植物研究雑誌 J. J. Jpn. Bot. 66: 66: 319-351 (1991)

Anatomical Structure of Fruits and Evolution of the Tribe Sorbeae in the Subfamily ()

1 Hiroyuki IKETANI ) and Hiroyoshi OHASHI

Biological Biological Institute ，Faculty of Science ，Tohoku University ， Sendai ，980 ，JAP AN

パラ科ナシ亜科ナナカマド連における果実の解剖学的構造と進化

池谷祐幸 1) ，大橋広好

東北大学理学部生物学教室 980 仙台市青葉区荒巻字青葉

(Received (Received on July 6， 1991)

Structures Structures of fruits of 61 species of 17 genera of the tribe Sorbeae (Rosaceae-Maloideae) 訂 e anatomically anatomically examined. Structure of the pulp is different at the generic level in Sorbeae. Nearly half of of the genera examined have a homogeneous pulp structure consisting of parenchyma and few or no stone stone cells. In contrast ， the other genera have several kinds of heterogeneous pulp structure consisting of of parenchyma ， stone cells ， and pigment cells. The inner epidermis does not be sclerified in most of genera ， but in some genera it becomes sclerenchymatous. Hypothetical Hypothetical evolutionary trends of the structure of the fruits in Maloideae are proposed. The fruit fruit of Crataegeae is inferred as more primitive than that of Sorbeae. Pulp structure of Sorbeae may have have been evolved through loss of sclerenchymatous cells and/or through specialized distributional pattem pattem of sclerenchymatous cells and pigment cells ，e. g. ，clusters of stone cells in the center of the flesh flesh (Pourthiaea) ，clusters of large sized pigment cells and/or stone cells in the whole part of the flesh (Micromeles (Micromeles and subgenera Aria and Chamaemespilus of Sorbus) ， and clusters of stone cells in the inner inner flesh (subgenus Torminalia of Sorbus). Phylogenetic relationships of the genera of Sorbeae are inferred inferred based on these evolutionary trends.

Introduction Introduction Decaisne (1874) first recognized the diversity of The subfamily Maloideae is divided into two pulp structure in Sorbeae. He discovered the tribes ， Crataegeae and Sorbeae ， based on heterogenωus nature of the pulp with normal-sized

characters characters of the fruit (Koehne 1890 ，Schluze 回 Menz parenchyma and large cells in Pourthiaea ， 1964 ，Kovanda 1965 ，Phipps et al. 1990). In Micromeles ， and subgenus Aria (including Crataegeae almost all parts of the carpel become Chamaemespilus) of Sorbus. This was not bony in the mature fruit and the fruit becomes evaluated as a taxonomic character until the middle drupe-like. drupe-like. This does not occur in the fruits of the of this century. Gabrielian (l958) and Kovanda tribe tribe Sorbeae. (1 961) re-examined anatomical characters of fruits

-319 一 320 320 植物研究雑誌第66 巻第6 号 平成 3 年12 月 of of European species of Sorbus and concluded that distribution of stone cells ，and structure of the pulp pulp structure is very useful for distinguishing the inner epidermis and neighboring cells of the frui t. subgenera subgenera of Sorbus. However ，since they did not Based on these data we w i1l attempt a systematic investigate investigate Pourthiaea ，Micromeles ， and the As iatic evaluation of these characters in this tribe. species species of Sorbus ，it was not clear whether the pulp structure structure is truly distinctive 町nong the subgenera Materials and methods of of Sorbus or not ，and whether the heterogeneous Seventy-three species belonging to 24 genera of nature nature of the pulp of these three taxa is homo- the Maloideae ， of which 61 species of 17 genera logous logous or no t. in the tribe Sorbeae and 12 species of 7 genera in It It is well known that the degree of density of the tribe Crataegeae ， were studied. Fruits examined stone stone cells in the pulp is a useful character for were either fresh or dried ， or from herbarium distinguishing distinguishing some genera of Sorbeae. For specimens. Sources of materials and voucher example ，there are many stone cells in the pulp of specimens are listed in the appendix.

Pyrus ， but only a few or no stone cells in that of Fresh 仕凶 ts we 児 fixed inFAAσive parts stock M alus ， with the exception of sections Docyniopsis fOFmalin; five parts glacial acetic acid; 90 parts and Eriolobus 侭ehder 1940). Structures of the 70 0/ 0 ethanol). Dried fruits were softened with a innermost innermost part of the fruit flesh have also been solution containing equal volumes of glycerol and recognized recognized as different 出 nong genera. It has been 10% aerosol-OT for about a week at room described described as cartilaginous ， membranous ， or temperature. After these treatments all materials leathery leathery (e.g. ， Hutchinson 1964 ， Robertson et al. from fresh or dried fruits were softened with 15% 1991). 1991). However ， except for some species with hydrofluoric acid for about two weeks at room edible edible fruits ， these characters have not been well temperature. Then they were washed with running investigated. investigated. water and dehydrated through 担任buthanol series As mentioned above ，anatomical characters of and embedded in paraplast (melting point the the fruit of Sorbeae have been recognized as useful 57-58 0 C) for microtoming. Some exceptionally taxonomic taxonomic characters for distinguishing genera hard or large specimens were dehydrated with an

since since the last century ， but they have not been well ethanol 凶 ethyl ether series and embedded in investigated. investigated. Al most all anatomical studies of the celloidin ， or double embedded in celloidin- fruit fruit have been performed in Malus domestica paraplast (Johansen 1940). Paraplast sections and Borkh. Borkh. (e.g. ，MacArthur and Wetmore 1939 ， double embedded sections cut 15-20μm in MacDaniels MacDaniels 1940 ，Tukey and Young 1942 ，Roth thickness were either triple stained with Heiden- 1977) 1977) and Pyrus communis L. (e.g. ， MacDaniels hain's Hematoxylin ， Safranin 0 and Fastgreen 1940 ，Sterling 1954 ，Roth 1977). Except for FCF ， or double stained with Safranin 0 and Gabrielian's Gabrielian's (1 958) and Kovanda's (1 961) works Fastgreen FCF. Celloidin sections cut 50-100μm in in Sorbus ，no anatomical works on the fruits of in thickness were stained with Safranin 0 and Sorbeae Sorbeae have been performed from a systematic Fastgreen FCF. All stained sections were mounted point point of view. in Entellan New (Merck Co. Ltd ふ In In this study ， we report the anatomical Prepared specimens were observed with a characters characters of the fruit of the genera of Sorbeae ， normal optical microscope. Specimens were also with with special attention to the structure of the pulp ， observed in crossed polarized light for the presence Decernber Decernber 1991 Journal of Japanese Botany Vo l. 66 No. 6 321 of of sclerenchymatous cells.

Results Results Terminology The so-called fruit of the sub- f出 nily Maloideae is a fal 田 fruit called a pome that consists consists of carpels and a fleshy enlarged hypan- thium (somet 卸les called calyx tube) surrounding 10 or or fused to them (F ig. 1). Bony p紅白 in the mature 9 fruit fruit of the tribe Crataegeae are sometimes recognized recognized as the endocarp ，and the fruit is called a drupe or pyrene (Lindley 1821 ，Kovanda 1965 ， Fig. Fig. 1. Explanatory scherne of the general structure Kalkman 1973). Actually ，however ， not only the of the fruit of Sorbeae. 1: style. 2: apical part of innermost innermost part of the carpel (true endocarp) but the inner flesh. 3: sepal. 4: a portion of free hypanthiurn. hypanthiurn. 5: inner epiderrnis. 6: dorsal bundle also also almost the other part of carpel becomes bony of the carpe l. 7: innerrnost prorninent bundle of as as described by Sterling (1964) ，and the hypan- the hypanthiurn. 8: epiderrnis. 9: inner flesh. 10: outer outer flesh. Broken line indicates the border thium becomes fleshy (F igs. 2-6). So ，this kind of between the part derived frorn the carpel (inner fruit fruit is not the same as that of true drupe ， as in 目白 h) and that derived frorn the hypanthiurn (out 釘 flesh). flesh). Prunus. Prunus. In the tribe Sorbeae generally both parts of of the fruit derived from the carpel and the hypan- thium become completely fleshy and cannot be separated separated histologically. Fahn (1 990) recognized the the border betw 民 n the dorsal bundles of the carpel and the innermost prominent bundles of the hypanthium. We follow him and we called the outer outer part of this border “ outer flesh" and the Fig. Fig. 2. Schernatic drawing of the fruit structure of inner inner part “ inner flesh" as used by Olson and Crataegeae Crataegeae and Prunus. A: Crataegeae ， B: Prunus. Steeves Steeves (1982). Hy: part derived frorn the hypanthiurn. C: part derived derived frorn the carpe l. L: locule. The dotted part In In true fruits ，exocarp ，mesocarp ，and endocarp represents sclerenchyrnatous tissue. are are often distinguished in the pericarp. In many fruits fruits of Sorbeae ，it is impossible to distinguish pattem varies in the genera. The shape of sclereids mesocarp and exocarp ， because these two parts are isodiametric (= stone cells) in the most part of become fleshy as outer flesh. Moreover ， these three flesh ， but in the inner epidermis and neighboring terms terms are sometimes confused in the fruits of cells it is sometimes radially elongated. In many

Maloideae ，and ， in some cases ，mesocarp is used cas 邸 the apical pa 此 of the inner flesh and the inner as as the whole fleshy part of the pome (Camefort epidermis and neighboring cells have some and Boue 1980). So ， in this study ，we don't use different pulp structure from the other part of these these three terms. frui t. Therefore ，we describe these parts separately. Description Description of the structure of the fruit Degree Hence ，a fruit is divided into the following six of of density of sclereids and pi 伊 lent cells in the flesh p 紅 ts; inner epidermis and neighboring cells ， inner differs differs in genera of Sorbeae. Also ，distributional flesh (except for the apical part of the inner flesh) ， 322 322 植物研究雑誌第66 巻第6 号 平成 3 年12 月

Figs. Figs. 3-6. Fruit of Hesperomeles heterophylla (Gillis et Powman 1023 4). Figs. 3 ，4: Longitudinal section. All parts of the the inner flesh (the part derived from the carpel) become sclerenchymatous ， but the outer flesh (the part derived from the the hypanthium) does no t. X 29. Figs. 5 ，6: Longitudinal section through apical part of the carpe l. Nearly all parts of of the carpellary wall except for the epidermis and neighboring cells become sclerenchymatous. X 29. Figs. 4 and 6: same same field as Figs. 3 and 5，respectively ， seen in crossed polarized ligh t. 0: outer flesh ，i: inner flesh ， cw: carpellary wall ，s: seed.

