294 SCIENCE PROGRESS 



figures for gases do not tend to discredit it. The tendency 

 certainly exists. It is most marked in the unsaturated nitrogen 

 compounds, but the presence of the same tendency in NH 3 and 

 HN0 3 shows that unsaturation is not the only cause. 



There can be no doubt that, if some good explanation could 

 be found for this group of facts, we should have a valuable 

 means of attacking the problem of refractivity. But the facts 

 are puzzling, and until a large quantity of fresh data are available 

 for discussion it is not likely that the riddle will be guessed. 



The third problem on which it is probable that the study of 

 refractive indices may throw some light is the nature of the 

 atom itself and the part which electrons play in the structure. 

 It is in relation to this question that the study of dispersion 

 assumes importance. So little work has, however, been done 

 on comparative dispersions that it would be useless to devote 

 much space to an examination of a subject by no means simple, 

 and a very brief and elementary account of the position must 

 suffice. 



In a vacuum light of all wave lengths travels with the same 

 velocity. In material media not only are all wave lengths 

 retarded, but the retardation of the long red waves is generally 

 about 1 to 2 per cent, less than that of the short violet waves. 

 In a few cases, however, the retardation of the shorter waves 

 is less than that of the longer, and the dispersion is then said 

 to be anomalous. To explain and correlate these facts is the 

 problem attacked in the study of dispersion. The theory which 

 finds most approval at the present day, and which is strongly 

 supported by many lines of evidence, is that the retardation of 

 light in passing through a material medium is caused by the 

 expenditure of energy in setting and keeping in motion the 

 electrons which form part of the atom, or, possibly, the atom 

 itself. Whatever the configuration of the parts of the atom may 

 be, they are so balanced by the forces acting upon them that 

 they have a natural period of vibration of their own. If upon 

 a system of such atoms, forming a gas, a beam of white light, 

 which contains waves of all wave lengths, is allowed to fall, 

 the waves whose frequency is very different from the natural 

 frequency of the parts of the atom are little affected by the 

 vibrations of those parts, and lose little of their energy and of 

 their velocity. But as the period of the light in question 

 approaches that of the parts of the atom, its energy is more and 



