TERTIARY WORK ON THE CRETACEOUS OVERLAP. 577 



neglecting low country by which the mountains could be avoided serve to 

 emphasize the need of explanations for these anomalies based on the location 

 of the rivers during the Cretaceous cycle of denudation, when the hard beds 

 of the mountains had a different relation to baselevel from that which now 

 obtains. This may be seen in the Eomney, Virginia, sheet ; Mill creek here 

 cuts Mill Creek mountain to join the South branch of the Potemac, although 

 the river could now be joined by running around the end of the mountain on 

 low ground ; and again on the Maynardsville, Tennessee, sheet, where Clinch 

 river bisects a short ridge, known as Lone mountain, instead of running 

 around it. As an example of adjustment of smaller streams during the 

 Tertiary cycle, no better illustration can be given than one presented on the 

 Piedmont sheet of West Virginia, where New creek has beheaded a small 

 stream that once rose west of New Creek mountain, leaving a " wind gap " 

 in place of its Avater gap ; two other small streams still maintain water gaps 

 through the mountain. 



Tertiary Baseleveling of the Cretaceous Overlap. — Frequent mention has 

 been made of the former inland extension of the Cretaceous formation 

 beyond its present bevelled edge, and special features have been found in the 

 drainage of that part of the peneplain once covered by the inland extension 

 of this formation. It has been noted also that while the harder rocks of the 

 old peneplain still retain in great measure the form that they had when they 

 were reduced to a lowland of denudation in the Cretaceous cycle, the softer 

 rocks of the old peneplain have been reduced to a new lowland of denuda- 

 tion in the Tertiary cycle. The same is true of the remnant surface of the 

 Cretaceous formation. Being generally of small resistance, it has been re- 

 duced to what I have called the " Tertiary baselevel," and thus is now to be 

 classified, so far as date of origin of its surface is concerned, with the low- 

 lands between the Appalachian ridges, the great valley, and the Triassic 

 lowlands. Here and there some of its harder sandstones hold their edges at 

 a slight height over the surrounding surface, and thus form a series of low 

 hills on the line of their strike. Several of these are shown on the New 

 Jersey maps. It is probable that the Long island hills, on which the rocky 

 terminal moraines are laid, are homologous with these hills in New Jersey, 

 and that similar hills occur in the south. I have not yet succeeded in tracing 

 the southeastern border of this baseleveled area of the Cretaceous, for in 

 that direction it approaches the sea-coast closely, and there, as has been ex- 

 plained in the introductory statement, the cutting and filling resulting from 

 comparatively brief and trivial elevations and depressions make a record so 

 complete and so complicated that its details encumber the problem and place 

 its solution out of reach for the present. In the southern Atlantic states, 

 where the fuller discussion of this part of the work will be possible, there is 