outer outer flesh ，hypodermis ， apical part of the inner the hypodermis. A portion of free hypanthium and

flesh ，and a portion of fr 民 hypanthium and sepals sepals become fleshy except for the tip of sepals. (Fig. (Fig. 1). Results are summarized in Table 1. Sorbus subgenus Torminalia (Figs. 11-13) Sorbus Sorbus subgenus Sorbus (Figs. 9， 10) Inner epidermis and the neighboring cells are In ner epidermis is not generally sclerified. Stone not generally sclerified. Large clusters of stone cells cells cells appear sparsely or rather densely in the inner appear densely in the inner flesh and in the apical flesh flesh and outer flesh ，and usually appear rather part of the inner flesh. The size of each stone cell densely densely in the apical part of the inner flesh. Pig- is not clearly larger than that of parenchymatous ment cells are generally absent or rare except for cell. Pigment cells are few in the inner flesh but

t + 十+ +十十 Fig. Fig. 7. Schematic drawings of the anatomical structure of inner epidermis and neighboring cells ，classified into five types. types. -: Inner epidermis does not become sclerenchymatous. t: Inner epidermis becomes partly sclerenchymatous. +: Inner epidermis becomes wholly sclerenchymatous. + +: Inner epidermis and neighboring cells become sclerenchymatous. sclerenchymatous. ++ +: In addition to ++ ，inner flesh also becomes completely sclerenchymatous. December December 1991 Journal of Japanese Botany Vo l. 66 No. 6 323

+ ++

③ @⑤ 。 Fig. Fig. 8. Schematic drawings of the fleshy part of the fruit ，classified into eight types. -: Stone cells (or pigment cells) absen t. t: Stone cells (or pigment cells) rare. +: Stone cells (or pigment cells) scattered. ++ : Stone cells (or pigment cells) cells) dense. ++ + : Tissue almost exclusively composed of stone cells (or pigment cells). +: Clusters of stone cells (or (or pigment cells) presen t. + +: Large clusters of stone cells (or pigment cells) presen t. 0: Large clusters of large sized sized parench 戸natous cells presen t.

appear appear densely in the outer flesh. A portion of free a few or rather many stone cells. A portion of free hypanthium partly becomes fleshy but sepals hypanthium and sepals become fleshy except for become shriveled upσig. 13). the tip of sepals ， but in Himalayan species (S.

80rbus 80rbus subgenus Chamaemespilus (Fig. 14) and cuspid α， ta and S. lanata) ，which have almost com- subgenus Aria (Figs. 15-20) pletely inferior ovary ， sepals are persistent but Inner Inner epidermis is not generally sclerified. become scarcely fleshy. Clusters Clusters of heterogeneous cells appear in the inner Micromeles (Figs. 21-24) flesh flesh and outer flesh. In 80rbus chamaemespilus ， Inner epidermis is not generally sclerified.

S. S. hayast α， na and S. kusnetzovii ， these clusters Clusters of stone cells appear in the inner flesh consist consist of parenchymatous cells or parenchymatous similar to those of subgenera Aria and Chamae- pigment cells (Figs. 14-16) or only a few mespilus of 8orbus. In the inner flesh these clusters

stone stone cells appear 泊 these clusters of the a問 usually composed of stone cells which usually inner inner flesh besides pigment cells and parenchyma. have vacuoles rich in pigments and in the outer In In the Asiatic species of Aria (8. cuspidat α， flesh they are mainly composed of sclerenchy- 8. 8. lan αtαand S. xanthoneura) ，clusters of the matous pigment cells. A portion of free hypan- inner inner flesh consist mainly of stone cells and thium and sepals fall off after anthesis.

those those of the outer flesh consist mainly of pigment Aroni α(Figs.， 25 ， 26) cells cells (8. xanthoneura; Figs. 17 ， 18) or sclerenchy- Inner epidermis is not sclerified. Stone cells are matous pigment cells (S. cuspidat αand 8.1 α，nat α;， sparsely or rather densely distributed in the inner Figs. Figs. 19 ，20). The apical part of the inner flesh has and outer flesh and rather densely distributed in 324 324 植物研究雑誌第66 巻第6 号 平成 3年12 月

Table Table 1. An atomical characters of the fruit of taxa examined*). Species Species IE 1F OF HD AI F FH Sorbus Sorbus subg. Sorbus 一"， t + S. S. commixta Hed l. + ++ + + t "，+ ++ ++ + ++ ++ S. S. esserteauian αKoehne ++ + + ++ ++ S. S. gr α，cilis C. Koch ++

S. S. matsumurana Koehne + t "，+ + + ++ + S. S. pohu αsh αnensis Hed l. 十+ + + ++ ++ ++ 十 S. S. sambucifoli αRoem. + t + t"，+ ++ ++ ++ + S. S. scαlα ris Koehne ++ + + Sorbus Sorbus subg. Torminalia EDED EDED + not S. S. torminalis C r. ++ t + +++ +++ + fleshy Sorbus Sorbus subg. Chamaemespilus (ED) (ED) - ND ND S. S. ch αmaemespilus Cr. @ @ ++ ++ ++ Sorbus Sorbus subg. Ar ia @ S. ωspid αtαHed l. EDED @ ++ @ ED ED 0 。 ++ 。 S. S. hαyα st α na Gabτ.

(ED) (ED) 0 。 。 S. S. kusnetzovii Zinser l.

一"， t @ + ++ S. S. lanat αC.， Koch @ @ ++ EDED EDED ++ ++ S. S. xanthoneura Rehder EDED EDED ++ + + Micromeles Micromeles EDED EDED ++ M. alnifolia C. Koch deciduous EDEB EDEB @ + t @ ND M. caloneura Stapf deciduous @ EDED 十+ ND ND EDED + ND M. corymbifera Kalkman deciduous ND EDED + ND * For key to abbreviations and symbols ， see end of Table 1. ecern D ecern ber 1991 Journal of Japanese Botany Vo l. 66 No. 6 325

Table Table 1.-continued.

Species Species IE 1F OF HD AIF FH EBED EBED 一"'EBEB ++ M. folgneri Schneid. deciduous $ED 十+ EDED EDED + ++ M. japonica Dcne. deciduous EBED EBED + + Aronia Aronia +rv+ 十+ ""'+ + ++ + A. αrbutifolia Elliot 十 十+ ++ ++ ++ t ""，+ t"，+ - "'+ ++ + A. melanocarpa Elliot ++ ++ ++ ++ "'" + "'" ++ t "'+ + 一'"'-'+ (+) + A. prunifolia Rehder ++ ++ ++ ++ Photinil α +'"'-'++ +'"'-'++ ++十 + + +++ + P. P. davidsoniae Rehd. et Wils. + ++ + + (+) (+) P. P. glabra Maxim.

+"'+十 +++ ++ +++ ND integrifolia P. integrifolia Li nd l. ++ 十 ++ 一"'+ + t + + P. P. prunifolia Li nd l. 一"'++ 一"'+ + '" + + -...... ，+ + + ++ ++ serratifolia P. serratifolia Kalkman + +"'++ (十) P. P. stenophylla Hand.-Mazz. + 十 一~十 P. P. wrigh Ui αnαMaxim. ++ 十

SUiα nv α'esi α， 十+ ++ ++ 十+ ++ S.α mphidoxa Schneid.

+ (+) S. S. dαvidi α，nαDcne.

-"'+ +++ +++ + S. S. nussi αDcne.，

(+) (+) S. S. undulat αDcne

Heteromeles t '" + t "，+ +十 arbutifolia H. arbutifolia Roem. + "'+ + 一~十 +++ 十+ ++ ++ 326 326 植物研究雑誌第66 巻第6号 平成 3年12 月

Table Table l-continued. Species; Species; IE 1F OF HD AIF FH Pourthiae α P. αrguta Dcne. (ED) ED +++ ++ var. var. salicifolia Iketani et Ohashi +十 @ +++ +"'ED P. P. bel αuverdiana Hatus. 十 + 一"'+ @ +++ ED rv+ + 一一一一 va r. notabilis Hatus. +"'++ + P. P. impressiven α @ @ @ Iketani Iketani et Ohashi ++ @ ++rv Efj @ P. P. villo s， αDcne. + 十 一一一一 var. parvifolil α @ + rv+ + +rvED Iketani Iketani et Ohashi 一"-'t -"-'+ ++ -rv+ Eriobot りIa t ++ ++ ++ E. E. bengalensis Hook. f. 一""'+ ++ + + + + + ++ ++ 一'"'-'t ++ + E. E. cavaleriei Rehder + + + + + + + + E. E. deflexa Nakai ++ 十 + ++ + + + ++ + E. E. frl α grans Champ. ++ ++ ++ + t + + E. E. japonic αLind l. ++ t ++ ++ + +'"""'++ +'"""'++ +rv++ ++ +rv++ E. E. salwinensis Hand.-Mazz. + +"'++ +rv++ ++ ++ + Rhaphiolepis R. R. indica Li nd l. + ++ + ++ deciduous deciduous var. var. umbellata Ohashi + ++ + + + ND R. R. ferruginea Metcalf deciduous + + ++ ND Ch αenomeles ++ ++ ++ C. 伊 eciosa Nakai deciduous + ND Pseudocydonia Pseudocydonia @ @ ++ P. P. sinensis Koehne deciduous + ++ ND De 氾ember 1991 Journal of Japanese Botany Vo l. 66 No. 6 327

Table Table l-continued. Species Species IE 1F OF HD AIF FH Cydonia Cydonia @ @ ++ not C. C. oblonga 恥1ill. rv+ rv+ + fleshy Docynia Docynia ++ @ @ + + ND D. indic αDcne. + 十 ++ + ND Pyrus Pyrus @ ED ED rv+ + + P. P. pyrifolia Nakai deciduous + + ++ + Malus M. b αccata Des f. ND ND deciduous deciduous var. var. sibirica Schneid. ND ++ + ND ++ not M. domestica Borkh. 'AKANE' + fleshy 十+ + ND M. floribunda Sieb. deciduous + ND ++ + not M. prunifolia Borkh. + fleshy + not M. toringo Vriese fleshy fleshy ++ EDrv++ EDrv++ +rv++ ++ not M. tschonoskii Schneid. + 十 ++ 十十 fleshy Amel αnchier trv+ + +'"'-'+ + + + A. asiatica Walp. ++ trv + + + A. sanguinea DC. trv+ ++ t

一~十 + + A. spicata C. Koch 十+ Peraphyllum Peraphyllum t t + ND P. P. ramosissimum Nut t. + + ND

