Qeology and Mineralogy. 227 



the very surface of the glacier. In the process of ages, as the ice 

 iriay be supposed to have gradually diminished, through evapora- 

 tion, if not through thawing, the superficial earthy material, which 

 never evaporated, must have accumulated to a large extent. How- 

 ever we account for this fact, every one knows that human bodies 

 or other objects, accidentally lost in the glaciers of Mont Blanc, re- 

 appear at the surface after a series of years, at points some thou- 

 sands of feet below. 



I infer, therefore, that the material moved by the diluvial waters 

 may have been afforded by some of the interior portions and even 

 the surface of the glacier, as well as by the subjacent rock-rubbish. 

 I will only add that some portions of the material in and upon the 

 ice may have been let down in situ by the slow disappearance of 

 the glacier, without having been subjected to the assorting action 

 of the glacial torrents.* 



This process is impressively illustrated along the borders of the 

 Mer de Glace and other Alpine glaciers ; and more instructively 

 still in the buried glacier stumps^ound in the gulches of the Sierra 

 Nevada, and elsewhere in the Pacific States. 



I know of no certain evidences, in Michigan, of a Champlain de- 

 pression of such extent as to bring the surface of any portion of the 

 State below the sea-level. In a district so nearly horizontal, how- 

 ever, there must have been a period, before the erosions of the mod- 

 ern drainage courses had begun, during which the drainage was ex- 

 ceedingly obstructed and slow. The supply of water from the dis- 

 solving glacier was greater than could be discharged through the 

 forming outlets; and'the extensive areas must have lain submerged 

 until the deepening of the outlets permitted their drainage. But 

 this period was, by hypothesis, that when a geologic winter was 

 merging into a geologic spring. There was not yet a summer cli- 

 mate ; and the annual winter must have congealed the surfaces of 

 the surrounding lakes, and arrested the superglacial torrents, if it 

 did not materially diminish the flow of the subglacial ones. 



I think the steps of this reasoning safe and sound. But we have 

 here an overlooked condition of glacial agency in the natural order 



positions of'rock-m'asses to which I have refen-ed. There were re- 

 gions in these lakes w-here rocky formations rose nearly to the sur- 

 face, or projected to a slight extent above it. On the freezing of 

 the watery surface, these would be firmly embraced in the ice. 

 Meantime, as the supply of water is diminishing through the ad- 

 vance of the annual winter, the lake subsides, and the frost takes 

 holil of the exposed rock at a greater depth. But the annual spring 

 and summer return. The supply of water increases, the surface of 

 the lake rises and the floating field of ice lifts sheets of previously 

 half-disjointed limestone, and floats them in the direction whither 

 the current sets or the wind blows. They may be dropped some 

 „v,*n'^l^ ^^^^ ^as first impressed upon my attention by my brother, N. H. Win- 

 chell, who has studied the Drift with much assiduity. See his papers in Proc. Amer. 

 Assoc, Dubuque Meeting, 1872, and in the "Popular Science Monthly" for June 



