Bulletin of the British Museum (Natural History) Botany

Si

yt yA

aie Has

ey Hi Seitet

’ sed!
i

jitite

i
He

"a

a

ie)
Neate een
it fat

i

Sip tere ah

AS

te)

in hy

A TAP eos
? 7 APR RA
A ” »

Bulletin of the een
British Museum (Natural es)

A revision of the Morinaceae
(Magnoliophyta-Dipsacales)

Margaret J. Cannon & John F. M.
Cannon

Botany series Vol 12 No 1 26 April 1984

The Bulletin of the British Museum (Natural History), instituted in 1949, is issued in four
scientific series, Botany, Entomology, Geology (incorporating Mineralogy) and Zoology,

and an Historical series.

sin the Bulletin are primarily the results of research carried out on the unique and
ever-growing collections of the Museum, both by the scientific staff of the Museum and by
specialists from elsewhere who make use of the Museum’s resources. Many of the papers are

works of reference that will remain indispensable for years to come.

Parts are published at irregular intervals as they become ready, each is complete in itself,
available separately, and individually priced. Volumes contain about 300 pages and several
volumes may appear within a calendar year. Subscriptions may be placed for one or more of
the series on either an Annual or Per Volume basis. Prices vary according to the contents of

the individual parts. Orders and enquiries should be sent to:

Publications Sales,
British Museum (Natural History),
Cromwell Road,
London SW7 5BD,
England.

World List abbreviation: Bull. Br. Mus. nat. Hist. (Bot.)

© Trustees of the British Museum (Natural History), 1984

The Botany series is edited in the Museum’s Department of Botany

Keeper of Botany: Mr J. F. M. Cannon
Editor of Bulletin: Mr J. R. Laundon
Assistant Editor: Dr A. J. Harrington

ISBN 0 565 08000 8
ISSN 0068-2292

British Museum (Natural History)
Cromwell Road
London SW7 5BD

Botany series
Vol 12 No 1 pp 1-35 '

Issued 26 April 1984 ;

a ;
Ss. a

A revision of the Morinaceae (Magnoliophyta — ~
Dipsacales) GENE

Margaret J. Cannon,and John F. M. Cannon
Department of Botany, British Museum (Natural History), Cromwell Road, London SW7 5BD

Contents

CPILG VERS cco Sncksceegeadeedonon scr eRoke Rcpr odie AUC can ecn ok eh ase Loune EACH cea eeane Ee ta teen 1
COC STE Se A ee gp ec oor Cr Aer aro Sot Onc ee ne ee See ee Ps
LEE STROTTGzI NS Dna aaa eaaseeenaceccade nce ccouen, catea meade cnet: Gaeet Ore ecac eter net Cone ee ee eee eee 2
Relationships of Morina sensu lato (Morinaceae) within the Dipsacales.................... 3
ST OCLC GED LO LOOM eect sa keer actA er eer nace ae MGM Cds Scinavnube anes « satsasiemesanes 3
PEATE CNL) OLY RM SER eral toe esas cies efoeie ede be Sila oine sais sae sles hiiera ne sine sea eeisca 4

3h, (COUGIUS/51- seewocpriedd Gade AGERSBEBHSES a8 dag ea So RO moo Nen ar Mar ene Oana 4
SOMES PUNY So EMIACKIMIORE | Verve orcct-s< spon d-qeemetanadaneussscncccendGcasecccnsdeescns 5

Bg SOMCLIGIONS tea en namerca mene ce dae cee seat ete nic saenesireatiew hoe didaesedsescbaoscaces 7
Peek AO MUMC YL OUIMACCA CE Merten eit eeerat a ceth ackeindeanardnsicsntcagees destabicemacterdagowassebas 7
GU STREEIG TV OCIUTTIEINY So daddngmaceldiesv deg: gar mS eee eL SURO ScnCEREE EC =: aECr eee Sean tate eee ae aeree 8
MI@IAIAC WEEE Ass. dbopd ink BbaocucdoSac6 Ade: Ua HasOre CREE oe ae Coot an ase cE eRe PERE ene ee ee eeee eee 8

aN aN CLD MME AEE US LR ere artes ens 0 pet ndins Ge ke vice rine wen wane wae $e nelna coe ron awnndecls 9

eA CATH LO CALY Xa (Eh) Gr) Mei GATIMOM (.o-ches decche= seiacecvewenmecsossaedsaceadecesssaneeaees 9
ISEVILOM MEN PCCLC SR Manereeen ateecnr cr otrckive Sarees oivccewtedsetiaceceer ders: ovctse sues’ 11
epARnepAlensts (DEMO IMs CannOM srs scc.ceccsecs-eaeeeaesas cere feencedecoavevnesneee 12

Pe LOVEE (ERAMEHET) Rl: CAMMOMN \.....cascnncsseasscorse de clsed ccd aviecs+oese dain 12
SerAndiog (Hand WMazza)iMi. (Cannons. ase: saqees fca-wgceonsoceeasccase sons ssseeeeccene 14

Mim Cryprotiiadia (Bunge) My Cann) renee. dec teecssaaeweanwerssseaduadesesscenesdeass 15

ING VALOIIME SD ECICS MeeeE ener tn eee) naire tn. stilt agents dn cotivedanseevaasessenbesas 17

ee Gat CHIT CILSES (EAT) Wen @ ATT O Di tyson ciate adaniaijainom da don aude abeeisce bc Wetewnsileedeic'ese 17

oe ear ora (cate iin oc Wit. YIM. CanMOM. 0. ...c. 2. cinnnccascocceecanecersnsennans 18

eC aKOKOMOTICA (TAO) Wis CANNON sees sa cieeieccies sn anasiecess vee: aeatiecoasseseee ne vnes 19

POAC ORARIBG (ICIS Vis CANTON oacc- sce scdeccsssccassbacercnoudadentssacadescaceeus 19

Sa Geporypniyua( Wallichex DCM: Cannon... 0. ccccccc.-ccsce+e+ceeceeceeceee cones 20
ere MUP E WM CMMI 0 ov asics canecnen dete ve het sey stactuut don fodtawedaeccklvesseenszas Ze

LUN, MICTALUE We odicacied dotgaddedeec Di uecatede Ortde- taeac a sEcE ee RAT OeE cere ECC Re CEE ee eacEr a erie 24

ECS Val OLIEISPECIE Sieepmene se naereeter entrar caer siecate sn eearacScio watts sueaieewonejcelteredvess 24

Re VURICO KRUG IIR GDC leer penta as se een at. ctaseo meet wn tania oes catinictaiials ¢- stedhadhethwedieeis 26

Dem VIRC OTL CLEATLOMN ONG a RAs edna rtse seid oes oes Wan» « Geis n ac deGumanticaateed 26

Bye NC IST GBI ese ta sec eee er ese eels ete etch iota sienera stein wlndleive'seeaaefasiojqeindinate «aren 28

BD VIPLOTISHTO LEN ALLL Chl © Xs) ©. nem PEPE Aire ie eysoiiae sce alee aeardad « scaisisn me sieeiieameoeeenosts 30

CE CUTTS S SUTIN SG [Sage hed essdtodaccseconduseeseec aL CBeedCoCuE EEL GCE meee CORe eee ater eee ne seee 31
ALS SOME GS ENMSTNG s cancdeababeena asad. ocdonesasancesoan@ocue Gods eC OC OREE ee ese ye eRe ane eee eereneenen 34
[RENE TECOS' coo ndecauncogeBorHeRt@eeadaeococo500 Gs0c uo Cen aU cop eE Eee Se ea aan eee a aCe ae aee renee 34
SRNR INC ANNRTCP ae Me fe | ee ree EEE Se PE EE Nee ee asecunceectecentencececesteros snes soene 35

Synopsis

The group of species comprising the genus Morina is reviewed, and support is established from the fields of
comparative morphology, embryology, cytology, and palynology, to justify their recognition as a distinct
family, Morinaceae J. Agardh, related to the Dipsacaceae. Three clearly defined groups are recognised

Bull. Br. Mus. nat. Hist. (Bot.) 12 (1): 1-35 Issued 26 April 1984

9) M. J. CANNON & J. F. M. CANNON

within Morina sensu lato and, as a consequence, the sections Acanthocalyx DC. and Cryptothladia Bunge
are elevated to generic status. Three new combinations under Acanthocalyx and five under Cryptothladia
are made, and a new species C. /udlowii is described. Thirteen species are recognised as occurring in the
family.

Introduction

This study was begun by one of us (J.F.M.C.) and, after a lapse of some years, was resumed by
both authors. Much of the detailed examination of specimens, measurement of organs, and
description of taxa was done by one author (M.J.C.), but the conclusions and consequent
taxonomy are the result of close collaboration and joint decision. So that future botanical
literature will not be unnecessarily burdened by dual author citations, we have decided that only
one of us (M.J.C.) will appear as the author of the new taxa and combinations. Unless otherwise
stated, all the specimens cited have been seen by the authors.

The Morinaceae includes a range of species that present fascinating problems in relationships
both within and outside the group. It is hoped that this revision will not only stimulate further
research, but will also encourage discerning plantsmen to grow a range of species beyond the
relatively well-known and quite widely cultivated Morina longifolia.

Historical summary

The genus Morina was first described by Tournefort in 1703, in honour of Ludovic Morin, a
doctor of medicine of Paris and a member of the Academy of Sciences of that city. It was
subsequently renamed Diototheca —a reference to the ear-like lobes of the calyx — by Vaillant in
1724 (subsequently misspelt by later authors Diolotheca, Diodotheca, and Dictotheca), and
placed by him in his class Dipsacées, between Pterocephalus and Valeriana. In both cases the
concepts were based on Morina orientalis carlinae folio, the only species known at the time and
now known as Morina persica L.

The genus passed into the era of binomial nomenclature in the first edition of Species
Plantarum (1753), when Linnaeus based his concept on his treatment in Hortus Cliffortianus
(1738), and beyond that on Tournefort’s (1703) original publication. In 1820, Rafinesque
separated Morina and Diototheca (misspelt by him as Diolotheca) from the Dipsacaceae as the
new family Morinidia, and considered this to be more nearly related to the Valerianaceae.
Coulter in 1824 included Morina in the Dipsacaceae, although he recognised many differences,
and de Candolle in 1830 continued to regard it as a member of that family, a view that was
followed by Bentham & Hooker in 1873.

Spach in 1841 classified Morina in the Dipsacaceae, but in a footnote says it has many affinities
with the Valerianaceae, and ought to be regarded as a monogeneric family between the
Valerianaceae and the Dipsacaceae. Bunge (1852) divided the genus into three sections:
Cryptothladia Bunge — M. parviflora only; Diotocalyx DC. — comprising M. persica, M. elegans,
M. longifolia, M. polyphylla, M. lehmanniana, and M. coulteriana; and Acanthocalyx DC. —
consisting of M. nana only. Van Tieghem (1909) recognised the Morinaceae J. Agardh as a
separate family, and distinguished it from Dipsacaceae by a number of morphological charac-
ters, pointing out that this left the Dipsacaceae as a much more homogenous family.

In 1938, Pai reviewed the Chinese species of Morina, recognising six species in two sections
(Acanthocalyx and Diotocalyx). He discussed the relationships of the additional species that had
been found as a result of the exploration of western China, notably by Forrest, Handel-Mazzetti,
and the French missionary collectors.

Embryological and cytological studies by Pouques in 1949, Vijayaraghavan & Sarveshwari in
1968, and Kamelina in 1976 and 1977 provide strong evidence for the retention of the
Morinaceae as a separate family, as do recent palynological studies by Blackmore & Cannon
(1983). In her thesis Verlaque (1976) separated the Morinaceae from the Dipsacaceae on
cytological grounds, and considered its origins to lie within the Valerianaceae, while in her later
publication (Verlaque 1977), she tabulates the differences between all three families in
morphology, cytology, and phytogeography.

REVISION OF THE MORINACEAE 3

Authors of major reviews of the flowering plants during the last 20 years have remained
divided as to whether or not to recognise the Morinaceae. Wagenitz (1964), Hutchinson (1973),
Heywood (1978), and Cronquist (1981) all include Morina in the Dipsacaceae, but agree that
many anomalous characters are involved. Thorne (1976) prefers to retain the Morinaceae, and
Takhtajan (1954) notes that it is a distinct family and refers to the work of Vijayaraghavan &
Sarveshwari (1968) and Kamelina & Yakovlev (1974), stating that the work of the last authors
demonstrates that it is near the Caprifoliaceae.

Relationships of Morina sensu lato (Morinaceae) within the Dipsacales

1. Morphology

It is generally agreed that Morina sensu lato must be placed in the Dipsacales, probably in the
Dipsacaceae, or in the Valerianaceae, or between them. There are, however, many difficulties
in placing it in either family, or even in regarding it as an intermediate. Many authors have
placed Morina in the Dipsacaceae, the most outstanding similarity being the possession by both
of involucels. Van Tieghem (1909) and others have pointed out that this character is not confined
to the Dipsacaceae, and occurs also in some genera of the Valerianaceae, some Compositae, as
well as in the Nyctaginaceae. Morphologically, Morina sensu lato is set apart from the
Dipsacaceae by the possession of a verticillate inflorescence, whorled leaves (not in Acantho-
calyx), the nature and form of the involucel and calyx, the aestivation of the corolla, the
insertion of the stamens (and staminodes) in 2 ranks, the ovary with 3 united carpels with 2
obsolete, and the 6-veined achene. It resembles the Valerianaceae in the number of carpels, the
zygomorphic corolla, and in some cytological characters (see account of cytology). The
Valerianaceae and Dipsacaceae are usually separated by the reduction or sometimes lack of a
conspicuous involucel in the Valerianaceae, the type of inflorescence, and the tricarpellate or
bicarpellate ovary. Table 1 summarises the differences between the two families and Morina
sensu lato.

Table 1 Summary of morphological differences between Morina sensu lato and the Dipsacaceae and
Valerianaceae.

Morina sensu lato

Leaves often whorled
(not Acanthocalyx)

Inflorescence verticillate

Involucel 12-nerved
(formed from fusion of
4 bracteoles)

Calyx zygomorphic

(2-lobed or sheathing)
Corolla lobes in bud with
posterior lobes overlapping
the 2 lateral lobes of anterior
lip

Stamens 2 (with 2 staminodes)
or 4 of 2 differing lengths

Ovary of 3 united carpels
—2 obsolete

Achene 6-veined

Nectaries at base of
filaments

Dipsacaceae

Leaves opposite

Inflorescence capitulate

Involucel 8-nerved
(formed from fusion of
2 bracteoles)

Calyx actinomorphic
or obsolete

Corolla lobes of bud with
2 lateral lobes of anterior
lip overlapping the
posterior lobes

Stamens 4 — all equal

Ovary of 2 united carpels
— 1 obsolete

Achene 8-veined

Nectaries round base of
style

Valerianaceae
Leaves opposite
(sometimes all basal)
Inflorescence cymose

Involucel usually
absent (8-nerved in
Triplostegia)

Calyx actinomorphic
or obsolete

Stamens 1, 2, 3 or 4
equal

Ovary of 3 united
carpels — 2 reduced
and empty or obsolete

Achene 8-veined

Nectaries often in spur
or sac at base of corolla

4 M. J. CANNON & J. F. M. CANNON

2. Embryology

The relationships of Morina sensu lato with the Dipsacaceae was extensively studied by
Vijayaraghavan & Sarveshwari (1968). They concluded that M. longifolia resembles the
Dipsacaceae in only a few embryological features, such as the dicotyledonous type of anther wall
development, the anatropous ovule, a Polygonum type of embryo sac, and the cellular
endosperm. It differs in many other characters, such as a persistent middle layer in the anther,
secretory anther tapetum, pollen grains shed at the 2-celled stage, absence of integumentary
vascular bundles, persistent pollen tube, ephemeral uninucleate antipodal cells, ruminate
endosperm, transverse division of the zygote, and Solenad type of embryogeny. They consi-
dered that there were so many major differences that Morina could not be considered as a genus
of the Dipsacaceae. They made no reference to possible relationships with other groups of the
Dipsacales. However, Crété (1963) stated that both Dipsacaceae and Valerianaceae could be
referred to the Seneciad type of embryogeny, which is far removed from the Solenad in his
classification. Table 2 summarises the embryological differences between Morina longifolia and
the Dipsacaceae.

Table 2 Summary of embryological and related characters (after Vijayaraghavan & Sarveshwari 1968).

Feature Morina longifolia Dipsacaceae

Anther wall Dicotyledonous type Dicotyledonous type
Endothecium Fibrous Fibrous

Middle layer Persistent Ephemeral

Tapetum Secretory Periplasmodial
Microspore tetrad Tetrahedral or isobilateral Tetrahedral

Shedding stage of pollen 2-celled 3-celled

Staminodes Present Absent

Ovule Anatropous, unitegmic and Anatropous, unitegmic

Integumentary vascular
bundles

tenuinucellar
Absent

and tenuinucellar
Present

Hypostase Present but no cavities Present, cavities contain
yellow liquid
Embryo sac Polygonum type Polygonum type
Antipodal cell 3-nucleate 3, uni or binucleate
Endosperm Cellular, ruminate Cellular, non-ruminate
Pollen tube Persistent Not persistent
Zygote Divides by transverse wall Divides by vertical wall
Embryogeny Solenad type Piperad type
Testa Only outer epidermis persists Not known
as a flimsy layer
3. Cytology
The following chromosome counts have been recorded for Morina sensu lato:
Morina persica L. 2n = 34 Kachidze (1929) and
Morina longifolia Wallich ex DC. 2n = 34 Verlaque (1976 & 1977)
Morina kokanica Regel 2n = 34 Verlaque (1976 & 1977)

In the Dipsacaceae, base numbers of x = 5,7,8,9, and 10 are known with various polyploids, but
2n = 34 appears nowhere in the lists. A similar situation is recorded for the Valerianaceae, with
base numbers of x = 7,8,9, and 11. The early count noted by Risse (1929) for Morina longifolia

REVISION OF THE MORINACEAE 5

as 2n = 16 has not been substantiated, and must be regarded as dubious in the absence of further
support. Poucques (1949) considered Morina to differ from the Dipsacaceae in the
homogenous, euchromocentric nucleus, and in the number and form of the chromosomes,
remarking on their very small size. He recorded the resting nucleus to be 6-7u, with one
nucleolus. Verlaque (1976 & 1977) studied three species of Morina and confirmed the findings of
Poucques (1949). She found the chromosomes at metaphase to be 0-7—1-5u, differing markedly
in size from the Dipsacaceae (1-5) and the Valerianaceae (0-7-4u). The centromere was
scarcely visible (more or less visible in the Valerianaceae and strongly marked in the Dipsa-
caceae). The resting nucleus was homogenous and euchromocentric in both Morina and the
Valerianaceae, but granulo-reticulate in the Dipsacaceae. She concluded that the origin of
Morina should be looked for in the Valerianaceae. Cronquist (1981) referred to fibrous
thickenings being present in all layers of the anther and connective in Morina, but only in the
endothecium and sometimes the endodermis in the Dipsacaceae, and the sperms of Morina as
having a little cytoplasm, whilst those of the Dipsacaceae are naked.

4. Palynology*

by S. Blackmore

Morinaceae pollen, which has been studied by Erdtman (1945, 1960a), Vinokurova (1959), and
Blackmore & Cannon (1983), is morphologically distinctive and diverse. Three main pollen
types, based on light microscopy of acetolysed (Erdtman, 19605) pollen, are distinguished here

(Fig. 1).

la. Pollen without prominent equatorial protrusions, ectoapertures colpate. Acanthocalyx-type
1b. Pollen with three prominent equatorial protrusions, ectoapertures porate:
2a. Equatorial protrusions domed, internal outline of pollen hour-glass shaped. Cryptothladia type
2b. Equatorial protrusions funnel shaped, internal outline of pollen similar to external. Morina type
Acanthocalyx type

Pollen tricolporate; triangular in polar view with the colpi in the rounded angles, subrhomboidal

in equatorial view. Ectoapertures broad colpi one third as long as polar axis, endoapertures

transverse colpi fused into a continuous equatorial band (zonorate). Exine thick, 6 um at poles

increasing to 10 um at colpus margins; sexine spongy, without distinct columellae; tectum

smooth with sparse microperforations. Polar axis 120-150 um long, equatorial axis 100-125 um.
Occurs in all species of Acanthocalyx.

Cryptothladia type
Pollen tripororate; rounded to triangular in polar view, prolate in equatorial view with
prominent domed equatorial protrusions. Ectoapertures pores situated in the equatorial
protrusions, endoapertures transverse colpi fused to form a narrow equatorial band (zonorate)
flanked by very marked thickening of the nexine. Exine thick, 5 um at the poles, 12 um at the
apertures; sexine spongy, thinner than nexine except at poles. Tectum smooth with scattered
microperforations towards the equator. Polar axis 100-120 um long, equatorial axis 55-70 um
(including protrusions).

Occurs in all species of Cryptothladia with some variation in the thickening of the sexinous
walls of the protrusions and the size of pores.

Morina type

Pollen tripororate; rounded to triangular in polar view, prolate to nearly cylindrical in equatorial
view with large funnel shaped equatorial protrusions prominent in either orientation. Ectoaper-
tures pores at the mouth of each funnel shaped protrusion, endoapertures pores, sometimes
with a very slight thinning of the inner nexine surface between them. Exine thick, 6 um at poles
increasing to 12 um at equator; sexine and nexine equal at poles, nexine much thicker at
equator. Tectum smooth at poles but often with two parallel bands of verrucae between each
*While this paper was in press, a paper by Dr R. Verlaque entitled ‘Contribution a!’étude du genre Morina L.’ appeared

in Pollen Spores 25: 143 — 162 (1983). This presents a review of the palynology, and although we cannot comment on it
here, the reference is included for the convenience of later workers.

M. J. CANNON & J. F. M. CANNON

Fig. 1 Scanning electron micrographs of pollen types. Scale lines 20 um.

. Acanthocalyx-type, A. nepalensis (Polunin, Sykes & Williams 4536), slightly oblique equatorial view
showing a colpate ectoaperture.

2. Morina-type, M. persica (Davis, Dodds & Cetik 19068), slightly oblique polar view showing verrucate

ornamentation at equator.

Cryptothladia-type, C. chinensis (Licent 4548), equatorial view showing domed apertural protru-
sions.

4. Morina-type, M. coulteriana (Vassiljeva 5487), equatorial view showing funnel shaped apertural

protrusions.

REVISION OF THE MORINACEAE 7

protrusion, microperforations more numerous towards the equator. Polar axis 150-270 um
long, equatorial axis 120-160 um (including protrusions).

Occurs in all species of Morina sensu stricto with some variation in ornamentation, size, and
shape.

Discussion

Pollen grains of Morinaceae differ from those of the Dipsacaceae in having spongy rather than
columellate sexine, lacking operculae over the apertures, and in their unusual pre-germinative
processes (Blackmore & Cannon, 1983). The zonorate endoapertures of two of the Morinaceae
pollen types have no counterpart in the Dipsacaceae and are probably secondarily absent from
the third pollen type. Similarly the domed and funnel shaped protrusions of Cryptothladia and
Morina sensu Stricto are unlike any apertural structures in the Dipsacaceae.

5. Conclusions

Major differences in morphology, palynology, embryology, and cytology have been shown to
occur between Morina sensu lato (now regarded by us as three quite distinct genera, Morina,
Cryptothladia, and Acanthocalyx), and the other members of the Dipsacales. The verticillate
flower arrangement, the nature and form of the calyx, the androecium, and the zygomorphic
ovary and achene are perhaps the most significant morphological features in this respect. The
pollen is most unusual, showing few similarities with any other, either within the Dipsacales or
with that of any other order of flowering plants. By palynological standards, these features alone
strongly suggest recognition at the family level. Embryological and cytological studies have been
made, so far as is known, only on members of the genus Morina sensu stricto, but it seems
probable that Cryptothladia would resemble Morina in these characters, while the position
of the less-closely related Acanthocalyx is not clear. Further wide-ranging research on these
to genera would be most welcome. The weight of evidence from these four fields is very much
in favour of the recognition of the Morinaceae as a distinct family within the Dipsacales.

Genera of the Morinaceae

The genus Morina has often been separated into two or three taxa (de Candolle (1830), Bunge
(1852), van Tieghem (1909)) solely on grounds of morphology. Our studies on the comparative
morphology of the group, and especially of its remarkable palynology, have convinced us that
the Morinaceae is best considered to comprise three quite distinct genera.

Acanthocalyx differs from the other two genera in so many ways, that differences are easier to
list than similarities. The most striking features are leaf arrangement, calyx shape, corolla shape,
stamen number, and achene shape. Table 3 summarises these distinctions.

Table 3 Characters distinguishing Acanthocalyx from Cryptothladia and Morina sensu Stricto.

Acanthocalyx Cryptothladia and Morina sensu stricto
Leaves opposite Leaves whorled

Flowers in close verticels to Flowers in well-separated verticels or in
subcapitate + cylindrical spikes of verticels

Calyx mouth obliquely cut, often Calyx distinctly 2-lobed, laminate

with 2 lateral and 3 (rarely 1)
posterior teeth, not laminate

Corolla + 2-lipped, lobes often Corolla markedly zygomorphic, 2-lipped,

nearly equal lobes unequal

Stamens 4 Stamens 2, with 2 sterile stamens or minute
staminodes

Nectary dome shaped, regular Nectary with 3 lobes or with 2 irregularly

lobed nectaries

8 M. J. CANNON & J. F. M. CANNON

Diotocalyx, the section recognised by de Candolle in contrast to his section Acanthocalyx, is
here divided between the genera Morina sensu stricto and Cryptothladia. These genera are
separated by the size and shape of the corolla, the shape and position of the staminodes or sterile
stamens, the number and shape of the nectaries, and the pollen characters already described in
detail. Table 4 summarises the major morphological differences.

Table 4 Morphological characters distinguishing Morina sensu stricto and Cryptothladia. See Fig. 2.

Cryptothladia Morina sensu stricto

Corolla equalling or shorter than the Corolla much exceeding the calyx

calyx

Corolla 24 lobed (sometimes scarcely Corolla clearly 5-lobed, limb expanded,
5-lobed), limb scarcely expanded, posterior posterior lip markedly 2-lobed, anterior
lip entire, slightly toothed or barely lip markedly 3-lobed

2-lobed

Sterile stamens 2, lobes equalling or Staminodes 2, minute, much smaller than
exceeding those of the fertile stamens, the fertile stamens, attached in the

at the base of the corolla tube or a corolla throat just below the fertile
short way above stamens

Flowers probably cleistogamous Flowers not obviously cleistogamous

Fig. 2 Diagram showing relative positions of fertile stamens, sterile stamens, and staminodes in (A)
Acanthocalyx, (B) Cryptothladia, and (C) Morina.

Systematic treatment

Morinaceae J. Agardh, Theoria systematis plantarum: 234 (1858).
Morinidia Raf. in Annls gén. Sci. phys. Brusc. 6: 88 (1820).
Type genus: Morina L.

Perennial herbs with rootstocks often covered with the remains of old leafbases. Leaves
opposite or whorled, exstipulate, often spiny, the petioles often connate, forming loose

REVISION OF THE MORINACEAE 9

cylindrical sheaths. Flowers hermaphrodite, zygomorphic, in verticillasters or subcapitulate
heads of biparous cymes. Involucel (epicalyx) of 4 fused bracteoles, smooth, with few —
numerous spiny teeth or setae, 12 nerved. Calyx epigynous, cupular, with an obliquely cut
mouth, or 2-lobed. Corolla tubular, scarcely to markedly 2-lipped, lobes 5, spreading, or 2-5
scarcely opening, the 2 lateral lobes of the anterior lip overlapping the median and the 2
posterior lobes overlapping them in bud. Stamens 4, all fertile, borne near the corolla mouth,
one pair above the other; or with 2 fertile borne near the corolla mouth or halfway up the tube,
and 2 sterile stamens (staminodes) borne at the base of the corolla or halfway up the tube.
Anthers introrse, 2-celled, the lobes subequal or markedly unequal, opening longitudinally.
Ovary inferior, 1-locular, formed by the fusion of 3 carpels, 6-veined, adherent to the calyx tube;
ovule solitary, pendulous, style slender, stigma simple. Fruit dry, indehiscent, enclosed in the
involucel and surmounted by the persistent calyx.

Morinaceae differs from Dipsacaceae in its often whorled leaves, the flowers in cymose
verticillasters (sometimes subcapitulate, but almost always with at least one detached lower
verticell), the venation and form of the involucel, aestivation, and the 6-veined ovary. It also
differs in many detailed palynological and embryological features and in chromosome numbers.
These are detailed in the appropriate introductory sections.

Key to the genera

la. Stamens 4; calyx limb oblique; plants not thistle-like ..................cseeeeeeeeeeee es I. Acanthocalyx (p. 9)
1b. Stamens 2 plus 2 staminodes; calyx limb 2-lipped; plants thistle-like:
2a. Corolla up to 1 cm, almost hidden by the much longer calyx, weakly lobed with 2 or 4 major
lobes; staminodes at base of corolla tube (4 way up the tube in C. chinensis) ............-002002000+
«ic con cud copes habdeloetle sored Set nig ne SU near Od coceIC eat Senttdites de Haseena lane ae anne arr II. Cryptothladia (p. 15)
2b. Corolla 2-5 cm long, well exserted from the calyx, strongly 5-lobed; staminodes borne 2 way
REE ON O MAUL Cra eectosemegteetaeacne Rarerae res atesace ecg somcecntsi seeaieGle sles ate bincsleaseve III. Morina (p. 24)
See Figs 4, 6, and 9.

I. ACANTHOCALYX (DC.) M. Cannon, stat. nov.

Morina Section Acanthocalyx DC., Prod. 4: 644 (1830).
Type species: Acanthocalyx nepalensis (D.Don) M. Cannon

Plants perennial, with a woody rootstock, often covered with the remains of old leaf bases.
Leaves opposite and decussate, linear, lanceolate or ovate, with or without spines or short setae.
Petioles joined to form a sheath, usually with 2 lines of hairs continued down the internode.
Uppermost leaves often more spiny at the base. Inflorescence subcapitulate, often with one or
more whorls of flowers below the primary head. Bracts free, connate or sheathing at the base.
Flowers sessile, shorter than the bracts, at least at the base. Involucel campanulate, very fragile
and papery at anthesis, enlarging and becoming more coriaceous in fruit, persistent, with few to
numerous teeth. Calyx cylindrical, the mouth obliquely cut, with a ventral fissure, often with 2
lateral and 3 posterior subspinose teeth, the laterals sometimes overlapping the fissure and
becoming sheathing. Corolla cylindrical, somewhat swollen below the limb with two posterior
and 3 anterior spreading lobes. Stamens 4, + equal, inserted just below the swollen part of the
corolla tube, scarcely emergent from it, anthers introrse, sub-equal. Nectary small, spherical at
the base of the corolla tube. Style equalling or exceeding corolla tube, stigma disc-shaped. Ovary
unilocular, ovule pendulous. Achenes smooth or rugose, apex somewhat cup-shaped.

Three species of Acanthocalyx are recognised by us: A. nepalensis, widely distributed in
Nepal, but spreading into Bhutan, Yunnan, Sichuan, and Xizang, A. alba (including Morina
leucoblephara Hand.-Mazz.), occurring mainly in western China and Xizang, and A. delavayi,
distributed mainly in Yunnan, Sichuan, and Gansu. See Fig. 3.

(7) soverpauts9yut
pue ‘(m) vqjv ‘y ‘(w) iMvavjap “y ‘(e) sisuajodau “y :saioads xX{jv20yjuvoy Jo uonNqiysiq ¢ “sIy

o0E

oAOL

008

REVISION OF THE MORINACEAE 11

Key to the species
la. Corolla pink or purplish, straight or slightly curved:
2a. Corolla tube hairy beneath, often somewhat curved, ovary glabrous ........... 1. A. nepalensis (p. 12)
2b. Corolla tube nearly glabrous beneath, straight, ovary pubescent ................... 2. A. delavayi (p. 12)
ibee@orollawhite or cream, markedly CURVE a.....0.-..scre-soesaeessnoseoececorscecssneoesaeneses 3. A. alba (p. 14)

The species of Acanthocalyx are difficult to distinguish and Table 5 of diagnostic characters is
provided to supplement the key.

Table5 The species of Acanthocalyx compared.

alba nepalensis delavayi
Calyx length (mm) (S—) 6-7 (-12) (9-) 11-12-5 (-14) (7—-) 10-14 (-15)
Calyx tube length (mm) 12:5 1-6-5 0-5—5-5 (-9)
Corolla curve +t +-(—) 0
Corolla hairs below 0-few sr few
Corolla hairs above =F af few to +
Corolla colour white (or cream) pink pink
Width of corolla tube (mm) 1-1-1-7 (1-4-) 1-8-2-1 (2-) 2:54
Ratio width/length 11-8-20 9-11 6-25-9 (-10)
Hairs inside involucel 0-very few on top 0-5-1:5 mm (0) top 0-5—
on veins) 2 mm (—base)
Involucel teeth (14-) 16 (-17) 9-15 (-16) 20-22 (-30)
Ovary hairs + 0 +
Leaf shape lanceolate (rarely linear to narrowly (broadly) ovate
linear lanceolate) lanceolate — lanceolate to
lanceolate

A number of specimens are difficult to allocate to the three species recognised. By far the
largest number of these are found in gatherings from Yunnan and Sichuan, e.g. McLaren’s
collectors 309D, B73, Potanin on 29.6.1983, Pratt 45, Rock 17725, and most collections of this
genus made by Maire. These gatherings appear to correspond with A. delavayi, except for the
ovaries which are glabrous. A few are similar in most respects to A. nepalensis, but have
pubescent ovaries (e.g. Forrest 7155) or two lines of hairs on the ovary (Ludlow, Sherriff &
Hicks 16573). A few gatherings from Yunnan and Sichuan include plants with both pink and
white flowers, e.g. Pratt 232; these are probably colour variants within the circumscription of A.
alba. Certain collections from Xizang, e.g. Soulie 71 and Dr King’s collector 8, probably have
specimens of A. alba and A. delavayi mixed on the same sheet. A gathering from Xizang
(Morton? 232) resembles A. nepalensis in most respects, but is recorded as having white flowers.
This may be an albino form of A. nepalensis or may involve a label error. The large number of
gatherings of intermediates and mixed collections suggest that different species often occur in
close proximity. This could imply that hybrids are frequent, and/or that the group is still
incompletely differentiated and that speciation is still continuing. However, until the situation
can be properly analysea using a range of experimental techniques on living plants, especially
from areas where ranges overlap, we prefer to draw attention to the variation by recognising
three species within Acanthocalyx, rather than to obscure the evidence by taking the easier
option of recognising one taxon only at the species level.

Pai (1938) recognised four species, but this approach does not seem to be supported by our
own observations. In his key characters used are the ciliate-spinose or densely ciliate nature of
the lower leaves, but this seems to us to be completely variable in all species; the corolla lobes are
said to be unequally lobed in Morina betonicoides (= Acanthocalyx nepalensis) and equally
lobed in Morina alba. Both species seem to us to have more or less unequally lobed corollas, the
upper two lobes being somewhat smaller than the lower three, and the two lateral lobes of the
lower lip being rather narrower than the central lobe. M. betonicoides is said to have emarginate

12 M. J. CANNON & J. F. M. CANNON

corolla lobes, whilst M. alba has undulate-crenulate lobes, but this does not seem to be the case
in the many specimens we have examined.

1. Acanthocalyx nepalensis (D. Don) M. Cannon, comb. nov.
Fig. 4.

Morina nepalensis D. Don, Prod. Fl. Nepal.: 161 (1825). Type: In Gossainthan Nepalensium, Wallich
[presumed to be Wallich 424(BM; E; K - type collection).

Morina nana Wallich ex DC., Prod. 4: 645 (1830). Type: Gossain Than, 1824, Wallich 424 (BM; E; K -
type collection).

Morina betonicoides Benth., in Ic. Pl. 12, t.1171 (1873). Type: Sikkim, 3400-3600 m, 12.7.1849, Hooker
s.n. (K— holotype; BM — isotype).

