HYLOCEREUS. 189 



Collected by Brother Leon on a wall, Jata Hills, near Guanabacoa, Province of Habana, 

 Cuba, July 14, 1913 (No. 3719). Living specimens were introduced into the New York 

 Botanical Garden which flowered in September 19 17. We are disposed to refer here J. A. 

 Shafer's No. 13931 from lime rocks at Portales, Province of Pinar del Rio, Cuba. A plant 

 from the Isle of Pines sent to us by 0. E. Jennings probably belongs here, but the poor 

 specimen which we have seen does not enable us to definitely refer it to this species. 



Figure 265 shows a section of a branch of the type specimen. 



9. Hylocereus lemairei (Hooker) Britton and Rose, Contr. U. S. Nat. Herb. 12: 42S. 1909. 



Cereus lemairei* Hooker in Curtis's Bot. Mag. 80: pi. 4814. 1854. 

 Cereus trinitatensis Lemaire and Herment, Rev. Ilort. IV. 8:642. 1859. 



A somewhat slender, high-climbing vine; joints 3-angled, freely rooting on one side, 2 to 3 cm. 

 in diameter, plain green; margins with slight elevations at the areoles; areoles 2 to 2.5 cm. apart; 

 spines usually 2, very short, swollen at base, brownish; flower-buds elongated, acuminate; flower 

 about 27 em. long; tube, including ovary, 15 cm. long; scales on ovary and lower part of the tube 

 ovate, dark green, with the margins and tips deep purple; scales on upper part of the tube much 

 elongated, but marked like the lower ones; outer perianth-segments about 20, 12 cm. long, 1 cm. 

 wide or less; edges slightly upturned, widely spreading or reflexed, yellowish green, sometimes a 

 little purplish at the tip and the inner one somewhat rose-colored at the base; inner perianth-seg- 

 ments about 15, mostly oblanceolate, 3.5 cm. broad at the widest portion, acute, the lower portion 

 pinkish, above nearly pure white; filaments cream-colored, about three-fourths the length of the 

 inner perianth-segments; style thick, nearly as long as the inner perianth-segments; stigma-lobes 

 cleft to the middle and the branches often notched at tip; flower somewhat odorous, not very 

 pleasing; fruit purple, oblong, 6 to 7 cm. long, when mature splitting down the center almost to 

 the base into 2 nearly equal parts and exposing the white flesh and black seeds. 



Type locality: Not cited. 



Distribution: Trinidad and Tobago. Perhaps also Surinam. 



The above description was based upon specimens sent by Mr. Wm. Broadway in 1907 

 from Trinidad, which flowered in the New York Botanical Garden in July and August 

 1912 (No. 27689). Our reference (Contr. U. S. Nat. Herb. 12:428) of this species to 

 Antigua and doubtfully to Culebra and Porto Rico, in which we followed previous authors, 

 can not be supported by specimens in our collections. 



This is a very beautiful species which has long been in cultiva- 

 tion, but the native home of which, until recently, has not been 

 known. In 1909, Mr. Broadway sent specimens from Trinidad 

 which soon flowered, enabling us to identify it definitely. Sir 

 Joseph Hooker, under Cereus lemairei in Curtis's Botanical Mag- 

 azine, volume 80, plate 4814, says, "Nothing is positively known 

 of its native country; but it happens that I have in my possession 

 a drawing made in Antigua, undoubtedly of this species ; so that 

 it is probably a native of that island." A copy of this drawing is 

 now 7 in the United States National Herbarium, and shows quite a 

 different species from Cereus lemairei, and may represent the 

 Hylocereus collected in the spring of 1913 on Antigua by Dr. Rose F % 2 ten^i gma x 1 o. b 7 eS ° f 

 (No. 3297), of which we have both herbarium andliving specimens, 



but the drawing is without stem and Dr. Rose's specimens were without flowers; however, 

 it may be that Hooker's drawing is of a flower of the commonly cultivated H. undatiis. 



This is one of the few species of cacti having bifid stigma-lobes. 



Illustration: Curtis's Bot. Mag. 80: pi. 4814, as Cereus lemairei. 



Plate xxxi is from Mr. Broadway's Trinidad plant which flowered in the New York 

 Botanical Garden. Figure 266 shows its style and stigma-lobes. 



*Cereus lemoinei (Mollers Deutsche Gart. Zeit. 6: 92. 1891) may be only a misspelling of this name. 



