Correspondence. 26 1 
the presence of the buried loess at Afton Junction should not be for- 
gotten. This loess is of the old type, and if, as seems probable from 
several lines of evidence, the older loess or white clay owes its peculiar 
properties as much to secondary change as to conditions of original 
deposition, it alone would show a considerable time interval. At the 
Grand river exposure it will be recalled that the upper surface of the 
lower drift showed apparently no signs of either loess or weathering. 
One would hesitate long before basing any argument upon a local dis- 
tribution of such loess as occurs in northeastern Iowa, but it is not so 
hazardous to use such an argument when discussing the older loess. 
The latter is uniformly widespread over the surface of the Kansan and 
Illinoian in southern Iowa. Its character gives one some confidence in 
assigning to water a considerable part in its formation and, inasmuch 
as the buried loess is of the same type as that now found over the up- 
land, it seems well in accordance with what conservatism demands to 
expect it to have at least a considerable distribution. Certainly we 
would look for its presence in the Grand river exposure scarcely a mile 
away. Its absence then becomes a legitimate argument favoring 
erosion before the gravels were laid down. One might suggest that 
this erosion was due to the ice except that in that event one would 
expect till and not water-laid beds to be the first deposits. Further- 
more, while we are becoming able to understand how a glacier may 
deposit over soft beds without disturbing them, we have as yet no case 
of glacial erosion of unconsolidated beds leaving as sharp and un- 
marked a surface as that of the top of the till at the point in question. 
If then erosion be granted it must be hejd to have been pre-Kansan, and 
in view of the freshness of the underlying till, it must have been consid- 
erable. Upon the whole this is believed to be tlie best explanation of 
the phenomena. 
Sixth. It has been shown that there are exposures in the region 
of a drift of peculiar physical type; that this drift is wholly unlike any 
known phase of the Kansan. and that in every in.stance there are some 
independent phenomena favoring the hypothesis that it is distinctly older 
than the Kansan. Whatever one may think of correlations based upon 
physical characters, these facts are certainly of some significance. 
Furthermore the same facts are true of the known exposures of the 
presumed pre-Kansan drift at Muscatine. Oclwcin, Albion, and indeed 
throughout the state. 
General co7icli(sion. It is believed that the argument for a pre-Kan- 
san drift sheet derived from erosion is strong and that it has independent 
value. The arguments from other sources tend to greatly strengthen 
it, and the cumulative force of the whole is believed to be suiTicient to 
put the burden of proof upon those, if any, who would attempt to deny 
the existence of nre-Kansan drift. All would, however, probably agree 
to the statement which the writer believes warranted by the evidence 
in hand, and which he expects future investigations to amply confirm, 
)nit for anything beyond which there is jjrobably as yet no sufficient 
