These observations having been made with Daniell's hygrometer, we might imagine that the want of coincidence in these values depends on the difference of the instruments employed. To be certain on this point, I made simultaneous observations for several months, during the summer of 1836, with the psychrometer and Daniell's hygrometer: their indications agreed very well. We are hence compelled to admit, that their want of coincidence is due to differences of climate; whilst, at Halle, the quantity of vapour goes on diminishing from morning till mid-day during the whole summer, we find no indication of this diminution at Apenrade, where it augments and diminishes with the temperature, and where the differences between the maxima and the minima are much more considerable than at Halle.

In order to give a value to these differences, we must know the laws that these variations obey; but, the number of observations being too limited, we must confine ourselves to a few general remarks. All these differences are connected, either with the ascending currents or with the resistance that the air opposes to the transference of vapours. When evaporation commences in the morning with the increase of temperature, the vapour, by virtue of the resistance of the air, accumulates at the surface of the soil, as observations made at all parts of the globe shew. This stratum of vapour does not attain a great thickness; but, as soon as the ascending current commences, especially in summer, the vapours are drawn away toward the upper parts of the atmosphere, with a force that continues increasing until mid-day. The evaporation from the soil is then more active on account of the increase in temperature; nevertheless, the ascending current carries away the greater portion, and there is a diminution in the quantity of vapour. Towards evening, when the temperature begins to fall, the ascending current diminishes in force, or even ceases altogether; then, not only does the vapour accumulate in the lower parts, but it even descends from the higher regions; and, on this account, we observe towards evening a second maximum, which is not sustained, because, during the night, the vapour precipitating in the form of dew or white-frost, the air necessarily becomes drier.

The justice of these remarks is fully confirmed by observations made on a high mountain. Whilst at a little height, as that of Halle, the quantity of vapour diminishes toward mid-day, we find, on elevated points, a rapid increase in the course of the day, and, in the evening, a no less rapid diminution. These two phenomena are the more marked

as the point is more elevated. In June 1832 and 1833, I observed the hygrometer on the Rigi (1810), and during the months of September and October in the same year, on the Faulhorn (2672), while M. Horner did the same at Munich. The following table gives the mean of the observations of the two years.

TENSION OF THE VAPOUR OF WATER AND RELATIVE HUMIDITY, AT ZURICH, ON THE RIGI, AND ON THE FAULHORN.

Means. 10,97 74, 6 6,85 84,3 9,25 74,8 4,13 74,4

(Vide Appendix, fig. 10.)

Let us compare the absolute tension of vapour at Zurich and on the Rigi; the pressure of the atmosphere of vapour is at its minimum at sunrise. On the plain, it presents a

maximum about nine o'clock in the morning; then comes a slight diminution, and we find, about three o'clock a value lower by about 0mm,2 than that of the morning. The quantity of vapour then increases progressively; and, about sunset, attains a second maximum, to diminish then with equal regularity until next morning. On the Rigi, situated at a little distance, and elevated about 1402 metres above the lake of Zurich, the mid-day minimum is entirely wanting, on account of the vapours, which incessantly rise from the plain; towards evening these vapours fall very rapidly below its summit. The observations at Faulhorn are still more decisive; on this mountain, the greatest tension of vapour does not take place until several hours after noon.'

*

These notable differences in the range of the absolute quantity of the vapour of water lead to still greater in that of relative moisture. De Saussure and Deluc had previously pointed them out to observers, notwithstanding the imperfection of their instruments. The ascending current of the morning draws the vapours toward the upper regions;

*We as yet possess but a very few series of hygrometric observations made upon high mountains. In 1841, M. BRAVAIS studied with me, and, in 1842, with M. PELTIER, day and night, the moisture of the air, by means of the psychrometer, on the summit of the Faulhorn. The mean of thirty days' observations gave us the following results :

The range of the absolute tension of vapour, and the relative humidity, is almost the same as that given in M. KAEMTZ's table. During our observations, the relative humidity was sensibly stationary from ten o'clock in the evening until the hour of sunrise; according to him, the said humidity slowly decreased during the same period. This difference might arise from M. KAEMTZ having obtained the relative numbers of the night-period by interpolation. The amplitude of oscillation is a little greater in our observations than in his.

With regard to absolute tension, the hours of maximum and minimum, as well as the amplitude of variation, are almost identical.-M.

the air becomes relatively drier than it would be, in consequence of the increase of temperature alone; whilst these vapours are rising, the dryness increases less rapidly, especially as we consider that, in the higher strata of the atmosphere, the temperature changes less than in the lower strata. It may even happen that, on very elevated points, the air becomes relatively moister during the course of the day, whilst its dryness increases toward evening, when the vapours descend to the plains. My observations prove the possibility in the most evident manner; for, although the relative moisture at Zurich and on the Rigi follows a rate depending on the hour of the day, the differences are, however, much less sensible on the Rigi. At Zurich, between sixteen hours and two hours, we find a difference of 31,4 per cent; on the Rigi, between sixteen hours and one hour, a difference of only 9,3 per cent. On the Faulhorn, which is 870 metres higher, the rate is almost the reverse; in the morning about eight or nine o'clock, some hours after the maximum humidity at Zurich, the air is driest. In like manner in the afternoon, the relative humidity is very great on the Faulhorn, whilst, in the plain, the air attains its greatest degree of dryness. There must, therefore, exist at a certain height a point where the hygrometer is stationary during the twenty-four hours.

I have insisted on these differences, as they serve to explain several interesting phenomena. We may easily deduce from them the causes which operate so that the diurnal rate of relative moisture is different on the coasts and in the interior of continents. We have seen that, on the seacoast, the quantity of vapour goes on increasing regularly from sunrise till about two or three o'clock in the afternoon; this happens, because the sea-breeze rises precisely at the moment when the ascending currents draw the vapour toward the higher regions. This breeze brings vapours from the sea, and the air becomes moister during the afternoon than it is in the middle of the continent.

ANNUAL VARIATIONS IN THE QUANTITY OF VAPOUR OF WATER.-Vapour, being the result of the action of heat on water, it is evident that its quantity must vary in different seasons. This fact is even proved by the very few continued series which we possess. Let it at present suffice to give a few results, shewing what the absolute and relative moisture is at Halle, during the different months of the year.

TENSION OF THE VAPOUR OF WATER AND RELATIVE HUMIDITY IN THE DIFFERENT MONTHS, AT HALLE.

In January, the coldest month of the year, the quantity of vapour attains its minimum; at the same time, the relative moisture is at its maximum. In proportion as the temperature rises, evaporation becomes more active, and the quantity of vapour increases, at first slowly, because the east winds, which commonly blow during this season, bring dry air from the interior of the continent. However, we must not deny, that the numbers for winter and spring differ probably much from means furnished by series embracing a greater number of years; for the latter winters have been warmer and the springs colder than they generally are. So that the numbers corresponding to winter are too high, and those to spring too low. The quantity of vapour attains its maximum in July, the month in which the air is driest. At the approach of winter, when the heat diminishes, the quantity of water precipitated in the form of rain, dew, and hoar-frost, greatly exceeds that which passes into the state of vapour. Its quantity, therefore, goes on diminishing, although the humidity is continually increasing, and is greater in November and December than in the month of January. This is the origin of the damp cold which characterises these two last months.