parison with the bright thermometer, the degree of cold produced by the radiation of this substance. By means of instruments thus prepared, M. Melloni has assured himself that the leaves of plants, glass, varnish, and lampblack, are always cooled in clear and still nights, from one to two degrees (Centigrade*) below the temperature of the surrounding air.

In observing the small amount of this cooling, it might seem that the decrease of temperature of seven or eight degrees, reported by Wilson and Wells, was a great exaggeration. But considering that the differences obtained by these gentlemen contained an element foreign to the question, that the thermometers used to obtain the ambient temperature were elevated four or five feet, while the instruments enclosed by the radiating substance were very near the ground, it is easy to understand the reason why there is so great a difference between the results obtained by them and those of the writer.

M. Melloni then refers to the experiments of Peclet, which showed, long ago, that the temperature of the air, in calm and clear nights, decreases rapidly as we approach the surface of the earth. He observes that the simple fact of this decrease ought to render the temperature of a radiating substance, placed near the ground, inferior to that of the air surrounding a thermometer elevated four or five feet; so that, in this arrangement, the difference of the two instruments by no means shows the desired result, namely, the cooling of the body below the temperature of the ambient medium.

There is, however, one experiment of Wells, in which a thermometer, wrapped in wool, being placed on the same level with a free thermometer, showed a decrease of temperature amounting to 5.3 degrees. Here, the wool was cooled to a degree two or three times greater than the lampblack used in the experiments of M. Melloni, who remarks that he knows, notwithstanding, that the emissive power of wool is not greater than that of lampblack. In order to investigate the cause of this extraordinary discrepancy, M. Melloni enveloped the coating of one of his thermometers with wool, and exposed it to the air with two others of the same size, one of which was covered with lampblack, and the other retaining its metallic brightness. In a few minutes the descent in the instrument wrapped in wool was twice as great as that in the blackened thermometer. A fourth instrument was then enveloped in an equal quantity of wool, which was closely pressed and confined around the metallic casing by means of a thread of the same material: this showed a cooling intermediate between the two former indications. He next enveloped a fifth thermometer with a double coat of fine flannel, which fell still less than the fourth. The experiments were repeated by substituting cotton for wool, and results altogether analogous were obtained. I then comprehended, says M. Melloni, that the superiority of cotton and wool over lampblack, in the phenomena of nocturnal cooling, was due to a certain modification effected in the emissive power of these bodies by the presence of the air contained within their interstices.

*A degree of the Centigrade thermometer, which is used by the French philosophers, is equal to one and four-fifths degree of Fahrenheit.

The agency of air in thus augmenting the cold produced by radiation, he explains as follows: It has been known, for a number of years, that the nocturnal cooling of bodies does not vary with the temperature of the atmosphere. Thus, Captains Parry and Scoresby found that, during calm and clear nights, in the polar regions, the snow cooled about nine degrees below the stratum of air at four or five feet high, when the atmosphere was at 25 or 30 degrees below 0, and the same when the temperature was very near O. M. Pouillet has seen the down of the swan fall about seven degrees in temperature at 0, and at 2.5°. And, on his own part, M. Melloni assures us that the fall in his blackened, or varnished, thermometers is a constant quantity, whatever may be the temperature of the night. Now it may be conceived that the tufts of cotton, or wool, spread over the upper part of the thermometric bulbs under the action of a clear sky, after being cooled by radiation, would communicate the cold thus acquired to the surrounding air, which, becoming heavier, descends within, and falls towards, the ground; but there is always required, by this condensed air, a certain time for its disengagement from the obstacles which arrest it among the threads. These are, then, surrounded by an air colder than it was at the commencement of the experiment; and as their decrease of temperature below the ambient medium must continue invariable, it necessarily follows that they must cool still further. This increase of cold will produce a further decrease of temperature in the medium; the fall of temperature in the medium will react by still further cooling the radiating body, and so on, continually, until the weight acquired by the condensed air delivers it from the obstacles which opposed its passage from the enveloping substance.

