ree.
The upper leaves radiate freely towards the stars and the cold
inter-stellar spaces, while the lower ones radiate to those above and
receive less heat in return; consequently, during the absence of the
sun, each tree cools gradually downward from top to bottom. Hence we
must take into account not merely the area of leaf-surface actually
exposed to the sky, but, to a greater or less extent, the surface of
every leaf in the whole tree or the whole wood. This is evidently a
point in which the action of the forest may be expected to differ from
that of the meadow or naked earth; for though, of course, inferior
strata tend to a certain extent to follow somewhat the same course as
the mass of inferior leaves, they do so to a less degree--conduction,
and the conduction of a very slow conductor, being substituted for
radiation.

We come next, however, to a second point of difference. In the case of
the meadow, the chilled air continues to lie upon the surface, the
grass, as Humboldt says, remaining all night submerged in the stratum of
lowest temperature; while in the case of trees, the coldest air is
continually passing down to the space underneath the boughs, or what we
may perhaps term the crypt of the forest. Here it is that the
consideration of any piece of woodland conceived as a solid comes
naturally in; for this solid contains a portion of the atmosphere,
partially cut off from the rest, more or less excluded from the
influence of wind, and lying upon a soil that is screened all day from
isolation by the impending mass of foliage. In this way (and chiefly, I
think, from the exclusion of winds), we have underneath the radiating
leaf-surface a stratum of comparatively stagnant air, protected from
many sudden variations of temperature, and tending only slowly to bring
itself into equilibrium with the more general changes that take place in
the free atmosphere.

Over and above what has been mentioned, thermal effects have been
attributed to the vital activity of the leaves in the transudation of
water, and even to the respiration and circulation of living wood. The
whole actual amount of thermal influence, however, is so small that I
may rest satisfied with mere mention. If these actions have any effect
at all, it must be practically insensible; and the others that I have
already stated are not only sufficient validly to account for all the
observed differences, but would lead naturally to the expectation of
differe