72 
SILUKIA. 
[Chap. IV. 
elsewhere. In Carmarthenshire and Pembrokeshire the Llandeilo Flags, 
with their large Asaphi and Ogygise, are surmounted, as before stated, by 
sandy beds with impure limestone, containing the same species of fossils 
as those of the Shelly Sandstones of Caer Caradoc. 
Seeing that in Carmarthenshire there was no apparent unconformity towards 
the east and south between the Llandeilo Flags and the Upper Silurian rock 
under which they dipped, as exhibited in the Sections, pp. 55 & 56, 1 formerly in- 
ferred that the equivalent of the Caradoc Sandstone of Shropshire had locally 
thinned out to the south and south-east of Llandeilo, or was there represented 
by certain coarse sandy beds or grits only. Again, detecting scarcely any trace 
of organic remains to the west and north of the zone in which the Llandeilo 
flags were dominant, I drew a line of limit on my original map, and termed all 
the rocks ' Cambrian ' which lay beyond the known fossil-bearing zone, stating 
that the separation of the Silurian from the then supposed Cambrian of South 
Wales had been effected by assigning to the former those beds which contained, and 
to the latter those which did not contain, the characteristic fossils (Sil. Syst. 
p. 360). And yet such sections as were made towards the west and north-west 
indicated that the strata of Carmarthenshire, on the right bank of the Towy, 
folded over to the north and west, and constituted a mass of arenaceous and in- 
coherent slaty schists, which were spoken of as lithologically inseparable from 
the Lower Silurian rocks (Sil. Syst. p. 360). 
It is not my object in this volume to describe the mines which occur in the 
Silurian or other Palaeozoic systems ; but it may be noticed that the chief mines 
in Carmarthenshire, or the lead-ores at Nant-y-Moen, north of Llandovery, the 
property of Earl Cawdor, occur in these quartzose rocks, which alternate with 
slaty schist. They are now ascertained to be in part the equivalents of the 
Caradoc Sandstone of Shropshire, and in part of the Llandovery rocks to be de- 
scribed in the ensuing Chapter. In Caradoc quartzites also are situated the old 
Roman gold mines of Gogofau, in the parish of Llanpumpsant, ten miles west of 
Llandovery *. In describing these rocks of slaty quartzose grit and sandstone, I 
accurately assigned to them a dip to the north-west f ; and that fact alone 
ought to have led me to class them as Silurian, since they appeared to overlie 
the Llandeilo flags. But the slaty condition of the rocks, and the absence of 
fossils, induced me to refer them to a system unknown to me. A great fault was, 
indeed, supposed to intervene, to account for this apparent inversion. These rocks 
were further noticed as containing coarse grits, sandstones, and conglomerates — 
the latter often appearing as huge lenticular masses. 
The only organic remains detected during my partial survey of such 
strata were a few fragments of Crinoids and Shells, and the Annelide- 
markings or tracks of marine Worms found near Llampeter, such as the 
Nereites and Myrianites %. These fossils are now known to be common 
to Silurian rocks in Scotland, Germany, and other parts of the world. 
In their laborious and accurate researches, the Government Surveyors 
established clear proof that, whilst on the left bank of the Towy a great 
* See Sil- Syst. p. 367, and Mr. Warington work. Greinitz, of Dresden, who has described 
Smyth's account of these mines, Eecords of Geol. these forms, considers them to be soft-stemmed 
Survey, vol. i. p. 480. creatxires of the same family as Graptolites (Ne- 
t See the woodcut, Sil. Syst. p. 368. reograpsus &c). They may even be due to the 
[ See Sil. Syst. pp. 363, 6U9, and Chap. X. of this superficial burrowings of Crustacea. 
