124 GLACIERS OF THE CANADIAN ROCKIES AND SELKIRKS. 



surface, began before the cover was completely formed and hence before 

 such pressure could have come into operation. As pointed out by Miigge 

 they also assume this position when an opening through the ice cover is 

 artificially maintained. Miigge believes that the plates are simply floating 

 in their position of equilibrium and that the pressure has nothing to do 

 with the orientation of the plates. That this, however, is not the cause of 

 the orientation is shown by the position of the columns in the ice dykes 

 above described, where the plates have formed in a vertical position, while in 

 the case of the ellipsoidal water-filled cavities they have formed at all angles 

 between the vertical and the horizontal. In the case of the ice dykes the forma- 

 tion of an ice cover would have given rise to a lateral pressure, as well as an 

 upward one, and the position of the plates in the horizontal columns would have 

 been in harmony with the view of Drygalski. No trace of this cover, however, 

 was seen, and it seems probable that the columns would still have formed at 

 right angles to the cold walls of the crevasse, under none other than hydrostatic 

 pressure. 



Some experiments still in progress in the freezing of water in variously shaped 

 vessels lead the author to believe that the basal plates are placed parallel to the 

 surface of refrigeration, independently of pressure or position of equilibrium. 

 The actual congealing temperature enters quiet water at right angles to this 

 freezing surface, regardless of its position, and each successive plane of mole- 

 cules in turn feels the effect of the crystallizing force. The result is that sheets 

 of molecules are successively frozen parallel with the requisite isothermal surface 

 as it slowly works its way into the body of the water. The orientation of the 

 plates is facilitated by the fact that in making the ice crystal the molecules ar- 

 range themselves more readily (because of the superior crystallizing force) in 

 the plane of the secondary axes than in the direction of the principal axis. This 

 is shown by the form of the snowflake which has been produced supposedly under 

 conditions in which the crystal was free to grow in any direction, so far as the 

 supply of moisture and suitable temperature are concerned. As is well known 

 the molecules are arranged mainly about the short main axis in the plane of 

 the secondary axes. The principle is illustrated further by the frost crystals which 

 form upon the window-pane, with cold air upon one side and a relatively warm, 

 moist atmosphere upon the other. At first only a very thin layer of moisture, 

 parallel with the surface of the glass, can congeal, and in this layer the molecules 

 at once arrange themselves in the plane of the secondary axes. As the atmos- 

 phere supplying the moisture becomes cooled for some distance back from the 

 glass the crystals may grow more or less irregularly. That the cohesive force in 

 the ice crystal is much more powerful in the direction of the basal planes than 

 in the direction of the principal axis, is demonstrated in the experiments to be 

 noted later (p. 130). Pressures in a direction at right angles to the main axis 

 will cause the basal plates to slide over one another, as in a bunch of tickets, 

 but no such shearing action can be secured when the direction of pressure is par- 

 allel with this axis. According to the view of the writer the temperature con- 



