Vol. 58.] GLACIER-LAKES IN THE CLEVELAND HILLS. 541 
the ice-edge retreated, the formation of two other marginal lakes 
similar to that in the valley of Kirk-Moor Beck, and each of these 
had its own separate overflow-channel cutting through the narrow 
watershed between the Robin-Hood’s Bay drainage-slope and the 
great intercepting-drain of Jugger-Howe Beck. These overflows 
doubtless originated in the channels produced by streams running 
directly off the ice-front itself as it slowly retreated. 
Some channels such as ] describe were thrown out of action, as soon 
as the ice shrank well away from the watershed, and lateral communi- 
cations were opened up. One good example of these temporary over- 
flows isseen running from the cross roads near Evan-Howe farmhouse 
towards Evan-Howe Slack. 
We now have the recess of Robin Hood’s Bay occupied by three 
Jakes—Kirk-Moor Lake on the north, Evan-Howe Lake in the 
centre, and Blacksmith-Hill Lake on the south. 
The overflow of the first of these by Foulsike I have already 
described. The second, Evan-Howe Lake, was drained by a gigantic 
trench, Evan-Howe Slack, a valley about half a mile long, 100 yards 
wide across the floor, and nearly 100 feet deep. There is at present 
no stream in the upper two-thirds of its length, and only a dimi- 
nutive rivulet in the lower part; its broad intake 1s probably at 
about 535 or 540 feet above sea-level. 
During the shrinkage of the ice-lobe, the spur separating this lake 
from Kirk-Moor Lake to the northward was gradually uncovered, 
and, as the outlet of Kvan-Howe Lake was lower than that of 
its neighbour, channels were cut across the lower parts of the spur 
whereby Kirk-Moor Lake was drained into Evan-Howe Lake. 
Therefore the Foulsike overflow now became functionless. ‘Two of 
these lesser channels are traceable at Moor-Close Plantation and 
Chapel Garth respectively. The former has its intake somewhat above 
550 feet O.D., and the latter is a little lower. Iam unable to explain 
why the levels should be arranged as they are. If the Chapel- 
Garth overflow, which is the farthest from the ice-front, were a deep 
and strong cut, I should be inclined to suggest that it had been re- 
deepened in consequence of the closure of its rival by a forward 
oscillation of the ice, but in fact it 1s a very shallow and ill-defined 
valley. 
The southern lakelet drained by the Blacksmith-Hill overflow, 
which is a gorge almost exactly comparabie in its dimensions with 
Evan-Howe Slack. It is about 1000 yards long, and 50 to over 
100 feet deep; a small stream occupies about two-thirds of its 
length. The intake is about 510 feet above O.D.; but, as in all 
other large channels, there is a good deal of peat in the floor. The 
lake was probably from its inception fed by the drainage previously 
flowing down Burn-Howe Dale, which channel was consequently 
abandoned. ‘Ihis is indicated clearly by the levels—the intake of 
Burn-Howe Dale is above the 600-foot contour, but the watershed 
upholding the Blacksmith lakelet for more than half a mile les 
below that level. The retreat of the ice-front threw the central and 
southern lakes into confluence ; and as the Blacksmith-Hill outlet was 
