~Vol. XVI ， No. 6. Jnne 1940

示直牛勿~子安芯奈佐言志 rTHE rTHE JOURNAL OF JAPANESE BOTANY

第十六巻第六披 (通巻第百六十六競) 昭和十五年六月後行

Oritical Oritical 110tes 011 some ]lormosan Euphorbiaceae (1)

By

Leon CROIZ .A T and Hiroshi H .A R .A

クロイツアー・原 寛: 墓湾産たかとうだい科植物エ就テ〈其ー〉

A g 1' eat cleal of confusion has so fa 1' resultecl from the fact that botanic work in Chiila ， Inclo- ancl Formosa has been concluctecl along independent lines lines by. botanists who. have studied in particular any one of these 宣oras.

:Such :Such confusion has been particula 1' ly great in the case of the Euphorbiaceae ， whiσh in the 1' egion of the Yellow. Sea presents very cli 盤cult p 1' oblems o~

~classi :fi cation and of dist 1' ibution. It has been our pu 1' pose to b1' ing togethe 1'

~and to interp 1' et at least some of the Formosan species in function of thei 1' allies allies 0 1' p 1' esumed alli~s endemic on the Asiatic .mainland. Our work has been been macle possible by the cooperation of P 1' of. T. N .A K .A I) Director of the

Botanical Botanical Institute of the Faculty of Science of Tokyo Impe 1' ial Univetsity ， who has generously consented to loah the types of these species ， thus making it it possible to b 1' ing. together for a critical comparison the historic material of the the herbarium of Tokyo Imperial Unive 1' sity ancl the rich collections of the he 1' barium of the Arnold Arboretum. The junior author in addition wishes to acknowledge with deep appreciation the kind interest of Prof. E. D. .MERR ，!LL J A c1 ministrator of the Botanical Collections of Harvar c1 U niverstiy an c1 Prof.

M. L. FERN .A LD) Director of the Gray Herba 1' ium of Harvard University to whom he stan c1 s in c1 ebte c1 for the extension of the full facilities available in herbaria herbaria ancl libraries of the University. The origin of the specimens cite c1 is in c1 icated as follows: TK (Herbarium

⑪ f the Botanical Institute of the Faculty of Science of Tokyo Imperial -1- 314 314 植物研究雑誌、 第十六巻焦六競

University); University); AA (Herbarium of the Arno1d Arboretum of Harvard Univer~ sity) sity) ; GH (Gray Herbarium of Harvard University) ; NY (Herbarium of the' N ew Y ork Bota ;n ica1 Garden).

孟ntidesma pentandrum (BLANCO) MERRILL in Phi1ipp. J ourn. Sci. 1玄，. p. p. 462 (1914); Enum. Phi1ipp.- P 1. II ， p. 但416 (1923) 一， PAX &， 宜OF 町、胃'F 町百官亦YM山'[A 却NN in

ENG 低LER 九) P ι. 丑イ.-1'

M旺[A 却SAMUNE 叫) Short F l. Fo 凹rmos .p.118 (1936). Syn. Syn. C αnsjer αpent αndr αBLANCO) F 1. Fi1ipp. p. 73 (1837).

Antidesm αkotoe ηsis KANEHIRA) Formos. . p. 472 (1918); ed. rev. p. p. 329 (1936)-YAMAMoTo in Journ. Soc. Trop. Agric. VI ， p. 558. (1934)-M .A sAMUNE ，1. c. (1936).

Antides 慨 α仰向町lisep α初冊 HAYATA) Icon. P l. Formos. IX ， p. 98 (1920) ← Nom. Jap. Koto-y αm αhα dzu.

Specim. Specim. Exam. Formosa. Taisenseki ，Koshun (HAYATA ， Jun. 1912 in TK.); ， Bankinsing (HENRY) no. 176 in AA.); South Cape (HENRY) no. 91 Ci in AA.); Takao (HENRY) no. 780 ，114 壬& 1885 in AA.). Liukiu. Liukiu. Loo-Choo (WRIGHT) no. 283 ，185 生 in AA.). The Formosan material seen agrees well with that of the Philippine Islands; which is ident i: fied as A. pentandnlm by MERRILL and PAX & HOFFMANN. To our knowledge the type-specimen of Blanco's species is not extant.

Bridelia Balans ぉ TUTCHER in J ourn. Linn. 80c. XX 玄VII ，P ・66 (1905)- -GEHRMANN in ENGLER) Bot. J， ahrb. XLI ，l. Beib 95 ， p. 37 (1908)-JABLON- SZKY in ENGLER) Pf l.・ .reich IV-147-VIII ， Ht. 65 ， p. 72 (1915).

8y11. 8y11. Bridelia pachinensis HA Y.ム TA in MATSU~主URA et HAYATA) Enum ・ P l. ]lormos. ]lormos. p. 362 (1906) nom. nud. Brideli αKα w αkα mii HAYATA) 1. c. (1906) 110m. 11ucl.

Bridelia Bridelia ov α ta (non DECAISNE) HAYATA) Mater. F l. Formos. p. 263 ，

(1911) (1911) ; Gen. ，Ind. F 1. Form. p. 65 (1917)-8. 8UZUKI In Sy1via IV t

p. p. 135 (1933)-IuNEHIRA) Formos. Tree. ed. 1' ev. p. 333 ，fig. 28&;

(1936) ー-MASAMUNE) Sho 1' tF l. Formos. p. 118 (1936 ト Brideli α pubescens (no11 'KURZ) JABLONSZKY ，1. c. p. 73 (1915) quoaa specim. specim. Formosa.

