countries situated within the tropics, because a much greater regularity is observed there than in our own climates.

In all places where the trade-wind blows constantly seaward, it does not rain; the sky is always serene, especially when the sun is in the other hemisphere; but it often rains in the region of calms. The ascending current draws with it a mass of vapours, which condense as soon as they arrive at the line of junction between the upper and lower tradewind. The sun almost always rises in a clear sky; toward mid-day, isolated clouds appear, which pour out prodigious quantities of rain. These showers are accompanied with violent gales. Towards evening the clouds dissipate, and when the sun sets the sky is perfectly clear. Thus the masses of air discharge the water they contain, into the very regions from which they rise; and hence arises the absence of rains in countries more distant from the equator, where the east wind regularly blows.

On shore, between the tropics, we find during a part of the year disturbances in the direction of the trade-winds; and the year is divided into two seasons,-the wet and the dry season. Europeans have found this climacteric division adopted by all indigenous people, and it is the more characteristic, as entire months frequently pass away during the dry seasons without a single cloud having been seen in the sky.

Notwithstanding local differences a great regularity is every where observed in the succession of phenomena, but I will content myself with pointing them out according to M. de Humboldt, inasmuch as his researches have thrown a bright light on the causes of the variations that are observed in our climates.

In the part of South America situated on the north of the equator, the sky is perfectly serene from December to February; the wind blows from the E. or E.N.E.; the air is dry, and vegetation is leafless. Toward the end of February, and at the commencement of March, the blue of the sky becomes deeper, the hygrometer denotes more moisture in the air, and the leaves of the trees begin to burst forth. A slight curtain of vapour dulls the twinkling of the stars, which is dense toward the zenith, where it is sometimes visible. The trade-wind blows less violently, and the air is occasionally entirely calm. Clouds resembling mountains gradually collect in the S.S.E., and sometimes traverse the sky with incredible velocity. Toward the end of March, lightnings shine in the heavens at the south; and the wind passes for several hours to the W. and W.S.W. The atmo

spheric electricity becomes stronger, especially at sun-set; and this is a certain sign of the approach of the rainy season, which, on the banks of the Oronoko, begins at the end of April. The sky is troubled, and, from its former blue, becomes grey. In the afternoon, at the moment when the heat is at its maximum, a storm, accompanied by heavy showers, rises in the plains. At the commencement, the clouds and rain are only formed during the burning hours of the afternoon, and disappear toward evening. But in proportion as the season advances, especially when the sun is at the zenith, they both commence being manifest in the morning, but at the end of the season they again appear in the afternoon.

In many countries, the night is almost always serene; in others, it rains in the night also, and even more so than in the day; but it is probable that this difference is due to the neighbourhood of large chains of mountains. M. Boussingault determined this on the table-lands and in the valleys of the Andes, at Peru;* Lyall, at Madagascar; Admiral Roussin, at Cayenne. Other travellers have confirmed these facts by isolated observations.

All these phenomena tend to prove that the ascending current, which is, for the most part, very strong in the place that has the sun in its zenith, causes a disturbance in the atmosphere. Hence there is first the scintillation of the stars, and then a change in the direction of the winds. The evaporation of the water that falls over night so saturates the air with vapours, that, even in Africa, cloaks, shoes, in a word, all things which are not placed near the fire, become moist; and the inhabitants live in a kind of perpetual vapour-bath. This is the epoch of the endemic diseases so fatal to Europeans. In Africa, the approach of the rainy season is also announced by changes in the direction of the winds.

These rains being a consequence of the ascending currents, the place where they fall changes with the declination of the sun, the presence of which determines the cur

In the neighbourhood of the gold mines of Marmarto, lat. 5° 27' N.; long. 5 11 W., absolute elevation 1426 metres, mean temperature 20°, 4, this philosopher obtained the following results:

(Vide Comptes rendus de l'Acad. des Sciences, t. ii. p. 109. 1836.)-M.

rent. In Africa, for example, near the equator, the season of rain commences as early as April. Between 10° of N. lat. and the tropics, principally in the countries watered by the Senegal, it lasts from the commencement of June to the commencement of November. It is the same in the interior of countries, as may be seen by the recitals of Mungo Park, Denham, Browne, Bruce, and others. In like manner, at Panama, on the western coast of America, the rain commences in the early days of March; and at San-Blas, in California, it rarely rains before the middle of June. As the sun passes twice over the zenith of each place, we find that in those which are near to the tropics a very considerable quantity of rain falls twice a-year, and at very near intervals. In countries situated near the equator, where the times of the passage of the zenith are separated by a longer interval, there are two rainy and two dry seasons.

