ound, yet all were at variance.

Consider, for example, that part of the river lying between Vicksburg
and the mouth. Here, quite aside from the problem as to the hands in
which river-control work should be vested--a very great problem in
itself--three separate and distinct physical problems are presented.

From Vicksburg to Red River Landing there are swift currents which
deposit silt only at the edge of the bank, or on sand bars. From Red
River Landing to New Orleans the problem is different; here the channel
is much improved, and slow currents at the sides of the river, between
the natural river bank and the levee, deposit silt in the old "borrow
pits"--pits from which the earth was dug for the building of the
levees--filling them up, whereas, farther up the river, the borrow pits,
instead of filling up, are likely to scour, undermining the levee. From
New Orleans to the head of the Passes--these being the three main
channels by which the river empties into the Gulf--the banks between the
natural river bed and the levees build up with silt much more rapidly
than at any other point on the entire stream; here there are no sand
bars, and the banks cave very little. In this part of the river it is
not current, but wind, which forms the great problem, for the winds are
terrific at certain times of year, and when they blow violently against
the current, waves are formed which wash out the levees.

This is the barest outline of three chief physical problems with which
river engineers must contend. There are countless others which have to
be met in various ways. In some places the water seeps through, under
the levee, and bubbles up, like a spring, from the ground outside. This,
if allowed to continue, soon undermines the levee and causes a break.
The method of fighting such a seepage is interesting. When the water
begins to bubble up, a hollow tower of sand-filled sacks is built up
about the place where it comes from the ground, and when this tower has
raised the level of the water within it to that of the river, the
pressure is of course removed, on the siphon principle.

As river-control work is at present handled, there is no centralization
of authority, and friction, waste, and politics consequently play a
large part.

Consider, for example, the situation in the State of Louisiana. Here
control is, broadly speaking, in the hands of three separate bodies: (1)
the United States army engineer, who disburses the money a