136 
AMERICAN JOURNAL OF BOTANY 
[Vol. io. 
The surface was sealed in this case in order to prevent the formation of a hard 
crust as the result of excessive evaporation from the surface and a break in 
the capillary movement from below. 
Koehler (i I), working earlier in this laboratory, found in quantitative 
experiments that growth with low soil moisture is not affected by sealing 
the soil surface, but that with higher soil moisture this sealing brings about 
a marked inhibition of growth. He also found that with very dry soil, in 
containers, there were great differences in moisture content at different soil 
levels, unless the surface was sealed. With the higher moistures, where 
capillary movement was very free, this did not hold except to a very slight 
extent. 
In the manipulation of the soil to obtain uniformity in soil-moisture 
content, an important factor is the degree of compactness of the soil mass. 
Various writers have noted the possibility of error arising from differences 
in this respect. The method finally adopted in these experiments, in order 
to obtain relative constancy, was to pack the soil firmly into the crocks and 
set in the plants. These crocks were then allowed to stand for a month or 
more, by which time the plants were well established and the soil was very 
compact, as would be the case in the field a month or more after planting. 
Then, just prior to the beginning of the actual experiment, one or more of 
the crocks was saturated, water being added until a slight excess of free 
water remained on the surface. The value secured in this way was taken 
as the saturation point. This saturation point varied with the height of 
the containing crock and with the degree of compactness of the soil. In 
series II, low crocks freshly filled with soil gave a saturation value of 40 
percent. In series III and IV, the same soil, in tall crocks which had been 
allowed to stand and settle for six weeks after they had been filled, had a 
saturation value of 35 percent. When the soil was actually saturated, 
whether this required 35 or 40 percent of moisture, the development of wilt 
disease was inhibited, while with the soil moisture just below saturation, 
whether the percentage content was 32 percent or 37 percent, the wilt 
developed rapidly. All percentages of moisture are expressed in terms of 
wet weight. 
Experimental Results 
Experiments I and II. These were in the nature of preliminary experi¬ 
ments and were not satisfactory in certain respects. In experiment I, con¬ 
ducted in the spring of 1919, the temperature was not sufficiently high to 
permit development of the wilt in a virulent form. The crocks were carried 
at low, medium, and high moisture, the actual percentages maintained 
being as follows: 
Soil high in organic content: 
18 percent, 27 percent, 40 percent (saturation). 
Soil low in organic content (a sand loam): 
11 percent, 20 percent, 30 percent (saturation). 
