ARBUDAS BRAZILIAN PLANTS. 249 ceding species, but is ligneous and flexible, and can easily be broken; it is covered with a layer of astringent pulp. The almond is a seed composed of two oily cotyledons of a disagreeable taste, but abounding in oil, of which some use is now made." — Koster, 500. Sir Joseph Hooker {I. c.) says, "Oiti-cica, PL umbrosissimam e deserto Pernam- bucano ad descriptas species reducere non audeo." PsiDiuM piGMEUM. Mavangoba. Koster, 492. " This is a species of goiaha plant which does not attain more than two or three feet in height ; it abounds in the chapada of the Serra Araripe of Cariri Novo." — I. c. This may be P. humile Veil., but in the absence of material it cannot be decided, especially as the name " Marangaba" is not taken up by Miers or in the books. It seems, however, to be known to travellers : Gardner (Trav. in Brazil, 193) says he found " on the top of the Serra a species called Marangaba ; it is the Psidium pigmeum of Arruda, a shrub from one to two feet high, the fruit of which is about the size of a gooseberry, and is greatly sought after on account of its delicious flavour, which resembles that of the strawberry." Mr. Jackson cites '^ pygmaum Veil, ex Steud. = pigmaBum," and this suggests a remark on the way in which bogus names get into circulation. For, in the first place, Steudel does not say pyg- maeum Veil.," but pygmaeum Arruda," and by so doing he merely means to correct Arruda's spelling — just as Mr. Jackson himself does when he writes "pigmaeum Arruda," for Arruda wrote "pig- meum." I do not say that Steudel' s speUing should not have been quoted, but I am curious to know what will be done in the future with " pygmaeum Veil." Vellozo never used the name : will future index-makers have to enter " pygmaeum Veil, ex Jacks." ? RiBEiREA soRBiLis. Mangahdva. Koster, 499. I have already cited Arruda's account of his dedication of the plant. Jackson doubtfully refers it to Rosacece, but Miiller (Fl. Bras. vi. i. 186), Miers, and others identify it with Hancornia speciosa B. A. Gomes in Mem. Acad. Scieuc. Lisboa, iii. (Mem. dos Corresp. 51, t. i.), and this identification is clearly correct. If the date of the publication of Gomes's paper were, as accepted by Jackson and by authors generally, 1812, Arruda's name ^ibeirea (1810) would have to replace Hancornia (1812). But it appears that Gomes's paper appeared first as an independent publication in Lisbon in 1803, f and was subsequently reprinted in the Memorias.l The date to be assigned to Hancornia must therefore be 1803, not 1812 as given in Index Kewensis. Hancornia commemorates an Englishman who finds no place in the Biogr. Index Brit, Botanists^ and it may be worth while to
- I have before expressed my opinion (Journ. Bot. 1894, 375-6), which
experience confirms, that Mr. Jackson should have given the names as spelt by their authors, and should not have attempted to make them conform to more accurate usage. t It originally appeared in two parts, which explains the two series of numbers on the plates in the Memorias. X See Pritzel, no. 3465 ; Colmeiro, Bot. Penins. Hispano-Lusitana, 199.