ー- <) '_' CROIZAT CROIZAT and HARA: Critical notes on some Formosan Euphorbiaceae (1) 3お

Nom. Jap. M α'ru 抑制α-k αnko (HAYATA 1917) ，K αw αkα 例 i':'k αnko.

Specim. Specim. E 玄am. Formosa. Chosokei (N.4: KAHARA) no. 156 ，Ju l. 1905 in TK.); Shirin (TANAKA & SHIMADA ， no. 11058 & no. 11164 ， 1932 in AA.); Vicinity Vicinity of Taihoku (TANAKA) no. 101 ，Ju l. 1929 in AA.); sine loco speciali (OLDHAM) no. 498 ，1864 in GH.).

Liukiu. Liukiu. 1s. 1riomote (J. L. GR ，ESSITT) no. 565 ，Aug. 1934 in AA.). Dist. Dist. Liukiu ，Formosa ，China (Kwangtung ，Kwa 昭 si ，Haina 叫 ) and 1ndo-China. This This is most closely related to B. pubescens 'KURZ) which has leaves pubescent pubescent along the nerves beneath. The leaves of B. B αlαη sa are glabrous or or very minutely adpressed-hairy on the lower surface. Bt'ide1i α0倒的 DECAISNE is a quite distinct species which has depressed globular fruits with 2 locules and glabrous calyces.

We retain G1ochidion basing the generic name upon a concept which is purely purely that of 'a natural group うbecause we are satisfied that the so called “ generic characters" used to segregat.e this genus from Phyll α nthus are ephemera l. The 1ndian members of the genus are indeed an homogenous unit ，which differs from Phyll α nthus In ・"the total .a bsence of a disk ， in habit ， .a nd in the singular modification of its styles and stigmas" (HOOKER fi l. F l. Brit. Brit. 1ndia V ， p. 306 ，1887) ， but this is far from being the case in other regions ，e.g づ Australasia ， in which G1ochidion is. common ，a p. d in which GZochidion GZochidion probably originated. Discrete styles are tresent in G. sericeum) for for instance ， which. J. D. H o6K ER 五1. lists (op. cit づ p. 326) ，and controverts J. J. D. HOOKER'S very note that styles of the genus are “singularly modified" by comparison with Phυ 11 ω~thus. The habit of S αurop u' s L1anosi is so com- pletely pletely that of a that it has for a Iong もime been classified under

Glochidion) Glochidion) which obviously renders the habit. character untenab~e as generic.

The absence of a disk does not separate Glochidion from Phyll仰~thus) as J. J. D. HOOKER and others believe ， because rudimentary 'disk-glands occur in the 白rst three series of Phyll α nthus) sensu MUIι.:ARG ・) that. include the true Glochididn Glochididn (cf: DC. Prodr. XV-2 ， p. 275 ， 1866). The mistaken belief that Glochidion Glochidion can be tI・ eated otherwise than as a 'purely natural group' (i.e. ， as

-_ 3 ー 3M 植物研究雑誌 第十六巻第六君主

an aggregate which in the Iast analysis is 0 凶 standi 昭 merely hy inta 時 ibles) has has led to severaI gTave errors ，one of which directly concerns a Japanese plant ，Phyll ω'lJ thus fiexuos~~s (8IEB; 凶 Zucc.) MULL.-ARG. This species is treated treated by P .A X & HOFFMANN (Nat. P :tl・ -fam. 2 Auf l. 19 c， p. 58. 1931) as a Glochiclion ，-w hich is against every one of its natural a缶nities and involves a very confuse concept of the classification of the entire phyllanthoid alliance.

On the mistaken 部 sumptioh that the absence of a disk and. that undivides styles styles are the generic characters of Glochidion ，L 並ANDRI has. published a Phyll α nthus as G. Perrieri (Bull. 8oc. Bot. France LXX 玄1， p. 606 ， 193 .;1:) which needlessly confuses the range of all the natural groups in the vicinity of of Phyllanthus. We believe that the best ar 伊lment ever introduced for the maintenance maintenance of Glochidion as a distinct genus is that of HOOKER (op.' cit づ p. p. 306) ，who acknowledges his keeping the genus as distinct to “comply with the the wishes of 1ndian botanists ，whose opinion it is of importance to consult . in in regard to the nomenclature of so very large and universall y. distributed an 1ndian genus". This frank and authoritative statement of the uIt imate reasons reasons why a genus should be maintained or. rejected is accepted by us as the the besf answer to be given. when certain genera of the Euphorbiaceae are challenged challenged because some of their characters happen to be transitionaI with those those of other genera.

Glochidion Hayatae CROIZAT et 1王ARA ，nom. nov.

Hyn. Glochidion bicolot ，HAYATA ， Rev. E 叩 horbiac. Jap. p. 18 ，t. II ，E (1904) exc l. syn. ，quoa c1 specim. et descr.-MATsUMURA ， 1nd. P l. Jap. 11-2 ， p. p. 306 (1912)-HAYATA ， Gen. 1n c1. F l. Form. p. 66 (1917). Glochidion Glochidion hypoleucum (non BOERLAGE 1900) HAYATA ， 1con. P l. Formos. 1X ， p. 95 (1920)-YAMAMoTo in Journ. 8oc. Trop. Agric. VII1 ， p.

