additional Heat, rife up to the upper furface of the
fluid water, and give off their fenfible Heat to the
under furface of the ice, never return to the
bottom, this communication of the Heat which
exhales from the earth produces very little motion
in the mafs of the water ; and this circumftance is r
no doubt, very favourable to the preservation of the
Heat of the water.

When a ftrong wind prevails, and the furface of
the water is much agitated, ice is not formed,
even though the whole mafs of water mould, by a
long continuance of cold weather, have been pre-
vioully cooled down to- that point to which it is
neceiiary that it fliould be brought, in order that
its internal motions may ccafe, and it may be dif-
pofcd to congeal ; for though the particles at and
near the furface may no longer have any tendency
to defcend, on being farther cooled, yet, as they
have fo confiderable a quantity of fenfible Heat
(eight or ten degrees) to difpofe of, after their
condenfation with cold ceafes, and as the agitation
into which the water is thrown by the wind does

not



in Fluids. 595

not permit any particle to remain long enough in
contact with the cold air to give off all its Heat at
once, there is a continual fucceffion of frefh parti
cles at the furface, all of which give off Heat
to the air ; but none of them have time to be
cooled fufficiently to be difpofed to form ice. The
water will lofe a vail quantity of Heat, and as foon
as the wind ceafes, if the cold ihould continue, ice
will be formed very rapidly,

But it is not merely the agitation of the water
which renders the communication of the Heat very
rapid, the agitation of the wind alfo tends to pro
duce the fame effect,

On the return of fpring, the mow melting
before the fun as he advances and his rays become
more powerful, all the Heat which the earth
exhales is employed in diffolving the ice at its
under furface, while the fun on the other fide
acts ftill more powerfully to produce tjie fame
effect.

Though ice is transparent, yet it is not perfectly
fo, and as the light which is flopped in its paffage
through it cannot fail to generate Heat when and
inhere it is flopped, or abforbcd, it is by no means
furprifmg that fnow fhould be found to melt when
expofed in the fun's rays, even when the tempera
ture of the air in the fhade is confiderably below
the point of freezing. Snow expofed to the fun
melts long before the even furface of ice begins to
be fenfibly foftened by its beams, and it is not till
fome time after ail the hills are bare that the icq
on the lakes and rivers breaks up.

The



2 9 6 Of the Propagation of Heat

The rays which penetrate a bank of fnow being
often reflected and refracted, defcend deep into it,
and the Heat is depofited in a place where it is not
expofed to be carried off by the cold air of the at-
mofphere ; but the rays which fall upon the hori
zontal and fmooth furface of ice, are moilly re
flected upwards into the atmofphere ; and if any
part of them are flopped at the furface of the ice,
the Heat generated by them there is inftantane-
oufly carried off by the cold air, and a particle of
water is no fooner made fluid than it is again
frozen.

Hence we fee that the fnow which in cold
countries covers the ice that is formed on the fur-
face of frefli water, not only prevents the Heat of
the water from being carried off by the air, during
the winter, but alfo affifts very powerfully in thaw
ing the ice early in the fpring,

Let us now fee what the confequences would
have been had the condenfation of water with cold
followed the lav/ which obtains in regard to all
other Fluids.

As the internal motion of the water could not
have failed to continue as long as its fpecific gravity
continued to be increafed by parting with Heat,
ice would not have begun to be formed till the
whole mafs of water had arrived at the tempera
ture of 32 of Fahrenheit's thermometer.

To fee what an enormous quantity of Heat would
be* loft when the water is deep in confequence of
its whole mafs being cooled in this manner, we
have only to compute how much ice this Heat

would



in Fluids. 297

would melt, or how much water it would heat
from the point of freezing to that of boiling.

It has been mown by experiment, that any given
quantity of ice requires as much Heat to melt it
as an equal quantity of fluid water lofes in cool
ing 140 degrees, confequently the quantity of ice
which might be melted by the Heat given off by
any given quantity of water in cooling any given
number of degrees, is to the given quantity of
water, as the number of degrees which it is cooled,
to 140 degrees.

Hence it follows that when the temperature of
the water is 8 degrees above the freezing point, it
gives off in cooling down to that temperature as
much Heat as would melt r -~ or -fj of its weight
of ice ; the water, therefore, which is cooled from
the temperature of 40 to that of 32, if it be
3 5 feet deep, will give off as much Heat in being
fo cooled as would melt a covering of ice z
feet thick.

