TEMPORARY DISPLACEMENT OF THE MISSISSIPPI. 95 



As the base of this exposure is 10 or 15 feet above the level of the 

 river, the level of the top of the bank or bluff is about the same as at 

 Warsaw. It differs from the Warsaw section chiefly in carrying two black 

 earthy deposits resembling a soil, one being at a higher elevation and the 

 other at a lower elevation than the soil exposed at the Warsaw section. 

 The sand bed which immediately underlies the loess at Warsaw is not 

 present at the "Yellow banks." The great thickness of the dark earth at 

 the "Yellow banks" exposures seems to justify the interpretation that it is 

 a flood-plain deposit, for a humus stain produced by vegetation rarely 

 extends to such a depth (12 feet). There ma)', however, if it is a flood- 

 plain deposit, have been a time interval of considerable length between the 

 deposition of the upper bed of black earth and that of the loess which 

 overlies it. It seems not improbable that the lower black earth constituted 

 a flood plain, or perhaps a second bottom or terrace, prior to the main filling 

 of the valley. In that case the portion of the section above the lower 

 black earth may be the full equivalent of the fluvial material at the War- 

 saw section. It seems not improbable that the difference in the sections is 

 referable to their relation to the channel of the stream. The difference in 

 level between the bed of a channel and the terraces or flood plains on its 

 border may' easity amount to as much as the differences displayed by these 

 sections. 



A bowlder bed at Keokuk described by Gordon 1 merits notice in con- 

 nection with these exposures at Warsaw and Yellow banks. Between the 

 foot of Main street and the mouth of Soap Creek in Keokuk the rock bluff 

 is but 50 to 60 feet above low water, and is capped by a bed of bowlders 

 about 20 feet in thickness, a view of which is given in PI. VII. In the dis- 

 cussion of this bed Gordon presents three interpretations : (1) That it was 

 formed by river action alone — i. e., as an alluvial bar; (2) that it is due to 

 the cutting down of a till sheet, the coarse material being left as a residue; 

 (3) that it is an accumulation formed at the edge of the ice sheet at the 

 Illinoian stage of glaciation. Of these interpretations the second seems to 

 the present writer, as well as to Gordon, the most applicable, especially 

 when the Warsaw bowlder bed and overlying deposits are taken into con- 

 sideration. The Keokuk bowlder bed, like the Warsaw, is capped by 



'Geology of Iowa, Vol. Ill, 1893, pp. 252-255. For an earlier notice of this bed see paper by 

 S. J. Wallace, Proc. Am. Ass'u Adv. Sci., Vol. XVII, 1869, p. 344. 



