COERASION BY GRAVITY STREAMS. 315 



velocity is not basal but at some varying height above the 

 base and along the main channel. Nevertheless such 

 maximum velocity of a stream does not concern us when 

 the question of the location of maximum stream corrasion 

 only is under consideration. For such stream portion does 

 not come into contact with the channel base or sides. The 

 question which does concern us however when considering 

 the relative corrasive power of various portions of a stream 

 is — "where does the maximum velocity of the stream 

 portion which happens to be in actual contact with the 

 channel structures occur"? The answer to this is — the 

 basal portion of the channel. (See Fig. 2, Part I and 

 descriptive text). The sides and base of the young valley 

 incised into the upland surface will therefore suffer heavy 

 corrasion while the associated upland will experience a 

 relatively weak action. 



Upon the shrinkage of the ice cap the valley glaciers 

 will first readjust their channel grades and at a later stage 

 will carry on the work of excavation. Or again the flood 

 stream will seize upon any occasion for local increase of 

 velocity in the matured surface it may be traversing, and 

 then under favourable conditions it will excavate at this 

 spot a basin which will recede with a "step" or cirque 

 head thus forming a youthful valley. Valleys thus may 

 be formed by an ice cap in an upland and channels 

 may be cut back on declivities. This is the analogy with 

 the action of an ordinary stream which has cut a channel 

 into the bye-wash of a dam. 



The Blair Athol occurrences admit of explanation in this 

 manner. An ice cap swarmed over the upland valley level 

 and gave rise to some corrasion there, but along the younger 

 stream valleys it acted much more fiercely. At certain 

 points it excavated "steps" and "treads" along definite 

 lines which under the continuous ice scour receded to form 



