Januabt 4, 1907] 



SCIENCE 



11 



equally striking. Rydberg^" has pointed 

 out that certain spectroscopic properties of 

 the elements recur with exactly the same 

 periodicity as that discovered by Mendele- 

 jeff. 



But it is important to observe that in 

 none of these laws is the question of atomic 

 structure involved, but only that of atomic 

 mass. Not a single one of these phenomena 

 has been predicted from previous notions 

 concerning the atom, and indeed all our 

 views concerning the Saturnian atom, with 

 the single exception of the Zeeman phe- 

 nomena, appear to be singularly devoid of 

 that spirit of prophecy which characterizes 

 all sound theory. 



8. When confronted with the fact that 

 many spectral lines show self-reversal, the 

 atom which is now on the witness stand 

 replies that its explanation of this phe- 

 nomenon is identical with that of any other 

 atom which consists of a vibrating mech- 

 anism. 



The essential feature of a line-reversal 

 appears to consist of a source within the 

 arc or spark which emits, from whatsoever 

 cause, a relatively wide line — thus approxi- 

 mating in small degree the incandescent 

 solid first used by Kirchhoff. The outside 

 region of the arc or spark emits, from 

 whatsoever cause, a relatively narrow line, 

 and its radiant atoms are, therefore, capable 

 of absorbing only certain periods from 

 those which appear in the broad or inner 

 source. 



Accordingly, the phenomenon of reversal 

 is, in a certain very true sense, not an 

 atomic problem. The fundamental ques- 

 tion here involved is, however, the follow- 

 ing, namely, how does it happen that, in 

 the inner source, some atoms have their 

 frequencies slightly increased while others 

 have theirs slightly diminished? And this 

 is an atomic problem. 



^ Rydberg, ' Rapports Congrfes de Physique ' 

 (Paris), II., 217. 



We may say that it is due to 'increased 

 density' of the luminous vapor, but we are 

 little wiser for that. In this emergency, 

 the Cambridge atom with its entourage of 

 stray corpuscles shows itself very capable; 

 for it is exactly in such a region, as the 

 interior of an arc, where the electric field 

 is strong, and where collisions are frequent, 

 that one might expect this 'swarm of cor- 

 puscles' to be varying largely — thus alter- 

 ing slightly the period of the radiant 

 source, whatever that may be, both by 

 change of inertia, and by change of electric 

 field— acting, so to speak, both upon the 

 numerator and denominator of the expres- 

 sion for the period. The case of double 

 reversals has been reduced to that of single 

 reversals by the clever experiments of Dr. 

 Humphreys.^^ 



The case of Wolf-Rayet stars^^ where H " 

 and PI^ are bright while the lemaining 

 hydrogen lines are dark is, so far as I am 

 aware, an unsolved puzzle. 



9. Regarding the last of the nine criteria 

 which have been cited, namely, the well- 

 nigh impossible feat of securing a line- 

 spectrum from a gas by means of heat 

 alone, the situation seems to be as follows. 



If one assumes that ordinary tempera- 

 tures are due to the translational energy of 

 the atom, while light consists in dissipation 

 of vibrational energy in the atom, then 

 Jeans^^ has proved, on dynamical grounds, 

 that it is possible only in a minute degree 

 to transfer energy 'from the principal de- 

 grees of freedom to the vibratory degrees 

 of freedom'— at least with any tempera- 

 tures which are encountered on this planet. 

 In the ease of ordinary collisions, the trans- 

 fer is infinitesimal on account of the high 

 frequency of the vibrations as compared 

 with the duration of the collision; in the 

 case of those rare collisions in which the 



"Humphreys, Astroph. Jour., 18, 204, 1903. 



=2 Campbell, iUd., 2, 177, 1895. 



'^ Jeans, ' Dynamical Theory of Gases,' Ch. IX. 



