268 REPORT ON THE DENEHOLE EXPLORATION
who live in a degree of greater security, the silo is more carefully
prepared, that it may contain that extra quantity required by possible
bad seasons. With the necessity of longer keeping, more suitable
spots of dry soil are selected than the damp field in which the corn
grew. Several families living near each other have their hoards close
together, partly for convenience in finding them, and partly to pre-
vent mutual robbery, a difficult feat where all eyes would be watching.
The size of these grain pits varies considerably. It is plain that
a very small hole comparatively would be sufficient for one family,
while many small, or one large pit, would be required by the man
whose needs were greater or whose aim was wealth.
Again, the burying just beneath the surface in the damp field is a
sadly hasty and risky mode of preservation, and one which it is
clear would be used only by the more wretched peoples. The
effort to keep out the damp is a constant trouble. It is accom-
plished in a measure by beating hard the sides of the pit if of clay;
by enveloping the deposit of corn with straw, reeds, fern, or bavins ;
by endeavouring to harden the sides by burning with fire, which could
only be done in very shallow pits; by erecting wooden walls and
floors; by covering with mastic or cement; and, finally, by building
the interior either with stones, finished masonry, or with bricks and
terra cotta. All these modes are practised in Europe, and the improve-
ment accords largely with the civilization of the country. The more
modern improvements proposed, though retaining the name "Ensi-
lage," are, however, such as to do away with the meaning of the word
in its simplicity. Where the more primitive form of hoarding grain
is in vogue, there are no stacks or ricks in the open, although the
use of barns and warehouses in towns is concurrent with pits in the
country. The reason is obvious. Ricks are exposed to many
dangers, and, from their size, are guides to the robber; but the pit
is secret, sure, and as free from vermin as from men.16
16 It is impossible that any but small and shallow depressions can be formed in light and alluvial
soils ; so shallow are these sometimes that the pit is only partially subterranean, while in the
opposite case the same effect may happen when the soil is hard rock, difficult to penetrate. Rocks
which are easily broken, as a rule, make good grain pits; but where the rock is very hard and
much severed by joints, the site is unsuitable from the impossibility of controlling the flow of
water. Limestones of various kinds are the most favourable of rocks to excavate, and those
which are easily broken, and porous, not crystallized in their texture, best of all. The chalky
limestones of the East and of Africa, soft tufa, and the chalk of Western Europe, are perhaps the
molt suitable. Unless below the permanent drainage line, no water stands in pits sunk in such
strata, however damp they may be ; and chalk will hold tenaciously an amount of water equal
to its own weight, without showing any moisture on its surface, and the formation of caves
with domed roofs is attended with greater permanence in limestones than in hardened sands.
As a rule too, loose sand, unless very consolidated, while making very good vertical sections
for the sides and shafts of caverns, is unsuited to horizontal spaces of considerable area and
especially to the rapid widening out which is seen in the caves in the chalk. The tertiary strata
which cover the Chalk of Western Europe, make very unstable roofs, and in the case of the