250 NICHOLS: THE VEGETATION OF CONNECTICUT 
hold true in the future. From a standpoint of present-day plant 
geography, therefore, the climax vegetation of rock ravines may 
be looked upon as practically permanent. 
Although the fact cannot be overlooked that ordinarily the plant 
societies of rock ravines are associated with a definite phase of 
stream development, it should also be recognized that while very 
commonly the formation of a ravine has been due to the activity 
of streams which are still operative, this is by no means always 
the case. There are many rock ravines whose formation cannot 
be accounted for by contemporaneous factors at. all. Often, 
as in the Devil’s Gulch, the streams now present in such 
ravines can have played little part in their formation. Many 
streams have been superposed, so to speak, on the topog- 
raphy. They have found rather than made their channels. Not 
infrequently, as in the Wolf Den, ravines have been developed in 
other ways than by stream erosion. It is largely due to the 
prevalence of this preérosion type of topography, which has been 
moulded by physiographic forces of the geologic past, that rock 
ravines are so much more highly developed in the Highlands than 
in the Lowland. 
RIVER AND StrEAM BLUuFFs 
The later phases in river activity may be observed along most 
of the larger streams throughout the state, As the result of 
lateral cutting, the ravine once present has been replaced by 4 
broad valley. ‘As a ravine widens out, the exposure to wind, sun, 
and changes of temperature increases, and the moisture content 
of the slopes is appreciably modified. The effect of these environ- 
mental changes on vegetation can be seen by comparing the flora 
of a river valley with that of a ravine. Inageneral way, the vege 
tation of stream valleys can be treated under two heads: BLUFFS 
and FLoop PLains.. The term Bluff, as used here, includes not only 
the relatively steep slopes which frequently demarcate the valley 
from the upland, but also the gentler slopes which commonly 
occupy most of the intervening valley floor. In other words, it 
embraces all parts of the valley which, in contrast to flood plains, 
have been formed by erosion rather than by deposition. 
The Vegetation of Bluffs in Unconsolidated Rocks —In the 
