SILT DIVIDES 171 



miles to the south of Oakland City, there being, in fact, not less than 5 

 or 6 such divides in the 8-mile stretch between the high hills south of 

 Francisco and the south fork of Patoka river. The elevation of the 

 divides is determined by the silts with which they are covered, though 

 the rock is usually not far from the surface. While these elevations may 

 possibly represent a local baselevel of a middle or late Tertiary erosion 

 cycle, it seems more likely, because of their agreement in altitude with 

 the marl-loess limit and with the silt flats both on the till area and on 

 the late Patoka deposits, that they should be referred to some agency 

 limited upward by the altitude mentioned. 



Probable conditions of accumulation. — Notwithstanding the distinct dif- 

 ferences between the physical character and chemical composition of 

 the common and the marl-loess types, it is believed that the evidence of 

 the flats and silted divides at the uniform altitude of 500 feet, the asso- 

 ciation of at least one point with fossiliferous marl-loess, and the probable 

 occurrence of pebbles, points on the whole to the aqueous hypothesis as 

 the one best explaining the features mentioned, and appears to warrant 

 the provisional reference of the portions of the loess of the common 

 type possessing these peculiar features to an aqueous origin. This sup- 

 posed aqueous division of the common loess can probably be traced only 

 along a relatively narrow belt just outside the limits of the marl-loess, 

 though an inconsiderable and unrecognizable portion of the silts for a 

 considerable distance back from the marl-loess belt may be of this type. 

 The great mass of the material lying beyond the marl-loess limits is not, 

 as will be explained, of this origin. 



The absence of visible stratification in the aqueous division of the 

 loess of the common type may be considered as resulting from the prev- 

 alence of uniform conditions, the locations of the deposits being such 

 that they were in general unafi'ected by the currents of the White, 

 Wabash, and Ohio rivers, and received additions only from the exceed- 

 ingly fine and slowly settling silts which still remained in suspension 

 after the deposition of the coarser marl-loess near the rivers. During 

 the deposition, if our view of the origin is correct, there were doubtless 

 many fluctuations of the fluvio-lacustrine level, by means of which the 

 slowly accumulating silts were exposed to the weather and oxidized and 

 the fossils, if ever present, removed by solution. In the case of the marl- 

 loess, on the other hand, accumulation was certainly far more rapid, suf- 

 ficiently, it is thought, so that a given layer of silt or shells was covered 

 and protected before oxidation or solution had time to materially affect 

 them. 



It seems almost certain, however, that parts of the loess of the flats 

 and divides, though accumulating in water, must have been derived, at 



