192 REPORT OF THE HORTICULTURIST OF THE 
in a sample of saturated soil takes place, when a quantity of water 
is added at the surface equal to that already contained by the 
sample, and that diffusion takes place very slowly within the soil. 
As reasons for this belief the following may be mentioned : 
1. When water is added to the upper surface of a column of 
saturated soil, percolation commences almost immediately from 
the bottom, the percolation continuing as long as any water 
remains above the surface of the soil and no longer. In pure sand 
of moderate size grain, the time that elapses between the adding 
of the water and the commencement of the percolation is so short 
as to be scarcely perceptible, even when the column is of consid- 
erable length. In this case, unless the diffusion is instantaneous, 
a more or less complete displacement must occur. 
2. A glass tube, three feet long and one and a half inches in 
diameter, was closed at the lower end with a bit of filter paper, 
which was held in place by a piece of fine wire cloth tied over the 
end of the tube. The tube was then filled nearly to the top with 
pure white quartz sand, of which the grains were something less 
than one hundredth of an inch in diameter. After saturating the 
sand with clear water, one hundred cubic centimeters of a very 
dilute solution of potash, colored a brilliant crimson by means of 
the indicator, phenolphtalein, were added at the top. The colored 
- liquid entered the column at its upper end and passed downward 
rapidly, until all of it had disappeared below the top of the sand, 
when the downward movement suddenly ceased. The moment the 
colored water began to enter the sand at the top, the clear water 
commenced to percolate from the bottom, and as soon as the 
former ceased to move downward, the percolation came to an end, 
One hundred cubic centimeters of clear water were next added, 
when the colored liquid at the top of the column moved downward, 
being followed by the clear water, until the water had all entered 
the sand, when the movement again stopped. Next, the colored 
liquid was again added with the result noted at first. The liquids — 
were alternated in this way until the column was filled throughout 
with the red and white bands formed by the differently colored 
liquids. The boundaries of these bands were not very sharply 
defined, though by no meansobscure. After this, the column was left 
undisturbed for three days, and at the end of this time the bands 
=e te : 
