THE GLACIAL GEOLOGY OF NORTHUMBERLAND 103 



According to the views advocated especially by Kendall, 

 these old watercourses or dry valleys are the overflow-channels 

 of ice-dammed lakes which had a temporary existence during 

 the last stage of the glacial period. In a former paper [18] 

 the old Northumbrian word "swire" was proposed as a suit- 

 able term to connote anomolous slacks or valleys having the 

 characteristics briefly sketched above. As the word is short 

 and has other recommendations, it will be frequently employed 

 in the sequel. 



If Kendall's hypothesis of the origin of the swires be 

 accepted, then it is evident that their position and grouping, 

 taken in conjunction with the evidence derived from striations 

 and rock-transport, throw important light on the extension of 

 the ice-sheets during the last stages of their existence and the 

 progress of their retreat. In addition, some peculiarities of 

 the swires themselves find explanation. Thus, the frequency 

 with which large erratics occur near the heads of the swires 

 suggests that they have floated in from the lakes on ice-rafts, 

 which have become stranded in the shallowing waters of the 

 stream. Again, the frequent occurrence of a multiple intake 

 may be explained by slight movements, either forward or 

 backward, of the ice-dam; and a peculiarity often noted in 

 the trenches which run along hill-sides, viz., the absence at 

 fairly regular intervals of a bank on the down-hill side of 

 the trench, is explained by assuming that the ice itself formed 

 the bank at these places. 



The interpretation of the evidence furnished by the swires 

 will now be attempted. For the correct understanding of the 

 matter it may be well to mention that the evidence so far 

 adduced all points to the existence of two great ice-sheets, the 

 western and the northern, during the latest stages of the glacial 

 period. Small, but still important factors, are the two local 

 ice-flows from the Cheviots and Carter and Peel Fells. The 

 positions occupied by these have been briefly indicated. It 

 may be assumed that on melting, land would first be un- 

 covered where the ice was thinnest, that is as far as possible 

 from the supply of ice. When two or more sheets were con- 



