E. V. 113,2; E. V. after acetylation 160,2. Nossi-Be5): d15o.0,9622 to 0,9737; aD - 39°15' to -43°14'; nD20o1,50689 to 1,50974; A. V. 0,9 to 1,4; E. V. 134,5 to 159,1; soluble in 0,5 to 1 vol. of 90 p. c. alcohol. The addition of more solvent causes turbidity. Seychelles3)6): d15o0,9201; aD - 29°6'; nD20o 1,49157; A. V. 1,8;

E. V. 40,2; also d15o 0,924; aD - 18°46' and d15o 0,958; aD - 45° 27'. Arnani7): The following constants were observed in connection with an oil distilled from fresh flowers, i. e. one obtained in a normal way: d15o 0,9366; aD - 17°0'; nD20o 1,48451;

A. V. 1,1; E. V. 136,3; soluble in 8 vol. and more of 70 p.c. alcohol with opalescence. Mayotte.): d15o 0,9324 to 0,9651; aD - 4° 4' to - 53° 56'; nD20o1,48451; A. V. 1 to 1,4; E. V. 129,7 to 136,3; E. V.

It should be added that Bacon9) has made experiments with the object of extracting ylang-ylang oil with various volatile solvents and that he obtained the best results with petroleum ether. The yield varied between 0,7 and 1 p.c. A rather dark oil containing considerable resin resulted. Undiluted, it did not have an extraordinarily pleasant or strong odor, but when greatly diluted the floral odor became very marked. The constants of this oil were: d30/4o0,940; nD30o1,4920; E.V. 135; E.V. after ace-tylation 208. It will be seen that these properties agree with those of first class oils. The somewhat different odor of the extracted oil (which is said to have the advantage that it cannot be reproduced synthetically and, therefore, should fetch a higher price) is attributed to the presence of small amounts of unknown constituents that are readily decomposed by heat. If the extracted oil be shaken with water fairly considerable amounts of resin are separated which have the characteristic odor of the flowers, whereas the odor of this oil itself then reminds of that of p-cresol methyl ether.

1) Comp. also Berichte von Roure-Bertrand Fils April 1910,67; E. Tassilly, Bull. Sciences Pharmacol. 17 (1910), 1; Report of Schimmel & Co. April 1910, 122.

2) Report of Schimmel & Co. October 1890, 60.

3) Observations made in the laboratory of Schimmel & Co.

4) Report of Schimmel & Co. October 1908, 135.

5) Observations made in the laboratory of Schimmel & Co.

6) Bull. Imp. Inst. 6 (1908), 110; Report of Schimmel & Co. October 1908, 135.

7) Report of Schimmel & Co. October 1911, 102.

9) Berichte von Roure-Bertrand Fils April 1910, 66; October 1911, 43.

10) Philippine journ. of Ss. 4 (1909), A, 127.

Properties of Cananga Oil

As already pointed out, the name cananga oil is applied both to higher fractions which are obtained as a by-product in the distillation of cananga oil (Manila), and to the total distillate of the flowers (Java). Hence it will be readily understood that the properties of the oil vary even more than those of the ylang-ylang oil: d15o 0,906 to 0,950; aD - 17 to - 55°; nD20o1,495 to 1,510; A.V. up to 2,0; E.V. 10 to 35. In 90 p.c. alcohol most of the cananga oils are not completely soluble: in 95 p.c. alcohol they dissolve-at first, but the addition of more than 1 to 2 vol. commonly produces turbidity.

In connection with 36 samples of second grade ylang-ylang oils from Manila, Bacon1) observed the following constants: d30/4o0,896 to 0,942; aD80. - 27,4 to - 87°; nD80.1,4788 to 1,5082; E.V. 42 to 94; E.V. after acetylation 96 to 141. Dried cananga flowers from Samoa2), which are known as Mosoi, yielded upon distillation 1 p.c. of oil, which, while it differed as to odor from the oil distilled from fresh flowers, nevertheless had the same general character. Like the ordinary cananga oil it contained benzoic acid: d15o0,922.

•1) Loc. at.

2) Report of Schimmel & Co. October 1890, 60.

The distillate varies according to whether "green" (not fully developed) or yellow (fully developed) flowers are used. Experiments along this line were conducted in the municipal pharmacy in Samarang1).

1. Oil from green flowers: d15o0,930; aD - 19°21'; S.V.24,31; insoluble in 10 vol. of 95 p.c. alcohol.

2. Oil from yellow flowers: d15o0,956; aD - 25°11'. Solubility same as under 1. It contained 12 p.c. eugenol.

Oil from Bangkok2). Distilled from fresh and dried flowers: d15o0,920; aD - 51°40'; nD80o1,50510; A. V. 1,82; E.V. 34,17. Oil from the Seychelles3): d15o0,954; aD - 43° 10'.

