Vol. 69.] PETROGKAPHl' OF BARDSEY ISLAND. 529 



The various specimens selected for slicing are so like one another 

 in composition and structure, that there is no difficulty in believing 

 that they are all of closely allied origin ; they ma}-, in fact, 

 represent one sill repeated by folding. 



Both muscovite and biotite are present in the slides, the latter 

 being the more abundant, though it is practically always replaced 

 by chlorite. Part of the muscovite is primary, but there is also 

 secondary white mica developed from the felspar. The rock has 

 contained much orthoclase, though this is seldom in good preserva- 

 tion, being mostly converted into ' shimmer-aggregates ' of sericite. 

 A fresh polysynthetic felspar is also common, and proves on being 

 tested always to be albite. Zircon, iron-ores, and sphene are 

 the accessories, and, in addition to sericite and chlorite, some- 

 times epidote, rutile, limonite, and carbonates occur as secondary 

 minerals. 



The structure is cataclastic in a high degree, the quartz being 

 crushed, the felspar broken and sericitized, and the mica drawn out 

 into irregular wisps and streaks. Some of the slides show a good 

 •deal of fine granulitic material, consisting of quartz, felspar, and 

 white mica. The granitic structure is often sufficiently clear how- 

 ever, especially at Ogof Hir [9*293] and Bau y Rhigol [9288]. 



Towards the margins of the sills there is a development of a 

 porphyritic facies, which has an abundant microcrystalline or 

 felsitic matrix with phenocrysts of original muscovite, in addition 

 to quartz and felspar [9289, north-western corner of the island ; 

 .9291, Ogof Hir]. 



The question arises whether the granite has any connexion 

 with the pillow-lavas and diabases. In several parts of Britain 

 (Tayvallicb, 1 Porthallow ') granitic rocks occur intrusive into pillow- 

 lavas, in such a way as to suggest that they are of kindred origin. 

 These granites are generally rich in soda, and contain much albite. 

 The Bardsey granite has a fair amount of albite, though not so 

 much as the ' soda-granite ' of Porthallow. It seems quite likely, 

 although in the circumstances it cannot be definitely proved, that 

 in Bardsey Island also the pillow-lava eruptions terminated by the 

 intrusion of an acid magma, now represented by the Bardsey 

 granite. 



(2) The Pillow-Lavas and their Tuffs. 



The rocks of this group are much altered by decomposition and 

 by pressure. Their felspars are replaced by sericite and carbonates ; 

 their femic minerals by carbonates and chlorite ; their iron-ores by 

 leucoxene and limonite. Little of their original structure remains, 

 ■except traces of phenocrysts and of amygduloidal cavities filled with 

 calcite and chlorite. Crushing has gone so far as to produce a well- 

 marked cleavage, and in some cases an irregular schistosity. Many 



1 ' The Geology of Knapdale, Jura, & North Kin tyre ' Mem. Geol. Surv. Scot. 

 1911, p. 93. 



2 ' The Geology of the Lizard & Meueage' Mem. Geol. Surv. 1912, p. 186. 



