444 ME. A. HAEKEB ON THE EEUPTIVE EOCKS IN THE 



Despite the superficial deposits, however, and notwithstanding 

 the absence of exposures in some of the critical localities, it is pos- 

 sible to draw conclusions regarding the relations of the rocks, some 

 of which may be considered reliable, while others are at least pro- 

 bable. 



The age of the sedimentary strata is naturally the first question, 

 and here unfortunately the evidence is very meagre. The fossils 

 recorded by Mr. Tawney and Sir A. Eamsay from Penarfynydd and 

 other localities near Llanfaelrhys are taken to indicate a horizon in 

 the Upper Arenig, and, judging by the observed strike of the strata, 

 there is no reason why the similar rocks (black shales, with occa- 

 sional beds of sandstone) seen to the east of the granitic mass 

 should not be referred to the same stage, although no organic remains 

 have been found. If not Arenig, these strata must be referred to 

 the lower part of the Bala series. The granite is bounded on the 

 north and north-west by the problematical green schists, which ex- 

 tend from Bardsey to Porth-dinlleyn ; but as the boundary is almost 

 certainly a faulted one, it is needless to discuss here the age of those 

 remarkable rocks. Por our purpose it is sufficient to know that 

 the beds in unfaulted contact with the eruptive masses are of Arenig 

 or, at the latest, Lower Bala age. 



II. The Gkanite and Gneissic Geanite. 



These rocks, the massive type largely predominating, cover a large 

 area on the map. The granite builds the round hill of Mynydd 

 Cefn-amwlch to the north-west of Sarn, and underlies the low 

 ground towards Bryn-croes, ranging as far south probably as Meil- 

 lionydd-bach, a distance of four miles. Around and to the west of 

 Bryn-croes, where the Survey map shows " greenstone," there are 

 no exposures ; but, judging by the form of the surface, the whole is 

 probably granite, like that to the north and south, and the two 

 " syenite " areas of the Geological Survey are united. The outlying 

 patch at Pyllau-giach, near Cefn-amwlch, also coloured as " green- 

 stone," is granite, and a tongue of the same rock extends from half 

 a mile south of Bryn-croes to the edge of the valley north of ITuriau. 

 The only place where the gneissic type of granite is exposed is in a 

 quarry at Meillionydd, in the extreme south of the mass ; even there 

 the character is only locally marked, and in the little hill to the 

 east the rock is the normal granite of the district. 



It is a biotite-granite or granitite, the mica partially giving place 

 to hornblende in a few localities only. The minerals composing the 

 rock are apatite, magnetite, pyrites, biotite, two felspars, and quartz, 

 with a chloritoid substance, epidote, actinolite, kaolin, leucoxene, 

 and ferruginous matter. Apatite and original iron-ores are sparingly 

 present. Biotite occurs in ragged flakes, often bent, and sometimes 

 showing " Gleitfldchen.'" It is normally deep brown, with the usual 

 characters of dichroism, but changes to green, giving then a bright 

 grass-green colour for vibrations parallel to the cleavage-traces, and 



