1821.] l^hanomena of Heat y Gases, GravitatioriyS^c, 407 



the goodness to send me the results of these and any accurate 

 experiments of the kind. 



The same principles might easily be extended to sohds ; but 

 sufficient has been said for the hmits of a mere memoir on this 

 part of the subject. I shall, therefore, now briefly, offer one or 

 two observations for which a distinct hypothesis has been con- 

 trived ; namely, the hypothesis of " latent heat." It is generally 

 admitted, 1 believe, that when gases or vapours become fluids, 

 two or more of their atoms or particles unite together to form a 

 particle of the fluid ; and that Vv'hen fluids become sohds, a still 

 further aggregation of the particles takes place ; so that fluids 

 are solids resolved into simpler elements, and gases sohds 

 resolved into still simpler, or, perhaps, their simplest elements. 

 This being granted, it follows, that if two or more atoms or 

 parts of the gas unite together into one, the motion of this one, 

 being compounded of the motions of the others which compose 

 it, will be greater than the motions of either of the constituent 

 parts before the aggregation. Therefore, if the temperature of 

 the bodies in each of the three states be measured by the motions 

 of their constituent parts, then when a gas hy any means 

 becomes a fluid, the aggregated parts of the fluid will have a 

 greater motion than the parts of the gas ; and consequently its 

 temperature for a little while will be higher. For the same 

 reason the temperature will be raised in passing from a fluid to 

 a solid state ; and in all cases the increase of temperature in 

 passing from a gaseous to a fluid, or from a fluid to a solid state, 

 will be proportional to the aggregation of the particles ; so that 

 in either change of state, the greater the number cf particles 

 that unite, the greater the rise of temperature. And conversely^ 

 an augmentation of temperature being produced by a union of 

 particles, a diminution must be produced by a separation of them ; 

 and this diminution will be the greater, the greater the number 

 of parts into which a particle is divided. Therefore the consti- 

 tution of things being such as I have supposed, a diminution of 

 temperature will accompany a change of state from a solid to a 

 fluid, or from a fluid to a gas or vapour. But though with these 

 views an increase of temperature generally accompanies a 

 change of state, when by that change the solidity is increased, 

 or vice versa, yet this is not of necessity a law of nature. Cir- 

 cumstances might be easily varied, so that an augmentation of 

 solidity by a change of state may be attended with a diminution 

 of temperature, or the contrary ; and thus the phsenomena of the 

 increase of temperature at the explosion of gun and all fulminat- 

 ing powders, as well as other anomalies, might be accounted for. 

 These, however, are things which the importance of others will 

 not allow me now to stop to explain. The same reason prevents 

 me from entering into an explanation of the cause of that appa- 

 rent loss of temperature in the experiment respecting the admix- 

 ture of equal parts in weight of water at 172°, and snow or ice 

 at the temperature of water freezing, and other similar pha&no- 



