As regards reflection from curved surfaces, the identity also holds good. Receiving the beam from our electric lamp on a concave mirror (_m_ _m_, fig. 49), it is gathered up into a cone of reflected light rendered visible by the floating dust of the air; marking the apex of the cone by a pointer, and cutting off the light by the iodine solution (T), a moment's exposure of the pile (P) at the marked point produces a violent deflection of the needle. The common reflection and the total reflection of a beam of radiant heat may be simultaneously demonstrated. From the nozzle of the lamp (L, fig. 50) a beam impinges upon a plane mirror (M N), is reflected upwards, and enters a right-angled prism, of which _a_ _b_ _c_ is the section. It meets the hypothenuse at an obliquity greater than the limiting angle,[23] and is therefore totally reflected. Quenching the light by the ray-filter at F, and placing the pile at P, the totally reflected heat-beam is immediately felt by the pile, and declared by the galvanometric deflection. [Illustration: Fig. 50.] § 7. _Invisible Images formed by Radiant Heat._ Perhaps no experiment proves more conclusively the substantial identity of light and radiant heat, than the formation of invisible heat-images. Employing the mirror already used to raise the beam to its highest state of concentration, we obtain, as is well known, an inverted image of the carbon points, formed by the light rays at the focus. Cutting off the light by the ray-filter, and placing at the focus a thin sheet of platinized platinum, the invisible rays declare their presence and distribution, by stamping upon the platinum a white-hot image of the carbons. (See fig. 51.) [Illustration: Fig. 51.] § 8. _Polarization of Heat_. Whether radiant heat be capable of polarization or not was for a long time a subject of discussion. Bérard had announced affirmative results, but Powell and Lloyd failed to verify them. The doubts thus thrown upon the question were removed by the experiments of Forbes, who first established the polarization and 'depolarization' of heat. The subject was subsequently followed up by Melloni, an investigator of consummate ability, who sagaciously turned to account his own discovery, that the obscure rays of luminous sources are in part transmitted by black glass. Intercepting by a plate of this glass the light from an oil flame, and operating upon the transmitted invisible heat, he obtained effects of polarization, far exceeding in magnitude those which could be obtained with non-luminous sources. At present the possession of our more perfect ray-filter, and more powerful source of heat, enables us to pursue this identity question to its utmost practical limits. [Illustration: Fig. 52.]