which is proportional simply to the deviation. At the present time, many physical cabinets in Europe and this country are adorned with an apparatus made after this model, which is known throughout the scientific world as Melloni's thermo-electric thermoscope. Ruhmkorff, of Paris, has acquired an extensive reputation for the manufacture of this apparatus, which, as it leaves his hands, is an elegant instrument, displaying beauty, compactness, and unity of plan, admirably arranged for the convenient repetition of the manifold experiments of Melloni on radiant heat, which we now propose to recapitulate. The apparatus consists of a wooden platform, with levelling screws. On two uprights is supported a graduated beam of massive metal, a metre in length. The movable parts, used in the various experiments, can be attached by binding screws to this beam, and the distances easily arranged. The instrument is furnished with four sources of heat, - the alcohol flame, the Locatelli lamp, a spiral of platinum made incandescent by a flame, and a cubical box of metal for boiling water. In the delicate experiments on polarization, Melloni inclosed the whole apparatus in a case with double walls. He dispenses with the microscope, used by Forbes to read the deviations of the needles.

As a luminous body emits rays of light in diverging lines, so does a hot body project rays of heat. It is not necessary to decide now whether these rays, mechanically considered, are to be represented by delicate projectiles, or by waves propagated through a disturbed ether. Radiation is easily distinguished from conduction and convection of heat. The former requires no other vehicle than the attenuated ether; the latter suppose some more substantial medium, which, in the case of convection, must be fluid. Radiation goes on, not only through bodies which are poor conductors of heat, but also in the most perfect vacuum that human ingenuity can produce, as the experiments of Davy, Dulong, and Petit attest. Moreover, radiant heat conforms to the general rule of radiant forces; namely, that the force exerted is proportional to the inverse square of the distance. In conduction, the heat diminishes in a geometrical ratio, while the distance increases by an arithmetical progression. The most intense heat that can be applied to one extremity of a bar of iron six feet in length, even if sufficient to melt it, will not be able sensibly to warm the other end. Again, the conduction of heat is slow, whereas its transmission by radiation is almost instantaneous. When Pictet had placed a heated body in the focus of one mirror, he observed that the effect was instantly perceived at the conjugate focus of another, sixty-nine feet distant. Wrede has remarked, that, if the velocities of solar light and heat are different, the amount of aberration will not be the same for the two elements; and the luminous and calorific images of the sun will not be superposed. He thinks he can detect a displacement of this kind, by which the temperature of the sun's eastern limb exceeds that of the western by a fraction of a degree. The amount of displacement indicates an excess of aberration of the heat above the light of twenty-five per cent., and consequently a velocity so much less. If the experiments hitherto made are insufficient to settle so nice a point as

this, they prove, in general, that the velocity of solar heat is of the same order as that of light, and that radiant heat cannot be confounded with the sluggish process of conduction. Conductors of heat allow it to spread in all directions through them, and are themselves heated. Radiation travels in straight lines from its source, and passes through media without raising their temperature. Still less can radiation be mistaken for convection, which is only upward.

When the rays of heat which have radiated from their primitive source fall upon a body, they are apportioned into three divisions;—1st, the absorbed rays; 2d, the reflected rays; 3d, the transmitted rays. The rays which are absorbed are again radiated, as from a new centre. The rays which are transmitted are refracted and dispersed. The rays which are reflected or transmitted are polarized. The whole subject of radiant heat may be very properly divided into five parts: -1. Absorption, including secondary radiation; 2. Reflection; 3. Transmission; 4. Refraction and Dispersion; 5. Polarization.

1. Absorption. — With_regard to all those bodies which are incapable of transmitting heat, the absorbing power is the reciprocal of the reflecting power. It has been known since the experiments of Hooke and Franklin, that the absorption of the solar heat increases with the darkness of the color of the body; that is, with the absorption of light. The solar beam contains, we now know, both dark and luminous rays. Black bodies absorb both, and are much heated. White bodies absorb only the first, and are less heated. The luminous part of terrestrial rays is subject to the same law of absorption; but the dark rays, like the dark rays of the sun, are absorbed more or less, not according to the color, but the texture of the body on which they fall. What is known with respect to the connection between texture and absorption, we shall consider under the head of reflection. We will only remark here, that Melloni* found that paints ground to different degrees of fineness did not, if of the same color, alter the relative proportion of the absorbed and reflected rays. It also appeared that the proportion of heat absorbed was not the same when it was received directly from its source as when it was stopped by one body and then thrown by secondary radiation into another body.t

