Pomona College Journal op Economic Botany 



373 



middle ; its wall is very thin, 0.5-1 mm. thick ; it contains only one seed, which 

 is globular, slightly flattened, very superficially lobed, 10 mm. in diameter, 

 7.5 mm. thick. Albumen not very hard, white, slightly oily ; embryo basal. 



I have based my description chiefly on the specimens of the "Plantae 

 Cubenses Wrightianje ", n. 1465 in Herb, de Candolk. In the Berlin Hierba- 

 rium I have also seen a fruiting specimen collected by Gundlach in Oriental 

 Cuba, one from the New York botanical Garden collected by Shafer (n. 3555) 

 on Sierra Nipe near Woodfred, Oriente, Pinelands, 500-650 metres altitude, 

 and another (Shafer n. 1750) from the Pinales, S. E. of Paso Estancia, Oriente. 

 Very slightly differing from the above is one I have received from Prof. Baker 

 (n. 3051) collected by W. T. Horne at Baracoa, Prov. Santiago d€ Cuba. In 

 this specimen the fruiting perianth is slightly smaller than in Wright's type 

 specimens (6 mm. in diameter) and is not split into three parts, or is at most, 

 only superficially three-lobed; the leaflets bear only a few bristles here and 

 there near the upper end, and on the mid-costa the spiculae are present only 

 near the base, and a few scattered near the apex; the upper leaflets shorten 

 gradually and have the apex briefly and almost equally obtusely two-toothed. 

 A very young plant collected by Eggers in Cuba (n. 5239) at La Prenda (800 

 m.) has primordial leaves which are deeply bilobed, acutely cuneate at the 

 base, with the lobes acuminate, and 6-7-costulate ; the petiole and the rachis 

 are powerfully armed with unequal needle-like, black, nitent spiculae, of which 

 the largest are 2-3 cm. long; all the other nerves are smooth; the margins are 

 irregularly spinulous. This palm is known in Cuba under the name of * ' Coco- 

 Macaco ' '. 



I have not seen specimens of this palm from any other of the Antilles. I 

 think that Grisebach had correctly identified the Bactris growing in Cuba 

 with the B. phimeriana of Martius, who applied this name to the Palnva 

 dactylifera aculeata minima represented in Plumier's unpublished plates 

 XLIII, XLIV, XLV. Of these plates I have seen a copy in the Berlin Her- 

 barium and about them I have to observe that the spadix represented in plate 

 XLIV bears fruits 10-11 mm. in diameter, or slightly smaller than those of 

 the plant from Cuba. 



Other species of Bactris are mentioned as proper to the Antilles, but I 

 have not as yet made a special study of them. 



SABAL Adans. 



Adans. Fam. nat., II, 495; Becc. in Webbia di U. Martelli, II (1907), 

 10. Sabal and Inodes O. F. Cook in Bull. Torrey Bot. Club, 1901, 529. 



A. All the primary divisions or segments of the leaves unieostulate and 

 deeply bifid. 



Divisions of the segments having a very acuminate, but rigid apex. 

 Flowers oblong-ovoid, very small, 3 mm. long, 1.8 mm. thick. Fruit 

 spherical, rounded at its base, 1 em. in diameter. S. parviflora 



Divisions of the segments having a very elongate filiform flaccid 

 apex. Full-grown flower-buds narrow and elongate, 5 mm. long, 2 mm. 

 thick. Fruit 8. fiorida 



