42 
Transactions Texas Academy of Science. — 1907. 
To determine the needs of the soil for plant food is not, however, a 
simple matter. The chemical analysis of a soil, while it affords indi- 
cations, is not certain, especially in soils long in cnltivation. For 
example, the rice soils of Texas contain from 500 to 800 pounds of 
phosphoric acid per acre in the surface foot, but the application of 
twenty-eight pounds phosphoric acid in acid phosphate has a great 
effect upon the yield. There is plenty of phosphoric acid in the soil, 
but not sufficient of it is in such forms that plants can take it up, hence 
the decided effect of the addition of a comparatively small quantity of 
jihosphoric acid that plants can take up. Ordinarily chemical analysis 
shows the amount of plant food in the soil, but it does not distinguish 
between that which the plant can take up and that which it can not. 
Hence while this chemical analysis shows the possibilities of a soil under 
good treatment, it does not show the immediate needs of the soil. 
Another method of ascertaining the need of a soil for plant food is 
by means of field tests. A field is selected uniform in all respects, and 
it is divided into plots, say one-tenth acre each in size. These plots are 
treated alike in all respects, excepting that they receive different fer- 
tilizers. The yield of the crop planted is taken as a measure of the 
effect of the various applications. 
Plot experiments are subjected to vicissitudes of the weather, the 
depredations of insects, and other variable conditions. A soil which 
appears uniform may really be variable in depth or chemical composi- 
tion. Heat and cold, moisture and drought, affect the crop, and the 
controlling influence in the development of the plant may be, at various 
periods during the season, other than the plant food in the soil. For 
these reasons the results of plot experiments are often irregular, con- 
tradictory and unsatisfactory. Yet, if carried on for a sufficient number 
of years, and when properly planned and conducted, plot experiments 
give results which can be secured by no other means of experimenta- 
tion, and the ultimate test of laboratory conclusions must be the field. 
Plot experiments are expensive, require considerable time, and the 
results are applicable only to the kind of soil under experiment. The 
results of such experiments should be correlated with the various types 
of soil before general application is made of the results. Too often 
field experiments on one kind of soil are applied to soils entirely 
different. 
Another method of ascertaining the need of the soil is by pot experi- 
ments. In these experiments, the soil is mixed thoroughly and portions 
of ten pounds or more placed in pots, and plants grown therein. For 
example, if we desire to ascertain whether or not a given soil is defi- 
cient in phosphoric acid, we would compare the crop produced by com- 
plete plant food with the crop made in the same soil without phos- 
