138 
Journal of Agricultural Research 
Vol. XXV, No. 3 
portions being discharged through the waste conduit)', and the color of 
this solution is compared with standard color solution A at frequent 
intervals until the two colors appear alike. When this occurs, the start¬ 
ing time of the experiment is recorded. The gas circulation continues 
at the established rate, and the indicator solution becomes gradually 
darker, until it matches standard color solution B. Then the time the 
experiment ends is recorded. The difference between these two time 
records gives the number of minutes required for the indicator solution 
to take up enough oxygen to alter its color from that of standard color A 
to that of B. Since all of this oxygen has come through the walls of the 
absorbing cylinder the time required for this color change may be taken 
as inversely proportional to the rate of entrance of oxygen into the 
cylinder. It is of course necessary to allow the gas circulation to go on 
for a considerable time—at least 3 to 5 hours—after every change in the 
exposure of the absorbing cylinder, in order to allow the cylinder and 
the gas spaces of the apparatus to come into dynamic equilibrium with 
the new surroundings. In practice, readings were generally taken before 
this equilibrium had become established, and the numerical results indi¬ 
cated the transition from the old to the new conditions of exposure. In 
order to avoid the accumulation of oxygen in the cylinder it has been 
found best to maintain the gas circulation through the system at all 
times, whether readings are being taken or not. 
According to several approximate determinations the color change 
just mentioned required about 0.015 cc. of oxygen under the ordinary 
pressure and temperature conditions of the laboratory air. This constant 
of the apparatus has not yet been precisely determined, however, since 
its exact value does not enter into the problem dealt with, as will appear 
below. As far as tests have gone, with the various features of the appa¬ 
ratus as they were, it appears that small variations in the temperature 
of the laboratory were without influence upon the results. 
EXPERIMENTATION WITH SOIL 
In the following paragraphs will be described a series of experiments 
made with a box of garden soil in the laboratory. The results are to be 
regarded more as an illustration of the way in which the apparatus may 
be used than as quantitative data on the soil used. They do furnish 
interesting indications as to how the oxygen-supplying power of the soil 
may be expected to differ with different depths, states of packing, and 
moisture contents. The soil used was a loam, with considerable admix¬ 
ture of organic matter, and water enough to make it moist but not wet, 
such as is commonly used for potting greenhouse plants. The box used 
was of wood, 30 cm. wide, 40 cm. long, and 30 cm. deep, paraffined on 
the inside. 
The absorbing cylinder with its two lead tubes (1.5 m. long) connecting 
it to the rest of the system, was first placed upright in a glass jar con¬ 
taining enough mercury to submerge the absorbing portion. The gas 
stream flowed for 48 hours without any alteration in the color of the 
indicator solution, showing that there were no appreciable leaks in the 
system. The mercury was then withdrawn from the jar, exposing the 
external surface of the absorbing cylinder to the air of the room. The 
rate of oxygen absorption through the walls of the cylinder under these 
conditions was rapid enough so that the color change from color standard 
A to color standard B required 40 minutes. From this it appears that 
