全 文 ：广 西 植 物 Guihaia 30(3)：316— 323 2010年 5月
湖北双蝴蝶小孢子发生及雄配子体发育
黄衡宇，龙 华，易婷婷，李 鹂
(吉首大学 植物资源保护与利用湖南省高校重点实验室，湖南 吉首 416000)
摘 要：首次报道了湖北双蝴蝶小孢子发生和雄配子体发育。主要结果如下 ：花药四室 ；药壁发育为双子叶型；
绒毡层异型起源，属腺质型绒毡层，药隔处的绒毡层细胞形成类胎座 ，其余部位的绒毡层细胞仍为一层细胞；花
药成熟时，药室内壁纤维状加厚且柱状伸长 ，表皮细胞减缩退化，纤维状加厚不明显 。小孢子母细胞减数分裂为
同时型，四分体排列方式主要为四面体形，少数为十字交叉形；成熟花粉多为 2一细胞型 ，偶见 3一细胞型，具三萌
发孔。
关键词 ：湖北双蝴蝶；小孢子 ；雄配子体 ；龙胆科
CLC Number：Q944．4 Document Code：A Article ID：1000—3142(2010)03—0316 08
Genesis of microspore and the development of male
gametophyte in Tripterospermum discoideum
HUANG Heng-Yu，LONG Hua，Yi Ting-Ting，LI Li
． (Key Laboratory of Plant Resources Conservation and Utilization(Jishou
University)．College of Hunan Province，Jishou 416000，China)
Abstract：The present paper firstly reports the microsporogenesis and the development of male gametophyte in
丁r pterospermuⅢ disc0ide“7n．The main results can be concluded as folows：Anthers are tetrasporangiate．The de—
velopment of anther wals conforms to the Dicotyledonous type and comprises of epidermis，endothecium，one or two
middle layers and tapetum at the mature stage．The tapetal cels have dual origin and belong to the glandular type．
The tapetal ce1ls on the connective side show radial elongation or periclinal division and intrude into the anther locule
t()form placenoids；The endothecium persists and its cells become pillar and fibrous，and the epidermis degenerates．
Cytokinesis at meiosis of microsporocytes is of the simultaneous type and most of microspore tetrads are tetrahedral，
tt1ere are stil a few other types，such as dilatera1．Pollen grain is mainly 2-celed type when shed，occasionally 3-celed
type，and has three apertures．
Key words：Tripterospennum discoideum ；microspore；male gametophyte；Gentianaceae
Tripterospermum discoideum belongs to
Tripterospermum of Gentianaceae． This twining pe—
rennial herb distributes primarily in Hubei，Shanxi and
Hunan province．and the west of Hunan is the mostly
station(He et a1．，1988)． discoideum has been
widely used in curing furuncle，mastitis，trauma，bleed—
ing，and fracture at civilian of some provinces that in
the middle and west of China，such as Hunan，Hubei，
Shan~ and Guizhou． In recent years，the wild re—
sources decrease annually with the rising colections by
some medica1 factories in the east area．Many factories
in the west of Hunan province have been trying to cul—
tivate discoideurn by artificial introduction．Howev—
er，it is always defeated at last on account of having lit—
tie knowledge about its life history，gexual reproduction
and asexual reproduction． Though there are lots of
Received Dale：2008—08—0 l Accepted date：2009—10-06
Foundation item：Supported by the National Natural Science Foundation of China(30360009}
Biography：Huang Heng—Yu(1970一)。Male。Ph．D．．Adjunct professor．Major in Plant DeVelopmental Biology-(E-mail)hhyhhy96@163·corn·
Author for coresI)ondence．E-mail：lilyjsu@126．corn
3期 黄衡宇等：湖北双蝴蝶小孢子发生及雄配子体发育 317
studies on embryo of Gentianaceae in present(Zhu et
“Z．，1989；Li et“Z．，1994，2006；Liu et a1．，1996a，b，
1997，l998；He et a1．，1999a，b，2000；Xue et a1．，
l999a，b，2002a，b；Xue，2005)，but there are little in—
vestigations on Tirpterospermum in China(Chen et
“ ．，1999；2000)and this genus has never been embryo—
logically investigated．The objective of the present pa—
per is to accumulate embryological data of Tirptero—
S]),~TITIIA．ITI in order to offer evidence for systematic tax—
onomy and conservation biology by studying on the mi—
crosporogenesis and the formation of male gametophyte
of丁．discoideurn． Furthermore，it wil1 provide basic
data that is beneficial for introduetion and artificia1 eul—
livation of this genus．
1 Materials and Methods
Malerial investigated for the present study was
colecl ed from Gaowangjie Forestry Centre of Guzhang
County of Xiangxi Tugia and Miao Autonomous，Hu—