But even this is not all, for as the particles of
water on being cooled at the furface would in con-
fequence of the increafe of their fpecific gravity,
on parting with a portion of their Heat, immedi
ately defcend to the bottom, the greatefl part of
the Heat accumulated during the fummer in the
earth on which the water repofes would be carried
off and loft, before the water began to freeze ; and
when ice was once formed, its thicknefs would
increafe with great rapidity, and would continue
increaiing during the whole winter ; and it feems
very probable that in climates which are now tem
perate,



298 Of the Propagation of Heat

perate, the water in the large lakes would be frozen
to fuch a depth in the courfe of a fevere winter
that the Heat of the cnfuing fummer would not be
fufficient to thaw them ; and fhould this once hap*
pen, the following winter could hardly fail to
change the whole mafs of its waters to one folid
body of ice, which never more could recover its
liquid form, but muft remain immoveable till the
end of time.

In the month of February, after a froft which
had lafted a month, the temperature of the air being
38, M. DE SAUSSURE found the temperature of
the water of the Lake of Geneva, at the furfaco, at
41, and at the depth of 1000 feet, at 40.
Had the frofl continued but a little longer, ice
would have been formed ; but had the conftitu-
tion of water been fuch that the whole mafs of
that Fluid in the Lake muft have been cooled down
to the temperature of 35 before ice could have
been formed, this event could not have happened
till the water had given off as much Heat as would
be fufficient to melt a covering of ice above 57 feet
thick !

This quantity of Heat would be fufficient to heat,
to the point of boiling, a quantity of ice-cold water
as large as the Lake, and 49 feet deep,

We cannot Efficiently admire the fimplicity of
the contrivance by which all this Heat is faved. It
well deferves to be compared with that by which
the feafons are produced ; and I muft think that
every candid inquirer, who will begin by diverting
iiimfelf of all unreafonable prejudices, will agree

with



in Fluids. 299

with me in attributing them both TO THE SAME

AUTHOR*

When we trace flill farther the afloniftiing effects
which are produced in the world by the operations
of that iimple law which has been found to obtain
in the condenfation of water on its being deprived
of Heat, we fliall find more and more reafon to
admire the wifdom of the contrivance.

That high latitudes might be habitable, it was
neceffary that vegetables mould be protected from
the effects of the chilling frofts of a long and
fevere winter : but if it be true that watery liquids
do not part with their Heat but in confequence of
their internal motions ; and if thefe motions are
occaiioned merely by the change produced in the
fpecific gravity of thofe particles of the liquid which
receive Heat, or which part with it, who does not
fee how very powerfully the fudden diminution
and final ceflation of the condenfation of water in
cooling, as foon as its temperature approaches to
the freezing point, operates to prevent the fap in
vegetables from being frozen ?

But if, for the purpofes of life and vegetation,
it be necefTary that the ground, the rivers, the
lakes, and the trees be defended from the cold
winds from the poles, it may be afked how this
inundation of cold air is to be warmed ? I an-
fwer by the waters of the ocean, which there is
the great eft reafon to think were not only defigned
principally for that ufe, but particularly prepared
for it.

Sea



300 Of the Propagation of Heat

Sea water contains a large proportion of fait
in folution, and we have feen that the condenfa-
tion of a faline folution, on its being cooled, fol
lows a law which is extremely different from that
cbferved in regard to pure water ; and which (as
may eafily be fliown) renders it peculiarly well
adapted for communicating Heat to the cold winds
which blow over its furface.

As fea water continues to be condenfed as it
goes on to cool, even after it has palled the point
at which frefh water freezes, the particles at the
furface, inftead of remaining there, after the mafs
of the water had been cooled to about 40, and pre
venting the other warmer particles below from com
ing in their turns andgiving off their Heat to the cold
air, (as we- have feen always happens when frefh
or pure water is fo cooled) thefe cooled particles
of fait 'water defcend as foon as they have parted
with their Heat, and in moving downward force
other warmer particles to move upwards ; and
in confequence of this continual fucceilion of
warm particles which come to the furface of the
fea, a vaft deal of Heat is communicated to the
air ; incomparably more than could poffibly be
communicated to it by an equal quantity of frefh
water at the fame temperature, as will appear by
the following computation.

