200 



Tlic Clinton Tcrraca lias not l)eeii fully sirdicd. Imt is kiidwii (<> oxtfiid 

 from near Hillsdale to tlic mouth of Brouillet's c-reek. a distance of IL' 

 miles. It is two miles or more in widtli and lies at two levels, the lower 

 in the southern half about 3(i feet above flood plain or 4'.tr> feet A. T. Near 

 Summit Grove it rises by a distinct bench 20 feet higher. 



Island Terraces. Between Clinton and Montezuma there is a line of 

 mid-valley terraces divided by cross-depressions into five islands. They 

 are generally flat-topped with a rather sharp border 20-30 feet high on the 

 river side and a less definite boundary on the bluff side, where they sloi)ej 

 gently to a depression lower than the river banks. (Fig. 4.) The southern- 

 most member of the series is unique. Its surface iiresents a confused' 



Fisurc 5. Clinton Kame Termre. 



as.semblage of hills and hollows. Ilie highest point rising -Id feet above 

 flood plain. (Fig. T).) The surface is fine clean sand whicli may be a wind 

 deposit, a conclusion which the topography alone would justify. The sand 

 is underlaid at a depth of S to 12 feet by coarse gi'nvel. If the topography 

 is detennined b.v gravel deposits they are not alluvi;il. luil c.in be classiHed 

 only as a kame with an eolian veneer. If the island is wholl\- eolian it is 

 ciillicult to imagine why (he wind should be .'~o ellicient there and without 

 effect anywhere else. If it is glacio-Huvial, at core a kame. it means that 

 when the ice sheet covered the valley Itottom, this iminl was the mouth of 

 a sub-glacial stream. In any case, it is one of the most interesting features 



