GLACIER THEORY. 71 



"wide, and so did the supposed ancient glaciers, of which traces now exist. But the 

 North American glacier must have extended uninterruptedly almost over hill and 

 valley, for at least 2,000 miles; nor even with that width could it have found 

 higher ground on its borders, unless it were the Rocky Mountains on the west, con- 

 cerning whose drift phenomena we know but little. 



Again, all known glaciers are situated upon slopes, greater or less. Indeed, 

 how could they advance, if not upon slopes ? For though expansion by freezing 

 might have some influence in urging them forward, as maintained by authors, yet 

 the facts and reasonings of Prof. Forbes seem to show very conclusively, that 

 gravity is the principal cause of their onward march. At any rate, I know of no 

 example where a glacier does advance upon a level surface — certainly where hills 

 oppose its progress. It is surely, then, a great demand upon our faith, to ask us to 

 believe that the broad North American glacier has crowded southerly 500 or 600 

 miles, over a highly uneven but not sloping surface, and that simply by expansion. 

 Even should it be proved that we have centres of dispersion in the White Mountains, 

 or the mountains of northern New York, we must still admit a great movement from 

 the north sweeping the whole country, save a few peaks. Nor does it relieve the 

 difficulty to suppose an enormous thickness of the sheet of ice in the arctic regions, 

 from which the great glacier proceeded ; for its movement was on the surface of the 

 earth ; and this had no greater average height to the north than in the United 

 States. 



As to those supposed traces of ancient glaciers, to be described in my paper on 

 that subject, as occurring in New England, the probability is, that they were made 

 earlier than the drift scratches. At any rate, the latter are altogether the most 

 obvious phenomenon, and the principal thing to be accounted for ; and it is their 

 characteristics that are reconciled with so much difficulty with the effects of 

 glaciers. 



5. I find some difficulty in reconciling to the glacier theory, the diversity of 

 direction taken by the drift agency in different parts of the country. Over the 

 mountains of New England the course was south and southeasterly. But in the 

 valley of Lake Superior, it was nearly southwest. What could have determined 

 different glaciers in directions so diverse, especially as they must have ascended 

 rather than descended, both in New England and to the southwest of Superior, I 

 am unable to conceive. But supposing icebergs to be driven forward by currents 

 in the ocean, and there is no difficulty in accounting for such diversity of direction 

 in the striae and boulders. 



Upon the whole, those difficulties seem too formidable to admit of the adoption 

 of the unmodified glacier hypothesis. I lean, therefore, at present, toward that 

 which imputes most of the work on this continent to immense icebergs, icefloes, 

 and shore ice; not because that view is free from difficulties; for I acknowledge 

 them to be many; but they appear less to me now than in the other hypothesis. 

 Perhaps, however, the iceberg hypothesis, as I have stated it, falls but little short 

 of that of the glaciers. For I agree with Professor Agassiz, that to sustain the 

 former, " we must assume an ice period— nothing less than an extensive cap of ice 



