102 
Eeport of the Horticulturist of the 
THE INFLUENCE OF ATMOSPHEEIC PKESSUBE UPON 
PERCOLATION. 
In an experiment made in the year 1885, planned for the purpose of 
studying the influence of stirring the surface soil upon drainage, it 
was observed incidentally that the flow of water was much greater on 
some days than on others; and on comparing the amount of flow with 
the barometric changes, a rather marked correspondence could be 
traced. It appeared that increased drainage accompanied a fall in the 
barometer, and vice-versa. This observation suggested the following 
series of experiments in December, 1886. 
In the first trial, a five-inch " deep " flower pot was filled with clay 
soil, and after being submerged eighteen hours in a water tank to 
secure saturation, was placed upon a wire stand, within the receiver of 
an air-pump. When dripping from the soil had ceased, a part of the 
air was exhausted from the receiver, when the flow immediately recom- I 
menced, and when this had ceased, another part of the air was exhausted, 
when drainage again commenced. A third trial was accompanied 
with a like result. 
The next trial was designed to show whether the dripping was due 
to a certain degree of exhaustion of the air, or was only the result of a 
change in atmospheric pressure. The air was admitted to the receiver, 
and a column of mercury connected with the latter so that the degree 
of exhaustion could be measured. The air was then rarefied until the 
water began to drip freely from the pot of soil. When this drainage had 
ceased, the air was let into the receiver once more, and after an inter- 
val, exhausted to the same degree as before. The dripping again com- 
menced, and the test was repeated with the same result, indicating 
that the percolation was due to a change from a heavier to a lighter 
air pressure, rather than to the reduction of the air pressure to a given 
limit. 
On considering the cause for these phenomena, the one first sug- 
gested was that the air contained in the interstices of the soil, and in 
the water with which the soil was saturated, would naturally tend to 
expand on the removal of a part of the atmospheric weight, and would 
thus exert a corresponding pressure upon the water held in the soil 
by capillary attraction. This would in effect decrease the force of 
capillarity, and a portion of the water would be set free. 
As a test for the above hypothesis a third trial was made. A bottle 
of about 400 c. c. capacity was taken, and through its cork, which was 
of rubber, a slender glass tube was inserted to such a distance as to 
nearly reach the bottom of the bottle. The cork was then removed, 
