W. Gimn — Suhaerial Denudation v. Glacial Erosion. 103 



scribed, where several feet of sliale are interposed between two hard 

 beds of limestone or sandstone, forming two waterfalls pretty close 

 together, we often find that some distance below the lowest waterfall 

 there are two separate scars with a little terrace between them. 

 How has this been formed ? Clearly not by the stream alone, for 

 it most likely did not run at all where we now see the little terrace 

 on either side ; but the upper hard bed, having been exposed longest 

 to subaerial action, is worn back further than the lower one, while 

 the soft shale between, being worn away faster than either, has left a 

 terrace. Again, where a hard bed dipping at a high angle crosses 

 the course of a stream, we see that atmospheric denudation has so 

 acted on the banks of the beck as to leave the hard bed in relief as 

 a ridge, the upper slope of which, from the weathering away of the 

 soft bed above, tends to coincide with the dip, and thus is formed a 

 terrace sloping at a considerable angle from the horizontal. These 

 examples show how scars and terraces may be initiated by subaerial 

 denudation. Thus the fact that " where the rocks are much disturbed, 

 the characteristic terraces usually keep to the same bed through all 

 its variations of position and inclination " (pp. 324, 328), is in my 

 view in favour of their being in the main the work of rain and frost, 

 and against their being formed by erosion of ice, which could not act 

 equally on the same bed in all positions. 



I have, I think, before shown that limestone, as a rule, is not 

 more easily weathered than the sandstones and shales with which 

 it is associated; and as it is by far the most important rock (p. 324), 

 and forms the thickest and most massive beds (and according to 

 my view generally the hardest), we need not be surprised that the 

 scars formed by it are usually more perfect than those of the sand- 

 stones. Of course there are exceptions to this rule. The sandstones 

 are generally laminated, and do not form thick beds ; but where it 

 happens that a thin limestone is superimposed on a thick hard bed 

 of sandstone, as near Preston in Wensleydale, the limestone and 

 sandstone in places form separate features, the sandstone being most 

 prominent. It seems strange that in these articles so much is said 

 about springs and streams, which can only act along lines, and so little 

 about rain and frost, whose action is universal. (See Greenwood's 

 "Eain and Eivers," and Lyell's "Principles.") 



It is unnecessary for me to show how scars once formed will 

 recede as the result of these agencies. Their effects have been well 

 shown by many geologists, of wliich a pretty full list is given in Mr. 

 Whitaker's admirable paper (Geol. Mag. Yol. IV. pp. 448 — 450). 



That there was extensive denudation during the Glacial Period all 

 must admit ; and that the ice-sheet would sweep away loose surface 

 debris, and here and there a talus temporarily accumulated at the 

 base of a scar,^ is not hard to see ; and it may have partially stripped 

 of shale some of the terraces. But if it had worn back or formed the 



^ It is possible that a scar of hmestone or sandstone, however first originated, may 

 in certain situations accumulate debris faster than subaerial agencies can remove it, 

 and that old pre-glacial scars were imcovered by glacial erosion. On the other hand, 

 as one of the results of the Glacial Period, we have, as seen in railway cuttings, old 

 scars covered up by glacial drift, and being gradually uncovered by rain and rivers. 



