As fuel the exhausted grass is used after it has been dried in the sun. Inasmuch as the Southern Province is poor in wood for fuel, the campaign comes to an end with the rainy season, since the exhausted grass can no longer be dried.

An apparatus 7 ft. high and 4 [ft ft. in diameter yields about 16 to 20 flasks of 22 oz. each, hence 360 to 440 oz. per day.

Elevation

Fig. 28. Floor plan.

Fig. 28. Floor-plan.

Fig. 29.

Fig. 29.

1. Chimney. 2. Boiler. 3. Stills. 4. Condenser. 5. Water basin (condenser). 6. Locked cellar for the storage of the distillate. 7. Roof. 8. Water supply pipe for boiler. 9. Basin for the storage of the distillate.

The grass is never weighed. The still being charged the distillation is begun. Hence more exact figures are not available.

In addition to the steam distillation outfit described above, apparatus involving direct heat are said to be used in a few districts. However, the bulk of the citronella oil is nowadays obtained by steam distillation. Wherever direct heat is employed, water must naturally be added to the grass. It remains to be stated that the grass is not specially dried previous to distillation. However, it is never moist when placed into the still since several hours elapse between the time of cutting and the time of distillation. At a temperature of about 65 to 70° C. in the sun this suffices to remove a large part of the moisture.

The districts in which citronella oil is distilled are graphically outlined on the accompanying map. The number of distilleries in operation in Ceylon is estimated at between 500 and 600.

According to the estimate of N. Wickremaratne1), the acre, when cut four times, yields a total of 18000 lbs. of grass per annum and, if the grass be in good condition, this yields 68 lbs. of oil.

Statistics. The exports of citronella oil from Ceylon are recorded in the following table.

1905 ....

1282471 lbs.

1906 ....

1107655 „

1907 ....

1230159 „

1908 ....

1276965 „

1909 ....

1512084 „

1910 ....

1747934 „

1911 ....

1524275 „

The exports of oil from Ceylon from 1892 to 1912 and the variations in price during this period are graphically represented by the accompanying curve.

Composition. Ceylon citronella oil is a very complex mixture of hydrocarbons and oxygenated constituents. Of terpenes it contains about 10 to 15 p, c.

Fraction 157 to 164° contains camphene-). When hydrogen chloride is passed into an ethereal solution of this fraction, there is obtained, in addition to liquid products, a solid chloride from which camphene can be obtained quantitatively by heating it to 100° with water under pressure. When the same fraction is heated with glacial acetic acid and sulphuric acid, isoborneol (m. p. 212°) results1). If fraction 158 to 162° is oxidized with mercuric acetate2), a mercury compound is formed which, when decomposed with sulphuretted hydrogen, yields solid camphene 3). Somewhat later Schimmel & Co.4) succeeded in isolating solid camphene directly from the oil. Fraction 160 to 163° as well as fraction 164 to 168°, when oxidized with permanganate in alkaline solution, yielded camphene-camphoric acid melting at 142°. That portion of the oil which was not changed by the oxidation, when distilled with steam, solidified in the condenser. After purification by means of alcohol it revealed all of the properties of ordinary camphene. It melted at 49 to 50°; [a]D - 173° 22' in 10,2 p. c. chloroform solution. Inasmuch as the lower fractions contain neither a- nor B-pinene nor sabinene, it must be assumed that they contain considerable amounts of another terpene. This is probably an isomeric camphene which, under certain conditions, may be converted into derivatives of the ordinary camphene.

1) Tropical Agriculturist April 1911; Chemist and Druggist 79(1911), 443. 2) Bertram and Walbaum, Journ. f. prakt. Chem. II. 49 (1894), 16.

Elevation 38Elevation 39

Of other terpenes, the presence of dipentene has been demonstrated (m. p. of tetrabromide 124 5) to 125°")), also that of l-limonene4) (tetrabromide; m. p. of nitrolbenzylamine 92 to 93°). In the course of a fractionation, terpene mixtures of very low specific gravity have been observed. This would seem to indicate the presence of a new terpene, the isolation of which, however, has not yet been successful.

Of oxygenated constituents, methylheptenone (m. p. of semi-carbazone 134 to 135°), of which traces occur in the lower fractions, should be mentioned.

Citronellal is largely responsible for the odor of citronella oil, hence may be regarded as its principal constituent. It is an aldehyde of the formula Cl0H18O. It first attracted the attention of the investigators, though contained in the oil to the extent of from 6 to 10 p. c. only. Because of the instability of this aldehyde, the ideas concerning its composition have undergone several changes in the course of time.

1) Report of Schimmel & Co. October 1899, 13.

2) L. Balbiano and V. Paolini, Berl. Berichte 85 (1902), 2995.

3) Report of Schimmel & Co. April 1903, 84. 4) Ibidem April 1912, 40.

5) Bertram and Walbaum, Journ. f. prakt. Chem. II. 49 (1894), 16. 6) Report of Schimmel & Co. October 1899, 18.

). H. Gladstone1) found the "citronellol", as he called the substance, to boil between 199 and 205° and assigned to it the formula C10H16O. C. R. A. Wright-) assigned to it the boiling point of 210°, and the formula C10H18O. By shaking out the aldehyde with alkali bisulphite and regenerating it with acid, Kremers 3) obtained a product that had no constant boiling point whatever, which he regarded as heptaldehyde. Schimmel & Co.4) designated as "citronellone" the substance which they regenerated from the acid sodium sulphite addition product with soda and which boiled between 205 and 210°. However, they left undecided whether the substance in question was a ketone or an aldehyde. Dodge5) favored the idea that it was an aldehyde but supplied no proof for the aldehyde nature of the "citronella aldehyde" as he called it. This was supplied by Semmler 6) who oxidized the "citronellone" to an acid with the same number of carbon atoms, the citronellic acid, C10H18O2. Hence this substance is now more rationally designated "citronellal", which name indicates its chemical character. The citronellal of citronella oil is dextrogyrate. For properties and derivatives see vol. I, p. 414.

1) Journ. chem. Soc. 25 (1872), 1. - Pharmaceutical Journ. III. 2 (1872), 746; Jahresb. f. Chemie 1872, 815.

2) Pharmaceutical Journ. III. 5 (1874), 233. Both Gladstone and Wright give Andropogon Schcenanthus as the botanical source of citronella oil. From their records, however, it becomes apparent that they worked with citronella oil and not with the oil of A. Schcenanthus. However, the „Unter-suchung des Ols von Andropogon Iwarancusa" of Stenhouse (Liebig's Annalen 50 [1844], 157), which is designated East Indian grass oil by him was, in all probability, not palmarosa oil, but citronella oil. Comp. „Uber Pa/marosao/" by E. Gildemeister and K. Stephan. Arch, der Pharm. 234 (1896), 323.

3) Proceed. Americ. pharm. Ass. 35 (1887), 571; Chem. Zentralbl. 18S8,898.

4) Bericht von Schimmel & Co. October 1888, 17.

5) Dodge proposed the correct formula C10H18O. Americ. chem. Journ. 11 (1889), 456; Chem. Zentralbl. 1890, I. 127.

6) Bed. Berichte 26 (1893), 2254.

Elevation 40