506 PROCEEDINGS OV COLUMBUS MEETING. 



to obtain the underlying sand, but nowhere did we see the stratum upon which the 

 sand rested, so thai we were unable to speak from observation of its nature; but 

 from the distribution of Welsh bowlders in the vicinity o'f Birmingham already 

 mentioned it is absolutely certain that Welsh ice had moved over this area previous 



to the invasion of glaciers which started from southern Scotland. They are there- 

 fore without doubt what would properly be called interglacial beds. Their eleva- 

 tion above the sea. as given me by Dr. ( !rosskey and Mr. F. W. Martin, who accom- 

 panied us on the trip and conducted us to the locality, is in round numbers 500 

 feet. The Wrekin, two or three miles away, is a solitary peak in the Severn valley, 

 rising 1,335 feet above the sea. In other localities of the vicinity Dr. Crosskey had 

 found shells having a more arctic character than these in glacial deposits of similar 

 character about 700 feet above the sea. 



In endeavoring to account for these shell-beds it will be necessary to take a still 

 more general view of the situation, and it is the more important to do this since 

 the explanation of this deposit is doubtless closely connected with that of similar 

 deposits found at still higher levels, namely, at an elevation of about 1,100 feet at 

 Macclesfield, a feu miles south of Manchester, and at 1,400 feet near Moel Tryfaen, 

 on the northwestern flank of Snowdon in Wales. On glancing at an orographic 

 map of England, it appears that between the northern part of the Welsh high- 

 lands and the southern projection of the Pennine chain in England there inter- 

 venes a valley, about 70 miles wide, known as the vale of Chester, running nearly 

 north ami smith, which is nowhere more than 500 feet above the sea. The shell- 

 beds under consideration occur near the head of the Severn valley at just about 

 the same height as the water-parting between the valleys of the Severn and the 

 Dee. A careful collection of facts made by Professor Percy F. Kendall concerning 

 the distribution of bake district and Scottish bowlders makes it clear that the vale 

 of Chester was occupied by the eastern branch of a confluent glacier which filled 

 the Irish sea, receiving vast contributions of ice from the two remaining centers 

 of glacial dispersion referred to above, namely, (3) the southwestern portion of 

 Scotland and northern England, and (4) Ireland. 



bowlders from the Lake district in England moved westward into Morecambe 

 bay, where they were met by the movement from Scotland; while in the mean- 

 time glaciers from Ireland pushed eastward into the Irish sea until the whole basin 

 north of Wales was at length tilled with ice under sufficient head to abut against 

 the Welsh mountains and to push upward upon their northern flanks to a height 

 of more than 1,400 feet. But the main mass of ice was divided by the obstruction 

 and flowed in two streams, the one over Anglesea on the west into Saint George 

 channel to an indefinite distance, the other on the east through the vale of ( Jhester 

 almost to Birmingham, occupying the area already described as covered by bowlders 

 from northern EnglancLand Scotland. It was this movement which deposited the 

 till at Ketley and which, I believe, brought along from the bottom of the Irish sea 

 the shells which were there found by Mr. Baldwin. This is the theory advocated 

 by the late Professor Henry Carvill Lewis to account for the shell deposits found 

 at Macclesfield and Moel Tryfaen. 



The considerations supporting this view are numerous : First, such shell-beds in 

 glacial deposits are strictly confined to areas known to have been occupied by glacial 

 ice which had previously moved over shallow sea-bottoms. At Ketley the bowlders 

 in the upper till all came from southwestern Scotland or the Lake district in Eng- 

 land by way of the Irish sea. Similar shell-beds found in the glacial deposits of 



