48 
Research Bulletin No. 3 
As only three different soils were used in the 22 cylinders 
and as these had been thoroly mixed, the determination of the 
hygroscopic coefficients was a simple matter. The most inter- 
est attaches to the subsoil and of this a very large number of de- 
terminations were made, the average being 5.6. The values for 
the first 6-inch and the second 6-inch samples are 9.3 and 11.0 
respectively. 
FILLING THE CYLINDERS. 
In all the 22 cylinders when filled there were only three differ- 
ent soils, the soil below the first foot being uniform, except for 
a layer of coarse sand weighing 400 grams and about three- 
fourths of an inch in thickness, which was placed in the bottom 
of each cylinder in order to facilitate drainage. 
The method of applying water used in 1908 and 1909 was 
not employed in this experiment for the reason that it left in 
the soil after seepage had ceased so very much more water than 
the same soil under dry-land field conditions would carry. Ten 
of the cylinders were filled with air-dry soil, after Avhich enough 
water was added to give the desired total amount. The other 
12 were filled with moist subsoil, water being added, after the 
filling, only to moisten the surface foot. Germinated seeds were 
planted in all but 2 of the cylinders on the same day. these 2 
being left implanted. 
Ten, I to X, were filled with dry soil in much the same way 
as in 1909, all except the surface inch being added with tamping. 
The weight of each of the three kinds of soil used for each cylin- 
der, as well as the percentage of moisture in each, was determined 
and from these data there was calculated the amount of water 
contained in the air-dry soil of each cylinder; then enough water 
was added to each to make the total amount present equal to 25 
per cent of the dry weight of the surface foot and 18 per cent of 
the dry weight of the subsoil. During the adding of the water, 
which lasted over several days, the tops of the cylinders were 
kept covered to prevent evaporation and the glass plugs were 
removed from the one-hole rubber stoppers. As soon as all the 
added water had soaked into the soil, the glass plugs were re- 
turned, the cylinders were weighed and placed in the pit in the 
greenhouse, germinated seeds were planted one inch deer) in the 
moist soil, and the final inch of dry surface soil added. No seep- 
age had occurred in the case of any of these before the plugs were 
inserted and doubtless none would have occurred if the plugs had 
been left out. 
For the 12 cylinders in which the subsoil was moistened be- 
fore being placed in the cylinder, eisrht lots of prepared subsoil 
were used. Each lot of moist subsoil was prepared by placing a 
