Distribution of the Loess. — Wright. 221 
(4) By the supposition of a reduction of the gradient of 
the Missouri river by reason of corresponding high water in 
the Mississippi ; for, at the time of the discharge of these im- 
mense vokimes of water into the valley of the Missouri, there 
was an equal, if not greater discharge, into the Mississippi 
valley proper. Whatever rise would occur in the Mississippi 
w^ould tend to lower the gradient of the Missouri to that 
amount. 
In considering this matter of the gradient of streams at 
flood time, it is to be remembered that a rise of a given amount 
in the upper portion of the stream is distributed over a long 
distance in its outflow. At the time of the remarkable flood in 
the Missouri river on the 8th of June, 1903, I was at St. Louis, 
when the water in the Mississippi had risen to a hight of 37 
feet and was overflowing the flood-plain for a width of about 
five miles, and in some places of ten miles. The announcement 
of the Weather Bureau led us to expect that the flood would 
begin to subside on that day, but four days after, upon the nth, 
the water, instead of subsiding,' had risen another foot, mak- 
ing 38 feet, though there had been no rain in the meantime. 
The truth was that the supply of water from a limited area 
in the valley of the Kansas river had been such as to produce a 
lake in the Missouri valley several hundred miles long, and 
from five to ten miles wide, furnishing a mass of watet whose 
movements baffled the calculations of the Weather Bureau. 
At Kansas City, the Kansas river, which brought in the larg- 
er amount of the waters of the flood, pushed out such a strong 
current into the Missouri valley that it carried railroad cars 
completely across the main current of the Missouri, and land- 
ed them upon the other side of the valley. This current of the 
Kansas river across the Missouri acted like a dam, producing 
slack water in the Missouri for many miles above. 
(5) Finally, the Missouri river for a distance of more than 
100 miles both above and below the mouth of the Osage river 
flows in a gorge 300 feet deep, and less than three miles wide; 
while at Hermann, 25 or 30 miles below the mouth of the Osage 
river, it is barely two miles in width between the upper rock 
blufifs, and would average at this point considerablv less. We 
have, therefore, all conditions necessary for the production of 
such variations in the water levels of the middle Missouri val- 