153 153 (1936)-KAN 阻 IR.A ，Formos. Tree ed. rev. p. 346 ，fig. 301 (1936) -MASAMUNE ，8hort ， F l. Formos. p. 120 (1936).

Nom. Jap. Ur αjiro-k αnkonoki (HAYATA 1904) ，T αkαsα go ・kek αη konoki ， Kob αno-k α侃 onoki. Specim. Specim. Exam. Formosa. Tamsuy (OLDHAM ， no. 499 ，1864 in AA.); in mo 凶 ibus Kelung (FAURIE ， no. 435 ，Mai 1903 in AA); Tikusiko ，Taihoku 一- 4 - CROIZAT CROIZAT and HARA: Critical notes on some Formosan Euphorbiaceae (1) 317

(8ASAKI) (8ASAKI) Oct. 1923 in AA.) ;' Giochi ， Holisha (HAYATA) Apr. 28 ，1916-type of of G. hypoleucum HAYATA in TK.); Pak-kang-khoe ，' Tsuisia ， Polisia (K. MIYAKE'; MIYAKE'; Jan. 21 ，1898 in TK.). Liukiu. Liukiu. Liukiu-honto (MATSUMURA) Mai 1897 in TK.).

Kyushu. Is. Amami-Oshima ， Mt. Uwan (8 ，. 8AITO) Nov. 1927); Inter Nishi- nakama et Higashinakama (S. SAITO) Nov. 1927). China. China. Kwangtung: Wat Shui Shan (W. Y. C :a: UN) no. 7401 ，Nov. 9， 1928 in AA.); Tung Koo Shan ，Tapu Dist. (W. T. TSANG) no. 21756 ， Sep. 1932. 1932. in AA.); Fan 8hiu Shan ，Wung Yuen Dist. (8. K. LAU; no. 2782 ，Nov. 1933 in AA.).

Glochidion dasyphyllum K. KOCH) Hort. Dendr. p. 85 (1853)-REHDER in in Journ. Arnold. Arb. VIII ， p. 30 (1926)-YAMAMoTo in Journ. Soc. Trop.

Agric. Agric. VIII ，p..1 路52 (1936) pro parte- 一一一一-一一白-KANE 回H王IR. 烏)A Fo ぽrn 乱I 宣勾 g. 297 (1936)-MAsAMuNE) Short. F 1. Formos. p. 120 (1936).

8yn. 8yn. Glochidion 伽 olle (non BLU]¥ '[E 1825) HOOKER et ARNOTT) Bot. Beechey Voy. p. 210 (1836).

Glochidion Glochidion Arnotti α'.num M むLL.-ARG. i1.1 Li 1.1 11aea XX 玄II ， p. 60 (1863) 一 FORBES &. HEMSLEY i11 J our11. Lin11. Soc. 玄玄VI ， p. 424 (1894)- HENRY) List. P l. Formos. p. 82 (1896) pro part ←HAYATA) Gen. Ind. l. F l. Form. p. 66 (1917); 1con. P 1. Formos. IX ， p. 94 (1920). Phyll αnthus A1'nottianus M 芯LL.-ARG. in Flora 玄LVIII ， p. 370 (1865); in in DC. ， Prodr. XV-:-2 ， p. 279 (1866).

Glochidion Glochidion mohwc α制 ，t， m (11011 M UI ゐ .-ARG.) var. HENRY) 1. c. (1896)-

MATSUMURA et HAYATA) E11um. P l. Formos. p. 361 (1906)-MAsA 回 MUNE) 1.仏、 p. 121 (1936). Glochidion Glochidion hirsutum M むLL.-ARG. sensu HAYATA) Rev. Euphorbiac. J ap. p. p. 17 ，t. 11. D (1904) ; Gen. 111d. F 1. Form. p. 67 (1917)- M; ATsuMuRA)

Ind. Ind. P 1. J 乱p. II-2 ， p. 306 (1912). Nom.Jap. 、。 bα no-kek αnkonoki (HAYATA 1904) ，Ak α -kanko (HAYATA 1917). Specim: Specim: Exam. Formosa. Tamsuy (O Li:>H AM) 110. 491 ，1864 in AA.) ; Giran

(E. (E. H. WILSON) 11O. 10158 ，Mar. 1918 in AA.); Takao (HENRY) 1.1 0. 1929 ，110. 2038 2038 in NY.).

ー- ;) - 318 植物研究雑誌、 第十六巻第六競

The determ In ation of ， the limits of this species ，both as to nomenclature an c1 as as to actual range ， presents at this hour difficulties which are overwhelming.

We accept G. clasyphylhL 1n as valid ， because this binomial 5s of well know l1 application application and has been current on the l11 ajority of the collections from coastal coastal China. We are not certain ，ho ，，"ア ever ， that this name has priority ， because because it is pdssible that G. dasyphyll α1n and G. hu"s 切れ t1n (ROXB.) VorGT m'e m'e the sa l11 e entity ， the latte 1' bino l11 ial being the earlie 1'.

1有The l1 publishing Bnfdleia hiTS~Ltα) ROXBURGH 1' efe 1'1' ed (F l. Ind. III ， p. 699 ，

1832) 1832) to a speci l11 en f1' om the Prince of ¥Vales' Islancl (Penang) ，which ，there- fore ， is technically the type-locality. MULL.-ARG. cites un c1 er P. hirsutus (in

DC づ Prodr. XV-2 ， p. 282 ， 1866) only WALLIOH 7861 B ，“ herb. ROXB." J. D. HOOKER) on his part ， states (F l. BriL In c1. V ， p. 311 ， 1887) that the specimen was sai c1 from Ro 玄BUR G1王 to have been collected near Penang ，but was believe c1 by W ALLIOH to have CO l11 e from China. He ac1c1 s that WALLICH ha c1 two plants ca l1 ed B. 1n ollis) na l11 ely 7858 ，which l11 ay be G-. ze υIα nicu 1n) and 7859 which is G~ hirsut~L伽.

