September 1957 Journ. .J ap. Bo t. Vo l. 32 No. 9 275

Tuguo TATEOKA*: MisceUaneous papers on the phylogeny of (10). Proposition of a new phylogenetic

system of Poaceae 料

館岡亜緒ホイネ科の系統分類に関する雑記(1 0) イネ科新分類系の提案林

Based on the considerations described in the previous papers of this series ， the classification classification and genealogical tree of Poaceae are presented in Table 1 and Fig. l. The relationships among tribes according to my conjecture are more clearly shown in in Table 2. Table 3 is the list of ]apanese grasses classified according to this view. The importance of chromosome con 白guration in the systematics of grass groups was for the first time revealed by Avdulov (1931) ，who gave an impulse to syste- matic investigations of grasses bγintroducing the characteristics of chromosomes as a systematic criterion. While Avdulov's work is monumental ， in the light of our present present knowledge his belief in stability of chromosomes as to number and size should make place to a more flexible view. Variation of the karyotype among individuals individuals of the same species ，spontaneous chromosome breakage ， variation of chromosomes within the same individual ， the presence of B-chromosomes ，etc. ，in- dicate dicate the flexibility of the chromosome numbers and sizes. This makes it neces- sary sary to revise Avdulov's views in several instances. For example ，Avdulov (1. c.) includes includes tbe Brachypodium in Poatae-Phragmitiformes which is clearly dif- ferent ferent from Festuceae-Bromeae ，on the ground that this genus possesses small chromosomes and the basic number is not 7. According to my grouping ， this systematic systematic treatment is inadequate. The nuclear plates of Poa annua Linn. (2n= 28) ，a member of Festuceae ， include several small chromosomes similar to Brachy- podium chromosomes ，and such small chromosomes are also found in the chromo- somes sets of. P. .n 争ponica Koidz. (2n=42 ， 56) which is closely allied to P. annua Linn. Linn. Most belonging to Festuceae have large- to medium-sized chromo- somes ，and it is quite possible that the small chromosomes in Brachypodium have developed from the chromosomes of the former type. If the chromosomes number

and size we 閃 so stable as maintained by Avdulov ，Brachypodium)hould be included in in Phragmitiformes. However ， in view of various other features possessed by

ホ National Institute of Genetics. Mishima. Shizuoka Pre f. 国立遺伝学研挺所.

林 Contτibutions from the National Institute of Genetics. No. 215.

- 19- 276 植物研究雑譜第32 巻第9 号 昭和 32 年 9 月

Brachypodium ， this genus should be most naturally referred to Festuceae ，and it~ chromosome configuration should have been derived from some ancestral chromo- some type which resembled that of the typical festucoid genera. Subsequently ， karyological karyological investigations of grasses have been carried out by many cytologists ， although no noteworthy study ot ・karyosystematics of Poaceae have been published. These data should be reexamined from a more recent view point. The characteristics in leaf structure are another valuable criterion in the syste- matics of Poaceae. Avdulov (1931) has reviewed the studies of leaf structure in grasses ，and divided it into two tγpes ，1 and II. Prat (1936) ，based on his OW Jl

Table 1. The classification o[ Poaceae according to the present author.

Subfam. Beet1e 持 Trib. Danthonieae C. E. Hubbard 呪ー

Trib. Trib. Anomoch10eae C. E. Hubbard Trib. Garnotieae Tateoka 持持長持

Trib. Trib. Streptochaeteae C. E. Hubbard Trib. Arundinelleae Stapi ・ Trib. Trib. Bambuseae Nees Subfam. A. Br. Trib. Trib. Olvreae Kunth Trib. Festuceae Dumor t. Trib. Trib. Oryzeae Kunth Trib. Monermcae C. E. Hubbard Trib. Trib. Phyllorachieae C. E. Hubbard Trib. Triticeae Dumort.

Trib. Trib. Pariancae D. E. Hubbard Trib. Agrosteae Nees -K-時長兼 Trib. Trib. Micraireae Pi1ger Subfam. Eragrostoideae Pi1ger

Subfam. Arundoideae Tateoka 長持 Trib. Spartineae C. E. Hubbard 骨 Trib. Trib. Coleantheae Ascher. & Graebn. Trib. Chlorideae Kunth Trib. Trib. Brachelytreae Ohwi Trib. Lappagineae Link Trib. Trib. Stipeae Nees Trib. Pappophoreae Kunth Trib. Trib. Hubbard & Vaugham Subfam. A. B r. Trib. Trib. Meliceae Fries Trib. Lecomtelleae Pilger

Trib. Trib. Glycerieae C. E. Hubbard 長 Trib. Bovinelleae A. Camus Trib. Trib. Nardeae Reichenb. Trib. Anthephoreae Pilger Trib. Trib. Lygeeae Lange Trib. Melinideae Hitchcock Trib. Trib. Centotheceae Roshcv. Trib. Trachyeae Pilger Trib. Trib. Streptogyneae C. E. Hubbard Trib. Isachneae Benth.