Abbreviations. Abbreviations. IE: inner epidermis ， IF: inner flesh ，OF: outer flesh ，HD: hypodermis ，AIF: apical part part of the inner flesh ，FH: a portion of free hypanthium. In In each taxon ， upper symbols indicate the distributional pattern of stone cells and lower symbols indicate indicate that of pigment cells. The symbols correspond to those of Fig. 7 and Fig. 8. Columns with “ND" are not observed mainly due to bad preparation of samples. 328 328 植物研究雑誌第66 巻第6 号 平成 3 年12 月

Figs. Figs. 9- 14. Fruit of subg. Sorbus ， subg. Torminalia and subg. Chamaem 田 pilus of Sorbus. Figs. 9， 10: S. sambucifolia (l ketani 1215). Longitudinal section. x 19. Figs. 11-13: S. torminalis (仕om Bo t. Gard. Vrije Univ.). Longitudinal section section (Figs. 11 ， 12) and the apical part ofthe inner flesh and style (Fig. 13). The separation between epidermis and flesh flesh is an artifact (Figs. 11 ， 12). Sepal is persistent but shriveled up (Fig. 13). x 19. Fig. 14: S. chamaeme. 伊 i1 us (from ard. Jard. Bot. alpine “La Jaysinia"). Longitudinal section. The separation between epidermis and flesh is an artifact. x 19. Figs. 10 and 12: same field as Figs. 9 and 11 ，respectively ， seen in crossed polarized ligh t. e: epidermis ，i: inner epidermis ，c: seed coat ，s: style ，1: locule ， sp: sepa l. the the apical part of the inner flesh. Pigment cells are inner flesh ， but in P. davidsoniae and P. integri- few few in the inner flesh but rather many in the outer foU αtheyapp 伺 r densely (Figs. 29 ，30). Stone cells flesh. flesh. A portion of free hypanthium and sepals are absent or few in the outer flesh ， but they app 伺 r become fleshy except for the tip of sepals. densely in P. integrifolia. Pigment cells 訂 e usually

Photini α(Figs.， 27-36) absen t. The apical part of the inner flesh have no ， Inner Inner epidermis and neighboring severallayers a few or rather many stone cells. A portion of free of of cells are sclerified except for glabra hypanthium and sepals becomes fleshy except for (Fig. (Fig. 31). The shape of these sclereids is isometric the tip of sepals. σigs. 27 ，33 ， 34) or radially elongated (Figs. 35 ， 8tranvaesi α(F igs. 37-40)

36). 36). Stone cells 紅 e usually absent or few in the Inner epidermis is not sclerified (8.α mphidoxa) ， December December 1991 Journal of Japanese Botany Vo l. 66 No. 6 329

Figs. Figs. 15-20. Fruit of subg. Aria of Sorbus. Figs. 15 ， 16: S. hayastana (Karaprtjen s.n.). Longitudinal section. x 19. Figs. Figs. 17 ， 18: S. xanthoneura (1 980 Sino ・Americ. Exped. 719). Longitudinal section. x 19. Figs. 19 ， 20: S. lanata (K itamura s.n.). s.n.). Longitudinal section. x 19. Figs. 16 ， 18 and 20: same field as Figs. 15 ， 17 and 19 ，respectively ，seen in crossed polarized polarized ligh t. e: epidermis ，i: inner epidermis ，c: seed coa t. partly partly sclerified (8. nussi ι 8. undulata) ， or com- Heteromeles (Figs. 41 ， 42) pletely pletely sclerified (8. davidiana). The shape of these Inner epidermis is not sclerified. Stone cells are sclereids sclereids is either isodiametric or radially elongated. absent or appear sp 紅 sely in the flesh. Pigment cells

Stone Stone cells are usually absent or appear sparsely appear densely in almost all pa 此 of the flesh. A in in the inner flesh ， but in 8. amphidoxa and 8. portion of free hypanthium and sepals become nussia nussia stone cells appear densely in the inner flesh. fleshy except for the tip of sepals. In In the outer flesh stone cells are absent or appear Pourthiaea (Figs. 43-48) sparsely ， but in 8.α mphidoxa they appear densely. Inner epidermis is not sclerified. Clusters 0 f Pigment cells are usually absen t. A portion of free stone cells appear at about the middle part of the hypanthium and sepals become fleshy except for flesh. The size of each stone cell is not clearly larger the the tip of sepals. than that of parenchymatous cell. Stone cells don't 330 330 植物研究雑誌第66 巻第6 号 平成 3 年12 月

Figs. Figs. 21-26. Fruit of Micromeles and Aronia. Figs. 21 ， 22: M. alnifolia (Iketani 201 の. Transverse section. x 19. Figs. Figs. 23 ， 24: M. japonica (l ketani 1663). Longitudinal section (Fig. 23) and the apical part of the inner flesh and style style (Fig. 24). Arrow in Fig. 24 indicates the top of hypanthium after the falling of the free hypanthium and sepals. x 19. Figs. 25 ， 26: A. prunifolia (Iketani 2019). Longitudinal section. x 19. Figs. 22 and 26: same field as Figs. 21 and 25 ，respectively ， seen in crossed polarized ligh t. e: epidermis ，i: inner epidermis ，c: seed coat ，s: style ，sd: seed ， 1: locule.

appear appear in the other pa 此 of the flesh. Pigment cells hypanthium usually becomes fleshy but sepals in are are few or absent and do not appear 担 the clusters some species don' t. of of stone ce11s. The apical pa 此 of the inner flesh Rhaphiolepis (Figs. 51 ， 52) usua11y usua11y has many stone ∞l1 s. A portion of free In ner epidermis is not sclerified. Stone cells and hypanthium and sepals become fleshy except for pigment cells are absent or appear sparsely in the the the tip of sepals. inner flesh and outer flesh. The apical part of the Eriobotrya Eriobotrya (Figs. 49 ， 50) inner flesh has many stone cells. A portion of free In ner epidermis is not sclerified. Stone cells and hypanthium and sepals fall off after anthesis. pigment pigment cells are absent or appear sparsely or Ch α enomeles (Figs. 53-56) rather rather densely in the inner flesh ， outer flesh and Inner epidermis is not sclerified. Small clusters the the apical part of the inner flesh. A portion of free of stone cells appe 紅 rather densely near the dorsal December 1991 Journal of Japanese Botany Vo l. 66 No. 6 331

Figs. Figs. 27-36. Fruit of Photinia. Figs. 27 ， 28: P. serratijolia (Iketani 202 乃. Longitudinal section. x 45. Figs. 29 ， 30: P. P. davidsoniae(Tang20660). Longitudinal section. x29. Fig. 31: P. glabr α(Iketsu 117 4). Longitudinal section. x 45. Fig. Fig. 32: P. serratijolia (Iketani 202 乃. The apical part of the inner flesh and style. x 45. Figs. 33 ， 34: P. stenophylla (Tagawa et al. T51 4). Inner epidermis and neighboring cells. x 88. Figs. 35 ， 36: P. wrightiana (Yamazaki 1675). Inner

epidermis epidermis and neighboring cells. x 88. Figs. 28 ，30 ，34 and 36: s担 ne field as Figs. 27 ，29 ， 33 and 35 ，respectively ， seen seen in crossed polarized ligh t. e: epidermis ，i: inner epidermis ，s: style ，m: mucilaginous cell of the seed coa t. bundles bundles of the ovary. Stone cells are absent from many stone cells. A portion of free hypanthium other other parts of the inner flesh and outer flesh. and sepals fall off after anthesis. Pigment Pigment cells and a few stone cells appear in the Pseudocydonia (Figs. 57-60) hypodermis. hypodermis. The apical part of the inner flesh has Inner epidermis is not sclerified. No stone cells 332 332 植物研究雑誌第66 巻第6 号 平成 3 年12 月

Figs. Figs. 37-42. Fruit of Stranvaesia and Heteromeles. Figs. 37 ， 38: S. davidiana (l ketani 202 の. Transverse section. X 19. Figs. Figs. 39 ， 40: S. amphidoxa (Teng 90474). Longitudinal section. X 19. Figs. 41 ， 42: H. arbutijolia (Hiroe 10044). X 49. Figs. Figs. 38 ， 40 and 42: same field as Figs. 37 ， 39 and 41 ，respectively ， seen in crossed polarized ligh t. e: epidermis ，i: inner inner epidermis ， m: mucilaginous cell of the seed coa t. are are found in the inner part of the inner flesh ，i. e. Inner epidermis is not sclerified. No stone cells the the ar 伺 inner than the dorsal bundles of the carpe l. are found in the inner part of the inner flesh ，i. e.

But outside of this area clusters of stone cells the ar 伺 inner than the dorsal bundles of the 伺 rpe l. appe 紅 in almost all pa 此 of the inner flesh and But outside of this area ，clusters of stone cells outer outer flesh. These clusters are large and app 伺 r appe 訂 in almost all the parts of the inner flesh and densely densely in the inner part and decrease toward outer flesh. The size and distributional pattern of outside outside of the frui t. Pigment cells appear only in these clusters are similar to those of Pseudo- the the hypodermis. The apical part of the inner flesh cydonia. Pigment cells appear only in the hypo- has has many stone cells. A portion of free hypanthium dermis. The apical pa 此 of the inner flesh has many and sepals fall off after anthesis. stone cells. A portion of free hypanthium and Cydonia Cydonia (Figs. 61-64) sepals are persistent in the mature fruit but doesn't December December 1991 Journal of Japanese Botany Vo l. 66 No. 6 333

Figs. Figs. 43-48. Fruit of Pourthiaea. Figs. 43 ， 44: P. villosa (l ketani 1662). Transverse section. X 29. Figs. 45 ， 46: P. impressivena impressivena (Taam 14 6). Longitudinal section. X 49. Figs. 47 ， 48: P. villosa var. parvifolia (Young 219). Longitudinal section. section. X 19. Figs. 44 ， 46 and 48: same field as Figs. 43 ， 45 and 47 ，respectively ，seen 恒 crossed polarized light. e: e: epidermis ，i: inner epidermis ，c: coat seed ， 1: locule. become fle ぬy. Pyrus (F igs. 69- 72) Docynia (F igs. 65-68) Inner epidermis is not sclerified. Clusters of Inner Inner epidermis and neighboring cells are stone cells appear in the inner flesh and outer flesh. sclerified. sclerified. The shape of these sclereids is isometric. The size and distributional pattern of these clusters Clusters Clusters of stone cells appear rather densely in the are similar to those of Pseudocydonia. Pigment inner inner flesh and outer flesh. The size and density cells appe 紅 sparsely in the flesh. The apical part of these clusters are not differenf among the part of the inner flesh has many stone cells. Species of the flesh. The apical part of the inner flesh has examined is Pyrus pyrifolia in this study ， but many stone cells. Pigment cells appear sparsely in almost same results of observations are reported the the whole part of flesh. A portion of free hypan- by MacDaniels (1940) ，Sterling (1 954) ，and Roth thium and sepals become fleshy in the matu 問 fruit (1 977) on P. communis. except except for the tip of sepals. M α lus (Figs. 73-80) 334 334 植物研究雑誌第66 巻第6 号 平成 3 年12 月

