described by Sir R. Barker, and afterwards by Mr Williams. It was a profitable business, and three hundred persons were employed in it. A piece of ground, nearly level, was divided into square plots, from 4 to 5 feet wide, which were surrounded by little mounds of earth, 4 inches high. In these enclosures, previously filled with dry straw or canes, many broad shallow unglazed earthen pans were placed, containing pump-water. When the air was still much ice was formed; wind prevented its formation altogether. The object of the excavations and little mounds of earth was to increase the stillness of the air. Since, then, the evaporation must have been small when ice was formed most plentifully, the cooling of the water and the formation of the ice was caused by terrestrial radiation. The object to be obtained in using dry straw was to secure for the pans a badly conducting substance; when the straw was wetted, and its conducting power thus increased, little or no ice was formed.

211. Effects of Terrestrial Radiation on Land.—One of the chief effects of terrestrial radiation on the air superincumbent over land, as compared with that over the sea, is to increase the range of the temperature. Thus, while the mean daily variation in the temperature of the air over the sea in the west of Scotland is about 6°.0, on the land it is about 12°.0; and in extreme cases the variation on land is three or four times the variation on the sea.

212. Increase of Temperature with the Height during Cold Weather. This takes place invariably in dry, calm, clear weather, during the night, and in winter, when the temperature of the air is lowered by contact with the chilled surface of the earth. Between the temperature of the air in contact with the surface of the ground, and the air 4 feet above the ground, the difference is frequently 15° or 20° during the night, but above 4 feet the differences are comparatively small. On the 8th April 1844, at 8 P.M., Mr Glaisher observed a difference of 25°.0 between two thermometers, one placed on raw wool over long grass, and the other at a height of 4 feet; and at 8 feet high the temperature was

3°.5 warmer than at 4 feet. Hence the temperature of the air on the ground was 28°.5 colder than at a height of 8 feet at the same time. From an extensive series of observations made by the same indefatigable meteorologist with thermometers fully exposed to the sky, the means, as compared with those on long grass, were as follow :—

A thermometer at 4 feet above long grass, protected from sixtenths of the sky, stood, on a mean, 8°.39 higher. Hence the necessity of having all thermometers in shade placed at the same height above the ground, for otherwise the results are not comparable. From experiments made by M. Ch. Martins in the south of France, he found, on a mean of 88 nights, that the temperature at a height of 6 feet as compared with 2 inches, was 1°.5 warmer; that for every yard between 20 and 96 feet, it was 0°.12 warmer; and between 96 and 162 feet, it was 0°.04 warmer. Professor Wheatstone's thermometer, described by him at the meeting of the British Association held at Dundee in September 1867, by which the temperature at the bottom of the sea might be read continuously, might, by means of fixed or captive balloons, be turned to excellent account in solving the important problem of the decrease of temperature with the height.

213. But the rate at which the temperature increases with the height is modified, to a very great extent, by the physical configuration of the surface-that is, whether that surface be level, undulating, or mountainous. To understand how this happens, let us suppose an extent of country diversified by plains, valleys, hills, and table-lands, to be in circumstances favourable to radiation, and each part under the same meteorological conditions, except in the single point of

position. Radiation will proceed over the whole at the same rate, but the effects of radiation will not be felt everywhere in the same degree and intensity. For as the air in contact with the declivities of hills and rising ground becomes cooled by contact with the cold surface of the ground, it acquires greater density and weight, and consequently slips down the slopes and accumulates in the low-lying ground at their base. Hence places situated on rising ground are never exposed to the full intensity of frosts; and the higher they are, relatively to the surrounding ground, the less are they exposed, because they are protected by their elevation, which provides, as it were, an escape for the cold almost as fast as it is produced. On the contrary, valleys more or less environed by hills or eminences, not only retain their own cold of radiation, but also serve as reservoirs for the cold of neighbouring heights. Hence low-lying places are peculiarly exposed to intense cold. Plains and table-lands are simply affected by their own radiation.

