and furnished with a wire-work cover. The amount of evaporation is ascertained by filling the dish to the point of overflowing, which is, let us suppose, 3 inches of water. An ordinary rain-gauge goes along with it for the purpose of ascertaining the amount of rain which may fall. In making an observation the water remaining in the dish is measured, together with any that may be in the receiving bottle, and also the depth of rain fallen. Add the rainfall to the 3 inches, then add the water remaining in the dish to that contained in the bottle; and the difference of the two sums is the amount evaporated. The best evapometer is one invented by Dr Arthur Mitchell. It consists of an evaporating dish, the losses in which are supplied as they occur from a reservoir, whose area is ten times, or any number of times, less than that of the evaporating dish, the principle being the same as that involved in the fountain inkstand, or the fountain drinking-cup of the bird's cage. The water in the dish is thus always kept at the same level, and any rain that falls is at once carried off by an overflow pipe, so constructed as to prevent any error arising from the wavelets raised by the wind on the surface of the evaporating dish. The advantages of this instrument consist in the rain and dew being at once conveyed away, and the facility with which minute quantities can be read off. As the reservoir is liable to be broken in the time of frost, a modification of this evapometer has been devised by Mr James Procter, Barry, one of the ablest observers of the Scottish Meteorological Society. His evapometer presents a surface of 10 inches square, with an overflow pipe at the level of the zero line. The small quantity daily evaporated is measured by a scale of brass divided into 10 equal parts placed diagonally in the evaporation-vessel, so that during the evaporation of one inch of water, the line of contact of the surface of the water with the diagonal scale will traverse the whole length of the scale. Since each of the 10 parts may be subdivided by the eye into 10 parts, the 100th part of an inch may be easily read on the scale.

310. There is another evapometer, or Atmometer, fig. 23, which from its simple construction will be found to possess some practical value. It consists of a long glass tube graduated into inches, having attached to the bottom a hollow ball of porous earthenware similar to that used in water-bottles. In using it, water is poured in at the top till it rises to the zero point of the scale. The outside of the porous ball being always covered with dew, the more rapid the evaporation, the more quickly will the water fall in the tube.

311. There is no class of observations which show such diversity, we may almost say such contrariety, of results, as those made by different observers on evaporation. This arises from the different methods employed. The object sought to be obtained is the drying power of the atmosphere, a question of prime importance in its relations to the animal and vegetable kingdoms. And since none of the ordinary meteorological observations enable us to arrive at this element with any approach to Fig. 23. exactness, it is hoped that some one evapometer, such as Dr Mitchell's, will come to be generally used among observers.

312. Loss of Heat by Evaporation.-One of the most important consequences of evaporation is the loss of heat which accompanies it. During the conversion of a liquid into the gaseous form, a very large quantity of heat disappears; and since it becomes imperceptible to the senses or to the thermometer as long as the gaseous state is retained, the heat is said to become latent. It must be kept in mind that the heat which thus appears to be lost, is not destroyed, but may be recovered or made evident at pleasure by bringing back the vapour to its original liquid state. In the gaseous state the force of the heat is expended in keeping the particles of vapour further apart than they were in the liquid state, and hence the thermometer is not affected by it. The

change of water to vapour by evaporation being thus productive of cold, and the conversion of vapour to water productive of heat, one or two important consequences follow. The ocean loses more heat from evaporation than the land, because the quantity evaporated from its surface is much greater. Again, since more rain falls on land than on sea, especially in hilly and mountainous countries, the temperature of the air over the land will be still further raised by the latent heat thus given out. It is for this among other reasons that the mean temperature of the northern hemisphere is higher than that of the southern hemisphere.

