478 THE CHEMISTEY OF ESSENTIAL OILS
When African copaiba is examined in this manner, the rotations of
all the fractions are dextro-rotatory, and the rotations of the successive
fractions increase, but to a much greater extent than with the South
American copaiba, in consequence of which the difference value is much
greater than copaiba and is a positive figure. The figures also show a
curious feature in that the tenth fraction has a considerably lower rota-
tion than the ninth. As would be expected from the fact that the range
of boiling-points of the constituents of the volatile oils from the two
varieties are practically identical, a mixture of the two will distil over
containing proportional parts in each fraction, and the presence of
African balsam will be shown at once by the difference value being
positive. In some cases where only a small percentage of the adulterant
is present, all the fractions will be laevo-rotatory but the difference value
will be positive. When the same process is applied to gurjun oil, like
copaiba, it gives laevo-rotatory fractions, but, unlike it, they successively
decrease instead of increasing, and thus give a positive difference value,
similar to African copaiba.
With the true copaibas the rotation of the first fraction is in every
case lower than that of the original oil, but in the adulterated samples it
is higher. It is important that the distillation of the oil should be con-
ducted in vacuo, since, if carried on under ordinary pressure, the higher
temperature necessary causes some decomposition, which entirely alters
the optical rotation.
The principal adulterations of copaiba and its oil are the corresponding
balsams and oils known as African copaiba (probably obtained from
Oxystigma Mannii and gurjun balsam).
The author dissents from this view entirely, and considers that there
is considerable objection to fixing this arbitrary standard, based on the
examination of a few samples, and then jumping to the conclusion that
samples of a product of known variability to a very high degree must
necessarily be adulterated. The tablel on the opposite page illustrates
the behaviour on fractionation, and the general characters of a number
of samples of balsams and oils, of which samples 0 to 5 are stated to
be apparently genuine, whilst samples 6 to 12 are assumed to be
adulterated, and the remaining few being quoted as illustrations of the
difficulty of drawing an opinion on fractionating results
The last column in the opposite table, headed " difference value,"
indicates the optical rotation of the last 10 per cent, of the oil left as a
residue on distillation, minus that of the first 10 per cent, distilled, which
should show a minus figure if Cooking's deductions are correct. In re-
gard to this point the authors above quoted remark as follows :—
" That generalisation which is conveniently referred to as Cooking's
(its most persistent advocate) for the detection in particular of admixed
African oil, whereby such dextro-rotatory substances become evident on
fractionation through their influence on the optical rotations of the first
and last 10 per cent.—the positions of maximum divergence owing to
opposite progressions—would appear to be under some suspicion.
"This method of analysis has again been of much service, and in the
experience of this laboratory, has been sufficiently reliable for deciding
commercially the purity of all balsams except that from Central America
(Maracaibo), which variety, as pointed out a year ago, is at times subject
(^1) Evans, Analytical Notes (1912), 26.