Low perennial herbs, with short woody rhizome, covered with remains of old leaf bases. Leaves
of sterile shoots entire, narrowly lanceolate or linear, glabrous or green above, glabrous and
paler beneath, with midrib and 2-4 nearly parallel raised veins, with subspinose setae (rarely
absent) on the margin, up to 20cm. Fertile shoots erect, up to 35 cm (rarely to 50 cm), with leaves
similar to those of the sterile shoots, except for the lowest pair, which are often small, nearly
ovate or spathulate, membranous and spineless. Upper stem leaves usually shorter than those of
the sterile shoots, opposite and decussate, with sheaths formed by the fusion of the two petioles;
with 2 lines of hairs continued down the stem below the sheaths. The uppermost pair of leaves
more spiny at the base, connate but not sheathing. Bracts ovate, spinose or ciliate, concave or
folded. Inflorescence + spherical, sometimes with one or two separated clusters below the
primary head. Jnvolucel cylindrical — campanulate, membranous at first becoming coriaceous in
fruit, with short, often reflexed hairs inside, 0-5—1-5 mm below the rim, with 9-15 (or rarely
more) setae, + equal in length and with numerous hairs around the rim. Calyx + cylindrical,
mouth obliquely cut, usually with 2 lateral and 3 posterior teeth, the latter sometimes connate
and with 1, 3 or more spines, the lateral teeth often spiny with 0-several setae. Calyx teeth (9-)
11-12:5 (-14) mm, the tube 4-6-5 mm with crisp hairs on the edge and sometimes also a few
inside below the posterior teeth, rarely to the base. Corolla pink, or purplish, curved, the tube
(1-4-) 1-8-2-1 mm wide, slightly swollen below the mouth with whitish spreading hairs all round.
Lobes of the limb spreading, 5, subequal, the lower 3 rather broader than the upper 2, the
middle lower lobe slightly broader than the 2 laterals. Stamens 4, all fertile, anthers introrse,
subequal, scarcely emergent from the corolla tube, 2 inserted at the base of the swelling, the
other 2 slightly lower. Stigma disc-shaped, scarcely papillose at the edge; style thread-like with a
tuft of hairs just below the stigma. Ovary glabrous or rarely with a few long hairs, unilocular,
with a pendulous ovule. Achenes rugose.

Distribution and ecology: A. nepalensis is the most commonly collected species in Nepal,
Sikkim, Bhutan, and southern Xizang. There are fewer records to the east of Lhasa and in its
immediate district, and it is thinly spread in Yunnan and Sichuan. In common with the majority
of species in the Morinaceae, A. nepalensis is a plant of high altitudes and has been recorded at
altitudes between 2500 and 4900 m, although the majority of records are in the ranges 3250-4750
m. It is recorded from a wide range of habitats, from rock ledges, alpine meadows, and dry
slopes to woodland, and is often described as being abundant in the places where it occurs. The
flowering season is mainly from June to July, but plants have been observed flowering as early as
May and occasionally as late as in August, September, and October.

2. Acanthocalyx delavayi (Franchet) M. Cannon, comb. nov.

Morina delavayi Franchet in Bull. Soc. bot. Fr. 32: 9 (1885). Type: In monte Hee Chan-men prope
Lankong, 2.6.1884, Delavay 52 (P — holotype).

Morina bulleyana Forrest & Diels in Notes R. bot. Gdn Edinb. 5: 208 (1912). Type: Chung Tien Plateau,
3400-3600 m, 1904, Forrest (E-—lectotype). Mountains near Tali [Mt. Tsang Chan], 20.6.1884, Delavay
90 (P —isotype).

Barleria crotalaria A. Léveillé in Reprium Spec. nov. Regni veg. 12: 285 (1913). — Lauener in Notes R. bot.

REVISION OF THE MORINACEAE 13

FLORA OF TIBET 1950
Coll. F. Kingdon-Ward

Reantho cody. ME pPAlensis QD. dee)

a a CE eee

Determinavit--—'-

Fig.4 Acanthocalyx nepalensis (D.Don) M. Cannon - the type species and a typical member of the genus
Acanthocalyx.

14 M. J. CANNON & J. F. M. CANNON

Gdn. Edinb. 32: 116 (1972). Type: Tong tchouan, Yunnan, 2700 m, Maire s.n. and Lou Pou, Yunnan,
3000 m, —.7.1912, Maire s.n. (E—syntypes).

Low perennial herbs, with short woody rhizome, covered with remains of old leaf bases. Leaves
of sterile shoots entire, linear-lanceolate to ovate-lanceolate, slightly spiny or spineless, up to 16
x 3-2 cm, fusing to form pale coloured sheaths. Fertile shoots erect, 10-35 (45) cm, lowest pair
of leaves ovate, with the sheath equalling or exceeding the lamina. Lower leaves spiny or +
spineless. Upper stem leaves ovate-lanceolate to broadly ovate, deep green above, somewhat
glaucous below, with 3-5 parallel veins, the petioles connate or sheathing, with 2 rows of white
hairs continuing down the stem from the join. Bracts broadly ovate, the edges spiny, becoming
more spiny towards the base. Inflorescence + spherical, sometimes with a few flowers in the axils
of the uppermost pairs of leaves. Involucel cylindrical-campanulate, fragile, papery at first,
becoming thicker and larger in fruit, teeth + uniform in length (16—) 20-30, hairy inside the top
0-5—1-5 mm, rarely to the base, very rarely without hairs. Calyx tube cylindrical-campanulate,
(7:5) 9-13 (-15) mm, mouth obliquely cut, usually with 2 lateral and 3 posterior teeth,
sometimes deeply cut, rarely quite entire, or the whole limb entire, terminating in a spine, the
outside usually glabrous, rarely with a few hairs at the tip. Corolla magenta, red-crimson or
purple, tube broad (2—) 2-5-4-5 mm, straight or + curved below the region of the swelling,
sparsely hairy to glabrous below and sparsely hairy to hairy above; lobes of the limb 3-5-6 mm.
Stamens 4, all fertile, the anthers introrse, the lobes subequal, scarcely emergent from the
corolla tube, 2 inserted at the base of the swelling, 2 just below. Stigma disc-shaped, somewhat
papillose at the edge. Style thread-like, with a tuft of hairs just below the stigma. Ovary
pubescent. Achenes rugose.

Observations: The syntypes of Barleria crotalaria A. Léveillé have glabrous ovaries and are
somewhat intermediate between A. delavayi and A. nepalensis.

Distribution and ecology: A. delavayi occurs in western China, mainly in Yunnan and
Sichuan, spreading northwards into Gansu and into the Xizang—Burma border region. It is a
plant of high altitudes and has been collected between 2500 and 4000 m. Collectors record it
from a variety of habitats, from dry stony ground to glacier meadows and the edges of pine
woods. It appears to flower a little earlier than A. nepalensis, with the majority of records for
May and June, with one single example from April and a few from July to October. However,
the earlier flowering span may be merely a reflection of the somewhat lower altitudinal range.

3. Acanthocalyx alba (Hand.-Mazz.) M. Cannon, comb. nov.

Morina alba Hand.-Mazz. in Sber. Akad. Wiss. Wien, Math.-Nat. 62: 68 (1925). Type: Likiang, Yulung
Shan, N.W. Yunnan, 3500-4000 m, 1914, Handel-Mazzetti 3799 (?W — not seen).

Morina leucoblephara Hand.-Mazz. in Sber. Akad. Wiss. Wien, Math. -Nat. 62: 68 (1925). Type: Yunnan,
1914, Handel-Mazzetti 3536 (?W — not seen).

Low perennial herbs, with short woody rhizome, covered with remains of old leaf bases. Leaves
of the sterile shoots entire, linear-lanceolate, rarely slightly ovate-lanceolate, with short hairs on
the margins, with or without spinose setae, lamina glabrous above and below. Fertile shoots
erect, 20-35 (rarely to 45) cm, lowest pair of leaves ovate, thinly coriaceous, deeply sheathing,
usually spineless, the sheaths as long as or longer than the lamina. Upper stem leaves entire,
opposite and decussate, linear to linear-lanceolate, green above, somewhat glaucous below, 3-5
veined, the sheaths becoming shorter near the top of the stem, with 2 lines of white hairs below
the fusion point of the petioles. Uppermost pair of leaves linear-lanceolate, slightly sheathing or
connate. Bracts broadly ovate, spinose, very prickly at the base. Inflorescence + spherical,
sometimes with one extra whorl in the axils of the uppermost pair of leaves. nvolucel cylindrical
—campanulate, persistent and enlarging in fruit, teeth 14-16 (-17), glabrous or with a few hairs
on the veins within the tube, with numerous hairs on the rim. Calyx tube cylindrical, 1-2-5 mm,
mouth obliquely cut, limb (5—) 6—7 (—12) times as long as the tube, with 3 posterior teeth and 2
lateral, sometimes deeply cut and sometimes spiny. Corolla usually white, cream or yellowish,
or shading to puce, the tube markedly curved, very hairy above, with few to 0 hairs below. Tube

REVISION OF THE MORINACEAE 15

12-20 x 1-1-1-7 mm, limb 5-lobed as in A. nepalensis. Stamens and stigma and style as in A.
nepalensis. Ovary pubescent. Achenes pubescent, at least when young, often at maturity
somewhat rugose.

Observations: Flower colour varies considerably. Ludlow & Sherriff 5017 records the colour
as creamy-yellow, ranging to puce. This seems to correspond to a number of gatherings by
several collectors who label their specimens as having white flowers, but in the gathering include
individuals which appear to have much darker flowers than the average for the sample. The calyx
is often flushed or almost completely coloured purple, both in the deep coloured specimens and
in those with obviously white flowers.

In his review of Chinese Morina species, Pai (1938) includes Morina leucoblephara as a
distinct species. Three gatherings recognised by him as M. leucoblephara have been seen by us
(Kingdon Ward 5894, Dungboo 4634 and Forrest 28444) and appear to be identical in all respects
to Acanthocalyx alba apart from their smaller stature. Handel-Mazzetti (1925) separates Morina
leucoblephara from M. alba by size, the deep ventral fissure of the calyx and the 3 posterior teeth
of the calyx, said to be totally connate in M. leucoblephara. Examination of the calyces of many
plants of all Acanthocalyx species shows a great variety in the depth of the three posterior teeth;
even in one plant there may be 3 or 5 deeply cut teeth, 3 closely connate teeth, or only one single
tooth. There seems therefore to be no recognisable distribution pattern of this character in the
populations. The depth of the ventral fissures in the Forrest specimen is 1-5 mm, which is within
the range shown by many of the A. alba specimens measured by us.

Distribution and ecology: A. alba has a similar distribution to A. delavayi, occurring in Gansu,
Sichuan, and Yunnan, but it does not extend so far to the west as that species. One specimen has
been seen from northern Assam and one from eastern Xizang. A. alba is also a high altitude
plant and has been collected from 2500-4250 m. It is noted as being abundant or extremely
common by a number of collectors who describe it as a plant of varied habitats from open, dry
pine forests to alpine meadows or even swamps. It flowers mainly between June and August,
with occasional records from May and October.

II. CRYPTOTHLADIA (Bunge) M. Cannon, stat. nov.

Morina Section Cryptothladia Bunge, Beit. Kentniss Flor. Russlands: 321 (1852).
Type species: Cryptothaldia parviflora (Karelin & Kir.) M. Cannon

Plants perennial, often tufted, with a woody rhizome covered with the fibrous remains of
previous year’s leaves. Leaves linear to lanceolate, whorled or rarely paired, the margins often
irregularly sinuate dentate and very spiny. /nflorescence verticillate, with several to many
verticillasters, flowers sessile or short stalked; sometimes flowering at ground level and
subsequently elongating to full height. Bracts free, connate at the base or fused for up to half of
their length, forming shallow cups. /nvolucel tubular to campanulate, enlarging and becoming
more rigid in fruit, with few to numerous teeth. Calyx cylindrical to campanulate, 2-lipped, each
lip 2- (or rarely more) lobed. Corolla tubular, scarcely emergent from the calyx, 4- (or scarcely 5)
lobed, rarely 2-lobed, the upper lip entire or scarcely 2-lobed, the lower lip often with 1 large
central lobe and 2 much smaller lateral lobes; the lips scarcely separating and the flowers
probably cleistogamous. Fertile stamens 2, inserted halfway down the corolla tube (one-third of
the way in C. chinensis), anthers introrse, subequal. Pollen yellow or rarely purple—violet. Sterile
stamens 2, inserted at or near the base of the corolla tube (4 of the way up in C. chinensis),
sterile anther lobes equalling or exceeding the fertile anther lobes. Nectaries 2, at the base of the
corolla tube, at or below the insertion of the sterile stamens. Style a little shorter than the corolla
tube or just equalling it. Ovary unilocular, ovule solitary, pendulous. Achenes rugose, with
obliquely truncate apex.

Six species of Cryptothladia are recognised by us: C. parviflora from the Tian Shan and Pamirs
of south-east U.S.S.R. and its borders with China, C. polyphylla from south of the Himalayas,
mainly in Nepal, but also occurring in Bhutan and south-eastern Xizang, C. kokonorica from
north of the Himalayas in Xizang, C. chinensis from north-western China, C. chlorantha

‘(o) NmMojpny ‘D pure ‘(Ww ) ydyddjod -D ‘(7 ) vyruvs0zyo
‘OD ‘(a) va1sou0yoy ‘Dd ‘(m) véopfiasnd *D ‘(e@) sisuauiys ‘D> :saieds vipyjyjojdduD Jo uonNqmsIq g ‘314

o0O0L

008

REVISION OF THE MORINACEAE 17

occurring to the south of the range of C. chinensis in China, and C. /udlowii from Xizang,
Assam, and Bhutan. See Fig. 5.

Key to the species
la. Bracts connate, forming shallow cups throughout the inflorescence, teeth of the involucel much
shartontmanthe tube 2.as lone OPleSS), uc cnecennes nes deo osenenaneenavensenseadi te 5. C. polyphylla (p. 20)
1b. Bracts free (fused at the top of the inflorescence in C. parviflora), teeth of the involucel at least
Eee SASIELIC RUDE S20. ee cc seek Cuca Syren a foc ds cee tidanc Dee awrmwereeneschsewacnticnaateesmescdencte su deans 2
PrMmPCAVE\CItINS OMMEATIYISO! oxtecccec nc ecsaceeep ier: Geten onl does emacs hove scllecianbe anicle ciate cUeineideceesteeee ps

3a. Bracts softly leafy with long acuminate tips, lower stem leaves not sheathing ...................
ca nS OgH ORD CH OBB ESE GREEEC Do CORED GCC AUREUS SEE CECE CTO TSEC GoCREn CRT On Ee CnncCeP Re Cr oAceee 4. C. chlorantha (p. 19)

3b. Bracts small, tri-lobed, or irregularly palmate -dentate, lower leaves deeply sheathing ......
RTs See nas Cas bau dame ecioron mane eemccian cece pace akin cuba ota sa tea baoe de aesioosonerees 6. C. ludlowii (p. ~,

em CACCS COATSECLY SIMMAte-CeMtatesr.. ccd teas eatsantant.caksivaducdaeancpesdbagantiorcts cbnoeduudaseewanteoe
Sees TACES COMMALE INSNE UDPEL WHOS 20. cnneqisaaantis dasa nvcesnnsasamasion aivaensicenaans 2. C. parviflora (p. S
BET ACTS MOtCOMMALE IID UPON, WAN OLIS er crap erecta cteaice ie ease ae ieee eae le ctclneajeeie siete olsen elvecte -tasieeins clei» 5

5. Staminodes 3 of way up corolla tube, calyx lobes 4, shallowly notched (up to 1 mm)........
Mee tee test ecesios sce cseiserecsre@ates senacedarecde ssedeaspniccekcsiesee seciees ties 1. C. chinensis (p. 17)

5. Staminodes at base of corolla tube, calyx lobes 5 (rarely 4), deeply notched (up to7 mm)
eee ce aac eet oct deauaae ice asbipa deus secon tae ataas cucavavvavessastessmesstcecate 3. C. kokonorica (p. 19)

1. Cryptothladia chinensis (Pai) M. Cannon, comb. noy.

Morina chinensis Paiin Reprium Spec. nov. Regni veg. 44: 122 (1938). Type: Tao river basin in meadows of
Maerhku, 2700-3000 m, 25.7.1925, Rock 12952 (E; K —lectotype).

Morina parviflora var. chinensis Batalin — referred to by Diels in Notes R. bot. Gdn Edinb. 5: 208 (1912) —
nom. nud.?

Herbs, with short rhizomes covered with old + complete leaf bases. Lower stem leaves up to 10
(20) x 8 (12) cm, whorled or paired, linear in outline, irregularly sinuose, spinose-dentate; the
teeth 2-3 fid with triangular lobules and apical spines and longer retrorse basal spines; the
petioles fused to form loose sheaths, glabrous. Upper stem leaves similar but smaller, up to 6 ina
whorl, free or scarcely connate, rarely with a very short sheath. Flowering stems up to S50 cm with
4 rows of hairs in upper parts, bearing several somewhat separated whorls of up to 20 flowers; the
stem visible between at least the lower whorls. Bracts leafy, free, up to 25 mm, the lower part
broadly ovate, markedly whitish and strongly net-veined; apices narrowly acuminate, spiny,
greenish, the blade spinose-dentate with many longer and often retrorse spines at the base, hairy
below and on the margin. /nvolucel narrowly campanulate to cylindrical, villous outside, tube
6-8 mm with 9-12 irregular spiny teeth (teeth not laminate at the base), often exceeding the
calyx, 4-5 mm, 3 or 4 teeth often longer than the rest. Calyx 6-8 mm, cylindrical, with a 2-lobed
limb, the lobes shortly bifid, the lobules slightly rounded or cuspidate, tube 3-4 mm with few
hairs outside, villous and glandular within. Corolla cylindrical, the throat somewhat swollen,
scarcely visible within the calyx lobes, limb more or less 2-lipped; lower lobe with 2 very small
lateral lobes, upper lobe entire or rarely slightly divided; lips scarcely open at anthesis,
pubescent and with sessile and stalked glands both within and without. Fertile stamens borne 7
of the way up the corolla tube, filaments short with a tuft of hairs just below the anthers. Anther
lobes unequal, one up to 1-5 times as long as the other. Sterile stamens with short filaments and
lobes as long as those of the fertile stamens, borne %3 of the way up the corolla tube, with 2
glandular nectaries at the base of the tube. Style about as long as the corolla, stigma disc-shaped.
Mature achenes surrounded by the enlarged involucel and surmounted by the enlarged calyx;
3-4 x 2-2-5 mm, plano-convex, furrowed on the abaxial face, the top edge of the adaxial side
rounded, smooth, not crenate.

Observations: The flowers are referred to as green on the labels of two specimens collected by
Rock. This may refer to the colour of the calyx, as the corolla is scarcely noticeable within the
more conspicuous calyx. Herbarium specimens appear to have whitish corollas.

18 M. J. CANNON & J. F. M. CANNON

A specimen collected at Sungpan hsein in Sichuan (Fang 4270) is probably best placed here. It
obviously has affinities with a specimen for which the collector’s name cannot be transliterated —
4484, which is similar in many respects to C. parviflora, but the shape of the calyx lobes is much
more like those of C. chinensis, as are the achenes.

Distribution and ecology: C. chinensis is a high altitude plant; most collections are from
localities at 3000-4000 m. It occurs in China in Xizang, Sichuan, Gansu, and Quinghai. Its
habitats are variously noted as alpine meadows, mountain tops, and roadsides, flowering in June
and July.

2. Cryptothladia parviflora (Karelin & Kir.) M. Cannon, comb. nov.

Morina parviflora Karelin & Kir. in Bull. Soc. Nat. Moscou 15: 373 (1842). Type: Alpis Alatau ad sinistram
ripam fl. Sarchan jacentis, Karelin & Kirilow (LE, not seen).

Herbs forming dense tufts, with sterile and fertile shoots. Rhizomes and lower parts of stems
covered with fibrous sheathing bases from old leaves. Lower stem leaves and leaves of sterile
shoots glabrous, linear with irregularly cut inciso-dentate spiny margins, up to 15 Xx 1 cm, in
whorls of up to 6; petioles expanding abruptly into broad, fused bases. Upper stem leaves similar
to those of sterile shoots, very shortly sheathing or barely connate, the uppermost whorl free;
covered with spreading crisped hairs, with stalked and sessile glands on the upper surface,
especially on the margins, the lower surface with fewer hairs and glands. Flowering stems up to
35 cm when mature, but very short at first and elongating during flowering; ridged and glabrous
below, thickly covered with spreading hairs and glands above, square in section between the
whorls of flowers, which are separate below but confluent above. Bracts up to 20 mm, free in the
lower whorls, connate or fused above; the bases broadly ovate and conspicuously net-veined,
the apices spiny, narrowly acuminate with long, narrowly-triangular spiny teeth densely covered
with spreading hairs and glands on the upper surface, fewer below. Jnvolucel cylindrical to
campanulate, villous and with occasional stalked glands at anthesis, tube 5 mm, with 4-5
broad-based laminate teeth; 2 teeth subequal, c. 4 mm, much longer than the others and
becoming rigid and spiny when mature. Calyx + campanulate, conspicuously net-veined, villous
and glandular, limb 2-lipped, each lip deeply bifid, lobes acute, sometimes spiny tipped. Corolla
pinkish, cylindrical, the throat somewhat swollen, just emergent between the calyx lobes; with
4-5 fimbriate lobes, the upper lip emarginate or bilobed, the middle lobe of the lower lip +
equalling the lateral lobes. Fertile stamens 2, borne halfway up the corolla tube, anther lobes
subequal, the shorter about % the length of the longer; filaments villous or with a tuft of
spreading hairs. Sterile stamens borne just above the base of the corolla tube, less than 3 from
the base; sterile anther lobes longer than the lobes of the fertile anthers, very narrow, somewhat
pigmented. Nectaries lobed, at the base of the corolla, extending upwards as far as the base of the
filaments of the sterile stamens. Style about as long as the corolla tube, stigma disc-shaped.
Mature achenes 6 X 4mm, rugose, shallowly furrowed and slightly convex on the abaxial face,
the apex obliquely truncate, tapering to an acute point.

Observations: The fruiting specimen 4484 (collector’s name not readable) on loan to us from
Peking is in habit very much like C. chinensis, especially the specimen Fang 4270. However, the
calyx lobes are acute, and the seed is much larger than that of typical C. chinensis. Pai considered
it to be Morina kokonorica (= Cryptothladia kokonorica), but it has only four calyx lobes, and
the plant is less spiny and the whorls of flowers are more separated than those of any specimen
we have seen of C. kokonorica. It corresponds to C. parviflora in seed, habit, and the shape of
the calyx lobes, and for these reasons we prefer to place it here. It has not been possible to
translate the blurred chinese characters on the label. The superficial similarity of this specimen
with C. chinensis may explain, in part, early references to Morina chinensis as a variety of M.
parviflora, but apparently without valid publication.

Distribution and ecology: This species is known from Turkestan, from the borders of
Kazakhstan with China (Dzungariau Ala-Tau), and from Kirgizia. It is stated to occur in the

REVISION OF THE MORINACEAE 19

alpine zone between 3000—4000 m, and to flower in June. Some collectors mention an unpleasant
odour, reminiscent of some Labiateae.

3. Cryptothladia kokonorica (Hao) M. Cannon, comb. nov.

Morina kokonorica Hao in Reprium Spec. nov. Regni veg. 40: 215 (1936). Type: Kokonor in den
Wustentalern, 12.9.1930, Hao 1268 (not seen).

Herbs forming tufts from short woody rhizomes densely covered with the fibrous remains of old
leaf bases. Lower stem leaves and leaves of sterile shoots paired or whorled, linear, coarsely and
irregularly sinuate-dentate, with broadly triangular spinose lobes, and with broad petioles fused
to form sheathing bases. Upper stem leaves similar, 5 or more whorled, up to 20 x 1 cm, free or
scarcely connate, with numerous retrorse and erect spines at the leaf base, sessile. Flowering
stems up to 40 cm, ridged and with 4 rows of hairs in the furrows above, terete and glabrous
below, bearing up to 7 whorls of 16-20 flowers, the stem becoming visible at least between the
lower whorls after flowering. Bracts 7-8 whorled, very spiny, often recurved and folded,
becoming very stiff after flowering; coarsely toothed and with short stiff yellowish spines,
broadly ovate at the base with a long acuminate spiny tip, glabrous. /nvolucel cylindrical-
campanulate, tube S—7 mm, often with 16-18 teeth, 2 (3-4) teeth much larger than the rest, often
with several lateral spines or teeth, the longest as long as or exceeding the calyx. Calyx tube 4-5
mm with few hairs; limb 2-lobed, each of which is very deeply divided almost to the tube, to form
(4) 5 (or rarely more) subequal lobes; the lobes ovate, acute to acuminate and often spiny
tipped, greenish and leafy at anthesis, becoming dry and spiny with inrolled edges in fruit, and
with a conspicuous tuft of hairs at the base of the tube. Corolla much shorter than the calyx, +
cylindrical with a few very short hairs on the outside; 4-lobed, the mouth scarcely opening.
Fertile stamens borne halfway up the corolla tube, with short filaments, anthers subequal. Sterile
stamens stalked, at the base of the corolla tube, with a nectary at the base of the filaments. Style
nearly equalling corolla tube, stigma disc-shaped. Mature achenes rugose; convex on the adaxial
face concave and deeply furrowed on the abaxial face; apex obliquely truncate, tapering to an
acute point, 5-5 x 3-3-5 mm.

Observations: One gathering at BM (Richardson 124) consists of very small, low-growing
plants, while others show individuals of varying heights. It seems probable that, in common with
many other alpine plants, this species may begin flowering very close to the ground, after which
the inflorescence elongates very rapidly to reach its maximum height, while flowers are still
present in the upper part of the inflorescence. Several collectors describe the corolla colour as
white.

Distribution and ecology: C. kokonorica is a high altitude species and collections seen range
from 3750-4725 m. All specimens seen by us are from Xizang, but the type locality is from
Kokonor in Quinghai. Its habitat has been recorded as stony ground and open hillsides, where it
flowers in June and July.

4. Cryptothladia chlorantha (Diels) M. Cannon, comb. nov.

Morina chlorantha Diels in Notes R. bot. Gdn Edinb. 5: 208 (1912). Type: East flank of Lichiang Range,
Yunnan, 3000-3300 m, —.6.1906, Forrest 2482 (E — holotype; K — isotype).

Morina chlorantha var. subintegra Pax & K. Hoffm. ex Limpr. in Beih. Repert. Spec. nov. Regni veg. 12:
497 (1922). Type: Ost. — Tibet, Ta Tsien lu, Sheto, steinige Halden des Tales vor dem Laniba, 4000 m,
Limprict 1834; Ta tsien lu, Dawo, Gata, auf den Grassmatten des Passes Hai tse schan am Dshava, 4350
m, Limprict 1899 (not seen).

Herbs with stout rhizomes densely covered with remains of old leaf bases. Lower stem leaves
ovate to obovate, 13-25 x 2-5-4 cm, somewhat toothed, especially near the base of the stem,
with short spines around the edge; gradually tapering into petioles which become joined to form
sheaths round the lower part of the stem, glabrous. Upper stem leaves entire, whorled, not
connate; sessile, lanceolate or broadly lanceolate, sometimes with small spines. Flowering stems

20 M. J. CANNON & J. F. M. CANNON

35-50 cm, hairy all round above, glabrous and somewhat ridged below, bearing up to 7
somewhat separated whorls of 12—20 flowers, the stem often visible between the whorls. Bracts
softly leafy in whorls of (3—) 4(-6), not connate, spreading and occasionally slightly recurved,
broadly ovate, tapering gradually into a long-acuminate leafy tip or into a short spine, usually
hairy on the margin and below. /nvolucels cylindrical campanulate, densely covered with hairs
and sessile and stalked glands, with 11-14 spiny teeth, the bases only rarely laminate, slightly
shorter than the tube, exceeding the corolla. Calyx, 8-11 mm, tube cylindrical 2-5-4 mm, with
hairs and sessile and stalked glands more numerous within than without; with a 2-lobed limb,
each lobe deeply notched, the lobules with rounded or spiny tips. Corolla hidden within the
calyx lobes, white, cylindrical, with a somewhat swollen throat and with numerous hairs and
sessile and stalked glands; 2-lipped, the lower lip with a large central lobe and 2 much smaller
laterals, the upper lip + entire or shallowly 2-lobed, the lobes shallowly crenate, the lips scarcely
separating at anthesis. Fertile stamens borne halfway up the corolla tube, anther lobes very
unequal, one sometimes twice the length of the other. Sterile stamens at the base of the tube,
with short filaments, and lobes as long as those of the fertile stamens, and irregularly lobed
nectaries at the base of the filaments of the sterile stamens. Style equalling corolla tube, stigma
disc-shaped. Mature achenes surrounded by the enlarged involucels and surmounted by the
enlarged calyx, 5 x 4mm; subglobular, convex on the abaxial face, scarcely furrowed, the apex
somewhat obliquely truncate, the top edge obtuse, entire or slightly crenate.

Distribution and ecology: C. chlorantha is distributed in Sichuan and Yunnan to the south of
the range of C. chinensis. It occurs in alpine meadows, grassland, and on moist shady cliff ledges
at altitudes between 3000 and 4425 m. It is in flower between May and July and is described as
having a rank odour.

5. Cryptothladia polyphylla (Wallich ex DC.) M. Cannon, comb. nov.
Fig. 6.

Morina polyphylla Wallich ex DC., Prod. 4: 644 (1830). Type: Nepal, Gossain Than, Wallich 425 (BM; K;
LE — type collection).

Tufted herbs, the rhizome thickly covered with the fibrous remains of several previous years’ leaf
bases. Lower stem leaves whorled, with long narrow petioles fused to form sheaths at least 3 the
length of the lamina, rarely equalling it, linear to linear-lanceolate, up to 40 x 1-5-2 (-3-5) cm,
very coarsely incisodentate with 3-5 fid, broadly based triangular lobules with apical spines,
glabrous or slightly pubescent. Upper stem leaves in whorls of (3—) 4-6, similar to the lower, but
smaller, the bases much less deeply sheathing, rarely almost free. Flowering stems up to 45 cm,
terete or slightly furrowed, with numerous spreading whitish hairs above, glabrous below,
bearing several whorls of up to 16 sessile or short stalked flowers in a dense cylindrical spike, the
lowest whorl rarely slightly separated from the rest. Bracts in whorls of (3—) 4-6, deeply connate
at the base forming a pale cup-shaped structure with conspicuous green net-veining, the free tips
of the bracts greenish, linear, with lateral spines tapering to a robust often yellowish spine.
Involucel (6—) 9-13 mm, shorter than the corolla, but becoming larger and more rigid in fruit;
cylindrical, toothed, usually pilose, rarely nearly glabrous; with many sessile and a few stalked
glands; teeth 8-10 or fewer, 1-3 mm at anthesis, with a broad laminate base, tapering to a
terminal bristle or spine, 1 or 2 longer than the rest. Calyx cylindrical to campanulate, 7-12 mm,
tube 4-5 mm, usually very pilose and glandular within; limb expanded with 2 spreading white or
lavender lobes, each deeply divided for at least half of its length, apices usually rounded, rarely,
spiny tipped. Corolla just visible within the calyx lobes, pink or white, equalling the involucel, +
cylindrical with 4-lobed scarcely expanded limb, the lips scarcely opening, dentate-fibrillate at
the edges. Fertile stamens 2, borne halfway up the corolla tube, anthers introrse, lobes subequal.
Pollen purple violet. Sterile stamens at base of corolla tube, short-stalked, lobes long and
tapering to a point, violet coloured, with a 3-lobed nectary at the base of each. Style equalling the
corolla tube, stigma disc-shaped. Mature achenes 4-75-5:5 x 2-75-3 mm, plano-convex,
obliquely truncate, the top edge of the adaxial face obtuse, crenate, scarcely furrowed on the

REVISION OF THE MORINACEAE 21

vi

“eG

FLORA OF NEPAL

Morina polypnylia Wall. ex OC.

Teghua, 8. of Twoicte, Kal! Gendaki.

Open gress sloper. Infleresoence re¢ end Light ereer

— 12,000 ft. Date 12.7.195h.

SEMNTON, Sykes & WiLtrams No, 1763

~¥ Cannerr

v Cyyprethladia, Pelyohyila (wats «bd e.)
™

- N -—-
Determinavit..-—.-- Se pn igs Se ot

Fig. 6 Cryptothladia polyphylla (Wallich ex DC.) M. Cannon - a typical member of the genus
Cryptothladia.

22 M. J. CANNON & J. F. M. CANNON

abaxial face, surrounded by the rigid expanded involucel and surmounted by the enlarged calyx.

Observations: The colour of the inflorescence has been variously described by collectors.
Most describe the ‘flowers’ as red, bluish-red, pink, white, or pale violet. As the corolla is so
small it is likely that it has been overlooked in a number of cases. In the specimens examined by
us the young flowers are generally much paler in colour than the mature ones. It seems very
probable that, as in the more familiar Morina longifolia and M. persica, the corolla changes
colour from white to deep red as it matures. The calyx is described as pale green or white, the
bracts being white or with pink margins. Many references to ‘flower colour’ probably refer to the
conspicuous calyx rather than the corolla. The whole plant is recorded as having an unpleasant
smell when bruised, and the leaves as being ‘strongly aromatic’.

Distribution and ecology: C. polyphylla has been collected mainly from Nepal, with a few
records from Bhutan, and two from the southernmost part of Xizang. It ranges in altitude from
2600 m to 4700 m. The flowering period appears to be mainly June and July, but it may extend
from April to September. It is recorded as being abundant in some localities and appeays to
prefer a variety of mostly open habitats, such as grassy hill-slopes, alpine meadows and the like,
although it has been collected from ‘the forest floor’ at 2600 m, which must be near the lower
limit of its altitudinal range.

6. Cryptothladia ludlowii M. Cannon, sp. nov.
Fig. 7.

Ab altis speciebus Cryptothladiae spica cylindrica et densa, bracteis trilobis differt.

Herbae caespitosae, caulibus floriferis anthesis initio brevibus tum celeriter per anthesis ad 47 cm altis
extendentibus. Folia lanceolata, 4-verticillata, integra vel paucidentata; petioli generatim connati vaginas
laxas facientes, eas caulis basi profundissimas eas super minores, petioli supremitamen vix connati. Spika
densiflora, cylindrica verticillis 20 ut minimum, in quoque 8—12 floribus. Bracteae parvae, imbricatae,
saepe rubrae, verticillo infimo lanceolato superioribus trilobis, lobis spinoso-dentatis medio eis lateralibus
maiore, apicibus acutis vel cuspidatis. Involucellum dentibus 4-8 (raro-12) uno quam ceteris saepe
longiore latioreque, basi laterali, laminaribus anthesis initio mollibus et fragilibus, grandioribus rigidisque
in maturitate. Limbus calycis bifidus, lobis breviter bifidis, lobulorum apicibus obtusis vel cuspidatis.
Corolla parva, quadriloba, lobo posteriore fimbriato, anteriore integro super posteriorem arcuato.
Stamina fecunda 2, medio corollae tubo inserta; staminodia imo inserta, lobis sterilis eos staminum
fecundorum aequantibus vel superantibus. Achenia parva, 3 X 1-75 mm. subglobosa.

Type: Bhutan, Ju La, Mangbe Chu, 4250 m., Ludlow, Sherriff & Hicks 16903 (BM — holotypus; E —
isotypus).

Herbs, the stem base and top part of the rhizome thickly covered with the fibres of old leaf bases.
Lower stem leaves and leaves of the sterile shoots, in whorls of 4, lanceolate, entire or slightly
toothed, with rather broad petioles fused together into loose sheaths, 5—7-5 x 10-12 mm, the
margins with hairs and short spines, slightly glandular — pubescent above. Upper stem leaves
similar to the lower, with broad based petioles forming shallower sheaths, the uppermost whorl
often barely connate. Flowering stems very short at first, 2-3 cm, elongating rapidly up to 47 cm,
terete or slightly grooved, densely covered with whitish, spreading, mostly retrose hairs.
Inflorescence cylindrical, dense, up to 20 cm, often with 20 or more close packed whorls, 8-12
flowers in each whorl. Bracts small, leafy, regularly opposite and decussate, imbricate, lowest
whorl rather similar to leaves but smaller; usually 3-lobed, lobes irregularly spinose-dentate,
acute or cuspidate, the centre lobe larger than the laterals, often reddish or deeply coloured.
Involucel campanulate — cylindrical, pilose, often with stalked or sessile glands; teeth 4-8 (-12),
one often larger than the others, broad at the base, laminate, soft and fragile at anthesis,
becoming larger and rigid in fruit. Calyx tube cylindrical-campanulate, 2-5-4 mm at anthesis;
limb bifid, the lobes shallowly divided, lobules rounded to apiculate, glabrous without, with long
white hairs and glands within. Corolla scarcely visible within lips of the calyx, + cylindrical with
a scarcely enlarged throat, glabrous without, 4-lobed, the posterior lobe + fimbriate, the
anterior lobe entire, markedly exceeding the posterior and arched over it, with strong erect hairs
on the inner surface, with 2 very small lateral lobes; the whole closed or scarcely separating at

REVISION OF THE MORINACEAE

FLORA OK! Krad man No. 16903

a ~ oka At. (4g BBQ Date ee ~¥
: € -{

A ama de am Lat Tonk

he Nt nent TN, ate eae)
ee Jw NAN ottin
qk
TYPE SPECIMEN %
ePine PRS Lla elbe wh DS Coll. P. Lupcow, G. Suegsiee & J. H. Hiexs

Cxptoti facia Mmdlloum MT Conn:

Fig. 7 Holotype specimen of Cryptothladia ludlowii M. Cannon.

