360 TRANSACTIONS OP SECTION B. 



acts partly as a constraint on the vibrations and partly as a barrier to the number 

 of encounters and partly as an absorbent of the radiant energy, so that the narrow 

 bands of the vapours are usually replaced by wide diffuse bands. In very thin 

 films the movements of the molecules are further restricted as a consequence of 

 the closer packing : but the selective absorption is not unlike that of the solu- 

 tions ; the chief difference being that the bands of the thin films are shifted more 

 towards the less refrangible regions. 



4. Absorption Spectra and Refractive Power of Metallic Vapours. 

 By Professor P. V. Bevan, M.A., Sc.D. 



More attention has lately been devoted to the absorption spectra of metals for 

 various reasons. The importance of the subject from the point of view of astro- 

 physics has been more generally recognised, and also the interest of the phenomena 

 from their character as a pure physical or chemical manifestation of the properties 

 of the molecule or atom. Research in these lines has been greatly stimulated by 

 the beautiful results obtained by R. W. Wood in the case of sodium. His ob- 

 servation of forty-eight of the lines of the principal series for sodium was a 

 remarkable extension of our knowledge of series spectra. My own work in this 

 part of the subject has been simply to extend the method of Wood to the other 

 alkali metals, and to find that in the cases of all of them a similar series can be 

 obtained. These series, the first and second members of which are the familiar 

 lines characteristic of the alkali metals, all appear as absorption lines when white 

 light is passed through the vapours of the metals. With increase of density of 

 the metal more lines come into view, and to obtain a much larger number than 

 have been measured yet requires only greater dispersion in the Spectrograph 

 used. The lines form a series getting closer and closer together at the ultra-violet 

 end of the spectrum, and more powerful instruments are required to resolve more 

 of them. (Slides shown to indicate the character of the spectra.) Up to the 

 present I have been able to extend the series to 41 members in the case of lithium, 

 24 for potassium, 30 for rubidium, and 31 for caesium. When it is remembered 

 that in the case of emission spectra only 7 members of the series were known for 

 sodium, 9 for potassium, 5 for rubidium, 9 for caesium, and 9 for lithium, it can 

 be seen what an extension of the field has been made by the new method. Wood 

 has also opened a wide field in the relationship between these absorption spectra 

 and other banded spectra which appear with them and fluorescent and magnetic 

 rotation spectra. The complexity of these spectra is very great, and no more 

 can be said than to refer those who are interested in the matter to Wood's great 

 paper on sodium spectra. 



The question of applicability to the series spectra of formulae based on the 

 work of Rydberg is of great interest. A great deal of work has been done in 

 this region by Hicks and Ritz especially, and formulae have been proposed which 

 express the series and the relations between different series with the use of sur- 

 prisingly few constants. 



The phenomena of dispersion in metallic vapours is of great interest because 

 of their bearing on optical theory and on the views we can derive as to the 

 nature of the atom and the vibrating systems that give rise to spectrum lines. 

 Anomalous dispersion has been observed in the cases of all the alkali metals and 

 several others. 



I can only refer to the work of Kundt, Becquerel, Ebert, Schon, Puccianti, 

 Luminer, and Pringsheim in this region. The later work of Wood and others 

 has given us a quantitative measure of the effects, and from these we can gain 

 some information as to the nature of atoms. A great deal of work has been 

 done on the dispersion in ordinary gasefe, but in these cases the region investi- 

 gated is far ivom the absorption lines. The case of hydrogen is different, for 

 Ladenburg has shown that when hydrogen is itself excited by an electric dis- 

 charge it can produce anomalous dispersion effects. This brings it into relation 

 with the alkaline metals, and is strongly in support of the theory that specialised 

 molecules take part in optical phenomena. The difficulty in experiments with 

 metallic vapours is to arrive at absolute measurements; my own work has been 

 concerned with relative measurements at different lines of the principal series. 



