Ixxiv PROCEEDINGS OF THE GEOLOGICAL SOCIETY. [vol. lxxvi, 



a preponderance of washed drift at the surface, yet the sands and 

 gravels of this origin are relatively circumscribed in extent, and are 

 mostly concentrated in mounds, ridges, or open delta-like patches. 

 If, however, the advance of the waxing ice was spasmodic and 

 temporarily rapid, with quiescent pauses ; and if its maximum 

 extension Avas reached by a final forward spurt, which may have 

 happened at different times in different places, the conditions for 

 producing the sheets of bare boulder-clay, as well as the irregular 

 features of the sands and gravels, can be readily grasped. Around 

 its periphery the ice, after its latest spurt, sank into the ' dead ' 

 condition and never again revived, for meantime the ameliorating 

 circumstances which marked the close of the Glacial Period had 

 become effective. Where the ice before its relapse had surmounted 

 the low watersheds of its principal basins, so that the bare land in 

 front of it had a downward slope, as for example in the catchment 

 areas of some of the northern tributaries of the Thames and of 

 the Severn, the copious thaw-water escaped for a time outward, 

 spreading its burden of solid waste in great masses beyond the 

 invaded area. 1 But, where the ice came finally to rest on rising 

 ground, or where at any time during its waning it was brought into 

 this position by shrinkage, the drainage during the thaw would be 

 in the main backward, on or under the ice, and would be unable to 

 transport any material outward on to the bare land. In this manner 

 the clays contained in the dissolving fringe of the ice-sheet, along 

 with those underlying it, would remain comparatively unaffected, 

 and would settle quietly down to their permanent resting-place. 

 The drainage conditions in many places would, in fact, be similar 

 to those which we saw around the detached end of the Sefstrom 

 Glacier in Spitsbergen, where a thick mass of ice-transported clay 

 was being left at the surface with hardly a trace of stream-borne 

 material upon it, and only very little at its margin. 



It is, of course, certain that in some places the presence of 

 boulder-clay at the existing surface is due to the sweeping-away of 

 its cover by post-Glacial denudation ; but I believe that all the 

 broader plateau-like spreads of our lowlands are original features, 

 produced more or less in the manner described. 



Besides the areas shown on my map, where the drifts are con- 



1 For a recent summary and discussion of the evidence around the south- 

 western fringe of the Midland area of glaciation, with good bibliography, see 

 ' Notes on the Cotteswold- Malvern Region during the Quaternary Period ' by 

 J. W. Gray, Proc. Cotteswold Nat. F. C. vol. xx, pt. 2 (1919) pp. 99-141. 