Adulteration. Adulteration with cocoa-nut fat has been observed repeatedly in connection with cananga oil4). Such an addition is readily recognized. While both specific gravity and optical rotation are influenced but little, the saponification value is greatly increased and the oil is no longer soluble in 95 p.c. alcohol. Cocoa-nut fat can also be recognized readily by placing a sample of the questionable oil into a freezing mixture of ice and salt. Whereas the pure oil remains liquid, an oil to which much cocoa-nut fat has been added congeals to a butyraceous mass.

Smaller amounts of fat, also fatty oils that do not congeal in the cold, remain in the residue when the oil is distilled with steam and can thus be detected. It should be noted, however, that even unadulterated oils leave a residue of up to 5 p.c. when distilled with water vapor. Hence, in an approximately quantitative determination of the added fat or oil 5 p.c. should be deducted from the amount of residue found. Comp. also vol. I, p. 613.

Among other adulterants occasionally employed in the Philippines, Bacon mentions turpentine oil, alcohol and petroleum.

For this purpose the turpentine oil is sprinkled over the flowers and distilled along with them. Inasmuch as pinene is a normal though minor constituent, the detection of such an adulteration, when practised moderately, is scarcely possible. Moreover, the presence of terpenes in ylang-ylang oil may be due to the utilization of immature flowers. In 100 cc. of an oil that had been distilled from good flowers, Bacon could not detect pinene or other terpenes. The distillation of imperfectly developed flowers, however, yielded an oil the odor of which resembled that of turpentine and bananas and which, without doubt, contained terpenes. A turpentine oil that had been obtained for the purpose of adulterating ylang-ylang oil revealed itself as a carefully purified dextrogyrate product to which a trace of peppermint oil had been added. The addition of appreciable amounts of turpentine oil imparts a pungent odor to ylang-ylang oil. In addition the specific gravity, optical rotation and index of refraction are lowered. If upon fractionation of 100 cc. of oil under 10 mm. pressure more than 1 cc. passes over below 65°, the addition of turpentine oil or some other low boiling substance may be assumed. The identification of pinene is accomplished in the usual manner by means of its nitrosochloride.

1) Report of Schimmel & Co. April 1899, 9.

2) Ibidem April 1904, 18.

3) Bull. Imp. Inst. 6 (1908), 111.

4) Report of Schimmel & Co. October 1897, 8; April 1905, 19.

For the detection of alcohol and petroleum see vol. I, pp. 612 and 614.

Physiological Action. P. Kettenhofen1), who has investigated the action of ylang-ylang oil on micro-organisms, on the white blood corpuscles, upon cold-blooded and warm-blooded animals, its influence on the amplitude of respiration and on the blood pressure, also its action during normal and stimulated irritability, has arrived at the following conclusions. Ylang-ylang oil harms or kills micro-organisms and prevents decay and fermentation as a result of its paralysing action on the protoplasm of organisms of putrefraction and decay. Toward the white blood corpuscles it exerts an analogous action; the corpuscles are paralysed and are prevented from leaving the blood vessels, thus suppressing any incipient suppuration. In the organism of cold-blooded animals, ylang-ylang oil produces, even in small doses, a general paralysis, which results in death when the dose is increased. In warm-blooded animals a temporary diminution of the functions results, however, without appreciable detriment to the organism. The frequency of pulse and breathing are diminished, both breathing and blood pressure become fainter. The animals show a somewhat apathetic behavior. If the reflex activities be stimulated by cramp poisons, they are moderated by ylang-ylang oil and the cramps cease. If inferences from these results be drawn with regard to man, Kettenhofen assumes that with moderate doses, aside from the symptoms mentioned above, no serious illness will be caused. The normal reflex activity is at first diminished and, with sufficiently large doses, ceases altogether.

1) Inaug. Dissert. Bonn 1906.

In the treatment of malaria it might be applied as a substitute for quinine wherever this is incompatible or cannot be administered for other specific reasons.

Production and Commerce. According to the official U. S. A. statistics, the values of ylang-ylang oil exported from Manila are as follows1):

1902/3...........................................

103789

1903/4...................................

103247 „

1904/5..................................

100349 „

1905........... ...........................................

93918 „

1906 ............................................... ....

99009 „

1908.........................................................

181 638 „

According to English consular reports, the export diminished during the following years until the normal had been reached. The quantities by weight recorded in the American statistics for ylang-ylang oil are gross weights, hence are misleading, since in the case of ylang-ylang oil the net weight is about one tenth of the gross weight.

According to French statistics the following amounts were exported from Reunion: -

1906 ....

118 kg.

at a value of 51630 fr.

1907 ....

1908 ....

1909 ....

523 „ 1126 „ 1792 „

an annual average of: 1237 kg. at a value of 418437 fr.

1910 ....

2373 „

at a value of 579094 fr.

1) The statistical sources do not indicate whether the last three years are calendar years or fiscal years which run from July 1 to June 30.

As regards exports of cananga oil from Java, no statistical data are available.