FLORA OF OLDER POTOMAC FORMATION. 377 



then examined the eastern margin of the belt all the way to the Poto- 

 mac River and found the argillaceous sands and white stratified clays 

 uniformly overlying the sandstone and overlain in turn by the marls. 

 The conclusion became irresistible that for this entire region this is 

 the normal order of deposition. This view is abundantly confirmed 

 by the plant remains found respectively in the lignitic clays on Potomac 

 Creek west of the Telegraph road and in the upper clays near Aquia 

 Creek, which differ widely in character and indicate a great time inter- 

 val between the earlier and the later deposits. The beds in this region 

 are thicker than in the valley of the Rappahannock, the erosion having 

 been less. Measuring as carefully as possible, we arrived at the follow- 

 ing approximate section for any line drawn across the belt perpendicular 

 to the strike, as, for example, on Potomac Creek or Accokeek Creek: * 



Section of Accokeek Creek. 



Feet. 

 3. Loose sands interstratified with white laminated clays carrying plants of high rank, such as Sapin- 



dopsis and other undoubted dicotyledonous genera 100 



2. Coarse feldspathic sandstone becoming workable freestone 1.50 



1 . Lignitic clays carrying the older types of plants (ferns, cycads, conifers, and archaic dicotyledons ) .50 



Total exposure 300 



Attention was next turned to the northern extension of the 

 Potomac beds, and two months were spent in their systematic study. 

 Following first the landward margin in the District of Columbia and 

 Maryland, we soon discovered that less difference exists between the 

 beds here and those of Virginia than had been supposed. The old 

 idea of an "Upper Clay member" in Maryland, as opposed to a "Lower 

 Sandstone member" in Virginia, was now whoU}^ dispelled, the Virginia 

 beds having been found to begin and end as clays and the sandstone 

 to occupy an intermediate position. It was now found that in Mary- 

 land also wherever the deposition is normal (i. e., no transgression of 

 higher beds) the basal member is clay and the succeeding one is, if not 

 sandstone, at least very arenaceous and often lithified. Moreover, this 

 second member in Maryland, although usually reddish from iron oxida- 

 tion that has filtered into it from the overlying iron ores or from iron 

 constituents of its own, usually contains casts and molds of stems, logs, 

 and plants wholly similar to those ^f the Rappahannock sandstone, 

 and these beds must be stratigr'aphically the same in both States. This 

 condition of things with slight variations extends entirely across the 



