OPTICALLY ACTIVE COMPOUNDS 



^7 



by piecing together, as above, evidence of different kinds, what 

 relationship will be maintained amongst these curves if the 

 examination of the substance could be extended over a further 

 temperature-range of, say, 200° ? 



As has already been explained, from the behaviour of ethyl 

 tartrate in quinoline solutions, and the rotation of ethyl di- 

 trichloracetyltartrate, it is to be expected that each curve will 

 tend towards a minimum, and it is of much interest to ascertain 

 the sequence, in that region of the T-R c.urves, of the absolute 

 values of the rotation for the various colours of light. At the 

 maximum already discussed (H in Figs, i and 6), the rotation- 

 value is greatest for violet and least for red ; at the minimum 

 (S in Fig. i), it might perhaps be expected that the opposite 



Fig. 4. 



should be the case. But if this were the fact it would be 

 necessary for the T-R curves to cut one another again between 

 the maximum and the minimum values. Since these curves 

 actually do intersect in this manner at medium temperatures, 

 before the maximum at H (Fig. 6) is reached, there is nothing 

 inherently improbable in the suggestion that they may cut 

 one another again after the maximum is passed and before the 

 minimum is reached, and thus give rise, perhaps, to a second 

 region of anomalous dispersion. This possibility is indicated in 

 Fig. 4. In order to determine the question an attempt was made 

 to follow the T-R curves for ethyl tartrate dissolved in quinoline 

 to as high a temperature as possible, but although the curves 

 approached one another towards the minimum, they showed 

 no signs of actual intersection. The scale diagrams will be found 

 elsewhere {J.C.S., 1916, 109, 1145, 1147 ; Proc. R.S.E. 191 8-19, 



