VOLCANIC GROUP OF ST. DAVID’S, 297 
into contact with and has been pushed over the slate *. In the bay 
below Nun’s Chapel we have beneath the conglomerate, first, the 
fine ash, porcellanized in bands, then the coarser ashes, all nearly 
vertical, and finally in the recesses of the caves the massive spheru- 
litic felsite without proof of intrusion; and all this is included in 
less than a hundred yards. When next, however, we come across 
a section in Caerbwdy valley, a red ash underlies the conglomerate, 
which is partly composed of its fragments; then is seen a massive 
. felsite; below comes a wide space containing thin-bedded porcellanites, 
next ashes, and finally the massive felsites of the cave, separated by 
nearly a quarter of a mile from the conglomerate. All the upper 
part of this series has been introduced here between the surfaces of 
junction, we can scarcely suppose by original deposition. Here then 
is unconformity or fault, or both, but certainly independence. 
The country hence to Solva valley is obscure, and what justifica- 
tion there may be for the boundary-line I cannot say. In that 
valley the mass marked as felspar porphyry is overlain both on the 
north and on the south by a conglomerate, or rather, a very coarse 
_ grit, which occupies the same place in the series as the other con- 
glomerates. These felspathic rocks I take therefore to be still part 
of the volcanic series, but certainly different in character from the 
beds already seen next the conglomerates. Further up the river, 
where the valley runs east, it coincides with-the line of junction. 
Yet on the north side we pass successively from felsites to ashes 
and agglomerates ; on the other from the purple slates through the 
Menevian to the Llandeilo. The break here, again, is at the same 
place in the series. 
Of the northern boundary I know nothing; it has always been 
marked as a fault. On the west, in the upper reaches of the Alan, 
near the farm of Trehenliw, we have, on one side, silvery ashes of 
peculiar character not seen elsewhere, on the other Cambrian slates ; 
again no regular sequence. On the coast the junction is first seen 
at Ogof Golchfa, where silky ashes associated with dark-coloured 
crystalline rocks strike at the conglomerate and are nearly parallel 
to the Cambrian slates, appearances which I interpret as due to a 
forked fault (fig. 3); but the locality will doubtless be more fully 
described by Dr. Hicks, who directed my attention to it. The con- 
glomerate here is of a different character compared with that on the 
eastern side, being well supplied with porcellanite as well as quartzite 
pebbles, and having a highly schistose matrix, possibly derived from 
underlying silvery schists which have been destroyed. South of 
this spot the conglomerates nowhere reach the coast till near 
Castell, but are cut out by repetitions of the overlying purple 
beds, as indicated by Dr. Geikie’s map. The nature of the junction 
is not directly observable along this line; but when the conglomerate 
is seen on the coast its strike and that of the overlying beds is about 
45° inland, or directly at a boss of lavas and ashes which is not 40 
* This account of what is seen there is, of course, at present assumed to be 
the true one. 
