freezing, and also by the end faces of the bar remaining 



parallel to each other. When the optic axis was longitudinal, the bar 



>ent indeed, but not very readily, and the general behaviour was 



nore obscure. Still, this case, too, was in satisfactory agreement 



vith the analogy mentioned above. 



Let us state this analogy more fully. The sheets of paper offer no 

 esistance to bending, but utterly refuse to stretch except, of course, 

 lastically. Initially they are plane and perpendicular to the optic 

 xis, and, after they have been deformed by bending, the optic axis jit 

 ny point is still normal to the sheet at that point. They are of 

 niform thickness, whence it easily follows that the directions of the 

 tic axis in any crystal form a series of straight, though not parallel, 



