1819.] and on the Laws of the Communication of Heat. 177 
ties observed by quantities, which scarcely amount to the hun- 
dredth of a degree. 
Now if liquids so different in their nature, their density, and 
their fluidity, exhibit laws of cooling absolutely similar, is it not 
natural to draw the same consequence to which we were already 
led by a comparison of the cooling of unequal masses—That 
within the limits of our observations, the cooling of a liquid mass 
is subjected to the same law as a body of infinitely small dimen- 
sions ? 
It remains now to examine the influence of the nature and 
shape of the vessel. 
We, in the first place, compared the cooling of two spheres ; 
the one of glass, the other of tin plate, both filled with water. 
(The diameter of the tin plate sphere was a little greater than that 
of the glass sphere.) 
Excess of the tem- 
perature of the Velocity of cooling|Ditto of the tin'Ratio of these velo- 
body. of the glass sphere.} plate sphere. cities. 
60° 1-39° 0-90° 1-54 
50 1-13 0°73 1:55 
40 “0°85 0°54 1:57 
30 0-62 0:38 1-63 
20 0°37 0°21 1:76 
Here the ratios in the fourth column vary always the same way, 
and show us that the law of cooling is more rapid in the tin 
plate sphere than the glass sphere. Mr. Leslie obtained the 
same result, which he has generalized by admitting that this law 
changes with the nature of the body, and that it is most rapid in 
those bodies that radiate least. This proposition is true in the 
portion of the scale to which Mr. Leslie’s experiments were con- 
fined ; but, by a very remarkable casualty, the contrary effect 
takes place at high temperatures ; so that when we compare the 
laws of cooling of two bodies with different surfaces, that of the 
two laws which is most rapid at the lower part of the scale, 
becomes the least rapid at high temperatures. Thus in the 
series given above, the ratios, inserted in the last column, dimi- 
nish in proportion as we consider greater excesses of temperature ; 
they should increase ; and as is the case with all quantities which 
change their sign, these ratios remain nearly the same during a 
considerable extent of the thermometric scale. This is one of 
the most important points of the theory of cooling. If we do not 
deceive ourselves respecting the accuracy of our observations, a 
very simple explanation will be found in the subsequent part of 
this memoir of this remarkable fact, which can only be observed 
by making experiments, as we have done, on the cooling of bodies 
raised to a high temperature. 
Vou. XII. N° III. M 
