indicates the result sought for, namely, the cooling of bodies below the ambient medium.

There is, however, one experiment of Wells' in which a thermometer, wrapped about with wool, placed at the same level with a free thermometer, gave a lowering of temperature of 50°3. Here the wool was cooled two or three times more than the lamp-black in my experiments; but I know that the emissive power of wool is not greater than that of lamp-black.

In order to determine the cause of this extraordinary cold observed by Wells, it was necessary, in the first place, to put the fact beyond doubt. With this view, having enveloped one of my thermometers in a tuft of wool, I exposed it along with two others of the same dimensions, one of which was covered with lamp-black, the other retained its metallic brilliancy, the instrument descended, in a few minutes, double the extent of the blackened thermometer. A fourth instrument, enveloped in an equal quantity of wool, condensed and pressed around the metallic cover by means of some rounds of thread of the same substance, gave an intermediate cold between the two preceding. I lastly prepared a fifth thermometer with a double covering of fine flannel, and it sunk still less than the fourth. I repeated these experiments by substituting cotton for wool, and obtained results in every respect analogous. I then perceived that the superiority of cotton or wool over lamp-black, in the phenomena of nocturnal refrigeration, was owing to a certain modification introduced into the emissive power of these bodies by the presence of air among their interstices.

But how does air increase the cold arising from radiation ? The answer is simple and obvious. We have known, for many years, that the nocturnal coldness of bodies does not vary with the temperature of the atmosphere. Thus, Captains Parry and Scoresby have found, that during the calm and pure nights of the polar regions, the snow cooled about 9 degrees below a stratum of air raised from 1m 30 to 1m 60, when the atmosphere was from -25 or -30 degrees, and when its temperature is very near zero. M. Pauillet has seen swan-down descend to about 7° under the temperatures of

0° and of 25. For my own part, I am convinced that blackened or varnished thermometers descend with a constant quantity, whatever be the temperature of the night. Now we may conceive that tufts of cotton or wool, placed on the upper part of thermometrical bulbs under the action of a clear sky, after having become cold by radiation, will communicate the cold they have acquired to the surrounding air, which, becoming heavier, will descend in the interior afterwards to fall to the ground; but this condensed air will always require a certain time to disengage itself from the obstacles which arrest it among the threads. The latter, therefore, will be surrounded with a colder air than at the beginning of the experiment; and, as their sinking of temperature below the ambient medium must continue invariable, it must necessarily happen that they will cool the more. This increase of cold will cause an additional depression of temperature in the medium; the latter, in its turn, will occasion an additional cooling in the radiating bodies, and so on consecutively, until the weight acquired by the condensed air release it from the obstacles which opposed its escape from the envelope.

What takes place with small quantities of cotton and wool artificially placed round thermometers, ought to occur naturally in many circumstances. In point of fact, plants with downy leaves are much colder than such as have smooth leaves. The temperature of grass and that of other low plants which cover fields, descends, in consequence of this frigorific reaction of the air, much below that of elevated bodies, on account of the neighbourhood of the ground which supports the ambient medium, and forces it to remain in the presence of radiating surfaces. In reality, the stratum of air in which meadow-grass is placed, is by no means immoveable; on the contrary, it whirls about exactly like the water in a vessel placed upon the fire; the aërial particles, condensed by the cold of the tops of the grass, descend to the surface of the meadows, are warmed by contact with the earth, ascend again towards the higher parts of the grass, and so on; but it is evident that, notwithstanding this state of agitation, they at last become colder, and that, in order to maintain itself constantly in a colder state than they, the grass must cool more

and more, and this circumstance will induce a gradual refrigeration, and an increasing humidity in the stratum of air.

I do not here enter into the details necessary to shew how the principle of the frigorific reaction of the air serves to explain all the facts which precede and accompany the appearance of dew in meadows, and a multitude of phenomena of which a proper account could not hitherto be given. But your own discernment will supply this want; and I shall conclude by indicating the principal questions treated of in the memoir, which I shall soon have the honour to present to the Academy.

