temperature.
The proportion of Fluids to Solids is much
greater in animals than in vegetables ; and in order
to preferve in them the great quantity of Heat
which is neceffary to the prefervation of life, they
are furnifhed with lungs, and are warmed by a
procefs fimilar to that by which Heat is generated
in the combuflion of inflammable bodies.
Among vegetables, thofe which are the moft
fucculent are annual* Not being furniflied with
lungs to keep the great mafs of liquids warm,
which fill their large and flender veflels, they live
only while the genial influence of the fun warms
them, and animates their feeble powers ; and they
droop and die as foon as they are deprived of his
fupport.
There are many tender plants'to be found in cold
countries, which die in the autumn, the roots of
which remain alive during the winter, and fend off
freih {hoots in the enfuing fpring. In thefe we
fhall conftantly find the roots more compact and
denfe than the ftalk, or with fmaller veiTels, and a
fmaller proportion of Fluids.
Among the trees of the foreft, we fhail con
ftantly find, that thofe which contain a great
proportion of thin 'watery liquids^ not only ihed
their leaves every autumn, but are fometimes
frozen, and actually killed, in fevere frofts.
Many thoufands of the largeft walnut-trees were
killed
Of the Propagation of Heat
killed by the froft in the Palatinate, during the very
cold winter in the year 1788 ; and it is well
known that few, if any, of the deciduous plants of
our temperate climate would be able to fupport the
excefllve cold of the' frigid zone.
The trees which grow in thofe inhofpitable
climates > and which brave the colds of the feverefl
winters, contain very little watery liquids. The
fap which circulates in their veflels is thick and
vifcous, and can hardly be faid to be fluid. Is there
not the ftrongeft reafon to think, that this was fo
contrived for the exprefs purpofe of preventing their
being deprived of all their Heat, and killed by the
cold during the winter .?
We have feen by the foregoing Experiments,
how much the Propagation of Heat in a liquid is
retarded by diminiming its fluidity ; and who
knows but this may continue to be the cafe,
as long as any degree of fluidity remains ?
As the bodies and branches of trees are net
covered in winter by the fnow which protects
their roots from the cold atmofphere, it is evident
that extraordinary meafures were neceflary to pre
vent their being frozen. The bark of all fuch trees
as are defigned by nature to fupport great degrees
of cold, forms a very warm covering ; but this
precaution alone would certainly not have been
fufficient for their protection. The fap, in all trees
which are capable of fupporting a long continuance
of froft, grows thick and vifcous on the approach
of the winter. What more important purpofe could
this change anfwer, than that here indicated ?
And
m Fluids. 229
And it would be more than folly to pretend that it
anfwers no ufeful purpofe at all.
We have feen, by the refults of the foregoing
Experiments, how much the fimple embarraflment
of liquids ift their internal motions tends to retard
the Propagation of Heat in them, and confequently
its paflage out of them ; and when we coniider
the extreme finallnefs of the veffels in which the
fap moves in vegetables, and particularly in large
trees 5 when we recollect that the fubftance of
which thefe fmall tubes are formed, is one of the
beft non-conduftors of Heat known ;* and when
we advert to the additional embarraffments to the
paffage of the Heat, which arife from the increafed
vifcolity of the fap in winter, and to the almofl
impenetrable covering for Confining' Heat, which is
formed by the bark, we mall no longer be at a lofs
to account for the prefervation of trees during the
winter, notwithftanding the long continuation of
the hard froft to which they are annually expofed.
On
* I lately, by accident, had occaCon to obferve a very ftriking proof of the
extreme difficulty with which Heat pafles in wood. Being prefent at the
foundery at Munich, when cannons were cafting, I obferved that the
founder ufed a wooden inftrument for ftirring the melted metal. It was a
piece of oak plank, green, or unfeafoned, about ten inches fquare and
two inches thick, with a long wooden handle, which was fitted into a hole
in the middle of it. As this inftrument was frequently ufed, and fometimes
remained a confiderable time in the furnace, in which the Heat was moft
iritenfe, I was furprifed to find that it was not confumed ; but Iwas ftill
more furprifed, on examining the part of the plank which had been
immerfed in the melted metal, to find that the Heat had penetrated it to fa
inconfiderable a depth, that, at the diflance of one-twentieth of an inch
below its ftirface, the wood did not feem to have been in the lead afie&ed
by it. The Color of the wood remained unchanged, and it did not appear
to have loft even its moifture.
