368 



A. W. Ewell — Rotatory Polarization. 



Twist 







120 



— 120 



120 



— 



120 



120 



— 150 



150 





78 



88 



91 



92 





83 



74 



78 



85 





79 



86 



69 



91 





63 



89 



69 



86 



Readings in 



82 



97 



79 



97 





.83 



95 



76 



90 



the 8 azimuths 



83 



89 



76 



86 





75 



98 



79 



92 





86 



89 



81 



88 





90 



89 



81 



87 





82 



90 



70 



94 





70 



89 



79 



85 





79 



98 



71 



95 





82 



91 



73 



86 





78 



91 



78 



92 





72 



96 



73 



83 



Mean position 





















of analyzer 



81 



91 



76 



92 





78 



91 



76 



87 



Rotation due 





















to twist 





-10 



5 



-11 





3 



-10 



5 



-6 



Twist .._ 





-150 



150 



— ] 



L20 





L20 - 



-150 



150 







82 



87 





94 





86 



77 



85 







79 



91 





70 





89 



75 



87 







77 



91 





77 





93 



79 



90 







82 



83 





79 





93 



79 



86 







83 



84 





95 





85 



77 



85 







75 



88 





69 





89 



83 



90 







75 



86 





87 





85 



76 



87 







74 



89 





76 





94 



74 



85 



Mean position 



of 



















analyzer 





78 



87 





81 





89 



78 



87 



Rotation 





3 



— 6 











-8 



3 



— 6 



The irregularity is not surprising when one considers the 

 strains which are unavoidable in these independent cylinders 

 and the dependence of the rotation upon certain of these 

 strains, as shown later. These and similar observations, some 

 of which are cited below, show qualitatively, that simple jelly 

 cylinders, without any envelope, acquire when tioisted rotatory 

 polarization in the opposite direction to the twist. It might 

 be suggested that possibly the rotation is due to a twist of the 

 jelly within the rubber tubing. This is scarcely possible, for 

 the rubber is so much more rigid than the jelly that the tub- 

 ing which tightly grasps the jelly hardly twists at all. 



A lateral envelope greatly increases the rotation for a given 

 twist. 



Dec. 9, '01. Independent jelly cylinder, length = 8'0, diam. =2*0. 



Twist —170 190 



Rotation 1 — 1 



Soft rubber tube slipped over jelly (internal diam. = 1* 76). 



Twist _ 90 —90 



Rotation —6 9 



