T5Y Hran millke. 



343 



It is not within the scope of this paper to consider swallow- 

 holes in their relation to underground drainage. Having de- 

 scribed their formation I pass from them, with the general 

 remark, that in multiplying local watersheds, and thus increasing 

 the unevenness of the furrows between escarpments, they may 

 have had a considerable influence on the minor configuration of 

 the ground. 



4. Aerial Pot-holes. — There is scarcely an escarpment in an ex- 

 posed situation that does not exhibit some of its surfaces drilled 

 into little pot-shaped cavities. If the water collected in one of 

 these be watched on a gusty day, it is seen to be in great commo- 

 tion, restlessly beating the sides and swirling unsteadily round. 

 Let a heavy shower come driving before the wind and its slant- 

 ing drops — beating on one side of the cup only — lash the water 

 into still more excited motion, and the attentive eye will discern 

 numbers of granules scouring round and round against the smooth 

 sides. The wind and the rain are making use of these to drill 

 small potholes, like those larger ones that the eddies of a stream 

 hollow among its rocks. 



On points of rock well in the play of the wind pot-holes are 

 usually small in proportion to their depth and set close together. 

 In such positions surfaces may be literally honeycombed with 

 pots of various shapes and sizes, some bowl-shaped, some under- 

 cut into wide mouthed flasks, some so merged and compound as 

 to resemble nothing that I know, so much as casts of the ' ^finger- 

 and-toe^'' in turnips. On a sandstone boulder, perched upon the 

 Queen's Crag, this close-set kind of pot-hole may be seen to per- 

 fection, one of them being large beyond the average, one foot 

 three inches long, by eight inches broad, and seven inches deep. 

 Around this boulder the wind can play freely, and the holes fret 

 its wliolc upper surface ; but on the angular projections of a crag 

 they arc apt to group in a sort of procession (Woodcut No. 10). 

 In front, where the wind first breaks and eddies, is the chief, 

 while behind, ranged in diininisluiig order, come the followers. 

 Wind has littk) or no influence in making the j)ot-hok^s shift 

 before it. Wlien liollowed upon a plane surface, unless biassed 