nan prox，ince，China(110。04 E，28。38 N，Ah：(792±13)
m)． The voucher(Long hua1 78)is deposited in the
Herbarium of Biological Resources and Environmental
Science Department of Jishou University．Specimen i—
dentifier is Zhang Dai—Gui(Biological Resources and
Environmental Science Department of Jishou Universi-
ty)．
Flower buds and flowers at different stages of de—
velopment were collected and fixed in the modified
FAA(5O％ alcohol：glacial acetic acid ：formalde—
hyde= 89 ：6 ：5)．The fixed materials were stained
in Ehrfich’s hematoxyfin． The materials were embe—
ded in paraffin after being washed with water，dehydra～
ted in ten grades of alcohol and transparentized with
five grades of xylene，and serially sectioned at the
l hickness of 5——8 ／xm by M icrom． Sections were
mounted on slides in Neutral balsam，observed and
photographed with Leica DM2000．
2 ()bservations
2．1 M icrosporogenesis
Flower of Tirpterosperraurn discoideum is bisex—
ual and has five stamina． Anthers are tetrasporangiate．
In the early phase of budding，the primordium of flow—
ers forms perianth and staminate primordium in se—
quence．From observations in the transverse section of
a young anther derived from staminate primordium，the
surface of anther is a layer of epidermal cells，inside
which there are a group of cells which divided actively．
W ith the development：of anther，four patches of tissue
are differentiated from the main mass of cells．The ar—
chesporial cells in each patch，which are differentiated
below the epidermis，are recognizable by their large
size，conspicuous nuclei and radial elongation． These
cells divide perielinaly(Plate T．2)to produce~rimary
parietal cels towards outside and primary sporogenous
cells towards inside(Plate I：3)．
2．1．1 The formation and differentiation of the anther
wall The primary parietal cells undergo periclinal di—
vision to produce outer layer and inner layer forming
three layers including the epidermis． The outer cels
formed a subepidermal endothecium and one middle
layer by periclinal division(Plate I：4)．At the time
when secondary sporogenous cells appear，the mature
anther wall consists of four layer cells：epidermis，endo—
thecium，one middle layer and one tapetum(Plate I：5)，
but a few(about 5 )consists of five layer cels：epider—
mis，endothecium，two middle layers which are botb
from the outer cels and one tapetum(Plate I：6)．It in—
dicates that the outer layer derived from primary parie—
tal cells to produce endothecium and middle layer by
divisions whereas the inner layer directly matures into
tapetum．According to Davis’s(1966)compartmental—
ization，the development of micrOsporangial wall con—
forms to the Dic0tyledonous type．
The innermost layer of anther wali is tapetum
with dual origin that mainly from the inner cells and
partly from the connective cells．Tapetum starts to dif—
ferentiate at phase of early sporogenous cels and its
cells gradually augment while its protoplasm gradualy
manifold in the course of development of microspore
mother cells． The tapetal cells increase attaining the
maximal development at phase of microsporocytes wi th
the feature of glandular cel1．Here，the tapetal cells are
much larger than those of the outer and present close
318 广 西 植 物 3O卷
Plate 1 EP ~tdcrmis；EN．Endothelium；M．Middle layer；PL．“Placentoid”originating fronl the tapetum cels；PP．Primary parietal cel；I】s．P rj—
mary s[Iorog,．!lous cel；SS．Secondary sporogenous cel；T．Tapetum．1．Male archesporium(十)×1000；2．Microspore archesporial cels dividing(十)
×1000；3．Primary parietal cels(PP)and primary sporogenous cels(PS)×1000；4．Primary parietal cel dividing(十)×1000；5．Four layers anther
wall in the time of secondary sporoginous cel1×1000；6．Five layers anther walI in the time of secondary sporoginous cell×1000：7．Tapetum cells in
