361 [Shaler. 



prevent, to a great extent, the escape of heat which is constantly- 

 passing from the interior to the surface of the earth. Therefore the 

 immediate result would be to bring the isogeothermal lines nearer the 

 original surface. In the uncovered condition of the land the isogeo- 

 thermal lines correspond in their curves with the principal irregulari- 

 ties of the surface rising beneath the ridges and curving downwards 

 under the valleys. As soon, however, as the glacial matter had filled 

 any considerable valley, the isogeothermal lines beneath would begin 

 to become parallel to the new surface, ceasing to have the original 

 conformity with the rock surface. Let further accumulation take 

 place, and it is plain that in time the isogeothermal of the melting 

 point of ice must invade the glacial mass. It is furthermore evident 

 that the melting arising from this ascent of the hues of equal heat 

 will take place first at the lowest points, as in the deeper river valleys, 

 and would, over such basin regions as New England, be particularly 

 active in the deep lake cavities. Beginning at the lowest points, and 

 over small areas, the points where melting could take place would in-., 

 crease in number and the areas enlarge until the removal of the ice 

 from this and other causes balanced the accumulation going on upon 

 the upper surface of the glacier. In case the increase in thickness 

 was gradual, it seems likely that the equilibrium between the deposi- 

 tion on the surface and the melting at the base of the glacier might 

 be easily established. If, however, the accumulation was rapid it 

 seems very likely that a very sudden melting of the base might occur, 

 and several oscillations in the position of the isothermals take place 

 before this equalization of the actions of deposition and melting 

 was brought about. 



In this local melting at the foot of the glacier it is believed we 

 have the agent which has produced the local erosion we find in gla- 

 cial basins. The melted water can not escape to the surface, as it 

 would there be frozen and the mass in no way diminished. Besides 

 this, the hydrostatic pressure would operate with great force to com- 

 pel a lateral movement along the base of the glacier. The most 

 probable means of escape would be over the lowest rim of the de- 

 pression in which it was melted in the direction of lower levels. The 

 existence of sub-glacial streams flowing in channels excavated in the 

 ice in the continental glaciers of Greenland, renders it likely that this 

 water would excavate a conduit in the direction of the drainage of 

 the country, and that from the many melting points connecting, might 

 arise a river system similar to that found in these regions when freed 

 from the ice. Where, however, the glacial sheet thinned off in any 

 direction, it is possible that the influence of pressure in determining 

 the course of the streams might be greater than the influence exer- 

 cised by gravitation, and that thereby the water might be forced 



