228 



NA TURE 



[January 5, 1899 



exist in the liyht radiated from the source, but is imposed 

 on the spectral lines by the apparatus used, namely, the 

 interferometer. Be this as it may, the application of 

 this instrument to the study of radiation phenomena in 

 the magnetic field is highly interesting. In his first 

 e.vperinients Michelson merely observed a doubling of 

 the spectral lines both along and at right angles to the 

 lines of force (I), but subsequent observations proved that 

 tripling occurred across the field of force, and that the 

 constituents of the triplets were themselves multiple 

 lines (see Astro. Phys. Journal, vol. vi. p. 48, 1897 ; 

 vol. vii. p. 131, 1898 ; vol. viii. p. 43, 1898). But this is 

 accompanied by the most surprising statement that the 

 separation of the lines in the triplets produced by the 

 magnetic field is independent of both the spectral line 

 and the substance. In other words, that the separation 

 is the same for all lines and all substances I Now, in all 

 observations with ordinary grating or prism spectro- 

 scopes the separation of the components produced by 

 the magnetic field varies very considerably for the 

 different spectral lines of the same or of different sub- 

 stances. Even in the case of lines of nearly the same 

 wave-length the difference is often very marked. The 

 separation not only differs for different substances, but 

 it is some complex function of the wavelength for any 

 one substance. That the interferometer has led to such 

 a law as that announced by Prof Michelson, shows that 

 there is some peculiarity of the instrument not yet taken 

 into account— or else that by chance Prof Michelson 

 has happened to confine his observations to lines which 

 give approximately the same separation ; yet this latter 

 could not be easily done. Be this as it may, Michelson 

 has examined these phenomena by aid of another new 

 instrument of his own design — the Echelon spectro- 

 scope (Astro. Phys. Journal, vol. viii. p. 43, 1898). With 

 this instrument he states that the results previously 

 obtained by aid of the interferometer, and the visibility 

 curve, were confirmed. .And this is striking, for if it 

 confirms the general law stated by him in regard to the 

 separation of the components, then the interferometer 

 and the Echelon spectroscope are at variance with all 

 other forms of spectroscope. 



With apparatus which reveals structure or multiplicity 

 in the ordinary spectral lines, it is to be expected that 

 multiplicity would be readily revealed in the constituents 

 produced by the magnetic field ; yet in the case of some 

 lines, the amount of finer structure revealed does not 

 appear to be as great as that observed with a good 

 grating, and this with other discrepancies require clear- 

 ing up. If we suppose that an ordinary spectral line 

 really consists of two or more very close lines, not separ- 

 ated in ordinary spectroscopes, and if we suppose that 

 this multiplicity is produced by small perturbations 

 caused by events inside the molecule, then it is clear 

 that the further perturbations (if any) brought about by 

 the magnetic field, may either increase, or diminish, or 

 possibly reverse, those previously existing in the free 

 field. And from this point of view the following most 

 interesting observations made by Michelson Hoc. cit.) 

 become intelligible. " A very remarkable eftect is 

 observed in the case of the yellow copper line. This 

 line without the field is a close double, the distance 

 being 1/ 150th of the distance between the I) lines, or 

 004 A.V. As the field increases the lines merge 

 together without broadening, and with a strong field 

 there is but a single narrow line." 



"The behaviour of the yellow-green line of manganese 

 is even more striking. The line is a quadruple line, just 

 resolvable. In a weak magnetic field the light accumu- 

 lates in the centre of the group, the lines becoming in- 

 distinct and merging together. In a strong field the 

 quadruple band is reduced to a single fine line at the 

 centre of the group." 



In conclusion, it is necessary to mention briefly some 



NO. 1523, VOL. 59] 



ingenious methods which have been devised to exhibit 

 the existence of the Zeeman phenomena in comparatively 

 weak magnetic fields. The first of these chronologically 

 was devised by M. Cotton (Coinpfcs remius, t. 125, p. 865) 

 in 1897, and depends upon the fact that if a small sodium 

 flame, .\, be placed in front of a larger one, 11, and viewed 

 against it, the outer edges of the small flame appear dark. 

 This arises, as is well known, from the absorption which 

 takes place in the outer sheath of the smaller flame. If, 

 however, the flame n be placed in the magnetic field, 

 the dark border around .\ disappears. This arises from 

 the fact that the magnetic field induces new periods of 

 vibration in v. (the side lines) which are not possessed 

 by .\, and therefore not absorbed. 



The next experiment to be mentioned is one of special 

 elegance, devised by Prof. Auguste Righi {Comptes rendus, 

 t. 127, p 216, 189S, and Rend, delta R. Accad. del Lincei, 

 July 1898). If a plane polarised beam of light from a 

 powerful source, such as an arc lamp, be transmitted 

 through an absorbing vapour, such as a sodium flame, or 

 sodium vapour in a tube, and if the light, after passing 

 through the vapour, be transmitted through a nicol's 

 prism, and then received on the slit of a spectroscope, a 

 continuous spectrum will be observed in which dark 

 lines occur corresponding to the absorption lines or 

 bands of the vapour. If the analysing nicol be rotated 

 till its principal plane is perpendicular to that of the 

 polariser, then all light in the spectroscope will be extin- 

 guished.' Now suppose this to be so arranged, and 

 suppose, further, that the absorbing vapour is between 

 the pole-pieces of a magnet so as to be subject to the 

 action of the magnetic field, and suppose that the light 

 passes through this vapour along the lines of magnetic 

 force by passing through axial holes pierced in the pole- 

 pieces, then under these circumstances, if the magnet be 

 excited, bright lines appear in the spectroscope corre- 

 sponding to the absorption lines of the vapour. At first 

 sight it appears as if the magnetic field caused the vapour 

 to emit its own vibrations as if it were highly luminous. 

 It is not so, however. The explanation is that the mag- 

 netic field so aftects the vapour, that if it were self- 

 luminous any spectral line appertaining to it of fre- 

 quency N is converted into two other vibrations of 

 frequency N -t- // and N - /; respectively : and these two, 

 along the lines of force, are circularly polarised in opposite 

 senses, and consequently the vapour when cold possesses 

 the power of absorbing vibration of frequencies N -h « 

 and X - //. Now the beam from the electric arc passing 

 through the vapour being continuous, possesses vibrations 

 of frequency X -I- n and also of frequency N - n. These 

 vibrations in the arrangement, described above, are 

 plane polarised, and any plane polarised vibration is 

 equivalent to two opposite circular vibrations. The 

 result is that the vapour absorbs one of the circular 

 components from the rectilinear vibration X -t- «, and 

 transmits the other. In the same way it also absorbs 

 one of the circular component vibrations from the vibra- 

 tion N - /;, and transmits the other. These transmitted 

 circular components are very intense .having evidently 

 half the intensity possessed by the arc light), and they 

 cannot be extinguished by the analysing nicol, so they 

 consequently appear in the spectroscope. If the mag- 

 netic field is not verv' strong, the vibrations N 4- « and 

 N - /; practically coincide with N, and what is presented 

 to the eye is that the absorption lines of the vapour 

 become bright when the magnetic held is excited. This 

 can be observed in fields of very small intensity. 



Prof Righi mentions that the phenomenon observed 

 in the foregoing experiment does not occur when the 

 light traverses the vapour in a direction perpendicular to 

 the lines of force. This is a result which differs from 



1 If a. sodium f\, 

 lines will slill 

 be bright. 