This process, which takes place when tufts of cotton, or wool, are artificially placed around thermometers, must occur naturally under many circumstances. Thus plants which have downy leaves are colder than those whose leaves are smooth and glossy. The temperature of grass, and other low plants, which cover the fields, falls, by reason of this frigorific reaction of the air, considerably below those of more elevated growth, on account of the immediate proximity of the earth, which sustains the ambient medium and compels it to remain about the radiating surfaces. In fact, the stratum of air which envelopes the grass of a meadow is not maintained at rest; but, on the contrary, is subjected to agitation similar to that of water in a vessel placed over the fire: the particles condensed by the cold of the pointed leaves of grass descend to the ground, are warmed by contact with the earth, ascend again among the herbage, and so on, continually. But it is evident that, notwithstanding this state of agitation, the temperature must be finally decreased; and that the grass, to be kept continually cooler than the air by a constant ratio, must itself become cooler and cooler, which will produce a gradual reduction of temperature and an increasing humidity in the stratum of air.

M. Melloni concludes his second letter by announcing that he will shortly submit to the Academy a memoir, in which he will treat of a number of interesting and important questions connected with the nocturnal reduction of temperature, and the production of dew.

In the "Comptes Rendus," for October 11, 1847, is a third letter from Melloni to Arago, complaining that certain journals had misrepresented his views as expressed in the former letters, and making further remarks upon the subject under discussion. This letter we transate as follows:

My studies upon dew seem to me to have placed beyond doubt this fact, that if the principle of Wells is true, the theory which goes by that name is erroneous, or, at least, altogether incomplete. I thought I had so clearly defined this proposition that no mistake could arise; but in looking over several periodical publications, I find that it has been altogether distorted. In effect, the conductors of these journals adopting, perhaps, the opinion of him among them who first gave an account of it, cite the first part of my theorem and say nothing of the second; the reader is thus led to believe that my labors tend solely to confirm the theory of Wells, such as it is developed in all the physical and meteorological treatises; though my experiments lead to a directly opposite conclusion. I shall endeavor to make myself better understood, starting from the data themselves which form the basis of this theory.

Let us suppose two pairs of thermometers, enclosed in their metallic cases, and suspended by means of thread, or metallic supporters, according to the method described in my first letter. Suppose each of these pairs to be composed of one thermometer with a bright case, and another blackened. And suppose, finally, that, during a still and clear night, one of these pairs should be fixed very near the grass in an open meadow, and the other at four or five feet high, so that the two thermometers of each pair shall be at the same level.

After being exposed thus for a few moments, the blackened thermometers will be seen to descend about 1·5° (Centigrade) below the metallic thermometers placed by the side of them. The temperatures, however, indicated by the lower couple will be very different from the temperatures marked by the higher two; this difference will reach to five or six degrees in calm and clear weather; and as the lowest indications will always be shown by the pair lowest in position, we infer from this that the differences observed between the indications of the two pairs of instruments, proceed solely from the unequal temperatures of the different atmospheric strata in which they are placed, and that consequently, during calm and serene nights, the temperature of the air decreases rapidly as we approach the surface of the ground. Now the experiment upon which the theory of Wells is founded, consists in the often repeated observation that a common thermometer, placed in contact with the grass, indicates a considerably lower temperature than one elevated four or five feet from the ground; from which it has been concluded that the grass was cooled several degrees, by radiating towards the sky. But it is easy to satisfy one's self that this deduction is by no means correct. Place the glass bulb of one of your thermometers in contact with the grass, and hold the other freely Suspended in the air, at the same distance from the ground, and you will find that the two instruments mark the same degree. Now no one will deny that this is not the way we should operate, to demon

strate, according to the old method, the reduced temperature of the grass below that of the surrounding medium. We are, therefore, compelled to admit that the fundamental data of the theory of Wells are inconclusive:-1st, Because the surfaces of the thermometers used have as much radiation as the leaves of the grass. 2nd, Because the thermometer intended to measure the temperature of the air was placed in an atmospheric stratum much warmer than that which surrounded the vegetable leaves placed in contact with the other thermometer.

The principle of the formation of dew, in consequence of the cold produced by the radiation of bodies is, I repeat, entirely just, and the theory of Wells inaccurate. The cause of this inaccuracy is evidently in consequence of having entirely neglected to consider the influence of the air in the production of the cold which is successively developed near the surface of the earth. It has, indeed, been said, vaguely enough, that radiating bodies, situated at a certain height, will not have their temperature lowered as much as those placed near the ground, on account of the descending currents which are formed around the former, and cannot exist around the latter. But that is insufficient to show the real agency which the air has in the formation of dew.

