282 
On the Measure of Temperature , fyc. 
[Sept 
In looking over the second and third columns of this table, we see that the tem- 
peratures deduced from the expansion of mercury and of air, are far from aereeins- 
as they ought to do, if Mr Dalton’s hypothesis had any foundation. The discor- 
dance, nevertheless, in the higher indications does not appear so great as it really 
is. for as the relative scales of mercury and air have two terms in common, the 
boiling and freezing points of water ; the great discrepancy which appears in 
the lower part of the scale has no influence on the upper temperatures. It is, in fact, 
as if the two scales commenced from different points— but in making them corn- 
mence from the same indication the discordance in the lower terms would he felt 
througnout the scale. So that even in measuring temperature by Mr Dalton’s 
new scale, the two first laws we have quoted would be found to represent in no way 
the actual phenomena. 3 
Again, if we would take into consideration the third law relative to the capacity for 
heat of bodies, we shall see that Mr. Dalton’s thermometer is still farther removed 
from that perfection which he attributes to it. 
In comparing the numbers in the first column of the preceding table, with those 
corresponding to them, in the 2nd and 3rd, we see that, with the exception of the 
terms 0 and 100, the accordance of which is assumed by the terms of the question, 
the indications of the ordinary air thermometer are always higher than the others. 
But we have shown that, even on that scale, the capacities of solid bodies increase 
more rapidly than their volumes— a fortiori then they would not appear to be con- 
stant, when relerred to a scale having a less rapid divergence. 
Finally, to show in a tew words that Mr. Dalton’s fourth proposition is also con- 
tradicted by experiment, it may be sufficient to say, that the law of cooling in the 
air is not the same for all bodies ; and that, therefore, no thermometric scale can sa- 
tisfy the condition of occasioning the loss of heat in each body to appear propor- 
tional to the excess of the temperature. 
Although the propositions we have just been discussing have not that foundation 
suposed by Mr. Dalton, they at least prove that the insufficiency of the doctrines 
generally received, had not escaped the penetration of this celebrated philosopher. 
The greater part of the phenomena of which he had perceived the irregularity, do 
taiy in the sense he has supposed \ but he ivanted the necessary data to verify his 
ingenious theory. The researches of which we have -riven some account, enable 
us to give a much more precise account of the measure of temperature, and to re- 
solve several difficulties which have been supposed to belong to the subject. It is 
evident, from wliat we have said on the variation of capacity, that no thermometric 
scaie can indicate directly the increments of heat, corresponding to a determinate 
rise ot temperature ; for— supposing one were found which had this property, consi- 
dered m relation to one particular substance, it could not be applied to others, as 
the capacity of different bodies varies according to different laws. 
In comparing together, therefore, the different thermometric scales, we may be 
assure that there is not one in which the expansions of all bodies will be expressed 
according to simple laws. Add to which, these laws would vary according to the 
scale adopted. So in taking the air thermometer as the standard, the rate of ex- 
pansion for all bodies, increases with the temperature— with iron a^ain, it would ap- 
pear to decrease ; while with mercury, freed from the irregularity occasioned by the 
c.ian D t of \ olume in its containing substance, iron and copper would appear to 
a \ e an j nci easing rate ot expansion ; and platina and the gases a decreasing one. 
And although, in the present state of the question, we cannot assign any decisive 
reason tor the exclusive adoption of any of these scales, we may yet sav, that the 
generally recognised uniformity of the principal phvsical properties of 'the gases, 
and chiefly the perfect sameness of their rate of expansion, render it probable that 
tiie sources of irregularity have in these bodies, perhaps, less infiuence than in 
soiidsanc liquids; and, consequently, that changes of volume, produced by heat, have 
mou daect dependance on the cause of their production. It is then the more 
•’ la ^ ~ e greater number of the phenomena of heat may present themselves 
un ei moie simple relations, when referred to the indications of an air thermome- 
ter. At east we have been determined, by these considerations, to employ this 
stale ill the prosecution of those researches which will form the subject of the 
econ pai tot this memoir. And the success which has crowned our enquiries 
nnt^ i >e °“ eiet as an additional argument in favor of the view here taken. We do 
case 10 " e . ver ’ mean to Sa y, that the other scales are to be excluded in every possible 
under • 1S ’ • instance > not impossible that certain phenomena should appear 
deduced ? ore8I “P le form, in referring the temperatures to the thermometric scale 
Join the expansion of the particular body which formed the subject of 
