IOWA ACADEMY OF SCIENCES 
189 
doubtless includes the deposition in several basins at different levels as 
well as in connecting channels, and these of different ages. The ’’flats” of 
Lake Dakota are extensive because young, but traces of similar flats may 
be found, particularly in east central Missouri and in northwestern Iowa 
and eastern Nebraska, remote from larger streams. Even in southwest- 
ern Iowa clear examples of them are found. Tabor, low^a, is upon one 
including nearly a square half mile. On the same level as shown by rail- 
road surveys is a similar area two miles north and again north and 
south of Hilldale. Cases like these could be multiplied from the same 
and other divides of that region. (See also the similar cases recognized 
and described by Dr. Udden, la. Geol. Survey, Vol. XIII, p. 128.) These 
areas present true flats wuth imperfect surface drainage, and some with 
abrupt shoulders and abrupt descents into ravines. The altitude of this 
level is about 1250 feet above the sea and considerably lower than points 
a few miles northeast next to the Missouri river bottom land, which are 
doubtless built up considerably by wund action. This is evident from 
the surface form and its position southeast of a wider area of bottom 
land, a location analogous to that of sand dunes built on the' general level 
of the loess plain east of West Point, Neb. 
It is sometimes urged that in fluvio-lacustrine deposits there would 
necessarily be much coarse material mingled with fine and this has been 
in the minds of some an argument against the aqeous origin of loess, 
but Lake Dakota shows that there is no such necessity for the central 
portions of the deposit. There is more or less coarse material near the 
margins and in the low^er portion, but apparently coarse material becomes 
stuck in the mud at the bottom before it has gone very far. Exceptions 
would of course exist where there was floating ice, but evidently this was 
absent w^hile most of the loam was being deposited. 
The lacustrine loam of Lake Dakota, therefore, furnishes a strong 
argument, by analogy, for the aqueous origin of the mass of the Missouri 
River loess. 
THE TESTIMOIHY OF KIVEK TEEKACES. 
We next proceed to show’ how the analogy of common river terraces 
points in the same direction. 
In the cases already considered there is quite an abrupt change from 
, coarse material upward into massive loam, sometimes wdth considerable 
interstratification of coarse materials between. A similar order is the 
I rule in the structure of river terraces generally. 
It may be seen in the flood plain of any stream of size carrying silt 
in quantity, like the Missouri, Platte, and most streams of the loess reg- 
ion, wherever there is enough coarse material obtainable sufficient to 
form the foundation. 
The coarse material attests the velocity of the stream in its rising 
stage and while it retains its channel. When it overtops its banks and 
covers more or less of its flood plain, especially when it spreads from 
bluff to bluff it attains a lacustrine character and deposits its suspended 