There appears that as ea 1' ly as 1837 there was some c1 0ubt as to the p 1' o- per application of the binomial' of ROXBURGH. The senior author saw in the the herbal'ium of the Museum of Natural History of Paris a specimen ， GAUDI Cl王 AUD) 110. 508 ， bearing the label “ Calcutta: plantes donnees par :M r.

W ALLICH) 1837". This speci l11 en is labellecl “B1"adleia hiTSU旬、 ROXB.~' HBC." in in an handwriting which a note of SPACH in sche c1. in c1 entifies as that of

W ALLIOH. Later on the bino l11 ial was misapplie c1 to speci l11 (ms such as

GRIFFITH no. 4847 from Birma ， to specimens fro l11 J ava an c1， gene 1' ally speaking ， to a host Glochidion which ha a. hairy leaves. The matter was

probably l11 ac1 e wo 1' se by HOOKER'S publishing 壱 as G. villic α ule an. entity which

is is synony l11 0us with G. er'ioc αrpU 1n and' beca l11 e confused in 七he l11 in c1 s of l11 any

taxono l11 ists with G. hirsutu 1n which is an altog'ethe 1' different species. Othe 1'

“ hairy" species published in the l' egions that al' e，0 1' 'l11 ay be the l' ange of

G. G. hirs 悦 um) and G. mish 叫 ，iense HOOK. f. ，Phyll αnth~Ls silhetictLS Mi もLL.-ARG.

ancl ancl Phyll αnthus α sperus' MULL.-ARG. The material of. all these species avail 四 able able in herbariu 円 is scanty ， often restricted to_ the typic collection ，' so that it

is is il11 possible to deter l11 ine with any c1 egree of precision how and where one _' 6- CROIZAT CROIZAT and HA :U: Cさti 叫 notes on some Formosan Eupho 的 ceae (工) 319

entity entity me 1' ges with ，0 1' bounds upon anothe 1'. We have deGidedp GomequeFtl 目白川ain G. dαsyph クllurri ，our concept of

which we exemplify for the present 011 the basis 、 of the following specimens

もcollected outside oI Fo 1' mosa: Hongkong (WILFORD)'; Hongkong (FORD); Hongkong (BODINIER ， no. 59.6); Hongkong (CHUN ，no ・6405) ; (LAU ， no. 110).

Glochidion puberum (L.) HUTCHINSON in SARGENT ， P l. ¥iV ilson. II ， p. 518 (1916). Syn. Agynej αpttbe1' αLINNAEUS ，Mant ，II ， p. 296 (1771).

Br αclleia sinic αGAERTNERj F 1' uct. II ， p. 127 ，t. 109 (1791). Glochidion Glochidion sin'icum (GAERTNER) HOOKER et ARNOTT ， Bot. Beechey. Voy. p. p. 210 (1836). Phyllantht(s Phyllantht(s pt(betfus (L.) MULL.-ARG; in Flora XLVIII ， p. 387 (1865) ; in in DC. ，Pro c1 r. XV-2 ， p. 307 (1866).

Glochiclio η erioc α rpt(m (no Il CHAMPION) HAYATA ， Icon. P l.' Formos.

IX ， p. 95 (1920)-KANEHIRA ，. Formos. Tree. ecl. 1' ev. p. 344 ，fig. 298 (1936)--:MASAMUNE) Sho :r tF l. Formos. p. 120 (1936). Nom. Jap. Ak αge-k αnkon 。μ，Kob α-kek αnkonok ι

Specim. Specim. E 玄乱m. Formosa. Musha (TANAKA ， no. 476 ， Oct. 1929 in AA.);

'Suis~列 Taityu (SASAKI ，Ju l. 29 ，1922 I11 AA.).

Dist. Dist. China and Fo 1' mos 乱.

G. G. erioc αrpum CHAMPION is readily distinguished fI 勺 m this plant by having branches ，leaves ， pedicels and fruits velutinous-tomentose with patent hairs ，

'and 'and ovate 0 1" ovate 句 lanceolate leaves roundish at the base.

Glochidion philippicum (CAVANILLES) C. B. ROBINSON in Philipp. Journ. Journ. Sci. IV ， p. 103 (1909)-MERRILL ，Enum.. Philipp. P l. 11 ， p. 401 (1923) -YAMAMOTO in Jonrn. Soc. Trop. Agric. VIII ， p. 154 (1936)-MAs~MuNE ， Short F l. Formos. p..121 (1936). Syn. Bradlei αphilippic αCAVANILLES ， 1con. et Deser. 'P l. p. 48 ，t. 371 (1797). Bradleia Bradleia philippensis WILLDENOW"Sp. P l. IV ， p. 592 (1805). Actinocl αphne stehold 印刷 MIQUEL in ZOLLINGER ， Syst. Verz. 11 ， p. 116 (1854)-M 田町田in DC. ， Prodr. XV-1 ， p. 215 (1864) ;印 e KOIDZUML

- 7 ー 320 320 植物研究雑誌第十六巻禁六麗

Glochidion Glochidion philippinense BENTH .A:M:) F 1. Hong k. p. 314 (1861).

