HARSHBERGER : PHYTO-GEOGRAPHIC SKETCH 131 
dition of widely separated colonies was brought about. In periods 
of elevation the reverse process must have taken place, and the 
many distinct areas must have been merged again into one. As 
we find /soetes saccharata Engelm. nowhere else than in Chesa- 
peake bay and J/svetes riparia Engelm. nowhere else than in Dela- 
ware bay, it seems fair to assume that the physiographic changes 
mentioned above as controlling the distribution of the former spe- 
cies have influenced the distribution of /scetes riparia along Dela- 
ware bay in a somewhat similar manner. 
We can picture to ourselves the manner in which the area up- 
lifted from the water was tenanted by various species of dry-land 
plants. The process may have been gradual, permitting the 
gradual adjustment of the newly implanted vegetal covering, or 
it may have been sudden, followed by an inrush of species. 
Lange * describes an interesting case, illustrating the latter process. 
The establishment of vegetation upon the flood-plain of the 
Delaware river may best be studied on islands formed by the 
deposition of silt about some obstacle in the river. A bar origi- 
nates. The first vegetation consists largely of annuals, then come 
willows, and finally a characteristic flood-plain forest. The red 
maple (Acer rubrum) appears, then poplars and the ash (/raxinus 
Americana). Gradually the plain becomes dry enough to permit 
the development of a true mesophytic forest. Some such stages 
were passed through by the vegetation of southeastern Pennsy]- 
vania with each uplift following a process of depression. The 
Same forces are at work now that were active in past geologic 
time, and the same laws are in operation which then controlled 
the distribution of plants. 
The history of the upland plant associations is somewhat differ- 
ent from that of the lowland. In tracing the genetic development 
of these associations, we must consider especially the vegetation 
Cowles + has described the origin of the 
vegetation of clay ravines. None of this class of ravines are found 
typically developed in southeastern Pennsylvania. Rock-gorges 
are common and are occupied by all of the streams that drain the 
The physical nature of the gneissic 
of the creek ravines. 
region under consideration. 
*Lange, D. Revegetation of Trestie Island. Minn. Bot. ated. 2: 621. 1901. 
¢ Cowles, H. C. The physiographic ecology of Chicago and vicinity. Bot. Gaz. 
: 88, i. 
3r 190 