the time of microspore mother cels(T)× 1000；8．Tapetum cels protruding the anther chamber in the time of mierospore mother cels×400：9．Ta—
petum cells whh two nuclei(T)×1000；10．Disintegrating tapetum cels at the telophase lI of meiosisX 1000；l1．Stage of microspore tetrads，disinte—
grated tapetum×400；12．Trails from disintegrated tapetum(+)×1000．
squareness(Plate I：7)．Cells of the tapetum on the
connective side show periclinal divisions and intrude in一
10 the anther locule．At this region where division OI2一
curs，the tapetum becomes two or more layers and ap—
pears as“quasi—placenta”or“cross-grid”(Plate T．8)．It
is observed that most of tapetal cells are uninucleate，
3期 黄衡宇等：湖北双蝴蝶小孢子发生及雄配子体发育 319
while it is also observed that minority are binucleate
(Plate I：9)．Tapetal cells start to separate from each
other when the development of microsporocytes are ap—
proximately over，degenerate obviously when miosis of
microsporocytes are approximately over(Plate I：1 0)
and disaggregate intensively at their original site in the
time of tetrads．The degenerated nuclei of tapetal cells
in the anther chamber are formed by the same layer
cells that enter it at early time but they are not formed
by degenerated tapetal cells of circumference that enter
it(PlateⅡ：11)．It is not found that amalgamation oc—
curs and periplasmodium appear when tapetal ceils dis—
aggregate from large numbers of sections．The tapetal
cells are single till disappearing．At late period of uni—
nucleate microspore，there are only few degenerated
vestiges left(Plate I：12)．Therefore，the tapetum is
similar to the glandular type．
The middle layer accomplishes differentiation at
time of secondary sporogenous cells，including one or
two layers，of which cels are of smaller size，bigger nu—
cleus and flat rectangle(Plate I：5，6)．The middle layer
get s degenerated during meiot ic division of microsporo—
cytes(Plate lI：13)．
The endothecium comes into being at time of sec—
ondary sporogenous cells while cells arrange close and
present strip shape(Plate T．5)．The endothecial cells
are most developed when microsporocytes formed
(Plate T：7)and develop fibrous thickenings from their
inner tangential walls anatony and ectad in mature an—
ther at late period of the uninucleate pollen．At time of
polen grains，the fibrous thickenings is very obvious
while the outer tangential thin wal is persist(PlateⅡ：
14)and this characteristics of the endothecium are
helpful for anther dehiscence．The anther wall dehisces
and the mature polens emit when shed with the cels
keep on losing water．The reason is that the outer tan—
gential wall without thickening of the cells of endothe—
cium shrinks and caves in．
Tbe epidermis only includes single-layer cells that
arrange closely． At time of secondary sporogenous
cells，the epidermal cells reach the highest level of de—
velopment(Plate T：5)and they begin to degenerate at
time of microspor0cytes(Plate I：8)．The epidermis on—
ly leaves the vestiges at time of anther dehiscence(PIate
11：15)．
2．1．2Ⅳ【icrosporogenesis The primary sporogenous
cells undergo mitosis resulting in secondary sporoge—
nous cells(Plate I：5，6)．In comparing with surround—
ing wal cels，the secondary sporogenous cells have lar—
ger size，greater nuclear-cytoplasmic ratio，dense cyto—
plasm，and deeper color；moreover，they arrange closely
and don’t have obvious vacuoles and present polygon．
The polygonal secondary sporogenous cels turn into
circular mi crospore mother ceils with development
(Plate T_9)．The microsporocytes show distinct differ—
ence from the other anther wal cells around，which
have larger size and cell nucleus，dense cytoplasm；Mo—
reover，they arrange closely and don’t have obvious
vacuoles．The microsporocytes undergo meiosis 1 with—
out cytokinesis(PlateⅡ：16—18)，but cytokinesis take
place in meiosis 1]of the microsporocytes and the callose
