358 HALL AND SARDESON — EOLIAN DEPOSITS OF ^flXNESOTA 



buff limestone bed. Therefore in tlie retreat of tlie falls the lowermost 

 or buff limestone blocks alone are piled in the ])ed of the river. The 

 same character of river debris continues from the present falls to and 

 beyond the point where the ancient river gravels under discussion now 

 lie, and it is believed for that reason that the condition of the falls has 

 long been the same, and, further, that the falls were far below this point 

 at the time the ancient channel was abandoned, leavinc^ the gravels which 

 are now found in place. 



The old channel or part of a channel borders the northeastern crest of 

 the present river gorge. Beyond the crest is the terrace which was once 

 the river bank ; still beyond that is seen a higher and probably " glacial " 

 river terrace. The channel and lower terrace are now cut across ob- 

 liquely by a stone quarry, and the succession of deposits above the lime- 

 stone is well exposed (see plate 33, figure 3). There is seen a gravel derived 

 from the drift and bearing numerous fossil shells identical with those 

 now found in Minnesota lakes and rivers, namely, Pahcdina, Limnssus, 

 Sphssrium, and Unio. The gravel is about two feet deep, yet varjdng, 

 and rests directly on the limestone bed of the then channel. Traced up 

 the terrace, till is found gradually to intervene between the gravel and 

 the limestone. The gravel rises on the surface of the 15-foot bank or 

 terrace of till and merges into a lag gravel of less thickness lying on an 

 eroded till surface. The shells which have been named occur half way 

 up the sloping bank — that is, about 8 feet above the floor of the channel. 



On the gravel in the old channel rests one foot or less of partly strati- 

 fied sand with fragments of shells. This sand graduates up the terrace 

 into a dune sand 3 or more feet thick, as now seen, which extends far- 

 ther over the adjoining plain and northwestwardly man}" miles. Over 

 the sand in the channel at the foot of the terrace there lies a bed of peat 

 of impure composition and varying texture, from 3 to 4 feet thick. The 

 peat thins out as it rises the slope and merges into the soil of the plain. 



In this section is seen sufficient evidence to prove that the Mississippi 

 river existed as such when the sands on the plain were being deposited. 

 It had already formed a narrowed channel and flood-plain like those 

 which now obtain above Saint Anthony falls. The shells abundantly 

 scattered through the river gravel and up the side of the terrace show 

 the extent and height of the river's domain at the time the shells were 

 inhabited ; at least, they prove the molluscan fauna to have been here 

 before the falls of Saint Anthony had reached Minneapolis, and give 

 special import to the total absence of like shells in the gravels and sands 

 of the plain. Had these deposits above the bank or terrace been found 

 within the river's flood-plain they should bear shells up to and espe- 



