rose to the surface, surrounded with a thick coat of ice. A cable also, three and a half inches thick, and about thirty fathoms long, which had been lost during the preceding summer, in a depth of thirty feet floated to the surface, enveloped in ice to the thickness of two feet. An anchor which had rested only an hour at the bottom, became so encrusted with ice, that it required not more than half the usual power to haul it up.

On the river Aar, a scientific observer noticed a quantity of floating ice appearing on the surface at a time when there was no sign of freezing, either along the banks, or in shady places, where the water was calm. It could not therefore be said that the floating masses were formed at the banks; nor could they have floated down the stream, because, higher up, the river was remarkably free from ice. Flakes of ice were also seen to ascend and form ice islands in the centre of the stream; so that in all, there were twenty-three of these islands, the largest being upwards of two hundred feet in diameter. These were visited in a small boat, and the surface of each island was found to be a layer of compact ice, resting on a sort of inverted cone of spongy ice, which was fixed in the bed of the river.

The Rev. James Farquharson describes, in the Philosophical Transactions of the Royal Society, for 1835, the formation of ground-ice in the rivers Don and Leochal, in Aberdeenshire. He found that when it begins to form on the bottoms of the

streams, it has a certain kind of orderly arrangement, reminding one of little hearts of cauliflowers shining with a silvery lustre; at first they are scattered irregularly, and in small numbers, but soon increase both in size and numbers, till the whole bed of the stream is covered. If the frost continue to be severe, they grow in height, but in a very irregular manner, until the streams are raised high above their former levels, so as frequently to overflow their banks. The first appearance of the ground-ice is that of lumps of snow floating in water, which is called gru by the country people; the ice at the bottom of the stream just described, is therefore called groundgru, and is justly dreaded in a district where streams are common, and many of them passable only by fords. Many horses refuse to enter a stream even slightly impeded by it, being greatly alarmed by the pieces which break and float up from the bottom by the action of their feet. This substance does not begin to form until the whole mass of the water is at, or about the freezing point, the air at the time being much colder. A clear state of the sky is most favourable to its formation. If the sky continue to be clouded during many hours, the ice becomes detached from the bottom, and floats down the stream.

The Rev. Mr. Eisdale has noticed* the formation of ground-ice in the neighbourhood of Perth. According to him, it is well known in all northern * Edinburgh New Philosophical Journal, vol. xvii. First Series.

climates, by its annoying effects in obstructing works which are carried on by the impelling power of water; for the streams are sometimes gorged up from the very bottom. In the south of Scotland, this kind of ice is called lappered-ice; the word "lappered" being commonly applied to the natural coagulation of milk. A miller, in the western part of Scotland, related to Mr. Eisdale a curious instance of the formation of ground-ice. During a severe frost, when the mill-lead was entirely free from any kind of ice, the miller had occasion one day to lop some branches from a tree, which overhung the lead; one of them fell into the water and was left there, as he did not apprehend any ill consequences from so trifling an occurrence. Next day, however, to his astonishment, the water was turned entirely out of the lead, and had flooded an adjoining meadow. On proceeding to ascertain the cause, he found that a solid barrier of ice had been formed across the lead, where the branch had fallen in, so as completely to prevent any water from passing, while the rest of the lead was free from ice.

The beds of rivers often become greatly worn and deepened by the formation of ground-ice, and large portions of the soil are floated away, and deposited in places, where it would otherwise be difficult to explain their presence. Mr. Weitz has noticed this action in traversing some of the rapid rivers of Siberia, at the beginning of winter. These rivers flow with great rapidity over a sandy,

or stony bed; and although much snow falls, and the cold is severe, and of long duration, they continue to flow, bearing along vast quantities of floating ice. The great transparency of the water enables any one to see the ice at the depth of fourteen feet. It had a greenish tinge, and looked not unlike moss. When it separated from the bottom, and rose to the surface, it soon grew more compact by contact with the cold air, and floated. away with the flakes detached from the sides. The pieces which rose from the bottom brought up sand and stones, which were thus transported by the current. On arriving at those parts of the river where, from the small slope of the bed, the motion of the water is slow, and the surface sometimes frozen over, the floating masses of ice collect and get fixed; thus forming a barrier to the icemeers that come down, and contributing by this means to the congelation of the whole surface of the river. When a thaw sets in, the ice lets fall the gravel and stones, in places far distant from those whence they came

The cause of ground-ice has been a good deal discussed. According to Dr. Farquharson, it is the result of radiation, and consequently depends upon the same principle as the formation of dew. It was objected to this theory, that as wind and shade are opposed to radiation, so a body of moving water, so deep as to be impervious to light, (and in the case of the river Neva, covered with a sheet of ice three feet thick, and a deep layer of snow

H

upon that), could not radiate heat from its bed. To this it was replied, that although there may be no radiation of heat into space from the bed of the river, there may be radiation from the bed of the river to the icy covering above. Thus, Dr. Hooker, while travelling in the narrow upper valleys of East Nepal, at an elevation of 8,000 feet, found the nights so brilliant, and the radiation from the earth so powerful, that the upper blanket of his bed was coated with dew, in consequence of the rapid abstraction of heat, by its radiation to the tarpaulin of his tent, the outer surface of which was covered with hoar frost, by its radiation to the sky. In this case, the tarpaulin did not prevent the blanket from cooling by radiation, and condensing the moisture of the air upon it. Had the temperature been somewhat lower, the dew on the blanket would have become hoar frost, and then we should have had a parallel case to that of the Neva, and a good illustration of Dr. Farquharson's theory.

Mr. Eisdale refers the formation of ground-ice, to the frozen spicule of the atmosphere falling into the river and forming nuclei, around which the water freezes at the bottom. This explanation is full of difficulties. M. Arago refers the formation of ground-ice to the fact, that water at 32°, at rest, or in very slow motion, as at the bottom of rivers, will freeze. He supposes that the motion of the current mixes up the various strata of water, which, in a lake or pond, would remain quite distinct, the