It was necessary to demonstrate, as I believe myself to be the first who has done, that, notwithstanding its incapacity of cooling by radiation, the air which rests very near the earth contributes greatly to lower the temperature of plants surrounded by it, by means of a series of actions and reactions, of which the cause and the effects are clearly defined, if I mistake not, in the second of the two letters which form the subject of this discussion. You, who have perfectly comprehended their true meaning, will doubtless permit me to spare you the fatigue of a useless repetition, and to recommend to the conductors of the journals just mentioned, a somewhat more attentive perusal of the numbers of the Comptes Rendus, in which they were inserted. After which they shall have full and entire liberty to prove that I am wrong; but they will first admit, at least I hope so, that they have wrongly. informed their readers with regard to the consequences resulting from my investigations of the phenomenon of dew.

New Mode of Making Artificial Magnets.

SIR, AS magnetism is now applied to many purposes of practical utility, as well as in many instructive and amusing experiments, and as artificial magnets can easily be made of greater power than loadstones, I presume it will be acceptable to many of your readers to know how artificial magnets may be made without any risk of fail

ure.

Makers of magnets and compass needles know that if the most careful and skilful workman be employed in preparing a number of magnets from the same steel bar, the magnetic power of the magnets, when compared with each other, will greatly vary; although every possible care may be taken in forging, tempering, and magnetising, in

uniform way. Experience in these matters convinced me that discrepancies in the magnetic powers of magnets of the same length, weight, and quality of steel, arise from the tempering alone; for if the metal be heated in a furnace, or coal fire, one par of the bar may be in contact with glowing coal, another part in flame, a third in heated air, a fourth in contact with coal in a state of ignition, &c. ; consequently, the metal is not in all its parts raised to the same temperature, when suddenly removed from the fire and plunged into a cooling fluid. The relation between the degrees of heat in the heating and cooling mediums is absolutely unknown; magnets tempered in this uncertain way will possess different degrees of hardness throughout their length, and their capacity for magnetism in all their particles will be unequal and uncertain.

Reasoning in this way, it appeared to me, that in order to make compass-needles and steel magnets successfully, we require specific heats in the warming as well as in the cooling process of tempering; in order to insure the same degree of hardness throughout the steel bars; that is to say, that the metal when cold should be a homogeneous mass, and possess a uniform capacity for the reception and retention of the magnetic virtue.

Metalurgists know that lead melts at a low temperature, but if left on a good fire, it gets first a red, and then a white heat; that continuing to absorb caloric, it nltimately boils at a uniform heat, which melts. gold or silver, as is evident in the process of "cupelation." Now here we have a specific heat at probably 5000°, and we have also a specific heat of boiling water at 212°. It therefore struck me that by heating my needles in boiling lead, and cooling them in boiling water, every particle of the steel would be first raised to, and then cooled down to the same temperature and degree of hardness. experiments have been made with complete success; and more powerful magnets have been made in this way than were ever made before, without risk of failure.

The

Magnets weighing 600 grains, and 6 inches in length, have held in suspension 14 times their own weight; and compass-needles have given by deflection 30°, at twice their length, from a test-needle. I find that magnets tempered in this way are not liable to break, but possess with great hardness a toughness, derived probably from the boiling water. Their power of retaining the magnetic energy has for four years remained unimpaired, although left without "keepers." To manufacturers of magnets and makers of compass-needles, a knowledge of this mode of tempering steel, at a specific temperature of the heating and cooling mediums, will enable them to make articles of a superior quality without risk of failure, or needless expense; and as I have no other object in view than the advancement of useful knowledge, I now send you these remarks for publication.

In heating the steel, the bars require to be pressed under the surface of the boiling lead (as the steel would otherwise float on its surface), and the magnet should be suddenly shifted from the lead to the boiling water, the instant it has acquired the temperature of the boiling lead; for to leave it longer in the lead would spoil the smooth surVOL. XV,-3RD SERIES.-No. 1.—January, 1848.

6