STEVENSON— INTERRELATIONS OF THE FOSSIL FUELS. 91 



solidated river drift, which where unprotected is attacked energeti- 

 cally by the current. The trees along the lower portion of the bank 

 have roots almost horizontal, as the wet ground is little more than 

 a foot below the collar. In very many cases the roots are exposed 

 to a distance of two to three feet from the trunk, the loose material 

 originally surrounding them having been removed. Several of these 

 trees have been observed each year, during ten years the exposed 

 portion has increased steadily, but the trees have continued to grow 

 and apparently they are as solidly fixed as at first. 



But in the sands under and over peat deposits as well as in rocks 

 contemporaneous with such deposits, one finds logs, even tree stems 

 with attached roots. Rivers undercut their banks, trees and plant 

 debris fall into the water and are transported. Some of this material 

 is carried to the sea, there to decay, but some is dropped in shallows 

 or stranded on the river plain during subsidence of the flood, to be 

 covered by deposits brought by succeeding floods. 



Contemporaneous Erosion. — Little rivulets are seen in the smaller 

 bogs, but great swamps, in which peat is accumulating, are more or 

 less imperfectly drained by rivers with sluggish flow. The streams 

 are subject to floods, during which they bring down more or less 

 organic material mingled with plant debris. Much of this is de- 

 posited in the channel way and much of the rest is spread over the 

 flooded portion of the swamp. Sometimes an obstruction dams the 

 stream and diverts its course, leaving below the dam a stagnant pool, 

 which in time becomes concealed by peat ; but the story is revealed 

 when a drainage canal is cut, for the half-filled channel way is shown 

 by a " roll " in the underclay. The drainage system is often distinct 

 in a buried bog. Lorie's^-^ observations in the peat region of Hol- 

 land-Belgium prove that the channels of large rivers have been filled 

 with sediment and that these are traceable easily when the records 

 of borings have been platted. 



Banks of the intersecting streams are irregular, as plants spread 

 out into the water, often becoming floating fringes. When the 

 channelway is filled gradually by deposit of inorganic matter, the 

 fringes are not torn away but are enveloped in approximately normal 



1-3 Cited in "Formation of Coal Beds, II.," Proc. Amer. Phil. Soc, Vol. 

 L., pp. 617-619. 



