113 
tions showed no necessary relation between the amount of plant food as determined 
by our methods and the yields of crops; that is to say, that high yields were nol 
associated necessarily with high amounts of plant food, and the reverse; that the use 
of fertilizers, which we all admit are beneficial to the soil, could not under thin rea- 
soning be for the amount of plant food they added to the soil, hut it has seemed evi- 
dent that the effect must be due to some other cause, which, as the coroner's jury 
would say, "is unknown at present to the jury." Further, it has seemed evident to 
us that the effect was probably associated with the physical condition of the soil, as 
there were certain evidences that would lead us to think that the trouble, in many 
cases at least, was a lack of a suitable moisture supply for the needs of the crop. 
For example, on our Susquehanna clay which we find in large area- between here and 
Baltimore, the vegetation has the desert characteristics; the leaves protect themselves 
against evaporation, and only certain classes of plants are found, although we know 
that the soil has an ample supply of moisture as measured by the moisture determina- 
tions made in the laboratory. We have long recognized that the infertility of this 
soil is not associated with a low food content, as analyses had shown no difference either 
in the physical or in the chemical properties as compared with the limestone soils of 
Pennsylvania and western Maryland. 
Starting from this point, therefore, where Bulletin 22 left off, we made a thorough 
investigation of the physical properties of the soils with particular reference to the 
movement of water, believing, as we did at that time, that fertility was at least 
largely dependent upon the movement of water with its dissolved salt content to 
the roots of the plants. We thought we should find that certain soils were unable 
to supply the necessary amount of plant food, simply because they were physically 
unable to deliver to the plant an adequate amount of this nutrient solution. After a 
very thorough investigation of the problem, we found, to our surprise, that there is 
practically no difference in the rate of movement of water in soils, even of very dif- 
ferent texture, when you have the amount below the optimum and considerably 
below the point of saturation of the soils. We had to devise new methods for the 
study of this subject, for the old methods of investigation, where we allow water to 
percolate through the soils or where we allow water to climb up by capillary attrac- 
tion through dry or moist soils, give us conditions that are entirely unlike field con- 
ditions. We never grow an agricultural plant under such conditions at all. It is, 
for the most part, of little interest to us to know what amount of water the soil can 
move if there is an optimum maintained at all times. The important thing was what 
the soil would deliver after it had become partially dried — that is, after it was 
approaching the drought limit. There was the place where we would expect to find 
the individual characteristics of the soil that would affect crop production, so that 
our aim was to study the movement in soils far short of saturation and below the 
optimum water content for the plant. 
It will be impossible, in the short space of time that I have now, to go into the 
details of the investigation. That will all be presented in sufficient detail in the 
publication that we "will probably issue in a short time. Suffice it to say that we 
have found that, in a soil below the point of optimum water content, the water has 
very different properties from what we recognize in water in mass. It does not obey 
the ordinary physical laws as we recognize them in capillarity, nor does it obey the 
laws of electricity as we recognize them in solution in mass. In other words, water 
in a slightly moist soil has lost some of the properties that we attribute to water in 
mass. Now, why this is we do not know. The electrical conductivity is very much 
higher in this water; it is fifteen times as high in a moderately moist sand as it ought 
to be. We do not know where it has gone; we do not know what has become of it; 
we do not know what has happened to it. It may be due to the thinness of the 
film. We have not been able to get these results with soap bubbles, however thin 
we may blow them. We believe that the film in the soil is much thicker than in 
the bubble with which we have compared it. At any rate, all I want to tell you 
now is that the electrical properties of the water in a moderately moist soil are 
different from those in the saturated soil, and different from the properties of liquid 
in mass. Another thing is that moisture in a moderately moist soil that is below the 
optimum quantity for plant growth does not obey the ordinary laws of capillary 
movement. There is a change. What it is due to we do not know, but the move- 
ment in these moderately moist soils is entirely different from the relative rate of 
movement of water in percolation experiments or in the capillary rise through moist 
or dry sands. 
As, however, we have found no relation between the delivery of water from a dry 
and moderately moist soil, whether it be a light sand or an apparently impervious 
brick clay, it was obvious that it was not in this respect that we should look for the 
21736— No. 142—04 8 
