484 THE GEOLOGY OF MINNESOTA. 



Chains of lakes. 



been their original condition, if they are parts, as thus indicated, of ancient 

 water-courses. Nowhere else in my exploration of the glacial drift, have 

 similar chains of lakes been found, bordered and occasionally divided by 

 areas of till, without notable deposits of modified drift, and not occupying 

 distinct valleys of former streams. Yet these plainly connected series of 

 lakes, converging, and one of them receiving a tributary branch, in their 

 course toward the south, are related to each other like confluent rivers. 

 Their origin cannot be referred to the ordinary causes and conditions, 

 already reviewed, which produced the irregularly scattered lakes of drift- 

 covered areas: but, excepting this arrangement of its lakes. Martin county 

 is not distinguishable from the surrounding region of drift. 



The explanation of these series of lakes, which seems most probable, 

 is that they mark interglacial avenues of drainage, occupying portions of 

 valleys that were excavated in the till after ice had long covered this region 

 and had deposited most of the drift-sheet, but before the last glacial epoch, 

 which again enveloped this area beneath a lobe of the continental glacier. 

 partially filling these valleys, and leaving along their courses the present 

 chains" of lakes. Fossiliferous beds are occasionally found in this and ad- 

 oining states, and, significantly, at a few places within the basin of lake 

 Agassiz, intercalated between thick deposits of till. Some of these inter- 

 glacial beds, doubtless including those in the Red river valley, since cov- 

 ered by lake Agassiz, were formed after an ice-sheet had extended to the 

 extreme southern limit of the glacial drift. They prove that the long, very 

 severely cold period in which ice-tields reached south to northeastern 

 Kansas, St. Louis, and southern Illinois, was succeeded by a milder climate. 

 under which the ice was melted from Minnesota and even as far northward 

 as to Hudson bay, again permitting plants and animals to occupy the land. 

 The terminal moraines of the Northwest, formed by a later ice-sheel, show 

 that another great epoch of cold once more buried the north half of the 

 continent under ice. which, however, did not extend so far south as before. 

 This ice was divided at its border into vast lobes, one of which, about thru*; 

 hundred miles long and one hundred miles wide, and probably from a tenth 

 to a half of a mile thick, was accumulated upon the area that stretches 

 from the head of the Minnesota river southward to central Iowa, including 

 Watonwan and Martin counties, its width at this latitude being from 