Phyll α nthus pMlippinensis (BENTl 王.A:M:) MULL ・同 ARG. in Floi'a XL VIII ，. p. p. 376 (1865); ill DC. ， Prodr. XV-2 ， p. 295 (1866).

Glochidion Glochidion 'A 1' notti αη um (non M むLL.-ARG.) HENRY) 'List. P l. Formos. p. p. 82 (1896) pro parte.

Glochidion Glochidion fo 1' 例 os αmtm H .A Y.A TA) Rev. Euphorbi a: c. J ap. p. 20 ，t. 11. G

(1~04)-M.A TSU :M: UR.A & H .A Y.A T .A) 1. c. p. 360 (1906). Glochidion αlbum (non MULL.-ARG.) H .A Y .A T .A) Mater. F 1. Formos. p，: 264 (1911)-K .A NEHI R. A) Formos. Tree. ed. rev. p. 342 ， f. 296 (1936) 一- M .A S.A:M: UNE) 1. c. p. 120 (1936).

Glochidion Glochidion Sieboldia 制 tm (MIQUEL) KOIDZUMI) F l. Symb. Or.-Asia. p. 25. (1930) (1930) .

Glochidion Glochidion dαsγ pl ゅlltt 伽 (non K. KOCH) Y .A M .A]旺 OTO in ， J ourn. Soc.. Trop: Trop: Agric. V1II ， p. 152 (1936) pro parte.

Nom. Jap. Ke-kanko η oki (H .A Y.A T .A 1904) ，N αnb αル kα nko 旬。 μ(Y .A M .A MOTCT 1936) 1936) . Specim. Specim. Exam. Formosa. Takao (HENRY) no. 713 in AA.); Bosyoto ，.

Kuraru ，Koshun (E. H. WILSON) 110. 10990 ， Nov. 1918 in AA.); Formosa (It er W .A RBURG) no. 10393 in AA.) Thian ，Pi-lam ， Taito (K. MIY .A KE) Dec. 23 ，

1899 1899 in TK.-Type of G. fo 1' mos α側刑) ; Ako (E. M .A TUD .A) Aug. 1915 in AA.) ;:

Garanpi ，Koshun (S. S .A ITO) Nov. 1924) ; Bankinsing (F .A U R.I E) no ・ι202 in NY.).

Dist. Dist. FOTmosa ，China ， Philippine and ， Malaysia.

Glochidion Glochidion assamicum (MULL.-ARG.) HOOKER 宜1.， F 1. Brit. Ind. V ， p._ 319 319 (1887).

Syn. Syn. .Phyllant l， 附 αss αmicns 1¥1[む LL.-ARG. in Flora 玄LVIII ， p. 378 ~1865) ;血

DC づ Prod. 玄V-2 ， p. 297 (1866).

var. var. magriicapsulum CROIZ .A T et H .A R .A) var.- nov.

Syn. Syn. Glochidion sp. HENRY) List. P 1. Formos. p. 82 ， no. 937 (1896) pro p 紅白F

Glochidion Glochidion 1仰 ceolat~tm (non H .A Y.A T .A) Y .A M .A MOTO in Journ. Soc. Trop~ Agr. Agr. V1II ， p. 153 (1936).

Capsula major 8-10 111m. in diametro pubescens. Calyx florum' ~ ca. 2.5 惨 mm. in diametro.

一唱 8 - CROIZAT CROIZAT and HARA: Critical notes on some Formosan Euphorbiaceae (1) 32];_

Nom: Jap. Henry-kankonoki (nov:).

SpeQ. SpeQ. Exam. Fo 1' mosa. Bankinsing (HENRY) no. 117 ，fr.-type .i n AA. ，

・ isotype_ in NY:); ibid. (HENRY) no. 415 ，O 丑. in AA. & .NY.).

1n dete 1' mining this.entity we have used the material.at hand which ， so fa 1':

as as Formosan collections ，inc h:山 s on1y one specimen in f1' uit (HENRY ， no.

117). 117). 1n this specimen the capsu1e and the ca1yx of the fema1e fl. owe 1' a 1' e'

la 1' ger than those of c1assic 1ndian specimens such as Assam (M .A STERS).

This. This. may be interpteted as an indication that the Formosan plant actually 戸

siffers siffers f1' om th~ 1ndian species in 1' egard to some details of the f 1' uit. ，lt may ，

conve 1' se1y ，13e merely an individual variation ，0 1'. the result of a miero- and

a mac 1' ocarpous fo 1' m's being endemic both in Assam and i l1 ForI I1 osa. Fee1ing “

unab1e unab1e to sett1e the issue with the mate 1' ial now avai1ab1e we have p 1' eferred to ，

treat treat the Formosan p1a l1 t as cO l1 specific with the 1ndia l1 collectio l1 s. '¥Ve have 司

been been encouraged to.do so i l1 view of the fact that the ra l1 ge of G. αss α伽 icum).

as as he 1' e defi l1 ec1， is conti l1 uous from Assam to Hainan al1 d to Formosa ，and~

both both mic 1' o- a l1c1 mac 1' ocarpous fo 1' ms OCCU l' in Hainan. We suspect that thc'

dist 1' ibution of Glochidion in the tropica1 Fa 1' E 乱st has take l1 place in 1' ough1y-

pa 1' alle1 broad zones which run without ，0 1' a1most tvithollt b1' eak from 1ndia

to to 1ndo-China and adjacent is1a l1 ds ， in pa 1' t. inc1uding Formosa. .On the.

signi 宜cance of .this distribution we may me 1' e1y specu1ate at this hou 1'; the' possibi 1i t5: i8 not exc1uded that it may be connected in some undetermined

manner with preexisting sea-shores 0 1' other g'eographical peculiarities which

have matel 匂 11y changed after ・the original distribution of the ge l1 US ha c1 been e:ff ected approximately within the existing range.