I may say, then, that besides the difficulties of which I have given a distinct solution in these two letters, my new experiments on the nocturnal cooling and on dew have enabled me to understand perfectly, 1st, The distribution of temperature among grass, which is found to be colder in the night among it than at the surface of the meadow. 2d, The inversion of the ordinary temperatures of the atmosphere near the earth's surface. 3d, The great humidity of the air near plants, from the first instant that the dew begins to appear. 4th, The injurious action of the least breath of wind. 5th, The formation and accumulation of dew during the whole course of the night. 6th, Its successive propagation from below upwards. 7th, The small quantity of dew on trees when compared with grass and low field plants. 8th, The disappearance of small drops of dew, which sometimes takes place on the lower parts of plants, while they are forming on the upper parts. 9th, The variable proportion of the meteor in the different seasons of the year. 10th, Its general distri

11th, The great differ

bution over the surface of the globe. ence between the diurnal and nocturnal temperatures of the torrid zone. 12th, The absence of dew in the small islands of Polynesia, and on vessels sailing in the midst of large seas. 13th, Its abundant formation when the vessels approach certain shores of continents. 14th, The sharp cold produced in the night in the sandy plains of central Africa. 15th, The natural and artificial congelation of shallow waters, when the temperature of the atmosphere is from 5° to 6° above zero, by taking into account the indisputable fact that water does

not cool more than 1°.5 in consequence of its direct radia

tion.

I may finally add, that the part performed by stagnant air in the phenomena of nocturnal refrigeration, appears to me to have modified certain experimental data which have been founded on for calculating the temperature of space.

On the Relations which exist between the Phenomena of Erratic Blocks in Northern Europe and the Elevations of Scandinavia. By M. DESOR.*

There is a point in the phenomena of boulders not hitherto presented, which appears to me capable of throwing much light on the question that now engages our attention: it is the examination of the connection which subsists between the erratic blocks of Northern Europe and the elevation of upraising of Scandinavia. These relations are the more important, as they admirably explain some circumstances which are peculiar to the erratics of the north, and of which there is no example in Switzerland. These peculiarities are 1st, the occurrence of polished and grooved surfaces beneath the present level of the sea; 2d, the existence of marine shells attached to the polished rocks at a height much above the present level of the sea; 3d, the presence of marine shells in the midst of the diluvium, even at an elevation of 800 feet; 4th, the osars, or ridges of boulders and stones which contain the shells of the Baltic.

Among the phenomena which prove so fully the instability of the Scandinavian soil, there are some facts which indicate the rising of the land, while others, on the contrary, attest the subsidence. Thus we cannot have less equivocal proof of an elevation of a country, than the occurrence, at a great height, and at a considerable distance from the coast, of shells now inhabiting the adjacent seas, and whose perfect state of preservation leaves no doubt that they lived where they now occur; for, had they been transported by a current

* Communicated, by the Author through Prof. Agassiz, to Editor of the American Journal of Science and Arts, May 1847.

or any other violent agent, they would have been broken, or at least much worn. But, should we refuse to admit the evidence which these shells offer, we cannot deny the proof afforded by the serpulas of Christiania and the barnacles of Uddevalla, whose shells still adhere to the rocks far above the sea.

On the other hand, the fact that the striæ and furrows are continued beneath the waters of the sea, attests no less strongly that at a certain epoch the land must have been more elevated than now. In fact, it is a point on which the partizans of different hypotheses are nearly agreed, that the phenomena of erratics took place over a submerged country. Glaciers can advance only as far as the limits of the land. We learn, from the observations of M. Martins, that even the glaciers of Spitzbergen do not project beneath the sea; for, as the temperature of the water is above that of ice, it melts the glaciers by its contact, and a considerable space, equal to the height of the tide, separates the glacier from the water.*

But if, as I believe I have sufficiently proved, the polished surfaces of the north have been occasioned by immense glaciers, which have transported from afar the erratic blocks of Scandinavia, and furnished the materials of the diluvium and of the osars, it follows, that the whole country which bears traces of scratches, must have been out of water when the glaciers produced this polishing, and produced the striæ and furrows which we now see there. If these striæ were exactly at the level of the sea, we might suppose that Scandinavia was then at the same elevation as at the present day. But we have seen numerous cases in that island, in which the furrows are found under the sea; from which facts we must conclude, according to the principles laid down, that the land at that epoch was as much above its present height as the

* In order that glaciers should advance upon the bottom of the sea, it is necessary that the temperature of the water should be below zero the year through; but such a climate would render the formation of glaciers impossible. A humid atmosphere, rather than a severe temperature, is necessary to the existence of glaciers, and the former is incompatible with the temperature of a sea constantly below zero.