The northern limit of these periodical rains is not exactly known. At Havanna, in the island of Cuba, and at Rio de Janeiro, climacteric conditions have been noticed, which have some analogy with those of high latitudes. In the desert of Sahara the limit appears to be about 16° of N. lat., but on the two seas that wash the coast of Africa it is some degrees more northerly.

In India, the alternation of seasons, compared with that existing between the tropics, is no less anomalous than the direction of the winds. The west coast of this peninsula has its season of rains during the S.W. monsoon; whilst its dry season prevails during the N.E. monsoon. When the wind that blows from the S. W. is forced to ascend along the flanks of the Ghauts, the vapours condense on their summits, and there are violent storms almost every day. In the interior of the country the rains are rare, and on the eastern coast the sky is serene. The rains are most abundant in July. During the N.E. monsoon the same succession is noticed on the coast of Coromandel; but the mountains not being so steep, the rains are not so heavy. During this time, the sky is perfectly serene on the west coast. The table-land of the Deccan partakes in the climate of the two coasts. The distribution of rain, during the seasons, depends on the distance of the different points from the sea. According as they are nearer the western or the eastern coasts, the course of the seasons is analogous to that of the corresponding coast. Some places, situated to the middle of the peninsula, have partial rains throughout the year; or else they have two maxima in the year.

The quantity of water that falls in these countries du

ring the space of each month, is more considerable than that of the whole year with us. In places situated near the sea, we may admit that from 190 to 320 centimetres of water fall during the year. Let us add that it never rains but in certain months, and only during one or two hours of the day, which renders the contrast still more striking. The drops of water are enormous, very close together, and reach the earth with great violence. But if we penetrate into the interior of countries, or ascend considerable heights, the quantity of rain diminishes. At Seringapatam, in India, and at Bogota, in America, it is hardly greater than that observed in Ger

many.

RAINS IN HIGHER LATITUDES.-The periodicity of rain disappears as we go further from the equator. Yet we are in want of certain facts in order to determine, in a positive manner, the transition from one system of climate to another. While, between the tropics, the greatest quantities of rain fall when the sun is at the zenith, that is to say, in a season corresponding to our summer; north of the tropics, it rains more abundantly in winter. If we designate the annual quantity of rain by 100, we have, for the few places hitherto observed:

So that, under this parallel, it rains most in winter, and the quantity of rain that falls in summer is altogether insignificant. The same relation is found on the N.W. coast of Africa, and in the Canary Isles. This contrast between climates situated on the two sides of the trade-winds is very remarkable; it is a sudden, not a gradual transition, as might, à priori, be imagined. I insist on this fact, to shew how much it is in contradiction to the assertions hazarded by ancient meteorologists.

This sudden change is easily deduced, from what I have said in general of the precipitation of aqueous vapours. The cause may most frequently be recognised in a mixture of strata of air of unequal temperatures. Now, variable winds

RAINY WINDS IN EUROPE.

often bring about this conflict. In summer, the regular east wind extends as far as the coast of Portugal (p. 47). Hence, there are less disturbances in the equilibrium of the atmosphere. The formation of clouds is then much more rare than during the variable winds of winter.

RAINY WINDS IN EUROPE.-On collecting all that is known in the different climates of Europe, we are led to establish three hyetographic regions: 1st, that of England and the west of France, which extends in a modified form even into the interior of the Continent; 2d, that of Sweden and Finland; 3d, that of the coasts of the Mediterranean. The limits of these regions are not always rigorously defined; they are not clearly recognised, except in points where they are marked by great chains of mountains. Every where else the transitions are found to be very orderly. The differences of these three groups consist in the different direction of the rainy winds, and of the distribution of the quantity of water which falls each year.

Let us consider the part of Europe north of the Alps and the Pyrenees; the predominance of west winds, a vast ocean on one side, a great continent on the other, are the determining circumstances of the distribution of rains. If the N.E. wind always prevailed, even at a considerable height, it would never rain; for it passes over lands before arriving at the low latitudes, where the elevation of temperature removes the vapours from their point of condensation. If the S.W. wind, on the contrary, blew without ceasing, it would always rain; for, as soon as the moist air gets cool, the vapour of water is precipitated. In spite of their alternations, these winds always preserve their relative character. If we inquire, with M. de Buch, how many times each wind brings rain, these results become evident. In 100 showers, which fell at Berlin, the different winds blew in the following proportions:

N.

N.E.

E.

S.E.

S.

S.W.

W.

N.W.

4,9 4,9 10,2 32,8

24,8 14,4

4,1 4,0 Thus, scarcely any rain fell with the N.E. wind, whilst at least half are brought by the W. and S.W. winds. But the winds do not all blow an equal number of times in the course of the year. The number of times that each wind has blown must, therefore, be divided by the number corresponding to each wind in the preceding table. We then obtain the following numbers :

E.

S.E.

E.

S.W.

W.

5,8 8,1

4,2

N.W. 4,5