MAGNETIZATION IN CRYSTALS — DILLON 391 



say, parallel to the line of sight. Wlieii this was done, and the angle 

 of rotation per unit thickness measured for light of various colore, the 

 specific rotation curve of figure 2 was obtained. This also shows some 

 structure and it is clearly related to that of the absorption cui-ve. The 

 rotation is least in the red, and greatest near the absorption edge in 

 the green. 



It might be remarked parenthetically that a great deal of the basic 

 scientific interest in these transparent crystals arises from the fact 

 that we are able to measure the structure in the absorption and rota- 

 tion curves out to the middle of the visible spectrum. The peaks in 

 a here and others in the near infrared represent electron transitions 

 involving the electrons responsible for the magnetization of tliese 

 crystals. The rotation very clearly demonstrates the involvement of 

 the magnetization. It is believed that the transitions involved in both 

 the rotation and the absorption are between levels of the iron ions in 

 octahedral sites. It is not the total magnetization that is unportant 

 in determining the sign of rotation, but the direction in which the 

 magnetization of the octahedral Fe+ + + lattice lies. 



DOMAIN VISIBILITY 



How, precisely, do we see the distribution of the magnetization in 

 one of our little transparent samples ? [10] Figure 3 illustrates this. 

 Consider a hypothetical crystal in which the magnetization is dis- 

 tributed in three sections: parallel, antiparallel, and perpendicular 

 to the line of sight, as shown in the figure. Before passing through 

 the crystal, the light goes through a polarizer. Consider the red light 

 and the green light separately. Passing through regions (a) and (6), 

 the planes of polarization will be rotated in opposite senses, and the 

 net rotation for the green will be much greater than that for the red. 

 The light which passed through (c) will not be rotated at all. How- 

 ever, in every case the green will be considerably more attenuated than 

 the red. 



Now compare three settings of a second polarizer, traditionally 

 called the analyzer, and the corresponding appearance of the magnet- 

 ization. If the analyzer is set at the angle which corresponds to 

 extinction without the sample, light passing through region (c) is not 

 rotated at all, and thus (c) appears black. However, the red light 

 going through (a) and (h) is rotated slightly, and the green light 

 about three times as much. Because of the greater angle for the green, 

 we see these regions as bright gi'een. However, in the narrow band be- 

 tween (a) and (&) the magnetization makes a smooth transition be- 

 tween parallel and antiparallel to the line of sight. In the center of 

 this the magnetization lies in the plane, and thus it, like region (c), 



579421—61 30 



