alternations of heat and cold are still further intensified by the great dryness of -the air over extensive tracts of sand. In warm countries the surface temperature of sandy deserts often rises to 120, 140, or even to 200, and the shade temperature has been observed as high as 125. It is this hot air, loaded with particles of sand still notter, and driven onwards by furious whirlwinds, which forms the dreaded simoon of the desert ; and the irritating and enervating sirocco of the regions bordering the Mediterranean is to be traced to the same cause. It is in the deserts of Africa, Arabia, Persia, and the Punjab that the highest tempera ture on the globe occurs, the mean summer temperature of these regions rising to and exceeding 95. The extreme surface of loam and clay soils is not heated during day nor cooled during night in so high a degree as that of sandy soils, because, the former being better conductors, the heat or the cold is more quickly conveyed downward,
and therefore not allowed to accumulate on the surface.When the ground is covered with vegetation the whole of the sun s heat falls on the vegetable covering, and as none of it falls directly on the soil its temperature does not rise so high as that of land with no vegetable covering. The temperature of plants exposed to the sun does not rise so high as that of soil, because a portion of the sun s heat is lost in evaporation, and the heat cannot accumulate on the surface of the leaves as it does on the soil. Hence the essential difference between the climates of two countries, the one well covered with vegetation, the other not, lies in this, that the heat of the day is more equally distributed over the twenty-four hours in the former case, and there fore less intense during the warmest part of the day.
But the effect of vegetation on the distribution of the temperature during the day is most markedly shown in the cuse of forests. Trees, like other bodies, are heated and cojled by radiation, but owing to their slow conducting power the times of the daily maximum and minimum temperature do not occur till some hours after the same phases of the temperature of the air. Again, the effects of radiation are in the case of trees not chiefly confined to a surface stratum of air a very few feet in thickness, but as already remarked, are to a very large extent diffused through a stratum of air equalling, in thickness at least, the height of the trees. Hence the conserving influence of forests on climate, making the nights warmer and the days cooler, imparting, in short, to the climates of districts clad with trees something of the character of insular climates. Evaporation proceeds slowly from the damp soil usually found beneath trees, since it is more or less screened from the sun. Since, however, the air under the trees is little agitated or put in circulation by the wind, the vapour arising from the soil is mostly left to accumu late among the trees, and hence it is probable that forests diminish the evaporation, but increase the humidity, of climates within their influence. The humidity of forests is further increased by the circumstance that when rain falls less of it passes immediately along the surface into streams and rivers; a considerable portion is at once taken up by the leaves of the trees and percolates the soil, owing to its greater friability in woods, to the roots of the trees, whence it is drawn up to the leaves and there eva porated, thus adding to the humidity of the atmosphere.
Much has been done by Ur Marsh and others in elucidation of the influence on climate of forests and the denudation of trees, in so far as that can be done by the varying depths of lakes and rivers and other non- instrumental observations. Little comparatively has been done anywhere in the examination of the great practical question of the influence of forests on climate, by means of carefully devised and conducted observations made with thermometers, the evaporating dish, or the rain gauge. The most extensive inquiry on the subject yet set on foot has been for some years conducted in the forests of Bavaria under the direction of Professor Ebermeyer, and a like inquiry was begun in Germany in 1875, the more important results being that during the day, particu larly in the warm months, the temperature in the forest is considerably lower than outside in the open country, there being at the same time a slow but steady outflow of air from the forest ; and that during the night the tempera ture in the forest is higher, while there is an inflow of air from the open country into the forest. The mean annual temperature in the forest increases from the surface of the ground to the tops of the trees (where it is observed to approximate to what is observed in the open country), a result evidently due to the facility of descent to the surface of the cold air produced by terrestrial radiation, and to the obstruction offered by the trees to the solar influence at the surface. The mean annual temperature of the woodland soil from the surface to a depth of 4 feet is from 2 to 3 lower than that of the open country. A series of observations was begun at Carnwath, Lanarkshire, in Ih73, at two stations, one outside a wood, and the other inside the wood in a small grass plot of about 50 feet diameter clear of trees. From these valuable results have been obtained relative to the differences in the daily march of temperature and the different rates of humidity, the most important being the substantial agreement of the mean annual temperature of the two places. The estab lishment of a station, with underground thermometers, which it is proposed to erect under the shade of the trees close to the station in the cleared space, will furnish data which will not only throw new light on the questions raised in this inquiry, but also on the movements and viscosity of the air and solar and terrestrial radiation.
When the sun s rays fall on water they are not as in the case of land arrested at the surface, but penetrate to a considerable depth, which, judging from observations made by Sir Robert Christison on Loch Lomond, and from those made on board the " Challenger," is probably in clear water about 600 feet. Of all known substances water has the greatest specific heat, this being, as compared with that of the soil and rocks composing the earth s crust, in the proportion of about 4 to 1. Hence water is heated much more slowly by the sun s rays and cooled more slowly by nocturnal radiation than the land. It is owing to these two essential differences between land and water with respect to heat that climates come to be grouped into the three great classes of oceanic, insular, and continental climates.
The maximum densities of fresh and salt water, which are respectively 39"1 and 26 2 (when the sea-water is the average degree of saltness), mark an essential distinction between the effects of sheets of fresh and salt water on climate. The surface temperature of sea-water falls very slowly from 39 0l l to 28 4, its freezing point, because as it falls the temperature of the whole water through its depths must fall ; whilst from 39 l to 32 the surface temperature of fresh water falls rapidly because it is only the portion floating on the surface which requires to be cooled. If the bottom temperature of fresh water exceed 39 l the cooling takes place also very slowly, since in this case the water through all its depth must be cooled down to 39 l as well as that of the surface.
which is practically constant at all seasons, is not 47 8, the mean annual temperature of that part of Scotland, but 42", which happens to be the mean temperature of the cold .half of the year, or that half of the year when terrestrial radiation is the ruling element of the tempera ture. Thus, then, there is an immense volume of w T ater
at the bottom of this lake at a constant temperature 5 8