27 
1918-19.] Researches in Optical Activity. 
the curves obtained being also shown in the diagram in fig. 3. But in 
this we were disappointed, for although the solvents altered the rotation 
to some extent, the change is not very great. In cinnamic aldehyde 
the general character of the curves appears to be almost the same as 
in the homogeneous ester ; in ethylene bromide there appears to be 
some slight difference in form, which we think can hardly be due to 
experimental error, but to which in the meantime we can do no more than 
direct attention. 
Since the T-R curves for homogeneous 'isobutyl tartrate and isobutyl 
dibenzoyltartrate thus show, directly, no common feature such as, for 
example, a similar maximum or minimum, or the intersection which is 
known as anomalous rotation-dispersion, their relationship is not so obvious 
as might otherwise be the case. That they are connected with one another 
is rendered clearer by the application of Armstrong and Walker’s character- 
istic diagram ( Proc . Roy. Soc., 1913, [A], 88, 392), the interpretation of 
which, from the point of view of the T-R curves, has already been dis- 
cussed by one of us ( J.C.S. , 1916, 109, 1180, 1195). From what was there 
said it is to be expected that if the T-R curves for isobutyl dibenzoyl- 
tartrate and those for isobutyl tartrate or ethyl tartrate are related, as 
suggested above, then the experimental data corresponding to the region 
y z a b c of fig. 1 should lie along a common line in the characteristic 
diagram. The points on the curve for ethyl tartrate in ethylene bromide, 
which clearly belong to this region, should lie along the same line, whereas 
points belonging to the T-R curves for 'isobutyl dibenzoyltartrate at 
temperatures below that at which the minimum occurs — that is to say, 
in the region of x (fig. 1) — should not necessarily be expected to lie along 
the same line on the characteristic diagram. It will be seen from fig. 4 * 
that this is actually the case. The points marked 1, 2, and 3 (shown also 
in fig. 3), for 'isobutyl dibenzoyltartrate, obviously do not lie on the line 
A B in fig. 4. But the points 4 and 5, which are on the part of the curve 
corresponding to z y in fig. 1, do lie on the line A B, and this has almost 
the same direction as that joining the two points for ethyl tartrate in 
ethylene bromide C D, both being close to that for ethyl tartrate itself, 
E F. Considering that we are comparing the behaviour of ethyl tartrate 
* The diagram is drawn according to the author’s modification of Armstrong and 
Walker’s method. Kotation values for Hg g are plotted along the horizontal axis (the 
reference line), the differences between the rotation values for Hg v and Hg g , and Hg y and 
Hg g respectively being then plotted, according to sign, vertically above or below the 
corresponding values for Hg g . The actual rotation value at any point on the diagram 
is the horizontal distance from the zero-point, plus (or minus) the vertical distance from 
the reference line. (See J.C.S. , 1916, 109, 1181.) 