214. This explains why vapour becomes visible so frequently in low places, whilst adjoining eminences are clear; and the same fact instinct has made known to cattle and sheep, which generally prefer to rest during night on knolls and other eminences. Along most of the watercourses of Great Britain, during the memorable frost of Christmas 1860, laurels, araucarias, and other trees growing below a certain height were destroyed, but above that height they escaped; thus attesting, by unmistakable proof, to the great and rapid increase of the temperature with the height above the lower parts of the valleys.

215. Distribution of Cold in Mountainous Countries during Winter. From the above remarks, it is evident that the distribution of temperature over the surface of a mountainous country during calm weather in winter, will be regulated by a different law than that of height above the sea. For the illustration of this point, it is fortunate that sixty-nine meteorological stations were established in Switzerland in 1863; and as one of the firstfruits of this Society, a paper has been

lately published by Professor E. Plantamour, on the Distribution of Temperature on the surface of Switzerland during the winter of 1863-64. Whenever the soil is colder than the air above it, the superficial layers become cold by contact, as already explained, and a system of descending air-currents sets in over the whole face of the country. The direction and intensity of these descending currents are modified by the irregularities of the ground, and, like currents of water, they tend to converge and become united in the gorges and ravines, down which they flow like rivers in their beds. These currents give rise to counter-currents flowing over them to supply their place.

216. When the station is on the top of a mountain, as the Righi, the counter-current comes from a great height above the ground, and being therefore warmer, the temperature of such stations is comparatively high. At places situated on the sides of mountains, the influence of the counter-current tends to raise the temperature, though in a less degree than at the top, on account of the descending current from the heights above mixing with it. The Swiss villages being generally built on eminences rising out of the side of the mountains, and bounded on both sides by gorges and ravines, are admirably protected from the cold of winter. For the descending currents flow aside into the gorges, and the counter-currents are constantly supplying warmer air from the upper regions of the atmosphere.

217. Though the space occupied by the current of cold air in the bottom of a valley is of greater extent than the bed of a river, it is nevertheless limited, and on all occasions tolerably defined, so that on rising above it in ascending the slope, an increase of temperature is readily perceptible. The gradual narrowing or contracting of a valley has a very appreciable influence in lowering the temperature; for the valley is thus transformed into a basin almost closed, into which cold currents of air descend from all sides. On such occasions, a cold wind rushes impetuously down the narrow gorge, which serves as an outlet to the basin; and it is on this principle that many

of the sudden gusts and breezes peculiar to mountain districts, such as vent du Mont Blanc, are to be explained. When the basin is a deep lake, the cold which is poured down on its surface, having cooled the surface-water, is thereby conveyed to greater depth, and has therefore scarcely any effect in lowering the temperature of the air resting over the lake. Hence lakes are a source of heat during winter, and places situated at their outlet are not exposed to such gusts of cold wind as those referred to above.

218. Influence of Forests.-The temperature is found to be warmer at the base of a mountain, and up its sides, when the slopes above are covered with trees. The beneficial influence of forests is exerted in two ways-viz., in the diminished radiation from the surface protected by the trees, and in the obstacle they oppose to the descending currents of cold air.

219. On the contrary, the cold of winter is more severely felt in those localities where the slopes above are destitute of vegetation, and consist only of bare soil and rocks, or of

snow.

220. This peculiar distribution of the temperature only takes place during comparatively calm weather; during windy and stormy weather the law of the decrease of temperature with the height takes effect.

221. Situations which afford the best protection against the Cold of Winter.-In countries such as Great Britain, and, indeed, in most temperate countries, the majority of the deaths which occur are occasioned, or at least hastened, by low temperatures. In the tables of mortality and temperature, published weekly by the Registrars-General for England and Scotland, we have constant proof of this statement. For when, during the cold months of the year, the temperature happens to fall a few degrees, the death-rate at once rises to a height proportioned to the depression of the temperature. It is thus a matter of most vital importance, especially to invalids, to know the local situations which afford the best protection from low temperatures. From what has been already said regarding the increase of temperature with the