Bodies that absorb heat become hot, and form independent centres of radiation. As several bodies, if exposed to the same heat, become heated to different degrees, so these same bodies, if brought to the same temperature, will not radiate it out again with equal facility. Ritchie's ingenious experiment is familiar, by which he showed that bodies radiated heat with the same facility as that which governed the absorption; therefore, the best radiators are the poorest reflectors. Some kinds of surface radiate seven or eight times better than others. Rumford proved that radiation went on, not at the surface merely, but from underneath. He added from one to four layers of varnish to a body, and improved the radiation

* Compt. Rend., 1840, XI. 659 - 678; XII. 375.

† Ibid., 1838, VII. 298.

by each additional coat. Leslie persevered in this till he had shown there was a limit of thickness, beyond which the radiation did not increase. Melloni* undertakes to discover whether this limit is at the same thickness, whatever the material. He applies to a body 19 coats of varnish, and finds that each addition is an advantage up to 16, when he obtains the maximum radiation of 40.9, that of a single coat being 9.3. He finds that the thickness of the 16 coats equals .044 of a millimetre. Gold-leaf is then added to the varnish, and it appears that, with this substance, the parts that radiate from beneath are not so far below the surface as for varnish. Melloni considers this experiment as a refutation of the theory of Prevost, Fourier, and Poisson, all of whom supposed the failure to radiate arose from internal reflection. For why could not the heat leave the varnish and come into the gold? Melloni also refers to an experiment exhibited by him in presence of our own countrymen, Henry, Bache, and Locke.

2. Reflection of Heat.-It had been frequently demonstrated by various experiments, previous to the labors of Melloni, that the law of reflection of light held good for heat also. The trial had been made upon luminous and dark heat, with plane and parabolic mirrors, and also with the frustum of a cone. Something was determined also in regard to the comparative reflecting power of different materials. No one, however, previous to the publication of Melloni's Memoir in 1835,† had aimed to show the precise number of reflected rays of heat as compared with the incident beam. Melloni selected two plates of the same substance, and so nearly alike in thickness that one transmitted just as much heat as the other. The excess of matter was split from the thickest plate, and placed at a little distance behind it. Since this thin portion produced no sensible absorption upon heat which had escaped from the principal portion, the heats transmitted and reflected by the thin portion were complementary to each other. If each was measured, the proportion of the transmitted or reflected portion to the incident beam could be assigned. It appears that the amount of reflected heat is sensibly the same at a perpendicular incidence, and for 25° or 30° from it; and even at larger angles of incidence it is not much increased.

The reflecting power of bodies which will not allow heat to pass through them is found by comparing the heat which they reflect with that which is reflected by a body which does transmit it; the incident beam being the same in both experiments. Melloni ascertains from his experiments, that a lens of rock-salt will have a focus twice as hot as the best metallic mirror of the same aperture. Few bodies reflect much heat, except the metals. Nobili and Melloni both state that polish produces much less effect on the reflection of heat than has been supposed, and none except with metal reflectors. Even with the metals, as we learn from Melloni, the improved reflection is not a consequence of the roughness and smoothness as such, but of the increased hardness of surface which the polishing has effected. In elastic bodies, as marble, ivory, and amber, whose density * Compt. Rend., 1845, XX. 575, 1796.

† Ann. Ch. Ph., LX. 402.

cannot be altered by attrition, smoothness does not improve the power of reflecting heat. Rough metal oxidates more easily than smooth metal, and the oxide absorbs better and reflects more poorly than the clean metal; hence the impression, that smooth metals reflect most perfectly. Melloni guarded his experiments by the use of gold and platinum, and roughened the surface by marking it with a diamond, that the substance used in polishing might not alter its metallic character. If the metal was hammered, the external density was diminished by scratching, and the surface did not reflect so well. But if the metal was cast, the scratching hardened the surface, and it reflected better when rough. This accounts for a fact noticed before by Dulong, that some new specula, furnished to the Polytechnic School, Paris, and of cast metal, did not work so well as the old ones, which, though smaller, were hammered.