Trib. Trib. Ehrhau:ae Tateoka" ー是非 Trib. Paniceae R. Br. Trib. Trib. Phaenospermeae Roshev. Trib. PresL Trib. Trib. Arundineae Reichenb. Trib. Maydeae Mathieu Trib. Trib. Thysanolaeneae C. E. Hubbard

- 20- 円、 LU lQJ 57 υeD&EL en 、a er 且 .J ourn. Jap. Bot. Vo l. 32 No. 9 277

observations ， divided the leaf structure of grasses into the Festucoid type and the

/~anicoid type including the Panicoid subtype ， Chloridoid subtype and ..Bambusoid subtype. subtype. Avdulov's Type 1 and Prat's Panicoid tγpe are seemingly similar ，though rather rather dissimilar in reality. The same is true of Type II and the Festucoid type. Prat's Prat's Panicoid type includes Avdulov's Type ! and a part of Type II ，and Prat's Festucoid Festucoid type corresponds with Avdulov's Type II partly. The members whose leaf leaf structures have been referred to Type II by Avdulov and to Panicoid type by Prat Prat are Phragmites ，Arundo and Uniola. While in the typical Panicoid type the chloroplasts chloroplasts are localized in an outer bundle sheath surrounding the vascular bundles ， the bundles in Phragmites ，'Arundo and Uniola are surrounded by an

持 Sine descrip 1. la 1. Subfam. Pharoideae Beetle in Bul l. Tor r. Bo t. Club 29: 360 ，

1950. 1950. T 1' ib. Glyce 1' ieae C. E. Hubba 1' d in Grasses ， p. 401 ， 1954. Trib. Danthonieae c. c. E. Hubbard in J. Hutch.'s Brit. F l. P 1.， p. 307 ， 1948. Trib. Spartineae C. E.

Hubbard in G 1' asses ， p. 401 ， 1954.

州 Subfamilia Arundoideae Tateoka ，sub f. nov. ・-Spicula uni- vel plu 1' iflo 1' a， saepe

floribus floribus superio 1' ibus reductis sed in speciebus paucis eis inferioribus impe 1' fectis ，

stig ll.J. atibus saepe dense et minute plumosis ， staminibus 3 vel paucio 1' ibus interdum 6. 6. Numerus chromosomatis saepe b= 12 ，11 ， vel 10. Characte 1' istics of t1' ansverse

leaf leaf section are mostly festucoid or bambusoid ，a few panicoid. Epide 1' mal features

are are panicoid 0 1' festucoid.

特長特 T 1' ibus Ehrharteae Tateoka ，t 1' ib. nov.-Spiculae pedicellatae ，lateralite 1'

compressae ，t 1' iflo 1' ae ，rach i1l a supra glumas non articulata; 臼o1' es inferiores 2 steri-

1e s; eius summus tantum fe 1' tilis. Lemmata sterilia quam glumae longio 1' a. Lem-

ma fertile interdum basi appendiculatum; appendice subnodifo 1' mi cum ea lemmatis

supe 1' io 1' is ca 1' donem fo 1' mante. Paleae angustea ，costatae ，1・ ve12 ・nervosae. Stamina 6，4. ， 3， 2， vel 1. Genera 3 includenda: Ehrharta Thunb. (Typus) ，Jt licrolaena R.

Br. Br. et Tetr αγ， rhena R. Br.

州制 Tribus Garnotieae Tateoka ， trib. nov ，-SBiculae uniflorae; pili calli basi spiculae spiculae siti recti ， breves longive ，plerumque torti; rachilla pone paleam non pro-

jecta. jecta. Glumae 2，memb 1' anceae interdum rigidio 1' es 司 trinervatae ， apice acutae usque

acuminatae ra 1' o obtusae. Lemmata glabra ，laevia ， raro ad apicem minute scabrida ，

convexa ， tenuiter 3・nervo 叫 apice a 1' istata (sed rarissime mutica). Paleae hyalinae ，

子nervatae ，secundum nervum carinatae. Flores he 1' maphroditi ， lodiculus parvis ， antheris antheris 3， ovario glabro ， stylo distincto ， stigmatibus 2 plumosis. Caryopsides fere lineari-oblongae ， scutello ad 1/3 caryopsidis extenso. Genus unicum includendum: Garnotia Brongn.

