FLORA OF OLDER POTOMAC FORMATION, 379 



We paid special attention to the contact of the Potomac with the 

 crystaUine and Paleozoic rocks in southeastern Pennsylvania and along 

 the Delaware State line, because it was here that Mr. McGee had 

 reported basal Potomac gravels and arkose. Near Brandywine Summit 

 are extensive beds of feldspathic rock which are worked for kaolin. 

 This occurs both as lithified and as decomposed feldspar, and the latter 

 might easily be confounded with Potomac arkose, especially when 

 partly mixed with Bryn Mawr and Chestnut Hill gravels, as it is in 

 the region above Media, Bryn Mawr, and Conshohocken. Northeast 

 of Conshohocken at a village called Cedar Grove, there are extensive 

 gi'avel, sand, and cla}^ pits. The Chestnut Hill gravel is here quite 

 thick and pure. It usually rests on the crystalline rocks, which are 

 often decomposed, forming pure kaolinic sands or clays of great thick- 

 ness. At a few points, however, the gravels rest on mottled clays, 

 which are probably Potomac and possibty basal Potomac. Sometimes 

 pure sand extends below these which seems to be basal Potomac sand. 

 At one place the lowest clays exposed were dark blue-black and full of 

 small pieces of lignite. The Trenton marble crops out at Cedar Grove 

 and is quarried there, and mottled clays were found resting on the lime- 

 stone. That these beds represent the Older Potomac seems tolerably 

 certain. 



We also made a somewhat careful examination of the "yellow rocks" 

 on the left bank of the Delaware above Trenton in New Jersey. They 

 are conglomeratic and, except in color, appear to be identical with the 

 conglomerates of the Trias as exposed in many places from Saltenstall 

 in Connecticut to Culpeper, Va. In a deep ravine they were found shading 

 off into regular red Triassic sandstone or brownstone, and there can be 

 little doubt that they are wholly Triassic and not at all Potomac. 



We were equalh^ unsuccessful in our search for Older Potomac 

 materials at the well-known "Sand Hills" of New Jersey. These hills 

 consist of a trap ridge overlain by a superficial deposit of varied character, 

 but that it can not be Potomac is proved by the fact that at lower levels, 

 and especially at Tenmile Run Corners, it was found resting unconform- 

 ably upon the plastic clay of the Raritan formation. 



On September 18, 1892, two plant beds were found in the new reser- 

 voir at Washington, yielding numerous ferns and conifers. This reservoir 



