73 



it has been passed without a scientific name. Pyrus gracilis Sieb. et 

 Zucc. Flor. Jap. Fam. Nat. n. 72 (=Sorbus gracilis C. Koch in Ann. 

 Mus. Bot. Lugd.-Batav. I. p. 249 ; Pyrus sambuci/olia Miq. Prol. FI. 

 Jap. p. 229, non Cham, et Schlecht.— Nom. Jap. Koba-nanakamado , 

 Nankin-nanakamado.) is clearly to distinguish from my plant by the 

 apparently 3-celled ovary and the presence of the peculiar stipules ; 

 Pyrus aucuparia Gasrtn. var. japonica Maxim ex Franch. et Sav. Enum. 

 PI. Jap. I. p. 140 (=Sorbus aucuparia Linn. var. japonica Maxim, in 

 Mel. Biol. IX. p. 170 ; Pyrus americana var. microcarpa Miq. loc. cit. 

 p. 229, non DO. nec Torr. et Gray; Sorbus americana 0. micrantha 

 Koch loc. cit. non Willd. nec. Dura- Nom. Jap. Miyama-nanakamado, 

 Yama-enzyu, Fudiki, Oyama-zansho, Nanakamado.) differs from my 

 plant by having the 3-4-celled ovary, the evidently acuminate and 

 subconcolor leaflets, and the ample and multiflorous corymbs; and 

 Pyrus sambuci/olia Cham, et Schlecht, in Linmea II. p. 36 (=8orbus 

 sambuci/olia Maxim, loc. cit. p. 172.— Nom. Jap. Takane-nanakamado.) 

 by having the concolor and thinly villose leaflets, and the larger and 

 less flowered corymbs, though it has 5-celled ovaries as in my plant. 



. Pyrus micrantha Francli. et Sav. loc; cit. II. p. 351 is probably 

 mere form of the Pyrus aucuparia var. japonica Maxim. ; because we 

 have in the Herbarium of the Imperial University of Tokyo a specimen 

 from cultivation in the Botanic Garden of the University which is not 

 else than a form of Pyrus aucuparia var. japonica Maxim., and at the 

 same time exactly coincides with Frauchet and Savatier's description 

 of Pyrus micrantha. 



Leaves of Pyrus aucuparia var. japonica Maxim., are not always 

 glabrous as described by Maximowiez, hut often rufo-villose. 



{To be continued.) 



