L. C. Miall — Formation of Swallow-holes. 517 



by the collection of the rainfall into one channel. Its sides become 

 less uniformly vertical, the frost detaches blocks of stone, and we 

 get a shallow pit filled with loose rocks and pebbles, and often 

 grass-grown. 



Subterranean channels are exceedingly common in a Mountain-lime- 

 stone district. They are generally, and no doubt correctly, assigned 

 to the action of water passing along fissures of the rock. These 

 concealed water-courses appear to be necessary to the formation and 

 preservation of every sort of swallow-hole. They carry off the 

 grains of sand and particles of mud which would otherwise accumu- 

 late so as to check further excavation, and may help to enlarge the pits 

 by their undermining action, causing an occasional slip or subsidence. 



It remains to consider why these singular cavities occur only in 

 Mountain limestone. It will be seen that three conditions at least 

 are essential : — 1st. Abundant jointing, which facilitates the pro- 

 duction of fissures ; 2nd. Eeady escape of water, which is partly 

 efi'ected by the fissures, and partly by the power of rain-water, to 

 excavate calcareous rock ; and, 3d. Alluvium or drift to discharge 

 the collected rain in a suitable manner. The last of these conditions 

 may be present in any place, irrespective of its stratigraphical 

 character; the fissured structure occurs in many rocks other than' 

 limestones, but is rarely so conspicuous as in Mountain limestone, 

 which is also peculiarly porous, owing to the solvent action of rain- 

 water, charged with carbonic acid, upon carbonate of lime. The 

 homogeneous character of that rook, too, enables it to be cut readily 

 into definite shapes. For the same reason that a crystalline marble 

 yields to the chisel more accurately than a stratified sandstone or 

 flagstone, it obeys the abrading forces of nature more completely, and 

 exhibits upon its hard and compact surface sharp grooves which have 

 been cut in it by water or agitated pebbles. I show here a specimen 

 of limestone from the brook below the Buttertubs, in which we see 

 cut distinct and sharp the channels of the little streams which have 

 been deflected right and left by a minute projecting fragment of 

 shell or encrinite. Hard as it is to wear, the absence of definite 

 arrangement in its texture renders it ready to assume the precise 

 form which the slight inequalities of pressure and friction tend to 

 impose. Granite does not produce such swallow-holes as these, be- 

 cause it is too impervious to water ^ — its cavities (unless they are 

 merely enlarged fissures) are occupied by pools, and not by cascades. 

 Chalk is too crumbly to preserve a vertical face against falling water,^ 

 while sandstones, flagstones, and slates are fissile, and conduct the 

 rills along channels which take the direction of the planes of readiest 

 decomposition rather than of the lines of quickest descent.^ 



^ In the discussion which followed, Prof. Ansted referred to swallow-holes occurring 

 in granite in Sark and Cephalonia. Enlarged fissures and pits, which receive streams, 

 are, doubtless, to be met with in strata of all sorts, but swallow-holes of the first class 

 described in the paper, whose characteristic features are vertical sides, fiuted surfaces, 

 and isolated pillars, are believed by the author to be restricted to a few formations only. 



'^ See Prestwich's paper on swallow-holes on the Chalk hills near Canterbury. 

 Quarterly Journal Geological Society, x. p. 222. 



3 In the discussion Mr. H. "Woodward cited the case of underground rivers in the 



