1821.] Further Remarks upon Mr. Herapath's Theory. 303 



for his accurate adherence to fact, and rejection of theory, taking 

 up a subject, which at first sight appeared to me utterly incapable 

 of definition, and, therefore, improper for investigation ; and still 

 more surprised was I to find the names of such distinguished 

 philosophers as those whose researches are there described,, 

 engaging in a speculation, which, it appeared to me, a simple 

 attempt at definition would at once set at rest. Dr. T.'s remarks 

 at the end of that chapter put a satisfactory termination to such 

 inquiries ; and I felt much satisfaction on looking into the last 

 edition of his incomparable work to find the chapter wholly 

 omitted. 



We cannot take Mr. H.'s law of temperature as the true 

 law, unless we are sure that it holds good at all points in the 

 scale; but of this we cannot be sure, any further than within 

 those limits at which experiments have been tried. How do we 

 know that beyond those limits the law of expansion may not be 

 modified, or some totally different law prevail 1 



I will now proceed to one observation respecting Mr. H.'s opi- 

 nion of the doctrine of capacity. He states (see Ainials for Sept. 

 p. 204), that if the present doctrine of capacity were true, the 

 results of mixtures, when the higher temperature belonged 

 respectively to each of two bodies, should be equidistant from 

 the arithmetical mean. 1 beg leave to suggest, that, according 

 to the common doctrine, it by no means follows that if tempera- 

 ture be measured by the equable expansion of mercury, then the 

 temperature of a mixture should be an arithmetical mean between 

 the previous temperatures of the two bodies ; for we do not 

 know that the temperature of the bodies employed will be mea- 

 sured by their expansion. Heat does not in all cases cause 

 expansion, and, therefore, we are not certain that part of the 

 heat communicated by one body to the other may not have other 

 effects without causing expansion. 



Also if a certain degree of heat in one volume of a body be 

 effective in expanding mercury to a given degree; and a certain 

 degree of heat in another equal volume of the same body be 

 effective in expanding mercury to another different degree ; we 

 may assume hypothetically, that when the two volumes are 

 mixed, the mean of the two previous temperatures will, in some 

 way, be effective in the mixture, but we can by no means be sure 

 that it will be in such a way as to be displayed by a proportional 

 expansion of mercury. Philosophers would be very properly 

 employed in observing what the resulting temperatures are in 

 such mixtures, and, consequently, what part of the heat is, after 

 mixture, effective in expanding mercury ; but to talk of the real 



Suantity of heat in bodies is talking in terms which cannot be 

 efined, and, therefore, the investigation is to no purpose. 

 1 must also remark, that if Mr. H.'s theory be considered as 

 established, there is no opposing theory exploded, as far as the 

 experimental doctrines of capacity and latent heat are concerned ; 