Without taking into the account that very great
advantage which fea water poiTeffes over frefh
water, confidered as an equalizer of the tempera
ture of the atmofphere, which arifes from the com
parative lownefs of the point of its congelation ; fup-

pofing



in Fluids. 301

poflng even fea water to freeze at as high a tem
perature as frefh water, namely, at 32 ; and fup-
pofmg (what is flrictly true) that as foon as either
fea water or frefh water is frozen at its furface,
and this ice covered with fnow, that the commu
nication of Heat from the water to the atmofphere
ceafes almoft entirely; we wi3 endeavour to deter
mine how much more Heat would, even on this
fuppofition, be communicated to the air by fait
water than by frelh water, after both have arrived
at the temperature of 40.

When frelh water, in cooling, has arrived at this
temperature, it ceafes to be farther condenfed with
cold, and its internal motions (which, as we have
already more than once obferved,are caufedyc/^/yby
the changes produced in the fpecific gravity of its par
ticles) ceafe of courfe, and ice immediately begins
to be formed on its furface ; but as the condenfation
of fait water goes on as its Heat goes on to be di-
minifhed, its internal motions will continue ; and
it is evidently impoffible for ice to be formed at
its furface till the whole mafs of the water has be
come ice-cold, or till its temperature is brought
down to 32. It would therefore give off a quan
tity of Heat equal to 8 degrees, at leaft, of Fahren
heit's thermometer, more than the frefh water would
part with before ice could be formed on its fur-
face.

To be able to form an idea of this enormous
quantity of Heat, we have only to recoiled: what
has already been {aid, and we fiiall find reafon to
conclude that it would be fufficient to melt a cover
ing of ice equal in thicknefs to -g- of the depth of
VOL. II. Q^q the



30-2 Of the Propagation of Heat

the fca. It would therefore be fufficient in that
part of the North Sea (lat. 67) where Lord Mul-
grave founded at the depth of 4680 feet, to melt a
cuke of ice 265 fe'et thick !

But the Heat evolved in the formation of each
fuperficial foot of ice would be fufficient to raife the
temperature of a fftatum of incumbent air 2220
times as thick as the ice, (confequently in the cafe
in queition 265x2220 feet, or 869 miles thick)
2" 8 degrees, or from the temperature of freezing
water, to that of 50 of Fahrenheit's thermometer,
or to the mean annual temperature of the northern
parts of Germany I

The Heat given off to the air by each fuperficial
foot of water in cooling one degree is fufficient to
heat an incumbent itratum of air 44 times as thick
as the depth of the water, 10 degrees. Hence we
fee how very powerfully the water of the ocean,
which is never frozen over, except in very high
latitudes, mu'ft contribute to Warm the cold air
which flows in from the polar regions.

But the ocean is not more ufcful in moderating
the extreme cold of the polar regions, than it is in
tempering the exceffive heats of the torrid zone ;
and what is very remarkable, the fitnefs of the fea
water to fcrvc this laft important purpofe is owing
ta the very fame caufe which renders it fo pecu
liarly well adapted for communicating Heat to the
cold atmofphere in high latitudes, namely, to the
fait which it holds in fohftion.

As the condenfation of fait water with cold con
tinues to go on even long after it has been cooled to
the temperature at which frelh water freezes, thcfe

particles



in Fluids, 303

particles at the farface which are cooled by an
immediate contact with the cold winds mult de-
fcend, and take their places at the bottom of the
fea, where they muft remain, till, by acquiring an
additional quantity of Heat, their fpecific gravity
is again diminished, But this Heat they ne-vcr can
regain in the polar regions^ for innumerable expe
riments have proved beyond all poffibility of doubt
that there is no principle of Heat in the interior parts
of the globe, which by exhaling through the bottom
of the ocean, could communicate Heat to the water
which refts upon it.

It has been found that the temperature of the
earth at great depth under the furface is different
in different latitudes, and there is no doubt but
this is alfo the cafe with refnecl to the temperature
at the bottom of the fea, in as far as it is not influ
enced by the currents which flow over it ; and this
proves to a demonflratian that the Heat which we
find to exift, without any fenfible change during
fumrner and winter, at great depths, is owing to the
Action of the fun, and not to central fires, as fo.me
have too haiKly concluded,

But if the water of the ocean, which, on being
Deprived of a great part of its Heat by cold winds,
defcends to the bottom of the fea, cannot be.
warmed where it defcends, as its fpecific gravity is
greater than that of water at the fame depth in
warmer latitudes, it will immediately begin to
fpread on the bottom of the fea, and to flow to
wards the equator, and this muft necefTarily pro
duce a current at the furface in an oppolite direc
tion j



304 Of the Propagation of Heat

tion ; and there are the moft indubitable proofs
of the exiftence of both thefe currents.

