352 
Journal of Agricultural Research 
Vol. XXXI, No< 4 
the difference in color between the soil dampened by the water applied 
and that not yet so dampened. The penetration of mercury and of 
formaldehyde in water solutions was found to agree largely with 
this visible penetration of the water carrying them. Hence the 
approximate penetration of given amounts of these solutions could 
be determined by noting their visible penetration in soil tubes or 
by determining the visible penetration of like amounts of water. 
Wart spores have not been shown to be present below the 8-inch 
level in soil, and this depth is usually reached by applications of 
1 gallon per square foot, the amount adopted as a standard applica¬ 
tion in our experiments. It seems quite possible that deeper pene¬ 
tration would sometimes be desirable and in very dry soil the amount 
of solution applied might need to be increased to insure a uniform 
penetration of 8 inches. If the depth of penetration should be found 
to be in direct proportion to the amount of the application, the 
required amount of increase in application any time the penetration 
was found to be inadequate could be computed immediately instead 
of awaiting the data from experimental tests. Since the pene¬ 
tration of mercury and formaldehyde solutions seems to approx¬ 
imate the penetration of water, water alone was used in these tests. 
In order to determine whether penetration was directly propor¬ 
tional to the amount of liquid added or whether it became greater or 
less (in proportion to the quantity applied) as the quantity applied 
was increased, 36 tubes of Leonaratown silt loam containing 5.73 per 
cent moisture were used. Applications of water at the rate oi 1, 
2, 3, 4, 5, and 6 quarts per square foot were made, using 6 tubes for 
each lot. At the end of one day the penetration in each tube was 
determined without disturbing the tubes by measuring the distance 
from the surface to the very distinct water line. In tubes treated 
with 6 quarts per square foot the water was found to have penetrated 
to the bottom of the tubes. An additional quart per square foot was 
added to each of the other tubes immediately after the measuring 
was completed. One day later the penetration in these tubes was 
again measured. The average increase in penetration, as shown by 
the last column in Table XVIII, is relatively uniform. The range 
of increases, however, in the next to the last column shows a wide 
variation, from to 3% inches per tube. It was thought possible 
that the tubes showing the greatest penetration this second day 
might be those that showed penetration below the average on the 
first day, or that the conditions that caused more than average pene¬ 
tration on the first day would cause an increased penetration on 
the second day in the same tube. Inspection of the figures failed 
to bear out either hypothesis. A curve plotted from any one of the 
columns of averages would be as regular as could be expected with 
such a small number of tests. The maxima in the first two range 
columns also show almost regular curves. In practical work the 
minimum penetration is the important consideration. The data 
of this table indicate that the minimum varies greatly and the 
average penetration is sometimes nearly an inch greater than the 
minimum. Due allowance for this would have to be made in plan¬ 
ning treatments to secure penetration to a definite depth. 
The penetrations secured in this experiment are much higher than 
were secured in regular treatments oi Leonardtown silt loam of low 
moisture content (Table X). 
