278 Froceedings of the Royal Pliysical Society. 



Towards the rio'ht the river-course is fixed in rock, and the 

 lower loop may be regarded as relatively persistent. On the 

 left the river flows upon softer glacial deposits with some cover- 

 ing of alluvium, and the upper part of the loop (with the slight 

 bend above it) may be regarded as relatively shifting. The 

 shifting curve is endeavouring to worm itself past the per- 

 sistent one, along the line of a buried pre-glacial channel 

 occupied by till. The limb of the stream above the gorge 

 has probably worked within its trapezoidal strip of alluvium 

 somewhat like the limb of a piston. At present this move- 

 ment is taking the form of a slow progress to right on the 

 part of a gentle curve above the double loop, which, exhaust- 

 ing itself at the gorge, may temporarily straighten the upper 

 half of it, until such time as the limb of the piston may come 

 down again along the bank upon the left. We have seen 

 such movements as this latter, very beautifully recorded in 

 the age of the vegetation clothing the bank left by the advan- 

 cing bend. Bushes and saplings grew nearest the fresh scar 

 beside the deflection-pool ; these gradually rose into trees of 

 a larger growth further up valley, and massed into a bank of 

 hanging wood.* 



Planation. 



These methods of river- action result in an intermittent 

 planing-out of the strata along the line of the valley. Eacli 

 river-curve, forced horizontally like a curved knife with the 

 blade laid flat, may leave as it travels, not only a plane over 



it is simply a diverting of the surface layer of water, but this latter may 

 channel a course which shall divert the under current also and supersede the 

 roundabout course. Some apparent shiftings, of course, are due only to the 

 occupation by the water of secondary cliaiinels, which will be afterwards re- 

 ferred to. 



* When, by these piston-like movements, the shifting bends shall have suc- 

 ceeded in cutting out the persistent one by breaching the lower neck of the 

 double loop, a rocky knoll will of course lie in the valley, as in the case of the 

 famous hills of Dunipace. An interesting case of loop-digging and its results 

 is supplied by the river Irthing. This river forms, for 14 miles, a winding 

 north-and-south boundary between Northumberland and Cumberland. It 

 then bends west at Gilslan<l. At this point, nearly on the watershed of 

 England, it had almost its choice of two valleys. It seems first to have chosen 

 the eastern one, and to have turned sharply off towards the Soutli Tyne and 

 North Sea. The arrangement of its terraces now seems to indicate that it 