Figs. Figs. 49-56. Fruit of Eriobotrya ，Rhaphiolepis and Chaenomeles. Figs. 49 ， 50: E. salwinensis (Kanai et al. s.n.). Transverse Transverse section. X 19. Figs. 51 ， 52: R. indica (Iketani 204 乃. Longitudinal section. x 49. Figs. 53-56: C. speciosa (l ketani 2363). Figs. 53 ， 54: Transverse section through outer part of frui t. Figs. 55 ， 56: Transverse section through inner inner part of frui t. Arrow indicates the dorsal bundle of the carpe l. x 19. Figs. 50 ，52 ， 54 and 56: same field as Figs. 49 ，51 ， 53 and 55 ，respectively ， seen in crossed polarized ligh t. e: epidermis ，i: inner epidermis ， c: seed coa t. December 1991 Journal of Japanese Botany Vo l. 66 No. 6 335

Figs. Figs. 57-64. Transverse sections through fru It of Pseudocydonia and Cydonia. Figs. 57-60: P. sinensis (l ketani 201 4). Outer Outer p訂 t (Figs. 57 ， 58) and inner part (Figs. 59 ， 60). x 19. Figs. 61-64: C. oblonga (Purchased ，Sendai). Outer part part (Figs. 61 ， 62) and inner part (Figs. 63 ，64). x 19. Figs. 58 ，60 ， 62 and 64: same field as Figs. 57 ，59 ， 61 and 63 ， respectively ， seen in αossed polarized ligh t. e: epidermis~ i: inner epidermis. 336 336 植物研究雑誌第66 巻第6 号 平成 3 年12 月

Figs. Figs. 65-72. Transverse sections through fruit of Docynia and Pyrus. Figs. 65-68: D. indic α(Hara et al. 630181 の. Outer Outer partσigs. 65 ， 66) and inner partσigs. 67 ，68). x 19. Figs. 69-72: P. pyrifolia (Iketani 1699). Outer partσigs. 69 ， 70) and inner part (Figs. 71 ，72). x 19. Figs. 66 ，68 ， 70 and 72: same field as Figs. 65 ，67 ， 69 and 71 ，respectively ， seen seen in crossed polarized ligh t. e: epidermis ，i: inner epidermis ，c: seed ∞ at. December 1991 Journal of Japanese Botany Vo l. 66 No. 6 337

Figs. Figs. 73-80. Transverse sections through fruit of Malus. Figs. 73 ， 74: M. floribunda (from Nat. Bo t. Gard. ，Glasnevin). x 29. Figs. 75 ， 76: M. domestic α‘ Akane' (Purchased ，Sendai). Inner par t. x 19. Figs. 77-80: M. tschonoskii (Iketani 2155). 2155). Outer part (Figs. 77 ， 78) and inner part (Figs. 79 ，80). x 19. Figs. 74 ，76 ， 78 and 80: same field as Figs. 73 ， 75 ， 77 and 79 ，respectively ， seen in crossed polarized light. e: epidermis ，i: inner epidermis ，c: seed coa t. 338 338 植物研究雑誌第66 巻第6 号 平成 3 年12 月

Inner Inner epidermis and the neighboring cells are Amelanchier (Figs. 81-83 ， 85) sclerified sclerified and usually the shape of these sclereids In ner epidermis is not sclerified. Stone cells is is radially elongated. Stone cells are or absent only appear sparsely or rather densely in the inner and a few in the flesh except for Malus tschonoskii ， outer flesh. Pigment cells appear only in the in in which small clusters of stone cells appear hypodermis. The apical part of the inner flesh has sparsely sparsely or rather densely in the flesh. Pigment cells no ，a few or rather many stone cells. A portion are are absent or appear rarely in the flesh except for of free hypanthium and sepals become fleshy M. tschonoskii ， in which there are a few pigment except for the tip of sepals. Olson and Steeves cells cells in the flesh. A portion of free hypanthium (1982) reported fruit structure of A. alnifolia and and sepals fall off after anthesis or they are per- their results are almost same as our observation. sistent sistent in the mature frui t. Peraphyllum (Figs. 84 ， 86)

Figs.81-86. Figs.81-86. Fruit of Amelanchier and Peraphyllum. Figs. 81 ，82 ， 85: A. sanguinea (from Conce r. Jard. Bo t. Nancy). Longitudinal Longitudinal section (F igs. 81 ，82). X 19. Inner epidermis and neighboring cellsσig.85). X Fig. 190. 83: A. asiatica (I ketani 196 の. Longitudinal section. X 19. Figs. 84 ， 86: P. ramossisimum (Matsumura s.n.). Longitudinal section (Fig. (Fig. 84). X 19. Inner epidermis and neighboring cells (Fig. 86). X 190. Fig. 82: same field as Fig. 81 ，respectively ， seen seen in crossed polarized ligh t. e: epidermis ，i: inner epidermis ，c: seed coat ，m: mucilaginous cell of the seed ∞ at. December December 1991 Journal of Japanese Botany Vo l. 66 No. 6 339

Inner Inner epidermis is not sclerified. Stone cells are Main components of clusters in Micromeles are absent absent or rare in the flesh. Pigment cells appear mainly pigment cells in the outer flesh and stone only only in the hypodermis. The apical part of the cells in the inner flesh. But there are continuous inner inner flesh has stone cells only at the under part forms intermediate between these two cell types. of of styles. A portion of free hypanthium and sepals Vacuoles of stone cells in the inner flesh are usually become fleshy except for the tip of sepals. rich in pigment (Fig. 90) and pigment cells in the

M alacomeles and subgenus Cormus of Sorbus out ぽ flesh have thickened walls (Fig. 89). The flesh were were not examined in this study. Kovanda (1961) of the Asian species of Aria is similar to that of described described the fruit structure of Cormus as follows: Micl ゅmeles (Figs. 87 ，88) ， but the difference “ the pulp is homogeneous with numerous stone between the outer flesh and the inner flesh is

cells ，and the endocarp is very thin and mem 四 so m， ewhat obscure. The flesh of the Caucasian and branous". branous". From this description ，we think that the European species of Aria and Chamaemespilus pulp pulp of Cormus is homogeneous like that of have few or no stone cells ，and most or all cells subgenus subgenus Sorbus of Sorbus ，and that the inner in the clusters are large sized pigment cells or epidermis epidermis and neighboring cells are not sclerified. parenchymatous cells without pigments.. Th erefore ， And we surmise the that pulp structure of M αla- variation of the heterogeneous pulp is morpho- comeles comeles is similar to Amelanchier 如 d Peraphyllum logically and geographically continuous among on the basis of the description of Robertson et al. these three taxa ，and we think this character is (199 1). homologous in them. Systematic Systematic relationships among Micromeles and Discussion Discussion the subgenera of Sorbus are still obscure (Challice Decaisne (1 874) first described the and Kovanda 1978). Challice and Kovanda heterogeneous heterogeneous pulp of subgenus Ar ia of Sorbus. (1978) proposed two phylogenetic schemes for the He reported that it is similar to that of Pourthiaea origin of the subgenera of Sorbus mainly based on and Micromeles. Kovanda (1961) reinvestigated the distribution of phenolic compounds. In their this this structure with the Chekoslovakian species of schemes ， Torminalia evolved from primitive Aria Aria and described the heterogeneous pulp as con- Crataegus ， and Aria and Chamaemespilus sisting sisting of basic parenchymatous cells and large developed from Torminalia (Fig. 4 in Challice and

sized sized cells. He stated that sclereids are not found Kovanda 1978) ，or ，otherwise ，Aria ，Chamaemes 田 in in this heterogeneous pulp in the species he pilus ，and Torminalia evolved independently from examined examined except inter-subgeneric hybrids. But primitive Crataegus (Fig. 5 in Challice and Gabrielian Gabrielian (1958) observed clusters of pigment cells Kovanda 1978). In both schemes ， Sorbus (sens. and stone cells in the pulp of Caucasian species of st r.) and Cormus derived from Aria. Aria. Aria. In the present study ，a few stone cells were It must be possible to suppose either of two observed observed in clusters in the inner flesh in Sorbus hypotheses when we follow their schemes; i. e. ， the hayastana hayastana and S. kusnetzovii; both species are heterogeneous pulp resulted through parallel evolu- distributed distributed from the Caucasus to eastern Europe. tion ， or the heterogeneous pulp is a primitive state In In the Asiatic species of Ar ia stone cells and that have already appeared in primitive Crataegus sclerenchymatous sclerenchymatous pigment cells are present in the from which the pulp structure of Torminalia inner inner flesh or throughout the flesh. evolved. In both hypotheses the pulp structure of 340 340 植物研究雑誌第66 巻第6 号 平成3 年12 月

Figs. Figs. 87-92. Stone cells and pigment cells of Sorbus subg. Aria ，subg. Torminalia ，Micromeles ， aqd PO μrt /;l (aea. Fig. Fig. 87: S. cuspidat α(Stainton et al. 8619). Pigment ce11s in the outer flesh. Ce11 walls are slightly thickened and lignified. Fig. Fig. 88: Ibid. Pigment cells of the inner flesh. Cell wa11s 紅 efairly thickened and lignified. Fig. 89: M. japonica (Iketani 1663). 1663). Pigment cells and stone ce11s of the outer flesh. Stone ce11s are fi11ed with pigmen t. Cell wa11 of the pigment cells cells are thickened and lignified. Fig. 90: 肋 id. Stone 田l1 s of the inner flesh. Stone cells are rich in pigment. Fig. Fig. 91: S. tormina /i s (from Bo t. Gard. Vrije Univ.). Pigment ce11s and stone cells of the inner flesh. Stone Stone cells are filled with pigmen t. The size of the stone ce11s is not larger than that of other ce11s. Cell wall of the pigment pigment cells is not thickened nor lignified. Fig. 92: P. impressivena (Ta 創 n14 の. Stone cells. They do not contain pigment pigment and they do not differ in size from other cells. All x 140.

