Magneto -optics of Iodine Vapour. 1017 



the pump. After the exhaustion is complete and the tube 

 sealed off, the cooled portion of the tube should be cut away 

 from the pump before the iodine vaporizes. 



The tube is mounted between the pole-pieces of the magnet 

 as shown in fig. 5, and arc or sun light concentrated at its 

 axis with a large lens. Observations are made through the 

 bulb, and it is advantageous to paint the lower portion with 

 black varnish to secure a dark background. The maximum 

 effect is obtained if the lower portion of the tube is cooled 

 to 0° in ice. Under these conditions we estimated the 

 reduction of intensity to amount to fully 90 per cent, with a 

 field of 30,000 gauss, and it is probable that with a field of 

 50,000 the fluorescence would be practically destroyed. At a 

 tension corresponding to room temperature, the reduction of 

 intensity is much less, and at 35 or 40 degrees scarcely 

 noticeable, though the fluorescence still remains fairly 

 bright, in the absence of a magnetic field. 



No obvious explanation of the effect of the field in reducing 

 the intensity of and ultimately practically destroying the 

 fluorescence suggests itself. We made numerous experiments 

 to determine whether the vapour at very low pressures was 

 thrown out of the field, but these all gave negative results. 

 They were based for the most part upon the principle of 

 allowing iodine to distil from a bulb at 0°, through tubes 

 of the same size into two small bulbs cooled with solid C0 2 , 

 one tube passing through an intense magnetic field, the 

 other well outside of it. The phenomenon may result from 

 orientation, but we obtained no evidence of this, though we 

 passed the light through the bulb both parallel and perpen- 

 dicular to the field and observed the fluorescence in the same 

 way. 



The absorption spectrum, as we have stated, shows no 

 change as a result of the field, but it must be remembered 

 that our observations were made with a vapour density 

 corresponding to 30 or 40 degrees, and the reduction in the 

 intensity of the fluorescence is almost imperceptible at this 

 pressure. To observe the absorption at 0° or even at room 

 temperature, it would be necessary to observe with a larger 

 bulb and the field would be less intense as a result. 



Conclusions. — Up to the present the magnetic rotations of 

 the plane of polarization in the vicinity of absorption bands 

 may be divided into two classes. (1) Anomalous rotations, in 

 which the sign changes in crossing the band, as observed by 

 Cotton for certain solutions and hy one of us for a solid film 

 of a neodymium salt. Rotations of this nature appear to 

 obtain in cases in which there is no change in the position of 