1n Formosan specimen (HENRY ， no. 415) the pedicels 'of the o 宜owe l' s: ，

are are capillary ，att 乱ining to 1.5 cm long ， but they appear to be variab1e in 1ndian 1ndian specimens ， being quite capillary and. manifest1y 丑exuose (P .A RKER ，.

L .A HORE) 0 1' sho 1' t and somewhat sti 自y borne (M .A STERS). The characters of

venation venation an c1 of vegetation are shown to be reasonab1y constant by the 1arge.

collection collection which we have been ab1e to consu1t. As stated ， the on1y eharaeter 司 of of which we have know1edge and might be significant is the size of the capsule. and ，perhaps ， the tendency of the capsu1e of 1ndian specimens to beéome~ more quickly glabrous. -9- ""3 22 植物研究雑誌集十六巻集六披

The following specimens ， all from the Arnold Arboretum ， are cited to illu 司 .'strate' .'strate' our concept of the species outside of its Formosan range: India. India. Upper Assam (JENKINS) no. 530); Assalll (MASTER .s); N. Bengal

(PARKER) ; Lahore ， Govt. Hort. Gardens (PAR 回 R). China. China. Yunnan (HENRY) no. 11717 & no. 11717 C; ROCK) 110. 2581);

:H a: inan (L 却) no. 3471; LIANG) 110 ・65034; How ，110. 71680). Indo-China. Indo-China. Tonkin (P 並TELOT ，110. 6513 & 110. 6529).

Glochidion Fo :r tuni HANCE i11 Ann. Sci. Nat. ser. 4，XVIII ， p. 228

(1862)-FoR 田.BES & HEMSLEY in Journ. Lin11. Soc. XXVI ， p. 424 (1894)-

HENRY) List. PL Formos. p. 82 (1896)-HAY~TA) Mater. F 1. Formos. .p. 264 (1911) (1911) ; Gen. Ind. F 1. Form. p. 66 (1917) ; Icon. P 1. Forlllos. IX ， p. 95 (1920) --YAMAMOTO in Journ. Soc. T.rop. 立gric. VIII ， p. 152 (1936)-KANEHIRA ， _Forlllos. _Forlllos. Tree. ed. rev. p. 344 ，f. 299 (1936).

， Syn. Ph υlla1 1l thus p αbe 1"l AS CI. ・Fortunei (HANGE) MULL.-ARG. in DC. ， Prodr.

XV-2 ， p. 307 (1866) quo 乱d syn. tantum. Phyll α nthus obovatus (non WILLD.) l¥!I ULL:-ARG.) 1. c. (1866) quoad p1. Chin. Chin. Glochidion Glochidion obov αれAm (non SIEB. et Zucc.) HAYATA ， Rev. Euphorbiac.

Jap. Jap. ト.、p .19 (ο19ω04 め) 伊quoa 刈dpμ 1. Forn 110 工 Pl. Pl. Form. p. 360 (1906). Nom. Jap. Hir αm ι.k ω llkonoki (HAYATA) 1917).

S pecim. .Exam. Formo 紙 sine loeo speciali (OLDHAM ， no. 502 ， no; 503; ~FAURIE ， 110 ・8198) ; Kelung (It er. W .A RBURG ，110 ・9964); Sozan D~iton range ， 'Taihoku 'Taihoku (E. H. WILSON ，110 ・11217); Hokuto (FAURIE ，110 ・417; 110 ・1162); Takao (HENRY ，110 ・105.3 ，110. 1216 ，110. 1793; E.' H. "¥可 ILSON ，110 ・9854); in m011te m011te Kwa11ni11 (F .A URIE ，110. 1195); omnia in AA. We much regret that we have not bee11 able to e:x amine the type of G. F01'tuni F01'tuni which is unav~ilable on a.ccount of the present inter11ational situation. We have accepted ，consequently ，a concept of the species which agrees with -t he concept generally entertained by taxonomists who have worked upon the

.. .. Glochidion of the southern coast of China. Our cO l1 cept ， at l~ast pro tempore ， is is exemplifiecl by such specime11 as DUNN ，110 ・3509 (Hong k. Herb.) from

- 10 -，- CROIZAT CROIZAT and HARA: C l' itical. I，l otes on some FO l' mosan Euphorbiaceae (1) 323

Fukien ，which was probably determinecl fro Iil authentic matetia l. The Formosan plant varies a great deal in the degree of hairiness 、on its branches branches and pedicels as well as in the shape of leaves ，which are always glabrous glabrous even when young. G. obo 叩 tum SIEB. et ZUCC. from J apan .proper

，di :ff ers from G. Fo付~Ln i by having smaller fruits ， generally 6-8 mm. in dia- me 同 shorter styles less than 0.5 mm. 凶1O白∞ 略n1 e引unea 叫te base. Specimens from the Liuki l1 1slancls ，however ，show somewhat intermediate intermediate characters between the two. species.

