BY HUGH MILLEK. 339 



therefore, may safely be classed Crag-talus, Siv allow -holes in lime- 

 stone, and what I may call Aerial PoUlioles drilled in the rock 

 surfaces. To these there must be added a fourth, less evidently 

 connected with escarpments, viz., some of the Denes or gorges 

 terminated by waterfalls. The last named may be taken first. 



1 . Waterfalls and Denes. — After the close of the glacial period 

 many streamlets found themselves ousted from their old channels 

 by glacial deposits. So choked and concealed did some conside- 

 rable vales become, that their sites are now discoverable only by 

 mining operations. The main valleys, although themselves much 

 blocked up with till, remained sufficiently open, as a rule, to 

 gather in theii^ drainage as before, but the minor conduits were 

 largely changed, and in picking their courses afresh down the 

 valley slopes, streams often fell over terraces or ledges of out- 

 crop lying in their way. An operation has proceeded ever since, 

 singularly like — in the typical cases — the notching of a plank by 

 a saw. The rock-bed is the plank, the terrace its convenient 

 edge, and every one who is familiar with the typical Linn 

 within its cul de sac will recognise its resemblance to the slit in 

 the wood lengthened inwards by the vertical sawing. Such of 

 them as I now refer to are recognised by having had, as it were, 

 the planh laid ready for the saw. The Holywell Burn near Birt- 

 ley, for instance, found two terraces laid across its line of flow ; 

 the Dinley Burn, before referred to, fell over another ; and here 

 and there among the dales, whenever a distinct preglacial terrace 

 is crossed by a post-glacial stream, the indenting process has 

 come into play.^' In numbers of instances, indeed, waterfalls 

 have not been formed thus at a stride. Given, a bank with a 

 variety of beds cropping out on its slope, and the harder ones 



* This process, even in our climate, would be rapid in its early stages. At Cawburn 

 llig, between Haltwhistle and Greenlea Lougli, a drain was led over the face of an es- 

 carpment-bank within the memory of a man who died in 1850 at the a{je of ninety-six. 

 It has now cut for itself a notch fifty-two feet broad and twenty-two feet deep, contain- 

 ing a cascade over sandstone steps, and a waterfall some six feet high. For its history I 

 am indebted to Mr. Armstrong of Linucres, near Warlt, whose grandfather it was that 

 made the drain. 



I omitted to say in the proper place that the shale in the sides of Dinley Gorge, except 

 near the waterfall, is prevented from iiollowing out by wcatl\ering It is not preserved 

 l)y vegetation, like tli.it under escarpments, but by its own shivery debris. 



