TRANSVERSE VALLEYS AND ROCK GORGE... - 177 
At the eastern end of each, where they debouch into Butternut valley, 
the valley wall is composed of a large amount of fragmental material 
arranged in terraces on the slopes toward the valley. This consists in 
very large part of sharp edged or imperfectly rounded pieces of Devonian 
rocks, such as occur in the immediate neighborhood, with which are min- 
eled a few well rounded cobbles and boulders, evidently exotics of glacial 
origin. The rock floors of the valleys are concealed to some extent by 
accumulations of vegetal and of lacustrine origin, but one looks in vain 
for the characteristic boulder-clay which is found so plentifully every- 
where else. Not only is this so, but the plateau surface between James- 
ville gorge and Rock gorge, an interval of 1,200 to 1,400 feet, is swept 
clean of glacial deposits. The only trace of glacial visitation found con- 
sists of a wellrounded quartzite boulder, varying in diameter from 23 to 
3 feet, which rests directly on the country rock about midway between 
the two gorges. It is striking on account of its unusual size. 
Rock GorGE 
Rock gorge (see plate 14, figure 1) has some special features of interest. 
It dissects the plateau the entire distance from Onondaga to Butternut 
valley. Its floor, which is quite level from side to side and from end to 
end, lies about 120 feet above the floors of Onondaga and Butternut 
valleys. It is noteworthy that the height of the floor of Rock gorge 
corresponds to the height of a well marked gravel terrace, apparently an 
old lake beach, which can be traced for several miles on both sides of 
Onondaga valley. 
Rock gorge does not run straight across from one main valley to the 
other, but trends shghtly southward as we pass from either end inward. 
It is thus, in the main map view, in the form of a very widely opened V, 
with its angle pointing southward. Atthe angle of the V—that is, about 
midway its length—and on the south side of the gorge the wall of the 
gorge is cut back in the form of an amphitheater, which is semicircular 
in outline and about 125 feet deep by 250 feet wide. The walls, as else- 
where in the gorge, are nearly perpendicular, with their bases concealed 
by recent talus accumulations. At the foot of the walls the floor of the 
amphitheater is depressed in a singular manner, or rather there is a series 
of pit-like depressions, such as might be found at the foot of a waterfall. 
The idea which suggested itself on my first visit to this place, that there 
had actually been a waterfall here, was enforced by finding later evi- 
dences of an old channel above, leading to the edge of the amphitheater. 
Near the eastern end of the gorge, and with its head lying about in the 
middle of the gorge floor, is a trench, evidently of much more recent origin 
