254 
Journal of Agricultural Research 
Vol. XXXI, No. 3 
of silica removed by the crop bear little or no relation to the yields 
of the differently treated pots given in Table I. The quantities of 
phosphoric acid removed, however, approximate fairly closely the 
relative yields produced by the various treatments. It would thus 
appear that the silica in the plant was not a determining factor in the 
yield and that the phosphoric acid assimilated probably was the lim¬ 
iting condition of growth. 
The last column in Table II shows that the application of silica 
gel promoted the assimilation of phosphoric acid from rock phosphate 
to a marked extent, while it increased the assimilation of phosphoric 
acid from acid phosphate comparatively little. 
In a complex medium such as a soil, silica gel might promote the 
availability of rock phosphate without the increased availability 
being apparent through any test except that of the growing crop. 
However, in the simple medium of this experiment—quartz sand and 
soluble salts—an increased availability ot rock phosphate should be 
revealed by an ordinary chemical analysis, since the quartz sand, 
except for slight impurities of iron and possibly other substances, 
would be without action on phosphoric acid, brought into solution. 
Laboratory tests were conducted on the solubility of rock phosphate 
simply with the nutrient salts used in the basic fertilization. 
EFFECT OF COLLOIDAL SILICA ON THE DECOMPOSITION OF 
ROCK PHOSPHATE 
The concentration and relative proportions of the nutrient salts 
added to the pots (basic fertilization) in the experiment just described 
of course underwent marked alteration with the growth of the plants. 
It was impossible, therefore, to determine the solubility or decompo¬ 
sition of rock phosphate in a solution representative of that present 
in the pots during all stages of the growth of the crop. However, 
the effect of silica gel in promoting the growth of plants supplied 
with rock phosphate was apparent when the plants were only a few 
inches high, and at this time the composition of the nutrient salts 
must have been essentially the same as at the start of the experiment. 
It was not expected that the quantities of phosphoric acid found in 
solution on shaking up rock phosphate with the nutrient salts and 
silica gel would correspond quantitatively to the quantities present 
in the pots at all times, but it was thought that the results obtained 
would mdicate qualitatively the effect of silica gel on the solution 
or decomposition of rock phosphate in the pots. 
In the first test, 0.5 gram or rock phosphate or 0.3 gram of acid 
phosphate was shaken up with a mixed salt solution or with a salt 
solution plus silica gel. The salt solution contained the equivalent 
of 1.25 grams of anhydrous salts in 250 cubic centimeters of water, 
the same salts being present in the same proportions as in the mixture 
used for the basic fertilization of the pot experiment. The silica 
gel added was the same as that used in the pot experiment. The 
mixture was shaken at intervals during seven days standing. It 
was then filtered through a Pasteur-Chamberland filter and P 2 0 5 was 
determined in an aliquot of the filtrate, which was perfectly clear. 
Prior to precipitating the phosphorus, special precautions were taken 
to rid the filtrate of silica which might contaminate the phosphorus 
precipitate. Three dehydrations with HC1 were made with inter¬ 
vening filtrations. The P 2 0 5 was determined as Mg 2 P 2 0 7 . The 