Glochidion hongkongense MULL.-ARG. in Linnaea XXXII ， p. 60 (1863)

- FORBES & HEMSLEY in JO l1 rn. Linn. Soc. X 玄V1 ， p. 42 生(1894)-HENRY) List. List. P 1. Formos. p. 82 (1896)-MATsuMuRA & HAYATA) 1. c. p. 361 (1906)- HAYATA) Mater. F l. Formos. p. 264 (1911); Gen. 1nd. F l. Form. 67 (1917) ;

1con. 1con. P 1. Formos..1X ， p. 95 (1920)-YAJVIAMOTO in JO l1 rn. Soc. Tiop. Agric. VII1 ， p. 152 (1936)-KANEHIRA) Formos. Tree. ecl. tey. p. 346 ，fig. 300 (1936) -MASAMUNE) Short F 1. Formos. p. 120 (1936).

Syn. Syn. Glochiclion littor αle (non BLUME) BEN 四 AM) F l. Ho 時k. p. 314 (1861).

Phyllanth~Ls hongko η gensis M むLL.-ARG ・in Flora XLVII1 ， p. 37 1. (1865); in in DC. ， Prodr. XV-2 ， p. 282 (1866).

Glochiclion Glochiclion zeylanicu 例 JUSSIEU sensu HAYATA) Rev. Euphorbiac. Jap. p. p. 17 (1904); Gen. 1nd. F l. Form. p. 67 (1917)-MATsuMuRA ‘& HiYATA) Emun. P l. Formos. p. 360 (1906)-MATsuMuRA) 1nd. P l.

Jap. Jap. II-2 ， p. 306 (1912)~ Glochiclio ω sphaerostigm~l例 HAYATA) 1con. P l. Formos. TX ， p. 96 (1920) pro minor. part.

Nom. Jap. Kakib αno-k αnko 仰 oki(HAYATA) 1904) 、; Honkon-k αnkonoki S ei-

ron 北側初旬oki (HAYATA 1917). Speeim. Speeim. Exam. Formosa. Kelung' (FAURIE) no. 1280 ， .Nov. 1914 in AA.);

Taihoku (F AUR .IE) no ・1832 ，Apr. エ915 in AA.) ; Vicinity of Taihok l1 (TANAKA)

110 ・35 ， Jun. 1929 in AA.); Takao (HENRY ， no. 707 in AA.); Hokuto (FAURIE)

110 ・203 ， Dec. 1913 in AA.); ibid. (FAURIE) no ・1657 ， Feb. .1915 in TK.--

Paratype of G. sph αθ rostigmum HAYATA). Liukiu. Liukiu. 1s. Okinawa: Nago (E: H. WILSON) no. 8068 ，Mar~ 1917 in AA.) ・ ー- 11- 324 植物研究雑誌 第十六巻第六銃

Kyushu. 1s~、 Amamí-Oshima: in f1' uticetis A 1' busta (FAURIE) no. 4188 ， Ju l. 1900 in AA.).

1n accepting G. hongkongense fo 1' this species we have followed the same'

line line of thought which has suggested that we retain G. dαsyphyllt 仰 V¥ Te a1' e

far f1' om certain that G. honglcongense is distinct f1' om G. ze ，ylanicum of

JUSSI~;U. It is impossible ， hpweve 1'， to determine with accur a， cy whether the

Formosan plant is actually ，distinct from the 1ndian ，entity which exampli 宜es ，

the the ，Jussieuan use of the binomial. We have found that a great deal of va 1' ia-

tion tion occurs in forms that may be p 1' esumed to rep 1' esen 七 G. zeylanicum and G. G. honglcongense) both in the length of the pedicels ， in the :r: elative size of

the the inflorescence and in deg 1' ee of evolution of thi styles ， 'but at the same

time time we have not been able to connect these va 1' iations with any factor that

might suggest eithe 1' a geog 1' aphic 0 1' a definite mo 1' phological segregation.

We ，suspect that seve 1' al local fo 1' ms o~cu1'， but ，we do not know whethe 1'

such such forms a1' e me 1' ely due to the in 丑uence of edaphic factors or represent

established established echotypes which dese 1' ve to be accepted as distinct ， either speci-

宣cally 0 1' varietally. Not unlike G. dωυ phyllum) the p 1' esent entity occurs

th 1' ouglwut the range f1' om 1ndia to Formosa ， exten c1 ing to the south into Malaysia. Malaysia. We base our understanding of G. zeylanicum 早ot upon Br α dleia

zeylanic αGAERTN サ which was cited with 、c1 0ubt by J USSIEU) and is illust 1' ated

in in a ， manner that precludes a definite' ic1 entification. We understand G.

ze γ1α nicu 伽 on the basis of the specimen in the herbarium of JUSSIEU) which

has been available to us ，al1 d of other materials cited by Mur ム -ARG .

.Gloohidion lanoeolatum HAYATA) Rev. Euphorbiac. Jap. p.'16 パ. II ， C.

