VOLCANIC GROUP OF SI. DAVID’S. 307 
ordinary light, the characteristic cracks of perlite are perfectly 
visible. Of the character of this rock there cannot be the shadow of 
a doubt; it is the oldest perlite known*. Towards the summit its 
glassy character is less marked, and a similar flow from the northern 
summit is also more devitrified in aspect ; but I have not examined it 
microscopically. 
In the hollow known as Pwll Hendro the scene is geologi- 
eally superb. On the remoter southern side, accessible only to the 
thousand sea-birds who here find habitation, are massive columnar- 
jointed rocks of, probably, the porphyry ; ; but on the rugged 
precipitous slopes which le on the north, the geologist scrambles 
over apparently vertical beds of finely stratified rock. Each several 
band, though varying in compactness, consists of a series of sphe- 
rules of greater or less size, very clearly and beautifully seen on the 
weathered surface; the finer bands have almost the aspect of a 
sandstone, but the structure is the same. These rocks, too, are lost 
in a point as they are traced up the hill, and I doubt whether 
they are stratified at all. I expected, indeed, to find that these 
spherules would show a radiate crystallization; but in that I was 
disappointed. The spherules consist of the same crystalline material 
as the remainder ; but the individual crystals are much larger and 
resist the weather better. Nevertheless, I think it may be a de- 
vitrified rhyolite, as no such structure, I believe, is known in meta- 
morphic rocks, and it would be difficult to account for if it were. 
There seem, however, to be fragments of crystals included in some 
portions of the mass here developed, which renders its origin 
most perplexing. The non-spherulitic rock is rather wide-spread, 
and is divided by a great vertical slice of conglomerate almost 
made of it, and let in by faults. If this be the Cambrian con- 
glomerate, the rocks of which it is composed must be pre-Cam- 
brian ; and if of more recent date they may be so still. One other 
amongst several interesting patches of rock here found must be 
Hobiced: It is brown in colour, and similar to rocks on the N.-E of 
St. David’s. This is the most basic rock I have seen unassociated 
with the ashes. Itis composed of large crystals of either kind of 
felspar with little, if any, interstitial matter; quartz is almost 
absent, but there is no inconsiderable quantity of a highly polarizing 
acicular mineral, probably representing the magnesian constituent. 
This is most like the rock which near Whitchurch lies below the 
conglomerate, and is generally described as a felspar porphyry ; : 
but this latter contains far more of the highly polarizing ingredient, 
and even in some respects suggests the question whether it was 
not originally a tuff. 
In the whole of this series the only doubtful rock is the banded 
spherulite of Pwll Hendro, the bands of which, by the way, run due 
K. and W., and not N. and S., as ‘‘ Arvonian” rocks are supposed to 
do; but even this in no way removes the series from the essentially 
volcanic group, both in the lie and in the ultimate structure of the 
rocks. It must be admitted that as yet no sign of the granite has 
* Unless those described by Mr. Allport be of the same age. 
x2 
~ eee SS 
ee a 