Sorbus Sorbus and Cormus are supposed to have evolved actually ， the inner epidermis and neighboring cells from that of Aria. In this case we must suppose are not sclerenchymatous ，and clusters of stone reversal reversal evolution from the heterogeneous pulp to cells are distributed only in the outer part of the the the homogeneous pulp. Challice and Kovanda inner flesh. This structure is also unique in the (1978) (1978) assumed the relationship of Torminalia and subfamily ，and we also think this ぉ a derived state. Crataegus Crataegus by the resemblance of structure of the Therefore ，it is difficult to assume a close relation-

innermost innermost part of the pulp ， because they con- ship between Crat α， egus and Torminalia. Whether sidered sidered that of Torminalia as thick and tough. But the evolution of pulp is from the Torminalia type December December 1991 Journal of Japanese Botany Vo l. 66 No. 6 341

to to the Aria type or vice versa is also difficult to is not similar to that of Pourthiae α. ， So ， while assume. assume. Phylogenetic schemes of Challice and Aronia is congeneric with Photinia as stated by

Kovanda (1978) ぽ e not supported from our results Robertson et al. (1991) ，Aronia may have in- on the pulp structure. dependently evolved from the ancestor of Photinia ， Micromeles Micromeles is usually treated as congeneric with and is not monophyletic with Pourthiaea. Sorbus Sorbus (Hedlund 1901 ，Rehder 1940 ，Hutchinson Photinia and Stranvaesia display great diversity 1964 ， McAllister 1986 ， Phipps et al. 1990) ， but as regards to the distribution of sclerenchymatous sometimes treated as a separate genus (Decaisne cells in the pulp structure. This variation can be 1874 ，Koehne 1890 ，Kalkman 1973 ，Kovanda and conveniently classified into the following three Challice Challice 1981). Kovanda and Challice (1981) types: 1) Almost the entire inner flesh becomes postulated postulated the close relationship of Micromeles and sclerenchymatous and in some cases the outer flesh Rhaphiolep お in spite of many morphological also becomes sclerenchymatous (P. davidsonia 乙 differences. differences. They denied the relationship between Figs. 29 ， 30; P. integrifolia; S. αmphidoxa ， Figs. Micromeles Micromeles and Ar ia because free part of hypan- 39 ， 40; and S. nussia). 2) Only the inner epidermis thium and sepals never fall off after anthesis in the and neighboring cells become sclerified (P. pruni- latter. latter. But this character appears in many genera folia; P. serratifolia ，Figs. 27 ， 28; P. stenophylla ， of Maloideae. Especia11y in Malus 如 d Pyrus ， both Figs. 34 ， 35; P. wrightiana ， Figs. 35 ， 36; S. persistent persistent and deciduous forms appears within one davidian α" Figs. 37 ， 38; and S. undulata). 3) Few genus. genus. So ， in this family ， this character is not or no stone cells are present in the pulp and inner necessarily necessarily a synapomorphy and it c組 be assumed epidermis (P. glabra ， Fig. 31). We think these that that Micromeles evolved from an ancestral type of variations have resulted from parallel evolution of Himalayan Aria species which had an almost loss of stone cells ，and do not indicate hetero- inferior inferior ovary and fused styles at the base. geneity of genera. The fruit of Stranvaesia is often Micromeles Micromeles and Ari a may be monophyletic and we described as dehiscent at maturity ，and ， in this agree agree with the treatment of Ar ia including Micro- character it is separated from Photinia (Rehder meles meles (Robertson et al. 1991). Chamaemespilus 1940 ，Hutchinson 1964). Kalkman (1973) denied may have evolved from the ancestor of Aria the dehiscence of the fruit of Stranvaesia ，and we through through loss of sclerification of the large sized ce11s. have also observed no dehiscence in S. dl α vidiana Non discoid homogamous flower and other dis- cultivated in J apan. In fact ， there are no structures tinctive tinctive characters of Chamaemespilus (Robertson such as an abscission layer that indicate the et et al. 1991) may be autoapomorphies. dehiscence in fruits of Stranvaesia. Consequently ， The pulp structure of Pourthiao αshows a these two genera cannot be distinguished by the distinct distinct difference from that of the Aria type. In fruit structure ，nor by any other characters. So ，

Pourthiaea Pourthiaea stone ぼ 11s become clusters ， but pigment we agree with the treatment of Strl αnvaesia included ce11s ce11s or sclerenchymatous pigment cells do not in Photinia ぽalkm 組 1973 ， Robertson et al. 1991). appear appear in these clusters ，and the size of the stone Chaenomeles ，Pseudocydonia ，Docynia ，and cells cells is not much larger than the parenchyma (Fig. Cydonia are recognized by Hutchinson (1 964) ，Y 註 92). 92). So ， the structure is not the same as the hetero- (1 974 ，1984) ，and Kalkman (1 988) as closely related geneous geneous pulp of the Aria type. On the other hand to each other ， or the former three genera are the the pulp structure of Aronia is homogeneous ，and recognized by Koehne (1 891) and Robertson et al. 342 植物研究雑誌第66 巻第6 号 平成 3 年12 月

(1991) (1991) as related to Malus while the last Cydonia rootstock of Maloideae (Gladkova 1972). From the to to Pyrus. In fruit anatomy ， Chaenomeles differs above hypotheses ， the following hypotheses of from the others in the distribution of stone cells evolutionary trends in the fruit in Maloideae may in in the pulp. This observation agrees with the be deduced. description description of Weber (1 964) ，who pointed out that One supposed ancestral type may be an achene Chaenomeles Chaenomeles is similar to the most of Malus in derived from a follicle through the indehiscence of having having few or no stone cells in most part of the the ventral suture (Fig. 93: A). Al l parts of the pulp. pulp. However ， Chaenomeles has many stone cells carpels became sclerified at maturity of the frui t. around around the dorsal bundles of the carrels ，and in Then the hypanthium became f1 eshy and enlarged ， this this respect it differs from Malus. enclosing the carpels. The fruit of Pyracantha is Docynia differs from the others in having a the most primitive in the Maloideae ，consisting of sclerenchymatous sclerenchymatous inner epidermis and neighboring five almost apocarpous carpels fused to a f1 eshy cells. cells. Malus has also the same character; in parti- calyx only half way from the base σig. B). 93: The cular ，M. tschonoskii is very similar to Docynia fruit of Crataegeae is more primitive than that of in in having clusters of stone cells in the pulp. M α lus So .r beae. Robertson et al. (1 991) denied to sec t. Docyniopsis ，where M. tschonoskii belongs ， distinguish the tribes in the Maloideae ，and they and sec t. Eriolobus have many stone cells in the inferred that Crataegus ，Mespilus ，Hesperomeles ， pulp pulp (Rehder 1940). From this point these two and Osteomeles are closely related to Pyrus and sections sections were treated as distinct genera (Koidzumi Malus ，and that Cotoneaster and Pyracantha are 1934 ， Robertson et al. 1991) or as members of similar to Sorb ωand Photinia. Phipps et al. (1991)

Docyni α(Schneider， 1906). inferred that the very hard pyrenes of Crataegus Pyrus Pyrus and Malus have been recognized as and others have been derived from a hardening closely closely related and sometimes they were united into core (that corresponds to our sclerified inner one one genus (F ernand 1947 ， Robertson 1974). In the epidermis and neighboring cells in this paper) like pulp pulp structure ， Pyrus is similar to M. tschonoskii ， that of M α lus rather than a soft pyrene like that but but Pyrus does not have a sclerenchymatous inner of Cotoneaster. But so far as we observed ， the epidermis. epidermis. Pyrus is rather similar to Cydonia and carpels of the fruit of the Crataegeae cosnist of Pseudocydonia Pseudocydonia in anatomical characters of the sclerified isometric cells g. (e. ，Figs. 3- 6)組 d no frui t. histological differences were observed between the Amelanchier Amelanchier and Peraphyllum are considered soft and the hard pyrenes of their sense. We could to to be related through several characters (Jones recognize a difference of thickness of carpel 1945 ， Robertson et al. 1991). They have similar between both pyrenes. So we think it is not im- characteristics characteristics in the pulp structure. possible to think an evolution from soft pyrene to Evolution Evolution of 曲e pulp structure There are two hard pyrene. If adopting the premise of evolution convincing convincing hypotheses on the origin of Maloideae. from “ hardening core" to “ hard pyrene" by One is that the ancestor of Maloideae is an Phipps et al. (1991) ，we must think such a reverse allopolyploid allopolyploid between a spiraeoid-ancestor and a of evolution carpels as once sclerified to fleshy then amygdaloid-ancestor amygdaloid-ancestor (Sax 1932; Stebbins 1950; f1 eshy to sclerified ， so far as thinking spiraeoid Challice Challice 1974 ， 1981; Phipps al. et 1991). The other follicle as the ancestral type. Therefore we think is is that only a spiraeoid ancestor is the original that Crataegeae may be a paraphyletic group December December 1991 Journal of Japanese Botany Vo l. 66 No. 6 343 sharing sharing the sclerified carpels of the fruit as a of Rosaceae. In the cladogram he considers the symplesiomorphy ，and that Sorbeae may be an most plausible (Fig. 1 in Kalkman 1988) ， the monophyletic monophyletic group sharing incompletely or not Maloideae ( and Cydonia group as sclerified sclerified carpels of the fruit as a synapomorphy. Kalkman's operational units) forms a Am ong Sorbeae ， the most primitive fruits are those monophyletic group with Osmaroniae and Kerriae ， in in some species of Photinia and Stranvaesia in and a synapomorphy of this group is the f1 eshy or which which almost all parts of the inner f1 esh and inner juicy pericarp. In this case ，reverse evolution must epidermis epidermis remain sclerenchymatous (Fig. 93: C). be assumed also for the evolution of the fruit of

