334 Messrs. J. C. McConnel and D. A. Kidd. [June 21, 



neighbouring grains seem to be arranged quite at random. Owing to 

 the structure being so complex, we failed to trace any relation 

 between the arrangement of the crystals and the rapidity of extension. 

 It is true that the most rigid piece of the three was composed of 

 small crystals, while the most plastic contained one very large crystal ; 

 but this was perhaps accidental. Fortunately, we were able to 

 obtain ice of a more regular structure, which has already thrown a 

 little light on the action at the interfaces of the crystals, and offers an 

 attractive field to further investigation. 



Some of the ice of the St. Moritz lake is built up of vertical 

 columns,* from a centimetre downwards in diameter, and in length 

 equal to the thickness of the clear ice, i.e., a foot or more. A hori- 

 zontal section, exposed to the sun for a few minutes, shows the irregular 

 mosaic pattern of the divisions between the columns. The thickness 

 of each column is not perfectly uniform. Sometimes indeed one thins 

 out to a sharp point at the lower end. Each column is a single 

 crystal, and the optic axes are generally nearly horizontal. Some 

 experiments on freezing water in a bath, lead us to attribute this 

 curious structure to the first layer of ice having been formed rapidly, 

 in air, for instance, below — 6° C. We found that if the first layer had 

 been formed slowly, and was therefore homogeneous with the axis 

 vertical, a very cold night would only increase the thickness of the 

 ice, while maintaining its regularity. 



We applied tension to a bar of lake ice carefully cut parallel to the 

 columns. It stretched indeed, but excessively slowly. During seven 

 days it stretched at the rate of only 0*0004 mm. per hour per length 

 of 10 cm., though at one time the temperature of the surrounding 

 air went up above zero. The tension was 2 kilos, per sq. cm. This 

 slight extension may well be attributed to the tension not being 

 exactly parallel to the interfaces of the columns. This experiment 

 corroborates our first result, that a single crystal will not stretch at 

 right angles to its optic axis. We next cut a bar at about 45° to the 

 length of the columns, and the difference was very manifest. During 

 80 hours under a tension of 2" 75 kilos, per sq. cm., it extended at the 

 rate of 0*015 mm. per hour per length of 10 cm., nearly 40 times as 

 fast. 



An icicle is an example of ice formed of very minute crystals 

 irregularly arranged. We found that an icicle under a tension of 

 2*2 kilos, per sq. cm. stretched at the rate of 0*003 mm. per hour per 

 length of 10 cm. This is very slow, especially as the temperature 



* This was the case in all pieces obtained from one end of the lake, where men 

 were cutting ice for storage purposes, whether new ice or old. In a part, however, 

 which had frozen a few days earlier, further out from the shore, we found mucli 

 larger crystals with the axes nearly vertical but not quite parallel to each other. — 

 July 6, 1888. 



