16» «B. W. Hilgard—Silt Analyses of Soils and Subsoils. 
of aggregates, the molecular properties of a clay or subsoil 
will never correspond exactly to the mean resulting from a 
mere consideration of the molecular coefficients of each one, 
multiplied into its percentage. How far this difference extends, 
is a question involving a previous investigation of those coeffi- 
cients. 
Among the latter, that of absorption of aqueous vapor is of 
no mean importance, since it determines, in a great measure, the 
resistance of the soil to drought. As Piao stated,* I find 
that at temperatures between +7° and +21°, the amount of 
aqueous vapor absorbed by a thin layer of a clay, or soil not 
unusually rick an humus, in a saturated atmosphere, is sensibly 
constant; the variations being within the limits of errors of 
observation, and indiscriminately either way. A glance at the 
data given in the table, opposite the heading “hygroscopic 
rend shows that while in general, as is well known, clay 
soils are more absorbent than sandy ones, yet there exists no 
seat inulin) relation between the amount of clay present 
and the absorbing power. Not only is that of the typical 
white pipe-clay (No. 238) scarcely greater than that of an ordi- 
nary loam subsoil (Nos. 897 and 219), but it is not half as 
— that of the clay soil 246 (with 40 per cent of “ clay”), 
ried ay ps cent of clay). Pipers 230, with 25°5 per cent 
over, the finer silts themselves possess a considerable absorbing 
ower. Again, the presence of hydrated ferric oxide materially 
influences this power; so much so that no general conclusion 
concerning the hygroscopic effect of “clay” can be reached, 
unless the amount of iron present be taken into account. 
am esa as yet, to furnish this datum for all the soils on the 
table, save as regards, for most of them, the percentage in the 
Sane: saubedtiass That the hy dro-ferric oxide accumulates 
mainly i in the “clay” obtained in silt analysis, I have already 
stated ; and hence the percentages given at the bottom of the 
table may measurably serve to form an estimate of its influ- — 
ence on the hygroscopic properties. In some cases, however, the — 
ferric oxide obtained in aby was almost altogether present — 
- the shape of bog-ore grains; these are placed in parentheses, 
t being obvious that the “ white” soils, to which these deter- 
sein belong, do not co exceeding 05 per cent of : 
the oxide in the finely-divided, oh aeons effective con- 
* Proc. Am, Assoc. Adv. Science, Dubuque meeting, 1872, p. 73. 
