116 
Research Bulletin No. 3 
than does that in humid soils. This may be due simply to the 
more complete penetration of the soil mass by the roots in the 
case of the semiarid and arid soils, and this in turn be due to 
the flocculation of the clay in the latter. It is well known that 
the arid and semiarid soils are strongly calcareous compared with 
humid soils, a good illustration being furnished by the loess 
soils of the Transition Kegion. 1 
Briggs and Shantz 2 consider that a statement of the free 
water may be very misleading in comparisons of the available 
moisture in soils. The most of the experiments described above 
furnish little or no evidence as to the correctness of these views, 
the cylinders in most cases not having been opened until long 
after growth of the contained plants had ceased. The few excep- 
tions include cylinders I and IV with wheat in 1909 and cylin- 
ders I and II with beans in 1910. The former ripened seed nor- 
mally and as soon as the plants appeared quite ripe the cylin- 
ders were opened; in both the moisture of the subsoil had been 
quite uniformly reduced to more than 2 per cent below the wilt- 
ing coefficient. Much the same was true for the two cylinders 
with beans in 1910. 
Some evidence is furnished by the numerous cylinders in 
which when the plants died the moisture content of the upper 
portion of the subsoil, where roots were the most numerous, was 
1 Alway, F. J. Composition of the Loess Soils of the Transition Region, 
Eighth International Congress of Applied Chemistry, 1912, Report of Sec- 
tion VII, p. 11. 
2 They have incorrectly stated the writer's earlier views as to the 
basis for comparing the available moisture in soils. Tbey state: "In the 
absence of a more definite relationship between the wilting coefficient 
and the hygroscopic coefficient, Alway has advocated deducting the hygro- 
scopic coefficient from the field soil-moisture determinations as a basis for 
comparing the available moisture in soils." (Bui. 230, Bureau of Plant 
Industry, p. 66.) On the contrary the writer had stated that the "free water," 
obtained by deducting the hygroscopic coefficient from the total water 
(= field moisture determination), "does not indicate the amount of water 
that is available in such form that the plant can continue a normal 
growth. * * * It is probable that the normal growth ceases when 
the percentage of free moisture falls to from 4 to 8 per cent according 
to the soil." (Bui. 130, Bureau of Plant Industry, p. 38.) "A comparison 
of the fields on the basis of the amount of free water in the first six feet 
assumes that a definite value is to be attached to a certain percentage of 
free water, independent of the kind of soil * * * in which it occurs. 
It is evident from Hilgard's work, however, that a higher percentage of free 
moisture is required on clay than on sandy soils in order that plant 
growth may continue. * * * "It has been assumed that the lower limit 
of free moisture in the boulder clay is 4.5 and in the lacustral clay 6.0 
per cent. The free water less this 4.5 or 6.0 per cent is for convenience 
at present designated the 'x water.'" (Jour. Agr. Sc., vol. 2, pp. 338-340.) 
In the earlier publication this "x water" had been referred to as "water 
probably available for normal plant growth." (Bui. 130, Bureau of Plant 
Industry, p. 37.) 