The density of stone cells in the f1 esh were the Crataegeae. In Kerriae the fruit of N r;? veusi α， diminished diminished from these fruits. Then clusters of stone is a drupe ， but that of Kerria is an achene ， so the cells cells became scattered in the f1 esh. The size of the same assumption is needed. clusters clusters further became smaller ，and solitary stone In comparing the two hypotheses ， the first is cells cells became scattered in the f1 esh or th 句r are nearly the more preferable from the principle of parsi- absent absent 93: (Fig. D). The pulp structure of the mony (Wiley 1981). Therefore ，we hypothesize Pourthiaea Pourthiaea typeσig. 93: E) ， the Micromeles type evolutionary trends of the fruit of Maloideae as σig. 93: F) ，and the Torminalia typeσig. 93: G) in figure 93. are are considered to be apomorphic state. Inner Relationship of the genera of Sorbeae The epidermis epidermis and neighboring cells are supposed to conventional system of Maloideae dividing it into have have evolved from sclerenchymatous to f1 eshy as two tribes ， Crataegeae and Sorbeae ， has not been well well as the f1 esh ， but they may have evolved convincingly proved ye t. But ， as discussed above ， relatively relatively independently from the f1 esh. In many we think Crataegeae a paraphyletic group and cases cases the loss of stone cells in the f1 esh was Sorbeae an monophyletic group. In the Sorbeae correlated correlated with that of inner epidermis and neigh- the fo l1 owing three groups are evidently inferred boring boring cells ， but in some case ， the f1 esh has no as monophyletic from respective characteristics stone stone cells as against sclerenchymatous inner that are considered to be synapomorphies: 1) epidermis epidermis (e.g. ， Malus). Amelanchier ， Peraphyllum ， and Malacomeles The other possib i1i ty for the ancestral type is having false septa in the fruit; 2) subgenera Aria a drupe. Only the innermost part of the carpels and Chamaemespilus of Sorbus and Micromeles became sclerenchymatous in drupes. But if a drupe having heterogeneous pulp structure; and 3) were were adopted as the ancestral type ，it must have Cydoni α， Pseudocydonia ， and Chaenomeles evolved evolved from an achene. In this case ，reverse evolu- having multiovulate carpels. In the last group ，loss tion tion is assumed in the carpels of the Crataegeae. of the clusters of the stone ce l1 s and fusion of floral Even if we accept that a amygdaloid-ancestor took organ ，i. e. ， adnation of carpels to the hypanthium part part in the origin of the Maloideae ，we can assume and connation among carpels and styles ，may have a capsule as in Exochorda ，which ， as suggested by been correlatively occurred. So ，Cydonia may be Goldblatt Goldblatt (1976) ， should be transferred to the the most plesiomorphic ， but Pseudocydonia may ， as the ancestral fruit type in that be intermediate. Chaenomeles must be the most subfamily ， which assumes that a drupe was not apomorphic. Ch α enomeles is considered to be necessarily necessarily the. primitive type. related to Mα lus by apparent similarity of the Kalkman (1988) performed a cladistic analysis structure of fruit (Koehne 1891 ， Robertson et al. 344 植物研究雑誌第66 巻第6 号 平成 3 年12 月

ーーー-・・ー

Fig. Fig. 93. Inferred evolutionary trends in the fruit of Sorbeae. A: supposed ancestral type ， B: most primitive type in Maloideae Maloideae (e.g. ，Pyracanth α') ， C: primitive type in Sorbeae (e.g. ，Photinia ，Str αnv αesia) ， D: derived type ， E: derived type type (Pourthiaea) ，F: derived type (e.g. ，Micromeles) ，G: derived type (Sorbus subg. Torminalia). Black dots represent sclerenchymatous sclerenchymatous parts. Irregular-shaped circles in F represent clusters of sized large cells. For further detailed explanation ，see tex t.

1991). 1991). However ，we cannot support this view based multiovulate characte r. on .the anatomical differences. Relationship of Pourthil αeαand subgenus Torminalia of Sorbus Docynia to other genera is still obscure. Docynia have very distinctive characters inferred as has has three to ten ovules per locule (Robertson et al. autoapomorphies respectively. Pourthiaea is c1 early 1991) 1991) against Chaenomeles and others having distinct and it has evidently evolved from Photinia. much more. As described above ， the anatomical To Torminalia the most closely related group in structure structure of the 仕凶tofDo のmia is similar to Mali ω the Sorbeae may be Aria though there are many especially especially to sect. Docyniopsis (and maybe to sec t. differences between both taxa. Torminalia is con- Eriolobus). Eriolobus). The similarity of the flavonoid sidered to be differentiated from the ancestor of chemistry chemistry of Docynia and Docyniopsis of Malus Aria before the latter acquired the heterogeneous was reported by Williams (1982). We think that pulp. Docynia and Docyniopsis are closely related ， but To infer more deta i1 ed and complete phylo- the the similarity of the pulp structure of sclerenchy- genetic scheme in this tribe is very difficult. As matous inner epidermis is inferred as plesio- pointed out by Phipps et al. (1991) ， the evolution morphy ， so the monophyly of Docynia and of the genera of Maloideae hぉ a possib i1i ty of Docyniopsis Docyniopsis has not been proved. pyrus and polychotomous (reticulate) origin. In this situation Cydonia have a very similar structure of the fruit. normal methodology of reconstruction of phylo- Pyrus may have been branched from the ancestor geny (e.g. ，parsimony method) is not applicable of Cydonia before the latter acquired the without excluding the taxa of hybrid origin before December December 1991 Journal of Japanese Botany Vo l. 66 No. 6 345 analyzing analyzing (Hull 1979 ，Funk 1981). It is necessary Fahn A. 1990. anatomy ， ed. 4. Pergamon to to make clear the existence and pattern of reticulate Press ，Oxf()rd. evolution evolution of genera to infer the phylogeny of Fernald 恥1. L. 1947. Minor transfers in Pyrus. Maloideae Maloideae in the future study. Rhodora 49: 229-233. Funk V. A. 198 1. Special concerns in estimating We thank Drs. K. Sohma ，H. Tohda and Y. plant phylogenies. In Funk V. A. and Brooks Tateishi Tateishi of Tohoku University for helpful discus- D. R. (eds よAdvances in cladistics: Proceeding sions sions about the structure of fruits i We thank also of the first meeting of the Willi Hennig Society. Professor Professor Y. Fukuda of Chiba University for 73-86. New York Bo t. Gard. ，New York. helpful helpful advice and criticisms on the manuscrip t. Gabrielian E. Ts. 1958. The fruit anatomy and We are grateful to D r. D. E. Boufford and Ms. floral morphology of the Caucasian repre- E. E. W. Wood of the Harvard University Herbaria sentatives of the genus Sorbus L. Izves t. Acad. for for reading the manuscript. We are grateful also Nauk. Armen. SSR. ser. Bio 1. 11 (7): 79-89 to to the curators of the following herbaria for giving (in Russian). us us the opportunity to 位置nine their specimens: BH ， Gladkova V. N. 1972. On the origin of subfamily KYO ，TI ，and TNS. Maloideae. Bo t. Z. ， Leningrad. 57: 42-49 (in Russian). Russian). Endnote Goldblatt P. 1976. Cytotaxonomic studies in the 1) 1) Present address: Division of Breeding ，Fruit tribe Quillajeae (Rosaceae). Ann. Missouri Bo t. Tree Tree Research Station ， Ministry of Agriculture ， Gard. 63: 200-206. Forestry Forestry and Fisheries ，Tsukuba ， 305 Japan. Hedlund T. 190 1. Monographie der Gattung 農林水産省果樹試験場 305 茨城県つくば市藤本 Sorbus. Svensk. Ve t. Akad. Handl. 35: 1-147. 2- 1. Hull D. L. 1979. The limits of cladism. Sys t. Zoo 1. 28: 28: 416-440. References References Hutchinson J. 1964. Rosaceae. In The genera of Camefort Camefort H. and Boue H. 1980. Reproduction flowering . 1: 174-216. Oxford Univer-

et et biologie des vegetaux superieurs. Doin ，P 紅 is. sity Press ， Oxford. Challice Challice J. S. 1974. Rosaceae chemotaxonomy J ohansen D. A. 1940. Plant microtechnique. and the origins of the Pomoideae. Bot. J. Linn. McGraw-Hill ，New York. Soc. Soc. 69: 239-259. Jones G. N. 1945. Malacomeles ，a genus of 一一一一 1981. Chemotaxonomic studies in the Mexican and Guatemalan shrubs. Madrono 8: family family Rosaceae and the evolutionary origins 33-39. of of the subfamily 恥1aloideae. Preslia 53: Kalkman C. 1973. The Malesian species of the 289-304. 289-304. subfamily Maloideae (Rosaceae). Blumea 21: 一一一一一 and Kovanda M. 1978. Flavonoids as 413-442. markers markers of taxonomic relationships in the genus 一一一一一 1988. The phylogeny of the Rosaceae. Bo t. Sorbus Sorbus in Europe. Preslia 50: 305-320. J. Li nn. Soc. 98: 37-59. Decaisne Decaisne M. J. 1874. Memoire sur famille des Koehne E. 1890. Die Gattungen der Pomaceen. pomacees. pomacees. Nouv. Ar ch. Mus. His t. Na t. Paris Wissenschaftliche Beitrage zum Programm des 10: 10: 113-192 ， p 1. 1-15. Falk-Realgymnasiums zu Berlin. N r. 95: 1-33 ， 346 植物研究雑誌第66 巻第6 号 平成 3 年12 月

Taf. Taf. 1-2. in the southeastern United States. J. Arn. Arb. 一一一一 189 1. Die Gattungen der Pomaceen. 55: 303-332 ，344-401 ，611-662. Gartenflora Gartenflora 40: 4-7 ，35-38 ，59-6 1. 一一一一一， Phipps J. B. ，Rohrer J. R. and Smith Koidzumi Koidzumi G. 1934. A synopsis of the genus Malus. P. G. 1991. A synopsis of genera in Maloideae. Acta Phytotax. Geobo t. 3: 179-196. Sys t. Bot. 16: 376-394. Kovanda M. 196 1. Flower and fruit morphology Roth 1. 1977. Fruits of angiosperms. In K. of of Sorbus in correlation to the of the Linsbauer ，Handbuch der Pflanzenanatomie ， genus. genus. Preslia 33: 1-16. Spez. Teil ，Bd. 10 ，Gebruder Borntraeger ， 一一一一 1965. On the generic concepts in the Be r1 in 恥1aloideae. Preslia 37: 27-34. Sax K. 193 1. The origin and relationships of the 一一一一 and Challice J. S. 1981. The genus Micro- Pomoideae. J. Ar n. Arb. 12: 3-22. meles revisited. Folia Geobot. Phytotax. 16: 一一一一 1932. The origin of the Pomoideae. Proc. 181-193. 181-193. Ame r. Hor t. Soc. 30: 147-150. Lindley Lindley J. 1821. Observation of the natural group Schneider C. K. 1906. Ill ustriertes Handbuch der of of plants called Pomaceae. Trans. Linn. Soc. Laubholzkunde 1. G. Fischer ， Jena. 13: 13: 88-106. Schulze-Menz G. K. 1964. Rosaceae. In Melchior

MacArthur M. 組 d Wetmore R. H. 1939. Develop- H. (ed よ A. Engler's Syllabus der Pflanzen- mental mental studies in the apple fruit in the varieties familien ， 2: 209-218. Gebruder Borntraeger ， Mclntosh Mclntosh Red and Wagener. J. Pomo l. Hor t. Berlin. i. Sc i. 17: 218-232. Stebbins G. L. 1950. Variation and evolution in MacDaniels MacDaniels L. H. 1940. The morphology of the flowering plants. Columbia Univ. Press ，New apple apple and other pome fruits. Memoir ， Cornell York. Univ. Univ. Ag r. Exp. Station 230: 1-32. Sterling C. 1954. Sclereid development and the McAllister McAllister H. A. 1986. The Rowan and its relatives texture of bertlett pears. Food Res. 19: (Sorbus (Sorbus spp ふUniversity of Liverpool Botanic 433-443. Gardens ， Ness. 一一一一 1964. Comparative morphology of the Olson Olson A. R. and Steeves T. A. 1982. Structural carpel in the Rosaceae. V. Pomoideae: changes changes in the developing fruit wall of Crataegus ， Hesperomeles ， Mespilus ，

