38 ROSACES. 



rotundatis ad basin 1^ nun. latis tnmcatis -forraibus pedicelliim anaplectan- 

 tibus margine glanduloso-ciliolatis, ciliolis apice obtusis ^ mm. longis ascen- 

 dentibus plus minus recurvis, bracteis rubescentibus ; pedicellis viridibus 7 mm. 

 longis totis subsequilongis hirsutis, pilis patentissimis. Calyx tubuloso-cam- 

 panulatus 8 mm. longus, tubo viridi 3 mm. longo sursum 2^ mm. lato subcylin- 

 drico vel sursum plus minus latiore intus extusque glabro, limbo late campa- 

 nulato cum lobis 5 mrn. longo ad 5-lobato, lobis triangulari-oblongis rube- 

 scentibus 3 mm. longis 2^ mm. latis apice acutis basi plus minus contractis 

 utroque pagine glabris, sinibus inter lobos angustissime obtusis. Petala subalba 

 vel leviter rosea obovato-oblonga oblonga vel ovato-oblonga vel rhomboideo- 

 oblonga 7mm. longa 4^-5 mm. lata apice ambitu obtusissima rotundata retusa 

 vel emarginata ad centrum apicis emarginati breve mucronata vel haud mu- 

 cronata basi subito cuneato-contracta vix vel haud unguiculata ad insertionem 

 J mm. lata glabra. Stamina ad partem limbi calycis haud lobatam circ. 

 1^ mm. longum multiseriatim inserta in longitudine variabilia, extimis longis- 

 simis 5-6 mm. longis cum petalis aequilongis vel quam petalis leviter breviori- 

 bus, filamentis glabris, antheris sub rotundatis utrinque emarginatis minus 

 quam ^mm. longis. Ovarium ovoideum in fundo calycis situm l^mrn. longum 

 plus quam 1mm. latum apice ad stylum abeims glabrum intus (ventral i) 

 tenuissime 1-sulcatum, sulcis per stylum usque ad stigma abeuntibus, stylo 

 8-9 mm. longo columnari-filiformi mm. crasso, stiginate capitato-infundibuli- 

 formi 1 mm. in diarnetro latere uno fissio ^ mm. alto. 



Very closely allied to, if not identical with, the Japanese species Prunus 

 Yamasalcura (a wild mountain-cherry). I do not think our plant is ever 

 distinct specificially from the named species. The only difference between them 

 is that the stipules of our plant are much narrower and serration of the leaves 

 is a little looser, than those and that of the Japanese. It may constitute a 

 subspecies or a variety of the latter. I am quite unwilling to treat our plant 

 as a species distinct from the Japanese cherry; yet, the latter is so veiy con- 

 fusing as to its proper name that it is at present as equally subject to altera- 

 tion to place the Formosan plant under one name as under others. 



HAB. Mt. Arisan : Tozan, ad 7500 ped. alt., leg. B. KANEHIRA, I. TANAKA, 

 et B. HAYATA, Aprili. 1914. 