The proof of the exiftence of one of them
would indeed have been quite fufficient to have
proved the exiftence of both, for one of them
could not poffibly exift without the other ; but
there are feveral direct proofs of the exiftence of
each of them.

What has been called the gulph ftream, in the
Atlantic Ocean, is no other than one of thefe cur
rents that at the furface which moves from the
equator towards the north pole, modified by the
trade winds, and by the form of the continent of
North America ; and the progrefs of the lower
current may be confidered as proved directly by
the. cold which has been found to exift in the fea
at great depths in warm latitudes ; a degree of
temperature much below the mean annual tempe*
rature of the earth in the latitudes where it has
been found, and which of courfe muil have been
brought from colder latitudes.

The mean annual temperature in the latitude of
67 has been determined by Mr. KIRWAN, in his
excellent treatife on the temperature of different
latitudes, to be 39 ; but Lord Mulgrave found on
the 2oth of June, when the temperature of the
air was 48^, that the temperature of the fea at the
depth of 4680 feet was 6 degrees below freezing,
or 26 of Fahrenheit's thermometer.

On the 3ift of Auguft, in the latitude of 69*,
where the annual temperature is about 38, the
temperature of the fea at the depth of 4038 feet

was



in Fluids. 305

was 32 ; the temperature of the atmofphere (and
probably that of the water at the furface of the fca)
being at the fame time at 59 \ .

But a ftill more ftriking, and I might, I believe,
fay an incontrovertible proof of the exiflence of
currents of cold water at the bottom of the fea,
fetting from the poles towards the equator, is the
very remarkable difference that has been found
to fubfift between the temperature of the fea at
the furface and at great depth, at the tropic,
though the temperature of the atmofphere there is
fo conflant that the greateft change produced in
it by the feafons feklom amounts to more than five
or fix degrees ; yet the difference-between the Heat
of the water at the furface of the fea, and that at
the depth of 3600 feet, has b^en found to amount
to no lefs than 31 degrees ; the temperature above
or at the furface being 84, and at the given depth
below no more than 53.*

It appears to me to be extremely difficult, if
not quite impofiible, to account for this degree of
cold at the bottom of the fea in the torrid zone,
on any other fuppofition than that of cold currents
from the poles ; and the utility of thefe currents
in tempering the exceflive heats of thofe climates
is too evident to require any illuftration.

Thefe currents are produced, as we have already
feen, in confequence of the difference in the fpeci-
fic gravity of the fea water at different tempera
tures ; their velocities muft therefore be in pro
portion

* PhiL Tranfa&ions, 1752.



3 e>6 Of the Propagation of Heat

portion to the change produced in the fpecific gra*
vity of water by any given change of temperature ;
and hence we fee how much greater they mud be
in fait water than they could poflibly have been
Jiad the ocean been compofcd of frefti water.

It is not a little remarkable that the water of
all great lakes is frefli, a.nd nearly fo in a.11 inland
feas (like the Baltic) in cold climates, and which
communicate with the ocean by narrow channels*
We {hall find reafon to conclude that this did not
happen without defign, when we coniider what
confequences would probably enfue ihould the
waters of a large lake in an inland iituation, in a
cold country, (fuch as the lake Superior, for inr
fiance, in North America) become as fait as the
fea.

Though the cold winds which blow over the lake
in the beginning of winter would be more warmed,
and the temperature of the air on the iide of the lake
oppofite to the quarter from whence thefe winds
arrive, would be. rendered fomewhat inilder thai*
it now is ; yet, as the water of the lake would give
off an immenfe quantity of Heat before a covering
of ice could be formed on its furface for its pro -
tection, it \vould, on the return of fpring, be found
to be extremely cold ; and as it would require a long
time to regain from the influence of the returning
fun the enormous quantity of Heat loft during the
winter, it would remain very cold during the
fpring, and probably during the greateft part of
the fummer ; and this could not fail to chill the
atmofphere, and check vegatation in the furround-



in Fluids i 307

lilg country to a very confiderable difiarice'. And
though a large lake of fait water in a cold country
would tend to render the winter fomewhat milder
on one Tide of it, namely, on the fide oppofite to
the quarter from \vhence the cold winds carne ;
yet this advantage Would hot only be confined to a
fmall tract cf country, but would not any where be
very important, and would by no means counter
balance the exteniive and fatal confequences which
would be produced in fummer by fo large a col
lection of very cold water <

