270 
The American Geologist. November, i898 
n 
a a. fr6- 
, 
(rtrd. cL 
r* \ 
'• <-; 
m 
w 
and approach each other, as water evaporates, to positions (b), 
and so on. 
Fig. 3 iUustrates a thin sheet of water (a) resting on a 
series of capillary tubes, in a vessel (b) for which it has no 
afftnity. Here surface tension of the water overcomes the 
pull of gravity. 
Fig. 4 illustrates, diagrammatically, the beginning of the 
process with conditions similar (o those in the experiments 
to be described on another page, (the vessel having been re- 
moved from the water supply). 
So long as water could be drawn up from below, while 
evaporation was going on above, there would be no change 
in the conditions except the gradual movement of water up- 
wards, and no necessary movement of the clay particles in any 
direction. 
When the supply of water is removed, surface tension at 
(a) and (b) still operates to elevate water to the original hight. 
If there were no points of weakness in the mass, that is, if 
the conditions were uniform throughout, and each particle 
(c) fixed in its place, there would result, of course, only a grad- 
ual subsidence of the water. If the particles, however, be con- 
sidered free to move in any direction, uniform as regards their 
size, shape and distribution, and if the attraction between 
the water and the sides of the vessel be exactly equal to that 
between the water and the cla}', a vertical shrinkage will take 
place. 
The capillary attraction capable, originally, of lifting water 
