Magneto -optics of Iodine Vapour. 1015 



the left, since the distance between 2 and 3 is greater that 

 the distance between 3 and 4. 



It is very probable that the same condition holds for the 

 vapour of sodium, at least in the green region. In the red and 

 orange portion of the spectrum it is probable that the + and 

 — rotations observed with the Fresnel double prism were 

 due to close groups of lines with rotations of the same sign. 

 It was observed that the luminous needles which penetrated 

 the dark bands were almost invariably found on one side or 

 the other of broad absorption lines, which were undoubtedly 

 unresolved groups of fine lines. Suppose we have aijgroup of 

 a dozen lines, the first four rotating the plane of polarization 

 in the same direction, while for the remainder the sign 

 changes in passing from line to line. It is clear that if the 

 spectroscope does not resolve the lines the Fresnel prism will 

 show a strong rotation in the vicinity of the first lines, that 

 is on one side of the group, and no rotation at all on the 

 other : in other words, we apparently have a broad line 

 which shows rotatory power on one side only, which was 

 exactly what was found in the earlier work with sodium. 



Examination for the Zee-man Effect. 



Since the selective rotatory power of the vapour in the 

 vicinity of absorption lines can be explained by a longitudinal 

 Zeeman effect, it was of interest to see whether, with the 

 high resolving power at our disposal, any evidence of such 

 an effect could be observed. 



We used for the purpose an arrangement employed by 

 one of us in a similar investigation of the vapour of bromine. 



A double circular analyser (two X/4 plates of mica, one 

 rotated through 90° with respect to the other) was mounted 

 between the iodine bulb and the analysing nicol (azimuth 45° 

 with respect to the neutral lines of the plates). One obtains 

 in this way tw r o fields of view separated by a fine line, one corre- 

 sponding to right-handed, the other to left-handed vibrations. 



If a longitudinal Zeeman effect exists, the absorption 

 line, which runs across both fields of view as a continuous 

 line in the absence of the magnetic field, should be displaced 

 in opposite directions in the two fields as soon as the magnet 

 is excited. We were, however, unable to detect a trace of 

 such a shift in the case of any of the iodine lines. 



Since the absorption lines 3 and 4 are separated by a 

 distance of about ^o A. U. we should have certainly been 

 able to detect a shift of 0*01 A. U. 



From this we must suppose, that if the Zeeman effect 



o 



exists, it is less than O'Ol A. U. for a field of 20,000 eauss 



3X2 > S • 



