part 4] PLiocEJSTE deposits of coenwall. 367 



Avitli the exception of two or possibly three of the mineral species 

 ■characterizing- these dej^osits, all occur in rocks known to-dav in 

 the vicinity of tlie outcrops under consideration : in other words, 

 85 per cent, of the species are probably of local origin. With 

 regard to xenotime, tliis mineral has not been recorded from the 

 acid igneous rocks of Cornwall, so far as I. am aware ; but there is 

 no reason why it should not be present in some of the granite- 

 masses, even though sparsely distributed. Staurolite is essentially 

 an ' Armorican ' mineral, with a t3q:)e-locality in Britann}^ to-day: 

 its origin may be more pertinently discussed hereal'ter (p. 870). 

 The kyanite is probably in the nature of a reman ie mineral, and 

 its crystallographic features point to derivation from a pre-existing 

 sediment, such as the Lower Greensand ; the nearest outcrop of the 



Lower Greensand, containing' kvanite similar to that of St. Asfnes 



. ... ~ 



and St. Erth, is in the north-west of Wiltshire : the significance 



of this fact will be ultimately apparent. 



The cumulative evidence to be obtained from a stud}^ of these 



minerals shows them to have belonged originall}^ to a distributive 



province which included acid, basic, and ultrabasic rock-types, 



with an associated suite of mel amorphic rocks : such a ])rovince 



clearl}^ is only indicative of Cornwall itself Avhere all such rock- 



facies are known to occur. Excluding, then, the staurolite and 



kyanite, the origin of which is discussed hereafter, the mainly 



local derivation of the deposits under investigation 



would seem to be most decidedly indicated, 



VIII. Pliocene Geouiiaphy of the South-West 



OF EXGLAKD. 



One of the most difficult problems facing the geologist in 

 Western Cornwall is the adequate explanation of certain out- 

 standing ph^^sical features, for which there is but little direct 

 geological evidence ; among such features that of the St. Erth 

 valley (see map, fig. 1, ]). 350) is probably the most striking, and 

 a close investigation of its nature and origin is a sine qua non in 

 ,any intelligent discussion of Pliocene palseogeography. 



The late Clement lieid, in a paper ' On the Probable Occurrence 

 of an Eocene Outlier off the Cornish Coast,' ^ drew attention to this 

 valley, and interpreted its origin in the light of his theory of the 

 existence of Eocene gravels on the floor of Mounts Ba3^ He in- 

 clined to the view that the St. Erth valley ^vas the site of an old 

 Eocene river flowing probably from north to south through this 

 valley, the river being responsible for the transportation and 

 accumulation of the material in the bay. Thence the inference is 

 that the initiation of this valley was the product of early Eocene 

 erosion, the material, composed of 86 per cent, of Chalk flint and 

 2 per cent, of Greensand chert, being derived from Cretaceous 

 rocks, presumably from the north-east. That the valley was not 

 originally formed in Pliocene times is apparent from careful 



1 Q. J. G. S. vol. Ix (1904) p. 113, 



