Mar., 1923] CLAYTON — SOIL MOISTURE AND FUSARIUM WILT 
135 
soil at frequent intervals, sufficient water being added at these times to 
replenish losses. With this method the moisture is not kept absolutely 
constant, the range of fluctuation depending on the frequency of weighing 
and the rate at which water is lost. In an effort to maintain soil moisture 
at constant, fixed values, and at the same time to do away with some of 
the laborious weighing, Livingston and his associates have devised the auto¬ 
irrigator. This makes good the loss of moisture from the soil as rapidly 
as this loss takes place. The auto-irrigator was given a trial, but it did not 
function properly with low soil moistures because of poor capillarity. The 
capacity of the auto-irrigator, furthermore, was insufficient to supply the 
needs of a large, rapidly transpiring plant. 
After both these methods had been used with various modifications, 
the following combination was finally decided upon. With dry soils, i.e ., 
those having a moisture content of 20 percent or below, the ordinary par¬ 
affin-seal method was adopted, the water being introduced through a glass 
tube leading from the surface to an inverted pot buried in the soil. The 
rate of loss of water from these low-moisture crocks of soil was slow, and 
the soil-moisture content was kept constant by weighing every two or three 
days and restoring the original weight by the addition of water. Measured 
amounts of water were also added between weighings. The auto-irrigators, 
as described by Livingston and Hawkins (12), were installed in those 
crocks which were to be run with a medium (23 percent) to high (35 percent) 
moisture content. Soils held at these moistures did not have the surface 
sealed, the only covering being a layer of non-absorbent cotton. The crocks 
were weighed every other day, and water was added to the surface 
of the soil when necessary. The gross weight necessary for a certain per¬ 
centage of soil moisture was known, also the weights of all materials other 
than soil. The method of moisture control, as outlined, was fairly satis¬ 
factory from the standpoint of manipulation and possessed certain other 
distinct advantages. The surface of the soil with a high moisture content 
was not sealed with paraffin, for in an early experiment sealing was found 
to exert an inhibitory effect on growth if the experiment was continued 
for two weeks or more. This was not found to be the case, however, with 
very low soil moisture. 
Bergman (2) has recently shown that shortage of oxygen is a factor which 
limits growth in the case of saturated soils. Obviously, then, if one of the 
deleterious effects of high soil moisture is the limitation of oxygen supply 
to the roots, any technique which tends still further to limit the oxygen 
supply will affect the results of a moisture experiment. Thus, a soil having 
the pore spaces filled with water until aeration had been reduced to a mini¬ 
mum would respond in a very marked way if the surface was sealed tightly, 
such treatment under these circumstances having the same effect as more 
moisture. Under low soil moisture, oxygen supply is not the factor limiting 
growth, and even with the soil surface sealed there is abundant aeration. 
