On the Causes of Glacier-Motion. 



215 



glacier may be so much, masked by this as not to be readily observ- 

 able. Lastly, it may be suggested as possible that a certain amount 

 of expansion by heat may have the effect of giving a set to ice, so that 

 it does not return to its original length when brought back to the same 

 temperature. If this be so the oscillations would be much less marked, 

 and at the end of the glacier would probably be indistinguishable. 



I may now draw attention to some phenomena of glacier action, 

 which are explained by the heat theory, but which do not seem expli- 

 cable on the gravitation theory. 



(1.) It is well known that glaciers, when they emerge from a narrow 

 gorge into a comparatively wide valley, spread out into a fan-shape. 

 The Rhone glacier is a well-known instance. A still better one is a 

 small glacier in Norway, mentioned by Professor Sexa, which spreads 

 out to five or six times its previous width, Now the effect of gravity, 

 acting on a mass as a whole, is to carry it in one single direction, that 

 of the steepest slope. The only way in which gravity can produce 

 such a spreading out is by the parts of the glacier shearing over each 

 other in the manner of a viscous solid. But the phenomena of ice 

 cliffs, as mentioned above, show that ice does not spread from this 

 cause, so that the fact seems impossible to explain by gravitation alone. 

 On the heat theory it is, of course, perfectly easy : the expansion and 

 contraction will take place in all directions where there is freedom to 

 move. 



(2.) Connected with this phenomenon is that of the longitudinal 

 crevasses seen near the edges of glaciers, and particularly where they 

 spread out in the manner just described. Now on the gravitation 

 theory, as remarked above, the only possible explanation of a crevasse 

 is that the ice is bending over a convex surface, and that its upper part 

 is thus placed in a state of tension, under which it breaks. Since, on 

 the gravitation theory, every part of a glacier is exposed to a severe 

 pressure from behind, this explanation does not fit very well even for 

 transverse crevasses ; but to longitudinal crevasses it is clearly 

 inapplicable, since the bottom of a valley is seldom or never convex in 

 the direction of its width. On the heat theory the explanation is 

 simple. We may suppose the heat energy communicated per square 

 foot of surface to be about the same, whether near the middle or edge 

 of a glacier. This energy is expended in producing an expansion 

 throughout the whole thickness of the glacier, as described above. 

 Hence the smaller this thickness, the greater will be the amount of 

 expansion, and the greater therefore the nett motion which results. 

 Hence the thinner parts of a glacier will always be tending to tear 

 themselves off from the thicker, and thus longitudinal crevasses will 

 frequently be found. 



(3.) The striae which are so marked a feature of glacier-worn rocks 

 become more easily explained on this theory. I have seen such striae, 



VOL. XXXIV. Q 