24 M. J. CANNON & J. F. M. CANNON

anthesis. Fertile stamens 2, with short filaments, borne halfway down the corolla tube, anther
lobes subequal, filaments with a tuft of hairs below the anthers. Sterile stamens 2, at the base of
the corolla tube, shortly stalked, deeply pigmented, the sterile anther lobes as long as or longer
than those of the fertile anthers, nectaries irregularly lobed, at the base of each sterile stamen.
Style a little shorter than corolla tube, stigma disc-shaped. Mature achenes 3 X 1-75 mm,
subglobular, with a shallow furrow on the abaxial face, apex slightly obliquely truncate, obtusely
pointed, somewhat rugose.

Observations: We take particular pleasure in naming this species after the late Frank Ludlow,
Honorary Associate of our Department, who was for many years an outstanding specialist on
Himalayan plants and who, with George Sherriff, made many important collections in this
region. A specimen from Bhutan Gould 486 (K) has a few long, narrow, spiny tipped bracts at
the base of the inflorescence, becoming smaller and similar to those of typical C. ludlowii
towards the apex. It closely resembles C. /udlowii in all other respects. r

Distribution and ecology: C. ludlowii has been collected at altitudes of 3500-4250 m in Bhutan
and its borders with Xizang and Assam. It is reported from a variety of habitats, such as amongst
shrubs, on rocky hillsides, and on cliff ledges. Flowering is from May to September. The flowers
are described as yellow, or yellowish green, but some obviously also have a red or violet
pigment. The flowers are said to be sweet scented.

Specimens seen: Bhutan: Ju La, Mangbe Chu, 4250 m, 19.7.1949, Ludlow, Sherriff & Hicks 16903 (BM —
holotype; E — isotype); Me La, south side, 4250 m, 29.5.1949, Ludlow, Sherriff & Hicks 20305 (BM);
Kantanang, Tsampa, 3850 m, 4.6.1949 Ludlow, Sherriff & Hicks 19050 (BM; E); Ridge of Tashigong
Kurted, 3700 m, 21.8.1915, Cooper 4525 (BM). Assam: Orka La, 4000-4250 m, 29.9.1938, Kingdon-Ward
14310 (BM). Xizang; Mago, 4000-4250 m, Kingdon-Ward 12383 (BM; K).

II. MORINA L.
Morina L., Gen. Pl. ed.5: 16 (1754).
Asaphes Sprengel, Syst. Veg. 4 (2): 222 (1827), nom. superfl. (Art. 63.1).
Morina Section Diotocalyx DC., Prod.: 4: 644 (1830), excluding M. polyphylla Wallich ex DC.
Type species: Morina persica L.

Plants perennial with woody rhizomes, often covered with the remains of previous year’s leaf
bases. Leaves linear to oblong-lanceolate, spinose, incisodentate to almost entire, whorled or
rarely paired, petioles often fusing to form loose sheaths. Inflorescence of several to many
verticillasters, flowers sessile or short stalked. Bracts free, variable in size, usually spiny at least
at the base. Involucel cylindrical to campanulate, with 8-16 teeth, 2 of which (usually opposite)
are markedly longer than the rest. Calyx campanulate, with a 2-lipped limb, the lips somewhat to
deeply 2 lobed, persisting at least until the fruit is mature. Corolla with a long narrow tube and
spreading into a + 2-lipped limb, the upper lip 2-lobed, the lower 3-lobed. Fertile stamens 2,
adnate to the throat of the corolla; staminodes 2, cordate, within the corolla tube, Nectary 1,
3-lobed, anterior, at the base of the corolla tube. Style usually exceeding the stamens, stigma
disc-like or slightly domed. Ovary unilocular, with a single pendulous ovule. Achenes rugose,
somewhat or markedly obliquely truncate.

Four species of Morina are recognised by us: M. kokanica from the Tien Shan, M. coulteriana
from the U.S.S.R., the western Himalayas, and Xizang, but absent from the eastern Himalayas,
M. persica spreading from the Balkans through to the western Himalayas, and M. longifolia
from the west and central Himalayas. See Fig. 8.

Key to the species
La. Leavesisimple 5.20.25... c05. snasciOnae sc oraemaeepeees cemecenat sae hac cceemnet es eemenseneaeee 1. M. kokanica (p. 26)
1b. Leaves coarsely dentate-serrate:
2a. Corolla yellow, calyx lobes deeply notched ...................ceceeeeeeeeeeeeeneeen 2. M. coulteriana (p. 26)
2b. Corolla white or pink, calyx lobes entire or shallowly notched:
3a. Filaments 10mm, teeth of involucel equalling or longer than the tube .......... 3. M. persica (p. 28)

3b. Filaments 1-2 mm, teeth of involucel shorter than the tube .................... 4. M. longifolia (p. 30)

‘(m) vy ofisuo]
‘W pure ‘(W) voisuad ‘Ww ‘(o) vuniuaynod ‘W ‘(e) vaIunyoY "Fw :seideds puluop Jo UOTNGINSIq g “Sly

26 M. J. CANNON & J. F. M. CANNON

1. Morina kokanica Regel in Bull. Soc. Nat. Moscou 40: 14 (1867). Type: Kokan, Turkestan,
1866, Sewerzow s.n. (LE, not seen).

Robust herbs with rosettes of sterile leaves and flowering shoots. Rhizomes woody, the top often
covered with the fibrous remains of petioles from previous years leaves. Leaves of sterile shoots
and lowest stem leaves glabrous, oblong lanceolate 25-35 x 2-4-5 cm, entire or rarely with
occasional small teeth or prickles, midrib prominent, sometimes elongated to form a small
spine; petioles free or fused to form short sheaths of 2-4 mm. Upper stem leaves similar to lower,
opposite or 3-4 whorled, petioles free. Flowering stems erect, somewhat square in section,
grooved and ridged, glabrous below, with well-separated whorls of numerous short-stalked
flowers; sometimes with short flowering shoots in the axils of the upper leaves. Bracts often
rather leaf-like in the lowest whorl, broadly ovate with an acuminate tip tapering into a spine,
the margin with a few rigid spines, somewhat villous below, the bases scarcely overlapping each
other. Involucel cylindrical, the tube at anthesis 5-6 x 3-4-5 mm, becoming twicé as long or
more at maturity, villous, with 12-16 teeth, two longer than the others, the longest shorter than
the tube. Calyx cylindrical-campanulate, 5-6 x 3-4-5 mm, often densely glandular hairy, limb
10-16 mm, 2-3 times as long as the tube at anthesis, lobes shallowly bifid, tips obtuse. Corolla
pale purple or pink with a darker area on the lower lip, the tube long and narrow, 35-45 x 1-3
mm, the throat somewhat swollen, limb deeply cut, the lobes 10-15 x 6-7 mm, densely
glandular hairy on the tube, somewhat less so on the interior and exterior of the limb. Fertile
stamens 2, filaments about 7 mm, with a tuft of hairs just below the anthers, anther lobes
subequal, the longer 2 mm, the shorter 1-7 mm. Staminodes 2, minute, + heart shaped, inserted
2-3 mm below the mouth of the corolla on the anterior side. Nectary 1, 3-lobed, anterior at the
base of the corolla tube. Style slightly exceeding the stamens, stigma disc-shaped. Achenes
rugose, + plano-convex with prominent diagonal veins, the apex obliquely truncate, somewhat
fluted.

Distribution and ecology: M. kokanica is apparently confined to the Tien Shan and Pamir-Alai
regions of Soviet central Asia at altitudes of 1600-3000 m. The recorded flowering period is June
to August.

2. Morina coulteriana Royle, J//us. Bot. Himal.: 245 (1835). Type: Kunawar, Royle s.n. (LIV —
holotype).
Fig. 9.

Morina breviflora Edgew. in Trans. Linn. Soc. Lond. 20: 62 (1846). Type: Pharkia near the Niti Pass in
Kumaon, Edgeworth 259 (K — holotype).

Morina lehmanniana Bunge, Beit. Kenntniss Flor. Russlands: 321 (1852). Type: In den Alpen des Karatau,
12.9.1841, Lehmann s.n. (P, not seen).

Robust herbs with rosettes of sterile leaves and flowering stems. Rhizomes woody, often
covered with old leaf bases. Leaves of sterile shoots and lower stem leaves 20-25 X 2:5 cm,
glabrous, pinnatilobate, with spiny simple or compound teeth, petioles soft, fusing to form deep
sheaths of 8-15 mm. Upper stem leaves similar but smaller, in whorls of 3-4 (—S), free or with
short sheaths up to 3 mm or rarely more. Flowering stems erect, up to 1-25 m, stems ridged and
glabrous below, square sectioned, villous and often deeply purple — pigmented above. Flowers
in up to 7 whorls, the upper 2-5 whorls confluent, with 2-4 separated whorls below, flowers 50 or
more per whorl, sessile in the outer part of the whorl, shortly petiolate (up to 2 mm) nearest to
the stem. Bracts broadly ovate with long acuminate tips, very spiny, often villous especially near
the stem, conspicuously net veined. Involucel + cylindrical, villous, 4-9 x 2-5—4 mm, with 9-12
teeth, 2 much longer than the others, the longest at least equalling or up to nearly twice as long as
the tube at anthesis. Calyx glabrous to villous, with a large tuft of hairs at the base, glabrous or
very shortly pubescent within, cylindrical-campanulate, 3-6-5 x 2-4 mm, usually more than
twice as long as wide at anthesis. Limb 2-lipped, villous, pubescent or glabrous without, glabrous
within; lips deeply bifid for 2-5 mm, apices acute or apiculate, with a short but prominent vein
within, often extended beyond the tip to form a spine. Mature calyx often equalling the corolla

REVISION OF THE MORINACEAE 27

ore emery s-commmenne neomtcarremanitme ei: th

FLORA OF CHITRAK

Loe. Alt Aisi
Bavor oon yoo
$2 $ Tide Haw

dey tog toe

tole. bn oh, Lh,
ictinn eoulteriana Royle
Goll, J.D, A. STAINTON No.2785

Fig.9 Morina coulteriana Royle — a typical member of the genus Morina.

28 M. J. CANNON & J. F. M. CANNON

tube. Corolla yellow, tube villous, without glandular hairs or with very few, long and narrow,
25-30 x 1-2 mm, throat somewhat expanded, limb 2-lipped, the upper lip 2-lobed, the lower
3-lobed, median lobe of the lower lip 5-5-10 x 3-5-6 mm. Fertile stamens 2, the filaments 2-3 mm
at anthesis, inserted just within the corolla tube, with a tuft of long hairs below the anthers.
Anther lobes unequal, one *% the length of the other. Staminodes minute, heart shaped, inserted
2:5—4-5 mm below the mouth of the corolla, below the anterior lip. Nectary 1, 3-lobed, anterior
at the base of the corolla tube. Style slightly exceeding the stamens, stigma disc-shaped. Achenes
with convex adaxial face, transversely rugose, abaxial face rugose, with very deep longitudinal
furrows; apex somewhat obliquely truncate, with conspicuously undulate margins.

Observations: Plants from Xizang have slightly smaller corollas than those from the western
end of the distribution range, the median lobe of the anterior lip of the corolla measuring 5-5-7
x 34 mm, the width of the lateral lobes 2-5-3 mm, and the other lobes being cérrespondingly
smaller than those of plants from further west. The length of the filaments also differ slightly,
those of the Tibetan plants being slightly longer (2-75—3-5 mm) than those of their western
counterparts (2-2-5 mm). A collection from Kashmir — Ludlow & Sherriff 9123 from the
Macchel Sapphire Mines, Kishtawar, has flowers similar in size to the Tibetan plants, but careful
measurements of many individuals from both east and west show no other significant differ-
ences. No formal recognition of this variation therefore seems appropriate. A gathering from
Kashmir — Robson 2013 consists of plants with rather narrow, almost spineless leaves, but
corresponds to typical M. coulteriana in all other respects.

The type of M. lehmanniana has not been seen by us. It was described by Bobrov (1957) as
from the upper Zaravshan, near the village of Fon and as being in the Paris Herbarium. A
specimen from the Bunge herbarium from Zaravshan has been seen; it is identical in all respects
with M. coulteriana. Specimens from this area tend to be conspicuously white-hairy above, but a
number of Afghan and Kashmir plants also share this character.

Distribution and ecology: M. coulteriana occurs in the western Himalayas, the Hindu Kush,
and central Afghanistan, and as far north as the Tien Shan, and also in eastern Xizang, but
without intermediate stations in Nepal, Sikkim, or Bhutan. This kind of disjunct distribution has
been referred to by several authors, e.g. Stainton (1972), Meusel (1971), and Kingdon-Ward
(1936), and is shown by a large number of genera. In many cases such plants have been described
as closely related vicarious species. In this instance, there seems to be no justification in
recognising the Tibetan plants as a distinct taxon. It is improbable that the species occurs in
intermediate Nepalese localities, as other similar species have been collected there and it seems
likely that a yellow-flowered Morina would have readily attracted the attention of the many
experienced collectors that have now travelled extensively in that country. There seems to be no
obvious reason for its absence from the central Himalayas, since it is a species of open habitats
and has been recorded from many substrates and habitats. It seems possible that the wetter
monsoon conditions of the eastern Himalayas do not suit it so well, although it has been
recorded as growing in a number of wet situations — streamside, wet sand, and moist pastures,
and many arid habitats are available in Nepal behind the main range. It is a plant of high
altitudes, rarely recorded below 2500 m and attaining 4000 m in Nepal; most records are found
between 3000-3500 m. Aitchison 746 from the Kurram Valley in Afghanistan has the fieldnote
‘at and above 11000-13000 ft (3300-4000 m) this yellow flowered species quite replaces M.
persica.’ It has been reported from a wide range of habitats, both sheltered and exposed,
possibly favouring dry conditions but also occurring in damp areas. The flowering period is
mainly from June to August, with some plants still in flower in September.

3. Morina persica L., Sp. Pl. 1st ed.: 28 (1753). Type: Habitat in Persia ad Hispaham, Herb.
Linn. 44/1 (LINN, isolated corolla only).

Morina orientalis Miller, Gard. Dict. 8th ed. (1768), nom. superfl. (Art. 63.1).
Morina verticillata Moench, Meth. Suppl.: 186 (1802), nom. superfl. (Art 63.1).
Morina wallichiana Royle, Illus. Bot. Himal.: 245, t.55 (1835). Type: Mussooree, Royle (LIV — holotype).

REVISION OF THE MORINACEAE 29

Morina aucheri Jaub. & Spach, Illustr. 5: t.429 (1854). Type: Persia Australis, Aucher Eloy (BM; K) &
Kotschy ?549 and ?2190 (BM; K; LE —syntypes).

Morina graeca Jaub. & Spach, Illustr. 5: t.429 (1854). Type: Greece, Sibthorp s.n.; Bory de St Vincent (not
seen); Heldreich s.n. (E; K; LE —syntypes).

Morina tournefortii Jaub. & Spach, Illustr. 5: t.429 (1854). Type: Armenia circa urbem Erzeroum,
Tournefort s.n.; Asie minore, Aucher-Eloy s.n.; Jaubert s.n.; and Pinard s.n. (syntypes, not seen).

Morina subinermis Boiss., Diagn. I1, 6: 94 (1859). Type: Turkey, in Bithynia. Duparquet (not seen).

Morina persica subsp. turcica Hal. in Oester. Bot. Zeit. 41: 409 (1891). Type: Macedonia prope Demirka-
pu, 19.8.1889, Farmanek s.n. and Tekir Dagh, in saxis promontorii Hodja Burnu, inter pago Panidos et
Kumbaos litoris Propontidis, 2.7.1890, Degen s.n. (not seen).

Morina turcica Degen & Hal. ex Celak. in Bot. Jb. 17: 396 (1893), nom. nud. (Art. 32.1).

Morina spectabilis Gontsch. nom. nud. (Art. 32.1) ?in Herb. LE, cited by Komorov in Fi. U.R.S.S. 24: 8
(1957).

Robust herbs often forming large clumps of sterile rosettes and flowering shoots. Rhizome
woody, the top often covered with the fibrous remains of old leaves. Leaves of sterile shoots and
lower stem leaves up to 25 X 2 cm, 3-4 per whorl, glabrous, + linear, with fairly regular 2-5
lobed, narrowly triangular teeth, petioles broad based, connate, forming short sheaths of 2-3
mm. Upper stem leaves smaller, similar to the lower leaves, but petioles scarcely connate or
free, 3-4 leaves per whorl. Flowering stems erect, up to 1:25 m. Stems densely villous and
glandular, square stemmed above, becoming terete or somewhat ridged and with fewer hairs
below. Flowers sessile or shortly-stalked, up to 50 or even more per whorl, in up to 8 (rarely
more) whorls, sometimes with small flowering shoots in the axils of the upper leaves. Whorls
usually well separated from one another, the uppermost sometimes confluent. Bracts very
variable, usually twice as long as the calyx whorls, linear with long acuminate tips, gradually
tapering into a long spine, or broadly ovate with a shortly spiny acute apex; almost glabrous to
densely villous and glandular, spiny, sinuate-dentate and similar to the leaves to almost entire
with few spines. Bracts of lower whorls free or slightly overlapping at the base, upper bracts free
or very rarely connate. /nvolucel cylindrical campanulate, tube 7-10 x 3-5-5 mm, densely
villous and with sessile glands and stalked glandular hairs to almost glabrous; teeth (8—) 10-14
(-16) of varying lengths, two usually much longer than the others, the longest much exceeding
the tube or rarely equalling it, often up to 14 mm or more at anthesis, the 2 longest teeth broad
based, laminate at anthesis, becoming inrolled and rigid at maturity. Calyx + glabrous, or very
shortly pubescent outside, with a tuft of long white hairs at the base, tube broad, often as wide as
deep, 3-5 mm long, lips often much longer than the tube, apex obtuse, nearly truncate or very
shallowly emarginate. Corolla white, becoming pink or red, the tube very long and narrow 35-45
mm, with or without simple and glandular hairs, the throat somewhat swollen; limb + 2-lipped,
the upper 2-lobed, the lower 3-lobed, lobes 12 X 6mm or more. Fertile stamens 2, filaments 9
mm or more, inserted at the top of the tube; anther lobes subequal. Staminodes minute, heart
shaped, inserted 4-5 mm below the mouth of the corolla, below the anterior lip. Nectary 1,
3-lobed, anterior at the base of the corolla tube, Style longer than the stamens, stigma
disc-shaped. Achenes rugose, adaxial face, + convex, somewhat longitudinally furrowed,
abaxial face furrowed, transversely rugose, apex obliquely truncate.

Observations: The protologue in the original Linnzan description refers directly with an
unmodified descriptive phrase to Hortus Cliffortianus. For this reason, a specimen in that
collection would be a preferred choice for typification. However, since there is no Morina
specimen in that herbarium, the fragmentary material in the Linnean herbarium (corolla only)
has been adopted as the best available alternative. The geography of the group and morphology
of the corolla leave little doubt as to the validity of the typification.

Plants from the Mussoorie area of India were described by Royle as M. wallichiana, nearly
allied to M. persica, but having spiny lobed leaves, the upper stem and corolla softly villous, and
the calyx lobes entire or shortly emarginate. The type specimen is a plant with rather ovate
bracts and few spines, and corresponds closely with a number of specimens seen by us from this
area. These tend to be greyer in colour and with fewer spines on the margins of the bracts than
typical M. persica. However, a range of bract variation occurs in M. persica from broadly ovate

30 M. J. CANNON & J. F. M. CANNON

to narrowly linear, with apices acute and shortly spined to narrowly acuminate and long spined,
the plants from Mussoorie being at one end of the variation range and not meriting the retention
of specific rank. Although many plants of this kind appear to occur in the Mussoorie area, similar
individuals have also been seen throughout the remainder of the range of M. persica, from
Afghanistan to Bulgaria. We do not, therefore, consider Royle’s taxon to merit retention as a
geographically based subspecies, and it is best considered as a rather extreme form within the
overall variation pattern of M. persica. An interesting specimen collected in the Muree Hills of
Pakistan (Prescott Decie s.n. 1915) is probably another extreme variant of M. persica, one
specimen on the sheet having leaves that are almost entire or with very small teeth. The bracts
are broadly lanceolate, entire or with a few simple teeth, the upper whorls are connate, forming
a small funnel-shaped sheath. The corollas have a dark coloured blotch on the lower lip; the
colour appears to be yellow, but there is no field note in support of this conjecture. The specific
rank of M. subinermis is maintained by Matthews (1972). It is known only from the type, which
we have not seen. It should probably be regarded as a synonym of M. persica, as the description
is in accord with the variant described above (Prescott Decie s.n., 1915).

Distribution and ecology: M. persica is distributed across western Asia and reaches south-
eastern Europe in Bulgaria, Romania, and Greece. It is widespread in Turkey, Iran, Israel, and
Lebanon. It is common in Afghanistan, Kashmir, and Pakistan, where its range overlaps those
of M. longifolia and M. coulteriana. This species occurs at lower altitudes than any other
members of the Morinaceae, being found at altitudes below 500 m in Turkey, although it has
been recorded from 3730 m in Afghanistan. It is a species of dry rocky places, steppe, and
semi-desert and is often found as a roadside weed. It appears to tolerate both acid and alkaline
soils, and is noted as growing on sandstone, dry granite, limestone, chalk, etc. Flowering
normally seems to occur from May to August, but it has been also reported in April, September,
and October. According to the notes with Chick 144, the peasants of Iran are said to eat the
seeds and describe them as being like rice.

4. Morina longifolia Wallich ex DC., Prod. 4: 644 (1830). Type: Gossain Than, Nepal, Wallich
426 (BM; K; E-type collection).

Morina elegans Fischer & Avé-Lall. in Index Seminum Hort. bot. imp. Petropol. 8: 67 (1841). Type: In
alpibus Himalaya, nota apud nos e seminibus a Cel. Candolleis acceptis (not seen).

Robust herbs forming clumps of sterile rosettes and flowering shoots. Rhizomes woody, the top
often covered with the fibrous remains of previous year’s leaves. Leaves of sterile shoots and
lower stem leaves glabrous, linear to linear-lanceolate, up to 40 x 4 cm, with fairly regularly
coarsely cut 4-5 lobed teeth, spiny tipped, the petioles fused to form deep sheaths of 25-45 mm.
Upper stem leaves similar, usually 3 per whorl, smaller, sheaths 8 mm or less, or petioles barely
connate. Flowering stems erect, up to 90 cm, somewhat square in transverse section and
pubescent to villous above, glabrous and slightly ridged to terete below. Flowers in (4-) 7-9
(-11) whorls, some shortly pedicillate others sessile; the upper whorls confluent, becoming more
separated as the inflorescence matures. Bracts broadly ovate to orbicular with an acute spiny tip,
margins often with numerous short spiny teeth, villous below and on the margins, often with few
hairs and glossy outer surface, the bases markedly overlapping. Jnvolucel + cylindrical,
glandular-hairy, sometimes densely so, 7-10 x 2-5-4 mm with (5—) 8-10 (—13) teeth, 2 much
exceeding the others, glabrous tipped and with adpressed hairs below, not or scarcely laminate,
shorter than or rarely equalling the calyx tube at anthesis. Calyx + glabrous or with short
adpressed hairs on the lips, lobes shallowly bifid, apices of the lobules rounded or rarely slightly
apiculate; tube 4-5—6 x 2:5—4 mm at anthesis Corolla white, becoming pink then deep red, the
tube with many glandular and a few simple hairs without, 20-30 x 1-2 mm, throat somewhat
expanded, lobes of the lower limb somewhat spreading, upper lip + patent, median lobe of
lower lip 7 X 3 mm. Fertile stamens 2, filaments 1-2 (-3) mm, with a tuft of hairs below the
anthers, inserted just inside the throat of the corolla. Anther lobes unequal, the shorter
sometimes % or less the length of the longer. Staminodes minute, inserted 2-3-5 mm below the

REVISION OF THE MORINACEAE 31

mouth of the corolla, under the anterior lip. Nectary 1, 3-lobed, anterior at the base of the
corolla tube. Style a little longer than the stamens, stigma disc-shaped or somewhat domed.
Achenes with abaxial face rugose with diagonal veins, adaxial face with deep furrow and
prominent longitudinal veins; apex somewhat obliquely truncate and crenulate.

Observations: The specimens of M. longifolia examined by us showed relatively little
variation. Size differences occur, notably a specimen from Bhutan, Ludlow, Sherriff & Hicks
1958, which is particularly robust, with leaves of up to 50 x 8 cm and very large leafy bracts.

Distribution and ecology: M. longifolia is mainly distributed in the Himalayas and Hindu
Kush, and is the only member of the genus to be found in Nepal and Bhutan. In the Punjab,
western Kashmir, and Pakistan its range overlaps those of M. persica and M. coulteriana. It is a
plant of high altitudes and we have seen no specimens from below 2000 m, the highest record, a
specimen from Nepal, was reported from 4250 m, while most specimens are collected from
between 3000 and 3750 m. It occurs in a wide range of wet and dry habitats, often on grassy
slopes or under shrubs. The flowering period is from June to September, sometimes extending
into October. This species is quite widely cultivated in Britain by discerning plantsmen, and in
the authors’ garden in Surrey it flowers in June and July, with occasional small basal shoots
flowering in October. The flowering sequence has been described by Kerner (1881). He
observes that the white flowers open at dusk, and are pollinated by night-flying moths, after
which they turn red and bend away from the rest of the flowers in the whorl in the course of a few
hours. This does not correspond to our observations as, in our plants, the flowers open in the
daytime, becoming pink by the following day, and do not turn red until the third day.
Examination of unopened flowers showed pollen already adhering to the stigmas, perhaps
indicating that cleistogamy may occur in this genus, paralleling the probable cleistogamy in
Cryptothladia.

Specimens examined

In the following enumeration all the specimens examined during the course of this study have
been listed, primarily by genera and species, and secondarily by collectors and specimen
numbers. The location of all specimens is shown using the standard international abbreviations
for the institutions concerned; a list of these will be found under the acknowledgements. Since
the number of taxa is not very large, it should be reasonably convenient for curators to use the
list as a means of checking identifications of further duplicates of the listed gatherings.

I.1. Acanthocalyx nepalensis (D.Don) M. Cannon

Beer, Lancaster & Morris 9419 BM. Bor & Kirat Ram 20444 K. Bowes Lyon 2058 BM; 3264 BM;
15126 BM; Bowes Lyons.n. BM. Cave 53 E. Chu 2785 E. Clarke 12913 K. Cooper 62 BM, E; 99 BM,
E; 258 BM, E; 1585 BM; 1699 BM, E; 2170 BM; 2533 BM; 2935 BM. Dobremez 337 BM; 380 BM.
Dungboo 55 E; 8539 LE. Einarsson, Skarby & Wetterhall 606 BM; 892 BM. Gamble 9520 K. Gould
904 K. Grikley s.n. LE. Grierson & Long 2592 E; 2669 E. Hanbury-Tracy 28 BM; Hara, Kanai,
Mutara, Togashi & Tuyama 2863 K. Hingston 181 K; 386 K. Hobson in 1897 K. Hooker in 1849 K.
King’s collector 647 K; in 1888 LE; in 1886 LE. Kingdon-Ward 11808 BM; 19582 BM; E;s.n. K. Lall
Dhwoj 167 BM, E. Lepcha 15 E. Lobbichlers.n. BM. Ludlow & Sherriff 170 BM; 422 BM; 1792 BM;
8780 BM; 8799 BM, E; 9756 BM, E; 11022 BM, E. Ludlow, Sherriff & Elliott 15184 BM, E; 15677
BM. Ludlow, Sherriff & Hicks 16379 BM, E; 19088 BM; 20293 BM. Ludlow, Sherriff & Taylor 4700
BM, E; 5518 BM, E. McCosh 322 BM. McLaren’s collectors AC140 E. Maire 49 E. Morton 232 K.
Polunin 329 BM; 627 BM; 1151 BM. Polunin, Sykes & Williams 78 BM; 1546 BM; 2329 BM; 4536 BM.
Pratt 45 BM. Ramesh Bedi 366 K; 1066 K; s.n. K. Ribu & Rhomon 5270 K. Richardson 245 BM.
Rohmoo Lepcha 787 E. Schilling 394 K; 995 K. Sharma E 416 BM. Shreltow 5140 BM. Shrestha &
Joshi 220 BM. Smith 120 BM. Spencer-Chapman 418 K; 652K. Stainton 875 BM, E; 557 BM, E.
Stainton, Sykes & Williams 3192 BM; 3237 BM; 4652 BM. Waddell 18 BM. Wager 285 K. Williams 792
BM. Wallich 424 K. Younghusband s.n. K. Yu 19845 E; 19730 E. Zimmerman 694 BM.

1.2. Acanthocalyx delavayi (Franchet) M. Cannon
Delavay 52 (298) P; 90 P; in 1885 P; in 1887 K. Forrest 393 E; 4653 E; 5638 BM, E, K; 5670 BM, E, K;
6104 BM, E, K. Handel-Mazzetti 2989 E; 9754 E. Kingdon-Ward 9632 BM. McLaren’s collectors LZA

32

Hee

II.1.

11.2.

33

11.4.

II.5.

II.6.

M. J. CANNON & J. F. M. CANNON

BM, E, K; N66 BM, E, K. Manbergs.n. K. Monbeig 62 E, K; 140 E;s.n. K. Potanins.n. K. Rock 3422
E; 3781 E; 16260 E, K; 18098 E; 23835 BM, E, K; 23942 E; 24610 BM, E, K; 24692 BM, E, K; 24980
BM, E, K. Schneider 1494 E, K. Sino—British Expedition 737 K. Soulie 71 pars K; 975 K. Tu 4493 PE.
Wilson 3781 pars BM, K. Yui 22152 E; 22750 E.

Acanthocalyx alba (Hand.-Mazz.) M. Cannon

Cunningham 471 E. Delavay 91 K, P;in 1886 P, K. Dungboo 6434 K; s.n. K. Fang 4008 E, K, PE; 6036
E, K, PE. Farrer & Purdom 215 E. Forrest 5832 BM, E, K; 2685 BM, E, K; 28444 BM, E; 28826 BM,
E. Gould 2133 K. Hanbury-Tracy 51 BM. King’s collector 547 K. Kingdon-Ward 5894 E, K; 7065 K.
Ludlow & Sherriff 8833 BM, E. Ludlow, Sherriff & Elliott 14077 BM, E; 15370 BM, E. Ludlow,
Sherriff & Taylor 5017 BM; 6180 BM, E. McLaren’s collectors AC163 E; LS0A BM, E, K; N104 BM,
E. Maire 2 BM. Potanin in 1885 K; in 1893 K. Pratt 202 BM; 232 pars BM, E, K; 696 K. Purdom 872 K.
Rock 4636 E; 5564 BM; 12617 E, K; 13119 BM; 14611 E; 16404 E; 17406 E, K; 24113 BM, E, K; 24885
pars BM, E, K. Schneider 3426 K; 3639 K; 3680 K. Soulie 71 pars K; 577 K. Wilson 3781 pars BM, K.
Yu 8671 BM; 13625 BM, E; 15252 BM, E.

Intermediates between A. nepalensis and A. delavayi

Chu 2785 BM. Forrest 7155 E; 14022 E; 27124 E. Gould 511 K. Kingdon-Ward 325 K. Lowndes L1125
K. Ludlow, Sherriff & Elliott 13894 BM. Ludlow, Sherriff & Hicks 16573 BM; 19017 BM. McLaren’s
collectors 309D BM, E, K; B73 E, K. Maire 49 E; 66 E; 69 E; 82 E; 86 E; 128 BM, E; 715 E; 729
E. Potanin s.n. K. Pratt 45 BM, K. Rock 17725 BM, E; 17823 E; 24072 BM, E, K.

Intermediate between A. nepalensis & A. alba
Gregory B99 K. King’s collector 8 K. Pratt 232 BM, E, K.

Cryptothladia chinensis (Pai) M. Cannon
Ching 539 E. Fang 4270 E, K, PE. Hao 1002 PE. Licent 4548 BM, K. Potanin in 1885 K; s.n. K.
Przewalsky s.n. K. Rock 12556 BM, LE; 12952 E, K. ? collector 1792 PE; 6129 PE.

Cryptothladia parviflora (Karelin & Kir.) M. Cannon
Gubanov 915 MW. Gukun s.n. BM. Karelin & Kirilov s.n. BM, K. Konsevyanova s.n. MW; Regel
21798 BM, E, K, LE; s.n. K, LE. ?collector 4484 PE.

Cryptothladia kokonorica (Hao) M. Cannon
Gould 2217 K. Ludlow & Sherriff 8691 BM. Morton 134 K. Richardson 124 BM. Spencer-Chapman
129 K; 649 K. Thorold 138 K. Wollaston 146 K. Younghusband 105 K. ?collector 6114 PE.

Cryptothladia chlorantha (Diels) M. Cannon

Ching 539 E. Forrest 2482 E, K; 5777 BM, E, K; 10212 E, K; 21255 E, K; 28801 BM, E. McLaren’s
collectors N9 BM, E, K. Rock 5129 E; 16499 E; 23834 BM, E, K. Sino—British Expedition 1641 K.
Wilson 3781A BM, K. Yii 15306 BM, E. ?collector 5482 PE.

Cryptothladia polyphylla (Wallich ex DC.) M. Cannon

Bowes Lyon 2210 BM. Cummins s.n. K. Dobremez 501 BM. Dungboo s.n. K. Einarsson, Skarby &
Wetterhall 209 BM; 1193 BM. Gardner 1419 BM. Grey Wilson & Phillips 741 K. King’s collector 402
K. Lall Dhwoj 0402 BM. Ludlow, Sherriff & Hicks 16243 BM, E; 17459 BM. Malla 9176 BM.
McCosh 381 BM. Polunin 668 BM; 1339 BM. Polunin, Sykes & Williams 4177 BM. Schilling 436 K.
Shrestha 5126 BM. Stainton 4325 BM. Stainton, Sykes & Williams 1763 BM; 2984 BM; 6138 BM.
Wallich 425 K, PE. Wigram 122 K.

Cryptothladia ludlowii M. Cannon
Cooper 4525 BM. Kingdon-Ward 12383 BM, K; 14310 BM. Ludlow, Sherriff & Hicks 16903 BM, E;
19050 BM, E; 20305 BM.

III.1. Morina kokanica Regel

Galubev s.n. MW. Gnezdillo 186 BM. Gomolitzsky s.n. E, K. Von Knorring 765 K. Linczevski &
Roshkova 39 K. Minkvitz 41766 BM, E, K. Neustrueva & Knorring 4176A BM, E, K. Neustrueva
157 BM. Pavlov 71 MW; 396 MW; 482 MW. Regel 6/1880 BM, K, LE; 7/1880 BM. Sovetkina 1050
BM. Tilchko s.n. BM. Vatolkina 447B E, K.

III.2. Morina coulteriana Royle

Aitchison 407 BM, K; 746 K, LE. Anders 10979 W. Battels.n. K. Bellews.n. K. Benham 1919 BM.
Bor 12289 K. Bowes Lyon 202 BM; 1095 BM, E. Bunge s.n. P. Capus 574 P. Carter 993 K. Chiddell

HT.3:

IIl.4.

REVISION OF THE MORINACEAE 33

88 BM. Clarke 28767 BM; 30871 K. Collett 46 K. Cooper 5298 E; 5536 E. Drummond 14204 K; 22526
K. Duthie 935 BM, LE; 5661 BM; 25645 LE. Eckburg 9577 E. Edelberg 1783 W. Edgeworth 259 K.
Ellis 272 K; 1348 K; 1881 K. Evershed s.n. BM. Furse 8048 E, K. Gibbons 607 E, K. Giles 58 K.
Gronbczewsky s.n. K. Gukuns.n. MW. Harbuckle s.n. K. Harriss 16234 BM. Hedge 449 K. Hedge
& Wendelbo 5228 E; 9444 E. Huggins 228 BM. Kaletkina 38 K. Kingdon-Ward 11996 BM. Koelz
13012 W. Komorov s.n. LE. Kurushkaya & Nicrasov 0887 BM. Lace 219a E; 1652 E. Lindbery 668
W; 917 W. Ludlow & Sherriff 1419 E; 1828 BM, E; 5719 BM, E; 8219 BM; 9123 BM. Ludlow,
Sherriff & Taylor 5481 BM. Munro 1324 K. Neubauer 662 W. Novitsky s.n. LE. Parmanand 598 E.
Pinfold 323 BM. Podlech 12348 E. Rechinger 31615 W. Regels.n. BM. Reid 111216 E. Rich B1082 K;
1095 K. Robson 2103 BM. Royle s.n. K. Schlagintweit 4181 BM. Sherriff 7350 BM. Siddigh 2269
BM. Singh 1771 K. Stainton 2785 BM, E. Stewart 3754 K. Stewart & Rahman 25376 BM. Strachey &
Winterbottom 3 BM, K. Thompson 210 K. Toppin 412 K. Vassiljeva 5487 BM, E, K. Watt 441 E.
Wendelbo 59 BM, K. Zaprialiev & Tekutvev 278 MW.

Morina persica L.

