362 NATURAL HISTORY BULLETIN. 



near Davenport, Iowa. Even the upper Princeton road is at 

 least one-half a mile from the center of section 20, and fossil- 

 bearing exposures were found neither along this road, nor 

 along the road leading north through the middle of section 

 20. The second exposure along the latter road north of the 

 south line of section 20, however, showed two distinct loesses, 

 separated from each other by an oxidized band, and the lower 

 resting on a buried soil grading below into a gumbo-like layer, 

 which in turn rests upon Kansan drift. Neither of these loesses 

 here showed any trace of fossils, but the lower, light-bluish 

 loess contained iron-tubules and a few lime-nodules, the latter 

 extending also into the buried soil (Yarmouth), and the gumbo 

 below it. 



The Green street cut in Muscatine is no longer avail- 

 able for study, as the banks have been sodded, but an excel- 

 lent similar section has been made recently about two blocks 

 farther west, facing Hershey avenue. In this section a light- 

 bluish fossiliferous loess, with tubules and nodules, lies between 

 Kansan and Illinoisan drift-sheets. In several places there is 

 loose sand under the loess, and in such pockets, near the loess, 

 occassional fossil shells are found. Several facts which may 

 throw light on this case here deserve special consideration. 

 This kind of loess has nowhere been found at the very top of 

 the hills. The presence of a relatively large number of 

 shells of Limiuea, as well as the somewhat peculiar mucky 

 texture of parts of this loess, suggest the presence of ponds, 

 or of seepy, soggy places, such as are not uncommon on hill- 

 sides. In such places springs often carry drift-sand out over 

 the muck of the surface, and produce a similar mixture of ma- 

 terials. There are also in this loess numerous land-shells, 

 such as exist on hillsides, often near such soggy places, and 

 whose shells could be carried out over them, and finally buried 

 in further accumulations of loess. In the deposit under dis- 

 cussion the mixture of shells and sand was noticed only for a 

 short distance below the loess, and this suggests the possibility 

 that underground streamlets, or those occupying deep gullies, 



