ROCKS OF PEMBROKESHIRE. 533 
materials at the junction, and no evidence of anything like a welding 
process having been setup. Asa result of the faulting also, different 
beds of the Cambrian series are brought in contact with the Dimetian. 
On the west side the beds belong to a mnch higher horizon in the 
succession than those on the east side, from having heen thrown 
back by a fault striking about N.N.W. ‘These actually strike up 
to the west edge of the so-called tongue. On the east side the 
beds are made to dip towards and as if under that edge, a very 
usual condition in faulted junctions, as I shall be able to show in 
other sketches, and one well known to all stratigraphical geologists. 
There is another point to which special attention must be called 
in regard to this sketch, as it would appear from a sketch pub- 
lished in Prof. Frazer’s paper already referred to (p. 14) that he, 
at least, when examining this section with Dr. Geikie, supposed 
that the Dimetian here cut into and enveloped a mass of the Cam- 
brian rocks. The supposed evidence for this is clearly shown by 
his sketch to have been based on erroneously taking the Cam- 
brian grits immediately to the north of the hut to be a part of the 
Dimetian. Better: proof, perhaps, of the fact so often insisted 
upon by us that the matrix of the conglomerate has been derived 
from the Dimetian could scarcely be wished for than that so 
experienced an observer should make such a mistake under the 
guidance of the Director-General and of Mr. Peach. In his revised 
paper the sketch remains as before, but it is cnly fair to him to 
give the following quotation from that paper, issued after the 
abstracts of Prof. Geikie’s paper and the reports of the discus- 
sions had been received by him :—‘ It would appear that neither 
Prof. Geikie nor Mr. Peach regarded the contact between the 
granite (containing some amphibole) and the sandstone (conglo- 
merate ?) at that point in the river Allan figured in the sketch, as 
indicating an envelopment of the latter by the former, since no 
allusion is made to it. This relation seemed to the writer so clear 
that he sketched it, and had a woodcut made of it, in spite of the 
fact that it gave a rude shock to his gradually increasing conviction 
of the practical parallelism of the rocks of the South Mountain and 
those of South Wales, by seeming to prove that the granite was of 
later origin than the clastic rocks which it environed. The writer 
has probably made a mistake here, though of what nature he is yet 
ignorant, and it is only fair to say that while regretting that the 
error is his, he is gratified to know that this abnormal position of 
the respective rocks does not exist.” The grit near the hut is 
described in Note 62, and the sandstone on the opposite side of the 
Dimetian in Note 63. Both show clear evidence of the presence of 
the dirty quartz, derived from the Dimetian, and of fragments of fel- 
spars, and there are no indications whatever of any alteration in the 
sediments such as would have been produced by the proximity of a 
great intrusive mass. In the sketch (fig. 6) it is necessary to state, 
that the beds marked 2, to the south of the hut, are really newer than 
the conglomerates and grits marked 1, though appearing in it to dip 
