308 REV. J. F. BLAKE ON THE 
been met with in association with the porphyries, which can clearly 
be massive enough without it. The character, therefore, of the 
granite stands, at present, on the St. David’s evidence alone. But 
in reality there is much more on the island to be seen yet. Mine 
was but a rapid and incomplete inspection. 
On the eastern side of St. David’s is another small area, almost 
a repetition of some parts of Ramsey Island. It is situated south 
of Pointz Castle, at the west end of Newgale Bay. On the 
western side it is brought against Cambrian slates by a fault, and 
there resembles somewhat the porphyry of Porth Hayog; but I have 
not examined it microscopically. Here also, in inaccessible spots, 
very good evidence of apparent bedding is visible; but it is on the 
eastern side that the most instructive development and section are 
seen. There we see great crags of banded felsite, looking, at a 
distance, like gneiss with the bands vertical. This, in structure, is 
quite comparable to the Pwll Hendro rock, and is equally enigmatical. 
On the shore, however, we have a second Ogof Llesugn, which has 
met apparently with a second misinterpretation. There is a deep 
cave sloping to the north, the upper face of which is Cambrian brown 
sandstone, the conglomerate not appearing here ; the under face is the 
banded felsite; the line of junction is a fault extremely brecciated : 
but no hardening has taken place, for there is no dyke. The mass 
to the south is very much bedded, but with no regularity, some being 
vertical, other parts in a synclinal; then again is another caye-pro- 
ducing fault, on the south of which the rocks are of an ashy character. 
This is an admirable confirmation of previous experiences. Accord- 
ing to Dr. Hicks, it forms part of a massive, of similar constitution, 
which stretches away from Roche Castle to Trefgarn. It was 
inevitable, therefore, to visit the Roche-Castle rock. I came away 
with the impression that it had nothing in common with the series 
under examination ; and if pre-Cambrian in age, this must be proved 
independently; and the same must be said of the Brawdy granite, 
which, to my eye, has little resemblance to that of St. David’s. The 
Roche-Castle rock, however, was so peculiar in appearance and 
so tough that I have examined it microscopically, and thus am 
confirmed as to its distinct character. It seems to have been 
originally a kind of andesite, as it consisted of large felspar crystals 
in an abundant ground-mass. Now, however, only the forms of the 
crystals are to be seen in ordinary light, while with crossed Nicols 
the boundaries become rather obscure, because the original felspar 
has all been replaced by quartz, which first lhmed the empty cavity ~ 
with hexagonal pyramids and then filled the rest with wre 
erystallization. The quartz-filling is thus exactly similar to that of 
the St. David’s rocks, though here in excess: but this is a subsequent 
alteration affecting alike rocks of very different character and possibly 
of very different age. 
Such is the evidence on which I depend in support of the view of 
the essentially igneous character of the whole series. They here 
seem united into a connected whole, and such I believe them to be; 
a belief rendered more probable and, indeed, induced by the known 
