Cardiganshire, Caermarthenshire, part of Pembrokeshire, and in parts of other adjacent counties is, I consider, of older date than the Carboniferous strata ; and I also consider that great part of that ancient Silurian tableland was overspread by these strata. On their removal by denudation, that tableland was attacked by all the ordinary powers of waste ; and now, a modified descendant of a pre-Carboniferous plateau, it is traversed by unnumbered valleys and streams that run towards every point of the compass, mostly across the strike of the strata, but some, like the Towey of Caermarthenshire, and the Teivi of Cardiganshire, along the lines of strike for long parts of their courses, where comparatively soft slates and Llandeilo flags form the surface, bordered by hills formed of harder rocks. Some of the rivers, like the Wye and the Usk of South Wales, run right through bold escarpments, in the manner of the Thames ; and similar cases in North Wales are not wanting. But this part of a very large subject I hope to return to in a future paper, should I find time to prepare it.
Discussion.
Mr. Hughes pointed out that in Wales and the Lake-district, which in this question might be considered as one, there were two plains of marine denudation — the one referred to by Prof. Ramsay averaging a little over 2000 ft., and the other about 3000 ft. above the sea. Such plains get eaten back and cut up into valleys; but their general level does not get much lowered by subaerial denudation. Therefore, in considering the western drainage-area of the ancient Severn, it was important to fix the age of these plains. He did not agree with Prof. Ramsay that the 2000-ft. plain was pre-Carboniferous, as the Carboniferous and Old Red hills of South Wales and, in a more marked way, those of West Yorkshire and the Lake-district were evidently cut down by the same denudation that planed off the top of the Silurian area, and their tops formed part of the same plain. He did not think that this plain could be even pre-Oolitic ; for the shingle-beach of the Trias, which might be regarded as the basement- bed of the Jurassic series, was evidently formed round the margin of that old land, whereas had this plain existed there would not have been land sufficiently high to arrest the Jurassic sea during the period of greatest submergence; and a conglomerate implies a shore near. The absence of a coarse shore-deposit at their base, and the character of the Cretaceous deposits, also would lead him to infer that the Chalk-sea probably washed no land so near as Wales ; but it was quite possible that the chalk was removed from the Welsh area when the 2000-ft. plain was formed ; and in that case we should refer the initial Severn to the time when the deposits of the sea that formed that plain were being eaten back, and not to the time when the Chalk was being removed. He asked where were the valleys when the drainage of the eastern area ran west into the Severn, as there was considerable difficulty in supposing that the main westerly drainage was in reverse direction along the
