RESEARCH METHODS IN STUDY OF FOREST ENVIRONMENT. 91 
It is desired to present another set of data obtained by Bates (105) 
to illustrate the need of establishing the wilting coefficient for the 
particular species in which one may be interested and, therefore, of 
establishing a specific relationship between the wilting coefficients and 
the capillary moistures of the same soils. This presentation also 
assists in showing, what has already been mentioned, that a measure 
of the capillarity or other moisture relation of the soil has an indirect 
value in permitting comparisons of the species under a variety of 
conditions. 
The tests as represented in Table 3 were performed on five distinct 
kinds of soil, varying as to origin (hence, chemically) and also con- 
siderably as to composition and water-holding capacity. With the 
exception of the prairie soil, which contained only 1 per cent of coarse 
sand and no gravel at all, these soils were prepared b}' passing 
through a sieve with quarter-inch meshes. 
The wilting coefficient determinations, moreover, were made with- 
out the Hse ^f paraffin. As the test was designed particularly to com- 
pare the four species which were grown in each soil, and it had be- 
come apparent that the rooting habit of each had a good deal of bear- 
ing on the stage in soil drying at which it succumbed, the effort was 
made to keep the upper layer of the soil well supplied with moisture 
by daily watering. As a result, the common drying of the steam just 
at the ground line was not appreciably in evidence and, indeed, so 
general was the drying that the determination of the end point was 
exceedingly difficult. It was based almost wholly on the flaccidity of 
the leaves. Whether because of this protection afforded the stems by 
surface watering, or because of the comparative shade in which the 
end points were approached, it is noteworthy that the ratios of wilting 
coefficients to capillarities are much lower, except for the heaviest 
clay, than in the results obtained under different conditions and 
already described. 
Another noteworthy feature of this test is that the seedlings were 
produced in each soil with the moisture brought daily to the moisture 
equivalent, so that the availability was, as nearly as could then be 
calculated, the same in all cases. When drying began, each soil 
was brought by easy stages to two- thirds of the moisture equivalent, 
and finallv to one-third. The seedlings attained an age about 6 
months before the test was completed. 
