The Sm-facc Tension of Water. — Ladd. 279 
through them so as to expose them beautifully in section, its 
banks being often many feet high. 
The singular physiographic feature prorkiced by the force 
under discussion, is the occurrence of promontory-like pro- 
jections beyond the vertical banks, over which minor streams, 
in falls and cascades, tumble into the river. 
Since flowing water naturally erodes such materials as 
marls and clays, I was puzzled to account for the absence of 
the recess usually found where one stream enters another. 
It seemed remarkable that a considerable body of water could 
be precipitated into the river, thereby increasing the destruc- 
tive force, at the given poipt. and still produce the efifect of 
locally retarding erosion, that is. leaving little residuary prom- 
ontories, while the clay walls retreated. But such the phe- 
nomena were. The material of the promontory was that of the 
bank in general, continuous with it, and in no way a deposit 
by the stream: nor did it appear to be in any case hardened 
by a local deposition of mineral matter. The explanation 
sought seemed to be found, in a general way, in the fact that 
the retreat ofthe banks was accomplished in part by some other 
means than the under-cutting of the river, or erosion by rains; 
and after careful observation a full detailed explanation seemed 
to be as follows: 
(i) Erosion by these lateral streams is at a minimum, 
owing to the absence in them of transported material, due to 
their flowing but short distances: and, in the main, across flat 
surfaces rank with vegetation, which filters out the sediment. 
(2) The clay material of the bank absorbs by capillary at- 
traction the water precipitated on its surface by night as dew. 
which falls very heavily in this region. This moisture is then 
rapidly evaported by day, through the heat of a semi-tro]Mcal 
sun. even though augmented by capillary waters from depths 
below the surface. The result is a consolidation and shrink- 
age on the face of the bank, which causes mud-cracks: and 
^.ince the extreme surface is the first to shrink, there takes 
place a curling up of the edges of the now isolated patches, 
and a peeling off, in fragments, of a thin layer of the clay or 
marl material. (3) Beneath the waters of the lateral stream 
the clay is continuou.sly in a wet state, and yet. though sufifer- 
in<2f from the under-cutting of the river and the erosion of the 
