W. J. McOee — Geology of the Mississippi Valley. 357 



and erratic, and very rarely, if ever, striated. The materials are 

 frequently stratified and assorted, when the member is thick. Some- 

 times the overlying formation graduates into it, but usually there is 

 a clear line of demarcation. More frequently, indeed generally, 

 it graduates into the subjacent member. Hence its thickness is 

 difficult to determine ; but from one to five feet may be taken as 

 the most common variations. 



Though found to almost invariably accompany member number 

 one in North-eastern Iowa, where there are large quantities of chert, 

 which forms a prominent material in the pebbly member, it has 

 been found absent in many localities. It can, therefore, scarcely 

 be considered a constant feature in the superficial deposits ; but its 

 occurrence is much too general to allow of its being disregarded. 

 It does not always present a common aspect. For instance, in Jersey 

 county, Illinois, it consists of nearly pure sand, often accumulated 

 in long strips or bars. Here the true loess of the Missouri covers it. 



Kames and asar are composed in a large part of this member, 

 considerably thickened. Like member number one, it frequently 

 exists at a higher level than any of the surrounding deposits of un- 

 modified drift. If these two could be mingled, the mixtui-e would 

 contain about as much gravel and sand as, though fewer boulders 

 than, the ordinary glacial drift. They are supposed to have had a 

 common origin ; and from their anomalous distribution the conclusion 

 seems inevitable that they were, at least in some places, deposited 

 before the complete disappearance of the ice of the last glacier. 



Member number three. — This forms the natural surface over most 

 of the region described. It is simply unmodified or slightly modified 

 glacial drift, and is distinguished by the same features here as else- 

 where. Its colour is yellowish to light brown, usually a dirty 

 yellow. Its materials are sometimes imperfectly assorted and strati- 

 fied, with intercalated layers and lenticular patches of gravel and 

 sand, but are commonly wholly devoid of arrangement. Boulders 

 of the Archaean rocks from far to the northward (few have travelled 

 less than 200 miles, and many much more) are abundant and 

 quite large, frequently reaching a diameter of 8, 10, or even 

 12 feet ; and a few are striated. In the visually boundless, and 

 almost perfectly level prairies of Central Illinois, the boulders are 

 but rarely found on the surface, though quite as frequently en- 

 countered in excavating wells as in Iowa. Over these vast cham- 

 paigns there seems to have been a greater assortment and vertical 

 distribution of materials, in the order of their fineness, than is 

 usually observed. Granite is the predominating material of the 

 boulders, but quartzites and other classes of rocks are repre- 

 sented. Hornblendic rocks are rare in this member in North- 

 eastern Iowa, but are frequently observed in some other localities. 

 Lumps and bands of blue clay, easily identified with the subjacent 

 member, are a common feature. The pebbles and gravel are made 

 up both of northern and local rocks. In the parts of Iowa and 

 Illinois in which the Niagara Limestone (the Mississippi Valley 

 equivalent of the Wenlock group) crops out, there are great quan- 



