THE STRUCTURE OF THE NUCLEUS. 



43 



beam, «<?, would remove white light. Hence absorptiou (which may be interfer- 

 ence) of some kind must be looked to, to explain these colors and it is not un- 

 reasonable to suppose that the central field is illuminated by tlie color a.xially 

 absoi'bed. 



21. Polarized light. — On looking through the single drum with an analyzer, 

 it appears that neither the axial colors (as a rule) nor the coronas are polarized. 

 Both are produced with the same vividness, caet. par., when the source of light is 

 polarized as when it is not so. No coronal light passes between ci-ossed nicols. 

 With the axial colors a peculiar result was obtained for the jjaiticular length of 

 column here used (77 cm.). The colors of higher oi'ders, if produced by polarized 

 light, seemed to pass with increased saturation through the crossed polarizer, while 

 the lower orders of colors were quite damped out. This was particularly the case 

 with the violet purple of the second and even of the third order, the latter of which 

 I had not seen before. The observation is brought out in the following table, the 

 results of which were obtained independently, twice. 



TABLE 17.— AXIAL COLORS, ETC., IN POLARIZED LIGHT. 



' Colors 5-12 seen through crossed nicols. 



Having obtained these suggestive results, I anticipated a similar intensification 

 of the axial colors, when produced by columns either longer or shorter. Accord- 

 ingly, the globe, giving a column 30 cm. in length, and the double drum, with a 

 column 180 cm. in length, were similarly adjusted between crossed nicols. In 

 neither case was any noteworthy result obtained. The dark polarized field re- 

 mained dark throughout. This thi'ows some doubt on the preceding investigations. 

 The importance of a possible depolarization produced by the particles themselves, 

 is, however, too great to justify a hasty dismissal of the subject and I have there- 

 fore recorded it. Axial colors reflected from water are usually more saturated, but 

 table 17 does not admit of references to merely subjective phenomena. 



22. Preliminary observations with the double drum.— The following table, 18, 



