22 



Th^ CACTACEA^. 



no other material of this species has been collected since Humboldt's time except that 

 in 1912 Dr. Weberbauer wrote that he had visited the INIarafion, at Humboldt's locahty, 

 and had collected a single specimen, which had been sent to the Botanical Museum at Berlin. 



17. Pereskia cubensis Britton and Rose, Torreya 12: 13. 1912. 



A tree, 4 meters high, with a trunk 2.5 dm. in diameter and a large, flat, much branched top; 

 bark brownish, rather smooth, marked here and there by black bands (representing the old areoles) , 

 these broader than high; young branches slender, smooth, with Ught-brown bark; spines from young 

 areoles 2 or 3, needle-like, brownish, 2 to 4 cm. long, from old areoles ver\' numerous (25 or more), 

 and much longer (5 cm. or more long) ; leaves several at each areole, sessile, bright green on both 

 sides, oblanceolate to oblong or obovate, i to 4 cm. long, 10 to 12 mm. wide, acute at both ends or 

 obtuse at the apex, fleshy, the midvein broad, the lateral veins very obscmre; peduncle very short, 

 jointed near the base, bearing i to 3 leaf-like bracts; flowers terminal and also axillar^^ solitary; sepals 

 5, obtuse or rounded, ovate-oblong to orbicular, unequal, 7 to 9 mm. long, the larger ones with broad 

 purple margins; petals 8, about 15 mm. long, deep reddish purple, obovate-elliptic, roimded; stamens 

 manv, about 6 mm. long; anthers light yellow; ovarj^ turbinate, naked, spineless; fruit not seen. 



Fig. iS. — Pereskia cubensis. 



-Pereskia cubensis. 

 X0.5. 



Type locality: In Cuba. 



Distribution: Near the southern coast of eastern and central Cuba. 



The tree is abundant on the plain between Guantanamo and Caimanera, Oriente, 

 where the type specimens were collected ; it also inhabits coastal thickets at Ensenada de 

 Mora, in southwestern Oriente, the plants of this colony bearing leaves with less acute 

 apices than those of the typical ones. A single plant was also observed on La Vigia Hill, 

 at Trinidad, province of Santa Clara, which had shorter and smaller leaves than either of 

 the other two. The description of the flower is from one of a plant collected by N. L. 

 Britton and J. F. Cowell at Ensenada de Mora, southern Oriente, Cuba, in 1912, and 

 brought to the New York Botanical Garden, where it flowered in IMay 191 7. 



Dr. Rose finds that the plant in De Candolle's herbarium which represents the Pereskia 

 portiilacifolia of the Prodromus is undoubtedly Pereskia cubensis. It was collected as 

 early as 1821. 



Illustration: Journ. N. Y. Bot. Gard. 10: f. 22. 



Figure 18 is from a photograph taken by Dr. M. A. Howe in the colony of this tree 

 at NuevaUches, near Guantanamo, Cuba, studied by Dr. N. L. Britton in 1909; figure 19 

 represents a leafy branch of the same plant. 



