L I Q 
um that part of the tube which rifes above the Hopper, 
and was divided according to that of Fahrenheit. The 
thermometer being fixed in the tube as above defcribed, 
furrounded by the fubftance the conducting power of 
which was to be afcertained, the inftrument was placed in 
thawing ice, and kept there till the thermometer fell to 
32 0 . It was then taken out and inftantly plunged into 
boiling water, and the conducing power of the fubftance 
under experiment was eftimated by the time required to 
make the heat pafs through it into the thermometer; the 
time being carefully noted when the liquid in the thermo¬ 
meter arrived at 40 0 , and alfo when it came to every 20th 
degree above it. A number of experiments were made 
with this inftrument, all of which tended to prove that 
the paflage of heat through water is much impeded by 
mixing other fubftances with it, whether they be fuch 
(eider-down for inftance) as merely embarrafs the water in 
its motions, or tliofe (mucilages, &c.) which lelfen its 
fluidity. 
The* time required to heat the inftrument from 32 0 to 
200, when the bulb of the thermometer was furrounded 
with 
1. 2276 gr. of water and 192 gr. of ftarch, was 1109* 
2. 2276 gr. of water and 192 gr. of eider-down, 94.9 
3. With water and 4.8 gr. or ■j^th of its bulk of 
eider-down ... . 765 
4. With dewed apples - - 10964 
5. With pure water .... 597 
The ftarch was boiled with the water with which it was 
mixed, as was alfo the eider-down ; the latter to free it 
from air, which adheres to it with great obftinacy. 
As count Rumford fuppofed, and juftly, that there was 
a poflibility of error when the refults were determined by 
the extreme points freezing and boiling , owing to the ilow- 
nefs of the procefs when the inftrument approaches the me¬ 
dium of the temperature in which it is placed, he thought 
they might be got more accurately by taking the times 
occupied in heating from So 0 to 160. They were as fol¬ 
lows ; 
With the j ft: - - 341" 
With the 2d - 269 
With the 3d - - 215 
With the 4th - - 335 
With pure water - 172 
The time the heat took to pafs out of the thermometer 
through the fame fubftances: 
from 200° to 40° 
from 160° 
through the ift - 
- 1548" 
468" 
the id - 
- 1541 - - 
- 460 
the 3d - 
- 1395 - - 
373 
the 4th - 
- J 749 « - - 
- 520 
Water - 
- 1032 
277 
As the refults of thefe experiments prove that the pro¬ 
pagation of heat in water is retarded, not only by thofe 
things which diminilh its fluidity, but alfo by thofe which, 
by mechanical means, and without forming any combi¬ 
nation with it whatever, merely obftrufil its internal mo¬ 
tions, it appears to the ingenious author, and we believe 
it will not’be denied, that this proves, that heat is propa¬ 
gated in water in confluence of its internal motions; or, 
that it is tranfported or carried by the particles of that li¬ 
quid, and that it does not fpreaa and expand in it, as in 
metals and other conductors of heat, as has generally been 
imagined. 
The fuccefsof thefe experiments encouraged the author 
to plan and execute others Hill more decilive and extremely 
interefting. It has been generally believed that water can¬ 
not be heated in contaft with ice; reflecting on the fub- 
jedl, he perceived that either this muft be a miftake, or his 
ideas refpefting the manner in which heat is propagated 
in that fluid erroneous. On the fuppofition that water 
is not a conductor of heat, according to the common ac¬ 
ceptation of that term, or that heat cannot pafs in that 
Vyr,. XII. No. 870. 
U I B. 777 
fluid, except when carried by its particles, (which, being- 
put in motion by the change it occafions in their fpecific 
gravity, tranfports it from place to place,) it did not ap¬ 
pear how ice, if, inftead of being permitted to fwim on 
water, it were confined at the bottom of it, or any where 
below the furface, could aft'eft the temperature of the fu- 
perincumbent water, or prevent its receiving heat from 
other bodies; but, on the other hand, the ice-cold water 
which relults from the melting of ice when on the furface 
of that fluid, cannot but defcend by its gravity to the 
bottom of the containing veffel in an uninterrupted ftream ; 
fo that, as long as this operation is going on, the mafs of 
water cannot be much heated. 
