On the Equilibrium and Motion of Liquids in Porous Bodies. 205 
number of beads to which motion is imparted is proportional to 
the magnitude of the pressure applied. Consequently the oppo- 
site extremity of the column only begins to be displaced when 
the difference between the pressures acting at its two extremities 
reaches a limit proportional to the number of beads in the 
column ; and if this number be increased indefinitely, the limit 
in question will also be indefinitely increased. In this manner 
a pressure of three atmospheres, acting incessantly for fifteen 
days at the extremity of a very fine tube containing a great 
number of beads, failed to produce the least visible displacement 
of the liquid. 
Inversely, when a partial vacuum is produced at one end of 
the tube the nearest bubbles of air dilate greatly, the interme- 
diate ones less, and those furthest distant remain unaffected so 
long as the diminution of pressure does not exceed a limit pro- 
portional to the number of bubbles or beads. To make the ex- 
periment, a very long tube containing a great number of beads 
may be cemented into the upper part of a barometer-tube. The 
mercury will then maintain precisely the same position as it 
would do if the tube were perfectly closed. 
This experiment shows that the pressure exerted at one ex- 
tremity diminishes abruptly by a constant quantity at each place 
where the continuity of the column of liquid is interrupted ; and 
this fact may be easily explained. 
For it is probable that the first effect of the pressure H! is to 
alter the form of the nearest bead of liquid, by hollowing out its 
anterior surface and increasing the radius of curvature of the 
meniscus which bounds its posterior surface. A portion, L, of the 
pressure being thus expended in the deformation of the first 
bead, a deformation which cannot exceed a certain limit, and 
which is the same for all the beads, the residual pressure H'—L 
is transmitted by it to the next succeeding air-bubble, and thus 
to the second bead, which, in becoming similarly deformed, again 
diminishes the pressure by the same amount as before. This 
action continues until the originally applied pressure has, by 
equal, successive decrements, one at each bead, become reduced. 
to H’—nL=H, the normal pressure in the tube, when, of 
course, equilibrium results. 
By generalizing this idea, it is easy to show that the chaplet 
may assume an infinite number of states of equilibrium, whose 
conditions may be calculated ; and experiment is found to verify 
the results of calculation. 
It will be at once seen that these properties must considerably 
modify the ascent of liquids in capillary tubes. There are in 
fact two cases to be distinguished. 
First. After raising the tube in the liquid in which one end is 
