MUSEUM OF COMrAKATIYE ZOOLOGY. 



190 



head upon the casteni ; neither could rock in pkce be detected in tlie 

 low banks of drift abreast the falls, nor along the carry. In the chan- 

 nel at the foot of the falls, the ledge was mostly hidden by accnninlated 

 granite bowlders, wdiile at the upper part the water was seen to pitch 

 over successive shelves of granite in place. 



Though it was hero that we first came upon fixed rock in the river 

 bottom or banks, we had seen, half-way between this and the next tall 

 below and some distance back from the west shore, two high forest-cov- 

 ered hills, wliich were the first considerable elevations met with since 

 wo left the lakes below. On the southern slope of each hill is a naked 

 clilf. Circumstances rendered it impossible to land and push through 

 the woods for a visit to the cliffs; and no second view of them occurred, 

 since my return from the region took place by another route. 



The passage through the dead-water of more than two miles to the 

 fourth lull showed no trace of rock in place, cither in the channel or along 

 the shores. Nor did the neighboring low hills exhibit any faces of 

 exposed rock. 



Pockwockamus Falls, the fourth, occupy about twenty rods of the riv- 

 er's length, beginning abruptly from still water above, and ending as 

 abruptly below in water of like character. The bed is an uneven floor 

 of granite ledge, measurably free from bowlders in the middle third, 

 but on cacli side covered with large blocks, mainly riven from the under- 

 lying fixed rock, which docs not differ materially from that of the third 

 fall. 



At the fifth or Aboljacarmegus Falls, a mile above the fourth, is the 

 next exposure of rock m situ. Tlic river makes here a sudden bend, 

 flowing from northeast to southwest, and is exceedingly narrow. The 

 lengt.h of the fall is ten or twelve rods only, and the underlying granite 

 differs strikingly from that of the third and fourth falls, and from that 

 wliich makes up Ktaadn. In places it is highly porphyritic, and occa- 

 sional patches of several square feet consist of massive feldspar or quartz, 

 not constituting veins. At the tliird and fourth falls the granite, being 

 highly jointed, has become divided into rhombohedral blocks of all sizes. 

 Worn at the angles in various degrees by attrition, these make a large 

 proportion of the bowlders that at both those falls conceal the rock in 

 place over much of its area. At the fifth fall the fragments split from the 

 solid mass in somewhat lenticular forms, never in rhombohedral blocks. 

 About the foot, the only pot-holes anywhere seen occur in considerable 

 number, but not of great size. Their presence here and absence from 

 the falls below must be explained by difference in the texture of the 



