422 THE GLACIAL LAKE AGASSIZ. 



uortheast part of section 14, townsliip 163, range 57, about 3 miles north- 

 west of Walhalla, the upper Campbell shores form such a terrace, which 

 falls from 1,075 to 1,035 feet; while a more moderate slope of sand and fine 

 gravel below, to 1,025 feet at the road running northwest from Walhalla, 

 probably represents the lower Campbell stage. 



Three miles farther northwest and about 1 mile south of the interna- 

 tional boundary a terrace of gravel and sand in the west part of section 

 34, township 164, range 57, marks the Campbell levels of the lake. The 

 front of the terrace rises steeply from 1,015 to 1,035 feet above the sea, 

 and its top has a further gentle ascent of 10 or 15 feet in its width of about 

 50 rods to Avhere it aliuts on the base of the lowest escarpment of the Pem- 

 bina Mountain, which rises from 1,050 to 1,100 feet. From the top of this 

 escarpment a terrace or plateau of till and underlying Cretaceous shale 

 extends across a width of three-fourths of a mile west to the principal 

 Pembina escarpment. The upper Campbell level probably passed along 

 the top of the sand and gravel terrace, near the elevation of 1,045 feet; the 

 second knel t)f the series was near the verge of this terrace, approximately 

 1,035 feet; and the third and lowest stage coincided with the lowest third 

 of its steep front. 



CAMPBELL SHORES IN WESTERN MANITOBA. 



(PLATES X.XX-XXXIII.) 



Along the course of the Cretaceous terrace, thinly covered with till, 

 which borders the base of the Pembina Mountain for at least 25 miles 

 northward from the international boundary, as described in connection with 

 the Tintah beaches, the upper Campbell sliore-line, there having an eleva- 

 tion of 1,045 to 1,050 feet, coincides generally with the low escarpment 

 which forms the east margin of this terrace. A portion of the sculpturing 

 of this escar})ment was doubtless done by the waves of the lake; but the 

 main outlines of the terrace as a bench intei'mediate between the expanse of 

 the Red River Valle}' and the high Pendiina escarpment seem clearly attrib- 

 utable to subaerial erosion before the Ice age. The first locality where I 

 observed a distinct beacli ridge of gravel and sand referable to this stage 

 is in section 3, township 4, range 6, a half mile west of Nelson, and thence 



