The radiations of a heated "black body" (q.v.) include theo retically the whole spectrum, but the only portions of measurable intensity with laboratory sources are the visible and infra-red. Planck's law giving the distribution of energy throughout the spectrum of black-body radiation at any temperature (see HEAT, QUANTUM THEORY) has been experimentally confirmed, within the limit of experimental error, from •5/.4 to 5o,u, the validity in the long wave region 4/4 to 52,u having been the object of a special research by Rubens (1921). The theoretical deduction of the formula, although Planck himself assumed the existence of reso nators or vibrators within the solid, need not involve any assump tion as to the nature of the vibrators involved in the proof, the reasoning being of a thermodynamical nature, which demands no precise mechanism. It is possible, as Einstein has shown, to deduce the formula on the assumption that the radiation is given out by atoms executing quantum jumps, as demanded by Bohr's theory, but as the distribution of radiation is quite independent of the atomic nature of the black body the precise character of the quantum jumps is immaterial. The radiation from a heated solid cannot be atomic, in the sense that the radiation from a gas is, for the behaviour of the atomic electrons must be influenced by the electric field and radiations of the surrounding atoms. It is, then, probably something which partakes of the nature of both an atomic and a molecular radiation, in that the atoms cannot be considered as rigid wholes, as they can for the infra-red bands, nor yet as independent entities within which quantum changes take place, as they can for the line spectra. Vibrating systems of every possible frequency are demanded by the theory, and they can be provided by some such scheme as that sketched, which is necessarily vague, as there is at present no precise knowledge of the mechanism of radiation from solids.
The generation of line spectra in the visible and ultra-violet region has been explained in terms of Bohr's quantum theory of spectra which is fully discussed under ATOM; QUANTUM THEORY, SPECTROSCOPY. Different stationary states of the atom are deter mined by the behaviour of the outermost electrons of the groups of electrons which surround the nucleus, and passage from one stationary state to another is accompanied by the emission of a radiation of given frequency. By this mechanism of quantum changes among the outer electrons a vast range of frequencies is emitted, for Brackett's hydrogen lines in the far infra-red, to which reference has already been made, are generated in this way, while Millikan's lines in the extreme ultra-violet are due to such changes in atoms which are already multiply ionised, that is, have lost outer electrons. The band spectra in the visible and ultra-violet are likewise due to quantum changes among the outer electrons in atoms constituting a molecule, the molecule executing changes of its own at the same time. The characteristic mechanism of the visible and ultra-violet is therefore an affair of the outer electrons.
The X-ray line spectra originate in similar quantum changes among electrons, but here it is the inner electrons of the atom that are involved, among which are associated with jumps much greater changes of energy than those among the outer electrons.
(See X-RAYS, NATURE OF: Röntgen Rays.) The 7-rays are not generated in the electron distribution round the nucleus, but pro ceed from the nucleus itself. The mechanism of their production is still uncertain. (See NUCLEUS.) The origin of the very pene trating cosmic radiation is still less understood; reference has already been made to Millikan's theory that it is due to atomic nuclei in the process of manufacture.
To sum up, then, corresponding to wide range of wave-lengths of the electromagnetic spectrum there is a range of generators of widely different sizes. The longest waves are generated by large installations of coils of wire, condensers and so on, and generators constructed on these principles can be reduced in size until the oscillator system consists of two minute metal filings. Then there are systems of vibrating molecules such as we have in crystals and in solid bodies in general, which generate radi ations in virtue of their heat motion. Another type of generator is the single molecule, effective in the gaseous state, which gives rise to infra-red radiations as long as the atoms can be treated as minute rigid bodies. When changes within the atom itself are taken into account we have another type of generator, and this type can be further subdivided according as the inner or the outer electrons are considered. When the outer electrons only are involved in the generation of radiation we can treat the nucleus and innermost electrons as forming a rigid system : when the inner electrons are in question we can still treat the nucleus as a minute rigid body. Finally, we have to consider the nucleus itself as a generator of vibrations, so that, descending the scale of wave-lengths, we start without the need of considering any atomic or molecular mechanism, and there in succession can treat molecule, atom and nucleus as rigid, finally being forced to take into consideration changes within the nucleus itself.
The ranges of wave-length generated by the types of system just outlined overlap one another widely, as has been already indi cated. Changes in the atom can generate lines in the far infrared: some nuclear 7-rays are of longer wave-length than the hardest X-rays, generated by the extra-nuclear electrons. From the purely empirical point of view it may be best to group the radiations according to wave-length, or according to methods of measure ment, especially as a sound theoretical basis is still to seek in some regions: from the theoretical point of view it seems preferable to group the radiations according to their method of origin. A rough analogy from acoustics, which must not be pressed too far, may be offered. Air vibrations might be grouped according to wave-length, or according to origin. In the latter case the division might be into inaudibly low notes originating in the vibration of planks : drum notes, piano notes, and whistle notes. Such a classification is quite without the theoretical meaning that attaches to the subdivision of electromagnetic radiations, accord ing to classes of natural vibration, but may serve to distinguish the two points of view.