417 
[Davis and Wood. 
southeast. The Cretaceous beds on which the drainage was first 
established have a southeastward dip in all cases ; and indeed Pro- 
fessor Cook has shown that the dip of these beds is about equal to 
the present slope of the reconstructed Schooley peneplain in the 
Highland area . 1 The southeastward drainage is a general feature 
of our whole Atlantic coastal slope, and was probably thus intro- 
duced over its entire extent. There are two ways in which the 
streams of a region might be reversed from this prevailing direc- 
tion. They might be reversed by capture, an inverted northwest 
course then being taken as the divide is pushed from the point of 
capture to its final stable position, and we have examples of such 
inversions on the inner slopes of Watchung mountains; or they 
might be reversed by a tilting of their drainage area, as appears to 
be the case in Ambrose’s brook above named. 
It is impossible to explain the northwestward course of the Mill- 
stone by regarding it as th'e inverted middle portion of a south- 
eastward river, whose northern head-waters were captured by the 
Raritan. In that case, the inverted stream must have had its head 
on the north side of the gap in Rocky Hill, near Princeton ; the 
rest of the original southeastward river would continue from the 
south side of the gap in its normal direction as a beheaded stream : 
the gap would be a wind gap, not a water gap. 
The other explanation seems more acceptable. Conceive the 
Millstone as one of man}^ southeastward streams, born on the Cre- 
taceous cover at the beginning of the Somerville cycle. It was 
unfortunate enough to sink on the Rocky Hill trap sheet, by which 
the down-cutting of its head-waters was retarded ; some of its head 
streams may then have been captured by the more fortunate Rar- 
itan, which happened to choose a soft course on its way to the sea. 
But we must suppose that enough drainage area was left to the 
Millstone north of Rocky Hill for it to cut down its gap in this 
ridge close to baselevel before the end of the Somerville cycle. 
This is not unreasonable ; for the perfection, of the baselevelling 
of the Central Plain could not have been attained before a good 
sized stream would have cut a deep channel even across a hard 
stratum. Now it is precisely in the old age of a stream, when its 
gradient is small, that a moderate tilting may suffice to reverse its 
slope and change its course. Hence the fully baselevelled channel 
1 Ann. Rep., 1883, 28. The Artesian water supply along the coast depends on this 
general southeastward dip . Ann. Rep., 1885, 109, 123. 
PROCEEDINGS B. S. N. H. VOL. XXIV 27 APRIL, 1890. 
