IIO DR. J. A. SMYTHE ON 



the Tweed Valley, and (3) either a sheet or pack-ice which 

 pressed from the North Sea area. It is evident that the last 

 touched land to a very limited extent, and the name 

 "Northern Sheet" has been given in the foregoing account 

 rather to those portions of the Tweed and local ice which 

 were impelled southwards by its agency. The problem to be 

 solved is the determination of the relative positions of these 

 sheets of ice at different stages. The evidence adduced points 

 to the importance of local ice at an early stage, the Cheviot 

 ice then spreading out in a broad sheet to the south and east, 

 its western margin being limited by the Carter and western 

 sheets, its eastern mingling near the coast with ice from a 

 northerly source. The pressure of the western sheet then 

 increased until it almost completely dominated the country, 

 overriding all but the highest hills. Towards the end of the 

 period pressure from the North Sea re-asserted itself, with the 

 result that the ice within 10 to 15 miles of the coast changed 

 direction from east to south, and the local ice from the 

 Cheviots, hemmed on all sides, was forced in a narrow stream 

 down the Aln valley and southwards along the coast. 



At the melting stage the lines of confluence of ice from 

 different sources seem, in many cases, to have been positions 

 of weakness (or thinness of ice). The sheets were sundered 

 along these lines and temporary lakes were formed when 

 favourable conditions of contour and ice-edge existed. These 

 conditions were frequently maintained during the retreat of 

 the ice, long after cleavage had taken place, so that various 

 stages in the retreat can be determined. 



It will be noted that the evidence of the direction of move- 

 ment of the various ice-sheets afforded by the striations, along 

 with that of direction and extension furnished by the character 

 of the drifted rocks, especially those in the gravels, agrees 

 with the evidence on these subjects derived from the position 

 of the dry valleys. The last line of evidence has the advantage 

 of enabling the positions of the ice-edges at melting to be 

 determined with some approach to precision. One result of 

 this is to show that the local ice of Cheviot origin had con- 



