mass of ice, over which a current of lava had flowed without melting it.*

The earth, as it turns round the sun, moves in a medium, the temperature of which, no doubt very low, is completely unknown to us. On the other hand, the stars, notwithstanding the infinite distances by which they are separated from us, send us rays both luminous and calorific. Some regions of the heavens also, being more rich in stars, the quantity of heat that reaches us from different points of space is not the same. But, like as the different indications of the thermometer, during the course of a year, are all reduced to a mean, so may we presume that the heat of the sky is uniformly spread through all the celestial vault. This heat, combined with that of the space in which the earth moves, gives us the temperature that Fourier named the temperature of space, and what he states to be from -50° to -60°; whilst M. Pouillet fixes it at-140°.† The

In the Annuaire for 1834, M. ARAGO published a notice on The Thermometric State of the Globe. He proved with his usual clearness :

1st. That there exists in the earth a central focus of heat.

2d. That, for 2000 years, the general temperature of the mass of the earth has not varied a tenth of a degree; and yet the surface has become cold in the course of ages, so as scarcely to preserve any sensible trace of its primitive temperature.

3d. He shews that the changes we have observed, or think we observe, in certain climates, are not connected with cosmical causes, but with circumstances entirely local; such as the clearing of woods and mountains, the drying up of morasses, extensive agricultural works, &c. &c. Thus, on comparing the thermometric observations made at Florence, according to the instructions of the Academy of Cimento, towards the close of the 16th century, with those comprised between 1820 and 1830, it was found that the mean remained sensibly the same. It would merely appear that the winters are not quite so cold, and the summers not quite so hot; a result probably due to the clearings that have been made since this epoch. In the United States, an analogous effect has been observed, in consequence of the vast clearings of which this country is the theatre. M. ARAGO then applied these notions to the climate of France; and he shews that there is nothing to prove its having undergone any other changes than those derived from the labours of man. With regard to the temperature of the terrestrial crust, at a depth of twenty-eight metres, which is that of the cellars of the Observatory, it has not changed for a century; for an observation made by MESSIER, in 1776, gives exactly the same cipher as in 1826, namely, 11°,8.-M.

In endeavouring to determine the temperature of space, M. POUILLET proposed an instrument which he called an actinometer. It is composed of four rings, of two decimetres in diameter, covered with swan's down, and resting on each other in such a manner that the swan's down shall not be compressed. The skin of the swan itself forms the base of the circle of each of the rings. This system is enveloped in a first cylinder, which is itself enveloped in swan's down, and contained in a larger cylinder. A thermometer rests in the centre of the upper swan's down; and the border of the exterior cylinder has such a height, that the thermometer can subtend only two-thirds of the hemisphere of the heavens. This border is pierced with holes, in order that the cold air may escape readily.

This apparatus being exposed in an open place, and on a serene night, to the radiation of the sky, its thermometer, and a neighbouring thermometer, suspended freely in the air, are observed from time to time. From the dif

difference of these two results shews how difficult this question is; and, moreover, the temperature of space appears to have but a feeble influence over that of the lower strata of the atmosphere.

Supposing a different temperature in the different regions of space, Poisson deduced the proper heat of the earth. For all our system being sustained in vacuo, it is possible that it may have traversed very hot regions. Hence the heat that is still observed in the deep strata of the earth, which have not yet had time to cool.

INFLUENCE OF HYDROMETEORS OVER TEMPERATURE. —Let us abandon theories to study causes, whose action is more powerful and more easy of demonstration. Among these causes hydrometeors occupy the first rank. We may, indeed, readily conceive that the state of the sky exercises an immense influence. When on a summer's morning the sky is calm, and the air serene, the temperature rises notably in a few hours. But if clouds cover the sky, and intercept the rays of light, the thermometer rises but little, or even falls considerably before the moment of the maximum of heat. The converse takes place when the sky is cloudy in the morning and serene in the afternoon. In winter, on the contrary, the thermometer

ference of these two thermometers or from the fall of that one attached to the actinometer, the zenith temperature is deduced.

Experiments made with this instrument gave M. POUILLET two limits for the temperature of space, -115° and 175°, the mean of which is 140°.

From these researches he deduces many consequences of great interest. The total quantity of heat which space sends to the earth and to the atmosphere, in the course of a year, would be capable of melting on our globe a bed of ice twenty-six metres thick.

We have seen that the quantity of solar heat is expressed by a bed of ice of thirty-one metres; so that the earth receives in all a quantity of heat represented by a bed of ice fifty-seven metres thick.

