The crumbling away of rocks and of soils from this cause has already been alluded to in the Introduction. Water is absorbed by various rocks by means of capillary attraction, and when winter comes on and the water freezes, it forms myriads of little wedges, which separate small fragments, and when a thaw comes, and the rain falls, these are detached and washed down into the lower valleys, where they form the rich fertilizing mud of rivers, and contain the mineral food that trees and plants require.

During a very severe winter, the sap of living trees may freeze, and the trunks burst asunder, with a loud cracking report, in consequence of the expansion.

The same cause which produces the crumbling or disintegration of rocks, produces the decay of buildings. The faces of the stones exposed to the weather absorb moisture, and this, by freezing, detaches portions from its surface, which fall or are washed off by the next thaw or rain, thus effacing all the chisel-marks, rounding the sharp angles of cornices and the well-defined outlines of ornaments, and roughening the stone, so that it retains moisture even more readily than before, and thus hastens the process of decay. That state of things may be prevented by selecting a stone which does not absorb moisture. Absorptive and nonabsorptive stones occur in the same quarry; even altering the position of a stone in the building from what it had in the quarry may render it absorptive.

To ascertain whether a stone is absorptive or not, M. Brard's plan is to boil specimens of the stone in a saturated solution of Glauber salts (sulphate of soda). The specimens are then taken out of the solution, hung up in the air, and as they dry the salt crystallizes, and pretty accurately represents the behaviour of freezing water in detaching particles of stone from the surface. Those stones which are absorptive of water, also absorb the solution, and will lose most in weight by the splitting off of fragments during the solidification of the salt. The less absorptive stones lose less from this cause, and of course are to be preferred in building.

The expansion of water in freezing causes the mortar of walls when fresh and moist to peel off; so that the operations of the mason and the bricklayer are suspended during frost; and if a frost be expected, they cover up fresh work with straw or matting, to prevent it from being spoilt. This force of expansion also loosens the flag-stones of our pavement, when water has got beneath them. It cracks our jugs and water-bottles, and bursts our service water-pipes—at least that part of them which is exposed to the air, since the part buried in the ground is protected from the action of the frost; hence it is usual to wrap up the exposed portion in a hayband, or other badly-conducting substance. Another precaution, during frosty weather, is to leave the tap a little way open, so that water may constantly trickle from it: in this way the pipe

may be preserved, on the principle that running water does not freeze so easily as stagnant.

When water-jugs and bottles are broken by the frost, the ice first forms on the surface, and prevents the water below from expanding; so that in many cases the fracture is occasioned by fluid pressure, as well as by the expansion of the ice.

When a pond is completely frozen over, the fishes are imprisoned in the water beneath, and are cut off from the supply of air which is generally supposed to be necessary to their respiration. When a hole is made in the ice, fishes will eagerly crowd to it, to obtain their necessary supply of air. But in ponds and lakes far from the abode of man, these holes cannot be made; and hence it has been doubted whether fishes really require such a provision to be made for them. Mr. Swainson, in his "Natural History of Fishes," says that, on the breaking up of the long and severe frost of the winter of 1837-38, he had the mortification of seeing the dead bodies of between thirty and forty fine tench floating on the surfaces of a pond in his garden. The pond was of rain-water, with a soft muddy bottom, from two to four feet deep, and fringed with aquatic plants, so that there was abundant shelter; but the very severe cold was supposed to be the cause of the mortality. The people about the place gave it as their opinion, that had holes been broken in the ice, the fish would not have perished. Mr. Swainson doubts the necessity of this in all cases, since

the tench of all such ponds as had not been opened would have been killed. On the other hand, Mr. Boccius, in his treatise "On the Management of Fresh-water Fish" (1841), gives particular directions for making such openings in the ice. He says:

"Should the winter prove very severe, when the surface of the store ponds is completely covered with ice, the best plan is to cut round holes, about three feet in diameter, and to place therein a bundle of withes or straight bushes, about six feet long, so that the points are uppermost, and about three feet out of the water; this will leave a sufficient space open for air, as the thickness of the bushes will not allow the centre to freeze, and thus the fish will become relieved: as air is very essential for the stock, all ponds ought to be ventilated on this plan during frost."

Apparently in opposition to this, Mr. Jesse relates the case of a golden carp being frozen hard in a glass vase in which it was kept; and yet, when the ice had been gradually thawed, the fish revived, and became as lively as usual. Dr. Richardson also states that the grey sucking carp of North America, when frozen hard enough to be brittle, will, if thawed, become quite active.

The fact is, there are great differences in the vitality of fishes, and their avidity for air. Cuvier states that some species eagerly inhale the air at the surface; and if prevented from getting to it by the interposition of a gauze screen, they die of suffocation. Mr. Yarrell is still more precise. He

says that those fishes that swim near the surface `have a high standard of respiration and a low degree of muscular irritability: they have a great need of oxygen, and die almost immediately when taken out of the water-their flesh also rapidly decays: such are the mackerel, the salmon, the trout, and the herring. On the other hand, those fishes that live at or near the bottom have a low standard of respiration and a high degree of muscular irritability: they have less need of oxygen, and live long out of the water; their flesh, also, will remain good for several days. Such are the carp, tench, eels, skate, and flat fishes generally. Indeed, in the markets of the North of Europe, eels and perch in a frozen state are constantly to be seen; and it is a familiar fact that they revive on being thawed.

It appears, then, that fishes vary greatly in their appetites for air; but it must be remembered, that in artificial fish-ponds there is a greater crowding of fish than in a lake where the fish are not preserved. Hence the pond may require to be kept

*

On visiting the market place of Cologne one summer's morning, we noticed tubs of water containing living carp for sale. The poor creatures were very much crowded, and were evidently suffering from want of air; they did not attempt to swim about, but kept in one position with open mouths, just above the surface of the water. We were informed that the fishes that were not sold were returned to the stews in the evening, where they remained until the next market day.

A friend of ours formed an aquarium of a tall glass jar, which was kept in the window of a room in a noisy street in London. The aquatic plants grew very well, but the fishes were constantly