38 Very — On a Possible Limit to Gravitation. 



parallel with them (though the discrete positive electrons 

 may be rotating in the opposite sense to the subjacent 

 negative ones), the magnetic field of the positive elec- 

 trons will have the opposite sign to that of the negative 

 electrons. The positive strands will repel each other, and 

 the positive shell will be strained almost to the bursting- 

 point, being retained in position solely by the attraction 

 of the enclosed sphere of negatives. In this view, the 

 sun-and-planets analogy does not hold in the atom, at 

 least not as to the gravitational properties of the nucleus, 

 but the positive and negative masses are presumably of 

 a like order of magnitude or even identical save in 

 aspect; and they are complementary in the sense that 

 one requires the other, whereas we can conceive of a sun 

 without planets. If the two spheres are not exactly con- 

 centric or not perfectly equal, they form the equivalent 

 of a doublet, 2 and the total electric value of the combina- 

 tion is that of a minute residual, or difference, while the 

 gravitational effect is that of the sum of the gravita- 

 tional forces which do not interfere. 



If the combined revolutions and potencies of positive 

 and negative electrons are exactly adjusted, there will 

 result a magnetically neutral atom ; but if there is a dif- 

 ference in the positive and negative revolutionary fields 

 within the atom, it becomes a magnet. 



The Question of Orbital Spircdity. 



Professor G. Johnstone Stoney has been the only one 

 I know who has attempted to apply the doctrine of spiral 

 orbits of electrons to explain harmonic series in spectral 

 lines, finding in the case of the A and B groups of the 

 atmospheric spectrum (first described in detail by Lang- 

 ley in 1878 in the Proceedings of the American Academy 

 of Arts and Sciences) that \- he could account for the 

 remarkable structure of these groups by compounding 

 oppositely rotating spirals of diminishing amplitude 

 which were supposed to give respectively the red and 

 violet components of the pairs of a train. The resultant 



2 Cohesion between atoms or molecules does not arise from gravitational 

 attraction, but follows a law of proportionality to the inverse fourth power 

 of the distance, or after the analogy of the gravitational model, is equal to 

 Am,m,/r 4 . Both Fessenden and Sutherland have shown that this follows as 

 a consequence if we assume the particles to be of the nature of electric 

 doublets. See especially W. Sutherland on The Electric Origin of Molec- 

 ular Attraction, Phil. Mag., Ser. 6, vol. 17, p. 657-670, May, 1909. 



