The Distribution of Safrol.


The prevailing constituent of the essential oil of sassafras root is Safrol, as will be seen in the text-books; for instance, in 'Pharmacographia,' second edition p. 536. In the crude oil, safrol is held in solution by the hydrocarbon safrene, C10H16, and may be separated either by fractional distillation or by cooling the oil. Safrol liquefies at 12° (53.6°F.) and yields very large and fine prisms, which I caused to be exactly investigated crystallographically by Professor Arzcuni, as mentioned in 'Pharmacographia.' The large crystals of safrol are very little softer than those of gypsum. Although they cannot be kept at a temperature exceeding their melting point, they were, curiously enough known in England a century and a half ago.

As to the constitution of safrol it has been shown by Eykman that by means of permanganate of potassium it yields piperonylic acid

C6H3( O

safrol, therefore, may be represented by the formula—

C6H3< O
\ C3H5

Piperonylic acid is obtained by oxidizing piperonal (now known, in perfumery, under the name of heliotropin)—

C6H3( O

and Poleck thinks that he has observed piperonal among the products of the treatment of safrol with permanganate of potassium, for the action of this salt on safrol is by no means very simple, 4 per cent. of piperonylic acid only having been obtained by Poleck.

Sassafras is not the only plant containing safrol. In the same natural order viz. that of Lauraceae, Mespilodaphne Sassafras, Meissner, a Brazilian tree, has a bark resembling safrol in odor. The same is also well known with regard to the Puchury nuts, or sassafras nuts, the cotyledons of two Brazilian species of Nectandra, a genus as yet very imperfectly known.

Again, the order of Monimiaceae, tribe Atherospermeae, closely allied to Lauraceae, would appear to be provided with safrol. Of this at least the aroma of the Australian 'sassafras bark' is strongly suggestive. This drug, which is not seldom seen in the London market, is the bark of Atherosperma moschatum, Labillardière, a tree indigenous to Australia and Tasmania. The bark of Doryphora Sassafras of New Caledonia, likewise of the order of Monimiacea—Atherospermeae, also smells of sassafras.

Although there can be but little doubt as to safrol really occurring in all those essential oils of the just-named plants the fact has not yet been proved.

This, however, has been most surprisingly done by the well-known house of Schimmel & Co., of Leipzig, with regard to the oil of the camphor tree Cinnamomum Camphora. Since 1885 the said house is manufacturing safrol from camphor oil to a very large extent. No doubt there is now much more safrol being made in the state of absolute purity at Leipzig than they are able to distil crude oil of sassafras in the United States.

Cinnamomum Parthenoxylon, Meissner and C. glanduliferum, Meissner, the former tree belonging to the forests of Penang, Sumatra and Java (Kayu-gadis of the Malays), perhaps also in Tennasserim; the second in Nepal, Sikkim, Bhootan and Khasia ("Sassafras of Nepal"), are also known for their odor resembling that of true sassafras. They would deserve a chemical investigation.

I am struck, lastly, with the very strong odor of the same kind displayed by the bark of an Australian tree, which has been described by Bentham (assisted by Ferdinand Müller) in the 'Flora Australiensis,' vol. v. (1870), p. 299, under the name of Nesodaphne obtusifolia. It is a large and handsome tree, growing in Queensland, Rockingham Bay, Fitzroy River, Rockhampton, Archer's Creek (according to Leichhardt), also in New South Wales, Clarence River. Hooker and Bentham, 'Genera Plantarum,' iii. (1880), p. 152, ultimately unite the genus Nesodaphne to Beilschmiedia; the tree under notice is, therefore, to be called Beilschmiedia obtusifolia, Benth. and Hook.

Dr. Joseph Bancroft, in his 'Contributions to Pharmacy from Queensland' (Colonial and Indian Exhibition of 1886, London), p. 11, states that the tree grows in the rich scrubs to the north of Brisbane. Its grey, rough bark, reddish-brown internally, has a strong aromatic odor and pleasant astringent taste, and is frequently used by bushmen to improve the flavor of their tea. The bark, according to Mr. Staiger, affords about 2 per cent. of volatile oil heavier than water, and 9 per cent. of tannin.

Being indebted to Mr. E. Merck of Darmstadt, for a good sample of the bark of Beilschmiedia obtusifolia, I may state that it agrees to some degree, in its microscopic structure both with the bark of cassia lignea and sassafras. The Beilschmiedia bark is as much as 15 millimetres (half an inch) in thickness, and shows the same exfoliation due to secondary cork bands (rhytidoma) as the bark of sassafras. Beilschmiedia bark is, on the other hand, much more fibrous than either of the above-named barks; its tissue being very rich in long thin fibres, and in its outer layer there are scattered, not in large number, sclerenchymatous cells, having comparatively thin walls. The oil ducts of Beilschmiedia bark are neither very numerous, nor considerably large. It remains to be proved that they really contain safrol as I venture to say.

In the natural system, the Magnoliaceae are not much distant from both Lauraceae and Monimiaceae. Eykman has shown that safrol also occurs in the essential oil of the fruit of Illicium religiosum, the false star-anise of Japan; this tree belongs to the order of Magnoliaceae. There the safrol is accompanied by eugenol, the formula of which


at once reveals its relationship to safrol as also to anethol

C6H4< OCH3

It would appear, therefore that at least eugenol C10H12O2 and safrol C10H10O2 may be in some generic relation. The former has been met with by Stenhouse in the leaves of the cinnamon tree. And, thirdly, anethol C10H12O, the chief constituent of the oil of true star-anise, Illicium anisatum, is there replacing safrol as occurring in the other variety.

It would be interesting to be able to convert one into the other of those three highly aromatic substances; their empirical formulae: C10H12O C10H10O2, C10H12O2, would apparently indicate the probability of such transformations, but not their structure—Phar Jour. and Trans. June 4th, 1887) p. 989.

The American Journal of Pharmacy, Vol. 59, 1887, was edited by John M. Maisch.