6 B. E. Moore 



was any constancy in the value (a — b)/2. Therefore, there 

 seems some reason to assume that the dissymmetries are due to 

 some other causes than those discussed in Voigt's equations. 



Dissymmetry has also been observed by Mr. Jack (Zeeman, 

 Phys. Zcit., vol. 9, p. 343). His spectrographs were taken partly 

 prior and partly subsequent to these thorium spectrographs and 

 with the same apparatus. Professor Voigt writes me that Mr. 

 Jack thinks he has traced the dissymmetry to changes in the azi- 

 muth of the polarized beam of light incident upon the grating. 

 This is important, if true. I think two experimental facts make 

 it implausible in thorium: firstly, a number of the observed dis- 

 symmetries are too large; secondly, before placing the plates in 

 the camera for an exposure the quartz condensing lens was often 

 shifted and readjusted to see that there was the best possible 

 illumination of the grating. (But the lens was never moved 

 during an exposure.) It does not seem probable that this lens 

 occupied the same position throughout the thorium exposures or 

 even the greater part of them. Consequently, any point on the 

 grating was illuminated at different times by light in a different 

 azimuth. If the effect had been appreciable, it would have pro- 

 duced inconsistencies in my measurements, which I have not 

 observed. 



In my previous investigation (/. c.) the />- and ^-components 

 were measured on independent plates. Small dissymmetries would 

 have remained undiscovered. It should be said, however, that 

 for these substances I had plates with both p- and ^-components 

 thereon, i. e. plates photographed without the intervening calcite. 

 As I passed from line to line comparing the intensities of the p- 

 and -.^-components, some of the pronounced dissymmetries herein 

 recorded would have been glaringly conspicuous without meas- 

 urements. Further, the dissymmetries are not only often large 

 but they are found also in lines having more than three compo- 

 nents up to and including a line of nine components. In these 

 also may be found multiple relationships among the unsymmet- 

 rical components, which could be only a coincidence unless these 

 dissymmetries were proportional to the field strength. Thorium 

 was also photographed without the intervening calcite, but the 



94 