wal form simultaneous(PlateⅡ：19-21)．Most of mi—
crospore tetrads are tetrahedral(Plate I：22)，there are
still a few other types，such as dilateral(Plate I1：23)．
2．2 M ale gametophyte
Because of the callosal dissolution，four micro—
spores separate respectively and are released to anther
chamber that permeate secretion of tapetum．The mi—
crospore j ust formed has a central nucleus and dense
protoplasm，but the cell wall has not been thickened
and there is no obvious vacuole in the cytoplasm．W ith
the development of microspores，many smal vacuoles
appear in the protoplasm(Plate I：25)，which get to—
gether to form a large central vacuole pushing the nu—
cleus to a peripheral position(PlateⅡ：26)．The micro—
spores divide mitotically so as to form two daughter
nuclei．Before long，cytokinesis between the two nuclei
follows，results in the form ation of a smaller lenticular
generative cell closed to polen wall and a larger vege—
tative cell with a large central vacuole by a vault ed cel
plate(PlateⅡ：27)．At the time，the germinal aperture
and germ furrow of the polen wall appear which
shows that extine starts to be thickened．The genera—
tive cell which is close to the pollen wall at the early
stage，moves into the cytoplasm of vegetative cel and
becomes a naked cel1 incursion in its cytoplasm with
32O 广 西 植 物 3O卷
Hate 1 13．Degraded middle layer at the telophase of meiosis×1 000；14．Anther wal of the polen spread×400；15．Anther wal of the time of pol—
len spread×1000；16．Microsporocyte atthemetaphase_IofmeiosisX1000；17．Microsporocyte atthe anaphaseI ofmdosisX1000；18．Microsporo—
cyte at the telophase 1 0f meiosisX 1000；19．Microsporocyte at the metaphase Iiof meiosis×1000；20．Microsporocyte at the anaphase lI of meiosisX
1000；21．Microsporocyte at the telophase I of meiosisX1000；22．Tetrahedral microspore tetrads×1000；23．Dilateral mierospore tetrads×1000；
24．Microspore had just formed from microsporetetrads×1000；25．Some smal vacuoles appear，entering themiddie of microspore×1000；26．Vacu—
olate period of uninucleate nficrospore× 1000；27．Early 2-celled pollen，showing obvious cell wal×1000：28．Mature 2-celleel pollen and three aper—
tures bourgeoning()x】000；29．Mature 3一 jed polenx l000．
3期 黄衡宇等：湖北双蝴蝶小孢子发生及雄配子体发育 321
the cell wall disappears，in a few minutes，generative
nuclear moves towards the vegetative nuclear． Up tO
NOW，development of the mature 2-celled pollen grains
is over(Plate 1I：28)．But there are also a few genera—
tive nucleuses of pollen grains that give rise tO two
sperms by further division at time of anther dehis—
cence．In that case，pollen grains are 3-celled(PlateI：
29)．
3 Discussion
Tripterospermum is classified tO two groups ac—
cording tO the types of fruit，sect．Platysperm“7” and
sect．Tripterosperm“772(He et a1．，1988；Wu，1984)．
The two groups can be differentiated from each other
merely at the stage of fruit and can not be differentia-
Table 1 Comparison of embryological characters between
discoideum and some species of TripterOspemnum
Table 2 The relationships between the development of antheral walls and the microspores
Note：“一”indicates that the structure has degenerated and disappeared．
ted at the stage of flowers．The table l shows the
comparision Of embryological characters between
T．discoideum and 丁．chinense(Chen et“Z．，l999)
of sect．Platyspermum and 丁．cordatum (Chen et
“￡．，2000)of sect．丁r pf ro p r， “7 ． There are
many similar embryological characters from table
1：tetrasporangiate anthers；I)icoty1edonous type of
anther walls development；glandular tapet um；pillar
and fibrous endothecium and degenerated epidermis
in the mature anther；simultaneous cytokinesis at
meiosis of microsporocytes；tetrahedral microspore
tetrads；but there are also some differences in em—
bryology：(1)in the middle layer，most of-／、．discoi—
deum has only one layer，which is similar tO 丁．
