■>>S A. W. Etcell — Rotatory Polarization. 



conclusive results were obtained owing to the great double refrac- 

 tion, introduced by the compression. Jelly is very incompress- 

 ible.* 



I do not see how in the above theory to account for the 

 effect of the envelope unless the envelope affords a constraint 

 preventing the jelly from yielding under the strains produced by 

 the twist, the more perfectly these strains are maintained the 

 greater being the rotatory polarization. 



Consider the above mentioned rhombus into which the 

 square has been distorted. The distortion will bring the 

 diagonal along which there is compression nearer the axis than 

 the other diagonal. Therefore longitudinal compression will 

 increase the compression along this diagonal more than it 

 will decrease the elongation along the other diagonal, and there- 

 fore we should expect the rotatory polarization to be increased, 

 which agrees with the experimental facts. Similarly we should 

 expect elongation to decrease the rotatory polarization, which it 

 does. On this theory we are not surprised that hydrostatic 

 pressure produces no effect, since it is perfectly symmetrical 

 and homogeneous. 



This rotatory polarization might be expected in a general way 

 according to many theories of rotatory polarization which 

 require rotatory polarization where there is such dissymmetry 

 present as in a twisted body. 



Conclusions. 

 The following are the most important facts added to those 

 established in previous papers. 



1. For a given twist, the rotatory polarization is very much 

 increased by an envelope about the jelly. 



' 2. The rotatory polarization of jelly twisted in a rubber tube 

 is very much increased by longitudinal compression, and 

 decreased by longitudinal elongation, closely following an 

 exponential law. 



3. It is not affected by hydrostatic pressure. 



4. It is, to some extent, dependent upon the previous history 

 of the jelly. 



5. When these disturbing influences are kept constant the 

 rotatory polarization is quite definite for a given twist and, 



6. The rotation is closely proportional to the fourth power 

 of the twist. 



7. The rigidity of the jelly increases with longitudinal 

 elongation. 



Finally, a theory has been suggested for explaining some of 

 the above, facts. 



I wish to acknowledge my indebtedness to Prof. A. Wilmer 

 Duff for many valuable suggestions and criticisms. 



Worcester Polytechnic Institute, Worcester, Mass. 



*P. von Bjerken, Wied. Ann., 1891, p. 817. 