A melanchier aln ぴo[i， α. Can. J. Bo t. 60: Osteomeles. Am er. J. Bo t. 51: 705-712. 1880-1887. 1880-1887. Tukey H. B. and Young J. O. 1942. Gross mor- Phipps Phipps J. B. ， Robertson K. R. ， Smith P. G. and phology and histology of developing fruit of the Rohrer Rohrer J. R. 1990. A checklist of the subfamily apple. Bo t. Gaz. 104: 3-25. 恥1aloideae (Rosaceae). Can. J. Bo t. 68: Weber C. 1964. The genus Chaenomeles 2209-2269. 2209-2269. (Rosaceae). J. Arn. Arb. 45: 161-205 ， 一一一一，一一一一， Rohrer J. R. and Smith P. G. 302-345. 199 1. Origins and evolution of subfam. Wiley E. O. 1981. Phylogenetics. 439pp. John Maloideae Maloideae (Rosaceae). Sys t. Bo t. 16: 303-332. Wiley and Sons ，New Y ork. Rehder A. 1940. Manual of cultivated trees and Williams A. H. 1982. Chemical evidence from the shrubs. shrubs. ed. 2. Macmillan ，New York. flavonoids relevant to the classification of Robertson Robertson K. R. 1974. The genera of Rosaceae Malus species. Bo t. J. Linn. Soc. 84: 31-39. December December 1991 Journal of Japanese Botany Vo l. 66 No. 6 347

YU T.-T. (ed よ 1974. Rosaceae (1). Flora Rei- 胞よりなる不均質な構造を示した.リンゴ属 Malus publicae publicae Popularis Sinicae ，Vo l. 36. Science ドキニア属 Docyni α， カナメモチ属 Photini αの Press ， Beijing (in Chinese). 一部の種，およびニイタカカマツカ属 Str αnu αesi α 一一一一 1984. ‘Origin and evolution of Rosaceae. の一部の種を除き，ナナカマド連では内果皮は厚 Acta Phytotax. Sin. 22: 431-444 (in Chinese). 壁化しないことを明らかにした. これらの観察結果を，サンザシ連やパラ科の他 要旨 の亜科の果実と比較して，ナシ亜科における果実 パラ科はふつう 4 亜科に分類され，その 1 つで の形質進化を推定した.ナシ亜科においては，心 あるナシ亜科 Maloideae は約 23 属1， 110 種を含み， 皮の全体が骨化するサンザシ連の果実はそのよう サンザシ連 Crataegeae とナナカマド連 Sorbeae にならないナナカマド連の果実よりも原始的であ とに分類されている (Phipps et al. 199 ①. Decaisne ると推定した.ナナカマド連においては，多くの (1874) はナシ亜科の果実を初めて解剖学的に体 属で果肉中の石細胞が失われる方向に進化したと 系的に研究し，不均質な構造の果肉をもっ群のあ 推定した.また，カマツカ属 Pourthi αω では石 ることを記録した. しかし， この解剖学的な差異 細胞が果肉の中央部に塊状に集合して分布すると についてはその後植物学的にはほとんど研究され いう特徴を示し，アズキナシ属 Micromeles およ ずに残されていて， Gabrielian (1958) とKovanda びナナカマド属 Aria 亜属と Ch αmα emespilus 亜 (1 961) がヨーロッパのナナカマド属を研究した 属では周囲の細胞より大きな色素細胞および石細 際に，果実の解剖学的形質がナナカマド属の分類 胞またはそのうちのどちらかが塊状に集合して果 形質として有用であることを報告したに過ぎない. 肉の全体に分布するという特徴を示し，ナナカマ パラ科では果実の形質は重要な分類形質とされて ド属 Torminalia 亜属では石細胞が塊状に集合し おり，特にナシ亜科は梨果(ナシ状果)をもつこ て果肉内層に分布するという特徴を示した. これ とが主要な特色である.そこで，われわれはナシ らの特徴はそれぞれ特殊化した進化であり，固有 亜科における果実の解剖学的形質を再検討し， こ 派生形質であると推定された. の形質がナシ亜科の分類形質として有用であるか 否かを判定するために，ナシ亜科の22 属74 種(そ Appendix のうちナナカマド連は17 属62 種， サンザシ連は7 Materials examined and voucher specimens Fresh Fresh materials with collecting data and voucher 属12 種)について果実の解剖学的形態を調べた. herbarium herbarium specimens are listed with an asterisk (*) この論文は，ナナカマド連の果実の解剖学的構造 after after the collection number. When no voucher を記載し，この形質の進化傾向を考察し，その結 specimens specimens are available ，collecting data are listed. 果に基づいてナナカマド連に含まれる属の系統関 Dried Dried materials were collected from herbarium 係を推定したものである. specimens specimens listed here and from botanical gardens 観察の結果，ナナカマド連では果実の解剖学的 or or institutes through seed exchange programs. 形質が属のレベルで異なることが判明した.多く V oucher specimens are preserved in the following の属では果肉は主として柔細胞よりなり，少数の herbaria: Biological Institute ， Faculty of Science ， 石細胞が散在するかあるいは石細胞が存在しない Tohoku UniversityσUS); Botanical Gardens ， という均質な構造を示した. これに対し，いくつ Faculty of Science ，Tohoku University (TUSG); かの属では果肉は柔細胞，石細胞，および色素細 University Museum ， University of Tokyo (TI); 348 348 植物研究雑誌第66 巻第6 号 平成 3 年12 月

NatIonal NatIonal Science Museum ，Tokyo (TNS); Depart- S.ω 伊 idat α(Spach)， Hedlund: Nepal ，An na- ment of Botany ， Faculty of Science ，Kyoto purna Hima l.， Modi Khola ，Stainton ，Sykes ， University University 侭YO); and Li berty Hyde Bailey and Williams 8619 (T1). Hortorium ， Cornell University (B H). S. h αryastana Gabrielian: USSR ，ArmSSR; 25 Oc t. 1958 ，R. Karaprtjen s.n. ，(B 町. Sorbus L. subg. Sorbus S. kusnetzovii Zinser l.: USSR ，ArmSSR ， S. S. commixta Hedlund: Japan ，Hokkaido ，M t. Daral'egis ， Alazsko'e ushchel'je ，140c t. 1959 ，

Kariba ，3 Aug. 1984 ，H. Hoshi et al. 1164* E. Gabrielian s. n リ (B 町.

(TUS). (TUS). In the c出 npus of the Faculty of Science ， S. lanat α(Don) K. Koch: Afghanistan ， Tohoku University ，cultivated ，2 Oc t. 1985 ，H. Nuristan ，Kushtaki to Pronz ，4 Aug. 1955 ， S. 1ketani 1ketani 1729* (TUS). Kitamura s.n. 区YO). S. S. esserteauian αKoehne: 1n the Royal Horti- S. xanthoneura Rehder: China ，W. Hubei ，

cultural cultural Society Garden ，England ，cultivated ， Shennongjia Forest ，7 Sep t. 1980 ， 1980 Sino ・ F.G. F.G. Meyer 3180 (BH). American Expedition No. 719 0α0). S. S. gracilis (Sieb. et Zucc.) K. Koch: Japan ， Micromeles Decaisne

Yamanashi Pref. ，Yamanakako ・mura. 100c t. M. alnifolia (Sieb. et Zucc.) K. Koch: 1n the 1985 ，H. 1ketani 1979* (TUS). campus of Tohoku University ，cultivated ，1 S. S. matsumuran α(Makino) Koehne: Japan ， Nov. 1985 ，H. 1ketani 2016* (TUS). Yamanashi Pref. ， Mt. Kitadake ，140c t. 1985 ， M. caloneura Stapf: China ，W. Hubei ， H. 1ketani 993* (TUS). Shennogjia Forest ，8 Sep t. 1980 ，1980 Sino- S. S. pohuashanensis (Hance) Hedlund: 1n the American Expedition No. 753 (KYO). Royal Royal Horticultural Society Garden ，England ， M.co η mbifera (Miq.) Kalkman: N. Thailand ， cultivated ， F.G. Meyer 3432 (BH). Prov. Chiang Mai ， C.F. von Beusekom and C.

S. S. sambucifolia (Ch 出 n. et Schlech t.) Roemer: Phengklai 2425 (KYO). Japan ，Yamanashi Pref. ， Mt. Ki tadake. 140ct. M. folgneri Schneider: China ， Kuling along M t. 1985 ，H. 1ketani 1215* (TUS). Slope ，25 July 1928 ，C.Y. Chiao 18696 (BH). S. S. scalaris Koehne: Cultivated in Windson China ， 13 July 1928 ，W.P. Fang 2177 (TNS). Great Great Park ，Collector unknown (BH). China ，W. Hubei ， Shennongjia Forest ， 11 Sep t.

Sorbus L. subg. Torminalia (DC.) K. Koch 1980 ， 1980 Sino ・American Expedition No. 1008 ( =( Torminalis Medik.) (KYO). National Botanic Garden ，Glasnevin ， S. S. torminalis (L.) Crantz: Botanical Garden ， Dublin ，泊 1985. Vrije Vrije Universiteit ，Nethe r1 ands ， in 1984. Hortus M. japonio αDecaisne: Japan ，Miyagi Pref. ， Botanicus Botanicus Camerinensis ，Universita del Studi ， Sendai Aobayama ，30 Oct. 1984 ，H. 1ketani It aly ， in 1985. 1663* (TUS). Sorb ωL. subg. Chamaemespilus (M edik.) K. Aroni αMedikus Koch A. arbutifoU α(L.) Elliot: T.R. Dudey 8100 S. S. chamaemespilus (L.) Crantz: Jardin (TNS). Hortus Botanicus ， Vrije Universiteit ， in Botanique Botanique Alpine “La Jaysinia" ，Samoens ， 1985. France ， in 1985. A. melanocarp α(Michx.) Elliot: Botanic Sorbus L. subg. Ar ia (Pers.) Beck Garden of the University of Ferburg ，BRD ，担 December December 1991 Journal of Japanese Botany Vo l. 66 No. 6 349

1984. 1984. H. arbutifolia (Ai t.) Roemer: M. Hiroe 10044 ，

A. prunifoU α(M 訂 sh.) Rehder: In the Ar itaki 22 Oct. 1955 ， (TNS). California ，1875 ，Y.R. Arboretum ，Koshigaya ， Saitama Pref. ，Japan ， Vasey s.n. ，σ1) Research Division ，L. A. State cultivated ， 15 Nov. 1985 ，H. Iketani 2019* and County Arboretum ， in 1985. (TUS). (TUS). Pourthiaea Decaisne

Photini αLindley， P. argu ω(Lind l.) Decne. var. sα licifolia P. P. davidsoniae Rehder et Wilson: China ，2-30 (Decne.) Iketani et Ohashi: Thailand ， Prov.

