HEAT. 65 



by the operation ; and we know farther that if the air 

 which melted a bit of ice can be sufficiently ratified or 

 attenuated, and the water that is converted into vapour 

 removed, the remainder of the water will become ice. 

 On the other hand, we know that if air be forced into 

 a smaller space, or made heavier, bulk for bulk, it will 

 become warmer ; and if the condensation be rapid and 

 extensive enough, the heat will kindle a match. We 

 know, in fact, that all substances increase in bulk by 

 the communication of heat to them, and diminish in 

 bulk when it is abstracted ; and that those results take 

 place even in cases where appearances are against 

 them. Water, which 'is heaviest at about the tempe- 

 rature of 39 J, expands as it is cooled towards freezing 

 as well as when it is heated above freezing ; and hence 

 one would at first be led to suppose that it both ex- 

 panded and contracted by heat, contracted from 32 

 to 39 1, and expanded afterwards. But the expan- 

 sion which goes on very slowly till it be within a single 

 degree of the freezing point, arises from the incipient 

 formation of the crystals of ice ; and the increase in 

 volume by cold is occasioned merely by the pores or 

 empty spaces between them. In the case of other 

 substances that become less in volume by the applica- 

 tion of heat, some part of them is separated, as the 

 juice of a fruit when it is dried in the sun. Each sub- 

 stance has generally a different degree of heat, at 

 which its component parts are separated, and its state 

 changed from solid to liquid, or from liquid to vapour 

 or gas ; but to understand the general principles that 

 enter into the natural history of the year, it is not 

 necessary to attend minutely to these: atmospheric 

 G 2 