制却制 Pilger (1954) included within Aveneae (sens. lat.) the genera which had

been treated as members of Agros[eae sens. st r. as well as the gene 1' a of Aveneae sens. sens. str. But ， in 1942 Ohwi has taken the same systematic grouping as Agrosteae Nees. Nees. According to the rule of nomenclature ，Ohwi's treatment must be adopted.

- 21 - 278 278 植物研究雑誌第32 巻第9 号 昭和 32 年 9 月 outer outer sheath lacking chloroplasts which are uniformly distributed throghout. the mesophy Jl. 1 (1956a) have reexamined the 'grasses in Prat's materials ， as well .as other other ]apanese and American grasses ，and obtained similar resUlts to those of the previous previous investigators except for the finding of peculiar features of leaf anatomγ in in Arundinella hirta Tanaka. While Prat's results are very interesting and impor- tant tant for the systematics of Poaceae ， the leaf structures of many species or species groups had remained unknown. Such species or species groups have been studied

〆 " ， ーー- 町.._ _- -ーーーーー' PRAROIDEAE

Fig. Fig. 1. Diagram showi 白g the phylogeny of.. Poaceae according to ~the present autho t"

- 22 ー September 1957 Journ. Jap. Bot. Vo l. 32 No. 9 279 bγvarious subsequent investigators ，such as Potztal (1952 ，1953) ，Hansen and Potztal Potztal (1954) ，Page (1947) ，de Wet (1954 ，1956) ，Tateoka (1956b ，k， 1， 1957c ，d ，e， f) ， etc. etc. These investigations have yielded discoveries which are valuable for the syste- matlcs matlcs of Poaceae ，and established the systematic significance of this character. The systematic importance of anatomy of embrγo in mature caryopsis which was pointed out by Van Tieghem has been verified by Reeder (1953) The features of of the first leaf in the seedlings are also useful for the systematics 0 1' Poaceae. Besides ， various other characters: the features of root-hair development ，endosperm starch.grains ， etc. are of systematic value (c f. Stebbins 1956). Some of these char- acters acters in various grasses of problematic position have remaiiled unknown. To me ， the two characters~ chromosomes and leaf structure ，seem to be especially useful for the the systematics of Poaceae at present. Thus ， the discussions are primarily based on these these two characters as well as the external morphology and distribution of the plants. plants.

Table 2. Relationships among tribes according to the present author Subfam. Pharoideae Anomochloiformes- Trib. Anomochloeae Streptochaetiformes- Trib. Streptochaeteae Bambusiformes- Trib. 0lyreae ， Trib. Bambuseae Oryziformes-Trib. Phyllorachieae ， Trib. Oryzeae Parianiformes Parianiformes ・-Trib. Parianeae Micrairiformes-Trib. Micraireae Subfam. Arundoideae Colean t. hiformes- Trib. Coleantheae Stipiformes-Trib. Brachyelytreae ， Trib. Stipeae ， Trib. Aristideae Glyceriiformes-Trib. Meliceae ， Trib. Glycerieae Nardiformes- Trib. Nardeae ， Trib. Lygeeae Arundiformes ・-Trib. Centotheceae ， Trib. Streptogyneae ， Trib. Ehrharteae ， Trib. Trib. Phaenospermeae ， Trib. Arundineae ， Trib. Thysanolaeneae ， Trib. Danthonieae Arundinelliformes-Trib. Garnotieae ， Trib. Arundinelleae Subfam. Pooideae (Eufestuciformes) ーTrib. Festuceae ， Trib. Monermeae ， Trib. Tri- ticeae ， Trib. Agrosteae Subfam. Eragrostoideae (Eragrostiformes)-Trib. Spartineae ， Trib. Chlorideae ， Trib. Lappagineae ， Trib. Pappophoreae Subfam. Panicoideae Paniciformes-Trib. Lecomtelleae ， Trib. Boivinelleae ， Trib. Anthephoreae ， Trib. Melinideae ， Trib. Trachyeae ， Trib. Isachneae ， Trib. Paniceae Andropogoniformes- Trib. AndrQPogoneae ， Trib.. Maydeae

一回一 280 280 植物研究雑誌第32 巻第9 号 昭和32 年 9 月

The essentials of my findings set forth in previous papers of this series may be summarized as fo11ows: ー 1) 1) Ehrharta ，Microlaena and Tetrarrhena should be removed from Phalarideae. 1t 1t seems most natural that these genera are treated as belonging to an independent tribe ，Ehrharteae (c f. Tateoka 1957f). Aveneae and Danthonieae are apparen t1 y different different phylogenetically. Danthonieae may be related to Arundineae. Aveneae and Phalarideae can be assembled into a tribe ， clearly related to Festuceae (sens. str.) ，Monermeae and Triticeae. Festuceae sensu Pilger (1954) represent a miscel- laneous group. The genera ha ving b = 7 and large (-medium) chromosomes and also also the Festucoid type of leaf structure are mixed with the genera possessing small .c hromosomes in which the basic number is not 7 and which have non-Festucoid type in leaf structure. Many members of Festucinae (sensu Pilger) ，Beckmannia ぅ ，Loliumand some of Glyceriinae ， Sesleriinae and Brominae can be referred

