Page:Popular Science Monthly Volume 83.djvu/248

244 1. Neutral "aggregates" of charged particles possessing, in general, translatory and rotatory energy. When undisturbed from without these aggregates would have little if any external electric field. When the equilibrium of such a system is disturbed by collisions or by electromagnetic waves, it may possess temporary fields that will serve as the source of heat and light radiation. This radiation may be due to a rapid oscillatory motion that may be radial, transverse or tangential and would probably be characterized by a definite period. On account of the magnitude of the forces necessary for stable equilibrium, the period of the radiation would probably be small. The spectroscopic models of Thomson, Nagaoka and others are of this type.

2. "Aggregates" may possess charged parts; these may be so far apart from each other that local fields of considerable intensity may exist. If such an aggregate were to rotate, an alternating electric field would result and radiations would be emitted. This radiation, depending on a central acceleration, would vary in period with each impact, so that the various periods emitted would vary about a mean, which would depend on the average rotational energy before impact and the nature of the impact.

3. Freely charged particles torn from neutral "aggregates" will radiate energy when their velocity is changed. The quantity of this radiation will vary with the velocity and the acceleration. The breaking up of the neutral "aggregates" may be called ionization if the resulting parts are charged. Ionization processes may take place within molecules and this is believed to be the condition existing in many kinds of organic compounds when they absorb light or heat.

The relation between ionization and luminosity is not yet clear. Some physicists believe that the two are related to each other and that luminosity becomes perceptible when the intensity of ionization is sufficiently great. It has been stated that a gas may become luminous when one molecule in every $$(10)^{7}$$ is ionized. This would mean that the (expenditure of about $$(10)^{-5}$$ ergs is necessary to excite luminosity.

There appears to be considerable evidence supporting the view that some band spectra such as those of bromine and iodine may be due to the dissociation of molecular systems or to a recombination of the dissociated parts. Ladenburg has found that luminous hydrogen gives an anomalous dispersion in the neighborhood of H_{\alpha} while this kind of dispersion is absent in ordinary hydrogen. The phenomena of dispersion indicate that different series of lines in a spectrum may be emitted by very different kinds of vibrating centers, while a particular center may emit only a single line of a series, depending on the manner of its excitation. Faintness in the intensity of lines may be due to the fact that there are very few light centers emitting the given line, or that the