48 
Journal of Agricultural Research 
Vol. IX, No. a 
been inverted. Any soil adhering to the tray was transferred to the 
cylinder. The cover was then put on, and both the opening between this 
and the cylinder wall and the holes in the bottom were sealed with paraffin. 
The cylinders were weighed and transferred to a basement room in which 
the temperature was practically constant. There they were placed in a 
large box and covered to a depth of 8 inches with moist soil to prevent 
any sudden or extreme fluctuations in temperature and also any drying 
of the soil in the cylinders, if by any chance openings should develop, as, 
for example, by the cracking of the paraffin seal. 
The cylinders were left undisturbed for io weeks, and most of them for 
4 to 6 weeks longer. As soon as each was removed from the box, it was 
weighed, in order to determine whether there had been any change in 
weight during its stay in the box. In the case of none did the scales, 
sensitive to half an ounce, show any gain or loss. As the cylinders were 
removed from the box, moist soil was put in their place. All had been 
placed in the box on 
the same day, but in 
some instances not all 
of those removed at 
the same time could 
be opened on the same 
day. In such cases 
the unopened cylin¬ 
ders were left in the 
constant temperature 
room until such time 
as they were opened. 
O M L /f J / hgeocb A Previous experiments 
Fig. a.—Diagram showing relationship between the final water content with Soil T had shown 
and both the hygroscopic coefficient and the moisture equivalent. . 4 , 
that after two months 
standing the distribution of water was but little affected by leaving them 
longer undisturbed (3, p. 270). 
The distribution of the moisture at the end of this period for the 12 
loams is given in Table XIV. That in the sand Q is treated separately. 
It will be seen that, except with B and G, the lower foot section had gained 
but little in moisture during the 69 to no days. Almost all the soils 
showed a very slight increase, even as far as the lowest portion of the 
cylinder; but only with B and G was this really distinct as deep as the 
twenty-first inch. The maximum depth affected in any case may be 
considered as the twenty-fourth inch. The final moisture content of the 
surface 3-inch section lay between 1.7 and 2.4 times the hygroscopic 
coefficient, the maximum in all cases being found in the surface inch. 
This relationship is shown in Table XV. Where the initial ratio was 1 
or 1.1, the final ratio in the surface 3-inch section varied from 1.7 to 2.3, 
that in the lowest 3-inch section remaining as low as 1.1 to 1.2. Where 
t 
/ 
\ 
y 
ftfS- 
*4 
\ 
— 
r .. 
A 
'A 
fATKt 
,4 
¥ 
J 
Y/ 
¥ 
CC 
err/ 
:/£W7 
4 
P 
/S/?Og 