J: o the festucoid genera ，although several genera ，such as Brachypodium ，showing the chromosome configurations difierent from those in the typical festucoid genera are found within them. On the other hand ，Fingerhuthia ，Aeluropus ，Uniola ， Distichlis ，Brachyelytrum ，Tetrachne ，Brylkinia ，Astrebla ，G:yceria ， the genera of Centothecinae ， etc. belong to the non-festucoid group. A part of these genera can be referred to Eragrostoideae ，while the large majority may be ascribed to Arun- .doideae. .doideae.

Table 3. The classification of Japanese grasses accordinξto the present present author

.s ubfam; ，Pharoideae

Trib. Trib. Bambuseae ー (omission) Trib. Trib. Oryzeae

Subtrib. Subtrib. Oryzinae-O つIza/*' Ic Leersia Subtrib. Subtrib. Zizaniinae-ZizanIa Subtrib. Subtrib. Chikusichloinae-Chikusichloa :S ubfam. Arundoideae

Trib. Trib. Stipeae-Achnatherum ，Orthorap .T1 iw 1l Trib. Trib. Brachy e1 ytreae-Brachyelytrum Trib. Trib. Meliceae- i'¥1 elica Trib. Glycerieae Subtrib. Subtrib. Glyceriinae- ， Subtrib. Subtrib. Brylkiniinae- Brylkinia

- 24 ，-- Septembey 1957 .l ourn. Jap. Bot Vol. 32 No. ，9 281

Trib. Trib. Centotheceae ー'Lophatherum Trib. Trib. Phaenospermeae Subtrib. Subtrib. Phaenosperminae-Phaeno ゆerma Subtrib. Subtrib. Diarrheninae- Di arrhena Trib. Trib. Arundineae Subtrib. Subtrib. Arundininae-Arundo ，Phragmites Subtrib. Subtrib. Moliniinae-Moliniopsis ，Hakonechloa Trib. Trib. Arundinelleae-Arundinella Subfam. Pooideae Trib. Trib. Festuceae

Subtrib. Subtrib. Festucinae-Festuca ，Briza; 'k Dac 砂lis ，徒 グ Poa ，Aulacol φ， is ， Puccinellia ， Torreyochloa Subtrib. Subtrib. Brominae-Brom ω Subtrib. Subtrib. Brachypodiinae-Brachypodium

Trib. Trib. Tritic 白ea 悶e一'Agro めyげ仰f仰'O 叫nzら，El かうymus ら，As; 中4ρ >en 問'ell μα ，Trit 枕iたcum ηzJ 戸+持持 Sec ωαJ たP Trib. Trib. Agrosteae-Avena 戸・持 Arrhenatherum 戸Helictotrichon ，Trisetum ，Descham- psia ，Holcus ，* Koeleria ，Milium ，Beckmannia ，Phalaris ，Anthoxanthum ， Hierochloe ，Phleum ，Alopecurus ，Polypogon ，Aira ，* Cinna ， ， Calamagrostis Calamagrostis Subfam. Eragrostoideae Trib. Trib. Chlorideae Subtrib. Subtrib. Chloridinae-Chloris 戸 Cynodon Subtrib. Subtrib. Eragrostinae-Eragrostis ，Dactyloctenium グ Eleusine ，Cleistogenes ， Di plachne ，Di nebra ，* Leptochloa ，Tl 争ogon Subtrib. Subtrib. Sporobolinae-Sporobolus ， Trib. Trib. Lappagineae-Zoysia Subfam. Panicoideae Trib. Trib. Isachneae-lsachne ， Coelachne Trib. Trib. Paniceae 一Cenchrus ，* Pseudora ρhis ，Pennisetum ，Setaria ，Sacciol φis ， Panicum ，Digitaria ，Pa ゆalum ，Eriochloa ， Oplismenus. Brachiaria ， Spinifex Spinifex Trib. Trib. Andropogoneae Subtrib. Subtrib. Dimeriinae-D 匡imeria Subtrib. Subtrib. Saccharinae-lmperata ，Miscanthus ，Eccoilopus ，sjμ diopogon ，Sac- charum ，Pogonatheruin ，Pseudopogonatherum Subtrib. Subtrib. Sorginae-Sorgum Subtrib. Subtrib. Andropogoninae-Arthraxon ，Bothriochloa ，Andr o. ρogon ，Cymb ゆ 0 ・ gon ，Themeda Subtrib. Subtrib. Ischaeminae-lschaemwn Subtrib. Subtrib. Rottboelliinae- ，Hemarth ，匂

Trib. Trib. Maydeae-Zea ，掛 Coix

発 Naturalized plants ， .，時 Cultivated plants.

