456 CONTRIBUTIONS FROM THE NATIONAL HERBARIUM. 



rather slender, their free part 3 to 4 mm. long; anthers rounded, their cells distinctly 

 bilobatej di.sk conical, muricate, about 3 mm. in diameter and 2.3 mm. high. Female 

 flowers not known. 



Fruit globose, about 5 cm. in diameter; skin verrucose, thick, lime-yellow outside, 

 exuding when sectioned a sulphur-yellow, resinous latex. Seeds 1 to 3, ovate, about 

 2 cm. long, wrapped in a white, mellow, sweetish-aeidulate, edible pulp. 



Colombia; La Manuelita near Palmira, eastern side of Cauca Valley, alt. about 



1,200 meters; male flowers and fruits, December 30, 1905, Pittier 843. 



Explanation of plates 92, 93.— PL 92, tree; pi. 93, fruits, one opened, and twig with terminal bud 

 and leaves. From photographs taken in the field by Pittier & Doyle. PL 93, natural size. 



Rheedia madru.no ovata Fittier, subsp. nov. Plate 94. 



Leaves mostly obovate-elliptic, long-cuneate at base, shortly acuminate or rounded 

 at tip; petioles longer than in type. Flowers unknown. Fruit ovoid or conical at 

 both ends, 6 cm. long, 4.5 cm. in diameter, with a long (2.5 cm.) pedicel. Seeds 3 

 cm. long. 



Type in the IT. S. National Herbarium, no. 531108, collected at La Manuelita near 

 Palmira, eastern side of Cauca Valley, Colombia, altitude 1,100 to 1,302 meters in 

 1905-190G, by H. Pittier (no. 916) . The type sheet shows only leafy stems. A second 

 sheet bears 3 photographs, one of them here reproduced. 



Explanation of Plate 94.— Twig with leaves, a fruit, another fruit dissected, and a seed. Photo- 

 graph by Pittier & Doyle. 



Rheedia madruno bituberculata Pittier, subsp. nov. Plate 95. 



Fruits found on the market at Cali (Cauca. Colombia). They have a basal aud an 

 apical nipple and are 7 cm. long and 5.5 cm. in diameter. The seeds are much longer 

 (4 cm.) than in the foregoing varieties. 



Type represented by a photograph in the U. S. National Herbarium, showing 4 

 fruits, 2 of them dissected. 



Explanation of Plate 95.— From the type photograph by Pittier & Doyle. 



In view of the several deficiencies in the descriptions of the Andean species of 

 Rheedia, I deem it preferable, notwithstanding the opinion of Vesque, to retain for 

 the present the specific division established by Planchon and Triana in the memoir 

 cited above. The identity of Rheedia acuminata and R. madruno of these authors is 

 by no means elucidated by Vesque's explanations, and the diversity in the shape' 

 and dimensions of the fruits, as observed by me in the Cauca Valley, points to the 

 coexistence there of several species of the group Verticillaria. 



The madronos of the Cauca Valley are during their season a staple article of con- 

 sumption among all classes, and are brought to the markets in large quantities. Their 

 taste is rather agreeable, but, on account of the pulp firmly adhering to the large 

 seeds, they feel rather awkward to the mouth. 



Symphonia globulifera L. f. Suppl. PL 302. 1781. 



Vesque a notes that he has not been able to observe a single fruit of this tree, 

 apparently widely spread in the tierra caliente of tropical America and well repre- 

 sented in most herbaria. He adds that Planchon and Triana have attributed to that 

 species a hairy seed, an assertion that is contradicted both by the testimony of the 

 original description of Linnaeus the son and by that of Bentham and Hooker.6 I 

 have had several opportunities of studying this Symphonia in the forests of Costa 

 Rica, where the Indians call it cerillo and tap it for the waxlike resin, which they use 

 for lighting purposes. The fruits, as recorded in my notebooks, are ovate or ovoid- 

 acuminate, 1.5 to 2 cm. long, 1 to 1.2 cm. in diameter, with a brown skin and a thin 

 mesocarp filled with laticiferous ducts; they bear at the tip the persistent stigmas 

 and the remnants of the stamens, and are almost without exception monospermous 



a In DC. Mono^r. Phan. 8 : 22S. *> Gen. PI. 1 : 173. 



