SIXTH REPORT OF THE DIRECTOR I909 185 



whatever in moving the mass of material now in the cave. If 

 other evidence were needed we could point out the fact that the 

 dentpits extended down by the side of this debris as far as we 

 made our excavations, and that the raised and sharp edges of 

 intersection of the confluent dentpits were in no manner cut 

 across or worn by any movement of the cave filling. Plate II 

 shows a considerable portion of this uncovered face of the wall. 

 The horizontal line a — a separates the formerly exposed surface 

 from that which we uncovered. The position of the camera was 

 about a half meter above the surface of the rilling and this will 

 account for the apparent inclination of the lower lines of dent- 

 pits. Not only had no marks been made on this wall by block 

 motion but the wall itself had been cut away from the filling 

 so as to separate one from the other. We hoped to penetrate 

 deep enough to see if this loose material rested on any undoubted 

 glacial deposit but the limit of time available led us to cease fur- 

 ther prosecution of this work for the season. 



The material taken from the cave filling has itself some valuable 

 evidence to offer. We have already seen that Lake Champlain 

 has had but little effect on its rocky shores so far as pure solution 

 is concerned. The material in this cave filling, at the depth 

 from which the specimen illustrated by plate 20 was taken, 

 shows a very marked amount of solution. An examination of 

 this figure w T ill reveal no trace of vortex action as an assistant 

 factor in the cutting but it does show that the rate of removal 

 depended entirely on the minutest differences in character of the 

 bed material. Apparently no better example of pure differential 

 solution could be desired. If there has been no help from me- 

 chanical abrasion, the filling at this depth is most certainly 

 older than Lake Champlain. The absence of movement in the 

 mass on this cave floor and the depth to which solution has acted 

 on the deeper portions of the filling can lead to no other con- 

 clusion than that these rock fragments were in the position in 

 which we now find them before Lake Champlain began to cut 

 its present series of cupholes and dentpits. 



The westernmost cave of the north wall of Paradise bay was 

 next examined. Plate 21 gives its appearance as seen from the 

 water edge at low water October 23, 1909. This cave is formed 

 on a minor fault plane and close to it, both to the east and west, 

 are parallel faults of small displacement. Freshly fallen blocks 

 from the face of the cliff at (a) reveal nearly horizontal slicken- 