313. Effect of Drainage on the Temperature of the Soil. -Theory should lead us to suppose that the temperature of drained land would be higher than that of undrained land, because, being drier, less heat is lost by evaporation. In order to bring this supposition to the test of experiment, the Marquess of Tweeddale in 1862 offered prizes to the amount of £80 for sets of observations on the temperature of drained and undrained land on mountain pastures, and on arable land under crops of turnips, wheat, and ryegrass. Two sets of observations, each extending over a year, were made, one at Otter House in Argyle, on arable land under a ryegrass crop, the soil being light and sandy and sloping 1 in 40 feet; and the other set at North Esk Reservoir, Pentland Hills, on hill pasture, the soil being clay mixed with decayed moss. From these valuable observations the following results were drawn: 1. The mean annual temperature of the arable land was raised nearly a degree (0°.8) by drainage. 2. The temperature of the hill pasture was also raised by drainage, but not to the same extent (0°.4). 3. During sudden falls of temperature and during protracted cold weather, such as when the soil was under a covering of snow, the cold passed more quickly and completely through undrained than through drained land. 4. When the temperature of the air was higher than that of the soil, drained land received more benefit from the higher temperature than undrained land, less of its heat being lost by evaporation. 5. Since, when

rain or sleet fell, the superfluous moisture soon flowed away from the drained land (it being sandy and the slope considerable), drainage tended to maintain in the soil a comparatively equable temperature; whereas the undrained land. was liable to considerable fluctuation, for when soaked with warm rain-water its temperature was temporarily raised, and when soaked with melted snow it was temporarily lowered. On one occasion sleet showers lowered the drained land 2° and the undrained land 4°. 6. The temperature of drained land was in summer occasionally raised above undrained land 3o, often 2°, and still more frequently 1°.5; and hence the beneficial effects of drainage are sometimes as great as if the land had been transported 100 or 150 miles southwards.

314. Since these different temperatures are chiefly caused by the different amounts of water evaporated from the land, it is evident that different results will be obtained from different soils, with different crops, and with different slopes and exposures.

315. In 1847 Professor James Elliot made a number of experiments which shed some light on this extensive subject. He found that peat-moss can absorb more than twice its own weight of water, dry clay nearly its own weight, dry earth or garden mould more than half its own weight, and dry sand little more than a third of its own weight. With equal times of drying under the same circumstances, peat-moss lost of all the water it contained, clay and earth each more than, and sand more than . Evaporation was greater from the surface of loose earth than from the surface of water, till the earth became so far dry as to be of a light colour. Evaporation from saturated moss was excessive during the first day, being far more than from the surface of water; but on the second day the water began to evaporate more, and on the third day very much more than the moss, although the moss was still wet 10 inches below the surface.

316. Three years ago D. Milne Home of Wedderburn made the following experiment: Two boxes of the same size were taken and filled, one with sandy loam, and the other

with strong clay.

Each was suspended at the end of a balance, and so adjusted that the one box was exactly equal in weight to the other. An equal quantity of water was poured into each box. Before a week had elapsed, the box with the sandy loam rose above the level of the other box, showing that more water had evaporated from it than from the strong clay, and during that time the temperature of the sandy loam was almost always lower than that of the clay. 317. From these experiments a few practical conclusions may be drawn. In all cases the amount evaporated from wet substances, and the consequent decrease of temperature, are proportioned to the number of evaporating points, or to the whole extent of the evaporating surfaces in contact with the air. This explains why evaporation is greater from wet moss and grass than from wet soils, and greater from wet soils than from a surface of water. But as evaporation proceeds and the substances begin to dry, the rate of evaporation is modified by the facility with which the water is drawn by capillary attraction from the interior of the substances to their evaporating surfaces. Thus dry sand parts with its moisture sooner than common mould, common mould sooner than clay, and clay sooner than peat-moss. In respect of evaporation, drainage affects the temperature of the soil in two ways: (1) By keeping the soil drier, and thus diminishing the evaporation, it maintains a higher temperature. (2) Since a dry soil is more friable than a wet soil, and thus presents more evaporating points, it is probable that on soils bare or nearly bare of vegetation, and therefore freely exposed to the influence of the weather, the effect of drainage will be to lower the temperature for some time after rain, or as long as the evaporation exceeds that of undrained land. It is evident that when the soil is covered with vegetation this peculiarity can only obtain to a very limited degree.

318. In regard to the experiments with moss, and grassy turf which resembles moss, and their bearing on the important question of climate, Professor Elliot remarks: "We cannot doubt that the destruction of the moss adds not only