314 Gerald M. Part— 



must be almost pure albite — a conclusion borne out by an ex- 

 amination of cleavage fragments. 



The ground mass is rather variable in character. It consists of 

 granular quartz, small laths of felspar, mainly albites about -Ol in. 

 in length, with interstitial chlorite. There are a few needles 

 of apatite, granules of secondarjf sphene, and occasional grains of 

 an iron-ore mineral. In one or two specimens there is a little 

 pyrites or limonite pseudomorphs after it. Vesicles are imcommon, 

 t)ut when they occur are filled with quartz. In structure the ground- 

 mass usually approaches the " micro-poecilitic " type so frequently 

 described in felsites, and their general appearance suggests that most 

 of these rocks have suffered more or less devitrification and some of 

 the quartz may be secondary. The granularity of the structure is, 

 however, modified by the small laths of felspar which frequently 

 exhibit a marked fluxional arrangement. In those cases where 

 the flow is not very pronounced in the ground-mass, there is usually 

 a distinct parallelism of the smaller phenocrysts. 



The chlorite occurs in grains, scaly or fibrous aggregates and 

 irregular patches lyiiig between the other constituents. Only rarely 

 has anything in the nature of well-formed pseudomorphs been noted, 

 and in the majority of cases there is no evidence as to what was the 

 original ferio-magnesian constituent of these rocks. In the rock 

 south-west of Banc-du there are numerous pseudomorphs of a pale 

 yellowish chloritic substance after some monoclinic mineral, and 

 usually having an irregular border of granular green chlorite (Fig. 2). 

 Some of these pseudomorphs have an external form which may 

 represent hornblende, as in the rocks of Coomb, Carmarthenshire.^ 

 In other cases it is possible that some of the mineral was a pyroxene, 

 and in a rock from Goetty Mountain there are a few small pseudo- 

 morphs in which the pyroxene form is rather more definitely 

 developed. In one or two cases sm,all jDleochroic haloes have been 

 observed in some of the more irregular patches of chlorite. It is 

 possible that mica occurred originally, but this does not seem to 

 have been more than a very minor constituent. In one specimen a 

 few relatively large zircons were observed, as m. the rhyolites of 

 Foel Trigarn. 



The rock exposed on the eastern side of Banc-du, forming the 

 Tipper surface of the mass, is now very different in character to the 

 rest, though it would appear to have been identical originally. 

 It is now a very fine-grained aggregate of qiiartz, white mica, and 

 a little limonite, with interstitial chlorite in the ground-mass serving 

 to distinguish this portion from the phenocrysts. There are a few 

 rare pseudomorphs in chlorite, similar to those found in the Coomb 

 rocks. Sections of the actual junction with the overlying slates 

 show the slight chilling at the surface, which has also been some- 

 what brecciated. This surface, however, is very irregular and the 



1 CantriU & Thomas, op. cit., p. 243. 



