Glacial Phenomena of Canada, 331 



partially covered with drift, and well wooded, the lower islets are 

 often only scantily clothed with grass and a few stunted trees and 

 shrubs. Some of them are formed of Laurentian gneiss, and others 

 of Potsdam sandstone. The Potsdam sandstone above the river- 

 bank at Brockville has been ground smooth, and in waving lines 

 passes under the river. The islands formed of Laurentian gneiss 

 or Potsdam sandstone present the same largely mammilated sur- 

 faces, rising from the midst of the river, which between Brockville 

 and the lake gradually increases to 9 or 10 miles in width. All of 

 them are moutonnees, somewhat like the islands of Loch Lomond; 

 and the surfaces of the little islets often slip under the water quiet 

 smooth and unbroken. 



This is one of those cases in which it might be contended that 

 the glaciation of these rocks may be due to the floating ice of the 

 river when it breaks up in spring - . But though it may produce 

 slight effects, there are several conclusive reasons why the greater 

 features should not be referred to this cause. The old glaciation 

 has passed up the country quite beyond the reach of the present 

 river, while the tops of most of the islands rise far above the ex- 

 treme height of the water ; and again, some of the islands with 

 well-rounded glaciated surfaces present vertical cliffs in the river, 

 sometimes 20 feet in height, where the rocks have split away at 

 the joints ; and on these cliffs I observed no sign of that glacia" 

 tion which we should expect to find if the river-ice exercised any 

 important influence. Further, it was observed by Sir Win. Logan, 

 that if the smoothing were produced by river-ice, many of the trees 

 of the islets would be shaved off by the yearly ice, — whereas, when 

 untouched by man, they grow to the water's edge. At the only 

 place I landed (a wooding-station), the rock had been too long 

 exposed to the weather to retain its striations ; but as we passed 

 the islands, I could see indications of striae ; and it is to be wished 

 that some one would settle the point by determining their exact 

 bearings, the chief directions of which, without presumption, I 

 venture to predict will be across the river, and approximately from 

 north to south. 



Drift and Strice in the Valley of the Hudson, including the Ca- 

 naan Hills. — On the banks of the Hudson, south of Albany, the 

 rocks frequently show the familiar mammillated surfaces, — the 

 striations, where I observed them, running nearly north and south. 

 The Highlands of the Hudson also, on a smaller scale, recall the 

 well-rounded outlines of the Laurentine Chain ; and at the mouth 