chinense，while there are also minorities have two
layers，which is the same as T．cordatum of sect．
Tripterospermum；(2)pollen grain is mainly 2一
celled type when shed in 丁．discoideum，but 3一
celled in T．chinense and 丁．cordatum．
A series of behaviors of anther wall，such as
322 广 西 植 物 3O卷
formation，development，disaggregation and disap—
pearance，shows great physiological value in丁．dis—
coideum．Disaggregation of each layer is beneficial
for the development of other layers．Table 2 shows
the relationships between the development of an—
ther walls and the microspores，from which we
know that at time of sporogenous cells，the four
layers of anther wall differentiate obviously． Be—
fore long，cells of t he middle layer have started to
be used and disappeared until the formation of mi—
crospores．Foster et a1．(1963)indicated that as one
or more layers of anther wall went further develop—
ment，it was stretched and spewed generally．W hen
the anther dehisced，it always was destroyed and
difficult to distinguish． However，t he other layers
of the anther wall are not stretched or spewed
when middle layers are disappearing，why?W hat’s
more．it is observed that cells of the middle layers
reduce and degenerate gradually． Thus，we believe
that degenerations of the middle layers are not only
t he result of stretching and spewing，but also sup—
ply nutrition for development of other layers of an—
t her wall，which proved that its behaviors have a
great of physiological value．
In recent years，with the help of electron mi—
croscope and histochemical method，many authors
have proved t hat t here is a cell wail formed be—
1 ween the generative cell and the vegetative cell af—
ter mitosis of microspores．But it has been repor一
1ed differently in various plants on the characters
of the wal1． In some genus，the cell wall is com—
posed of cellotetrose or pectin(M aruyama et a1．，
l965；W illiam et a1．，1968；Sanger et a1．，1971)，
while it has been considered as callose in some oth—
er genus(Heslop—Harrison，1968；Blackman，1983；
I iu et( f．，1997)．Similar to other genus of Genti—
anaceae．丁．discoideurn has a eel】wall composed of
callose when generative cell is formed，but it will
disappear when the generative cell moves into the
cytoplasm of vegetative cel1． It can be considered
as a barrier to isolate the generative cell and the
vegetative cell in short time for their development
in different ways． Callose makes great contribu—
tions not only to the formation of the male gameto—
phyte，but also to microsporogenesis． W ang Fuhsi—
ung(1993)considered that，in the term of sporoge—
nous cells，there were plasmodesmi existed both
between each layers of the anther(wall，tapetum ，
and sporogenous tissue)and neighbouring cells in
the same layer．But those connections were cut off
gradually when microsporocytes were formed and
started to undergo meiosis． At the same time，
there was callose collected around the microsporo—
cytes．The later make the microsporocytes separa—
ted from the cells of anther wal1． It ensures that
there are no diploid cells disturbing in the process
of the translation from diploid cells to monocaryot—
ic cells by meiosis． W ith collections of callose，the
primary parietal of microspor。cytes begins to de—
generate，and forms cytoplasmic channels among
microsporocytes，cytoplasm and organelles can pass
these channels freely． As a result，exchanging of
substances and genetic message among the micros—
porocytes can be more convenient．A1l of the mi—
crosporocytes surrounded by callose remain togeth—
er to form a symplasm．
References：
Blackman． 1983．Comparative anatomy of polinia and caudicle of
an orchid(Epicendrum)[J3．Bot Gaz，144：331—337
Chen SL，He TN，Liu JQ，et a1．1999．Mega，Microsporogenesis
and Female，Malegametogenesis of Tripterospermum Chinese
