2. The effect of heat upon the sensations, although the one which must have been first attended to, and that which constantly offers itself to our observation, is found to be so inaccurate as a measure of its quantity, that we always have recourse, for this purpose, to the second of its effects, which we enumerated above, expansion. According to one of the characters of heat, which we have already described, that of distributing itself equally among bodies whenever two substances are brought together, which differ in their tem perature, the one gives off a portion of its heat to the other, and that which receives it becomes expanded, or has its bulk increased in all its dimensions. This expansion con tinues as long as the body in question maintains its elevated temperature ; but when the heat is withdrawn, it begins to contract, and by degrees it acquires its former bulk. This expansion occurs, with very few exceptions, in all bodies ; but it differs extremely in degree. It is by much the great est in aeriform fluids of all kinds. It is less in quantity, although still considerable, in liquids ; while in solids, it is so small as not to be perceived, except by the intervention of an apparatus expressly contrived to render it visible. Not only these general classes of bodies, but many of the indi vidual substances comprised under each of them, have very different degrees of expansive power, which has been fre quently made the subject of experiment.
The different gases have been submitted to numerous trials, for the purpose of ascertaining their relative expan sion. Priestley paid a good deal of attention to the subject. It was farther investigated by Roy, Monge, and Saussure ; and lastly, with much apparent accuracy, by Guyton. (Ann. Chinn. t. i. p. 256. et. seg.) Their results, although vari ous, agreed in the main conclusion, that each of the gases had a specific power of expansion peculiar to itself, and which is uniform for each of the gases at the same tempe rature. But, notwithstanding the weight of these authori ties, a farther examination of the subject by Mr Dalton and M. Gay-Lussac has led to an opposite conclusion, that all the aeriform fluids suffer the same expansion from the same addition of heat. Mr Dalton's experiments indicate, that 100 parts of any gas, in passing from the freezing to the boiling point, increase to 137 parts, which is nearly for each degree : (Manch. Mem. vol. v. p. 598.) M. Gay-Lus sac's experiments Were published a short time subsequent to those of Mr Dalton ; but being made without concert or co-operation, they are to be regarded as equally original, and they lead to conclusions so nearly similar, as to afford the strongest evidence of their correctness. (Ann. Chim. t. xliii. p. 137. et. seg.) The source of error in the expe riments that had been performed, previous to those of Dal ton and Gay-Lussac, appears to have arisen from a quan tity of aqueous vapour that had been mixed with the gases. See EXPANSION.
The expansion of liquors by heat differs, in many re spects, from that of gases ; it is less in amount, it varies much in different fluids, and it is found not to be in propor tion to the quantity of caloric that is added to the heated body. If we observe the effects produced upon mercury, water, and alcohol, by the same addition of heat, and when they are all at the temperature of 50°, we shall find it to be, in the first of these fluids, equal of its volume, fir the second and in the last -A. The comparative changes in bulk are also found to vary for the same fluid at different temperatures, contrary to what has been observed with re street to aeriform bodies, the expansion proceeding, in a gradually increasing ratio, as the temperature advances. Thus, at the temperature of 50°, the expansion produced by 'a single degree of heat is less than what would be pro duced by a single degree of heat at the temperature of 100'. Many accurate experiments, of which we have given an ac count in our article Ex I.Ansion, were performed on this subject by Deltic. He found, that in fluids the rate of ex pansion is frequently irregular near the two extremes of freezing and boiling, and more equable In the middle of the scale ; and hence he deduced a practical rule, that those fluids are the best adapted for measuring the degrees of heat, which have the largest interval between the limits of freezing and boiling. The general result of our examina tion of the expansive power of different fluids leads us to conclude that it has no connexion with their density, but it seems rather to be related to the quantity of heat neces sary to convert them into the gaseous state. Thus ether is probably the most expansible, and mercury the least so, of any fluids, because they possess respectively the lowest and the highest boiling points. Recherches, t. i. p. 271, 311.
We have hitherto spoken of the expansion of fluids by heat, and their consequent contraction by cold, as of a pro perty possessed by all bodies of • this description ; but there is one very remarkable exception to the general law, that of water. Water, through a certain range of temperature, like other fluids, experiences the usual augmentation of bulk by an addition of heat ; but when it is cooled down to about 40°, it seems to arrive at its maximum of density ; and if the cooling process be farther continued, it begins to expand, and continues its expansion until it arrives at its freezing point, when it becomes solid. By employing cer tain precautions, particularly by avoiding every degree of agitation, it will continue fluid for several degrees below the usual point of congelation, and it still goes on increas ing in volume, until its actual solidification takes place. When a water thermometer, therefore, stands at 50°, it 'is impossible to know whether it be really at the temperature of 50°, or at that of 30°, 18° above, or two degrees below, its freezing point.