4388 #. W. Hilgard—Soil Analyses and their Utility. 
their having obviously been taken at an improper lone 
e. g., near a foot-path, by the side of a fence, on a partially 
denuded hillside or ravine, in the bed of a run, at the foot 
of a tree, ete. 
e question of depth must, in my view, be left to be deter- 
mined by the circumstances of each case, except in so far as 
the extreme depth to which tillage may cause the roots of crops 
to reach, must be within the limits of the samples taken. Of 
these, one should ordinarily represent what, under the usual 
practice of tillage, becomes the arable soil; another, the sub- 
soil not usually broken into; a third will in most cases be use- 
to show what materials would be reached were the land to 
From the fact that the atmospheric surface water must, in 
its course, inevitably have a tendency to bring about such 
agreem 
Such I find to be very decidedly the case; so much so, that 
habitually look to the former as the most reliable index, of & 
soil’s distinctive character. To this there can be no legitimey 
objection, when, as in all the upland soils now under consi 
eration, the surface soil is directly derived from the su me 
and its depth is less than thorough culture would give to 3° 
arable soil. 
