568 REPORT—1904. 
the Kiveton gorge seems to have presented a clear course for the overflow of the lake 
formed by the damming back of the drainage. The second glacier appears to have 
retreated north of the Aire before the overflows at the head of Calderdale were in 
full swing. The Don and Dearne valleys were, therefore, in all probability, clear 
of ice during the later part of the Glacial Period, and have been subjected to 
enormous denudation, both during the Glacial Period and since, which has cleared 
away the bulk of the boulder clay and only left relics of previously widespread 
deposits. 
ii. Glacial Lakes and Overflow Valleys. 
Such a series of glacier movements as has just been indicated would divert the 
normal drainage of the district and produce lakes in the valleys thus dammed up, 
The boulder clay at Ashfield’s Pit, and near the railway station at Conisborough, 
and at Cadeby, on the opposite side of the Don, shows that this gorge must have 
been filled with ice up to the 225-foot contour. The scattered patches of drift 
from Edlington to Clifton and Braithwell, reaching up to 400 feet, indicate that 
the gorge was entirely closed above the 350-foot contour. This is the general 
height of the Midland watershed of the Don system, and is only broken through 
at one point south of Conisborough, the Kiveton Valley (330 feet), near the middle 
of which one of the sources of the river Ryton takes its rise. These considerations 
warrant one in assuming the existence of a great glacial lake, rising to the level of 
the 330-foot contour to the west and south, and dammed back by ice from 
Jonisborough to Barnsley, This lake would overflow by the Kiveton gorge 
towards Worksop. One cannot expect to find abundant evidences of lake deposits 
in an area which has suffered so severely by denudation as this; but the geological 
surveyors map from 4 feet to 9 feet of brick earth and clay resting on gravel at 
Parkgate, and from 3 feet to 7 feet of brick earth near Wombwell. These 
indicate a lake both in the Don and Dearne Valleys, covering up the old river 
gravels. 
Following this line of argument, and taking the various patches of drift as the 
relics of moraines, and therefore as indications of periods of rest in glacial 
movement, I have attempted to map out the lakes that would be produced at the 
different positions of the ice-front, and have examined the watersheds to see if 
overflow channels existed such as would be necessary to drain such lakes. The 
whole has been plotted out on the 6-inch contoured maps, by which the results 
have heen carefully tested, and a series of lakes made out discharging successively 
over cols from 175 feet to 335 feet above O.D. These overflow valleys are not of 
the type so characteristic of Cleveland and the Cheviots. The long period of sub- 
aérial denudation to which they have been subjected has worn back their sides so 
that they are now V-shaped, but they are streamless either in whole or in part, 
and often the nearest streams cut across their ends. 
In spite of this weathering back there has probably been little alteration of 
their level, and their present levels may be taken approximately as those of the 
Glacial Period. Some of them are strike-valleys formed by the denudation of the 
shales between the outcrop of a bed of Carboniferous sandstone and the dip slope 
of a lower grit. The objections against such valleys as overflows have been 
carefully considered, but as the movement of the ice seems to have brought its 
margin parallel to the general strike of the Coal Measures of this area, it is natural 
that the deflected drainage should sometimes escape by such routes. In consider- 
ing the course of the first glacier, it seems probable that it would dam up the 
Dearne at Ardsley and form a lake overflowing by the Stairfoot Valley at 175 feet. 
A forward movement would carry it to the Wombwell ridge, and the overflow 
would be by the Wombwell and Swinton strike-valleys. Further south the ice 
would probably abut against the projecting spur of the 350-foot contour west of 
Rawmarsh, and hence would form a lake about that level stretching up to Elsecar, 
Cawthorne, and Bretton. In searching the watershed for a possible overflow for 
such a lake, a narrow cut through the 350-foot contour was found at the head of 
the Wentworth Woodhouse Valley, sloping back to the 400-foot contour on each 
hand, and with a little stream running across each end at right angles to the 
