Loss of Fertilizer Constituents in Drainage Water 
A. W. Blair, Chairman Committee on Fertilizers. 
Mr. President , Ladies and Gentlemen : 
The plot experiment, to test the effect 
of various fertilizers and fertilizing ma¬ 
terials on different crops, has been very 
generally adopted in Experiment Station 
work, both in this country and abroad. 
The value of the results obtained in this 
way is beyond our computation; and, on 
account of the knowledge already gained, 
this plan will continue to be followed 
with even greater success in the future. 
However, efficient as the plan is, it still 
fails to give us all the data that we should 
have. 
We measure the amounts of fertilizers 
applied, determine accurately the weight 
and character of the crop produced, deter¬ 
mine the amount of fertilizing materials 
removed by a given crop, and learn some¬ 
thing of the composition of the soil, and 
from these facts draw our conclusions. 
On the other hand, we know but little 
concerning the amount of fertilizing ma¬ 
terial that is carried away in the drain¬ 
age waters. We know in a general way 
that millions of tons of dissolved mate¬ 
rial are carried into the sea every year in 
Florida alone; we know that soils formed 
by the decomposition of limestone may 
finally, through the action of carbonated 
waters, have the lime so thoroughly 
leached out as to render them unproduc¬ 
tive until lime has been applied. We 
know that soluble nitrogen compounds, 
like nitrate of soda, are easily leached 
away. Our knowledge on these points, 
however, is not yet sufficiently accurate 
to enable us to draw quite definite con¬ 
clusions. 
Realizing the importance of more defi¬ 
nite knowledge along these lines, it has 
seemed to us worth while to supplement 
our plot experiments with some work 
that will enable us to control, to a greater 
degree than we can in the field plots, the 
various factors involved. 
PLAN OF THE EXPERIMENT. 
The plan adopted enables us to collect 
the drainage water, resulting from the 
natural rainfall, from a definite quantity 
of soil in which one orange tree is grow¬ 
ing. To do this, four heavy galvanized 
iron tanks, having a diameter of a little 
more than five feet, and a maximum depth 
of four and a half feet, were constructed. 
They were set in the ground forming a 
square so that the distance from center 
to center is about fourteen feet. At the 
middle of this square is a pit, about eight 
feet deep and six by six feet on the in¬ 
side, for collecting the drainage water. 
(Figure, i.) From the lowest point in 
the bottom of each tank (as indicated in 
