MORE LIGHT ON THE ORIGIN OF THE MISSOURI RIVER LOESS. 
BY J. E. TODD. 
My excuse for offering another paper on this well worn subject is my 
recently acquired familiarity with a deposit of Lake Dakota, which closely 
resembles loess, in structure and composition, and in its relations to 
underlying till and to recent channels. In short it seems to be a more 
recent and less extensive formation of the same sort. 
The resemblances are so striking and its fluvio-lacustrine origin so 
unquestioned that a similar origin for the loess of the Missouri is strongly 
suggested. 
THE TESTIMONY OF LAKE DAKOTA. 
Lake Dakota was a body of water related to the James river in late 
glacial times, somewhat as Lake Pepin is to the present Mississippi, but 
much larger. It was about 110 miles long and 20 to 25 miles wide, with 
a depth at its maximum of more than 50 feet. 
The glacial erosion had deepened and widened the valley toward the 
north so that in flooded stage the James river of that time, the muddy 
waters from the edge of the ice sheet a- few miles further north, and from 
the surrounding slopes, poured into the basin more rapidly than they 
cuuid escape through the narrower and stony portion of the valley near 
the south line of Spink county. 
Near the southern end the channels, by which the ice sheet was 
drained before it had vacated the basin, which are now tributaries of 
the James, had been cutting channels into the till of depth approaching 
in some cases that of the present streams. The maximum flood doubt- 
less attended the recession of the ice from the Antelope or Third, mor- 
aine, but was sustained or replenished from time to time during the 
occupation of the Fourth moraine. 
It probably fluctuated in level with the seasons and was quite as 
much a river or cluster of rivers as a lake, for it formed no beaches, 
and yet eventually much of the basin was filled quite uniformly to the 
level of 1200 feet above sea level. The filled portion w'as left level as a 
floor over many square miles, while other portions were 25 or 30 feet 
lower, and in or near the principal channels where w^e may suppose the 
motion of the water may have prevented deposition, or when the water 
subsided erosion may have been greater, the surface of the loam deposit 
is still lower. Besides these deeper channels there are smaller ones 
nearly filled, only narrow winding sags in the surface of the plain, which 
apparently represent the condition of some of the deeper ones before they 
were washed out in the final drainage of the lake. 
( 187 ) 