We shall undoubtedly be astonished that space, with its temperature of -140°, can communicate to the earth so considerable a quantity of heat, that it is found almost equal to the mean heat of the sun. But we should remark that, with respect to the earth, the sun occupies only the five-millionth of the celestial vault; that it must, consequently, send two hundred thousand times more heat in order to produce the same effect.

If the action of the sun were not felt on our globe, the temperature of the surface of the ground would be every where uniform, and at -89o. Now, since the mean temperature at the equator is 27°,5, we must conclude that the presence of the sun increases the temperature of the equatorial zone 116°,5.

To extend these calculations to other regions, we have merely to take account of the decrease of the temperature of the earth, in proportion as the latitude increases. (Vide Comptes rendus de l'Acad. des Sciences, t. vii. p. 53 [1838], and Eléments de Physique, t. ii. p. 538, and fig. 377.)

M. ARAGO having found, in the relation of the voyage of Capt. BACK, that at Fort Reliance the thermometer descended to 56°,7, concluded that the temperature of celestial space could not fail to be notably lower than -57. (Comptes rendus de l'Acad. des Sciences, t. ii. p. 575, 1836.)—M.

rises when the sky is clouded, and sensibly falls as soon as the clouds are dissipated.

The summary of observations accords with these isolated facts. If in an isolated month we take the mean of the serene days and of the cloudy days, we find a notable difference between these two numbers. In winter, the cloudy days are several degrees hotter ; in summer, it is the reverse. This difference between the two seasons results from what we have said of the absorption of calorific rays by the atmosphere, and (p. 24) from the range of heat in these two seasons. In summer, as in winter, the earth loses by radiation one part of the heat that it has received from the sun, but in summer it receives much more than it loses. Although the dark calorific rays are relatively much more absorbed than the others, yet the heat received is greater than the heat emitted; but if, during summer, the sky is clouded, there is a fall of temperature. In winter, on the contrary, the earth becomes generally cold, the loss due to nocturnal radiation at night being greater than the heating by solar action. But, as clouds oppose radiation, and reflect back to the earth a portion of the obscure rays that it emits, there is an elevation of temperature in cloudy weather. Add to this, that the vapours precipitated during winter are at a much less height than in summer, and that the latent heat, which becomes free at the moment of their condensation, may act on the ground.

The fall of temperature that is noticed in summer, when the sky is clouded, is still more considerable when it rains; then, not only are masses of water precipitated from the high and cold regions of the atmosphere, and notably reduce the temperature in virtue of their great capacity for heat, but this water, in evaporating, again absorbs a notable quantity of heat, which it takes from the earth and air that are in contact with it. Hence arises the cold that is observed after rain-storms. If, in like manner, we study a long series of observations, we shall find the evident differences of temperature that exist between the rainy and the dry months of winter and summer every one remembers the rainy and cold summers of 1833 and 1838, the serene and hot summer of 1834, the mild and rainy winter of 1833-4, as well as the clear and cold weather of that of 1829-30.

Within the tropics, the influence of the state of the sky over the temperature is especially remarkable. The meridian height of the sun varying but little in these climates, the rains are the more immediate cause by which the range of temperature is regulated,—a range totally different from

that which takes place in our climates. When the sun is very far from the zenith, that is to say, when it is in the northern hemisphere during the months of December and January, the temperature is relatively very low. In proportion as the meridian height of the sun increases, the heat increases also, and would go on increasing without cessation until the sun is at the zenith; but then the rain commences and the heat diminishes; and it is not until later, when the sun, having passed the zenith, is in the other hemisphere, that there is an increase in the temperature, which attains its maximum when the rain is about to cease, and then diminishes to attain the minimum of which we have spoken. Thus, while in our climates the temperature has one minimum and one maximum, two maxima and two minima occur in hot countries. The two latter are one in the middle of the dry, and the other of the wet season, when the zenith distance of the mid-day sun is as great as possible. The two maxima occur at the beginning and the end of the wet season. Each locality within the tropics presents a different range of temperature: the minimum is an instantaneous effect of rain, but which lasts only a short time, or else remains for several months, without any very notable maximum afterwards following; because, as the sun recedes from the zenith, the heat diminishes.

Among the great many places situate within the tropics, where I find the confirmation of what I have just said, let me mention three cities in India. In the following table I give for each of them the quantities of rain, and the mean monthly temperatures; the last column presents the sum of the monthly quantities of water, and the mean of the temperatures:

CALCUTTA.

28,5

28,6

33,84

30,5

16,24

29,7

26,6

22,56

31,2

575,24

29,3

29,8

338,38

28,1

99,26

29,3

311,31

28,3

26,8

27,7

42,86

27,2

25,9

20,30

23,0

26,6

0, 0

19,2