872 MR. H. B. MILNER OlS^ THE [vol. Ixxviii,. 



The evidence of the existence of an ancient Eocene drainage 

 from north to south has ah'eady been mentioned ; as a matter of 

 fact, the occiuTence of scattered Chalk flints is by no means un- 

 common in Cornwall, as, for example, at Newquay ^ and Padstow,^ 

 Avhere in the latter localit}^ it is associated with Grreensand chert 

 as at Ludgvan,^ the area examined by Clement Reid. The 

 significance of this association of chert and flint is too strong 

 to be ignored ; clearh^ it suggests derivation from the east or • 

 north-east, v^here the typical developments of Cretaceous strata 

 are known to occur. If we bear in mind the j^i'obably greater 

 westward extent of such strata prioi' to early Tertiary erosion, it 

 would seem not imreasonable to conclude that this material was 

 carried south-westwards b}^ an ancient river- sj^stem draining approx- 

 imately along the line of the present Bristol Channel, receiving 

 affluents possibly from South Wales on its northern bank and from 

 Exmoor Forest and the northern part of Dartmoor on its southern 

 bank, and finally emptying itself into the English Channel by way 

 of the St. Erth valley. Such a drainage would be quite compatible 

 with the little knowledge that Ave possess of the physiography of 

 the South- West of England in the Palaeogene epoch, before the 

 renewal of folding in Miocene times. There is little evidence to 

 show that this folding entirel}^ obliterated any western drainage 

 previously existing ; on the contrarj^, such a trend Avas accentuated 

 rather than opposed, for a north-and-south Avatershed Avas probably 

 Avell established w^est of the CotsAA^old Hills in early Tertiary times, 

 differentiating a main east-and-Avest drainage, the former into the 

 Thames basin, the latter into the Bristol- Channel and Exmoor 

 region, although the precise location of this Avatershed is at present 

 obscure in so far as our information goes.^ Miocene folding 

 merely cut short the soutliAA^ard extension of that AA^atershed by the 

 superposition of an east-and-Avest line of flexures, noAv indicated by 

 such tectonic features as the Kingsclere, PcAvsey, and Mendip 

 uplifts. Consequently^, the ancestral drainage of rivers draining 

 westwards, far from being obliterated, AA^as resculptured to form 

 the basis of the present river-system in this region, and material 

 continued to be borne AA^estAvards until the Pliocene submergence 

 set in. With this submergence came the gradual advance of the 

 sea over the Miocene land-areas of the South- West of England,, 

 the droAvning of pre-existing loAvland topography, and the limita- 

 tion of the distributive poAver of the sediment-bearing rivers. If 

 the apportionment of the kyanite-grains in the Pliocene deposits 

 is any criterion of these fundamental physiographical changes, then 

 Ave can readily appreciate the absence of the species at St. Keverne,. 



1 Mem. Geol. Surv. Sheet 346 (1906) p. 65. 



2 Ihid. Sheets 335 & 336 (1910) p. 93. 

 '^ Ihid. Sheets 351 & 358 (1907) p. 68. 



'^ Some light was thrown on this problem recently by Mr. W. D. Varney 

 at a meeting of the G-eologists' Association on May 6th, 1921, when a paper 

 on the ' Geological History of the Pewsey Vale ' was read (Proc. Geol. Assoc 

 vol. xxxii, p. 189). 