A cylindrical glafs jar 4-7 inches diameter and 14 inches 
high was nearly filled with boiling-hot water. A circu¬ 
lar cake of ice of nearly the fame diameter, 34 inches 
thick, and weighing io^oz. was then gently put on the 
furface of the water. In 2' 58" it was entirely melted. 
A cake of ice of the fame form and dimenfions with the 
above was afterwards fattened down on the bottom of the 
fame jar, by means of two (lender laths, a trifle longer 
than the diameter of the jar, forced down into it, and laid 
acrofis each other over the ice. Cold water to cover the 
ice to the height of } of an inch was then poured into the 
jar to prevent its breaking by the fudden application of 
boiling water with which it was next filled. To protect 
the ice while the hot water was pouring in, a circular 
piece of paper was previoufly laid on its furface, which 
was afterwards gently removed by means of a (tring fait- 
ened to one fide of it. The refult of the experiment 
proved, that the count was not miftaken in his ideas. 
When 20' had elapfed, the heat in the water at differ¬ 
ent depths was found to be—immediately above the fur¬ 
face of the ice, 40 0 ; at 4 inch above it, 46°; at 1 inch, 
130 0 ; at 3 inches, 159°; at 7 inches, 160 0 . At the end 
of 35', it was, at the furface of the ice, 40 0 ; at 4 inch, 
76°; at 1 inch, no 0 ; at 2 inches, 144 0 ; at 3, 148°; at 5, 
1484°; at 7, 149 0 . After 1 hour and 15', it was, at the 
ice, 40°; 1 inch above it, 82°; 2 inches, 106 0 ; 3, 123 0 . 
Thefe were taken near the fide of the jar; the following 
at the axis: At the end of 1 h. 30 m. at the furface of 
the ice, 40 0 ; at 1 inch above it, 84°; at 2 inches, 115 0 ; 
at 3, 116 0 ; at 7, 117 0 . At the end of 2 hours, itill 40 0 
at the furface of the ice; 76° at 1 inch above it; 94 0 at 2 
inches; io6°at3; io8°at4; io8J°at6; 1084 0 at 7 inches. 
An end was now put to the experiment; the water was 
poured off; and, the remaining ice being weighed, it was 
found that 5 oz. 6 gr. troy had been melted. Taking 
the mean temperatnre of the water at the end of the ex¬ 
periment at 106 0 , it appeared that 734 oz. of hot water 
employed was cooled 78°. As 1 oz. of ice abforbs juft: 
as much heat in being melted as 1 oz. of water loles in 
being cooled 140 0 , each of thefe 735 oz. would give off as 
much heat as would melt -j^ths of 1 oz. of ice ; that is, 
all of them together as much as would melt 4c -fa ounces 
of ice; but the quantity melted was only about 5 oz and 
hence it appears, that left than -feth part of the heat tojl by the 
mater mas communicated to the air ; or, ice melts eighty times 
f outer at the bottom of a mafs of boiling-hot mater than when 
fwirmning on its furface. 
But the ice was melted, though (lowly, at the bottom of 
the hot water, which the author accounts for in a fatif- 
fa< 5 tory manner without needing to abandon his hypothe- 
fis. Freffi water prefents an exception to the general lavr 
of condenfation by cold; it condenfes in cooling till it 
comes to 40°, but on cooling it lower it expands, and con¬ 
tinues to expand till it freezes and after it has become 
ice. (Salt water on the contrary continues to be con- 
denfed the more, the more it is cooled.) Water in con¬ 
tact with melting ice is always at the temperature of 32 0 , 
and therefore fpecificaliy lighter than that which is 8 de¬ 
grees warmer, or at 40 0 ; the latter will therefore defend, 
being heavier, and by' its greater heat melt a portion of 
the ice, be cooled, and then give place to the delcending 
currents of warmer water which l'ucceed it. It appears 
9 M too 
