
New York AGRICULTURAL EXPERIMENT STATION. 191 
out of the reservoir through the other slender tube into the 
drainage tube. Thus, whatever water passes out of the reservoir 
F represents the amount of bottom water absorbed by the soil of 
the lysimeter, and whatever flows out through the drainage-cock 
H represents the percolation. If the apparatus is properly 
adjusted, the movements of the soil water may thus be measured 
with a very great degree of accuracy. 
When it is desired to refill the reservoir F, it is only necessary 
to turn the cock G one-fourth of the way around, which shuts off 
the connection with the drainage tube, and at the same time opens 
two other orifices. Water now flows through the rubber tube from 
the supply bottle J and rises into the reservoir, while the air 
escapes from the latter through the opposite end of the stop- 
cock pin. 
Tur Progressive Movement oF Som, WATER DuriInG PERCOLATION. 
Whether rain water, falling on a saturated soil, mingles at once 
with the soil water, the two flowing off together through the 
drainage, or whether it remains in a layer above the water already 
contained by the soil, forcing the latter out and taking its place, 
is a question of some interest to scientific agriculture. It is a 
well-known fact that liquids in contact tend to diffusion when not 
hindered by the presence of insoluble matters. For example, if 
we add pure water to a solution of any soluble salt, the two liquids 
do not remain separate, but each quickly mingles with the other, 
and a more dilute solution is formed. 
But suppose that, before adding the water to this solution, we 
add to it as much of some finely divided insoluble powder as it is 
able thoroughly to moisten. How far, if at all, does the presence 
of this powder interfere with the diffusion process? 
The soil is a mixture of more or less finely divided and insolublé 
materials, more or less completely saturated with a very dilute 
solution of various substances. Water falls on the surface as rain. 
The question raised, therefore, may be applied directly to the soil. 
During the past year, a series of experiments has been conducted 
at this Station, intended to throw light upon this question. These 
experiments have gone so far that the results seem to warrant the 
‘belief that a nearly complete displacement of the water contained 