- 25- 282 植物研究雑誌第32 巻第9 号 昭和32 年 9 月

Pilge 1' (1954) advocates the following phylogenetic 1' elationships: AM WH 間 M ぺ・唱‘-eea Phala 1' ideae sens. str. Il-- Fe 雪tuceae- Festucoid genera Da 主 ∞ e Ehrharteae Festuceae-Non イestucoid genera

1 can not support such a view. 1 have poin 主ed out that the three groups indicated above ， together with T 1' iticeae and Monermeae ，should be considered to constitute a distinct phylogenetic group to which the name of Eufestuciformes has been pro- posed. posed. Ttis..group. is apparently monophyletic. The members of the group Eu- festuciformes festuciformes are characte fI zed by having large medium sized chromosomes of which the the basic number is nea l'匂 always ，7，and also by the festucoid leaf structure. Also ， all of the tribes included in this group have their centers of distribution in the the Medi~erranean and neighbouring regions. 2) 2) Cleistogenes ， Gouinia and Crin 争es should be placed in Chlorideae sensu Ohwi (1942) ，although they show resemblance in several respects of external morphology to Arundineae. While Neyraudia is referred to Eragrosteae by severa1 investigators ， it seems to be more naturally ascribed in Arundineae. Arundineae (記 nsu Pilger ，except for the above three genera); Danthonieae ，Thysanolaeneae ，

Phaenospermeae ，Centotheceae ，Stipeae ， etc. may be more or les 話 allied. Pappo- phoreae ，although they have 7・0 1' more-nerved lemmas ， should be regarded as a member of Eragrostoideae. The number of lemma nerves is an important character in in the systematics of Poaceae ，but it can not be used in some cases as a systematic criterion. criterion. C. E. Hubbard (1947) has pointed out that Drake-Brocknzannia having 5

(-7) (-7) nerved lemmas has a striking similarity in many features of external morpho ・ logy to l- Ieterocarpha having 3-nerved lemmas. Aristideae seem to be di 百erent from Eragrostoideae in showing the epidermal features of Panicoid subtype and the the awns being differentiated into columna and subula. This tribe may be refer- red to Arundoideae. Also ， Garnotia should be separated from Eragrosteae ，and may be treated as an independent tribe ，Garnotieae ，which has てelationship with

Arundinelleae (cf. Tateoka 1956k). The subfamily Eragrostoideae should be co 凶 ned to to Chlorideae sensu Ohwi (1942) (pro parte) ，Lappagineae ，Pappophoreae and Spartineae. Spartineae.

3) 3) Acco 1' ding to Brown (1951) ，Elyonurus tripsacoides Humb. & Bonpt. and

E. barbiculnzis Hack. have 2n=20 la 1' ge chromosomes in cont 1' ast to the small chromosomes of most other Andropogoneae. cylindrica Kuntze resembles

- 26 ー 円、 守 ..Qd可/ e YA bρ 5 U ρUph， en va ， L し a Journ. Jap. Bot. Vo 1. 32 No. 9 283

Elyonurus in external morphology a吋 its chromosomes (2n=18) are of medium size. size. Hemartkria sibirica Ohwi also has similar medium-sized chromosomes. These genera belong to Rottboelliinae-Rottboelliininae. ， ， etc. whose chromosomes are small like those of most other species in Andropogoneae are are included in the same subtribe. The difference in chromosome size between Elyonurus and Eremochloa ，Rottboellia ， etc. is similar to the di 百erence in chomo- somes found between Phalarideae sens. str. and Ehr ・harteae. 1 consider such chromosomal differences as suggesting a phylogenetic divergence in the case of Phalarideae ，but the same is not true of Andropogoneae. i) The chromosomes of Andropogoneae show a continuous variation from small to large; small chromoso :m es found in Eremochloa ，Rottboellia ，etc. ，medium sized ones shared by some species of and Manisuris and the large chromosomes observed in Elyonurus. On the the contrary ， in the case of Phalarideae ， the di 百erences in chromosome size are strikingly strikingly discontinuous. ii) In the case of' Phalarideae ， differences maγbe found in in their basic chromosome numbers :-Pealarideae sens. str. have b=7 ，6 or 5，while Ehrharteae are characterized by b= 12. But ， in the case of Andropogoneae ，sueh a difference can not be 、 found; Elyonurus ，Manisuris and Hemarthria show b= lO or or 91ike other Andropogoneae having small chromosomes. iii) Phalarideae sens. str.

and Ehrharteae. have differences in the featu 1' es of leaf structure ，which suggest

phylogenetic phylogenetic dive 1' gence between these two groups. But ， in the case of Andropo ・ goneae ，such differences can not be assumed. Elyonurus has Panicoid typ-Panicoid subtype in accordance with Eremochloa ，Rottbof ，z lia ， etc. iv) In the features of

external external morphology ， the genera of Andropogoneae closely 1' esemble one another.

