2/8 The Amaican Geologist. November, i898 
The theoretical possibiUty coincides with the phenomena 
as actually observed. This case is the exact reversal, in the 
position of material, of that of sand floating upon water, to 
be discussed under another general heading. Here, the water 
rests, as a sheet, upon a closely packed series of sand columns 
(not attracting the liquid), and its thickness is limited by the 
number of these, and by the strength -of its surface tension. 
The phenomenon of the sand dunes maybe a capillary phenom- 
enon, though the reverse of "capillary attraction;" and the con- 
ditions pertaining in the cases previously discussed, where 
wetness and an ultimate coherent mass resulted, are also re- 
versed. Surface tension, operating without adhesive attrac- 
tion, tends to preserve the mass in an incoherent state. 
Geologic and physiographic effects result from the com- 
parative rapidity with which such sands are transported by the 
winds. Prof. Shaler refers to this point and says: "In con- 
sequence of this peculiarity of dune sands, which retards their 
deep wetting in ordinary seasons, they are retained in march- 
ing order So that a strong wind may excavate and 
bear away large quantities of the material." 
\ study of sand dunes furnishes a valuable illustration of 
what would happen to the fine clay and rock dust materials, 
through wind erosion alone, if they were not consolidated by 
the processes described. 
The phenomena of drying and shrinking may in some spe- 
cial cases, and in a limited way, counteract the general tend- 
ency to retard erosion, thus seeming to accelerate it. 
"Mud cracks" result, as shown, from shrinkage. The pro- 
cess which forms them may, under certain conditions, accom- 
plish in a local way "this relative accelleration, and may effect, 
in detail, minor physiographic changes. 
I have seen very fine examples of this sort of work while 
making canoe trips down some of our southern rivers, the most 
beautiful cases being on the Chattahoochee. These occur on 
both the Georgia and Alabama sides at many points below 
Columbus, southward to the region of the swamps, near the 
Florida line. 
This interesting stream, within the limits mentioned, trav- 
erses a series of strata, consisting largely of clays, marls and 
sands. For a distance, perhaps, of two hundred miles, it cuts 
