There also seems to be a limited range of waves which can induce this fluorescence, and this range also depends upon the fluorescing substance. As a rule, the inducing waves are shorter in length than the induced fluorescence, but this rule has some very marked exceptions. The fact that only a limited range of waves produces fluorescence explains a noticeable characteristic of the phenomenon. If the fluorescing solutions are at all strong the fluorescence is confined to the region close to where the light enters the solution, thus showing that the rays which are responsible for inducing the glow become rapidly absorbed, whereas the remainder of the light goes on practically unabsorbed. +Phosphorescence.+--Sometimes the emission of the induced light continues for some time after the inducing waves are withdrawn, and then the phenomenon is termed phosphorescence, since phosphorus emits a continuous glow without rise of temperature. Sometimes the glow will continue for several hours after the exciting rays have been cut off, a good example of this being provided by Balmain's luminous paint, which is a sulphide of calcium. With other substances the glow will only continue for a very small fraction of a second, so that it is impossible to {60} say where fluorescence ends and where phosphorescence begins. In order to determine the duration of the glow in the case of these small times, an arrangement consisting of two rotating discs, each of which have slits in them, is set up. Through the slits in one of them the substance is illuminated, and through the slits in the other the substance is observed while the light is cut off. By adjusting the position of the discs with regard to each other the slits may be made to follow one another after greater or shorter intervals, and so the time of observation can be made greater or smaller after the illumination is cut off. All the bodies which have been observed to exhibit phosphorescence are solid. +Theory of Fluorescence.+--It is fairly simple to imagine a mechanism by which fluorescence might be brought about, as we might assume a relation between the periods of oscillation of certain types of electron in the substance and the period of the stimulating waves. Thus resonance might occur, and the consequent vibrations of the electrons would start a series of secondary waves. If, however, we assume resonance, it is difficult to see why there is a range of wave-lengths produced and another range of wave-lengths which may produce them. We should have expected one definite wave-length or a few definite ones producing one or a few definite wave-lengths in the glow, while if a whole range of waves will produce the effect it is difficult to see why all bodies do not exhibit the phenomenon.