220 The American Geologist. ^p^^'- i^^*- 
tion of the Tuscumbia boulders, where, as already remarked, 
it is easily perceived that 'the Osage river, flowing entirely out- 
side of the glaciated region, would not be subject to any ab- 
normal rise of its waters from causes operating in the upper . 
part of its basin. On the other hand, as the water in the Mis- 
souri rose an actual current of backwater would set up into the * 
Osage valley, which might easily carry ice floes loaded with 
boulders which had been brought down from the glacial tribu- 
taries of the Missouri, while such an upward current is ab- 
solutely necessary to get them into the positions where they are 
found. The one in Shut-in creek especially was not far from 
150 feet above the present level of the Missouri, which would 
involve a rise of about 200 feet in the Missouri to accomplish 
the result. 
Phenomena analogous to this are too familiar to need speci- 
fication. My own attention, however, to such a movement of 
water up a tributary stream was early attracted by an occur- 
rence upon Hubbardton creek, in Westhaven, Vt. This creek 
flows from the north into the lower portion of Poultney river, 
or East bay, as it was called, near where it enters the head of 
lake Champlain. At one time a series of terrific thunder show- 
ers passing over the headwaters of Poultney river raised the 
stream in its lower portions to a great bight, while there was 
no corresponding rise in Hubbardton creek. As a result, a 
milldam in the lower part of the creek was bodily carried up- 
stream by the back water which set in from East bay. A sim- 
ilar current of back water annually sets up the Williamette 
river in Oregon, upon the rise in the Columbia, due to the 
melting snows about its headwaters. 
This positive evidence of such a rise in the Missouri river 
is sufficient of itself to compel us to imagine conditions which 
would bring it about. In doing this we are aided by 
(3) The probable diminution of the gradient of the Mis- 
souri river by reason of a differential depression of the land 
towards the north. This is a phenomenon so generally wit- 
nessed to over the northern portion of the glaciated area that 
it needs no further reference. Such a dift'erential northern de- 
pression is not only possible, but is altogether probable in 
accordance with the analogy of a great number of known facts 
relating to the region. But we are aided also 
