mometer with its bulb blackened. As in the case of the maximum black-bulb, it is laid on grass with its blackened bulb freely exposed, and in such a position as to command a view of the whole of the sky, or as much of it as possible. If it does not command a view of the whole, the amount cut off by surrounding objects should be noted. It should always be placed just above the ground, never on an object at some height above the surface; it should also be laid on grass kept short; and the blackened bulb should project a little beyond the scale, so as to be freely exposed all round. The reason for these restrictions will soon appear.*

199. The degree to which the temperature falls depends on the radiating and conducting powers of the surface over which the thermometer is placed, being greater as the radiating power is greater and the conducting power less, and vice versa. We give below the relative cooling powers of a few of the more important substances, as determined by Mr Glaisher from experiments made by him,-long grass being 1000:

200. One of the most instructive examples illustrative of this subject that could be given is the result of Mr Glaisher's observations on the different temperatures of long and short grass. A thermometer placed on long grass was found to be on a mean 1°.1 lower than one on short grass, whilst the temperature of the soil under long grass was 1°.1 higher than under short grass. The temperature was thus the same amount in excess under the long grass as it was in defect over it. Hence the difference of temperature over the long and the short grass was entirely due to the greater quantity

* But it must be conceded that all the means yet proposed for determining solar and terrestrial radiation are only approximations to a solution of the problem, and that no satisfactory method has yet been devised.

of heat conducted from the soil to the top of the short grass over that conducted to the top of the long grass, and not to any difference in the radiating powers of the grasses. The experiments were extended, and it was found that the temperature varied with every variation of length, fineness, and closeness of texture of the blades of the grass.

201. We thus see why it is that the temperature of grass and other plants falls during the night so much below that of garden mould, sand, and gravel. The great difference is not owing so much to different quantities of heat radiated into space, but to the ease or difficulty with which heat is conducted from the soil to the radiating surfaces so as to supply the loss sustained by radiation.

202. Dew.-When a glass of cold water is brought into a warm room the surface of the glass is soon covered with small globules of water, provided the temperature of the water be below the dew-point of the air of the room. This phenomenon takes place on a grand scale over the earth's surface every night when the atmosphere is comparatively clear and calm. For, as the earth is cooled down by radiation, as soon as the temperature of the dew-point is reached by any body, the vapour of the air begins to be condensed into dew on its surface. The quantity of dew which is deposited is in proportion to the degree of cold produced, and the quantity of vapour in the air. Hence, from what has just been said in par. 199, more dew will be deposited on furs, wool, silk, flax, and cotton, than on grass and vegetable substances generally, and less on glass, mould, sand, and gravel. Thus, by a most beneficient arrangement, dew falls most copiously on the objects of the earth's surface which most require its refreshing influence. It is not deposited in cloudy weather, because clouds obstruct the escape of heat by radiation into space; nor in windy weather, because wind constantly renews the air in contact with the ground, and thus prevents the temperature from falling sufficiently low. Dew is rarely deposited on the surface of deep water, because its temperature scarcely ever falls below the dew-point. When the

temperature is below the freezing-point, the dew freezes as it is deposited, and hoar-frost is produced. See par. 13.

203. Effects of Radiation on Water.-Water being a good radiant, heat is freely given off from its surface as well as from land. But from various causes the effects of radiation on the surface of water are very different from those on land. Owing to its great specific heat, its temperature falls much more slowly than that of land. This conserving influence of water on the temperature is greatly increased by another peculiarity. For when the particles floating on the surface are cooled by radiation they become heavier and sink, and warmer particles from below rise to supply their place. Thus the surface can only cool as the entire body of water is cooled, and the change of temperature occasioned in a body of deep water by radiation from the surface during a night will be almost imperceptible; consequently the temperature of the air resting on this water will undergo comparatively little depression on calm nights.