Aitchison 420 BM, K. Alava 10654 E. Amsel s.n. W. Anders 3738 W; 10722 W. Anderson 49 E.
Archibald 2856 K. Ascherson 439 BM, E, K. Atchley 1034 K; 2239 K. Aucher-Eloy 754 BM, K.
Balansa 796 BM, K. Balls 514 E, K; 1183 BM. Baytop 14357 E; 19927 E. Behboudi 1293E W.
Bledencvii 5085 LE; 5723 LE. Bokhari & Edmundson 2036 E. Bornmuller 786 K; 4518 K; 9496 BM,
K; 11955 BM, E. Bourgeau 240 K. Borne 3694 K. Brant & Strangways 549 K. Calvert & Zohrab 388
BM, K. Carter 1309 K. Chick 144 K. Clarke 24494 BM; s.n. K; s.n. E. Constantintinon H1507 K.
Coode & Jones 2323 E. Cooper 4983 E. Crookshank 435 K. Darrah 467 E. Davis 10099 E, K; 16369
E, K; 20370 K; 46443 K; 47477 E, K. Davis & Coode 36591 E. Davis, Dodds & Cetik 19068 BM, E.
Davis & Hedge 30051A E, K; 31731 BM, E, K. Drummond 25808 E, K. Dudley 35872 E, K; 36089
E, K. Duthie 336 K; 15603 K; 21847 K. Edelberg 1108 W. Edmondson 580 E. Ehrenberg s.n. K.
Ekim 36 E. Falconer 534 K. Flemburg 851 E. Flemming 20 E; 61 E;s.n. K. Foxs.n. K. Furse 4098 K.
Gamble 4213A K; 4308A K; 22819 K; s.n. K. Gay s.n. K. Gerard s.n. BM. Gilli 3881 W; 3882 W.
Grant 17684 W. Grebenchikoff in 1936 K; in 1938 K. Gubanov & Pavlov 162 BM. Guichard T/42/60
K. Haghighi 6241T W. de Halacsy s.n. K. Haradjian 1248 E, K; 2360 E, K. Harsukh 15336 K.
Haussknecht in 1865 BM, K; in 1885 K. de Heldreich 46E, K, LE; 99 E;s.n. K. Hewer 998 K; 1352 E,
K; 2042 K. Hisbourg s.n. K. Kashkaulis.n. W. Kerstan 1447 W. Khan, Prance & Ratcliffe 276 E, K.
Koelz 11744 E, W. Kotschy 549 BM, K, LE. Kotte s.n. K. Lace 3869 E, K. Lack 4288 E. Lambert &
Thorp 559 K. Lammond 2478 E. Leonis s.n. LE. Liston s.n. E, K. Loftus s.n. BM, K. Madden s.n.
E. Manisadjan 56 K; 112 K; 145B K. Marten 16 E. Mill s.n. K. Mooney 4514 K. Moussari &
Transhahr s.n. W. Neubauer 202 W. Orphanides 101 E, K. Parkinson 7366 E. Parry 251 E. du
Pavillon 15 BM, K. Pichlers.n. K. Posts.n. BM. Price 916 K. Rechinger 5761 W; 16966 W; 31056 K,
W; 32262 W; 34355 W; 47483 K, W. Rich 179 K. Rix 86 E. St Lagers.n. K. Sawyer 53 E. Sibthorpe
s.n. K. Siehe 232 BM; 318 E. Smith 4107 BM. Stainton & Henderson 5452 K. Stanf 403 K. Stewarts.n.
E, K. Stribrny 66 E; 68 E; s.n. E, LE. Tebey 2007 E; 2669 E. Thompson 49 BM, E; 654 BM.
Thompson & Clarke s.n. K. Transhahr & Moussari s.n. W. Walne s.n. LE. Warr 10099 K. Watson
275 K. Watt 45 E; 90 E. Wendelbo 831 W. Wiedermann s.n. BM, K. Winter 308 BM. Woronov s.n.
LE. Zorab 404 K.

Morina longifolia Wallich ex DC.

Abel 78 BM; Bailey’s collectors s.n. BM. Aitchison 47 K;s.n. LE. Bhattacharya 24331 LE. Bis Ram
530 BM. Burtt 1215 E. Clarke 24238 K; 31086 BM. Cooper 4914 BM, E; 5106 E; 5597 E. Duthie 935
LE; 1157 LE; 13115 BM, E; s.n. BM, LE. Dobremez 521 BM. Einersson, Skarby & Wetterhall s.n.
BM. Falconer 533 K. Fuller 854 K. Gamble 1495A K; 5666A K; 6518C K; 26917 K. Giles 727 K.
Griffiths 2132 BM. Komorov s.n. LE. Lace 367 BM, E; 1497 BM, E. Lall Dhwoj 103 BM, E; 0608
BM, E. Lance 134 K. Lawrie 5448 BM. Lowndes 1376 BM. Ludlow & Sherriff 3567 BM; 7819 BM,
E; 8275 BM; 9226 BM. Ludlow, Sherriff & Hicks 19581 BM, E. Maclagan 711 BM. Nand 251 E.
Osmaston 67 K. Parker 3043 LE. Polunin 56/256 BM, E; 1449 BM. Polunin, Sykes & Williams 208
BM; 2608 BM; 4358 BM; 4423 BM. Ram 8942 E. Reids.n. E. Rich 1255 K. Ribu & Rhomon 5509 E.
Schlagintweit 10016 E. Schlich s.n. E. Sherriff 7373 BM. Stainton 4961 BM. Stainton, Sykes &
Williams 1964 BM; 3388 BM; 3475 BM; 7376 BM. Stewart 44 E; 45 E. Strachey & Winterbottom 1
BM, LE. Thompsons.n. E. Venning K141 K. Wallich 426 BM, K. Watt91 BM; 109 E; 445 E; 1944 E;
3300 E; 8833 E; 9597 E. Wigram 76 E. Young s.n. BM.

34 M. J. CANNON & J. F. M. CANNON

Acknowledgements

In addition to the material in the British Museum (Natural History) (BM), we have also made extensive use
of the collections preserved in the herbarium of the Royal Botanic Gardens, Kew (K). We are most
grateful to the directors and staff of the following institutions, who have made specimens available to us on
loan: Royal Botanic Garden, Edinburgh (E), Komorov Institute of the Academy of Sciences of the
U.S.S.R., Leningrad (LE), Biology Department of the Lomonsov State University of Moscow (MW),
Laboratoire de Phanérogamie, Muséum National d’Histoire Naturelle, Paris (P), Institute of Botany,
Academia Sinica, Beijing (PE), and Botanische Abteilung, Naturhistorisches Museum, Wien (W).

We are most grateful to our colleague Dr S. Blackmore for his co-operation with the palynological
aspects of this investigation, and especially for his personal contribution of the palynological section of this
paper. We are also very appreciative of the help afforded by Dr C. J. Humphries who, in the course of
extensive and most useful discussions, brought his wide experience of modern techniques in phylogenetic
systematics to bear on relationships in the Morinaceae, and so helped to clarify our own ideas. We are also
grateful to Dr J. R. Edmonson of the Merseyside County Museums for information on the Royle
Herbarium, to Miss K. P. Kavanagh for assistance with the Latin description, to Mr J. R. Press for advice
on mapping and to Mrs S. M. Burton for typing the final manuscript. Lastly, we acknowledge the help of
Mr A. O. Chater, Dr N. K. B. Robson and Mr J. D. A. Stainton, through much helpful discussion at
various stages during the course of the work.

References

Bentham, G. & Hooker, J. D. 1873. Genera plantarum 2: 158. London.

Blackmore, S. & Cannon, M. J. 1983. Palynology and systematics of the Morinaceae. Rev. Palaeobot.
Palynol. 40: 207-226.

Bobrov, E. G. 1957. Family CLV. Morinaceae (Dum.) Van Tiegh. In V. L. Komarov, Flora of the
U.S.S.R. 24: 3-9. (Trans. Lavoott, Jerusalem 1972).

Bunge, A. 1852. Beitrag zur Kentniss der Flor Russlands: 321-323. St Petersburg.

Candolle, A. P. de 1830. Prodromus systematis naturalis regni vegetabilis 4: 644. Paris.

Coulter, T. 1824. Mémoire sur les Dipsacaceés. Mém. Soc. Phys. Hist. nat. Genéve 2: 13-60.

Crété, P. 1981. Embryo. Jn P. Maheshwari (Ed.) Recent advances in embryology of angiosperms: 171-220.
Delhi.

Cronquist, A. J. 1981. An integrated system of classification of flowering plants: 1013-1016. New York.

Erdtman, G. 1945. Pollen morphology and plant taxonomy III. Svensk bot. Tidskr. 39: 186-191.

— 1960a. Notes on the finer structure of some pollen grains. Bot. Notiser 113: 285-288.

— 1960b. The acetolysis method. A revised description. Svensk bot. Tidskr. 54: 561-564.

Hemsley, W. B. & Pearson, H. H. W. 1902. The flora of Tibet or high Asia. J. Linn. Soc. (Bot.) 35:
254-257.

Heywood, V. H. (Ed.) 1978. Flowering plants of the world: 261-262. Oxford.

Hutchinson, J. 1973. Families of flowering plants, ed. 3: 584-585. Oxford.

Kachidze, N. 1929. Karyologische studien tiber Familie der Dipsacaceae. Planta 7: 484-502.

Kamelina, O. P. & Yakovlev, M. C. 1974. Development of the embryo sac in the genus Morina. Bot. Zh.
S.S.S.R. 59: 1609-1617. [In Russian. ]

— 1976. The development of microsporogenesis in representatives of the Dipsacaceae and Morinaceae.
Bot. Zh. S.S.S.R. 61: 932-945. [In Russian. ]

Kerner von Marilaun, A. 1881. The natural history of plants 2: 222 & 351-353. London. (Trans. Oliver
1894-95).

Kingdon-Ward, F. 1936. Sketch of the vegetation and geography of Tibet. Proc. Linn. Soc. Lond. 1936:
142.

Linnaeus, C. 1738. Hortus cliffortianus: 14. Amsterdam.

1753. Species plantarum: 28. Stockholm.

Matthews, V. A. 1972. Morina L. In P. H. Davis (Ed.) Flora of Turkey 4: 581-582. Edinburgh.

Meusel, H. 1971. Mediterranean elements in the flora and vegetation of the western Himalayas. Jn P. H.
Davis, P. C. Harper, & I. C. Hedge, Plant life of southwest Asia: 53-72. Edinburgh.

Pai, Y. Y. 1938. Die Chinesischen Arten der Gattung Morina. Reprium Spec. nov. Regni veg. 44: 114-124.

Poucques, M. L. de. 1949. Récherches caryologiques. Revue gén. Bot. 56: 114-115.

Rafinesque, C. S. 1820. Tableau analytique des Ordes Naturels . . . Annis gén. Sci. phys. Brux. 6: 88.

Risse, K. 1929. Beitrage zur Zytologie de Dipsacaceen. Bot. Arch. Berlin 23: 266-288.

Spach, E. 1841. Histoire naturelle des vegetaux: phanerogams 10: 313. Paris.

REVISION OF THE MORINACEAE 35

Stainton, J. D. A. 1972. Forests of Nepal: 138-169. London.
Takhtajan, A. L. 1954. Origins of angiospermous plants: 62. Washington. (Trans. Gankin & Stebbins,

1958).

— 1980. Outline of classification of flowering plants. Bot. Rev. 46: 225-359.

Thorne, R. 1976. Phylogenetic classification of Angiospermae. Evolut. Biol. 9: 64.

Tieghem, P. E. L. van. 1909. Remarques sur les Dipacacées. Annis Sci. nat. (Bot.) IX, 10: 148-200.
Tournefort, J. P. de. 1703. Institutiones rei herbariae: 48. Paris.

Vaillant, S. 1724. Classe des Dipsacées. Hist. Acad. Sci. Paris 1722: 184-243.

Verlaque, R. 1976. Contributions a |’étude cytotaxonomige des Dipacacées & Morinaceae du bassin

Mediterranean. Thesis — University de Provence.

—— 1977. Rapports entres les Valerianaceae, les Morinaceae & les Dipsacaceae. Bull. Soc. bot. Fr. 124:

475-482.

Vijayaraghavan, M. R. & Sarveshwari, G. S. 1968. Embryology and systematic position of Morina

longifolia. Bot. Notiser 121: 383-402.

Vinokurova, L. V. 1959. Palynological data to the systematic position of Dipsacaceae and Morinaceae.

Problemy Bot. 4: 51-67. [In Russian. }

Wagenitz, G. 1964. Jn H. Melchior & E. Werdermann (Eds.) Engler’s Syllabus der Pflanzenfamilien 12th

ed. 2: 477.

Taxonomic index

Accepted names are in roman and synonyms in italic; new names are in bold, as are principal references.

An asterisk (*) denotes a figure or map.

Asaphes Spreng. 24
Barleria crotalaria A. Léveillé 12,14
Acanthocalyx (DC.) M.Cannon 2, 3, 5, 6*, 7, 8*,

eo ,.11, 13

alba (Hand.-Mazz.)M.Cannon 9, 10*, 11,
14-15, 32

delavayi (Franchet)M.Cannon 9, 10*, 11,
12-14, 15, 31

nepalensis (D.Don) M.Cannon 9, 10*, 11,
128". 14, 15, 31
Cryptothladia (Bunge) M.Cannon 2,5, 6*,7,8*,
9, 15-17*, 31
chinensis (Pai) M.Cannon 15, 16*, 17-18,
20, 32
chlorantha (Diels) M.Cannon
19-20, 32
kokonorica (Hao) M.Cannon
18, 19, 32
ludlowii M. Cannon 2, 16*, 17, 22—24*, 32
parviflora (Karelin & Kir.) M.Cannon 15,
16*, 17, 18-19, 32
polyphylla (Wallich ex DC.) M.Cannon 15,
16*, 17, 20—22*, 32
Diototheca Vaillant 2
Monnal. 1,2,3,4,5,6*,7,8*,9, 24, 25*, 28
Sect. Acanthocalyx DC. 2,3,9
Sect. Cryptothladia Bunge 2, 15
Sect. Diotocalyx DC. 2,24
alba Hand.-Mazz. 9, 14
aucheri Jaub. & Spach 29
betonicoides Benth. 11, 12
breviflora Edgew. 26

IS) UY 5 WILE

155167 ds,

bullyana Forrest & Diels 12

chinensis Pai 17,18

chlorantha Diels 19

chlorantha var. subintegra Pax & K. Hoffm.

ex Limpr. 19
coulteriana Royle 2, 24, 25*, 26-28%, 30, 31,
32

delavayiFranchet 9, 12
elegans Fischer & Avé-Lall.
graeca Jaub. & Spach 29
kokanica Regel 4, 24, 25*, 26, 32
kokonorica Hao 18, 19
lehmanniana Bunge 2, 26, 28
leucoblephara Hand.-Mazz. 9, 14,15
longifolia Wallichex DC. 2,4, 22,24, 25*,
30-31, 33

nana Wallichex DC. 2,12
nepalensis D.Don 9, 12
orientalis Miller 28
parviflora Karelin & Kir. 2
parviflora var. chinensis Batalin 17
persicaL. 2,4, 22,24, 25*, 28-30, 31, 33
persica subsp. turcica Hal. 29
polyphylla Wallichex DC. 2, 20, 22, 24
spectabilis Gontsch. 29
subinermis Boiss. 29, 30
tournefortii Jaub. & Spach 29
turcica Degen & Hal. ex Celak 29
verticillata Moench 28
wallichiana Royle 28, 29

Morinaceae J. Agardh 1,2,3,8,9

Morinidia Raf. 2,8

2, 30

British Museum (Natural History)

Seaweeds of the British Isles

The synthesis of many years’ research carried out by the British Museum (Natural
History) and the British Phycological Society, this is the second in a series of books
that will be published under this title covering all the British and the majority of
northern Atlantic seaweeds.

Volume1 Rhodophyta
Part2A Cryptonemiales (sensu stricto), Palmariales,
Rhodymeniales.

Linda M. Irvine

Following the style and format laid down in Part 1, this book deals with the species
attributed to the orders Cryptonemiales, Palmariales and Rhodymeniales. Each
species description incorporates notes on ecology and distribution and is supported
by one or more line illustrations. Keys to aid identification are also included.

When complete, this title will be the standard work of reference in its field; it will
provide students and researchers with a digest of the most up-to-date and compre-
hensive information available on the marine algae of Great Britain and Ireland which
will be indispensable throughout the north Atlantic region and beyond.

ISBN 0 565 00871 4 128 pagesillustrated £13.00.

An introduction to fern genera of the
dian subcontinent

istopher R. Fraser-Jenkins

ma. ‘a

Bot any series Vol 12 No 2 25 October 1984

The Bulletin of the British Museum (Natural History), instituted in 1949, is issued in tonal ae
scientific series, Botany, Entomology, Geology (incorporating Mineralogy) and ae “
and an Historical series.

Papers in the Bulletin are primarily the results of research carried out on the unique Pa
ever-growing collections of the Museum, both by the scientific staff of the Museum pk
specialists from elsewhere who make use of the Museum’s resources. Many of the papers are ~
works of reference that will remain indispensable for years to come. MAS ee.
Parts are published at irregular intervals as they become ready, each is complete in itself, —
available separately, and individually priced. Volumes contain about 300 pages and several
volumes may appear within a calendar year. Subscriptions may be placed for one or more of |
the series on either an Annual or Per Volume basis. Prices vary according to the contents of
the individual parts. Orders and enquiries should be sent to: Bee,

Publications Sales,

British Museum (Natural History),
Cromwell Road,

London SW7 S5BD,

England.

World List abbreviation: Bull. Br. Mus. nat. Hist. (Bot. } 2

™
%.\
oy

ae

© Trustees of the British Museum (Natural History), 1984

The Botany series is edited in the Museum’s Department of Botany

Keeper of Botany: Mr J. F.M. Cannon
Editor of Bulletin: © Mr J. R. Laundon
Assistant Editor: Dr A. J. Harrington

ISBN 0 565 08001 6 . :
ISSN 0068-2292 Botany series
Vol 12 No 2 pp 37-7
British Museum (Natural History) a
Cromwell Road : ;
London SW7 5BD Issued 25 October 19

An introduction to fern genera of the Indian |
subcontinent

Christopher R. Fraser-Jenkins

c/o Department of Botany, British Museum (Natural History), Cromwell Road, London
SW7 SBD

Contents
ea ra ete cae ope Ped aan cos xc nein sa eee saeasacep se sieae ose done 37
AT RG CCl ITC NOTRE se ere ler ras ectectae pees opie g ee pats ddt se se tleae ican Spe eeieleeare wieelethainsaseivensinesoes 37
Somecousiderations forthe study offermsin Asia. .........0.0e0c0-0h-whaaesscstedascnscascasone 38
Asreview oulndian subcontinent fermi Studly .2...-.cm-200<-nc-scscnsteaveddanestescseccscecsssseses 42
ne RED OR MON cities ot fc Nl ake ela ce Bothalsa a4 aes dns «nid wig Parle UW china dn nk dein hina swnnienins 52
a Soca arta eA EEN a aan dena -rageBr npe hw Side ale» Sain nipcinninien «ninco 65
EACTRISLICCOTISIC ET ALLOUUS sc artry ova eee cies dyanntaen isaincee cu spot stained secinsinae acs ceensinieeaiclon 65
Geographical limits and major phytogeographic divisions within India.................. 68
DUS OM MLO ANG Ma PDIP OUSPECIES cesar set sccsceacucerastoessteaeidhy oaduceeesaedevsesasensacs 71
MPR ARTICLES Seep fice tached: oth fies deaes vate ce co cetdaacas io deinen ysticctis odie eadsccccvsiascietss 73

Synopsis

This paper is a general introduction to fern genera in the Dryopteridaceae and Athyriaceae (Aspidiaceae,
sens. lat.) of the Indian subcontinent and in particular to a forthcoming monograph of Dryopteris. Details
of the methods of investigation, history of fern study, an extensive bibliography, and the phytogeography
of the ferns concerned are included. The area is divided into 100 regions in four major divisions.

Introduction

This paper is intended to be a general introduction to representatives of the Aspidiaceous genera
Dryopteris and Polystichum in the Indian subcontinent. These large and complex genera have
often been ignored, or only scantily or fragmentarily dealt with, due to the taxonomic problems
they raise. The author has prepared a monograph of Dryopteris which will be published later.

The Indian subcontinent is here construed as extending from the extreme western limits of the
monsoon area, which are in eastern Afghanistan, throughout the Himalaya, through Pakistan,
India, Nepal, Sikkim, and Bhutan to Assam and southward through Bangladesh and all of India
and Pakistan to Sri Lanka (Ceylon). It forms a unit defined partly from a floristic point of view
and partly for geographical and historical convenience, in view of its previous treatment as a unit
by British botanists. The natural limits of the flora extend in reality beyond the east Himalaya
into at least western China, but at present it is impractical to include the Chinese part even
though it has been extremely important to consider Chinese species. The author is fortunate to
have been working at a time when it has once more been possible to communicate with Chinese
botanists and visit Chinese herbaria. It is clear from the high number of species and their
affinities throughout the world that the east-Himalayan and west-Chinese region is a major
centre of distribution of species of many genera, or at the very least an area where, as a result of
an equable climate and the presence of mountain refugia, by far the greatest number of species
have survived. Thus it is hoped that a detailed account, using modern knowledge and methods,
of the species in the area will be useful for the study of the genera concerned throughout the
world and especially in the rest of Asia.

Material studied for the monographs has been determined anew by the author and has come
from the major herbaria containing Indian subcontinental material. Apart from the author’s
extensive collections (in BM), all the material of the genera concerned has been studied at the

Bull. Br. Mus. nat. Hist. (Bot.) 12 (2):37-76 Issued 25 October 1984

38 CHRISTOPHER R. FRASER-JENKINS

following herbaria (abbreviations are according to Holmgren, Keuken & Schofield, 1981): K,
BM, E, MANCH, CGE, LINN, LIV, P, L, B, BR, G, Z, UC, MICH, PPFI, RAW, ISL, LAH,
PAN, PUN, DD, DBS, LWG, ASSAM, CAL, KATH, MH, KUN, IBSC, PE, SYS, & YUKU.
Reference has also been made, for some of the rarer species, to S, US, W, H, LE, KYO, TI,
TOFO, and KABA. All distributions given in the introduction and monograph are based
entirely on specimens seen and determined by the author and not (except in a few stated cases)
from records taken from literature, which are considered unreliable in the light of the large scale
revision which has now been necessary in these complex genera. All the synonymy, except in a
few stated cases, has similarly been investigated by examining types.

Some considerations for the study of ferns in Asia

The genus Dryopteris is possibly the most complex and difficult genus to study in the
subcontinent, not only on account of its large size, but also because it contains groups of closely
related species, often polyploids or apomicts, which are difficult to separate. It is closely
followed in this respect by the other genera Polystichum and Athyrium and indeed all three
suffer from their treatment in earlier years when quite distinct species were lumped together and
often mistakenly identified as being the same as species from Europe. Even recent literature
suffers from this problem. However, in recent years members of these genera have been
carefully and fully elucidated in Europe, western Asia, North America, and Japan, and it is now
possible to begin to apply the same detailed and critical methods of study to central- and
southern-Asiatic species and to compare them properly with species from other parts of the
world. It is important in this connection to appreciate the ranges of intraspecific variation, and to
distinguish where possible between such variation and differences arising from the occurrence of
distinct combinations of genomes, which in general delineate the different species at least in the
first two of the genera concerned. Hence population sampling when collecting, and frequently
the cultivation of variant plants side by side, are both considered to be extremely important to
avoid the erroneous description of new species based on only one or a few, perhaps abnormal
specimens.

Apart from the essential and basic recognition of the characteristic morphology of the various
species, another principal method of approach to this study in Europe has been to attempt to
gain an understanding of the systematics of the groups by cytological study of interspecific
hybrids leading to genome analysis. Useful references to world-wide cytological reports are to
be found in Love, Love & Pichi Sermolli (1977), though the original reports must be checked as
they are frequently misquoted. In some cryptic groups the results may be merely confusing, and
cannot be reflected in their taxonomy and nomenclature, as for example in the complex
Cystopteris fragilis (L.) Bernh. aggregate (see Vida, 1974 and Lovis, 1977) in the genus
Cystopteris; in others widespread genome homology may produce uninterpretable results, as in
a number of phanerogam genera. But in general several Aspidiaceous genera contain species
with distinct genomes clearly reflected in their distinct morphology, and are amenable to study
by such methods, particularly the genera Dryopteris and Polystichum, though Athyrium appears
to contain several cryptic groups. A complication that arises frequently in Dryopteris though,
and to a certain extent in Polystichum, is the occurrence of apomictic species which show much
variation, but in which individuals are unable to breed together, hence restricting gene-flow
through the population. This reproductive isolation allows the preservation of any variants that
occur, which may form slightly distinct populations by cloning. The plants are also more difficult
to study cytologically because chromosome pairing analysis is only possible in a restricted
number of spore-mother cells (those in the 16-celled sporangia), and since hybridisation with
sexual species merely adds on extra genomes, it may be difficult in natural or artifically produced
apomictic triploid or higher polyploid hybrids to ascertain from which genomes bivalents seen at
meiosis may have originated. It is worth emphasising here the usefulness of mitotic (root-tip)
counts in order to avoid possible (and all too frequent) confusion between 16- and 8-celled
sporangia at meiosis, which results in an incorrect ploidy being published; also a diploid apomict
and a tetraploid species may both have 82 bivalents and if 16-celled sporangia are in low numbers

AN INTRODUCTION TO FERN GENERA OF THE INDIAN SUBCONTINENT 39

and are not noticed in the apomict a report of a sexual tetraploid may be made in error. In
addition root-tip counts may be made at any time of the year when the plant is growing, whereas
meiotic counts can only be made on mature plants during the short time when sporangia are
developing. In cases where the cytology does not clarify the situation, help may come from
another method being increasingly used in modern studies, that of chemical analysis: for
example, of the phloroglucides present in the stipe-bases of most Dryopteris species. Though in
general less exact than genome analysis, this may supply the crucial information that allows
understanding of the origin of a taxon, and hence may help to decide if it is really a distinct
species or only a variant. Both these studies are merely beginning in much of Asia except Japan;
in connection with the present work, for example, Professor T. Reichstein, of The Institute of
Organic Chemistry, Basel, Switzerland, and Professor C.-J. Widén of the University of Kuopio,
Finland, are carrying out analysis on a large amount of dried rhizome material of Dryopteris that
the author was able to bring back from the Himalaya, south India, and China, and Dr M. Gibby,
of the British Museum (Natural History), London, has carried out cytological checks on
corresponding living plants, which it is hoped may be extended to genome analysis of certain
groups at some time in the future. A rather large number of cytological reports on Indian
subcontinent ferns have now been published, but it must be noted that as yet the nature of only a
single polyploid species in the Dryopteridaceae of the area has been elucidated by genome
analysis, and that (on Dryopteris filix-mas (L.) Schott) was carried out in Europe on European
material. A similar situation applies in the Aspleniaceae, with four species all elucidated from
Europe only, though work on the A. /aciniatum D. Don (= A. varians Wallich ex Hook. &
Grev., see Morton, 1973) aggregate is at present being carried out by Professor T. Reichstein at
Basel. By contrast, on the European mainland, eight of the 13 known polyploid species of
Dryopteris and Polystichum have been at least partially elucidated, six of them completely, and
all 18 known polyploid European Asplenium have been completely elucidated. The figures for
North America are now more complete still. Athyrium in Europe cannot be compared with Asia
as it has only two species, both diploid sexual. It is therefore to be hoped that in the future more
complete work may be carried out on some of the groups of Indian species, as a species has not
been ‘worked out’ merely because a meiotic chromosome-count has been carried out. In this
respect my study (Fraser-Jenkins, 1985) should be considered as a preliminary catalogue to aid
such future work, and may well need to be added to and altered accordingly.

Study of the morphology, size, and degree of fertility of spores is another method which
throws much light on the nature or identity of a species. Unfortunately this has been neglected
for much of Asia, or, when it has been carried out, has often been of a vague general nature,
covering one or two species from various groups instead of its somewhat more useful application
to a particular group of species. Spore ornamentation is sometimes characteristic of a species, as
may usually be readily observable with the light microscope, which may produce results as
diagnostically useful as those from stereoscan studies, the latter often being more difficult to
interpret. More importantly the mean length of the exospore (ignoring the perispore and also
ignoring unusually small spores or those standing on end in the preparation), even from as small
asample as 20 spores from an individual, as long as they are fully ripe, gives information relevant
to ploidy within a group, and often more reliable information than that from stomatal guard-cell
size, which may vary more markedly depending on environment. Thus a pair of closely related
species within one group (but not across differing groups with unrelated species, whose basic
diploid spore size may differ from case to case), where one is diploid and another tetraploid,
almost always, so far as has been discovered, shows some spore-size difference, though there
may be some overlap. The non-abortive spores of an apomict are also markedly larger than
those of a related sexual diploid species, being twice the ploidy of a normal spore, and the same
ploidy as the sporophyte plant-body. It is also highly useful to observe whether the spores are
partially abortive as in an apomict, or uniform as in a sexual species, or fully or almost fully

_ abortive (frequently + shapeless, or markedly irregularly shaped, with thickened or + irregular
_ perispores) as in a spontaneously originating sterile hybrid. But care must be taken to ensure
that the sample is fully ripe, with black sporangia in the process of dehiscence at the time of
| collection of the specimen, as young spores are smaller, and very young ones are small, without

|

|

40 CHRISTOPHER R. FRASER-JENKINS

perispore, and appear white or pale; samples that are too old and have long ago dehisced (with
brown sporangia) have lost the heavier, good spores first and may have an unusually high
proportion of abortive ones left over. It is therefore important to distrust occasional isolated
spores in samples with very few spores left, which could be contaminants, even after washing the
pinna under running water, though this reduces the number of loose spores considerably.
Sometimes breaking open individual undehisced sporangia, using two slides squeezed slightly
together, or a fine needle, or by placing in concentrated lactic acid on the slide, may produce
reliable results when there are very few spores left, at least if the spores are good and unaborted.
It is important, too, to distinguish between sporangial abortion due mainly to drought and
true spore-abortion due to hybridity, and dust in the sample must be distinguished (with
a dark background if necessary) from abortive spores so that clean samples are always prefer-
able. It is because of the usefulness of such spore studies that it is vital that cytological voucher
specimens have ripe spores wherever possible, even if plants have to be cultivated to allow
this.

All these studies need more emphasis in Asia in order to allow the taxonomic treatment of a
group to reflect more accurately the systematics. In some cases it is necessary to understand that
what has been seen as variation may really arise from the occurrence of a distinct combination of
genomes, which may then indicate that treatment as a distinct species may be warranted if it is
also upheld by the important artificial criterion that the plants can be readily morphologically
distinguished, not only by microscopic characteristics (in the Dryopteridaceae at least).
Alternatively, taxa described as different species may be superfluous due to the overlooking of
earlier names (often from different areas in Asia), or, a more frequent and much more
troublesome situation, may merely be variants described as species in error (often from too few
specimens) due to a previously unappreciated range of infraspecific variation, particularly that
which occurs in apomicts. Variation in allopolyploids may also cause some confusion which must
be sorted out; in Europe at least some have been observed to veer towards the morphology of
one ancestral species or the other from locality to locality, perhaps due to a particular set of
environmental factors bringing out a certain phenotypic response from the mixture of genes
originating from both ancestors. In other species local or widespread ecotypes or geographical
variants may form which can only be adequately treated (often as subspecies) once their nature
is understood.

An even more immediately important area of research which has been but poorly attended to
throughout Asia is that of typification, which allows consequent accurate identification. Without
this a publication may be highly confusing, especially if, as is often the case, material is cited
without numbering and thus cannot be properly or conclusively checked at a later date in a
herbarium. The problem arises partly from the fact that most of the type specimens are stored in
Europe and funds are often not forthcoming to allow visits to see them, while publications in
India seldom give clear descriptions and illustrations. Nevertheless typification has often merely
involved interpretation from the name or reading the protologue of the publishing author (or
accounts of later authors) rather than studying the relevant type specimens, which are for the
most part preserved and available in London (K and BM), Paris (P), or Berlin (B), with
south-east Asiatic types usually in Leiden (L) or Berlin (B), Chinese mainland ones in Peking
(PE) and Canton (IBSC), and Prov. Taiwanese specimens in Tokyo (TI) and Kyoto (KYO).
Otherwise many identifications depend on those carried out on herbarium collections some
years ago which were sent from the subcontinent to Alston, Ching, and others, and which were
generally not accurate enough for the modern post-cytological elucidation of complex groups or
confused species. In many cases it is not easy to select a representative type when an author may
have had mixed concepts, or cited a number of specimens, or inadvertently published a name
that was previously invalid. In this, as in any study, it has been necessary to unravel and
understand the validating author’s concept before selecting a lectotype. Only in this way can an
author be sure of the identity of a species he is dealing with. Typification has almost invariably
been carried out by reference to the original material in the present studies (Fraser-Jenkins,
1985), except in a few stated cases (where the type is cited without an exclamation mark after it)
and it is hoped that the photographic illustrations of the species and mention of their diagnostic

AN INTRODUCTION TO FERN GENERA OF THE INDIAN SUBCONTINENT 41

characteristics, following this typification, will provide a helpful basis for identification in future
studies.

Although keys are provided for each section (Fraser-Jenkins, 1985), it should be pointed out
that the species concepts of later workers should not be based upon the so-called ‘key-
characteristics’; these may be misleading, as keys for complex fern genera frequently do not hold
true, and falsely emphasise a particular characteristic which in reality is only a tendency. It must
be pointed out, too, that because in these genera many of the species vary considerably, it must
be expected that many individuals may not readily be keyed out with a dichotomising system
based on presence or absence of a few artificially weighted characteristics. Yet the species may
be readily separated by the human brain once actual material is seen with the correct name
attached; this is done by means of observing and partly intuitively summing up a combination of
characteristics. The author is firmly of the perhaps slightly unorthodox opinion that comparing
specimens with a photograph of each species is both more immediate and more accurate than the
use of a key. Reference may be made to both the illustrations and the text (to discover which
features to look for) in the monograph (Fraser-Jenkins, 1985) and checks can be made on the
spores for evidence of hybridity, apomixis, or tentative ploidy as indicated by spore-size, when it
is expected that accurate identifications will ensue. In cases where important material that has to
be identified seems not to fit the best course would be to carry out a meiotic chromosome count if
time and materials permit, and to make comparisons with material known to be correctly
identified in herbaria. For this purpose the author has set about determining material in many of
the relevant herbaria when time was available, or, more reliably, citing specimens in the text.

Much detailed attention has also been paid here to questions of nomenclature and the
application of the present /nternational code of botanical nomenclature. As a result of careful
study of questions of nomenclature and typification, there are several nomenclatural changes to
be made which have lain unnoticed or semi-dormant in recent literature. Many of these involve
names attributed to Wallich, who, with his large series of mixed collections with invalid names
sent out to numerous institutes and botanists of his day, unwittingly caused lasting trouble and
difficulties. Yet there are still many workers, particularly Indian botanists, who do not number
each plant separately in their collections of herbarium material and to a certain extent run some
of the same risks when their collections are utilised in their publications. Far too many mistakes
have been perpetuated by following previous authors uncritically, rather than studying each case
anew when compiling a work for publication. The author can only apologise in advance for any
inaccuracies or omissions he has failed to notice.

Finally some remarks on the making and labelling of specimens may be of value. It is
emphasised that a good specimen can be easily identified and its characteristics more easily seen,
whereas a poor specimen can often create considerable difficulties. With most of the larger ferns
undoubtedly the most useful features are the basal half of the frond and the scales near the stipe
base. Thus a specimen whose basal and lower-middle pinnae are properly spread out and visible
is of far more use than one where the basal pinnae (with their highly important lowest pinnules,
especially the lowest basiscopic one) are either folded up or obscured by the upper part of the
frond lying over them. An ideal large specimen should have the stipe bent sideways just below
the lamina and then again upwards shortly before the edge of the herbarium sheet is reached,
care being taken during collecting and drying never to handle the frond by the stipe, which
causes the often highly characteristic scales to be rubbed off. The bottom of the lamina can then
be placed at the bottom of the sheet. It also helps in all the dryopteridaceous ferns (at least) if the
upper surface of the lower part of the lamina is displayed so that its important surface texture and
colour are visible; the under surface is rarely as informative. However, an important feature of
the under surface is the sori and their indusia, and in order to display them the tip of the frond
may be bent under the rest of the frond from near the top of the sheet and deflected to one side.
With smaller specimens some fronds may be mounted with the top surface uppermost and others
the other way up on the same sheet. Particularly with the Polypodiaceous and epiphytic ferns
parts of the creeping rhizome with its helpful scales should also be preserved, attached to the
fronds, but if possible without earth, which contaminates the spores and makes them harder to
interpret. Whenever possible, specimens with ripe (black) sori should be selected and those with

42 CHRISTOPHER R. FRASER-JENKINS

young sori avoided unless an extra frond is deliberately included in order to show young indusia.
Once good specimens have been made it makes them of far greater value and use if full localities
are given on the label; all too often, especially with Kumaon and Nepalese localities, the label
merely indicates a very small and obscure village or stream and the Indian State concerned.
These obscure localities may be extremely difficult to trace and may involve hours searching
detailed maps; frequently they cannot even be traced by asking in local botanical institutes,
because, unless the original collector can be found, few of the collectors have any wide
knowledge of their country. A properly written label must indicate the approximate altitude,
direction, and approximate distance from a village and then place the village in relation to large
and well-known towns, etc.; in very remote regions a map reference may be of assistance if no
well-known towns are nearby. Separate plants should be numbered separately and the number
and date clearly indicated on the label; thus when they are cited in any publications the exact
specimen referred to can be traced by later workers.