Howeve 1'， Phalarideae sens. st 1'. and Ehrharteae show distinction in these features.

We thus can not attach the same importance to the ch 1' omosomal differences found among the genera of Andropogoneae as those found among Phalarideae. Consider a:- ble ble variation in chromosome size may have occurred during the course of evolution within within Andropogoneae. Elyonurus apparently falls within Andropogoneae ，although this this characterized by large chromosomes. 4) 4) Paniceae and Andropogoneae share many cornrnon features in respect to chromosomes and leaf structure. Also ， the spikelets in the two tribes are similar in in having a tendency to an aborted or rudimentary floret borne below the fertile floret ，the spikelets being dorsally compressed and having a disarticulation below the the glumes.ιThe fo l1 owing differences ，however ，may be found in their spikelet characterist Ic s.

-'2 7- 284 植物研究雑誌、第 32 巻第9 号 4 昭和 32 年 9 月

Paniceae Andropogoneae

Spikelets Spikelets usually not in 山 pairs ， or in Spikelets usuallγ 戸 being in pairs ，one paxrs. paxrs. of each pair sessile and the other pediceled. pediceled. First First glume shorter than the spikele t. First glume longer than the spikelet. Lower lemma similar in texture to Lower lemma usually similar to the glumes. upper lemma in texture. Upper lemma usually indurated ， often Upper lemma membranaceous or hya- chartaceous to crustaceous or at least line ， thinner than the glumes ， usually firmer firmer than the glumes ， usually with a geniculate awn from the entire awnless. awnless. tip or the sinus of the 2-1obed tip.

れihile Maydeae are distinguished from Andropogoneae in several characters 0 1' external external morphology ， the former may be considered to be the advanced group as

compa l" ed with the latte r. As for morphological di 百'erences beh¥'een Andropogoneae and Paniceae described above ，such supposition is untenable. In other words ， the phylogenetic phylogenetic relation between Maydeae and Andropogoneae is 'linear' but that between Andropogoneae and Paniceae may. be ‘dichotomous .' t1 is interesting to

find find di 百erences in the distribution of the members belonging. to ふ Andropogoneae and those belonging to Paniceae. Apart from some naturalized plants ， there are only a few native Andropogoneae in the New World ，and the.large majority of the genera of this tribe 冶re distributed in the Old World. On the contraη ， enclemic enclemic genera of Paniceae are abunclant in both the New and the Old World. Hartley (1950) mentions that Anclropogoneae have not yet attained full develop- ment in the New Worlcl ，and they may represent a natural tribe of relatively recent recent origin. In spite of such clifferences the two tribes are undoubtedly more or less less dosely relatecl. If the family Poaceae should not be divided into many small subfamilies ，Paniceae and Andropogoneae ，with other small tribes ，could be included in the same subfamily.

5) 5) Groups exhibiting various primitive features in external ~orphology ， such as as Anomochloeae ，Parianeae ，Bambuseae ，Streptochaeteae ，Olyreae ，Oryzeae and Phy I1 orachieae ， share many common features in their Ieaf structures. .The anatomy of the leaves of Bambuseae ，Olyreae ，Parianeae ， Streptocbateae and Anomochloeae

、¥vas carriecl out by various investigators ， especially by Page (1947). She has found that that these grasses share enlargecl mesophyll cells and the chlorechyma being ar-

ranged in layers above and below the enlarged cells. ~ (195 匂) ha v. e exa 叩ined the leaf leaf structure of Phyllorachis sagittata Trim. (Phyllorachieae). This species l;:!cks