VOL. II. G g
2 3 Of the Propagation of Heat
On the fame principles we may, I think, ac
count, in a fatisfa&ory manner, for the prefervation
of feveral kinds of fruit, fuch as apples and pears
for inftance, which are known to fupport, without
freezing, a degree of cold, which would foon
reduce an equal volume of fure water to a folid
mafs of ice*
At the fame time that the compact ikin of the
fruit effectually prevents the evaporation of its fluid
parts, which, as is well known, could not take
place without occafioning a very great lofs of Heat,
the internal motions of thofe fluids are fo much
obftrufted by the thin partitions of the innumerable
fmall cells in which they are confined, that the
communication of their Heat to the air ought, ac-*
cording to our hypothecs, to be extremely flow
and difficult. Thefe fruits do, however, freeze at
laft, when the cold is very intenfe j but it muft be
remembered, that they are compofed almoft en
tirely of liquids, and of fuch liquids as do not grow
vifcous with cold ; and moreover, that they were
evidently not deiigned to fupport,, for a long time,
very fever e frofts.
Parfnips and carrots, and feveral other kinds of
roots, fupport cold; without freezing, ftill longer
than apples and pears, but thefe are lefs watery,
and I believe the veflels in which their fluids are
contained, are fmaller ; and both thefe circum-
ftances ought, according to our aiTumed principles,
to render the paflage of their Heat out of them
more difficult, and confequently to retard their
congelation.
But
In Fluids. $31
But there is ftill another circumftance, and a
very remarkable one indeed, which, if our conjec
tures refpefting the manner in which Heat is pro
pagated in liquids be true, muft act a moft impor
tant part in the prefer vation of Heat, and confe-
quently of animal and vegetable life, in cold cli
mates. But as the probability of all thefe deduc
tions muft depend, very much, on the evidence
which is brought to prove the great fundamental
facl on which they are eftablifhed ; -that refpecl-
ing the internal motions among the particles of
liquids, which neceffarlly take place when they are
heated or cooled ; before I proceed any farther
in thefe fpeculations, I mail endeavour to throw
fome more light on that curious and interefting
fubject
Of the Propagation of
CHAP. II.
Farther Inveftigations. of the internal Motions
the Particles of Liquids , which necejjarily take
pla.ce when they are heated or cooled.- Defcription
of a mechanical Contrivance, by which thefe Mo
tions in Water were rendered vi/ible. An Account
of various amujing Experiments, which were
made with this new-invented InJlrument.They
lead to an important Difcovery. Heat cannot be
propagated DOWNWARDS in Liquids, as long as
they continue to be condenfed by Cold. Ice found y
by Experiment, to melt more than eighty Times
flower, when boiling-hot Water flood on its Surface,
than when the Ice was fuffered to fwim on the
Surface of the hot Water. The melting of Ice by
Water Jlanding on its Surface can be accounted for,
even on the Suppofition that Water is a perfect
Non-conducJor of Heat. According to the qffumed
Hy 'pot 'hefts , Water only eight Degrees of Fahren*
heit's Scale above the freezing Point, or at the.
Temperature of 40, ought to melt as much Ice^
in any given Time, when Jlanding on its Surface^
as an equal Volume of that Fluid, at any higher
Temperature, even were it boiling-hot. This
remarkable Foci is proved by a great Variety of
decifive Experiments. Water at the Temperature
of 41 is found to ?nelt e-ven MORE Ice, when ft and-
Ing on its Surface, than boiling-hot Water. The
Refuto
In Fluids. 233
Refults of all thefe Experiments tend to prove that
Water is, in fa5l^ a perfect Non-conductor of Heat ;
er that Heat is propagated in it, merely in confe-
quence of the Motions it occafans among the
infulated or folitary Particles of that Fluid, which,
among themfelvcs, have no Communication or Inter-
wirfe whatever in this Operation. The D.if cover y
(f this Fad opens to our View one of the grandeft
find mojl interejling Scenes In the Economy of
Nature*
AS the particles of water, as alfo of all other
Fluids, are infinitely too fmall to be feen by
human eyes, their motions muil of courfe be im
perceptible by us ; but we are frequently enabled
to judge, with the utmoft certainty, of the motions
of inviiible Fluids, by the motions they occafion in
vifible bodies. Air is an invifible Fluid, but we
acquire very juft notions of the motions in air, by
the duft, and other light bodies which are carried
along with it in its motions. Nobody who has
ever feen a whirlwind fweep over the furface of a
ploughed field in dry weather, can have any doubt
refpefting the nature of the motions, into which
the air is thrown on thofe occafions ; noUvithftand-
ing that they are extremely complicated, and
would be very difficult to defcribe.
