January 5, 1899] 



NA TURE 



marked modification. The general type, or characteristic 

 type, of this modification is that when the slit of the 

 spectroscope views the sources of Hght across the hnes of 

 magnetic force, each spectral Hne becomes a triplet, of 

 which the middle line has the same wave-length as the 

 original line ; whereas the side lines of the triplet have 

 wave-lengths, respectively, a little longer and a little 

 shorter than that nf the unmodified line, the difference 

 of wave-length being proportional to the strength of the 

 magnetic field. Further, the central line has its vibra- 

 tions parallel to the lines of force, whereas the side lines 

 of the triplet have their vibrations perpendicular to the 

 lines of force. Thus, if the axis of the magnetic field is 

 horizontal, so that the lines of force are horizontal, and if 

 the slit of the spectroscope looks horizontally across the 

 lines offeree, then in the central constituent of the triplet 

 the vibrations are horizontal, while in the side lines the 

 vibrations are vertical. Thus the central line is plane 

 polarised, and the side lines are also plane>polarised, but 

 in a perpendicular plane. This is the typical phenomenon 

 when the light is viewed across the hnes of force. When 

 the light is viewed along the lines of force — that is, through 

 axial holes pierced in the pole-pieces of the electro- 

 magnet, the modification is different. In this case, in- 

 stead of a triplet with plane polarised constituents, we are 

 presented witli a doublet, having circularly polarised con- 

 stituents. That is, each spectral line is broken up into 

 two lines of slightly difterent wave-length ; one con- 

 stituent being circularly polarised in one sense, and the 

 other in the opposite sense. As before, the difference of 

 wave-length, and therefore the separation of the con- 

 stituents of these doublets in the spectroscope, is pro- 

 portional to the strength of the magnetic field for each 

 line, but differs in amount for the different spectral 

 lines. 



In order to fix the ideas of those who are not familiar 

 with this department of physics, the phenomena described 

 above are represented diagrammatically in Fig. i. Thus 

 at .\ the upper single line is supposed 

 B to represent a bright spectral line of 

 some substance when the radiating 

 source is not influenced by the mag- 

 netic field. This line becomes con- 

 verted into three distinct lines, that is 

 a triplet, as shown underneath at .\', 

 when the source of light is subject to a 

 strong magnetic field, and the radi- 

 ation takes place across the lines of 

 force. If N be the vibration frequency 

 l.i,;. I. of .\, then the vibration frequencies of 



the members of the triplet .\', into 

 which .\ is converted, are N - //, N, N H- «, where n 

 is a small quantity depending on the strength of the 

 magnetic field. On the other hand, when the source of 

 light is viewed along the lines of force a bright spectral 

 line, B, becomes converted into a doublet, B', consisting of 

 two distinct lines which are circularly polarised in op- 

 posite senses. The constituents of the triplet .\' are, on 

 the contrary, plane polarised, the direction of vibration in 

 the middle line being horizontal, while that in the side 

 lines is vertical. 



The foregoing are the phenomena demanded by the 

 simplest form of theory, and they are the phenomena 

 actually yielded by experiment in the case of the vast 

 majority of spectral lines. Many lines, however, when 

 carefully examined in a sufficiently strong magnetic field, 

 yield phenomena which difl'er in a remarkable manner 

 from the simple theoretical expectation described above. 

 In some cases the middle line of the triplet becomes 

 resolved into a pair of lines so that the triplet becomes a 

 quartet, while in other cases each line of the triplet be- 

 comes a pair, and thus a sextet is produced ; and in some 

 cases the side lines of the triplet become resolved into 

 triplets, while the middle line becomes a doublet, and 



NO. 1523, VOL. 59J 



A 



A 



then an octet is produced, and so on. Thus generally, 

 when the light is viewed across the lines offeree, we may 

 say a single spectral line becomes resolved by the mag- 

 netic field into a system of lines consisting of a central 

 part bordered by two side parts. The central part may 

 consist of one or more lines, and is plane polarised, while 

 the side parts may each consist of one or more lines, and 

 are also plane polarised in a plane at right angles to the 

 plane of polarisation of the central part. 



On account of this opposite polarisation the central 

 part may be quenched and the sides examined separately, 

 or 7>tct' 7iersii, by means of a nicol's prism, and conse- 

 quently the existence of this plane polarisation enables us 

 to scrutinise the phenomena much more closely and 

 effectively than would be otherwise possible unless, 

 indeed, a magnetic field of any desired strength could be 

 produced so as to obtain complete and wide separation 

 of the various constituents of the modified line. But it is 

 not possible at present to produce a magnetic field for 

 working purposes of a strength exceeding 30,000 to 

 40,000 C.G.S. units. Hence the polarisation is of im- 

 portance for purposes of observation. The best way to 

 take advantage of it is not to use a nicol's prism (which 

 lets through only one of the two plane polarised beams), 

 but to use instead a double image prism, or a rhomb of 

 doubly refracting crystal, placed before the slit of the 

 spectroscrope, so that two images of the source are pro- 

 duced on the slit, one above the other (the slit being sup- 

 posed vertical). Of these images one consists of light 

 vibrating horizontally — that is, it consists of the light 

 which forms the central part of the triplet .\'( Fig. i), while 

 the other image consists of light vibrating vertically — that 

 is, the light which forms the sides of the triplet .\' when the 

 magnetic field is excited. These two images on the 

 slit give rise to two spectra in the field of view of the 

 spectroscope, one above the other : one consisting of the 

 lines which form the centres of the triplets, and the 

 other of the lines which form the sides. This is shown 

 in F"ig. 2, where .\ represents a triplet as seen in the 

 field of view, without the use 

 of any nicol or double image V 



prism, and A' represents what , , 



is seen when a double image 

 prism is used. The upper line 

 in a' represents the light vi- 

 brating horizontally, and is 

 what would ,be seen if a 

 nicol's prism were placed a 

 before the slit with its prin- ^ 



cipal plane vertical ; whereas j\ 



the two lines below in .\' are p,r.. ,. 



formed by the light vibrating 



vertically, and constitute what would be seen were the 

 nicol turned through a right angle. With the double 

 image prism, however, the upper and the lower lines 

 in .V are seen simultaneously, and so a great deal of 

 trouble is avoided, and much time is saved when the 

 phenomena are being photogra])hed. But the chief 

 advantage of separating the middle from the side lines, 

 as at .\'(Fig. 2), lies in the fact that in many cases the 

 difference of wave-length of the middle and the side lines 

 is so small, even in a very strong magnetic field, that the 

 width of the lines causes them to overlap, and so 

 obliterate the phenomena. It was for this reason that 

 in the earlier experiments made by Dr. Zeeman, merely 

 a broadening of the spectral lines was observed, and not 

 a tripling. In fact, it was not until theory pointed out 

 that tripling and plane polarisation should exist across 

 the lines of force, that Zeeman interposed a nicol's prism, 

 and found that the broadened line exhibitecl the polarisa- 

 tion required, and that the facts were not discordant with 

 the theory. It is to be observed, however, as 1 have pointed 

 out elsewhere, that the removal of the central part from 

 the broadened line by a nicol properly interposed (so that 