(Gentianaceae)[J]．Acta Biologica Plateau Sin，14：26—34
Chen SL，He TN，Liu JQ，ct a1．2000．Embryology of Triptero—
“Ⅲcordatum(Oentianaceae)E J]．Acta Bot Yannan．22
(1)：53—58
Davis Gwend L．1966．Systematic Embryology of the Angiosperms
EM3．New York：John Wiley
Foster AS，Gifford EM ． 1963．An introduction to the embryology
0f angiospern1s(cl1jnese Version by Li ZL)EM]．Beij ing：Science
Press：289
Heslop—Harrison J．1968．Synchronous pollen mitosis and the
formation of the generative cel in massulate[J]．J Cell Sci，
3：457— 466
He TN，Chen SL，Liu JQ，el a1．2000．Embryology of Gentiana
sgriata(Gentianaceae)[J]．Actu Bot Boreal-Oct ident Sin。20
(6)：960— 967
He TN，Liu JQ．1999．Embryology of Gentiana lawrencei var．
／arreri[J]．Acta Bot Boreal-Occident Sin，19(2)：234-240
He TN，Liu SW，Wu QR．1988．Gentianaceae[A]．Flora Reipu—
3期 黄衡宇等：湖北双蝴蝶小孢子发生及雄配子体发育 323
blicae Popularis Sinicae[M]．Beijing：Science Pres，62：257—258
He TN．Liu SW。Wu QR．1988．Oentianaceae[A]．Flora Reipu—
blicae Popularis Sinicae[M]．Beijing：Science Press，62：261-262
He TN，Xue CY，Liu JQ．1999．Embryology of Swertia erythros—
ticta Maxim[J]．Acta Bot Boreal-Occident Sin，19(1)：76—8O
Li L，Huang HY．2006．The genesis of microspore and the forma—
tion of male gametophyte in Swertia davidii Franch[J]．Bull
Bot Res，26(4)：452— 460
Li HJ，Wang YZ．1 994．Embryological studies in Gentiana macro—
phylla[J]．Acta Bot Boreal—Occident Sin，14(4)：243—248
I iu J Q．He TN．1 996a．The Embryological studies of Comastoma
Pulmonarium(Gentianaceae)[J]．Acta Phytotaxon Sin，34(6)：
577— 585
Liu JQ，He TN． 1996b． Embryological studies of Gentianela
aZUtearJ_．Acta Bot Yunnan，18(2)：l5l一158
Liu JQ，He TN． 1997． Embryology of Gentianopsis Paludosa
(FR．)MA[J]．Acta Biologica Plateau Sin，13：31-41
I iu JQ，Xue CY。He TN．1 998．Embryology of Sz~ertia franchi—
tiana．a famous Tibetan medicine[J]．J Northz~est Nonnal
U7dv：Nut Sci，34(4)：59— 66
Maruyama K．Gay H．Kaufmann BP． 1965． The mature of wal
between generative ceil and vegetative nuclei in gain of Trades—
cantia paludosuI-J]．AmerJ Bot，52：605—610
Sanger JM．Jackson WT．1971．Fine structure study of polen de—
velopment in Hacmanthus Katherinae Baker，I．Formation of veg—
etavia and generative cel[J]．J Cel Sci，8：289—301
Wiliam AJ．Donald BF，Mary EA．1968．Cotton embryogenesis：
The pollen cytoplam[J-]．Planta，81：206—228
Wang Fuhsiung．1993．Selected Works of Wang Fuhsiung[M]．
Beijing：Clnaesperanto-Eldonejo：336
Wu QR．1 984．Taxa Et Combinationes Tripterosperm i Et Craw—
furdiae Flora Sinica[J]．Bull Bot Res，4(3)：129—139
Xue CY，He TN，Liu JQ 1999a．Embryology of a Tibetan Medi—
cine Halenia ellipticu[J]．AttaBot Yunnan，21(2)：212—217
Xue CY，He TN，Liu JQ． 1999b．Embryology of Swertia tet—
raptera Maxim．(Gentianaceae)and its systematic implication
I-J]．Acta Phytotax Sirs，37(3)：259—263
Xue CY，He TN，Li DZ．2002K M egasporogenesis and female game—
togenesis development of the Tibetan medicine“Zang Yin Chen”一
Swertia mussotii(Gentianaceae)[J]．Guihaia．22(3)：249～25l
Xue CY，He TN，Li DZ．2002b． Embryology of S~ertia cincta
(Gentianaceae)and its systematic value[J~．Acta Bot Yunnan，24
(1)：75—8l
Xue CY，Li DZ．2005．Embryology of Megacodon stylophorus and
Veratrilla bailonii(Gentianaceae)：descriptions and systematic
implications[J]．Bot J Linnean Society，147：317—331
Zlm XH，Shen JH．1989．Microsporogenesis and megasporogene—
sis，and the development of the male and female gametophytes in
Gentiana manshurica K．[J]．J Harbin Normal Univ：Nat Sci
Edi，5：63— 73
井0斗0牛 斗0斗0斗e斗 斗 斗 斗 牛e牛e* 牛拳X,-(ol- 斗0牛 牛 牛搴牛 牛搴* 斗e牛拳蚌e斗0斗 斗0斗0斗 牛 *0牛 牛0斗 斗 井0斗(， 心 心 牛0斗 牛0*，，*0斗0
本刊外文 刊名“Gui haia的注释
本刊用“Guihaia”作为外文刊名，是取自“桂海”的汉语拼音，并予拉丁化缀以”a”字尾而成。
“佳海”一词，较早出现在我国南朝梁文学家江演的<杂体涛 ·袁太尉>：”文轸薄梓海”涛句吖1。以
“
南海有桂 ．敞 吲挂海”．仕海是泛指南方 近海地方 。 后 ．南术 人范成 火 曾仔静 _『fj=(腑 活今性林 if )
f¨J 南 路(今广西)地方长官，就其见I#1．追述广西山川、风物、花、 、草木的 蒋《仕海虞衡忠}，亦川
“l 海”～一词 ，概4旨』一西地区 。
本刊采用 写简便的“Guihaia”为外文刊名，仅为学术交流时 ，在外文文献上便于引 而已
≈、』 ‘ 物 编 部