312 THE CHEMISTKY OF ESSENTIAL OILS
The oil has been exhaustively examined by Schimmel & Co.^1
After repeated fractionation they succeeded in obtaining 11 grms. of a
body with boiling-point 158° to 168° and an odour of pinene. Treated
with nitrosylchloride a very small quantity of a nitrosochloride was
obtained, of which the benzylamine-combination melted at 120° to 123°.
The presence of a-pinene could be more readily proved by oxidation
into active d-pinonic acid. The acid melted at 68° to 69° and was
identical with d-pinonic acid from Greek oil of turpentine. It follows
that the oil contains i- and d-a-pinene.
The fractions with higher boiling-point contained, besides j9-cymene,
small quantities of terpene hydrocarbons. j0-Cymene was identified by
oxidising with permanganate (hydroxy-^sopropyl benzoic acid, melting-
point 155° to 156°). When the cymene fraction was oxidised it yielded
a small proportion of sodium nopinate. The free nopinic acid melted
at 125° to 126° and was laevo-rotatory. On oxidation with perman-
ganate it yielded nopinone, of which the semicarbazone melted at 187°
to 188°, thus proving the presence of /?-pinene. Wolpian, who had
isolated considerable proportions of hydrocuminene from the oil, was
probably dealing with a sample which was adulterated with turpentine
oil, and his hydrocuminene was certainly only a mixture of a-pinene,
/?-pinene, and ^-cymene,
Although it was the fraction boiling at 178° which showed the
smallest rotation (ao4 - 0° 35'), it absorbed about 10 per cent, of
bromine. With hydrochloric acid it yielded a solid dihydrochloride,
melting at 48°, which, when mixed with dipentene dihydrochloride,
suffered no reduction of the melting-point. From the same fraction a
nitrosochloride was also obtained, but only in very small quantities
and only separated out in the solid form upon the addition of water,
alcohol, and ether. These facts point to the presence of dipentene.
The presence of terpinene could not be proved. No fraction gave
the nitrite reaction, and any considerable quantity of phellandrene is
therefore out of the question. The results of the oxidation with per-
manganate indicate the presence of /3-phellandrene, which proves that
when small quantities of phellandrene are present, identification by the
nitrite reaction is very often a failure.
A small quantity of a-terpineol was isolated, together with a small
amount of cuminic alcohol.
The bisulphite compound, after being purified by washing with
alcohol and ether, was decomposed by boiling it with soda solution
and then distilled by steam. Its constants were as follows: boiling-
point 97° to 99° (7 mm.); d^ 0-9731; aD + 0° 3'. When oxidised with
chromic acid the aldehyde was converted into cuminic acid, melting-
point 113° to 114°. The melting-point of the semicarbazone was 210°
to 211°, of the oxime, 55° to 57°. Cuminic aldehyde (cuminol) which
has long been known as forming the principal constituent of oil of
cumin, was thus isolated in a state of great purity. It caused some
surprise, however, to find the first fraction of the aldehyde showing a
slight rotation to the right (OLD + 1° 13'), but oxidation always gave
only cuminic acid. In the course of fractional crystallisation of the
semicarbazone they succeeded in obtaining, in addition to cuminic
aldehyde semicarbazone, crystallising in flakelets (melting-point 210°
i Report, October, 1909, 49.