On rectification about 2 per cent, of acid water and some aldehydes came over below 1730 C. (corr.). Between 173-188°, 51 per cent. distilled; between 188-255°, 3° Per cent, came over, and between 255-272°, 7 per cent, distilled. These fractions gave the following results: -

First fraction, sp. gr. at 15o C.

=

0.8710; rotation aD - 7.5°; refractive index at 20° = 1.4811.

Second

"

"

"

=

0.9127; rotation aD - 21.20; refractive index at 20° = 1.4929.

Third

"

"

"

=

0.9333; rotation not taken; refractive index at 200 = 1.5040.

The left rotation of the first fraction was partly due to the phellandrene, while that of the second fraction was almost entirely owing to the aromadendral. The high refractive index of the first fraction, taken in conjunction with the low specific gravity, was evidently due to the influence of the cymene.

The cineol was determined by the resorcinol method in the first fraction; when calculated for the crude oil the result was II per cent.

For the determination and identification of the low-boiling constituents in the oil of this species another portion was distilled. The oil which came over below 1850 C. was added to that of the first distillation, and the whole refractionated, using a rod and disc stillhead. Although commencing to distil at 1570, yet only 7 per cent, came over below 1700 C. (corr.); but at 173°, 38 per cent. had distilled. The three fractions were separated as follows: - 157-173°, 38 per cent.; 173-1750, 31 per cent.; 175-179°, 15 per cent.; residue boiling above 179°, 16 per cent. The fractions and residue gave the following results: -

First traction, sp. gr. at 15o C.

=

0.8047; rotation aD - 4.5 ; retractive index at 200 = 1.4777.

Second

"

"

"

=

0.8676; rotation aD - 6.4°; refractive index at 200 = 1.4806.

Third

"

"

"

=

0.8714; rotation aD - 7.2°; refractive index at 200 = 1.4812.

Residue

"

"

"

=

0.9063; rotation aD - 17.8°; refractive index at 200 = 1.4880.

It was thus evident that the amount of pinene could be but small, and that it was either inactive or slightly lævo-rotatory. The high refractive index, and low specific gravity, suggests the presence of cymene. Phellandrene was detected in all three fractions.

Cineol was also detected in all three fractions, and as this constituent was small in amount it was removed by agitating each fraction with 50 per cent, resorcinol, and the uncombined portions washed and dried. The influence of the cineol is shown from the following figures, when compared with those given above: -

First fraction, sp. gr. at 15° C.

=

0.8613;

refractive index at 20°

=

1.4805.

Second

"

"

"

=

0.8649;

refractive index at 200

=

1.4842.

Third

"

"

"

=

0.8667:

refractive index at 20°

=

1.4856.

Cymene. - The determination of the results as recorded above suggested the presence of cymene, and to arrive at the identity of this hydrocarbon, the fraction of the rectified oil, thought to contain cymene in greatest quantity, was oxidised by an aqueous solution of potassium permanganate (12 grams KMnO4 in 330 grams water), as suggested by Wallach; about 2 grams of oil at the time were added, and the solution heated on a rapidly boiling water-bath. When the reaction was completed the precipitate was filtered off, the filtrate evaporated to dryness, the salt boiled out by alcohol, evaporated to small bulk, water added and acidified with sulphuric acid. The separated acid was purified from alcohol; it then melted at 155-156° C. From the method of formation, and its melting point, there seems little doubt but that the acid formed was p-oxyisopropylbenzoic acid. This result, together with other factors, indicates the presence of p-cymene in some quantity in the oil of E. rostrata.

It will thus be seen that this oil differs entirely from that derived from its variety borealis.

(For the determination of the aromadendral see the article in this work on "The cyclic (aromatic) aldehydes occurring in Eucalyptus oils.")