DILLENIACEAE. — ACTINIDIA 379 



floribus masculis sepala ovata, obtusa, 4-5 mm. longa, ciliolata, dorso 

 glabra v. rarius laxe puberula; petala ovalia v. oblongo-obovata, 7-7.5 

 mm. longe et 4-5 mm. lata, margine integra v. eroso-denticulata; sta- 

 mina dimidia petala aequantia, antheris oblongis 2 mm. longis loculis 

 basi divergentibus nigrescentibus; rudimentum pistilli staminibus plus- 

 quam duplo brevius, cylindricum, basi subito in discum dilatatum, apice 

 irregulariter laceratum; in floribus fertilibus sepala ovata, obtusa v, 

 acutiuscula, inaequalia, 5-7 cm. longa, fere glabra; petala late ovata v. 

 ovalia, concava, valde inaequalia, 8-12 mm. longa et 4.5-7 nmi. lata; 

 stamina ovario breviora; ovarium conico-eylindricum, apice attenua- 

 tum, glabrum, circiter 6 mm. longum, stylis radiatis circiter 15, stig- 

 mate subcapitato. Bacca ovoidea v. oblonga, glabra, apice rostrata, 

 2-2.5 cm. longa, purpurea, sapore grato; semina late ovoidea, 2-2.5 

 mm. longa, foveolata, flavo-brunnea. 



Western Szech'uan: southeast of Tachien-lu, thickets, alt. 

 1800-2000 m., June and October 1908 (No. 13 14; staminate flowers 

 and fruits, type); Mupin, thickets, alt. 1500-1800 m., June 1908 (No. 

 1314'*, fertile flowers); Yung-ching Hsien, Wa-wu-shan, alt. 1800-2250 

 m., September 14, 1908 (No. 893; fruits); without precise locality, 

 July 1903 (Veitch Exped. No. 3269; staminate flowers). Yunnan: 

 Mengtsze, alt. 2100 m., A. Henry (No. 1108). Western Hupeh: 

 without precise locality, July 1901 (Veitch Exped. No. 1165; staminate 

 flowers); without precise locality, A. Henry (No. 5622). Kiangsi: 

 Kuling, thickets, abundant, July 31, 1907 (No. 1512). 



This species is closely related to A. arguta Miquel ' which is chiefly distinguished 

 by the broader leaves lustrous above and more strongly serrate with spreading 

 teeth, by the larger flowers and by the greenish yellow subglobose fruit. Actinidia 

 purpurea Rehder is also closely related to A. melanandra Franchet which differs 

 in the glaucous under side of the leaves, in the short petioles, in the few-flowered 

 inflorescence and in the globose-ovoid darker fruit. 



Wilson's Nos. 893 and 1165 and Henry's No. 5622 differ from the type in having 

 the petioles and the veins or at least the midrib on the under side of the leaves more 

 or less tomentose, and Henry's specimen, moreover, has smaller and broader 

 leaves partly subcordate at the base while the specimen from Kiangsi which con- 

 sists only of leafy branches has the petioles and the midrib setose beneath. Henry's 



1 According to the International Rules A . arguta Miquel is the correct name of 

 the species, while Dunn follows Komarov and Finet & Gagnepain in using A . rufa, 

 because Trochosligmn rufa i)recedes T. arguta in the original enumeration. 

 The International Rules, however, do not recognize the so-called priority of posi- 

 tion, but rule that, if the names of the groups to be united are of the same date, 

 the author chooses and his choice cannot be modified (art. 46). In this case 

 Maximowicz in 1886 was the first to unite the two species, and he took up the name 

 A. arguta for the aggregate and consequently this name must be used for the species. 



