J". G. GoodchUd — On Glacial Erosion. 359 



of an ellipse, and then gradually become longer, nntil they extended 

 from the point where they were initiated right across the terrace to 

 the very inner margin. If the rate of recession of the scar at the 

 edge of the terrace were equal to that of the overlying bank of shale, 

 the first formed swallow-holes would soon be cut back too, and we 

 should then find a series of ravines extending across the whole 

 width of the terrace. 



It need hardly be repeated that nothing of the kind is to be found 

 in the part of the Dale District here referred to : perhaps it would be 

 difficult to find a single instance of such a series of ravine-like 

 swallow-holes in the entire Dale District. 



Now it is clear, as was pointed out above, that the existence of 

 these terraces has been determined, not by the capacity of the rocks 

 that form them to resist subaerial erosion, but by their relative 

 powers of resistance to erosion by mechanical means. 



Bearing these facts in mind, and reflecting upon the regular forms 

 of the scars ; their parallelism with others above and below ; their 

 frequent correspondence in form with others on the same horizon 

 across valleys a mile or more in width ; the existence of perfect scars 

 and terraces many hundreds of feet above where, under existing 

 physical conditions, it would be possible for any stream to produce 

 them ; the uniformly weathered surface of the limestone and the 

 restriction of the principal swallow-holes to the inner margin of 

 each terrace ; the general absence of much debris from the higher 

 beds ; and, finally, the existence of glacial markings close to the 

 inner margins of some of the widest terraces — we may well hesitate 

 to accept any theory whereby the origin of these characteristic rock 

 features is referred to Subaerial Erosion. 



At least as early as the summer of 1868 Prof. Hughes, when we 

 were together surveying part of the district referred to, expressed 

 a doubt whether any one of the Subaerial Denudation theories was 

 adequate to account for all the facts connected with the rock features 

 of that district ; at the same time he stated his belief that no small 

 share in the development of the surface characteristics of the Dale 

 Eocks must be attributed to the action of land-ice. Since then an 

 increasing acquaintance with the physiography of a large area of the 

 Upper Palaeozoic rocks adjoining the Dale District has increased my 

 conviction that the theory put forward by Prof. Hughes is the only 

 one that really sorts with the facts. 



In the communication referred to at the head of this paper I have 

 stated my reasons for believing the former existence in the Dale 

 District of an ice-sheet of such a thickness that it overtopped the 

 highest fells ; therefore its surface could not have been much less 

 than 2300 feet above the level of the sea. Nearly everywhere at 

 high levels the ice seems to have flowed away from certain pretty 

 well defined lines and centres without being much influenced by the 

 form of the underlying low ground ; while in proportion to its 

 nearness to the bottoms of the valleys it seems in general to have 

 been more and more guided by the configuration of the adjoin- 

 ing surface, in the manner so often spoken of by Prof. Kamsay. 



