448 



NA TURE 



[April 8, 1922 



constants. Drude himself did not apply his formula 

 to a single member of the vast array of optically 

 active liquids and solutions, which have been pre- 

 pared and studied more especially from the time of 

 Pasteur onwards, and he can, perhaps, scarcely be 

 blamed for this omission, in view of the fact that the 

 rotatory power of the great majority of these media 

 had been determined for one wave-length only. It 

 was therefore not until the problem of rotatory dis- 

 persion had been taken up afresh and new series of 

 exact measurements had been accumulated that the 

 unique merit of Drude's formula was established. 



The results of these new tests were most striking. 

 Fifty series of measurements of magnetic and optical 

 rotatory dispersion were made and classified into 

 groups with similar rotatory dispersion, in order to 

 minimise individual errors of observation. It was 

 then found (Lowry and Dickson, Trans. Chem. Soc, 

 vol. 103, p. 1067, 1913) that the whole of these readings 

 could be expressed within the limits of experimental 

 error by using a single term of Drude's equation, 

 involving only two arbitrary constants — namely, a 

 " rotation-constant," k, and a " dispersion-constant," 

 \^, as set out in the equation a.= 'k\{\^-'K^^. 



The substances examined at this stage were nearly 

 all compounds of simple structure — e.g. secondary 

 alcohols of the aliphatic series ; but the two methyl 

 glucosides, each containing five asymmetric carbon 

 atoms, were proved to obey the same simple law 

 (Lowry and Abram, Trans. Faraday Soc, vol. 10, 

 p. 112, 1914). A somewhat dramatic vindication of 

 Drude's formula, in the case of compounds of much 

 greater complexity, has, however, been provided by 

 the more recent work of Prof. Rupe, of Basel, who 

 published in 1915 (Ann. der Chem., vol. 409, p. 327, 

 1915) a series of measurements of the rotatory power 

 for four different wave-lengths of some forty compounds 

 of the terpene series. In order to determine the 

 mathematical form of the dispersion-curves he plotted 

 a against A, log a against A., log a against i/A, a against 

 i/A, a against i/A^ (to test Biot's equation and Stefan's 

 equation), and aA^ against i/A^ (to test Boltzmann's 

 equation) ; but in no case was there any indication 

 of a linear relationship. The results obtained by 

 plotting i/a against A^, in order to test the validity 

 of the one-term Drude equation (Lowry and Abram, 

 Trans. Chem. Soc, vol. 115, p. 300, 1919), are, however, 

 "most remarkable, since thirty-seven of the forty 

 substances studied by Rupe give straight Hues, and 

 only three show any marked curvature. It is, more- 

 over, noteworthy that two of these exceptional com- 

 pounds agree in containing the group, C : C(C6H5).2, 

 although it is not clear why this group should be 

 associated with the occurrence of abnormal optical 

 properties. 



Further work by Pickard and others has confirmed 

 the fact that the rotatory dispersion of a vast range 

 of organic compounds can be represented by the simple 

 formula a = ^/(A^ - Ao^)^ and that a satisfactory classi- 

 fication of optically active compounds can be made 

 by distinguishing between " simple rotatory disper- 

 sion," where this law holds good within the limits 

 of experimental error, and " complex rotatory dis- 

 persion," where marked deviations from the law are 

 found. 



NO. 2736, VOL. 109] 



3. — Complex and Anomalous Rotatory Dispersion. 



Amongst the substances which do not obey the 

 simple law of rotatory dispersion, tartaric acid and 

 its derivatives have been conspicuous ever since 

 Biot in 1837 directed attention to the peculiar be- 

 haviour of the acid in aqueous and in alcoholic 

 solutions. The principal anomaly noted by Biot was 

 the fact that the rotation, instead of increasing con- 

 tinuously with decreasing wave-length, rose to a 

 maximum in the green, and then diminished again 

 in the blue, indigo, and violet to values almost as 

 low as those observed in the red region of the spectrum ; 

 but the extreme sensitiveness of the rotatory power 

 of the acid to changes of temperature and concentra- 

 tion, as well as to the influence of solvents and of 

 chemical agents, was in Biot's opinion at least as 

 important an anomaly as the maximum in the curve 

 of rotatory dispersion. 



When, however, the quantitative basis for the study 

 of rotatory dispersion had been destroyed, attention 

 was no longer directed to the deviations from the law 

 of inverse squares (which were then recognised as being 

 universal), but to the qualitative peculiarities of the 

 curves, which alone were regarded as justifying the use 

 of the term " anomalous dispersion." The principal 

 anomaly thus selected for special attention was the 

 occurrence of a maximum ; but a reversal of sign or 

 a decrease of optical rotation with diminishing wave- 

 length were sometimes included as anomalies of similar 

 importance. The undue emphasis thus laid upon the 

 qualitative anomalies has had some curious results ; 

 in particular, Winther not only adopted the view that 

 the maximum is the sole criterion of anomalous rotatory 

 dispersion, but actually insisted that this maximum 

 must lie within the visible region of the spectrum. 

 He therefore speaks of a dispersion-curve which 

 " becomes normal in that the maximum passes into 

 the ultra-violet," whilst a curve which cuts right across 

 the axis is described as " normal with a maximum in 

 the infra-red." A definition of anomalous dispersion 

 which thus depends on the physiological properties of 

 the eye, instead of on the physical properties of the 

 medium, can scarcely be regarded as worthy of serious 

 consideration, but it provides a suitable anticlimax to 

 direct attention to the value of the more precise methods 

 of treatment which prevailed when rotatory dispersion 

 was first studied almost a century before. 



A complete solution of the problem of anomalous 

 rotatory dispersion has been found by returning to the 

 mathematical methods of Biot and applying similar 

 processes of analysis to curves plotted with the greater 

 accuracy which modem physical apparatus has 

 rendered possible. A series of dispersion-curves 

 (Fig. i) for aqueous solutions of tartaric acid of different 

 concentrations will illustrate the typical forms of the 

 curves that are encountered in studying the substances 

 of this group. 



These curves show clearly three principal anomalies 

 — inflexion, maximum, and reversal of sign — appearing 

 at various points on the experimental curves as the 

 concentration of the solutions is altered. 



Similar curves, but covering a wider range, are 

 obtained when the esters of tartaric acid — e.g. methyl 

 tartrate and ethyl tartrate (Fig. 2) — are examined as 



