448 WILSON EXPEDITION TO CHINA 



Involucra fructifera velutina et etiam setis pungentibus plus minusve dense 

 vestita (tantura in var. miti setis deficientibus), apice in tubum angustiorem 

 plus minusve longum contracta 8. C. Sieboldiana. 



ENUMERATIO SECTIONUM ET SPECIERUM ASIAE ORIENTALIS 

 HIMALAYAEQUE. 



Sect. I. ACANTHOCHLAMYS Spach in Ann. Sci. Nat. s^r. 2, XVI. 108 

 (1841). — A. De Candolle, Prodr. XVI. pt. 2, 129 (1864). — Mouillefert, Traite Arb. 

 &Arbriss. 11. 1143 (1897). 



The main character of this section is found in the spinose segments of the fruiting 

 bracts, the fruits resembling those of a Castanea. In the male flowers the two thecae 

 of the anthers seem to be mostly somewhat united at the middle in C. ferox Wallich, 

 while in C. tibetica Batalin the thecae are free, and only the filaments are more or 

 less united, although they are sometimes divided to the base. The color of the an- 

 thers seems to be almost purple, the filaments being whitish. In C. ferox Wallich 

 the anthers are more hairy at their apex than in the other species. 



1. Corylus ferox WalHch, PI. As. Rar. I. 77, t. 87 (1830). — Loudon, Arb. Brit. 

 111. 2030, fig. 1950 (1838). — A. De Candolle, Prodr. XVI.pt. 2, 129 (1864).— 

 Brandis, Fore.st Fl. Brit. Ind. 494 (1874); Ind. Trees, 624 (1906), exclud. var. 

 thibetica. — Hooker f., Fl. Brit. Ind. V. 625 (1888). — Mouillefert, Traite Arb. & 

 Arbriss. II. 1143 (1897). — Gamble, il/an. Ind. Timb. &S4: (1902). — Schneider, 

 III. Handb. Laubholzk. 1. 144, in adnot. (1904); II. 897, fig. 560 b, 561 c (1912). — 

 Winkler in Engler, Pflanzenr. IV.-61, 44, fig. 13 (1904), exclud. var. — Rehder in 

 Bailey, Stand. Cycl. Hort. II. 859 (1904), exclud. var. 



INDIA. Nepal : " ad cacumen Shedpore montis," 1821, A''. Wallich (No. 2797, 

 type, ex Wallich; tree 7 m. tall). Sikkim : "reg. temp., alt. 8-10,000 ped.," J. D. 

 Hooker (flowers and young fruits; mixed with a winter branchlet of C. Jacque- 

 montii in Herb. Gray). 



This is a well marked species of which I have seen so little material that I cannot 

 point out clearly its differences from the following nearly related species which 

 inhabits a different geographical area, neither of the two species apparently being 

 found in the region from I"Qiasia to southern Yunnan. 



2. Corylus tibetica Batalin. See p. 443. 



Sect. II. AVELLANA Spach in Ann. Sci. Nat. s^r. 2, XVI. 101 (1841), includ. 

 sect. Tubo-avellana. — A. De Candolle, Prodr. XVI. pt. 2, 129 (1864). 



This section includes all the other species now kno'mi. The shape and ap- 

 pearance of the involucres of the fruits of the different species are very variable, 

 but the segments or lobes are never spiny, although there is a group of species the 

 involucres of which are beset with bristly setose hairs. Spach described besides 

 sect. Acanthochlamys two other sections, but his Sect. Tubo-avellana cannot be 

 accepted as of the same taxonomic value as sect. Avellana. A. De Candolle has 

 already united these sections. He made two groups (§ 1 and § 2) under sect. 

 Avellana, separating the species with setose involucres from the others, but he 

 did not separate C. Colurna and its allies which, I believe, form a distinct sub- 

 section, while the other species may be united in another subsection. It needs 

 further investigation to decide if the setose species (C. rostrata Alton and its 

 allies) form a natural group of equal taxonomic value as C. Avellana Linnaeus and 

 its alUes or if they are connected with them by intermediate forms. 



