558 REPORT—1904, 
2, Hvidence in the Secondary Rocks of Persistent Movement in the 
Charnian Range. By Professor Percy F, KenpAtt. 
3. Rwver-capture in the Don System. By Rev. W. Lower Carter, IA. 
The river Don has a remarkable semicircular course. Rising in the Middle 
Grits west of Dunford Bridge at 1,500 feet above O.D., it flows eastwards to Peni- 
stone (700’), where it makes a bend to the south-east, quickly deepens its valley to 
500’, and at Wortley breaks through the great watershed (1,000’) of the Greno- 
side and Wharncliffe grits. It then receives the Little Don, the Ewden, and the 
Loxley, on its right bank, and falls into the valley of the Sheaf at Sheffield (150’). 
The Don then makes a rectangular bend to the north-east, following the old valley 
of the Sheaf to Conisborough, receiving the Rother on its right bank at Rother- 
ham (87’) and the Dearne on its left bank at Denaby (45’). It then traverses the 
Maguesian Limestone escarpment in a fine gorge and continues past Doncaster in a 
north-easterly direction to Thorne, where it bends northwards towards the Aire, 
which it formerly entered at Snaith. It has, however, been artificially diverted 
by the Dutch river to the Ouse at Goole. 
The history of the present river course is presumed to have commenced when 
the Pennine anticlinal rose from the Cretaceous sea, and the original consequent 
streams commenced to run down the dip-slope of the Chalk. Slack Beck (Broad- 
stone Dyke), which is diverted south-east at Ingberchworth by a tributary of the 
Don, is considered to be the head-stream of the brook that runsby Cawthorne, only 
a narrow dip in the watershed dividing them. The Don at Penistone (700’) faces 
a watershed of 700 feet, which forms a dip between Hoyland Swaine (900’) 
and Thurgoland (810’). Immediately beyond this watershed are the head- 
waters of the Dove, flowing eastward in direct continuation of the course of 
the Don above Penistone. The Dove is thus considered to be the beheaded 
remnant of the Don. The southerly bend of the Don and the cutting of the 
Wharncliffe gorge are explained as due to river-capture by a feeder of the Sheaf. 
This Wharncliffe stream, with a rapid fall to the Sheaf, was able to capture 
successively the Loxley, the Ewden, and the Little Don, and then the watershed 
at Wortley was attacked by a branch of this stream, and on the other side by a 
feeder of the Don, As the watershed was cut through, the Wharncliffe stream, 
by reason of its steeper fall, captured the Wortley feeder of the Don and then the 
Don itself. 
The Dearne.—At a very early date the Bretton stream must have been captured 
by the Darton feeder of the Cawthorne stream, as it flows straight at the Woolley 
Edge escarpment (527’), and therefore must have been captured before the land was 
reduced to this level. The Dearne flows eastwards, by Barnsley to Cudworth 
Common, where it males a rectangular bend southwards, and cutting through the 
Upper Chevet Rock (225’) at Darfield, enters the old valley of the Dove (100). 
This gorge at Darfield proves the extension of the 225-foot contour eastwards, 
towards Hickleton, forming the watershed between the Dearne and the Dove, and 
there is an old river valley at Frickley (200’) between Clayton and Hickleton, 
which was probably the original course of the Dearne, which flowed through 
Hampole gorge into the central plain. The Darfield gorge is a case of river- 
capture by a feeder of the Dove. The Dove itself had probably been captured by 
the Sheat at a period hefore the present level of the Magnesian Limestone esearp- 
ment was reached bv denudation. 
The Rother.—Vhe original cousequents of the Rother are Shire Brook, the Moss, 
and the Stayeley stream. The Shire and Moss probably coalesced and formed the 
head-waters of the Ryton. The two gorges (530’) uniting at Kiveton are plainly 
traceable, and have subsequently been used, in all probability, as a channel of 
glacial overflow, The Moss must have captured the Staveley stream before it was 
itself captured by the Rother. 
The whole inner Don system is thus explainable by a series of riyer-captures, 
