( 3S4 } 
exactly to what is observed in a direction normal to the lines of 
force 1 J . 
The described phenomena in so far as they are due to the reflected 
light, ordinarily are only just recognizable, the intensity of the light 
being rather small. The capillary used in the last mentioned obser¬ 
vations was of wide bore. With fine capillaries the intensities of 
the reflected components are considerably weakened. 
The diameter of the luminous filament, which emits the most 
intense light, and which is visible while the vacuum tube is sub¬ 
jected to magnetic forces, probably has as much influence as the 
width of the capillary in determining the intensity of the reflexion 
phenomena under consideration. The results of § 8 have been 
obtained with tubes exhibiting only slight perturbations due to 
reflected light. 
10. The middle component in one or two of the experiments 
appeared to exhibit only nearly plane polarization. The reason of 
the very small ellipticity of the reflected light might perhaps be due 
to the elliptic polarisation, which Voigt predicted, in the case light 
]s emitted in a direction inclined to the lines of force. 
Such obliquely emitted light undoubtedly contributes to intensifying 
the observed central line. 
A somewhat faulty position, however, of Fresnel’s rhomb also 
may be responsible for a slight ellipticity of the polarization. New 
experiments of course are wanted before the predicted elliptic polari¬ 
zation can be considered as proved by experiment. In my present 
observations only a very first trace of the effect may perhaps be 
operative. 
11. Before leaving the subject of reflexions in the tube, one point 
may be considered somewhat more in detail. Is it possible that 
the circularly polarized light of the components of the doublet, 
reflected back in the original direction may disturb the purity of 
the observed effect? 
The same electron, which emits right-handed circularly polarized 
light in the direction of the lines of force, which, let us assume, proceed 
to us, in the opposite direction sends left-handed circularly polarized 
light of the same period. By reflexion, however, the sense of the 
circular polarization is reversed, hence the direct as well as the 
reflected beams enter the eye right-handed circularly polarized. The 
X ) Zeeman, These Proceedings 30 Nov. 1907. p. 351. 
