A. W. Ewell — Rotatory Polarization. 377 



constant — under these conditions we should expect the rotation 

 for a given twist to be constant and definite. Such is approxi- 

 mately the case, as is illustrated by the following observations, 

 where these precautions have been carefully observed : 



Relation between Twist and Rotation. 



July 21, a. m. Tube D 3. Length of jelly = 3*0 cm . 1' delay 

 after twist was applied and commencement of readings of 

 rotation. 



Twist 



— 90 90 



— 110 110 



— 120 120 



— 130 130 



Rotation . . 



1 -4 



2 —13 



2 -18 



2 —26 



Sum 



5 



15 



20 



28 



July 21, p. m. After tube has been removed from the instru- 

 ment, turned end for end and replaced. 



Twist .. —60 60 90 —90 110 —110 120 —120 130 —130 

 Rotation -1 1—6 0—13 —21 1 —26 1 



Sum ... 2 6 13 22 21 



July 22, a. m. After tube has been taken out, turned end for 

 end, and replaced. 



Twist .. —60 60 —90 90 —110 110 —120 120 —130 130 

 Rotation 2—4 2—11 3 -17 4 -22 



Sum 6 13 20 26 



On account of change of zero due to relaxation the actual 

 twists of the jelly, when the rotation was read, were 56, 83, 

 101, 109, and 118 respectively. 



July 22. Tube D 4. Length of jelly = 3 cm . 2' delay between 

 twist and reading of rotation. 



Twist. 60 



Rotation —8 



Sum 10 



The actual twists were 56 and 83. 



July 22. Tube F 1.* Length of jelly = 4 cm . 1' delay between 

 twist and reading. 



Twist —60 60 —90 90 —110 110 



Rotation.. 1—1 3 —1 6 -3 



Sum 2 4 9 



Twist (cont.) —120 120 —135 135 —150 150 

 Rotation " 8 —4 11 —6 19 —15 



Sum " 12 IV 34 



* These readings differ widely from those given for this tube on previous 

 pages. The explanation is that in the earlier observations the lower clamp 

 rested on the platform e and the jelly tube was longitudinally compressed 

 relatively to what it was here when stretched by the freely hanging lower 

 clamp. 



60 



90 



— 90 



2 



-39 



48 



9 



