74 SCIENCE PROGRESS 



data group themselves meet very nearly in one point, which is, 

 however, not the zero of rotation. Now the actual rotation 

 value at the point i is BC 4- Ci, and the vfg dispersion co- 



BC + Ci 



efficient by the ordinary method of calculation is ^^f — • 



BL 



It is quite obvious that the coefficient calculated in this way 

 from values lying along the line for mercury violet cannot re- 

 main constant. It is equally obvious, on the other hand, that 



the ratios such as ^^ — must remain constant. But the 



point A, which has the rotation value — 0-9°, is exactly that at 

 which the temperature-rotation curves for these colours of 

 light for that particular substance would intersect, the point 

 which has already been suggested as the rational zero. The fit- 

 ting into the characteristic diagram of data obtained from de- 

 rivatives of some compound taken as mother substance is 

 practically the same thing as the tracing of parts of temperature 

 rotation curves ordinarily inaccessible, by examining the temper- 

 ature rotation curves of derived substances ; and the character- 

 istic diagram is thus also a very strong confirmation of the 

 suggestion that dispersion-coefficients should be calculated with 

 respect to a rational zero, since it is quite obvious that if, in fact, 

 the data for two given colours of light for an active substance, 

 for that substance in solution, for that substance at different 

 temperatures, and for derivatives of that substance, also under 

 these different conditions, all lie along two straight lines of a 

 characteristic diagram which intersect at some point not zero,^ 

 the rotation for the other colours must remain constant through- 

 out all that wide range, provided, and only provided, the rota- 

 tion values are calculated with respect to the point of intersec- 

 tion and not with respect to the actual zero. 



It will be seen, from what has been said above, that the 

 work of the last ten years has introduced with an almost 

 dramatic suddenness, into what was previously the very 

 confused and difficult field of optical activity, an astonishing 

 degree of uniformity and harmony. 



^ As far as we are aware, no instance is known at which such Unes do actu- 

 ally intersect at zero, in the case of permanent rotation. 



