sulphur. Alum, which is highly transparent, will not transmit heat enough to deflect the needles more than 6°, when a piece of smoky quartz, fifty times as thick, which we cannot see through at all, will transmit sufficient to produce a deviation of 19°. Gold-leaf so thin that a landscape can be seen through it obstructs the passage of heat. Metal, wood, and marble, which wholly stop the light, also intercept the heat. Bodies which are highly diathermanous transmit the most rays, and without being much heated. The radiation from a red-hot ball is felt at a considerable distance, though the temperature of the surrounding air is not raised. The heat of the sun, when concentrated by a lens on the delicate web of the spider, does not consume it. Melloni agitated fluid screens in various ways, and fiddled upon solid ones, without disturbing the transmission of heat through them. Biot and Dove made a similar observation on light.

It is interesting to know whether any distinction belongs to the rays proceeding from different sources of heat, or to the various rays from the same source analogous to the colors of light. There are four ways in which the colors of light are analyzed :—1. By reflection; 2. By refraction; 3. By absorption; 4. By interference. If the rays of heat possess specific characters corresponding to the colored rays of light, it will be difficult to verify the fact, as we have no eye for heat, as for light, which can perceive at a glance the difference in question. Red waves of light differ from violet waves, not only in the sensation they awaken, but in length and quickness of succession; and this is a difference of which the eye takes no direct cognizance. Moreover, a special refrangibility distinguishes, if not the simple colors, at least such as are found in the solar spectrum. Now, the amount of refraction and the length of the wave admit of measurement, and with the same facility for the waves of heat as those of light. All the fundamental mechanical variety on which color depends is as easy of conception in a wave of heat as a wave of light, and, if it exist in nature, is as easy of demonstration. We shall first consider the specific qualities of the rays of heat, as indicated by absorption. Some bodies, as air, glass, water, absorb the same proportion of all the colors, so that, when looked through by sunlight, they appear white. Others absorb more of one color than another, so that the light which traverses them is discolored, and the bodies themselves appear of the color which they transmit most freely. Now, diathermanous bodies manifest the same elective absorption for heat as for light. A few, such as rock-salt and air, transmit all kinds of heat in the same proportion; in respect to heat, they are white. Some bodies, however, which are white in reference to light, are colored in relation to heat; and some bodies may have one color for light, and another for heat. Alum, which is white to light, is tinged with violet as to heat; that is, it transmits only the most refrangible rays of heat. The same is true, in a less degree, of white glass. Colored glasses, with the exception of green and black, are not colored for heat. They transmit the same kinds of heat, as Melloni shows by sending the rays of heat first through them, and then through a piece of alum. The alum, though possessing so decided an

elective absorption, transmits the same quantity of rays from whatever glass they have emerged. The alum, however, will not transmit the rays that have passed through the black or green glass, because these are the least refrangible, or the red rays. A plate of alum, with one of green or black glass, is as impervious to all kinds of heat as a blue and a green glass united are to light. Most bodies transmit the more refrangible rays of heat in greater abundance than the less refrangible rays, so as to have more or less of a calorific tinge of violet. Melloni prepared an artificial medium, smoked rock-salt, which stopped the more refrangible rays and allowed the others to pass. The calorific color of such a body is red. Forbes states that laminated mica, split by heat, possesses the same property. As rocksalt is the only known substance, except the air, which transmits heat without discoloring it, it is indispensable in calorific experiments analogous to those optical experiments in which we use white glass. It is the true glass for heat, as Melloni expresses it; and lenses and prisms should be made of it. Tepid water placed in the focus of a lens of rock-salt sends forth a beam of parallel rays of heat, which will affect a differential thermometer with small bulbs at a great distance. It becomes a Pharos of

heat.

4. Refraction. In his second memoir on transmission, which Melloni presented in 1834, the refraction of heat is also discussed. A prism of rock-salt is mounted so as to be traversed by a parallel beam of rays, and the battery placed so that the refracted heat will fall upon it. He ascer tains that the rays from the most intense sources of heat have the greatest refrangibility; those from the sun being refracted nearly the same as the rays of light. All the rays which come from any one source have not the same refrangibility. The sunbeam may contain some rays like those of terrestrial origin; and artificial flames may emit a few kindred to those which proceed from the sun. The rays of different refrangibility for the experiments on transmission were obtained by using heat from sources of different intensity. In uncrystallized bodies, the diathermancy increases with the intensity of the heat, that is, with the refrangibility; but the same rule does not hold good for crystallized media.

