286 PROF. A. II. COX AND MR. A. K. WELLS ON THE [vol. lxxvi, 



This statement is completely justified by further evidence that has 

 been obtained during the examination of the district. 



The south-westward transgression on to higher beds at once 

 suggests that the sill had its origin at some point away to the 

 north-east, and that the fiow took place from the north-east where 

 the rocks have since been removed by denudation, towards the 

 south-west, in which direction the sill now dives deep underground. 

 Further evidence in support of this view may be adduced from the 

 •distribution of the minor granoplryric intrusions. 



The actual point of origin of the great sill remains undeter- 

 mined. The north-eastern extremity lies buried beneath the 

 alluvium of an old, matured river-valley. There is accordingly an 

 area having a width of up to a third of a mile where exposures 

 cease, and it is quite possible that the root of the sill is situated 

 under this alluvium in the neighbourhood pt Tal-y-waen. No other 

 intrusion of acid rock has been found between this point and the 

 boundary of the estuary, a mile to the north. Then comes another 

 stretch of alluvium half a mile to a mile wide, that might con- 

 ceivably conceal the duct through which the magma rose. North 

 of the estuary there is, so far as we are aware, no intrusion of 

 granophyre (or microgranite) comparable to that of the Crogenen 

 granophyre ; no such rock is present in the area examined by 

 Dr. A. fi. Andrew. 1 It is probable, therefore, that the duct is 

 situated under the sill itself, or under the alluvium that borders it 

 in the neighbourhood of Tal-3 r -waen. This supposition is supported 

 hy the fact that the minor granophyric intrusions are concentrated 

 near this lccalit} r . 



Basic marginal modifications of the Crogenen grano- 

 phyre. — Along both upper and lower margins the Crogenen 

 granophyre passes into darker basic rocks, which in their colour, 

 mode of weathering, etc. closely simulate a diabase. The marginal 

 zones are each from 50 to 100 feet thick, and within this 

 thickness there is usually a perfect transition from thoroughly 

 acid to thoroughly basic rock. These marginal rocks evidently 

 represent a differentiation-product of the main granophyre, and 

 the perfect transition between the two types and the general absence 

 of any pronounced brecciation suggest that the differentiation was 

 ■accomplished in place. This view is further strengthened by the 

 uniform character which the marginal zones, although so thin, 

 maintain over many miles. Such uniformity of comparatively 

 thin zones over great distances could not be expected if the basic 

 rocks represented an earlier and deep-seated differentiated product. 

 Exceptionally, intrusion-breccias do occur, a good example being 

 found at the western end of Pared-y-Cefn Hir. There the earlier 

 ■consolidated marginal basic product has been fractured, as the 

 result of some local movement, and the fractures injected with 



1 ' The Geology of the Dolgelley Gold- Belt ' Geol. Mag. 1910, p. 159. 