It was by the motions of the very fine particles
of duft, which by accident had been mixed with
the fpirits of wine in my large thermometer, and
which, when ftrongly illuminated by the direct
beams of the fun, became vifible, that I firft dii-
covered
Of the Propagation of Heat
covered the internal motions of that Fluid, whicfy
take place when it is cooling ; and ? availing myfelf
of this kind hint, I contrived to render the internal
motions of water equally viiible. This, I immedi*
ately faw, could be done with the utmoft facility,
if I could but fin4 any folid body of the fame fpe-
cific gravity as water, which would be proper to
mix with it ; that is to fay, that would not be
liable to be diffolved by it, or to, be reduced to
fuch fmall particles as to become itfelf invifible ;
but fuch a fubftance was not to be found. On re-
fledion it occurred to me, that it is very fortunate
that fuch fubftances do not abound ; for otherwife
we mould find great difficulty in procuring water
in a pure ftate.
Not being able to find any folid fubftance fit for
my purpofe, of the fame fpecific gravity as pure
water, I was obliged to have recourfe to the fol
lowing ftratagem.
Looking over the tables of fpecific gravities, I
found that the fpecific gravity of tranfparent yel
low amber was but a little greater than that of
water, being 1.078, while that of water is i.ooo ;
and it occurred to me, that by diflblving a certain
quantity of pure alkaline fait, I might augment its
fpecific gravity, or rather bring the fpecific gravity
of the folution to be precifely equal to that of the
amber, without impairing the tranfparency of the
liquid, or changing any of its properties, by which
the manner of its receiving and tranfporting Heat
could be fenfibly affected.
This
in Fluids. 235
This Contrivance was put in execution in the
following manner, with complete fuccefs. Having
provided myfelf with a number of glafs globes of
various fizes, with long cylindrical necks, I chofe
one which was about two inches in diameter, with
a cylindrical neck \ of an inch in diameter, and
twelve inches long \ and putting into it about
half a tea-fpoonful of yellow amber, in the form
of a coarfe powder, (the pieces, which were irreg
ular in their forms, and tranfparent^ being about
the fize of muftard-feeds) poured about ^ it a
certain quantity of diflilled water, which was at the
temperature of the air in my room (about 60 F.)
Finding, as I expected, that the amber remained
at the bottom of the globe, I now added to the
water as much of a fatuated folution of pure vege
table alkali, as was fufficient to increafe the fpecific
gravity of the water, (or rather of the diluted
faline folution) till the pieces of amber began to
float, and remained apparently motionlefs, in any
part of the liquid where they happened to reft.
As the glafs body was not yet as full as I wifhed,
I Continued to add more of the alkaline folution,
and of water, in due proportions, till the globe was
full ; and alfo till its cylindrical tube was filled to
within about three inches of its end ; and then
clofed it well with a clean cork.
Having Ihaken the contents of this glafs body
well together, I placed it, with its cylindrical tube
in a vertical pofition, on a wooden ftand, and left
it to repofe in quiet, in order to fee how long the
folid particles of amber (which appeared to be very
equally
Of the Propagation of Heat
equally difperfed about in the whole rfiafs of the
liquid) would remain fufpended.
Though the greater number of thefe particles
feemed at firft to have no tendency either to afcend
or to defcend,yet fome of them foon began to move
very flowly upwards, and others to move as {lowly
downwards 5 and as thefe particles were moving at
the fame time promifcuoufly in all parts of the fame
liquid, and even in the fame part of it in both di
rections at the fame time^ the afcending and dc^
fcending particles frequently paffing each other fo
near as to touch, I faw that thefe motions were in
dependent of any internal motion of the liquid,and
arofe merely from the difference of the ipecific grav
ity of the different fmall pieces of the amber, and of
that of the liquid. Some of the pieces of amber
being evidently heavier than the liquid, moved
downward, while others which were lighter, af-
cended to its furface*
Finding that there was fo much difference in the
fpeciiic gravities of the different pieces of amber,
I now added more of this fubftance to the liquid,
and fuffering it to fubiide after I had Ihaken it well
together, I gently poured off what hadrifen to the
top of the liquid, and retaining only that which had
fsttled at the bottom of it, I increafed the fpecific
gravity of the liquid by adding a little of the alka
line folution, till the fmall pieces of amber which
remained in the glafs were juft buoyed up and fuf
pended in the different parts of the Fluid, where
they feemed to have taken their permanent fta-
tions.
