542 HENRY B. KUMMEL 
Delaware River. Northwest of this line is a broad plateau, hav- 
ing an average elevation of about 600 feet. It extends west- 
ward into Pennsylvania, being dissected by the Delaware to a 
depth of 4oo to 500 feet. Its highest part is along the south 
and east, about a mile back from the top of the escarpment. 
Thence it declines in elevation very gently northward and west- 
ward. The escarpment is most marked in the vicinity of Flem- 
ington, where the contrast in hardness between the rocks of the 
plateau and of the low ground is the most marked. 
Sourland plateau extends from Lambertville on the Delaware, 
northeastward for seventeen miles. It has an average width of 
four and a half miles and varies in height from 450 to 560 feet. 
The backbone of the plateau is a belt of trap rock, a mile in 
width, but the hard sandstones and argillites on either side rise 
nearly to the same elevation. In the vicinity of Hopewell and 
northward, the plateau is separated from the low plain to the 
southeast by an escarpment varying from 200 to 400 feet in 
height. Other masses of trap rock forming minor hills and 
ridges rise from 200 to 500 or 600 feet above the general level. 
Along the northwestern boundary also there are several marked 
elevations due to massive quartzite conglomerates. 
THE ROCKS. 
It has been found possible to divide the sedimentary rocks 
of the Newark system into three subdivisions.* These are not 
based upon paleontological evidence, since fossils are too few to 
be used for this purpose, but on lithological differences, which 
permit the establishment of recognizable horizons. While fully 
aware of the dangers attending the use of lithological characters 
in correlation, the author is confident that in this case they have 
been reduced to a minimum, owing to the care with which the 
beds have been traced step by step. The beds of the three series 
* Practically all the outcrops and sections—many hundreds in number — have 
been examined and plotted. All the roads and nearly all the stream beds have been 
traversed. So monotonous are the beds that it is only by this detailed work that there 
is any possibility of detecting and tracing the structural complications. 
