136 WILSON EXPEDITION TO CHINA 



Salix cyclophylla Seemen in Bot. Jahrb. XXX. Beibl. LXVIl. (non Rydberg 

 1899) (1902); Salix. Jap. 69, t. 16, fig. e-f (1903). — Leveille in Bull. 

 Acad. Int. Geogr. Bot. IV. 210 (1904). 



JAPAN. Hokkaido: summit of Mt. Riishiri, July 25, 1899, U. Faurie (No. 

 3713; 9 type of S. cyclophylla ex von Seemen); same locality, "repens in caveis 

 ubi nix fusafuerit," June 27, 1891, U. Faurie (No. 3713; 9 , ex Leveille). Hondo : 

 prov. Shinano, on Yatsugatake, very rare, alt. 2800 m., August 1, 1910 (ex Herb. 

 Sakurai; 9 ); same locality, August 1907 (ex Herb. Yokohama Nursery Company; 

 9); same prov., on Siroumatake, August 8, 1908 (ex Herb. Tokyo sub nom. 

 S. arctica); Dailenyezan, August 1912, Nakamura (types of S. Nakamurana ex 

 Koidzumi). 



The leaves of this species are rather variable in shape, but there is otherwise no 

 real difference between Koidzumi's species and von Seemen's plant according to 

 Matsumura's plate, and I cannot separate these two forms. Von Seemen's name 

 having been used earlier by Rydberg for another species must be changed. The 

 specimens before me have almost entire or not deeply cleft stigmas, and the pedi- 

 cel of the ovaries is more or less distinct, while in von Seemen's plate the stigmas are 

 represented as deeply cleft and the ovaries almost sessile. 



Sect. DIPLODICTYAE Schneider, n. nom. 



Salix, sect. Arcticae ^ Rydberg, in Bull. N. York Bot. Gard. I. 263 (pro parte) 

 (1899). 



Frutices prostrati ramuhs ramisque radicantibus v. nani ramulis plus minusve 

 erectis. Folia Integra, superne ut videtur stomatibus paucis instructa, subtus pal- 

 lida, satis firma, saepissime distincte petiolata (petiolis saepe M~M laminae longis). 

 Amenta apice ramulorum foliiferorum multiflora, densiflora; flores cf glandulis 

 duobus (an semper?) instruct!, glandula dorsali saepe minima, antheris (an semper?) 

 rubescentibus; flores 9 glandula una ventrali; ovaria subsessiUa v. brevipedicel- 

 lata, plus minusve pilosa v. tomentosa, stylis distinctis apice saepe bifidis, stig- 

 matibus bifidis elongatis (? v. brevibus). 



Rydberg includes in this section (with some doubt, I believe) S. polaris Wahlen- 

 burg, see p. 139, and also S. glauca Linnaeus. Von Seemen (in Ascherson & Graebner, 

 Syn. Mitteleur. Fl. IV. 162 [1909]) says that *S. arctica Pallas, S. Brownei Lundstrom 

 and S. ovalifolia Trautvetter belong to sect. Myrtosalix. It is difficult to classify 

 the species No. 103-106 and their relatives. It is unfortunate that Rydberg did 

 not apparently pay any attention to the glands of the cT and 9 flowers. 



103. Salix arctica Pallas, Fl. Ross. I. pt. 2, 86 (1788). — Georgi, Beschreib. Russ. 

 Reich, pt. III. vol. IV. 1339 (1800). — Lundstrom in Nov. Act. Soc. Sci. Upsal. 

 1877, 31, t., fig. 1, 1-3 (Weid. Now.Semljas) (1877). — Bebb in Bot. Gaz. XIV. 115 

 (pro parte) (1889). — Rydberg in Bull. N. York Bot. Gard. I. 265 (1899). — Schnei- 

 der, III. Handb. Laubholzk. I. 41, fig. 20n-o,26 d (1904). — Koidzumi in Tokyo Bot. 

 Mag. XXVII. 97 (1913). — See also Herder in Act. Hort. Petrop. XI. 438 (1891). 



Salix Pallasii, var. diplodictya Andersson in De Candolle, Prodr. XVI. pt. 2, 

 285 (pro parte) (1868). 



* The name Arcticae (Arctica) was first used by Andersson (in Ofvers. K. Vet.- 

 Akad. Fork. 1858, 119 and in Proc. Am. Acad. IV. 59 [Salic. Bar. -Am. 13] [1858]) 

 for a section containing S. Hookeriana Barrat, S. speciosa Hooker & Arnott, etc. 

 which Andersson (in De Candolle, Prodr. XVI. pt. 2, 274, 275 [1868]) referred to 

 sect. Niveae. 



