88 STEVENSON— FORMATION OF COAL BEDS. [April i8, 



snags or towheads, which even the greatest flood possible on such a 

 river could not dislodge, as conditions along the Mississippi abun- 

 dantly show. It is impossible to conceive of any means whereby a 

 tree capable of carrying such a load could be floated away to deep 

 water, unless it grew on the wall of a fiord — where it could not 

 secure the water-worn fragments. 



The assumption that shales, sandstones and conglomerates were 

 deposited necessarily in deep water or in a permanent body of water 

 must be regarded as unsupported by any positive evidence. The 

 writer, during a tedious search through the literature, has not found 

 that authors think that the proposition needs evidence ; it seems to be 

 accepted as axiomatic. But evidence to refute the doctrine abounds 

 in the Tertiary and Quaternary and, in so far as the Appalachian 

 Coal Measures are concerned, the facts seem to indicate that they 

 are flood-plain deposits and reworked alluvial fans. This condition 

 may afford a clue to explanation for some of the occurrences. 

 Rivers, torrential in their upper reaches, flowed across the plain. 

 Rolled fragments of varying size were pushed along the beds. 

 Pebbles of quartz,^^ 5 inches in diameter, have been found in the 

 Sharon of southern Ohio at not less than 3Cmd miles from their 

 source. During a great flood, if the stream were dammed tempo- 

 rarily, the water would sweep over the '* bottoms " or break across 

 the necks of curves; a new channel would be cut, the old channel 

 above for a short distance would be scoured and its sand and pebbles 

 would be strewn on the river-plain. This happens only too often 

 along the Mississippi, as has been shown on preceding pages. In 

 such a rush of water, a block of 400 pounds would be gathered up 

 in the mass as readily as though it were a pebble ; but gravity would 

 act promptly and the coarse fragments in the load would be left scat- 

 tered on the surface while the finer materials would go far beyond. 

 Succeeding floods would cover the sands and gravels as well as the 

 larger fragments with finer materials in which the larger river-worn 

 masses would be widely separated, for the most part, though here 

 and there they would be grouped in smaller areas. One finds this 



^' E. B. Andrews, Ann. Rep. Geol. Surv. Ohio, 1870, p. 67. 



