110 Harper: Trip oN WARRIOR AND TOMBIGBEE RIVERS 
following counties alternately on the right and left: Greene, Hale, 
Sumter, Marengo, Choctaw, Clarke, and Washington (besides Bald- 
win and Mobile below the point where our journey ended). 
Like most rivers of gentle slope, this one meanders so that dis- 
tances by water are about twice as great as by air line, and forms 
bluffs on the convex sides of the bends in the usual manner. The 
bends are naturally shorter and sharper on the Warrior than on the 
Tombigbee, on account of the smaller volume of water ; and for 
some reason not altogether obvious, they seem to be most fre- 
quent in the region of the Eutaw formation, in the upper parts of 
Hale and Greene counties. 
The banks of the river present a variety of aspects, according 
to the material of which they are composed, and their position 
with respect to the meanders and to the sun. Lithologically they 
are sandstone, limestone, clay, sand, mud, or various intermediate 
conditions ; and there is often considerable variation in a single 
formation. The oidest material is the stratified Cretaceous and 
Eocene rocks, which in some places extend as much as 200 feet 
above the water, in steep bluffs, and elsewhere are entirely under 
water, or concealed by later formations. Capping the stratified 
rocks on bluffs which extend above high-water mark can often be 
seen the stiff reddish loam of the Lafayette formation, which is 
probably Pliocene. Bluffs not more than 30 feet high are often 
composed entirely of “second bottom ” deposits (Pleistocene), 
which are usually of loose buff-colored loam, and level on top for 
considerable distances. Recent alluvium is confined chiefly to low 
banks subject to frequent inundation. These are of comparatively 
limited extent, however, and there is very little swamp along any 
of the rivers in the Cretaceous and Eocene regions of Alabama. 
The alluvial banks often slide into the river, carrying down 
whatever trees grow on them (mostly willows in such places) and 
drowning them. This phenomenon is especially common in the 
region of the Tuscaloosa formation (which consists mostly of 
slippery clay, and lines the river-banks for the first 60 miles 
below Tuscaloosa), where rows of dead willows are in sight from 
nearly every point. In this same portion of the river, and appar- 
ently nowhere else on our route, are many cut-offs or ox-bow 
lakes, showing that the channel has been shifting there rather fre- 
