1925] WILSON, THE RHODODENDRONS OF EASTERN CHINA 165 



rounded, emarginate; stamens 10, of unequal length, shorter than corolla, 



filaments flattened and puberulous at base; pistil overtopping stamens, 



included, ovary oblong-ovoid, about 0.6 cm. long, furrowed, clothed 



with short red-brow r n hispid pubescence; style curved, glabrous or nearly 



so, stigma capitate. Fruit cylindric, curved, 1.5-2.5 cm. long, 0.5 cm. 



broad, furrowed, glabrescent or sparsely pubescent, slightly verruculose; 



seed ellipsoid, shining pale brown, about 2 mm. long, with minute wing 



at apex and base. 



Habitat: Formosa, forests on central rang between 2100 and 3200 m. altitude. 



This is the common Rhododendron of the Formosan forests above 

 the elevation of 6500 ft. when it grows to a large size. Very often it is 

 a wide-spreading shrub from 20 to 25 ft. high, occasionally as much as 

 30 ft., with a trunk 3 ft. in girth. Around Arisan and on the ranges 

 beyond it is plentiful but clings to the edge of the forests of mixed 

 broad-leaf trees and Conifers. I did not see it in bloom but judging from 

 herbarium specimens it is much less showy than many other east Asiatic 

 species. The flowers are white, spotted with brown-purple, numerous, 

 sometimes as many as twenty, in a rounded truss but they are small 

 for such a large and vigorous plant. It is most closely related to R. 

 pachytrickum Franch. of western Szechuan which though similar in 

 appearance of flower and foliage has usually densely hairy shoots. 

 Hayata distinguishes his R. pachysanthum by its leaves being densely 

 tomentose below, less reticulate on the upper surface and its glabrous 

 style. I do not find these characters so marked neither are they constant, 



the tomentum on the under surface of the leaves is floccose and may dis- 

 appear partially or wholly and early or late as is the case with other 

 species having this type of pubescence. 



Rhododendron Morii was named by Hayata from material collected in 

 1908 on Mt. Randai by U. Mori but appears to have been first discovered 

 the year previous by T. Kawakami on Mt. Morrison and wrongly referred 

 to the Japanese R. brachycarpum (J. Don by Hayata. It has since been 

 found on many mountains of the central range. On October 31, 1!)18, I 

 gathered ripe seeds on the summit of Arisan and sent them to the 

 Arnold Arboretum who distributed them among friends in Europe and 

 America. It is now growing in several gardens but I have not heard 

 that it has flowered in cultivation. 



6. Rhododendron formosanum Ilemsley in Kew Bull. Misc. Inform. 

 1805, 183. — Matsumura & Hayata in Jour. Coll. Sci. Tokyo, xxn. 218 

 (Enum. PL Formos.) (1906). — Matsumura, Ind. PI. Jap. n. pt. 2, 45!) 

 (1912).— Hayata, Icon. PL Formos. III. 132 (1913); ix. 65 (1920).— 

 Kanehira, Formos. Trees, 319, fig. 3, 320 (1917). — Hutchinson in Millais, 

 Rhodod. 168 (1917). 



Bush 2-6 m. or tree 10 m. tall with trunk 1 m. girth, shoots rigid, 

 purple-brown, when young clothed with pale gray scurfy tomentum, 

 soon glabrescent. Leaves crowded, oblong-lanceolate to oblong-oblanceo- 



