1830-3 
On Sensible Tenperature. 
li k 
heated but 5®. The proportion will be something more than twice, though less 
than four times. These considerations show that such a method of estimating 
sensible temperature is worse than useless. 
But though the law of cooling, as affected by difference of temperature, be such 
as to interfere with any such method of estimating sensible temperature, it is not 
so with the size or shape of the bodies to be cooled. These elements of the ques- 
tion follow a simple law, the rapidity of cooling being directly as the surface and 
inversely as the mass, i. e. in similar bodies, directly as the squares, and inversely 
as the cubes of the homologous sides. And in dissimilar bodies, particularly 
where one or both are very irregular in shape, though it might be difficult to as- 
sign what the proportion is, yet we may be sure it will always continue the same 
as long as the surface and mass of each body continue the same. If, therefore, 
we can place the thermometer in the same circumstances as the human body, (I 
mean as far as those circumstances affect the cooling process,) we shall be certain, 
notwithstanding the great difference as to surface and mass, that the results will 
always be proportional, and consequently comparable. Thus, if, in certain circum- 
stances, a thermometer is found to lose as much heat as sinks it 1°, while under other 
circumstances it will lose as much as make it sink 10o and in the same 
moment of time, it cannot be doubted but that the human body was losing, in 
these two cases, heat in the same proportion, i. e. (supposing the thermometric divi- 
sions to represent equal degrees of heating power,) ten times as much in the one 
case as in the other. For allowing that in the human body there was 1000 times 
less heat lost than in the thermometer, under the first supposed state of things, it is 
certain that in the other case it would also lose 1000 times less than the thermometer, 
because the surface and mass remaining the same, the. proportion which depends 
on them would remain the same ; and, therefore, the i° and 10°, being divided by 
the same number, their quotients would still be to each other as 1 : 10. 
Such being the case, it is only necessary to place the thermometer in precisely 
the same circumstances as we are ourselves in ; and by observing how many 
degrees it falls in a given moment of time, we have a comparative estimate 
of our sensations in numbers. That those numbers do not represent the heat lost 
by the human body, signifies little ; for the question we are concerned to resolve 
is, not how much beat is given off in such and such circumstances, but rather, 
whether more or less, and in what proportion under two different sets of con- 
ditions. Yet, although the absolute quantity is not wanted to render the re- 
sults comparable, we require that the proportion should be always the same ; 
in other words, that the thermometer, from the indications of which we draw 
our estimate, should be the same. This is a condition scarcely possible to be 
fulfilled, for we seldom can find even two thermometers of the same size and shape. ; 
but we may obviate the objection by reducing the results made with any thermome- 
ter, whatever its size, to a standard one. This may be done two ways ; either by 
measuring accurately the size and surface of the ball of the thermometer, and from 
these data reducing the indications to the hypothetical or standard instrument, by the 
application of the theorem already announced, viz. that the rapidity of cooling is di- 
rectly as the surface , and inversely as the mass ; or by taking a certain state of the at- 
mosphere, with regard I mean to all the conditions which affect the cooling process; 
then assuming any convenient sized thermometer as the standard, and observing 
what are its indications — those of any other thermometer may he reduced to it, by 
applying it to the same fixed or standard state of the atmosphere, which should be 
such as to be of frequent occurrence. The scale of the instrument actually used 
being, so to speak, fixed by either of these proceedings, the other results obtained 
with them become comparable amongst themselves. 
Thus, then, the first point is to establish a good set of standard experiments with 
which other experimenters might compare their results, and perhaps be enabled to 
deduce some useful truths. I was desirous of undertaking such a series, but regret 
that the want of leisure has prevented me ; and I must conclude this paper with a 
few meagre indications of the results I obtained, which I olfer moie with the view 
of fully explaining the subject, than as thinking them either useful or even curious. 
In fact, having neglected to take note of the size of the thermometei, they are ren- 
dered of no use whatever as terms of the comparison. Such as they are, they may 
serve to incite others to take up an enquiry, which is, I am persuaded, well worthy 
of being prosecuted, if only as a means of throwing light on a \ery curious 
physiological question. , . . o , 0 , .. 
The first experiment I tried was when the temperature ot the an was o i , and that 
of a moist thermometer 74°. In this state ot things I took a thermometer, covered 
the bulb with cambric muslin, and having dipped it into water of the temperature 
98®, I observed that it sunk in one minute. In this experiment* the cooling 
