part 4] PLIOCENE deposits op coenwaLl. S7S 



its appearance at St. Erth, and its still commoner occurrence at 

 St. Agnes, St. Keverne being the southernmost and St.- Agnes 

 the northernmost locality. The probable trend of advance of the 

 Pliocene sea was from south-west to north-east, and obviously as 

 this sea deepened, a constantly diminishing amount of material 

 would find its way from the north-east to the submerged parts of 

 Cornwall, and the influence of the drainage would be progressively 

 curtailed. Again, the occurrence of staurolite at St. Agnes would 

 seem to suggest the influence of south-westerly marine currents at 

 least as far north as that locality, which further explains the 

 absence of kyanite in the extreme south. 



Finally, the possibility of derivation of the kyanite from any 

 pre-existing Eocene deposits in South-Western Cornwall is nega- 

 tived by the entire absence of Chalk flints and Greensand cherts 

 from the Pliocene deposits, indicating the extensive erosion that- 

 the older deposits suffered in early and Middle Tertiary times. 



IX. Summary and Conclusions. 



The results of the investigation may now be summarized as 

 follows : — 



(1) The petrographical characters of the St. Agnes, St. Erth^ 

 and St. Keverne deposits have been shown to be substantially the 

 same, and they suggest derivation from rocks belonging to a 

 homogeneous ' distributive province ' such as Avould be furnished 

 by the Palaeozoic rocks of Cornwall. 



(2) The geological age of the St. Erth Beds has previously been 

 proved palseontologically to be early Pliocene. Although no f ossil& 

 occur in the St. Agnes or St. Keverne deposits, and since the latter 

 are the products of contemporaneous erosion, by petrographical 

 correlation their Pliocene age is established. 



(3) In the case of the alleged occurrences of Pliocene material 

 at Canons Town and Polcrebo, in neither instance is the evidence 

 sufficient to establish relationship with the other deposits, although 

 there is little doubt that, if an outcrop exists at the former locality, 

 it has a direct connexion Avith the St. Erth material on the 

 opposite side of the valle}^ The Polcrebo gravels are, in my 

 opinion, the product of much later erosion. 



(4) The topographical evidence furnished by the ' 400-foot ' 

 plateau has been used, in conjunction with the petrographical 

 investigation of the deposits associated with it, as a means of 

 reconstructing the early Pliocene geography of Cornwall. The 

 occurrence and distribution of the two species, staurolite and 

 kjT^anite, have been shown to be of specific value in this connexion,, 

 the former as connoting the existence of the ancient Armorican 

 land-mass, the latter as indicating the direction of drainage from 

 the north-east, at the close of Miocene and the commencement of 

 Pliocene times. ■ 