When the winter is once fairly fet in, ~ \vlien the
earth is well covered with fnow, and the rivers and
lakes with ice, and more efpecially when the ice
as well as the land is covered with that warm
winter garment, a few degrees more of cold in the
air cannot produce any lafting bad confequences.
It may oblige the inhabitants to ufe additional pre
cautions to guard themfelves, their domeftic ani
mals, and their provifions, from the uncommon fe-
verity of the weather ; but it can have very little
influence in the temperature of the enfuing fum
mer ; and even it is probable, if it influences it at
all, that it fends rather to make it warmer than
colder. Lakes of fait water could therefore be of
no real ufe in winter in cold countries, and in fum
mer they coulci not fail to be very hurtful ; while
frefh lakes, as they are frozen over almoft as foon
as the winter fets in, and long before the whole
mafs of their water is cooled down to the tempera
ture of freezing, they preferve the greater part of

their



308 Of the Propagation of Heat, &c.

their Heat through the winter, and if they are of
no ufe during the cold feafon, they probably do
little or no harm in fummer.

But I mufl take care not to tire my reader by
purfuing thefe fpeculations too far. If I have per-
fitted in them, if I have dwelt on them with pe
culiar fatisfaction and complacency, it is becaufe I
think them uncommonly interefting, and alfo
becaufe I conceived that they might be of real ufe
in this age of refinement and Jcefticifm.

If, among barbarous nations, the fear of a
God, and the practice of religious duties, tend to
foften favage difpoiitions, and to prepare the mind
for all thofe fweet enjoyments which refult from
peace, order, induftry, and friendly intercourfe,
a belief in the exiftence of a Supreme Intelligence^ who
rules and governs the univerfe with wifdom and
goodnefs, is not lefs eflfential to the happinefs of
thofe who, by cultivating their mental powers,

HAVE LEARNED TO KNOW HOW LITTLE CAN BE
J&NOWN.

PESgRIPTION






of In



JSofton Published h\ n.\\rrl -V'.' .iff (



( 39 )



DESCRIPTION OF THE PLATES.

PLATE I.

HTHIS Plate reprefents the cylindrical Paffage
Thermometer ufed in the Experiments, on the
conducting power of liquids with regard to Heat.

Fig. i. a, b, is a fedion of the brafs tube in
which the Thermometer c 9 with an oblong copper
bulb, is placed.

*>,/, is the glafs tube of the Thermometer, which,
for want of room in the Plate, is reprefented as
broken off at /.

, is a ftopple of cork by which the end of the
brafs tube, #, , is clofed ; and

h, is a circular difk of the fame fubftance.
The fpace in the brafs tube below this difk , fur-
rounding the bulb of the Thermometer, was occu
pied by the liquid whofe conducting power was
determined. The fpace between the difk and
the cork ftopper , was filled with eider-down.

Between the infide of the brafs tube and the
lower part of the bulb of the Thermometer are
feen the wooden pins which ferved to confine the
Thermometer in its place.

Fig. 2. This is an horizontal fedion of the brafs
tube, and a bird's-eye view of the Thermometer in
its place.
VOL. U. R r PLATE



3 1 Q Defer Iftkn of the Plates,

i

PLATE II.

Fig. 3. This Figure fliows the manner in which
the Experiments were made, in which a cake of ice
at the bottom of a tall glafs jar was thawed by hot
water Handing on its furface.

a, is an earthen bowl filled with pounded ice
and \vater, in which the glafs jar, b, was placed.

c, d, is the level of the upper furface of the ice
in the jar.

e, /, is the level of the furface of the water
{landing on the ice in the jar.



END OF THE SECOND ESSAY.



$3* THE Publijher of this American Edition of Count
Rumford's EJfays, informs Subfcribers thereto and the
Public, that, as no title-page has been given to thefs
two Effays in the London Edition, prefuming they are
not intended for a volume, he requejis the pojfeffbr not
to bind them until iue hear further from the Author*

AUGUST, 1798.



Plate.





Jbale of Inches



Ratio



\y &.Wr*t,JV: 56 Comhilt.



S S A Y VIL

PART II.

An Account of feveral NEW EXPERIMENTS,
with occafional Remarks and Obfervations,
and CONJECTURES fefpe&ing Chemical