286 
MESSRS. THOMAS MARTIN LOWRY AND PERCY CORLETT AUSTIN 
Table XVIII. {a), are in agreement with G-rossmann’s view. The rotations are positive 
and of the same order of magnitude as in the case of an alkaline solution of the potassium 
salt. The dispersion shows the same rough conformity with Biot’s law, and the same 
close approximation to the requirements of a simple formula with an abnormally small 
dispersion-constant, as in Table XV. (a), but the observed rotations are too small in 
magnitude and too few in number to provide conclusive evidence of the complexity of 
the dispersion. In the absence of an alkali, the formation of an arseno-tartaric acid is 
indicated by the increase of rotatory power, Table XVIII. (6), which is produced by 
saturating a solution of tartaric acid with arsenious oxide and filtering off the excess 
of arsenic. The dispersion-curve is not “ simple,” like that of tartar emetic, since 
the dispersion-ratio a 4358 /a 5461 — 1*476 is less than the minimum value 1*570 for 
A 0 2 = 0 ; it is probable, therefore, that the arseno-tartaric acid is already partially 
dissociated, and that the addition of an excess of alkali merely completes the hydrolysis 
already initiated by the water in the solution. 
The dispersion of the corresponding compounds of bismuth do not appear to have 
been investigated hitherto*, but we have been able to prepare an alkaline solution of a 
bismuthyl compound which shows all the characteristics of alkaline solutions of tartar 
emetic. Its rotatory power is strongly negative, though less than in the case of tartar 
emetic, and its rotatory dispersion, as set out for 21 wave-lengths in Table XIX., agrees 
very well with a simple dispersion formula [M] = — 37*414/(\ 2 —0*0645). The 
dispersion-constant in this formula agrees closely with the value 0*06275 given by an 
alkaline solution of tartar emetic ; it is therefore clear that in alkaline solutions 
bismuth yields a kevorotatory compound with tartaric acid of exactly the same type 
as the compound formed from tartar emetic, and that both compounds differ from 
the acid and from its simple salts in giving simple rotatory dispersion. The simple 
dispersion in these more complex tartrates is indeed the most striking discovery that 
we have made in the course of the present investigation. 
11. Boro-tartaric Acid. 
Boric acid was mentioned in Biot’s sealed communication of August 25, 1834, as a 
substance which (like the alkalis) combined with tartaric acid, giving dextrorotatory 
solutions which obeyed the general laws of rotatory dispersion as they had been estab¬ 
lished in the case of quartz and of several optically-active liquids (' ComptesRendus,’ 
1835, vol. 1, p. 458 ; compare ‘ Mem. Acad. Sci.,’ 1838, vol. 16, p. 271). The action of 
boric acid was described in detail ten years later (‘ Ann. Chim. Phys.,’ 1844, vol. 11, 
pp. 82-112 ; see also 1860, vol. 59, pp. 229-256) in a long memoir “ On the Employment 
of Polarised Light to Study Various Questions of Chemical Mechanics.” The acid 
produced a marked increase in the dextrorotatory power of tartaric acid ; and this 
could be represented by a hyperbolic formula tending to a constant value when the 
* For rotations of sodium light in potassium bismuthyl tartrates, see Rosenheim, Vogelsang, and 
Grossmann, * Zeitschr. fur anorg. Chem.,’ 1906, vol. 48, p. 209. 