Nov. 1932 ，W.T. Tsang 20660 (TI). Chaiyaphum ，Ban N 出 n Phrom ，R. Geesink ， T. P. P. glabra (Thunb.) Maxim.: In the Matsugaoka Hattin ，and K.C. Phengklai 6888 (KYO). Park ，Taira ，Iwaki ，Fukushima Pref. ，Japan ， P. beauverdiana (Schneid.) Hatusima: China ， cultivated ，4 Jan. 1987 ， J. Iketsu 1174* (TUS). K wantung ，Tinghushuan ， K. S. Chow et al. P. P. integrifolia Li ndley: Nepal ， near Kath- 78003 (TUS). mandu ， Sheopuri Lekh ，Stainton ， Sykes and 一一一一 va r. notabilis (Schn.) Hatusima: Williams Williams 6928 (T I). Formosa Ki rairei-Asahi ， E. Matsuda 280 ， ut P. P. P. prunifolia Lindley: China ， Gungdong ， kudoi (TI). Botanic Garden of University of Dinghusha ，G.Q. Ding and K. L. Shi 1656 ， Oxford ， in 1984. China ，Guangzhonm ， (TUS). (TUS). Hongkong ，Victoria Park ，5 Jan. 1964 ， Guangdog ，H.G. Yip 260 (TUS). Taiwan ， H. Hara and S. Kurosawa s.n. (TI). Taihoku ，Sinten ， K. Odashima 13666 (TI) ， ut P. P. serratifolia (Des f.) Kalkman: In the Photinia benthamina. Botanical Botanical Garden ， Faculty of Science ， P. impressiven α(Hayata) Iketani et Ohashi: University University of Tokyo ，cultivated ， 20 Nov. 1985 ， China ，K wangtong ， Sinfung District ，Y. W. H. Iketani 202 7* (TUS). Taam 146 ， ut Photinia euphlebia Mer r. et P. P. stenophylla Hand.-Mazz.: Thailand ，M. Chun. (KYO). Tagawa ， K. Iwatsuki and N. Fukuoka T514 P. villosa (Thunb.) Decne.: Japan ， Miyagi 0α0). Pref. ，Sendai ，Aobayama ， 29 Oct. 1984 ，H. P. P. wrightiana Maxim.: Japan ，Ryuku ， Is l. Iketani 1662* (TUS). Japan ，Yamanashi Pref. ， Kumejima ，Gushikawasan ， T. Yamazaki 1675 Aokigahara ， 16 Oct. 1983 ，H. Iketani 1248* (TUSG). (TUS). Japan ， Miyagi Pre f.， Sendai ， Stranv αesi αLindley Aobayama ，2 Nov. 1985 ，H. Iketani 1665* S. S. amphidoxa Schneider: China Kweichow ， (TUS).

S.W. Teng 90474 0α0). 一一ー var. parvifo /i， α (p ritz.) Iketani et Ohashi: S. S. dl αvidiana Decaisne: In the Botanical Garden ， Taiwan ， Taipei Co. ，Shih-tung ，K.C. Young Faculty Faculty of Science ， University of Tokyo ， 219 (TUS). T 剖wan ，M t. Daiton ， T. Sato 326 ， cultivated ， 20 Nov. 1985 ，H. Iketani 2026* ut Photinia kankoensis (TI). China ， (TUS). (TUS). Kwangtong ，T.M. Tsui 638 ， utP. subumbellata S. S. nussi α(D. Don) Don: Nepal ，Kathmandu R. et W. (TI) District ，D. H. Nicolson 2719σ1). Eriobotrya Li ndley S. S. undulata Decaisne: Botanic Garden ， E. bengalensis (Roxb.) Hooker f.: S. W. University University of Oxford ， in 1985. Thailand ， Prov. Kanchanaburi ， Geesink and Heteromeles Heteromeles Roemer Phengkahl 6178 0α0). 350 植物研究雑誌第66 巻第6 号 平成 3 年12 月

E. E. cavaleriei (Lev l.) Rehder: China ，Kwantung Pyrus L. Sin-fung Sin-fung district ，Y.W. Taam 305 (KYO). P. pyrifo /i. α(Burm. f.) Nakai: Aobayama ， E. deflexa (Hems l.) Nakai: Taiwan ，Nantou Sendai ，Japan ，cultivated ， 15 Sep t. 1985 ，H. Co. ， Y.F. Chen 3481 (TUS). Iketani 1699* (TUS). E. fragrans Champ.: China ，Kwantung ， south M alus Miller

of of Li nchow ，T.M. Tsui 614 (TI). M. baccata (L.) D 回 f. var. sibiric α(Maxim.) E. japonica (Thunb.) Lindley: In a private Schneider: Korea ， Inter Nansen et Arijyo ，7 garden ，Kawauchi ，Sendai ，Japan ，cultivated ， Sep t. 1902 ， T. Uchiyama s.n. (TUS). on 6 Aug. 1986*. M. floribund αSiebold:， National Botanic E. E. salwinensis Hand. ふ1azz.: Nepal ，Godawari- Garden ，Glasnevin ， in 1985. Phulchoki ， 23 June 1967 ，H. Kanai et al. s.n. M. prunifolia 何Tilld.) Borkh.: In the medical (TI). (TI). plants garden of the Faculty of Pharmacology ， Rhaphiolepis Rhaphiolepis Lindley Tohoku University ，cultivated ，3 Nov. 1985 ，H.

R. R. indic α(L.)， Lind l. ex Ker var. umbellat α Iketani 2013* (TUS). (Thunb. (Thunb. ex Murray) Ohashi: In the Botanical M. domestic αBorkh. ‘ Akane': Purchased at a Garden ，Faculty of Science ，Tohoku University ， market in Sendai ，on 29 Sep t. 1986*. cultivated ， 22 Dec. 1986 ，H. Iketani 204 7* M. toringo (Sieb.) Sieb. ex Vriese: In the (TUS). (TUS). Botanical Garden ，Faculty of Science ，Tohoku R. R. ferruginea Metcalf: China ，Kwantung ， University ，cultivated ，on 14 Sep t. 1986*.

Sin-fung Sin-fung district ，Y.W. Taam 167 (KYO). M. tschonoskii (M 鉱 im.) Schneider: Japan ，

Chaenomeles Chaenomeles Li ndley Miyagi ，Sendai ，Aobayama ， in the c出 npus of C. C. speciosa (Sweet) Nakai: In a private garden ， the Faculty of Pharmacology ， Tohoku Aobayama ，Sendai ，Japan ，cultivated ，7 Sept University ，3 Nov. 1985 ，H. Iketani 2155* 1986 ，H. Iketani 2363* (TUS). In a private (TUS). Japan ，Miyagi ，Sendai ，Aobayama ， in garden ，Ouchi ，Akita ，Japan ，cultivated ，on 15 the Botanical Garden ， Faculty of Science ， Sep t. 1986*. Tohoku University ，9 Nov. 1985 ，H. Iketani Pseudocydonia Pseudocydonia Schneider 1661* (TUS). P. P. sinensis (Thourin) Koehne: In the medical Amelanchier Medikus plants plants garden of the Faculty of Pharmacology ， A. asiatica (Sieb. et Zucc.) End l. ex Walp.: In Tohoku University ，cultivated ，5 Nov. 1985 and the experimental garden of the Biological 3 Nov. 1986 ，H. Iketani 2014* (TUS). Institute ， Faculty of Science ， Tohoku Cydonia Miller University ，cultivated ，1 Oct. 1985 ，Japan ，H. C. C. oblonga Miller: Purchased at a market in Iketani 1966* (TUS).

Sendai Sendai city ， Japan ，on 22 Oct. 1986*. A. sanguine α(Pursh)， DC.: Concervatoire et Purchased Purchased at a market in Sendai city ，Japan ， J ardin botanique de N ancy. on 28 Oc t. 1986*. A. spicata (Lam.) K. Koch: Botanisch Garten ，

Docyni αDecaisne， Munchen-Nymphenburg. D. indic αDecaisne: Sikkim ，Y oksan ，H. Hara Peraphyllum Nutall et et el;l. 6459 (TI). East Nepal ，Dumban ， P. ramossisimum Nut t.: U.S.A. ，Colorado ， Taplejung ，H. Hara et al. 6301816 (TI). 1963 ， Y. Matsumura s.n. ， (KYO). December 1991 Journal of Japanese Botany Vo l. 66 No. 6 351

Mespilus Mespilus L. Garden ，Faculty of Science ，Tohoku University ， M. germanicus L.: Botanisk Museum ，Hage cultivated ， 20ct. 1985 ，H. Iketa 凶 2048* (TUS). Univ. Univ. Bergen. Crataegus L. Dichotomanthes Dichotomanthes Kurz. C. monogyna Jacq.: In the Botanical Garden ， D. tristaniaecarpa Kurz.: Hortus Botanicus Faculty of Science ， Tohoku University ， Instituti Instituti Botanici nom V. L. Komarovi i. cultivated ， 20ct. 1985 ，H. Iketani 1715* (TUS). Cotoneaster Cotoneaster Medikus H esperomeles Li ndley C. C. frigidus Wal l. ex Li nd l.: National Botanic H. ω neata Lind l. Peru ， 14 Nov. 1965 ， F. Garden ，Glasnevin ， Dublin. 恥faekawa et al. 0693 (TI). Peru ， 19 Nov. 1965 ， C. C. horizontalis Decne.: In the medical plants F. Maekawa et al. 1259 (TI). garden garden of the Faculty of Pharmacology ， H. heterophylla (R. et P.) Hook.: Costarica ， Tohoku University ，cultivated ， 27 Oct. 1986 ，H. Cartago Provincia ， 14 Feb. 1971 ，Gillis and Iketani Iketani 2045* (TUS). Powman 10234 (KYO). C. C. moupinensis Franch.: National Botanic H. lαtifoU α(HBK.) Roemer: Peru ， 13 Dec. Garden ，Glasnevin ， Dublin. 1965 ， F. Maekawa et al. 1921ρt) (TI). C. C. rhytidophyllus Rehd. et Wils.: National Osteomeles Li nd l. Botanic Botanic Garden ，Glasnevin ， Dublin. O. anthylidifolia Li ndl.: Ryukyu ，Isle. Ohgami ，

Pyrcanth αRoemer Y. 恥fiyagi et al. 38360 (TUSG). P. P. crenulatl α(D. Don) Roemer: In the Botanical