386 A. W. Ewell — Rotatory Polarization. 



Dispersion. — JSTo success was attained in measuring the 

 dispersion, i. e., the variation in rotatory polarization for light 

 of different wave lengths. 



The biquartz could not be used since this is cut for some 

 particular color, in this case to rotate the mean yellow through 

 90°, giving rise to the "sensitive tint "when placed between 

 parallel Nicols. With simply two Nicols the dispersion was 

 evident. If the analyzer and polarizer were crossed and the 

 jelly introduced, on slightly turning the analyzer the field 

 became green, showing the extinction of the longer waves, and 

 on turning farther in became red, showing the extinction of the 

 shorter waves. When a twist was applied this dispersion was 

 still evident and the positions for the extinction of the dif- 

 ferent colors seemed farther apart, but no quantitative measure- 

 ments could be made. When the jelly was twisted the coloration 

 of the fields was irregular. Attempts were made to test the light 

 emerging from the analyzer in both a direct vision spectroscope 

 and an ordinary spectroscope. As the analyzer was rotated, 

 successive parts of the spectrum did not perceptibly fade out, 

 but near certain positions the whole spectrum gradually 

 diminished in brightness until a mininum was reached, and then 

 increased in brilliancy. The position for this minimum could 

 only be determined roughly, but it essentially agreed with the 

 portion of the plane of polarization of yellow light as deter- 

 mined with the biquartz. 



Aug. 6. Tube 31. 3 cm of jelly. 



Position of analyzer ; with biquartz — 225, without biquartz for 

 greatest darkness = 226, with the addition of a direct vision 

 spectroscope, for greater general darkness of spectrum = 225. 



Twist of 100. 

 Position of analyzer ; with biquartz = 243, without biquartz 

 = 242, with d.v.s. = 241. 



Twist of —140. 

 Position of analyzer ; with biquartz = 207, without biquartz 

 = 207, with d.v.s. = 205. 



Quartz plates introduced produced a dark band in the appro- 

 priate part of the spectrum, showing that the apparatus was 

 properly adjusted. 



Probably dispersion is present and very likely it follows the 

 law of inverse squares of the wave length. The biquartz 

 locates with great sensitiveness the mean position of the plane 

 of polarization of yellow light. Owing to the complicated 

 strains introduced with the twist, there is probably always 

 yellow light polarized in all neighboring azimuths, of only 

 little less intensity and similarly for other colors. Moreover, 

 the jelly appears much more transparent for yellow light than 



