94 SECTIONAL ADDRESSES 



(b) The Welsh Re-advance ' or Little Welsh Glaciation. 



The fourth and last glacier to reach our area was an extension of the 

 Upper Severn valley-glacier down as far as Shrewsbury to which White- 

 head has given the name Welsh Re-advance. There is strong evidence 

 that the lowest of the important Severn Terraces, the Worcester Terrace, 

 was being formed during this re-advance. It is the highest terrace to be 

 recognised on both sides of the Iron Bridge gorge, and its level there shows 

 that the gorge was then some 60 feet shallower than now. 



There is no conclusive evidence in the Midlands that the cold conditions 

 of the Welsh Re-advance were anything but a climatic oscillation in the 

 general amelioration that caused the gradual melting of the Irish Sea 

 glacier. Whitehead in his account of the Shrewsbury district writes as if 

 he considered that the Irish Sea ice-sheet possibly still existed not far to 

 the north, when the Welsh Re-advance glacier was at its maximum. 

 There is, however, some slight evidence for interglacial conditions in the 

 Midlands during the time between the deposition of the Main and Wor- 

 cester Terraces, which, taken in conjunction with evidence in other areas 

 (particularly the sealing by Irish Sea boulder clay of Upper Palaeolithic 

 caves in North Wales, and the upper boulder clay on the top of the Wrex- 

 ham delta-terrace, now being investigated by Miss D. S. Coates), suggests 

 the possibility that the Irish Sea ice sheet first withdrew completely and 

 then re-invaded the northern part of its old domain simultaneously with 

 the re-advance of the Upper Severn glacier. 



This problem is one among many relating to our glaciations that await 

 solution, and yet can never be solved by work in one restricted area. The 

 cry is always for accurate data in neighbouring areas. I close this address, 

 as I began it, by an appeal for amateurs who are willing to undertake con- 

 scientiously and scientifically the recording and co-ordinating of every 

 scrap of evidence in the district in which they live, whether it be a glacial 

 or a periglacial one. If this were done so carefully that no temporary 

 exposure escaped record, data would gradually, but I think quickly, accu- 

 mulate by which some at least of the many outstanding problems of glacial 

 correlation and interpretation would reach solution. 



' The map. Fig. 2, does not attempt to show the limits of this along the Welsh 

 borderland, as worked out by Dwerryhouse and Miller and by Charlesv/orth, since 

 they lie wholly outside the Midlands. 