I had
in Fluids. 237
I had now an inftrument which appeared to me
to be well calculated for the very interefting Ex
periments I had projected, and it will ealily be
imagined that I loft no time in making ufe of
it.
The firft Experiment I made with this inftru
ment was to plunge it into a tall glafs jar, nearly
filled with water alrnoft boiling hot. The refult
was juft what I expected. Two currents, in op-
polite directions, began at the fame inftant to move
with great celerity in the liquid in the cylindrical
tube, the afcending current occupying the iides of
the tube, while that which moved downwards oc
cupied its axis.
As the faline liquor grew warm, the velocity of
thefe currents gradually diminiflied ; and at length,
when the liquor had acquired the temperature of
the furrounding water in the jar, thefe motions
ceafed entirely.
On taking the glafs body out of the hot water,
the internal motions of the liquor re-commenced ;
but the currents had changed their directions,
that which occupied the axis of the tube being
now the afcending current.
When the cylindrical tube, inftead of being held
in a vertical poiition, was inclined a little, the af
cending current occupied that fide of it which hap
pened to be uppermoft, while the under fide of it
was occupied by the current which moved (with
equal velocity) downwards.
When the contents of the glafs body had ac
quired the temperature of the air in the room,
VOL. II. H h thefe
2 3$ Of ^ e Propagation of Heat
thefe motions ceafed, 'but they immediately re
commenced on expoiing the inftrument to any
change of temperature.
In all cafes where the inftrument received Heat,
the current in the axis of its cylindrical tube, when
it was placed in a vertical pofition (and that which
occupied its upper fide when it was inclined)
moved downwards. When it parted with Heat its
motion was in an oppofite direction, that is to fay,
upwards.
A change of temperature amounting only to a
few degrees of Fahrenheit's fcale, was fufficient to
fet the contents of the inftrument in motion ; and
the motion was more or lefs rapid as the velocity
was greater or lefs with which it acquired or parted
with Heat, and the motion was moft rapid in thofe
parts of the inftrument where the communicatior*
was not rapid.
A partial motion might at any time be produced
in any part of the inftrument by applying to that
part of it any body either hotter or colder than the
inftrument. If the body fo applied were hotter
than the inftrument, the motion of the faline liquor
in it in that part of it immediately in contact with
the hot body, was upwards , if colder, downwards;
and whenever a hot or cold body produced a cur
rent upwards or downwards, this current immedi
ately produced another in fome other part of the
liquid which flowed in an oppolite diredtion.
On inclining the cylindrical tube of the inftru
ment an angle of about 45 degrees with the plane of
the horizon, and holding the middle of it over the
flame
in Fluids. 239
flame of a candle, at the diftance of three or four
inches above the point of the flame ; the motion
of the Fluid in the upper part of the tube became
exceffively rapid, while that in the lower end of it
where it was united to the globe, as well as that
in the globe itfelf*, remained almoft perfectly at
reft.
I even found that I could make the Fluid in the
tipper part of the tube aftually boil, without that
In the lower part of it appearing to the hand to be
fenfibly warmed, Eut when the flame was direct
ed againft the lower part of the tube, all the upper
parts of it in contact with the liquid, and efpecially
that fide of it which was uppermoft as it lay in an
inclined pofition, where the afcending current was
moft rapid, where it impinged, againft the glafs,
Were very foon heated very hot.
The motions in oppofite directions, in the liquid
in the tube, were exceedingly rapid on this fudden
application of a ftrong Heat, and afforded a very
entertaining fight : but to afcientific obferver they
were much more than amufing. They detected
Nature, as it were, in the very act, in one of her
moft hidden operations, and rendered motions vifi-
ble in the midft of an invifible medium which
never had been feen before, and, which moft proba
bly had never been fufpecled..
Encouraged by this fuccefs, and confirmed in
my opinions refpecting the interefting fact I had
undertaken to inveftigate, I now proceeded with
confidence to ftill more direct and decifive Exper
iments.