The dispersion of solar heat, which has long perplexed physicists, was fully investigated by Melloni, and the results of his labors were published in 1844.* Since the time of Newton's experiments on the dispersion of light, it had been assumed that the heat of the different portions of the spectrum was proportional to their illuminating power. The experiments of Landriani, Rochon, and Sennebier, which placed the maximum calorific effect in the yellow, near the orange, seemed to confirm this view. About 1800, researches were made by Sir Wm. Herschel, which placed this maximum outside of the red. Malus and Berard performed some experiments in presence of Berthollet, which indicated that the heat extended sensibly beyond the red, though the maximum was in it. A host of physicists of

* Compt. Rend., XVIII. 39.

high reputation for delicate manipulation and intellectual gifts followed in the investigation, among whom we may enumerate the names of Leslie, Englefield, Wünch, Ritter, Ruhland, and Davy. As some disputed, while others defended, the conclusions of Herschel, the question still remained in as unsatisfactory a state as before. The difficulty seemed to clear up a little when Seebeck showed, in 1819, that prisms of water, sulphuric acid, alcohol, and crown and flint glass give the maximum in different positions. With the water prism, it is in the yellow; with the sulphuric acid and alcohol, in the orange; with crown glass and some kinds of flint, in the red; in other kinds of flint glass, outside of the red; and more in the English than in the Bohemian. The fidelity of the old experiments was thus vindicated, but not explained. After Melloni's experiments on the elective absorption of media, there can be no doubt as to the origin of this shifting maximum. The heat sent through rock-salt is the only true white heat; and the position of the maximum in a spectrum formed by such a prism is its true position. Melloni finds, that, in this case, it is as far on one side of the red as the yellow is on the other side. This is the normal spectrum of heat. Any other would be as imperfect as the luminous spectrum would have been had Newton performed his grand experiment of dispersion with colored glass. Melloni tests the accuracy of these views, by sending the calorific spectrum, derived from a prism of crown glass, through a layer of water; the maximum leaves the red, and marches towards the violet, because the water absorbs the less refrangible rays in larger proportion than the most refrangible. He displays the thermochroic influence of other limpid bodies, such as alcohol and crown and flint glass, by sending the normal spectrum through each. The maximum moves towards the violet, but more for the first than the last.

5. Polarization. Melloni's Memoir on the Polarization of Heat appeared in 1836,* and the sequel in 1837. Berard, of Montpellier, announced, as long ago as 1817, that heat, not only from the sun, but from terrestrial sources, luminous and non-luminous, was capable of double refraction and polarization. Berard employed Malus's contrivance of two mirrors, mounted in a tube, and detected, as he supposed, a difference in the intensity of the heat, depending on the relative azimuthal position of the planes of reflection. These experiments were not guarded by screens and otherwise. Still, as they were repeated in the presence of Berthollet and Dulong, the results were enrolled among the established truths of science, until doubts were cast upon their accuracy by the failure of Powell,+ in 1829, to realize them. In Melloni's second Memoir on the Transmission of Heat, he refers to an experiment which he made to polarize heat by crystalline absorption. He assures the reader that no difference was perceived in the effect upon the thermoscope, whether the axes of the tourmalines were crossed or parallel, although, in the former case, the light * Ann. Ch. Ph., LXI. and LXV. Edinb. Jour., VI. and X., 1829.

↑ Société d'Arcueil, III. 5.

was so much diminished, that the incandescent platinum was no longer visible. Nobili, though in possession of very delicate thermoscopes, failed to polarize heat by reflection. In 1834, Professor Forbes read a paper before the Royal Society of Edinburgh on the polarization of heat by tourmalines; also by reflection and refraction; and on depolarization and double refraction. He followed up the subject in a second and third series of experiments, accounts of which he read to the Royal Society of Edinburgh in 1836 and 1839. In one of these he speaks of a successful exhibition of the circular polarization of heat. In his first paper, Mr. Forbes states that the amount of polarization which he discovered was .29 for the heat of an Argand flame; .24 for that of a Locatelli lamp; .36 for that of an alcohol flame; .40 for that of incandescent platinum; .22 for copper heated to 400° Fah.; .17 for the heat radiated from an iron vessel containing mercury at 280°; and only .06 for that of boiling water. The polarizing apparatus consisted of two bundles of mica plates, each of ten laminæ, which produced a polarization of all the colors of light equal to 90 per cent.