204. The descent of the cold, and ascent of the warm, particles of water, is a much slower process than is generally supposed. From observations made by Captain Thomas, in the ocean west of Scotland, during September and October 1858 and 1859—that is, during the great annual fall of temperature-it was found that the mean temperature at a depth of 24 feet was 1° warmer than at 1 foot, and about 0°.8 warmer at 60 feet than at 6 feet. This is instructive, and shows the value of observation over theoretical deductions; for theory might have led us to suppose that the warmer water would have risen more readily to the surface. At this season the water of the sea is not in a state of equilibrium, a disturbing agent being at work (the cooling of the surface water), causing a continual descent of the colder and ascent of the warmer particles; and this process goes on as long as the surface is colder than any of the underlying watery strata. The difference between the temperature of the surface and that at several fathoms below the surface, shows, in a forcible manner, the length of time required to effect a complete

correspondence between the temperature of the sea and that of the air; for, being brought about by upward and downward currents, the difference of whose temperature is small, the motion of the currents and the interchange of temperature resulting therefrom proceed at a very slow rate.

205. The mean daily range of the temperature of the sea has been ascertained off the Scottish coast from Captain Thomas's observations, already referred to. On a mean of

Hence the temperature of the sea near the surface only varies on the average about 0°.6 in the day, while in Scotland the air varies 12° on the average. The greatest differences observed on any day in the temperature of the sea were 5°.6 and 5.3, amounts especially exceptional; whilst the temperature of the surface of the land varies not unfrequently as much as 100°. These figures present in perhaps the most striking light the conserving influence of the ocean on climate.

206. It is a well-known law of matter that bodies expand with heat and contract with cold; but to this law water is one of the remarkable exceptions. Water follows this natural law till it falls to 39°.2, after which, as the temperature falls, instead of contracting it begins to expand, and continues to do so till it reaches the freezing-point. Water is, therefore, at its maximum density at 39°.2. When it is just falling to 32°, or immediately before it freezes, it occupies as great a space as it did at 48°. But if common salt be dissolved in water, the temperature of its maximum density is lowered; and if it be brought to the average degree of saltness of seawater, its maximum density would be at 27°.2, if it did not freeze before reaching that point.

207. Since, then, sea-water of the average saltness follows the general law in contracting as it is cooled till it freezes, no ice can be formed on its surface till the temperature has

G

fallen through all its depths nearly to freezing. On the other hand, as soon as the temperature of fresh water has fallen to 39°.2 through all its depths, the surface-water becomes lighter as it gets colder, and consequently no longer descends, but floats on the surface. This circumstance marks an essential distinction between the effects of sheets of salt and fresh water respectively on climate. The surface temperature of sea-water falls very slowly from 39°.2 to 32°, because as it falls the temperature of the whole water through its depths must be reduced, whilst from 39°.2 to 32° the surface temperature of fresh water falls rapidly, because it is only the portion floating on the surface which requires to be reduced.

208. Difference of Shallow and Deep Lakes on Winter Climates. Owing to its great depth, Loch Ness, in Inverness-shire, is not known to freeze. Its temperature being therefore always high, even during intense and protracted frosts, the winter climate along its shores is mild and partially insular in its character. So high is the temperature of the lake that the Ness does not freeze in its short course to the sea; thus presenting so remarkable a contrast to the icebound streams on each side, that an oily element is vulgarly attributed to its water, preventing it from freezing. On the other hand, Loch Leven, in Kinross-shire, being shallow, is easily frozen over, and hence it has no effect whatever in moderating the rigours of intense frosts on its banks.

209. As these lakes are small, their influence operates but to a very limited extent. It is in the magnificent system of fresh-water lakes in North America that this influence is most strikingly displayed. Owing to the severity of the winters, these lakes are partially frozen for some distance from their shores during that season, and hence that region is curiously characterised by a winter climate of almost continental severity, and a summer climate of comparative insular cool

ness.

210. Advantage has been taken of terrestrial radiation to secure the formation of ice during the night in Bengal when the temperature of the air is above 32°. The process was first