246 NICHOLS: THE VEGETATION OF CONNECTICUT 
no means uncommon. They are developed along streams through- 
out the Central Lowland, while in the Highlands they may occur 
wherever there are deposits of glacial drift. More often than not 
they are relatively small and correspondingly. unimpressive; but 
in some cases, particularly in the northern half of the Central 
Lowland, they may assume considerable size and importance. 
One of the best examples which the writer has examined of a 
ravine in unconsolidated rock is located in the town of Windsor. 
It has been scoured out to a depth of perhaps thirty feet froma 
coarse, sandy substratum through the activity of a small brook 
which flows into the Farmington River from the north. The 
slopes of the ravine, forested with hemlock, yellow birch, sugar 
maple, beech, white ash and tulip, contrast sharply with the 
surrounding upland which is largely overgrown with white pine, 
oaks, and chestnut. The undergrowth includes many of the herba- 
ceous and shrubby species which have been listed as characteristic 
of rock ravines. But along with these grow Lycopodium obscurum, 
Oakesia sessilifolia, Corylus americana, Geranium maculatum, 
Chimaphila umbellata, Pyrola rotundifolia, and Erigeron pulchellus 
—all plants of relatively dry, open woods. As might be antici- 
pated, the bryophytic flora is poorly represented, The rock-face 
and crevice mosses and liverworts, which constitute such a striking 
feature of the vegetation in rock ravines, are absent. Along the wet, 
sandy banks of the stream are a few species, such as Pellia ept- 
phylla, Conocephalum conicum, Mnium hornum, and Catharinaea 
undulata; but aside from these about the only bryophytes present 
are a few forms which grow on rotten wood or humus, such as 
Mnium cuspidatum, Stereodon cupressiformis and Georgia pellucida. 
Ravines in unconsolidated rock compare neither in scenic nor 
botanic interest with rock ravines. To be sure, all gradations are 
found between extremely shallow and scarcely perceptible de- 
pressions, which ordinarily would hardly be classed as ravines, 
and ravines of considerable depth, with fairly steep sides, like the 
one at Windsor. As a result of the diversity in environments 
which it is obvious may thus be afforded by different ravines, 
all stages of transition may be found between the vegetation of 
swamps and uplands on the one hand and that of typical ravines 
on the other. On the average, however, ravines in unconsolidated 
