THE GRAVEL PIT. 23 



finding a temperature which will suffice to melt it all away be- 

 fore the next winter. So tongues of granular snow stretch 

 down the mountain valleys, and being, like our late spring 

 snow, exposed to increased action of warmth, these valley pro- 

 longations of the upper firn become completely changed into 

 solid ice. This is now a glacier. 



We may reason a step further from facts of observation. All 

 substances expand with increase of temperature, and contract 

 with reduction of temperature. The glacier is certainly at a 

 lower temperature in winter than in summer though it can 

 never be warmed above thirty-two degrees Fahrenheit, which 

 is the thawing temperature. The surface of the glacier is also 

 at a lower temperature during the night than during the day. 

 The glacier, therefore, must sometimes expand and some- 

 times contract. Now, when it expands, the whole expansion 

 will be developed at the free lower border, since the upper 

 border is frozen to the earth, and pressed also, by the snows 

 beyond. Also, if both were free, most of the expansion would 

 be developed below, because gravity aids motion downwards. 

 Next, when the glacier contracts, the lower border does not 

 retreat, because the ice is not strong enough to bear the pull 

 of the mass up the slope. The ice breaks in innumerable 

 little cracks. These are soon filled with water, which freezes, 

 and thus restores the complete solidity of the glacier. Thus, 

 when the next expansion takes place, the glacier takes another 

 slide down the valley. So the glacier travels. So, if a whole 

 state should become glacier-covered, the ice-sheet would have 

 a motion from higher to lower, and from colder to warmer. 

 Every thing on its surface would be transported ; every loose 

 object beneath it or in front of it would be pushed forward. 



Now, here are some hints toward an explanation of the 

 method of transportation of our millions of bowlders. If we 

 go to the Alps we find exactly such glaciers, on a small scale, 

 performing precisely such work. Thus our theory receives 

 confirmation. We can not pretend that glacier action ex- 

 plains all the phenomena of the Drift. Nor do we pretend 

 that any thing more than a hint has been given toward an ex- 



