ON OPTICAL ROTATORY DISPERSION. 
281 
from a RED = 15-4 to 10-90 (‘ Mem. Acad. Sci.,’ 1838, vol. 16, p. 280) ; but a concen¬ 
trated solution containing 
C 4 H, ; 0 (i 22 • 68 gram, S0 3 65 • 02 gram, H 2 0 95 • 90 gram 
gave a much lower rotation and a novel type of dispersion in which the maximum 
disappeared into the infra-red, whilst a reversal of sign occurred in the blue region, 
giving rise to laevorotations in the violet thus : 
red +2*200°, yellow +0*850°, green +0*550°, violet —4*950° 
(ibid., p. 301). Grossmann has recently examined solutions of tartaric acid in sulphuric 
acid of higher concentrations (including the anhydrous acid containing 100 per cent. 
H 2 SO,), and has shown that laevorotations are found for violet light of wave-length 
4620 from 45 to 75 per cent, of sulphuric acid only ; when sulphuric acid containing less 
than 35 per cent, of water is used as a solvent, the tartaric acid increases again in rotatory 
power, and finally gives a specific rotation from 3| to 7 times as great as in aqueous 
solutions ( £ Trans. Faraday Soc.,’ 1914, vol. 10, p. 67). This case is of special interest 
on account of the clear evidence which it affords of the formation of some new chemical 
compound of high dextrorotatory power, perhaps an anhydride : 
HO.CH.CO 
/ 
for which 
HO.CH.CO 
Hd = +124°, 
O, compare 
HO.CH.CO 
>NCH:s 
HO.CH.CO' 
[M] d = +281°, or a sulphate : 
/O.CH.CO.OH HO.CH.CO.O x 
S0 2 < I or I >S0 2 
O.CH.CO.OH HO.CH.CO.O 
Of special interest is the fact, which is disclosed by plotting 1/a against X 2 , that, 
whatever the nature of this compound may be, it is sufficiently “ fixed ” to give rise 
to a rotatory dispersion which, in the case of five readings out of six, appears to obey 
the “ simple ” law a = &/(A 2 — A 0 2 ). In view of the importance of these observations 
we made several attempts to repeat them, in order to test the validity of the “ simple ” 
dispersion law by means of fresh data extending over a wider range and including a 
larger number of wave-lengths ; but, up to the present, we have not succeeded in prepar¬ 
ing solutions sufficiently clear to use for such a test. 
In quite a different category must be placed Biot’s observation that aluminium 
tartrate, which is strongly dextrorotatory in dilute solutions, becomes laevorotatory 
when the solution is concentrated (‘ Comptes Rendus,’ 1835, vol. 1, p. 459 ; ‘ Mem. 
Acad. Sci.,’ 1838, vol. 16, Tables 12, 13, 14, at end of volume). It is possible that 
tervalent aluminium behaves in some respects like boron or antimony, and that concen¬ 
tration is accompanied by a change of structure analogous with that which is produced 
2 R 
vol. ccxxn.— A. 
