Effect of crossed electric and magnetic fields on the helium spectrum* By J. S. Foster, F.R.S., McGill University, and E. R. Pounder, Ph.D., Royal Canadian Air Force (Received 26 February 1946).
288 J. S. Foster and E. R. Pounder electric field directed at right angles to the applied magnetic field. This has been experimentally confirmed. It was not, however, until the introduction of quantum mechanics that uniformly sharp components were theoretically expected to appear in crossed fields of any strength. On this basis the theory of the present problem has been worked out at Professor Heisenberg’s Institute.* On the experimental side, the problem is to establish and maintain steady electric and magnetic fields in a source of sufficient strength to permit analysis by an optical system of high resolution. Provided the design of the source is such as to give a suitable magnetic field, there is no further difficulty in keeping the field steady. On the other hand, electric fields in all strong sources depend to some extent upon space charges which may vary during the exposure. A major part of the experi mental problem is therefore similar to that encountered in the application of electric fields alone.
Effect of crossed electric and magnetic fields on helium spectrum 289 the pole tips a cylindrical unit in which the pole faces have a fixed separation of 7 mm. Steel shoulders at either end of this part of the tube prevent any motion which otherwise might arise from unbalanced forces on the pole tips. A 0-06 in. hole drilled along the axis of this central structure clears the canal-ray beams which are limited to 0-04 in. diameter by perforated aluminium cathode disks. Owing to the small hole through the steel, as required on this plan, the magnetic field realized in the experiments is about 80 % of that obtained with solid poles. The aluminium disks not only limit the diameter of the beams—and thereby prevent bombardment of the canal wall and electric field plates—but also minimize sputtering. Resting on these disks, and closely fitted into the steel sleeve extensions from the pole tips, are the glass tubes which support the anodes.
290 J. S. Foster and E. R. Pounder On the way to a six-prism glass prism spectrograph the light passed through a Wollaston prism. A quarter-wave plate was inserted into the beam with vertical electric vector in order to reduce losses at the prism faces.
292 J. S. Foster and E. R. Pounder photographs cannot be compared exactly since the dispersion on various plates ranged between 14 and 27 cm.-1/mm. Further, the enlargement factor is not exactly constant for all prints.
Effect of crossed electric and magnetic fields on helium spectrum 293 there is interaction also between levels of different m values. These interactions produce separations between certain levels which are quite comparable to Zeeman separations. On this basis, normal Zeeman separations no longer are uniformly maintained. Even in high electric fields the somewhat complex structures continue to show rapid variations of relative displacements with electric field. The resulting structures in most cases are beyond complete resolution in these experiments. Parenthetically, it should be noted that the weakest member of the orthohelium triplet failed to appear.
294 J. S. Foster and E. R. Pounder Actually, only three examples of such components with suitably clear separations and intensities are found. The corresponding observations are set in heavy type in the tables. Each structure contains at least three members of different wave lengths. Except in the first example, the polarization is not the same for the three components. These observations mean that in all cases where clear examination is possible there are transitions in which m changes from + 2 to 0. It may be concluded that in crossed fields , . , _ , _ | Am I = 0,1,2.
Foster and Pounder Proc. Roy. Soc. A, volume 189, plate 6 kV/icm. 2 P-4 SDFP ( A4922) kV/cm. 2¥~43£/©F(A4472) S IB11II1I s 68-7 56*2 P P s S 70*4 78-8 P - * r S 116*1 120*2 P P l .1.1.1 D S 2 3P 5 3 SPDFG ( A 4 0 26) 41*4 P S 56*0 P 64*0 I ! **** vws /> Figure 3. H=52,800 oersteds in all cases. E values as indicated. In each set the arrow indicates the position of the undisplaced line.