The great want of diathermancy in tourmalines is a serious obstacle in attempting to show the polarization of heat by them. To economize his heat, without making the distance between the various parts of the apparatus so small as to introduce secondary radiation, Melloni placed the source of heat in the focus of a metallic mirror, and sent it out in a parallel beam. He next received it upon a lens of rock-salt, two and a half inches in diameter; then it passed through the tourmalines, and afterwards upon another lens of rock-salt, fourteen lines in diameter. The transmitted rays of heat went to the battery nearly parallel. The second lens served to disperse the rays of secondary radiation, should there be any. With these precautions, sufficient heat was obtained from a Locatelli lamp, at the distance of a metre, to produce a deviation of 60° or 80° in the needles. Melloni repeats now his old experiment with tourmalines, and produces a polarization of heat, varying with the color of the tourmaline, from 3.71 to 21.89 per cent. Melloni found, that, with some specimens of tourmaline, the proportion of polarized heat increased with the elevation of temperature, while, in other specimens, it diminished. He next took that pair of crystals in which the polarization was most complete, and sent the heat through them after it had been sifted by passing through various diathermanous bodies. Heat which had passed through distilled water was polarized 66 per cent.; that from alum, 95.81. Colored glasses did not alter the proportion, except black and green. The first reduced the proportion from 21.89 to 1.51, the second to 2.76.

In the second part of this memoir, Melloni discusses the polarization of heat by refraction and reflection. Bundles of thin glass, which answer for the polarization of light, cannot be used with advantage for heat, on account of their want of diathermancy. Rock-salt, which is the best material for

* Edinb. Phil. Trans., XIII.

this purpose, cannot be procured in abundance, and of a pure quality. Melloni selects plates of mica, as Forbes had already done. Forbes effected the cleavage of the mica into very thin plates by heat. Melloni splits it with the lancet, and then unites the parts again by glue, in their original position, having placed a rectangular frame of paper between them. He eliminates the doubly refracting structure of the mica by so placing it that the plane of refraction coincides with one of the neutral sections of the crystal, when it will act like glass, or any other amorphous body. The number of laminæ in a bundle varied from three to twenty. The results of Melloni differ widely from those of Forbes. The specific polarization of different kinds of heat by tourmalines is easily understood, as the specific diathermancy of the substance would have play. Indeed, there are certain rare specimens of tourmaline which polarize the colors of light unequally. Biot mentions one in his possession which did not absorb and polarize the red as it did the other colors. But Forbes's experiments show an equally remarkable difference in the polarization of various kinds of heat by refraction, which Melloni attributes to a defect in the experiment. Melloni thinks that the distance from the bundles of mica to the thermoscope (five and a half inches) was so small that the latter was influenced by secondary radiation. Moreover, when Forbes altered the distance from the thermoscope to the source of heat, to make the deviation of the needle independent of the intensity of the source of heat, he did not allow for the change in the amount of polarization which arose from that in the angle of incidence. The second series of experiments is less objectionable; the distance from the source to the bundles is three times as great as before, and is constant throughout the series. The improvement betrays itself in the results, which exhibit, not only a general increase, but a much closer approximation. The polarization of the Argand flame is 73 per cent., and of the boiling water, 44.

Melloni guards against the error that would be induced by a change in the angle of incidence, by placing the source of heat in the focus of a lens of rock-salt. Thus he can increase his distances, and still preserve an intense parallel beam of rays, sufficient to produce large deviations in the needles. The two bundles of mica are close together, and removed half a metre from the radiant centre, and two or three tenths of a metre from the battery. The deviation is rendered constantly about 35°, for the different intensities of the rays, by a mirror placed behind the heated body, and covered more or less with lampblack. Melloni's experiments conduct him to the following conclusions:

1. The proportion of polarized heat increases with the angle of incidence, measured from the normal.

2. With a large number of laminæ, a maximum is reached when this angle has been increased to a certain value, and the polarization remains constant for all larger angles.

3. The angle at which this constant maximum begins is less as the number of laminæ is greater.