272 
On the Scale of Temperature. 
[Sept. 
If it be then true, as this writer asserts, that not only mercury, but every other 
liquid resembles water in this respect, the fact can no longer be viewed as an ano- 
maly. In this case it is the indications of the thermometer that are in fault, and 
consequently our scale of temperature must be erroneous. This was Mr. Dalton’s 
opinion, who was one of the first to observe the fact with regard to water. His 
opinion was that the squares of the true temperatures would correspond with the 
expansions reckoned from the point of greatest density ; and that whatever the fluid 
of which the thermometer should be constructed, if its degrees were formed on this 
principle, they would agree. InRees’ Cyclopaedia, under the article Heat, it is further 
asserted thatthe volumes of a gaseous body, submitted to different temperatures, as 
measured by such a thermometer, form ageometric progression, when the tempera- 
tures vary by a common difference. Further, that the cooling of a body heated 
above the temperature of the medium in which it is placed, proceeds according to the 
law of Newton, i. e. is proportional to the excess of temperature. Lastly, that the 
elasticities of steam, when referred to the indications of such a thermometer, would 
form a geometrical series, the temperatures being in arithmetical progression. 
Were this a true statement of facts, Mr. Dalton’s scale of temperature would be 
clearly, if not the true one, infinitely the more convenient, and as such doubtless en- 
titled to general adoption. .With regard in fact to the question of which is the 
true scale, it is a dispute about words, for till we can settle what we understand by 
temperature, it is vain to inquire which is the true and which the false method of 
estimating its amount. Bnt if it can be shown ; that one scale more than another 
simplifies our consideration of questions connected with the propagation or commu- 
nication of heat ; if it occasiou anomalies to disappear, and enable us to express the 
known facts after a more concise, intelligible, and universal algorithm 4 there are good 
practical grounds for adopting such a scale, even though, from our ignorance of the 
nature of heat aud temperature, we fail to prove that it is the time one. If therefore 
it could be shown, that the adoption of Mr. Dalton’s scale would he attended with 
such a simplification of the phenomena of heat, there would be little question as to 
the value of the scale, whatever epithet we might attach to it. 
It is needless to say, thatsuch is not the case. When the volumesof agas are taken 
in geometrical progression, and the temperatures, as an arithmetical series corre- 
sponding; the elasticities of steam have an increasing ratio; as have also the decrements 
of temperature of a heated body, estimated as occurring in equal intervals of time. 
This scale, therefore, notwithstanding its promised advantages, has neve been adopted ; 
nor is it likely it ever will, although it has still one plausible argument in its favour, 
— the equal ratios of expansion in gases answering to equal differences of tempera- 
ture. Thus it supposes that 10* or 1', in whatever part of the scale, occasions such 
an increase in the volume of a gas, as uiust always hear the same proportion to the 
volume before expansion. This is a simplicity which the common scale wants, in 
which the same change of temperature may either triple the volume, or only occa- 
sion an expansion of r. 
In our attempts to form an idea of a scale of equal differences of temperature, we 
labourunder insuperable difficulties, owing to our ignorance of the nature of heat. The 
common method of avoiding this difficulty is, to consider the subject in the following 
point of view. If we take two equal portions of a substance having different tempe- 
ratures and mix them, it seems a reasonable inference that one portion will lose as 
much heat as the other gains; and consequently that the temperature of the mixture 
will he the arithmetical mean of the two original temperatures. If we apply this 
theorem to the mixture of different portions of water, the result does not confirm the 
opinion, that Fahrenheit’s scale measures equal increments of heat: the temperature 
of the mixture is always below the mean. The fact has been attempted to be ac- 
counted for by those who consider Fahrenheit’s scale to be the correct one, by say- 
ing, that water has a greater heating, and consequently a greater cooling power as the 
temperature falls. This is an example, amongst many that could be produced, of a 
mere change of enunciation in a proposition being mistaken for an explanation. 
For in reality, the supposition of an increasing specific heat in the water as its tem- 
perature falls, is but another manner of stating the above curious result. No reason 
has ever been assigned why this should be the case, — no attempt to connect it with 
other facts, or to refer it to a more generallaw. And if it be the fact, that ail other 
liquids resemble water in this respect, we shall he more convinced of the absurdity 
of the supposed explanation. If however it prove, that water is the exception and not 
the rule, it must he allowed, that, however inexplicable, the fact is not conclusive 
against the truth of our thermometric scale. 
This is what the advocates for that scale assert. And they add, that all liquids 
are more or less irregular in this respect ; mercury being the least so. They there- 
