450 Mr. Herapath on True Temperature, and the [Dec. 



Prop. XX. Prob. V. 



Let it be required to determine the ratio of the baromerins of 

 water and vapour, the result of some experiment on the mixture 

 of certain portions of these bodies being given. 



We may now be said to have arrived into the regions of expe- 

 rimental inconsistency and error. After what we have already 

 done on this subject, there can be no difficulty in the physical 

 or mathematical part of our present inquiries ; but the misfortune 

 is we know not what or whose experiments to prefer. In no one 

 subject, except, perhaps, in that of " capacities," do the expe- 

 riments of philosophers exhibit such incongruous and discordant 

 results, as on their favourite and much cultivated doctrine of 

 " latent heat ; " and particularly on that part of it which is of by 

 far of the most practical consequence, the determination of the 

 " latent heats " of vapours. A fair specimen of these discordan- 

 cies may be seen in the following table : 



Authors. Latent heat of aqueous vapour. 



'Black , . . 800° Fahr. 



Watt 950 



Lavoisier. .. „ 1000 



Southern.,... . ..«'.* 945 



Clement aiuLDesormes 990 



Rumford. ..... 1021 



Thomson. . . * a . . . 1016 



Tre 888 



N.B. Dr. fre computes the latent heat of water to be 967°; 

 but that his experiment gives only 888°. — See note to Prop. 22. 



The number ^.here set against Dr. Thomson's name I have 

 copied from page 70 of the first volume of his Chemistry ; but I 

 cannot tell whether it be the result of experiment or calculation. 

 If it be derived 'from experiment, the circumstances are by far 

 the most judiciously chosen of any that I have seen. In general, 

 philosophers have sought to determine the "latent heat" of 

 vapour by the condensation of a very small quantity of vapour 

 on a very considerable quantity of water. By this process, the 

 rise of temperature which the condensation occasions is very 

 inconsiderable, and consequently a very trifling error in measur- 

 ing this rise exposes the determination of the numerical value of 

 its cause to material errors. This objection applies with consi- 

 derable force against the experiments of Count Rumford, where 

 the augmentation of temperature is under 13° Fahr. and with 

 much greater against the experiment of Dr. Ure, who had an 

 increase of only 6i°. In the case mentioned by Dr. Thomson, 

 the rise is no less than 1 16-2-° Fahr. which is sufficient, if it be the 

 result of correct experiment, to enable us to determine, with all 

 reasonable precision, the elements of the problem in question. 

 For this reason, and because the result nearly coincides with that 



lflfefe«y...H 



