FLORA OP" tup: kurile islands. 229 



Georgi, ill his " Geograpliiscli-physikalische und naturhistorisehe Besehreibung des 

 Russischen Reichs," iir, 4, p. 102?), mentions the occarrence of Rahus frutlcosus, L., in 

 the Kurile Islands. It is most probable that he mistook some other black-berried species 

 for the phiiit, f(jr we know that Hub/is frutlcosus has thus far been found in a wild state 

 only in Europe and in western Asia as far east as the western temperate Himalaya. 



Growing in Yezo and in the alpine districts of the main island of Japan there is a 

 species of Ruhus with a fruit which is dark-purple when ripe. In its general characters 

 it approaches very closely Ruhus occidentalis of !N^orth America. It forms a shrubby 

 bush, with somewhat recurved steins, which are generally destitute of prickles and glau- 

 cous. Young branches and petioles are villosely tomentose, armed with more or less 

 recurved prickles, and often covered with glandular hairs, which are especially promi- 

 nent on peduncles and pedicels. The leav^es are pinnately 3-foliolate ; the leaflets broadly 

 ovate, coarsely doul)ly-serrate, canescently tomentose l)eneath, the terminal sometimes 

 subcordate at base, and the lateral distinctly short-petioled. The midribs ai-e armed be- 

 neath with a few recurved prickles, and the stipules are setaceous. Axillary peduncles 

 1-3-fiowered, the terminal genei'ally 8-10-flowered in a close rounded corymb; pedicels 

 once to twice the length of the cal)x lobes, and both calyx and pedicel destitute of 

 ])rickles. The fruit is roundish, glaucous, and dark-])ur[)lc when ripe. A partly ripened 

 fruit when dried hardens up close to the stones, showing the lacunose markings on the 

 surfiice. 



From the typical form of Ruhus occidentalis it is distinguished chiefly by its younger 

 branches being villosel}^ tomentose intermixed often with glanduhir hairs, and by its jied- 

 icels being destitute of prickles. The fruits are somewhat smallci' and much more closely 

 clustered. In other respects there are striking resemblances between these two plants. 



From Ruhus Idceus, var. strigosus it is easily distinguished by the petioled lateral leaf- 

 lets, by the absence of prickles on the calyx, by the shape of the inflorescence, and above 

 all ])y the color of the rij^ened fruits. 



This plant has generally been taken in Japan to be a variety of Ruhus Idceus, doubt- 

 less from the fact that its ripened fruits were not known to collectors. From the de- 

 scriptions given by Franchet & Savatier of their R. Idceus, var. exslcca, there is no doubt 

 that their plant is the same as ours. Our plant, however, is very closely related to R. 

 occlclentalis and should not be placed under R. Idceus. As the former is a variable plant 

 and is very widely dispersed through the continent of ]N"orth America,' I prefer to con- 

 sider our plant as its variety I'ather than as a distinct new species : — 



82. Rubus occidentalis var. japonicus, L. Sp. p. 493; Torr. & Gray, Fl. X. Am. i, p. 453. 



R. Idceus, var. exsicca, Fr. & Sav. Enum. ii, p. 334; Matsumura, Cat. PI. Herb. 



Imp. Univ. Tokyo, p. 57. 

 Hah. In the alpine woods of middle Jajyan: Komagatake, Prov. Shinano ( Yatabe) ; 

 Nikko (Savatier) ; " N'iiihon media" (Tschonoski!) ; Yczo: in rich alluvial woods, Sapporo 

 (K. M., J uly, 1880, fl.) ; in cleared forests, Samani Sando, Prov. Hidaka (K. M., Aug. 19, 

 1884, fr.) ; Kurile Islctnds ? (Geovgi) . 



83. Geum strictum, Ait. Iloi't. Kew. n, p. 217; Ledeb. Fl. Ross, n, p. 22; Torr. & Gray, 



' Koclu', l!at<);j;niiiliisclii' Abliaii.l. in Naliirw. \'i'i-fiii zu IJrrnii'ji iv, |i|;. 117 anil 105. 



