January 4, 1907] 



SCIENCE 



few definite positions in the scale of wave 

 lengths? I think it must be admitted 

 frankly that, in the picture of the atom as 

 given us, there is nothing which determines 

 the rate of rotation of the corpuscle in its 

 orbit, and nothing, therefore, which gives 

 the same periodicity of radiation for all 

 atoms of any one element. 



This difficulty has been stated in a very 

 forcible manner by Professor Jeans, and 

 again by Lord Rayleigh. Each of them 

 has proposed a method of getting around 

 the difficulty, one" by introducing a law of 

 electrical action different from the simple 

 law of inverse squares, the other^" by intro- 

 ducing mobile negative particles, which do 

 not revolve, but which vibrate about their 

 positions of equilibrium located in a rigid 

 positive charge. 



But the adoption of either of these sug- 

 gestions would completely change the en- 

 tire character of the atom. It may be that 

 we shall later be driven into the corner and 

 be compelled to accept some such mode of 

 escape. But at present these devices im- 

 press one as highly artificial, and too in- 

 consistent with other facts to warrant 

 adoption. 



The crux of the situation seems to be 

 just here, if one assumes that the fre- 

 quency of the radiation is identical with 

 the rate of revolution of the corpuscles, he 

 can not expect sharp lines in the spectrum. 

 If, on the other hand, one assumes that 

 light is due to the internal vibrations of 

 the corpuscles, then not only does he fail to 

 predict the Zeeman effect, but he is forced 

 to conclude that since the corpuscles appear 

 to be the same for all elements, the spectra 

 of all elements should be identical. To 

 derive the Zeeman effect from a rectilinear 

 vibration by substituting for it two circular 

 vibrations would seem to employ a purely 

 kinematic device instead of offering a 



"Jeans, Phil. Mag., 11, 607, 1906. 



"Rayleigh, Phil. Mag., 11, 118, 1906. 



physical explanation. Nor is this all : The 

 value of e/m obtained from the Zeeman 

 effect is, if not correct, at least in beautiful 

 accord with values determined in a variety 

 of other ways. 



Summarizing, one might say that the 

 adoption of the Saturnian atom would com- 

 pel us either to give up the Zeeman effect, 

 or to give up sharp lines in the spectrum. 

 On the other hand, I am not aware that it 

 has ever been shown that even in a gas- 

 spectrum the region between any two lines 

 is entirely free from radiation. Is there 

 any spectrograph so free from diffused 

 light as to make an experimental answer to 

 this question anything other than a more 

 or less rough approximation ? But even if 

 every spectrum is, to some slight extent, 

 continuous, the fact remains that spectral 

 lines are essentially sharp. 



2. Passing now to the second funda- 

 mental fact, which is that spectral lines are 

 not perfectly sharp, but (within limits not 

 yet resolved by any grating) possess a com- 

 plicated structure, Professor Michelson^^ 

 and Lord Rayleigh" have shown that in 

 the case of a gas at low pressure the chief, 

 if not the only, cause of widened lines is 

 motion in the line of sight, an effect which 

 depends as much upon pressure and tem- 

 perature as upon atomic structure. But 

 when it comes to the asymmetric distribu- 

 tion of intensity within these narrow lines, 

 i. e., a linear structure such as has been 

 revealed to us especially by the interfer- 

 ometer in the hands of Michelson, an effect 

 v/hich would appear to be a function solely 

 of atomic structure, then the chances of 

 explanation in terms of the Saturnian 

 atom appear even more remote than in the 

 case of perfectly sharp lines. 



3. The next query to be presented to the 

 Saturnian atom is what explanation can be 

 offered for the fact that very many lines in 



^ Astroph. Jour., 2, 251, 1895. 

 ^Phil. Mag., April, 1889. 



