ON THE SOURCES AND EFFECTS OF HEAT. 493 



freezing point ;* they then expand again almost in an equal degree as they 

 are still more cooled ; and provided that they be free from agitation, they 

 may remain liquid at a temperature considerably below the point at which 

 they usually freeze, and at which their ice always melts. Water may be 

 cooled in this manner to about 10 of Fahrenheit, and if it be then agitated, 

 or especially if a small particle of ice or snow be thrown into it, a certain 

 part of it will instantly congeal, and its temperature will be raised at once 

 to 32, in consequence of the heat which is always produced or extricated 

 in the act of freezing, f In most cases, although not in all, the solid 

 occupies more space than the fluid : thus, it is probable that ice, when per- 

 fectly free from air bubbles, is at least one 16th lighter than water at the 

 same temperature. A saturated solution of Glauber's salts, or sulfate of 

 soda, in hot water, may be cooled slowly to the temperature of the atmo- 

 sphere, when the pressure of the air is excluded, and may be made to crys- 

 tallize by admitting it suddenly, the liquor becoming at the same time 

 warm in consequence of the heat which is extricated ; and there is no 

 doubt but that the congelation of water, and perhaps of all other sub- 

 stances, is a crystallization of the same kind. 



The expansions of solid bodies appear to be more regular than those of 

 liquids or even of elastic fluids ; they vary little at any temperature, 

 although it is said that they do not always take place in their full extent 

 at the instant that the substance has become heated, and that a blow, or 

 the agitation produced when they are made to sound by the friction of the 

 bow of a violin, may sometimes be observed to cause them to assume the 

 state of equilibrium with greater rapidity. Brass expands about one hun- 

 dred thousandth of its length for each degree of Fahrenheit, copper and 

 gold a little less ; silver somewhat more ; glass and platina less than half 

 as much ; iron and steel about two thirds as much ; tin one third more, and 

 lead and zinc about half as much more. Wood and earthenware are the 

 least expansible of all known solids. The diminution of the elasticity of 

 iron and steel by the elevation of their temperature amounts to about 5 ' 

 of the whole for each degree ; but probably various substances are variously 

 affected in this respect. 



The liquefaction of solids, and their conversion into fluids by the opera- 

 tion of heat, is liable to fewer irregularities than any other of its effects ; 

 the change depending only on the temperature, and not being accelerated 

 or retarded by any accidental circumstances. When the temperature is 

 too low, or the pressure too small, for the existence of the substance in a 

 liquid form, it may still be converted into vapour, either mixed with air, 

 or in a separate state ; thus ice loses weight when it is exposed to a dry 

 frosty wind ; and camphor, benzoin, and ammonia are sublimed by heat 

 without being melted, although it is probable that a pressure sufficiently 

 strong might enable them to exist as liquids in elevated temperatures. In 

 all changes from solidity to liquidity or to elastic fluidity, a certain quan- 



* On the point of maximum density of water, see "Waller's Trans, of the Floren- 

 tine Exp. p. 77. Blagden on the Congelation of Aqueous Solutions, Ph. Tr. 1788, 

 p. 277, Hope, Ed. Tr. v. 379. Some substances contract in freezing : see Despretz, 

 in Pogg. Ann. xli. 498. 



f See Blagden, Ph. Tr. 1788, p. 125 ; Walker, ibid. 1788, p. 395. 



