rally, and I thought that heat in liquids
could be propagated only upwards.

MRS. B.

You do not take notice that the heat is imparted from one liquid to the
other, through the medium of the phial itself, the external surface of
which receives the heat from the water in the glass, whilst its internal
surface transmits it to the liquid it contains. Now take the phial out
of the hot water, and observe the effect of its cooling.

EMILY.

The currents are reversed; the external current now descends, and the
internal one rises. --I guess the reason of this change:-- the phial
being in contact with cold air instead of hot water, the external
particles are cooled instead of being heated; they therefore descend and
force up the central particles, which, being warmer, are consequently
lighter.

MRS. B.

It is just so. Count Rumford hence infers that no alteration of
temperature can take place in a fluid, without an internal motion of its
particles, and as this motion is produced only by the comparative levity
of the heated particles, heat cannot be propagated downwards.

But though I believe that Count Rumford's theory as to heat being
incapable of pervading fluids is not strictly correct, yet there is, no
doubt, much truth in his observation, that the communication is
materially promoted by a motion of the parts; and this accounts for the
cold that is found to prevail at the bottom of the lakes in Switzerland,
which are fed by rivers issuing from the snowy Alps. The water of these
rivers being colder, and therefore more dense than that of the lakes,
subsides to the bottom, where it cannot be affected by the warmer
temperature of the surface; the motion of the waves may communicate this
temperature to some little depth, but it can descend no further than the
agitation extends.

EMILY.

But when the atmosphere is colder than the lake, the colder surface of
the water will descend, for the very reason that the warmer will not.

MRS. B.

Certainly: and it is on this account that neither a lake, nor any body
of water whatever, can be frozen until every particle of the water has
risen to the surface to give off its caloric to the colder atmosphere;
therefore the deeper a body of water is, the longer will be the time it
requires to be frozen.

EMILY.

But if the temperature of the whole body of water be brought down to the
freezing point, why is only the surface frozen?

MRS. B.

The temperature of the whole body