146 THE VALLEYS OF SOME OF THE IRISH LAKES. 



ment in favour of their having been originally started 

 along joints or' fissures." 



A pot-hole formed in a sheet of limestone will have 

 a deposit of clay at the bottom of it, the insoluble 

 residue left after the rest of the rock has been carried 

 away in solution through the cracks and joints of 

 the rocks, while if the rock contains nodules of chert 

 their debris will be mixed with the clay. It is the 

 same with pot-holes under drift, except that not only 

 the bottoms, but also the sides, will have a clay 

 lining. This is as might naturally be expected. As the 

 holes became deeper and wider, the drift coming in 

 from above would force out the clay and form a coat- 

 ing round the walls ; or if the holes were filled by a 

 more recent accumulation of drift, during the sub- 

 sequent weathering the water percolating down over 

 the bed of limestone would carry the residue along 

 with it, and thus form the clay lining. It has been 

 suggested that the clay in the pot-holes in the chalk 

 came in from above. This may possibly be the case 

 in some instances, but in most it is impossible, as 

 the clay at the bottom of the holes and lining the 

 sides usually far exceeds in quantity the clay that 

 would be necessary to form a horizontal sheet over 

 the top of the hole, if of an equal thickness to the 

 layer of clay that all round the pot-hole intervenes 

 between the drift and the subjacent rock. In some 

 of the subterranean caves brought to light during 



