184 
K. F. Mather 
Moreover, the sag at K, the bench at M, and the channel at N* 
could not have been formed if the large valley had been occupied 
by a lake at that time. They must be the work of lateral drain- 
age between the front of the ice and the adjacent hill-sides. This 
stream, and the rest of the water from the melting ice must have 
had an unobstructed passage out of the valley, which, as has been 
shown, was occupied by the valley dependency. This outlet 
could only have been through the present Licking Narrows. This 
necessitates the placing of the capture of the Newark river by 
the Licking in pre-glacial times. 
Under this hypothesis the dilhculties at once vanish and all the 
phenomena of the region are satisfactorily accounted for. The 
tongue of ice occupying the larger valley would have formed a 
dam across the mouths of the tributaries J, B and C, and would 
have caused a lake to form in each of them. The connection 
across the sag K would have made the body of water continuous 
in valleys A and B and accounts for the correspondence between 
the elevations of the terraces in them. In valley C the water 
would probably stand at a different level and this corresponds with 
the observations of the heights of the present terraces. In fact, 
this is the only hypothesis that can correspond with the phenom- 
ena and account for two different lakes at different levels in val- 
leys B and C, and at the same time allow valley D to remain with- 
out fluvio-glacial deposits. 
The waters from the lake in valley A at first overflowed across 
the sag K into valley B. Here the water level rose until it reached 
the lowest place of exit, where the ice front abutted the flank of 
the hill at M. The channel was gradually cut down until it ( 
formed the present bench. The stream of glacial waters flowing 
along the face of the cliff from tributary B to valley L, between it 1 
and the ice front, has left faint traces of the work of degradation 
in the slope of the profile. This stream, reinforced by the waters 
of the lake in tributary cut, at the point N, the same sort of a 
bench as now exists at M. The stream here was much stronger 
than at the latter place, and the channel was cut much deeper. 
When the stream at M reached the level of the present bench 
a slight retreat of the ice occurred. This opened up the channel 
at L and separated the lake of valley A from that of valley B. 
The waters of the former then flowed across the bench at that i| 
point and carved it in the same manner as that already described | -j 
