its heat.
The experiments of Count Rumford, showing that heat is not propagated
downward in fluids, may be found at page 273. This is a principle too
important to be overlooked, especially in New England, where we need
every aid from Nature and Art, to contend successfully against the
brevity of the planting season. Soil saturated with cold water, cannot
be warmed by any amount of heat applied to the surface. Warm water is
lighter than cold water, and stays at the surface. In boiling water in a
kettle, we apply fire at the bottom, and no amount of heat at the
surface of the vessel would produce the desired effect. So rapid is the
passage of heat upward in water, that the hand may without injury be
held upon the bottom of a kettle of boiling water one minute after it
has been removed from the fire.

The following experiments and illustrations, from the _Horticulturist_
of Nov. 1856, beautifully illustrate this point:

"RATIONALE OF DRAINING LAND EXPLAINED.

"The reason why drained land gains heat, and water-logged land is
always cold, consists in the well-known fact that heat cannot be
transmitted _downwards_ through water. This may readily be seen by
the following experiments:

[Illustration: Fig. 97.]

"_Experiment No. 1._--A square box was made, of the form
represented by the annexed diagram, eighteen inches deep, eleven
inches wide at top, and six inches wide at bottom. It was filled
with peat, saturated with water to _c_, forming to that depth
(twelve and a half inches) a sort of artificial bog. The box was
then filled with water to _d_. A thermometer _a_, was plunged, so
that its bulb was within one inch and a half of the bottom. The
temperature of the whole mass of peat and water was found to be
39-1/2 deg. Fahr. A gallon of boiling water was then added; it raised
the surface of the water to _e_. In five minutes, the thermometer,
_a_, rose to 44 deg., owing to the conduction of heat by the
thermometer and its guard tube; at ten minutes from the
introduction of the hot water, the thermometer, _a_, rose to 46 deg.,
and it subsequently rose no higher. Another thermometer, _b_,
dipping under the surface of the water at _e_, was then introduced,
and the following are the indications of the two thermometers at
the respective intervals, reckoning from the time the hot water was
supplied: