so as to reach both rings, and be suspended in them while it remains warm.

If spirit of wine is contained in an oblong vessel, and immersed in hot water, the spirit will be observed to rise like the mercury in a thermometer.

A bladder, with a little air in it, when heated will expand so as almost to burst; but on removing it into the cold, it condenses and shrinks to its former state.

Rarer and lighter bodies expand more than the heavier and denser; but this proportion does not always exactly correspond to the respective density and rarity of matters; for metals expand more than glass.

As an exception to this doctrine it has been observed, that water suddenly swells in passing from a fluid to a solid state : but this is to be explained upon different principles. Boyle took a brass tube, three inches in diameter, and put some water into it; he then brought down into the tube a plug with a weight placed at the head of it of 74lb., exposing the tube to the cold, and the water freezing and expanding itself raised the 74lb. The Florentine academicians filled a brass globe with water, closing the orifice by a well-fitted screw, and immersed it in freezing water; but as the sides were too thick, it did not burst. They then pared off such a quantity of the brass as left the sides of the globe unable to resist the expansion of water; the force which

1

was required to burst the globe in this state was computed at upwards of 27,000lbs. Boyle says the expansion of water in freezing is about an 8th or 9th of its bulk. The bursting of leaden pipes placed near or within the earth's surface proceeds from the same cause. The pavement even suffers from the frost, which swells the earth and loosens the stones; nay, rocks have been known to burst in frosty weather. Frost is by some supposed to fertilize, by loosening the cohesion of the particles of earth.

As ice is never clear or transparent, and as we find several cavities in it, some have thought that the air insinuates itself, and in this way have endeavoured to account for the expansion. But this has been refuted by water being frozen under an exhausted receiver, and the same cavities being found in the ice; the ice, indeed, instead of being heavier, was lighter, and floated on the water. M. D. Mairan at length solved the difficulty. He says, the particles of water in freezing assume a different arrangement, are not in so close contact, and cut each other at angles of 60". If this is the case, then, as indeed is now generally agreed, we cannot say with propriety that the solid particles of water expand, but that from their crystallization into the form of ice they require more room, or occupy a greater space.

As salts are observed in crystallization to put on regular figures, it was thought that the starlike appearance of snow was owing to a salt

mixed with the water; but Margraaff proved that snow is in fact composed of the purest water.

Reamur observes, that melted cast iron, in passing from a hot to a cold state, expands. This effect is more sensible in this than in any other of the metals, on account of its platey texture. He found also that cast iron thrown among some of the same metal in a melted state swims upon the top. In this case of immediate expansion upon congealing, the iron seems to agree with water. But they differ in this, that the iron never expands by cold afterwards, whereas the ice, being exposed to greater degrees of cold, becomes still more bulky; the solid parts not being so closely connected form a particular arrangement, which renders the whole mass specifically lighter than before.

Denser bodies for the most part expand less than rarer; but this I observed is not an invariable rule, for metals expand more than glass or stones. The expansion of metals was found to be a great obstruction to the regular going of clocks, but is obviated now by the ingenious contrivance of making use of two different metals, which do not expand equally, in constructing the pendulums.

Of the three classes of bodies which compose the universe, solid bodies suffer the least expansion from the presence of caloric. Liquids are more expansible than solids; and aëriform fluids. are most expansible of all. By the accurate ex

periments of Mr. Dalton of Manchester, it was found that all gasses, or aëriform fluids, undergo the same expansion by the same degrees of heat, supposing the circumstances the same. It may

be useful to note, that he found by his experiments that 100 parts of air by being heated from 55° to 212' expanded to 132. The steam of water and the vapour of ether undergo the same expansion with air when the same addition is made to their temperature: hence all elastic fluids may be said to expand equally and uniformly by heat.

What is remarkable, bodies, though expanded by caloric, never suffer any increase of weight from its presence. Muschenbroeck says that iron when heated loses 1-100th part of a grain in the cubic inch, and the same quantity of lead loses 4 grains, but when cool they return to their former weight. This loss of weight, however, is rather owing to the rarefaction of the air round the scale; the pressure of the atmosphere on the scale being lessened, the metal weighs lighter; the heat might also expand the side of the beam, and render it false.

When we speak of the specific gravity of bodies, we ought to mention the degrees of heat; for the specific gravity is different in winter and in summer, and this is owing to the expansion of the body by the presence of a larger quantity of caloric. Boerhaave supposes that air, which

is the lightest body, might be made to assume the density of gold.

In the second place, I have to notice the distribution of caloric or heat. That substances, when heated above the temperature of those bodies which surround them (a bar of iron for instance), soon part with their superfluous caloric (in other words cool) is well known and in like manner a cold body introduced into a warm place soon acquires the temperature of that place. The distribution of heat is more rapid in proportion to the contact of parts; hence the excessive slowness with which heat is communicated to those bodies which are rare and spongy, and hence its celerity in hard bodies. If a rod of iron is put into the fire for a little time, the end which is at a good distance from the fire will almost burn the hand, but a stick will be burnt to ashes before the other end is heated. We find caloric greatly retarded by cork, and still more by feathers and wool. Hence arises the distinction of good and bad conductors;the iron is a better conductor of caloric than the wood, and the wood is still better than cork or feathers. The difference in the cooling of hot bodies depends much upon their goodness as conductors. When mercury and water, each heated to the same degree, are placed in similar circumstances, the mercury cools twice as fast as the water. Straw, flannel, and feather-beds are