(1904)-MATsm 町 JRA et HAYATA) Enum. P l. Formos. p. 360 (1906); Gen. 1nd. 1nd. F l. Form. p. 67 (1917)-MATsuMuRA) Ind. P l. J ap. 11-2 ， p. 306 (1912)-

Iむ NEHIRA ，Fo 1' mos. T l' ee. ed. 1' ev. p. 347 ，勾. 302 (1936)-MAsAMuNE) Short

-F l. F Ol~mos. p. 120 (1936). Syn. Syn. Glochidion hongkong 仰抑 (non MULL.-ARG.) FORBES et HEMSLEY in J ourn. Linn. Soc. 玄玄V1 ， p. 424 (1894) pro parte. Glochidion Glochidion sp. HENRY) List. P l. Formos. p.' 82 ，110. 937 (1896) pro parte. G.l oc 7i idion kotoense HAYATA) 1co11: P l. Formo 包. 1X ， p. 96 (1920) 一 -12 ー OROIZA' l' and HARA: OriticaI notes on some Formosan Euphorbiaceae (1) 325

MASAMUNE ，1. c. p. 120 (1936).

Glochidio η sphα erostigm~tm HAY~T.A) 1. c. (1920) pro major. part.- MAS .A MUNE) 1. c. p. 121 (1936).

~Glochidion kusukusensisH .A Y.A T.A， 1. c. (1920) ....，- MAsAMuNE ，1. c. p. 121 (1936) (1936) .

Nom. J[¥-p. ' Kiir ・仰もみαnkono μ (HAYATA 1904) ，Koto-k α励。ωoki) Nag αbα- kα nkonoki.

Specim. E 羽 m. Formosa. Kelung (MAKINO) Oct.' 31 ，1896 in TK.-Syntype

，of G. Zα 銘 ceolatum); ibid. (OWATARI) Nov. 1， 1896 in TK.'--'Syntype of G ，.

1仰 ceolatum); ibid. (It er W ARBURG 乃 no.' 9962 in AA.); Suisha ， Sintiku (in

'TK. ，Syntype of G. sph αerostig 例 um); Kaukaukei (T. SOMA) Apr. 25 ，1917 in in TK. ，，Syntype of G. sphaerostigm~tm); ， Kotosho (Type of G. 7c otoense in 'TK.); Bankinsing (HENRY) no. 829 i11 AA;); Formosa (OLDHAM ，110. 488 & 489 ，1864 in AA.).

Liukiu. Liukiu. Loo 田 Choo (WRIGH 'l') no. 266 ，1854 in GH.); Liukiu (Yokohama N ursery Co. ，1914 in AA.).

The Formosan species of Glochiclion enumerated in this paper are distin- guished as follows. G. ch αcle 伽 enosoc 仰 'pum HAYATA) G. suish αω se HAYATA)

.and .and G. ，longipediceZ 如何冊 YAMAMOTO which are imperfectly known are omitted.

1 ~ Rami foli 叩 ep 伽 sc 削 es. Oap 卵日飢1山 p伽 sce 合 .2

l Rami 白f01 五iaque glabr 吐i (i 担n G.Fo 併rt 加旬 ni rami 知Iii te 目rdum mi 担nu 凶1泊ti 均目日討i直me がpilos 必i) .0

(Oapsula 3 (..，. 4) -locularis 5-7 mm. in diametro. Folia sul 北u日 adpresse minute

I pilo 叫 saepe glauca. Sty1i breviter columnares ad apicem levite 1' inc 1' a，ssati 2i G.H α yatae

I Oapsula 5-7 -locularis 8-15 mm. in diamet 1' o. Folia 剖s叫u1b 予巾tus

¥.. ¥.. molli 抗te 町l' pubescen 凶1此ti 弘a，numquam 凶glauca. Styli 白cra 制ss 討i，'dep 伊1' esse glo bos 凶i ....3

3~ れ刊F O叫li 勾as 耐 us v 山 t世白伽担II n川 ome 凹n削 a，b糊 a制si saepe 附 undata G. dasyphyllum

¥ Folia subtus mollite 1' pubescentia 4

( Lobi calycis elliptici pistillo paullo b l' evio l' es. Oapsula 10-15 mm. in diametrm

4J 4J Folia minora ，ple l' umque

| 工Lobi 註ic伺al 訪yTcis 埼 ova 批t世i，'pist 凶i臼110 v羽叫alde brev 吋i01' 司e伺s. Capsula 8-10 mm. in diametro.

¥.. ¥.. Folia majo 1' a ...... G. phillipicum

-13 ー 326 植物研究雑誌第十六巻第六畿

( Capsula pubescens. Calyces 丑OrU ll1 ~ parvi ca. 2.5 1l11l1. in dia ll1 etro. Folia

5~ basi v向 o ，c問 ata. Pedicelli 丑orum O 五lifol' ll1叫 saepe 伽 19ati .曜匡

I ~αωss抑α伽4側旬伽7， V 乱町r. 伽 αd仰偽 4旬cα 仰ps 叫旬叫1加協悦'

'- Capsula glabra ，...... ，...... '. • ，• • • • • • • • • • • • • • • • • • • • • • • . • o • : • • .6

( Folia ll1 inora subobovata basi 土 cuneata. Styli ad apice ll1 paullo incrassat i.

6~ Ra ll1 i ll1 Inutiss Im e pilosi vel glabri ..，...... ・・・ょ..，...... ，.. .G.Fortuni

L Folia majora. Styli subconici vel he ll1 isphaerici. '. ， ， ...... 7

( Calyces majores 5 mm; in dia ll1 etro. Styli breviter conici. Folia basi vulgo'

7~ r伽 .ndata ...... ;...... '...... ;...... G. h仰 gた仰側6

I CaJyces minores 3m ll1. in diametro. Styli he ll1 isphaerici .leviter 6-1obat i.

l Folia basi saepe cuneata ...... ，...... o ，...... G.lanoeolat 仰 111

(to (to be continued)

一- 14-