188 THE GLACIAL LAKE AGASSIZ. 



gravel aud sand, extendiug' 8 feet above and at least as far below the gen- 

 eral level of the adjoining surface, and the visible width of the deposit is 

 about 30 rods. How much deeper it may extend, perhaps with increasing 

 width, is undetermined. Its gravel, which is nearly all limestone, contains 

 pebbles up to 6 inches in diameter. No bowlders occur in this excavation, 

 and they are rare upon the surface of this and other such comparatively 

 broad and high jiortions of this esker, none being sometimes seen along a 

 distance of several rods; but in its narrower and. shghtly lower portions, as 

 traced in its somewhat crooked course northward through the next 1 J miles, 

 it. often is found to be sprinkled with frequent bowlders up to 3 or 4 feet in 

 diameter, mostly Archean. They appear to ha^e been stranded, as at Birds 

 Hill, immediately after the ice walls inclosing the esker were melted, or 

 even during that process, and before the melting of the ice under this 

 gravel and sand allowed the water of Lake Agassiz to submerge the more 

 massive portions of the ridge. Only a small depth of water, probably not 

 more than 30 or 50 feet at the most, would be required for this; and after- 

 ward tlie melting of the underlying ice gave to the lake here a depth of 

 fully 500 feet. Farther to the north the esker sinks or is merged in the 

 moderately undulating till which there forms the surfoce. The crest of 

 this peculiar ridge, approximately 800 to 805 feet above the sea, undulates 

 3 to 6 feet within short distances, not showing so much uniformity in eleva- 

 tion and directness in its course as are characteristic of beach ridges; and it 

 is the only instance observed in all niv exjjloration of Lake Agassiz where 

 a gi-avel formation nearly resembling a beach bears bowlders on its surface. 

 Not a single bowlder has been anywhere found on or within the beaches of 

 this lake; nor have eskers like the Birds Hill group or like these of smaller 

 size and more stream-like courses been oljserved b}" me in any other part 

 of this lacustrine area, excepting the vicinity of Red Lake, in Minnesota. 

 But eskers doubtless exist here and there throughout the belt of modified 

 di-ift that extends upon this area from Red Lake by the Lake of the Woods 

 to Birds Hill and Burns Ridge, and probably tliey continue north-north- 

 westerly upon the country between Lake Winnipeg aud Shoal Lake. 



