383 
Another effect of these upheavals was to cause the accumu- 
’ lation of sediment, which in the case of the Cotteswold sea 
came from the North, against the Southern isthmus. The 
accumulation of this sediment soon raised the sea-bottom near 
the southern shore, and placed it in such a position that it 
received no further deposits. Thus we find that the limit of 
deposit during the Murchisone-zone is represented by a line, 
a—a, running from just above Bath, eastward of Cirencester, 
to Northleach and Rissington.* Southwards of that line was 
an area which received no deposit,+ thus increasing the barrier 
between the Dorset-Somerset seas and the Cotteswold sea. 
The limits of deposition during the Concavum-zone do not 
extend so far south. The line, b—b, running just below Stroud 
marks the southern limits of these beds. When we come to 
the time of the Humphriesianum-zone, we find no trace of it in 
the Cotteswolds; and my supposition is, that, if deposited 
anywhere, it was northwards of the present Cotteswold Hills, 
and has therefore all been removed by denudation. 
This retreat of the area of deposition towards the north is 
paralleled by a similar retreat southwards and eastwards in the 
Dorset-Somerset sea, for we find the part covered by the 
Humphriesianum zone is narrowed to an area extending from 
Burton Bradstock to. Sherborne and Milborne Wick, c—c. 
Assuming that Dundry was connected with the Dorset-Somerset 
area round the western end of the Mendips—and it is singular 
to find that even the lithology of the Humphriesianum-zone is 
very similar at Dundry, at Sherborne, at Burton Bradstock, and 
even at Bayeux in Normandy—we must suppose that the Dundry 
* The Parkinsoni-zone rests on the Midford Sands at Bath, and on Upper 
Lias at Rissington and Northleach. 
+ As remarked by the President, the absence of beds over certain areas 
does not always imply their non-deposition ; but it seems to me that this 
theory fits in with the facts of the case better than to suppose the deposition 
of these beds, and again, their subsequent total denudation prior to the time 
of the Parkinsoni-zone. My idea is that the accumulation of strata in the 
southern part of the Cotteswold sea continually forced the area of deposition 
further northwards, until the subsidence which allowed the sea of the 
Parkinsoni-zone to overflow all these areas. 
