Origin. According to E. M. Holmes3), the leaves of Barosma pulchella, (L.) Bartl. et Wendl., which are occasionally admixed with those of B. betulina and with which they are occasionally confounded, differ from the latter by their citronella-like odor and are smaller. They are about 7 to 12 mm. long and 4 mm. wide. The petiole is short, the leaf ovate lanceolate, serrulate, with a blunt tip and broadest at the base. Like the other species of Barosma this shrub is a native of Cape Colony.

1) Berl. Berichte 39 (1906), 1158; Chem. Ztg. 30 (1906), 1208. 2) Kondakow, Journ. f. prakt. Chem. II. 72 (1905), 186. 3) Pharmaceutical Journ. 79 (1907), 598; comp. also Chemist and Druggist 71 (1907), 702.

Properties. The oil has been distilled and examined by Schimmel & Co.1). The leaves used contained a small amount of those of B. pulchella var. major and of B. latifolia, (LA.) Rom. et Schult. Upon distillation they yielded 3 p. c. of a golden-yellow oil with a decided odor of citronella but accompanied by an unpleasant narcotic odor which prevents the use of the oil. d15o 0,8830; aD + 8°36'; nD20o1,45771 ; A. V. 18,5; E. V. 27,2; E. V. after acetylation 237,0 = 79,3 p. c. of C10H18O.

Composition1). The unpleasant odor referred to is due to a base which can be extracted from the oil with a 25 p. c. solution of tartaric acid. The base isolated from the tartrate solution by means of soda possessed an unusually strong, narcotic odor. Its boiling point was not constant, varying from 130 to 140° (5 mm.). The solution of its hydrochloride did not yield a crystalline double salt with platinic chloride.

The principal constituent of the oil is d-citronellal, which, after regeneration from the bisulphite addition product, has the following properties: b. p. 205 to 208° (atmospheric pressure) 73 to 75° (7 to 8 mm.); d15o 0,8560; aD + 13°6'; nD20o1,44710. The semicarbazone melted at 81 to 82°. From 500 g. of oil about 220 g. of aldehyde were isolated. The lowest fractions contained methyl heptenone (m. p. of semicarbazone 134 to 135°). Another ketone also appears to be present.

In fraction 75 to 82° (5 mm.) which showed an optical rotation of + 6°54', the presence of a small amount of d-menthone was demonstrated (m. p. of semicarbazone 178 to 181°). d-Ci-tronellol is also contained in the oil (b. p. 93 to 95° under 5 to 6 mm. pressure; d15o0,8723; aD + 2°14'; nD20o1,46288; m. p. of silver salt of acid phthalate 125°).

When acted upon with alkali, the original oil yielded an acid which, in all probability, is identical with citronellic acid, C10H18O2: b. p. 257 to 263° (ordinary pressure); 125 to 131° (5 to 6 mm.); d15o,0,9394; aD + 5°2'; nD20o1,45611. Inasmuch as the amide melted undefinitely at 87 to 88° the amide of citronellic acid was prepared according to the directions of Tiemann2) by means of the incomplete saponification of the nitrile of citronellic acid with alcoholic potassa. The citronellic acid amide1) from citronellal, recrystallized from petroleum ether or dilute alcohol, melted at 84 to 85°; the acid amide from the acid of the above-mentioned oil melted at 87 to 88°; a mixture of both derivatives melted at 80 to 82°. So far as appearance is concerned the two cannot be distinguished. It may be that the amide with the undefinite melting point of 87 to 88° is a mixture of the amide of citronellic acid with traces of the amide of another acid (possibly caprinic acid). It was further ascertained that the citronellic acid from citronellal also boils at 257° and its chloride, like that of the acid from the above oil, boiled at 122 to 123° (5 to 6 mm.). Hence they may be regarded as identical.

1) Report of Schimmel & Co. April 1909, 94; April 1910, 20. 2) Berl. Berichte 31 (1898), 2902.