A review of Indian subcontinent fern study

A general bibliography of the more important publications (unless they are specific to certain
genera) appears after the end of this account of the literature, but it is of use here to sum up the
work of the major authors and fern-workers in a chronological review, making note of botanical
advances.

The first Indian ferns were described by Linnaeus (1753) and later amplified by Swartz (1801
and 1806), but it was not until Professor D. Don [1799-1841] published his Prodromus Flora
Nepalensis (February 1825), that a spectrum of the more obvious Himalayan fern species was
described. Don drew on the collections of Hamilton (whose former name was Buchanan),
Roxburgh, (whose collections were published posthumously with the help of Griffith), Wallich,
and others. Unfortunately, Don’s species are often very difficult to typify, his descriptions are
usually rather too vague to allow identification by themselves, and his genera, such as
Polypodium, Nephrodium, Aspidium, and Asplenium, were broadly delimited by modern
standards. What remains of his original specimens are mostly to be found in BM. These are
almost all the collections of Hamilton, and have Hamilton’s rather large writing on them, with
the locality as given by Don; many of them (including a very few Wallich collections) also bear a
small label and name written in Don’s distinctive, small handwriting, with slits cut into the label
through which the specimens were originally inserted, and with a reference to Don’s Prodro-
mus. Hamilton collected in Nepal from 1802-03 for the East India Company. But most of Don’s
species had Wallich collections as their types, and only a few of these survive today with
annotations by Don; it is possible that, if Don ever had such specimens in his possession, which is
doubtful, they may have been destroyed in a fire long ago (Christensen via Ching 1980, pers.
comm.), though the author has been unable to verify this. Some other types also exist in Paris (P)
and Bruxelles (BR). At present the most accurate course of action when typifying Don’s species
based on Wallichian collections, which do not exist bearing Don’s label at the BM, is to draw on
the nearest to the original East India Company Wallich material. This is either that distributed to
Sir W. J. Hooker, now in the general herbarium at Kew, or preferably, the separate Kew-
Wallich herbarium, which contains a large proportion of the specimens left in the East India
Company herbarium after Wallich’s main distributions, and thus the bulk of the specimens Don
actually worked on. It is necessary to lectotypify the specimens at Kew by reference to Don’s
description and, if they exist, to Don’s names written on the sheets by other authors, and, most
importantly, to the sense in which Sir W. J. Hooker and other authors of the last century took
Don’s names, while ascertaining that the specimen concerned was collected before 1825, the
date of publication of Don’s work. It clearly requires a full knowledge of what specimens exist,
and of both the genus concerned and the sense in which other authors took the name, before
lectotypes are selected, and this should not be done only by reference to the somewhat scanty
material present in Indian herbaria. Morton (1967, 1973) has made some useful comments on
some of Don’s names, among others. Don came under severe criticism at the time (Lindley,
1825; Wallich, 1830) and his work was largely deliberately ignored mainly for his independent

AN INTRODUCTION TO FERN GENERA OF THE INDIAN SUBCONTINENT 43

nomenclature, because he generally did not use Wallich’s (or others’) names or numbers, even
for Wallich collections, thus causing endless confusion today. It appears that though he was
commissioned to work on Hamilton’s collections he unauthorisedly took over many of Wallich’s
too, while Wallich was still working on them in India before he died. Lindley (1825) was furious
that Don altered all Wallich’s names and substituted his own, and criticised Don as an
uneducated interloper, while Wallich (1830) refused to use Don’s names, though they were
legitimate. He was upset that Don rejected his names for no good reason and complained that he
had never in his ‘life experienced such conduct, or anything in the remotest degree like it, from
any other quarter’. Unfortunately, his contemporary’s refusal to acknowledge his work, though
justified, has contributed markedly to so little being known about many of Don’s species at
present. Some of his names have caused further confusion as it has been unclear whether they
were independent later homonyms of names published by Swartz, etc., or whether they were
really intended to be based on an earlier author, who is sometimes cited at the beginning of the
genus. From analysis of the 14 apparent later homonyms among the 64 (actually 66, or 72
including fern allies) new species of ferns in Don’s work (Smith & Fraser-Jenkins, 1981), it is
clear that Don’s later homonyms are indeed independent of earlier authors; they are only based
on earlier authors when he cites the authority in the account of the species concerned. This is the
case even when the two species are the same, which occurs with a few of the homonyms (e.g.
Aspidium paleaceum, a name which had long been under dispute). Don described three
Dryopteris, six Polystichum, and three Athyrium species, as well as species in other genera.

The extensive collections of Dr N. Wallich (formerly named Wolff and coming originally from
Denmark) [1786-1854], for the East India Company, made mainly in Kumaon (coll. Robert
Blinkworth), Nepal (from the surroundings of the Kathmandu valley) and Khasia, from 1815 to
1832, and later up until 1849, were drawn on by a number of authors. But unlike Don they used
the names that Wallich had given to his specimens and had circulated as invalid names, without
descriptions, on the herbarium sheets, or in his numbered catalogue lists of 1828-1832 and
1847-1849. Unfortunately, Wallich never described these ferns himself and much of his series of
collections, originally at East India Company House and later at the Linnean Society before
going to the Kew—Wallich herbarium, was distributed by him all over Europe and later to India,
and sometimes contained muddled labels, or more frequently contained more than one species
per sheet or catalogue number. Much of the material sent to European botanists was intended to
be on loan, but it was not until Clarke reorganised the Wallich herbarium that specimens were
returned and reunited with the remnants of Wallich’s material from the Linnean Society to form
the Kew—Wallich herbarium. It is therefore sometimes very difficult to ascertain the concept of
an author who first validated a Wallich name, and so find out to which species the name applies,
without taking account of which particular specimen of the number cited the particular author
had in mind. Also different authors often meant different things by the same name as many did
not realise that they should have referred to the earliest valid publication of a Wallich name (by
another author) as the protologue of that species, and have decided on the type accordingly.
Indeed in some cases a name attributed by an author to Wallich may really be typified from the
author’s concept, as shown in his description, as a species different from any of the numerous
Wallich specimens of the relevant catalogue number. Nomenclatural confusion and typification
problems persist to this day, not helped by the vague localities which Wallich often gave, such as
‘Napalia’ (Nepal). Thus although Wallich considerably increased the number of species known
at the time, another result of his work has been a great deal of confusion, all of which requires
clarification, species by species. Some species were given three or even four names by him, anda
confused modern researcher ensnared in nomenclatural questions may well be forgiven for
wishing that Wallich had collected in unicate and never named any of his specimens. In all he
collected about eight Dryopteris, seven Polystichum, and eight Athyrium species from the
Indian subcontinent, named under several different genera and under various specific names, or
overlooked among the collections. It is only thanks to the two Hookers’ and Clarke’s careful
work on his herbarium that sense can be made of it, and perhaps fortunately, rather few of his
names with their attendant typification problems stand today.

At about the same time as Wallich’s collecting, C. L. von Blume, whose herbarium is in

44 CHRISTOPHER R. FRASER-JENKINS

Leiden (the type specimens bearing his signature ‘Bl.’), published a large number of fern species
from Java in his Enumeratio plantarum javae (1828), several of which occur in the Indian
subcontinent, mainly in the east Himalaya, south India, or in Sri Lanka. Some have also been
mistakenly reported from the area due to misidentification, and it is possible that others may yet
have been overlooked. Hence the importance of the herbarium at Leiden for studies of Indian
ferns, some isotypes of Blume’s also being at Kew.

Following Wallich’s distributions a number of the important early German authors, whose
type specimens (including those that are Wallich collections) are mostly in Berlin, published
some of his or their own species, for example: Sprengel (1827), Presl (1836) and (1851) (without
description, so invalid), Kunze (1851, etc.), Kuhn (1869), and Mettenius (1856-69). Kunze and
Mettenius between them published a rather large number of species from all over Asia as well as
the Indian subcontinent and their works are therefore of considerable importance for the area.
Kunze’s types were mainly destroyed during the Second World War at Leipzig, but isotypes that
he saw usually exist in Berlin (with some at Kew), apart from his south-east Asian types which
are mainly at Leiden. Of considerable importance, too, was the upsurge of many, mainly
British, botanist-collectors (with a few German ones, mostly employed in India, and one or two
French voyagers-cum-botanists), who began in earnest last century, and continue to the present
day. The earlier ones included Beddome, Bourne, Freeman, Gardner, Hutchison, Meebold,
Perrottet, Schmid, Thwaites, Trimen, Walker, Wall, and Weigle, from south India and Sri
Lanka; Aitchison, Anderson, Atkinson, Bates, Beddome, Blanford, Bliss, Brandis, Cattell,
Cave, Clarke, Collett, Duthie, (and his collectors Harsukh and Inayat), Edgeworth, Falconer,
Gammie, Gamble, Gatacre, Haines, Harriss, Hooker (fil.), Hope, Jacquemont, Jerdon, Jukes,
King (and his collectors Kunstler, Scully et al.), Lace, Levinge, Macleod, the two Mackinnons,
Marten, McDonell, Meebold, Prain, Rawson, the Schlagintweits, Smith, J. L. Stewart,
Strachey, Thomson, Treutler, Trotter, Wight, and Winterbottom in the west Himalaya,
Darjeeling, and Sikkim; and Booth, Clarke, Griffith, Hooker (fil.), Hutton, Mann, Meebold,
Thomson, Sauliére, Simons, and Watt in the east Himalaya. A number of published works by
some of these collectors mention ferns, the less comprehensive ones including Aitchison
(1881-1882 & 1888), Anderson (1863), Blandford (1886, 1888, 1889), Cattell (1877), Collett
(1902 & 1921), Duthie (1906), Ferguson (1880), Griffith & Roxburgh (1844), Decaisne in
Jacquemont (1844), Marten (1909), Prain (1903), Strachey (1906), Thwaites (1864), Trimen
(1885), Trotter (1889), and Wall (1873). It should be noted here that Thwaites, who distributed
his specimens to many herbaria from Peradeniya Botanic Garden, Sri Lanka, used a system of
numbers (e.g. C.P. 1368, etc.) based on the catalogue of plants at Peradeniya and many of the
numbers were used for different gatherings and may contain mixed species, so that the numbers
cannot be treated like modern collection numbers. C.P. numbers were also attached to
collections of Gardner and some others, and are indexed in Thwaites’ book.

It is convenient to mention here too, in connection with the British botanist-collectors, Dr T.
Moore of Kew, who did not visit India but named a number of Indian species among his works,
often in his Index filicum (1857-1862). This was not completed before his death, only reaching as
far as Goniophlebium in alphabetical order, so that several of his species (in Lastrea, Polys-
tichum, etc.) were not validly published except for those which appeared in obscure places such
as horticultural works (e.g. Sim’s catalogues of 1859, Gardeners’ chronicle, etc.). J. Smith of
Kew in his Historia filicum (1875) also listed a number of Indian species, including Wallich’s
names, but without descriptions, his herbarium being at BM. Often overlooked, but of
importance, are the horticultural trade catalogues of R. Sim (1859, 1866, etc.) which have also
provided some early validations of Wallich names.

Of the British botanists, four deserve special mention as having produced outstanding,
detailed works, as yet unsurpassed in many respects, which together form a set of the most
important references, vital for any workers on Indian subcontinent ferns. The four are Hooker
(senior), Beddome, Clarke, and Hope.

The first of these, Sir W. J. Hooker (senior) [1785-1865], Director of Kew, never visited
India, but nevertheless published a number of important works of monumental scope, dealing
with ferns world-wide and placing and ordering the species in their genera more carefully than

AN INTRODUCTION TO FERN GENERA OF THE INDIAN SUBCONTINENT 45

had been done before; these include Hooker & Greville’s Icones filicum (1827-31), Hooker’s
Species filicum (1846-1864) in five volumes, and Hooker & Baker’s Synopsis filicum (1865-
1868). Altogether Hooker dealt with approximately 14 Dryopteris, 13 Polystichum, and 15
Athyrium species from the area, and, apart from Don, was the first botanist after Wallich to have
access to the rich East India Company collections. Wallich had also sent Hooker a set of his
specimens, starting in 1818, and Hooker published numerous details of this collection in his
various works. Hooker’s set of these ferns (combined with others) is in the Kew general
herbarium, usually bearing the circular stamp ‘Herbarium Hookerianum 1867’ on the sheets,
and should normally be referred to for the types when typifying names attributed to ‘Wallich ex
Hooker’, though if specimens are missing it should be borne in mind that he also worked on the
rest of the Wallich material from the East India Collection which is now best represented by the
Kew-Wallich herbarium. His son, Sir J. D. Hooker (fil.) [1817-1911], Director of Kew,
travelled extensively in the Himalaya, collecting mainly with Dr T. Thomson, both in the west
Himalaya and in Sikkim and the east, and deposited valuable sets of specimens at Kew (with
duplicates elsewhere) collected from 1847-1852 and given to Kew in 1867, bearing the same
Herb. Hooker stamp on the sheets. Details of his journeys can be found in his Himalayan
journals (1855, etc.), but though he went on to produce the Flora of British India he
unfortunately did not include ferns. His other major contribution was the partial reorganisation
of Wallich’s collections from 1847-1855 at the Linnean Society, including the sending of a partial
set to Kew (apart from that already sent by Wallich to his father) and to Calcutta (see Clarke
(1913), and anon. ‘Sir J. D. Hooker’, Kew Bulletin 1912: 5).

Lieutenant-Colonel R. H. Beddome [1829-1911] lived in southern India and was the
Conservator of Forests, Madras. He collected slightly later than J. D. Hooker, between 1857
and 1898, and published his two volume, The ferns of southern India and Ceylon (1863-64),
describing and figuring all the species known to him at the time, including numerous new
species. Though less familiar with the north than the south he went on to publish the two volume
Ferns of British India (1865-70), with a Supplement to both his works (1876), and the excellent,
Handbook to the ferns of British India (1883), which latter, together with the Supplement to the
handbook (1892), is the most comprehensive book on Indian ferns to date. In all, his
contribution to the understanding of Indian ferns was probably greater than that of any other
worker and he was able to recognise about 26 Dryopteris, 22 Polystichum, and 20 Athyrium
species. His types, annotated on the bottom of the sheet in his somewhat thin, rather sloping,
handwriting are mostly at the BM (including most of the specimens he illustrated) or otherwise
at Kew. Recently, a Companion to Beddome’s handbook by Nayar & Kaur (1974), has been
produced in recognition of the fact that there is no modern fern-flora of India. It has attempted
to bring up to date the nomenclature of Beddome’s work and is a useful publication, though
generic concepts were somewhat out of date or inaccurate in places. But it suffers from the
drawback that the authors did not see Beddome’s types (Kaur, pers. comm. 1979), but tried to
unravel the jumble of other authors’ opinions as to what Beddome meant, which has resulted in
much confusion and unreliability, the names not being identified, but many concepts indicated.

Beddome’s lead was followed by C. B. Clarke [1832-1906] who made extensive, detailed, and
carefully numbered collections, superior to those of many collectors even today, from all over
the Himalaya, including the now virtually inaccessible mountains of eastern Assam. He then
went to Kew and was responsible for numerically arranging (between 1881 and c.1887) the
Wallich collection according to Wallich’s list, then housed in the Linnean Society buildings in
London, and for adding many of the missing numbers which were returned from various
European herbaria at his request. These later became the separate Kew—Wallich (K-W)
herbarium of today (see Clarke, 1913) and may be taken as the types of names attributed to
‘Wallich ex Clarke’, though not, as Clarke believed, of all Wallich names, which depend on the
concepts of the validating authors and the specimens they saw, or which are most representative
of their concepts. Clarke’s major fern publication was his Review of the ferns of northern India
(1880), based mostly on his collections. In it he recognised many of the more critical and
previously unrecognised species, though he made occasional muddlings of names or, in one or
two cases, dealt with the same species under two names, and following earlier authors, tended to

46 CHRISTOPHER R. FRASER-JENKINS

reduce too many species to varieties of well-known European ferns, such as Athyrium
filix-femina, Polystichum aculeatum, and Dryopteris filix-mas. But his work is usually accurate
and is of fundamental importance for the area. Unfortunately, he did not select types and gave
only the ranges of his species rather than citing specimens or localities; but the best material from
which types of his species or varieties may be selected is represented by his specimens at Kew, or
otherwise at BM, though duplicates exist in several other herbaria.

The last of the ‘golden age’ of British fern collectors in India was C. W. W. Hope [1832-1904]
who collected mainly in the Simla area, and followed this up, on returning to England in 1896, by
careful work in herbaria. Those he worked through include DD, CAL, P, DBN, LIV, BM, K,
MANCH, and E. He developed an intuitively good ‘eye’ for different species, especially in
Dryopteris and other large and critical genera, probably being the most careful and detailed of
any fern-workers in the area to the present day, as his annotations at Kew and elsewhere show.
He frequently recognised an obscure species, not generally recognised until much later, as being
distinct, even if he had no name for it. Some new species were published in his paper, Three new
Lastreas from Assam (1890) and others in his Ferns of the Chitral Relief Expedition (1896), but
most were in his Ferns of the north western Himalaya (1899-1904). This latter contains perhaps
the finest fern drawings (by Brown, Fitch, and others) seen in any European or Asiatic literature
and which, unlike most, capture the look of the species and match in every detail the
corresponding BM specimens. It is also the most detailed and comprehensive work yet
published on north-west Indian ferns and deals with 17 Dryopteris, 16 Polystichum, and 15
Athyrium species from that area alone. It is unfortunate that his work did not extend to other
areas as well. He, like Clarke, did not select types, but he cites numerous specimens in detail and
it is usually possible from his discussions to select lectotypes with which the name is clearly
associated, at Paris, the British Museum, Kew, or sometimes from Blanford’s herbarium in
Manchester or the Forest Research Institute at Dehra Dun. The main part of his herbarium is in
Paris (P), with originals of his illustrations at BM.

At the turn of the century there was a lull in the contribution of British botanists to Indian
pteridology, but Dr Hermann Christ of Basel [1833-1933] filled this gap with his work on
Chinese ferns. His herbarium contains specimens sent from Chinese collectors and most of his
types and is in the Museum National d’Histoire Naturelle in Paris (P). Many of his numerous
papers recombined Clarke’s varieties as species, or gave them valid names, or described new
species and varieties, common to both sides of the Himalaya in many cases. He brought to
attention earlier works by Franchet, Diels, and others, and worked mainly on the collections of
Biondi, Bodinier, Cavalerie, David, Delavay, Ducloux, Esquirol, Faber, Farges, Franchet,
Faurie, Giraldi, Henry, Hugh, Léveillé, Maire, Martin, von Rosthorn (as published by Diels
from von Rosthorn’s original specimens at Oslo (O) and Berlin (B)), Scallon, Scortechini,
Soulié, Taquet, and Wilson. These were mostly French (with a few Italian, etc.) botanist-
missionaries who collected in China and whose material is mostly in Paris, with some duplicates
at Edinburgh, Kew, BM, and Berlin. Christ probably had a greater general knowledge of ferns
world-wide than anyone else before or since, and during his long life came to understand them in
detail similar to that of post-cytological work.

He later corresponded with another remarkable worker in his field, Dr C. F. A. Christensen
of Copenhagen [1872-1942], whose many publications began with the phenomenal world-wide
Index filicum (1905-6), with Supplements (1913, 1917, and 1934) (and see Pichi-Sermolli, 1965).
This listed all the world’s fern species and provided many new combinations for Indian species,
or reduced others to synonyms to create what is usually a more or less accurate synonymy of the
species. He thus included both Christ’s names (some of which were redescriptions of his own or
others’ species) and those of Rosenstock from S. China and S.E. Asia, which had remained
rather little-known. The types of Rosenstock’s species are mostly in Stockholm (S), or Leiden
(L), with some at Berkeley, California (UC) and duplicate specimens were well distributed
elsewhere. The index supplements were continued by Pichi-Sermolli (1965) and have been
compiled as card-indexes since then at Kew, a further publication being much needed, as it is
long overdue and its delay is somewhat inconvenient in the light of the large amount of modern
and often local work, which can be difficult to trace. It is now likely to be completed at Kew

AN INTRODUCTION TO FERN GENERA OF THE INDIAN SUBCONTINENT 47

before long by Parris etal. (in prep.). Christensen clearly recognised the modern genera, such as
Dryopteris (instead of Neophrodium and Lastrea), from which he separated the thelypteroids in
a major step forward (1938), Polystichum (instead of Aspidium), and Athyrium (instead of
Asplenium); all being confirmed by modern study (see: Christensen (1938), Holttum (1947) and
also Holttum’s subsequent work, including his series of papers on the thelypteroids of the Old
World, Sledge (1973), Lovis (1977), Pichi-Sermolli (1977), and Ching (1978) for recent details of
generic and family distinctions). Christensen’s herbarium, containing fragments or duplicates of
a great deal of important type material, is mainly at the BM, with duplicates at Copenhagen (C).
His shorter papers included several very useful ones dealing with a large number of collections of
Chinese species, whose types he always took the trouble to examine, and which he related to
Indian species.

More Chinese work, which continues in the fore-front of fern-work on the Asian flora today
was carried out by Professor Ching, Ren-Chang of Peking, in earlier days a colleague of
Christensen, who made extensive visits to western herbaria in the 1930s and is still working
actively at present. His publications, like Christensen’s, have been of major and general
importance, especially in terms of detailing generic limitations in a multitude of families, and his
Revision of Chinese and Sikkim Himalayan Dryopteris (1936-38) lists a large number of species
of Dryopteris and related genera with full details, including synonymy and distribution. It also
separates various other genera, including Crenitis (with its confusing C. apiciflora (Wallich ex
Mett.) Ching aggregate, often mistaken for Dryopteris). He has published numerous other
papers, describing, among many others, new species of Dryopteris, Polystichum, and Athyrium
from the immensely rich Chinese flora, many of which species are relevant to India too, though it
should be said that there are some typification problems, since a fair proportion of Ching’s
photographs of type specimens made in the 1930s do not in fact represent the correct type
specimens. Present major projects in which he has played a leading part are the Flora xizangica
(Tibet) and Flora reipublicae popularis sinicae, both now in active preparation, the ferns for the
former (Ching & Wu, 1982) being nearly complete. Both are of major relevance to the Indian
subcontinental flora. It is obvious, however, even though it is a matter of opinion, that much of
his work suffers from over-splitting at the generic and species level, many of the so-called species
being too readily described without further study from single herbarium sheets which often
represent juvenile plants or very slight phenotypic variants well within the species. This applies
even to his earlier work but is more markedly the case with his most recent publications such as
the Flora tsinglingensis; the Flora xizangica, in particular, suffers badly from it. Indeed it must
be said that continued publication of names of ‘new species’ in this way will cause many years
work for future botanists simply to be able to unravel and sink the taxa concerned into their
proper species, particularly as many of the types are poor specimens which are difficult to
identify and have been published at species rank instead of forma to which they usually more
properly belong, if indeed any taxonomic recognition is necessary. However, hidden among the
names there are good species as well, which perhaps compounds the difficulties, but makes his
work of great value and importance. For an example of the problems involved in studying at
Peking, Dryopteris in the Peking herbarium (PE) contains (1982) some 600 manuscript names,
several used more than once, which appear to be referable to only about a fifth or less of that
number of species (according to most western species concepts); most of the names are readily
referable to well known species, and effectively the bulk of the specimens in Peking herbarium
are thus unidentified and have never been fully used for mapping or other purposes. It is
unfortunate, too, that about three-quarters of the ‘species’, even those published some time ago,
still consist of only one collection, or sometimes two. A similar situation occurs with Polys-
tichum, Athyrioid ferns, Lepisorus and other complex genera, while some of the genera too may
be excessively split up and involve too rigid a reliance on a particular key characteristic to allow
all the species to be naturally placed. This is due mainly to following a species concept quite
different from modern biosystematic concepts elsewhere and it appears that each observable
variant is in general recognised as a species. The disastrous effects of the cultural revolution
(from c. 1965-1976), which temporarily prevented modern species concepts and cytological and
field studies from making their influence felt at a most crucial time, are also much to blame. Such

48 CHRISTOPHER R. FRASER-JENKINS

studies are now beginning on a remarkable scale and with considerable determination in China
and will certainly balance the excessively artificial species concept used there with a more natural
biological one in time. Ching’s monumental work with its many correctly described species, is of
great importance as some of his earlier species not only stand, but show a remarkable degree of
foresight; it is to be hoped that the present atmosphere and revitalised attitude towards
academic research in China, and contacts with foreign research workers, will help to overcome
the problems that have occurred, and that an excessive profusion of new ‘species’ and ‘genera’
will not continue to be described. There is no doubt that the standards, thoroughness, and
quality of Chinese work are indeed very good and that work of the highest quality will doubtless
ensue.

Despite the problems mentioned above Ching may be considered the foremost authority on
ferns alive today. His types are mainly at Peking (PE), Sian, and Canton (IBSC). Apart from
him there are well over a dozen active fern workers in China at present, many of them students of
his, including Wang, Zhong-Ren, Shing, Kung-Hsieh, Ling, You-Shing and Ying, Jung-Sen at
Peking; Hsieh, Yin-Tang at Sian; Wang, Jian-Zhong at Shenyang; Kung, Hsien-Shiu at
Chengdu; Wu, Su-Kung and Chu, Wei-Min at Kumming; Wang, Chu-Hao at Canton and Chu,
Pei-Shi at Shanghai. Much work is at present being done on a series of local floras, province by
province, including the provinces of Kiangsi, Tibet (Xizang), Szechuan, Yunnan, Kweichow,
Kwangsi, Hupeh, Chekiang, Fukien, Kwangtung, Hainan, and Shantung, which are now either
published or in active preparation; it is to be expected that many important new geographical
records will result, especially once superfluous species have been eliminated. But is also hoped
that more emphasis will be given to monographic studies, as in some ways local floras are
premature when the genera themselves are not well known: a few good generic monographs
covering all of China (incl. Prov. Taiwan) and taking account of the species of the Indo-
Himalaya, S.E. Asia, and Japan would be of great value.

It should be noted that in the new floras the orthography of species since the last two years in
China has been converted to Pin-yin spelling by governmental decree, which has resulted in
some names, particularly those containing the letters Q, X, C, Z, etc., becoming quite different
in pronunciation according to various European or Latin values of the letters, from the way
place names are pronounced in China. It is to be hoped that this may be changed as it is
somewhat against the spirit of the International code of botanical nomenclature (Voss et al.,
1983).

A somewhat overlooked area of work carried out in the first quarter of the century was
Hayata’s study of the ferns of Taiwan in his [cones plantarum formosanum (1914-19), followed
by Tagawa’s Studies on Formosan ferns (1940-1949), and more recently the less comprehensive
The flora of Taiwan by Li et al. and De Vol (1975), and some interesting papers by Serizawa
(1970-76), Tsi (1973), and Tsai & Shieh (1975, 1977). Though Christ (1904) drew on some of the
early collections of Faurie and Taquet from Taiwan, it is only now coming to light that there is a
very large and considerable floristic connection between the west and east Himalaya and
Taiwan, so that Taiwanese names and species may be highly relevant to the Himalayan fern
flora. The types of Taiwanese species are mostly in Japanese herbaria (Serizawa in TNS, Kurata
in TOFO, Hayata in TI, and Tagawa in KYO). There is a connection too between the Himalaya
and Japan, though markedly smaller than that with Taiwan, see Hirabayashi (1974), for
example, and Nakaike’s excellent and generally highly accurate Enumeratio pteridophytarum
Japonicarum (1975) and, for quick reference, Tagawa (1963) and Kurata & Nakaike (1979).

Another spate of work began on the north-eastern borders of the Indian subcontinent, in
China, in the first quarter of the present century with the collections of C. Schneider and H.
Handel-Mazzetti of Wien (specimens at W), Dr H. Smith of Uppsala (specimens in BM, PE, US
and UPS), Dr J. F. Rock (including the north-east border of Burma) (specimens at K, BM, and
US) and G. Forrest (specimens at E), up until the 1930s; and within the subcontinent, those of F.
Kingdon-Ward (specimens at BM) from east Tibet, Assam, and north Burma between 1924 and
1949 and, more importantly for ferns, those of R. E. Cooper (specimens at E) in Bhutan in 1914,
Dr N. L. Bor (specimens at BM, K, CAL and DD) in Assam from 1931-1936 and particularly
those of F. Ludlow, G. Sherriff and G. Taylor (see Fletcher, 1976, and Stearn, 1976) and others

AN INTRODUCTION TO FERN GENERA OF THE INDIAN SUBCONTINENT 49

who joined with them in Bhutan, Assam, and east Tibet (specimens at BM), whose excellent
collections, including many ferns, were made from 1933 until 1950, shortly after the closure to
foreigners of Tibet in 1947. All these areas are almost completely inaccessible today for political
reasons, though some brief collecting was also done in Bhutan between 1963 and 1965 by the
Botanical Survey of India (see Deb, Sen Gupta & Malick, 1968). In particular the fern flora of
north Burma, which is probably among the richest of the whole area, is very little known and
badly undercollected and the area cannot now be visited; only Rock, Kingdon-Ward and
Dickason have collected there to any extent, and in comparison with other areas their collections
were limited. The closure of the east however was partly compensated for by the opening up of
Nepal to foreigners in 1949 and the subsequent large collections of O. Polunin,
J.D. A. Stainton, W. R. Sykes and L. H. J. Williams (specimens at BM), albeit with very brief
information about localities that therefore take much finding, and of A. H. G. Alston, C. E. B.
Bonner and A. Zimmerman (specimens at BM and G). Capt. Lall Dhwoj also collected in Nepal
as early as 1927 and deposited specimens at BM. A considerable number of smaller collections,
such as those of J. F. Dobremez (University of Grenoble), H. Emery, Foster and E. W. Cronin,
from the Arun valley (American Peace Corps), T. Wraber (University of Ljubljana) and many
others, have been made in recent years, though most of the interesting part of the Himalayan
main range in west and east Nepal is now restricted again for foreigners, but this is partly
compensated for by the excellent new work of Nepalese botanists (see below). The Nepalese
collections began to open up large areas in the middle of the Himalaya that had been almost
blank, at least since Wallich’s rather restricted collection in south-central Nepal in 1820, so that
material became available which now allows the full ranges of species to be understood. Rather
little work has been done at Kew on the area over the last fifty years, except for Professor R. E.
Holttum’s outstanding and comprehensive studies of thelypteroid ferns (1969-1979), along with
some identification of submitted material. A. H. G. Alston at the British Museum (Natural
History) was able to identify much of the material sent from the area, if somewhat inaccurately
in critical genera, but nevertheless covering a wide spectrum and with a comprehensive
knowledge of the literature, which enabled him to identify incoming material and inform the
new schools of local botanists in India up until the time of his death in 1958.

The first active fern-worker in India, after a gap of some time, was the American botanist, Dr
R. R. Stewart, Principal of Gordon College, Rawalpindi, now Pakistan (specimens at the
National Herbarium, Islamabad (RAW), BM, and at Ann Arbor (MICH), Michigan, US and
PE), who built up a large collection of west Himalayan ferns, mainly from Pakistan and
Kashmir, areas that had been largely ignored beforehand. He also purchased the important
Trotter herbarium which, together with his own extensive collections, forms the basis of the
National Herbarium, Islamabad, formerly at Gordon College. He communicated with Alston in
London and published several useful papers on the area (Stewart, 1942, 1945, 1951, 1952, 1957,
1977). Before he retired to Ann Arbor, Michigan, U.S.A., in 1960, where he is still actively
working, he inspired a generation of local botanists whom he had taught at Gordon College, and
others whom he influenced elsewhere.

Another worker with whom Stewart had considerable contact was Professor P. N. Mehra of
the Punjab University, Lahore (formerly a co-worker of Professor S. R. Kashyap), who later
moved to Amritsar as a result of the partition of India and Pakistan in 1947, and then founded
the botany department of the new Panjab University at Chandigarh. Professor Mehra is
undoubtedly a founding figure in modern Indian botany and under his remarkable direction the
Panjab University probably became the most active school of botany in the subcontinent.
During the last forty years many of his students, including those who, since partition in 1947,
moved to new centres in India, have worked on the ferns of their own country with small mono-
graphic papers and local or regional, annotated lists. Under Mehra’s influence, the new
generation of local botanists has seen the revolution in botanical studies resulting from the
arrival of the new cytological approach pioneered by the work on ferns of Manton (1950, etc.).
In addition to more traditional botanical studies, which, it must be emphasised, are still every bit
as important now as before, Mehra was able to introduce some of the impact of the modern
cytological approach and some of the results of systematic studies into fern-work in the Indian

50 CHRISTOPHER R. FRASER-JENKINS

subcontinent. A large number of preliminary cytological reports have been published and an
increasing number of botanical publications are being produced from the increasing universities
and institutes there, including the nation-wide offices of the Botanical Survey of India, with its
headquarters at Calcutta. Among others, some of the more active centres in the subcontinent,
with herbaria and with their staff producing work relevant to these monographs, are as follows:

In Pakistan:

The Pakistan Forest Institute, Peshawar (PPFI), where Dr A. R. Beg (an ex-student of Stewart), and
others are working.

The University of Karachi (KUH), where Professor S. I. Ali (an ex-student of Stewart) is working with
Professor E. Nasir on the excellent emergent Flora of West Pakistan (Nasir & Ali (eds) 1970-79 and in
prep.) which will probably include ferns. There is unfortunately no such corresponding ‘Flora of India’,
though it is at present being discussed and planned by the Botanical Survey of India hopefully drawing on
the expertise of authors world-wide, group by group, and would be a much more major undertaking.

The National Herbarium and Stewart Collection, Islamabad (RAW), where Professor E. Nasir (an
ex-student of Stewart) is co-editing the Flora of West Pakistan and has set up a small but highly competent
botanical institute.

There is also a new Quaid-i-Azam University herbarium (ISL) at Islamabad, which has produced a new
journal (Pakistan systematics (1977- etc.)) and needs further encouragement and advice, having built up a
useful collection rapidly. It is to be hoped that their work will develop and continue.

In India, the main establishments which are important for fern study are as follows:

The Jammu and Kashmiri University, Srinagar (KASH), where Professor P. Kachroo has worked on
ferns.

The Panjab University, Chandigarh (PAN), where Professor P. N. Mehra, retired, has carried out much
fern work, and from where Professor S. C. Verma and Professor D. S. Loyal (both ex-students of Mehra)
have published numerous papers. The next generation of active fern workers there, who also graduated
under Mehra, are Dr S. P. Khullar, Dr K. K. Dhir and Dr H. S. Puri.

The Punjabi University, Patiala (PUN), where Professor S. S. Bir (an ex-student of Mehra), previously
at Chandigarh, has published and edited a large amount of important taxonomic fern literature.

The Forest Research Institute, Dehra Dun (DD), which has what is probably the finest fern-herbarium
in the subcontinent with many important, old collections, including those from the Saharanpur herbarium,
containing isotypes etc. from Clarke, Hope, Blanford and others, and where Dr K. M. Vaid (an ex-student
of Stewart) has published on ferns.

The University of Delhi (DUH), with Dr N. P. Chowdhury, who has worked on ferns.

The University of Poona, where Professor T. S. Mahabale, retired (now of the Maharashtra Association
for the Cultivation of Science-Research Institute) and others have studied the anatomy of certain groups of
ferns.

The National Botanical Research Institute and Gardens, Lucknow (LWG), most actively run by Dr
T. N. Koshoo with, among other active workers, Dr S. Kaur who has worked on ferns.

The Benares Hindu University, Varanasi (BAN), with Dr S. K. Roy, who studied under Professor
Manton at Leeds, and Dr J. B. Singh, who have worked on ferns.

The University of Patna, with Professor R. P. Roy and Dr B. M. B. Sinha, who have worked on ferns,
including those from Nepal.

The University of Kalyani, Kalyani, Nadia, West Bengal where Drs U. Sen and T. Sen have worked on
fern anatomy.

