PAPERS ON GEOGRAPHY AND GEOLOGY 197 



found to have the following positions; strike X. 180° 

 E., dip 40° W; strike N. 15° E., dip 39° S; and strike 

 N. 85° E., dip 87° S. A similar process is responsible 

 for some of the so-called rock " tanks" in arid regions 4 . 



Again it happens that the Killarney granite has a 

 natural conchoidal fracture in some places, and (Figure 

 5) under any stress, such as that, for instance, of a pass- 

 ing glacier heavily loaded with locally burdensome drift, 

 it is likely to break along curved surfaces. A piece which 

 breaks out of a plane face naturally leaves a concave 

 surface, and after such a surface has been glaciated, 

 that too is like a pot-hole. 



There are porphyritic phases within the granite and 

 the gneiss which are not of the same fine, even grain 

 characteristic of the rest of the rock, and the parts 

 which are weaker than others weather more readily and 

 leave hollows in the surface, which are likely to have 

 been smoothed off by giaciation. (Figure 6.) 



In the side of a nearly vertical surface facing north- 

 ward, on the north side of an island south of Philip Ed- 

 ward Island, there is a cavity, over a foot deep, which 

 penetrates a glaciated surface, and which seems to be 

 in part of age greater than the last ice sheet. This hole 

 is gradually growing larger by a sort of spherical scal- 

 ing within the hole. The hole is growing on opposite 

 sides of a narrow crack in the rock. The crack appears 

 to be the thoroughfare by which ground water works 

 into the rock, oxidizes the minerals, and causing them to 

 expand in volume, creates a pressure sufficient to crack 

 the rock. The shape and position of the hole is such 

 that pieces which break off fall out of the hole into the 

 lake ; thus fresh rock is repeatedly exposed to the agency 

 of disruption. Of course this is not a pot-hole, for it is 

 quite impossible for a pot-hole to form in the side of a 

 vertical cliff, yet it is a hole, and after giaciation such 

 a hole would be distinguished with difficulty, in some sit- 

 uations, from a real pot-hole made by swirling boulders. 

 (Figure 7.) 



♦Kirk Bryan, U. S. Geol. Surv. Water Supply, Paper No. 498. (1923), 

 p. 43. 



