378 Topographical Phenomena in Indiana: [April, 
nate in fine sand, with a pour over the end as shown in the rail- 
road embankment in Figs. 7 and 8, and more definitely in Fig, 10. 
In Fig. 10 the dotted line, a 4, shows the longitudinal section 
of the bar, as left by the last preceding flood. The line, a7, 
shows it as left by the next succeeding flood, and the lines a" 4” 
and a’” 6’ show the conditions after still later floods. Hence — 
bars and bottoms are always building down stream and being 
worn away at their up-stream ends. While they are, in fact d- 
posited in an oblique order, it is rarely that the eye can detect it 
in a longitudinal section of the bottom. | 
A bottom built in this order would be expected to have a sliding : 
tendency like the straw stacks and railroad banks before men- 
tioned. The fact that the trees which grow thickly over the bar 
duce the result we see. The trees do not get their ~ ar | 
from the pressure of the floods. These bottom e Besi 
idly, and in ten years are able to withstand the epee pidi | 
the current is always weak at the lower end of thé a edine 
generally prevail there, otherwise there would be 29 i 
dropped. 
When a growing tree is bent or inclined, it 