The Central National Herbarium of the Botanical Survey of India (CAL), at Howrah, Calcutta, where
Dr G. Panigrahi (who studied briefly under Manton at Leeds), Dr A. S. Rao, Dr K. Nag, and others have
worked on ferns. This herbarium is an important one with many old collections, though unfortunately
recent local workers have sometimes taken it upon themselves to cross out the more accurate and
important identifications of the original collectors, and many sheets can be found which have had their
original labels cut in half or even cut off altogether in order to fit them on to the new, smaller-sized sheets
now in use. Some skilful interpretation may therefore be necessary on occasions, and some specimens may
have been rendered useless. The regional offices of the B.S.I. are at Dehra Dun (BSD), Coimbatore
(MH), Shillong (ASSAM), Jodhpur (BSJ), (with B. V. Shetty, Malhotra and R. P. Pandey, who have
worked on ferns), Poona (BSI), and Allahabad (BSA).

The University of Calcutta (CUH), where Drs S. & N. Pal have worked on ferns.

The Jawarharlal Nehru University, Imphal, Manipur, where Dr J. Ghatak, formerly of CAL, who
studied under Manton at Leeds, has worked on ferns.

AN INTRODUCTION TO FERN GENERA OF THE INDIAN SUBCONTINENT 51

The University of Bangalore, where Dr S. N. Agashe has published on ferns.

The University of Kerala, Trivandrum, where Drs A. Abraham, K. V. Bhavanandan, C. N. Ninan, and
P. N. Matthew have worked on ferns and on the cytology of south Indian species.

The Botanical Survey of India, Southern Circle (MH), at Coimbatore, including what little remains of
the Old Madras Herbarium, with Dr N. C. Nair, who has worked on ferns.

The University of Calicut, where Professor B. K. Nayar and Dr K. K. Geevarghese have worked on
ferns.

In Sri Lanka:

The National Herbarium and Botanic Garden, Peradeniya, Sri Lanka (PDA), which contains an
important collection of Sri Lankan specimens, including many older collections by Thwaites, Trimen, etc..,
with Catalogue Peradeniya (C.P.) numbers, and which are frequently type material.

Numerous other Indian workers have worked on pteridophyte genera not relevant to the
present study. Many of the publications from the research workers mentioned above are the
only modern ones available for numerous genera and the lack of a fern flora of larger scope is
keenly felt. Unfortunately it is necessary to point out that much modern Indian work suffers
badly from numerous misidentifications, and there is a noticeable tendency to redescribe species
already known under earlier names. This is partly due to inability to obtain funds to examine the
important type specimens, which as a result of history are almost all in Europe, and mostly at
Kew and BM in Britain. Indeed the situation concerning lack of typification is so serious that few
Indian publications can be fully relied upon. A useful code of practice might well be not to
publish, where typification is important, until the genuine, correct type or a photograph of it has
been seen, as it is clear that in many ways it is better not to publish at all than to cause even
greater confusion merely because that most essential part of the work has not been carried out.
There is also too much local institute or university political pressure concerning promotion
prospects, etc., at present, to publish too frequently, and often a great deal of rivalry with little
communication between workers, or even the unauthorised taking over of work done by others,
from determinations made in herbaria, etc., which has caused the quality and aim or purpose of
work to suffer. Other noticeable problems are the poor herbarium specimens which are all too
often fragments, or lack stipe scales, or are twisted and folded up and brown, the untraceably
brief localities, the lack of numbering with identifiable cytological and other voucher specimens
(where each plant should be numbered separately and the numbers cited to back up each
cytological result), or vouchers with unripe spores, often due to cytological fixation being carried
out only in the field, rather than culturing plants to subsequent spore maturity. Some of the
problems with collecting are caused by the use of newspaper without flimsy papers and without
proper drying paper, and local botanists all too frequently concentrate on a few well-known,
low-level areas, even avoiding many of the far more interesting high-level areas that have roads
leading to them, and thus are often unaware of the many extra species occurring higher up, even
though they may have been clearly recorded in the past. Unfortunately private transport is not
generally available as it isin Europe, but more useful and constructive programmes for the use of
institute vehicles could easily be planned to allow improvements and to increase their workers’
topographical knowledge of a wider area of the Himalaya. It is to be hoped that these
‘teething-troubles’ in Indian work will eventually be eliminated, and especially that cytological
studies will go further than mere lists of chromosome counts; the development of a few small
gardens for the culture of Himalayan species at such amenable places as Simla and other hill
stations would allow experimental research, which is sadly lacking at present merely because
high-level species cannot be cultivated on the plains where the main institutes are based. For
comparison, and as a guide to what has been done in this field, an excellent summary of the
results gained from more far-reaching recent studies in Europe and North America, much of it
relevant too to the Indian subcontinent, is to be found in Lovis (1977), as well as in Manton
(1950). Both of these workers were at Leeds University, though Professor Lovis is now at the
University of Canterbury, New Zealand.

The advent of the Japanese botanists into the area with the Flora of the east Himalayas reports
(It6, Tagawa, Nishida & Iwatsuki, 1966; It6, Tagawa, Nishida & Iwatsuki, 1971; Iwatsuki, 1975)
has added a number of new records from their collecting excursions. However, the first two

52 CHRISTOPHER R. FRASER-JENKINS

reports have also included a considerable number of misidentifications in the ferns due to lack of
typification, with some redescription of previously known species, and erroneous records of east
Asian and Taiwanese taxa in error for comparatively well-known Himalayan species. These are
fortunately mostly corrected in the much more reliable third report. Their collections (in Tokyo
(TI), with duplicates in KATH, BM, and K, etc.) have clearly been of a high quality and are of
considerable use, often including new records of rare species due to their energy in collecting
from remote places at considerable altitudes, which have usually not been visited by Indian
botanists.

In Nepal, a new and active school of botanists has arisen, with a herbarium at Godavari
belonging to His Majesty’s Government, Department of Medicinal Plants and Herbs, Kathman-
du (KATH), which has produced preliminary fern-work of a good standard (see Gurung, 1974,
1976a, 1976b, and in press). The herbarium contains a large number of excellent specimens,
albeit rather poorly identified (based mainly on the earlier Japanese expeditions’ determinations
and those done by the Botanical Survey of India, Calcutta), and the range of their collecting
excursions into remote parts of the high Himalaya, accessible only by trekking, is a fine example
to other workers in the subcontinent. Other work has been carried out by Dr A. R. Sakya of the
Botany Department, University of Kathmandu (see Roy, Sinha & Sakya, 1971).

A considerable amount of collecting of Nepalese ferns was also carried out by the American
worker, Dr R. L. Fleming senior, of the United Mission to Nepal, and of Woodstock School,
Mussoorie, N. India, who has recently (in 1980) retired to Phoenix, Arizona, U.S.A., and has
worked with Dr R. R. Stewart and with the Nepalese botanists. His specimens are at Kew, BM,
Dehra Dun (DD), Ann Arbor (MICH), and Kathmandu (KATH).

It is unfortunate that restrictions exist once more in all the interesting western and eastern
parts of the Himalayan main range in Nepal so that the furthest east that foreign workers may
visit in all the Himalaya, apart from south-east Nepal, the Darjeeling area, and S.W. Sikkim, is
the upper Arun valley east of the Everest range. The east Himalayan ranges are thus closed to
the bulk of botanical specialists and collectors, including local botanists in all the far eastern
part.

Some important work by British botanists on the fern flora of Sri Lanka (Ceylon) in recent
years has been carried out by Professor I. Manton and Dr W. A. Sledge of Leeds University,
including also detailed cytological studies on the genus Adiantum throughout India (Manton,
Ghatak & Sinha, 1967; Manton, Sinha & Vida, 1970; Sinha & Manton, 1970), and surveys of the
cytology of Sri Lankan ferns (Manton, 1953; Manton & Sledge, 1974). Sledge has produced a
fine series of considerably detailed papers or monographs of fern genera in Sri Lanka (Sledge,
1960, 1962, 1965, 1973, 1981, 1982) and, while unfamiliarity with related Himalayan species is
sometimes noticeable and has caused some inaccuracies, his work is of remarkable scope and
fundamental importance for the area.

Recent major work on the Indian subcontinent flora includes Hara, Stearn & Williams (1978),
Hara & Williams (1979), and Hara, Chater & Williams (1982) working on the Nepalese flora at
the British Museum (Natural History), London, but excluding ferns, and Grierson & Long (in
preparation) working on the Flora of Bhutan at Edinburgh, under the sponsorship of the Royal
Bhutanese Government and Overseas Development Aid, and including a brief account of the
better-known ferns. A detailed account of the ferns and fern-allies of Pakistan west of the Indus,
Afghanistan, and Iran, is also in preparation by Fraser-Jenkins, Khullar & Reichstein (in
preparation) for Professor K. H. Rechinger’s far-reaching series Flora iranica, nearing comple-
tion from Wien, Austria.

General bibliography

This bibliography has emphasis on works which include Polystichum, Dryopteris, and Athyrium,
and are relevant to the Indian subcontinent and associated areas.

Abeywickrama, B. A. 1956. The genera of Ceylon pteridophytes. Ceylon J. Sci. A, 13: 1-30.

—— 1964, The pteridophyes of the Knuckles region. Ceylon J. Sci., Biol. Sci. 5 (1): 18-29.

Abraham, A., Ninan, C. A. & Matthew, P. M. 1962. Studies on the cytology and phylogeny of the

AN INTRODUCTION TO FERN GENERA OF THE INDIAN SUBCONTINENT 53

pteridophytes VII. Observations on one hundred species of south Indian ferns. J. Indian bot. Soc. 41:
339-421.

Aitchison, J. E. T. 1881. On the flora of the Kurram valley, etc., Afghanistan I. J. Linn. Soc. (Bot.) 18:
1-113.

— 1882. On the flora of the Kurram valley, etc., Afghanistan II. J. Linn. Soc. (Bot.) 19: 139-200.

— 1888. The botany of the Afghan Delimitation Commission. Trans. Linn. Soc. Lond. II, 3 (1): 1-139.

— 1889. A summary of the botanical features of the country traversed by the Afghan Delimitation
Commission during 1884-1885. Trans. bot. Soc. Edinb. 17: 421-434.

Alston, A. H. G. 1958. Pteridophyta. In M. Kgie & K. H. Rechinger, Symbolae Afghanicae IV. Biol. Skr.
10 (3): 7-12.

Alston, A. H. G. & Bonner, C. E. B. 1956. Résultats des expéditions scientifiques Genevoises au Nepal en
1952 et 1954 (partie botanique). 5. — Pteridophyta. Candollea 15: 193-220.

Anderson, T. 1863. On the flora of Bihar and the mountain Parasnath, with a list of species collected by
Messrs. Hooker, Edgeworth, Thomson and Anderson. J. Asiat. Soc. Beng. 32: 189-218.

Anon. 1912. Sir Joseph Hooker. Kew Bull. 1912: 5.

Awasthi, D. K. & Sharma, M. P. 1980. Ecological and phytogeographical observations on the ferns and
fern-allies of Nagpur block (Chamoli Garhwal), western Himalayas. Proc. Indian Acad. Sci. (Plant Sci.)
89: 307-313.

Baehni, C., Bonner, C. E. B. & Vauthier, S. 1952. Plantes recoltées par le Dr. Wyss-Dunant au cours de
L’Expedition Suisse 4 L’Himalaya en 1949. Candollea 13: 213-236.

Baishya, A. K. & Rao, R. R. 1982. Ferns and fern allies of Meghalaya State, India: 125-129. Jodhpur.

Baker, J. G. 1888. On two recent collections of ferns from western China. J. Bot., Lond. 1888: 225.

Banerji, M. L. 1955. Botanical explorations in east Nepal. J. Bombay nat. Hist. Soc. 55: 243-268.

— 1963. Outline of Nepal phytogeography. Vegeratio 11 (S—6): 288-296.

— 1972. A collection of ferns from eastern Nepal. Candollea 27: 265-281.

— 1973. Nepal, its forests and phytogeography. Jn P. K. K. Nair, R. K. Grover & T. M. Verghese (Eds),
Glimpses in plant research 1: 182-205.

Baroni, E. & Christ, H. 1897. Filices plantaeque filicibus affines in Shen-Si septentrionali provincia Imperii
Sinensis a Rev. Patre Josepho Giraldi collectae. Nuovo G. bot. ital. Il, 4 (1): 3-19. Boll. Soc. bot. ital.
1898: 27-32, 1898: 182-184, 1900: 260-263, 1901: 288-292.

— 1901. Filices Setciouenses a Rev. Patre U. Scallan collectae etc. in Monte Uo-Mi-San, 1899. Boll. Soc.
bot. ital. 1901: 293-297.

Beddome, R. H. 1863-1864. The ferns of southern India and Ceylon: 1-38, t.1-110 (1863). 39-88, t.111-271
(1864). Madras. 1873. (2nd ed.) 1-88, t.1-270. Madras. [Reprinted in Delhi 1970]

— 1865-6. 1866-70. The ferns of British India. 1: 1-120 (1865), 121-150 (1866). 2: 151-210 (1866),
211-255 (1867), 256-300 (1868), 301-330 (1869), 331-345 (1870). Madras.

— 1876. Supplement to the ferns of southern India and British India: 1-28, t.346-390. Madras.

— 1883. A handbook to the ferns of British India, Ceylon and the Malay peninsula. Calcutta. [Reprinted
in Delhi 1976]

— 1892. Supplement to the ferns of British India, Ceylon and the Malay peninsula. Calcutta. [Reprinted
in Delhi 1976]

— 1893. Notes on Indian ferns. J. Bot., Lond. 31: 227-228.

Bhavanandan, K. V. 1968. Studies on the cytology of sixteen species of south Indian ferns. Caryologia 21:
333-338.

— 1981. Studies on the cytology of south Indian Aspidiaceae. Cytologia 46: 195-207.

Bir, S. S. 1962. Cytological observations on some ferns from Simla (western Himalayas). Curr. Sci. 31:
248-250.

— 1963. Observations on the pteridophytic flora of Simla Hills (north western Himalayas). Bull. Bot.
Surv. India 5: 151-161.

— 1964. Taxonomic notes on some Himalayan ferns. J. Indian bot. Soc. 43: 556-572.

— 1965. Chromosome numbers in some ferns from Kodaikanal, south India. Caryologia 18:
107-115.

—— 1968. Pteridophytic flora of Simla Hills (north western Himalayas): introduction and general account.
Nova Hedwigia 16: 439-447.

—— 1973. Cytology of Indian Pteridophytes. Jn P. K. K. Nair, R. K. Grover, & T. M. Verghese (Eds),
Glimpses in plant research 1: 28-119. Delhi.

Bir, S. S. & Shukla, P. 1967. Cytology of some north Indian ferns. Cytologia 32: 24-30.

Bir, S. S. & Vasudeva, S. M. 1971. Pteridophytic flora of Kodaikanal (south India). J. Bombay nat. Hist.
Soc. 68 (1): 169-195.

54 CHRISTOPHER R. FRASER-JENKINS

—— 1973. Ecological and phytogeographical observations on the pteridophytic flora of Pachmarhi Hills
(central India) J. Indian bot. Soc. 51: 297-304.

Bir, S. S. & Verma, S. C. 1963. Ferns of Mt. Abu. Res. Bull. Panjab Univ. II, 14: 179-194.

Birdwood, H. M. 1887. A catalogue of the flora of Mahabaleshwar and Matheran. J. Bombay nat. Hist.
Soc. 2: 107.

—— 1897. A catalogue of the flora of Mahabaleshwar and Matheran [2]. J. Bombay Nat. hist. Soc. 10:
394.

Biswas, K. P. 1935. Ferns and fern-allies of Burma. Proc. Indian Sci. Congr. 22: 255.

Blanford, H. F. 1886. The silver ferns of Simla and their allies. J. Simla Naturalists Soc. 2: 13-22.

— 1888. An annotated list of the ferns of Simla. Calcutta.

—— 1889. A list of the ferns of Simla in the N.W. Himalaya between levels of 4,500 and 10,500 feet. J.
Asiat. Soc. Beng. 57 (2): 294-315, t. 16-21. [Reprinted in Delhi]

Blatter, E. & d’ Almeida, J. F. 1922. The ferns of Bombay. Bombay.

Blume, C. L. von. 1828. Enumeratio plantarum javae et insularum adjacentium 2: 99-274. Leyden. 1830.
2nd ed. 2: 99-274. Leyden.

— 1829-1851. Flora Javae. Leyden.

Boissier, P. E. 1884. Flora orientalis 5: 719-751. Genevae, Basilea et Lugduni.

Bor, N. L. 1938. A sketch of the vegetation of the Aka Hills, Assam. Indian Forest Rec. (Bot.) 1: 103-221.

Breckle, S.-W. 1971. Vegetation in alpine regions of Afghanistan. In P. H. Davis, P. C. Harper & I. C.
Hedge (Eds), Plant life of south west Asia: 107-116. Edinburgh.

Breckle, S.-W., Frey, W. & Hedge, I. C. 1969. Botanical literature of Afghanistan. Notes R. bot. Gdn
Edinb. 29: 351-371.

Burkill, I. H. 1910. Notes from a journey to Nepal. Rec. bot. Surv. India 4: 59-140.

— 1924-1925. The botany of the Arbor expedition. Rec. bot. Surv. India 10 (1): 1-154 (1924); 10 (2):
155-420 (1925).

Burrard, S. G. & Hayden, H. H. 1907-1908. A sketch of the geography and geology of the Himalaya
mountains and Tibet. Calcutta.

[Cattell, W.] (‘An Amateur’) 1877. A handy guide to the known ferns of the Himalayas of northern India.
Lahore.

Chadhuri, R. K. 1966. Annual report of the cytogenetics laboratory of the Department of Botany,
University of Calcutta. Res. Bull. Univ. Calcutta 1: 35.

Chandra, P. 1979. Ferns of Kedarnath, Madhyamaheshwar and Tungnath. J. Bombay nat. Hist. Soc. 74:
640-650.

— 1980. Botanical exploration in Tawang. Ferns and fern allies. Nova Hedwigia 32: 399-414.

Chatterjee, D. 1940. Studies on the endemic flora of India and Burma. J/ R. Asiat. Soc. Beng. 5: 1-69.

Chaudhri, M. N., Vegter, I. H. & De Wal, I. M. 1972. Index herbariorum II (3). Collectors (I-L). Regnum
veg. 86: 297-473.

Chiarugi, A. (ed.) 1960. Tavole chromosomiche delle Pteridophyta. Caryologia 13: 27-150.

Ching, R. C. 1930-1949. The studies of Chinese ferns 1. Sinensia 1: 43-45 (1930). 2. Bull. Fan meml Inst.
Biol. 1: 145-159 (1930). 3. Bull. Fan. meml Inst. Biol. 2: 1-14 (1931). 4. Bull. Fan meml Inst. Biol. 2:
15-28 (1931). 5. Bull. Fan mem Inst. Biol. 2: 185-223, t. 1-31 (1931). 6. Sinensia 1: 175-200 (1931). 7.
Sinensia 2: 9-36 (1931). 8. Sinensia 2: 131-156 (1932).9. Bull. Fan. meml Inst. Biol. 4: 47-116 (1933). 10.
Bull. Fan mem Inst. Biol. 4: 293-363, t. 1-2 (1933). 11. Contrib. Inst. bot. Nat. Acad. Peiping 2: 31-100,
t. 3-4 (1933). 12. Bull. Chin. bot. Soc. 1: 36-72 (1935). 15. Lingnan Sci. J. 13: 493-501 (1934). 20. Bull.
Chin. bot. Soc. 2: 85-106 (1936). 29. Sunyatsenia 6: 1-14 (1941). 30. Bull. Fan meml Inst. Biol. (Bot.) 10:
1-18 (1940). 31. Hong Kong Nat. 10: 194-204 (1941). 32. Lingnan Sci. J. 21: 31-37 (1945). 33. Bull. Fan
mem Inst. Biol. (Bot.) 10: 173-183 (1940). 34. Bull. Fan meml Inst. Biol. (Bot.) 11: 51-76 (1941). 35.
Bull. Fan mem Inst. Biol. (Bot.) 11: 79-81 (1941). 36 [‘35’]. Bull. Fan meml Inst. Biol. IT, 1: 267-317
(1949). [Nos 13, 14, 16-19, and 21-28 were not published]

— 1930-1958. Icones filicum sinicarum 1: 1-50 (1930), 2: 51-101 (1934), 3: 101-150 (1935), 4: 151-200
(1937), 5: 201-250 (1958).

—— 1933. The pteridophyta of Kiangsu province. Sinensia 3: 319-348.

— 1934. A revision of the compound leaved polysticha and other related species in the continental Asia
including Japan and Formosa. Sinensia 5: 23-91.

—— 1936-1938. A revision of the Chinese and Sikkim—Himalayan Dryopteris with reference to some
species from neighbouring regions. Bull. Fan meml Inst. Biol. (Bot.) 6: 237-352 (1936), 8: 157-268
(1938), 8: 275-334 (1938), 8: 363-507 (1938).

—— 1959. Pteridophyta (Ophioglossaceae-Oleandraceae). In S. S. Chien & W. Y. Chun, Flora reipubli-
cae popularis sinicae 2: 1-379. Peking.

AN INTRODUCTION TO FERN GENERA OF THE INDIAN SUBCONTINENT 55

— 1978. The Chinese fern families and genera: systematic arrangement and historical origin. Acta
phytotax. sin. 16: 1-37.

Ching, R. C. & Wu, S. K. 1980. The floristic characteristics of the Xizang (Tibet) pteridophyte flora in
relation to the upheaval of the Himalayas. Acta bot. yunnan. 2: 382-389.

— 1982. Pteridophyta. Jn Wu, Chen-Yih (Ed.) The series of the scientific expedition to Qinghai-Xizang
plateau. Flora Xizangica 1. The comprehensive scientific expedition to the Qinghai-Xizang plateau.
Academia Sinica. Peking.

Chowdhury, N. P. 1971. Researches on living pteridophytes in India, Burma and Ceylon. London.

— 1973. The pteridophytic flora of the upper Gangetic plain. New Delhi.

Christ, H. 1896, 1899, 1901, 1902, 1904. Filices Faurieanae [Japan & Formosa] Bull. Herb. Boissier 4:
664-675, 7: 817-824, II, 1: 1013-1021, II, 2: 825-832, II, 4: 609-618.

— 1898. Fougéres de Meng-tze, Yunnan mér. Bull. Herb. Boissier 6: 860-880, 956-973.

—— 1899. Fougéres de Meng-tze II. Bull. Herb. Boissier 7: 1-22.

— 1902. Filices Bodinierianae. Bull. Acad. int. Géogr. bot. 11: 189-274.

—— 1903. Filices Chinae centralis leg. Wilson. Bull. Herb. Boissier Il, 6: 508-514.

—— 1904, 1906, 1909, 1910. Filices Cavalerianae. Bull. Acad. int. Géogr. bot. 13: 105-120, 16: 233-246, 20
(Mém.): 169-178, 20: 137-143.

— 1905. Les collections de Fougéres de la Chine au Muséum d’Histoire Naturelle de Paris. Bull. Soc.
bot. Fr. 52 (Mém. 1): 1-69.

— 1906a. Filices Esquirolianae. Bull. Acad. int. Géogr. bot. 16: 247-252.

— 1906b. Filices Chinae Occidentalis auspiciis James Veitch et sons a E. H. Wilson 1903 et 1904
collectae. Bull. Acad. int. Géogr. bot. 16: 97-142.

— 1907a. Filices yunnanensis duclouxianae. Bull. Acad. int. Géogr. bot. 17: 129-140.

— 1907b. Filices chinensis esquirol et cavalerie 1905-1906. Bull. Acad. int. Géogr. bot. 17:
140-151.

— 1908. Filices coreanae novae. Reprium nov. Spec. Regni veg. 5: 284-285.

— 1909a. Fougéres d’extréme-orient: Filices faurieanae coreanae, and: Filices insulae sagalien. Bull.
Acad. int. Géogr. bot. 20 (Mém.): 146-168.

— 1909Db. Filices novae chinenses. Jn H. Lecomte (Ed.), Notulae Systematicae 1: 33-58.

— 1910a. Filices taquetianae coreanae. Bull. Acad. int. Géogr. bot. 20: 4-11.

— 1910b. Filices michelianae [Kouy-Tchéou]. Bull. Acad. int. Géogr. bot. 20: 12-16.

— 1911. Filices wilsonianae. Bot. Gaz. 51: 345-359.

— 1912. Filices. Annu. Conserv. Jard. bot. Genéve 15-16: 178-222.

Christensen, C. 1905-6. Index filicum 1: 1-384 (1905), 2: 385-744 (1906). [Reprinted in Koenigstein 1973]

— 1911. Pteridophyta in Insula Quelpaert a cl. P. Taquet anno 1910 lecta. Bull. Acad. int. Géogr. bot.
21: 69-72.

— 1913a. Index filicum Supplementum 1906-1912. Hafniae. (Supplements I-III reprinted together in
Koenigstein 1973)

— 1913b. Filices purdomianae. Bot. Gaz. 56: 331-338.

— 1913c. Filices esquirolianae 1910-1911. Bull. Acad. int. Géogr. bot. 23: 137-143.

— 1916. Spec. nov. Pterid. In H. Léveillé, Catalogue des plantes de Yunnan: 98-112.

— 1917. Index filicum. Supplément préliminaire. Hafniae.

— 1924. Plantae sinensis a Dre. H. Smith annis 1921-1922 lectae. 3. Pteridophyta. Acta Horti gothoburg.
1: 41-110.

— 1927. Ona small collection of Pteridophyta from the province of Kansu, China. J. Wash. Acad. Sci.
17: 497-501.

— 1931. Asiatic Pteridophyta collected by Joseph F. Rock, 1920-1924. Contr. U.S. natn. Herb. 26:
265-337.

— 1934. Index filicum. Supplementum tertium. Hafniae.

1938. Filicinae. In F. Verdoorn (Ed.), Manual of pteridology: 522-550. The Hague.

Christensen, C. & Tardieu-Blot, M. L. 1934-1939. Les Fougéres d’Indochine 3. Bull. Mus. natn. Hist. nat.
Paris Il, 6: 445-451 (1934). 4. Phanérogamie 5: 2-13 (1935). 6. Phanérogamie 5: 260-267 (1936). 11.
Phanérogamie 6: 135-149 (1937). 15, 16, & 17. Phanérogamie 8: 175-210 (1939). [See also Tardieu-Blot
& Christensen 1934-38]

Clarke, C. B. 1874. Cryptogamia miscellania. Jn W. Roxburgh, Flora Indica, 2nd ed.: 745-763.

— 1876. Botanic notes from Darjeeling to Tonglo. J. Linn. Soc. (Bot.) 15: 116-159.

— 1880. A review of the ferns of northern India. Trans. Linn. Soc. Lond. II (Bot.) 1: 425-611.
[Reprinted in Delhi 1973]

— 1889. On the plants of Kohima and Muneypore. J. Linn. Soc. (Bot.) 25: 90-107.

56 CHRISTOPHER R. FRASER-JENKINS

—— 1898. On the subareas of British India, illustrated by detailed distribution of the Cyperaceae in that
empire. J. Linn. Soc. (Bot.) 34: 1-146.

— 1913. The Wallichian herbarium. Kew Bull. 1913: 255-263.

Collett, H. See Hope.

Copeland, E. B. 1947. Genera filicum. Waltham, Mass.

Deb, D. B., Sen Gupta, G. & Malick, K. C. 1968. A contribution to the flora of Bhutan. Bull. bot. Soc.
Beng. 22: 169-217.

Decaisne, M. J. 1844. Botanique. Jn V. Jacquemont. Voyage dans L’Inde par V. Jacquemont, pendant les
années 1828 a 1832 4: 175-178, t. 176-179. Paris.

Devi, S. 1977. Spores of Indian ferns. New Delhi.

Devi, S. & Nayar, B. K. 1969. Spore morphology of Indian ferns IX. Ophiglossaceae, Marattiaceae,
Osmundaceae & Schizaeaceae. J. Sen Mem. Vol. Bot. Soc. Bengal: 209-222.

De Vol, C. E. & Kuo, C. M. 1979. Additional notes on Taiwan pteridophytes. Taiwania 24: 106-107.

Dey, A. C., Unniyal, M. R., & Shankar, V. 1969. Flora of the Bhillaganga valley of the erstwhile
Tehri-Garhwal State. J. Bombay nat. Hist. Soc. 65: 384407.

Dhir, K. K. 1980. Ferns of north-west Himalayas. Panjab University, Chandigarh. J. Cramer Bibliogr.
Pteridologica 1: 1-158. Vaduz.

Distribution patterns of Himalayan ferns. Nova Hedwigia [In press].

Dhir, K. K. & Datta, K. S. 1976. Ferns of Dharmsala Hills — 2. (North western Himalayas) Families
Woodsiaceae, Aspidiaceae and Athyriaceae. Nova Hedwigia 27: 393-424.

— 1977a. Ferns of Dharmsala Hills 1. (North western Himalayas). Ophioglossaceous and Schizaeaceous
Series. J. Bombay nat. Hist. Soc. 74: 459-480.

—— 1977b. Ferns of Dharmsala Hills — 3. (North western Himalayas). Thelypteridaceae, Aspleniaceae
and Blechnaceae. Nova Hedwigia 28: 137-154.

—— Ferns of Dharmsala Hills 4. (North western Himalayas). Family: Polypodiaceae. Nova Hedwigia. [In
press]

Dhir, K. K. & Sheera, P. S. 1975. Ecological and phytogeographical observations on the pteridophytes of
Dharmsala Hills (N. W. Himalayas). Nova Hedwigia 26: 353-371.

Dhir, K. K. & Sood, A. 1981. Fern-flora of Mussoorie Hills. J. Cramer Bibliogr. Pteridologica 2: 1-99.

Dickason, F. G. 1946a. A phylogenetic study of the ferns of Burma. Ohio J. Sci. 46: 73-108.

1946b. The ferns of Burma. Ohio J. Sci. 46: 109-141.

Diels, L. 1901 [‘1900’]. Die Flora von Central-China. Bot. Jb. 29: 169-211.

Dixit, R. D. 1979. Studies on pteridophytes in Botanical Survey of India. Bull. bot. Surv. India 19:
309-314.

Dobremez, J. F. 1972. Les grandes divisions phytogéographiques du Népal et de l’Himalaya. Bull. Soc.
bot. France 119: 111-120.

Dobremez, J. [F.], Vigny, F. & Williams, L. H. J. 1972. Bibliographie de Nepal 3 (2). Bot. Paris.

Don, D. 1825. Prodromus florae nepalensis: 1-19. London. (Reprinted in Delhi 1976)

Dudgeon, W. & Kenover, L. A. 1925. The ecology of Tehri-Garhwal: a contribution to the ecology of the
western Himalaya. J. Indian bot. Soc. 4: 233-284.

Duthie, J. F. 1893-4. Report on a botanical tour in Kashmir. Rec. bot. Surv. India 1: 1-18 (1893), 3: 25-47
(1894).

Duthie, J. F. & Strachey, R. 1906. Catalogue of the plants of Kumaon and of the adjacent portions of
Garhwal and Tibet, based on the collections made by Strachey and Winterbottom, 1846-9: 222, 232, 246.
London.

von Euw, J., Lounasmaa, M., Reichstein, T. & Widén, C.-J. 1980. Chemotaxonomy in Dryopteris and
related genera. Studia geobot. 1: 275-311.

Fabbri, F. (Ed.) 1963. Primo supplemento alle tavolo chromosomiche delle Pteridophyta di A. Chiarugi.
Caryologia 16: 237-335.

(Ed.) 1965. Secondo supplemento alle tavolo chromosomiche delle Pteridophyta di A. Chiarugi.
Caryologia 18: 675-731.

Fedtchenko, B. A. 1927. Vertikalnoe rasprostranenie paporotnikov Turkestanskoi goroi strani. Izv. Glav.
bot. Sada SSSR 26: 66-70.

—— 1938. Material for a flora of Afghanistan. Sbir. Rabot. Posv. Pamir. Akad. A.V. Fominizd. Akad.
Nauk. S.S.S.R. Kiev: 26-37.

Fée, A. L. A. 1852. Mémoires sur le famille des fougéres 5. Paris & Strasbourg. [For the date of this
publication see Pichi Sermolli in Webbia 9: 361-366, 1953]

Ferguson, W. 1880. Ceylon ferns. Colombo.

Fischer, C. E. C. 1938. The flora of Lushai Hills. Rec. bot. Surv. India 12: 75-161.

AN INTRODUCTION TO FERN GENERA OF THE INDIAN SUBCONTINENT 57

Fletcher, H. R. 1976. A quest of flowers: the plant explorations of Frank Ludlow and George Sherriff.
Edinburgh.

Fomin, AY. 1934. Pteridophyta. Jn V. L. Komarov (Ed.) Flora SSSR 1. Leningrad. [Translated from
Russian; Jerusalem 1968]

Franchet, A. 1887. Plantae davidiane ex sinarum imperio. 2. Plantes du Thibet Oriental (Province de
Moupine). Nouv. Archs Mus. Hist. nat. Paris I, 10: 115.

Fraser-Jenkins, C. R. 1980. Dryopteris affinis: a new treatment for a complex species in the European
pteridophyte flora. Willdenowia 10: 107-115.

— 1985a. A classification of the genus Dryopteris. Bull. Br. Mus. nat. Hist. (Bot.). [In press]

— 1985b. A monograph of the genus Dryopteris (Pteridophyta: Dryopteridaceae) in the Indian
subcontinent. Bull. Br. Mus. nat. Hist. (Bot.). [In press]

Fraser-Jenkins, C. R. & Khullar, S. P. 1983. Athyrium attenuatum — a confused Sino-Himalayan fern. Jn
S.S. Bir (Ed.), Aspects of plant sciences 6: 95-103. New Delhi.

Fraser-Jenkins, C. R., Khullar, S. P. & Reichstein, T. Pteridophyta. Jn K. H. Rechinger (Ed.), Flora
iranica Graz. [In preparation]

Freitag, H. 1971. Studies in the natural vegetation of Afghanistan. Jn P. H. Davis, P. C. Harper & I. C.
Hedge (Eds) Plant life of south-west Asia: 89-106. Edinburgh.

Fryson, P. F. 1932. The flora of the south Indian hill stations 1, 2. Madras.

Gamble, J. S. 1915-1935. Flora of the presidency of Madras 1-10. London.

Ghatak, J. 1961a. Problems involved in the proper identification of ferns. Bull. bot. Surv. India 3:
79-82.

—— 1961b. Some meiotic counts on ferns from the Khasia—Jaintia Hills. Proc. Indian Sci. Congr. 48 (3):
270.

—— 1962. Observations on the cytology and taxonomy of some ferns from India. Nucleus 5: 95-114.

— 1963a. Two new species of the Adiantum caudatum complex. Bull. bot. Surv. Ind. 5: 71-77.

— 1963b. Observations on the cytology and taxonomy of some ferns from India. Proc. Indian Sci.
Congr. 50 (3): 371-372.

— 1979 [‘1977’]. Biosystematic survey of pteridophytes from Shevaroy Hills, south India. Nucleus 20:
105-108.

Gray, W. 1886. The botany of the Bombay presidency. Bombay Gaz. 25: 376-377.

Greville, R. K. & Hooker, W. J. 1833. Enumeratio filicum. Bot. Misc. 2: 360-403, 3: 104-109.

Grierson, A. & Long, D. The flora of Bhutan. Edinburgh. [In preparation]

Griffith, W. & Roxburgh, W. 1844. The cryptogamous plants of Dr. Roxburgh. Calcutta J. nat. Hist. 4:
463-522.

Grimes, J. W. Index of Thelypteridaceae. Kew Bull. [In press]

Gupta, R. K. 1955-1962. Botanical exploration in the Bhillaganga valley of the erstwhile Tehri-Garhwal
State. J. Bombay. nat. Hist. Soc. 53: 581-594 (1955), 54: 878-886 (1957), 59: 486-512 (1962).

— 1962. Some observations on the plants of south India hill tops (Nilgiri and Palni plateaus) and their
distribution in the Himalaya. J. Indian bot. Soc. 41: 1-15.

— 1964. Bioclimatic types of the western Himalaya and their analogous types towards the mountain
chains of the Alps and the Pyrenees. Indian Forester 90: 551-577.

— 1966. Bibliography of Himalayan ecology, forestry and phytosociology. Excerpta bot. B, 7: 60-74.

Gurung, V. L. 1974. Enumeration of pteridophytes. Jn S. B. Malla et al. (Eds), Supplement to the flora of
Phulchoki and Godavari. Bull. Dept. Med. Plants Nepal 5: 1-38.

— 1976a. Pteridophyta. Jn S. B. Malla et al. (Eds), Flora of Langtang and cross sectional vegetation
survey (central zone). Bull. Dept. Med. Plants Nepal 6: 1-29.

—— 1976b. Pteridophyta. Jn S. B. Malla et al. (Eds), Catalogue of Nepalese vascular plants. Bull. Dept.
Med. Plants Nepal 7: 1-27.

— Pteridophyta. Jn S. B. Malla etal. (Eds), Flora of Kathmandu valley. Bull. Dept. Med. Plants Nepal 9
(in press).

—— InS. B. Malla et al. (Eds), Keys to the pteridophytic genera in Nepal. Bull. Dept. Med. Plants
Nepal (in preparation).

Haines, H. H. 1924. Pteridophytes. The botany of Bihar and Orissa.: 1183-1226. London.

