I9I3.] STEVENSON— FORMATION OF COAL BEDS. 87 



condition requires topographical features which did not exist. The 

 whole Coal Measures area of Ohio was a low plain; the nearest 

 highlands were in Canada, hundreds of miles toward the north, and 

 the Appalachians, hundreds of miles away toward the east. The 

 agency of ice must be set aside as in the highest degree improbable. 

 The majority of authors have supposed that uprooted trees 

 floated away carrying the masses entangled in their roots; but the 

 difficulties involved in this conception appear to be insuperable. 

 There can be no doubt that trees do seize such blocks and that under 

 proper conditions they could transport them. Any one, who has 

 seen the manner in which the white birch of the White Mountains 

 enwraps its roots about blocks of stone weighing half a ton or more, 

 recognizes that trees do seize large fragments. But that is not the 

 question. The observer is confronted at once with the problem of 

 conveying that tree and its load to deep water, sea or lake, where 

 the great tree, 75 or more feet high, may float in vertical position, 

 almost wholly submerged. Trees grow on the land, where alone the 

 fragments can be obtained. The transfer cannot be made by tor- 

 rents, as tree and load would be deposited at the first rapids. A 

 debacle, like that of Martigny or Johnstown, cannot be conceived of 

 as the agent, since a topography would be required such as did not 

 exist near any of the extensive coal fields whence large fragments 

 have been reported. Even had it existed, the terrific collisions, as 

 the flood dashed through narrow gorges and spread out in wider 

 portions of the valley, would have dislodged the fragments long 

 before reaching the open water. The bowlders cannot be relics of 

 floating islands, such as those of the Orinoco, Amazon or Congo, 

 since the origin of those islands forbids the suggestion.®^ Nor is 

 there any reason to suppose that trees growing on the seashore could 

 become the transporting agents, for, even though river-worn or 

 wave-worn fragments were abundant on the shore, the difficulty of 

 transferring the tree to deep water would still remain. If the trees 

 grew on the river banks, along the lower reaches of a great stream, 

 and were undercut, they would be stranded at the first bar to become 



""Formation of Coal Beds," these Proceedings, Vol. L., 1911, pp. 551, 



553, 554- 



