Vol. 50.] FEAGMENTAL KOCKS IN ST.W. CAEENAEVONSHIEE. 593 



that in the so-called line of unconformity, that above it, and that in 

 the possible banding below it. 



The alleged unconformity, therefore, seems not only unsupported, 

 but also actually contradicted by direct evidence. If further proof 

 were needed, it would be found in the continuation of the section 

 northward and southward. The dips in both directions and the 

 lithological characters give some evidence. On the north Mr. Blake 

 marks ' greenstone or green grit not properly examined ' (op. cit. 

 p. 445). We find here a greenstone which is partly compact, 

 partly porphyritic, and grits (2 & 7 in our fig. 2 ; see also map, 

 p. 594). The grits are distinctly banded, and continue the succession 

 above the conglomerate, with purple argillite-laminae of the typical 

 interbanded character. In addition, some bands show distinctly the 

 ' rain-spot ' breccia-type, proving the recurrence of this same variety 

 above the supposed unconformity. 1 



We have discussed this section at some length, because it is the 

 place of which Mr. Blake states that he ' cannot imagine a more 

 satisfactory proof of unconformity in a single section' (op. cit. 

 p. 446). If unconformity there be, we have never before seen one 

 so successful in concealing itself. 



VI. Stbatigbaphical Succession. 



Here, perhaps, a short summary may be given of the explanation 

 which we venture to suggest for some of the sections described 

 above. 



On the west of the lake the conglomerate next the felsite (and 

 largely derived from it) extends above the road above Tan-y-pant, 

 followed by a grit of similar material. Still higher up on the hill, 

 interbanded grit and argillite occur, with pebbly layers and a kind 

 of ' rain-spot ' breccia. The oblique direction of the outcrop of 

 these beds, and their general dip to a north-westerly point, i. e. 

 towards the felsite, indicate the probability of a line of fault along 

 the boundary of the igneous rock (F 4 in map, p. 594), as was 

 formerly suggested by one of us. 2 



The conglomerate, however, and other beds are abruptly termi- 

 nated near the little streamlet already mentioned, beyond which 

 a sudden change of dip is found, so that this also seems to be a 

 parallel line of fault (F 5 in map). 3 By it the reappearance of the 

 felsite towards the lake may be explained, as well as the crushed 

 and schistose condition of some of the rock in the small ' stream 

 quarry.' But this fault makes it more difficult to come to a 



1 Bands more or less similar occur at other places — in the syncline by this 

 railway, on the railway west of Llyn Padarn, and on the hill above Tan-y-pant. 



2 T. G. Bonney, Quart. Journ. Geol. Soc. vol. xxxv. (1879) p. 314. This fault 

 might account for the true felsitic conglomerate reappearing in small exposures 

 near Groeslon, and for the crushed condition of other beds near. 



3 A fault roughly near this seems to be marked on the Geological Survey 

 map. Possibly another parallel fault occurs, bringing up the felsite N.W. of 

 this, already mentioned. See note, p. 582. 