Handel-Mazzetti, H. 1929. Pteridophyta. Symbolae Sinicae 6: 1-52. Wien.

Hara, H., Stearn, W. T. & Williams, L. H. J. 1978. An enumeration of the flowering plants of Nepal 1.
London.

Hara, H. & Williams, L. H. J. 1979. An enumeration of the flowering plants of Nepal 2. London.

Hara, H., Chater, A. O. & Williams, L. H. J. 1982. An enumeration of the flowering plants of Nepal 3.
London.

58 CHRISTOPHER R. FRASER-JENKINS

ne B. 1911. Materials for a flora of Formosa. Cryptogamiae. J. Coll. Sci. imp. Univ. Tokyo 30:

10-451.

—— 1914-1919. Icones plantarum formosanarum 4: 129-257 (1914), 5: 252-349 (1915), 6: 154-163 (1916),
7: 95-102 (1918), 8: 136-156 (1919).

Hirabayashi, H. 1974. Cytogeographic studies on Dryopteris of Japan. Tokyo.

—— 1981. A study on intraspecific polyploidal species of the Japanese pteridophytes. 1. List of
intraspecific polyploidal species of the Japanese pteridophytes. Bull. Fac. Toho Gakuen School Music 7:
1-25.

Holmgren, P. K., Keuken, W. & Schofield, E. K. 1981. Index herbariorum. Part 1. The herbaria of the
world. 7th ed. Regnum veg. 106: 1-450. Utrecht.

eet R. E. 1947. A revised classification of the leptosporangiate ferns. J. Linn. Soc. (Bot.) 53:
123-158. ;

—— 1949. The classification of ferns. Biol. Rev. 24: 267-296.

—— 1955 [*1954’]. A revised flora of Malaya. 2. Ferns of Malaya. Singapore. 1968. 2nd ed. Singapore.

—— 1960. Vegetative characters distinguishing the various groups of ferns included in Dryopteris of
Christensen’s Index filicum, and other ferns of similar habit and sori. Gdns’ Bull., Singapore 17:
361-367.

—— 1969a. On the significance of some name changes in ferns. Brit. Fern Gaz. 10: 193-199.

—— 1969b. Studies in the family Thelypteridaceae. The genera Phegopteris, Pseudophegopteris, and
Macrothelypteris. Blumea 17: 5-32.

—— 1971. Studies in the family Thelypteridaceae III. A new system of genera in the old world. Blumea 19:
17-52.

— 1972. Studies in the family Thelypteridaceae IV. The genus Pronephrium Presl. Blumea 20: 105-126.

—— 1973. Studies in the family Thelypteridaceae. VI. Haplodictyum and Nannothelypteris. Kalikasan 2:
58-68.

— 1974a [‘1973’]. Studies in the family Thelypteridaceae V. The genus Pneumatopteris Nakai. Blumea
21: 293-325.

— 1974b. Studies in the family Thelypteridaceae. VII. The genus Chingia. Kalikasan 3: 13-28.

—— 1975a. Studies in the family Thelypteridaceae VIII. The genera Mesophlebion and Plesioneuron.
Blumea 22: 223-250.

—— 1975b. Studies in the family Thelypteridaceae. IX. The genus Sphaerostephanos in the Philippines.
Kalikasan 4: 47-68.

— 1976a. Studies in the family Thelypteridaceae X. The genus Coryphopteris. Blumea 23: 18-47.

— 1976b. The genus Christeila Léveillé, sect. Christella. Studies in the family Thelypteridaceae, XI. Kew

Bull. 31: 293-339.

1977. Studies in the family Thelypteridaceae XII. The genus Amphineuron Holttum. Blumea 23:
205-218.

— 1979. Sphaerostephanos (Thelypteridaceae) in Asia, excluding Malesia. Kew Bull. 34: 221-232.

Holttum, R. E., Sen, U. & Mittra, D. 1970. Studies in the family Thelypteridaceae II. A comparative study
of the type-species of Thelypteris Schmidel, Cyclosorus Link, and Ampelopteris Kunze. Blumea 18:
195-215.

Hooker, J. D. 1854. Himalayan journals. London. 1855. 2nd ed. London. 1891. 3rd ed. London. [Reprinted
in Delhi 1969, 1974, & 1980]

— 1904. A sketch of the flora of British India. London.

Hooker, W. J. 1846-1864. Species filicum 1: 1-245, t. 1-70 (1846). 2: 1-250, t. 71-140 (1858). 3: 1-291, t.
141-210 (1860). 4: 1-292, t. 211-280 (1862). 5: 1-314, t. 281-303 (1864). London. [Reprinted in Delhi
1970]

— 1857-1859. Filices exoticae t. 1-32 (1857), 33-88 (1858), 89-100 (1859). London.

— 1886-1887. Icones plantarum 7 (17): 1601-1700. London.

Hooker, W. J. & Baker, J. G. 1865-1868. Synopsis filicum. London. 1874. 2nd ed. London.

Hooker, W. J. & Greville, R. K. 1827-1831. Icones filicum 1: 1-120 (1827-28), 2: 121-240 (1829-31).
London.

Hope, C. W. 1890. Three new Lastreas from Assam. J. Bot., Lond. 28: 326-329.

— 1892. Indian ferns. Calcutta Rev. Jan. 1892: 1-35.

— 1896. The ferns of the Chitral Relief Expedition. J. Bot., Lond. 34: 122-126.

— 1899-1904. The ferns of north-western India. J. Bombay nat. Hist. Soc. 12: 315-325. 12: 527-538. 12:
621-633 (1899). 13: 25-36. 13: 236-251 (1900). 13: 443-461. 13: 657-671 (1901). 14: 118-127. 14:
252-266. 14: 458-480 (1902). 14: 720-749. 15: 78-111 (1903). 15: 415-419 (1904). [Reprinted in Delhi,
with altered pagination]

AN INTRODUCTION TO FERN GENERA OF THE INDIAN SUBCONTINENT 59

— 1902. Pteridophyta. Jn H. Collett, Flora simlensis: XXXTX—XLI. London. 1921. 2nd ed.: XXXIX-—
XLI. Calcutta.

Hope, C. W. & Trotter, E. W. 1890. Catalogue of ferns in the herbarium of the Government of India at
Saharanpur.

It6, H. 1935. Filices Japonenses I. Bot. Mag., Tokyo 49: 359-366. II. Bot. Mag., Tokyo 49: 432-437. ,

1936. Filices Japonenses. III. Bot. Mag., Tokyo 50: 32-39. IV. Bot. Mag., Tokyo 50: 67-72. V. Bot.
Mag., Tokyo 50: 125-128.

—— 1937. Filices Japonenses. VI. Bot. Mag., Tokyo 51: 709-714. VII. 51: 725-730.

— 1939 [‘1938’]. Polypodiaceae, Dryopteridoideae I. In T. Nakai & M. Honda, Nova Flora Japonica 4:
1-243. Tokyo & Osaka.

— 1944. Illustrated flora of Japan. Tokyo.

— 1959. Index speciminorum typicorum in herbariis Japonensibus. Pteridophyta: 1-58. Tokyo.

It6, H., Iwatsuki, K. & Tagawa, M. 1971. Pteridophyta. /n H. Hara (Ed.), The flora of eastern Himalaya.
Second report. Univ. Mus. Univ. Tokyo Bull. 2: 197-221.

It6, H., Tagawa, M., Nishida, M. & Iwatsuki, K. 1966. Pteridophyta. Jn H. Hara (Ed.), The flora of eastern
Himalaya: 453-500. Univ. Tokyo.

Iwatsuki, K. 1975. Pteridophyta. Jn H. Ohashi (Ed.), Flora of eastern Himalaya. Third report. Univ. Mus.
Univ. Tokyo Bull. 8: 166-205.

Jacquemont, V. 1844. Voyage dans L’Inde, par V. Jacquemont, pendant les années 1828 a 1832. 4, Bot.:
175-178, t. 176-179. Paris.

Kachroo, P. 1953. List of ferns of Assam. J. Asiat. Soc. Calcutta 29: 161-174.

— 1975. Fern flora of Assam with some phytogeographical notes. J. Indian bot. Soc. 54: 13-26.

Kanai, H. 1963. Phytogeographical observations on the Japano—Himalayan elements. J. Fac. Sci. Tokyo
Univ. 3, 8: 305-339.

— 1966. Phytogeography of eastern Himalaya with special reference to the relationship between
Himalaya and Japan. Jn H. Hara (Ed.), The flora of eastern Himalaya: 13-38.

Kashyap, S. R. 1932. Vegetation of western Himalaya and western Tibet in relation to their climates J.
Indian bot. Soc. 4: 327-334.

— 1932. Some aspects of the alpine vegetation of the Himalaya and Tibet. Proc. Indian Sci. Congr. 19:
13-53.

Khullar, S. P. 1978. The significance of the perine in ferns. Proceedings Second All India Palynological
Conference [Bangalore] Abstract 9: 6.

— 1979. Cytological studies of some ferns from Darjeeling area. Third Indian Geophysical Conference
[Lucknow] Abstract: 21.

Khullar, S. P. & Mehra, P. N. 1970. IOPB chromosome number reports. Taxon 19: 404-441.

— 1972a. Cytotaxonomy of W. Himalayan Ferns. 1. Schizaeaceous series. Res. Bull. Panjab Univ. 23:
189-204. [See also Mehra & Khullar]

— 1972b. Cytotaxonomy of W. Himalayan ferns. 3. Family Polypodiaceae (sensu stricto). Nucleus 15:
156-162.

Kihara, H. 1955. Fauna and flora of Nepal Himalaya. Scientific results of the Japanese Himalayan
expedition to the Nepal Himalaya: 1-390. Kyoto.

Kingdon-Ward, F. 1944-45. A sketch of the botany and geography of north Burma. 1. J. Bombay nat. Hist.
Soc. 44: 550-574 (1944). 2. 45 (1): 16-30 (1945).

Kitagawa, M. 1979. Neo-lineamenta florae manschuricae. Vaduz.

Koshoo, T. N. (Ed.) 1959. Medicinal ferns of India. Nat. Bot. Gard. Lucknow Bull. 29: 1-36.

Kuhn, M. 1868-1869. Reliquae Mettenianae. Linnaea 35: 385-395 (1868). 36: 41-169 (1869).

Kunze, G. 1846-48. In Filices Javae Zollingerianas aliasque ex Herbario Moricandiano Observationes.
Bot. Zeitung 4: 417-426, 441-447, 457-463, 473-478 (1846), 6: 113-122, 141-146, 172-177, 189-199,
209-216, 234-239, 258-263, 282-285, 304-306 (1848).

— 1848. Pteridographia Japonica. Uebersicht der Farrn der Flora von Japan, sowie der Inseln Loo
Choo, Peel und Bonin Sima. Bot. Zeitung 6: 489-496, 505-509, 521-526, 53/-543, 553-559, 569-573,
585-589.

— 1851. Filices nilagiricae. Linnaea 24: 239-299.

Kurata, S. & Nakaike, T. (Eds) 1979. Illustrations of pteridophytes of Japan 1. Tokyo.

Kuriachan, P. I. 1964. Cytology of some south Indian ferns. Proc. combined 51st and 52nd Sess. Indian Sci.
Congr. 3. Abstracts: 359.

— 1967. Cytological observations on some south Indian ferns. Cytologia 32: 500-506.

Landon, P. 1928. Appendix XIV. Flora of Nepal. In Nepal 1: 357-358. Kathmandu. [Reprinted in H. K.
Kulgy (Ed.), Bibliotheca Himalayica I, 16: 357-358 (1976) New Delhi]

60 CHRISTOPHER R. FRASER-JENKINS

Lanjouw, J. & Stafleu, F. A. 1954. Index herbariorum II (1). Collectors (A—D). Regnum veg. 2: 1-174.

—— 1957. Index herbariorum II (2). Collectors (E-H). Regnum veg. 9: 175-296.

Léveillé, A. A. H. 1914-1915. Flore du Kouy Tch’eou: 487-494. Le Mans.

1915-1917. Catalogue des plantes de Yun-nan: 1-112 (1915-16), 172 (1917). Le Mans.

Li, H.-L., Liu, T.-S., Huang, T.-C., Koyama, T. & De Vol, C. E. 1975. Flora of Taiwan 1. Taipei.

Limpricht, W. 1922. Botanische Reisen in den Hochgebirgen Chinas und Ost-Tibets. Reprium Spec. nov.
Regni Veg. 12: 298-301.

Lindley, J. 1825. Note. Bot. Reg. 11: sub. tab. 872.

Linnaeus, C. 1748. Flora zeylanica: 1-240. Holmiae.

1753. Species plantarum: 1061-1106. Holmiae.

Léve, A. (Ed.) 1965. IOPB chromosome number reports V. Taxon 14: 191-196. [See also Solbrig]

Léve, A., Léve, D. & Pichi-Sermolli, R. E. G. 1977. Cytotaxonomical atlas of the pteridophytes. Vaduz.

Lovis, J. D. 1967. Fern hybridists and fern hybridising I. The work of Edward Joseph Lowe (1825-1900).
Br. Fern Gaz. 9: 301-308.

1968. Fern hybridists and fern hybridising II. Fern hybridising at the University of Leeds. Br. Fern.
Gaz. 10: 13-20.

—— 1977. Evolutionary patterns and processes in ferns. Jn R. D. Preston & H. W. Woolhouse (Eds),
Advances in botanical research 4: 229-415.

Loyal, D. S. 1966. Apomixis in ferns with particular reference to Dryopteris Adanson and some
heterosporous members. J. Indian bot. Soc. 45: 44-53.

Loyal, D. S. & Verma, S. C. 1960. Ferns of Nainital. J. Bombay nat. Hist. Soc. 57: 479-490.

Macpherson, T. R. M. 1894. List of ferns gathered in North Kanara. J. Bombay nat. Hist. Soc. 5: 375.

Malla, S. B., Shrestha, A. B. et al. 1974-1980. See Gurung.

Mani, M. S. 1978. Ecology and phytogeography of high altitude plants of northwest Himalaya. New Delhi.

Manton, I. 1950. Problems of cytology and evolution in the Pteridophyta. Cambridge.

—— 1953. The cytological evolution of the fern flora of Ceylon. Symp. Soc. exper. Biol. 7: 174-185.

— 1955 [‘1954’]. Cytological notes on one hundred species of Malayan ferns. Jn R. E. Holttum, A revised
flora of Malaya 2: 623-628. Singapore.

Manton, I., Ghatak, J. & Sinha, B. M. B. 1967. Cytotaxonomic studies in the Adiantum caudatum complex
of Africa and Asia I. Parentage of A. indicum Ghatak. J. Linn. Soc. (Bot.) 60: 223-235.

Manton, I., Roy, S. K. & Jarrett, F. M. 1966. The cytotaxonomy of some members of the Cheilanthes
farinosa complex in Africa and India. Kew Bull. 18: 553-565.

Manton, I., Sinha, B. M. B., & Vida, G. 1970. Cytotaxonomic studies in the Adiantum caudatum complex
of Africa & Asia II. Autoploidy and alloploidy in African representatives of A. incisum. J. Linn. Soc.
(Bot.) 63: 1-21.

Manton, I. & Sledge, W. A. 1954. Observations on the cytology and taxonomy of the pteridophyte flora of
Ceylon. Phil. Trans. R. Soc. B, 238: 127-185.

Marquand, C. V. B. 1929. The botanical collections made by Capt. Kingdon Ward in the eastern
Himalayas and Tibet in 1924~25. J. Linn. Soc. (Bot.) 48: 149-229.

Marten, J. 1909. List of ferns in and around Mussoorie. J. Bombay nat. Hist. Soc. 19: 179-183.

Mason, K. 1934. A note on the Nepal Himalayas. Himalayan J. 6: 81-90.

— 1955. Abode of snow: a history of Himalayan exploration and mountaineering. London.

Matthew, C. G. 1911. Enumeration of Chinese ferns. J. Linn. Soc. (Bot.) 39: 339-393.

Meher-Homji, V. M. 1972. Himalayan plants on south Indian hills: role of pleistocene glaciation versus
long distance dispersal. Sci. & Cult. 38: 8-12.

Meher-Homji, V. M. & Misra, K. C. 1973. Phytogeography of the Indian subcontinent. Jn R. Misra, B.
Gopal & K. P. & J. S. Singh (Eds), Progress of plant ecology in India: 9-88. New Delhi.

Mehra, P. N. 1932. Apogamy in some Indian ferns. Proc. Indian Sci. Congr. 19: 306-307.

1939. Ferns of Mussoorie. Publs Univ. Panjab 29: 1-29.

—— 1958. Taxonomy of pteridophytes. Mem. Indian Bot. Soc. 1: 30-36.

— 1961a. Cytological evolution in ferns with particular reference to Himalayan forms. Proc. Indian Sci.
Congr. 48 (2) : 130-153.

(Ed.) 19616. Chromosome numbers in Himalayan ferns. Res. Bull. Panjab. Univ. I, 12: 139-164.

Mehra, P. N. & Bir, S. S. 1964. Pteridophytic flora of Darjeeling and Sikkim Himalayas. Res. Bull. Panjab.
Univ. II, 15: 69-181.

Mehra, P. N. & Dhir, K. K. 1968. Ferns and fern allies of Dalhousie Hills. Bull. bot. Surv. India 10:
296-308.

Mehra, P. N. & Khullar, S. P. 1977. Biosystematics of Adiantum lunulatum Burm. complex in India with
special reference to W. Himalayan taxa. Cytologia 42: 501-511.

AN INTRODUCTION TO FERN GENERA OF THE INDIAN SUBCONTINENT 61

— 1980 [‘1974’]. Cytotaxonomy of W. Himalayan ferns. 2. Gleicheniaceous series. Res. Bull. Panjab
Univ. Il, 25: 136-178. [See also Khullar & Mehra]

Mettenius, G. 1856-1861. Uber einige farngattungen I-VI. I. Polypodium: 1-138 (1856). 2. Plagiogyra:
265-275 (1856-1858). 3. Pteris: 276-284 (1856-1858). 4. Phegopteris and Aspidium: 285-420 (1856-
1858). 5. Cheilanthes and 16. Asplenium: 47-99, 100-254 (1859-61).

Meusel, H. 1971. Mediterranean elements in the flora and vegetation of the west Himalayas. Jn P. H.
Davis, P. C. Harper and I. C. Hedge (Eds), Plant life of south-west Asia: 53-72.

Meuseul, H. & Schubert, R. 1971. Beitrage zur Pflanzengeographie des Westhimalajas. Flora, Jena 160:
137-194, 373-432, 573-606.

Milde, J. 1867. Filices Europae et Atlantidis, Asiae minoris et Sibiriae. Leipzig.

Moore, T. 1857-1862. Index filicum. London.

Morton, C. V. 1967, 1973. Studies of fern types. Contr. U.S. natn. Herb. 38 (2): 29-83 (1967), 38 (6):
215-281 (1973).

— 1974. William Roxburgh’s fern types. Contr. U.S. natn. Herb. 38: 283-396.

Nair, N. C. 1964. On a botanical tour to Lahaul and Spiti (Punjab Himalaya). Bull. bot. Surv. India 6:
219-235.

— 1968. Nomenclature of some Indian ferns. /ndian Forester 94: 169-170.

Nakai, T. 1952. A synoptical sketch of Korean flora, or, the vascular plants indigenous to Korea, arranged
in a new natural order. Bull. natn. Sci. Mus. Tokyo. 31: 1-152.

Nakaike, T. 1975. Enumeratio pteridophytarum japonicarum. Filicales. Tokyo.

— 1982. An enumeration of the ferns of Nepal I. Polystichum Roth. In Y. Otani. Reports on the crypto-
gamic study in Nepal: 135-158. Tokyo.

Nasir, E. & Ali, S. I. (Eds) 1970. Flora of West Pakistan 1.

Nayar, B. K. 1961a. Ferns of India No. I. Adiantum L. Bull. natn. bot. Gdns Lucknow 82: 1-42.

— 1961b. Ferns of India No. II. Drynaria and Pseudodrynaria. Bull. natn. bot. Gdns Lucknow 56: 1-30.

— 196lc. Ferns of India No. III. Microsorium. Bull. natn. bot. Gdns Lucknow 58: 1-38.

— 1962a. Ferns of India No. V. Hemionitis. Bull. natn. bot. Gdns Lucknow 67: 1-14.

— 1962b. Ferns of India No. VI. Cheilanthes. Bull. natn. bot. Gdns Lucknow 68: 1-36.

—— 1963. Ferns of India No. VII. Actiniopteris. Bull. natn. bot. Gdns Lucknow 75: 1-14.

— 1964a. Ferns of India— XIV. Lemmaphyllum. Bull. natn. bot. Gdns Lucknow 106: 1-15.

—— 1964b. Palynology of modern pteridophytes. Jn P. K. K. Nair (Ed.), Recent advances in palynology:
101-141. Lucknow.

— 1970. A phylogenetic classification of the homosporous ferns. Taxon 19: 229-236.

Nayar, B. K. & Chandra, S. 1965. Ferns of India— XV. Pyrrosia Mirbel. Bull. natn. bot. Gdns Lucknow
117: 1-98.

Nayar, B. K. & Kaur, S. 1963a. Ferns of India— VIII. Microlepia Presl. Bull. natn. bot. Gdns Lucknow 79:
1-25.

— 1963b. Ferns of India— IX. Peranema and Acrophorus. Bull. natn. bot. Gdns Lucknow 81: 1-40.

— 1964a. Ferns of India— XI. Bolbitis. Bull. natn. bot. Gdns Lucknow 88: 1-74.

— 1964b. Ferns of India — XII. Some new taxa. Bull. natn. bot. Gdns Lucknow 59: 141-154.

— 1964c. Ferns of India — XIII. Egonolfia. Bull. natn. bot. Gdns Lucknow 100: 1-38.

— 1966. On the fern genera Acrophorus, Diacalpe, Lithostegia and Peranema. Bot. Notiser 119: 1-23.

— 1972. Companion to R. H. Beddome’s Handbook to the ferns of British India. New Delhi. 1974. 2nd
ed. New Delhi.

Nayar, B. K. & Kazmi, F. 1962. Ferns of India No. IV. Plagiogyria. Bull. natn. bot. Gdns Lucknow 64:
1-36.

— 1963. Ferns of India — X. Matteuccia. Bull. natn. bot. Gdns Lucknow 82: 1-18.

Nishida, M. 1966. Ferns of Nepal. J. Coll. Arts Sci. Chiba Univ. Nat. Sci Ser. 4: 571-581.

Numata, M. 1966. Vegetation and conservation in eastern Nepal. J. Coll. Arts Sci. Chiba Univ. Nat. Sci.
Ser. 4 (4): 1-18.

1967-1968. Notes on a botanical trip in eastern Nepal. J. Coll. Arts Sci. Chiba Univ. Nat. Sci. Ser. 5
(1): 57-74 (1967), 5 (2): 243-258 (1968).

Ogata, M. 1928-40. Icones filicum Japoniae. 1-8. Tokyo. 1981. 2nd ed. (rev. H. It6) 1-3. Tokyo.

Ohwi, J. 1957. Flora of Japan. Pteridophyta: 1-244. Tokyo.

Ornduff, R. (Ed.) 1967. Index to plant chromosome numbers for 1965. Regnum veg. 50: 1-128.

—— (Ed.) 1968. Index to plant chromosome numbers for 1966. Regnum veg. 55: 1-126.

— (Ed.) 1969. Index to plant chromosome numbers for 1967. Regnum veg. 59: 1-129.

Ovchinnikov, P. N. (Ed.) 1957. Flora Tadzhikskoi S.S.R. 1: 1-44. Moscow.

Pande, B. D. 1962. Some aspects of the vegetation of Nepal. Bull. bot. Surv. India 4: 137-140.

62 CHRISTOPHER R. FRASER-JENKINS

—— 1973. Pteridophytic flora of Ranikhet. Indian Forester 99: 49-52.

Panigrahi, G. 1958. A note on ecogeographical and cytological studies in the Pteridophyta. Mem. Indian
Bot. Soc. 1: 127-139.

— 1960. Pteridophytes of the eastern India. Enumeration of the species collected and their nomencla-
ture. Bull. bot. Surv. India 2: 309-316.

— 1962. Cytology and its contribution to fern systematics. Jn P. Maheshwari etal. (Eds). Proc. Summer
School in Botany, Darjeeling: 261-275. New Delhi.

— 1968. Indian ferns. Indian Mus. Bull. 3: 139-144.

— 1969. Some recent advances in the studies of Indian pteridophytes. Proc. seminar morphology,
anatomy and embryology land plants, Univ. Delhi: 13-14.

— 1975. Taxonomic notes on certain taxa of Asiatic ferns. Phytologia 31: 251-258.

Panigrahi, G., Chowdury, S., Raju, D. C. S. & Deha, G. K. 1964. A contribution to the botany of Orissa.
Bull. bot. Surv. India 6: 237-266.

Panigrahi, G. & Dixit, R. D. 1966. New records of ferns for Madhya Pradesh. Proc. natn. Acad. Sci. India
36 (2): 135-144.

Panigrahi, G. & Naik, V. N. 1961. A botanical tour to Subansiri frontier division (NEFA). Bull. bot. Surv.
India 3: 361-388.

Panigrahi, G., Singh, A. N. & Misra, O. P. 1969. Contribution to the botany of the Tarai forests of the
Bahraich district of Uttar Pradesh. Bull. bot. Surv. India 11: 89-114.

Parandekar, S. A. 1966. Some ferns from Kolhapur (Maharashtra State). Maharashtra Vidnyan Mandir
Patrika 1: 56-63.

Parris, B. S., Jarrett, F. M. et al. Index filicum supplementum, 1964-1975. [in preparation].

Paulsen, O. 1920. The second Danish Pamir Expedition. Studies in the vegetation of Pamir. Arb. bot. Have
Kgbenhaven 91: 1-132.

Peking Bot. Inst. (Palaeobot.) 1976. Sporae pteridophytarum sinicarum. Peking.

Pichi-Sermolli, R. E. G. 1965. Index filicum, supplementum quartum. Regnum veg. 37.

—— 1970. A provisional catalogue of the family names of living pteridophytes. Webbia 25: 219-297.

—— 1977. Tentamen pteridophytorum genera in taxonomicum ordinem redigendi. Webbia 31: 313-512.

Polunin, O. & Stainton, J. D. A. Flowers of the Himalaya. London. [In press]

Prain, D. 1903. Bengal plants 2: 1257-1270. Calcutta.

Presl, C. B. 1836. Tentamen pteridographiae seu genera filicacearum. Praha. 1845. Supplementum
Tentaminis pteridographiae. Praha.

— 1851. Epimeliae botanicae. Praha.

Raizada, M. B. & Vaid, K. M. 1952. Ferns of Nepal. Indian Forester 78: 576-581.

Rao, A. S. & Hajra, P. K. 1980. Fern allies and ferns of Kameng district. Indian Forester 106: 327-349.

Rao, R. S. 1963. A botanical tour in the Sikkim State, eastern Himalayas. Bull. bot. Surv. India 5: 165-205.

Rao, T. A. 1959. A botanical tour to Milam Glacier. Bull. bot. Surv. India 1: 97-120.

1960. A botanical tour to Pindari Glacier and Kumaon hill stations. Bull. bot. Surv. India 2: 47-57.

Rau, M. A. 1960. On a collection of plants from Lahaul. Rec. bot. Surv. India I, 2: 45-S6.

1961. Flowering plants and ferns of north Garhwal, Uttar Pradesh. Bull. bot. Surv. India 3: 215-251.

Rosenstock, E. 1915. Filices Formosanae novae a cl. Pe. U. Faurie anno 1914 collectae. Hedwigia 56:
333-348.

Roxburgh See Griffith.

Roy, R. P. & Pandey, S. N. 1962. Cytotaxonomical studies of the fern-flora of Parasnath Hills. Proc. Indian
Sci. Congr. 49: 333-334.

Roy, R. P., Sinha, B. M. B. & Sakya, A. R. 1971. Cytology of some ferns of Kathmandu valley. Br. Fern
Gaz. 10: 193-199.

Roy, S. K. & Holttum, R. E. 1965. Cytological observations on ferns from south China. Am. Fern J. 55:
154-158.

Roy, S. K. & Singh, J. B. 1973. Cytological and ecological notes on the ferns of Pachmarhi. PI. Sci. 5: 1-6.

Schelpe, E. A. C. L. E. 1954. Ecological observations on Pteridophyta in the Kangra Himalayas. Am. Fern.
J. 44: 49-65.

Schott, H. 1834-1836. Genera filicum: Nos 1-20. Vindobonae.

Schweinfurth, U. 1957. Die horizontale und vertikale Verbreitung der Vegetation in Himalaya. Bonn.

Scott, J. 1868. A list of the higher cryptogams cultivated in the Royal Botanical Gardens, Calcutta. Proc.
agric. hort. Soc. India II, 1: 200-250.

Serizawa, S. 1970-1975. Taxonomical notes on Asian ferns. J. Jap. Bot. 45: 117-122, 262-266 (1970), 46:
16-22, 278-282 (1971), 50: 15—20 (1975).

— 1974. The leaf-architecture of the dryopteroid ferns. J. Jap. Bot. 49: 273-284.

AN INTRODUCTION TO FERN GENERA OF THE INDIAN SUBCONTINENT 63

— 1976. A revision of the dryopteroid ferns in Japan and adjacent regions. Sci. Rep. Tokyo Kyoiku
Daigaku B, 16: 109-148, plates I-VI.

Sim, R. 1859. A priced catalogue with brief descriptions and cultural remarks etc. of ferns grown for sale by
Robert Sim, nurseryman 6. London. 1866. 7. London.

Sinha, B. M. B. & Manton, I. 1970. Cytotaxonomic studies in the Adiantum caudatum complex of Africa
and Asia III. Segmental allopolyploid origin of A. malesianum Ghatak. Bot. J. Linn. Soc. 63: 247-264.

Sledge, W. A. 1960. The Polypodiaceae and Grammitidaceae of Ceylon. Bull. Br. Mus. nat. Hist. (Bot.) 2:
133-158.

— 1962. The athyrioid ferns of Ceylon. Bull. Br. Mus. nat. Hist. (Bot.) 2: 277-323.

— 1965. The Ceylon species of Asplenium. Bull. Br. Mus. nat. Hist. (Bot.) 3: 235-277.

— 1967. The genus Elaphoglossum in the Indian peninsula and Ceylon. Bull. Br. Mus. nat. Hist. (Bot.)
4: 79-96.

— 1968. The Hymenophyllaceae of Ceylon. J. Linn. Soc. (Bot.) 60: 289-308.

— 1972. The tectarioid ferns of Ceylon. Kew Bull. 27: 407-424.

— 1973a. The dryopteroid ferns of Ceylon. Bull. Br. Mus. nat. Hist. (Bot.) 5: 1, 3-43.

— 1973b. Generic and family boundaries in the Aspidiaceae and Athyriaceae. In A. C. Jermy, J. A.
Crabbe & B. A. Thomas (Eds), The phylogeny and classification of the ferns. J. Linn. Soc. (Bot.) 67
(Suppl. 1): 203-210.

— 1973c. Native and naturalised species of Adiantum in Ceylon. Ceylon J. Sci. (Bio. Sci.) 10: 144-154.

— 1981. The Thelypteridaceae of Ceylon. Bull. Br. Mus. nat. Hist. (Bot.) 8: 1-54.

— 1982. An annotated check-list of the Pteridophyta of Ceylon. Bot. J. Linn. Soc. 84: 1-30.

Smith, A. R. & Fraser-Jenkins, C. R. 1982. Dryopteris paleacea is asynonym of D. wallichiana. Taxon 31:
326-329.

Smith, J. 1841-43. An arrangement and definition of the genera of ferns, with observations on the affinities
of each genus. J. Bot. 4: 38-70, 147-198 (1841). Lond. J. Bot. 1: 419-438, 659-668 (1842), 2: 378-394
(1943).

— 1866. Ferns, British and foreign. London. 1877, 2nd ed. 1896, 3rd ed.

—— 1875. Historia filicum. London.

Smith, W. W. 1913. The alpine and subalpine vegetation of south-east Sikkim. Rec. bot. Surv. India 4:
323-431.

Smith, W. W. & Cave, G. H. 1911. The vegetation of the Zemu & Lhonakh valleys of Sikkim. Rec. bot.
Surv. India 4: 141-260.

Solbrig, A. & Solbrig, O. 1964-1965. IOPB chromosome number reports. Taxon 13: 99-110, 201-209
(1964), 14: 50-57, 86-92 (1965). [See also Love]

Sprengel, K. P. J. 1827. Caroli Linnaei. Systema vegetabilium 16th ed. 4: 8-134. Gottingae.

Srinivasan, S. K. 1959. Report on a botanical tour to Bomdi-la, N.E.F.A. (May, 1955). Rec. bot. Surv.
India 17 (2): 1-38.

Stainton, J. D. A. 1972. Forests of Nepal. London.

Stearn, W. T. 1960. Allium and Milula in the central and eastern Himalaya. Bull. Br. Mus. nat. Hist. (Bot.)
2: 161-191.

—— 1973. Botanical Latin. 2nd ed. Newton Abbot.

— 1976. Frank Ludlow (1885-1972) and the Ludlow-Sherriff expeditions to Bhutan and south-eastern
Tibet of 1933-1950. Bull. Br. Mus. nat. Hist. (Bot.) 5: 243-268.

van Steenis, C. G. G. J. 1934-1936. The origin of the Malaysian mountain flora. Bull. Jard. Bot. Buitenzorg
III, 13: 135-262, 287-417 (1934), 14: 56-72 (1936).

— 1962. The mountain flora of the Malaysian tropics. Endeavour 1962: 173-189.

Stewart, R. R. 1916-1917. The flora of Ladakh, western Tibet. Bull. Torrey bot. Club 43: 571-590 (1916),
625-650 (1917). (Reprinted in Delhi 1973)

— 1942. The ferns of Mussoorie and Dehra Dun. 150th Anniv. Vol. Roy. Bot. Gard. Calcutta 2: 159-172.
Alipore.

—— 1945. The ferns of Kashmir. Bull. Torrey bot. Club 72: 399-426.

— 1951. The ferns of Pahlgam, Kashmir J. Indian bot. Soc. 30: 137-142.

—— 1952. Ferns of Rawlpindi district. Forest Rec. 2: 88-90.

— 1957. The ferns and fern allies of West Pakistan and Kashmir. Biologia 3 (2): 1-32.

— 1972. Pteridophyta. Jn E. Nasir & S. I. Ali (Eds), Flora of West Pakistan: 1-21.

— 1982. Missionaries and clergymen as botanists in India and Pakistan. Taxon 31: 57-64.

Strachey, R. 1882. Plants found in Kumaon, Garhwal and the adjoining parts of Tibet by Captain (now
General) Richard Strachey and Mr. Winterbottom. Jn E. [F.] T. Atkinson, Gazetteer, N.W. Provinces
10. The Himalayan districts: 642-654. Allahabad.

64 CHRISTOPHER R. FRASER-JENKINS

Strachey, R. & Duthie, J. F. 1906. Catalogue of the plants of Kumaon and of the adjacent portions of
Garhwal and Tibet, based on the collections made by Strachey and Winterbottom, 1846-9. London.

Sutton, S. 1978. Plant collectors in Nepal. Jn H. Hara, W. T. Stearn & L. H. J. Williams, An enumeration of
the flowering plants of Nepal 1: 13-21.

Swartz, O. 1801. Genera et species filicum. J. Bot. Géttingen 2 (1): 3-136.

— 1806. Synopsis filicum. Kiliae.

Tagawa, M. 1932-1939. Spicilegium pteridographiae asiae orientalis 1-20. Acta Phytotax. Geobot. 1:
88-94, 156-163, 306-313 (1932), 2: 14-24 (1932), 2: 189-205 (1933), 3: 28-37, 88-96 (1934), 4: 132-148,
202-206 (1935), 5: 101-115, 189-197, 250-262 (1936), 6: 89-100, 154-168 (1937), 7: 72-87. 184-191
(1938), 8: 20-31, 91-100, 164-176, 229-236 (1939).

— 1936-1938. Miscellaneous notes on the east Asiatic pteridophytes with special reference to the
Japanese species 1-7. J. Jap. Bot. 12: 486-495, 537-544, 746-755 (1936), 13: 180-190 (1937), 14:
101-112, 595-605, 705-712 (1938).

— 1940-1949. Studies on Formosan ferns 1—7. Acta phytotax. geobot. Kyoto 9: 87-96, 139-148, 203-215
(1940), 10: 193-208, 275-289 (1941), 11: 303-311 (1942), 14: 8-10 (1949).

— 1949-1961. Fern miscellany 1-11. J. Jap. Bot. 22: 160-165 (1948), 23: 76-79 (1949), 25: 113-116
(1950), 26: 19-22, 185-188 (1951), 27: 213-218, 319-324 (1952), 30: 279-282 (1955), 33: 92-96, 199-204
(1958), 36: 205-212 (1961).