3. Transmission. - Melloni's paper on the diathermancy of bodies, that is, their transparency to heat, was presented to the French Academy of Sciences in 1833. Before this investigation, the diathermancy of bodies to all kinds of heat was far from being satisfactorily established. The burningglasses of the Greeks tell us how long ago the penetration of glass by the sun's rays was known; and yet this material is often used as a screen against the intense radiation of a hot fire. The contrast herein betrayed between the heat from different sources was distinctly noticed by Mariotte,* who observed that, when a metallic mirror was placed five or six feet in front of a fire, the heat at its focus was very painful; but if a plate of glass was interposed, the focus, though as bright as before, was not sensibly warm. Lambert represents the focus in a similar experiment as scarcely warm. Scheele,t a century later, repeated the experiment with a similar result. Pictet, however, placed a candle or a vessel of boiling water in one of the foci of two conjugate mirrors, and changed the mercury of a thermometer in the other focus several degrees, though a very thin plate of glass was interposed. Sir Wm. Herschel § describes a successful experiment which he made on the transmission of heat, unconcentrated by reflectors. It has been objected to this experiment, that heat was absorbed at the surface of incidence, and then, passing by conduction to the opposite side, was sent to the thermometer by secondary radiation. Prevost, || of Geneva, contrived an experiment to elude this objection. He substituted for the plate of glass a layer of water, one fourth of a line thick, and spouted from a fountain through a jet of parallel plates. The heating and secondary radiation, above described, which require time, could not occur in this layer, constantly renewed. Yet Prevost found that a candle or heated ball on one side of the jet affected an air-thermometer on the other side.¶

* Traité de la Nature des Couleurs, Paris, 1636, Pt. 2, end of Introd.

↑ Traité de l'Air et de Feu, Paris, 1777, § 56.

Essai sur le Feu, 1791.

§ Phil. Trans., London, 1800.

Du Cal. Ray., 1809.

¶ Ritchie made a similar experiment on liquid held by threads. Phil. Trans., 1827.

*

De la Roche interposed between the heated body and the thermometer a plate of glass, and afterwards the same plate covered with Indian ink on the side towards the heat. The greatest effect is produced by the simple glass, though the absorption and secondary radiation must be then the least. Powell confirms the result of De la Roche, but he does not consider it as proving the transmission of heat. Leslie, in 1804, had tried two plates of glass covered on one side with tinfoil; as the effect was different according as the plates were in contact or otherwise, and according as the coated surfaces faced one another or not, it was referred to absorption and secondary radiation. This was the explanation adopted by Brewster, and received by Laplace, of all these ingenious experiments. Not much weight would be allowed, at the present day, to one of the arguments on which they confidently rested; namely, that a thick glass, though more transparent than a thin one, nevertheless intercepted the heat more effectually.

Thus perplexed had the subject of diathermancy grown, when Melloni applied himself to the study of it. He soon devised means for distinguishing the effect of conduction and secondary radiation from that of direct transmission. The heat absorbed, and then radiated, will vary in amount with every position which is given to the plate between the heated body and the thermoscope, and will be a minimum when it is midway between the two. The heat directly transmitted is independent of the position of the plate. When Melloni finds that the deviation of the rheometer remains constant in all positions of the plate, he concludes that the heat sent by absorption and secondary radiation is not sensible. He owes this exemption from an interference, so fatal to other experimenters, to the great delicacy of his thermoscope, which enables him to use large distances in all his experiments. For greater security, his rule is always to place the plate midway between the thermoscope and the source of heat. Still further, the effect of absorption and secondary radiation is slow to appear, and remains for a time after the original cause has gone. But in Melloni's experiments, the effect began and ended instantaneously, and could, therefore, be referred only to direct transmission.

Melloni's experiments were not confined to glass or water, but were made on every variety of substance, and with heat from the four sources elsewhere enumerated. The diathermancy of bodies is subject to the same variety, both in degree and kind, as their transparency. We must not infer from this that the most diaphanous substances are always the most diathermanous. Smoky quartz, which transmits heat as well as ordinary quartz, does not transmit light. Chloride of sulphur, which transmits 63 per cent. of the incident heat, intercepts most of the light, and appears of a deep red-brown color. Alum, on the contrary, which is very transparent, is highly adiathermanous. Spirit of turpentine transmits only 31 per cent. of heat; sulphuric ether, 21 per cent.; sulphuric acid, 17 per cent.; and water 11 per cent., though so much more diaphanous than the chloride of † Phil. Trans., 1816.

*Phil. Trans., 1826.