GEOLOGY OF THE NEW YORK CITY AQUEDUCT 69 



southeastern New York are still of this type — the W'allkill, the 

 Rondout and the lower portion of the Esopus. 



But the larger rivers, the great master streams, of the super- 

 imposed drainage system, in some cases were so efficient in the 

 corrasion of their channels that the discovery of discordant struc- 

 tures has not been of sufficient influence to displace them, or re- 

 verse them, or even to shift them very far from their original direct 

 course to the sea. They cut directly across mountain ridges be- 

 cause they flowed over the plain out of which these ridges have 

 been carved and because their own erosive and transporting power 

 have exceeded those of any of their tributaries or their neighbors. 

 They are superimposed streams (not antecedent), they have, with 

 their tributaries, settled down in the ancient plain, and, by their 

 own erosive activity, have carved the valleys deeper and deeper, 

 cutting the upland divides narrower and narrower until now only 

 here and there a ridge or a mountain remnant stands with its crest 

 or summit almost reaching up to the level of the ancient pene- 

 plain on which the work began. If the transported matter could 

 all be brought back and replaced in these valleys the old plain 

 might be restored, but the work would immediately begin all over 

 again. 



Of these great master streams the Hudson is the only local rep- 

 resentative [see Study of the Hudson River gorge in part 2]. 



Tertiary incomplete peiieplanation. Stich processes, if allowed to 

 continue on a stable continental region, would ultimately reduce 

 the land for a second time to a monotonous plain (complete cycle 

 of erosion). The beginnings of such a plain would be made in 

 the principal stream valleys upon reaching graded condition. Their 

 lateral planation and the development of flat-bottomed valleys 

 would begin at about the level that the plain would stand in the 

 final completed stage. The difference of elevation between the 

 ridge crests or hilltops and these flat valleys, i. e. between the old 

 peneplain and the new unfinished one would be an approximate 

 measure of the amount of the continental elevation that instituted 

 the new cycle. 



But judging from such remnants of this later plain as are to 

 be seen, the two, i. e. the old Cretaceous peneplain and the new 

 Tertiary peneplain are not parallel. Toward the southeast, toward 

 the sea, the older plain descends more rapidly than the younger 

 and intersects it. Both pass beneath sea level in that direction. 

 The dift'erence between them therefore varies with localitv from 



