HYLOCEREUS. 191 



12. Hylocereus extensus (Salm-Dyck). 



Cereus extensus Salm-Dyck in De Candolle, Prodr. 3: 469. 1828. 



Creeping and probably often climbing, bearing the usual aerial roots of the genus; joints green, 

 rather slender, 1.5 cm. in diameter, 3-sided, the obtuse angles not at all winged; areoles remote, 

 small, woolly and often setose; spines 2 or 3, rarely 4, very short and stout, dark brown, i to 2 mm. 

 long; flowers large and handsome; tube green, cylindric; scales of the ovary ovate; scales of the tube 

 rather short, becoming elongated above and passing into the narrow outer perianth-segments, 

 greenish yellow, tipped and margined with red; inner perianth-segments oblong to obovate, acute, 

 rose-red; style thick, longer than the stamens; stigma-lobes linear, entire; fruit not known. 



Type locality: Not cited. 



Distribution: Trinidad, according to Curtis's Botanical Magazine. 



The above description is based on the figure and description found in Curtis's Botani- 

 cal Magazine as below cited. This may or may not belong to the plant described by De 

 Candolle (Prodr. 3: 469), for he describes the radial spines as 10 to 12, pilose and white, 

 and the centrals as 2 to 4, small, rigid, and yellow; it is hardly the Cereus extensus of Pfeiffer 

 (Enum. Cact. 119), where the inner perianth-segments are said to be white and obtuse. 



Cereus subsquamatus Pfeiffer (Allg. Gartenz. 3: 380. 1835) is referred here by Pfeiffer. 



Illustration: Curtis's Bot. Mag. 70: pi. 4066, as Cereus extensus. 



13. Hylocereus napoleonis (Graham) Britton and Rose, Contr. U. S. Nat. Herb. 12: 429. 1909. 



Cereus napoleonis Graham in Curtis's. Bot. Mag. 63: pi. 3458. 1836. 



Stems much branched, light green, the joints with 3 acute angles and concave sides; angles 

 tuberculate, with repand intervals, not at all horny; areoles about 4 cm. apart; spines 4 or 5, rigid, 

 about 9 mm. long, with swollen bases; flowers 20 cm. long and nearly as broad; tube 7.5 cm. long, 

 green, bearing a few subappressed, deep red scales, gradually enlarging upward; outer perianth- 

 segments yellow, lanceolate, linear; inner perianth-segments pure white, spatulate-lanceolate, crenate 

 at apex; stamens numerous, yellow; pistil stout; stigma-lobes numerous, entire. 



Type locality: Unknown; described from a cultivated plant. 



Distribution: West Indies and southern Mexico, according to Schumann; but we 

 know it definitely only from the original illustration. 



The origin of this species has long been in doubt. It was described by Graham at the 

 time it flowered in the botanical garden at Edinburgh. The plant had then been in cultiva- 

 tion for about ten years, having been sent by a Mr. McKay of Clapton, but without any 

 record of its source. It is possible that this species should be referred to the true H. tri- 

 angularis, although Pfeiffer states in the most emphatic terms that they are very distinct. 

 According to Loudon (Card. Diet. 2: 65. 1827) Cactus napoleonis occurs in a list of new 

 plants offered by L. C. Noisette, a nurseryman in Paris. This was about nine years before 

 the name was published in the Botanical Magazine. 



Cereus triangularis major Salm-Dyck, Allg. Gartenz. 4:80. 1836 and Cactus napo- 

 leonis Hortus, unpublished names, are often given as synonyms. 



Cereus lanceanus (G. Don in Sweet, Hort. Brit. ed. 3. 285. 1839), C. inversus, and C. 

 schomburgkii are names of garden plants which are referred to this relationship by Forster 

 (Handb. Cact. 422. 1846). 



Plants from Santo Domingo resemble the original illustration in armament. We have 

 these in cultivation, both at Washington and at New York, but they have not flowered 

 (Rose, Nos. 3734, 3839, and 4147). Boldingh (Fl. Ned. West Ind. 296) records the plant 

 from Aruba. 



Pfeiffer (Enum. Cact. 117. 1837) referred here Burmann's plate of Plumier (pi. 199, f. 2) 

 which is perhaps the best disposal to make of it. The fruit, however, has spiny areoles and 

 in this respect resembles Acanihocereus pcntagoniis. Gosselin considered it an undescribed 

 species which he called Cereus pluniicri (Gosselin, Bull. Soc. Bot. France, 54: 668. 1907). 



Illustrations: Curtis's Bot. Mag. 63: pi. 3458; Loudon, Encycl. PL ed. 2. f. 17363, both 

 as Cereus napoleonis; (?) Plumier, PL Amer. ed. Burmann, pi. 200, f. i, as Cactus etc. 



