56 PROCEEDINGS COTTESWOLD CLUB 1901 
of these beds extend to the west about as far as Bridgend 
on the Welsh coast and to Porlock in West Somerset. The 
mountainous land of the South Wales Coalfield formed 
the rim of the valley on the north, running parallel to the 
Exmoor range on the south. The hollow between these 
rims must have been excavated by a river, for it was formed 
before the subsidence of the land permitted the access of 
the Triassic Sea. The drainage of the valley was 
apparently eastward. This is rendered probable by the 
absence of Mesozoic rocks to the west of the hypothetical 
valley, and the probability is strengthened by the con- 
tinuity of the Trias of the valley with the band of Trias 
that runs from Bridgwater Bay into South Devon. The 
Channel valley was doubtless widened by the waves of 
the Triassic Sea. Originally, it narrowed towards the 
west, and received a river draining high land in that 
direction, as well as streams from South Wales, Devon, 
and West Somerset. How far the Channel valley 
was continued to the west we are unable to determine, 
but we cannot safely carry it more than half-way between 
Weston-super-Mare and St. David’s Head. 
The excavation of the Channel valley took place along 
the syncline between the Devonian anticline of North 
Devon and West Somerset and the anticline which con- 
tinued the Mendip anticlinal axis to the west. If this 
axis is prolonged in a westerly direction, it is seen to pass 
along the northern side of the present Channel, not far 
from the southern lip of the great coal-field. So also the 
syncline between the Mendips and Bridgwater, if produced 
to the west, coincides with the axis of the Channel. The 
old Channel river must, therefore, have been a strike-stream, 
and must have been one of the tributaries of a primary 
river, draining Mendip land, probably on its southern 
watershed. We may suppose, with some probability, that 
this primary flowed along the line indicated by the Triassic 
a 
