Magneto-optics of Iodine Vapour. 1013 



and that a reversal of the direction of the field caused them 

 to become so fine as to be almost invisible, as a result of the 

 brightening of the regions bordering them. This brightening 

 was in some cases much stronger on one side of the line 

 than on the other. 



To obviate the necessity of reversing the field to observe 



these changes, a half-wave plate of mica was placed over 



one half of the slit of the collimator, the principal directions 



of the plate coinciding with those of the polarizing prism. 



As the echelon showed a trace of astigmatism, it was 



necessary to form an image of the edge of the half- wave 



plate at a distance of a few millimetres behind the slit, in 



order to have a sharp hair-line divide the two fields of view. 



The action of the half-wave plate is as follows : — 



Let OP (fig. 2) represent the direction of vibration of the 



-p.^ 9 light traversing the iodine bulb. The 



t ' principal directions of the mica plate 



; p oy and ox being parallel and perpen- 



k p J dicular respectively to OP, the light 



' traverses the plate without change. 



The analysing nicol is now turned 



through an angle a from the position 



of extinction, and it transmits the 



component of OP parallel to OA. 



Suppose now that a certain wave- 

 length A, is rotated by the magnetized 

 iodine vapour from the position OP 

 to OP', i. e. in the same direction as that in which the analyser 

 was turned. In that part of the field of vision given by the 

 light which has not traversed the mica, the wave-length X will 

 appear darker than before the excitation of the magnet. The 

 vibration which traverses the mica (wave-length X) is rotated 

 by the mica from the position OP' to OP", and is consequently 

 more copiously transmitted by the analyser than the wave- 

 lengths not rotated by the iodine. Consequently X appears 

 brighter in this part of the field of vision. The two 

 conditions seen with magnetic fields of opposite direction 

 are thus visible simultaneously one above the other, and in 

 exact coincidence. Any shift, due to anomalous rotation, 

 would thus be doubled, but no such shift was observed. The 

 arrangement of the entire apparatus is shown in fig. 3. It 

 was at once evident that the observation of a larger rotation 

 on one side of certain absorption lines than on the other, 

 made with the grating, was a correct one ; and the explanation 

 of the phenomenon was apparent as soon as a careful study of 

 the rotation produced by the various lines had been made. 

 Phil. Mag. S. 6. Vol. 27. No. 162. June 1914. 3 X 