It
240 Of the Propagation of Heat
. It is an opinion which, I believe, is generally
received among philofophers, that water cannot be
heated in contact with ice ; reflecting on the
fubjecl:, I immediately perceived that either this
muft be a miftake, or all my ideas refpecling the
manner in which Heat is propagated in that Fluid
muft be erroneous. I faw that as long as the ice
floats at the furface of water which is attempted to
be warmed over a fire, (or in any other way) the
ice-cold water which refults from the melting of
the ice, muft, according to my own hypothefis,
defcend, and fpreading over the bottom of the con r
taining veffel, and, before it has time to be much
heated, being in its turn forced to give place to
the ice-cold water which, as long as any ice re
mains, continues to defcend in an uninterrupted
ftream as long as this operation is going on, the
mafs of the water cannot be much heated ; but on
the fuppofition that water is not a conductor of
Heat, according to the common acceptation of that
term, or that Heat cannot pafs in that Fluid except
when it is carried by its particles, which, being put
in motion by the change it occaiions in their fpe?
cific gravity, tranfports it from place to place, it
does not appear how ice, if inftead of being per r
mitted to fwim on water, were confined at the
bottom of it, or at any given diftance below its fur-
face, could in any way affect the temperature of the
fuperincumbent water, or prevent its receiving
Heat from other bodies.
Were water a conductor of Heat, there is no
doubt but that the influence of the prefence of the
ice
in 'Fluids. 241
ke would be propagated in the water in all direc
tions.
The metals are all conductors of Heat, and
Profcffor PIETET found by an ingenious and de-
cifive Experiment,* that in a bar of copper 33
inches in length, placed in a vertical pofition, Heat
paffed downwards as well as upwards, and nearly
with the fame facility in both thefe directions ; and
if it can be mown that Heat cannot defcend in water,
that alone will, I imagine, be thought quite fufiicient
tp prove that water is not a conductor of Heat.
When we meditate profoundly on the nature
pf Fluidity, it feems to me that we can perceive
feme faint lights which might lead us to fyfpect
that the caufe^ and I may fay the very ejftnce of
fluidity is that property which the particles of
bodies acquire when they become fluid, by which
all farther interchange or communication of Heat
among them is prevented. But however this may
be, the refult of the following experiments will
certainly be confidered as affording indifputable
evidence of one important fact reflecting the man-
ner in which Heat is propagated in water.
Experiment, No% 15,
Into a cylindrical glafs jar 4.7 inches in diame
ter, and 14 inches high, \ fitted a circular cake of
ice nearly as large as the internal diameter of the
jar, and 3- inches thick, weighing lof oz.
Tin*
* Effaij de Phy%ue, tome i. Geneve 1790.
Qf the Propagation of Heat
This cake of ice being ready, I now poured into
the jar 61b. i^ oz. Troy, of boiling-hot water,
and putting the ice gently into it, I found that it
was entirely melted in 2 minutes and 58 feconds.
Having found by this Experiment how long the
ice was in melting at the furface of the hot water,
I now endeavoured to find out whether it would
not require a longer time to melt at the bottom'
of the water,
Experiment^ No. 16.
Into the fame jar which was ufed in the fore
going Experiment, I now put a cake of ice of the
fame form and dimenfions as that above defcribed,
but inftead of letting it fwim at the furface of the
hot water, I faftened it down on the bottom of th
jar, and poured the water upon it.
This cake of ice was faftened down in the jar
by means of two flender and elaftic pieces of deal
about - of an inch thich,' and \ of an inch wide,
which being a trifle longer than the internal diam
eter of the jar, were of courfe a little bent when
they were introduced into it in a horizontal po-
fition, and on being put down upon the ice, &
right angles to each other, ferved to confine the
ice, and prevent its rifing up to the furface when
the water was put into the jar upon it.
To protect the ice while the boiling-hot water
was pouring into the jar, its furface was covered
with a circular piece of ftrong writing paper, which
was afterwards removed as gently as poffible by
means
in Fluids. 243
means of a firing which was fattened to one fide of
it ; and to prevent the gla& jar from being crack
ed by the fudden application of the boiling-hot
water, I began by pouring a fmall quantity of cold
water into the jar, juft enough to fill up the in-
terftices between the ice and the glafs, and to
cover the ice to the height of about ~ of an inch ;
and in pouring the hot water into the jar, out of
a large tea-kettle in which it had been boiled, I
took care to direct the flream againft the middle
of the circular piece of paper which covered the
ice.
The jar with the ice and the hot water in it
being placed on a table near a window, I drew
away as gently as poffible the paper which covered