- 28 ー September 1957 Journ. .J ap. Bo t. Vo l. 32 No. 9 285 the the enlarged' mesciphyll cells ，but it has processes of cell merribrane in the assimila- tive tive parenchyma.' and :the outer bundle sheath without chloroplasts: these features are are very common in the members of the grou J) s mentioned above. Although Prat (1 936) has referred Oryza in its anatomical features of leaves to the Festucoid group ， they are not typically festucoid but resemble rather those found in leaves of Bain- busoids. busoids. Prat has exa m. ined only the genus Ory: 't a. Reports concerning the leaf structures structures of other Oryzeae by subsequent authors are rather numerous. All members of Oryzeae ，except for Chikusichloa aquatic αKoidz. ， possess cell wall pro- cesses cesses in the ass "I milative parenchyma; also the outer bundle sheath is generally well-developed ，and never contains chloroplasts ， similarly as the leaves of Bambuseae. Similar Similar features may be also found in Arundineae (c f. Tateoka 1956b) and Thysa- nolaeneae (c f. Tateoka 19561). Arundineae a問 believed to be of very ancient oogin because the fussil plants belonging to Arundo and Phragmites have been found in the the strata of the cretaceous period. The epidermal characteristics of the Panicoid type type is shared by all the plants belonging to the groups mentioned above ¥vithout exception. exception. All of these groupsa 1' e apparently of ancient origin ，and the Bambuseae- like like leaf structure ch a:- racterizes all of these groups.

Acknowledgemen.-:... Acknowledgemen.-:... It is a great pleasure to record here a debt of g 1' atitude to seve 1' al pe 1' sons for their helps during the course of the ptesent investigations. 恥fy gratitude gratitude is fitst due t6 Dr. J. Ohwi (Na t. Science Mus. ，Tokyo) for his interests and advices throughout the studies. To Dr. Y. Takenaka (Nat. Inst. Gen. ，_ Mishi- ma) 1 owe specical thanks for his guidance on cytology. My thanks are also due to to Pro f. Dr. H. Ono (Tokyo Met 1' opolitan Univ.) ，Dr~ E. Potztal (Bot. Mus. ， Berlin-Dahlem) ， Prof. Dr. F. Maekawa (Tokyo Univ.) ，Mr. K. Kasaki (Tokyo

Met 1' opolitan Univ.) and Mr. B. Sakai (Tokyo Metropolitan Univ.) for their kindness and advices on 'many matters. Furthe 1'， 1 am much indebted to M r. T. Koγama

(Tokyo Univ.) and Mr. M. Mizushima (Res. Inst. fo 1' Nat. Resources ，Tokyo) for their their helps ，on the matters of . Also ，1 wish to express he 1' e mγgratitude to to all persohs and ins tI tutions to whom 1 owe various seed samples.

訂正 1) 筆者は，第 6 報に沿いて， Aristideae を Eragrostoideae に含めたが，乙れ

は Arundoideae trC人れた方がより妥当のようである。 Aristideae が純粋の Eragro ・』 stoideae から違っている点として，葉の表皮の'2 細胞性の毛が棒状である乙と，染色体 基本数が 11 であるとと，外頴(l emma) のさが columna と subula にはっきりわかれて いること，をあげるととができる。 Stipeae とは葉の横断面の特徴にb いてはっきり具.

~， 29 ー 286 植物研究雑誌第32 巻第9 号 昭和32 年 9 月

在るが， Aristideae にみられるよう左，葉緑体を多分に含む outer bundle sheath は， Eragrostoideae ，Panicoideae のみならず， Arundoideae K 沿いても.Streptogyneae ， Arundinelleae ，Garnotieae などでみられている。 2) 第 9 報に沿いて，筆者は Lygeeae を非常に古いものの l つとして扱ったが，乙 れはそれ工りやや新しい Arundoideae の一員と考えた方が妥当の工うである。

Literature cited

Avdulov ，N. 193 1. Karyosystematische Untersuchung der Familie Gramineen. Bul 1. App l. Bot. Gen. etc. ，Supp l. 44: 1-428. Beetle ，A. A. 1955. The four sub- families families of the Gramineae. Bull. Torr. Bot. Club 82: 176-197. Brown ，W. V. 195 1. Chromosome numbers of some Texas grasses. Bull. Torr. Bot. Club 78: 292-299. De Wet ， J. M. J. 1954. The genus Dant 五onia in grass phylogeny. Amer. Jour. Bo t. 41: 20ι21 1. 1956. Leaf anatomγand phylogeny in the tribe DanthonIae. Ibid. Ibid. 43: 175-182. Hansen ， 1. and E. Potztal1954. Beitrage zur Anatomie und Sys- tematik tematik der Leptureae. Bot. Jb. 76: 251-270. Hartley ，W. 1950. The global dis.' tribution tribution of tribes of the Gramineae in relation to historical and enviro I1 mental factors. Australian Australian Jour. Agr. Res. 1: 355-373. Hubbard ， C. E. 1947. Drake-Brockmannia somalensis somalensis Stapf. Hook. Ic. P l. Tab. 3455. 1948. The genera of Bri- tish tish Grasses. In J. Hutch.'s British F l. P l. 1954. Grasses. Penguin Books ，A 295. London. Ohwi ， J. 1942. Gramina Japonica IV. Acta Phytotax.

