However, none of these investigations revealed a true insight into the composition of the oil. Wallach4) was the first to demonstrate that the bulk of the oil consists of caryophyllene, a sesquiterpene C15H24 (see Vol. I, p. 331) that also occurs in oil of cloves. If fraction 250 to 270° is treated with glacial acetic and sulphuric acid, caryophyllene hydrate, the handsome crystals of which melt at 96°, results.

From the caryophyllene of copaiba oil E. Deussen and A. Hahn5) succeeded in obtaining the nitrosochloride and the nitrosate of the inactive hydrocarbon designated a-caryophyllene.

In order to clear up the contradictions of the earlier investigators Schimmel & Co.6) examined the products resulting from the action of hydrochloric acid on the oil.

As already stated, the melting point of the hydrochloride is recorded by Blanchet as 77°, by Soubeiran and Capitaine as 54°, whereas Brix and Umney failed to obtain a solid chlor-hydrate.

The oil to be examined (150 g.; d16o0,9036; aD - 9°58') was fractionated into three equal parts:

1) Liebig's Annalen 69 (1849), 67.

2) Ibidem 148 (1868), 148.

3) Ibidem 242 (1887), 189. - Berl. Berichte 18 (1885), 3206, 3209.

4) Liebig's Annalen 271 (1892), 294.

5) Chem Ztg. 34 (1910), 873.

6) Report of Schimmel & Co. Oktober 1910, 184.

I. B.p. 113 to 115° (6 mm.); dl5o0,8989; aD - 8°7'

II. B.p. 115 to 119° (6 mm.); d15o 0,8960; aD - 10° 1'

III. B.p. 119 to 133° (6mm.); d15o 0,8968; aD - 11°58'.

The chlorhydrates were prepared according to the well-known method by passing the hydrogen chloride into the well-cooled ethereal solution. From the last fraction only was it possible to obtain a solid hydrochloride without great difficulty. The residues of the other two fractions had to be chilled thoroughly for a long time and the oily mass, permeated with crystals, had to be spread on porous plates. The melting points of the several chlorhydrates varied from 60 to 90°, hence did not differ materially from those obtained by Blanchet and by Soubeiran and Capitaine. Separation was brought about by fractional crystallization. The bulk of the crystals occurred as fine needles, melted at 113 to 115° ([a]D - 34°28' in a 2,5 p.c. chloroform solution) and the melting point was not lowered when mixed with \-cadinene dihydrochloride. The chlorhydrate that melted at 65 to 70° and which was obtained in smaller amount ([a]D +18°35' in a 2,1 p.c. chloroform solution) was identical with caryophyllene dihydrochloride. The yield of cadinene dihydrochloride amounted to 4 to 5 p.c, that of caryophyllene dihydrochloride about 1/2 p.c. Hence it becomes apparent that B-caryophyllene occurs in the oil in addition to inactive caryo-phyllene, first found by Wallach and characterized as a-caryo-phyllene by Deussen and Hahn, and l-cadinene. However, the presence of other sesquiterpenes has not yet been excluded.

Later E. Deussen and B. Egerl) verified the presence of B-caryophyllene in the several copaiba balsam oils by preparing what they call B-nitrocaryophyllene, a compound of the formula C12H19N806. On account of its insolubility in most organic solvents, this compound is well suited for the identification of B-caryophyllene. For its preparation 2 to 3 g. of the oil to be examined are dissolved in ether and oxides of nitrogen are passed into this solution which is not cooled but protected against light, until crystals no longer separate. After the precipitate has been washed with ether it is weighed. In this manner Deussen and Eger obtained from 9,5 to 16 p.c. of ^-nitrocaryophyllene from four Para copaiba oils, 3 to 6 p. c. from two Maracaibo oils and 8 to 9 p.c. from a Maturin oil. If the oils are adulterated with gurjun balsam oils the amounts are correspondingly less.

1) Liebig's Annalen 388 (1912), 136. - Chem. Ztg. 36 (1912), 561.

In a Surinam copaiba balsam oil, van Itallie and Nieuwland1) found cadinene (m. p. of the dichlorhydrate 116 to 117°; [a]D in chloroformic solution - 36° 5') also a sesquiterpene alcohol C15H260 that melted at 113 to 115° and which, when treated with anhydrous formic acid yielded a sesquiterpene with the following properties: b. p. 252° (759 mm.); d15o0,952; aD - 61,7°; nD16o1,5189. In addition to cadinene other sesquiterpenes occur in the oil, but caryophyllene could not be identified.

Adulteration. Copaiba balsam oil as such is adulterated but rarely, if ever. However, the adulteration of the balsam with gurjun balsam, or African copaiba or Illurin balsam has been observed. Gurjun balsam is distinguished from copaiba balsam by means of a number of color reactions. As in the case with most of these reactions, the results are more or less questionable. Of these tests the following has proven the most reliable. If to a solution of 4 drops of copaiba balsam in 15 cc. glacial acetic acid, 4 to 6 drops of concentrated nitric acid be added, a pure copaiba balsam remains unchanged, whereas in the presence of gurjun balsam the glacial acetic acid solution is colored a purplish-red 2). The oils of the two balsams behave in like manner. A further indication of the presence of gurjun balsam oil is obtained by means of Turner's reaction. According to Deussen and Eger3) one drop of the material to be examined is dissolved in 3 cc. of glacial acetic acid and 2 drops of a freshly prepared 1 p. c. solution of sodium nitrite are added. This solution is carefully poured over concentrated sulphuric acid. If within 5 minutes the glacial acetic acid solution is colored a deep violet the presence of gurjun balsam oil may be assumed.

1) Footnote 3, p. 594.

2) Comp. Utz, Chem. Zentralbl. 1908, II. 1212; Report of Schimmel & Co. April 1909, 41. - The color reactions of the copaiba balsams are discussed in detail by A. Eibner in a contribution entitled: Ober Copaivabalsame und Copaivaole. Technische Mitteil. f. Malerei 24 (1908), Nos. 22 and ff.

3) Chem. Ztg. 36 (1912), 561.

Exact proof of the presence of this adulterant, however, is supplied by the preparation of the gurjunene-ketone semicar-bazone. According to E. Deussen and H. Philipp1) that portion of 170 g. of oil which distils over below 145° under 10 to 12 mm. pressure is resolved into three fractions of 50 g. each, the residue above 145° being ignored. The acetone solution of each of these fractions is oxidized with potassium permanganate. To those that yield Turner's reaction semicarbazide solution is added. Aside from the melting point of 234° when re-crystallized from hot alcohol, the high optical rotation is characteristic of the semicar-bazone. Dissolved in concentrated aqueous chloralhydrate solution, the specific angle of rotation is +317°.

Inasmuch as gurjun balsam oil is strongly laevogyrate, it increases the rotation of the copaiba balsam oil to which it has been added as adulterant. In order to counteract this effect, the oil from the dextrogyrate copaiba balsam has been added. Moreover, gurjun balsam oil has a higher specific gravity than copaiba balsam oil. This fact also may assist in the detection of the former.

T. T. Cocking2) has observed that the oil of unadulterated balsam invariably has a slightly higher rotation than the first 10 p.c. distilled distilled therefrom in vacuum. Hence this may serve as a means to detect additions of gurjun balsam as well as of African balsam. Parry3) claims that this observation does not hold true, but Cocking4) maintains the correctness of his original position. It would seem, therefore, that this method should be further tested out in connection with material of unquestioned purity.