260 OBSERVATIONS ON SUPPOSED GLACIAL DRIFT. 



uncommon occurrence for the anchors of the nets of a " seal-fiahery " 

 on the north shore of the Gulf of St. Lawrence to be frozen to the 

 bottom at the depth of from 30 to 60 feet ; and when anchors are 

 then raised, they bring with them frozen masses of sand. But it is 

 in rapid rivers that the formation of anchor-ice is most remarkable, 

 and most effective in excavating these beds. It forms on the beds 

 of rivers above the head of a rapid, and frequently bursts up with a 

 load of frozen mud or shingle, or slabs of rock, which it has torn 

 from the bottom. This phenomenon is witnessed every winter in 

 the valley of the St. Lawrence, but it is best observed after a pro- 

 longed term of cold, when the thermometer indicates a temperature 

 considerably below zero. Anchor-ice has only been observed, as far 

 as my knowledge of the subject goes, in rapid currents in open 

 water ; and the sudden and apparently inexplicable rise of the St. 

 Lawrence during extreme cold is most probably due to this cause.* 

 It is not difficult to see how the rivers issuing from beneath the pre- 

 cipitous walls of glaciers, as described by Dr. Eink, may rapidly 

 excavate deep channels by means of anchor-ice, to be widened by the 

 subsequent operations of the glacier itself. Nor is it improbable 

 that by this means a glacier in very cold climates may increase from 

 the bottom upwards with a load of frozen mud and fragments of 

 rock, particularly near its base, when that does not meet the open 

 sea. The great lakes of North America, including Lake Winnipeg, 

 are excavated on the edges of the fossiliferous rock-basins ; and these 

 lakes may represent the boundary of a glacial mass similar to that 

 which now covers Greenland. 



§ 6. Parallelism of Escarpments in America . 



In 1860t I described the remarkable parallelism which exists be- 

 tween great escarpments in America north of the 40th parallel of 

 latitude. 



1st. The Niagara escarpment. 



2nd. The Riding, Duck, and Porcupine Hill escarpment, west of 

 Lake "Winnipeg. 



3rd. The escarpment of the Grand Coteau de Missouri. 



* See "Notes on Anchor-ice," by C T. Keefer, C.E., Canadian Journal, New Series, Yol. 

 vii. p. 173 (1S62 ) 



t See my 'Narrative of the Canadian Exploring Expeditions of 1857 and 1S58,' vol. ii» p^ 

 266, for a notice of these escarpments. 



