558 W. M. DAVIS — DATES OF TOPOGRAPHIC FORMS. 



mergence occurred which alh)wed the sea to reach the crystallines of the 

 highlands ; for if the baseleveling had been the product of the shore waves 

 cutting their way inland, then the lowest beds of the overlying Cretaceous 

 series must have been composed chiefly of Triassic waste ; and this is em- 

 phatically not the case. The peneplain must have been first completed over 

 the area of the weak Triassic formation ; then a slight submergence would 

 allow a rapid transgressionof the shore-line, halting nowhere until it reached 

 the margin of the crystallines; and in this way the lowest beds of the Cre- 

 taceous might reasonably be expected to be almost free from Triassic frag- 

 ments. 



Inland Extension of the Cretaceous Formation. — When the Atlantic reached 

 the edge of the crystalline plateau, it undoubtedly had some effect in cutting 

 its way into the mass, thus aiding the streams that ran out into it in furnish- 

 ing Cretaceous sediments. But the amount of inward cutting thus performed 

 could not have been great, for the sea was shallow and the rocks were hard. 

 More than this, the arrangement of the streams in the back, or inland, por- 

 tion of the highlands decides clearly enough that no Cretaceous beds ever 

 reached over their courses ; and in all the farther inland extension of the 

 peneplain over Pennsylvania, and perhaps beyond, the sea had no share. 



The argument from the arrangement of the streams is as follows : The 

 highlands consist of belts of hard and soft rocks, trending northeast and 

 southwest with the great Appalachian chain. The texture and structure of 

 these rocks leave no room for doubting that they were once deeply buried, 

 and that they are now revealed by long-continued denudation. Whatever 

 the arrangement of the rivers in the beginning of the long denudational 

 work, the smaller of them must have become closely adjusted to the internal 

 structure of the mass before the old age of the region — the' peneplain stage 

 in its development — was reached ; that is, only those small streams survived 

 that accepted the lead of the weaker rocks. This principle appears to be of 

 general application. Since the peneplain was uplifted, the streams should 

 retain the well adjusted courses that they had gained when it was a lowland ; 

 they should still follow the lead of the weaker rocks. But if the production 

 of the peneplain be ascribed to the cutting action of the sea-shore waves, no 

 such correspondence of stream-course and weak rocks need now be looked 

 for. As fast as the plain of marine denudation was produced, it must have 

 been strewn over with the waste from the remaining land farther westward. 

 When elevation came afterwards, the penejilain must have risen with the 

 Cretaceous cover on its back, and the streams then would have taken courses 

 simply consequent on the slope of the cover, just as the streams now flow 

 down the gentle slope of the Tertiary plains of southern New Jersey, in- 

 different to the crystalline structure that is buried beneath. Streams thus 

 located might later cut through the cover and thus become superimposed 



