PRE-CAMBRIAN ROCKS IN PEMBROKESHIRE, 287 
looks as if an incipient gneiss was being formed, the metamorphic 
action being incomplete, a kind of semi-metamorphism and softening 
haying taken place sufficient only to allow the particles to arrange 
themselves according to their natural affinities. Jn general, as well 
as in microscopical appearance, this rock is so like the ‘“‘ Halleflinta” 
of the Swedish geologists, that we propose for the present, in the 
absence of a satisfactory English term, to give that name to all the 
rocks of this type in our descriptions. Though the group may be 
said to be chiefly characterized by these hillefiintas, we yet find 
associated with them true quartz-felsites, evidently old lava-flows, 
and some volcanic breccias. It is a highly acid group on the 
whole; and much of the matcrial which has entered into the com- 
position of the hiilleflintas was probably derived either directly as 
fine dust from volcanos, or from the breaking up and denudation 
of acid voleanic lavas, combined, of course, with a certain propor- 
tion of ordinary marine sediment. 
The next area to be described is in a line with the former, though 
separated from it by a considerable thickness of Lingula-flags and 
other Cambrian beds, which have been dropped between by faults. 
This has a length of about five miles, with an average width through- 
out of over a mile. Along its north-western edge the Lower Cam- 
brian beds occur as in the former case, and between these and the 
true hialleflintas are altered schistose beds probably of Pebidian age. 
Its eastern edge is a line of fault, with Silurian rocks abutting against 
it. The southern line has in parts also some overlying Pre-Cambrian 
rocks (Pebidian) resting upon it, and in other places Silurian rocks 
brought against it by faults. The evidence here is conclusive as to 
the position which the group must occupy in the Pre-Cambrian for- 
mations, especially when combined with the information subse- 
quently obtained from other areas. We see at once that it is Pre- 
Cambrian from the way in which Cambrian rocks are everywhere 
brought down against and upon it. We also see that it is Pre- 
Pebidian, as these last he upon it, and fragments of the hilleflinta 
are frequently found in the Pebidian conglomerates. The evidence 
to be brought forward will also show that it is not of Dimetian age, 
but subsequent to that formation. It is clear, therefore, that it 
holds an intermediate position between the Dimetian and Pebidian 
formations, and is not of the age of either. or this reason I pro- 
posed at the last meeting of the British Association in Dublin to 
eive to the group a new name; and it was then thought that, as these 
rocks were also found by us to occupy some areas in Caernarvon- 
shire, the name “ Arvonian,” derived from the ancient name for 
Caernarvon, would be an appropriate one. The general character 
shown by the rocks in the area now described is similar to that 
found at Roch Castle. They are for the most part true hilleflintas, 
and excellent sections may be seen in the gorge through which the 
western Cleddau river flows, and the road from Haverfordwest to 
Fishguard passes. The “'Treffgarn Rocks,” very prominent wea- 
thered masses of these rocks, are seen for a considerable distance 
along this road, and yertical cliffs of considerable height occur close 
