[32 
52 
vejs slionld be.£:in at a base mark, six inches above the bottom of the 
outlet of that lake, at its briH^^e. The summit level line should run 
north as far as possible, towards the rid.2;e which divides Coiineaut 
lake from Conneaut creek of the lake, and south to the point, where it 
will be necessary to drop it. By this level we will n>oasure and cal- 
culate the deep cuts which will be i-equircd through the ridges \Ahich 
bound north or south the basin of the lake, and judge w hether the le- 
vel of that line should be raised to diminish the deptli of these cuts, or 
lowered, in order that Conneaut lake may be high enough for a re- 
servoir; in short, it will shew relatively to its southern section 
whether the summit level should be separated from the lake. 
Having fixed the summit level line, we should run it as far north as 
possible towaid the valley of Conneaut creek of the lakes, and de- 
scend the right shore of that valley- on levels as long as possible, to 
reach Lake Erie at the mouth of Elk creek. All the localities of 
this spot must be examined with care, to determine what work will be 
necessary to form a port in the Lake. 
To the south of the summit level, two routes may be tried to descend 
to Greenville: the one through Siienango creek, and the other 
through Crooked creek. From Greenville, the canal will descend 
Shenango Creek to the mouth of the Pymatuning, from whence it fol- 
lows the same route as the first which we have analyzed. At Ibis 
spot bench marks should be established to compare the levels of thes« 
two routes. 
, A feeder, with a slope of six inches per mile, should be traced from 
Conneaut lake to French creek, following the left shore of Conneaut 
creek and the right shore of French creek, directing itself towards 
Meadville. The surface of Conneaut lake, at its usual level, should 
be measured, and, in the supposition that its bed should be raised three, 
" six, or nine feet higher, the level of the surrounding ground should 
he taken on each of these hypothesis, in order to calculate the height, 
length, and dimensions of the dams, which would be required, itj or- 
der to confine this reservoir within the bounds, which would be ne- 
cessary. The elevation of the dam required to keep the waters of 
French creek at the elevation of the summit level, and its location 
below Conneaut creek, or below Little Sugar creek, will then be fixed. 
It* this elevation below Sugar creek was moderate, it would ])rocure 
on the summit level an open passage, through which the trade of Big 
Sugar creek and the Alleghany above Fianklin might pass to the ca- 
nal, and form, moreover, a vast reservoir to supj)ly a!l its wants. 
Lastly, we shonld examine whether a branch might not be run 
fiom the northern extremity of the canal to Erie, (Presqu* isle) either 
by following the banks of the Lake, or crossing Elk creek and Wal- 
nut crerk. 
Fonrlh route. From Pittsburg, it might ascend the valley of the 
Alleghany to Franklin, and then ascend French creek and Conneaut 
Citek to the summit level of the abovementioned route, following 
the same directions in its subsequent position. 
The summit level and northern section of this route would be the 
same as those of the former. Its southern section, afler descending' 