Geobot. 11: 261-274. Page ，V. M. 1947. Leaf anatomy of St れゆ的chaeta and the relation relation of this genus to the . Bul l. Torr. Bo t. Club 74: 232-239. Pilger ， R. 1954. Das System der Gramineae. Bot. Jb. 76: 281-384. Potztal ， E. 1952. uber die Blattanatomie der Isachneae. Bot. Jb. 75: 551- 回 9. 1953. uber die Anatomie von Micraira subulifolia F. Muel l. lbid. 76: 134-138. Prat ‘ H. 1936. La systematique des Graminees. Ann. des Sci. Nat. ，Bot. ，lO e serie ，18: 165- 258. 258. Reeder ， J. R. 1953. A 伍niti e- s of the grass genus Beckmannia Host. Bull. Torr. Torr. Bot. Club 80: 187-196. Stebbins ，G. L. 1956. Cytogenet Ic s and evolution of the the grass family. Amer. Jour. Bot. 43: 890-905. Tateoka ，T. 1956a. Reexamina- tion tion of anatomical characteristics of the leaves in Eragrostoideae and Panicoideae (Poaceae). (Poaceae). (J apanese) Jour. Jap. Bot. 31: 210-218. 1956b. On the systematic systematic position of the genus Cleistogenes with special reference to the characte- rIstics rIstics of leaf structure. (J apanese) Ibid. 31: 326-332. 1956c. Miscel- laneous laneous p叩町s on grass phylogeny (1). On the tribe Phalarideae. (J apanese) Ibid. 31: ト5. 1956d. Ditto (2). On the tribe Aveneae. (J apanese) Ibid.

- 30- September 1957 Journ. Jap. Bo t. Vo l. 32 No. 9 287

31: 84-90. 1956e. Ditto (3). On Fcstuceae (Part 1) ， Triticeae and Monermeae. (J apanese) Ibid. 31: 179-186. 1956 f. Ditto (4). On the tribe tribe Festuceas (Part 2). (J apanese) Ibid. 31: 241-248. 1956g. Ditto (5). (5). Phylogenetic considerations on the subfamily Festucoideae. (J apanese) Ibid. 31: 267-272. 267-272. 1956h. Miscellaneous papers on the phylogeny of Gra- mineae (6). A review of the subfam i1 y Eragrostoideae. (J apanese) Ibid. 31: 307- 315. 315. 1956 i. Ditto (7). A review of Panicoideae. (J apanese) Ibid. 31: 353-358. 353-358. 195 句. The place of the gcnus 1ヲiyllorachis in the system of of Gramineae. Bot. Mag. Tokyo 69: 83-86. 1956k. Notes on some gr 出 ses 1. Ibid. 69: 311-315. 1956 1. Ditto II. Ibid. 69: 336-339. 1957a. 1957a. Miscellaneous papers on the phylogeny of Poaceae (8).On the subfamily subfamily Andropogo 田 ae. (J apanese) Jour. Jap. Bot. 32: 11-17. 1957b.

Ditto Ditto (9). On the groups having primitive features of ex 1. ernal morphology ，es 圃 peciall y' ，Anomochloideae ， ， 01yroideae and Bambusoideae.

(J apanese) Ibid. 32: 42-49. 1957c. Notes on some grass 回 III. Bot. Mag. Tokyo 70: 8-12. 1957d. Ditto IV. Jour. Jap. Bot. 32: 111-114. 1957e. 1957e. Ditto V. Bot. Mag. Tokyo 70: 115-117. 1957 f. Ditto Ditto V 1. Ibid. 70: 119-125.

誤一行一同 IE IE 表 (Errata)

巻(号) I 頁| EL 誤 (Vo l. No.)1 (Page) I (For) I (R(Read) '!，

31 31 (1) 下から 11 行目 含有種類 含有種数 31 31 (6) 180 I Table 1 の Sesleriinae 。γ'e ochlloa Oreochloa " 185 I上から 2 行目 Maingurtus Merill 哩u: γus 31 31 (8) 241 I Table 1 の上から 3 ツ目 IFind 'e rhuthia Fingerhuthia (Table 1，3 froin top)

31 31 (10) 307 I ~本文，上から 5 行自 |回時 当時

JJ JJ グ I Table 1 の下から 2 行目 12n= 1O 2n=20 (Table 1，2 from bottom)

11 11 I 310 I 長時 I Brywn (1955) I Brown (1955) 31 31 (1 2) I 354 I Table 1 の Anthephoreae I 小穏は羽を |小穂は群を

11 11 I 355 I下から 13 行目 IOhwi IOhwi

11 11 I 356 I Table 1の Paniceae の 2ツ目 1 Lariacis 1: Lasiacis (Table 1，3 from top)

- 31-