ROCKS OF PEMBROKESHIRE. 523 
about midway between Nine Wells and the coast, so as to dip 
under the lowest Menevian beds. 
d. Pre-Cambrian Rocks in this Area.—Immediately under the 
Cambrian conglomerates in the Caerbwdy Valley are some reddish 
felsitic tufis and breccias belonging to the Pebidian series, marked 6 
on the map. ‘The conglomerates clearly overlie the latter uncon- 
formably, as may be seen from the way they creep over different 
beds in the Pebidian series. ‘The junctions are not very well seen 
here, but the evidence of unconformity becomes perfectly clear by 
following the line of strike of the Cambrian conglomerates. Below 
the series 6 are found the highly siliceous thin-bedded rocks which 
we have here called porcellanites. The microscopical characters 
of these rocks are not very satisfactory, but their sedimentary 
origin becomes thereby clear (Note 19). Associated with them 
are some fine breccias, consisting of fragments of quartz-felsites, of 
broken bits of quartz and felspar, and of basic lavas, all cemented 
together by a fine-grained material, probably derived mainly from 
decomposed felspar. Below the last-mentioned series the great 
agelomerates and breccias of Clegyr Hill are found. These were 
described in a former paper*, and their microscopical characters 
are given fully in the Notes to that. paper by Mr. Davies (p. 166). 
The presence in these of numerous fragments of spherulitic quartz- 
felsites, like those classed as Arvonian, and found in sitw in the 
ridge under St. David’s and elsewhere, is a fact of importance ; for 
many of these fragments show indications of having been rolled by 
water-action ; and the finding of basic fragments and of indurated 
argillites occasionally in the same beds proves that they were de- 
rived, partially at least, by denudation from pre-existing series. As 
very many of the fragments, however, are angular, it is probable 
that the main portions may have been cast out by volcanoes, and 
but partially rearranged afterwards by water-action. These volcanoes 
seem to have broken through the quartz-felsites of the Arvonian, 
and thereby to have caused the admixture of the materials found. 
When my former paper, in which these agglomerates were described, 
was published, the Arvonian rocks were included in the Dimetian, the 
other name not having been at that time suggested. It is necessary 
to state this, as though it is clear that minute fragments of the true 
Dimetian do occur in these agglomerates, the larger masses men- 
tioned were such as I should now class with the Arvonian. These 
agglomerates are exceedingly wel! exhibited in the hills on both 
sides of the valley near Clegyr Bridge, and in some rocky promi- 
*nences in the fields towards Clegyr Farm. Beyond these points, 
towards the N.E., they are hidden for the most part in moors, &e. 
They are diminished also in thickness by faults which drop them 
along with some of the Cambrian beds. In the valley to the north of 
Clegyr Bridge, at Trefynard and Trepewet, some few exposures of 
rocks occur, which show that at these places there are intrusive 
porphyritic quartz-felsites, in some quartz-felsite breccias. These 
* Quart. Journ, Geol. Soc. vol. xxxiv. p. 157. 
