Spicilegium Sal team Japonensium 

 novarum aut imperfecte cognitarum 



By 

 <x. Koidzimii 



The most important contribution recently made to our 

 knowledge on the Japanese Salices has been Von Seemen's 

 Monograph, in which, besides four imperfectly known and seven 

 doubtful ones, thirty-three species, with several varieties and 

 forms, are recognized as occurring in the Archipelago. After 

 that, Leveilee and Vaniot have described three from the 

 collections of Faurie and Kinashi ; but their Salix Kinashii 

 is identical with Salix triandra. My Salix hondoensis is a 

 distinct species having the characteristic leaves and the elong- 

 ated style ; I think it is convenient to put this, with Salix 

 koreensis Anders., into the new section of Heteradenias-Diandras- 

 Dolichostylse. However, the peculiar colouring of hair of the 

 catkin in Salix rulpina, is not so important a character as to 

 establish a special section, so I reject the section Vulpinaas of 

 Seemen, including it in the section Hastatas. It seems to me 

 that, Salix Sieholdiana, S. Buergeriana, S. Saidaiana, and 5. 

 Harmsiana are but forms or varieties of one species, and at 

 the same time, Seemen's Salix Matsutnuraei and 5. daisen- 

 iensis had better be considered as varieties of the S. vulpina. 

 In this paper three additional species have been descrided ; thus 

 bringing the number of species of distinctly known Salices to 

 thirity-two. 



1. Salix (Pentandrae) glandulosa Seem, in Engl. Bot. Jah- 

 rb. XXI. (1S96), Beibl. LIII. 55 ; Sal. Jap. (1903) 22 ;— Levl. 

 et Vnt. in Bull. Akad. Int. Geogr. Bot. XIII. (1904), 208 ; 

 XVI. (1906) 144, 148 ;— Diels in Engl. Bot. Jahrb. XXIX. 

 276 .;— Nakai Fl. Kor. II. 214 ;-Shirai in Tok. Bot. Mag. XVII. 

 (1903) 223, t. 4. 



