154 R. M. Deeley and G. Fletcher— Structure of Glacier-Ice, 



of the usual glacier grains, many of considerable size. On the 

 average they were, as near as can be remembered, as large as peas 

 or even beans. Indeed, in many instances where the neve was 

 examined the ice was coarsely granular. Regarded from a distance 

 the neve appears to be very finely stratified, layers of comparatively 

 pure blue ice alternating with white ones. On close examination 

 this stratified appearance is seen to be practically wholly due to 

 the distribution in layers of countless imprisoned air-bubbles. Each 

 granule of ice, it must be remembered, is a truly crystalline, if 

 not a crystal, particle. As McConnell and Kidd say, " the optical 

 structure of each grain is found under the polariscope to be perfectly 

 uniform," but the " bounding surfaces are utterly irregular and are 

 generally curved." In spite, however, of this irregularity of outline, 

 and random arrangement of the optic axes, the lines of bubble? 

 traverse the mass in definite layers. Indeed, it is clear that the 

 granular structure of the neve is entirely independent of the dis- 

 position of the air-bubbles. After a fall of snow, surface melting 

 leads to the production of a mass of more or less spherical granule^ 

 of ice, the interstices between which are occupied by air. Further 

 accumulations of snow leads to pressure, the granules are com- 

 pressed, and much of the air may be expelled. But under certain 

 conditions of weather a surface layer of snow may be melted, and, 

 freezing again, may form an impervious layer, and the adjacent air- 

 bubbles be unable to escape, even under the pressure resulting from 

 further falls of snow. Thus we have bands of air-bubbles parallel 

 •with the surface and alternating with strata of blue ice which are 

 comparatively free from air. Meteorological conditions will, there- 

 fore, have a great influence upon the volume of air imprisoned. In 

 the Grindelwald we found the ice remarkably pure and blue as 

 compared with the ice coming down from the Mont Blanc range. 



One would suppose that any internal molecular rearrangement 

 would lead to the displacement of the air-bubbles, but this is not 

 the case. The small granular particles of ice referred to undergo 

 a rapid metamorphosis. They grow, some of them increasing in 

 size at the expense of others, until they may exceed one or two 

 inches in diameter, and would probably increase indefinitely but for 

 the stresses existing in the glacier which cause their fracture. The 

 mode of growth, and also the effects of fracturing, are discussed 

 later. It is remarkable that these changes in the shape and size 

 of the ice-grains do not appear to affect the arrangement of the air- 

 bubbles, the layers of which are extremely regular, the transference 

 of water in a molecular condition from grain to grain failing to 

 afiect the position, in space, of the air-bubbles, which are seen to 

 traverse the crystalline grains. 



It is to Hagenbach that we owe the suggestion that the large 

 crystals slowly absorb the smaller ones. Forel had suggested that 

 the crystals all grow in size by infiltration and freezing. This 

 view, however, has been shown to be untenable. Against Hagen- 

 bach's theory it has been urged that the glacier grains in any given 

 district are virtually of the same size, whereas the theory requires 



