41 [32] 
and dce^i cuts which it would require. Their profile will he surveyed 
next season. This route would he longer than tlte other, and its sum- 
mit, level should he fed hy the waters of Casselman's River ahove 
Salisbury, led by a feeder to the western extremity of the tunnel. This 
feeder, following the eastern sideof Cassellman's valley, would receive 
the waters of its tributaries between Salisbury and Flaherty's Creek. 
At their lowest stage tlie.se tributaries gave, altogetlier, five feet in a 
second, and Cassellman's River above Salisbury, 15.33 cubic feet; 
total 20.33 cubic feet to feed the summit level. This quantity is not 
considerable when we consider, that, on a length of thirty miles from 
the summit level to Cumberland, the canal would have to dravv 
most of its water from Casselman's River: for Will's Creek is a tor- 
rent, which, in the greatest part of its course, gives but little water in 
summer. 
The length of this summit level, and of the route which the canal 
would thus trace, are less than by Deep Creek. As to their compa- 
rative heights, no survey was made in tJie season of 1824, to ascertain 
the difference. We shall now expose the reasons why the western 
branch of the canal was not led through the valley of the Mononga- 
hela (before concluding this part of our report.) 
We have already seen that the valley of Cheat River, through 
which it would be necessary to pass to the Monongahela, is divided 
from the Upper Youghiogeny by a ridge whose greatest depression, at 
the head of the two muddy creeks, is 226.77 feet above the level of the 
base mark. A tunnel would, therefore, be necessary to pass from 
the valley of the Youghiogeny to tliat of Cheat River. 
A single inspection of the map will show that the route of the canal 
would be very much lengthened by running its summit level from the 
heads of the north branch of the Potomac to those of Cheat River, and 
that it should be raised to a much higher level than on the route of 
Deep Creek. There is every reason to believe that the bed of Cheat 
River has a more rapid descent than tliat of the Youghiogeny, and that, 
where it forces through the Laurel Hill, it is already nearly on a leA-el 
with the Youghiogeny at Connellsville: for, at this gap, and a little 
above Furnace Run, it begins to be navigable. Its bed is here about 
150 yards wide. 
The highest floods in Cheat River do not rise above eight or ten 
feet at Furnace Run, and at its lowest stage in August and Septem- 
ber, it is very low at this place, and often'fordable. Indeed, Cheat 
River, to its junction with the Monongahela, receives no stream of 
any importance but the Big Sandy, whose supply is constant, but in 
the summer, is very trifling, even towards its mouth and in the lower 
part of its course. After descending along a rocky and very preci- 
pitious bed. Cheat River mingles its clear and limpid waters with 
tlie muddy stream of the Monongahela, whose bed and shores are all 
formed of alluvial soil. 
The Monongaliela has absolutely the same features as the Ohio; its 
shores are flat, but raised per])endicularly along both sides of the river 
to the height of fifteen or J wenfy-five feet above the line of water, fornu 