— 1950-1956. Occasional notes on Asiatic Pteridophyta 1-7. Acta phytotax. geobot. Kyoto 14: 44-48
(1950), 14: 91-95 (1951), 14: 191-194 (1952), 15: 14-16 (1953), 16: 48-51 (1955), 16: 71-78, 174-178
(1956).

— 1963. Coloured illustrations of the Japanese Pteridophyta. Osaka.

— 1967-1978. New or interesting ferns from Thailand 1-10 Acta phytotax. geobot. Kyoto 22: 97-103
(1967), 23: 48-56, 110-116 (1968), 23: 175-178 (1969), 24: 60-64 (1969), 24: 175-181 (1970), 24: 16-21
(1971), 26: 58-61 (1974), 26: 169-172 (1975), 29: 22-24 (1978).

Tagawa, M. & Iwatsuki, K. 1967. Enumeration of Thai pteridophytes collected during 1965-1966. South
East Asian Studies 5: 23-120.

Tardieu-Blot, M. L. 1941. Sur la repartition et les affinités des Fougéres d’Indochine. Bull. Soc. bot. France
88: 825-834.

Tardieu-Blot, M. L. & Christensen, C. 1934-1938. Les fougéres d’Indochine. A. Bull. Mus. natn. Hist. nat.
Paris I, 6: 287-290 (1934). B. Bull. Mus. natn. Hist. nat. Paris. 11, 6: 383-386 (1934). V. Phanérogamie
5: 165-173 (1936). VI. Phanérogamie 6: 1-11 (1937). VU, IX, X. Phanérogamie 6: 129-134 (1937). XII.
Phanérogamie 6: 161-176 (1938). XIII (‘XI’). Phanérogamie 7: 10-15 (1938). XIV. Phanérogamie 7:
65-104 (1938). [See also Christensen & Tardieu-Blot 1934-39]

—— 1939-1941. Fougéres. In H. Lecomte, Flore générale de Il’ Indo-Chine 7 (2): crypt. vasc., fougéres, 7
(6): 1-144 (1939), 7 (7): 145-288 (1940), 7 (8): 289-432 (1941), 7 (9): 433-544 (1941).

Tewari, R. B. 1973. Pteridophytic flora of Lucknow district (U.P.). Bull. bot. Surv. India 15: 129-130.

Thunberg, C. P. 1825. Florula Ceilanica: 10. Uppsaliae.

Thwaites, G. H. K. 1864. Enumeratio plantarum zeylandiae: 377-397, 437-439. London.

Trimen, H. 1885. Systematic catalogue of the flowering plants and ferns of Ceylon: 111-119. Colombo.

1887. Hermann’s Ceylon herbarium and Linnaeus’ Flora Zeylanica. J. Linn. Soc. (Bot.) 24: 129-155.

Trotter, E. W. 1889. The ferns of the Punjab popularly described. Lahore.

Tsai, J. L. 1972-73. Chromosome numbers of some Formosan ferns. J. Sci. Eng. 9: 123-132 (1972), 10:
261-275 (1973).

Tsai, J. L. & Shieh, W. C. 1975. Chromosome numbers of the fern family Aspidiaceae (sensu Copeland) in
Taiwan. J. Sci. Eng. 11: 321-324 (1975), 14: 91-104 (1977).

Vaurie, C. 1972. Tibet and its birds: 1-160, 345-380. London.

Vegter, I. H. 1976. Index herbariorum II (4). Collectors (M). Regnum veg. 93: 474-576.

Verma, S. C. & Khullar, S. P. 1976a. Adaptive significance of simultaneous cytokinesis in pteridophytes.
Phytomorphology 26: 96-102.

—— 1976b. Some considerations of spore morphology and evolutionary biology in ferns. JV Int.
Palynological Conf., Lucknow 1976-7. Late abstracts: 76. 1978. Abstr. IV Int. Palynol. Conf., Lucknow
1976-77. 1: 143-151.

—— 1978. Spore biology of eusporangiate ferns: an essay. Advances in pollen-spore research 4: 53-73.

—— 1980. Ferns of Nainital (western Himalayas): an updated list. Fern Gaz. 12: 83-92.

Verma, S. C. & Loyal, D. S. 1960. Chromosome counts in some ferns from Nainital. Curr. Sci. 29: 69-70.

Vida, G. 1974. Genome analysis of the European Cystopteris fragilis complex. I. Tetraploid taxa. Acta bot.
hung. 20: 181-192.

Voss, E. G. et al. 1983. International code of botanical nomenclature. Regnum. veg. 111.

AN INTRODUCTION TO FERN GENERA OF THE INDIAN SUBCONTINENT 65

Wall, G. 1873. Catalogue of the ferns indigenous to Ceylon. London.

Wallich, N. 1828-1849. A numerical list of dried specimens of plants in the East India Company’s museum,
collected under the supervision of Dr. Wallich: Nos 1—2159 (1828), 2160-4361 (1829), 4362-7683 (1830),
7684-9148 (1847-1849).

— 1830. Plantae asiaticae rariores 1: 42 (1830). London.

Wendelbo, P. 1964 [‘1963’] Pteridophyta. Jn I. C. Hedge & P. Wendelbo. Studies in the flora of
Afghanistan 1. Arbok Univ. Bergen Mat.-Naturv. Ser. 18: 31-32.

Willis, J. C. 1911. Catalogue of the flowering plants and ferns of Ceylon: 114-124. Colombo.

Wormald, G. 1976. An annotated checklist of the plants at Lumle Agricultural Centre. Lumle (Nepal).

Zimmermann, A. 1956. Resultats des expeditions scientifiques, Genevoises au Népal en 1952 et 1954.
Partie botanique. Candollea 15: 127-147.

Geography

Floristic considerations

The three genera Athyrium, Dryopteris, and Polystichum are all predominantly temperate,
probably with their main evolutionary centres in south-west China where by far the maximum
number of species occur today. Dryopteris and Polystichum have some apparent secondary
areas of evolution in Europe (based on the Caucasus and Atlantic regions), East Africa, North
America, the Far East (Japan and Taiwan) and south-east Asia. In Dryopteris only a few of its
approximately 215 species, mainly the south-east Asian element, are subtropical or tropical in
ecology and distribution. Polystichum however has a far larger number of lowland tropical or
subtropical species occurring in south-east Asia, Australasia (where Dryopteris is restricted to a
very few species) Africa, and South America, in addition to the large number of species
occurring in all the temperate regions of the world. It is therefore a more cosmopolitan genus. It
is worth mentioning here that the genus Arachniodes, with apparent relationships to Poly-
stichum, is predominantly sub-tropical. Athyrium by contrast has only two representatives in the
European type flora, several more in North America and Africa, and is absent from South
America and Australia. Its main centre of distribution is in eastern Asia, particularly in west
China and Taiwan, and in Japan, with a number of species occurring in south-east Asia. Related
genera such as Lunathyrium, Dryoathyrium, etc., are generally similarly distributed.

These genera in the Indian subcontinent contain species whose distribution patterns show that
they represent a number of different floristic or geographical elements within the flora. But the
ecology of all the species makes them temperate or subtemperate mesophytic plants, so that they
only show more or less temperate mesophytic distribution types, and do not show any of the
rather more common patterns, including those of tropical plants or xeromorphs, etc., that are
frequent in the Indian subcontinent. The bulk of species represent the deciduous or coniferous
forest elements of the extra-tropical humid ecological type of Meusel (1971), and occur or
frequently have close relations further east in south-west China, etc., and east to Taiwan. The
greatest number of species occurs in the highly oceanic eastern parts of the Himalaya, where the
summer monsoon (June—September) from the Bay of Bengal and Indian Ocean has its greatest
effect, and there are also fairly frequent winter rains. The mountains of southern India and Sri
Lanka, affected strongly by both a summer and winter monsoon, also contain some of these
elements, though they have a relatively greater proportion of species with south-east Asian
affinities. Other elements are also present in the Himalaya, some are probably sub-types which
have specialised from the extra-tropical humid type, but there is also a small temperate
European element in the drier west Himalaya. Most of these distribution types correspond with
one of the ten different patterns described by Stearn (1960). The three genera concerned are
absent from the low-lying parts of the plains of India, Pakistan, and Bangladesh.

The different phytogeographical types may be summarised as follows:

1. Sino—Himalayan species
These are species found in south and south-west China, south-east Tibet and the forest zones of
the Himalaya. Outside this area they often occur somewhat disjunctly in Taiwan (though several

66 CHRISTOPHER R. FRASER-JENKINS

species have in the past been thought not to be present across south China, but are only now
known to occur on mountains in scattered localities bridging the gap between Yunnan and
Taiwan) and sometimes as far east as the Philippines, and occasionally Japan too (see Kanai,
1963). Some species may also have more distant and ancient European or East African
relationships — for example, Athyrium attenuatum (C. B. Clarke) Tag. to A. filix-femina (L.)
Roth of Europe (See Fraser-Jenkins & Khullar (1982)); Polystichum piceo-paleaceum Tag. and
other species of the P. yunnanense Christ and P. makinoi (Tag.) Tag. agg. to P. fusco-paleaceum
Alston of East Africa and to P. setiferum (Forsskal) Woynar of Europe; P. luctuosum (Kunze)
T. Moore present in the west Himalaya and East and South Africa (and the almost certainly
synonymous P. tsus-simense (Hook.) J. Smith in China and the Far East). Within this major
group certain more-or-less distinct geographical sub-types occur, though these are really based
on a difference of degree of ecological tolerance and adaptations concerning oceanicity,
altitude, drought, etc., so that intermediates may occur. It should be noted though that there is
also a marked gradation effect depending on depth into the Himalayan range (i.e. outer versus
inner ranges), which markedly affects the rainfall and thus their distribution (as discussed by
Schweinfurth (1957) and later authors). These sub-types are:

(i) Widespread Sino—Himalayan species:

Some species in this group may not be common to both west China and the Indian side of the
Himalaya, but most of the widespread species are, and there is a particularly marked similarity
between the fern flora of most of Yunnan and south-west China and that of the eastern part of
the central Himalaya and of the east Himalaya. The widespread species have somewhat oceanic
requirements and occur in the high rainfall areas maintained by the summer monsoon. In the
Indian subcontinent they are common in the central and east Himalaya, and are also quite
frequent in Kumaon in the eastern part of the west Himalaya, though here often only in the first
major ranges of hills north from the plains (i.e. the outermost ranges), which receive the highest
rainfall and contain the thickest forest, providing they are high enough for the species
concerned. From here they often have scattered extensions westwards in the outermost ranges
at mid-levels, commonly as far as Chamba, but even, via the Pir Panjal Range, up to Kashmir,
the eastern part of Pakistan, or west of the Indus. Examples include Athyrium attenuatum (C. B.
Clarke) Tag. (west China to Yunnan and throughout the Indo—Himalaya to Afghanistan),
Polystichum squarrosum (D. Don) Fée (south-east Tibet and Assam, west to Pakistan east of
the Indus; also in south India), P. obliquum (D. Don) T. Moore (south-west China, west to
Chamba), and Dryopteris stenolepis (Baker) Christensen (south mainland China, Taiwan,
south-east Tibet, North Vietnam, etc., and west to Tehri Garhwal in the eastern part of the west
Himalaya).

(ii) East Himalayan west Chinese species:

These are markedly oceanic in requirements and may be defined as not occurring west of the
central Himalaya (i.e. Nepal) and generally only as far west as central Nepal. With more oceanic
genera, such as Rhododendron, the dividing lines between west and east Himalayan species are
slightly further east. Their westward range may be limited by the length and intensity of the dry
period between the monsoons; the monsoon beginning earlier and ending later in the east
Himalaya, and pre-monsoon spring rains being more frequent there (see Stainton (1972)).
Examples include Dryopteris marginata (Wallich ex Clarke) Christ (Taiwan and from mainland
south China to the Kathmandu valley in central Nepal; western records refer to D. carolo-hopei
Fraser-Jenkins) and Polystichum acutidens Christ (south and south-west China and west in the
east Himalaya to the Sikkim area).

(iii) West Himalayan species:

These are often endemic to the west Himalaya and have presumably become adapted to the
rather drier conditions present in the forests of the west Himalaya, seldom occurring east of west
Nepal and uncommon this far east. Several of them cross over and occur behind the Himalayan
line in places in south and south-east Tibet (north of east Nepal, Sikkim, Bhutan, etc.) where
some of the valleys pass through the main range allowing somewhat greater rainfall unlike the

AN INTRODUCTION TO FERN GENERA OF THE INDIAN SUBCONTINENT 67

situation in the central and west Himalaya, so that these west Himalayan elements may continue
eastwards as scattered populations in south and south-east Tibet and reach a second area of
distribution in south-west China. This is presumably due to the medium dry conditions in a few
scattered areas in a narrow belt behind the Himalayan line being similar to those in the west
Himalaya; a few such areas also occur in Bhutan (Stainton, 1972) and may also be expected to
contain some of these species. A similar situation has also been noticed in some species of the
genus Allium by Stearn (1960), and Stainton (1972) discusses this phenomenon in some detail,
with other examples from several genera. Some of the species may have close relations or
apparent bicentric occurrences east of the Himalaya, probably due to extinction of the linking
populations either by excessive dryness, or altitude in Tibet, or excessive oceanicity in the east
Himalaya. The close juxtaposition of these two limiting factors in the central and east Himalaya
appears to have been the cause of their obliteration in the area between the two zones of
occurrence. The links between the two populations probably existed at the time when the
Tibetan flora had not been mainly extinguished due to rain shadow caused by the tertiary uplift
of the Himalayan ranges (see Ching & Wu, 1980); however, there is as yet little definite evidence
that the Himalayan uplift was actually the cause of the drying up of Tibet, which could in fact be
part of a general change in atmospheric patterns which has led to world-wide desert formation.
As far as Dryopteris, Polystichum, and Athyrium are concerned, the west Himalayan species
are, from their relationships to other species, part of the Sino—Himalayan element even when
they do not occur in China. Examples include Dryopteris ramosa (Hope) Christensen (Afgha-
nistan to west Nepal; perhaps distantly related to the north and east Chinese, Siberian, and
Japanese D. goeringiana (Kunze) Koidz, and the Japanese D. shiroumensis Kurata & Nakaike),
D. blanfordii (Hope) Christensen (Afghanistan to Pithoragarh in the eastern part of the west
Himalaya, but apparently with a second and disjunct centre in south-east Tibet and west China,
as subsp. nigrosquamosa (Ching) Fraser-Jenkins, in similar ecological conditions), Polystichum
luctuosum (Kunze) T. Moore (South and East Africa, the Himalaya from Pakistan west of the
Indus to west Nepal, with the probably synonymous P. tsus-simense (Hook.) J. Smith occurring
again in south-east Tibet, south, central, and east China (including Taiwan), and Japan).

(iv) Tibetan species:

These are high altitude plants adapted to extremes of drought and cold, occurring in Tibet and
the higher, inner ranges of the Himalaya, often north of (behind) the Himalayan line, which is
the line beyond which the forests do not occur on the approaches to Tibet due to rain-shadow
(see Stainton, 1972). Examples include Polystichum glaciale Christ (west Bhutan, Tibet, and
high mountains in Yunnan and Szechuan) and Athyrium davidii (Franchet) Christ (= A. duthiei
Bedd.) (north Chamoli to Bhutan, south Tibet, Yunnan, and Szechuan).

2. European species

These have been referred to by some authors (e.g. Clarke, 1898; Fedtchenko, 1903-1947; Mani,
1978) as part of the central-Asian or middle-Asian component, occurring in the Pamirs, Tien
Shan, etc. Undoubtedly this is partly correct but their ‘central-Asian relationships’ are also, in
part, due to a continuing tendency for Russian botanists, somewhat isolated from elsewhere, to
think in local terms and refer to species as endemics when they are not. Because ferns are
mesophytes, almost the only species that have reached the Himalaya and those parts of central
Asia from anywhere other than the east and south, are a few widespread European species with
a greater measure of drought tolerance and robustness than usual, thus enabling them to cross
the desert gap of Iran, Afghanistan, and Kazakhstan, presumably before it became quite as dry
as it is at present. There is only a very small Soviet—central-Asian fern component, and it has not
penetrated the Himalaya and is probably of comparatively recent tertiary European origin.
Hence as far as ferns are concerned, the ‘middle-Asian-component’ may be more correctly
referred to as the European element. Almost all the European species in the Himalaya are
confined to the west Himalaya, being limited by the heat, or by the humid, subtropical, or
temperate, forest zone which allows excessive competition to these species further east, or by

68 CHRISTOPHER R. FRASER-JENKINS

the cold and harshness of the upper Tibetan zone. The west Himalaya are also affected by
considerably more winter rain or snowfall (and thus spring melt) than further east, somewhat
more like the European region (see Stainton, 1972). In the west Himalaya the European species
occupy either a Mediterranean-type belt at lower levels (e.g. in Cheilanthes), or a mid- to
high-level belt between the Tibetan and forest zones (e.g. in Asplenium and the examples
below). Examples include the only two European species in the genera Polystichum, Dryopteris,
and Athyrium: Polystichum lonchitis (L.) Roth (Afghanistan to east Kashmir, also the Tien
Shan, Altai, western Asia, Europe, Siberia, Far-East, Japan, and North America) and
Dryopteris filix-mas (L.) Schott (Afghanistan to south Baltistan, also the Pamirs, Kazakhstan,
Tien-Shan, Altai, west Siberia, Dzhungaria, western Asia, Europe, north-west Africa, North
and South America).

3. South-east Asian species

These are an important element in the fern-flora of Sri Lanka and southern India, extending up
the mountain ranges in places in central India to the Himalaya, where they behave like
widespread Sino—Himalayan species in the forest zone, but at rather low levels. They are present
in Malaysia, Thailand, Indo—China, the islands of south-east Asia, the Philippines, northern
Australia, and often southern Japan. The species present in the Indian subcontinent appear to
have a wide ecological tolerance and therefore cover a wide area. Some of the Sri Lankan and
south Indian species may not be south-east Asian species, but endemics, with their closest
relations being south-east Asian elements. Dryopteris deparioides (T. Moore) Kunze and other
species allied to D. sparsa (Buch.-Ham. ex D. Don) Kunze are examples. True south-east Asian
species include Polystichum subapiciflorum Hayata from central Nepal east to north Assam,
Khasia, Burma, Thailand, North Vietnam, south China including Taiwan, which is allied to P.
biaristatum (Blume) T. Moore of south-east Asia and Sri Lanka, Dryopteris sparsa Buch.-Ham.
ex D. Don) Kuntze (south-east Asia, northern Australia and Australian islands, south-east
Tibet, south China, Taiwan, the Philippines, south Japan, Sri Lanka, south India, mountains of
central India, Assam, the Himalaya from south-east Tibet and Assam west to the Simla region,
though rare west of Nepal), D. cochleata (Ham. ex D. Don) Christensen. (South-east Asia,
south-east Tibet, south China, the Philippines, south India, mountains of central India, the
Himalaya west to Chamba), D. hirtipes (Blume) Kuntze and subspecies (south-east Asia, Sri
Lanka, south India, the east Himalaya including south-east Tibet and Yunnan, south-east
China, Japan, the Philippines and New Hebrides), Athyrium nigripes (Blume) T. Moore
(south-east Asia, Sri Lanka, south India, south-east Tibet, Yunnan, the Himalaya west to Tehri
Garhwal), and Polystichum biaristatum (Blume) T. Moore (south-east Asia and Sri Lanka).

4. Cosmopolitan species

One species, Dryopteris wallichiana (Sprengel) N. Hylander, is probably worthy of a separate
phytogeographical category, though it may also be considered a sub-group of the Sino—
Himalayan species. Its origins appear to have been in west China and the east Himalaya, though
from its wide distribution and relationships elsewhere it must be ancient, with polyploid
apomictic pillar complexes built up from it in various parts of the world, including the D. affinis
group in Europe (see Fraser-Jenkins, 1980). D. wallichiana is distributed in subtropical regions,
in Central- and South-America, the West Indies, southern Atlantic Islands, south-east Africa,
Madagascar, the Indian subcontinent (the Himalaya, with a related species in south India and
Sri Lanka), south-east Asia (with related species), south-west, south, and south-east China
(including Taiwan), the Philippines, Japan, and Hawaii (with related species).

Geographical limits and major phytogeographic divisions within India

The area included in this study consists of east Afghanistan, Pakistan, India, Nepal, Sikkim
(India), Chumbi in south Tibet (China), Bhutan, Assam (India), Bangladesh, and Sri Lanka
(Ceylon), which forms the Indian subcontinent. The western limit to the Indian (part of the

AN INTRODUCTION TO FERN GENERA OF THE INDIAN SUBCONTINENT 69

Sino—Himalayan) fern flora is defined naturally by the line beyond which the influence of the
summer monsoon is no longer felt, which passes through south-western Pakistan and then
through Afghanistan approximately level with the high middle of the Hindu Kush and including
the wetter north-eastern region, Nuristan (previously known as Kafiristan), before running
northwards to the border of the U.S.S.R. on the Oxus (Amu Darya), south of Dushanbe.
Beyond the line, to the west, the harsh and dry desert climate of western Afghanistan,
Kazakhstan, and most of Iran forms a large natural gap between the Sino—Himalayan fern flora
and the European-type fern flora. The latter reaches its south-eastern limit in the eastern Elburz
mountains of north-eastern Iran shortly east of the Caspian Sea, along the southern shores of
which a narrow belt of forest extends eastward from Transcaucasia. The European flora also
passes east through the Russian to the Chinese Tien Shan (with some species present in the
Pamirs), west Siberia and the Altai to Irkutsk, where it gives way to the Far-Eastern fern flora.
Few species of ferns have been able to cross the gap from west to east into the Himalaya, or in the
other direction, apart from Dryopteris barbigera subsp. komarovii (Kossinsky) Fraser-Jenkins,
which reaches the Pamirs and Tien-Shan in the U.S.S.R. and China, and Polystichum wilsonii
Christ and P. lachenense (Hook.) Bedd. which reach as far west as the Chinese Tien Shan south
of Urumchi. In the early tertiary period, though, several species of flowering plants migrated
from the Indian subcontinent towards Europe (see Meusel, 1971), and probably one or two
ferns, such as the ancestors of Dryopteris wallichiana (Sprengel) N. Hylander, from which the
European D. affinis (Lowe) Fraser-Jenkins appears to have been derived (see Fraser-Jenkins,
1980). During the late tertiary the area between the Himalaya and the Caspian dried up almost
completely, including what was once a large shallow sea north of the Elburz and West Pamirs,
now represented only by the Caspian and Aral Seas. To the north-west of the Indian
subcontinent the great masses of the Pamirs and Karakoram have also formed a natural
boundary to the flora, reinforced by the desert areas of Kazakhstan and the Sinkiang to the north.
This barrier extends all along the north of the area, north of the Himalayan line, as the extensive
cold desert highlands of western and central Tibet, extending more or less as far east as the
region north of the Tsangpo, north of Sikkim on the western edge of the east Himalaya, so that the
western and most of the northern limits of the flora are clear and natural.

In the east Himalaya the Indian subcontinent flora joins up with that from the more equable
south-eastern Tibet and western-China, with few barriers. The Himalayan ranges are slightly
less high there and are penetrated by several deep valleys which allow them to be wet on the
Chinese and Burmese sides as well, where they come to an end in the east, the moisture probably
being due mainly to the far- reaching effects of the summer monsoon that turns east-north-
eastwards from the Bay of Bengal. The continuity of the east Himalayan and west Chinese flora
indicate the possible Chinese origin of many of the Sino—Himalayan ferns, the flora apparently
being much richer on the Chinese side. The rich flora of western Yunnan (and perhaps also north
Burma, though this area is too little known at present), constituting the centre of distribution of
many temperate genera, may well be due partly to the fact that the mountain ranges there
(constantly causing a moist climate at their edges) run north to south rather than east to west as
in the Himalaya proper. Thus as the climate cooled or warmed up they did not constitute a
barrier, but acted as a route for north and southward migration (and also allowed altitudinal
migration) of species, so that fewer were extinguished. A similar situation exists when
comparing the flora of Europe (with the horizontal Pyrenees, Alps, Carpathians, Balkans,
Pontic, Caucasian, and Elburz line) to that of North America (with its vertical Rockies, or
Appalachian lines). In several species their range extends eastwards well beyond the Indian
subcontinent to south China, including Taiwan and, to a lesser extent, to southern and central
Japan. The eastern and north-eastern boundary to this study is therefore artificial and is dictated
mostly by political and historical considerations, confining it to the Indian subcontinent and
excluding north Burma, China, and Tibet; the high peaks of the eastern part of the great
Himalayan range and the mountainous eastern border of Assam represent a boundary of
convenience. Knowledge of what happens to the fern-flora in montane Assam (except in Khasia

and, to a certain extent, Manipur) and north Burma is necessarily vague due to the political
: impossibility of even local botanists collecting in that little-known and perhaps somewhat

70 CHRISTOPHER R. FRASER-JENKINS

dangerous region; hence the importance of such collections as those of Ludlow and Sherriff, and
Kingdon-Ward, who were able to visit the far east Himalaya this century.

South of the Himalaya and Assam there are rather few ferns present in temperate genera,
except in the mountains of the south-Indian peninsula and in Sri Lanka. Due to the low altitude
and somewhat arid climate, the flora of the plains of central India is predominantly of a tropical
semi-arid type (Meusel, 1971) and excludes most species belonging to temperate or mesophytic
genera, except for a few subtropical ferns mostly representing south-east Asian elements, which
occur on the isolated low mountain ranges of Hindustan (central India). Sri Lanka and the
southern part of India were included in this study partly because they complete a political unit,
and because, although it is not very large, there is nevertheless a noticeable floristic connection
between the north and south, probably as strong as or stronger than those between the south and
any Other areas. The low hills of central-India may well have acted as stepping-stones in a
connecting land-bridge between the Himalaya and the south at some stages in the past (see
Burkill, 1924-25; Gupta, 1962, Bir & Vasudeva, 1973).

Within the Indian subcontinent four major floristic and geographical divisions are recognised:

1. The Himalaya

The Himalaya contains by far the richest flora and is the most important area for the present
study. As will already have been noticed, it has been geographically divided, following Hooker
(1904), Chatterjee (1940), and some later authors, into the west, central, and east Himalaya.
This division is both natural and convenient, considering the fact that the total length of the
Himalaya is approximately two thousand miles in a diagonal curve from the north-west to the
south-east, covering over 30° of longitude and also, importantly, some 10° of latitude (though
only some 250 km wide) and grading from the drier climate in the north-west to the wetter
conditions in the south-east. It is also reflected in the distribution patterns of species, including
those within the genera concerned, though this is not a case of hard-and-fast boundaries between
the three areas, but rather a gradation as might be expected. But there is a considerable
change-over of species occurring in about the region from East Kumaon (west of Nepal) to near
Jumla in west Nepal; the actual boundary between the west and central Himalaya being fixed for
reasons of convenience at the western Nepalese border. It is only since the opening up of Nepal
to foreign botanists in 1949 that the position of this boundary could be clarified. The boundary
between the central and east Himalaya on the other hand is often not yet completely clear except
for certain flowering plant genera (e.g. Primula, Rhododendron, Lilium) which, due to their
horticultural value, attracted considerable attention at an early date; more access to the east
Himalaya will have to be gained before it can be placed with accuracy. But for historical and
political reasons this has been placed at the eastern border of Nepal; Sikkim and Darjeeling,
which have been accessible for a considerable time until recently, being considered to be in the
east Himalaya. This is mostly borne out by the genera Dryopteris, Polystichum, and Athyrium,
though some of the east Himalayan species of Dryopteris may just extend to eastern Nepal (east
of the Arun) and the lower altitude plants may even reach the Kathmandu valley. Thus Nepal
and the term central Himalaya may be considered as synonymous, as first suggested by
Chatterjee (1940). Further study is perhaps required to examine the validity of this usage in
general, beyond the bounds of the present genera. It is also convenient and more-or-less natural
to consider Khasia and Assam as part of the east Himalaya. It should be noted that Stearn (1960)
did not use the term central Himalaya except in strictly geographical terms, as used here.
However, its phytogeographical distinctness is as an area where certain species do not occur,
rather than as a region with a distinct element of its own. Within the central Himalaya (Nepal),
which may also be referred to as the Nepal Himalaya in order to show that the term is not a
floristic one, various phytogeographical zones have been recognised. Mason (1933, 1955)
delineated five natural regions, whilst Stainton (1972) independently recognised seven, mostly
rather similar to those of Mason, and Dobremez (1972) condensed the zones into four. Though
these zones are clearly natural and useful they are not used here, though they have influenced
the present scheme of subdivision of Nepal into fifteen geographical areas; however for the

AN INTRODUCTION TO FERN GENERA OF THE INDIAN SUBCONTINENT 71

purposes of recording the distribution of species when spot-maps are at present impractical, at
least in most parts of the Himalaya, the natural phytogeographical zones within Nepal are
considerably too large. It should be noted that, for more general terms of reference, Stearn
(1960) proposed a useful subdivision of Nepal into three zones: west, central, and east Nepal,
bounded by the 83°E and 86° 30’E lines of longitude and based on the regions at which many
flowering plant genera demonstrate the furthest points of distribution of their east Himalayan
species towards the west (the 86° 30’ line) and of their west Himalayan species towards the east
(the 83°E line). This scheme is now in general usage, though, again, it cannot be utilised here in
the present more detailed context. As Stearn did not use the term central Himalaya phytogeo-
graphically (there being no such thing as a Central Himalayan species distribution pattern), he
pointed out that the 83°E line, between west and central Nepal, is the approximate dividing line
between the west and east Himalayan species, though this is not always strictly true in at least the
present genera, where the higher altitude east Himalayan species usually reach only into the
easternmost part of Nepal, or to Sikkim.

2. The plains of north and central India

The plains include the lowlands of Pakistan, the bulk of lowland India, and all of Bangladesh.
They are of considerably less interest than the other areas as far as ferns are concerned, though
they do contain, among others, the following low mountain ranges: Abu in Rajasthan,
Pachmarhi and Balaghat in Madhya Pradesh, Hazaribagh in the former Chota Nagpur region of
Madhya Pradesh and Bihar, and the interesting Parasnath Hill in Bihar. Some representatives of
the genera concerned occur in all of these hills.

3. The south-Indian peninsula

South India is considered here to include the low-lying north-western and central-western
Ghats, running through south Gujarat, Maharashtra, Karnataka, and north Kerala states; the
similarly low north-eastern Ghats run through south Orissa, Madhya Pradesh and north-eastern
Andhra Pradesh. Both of these extend down the peninsula of India along each coast and may
have been important as land bridges in the past; they also contain a few representatives of the
fern-genera concerned. Further south, in Kerala and Tamil Nadu (Madras) states are the high,
wet, and fern-rich south-western Ghats (Nilgiris, Anamalai, and Palni Hills), the south- and
central-eastern Ghats (including the moist and fern-rich Shevaroy Hills in Salem district), and
the southern Ghats, near the tip of India. The higher mountains in these areas are of
considerable interest, containing a large number of species, and being affected by both the
summer and winter monsoons; they contain both Himalayan and Sri Lankan/south-east Asian
species.

4. Sri Lanka

_ The island of Sri Lanka (Ceylon) shows floristic connections with south India, the Himalaya (in

some cases with the Himalaya but not south India), and south-east Asia. It also contains in
addition a rather high proportion of endemics, sometimes with related species in south-east

Asia. For this, as well as for political reasons and because it is a large island, it is considered a

_ separate major division. It contains an area of high mountains just south of the centre, and, like

_ the southern part of south India, is affected by both summer and winter monsoons.

Distribution and mapping of species

The detailed distribution of ferns in the Indian subcontinent has scarcely been considered. The

problems lie in part in the complex taxonomy of species in the large genera, and the
_misidentification that has so frequently occurred; the subcontinent also remains a little-studied
area because much modern work has been in the nature of small, local, annotated lists, rather
than a major flora of, for example, the Himalaya, or the whole of India. However, the

72 CHRISTOPHER R. FRASER-JENKINS

distribution patterns of species are of considerable interest and value, and a wealth of herbarium
material exists, both in herbaria in the subcontinent and in Europe and the U.S.A., and may be
drawn upon to produce maps, or descriptions of species distributions. Ideally, spot-maps could
be produced which would show detailed distribution of species as well as under-collected areas.
With this in mind I have compiled lists of all the specimens of Dryopteris, Polystichum, and
Athyrium seen in many herbaria and it is hoped that before long a full-scale mapping project
might come into being. In the meantime, in order to show the distribution of species, the
subcontinent has been divided up here into many, quite small, numbered regions, the number of
each area being listed under each species given in the taxonomic account (Fraser-Jenkins, 1985)
when it occurs in that area. It is emphasised that only actual specimens seen and identified by the
author have been considered for recording purposes, which is an advisable restriction. The
literature, especially for the subcontinent, abounds with incorrect records in any critical genus;
these may be useful for checking out possible new areas for a species, but are not to be relied
upon until the specimen cited has been seen and identified. Hence the value of numbering each
plant separately when collecting, and citing the specimen number when publishing new records.
The numbered areas are based partly on geographical considerations, mountain groups, etc.,
and partly on named political regions, as so many of the older specimens are labelled only with
the name of the region and modern collections are often too scarce from the less accessible areas.
Altogether 100 regions have been recognised, as follows below (Fig. 1). The author hopes to
publish detailed maps at a later stage.

1. Himataya: Afghanistan: 1 Badakhshan. 2 Nuristan. 3 Wakhan. 4 Central Hindu Kush.
5 West Kurram and Paktai. 6 Zabul, Katawaz-Urgun, and Ghazni.

Pakistan: 7 Baluchistan (Kalat). 8 Quetta and Sulaiman Mts. 9 Waziristan, Bannu, and Dehra
Ismail Khan. 10 East Kurram, Kohat, Tirah, and Khyber (this includes Aitchison’s specimens
from Kurram). 11 Peshawar and Mardan. 12 Malakand, Bajaur, and Dir. 13 Chitral and
Yarkhun. 14 Kalam, Swat, and Buner. 15 Gilgit and Tribal Territory. 16 Hunza (Baltit). 17
Karakoram. 18 Deosai and Baltistan. 19 Chilas, Wazarat, and Burzil. 20 Hazara and Murree. 21
Muzaffarabad. 22 North-west Pir Panjal, Kotli, and Mirpur.

India: 23 Poonch and Riazi. 24 West Kashmir. 25 North Kashmir. 26 East Kashmir and Maru. 27
South Kashmir and north Udhampur. 28 Jammu, Kathua, and south Udhampur. 29 West
Zaskar, Nunkun, and east Udhampur. 30 East Zaskar, Ladakh, and Nubra. 31 Rupshu and
Demchok. 32 Chamba and Pangi. 33 Dharmsala, Dalhousie, and west Kangra. 34 Mandi. 35
East Kangra, Rohtang, Kulu, and Parbatti. 36 Lahul and Spiti. 37 Simla and Hill States,
Mahasu, Sirmaur, and Chhachpur. 38 Kinnaur. 39 Jaunsar/Chakrata. 40 Dehra Dun, Mus-
soorie, and Nag Tibba. 41 Uttarkashi and Gangotri. 42 Tehri Garhwal and west Chamoli. 43
North Chamoli (inc. British Garhwal). 44 Lansdowne and Garhwal. 45 Central Kumaon (South
Chamoli, North Almora, and Pindari). 46 South Almora. 47 Nainital. 48 Pithoragarh.

Nepal: 49 Dandeldura and Api. 50 Siligarhi, Doti, and Saipal. 51 Rara, Jumla, and Dailekh. 52
West Nepalese Siwaliks. 53 Rukumkot and Dhaulagiri. 54 Inner Nepal, Mustang, and Mukh-
tinath. 55 Annapurna, Lamjung, and Pokhara. 56 Tansing (Tansen). 57 Gurkha, Manaslu, and
Ganesh. 58 Chitawan, Kathmandu valley, Shivpuri, and Bhadgaon. 59 Langtang, Ghumtang,
and Charikot. 60 Khumbu, Everest, and Chamlang. 61 Sindhuli Gari and Churia Ghati. 62
Dhankuta, Topke Gola, Taplejung, and West Kanchenjunga. 63 Hangsari and Ilam.

Sikkim and N.W. Bengal (India): 64 Sandakphoo, Phalut, Darjeeling, and Kalimpong. 65 North
Sikkim. (For convenience area 64 and 65 are referred to as Sikkim. )

Tibet (China): 66 Yatung and Chumbi.

Bhutan: 67 West Bhutan and Thimphu. 68 Punakha and Mangde Chu. 69 Bumthang. 70
South-central Bhutan. 71 Shingbe and Trashiyangsi. 72 Dewangiri, Sakden, and east
Bhutan.

Assam (India): 73 Duars. 74 Kameng Frontier Division and Tezpur. 75 Subansiri Frontier
Division and north Lakhimpur. 76 Siang Frontier Division and Dihang. 77 Luhit Frontier
Division. 78 Dibrugarh, Tirap Frontier Division, and south Lakhimpur. 79 Sibsagarh, Naga-
land, and Kohima. 80 Manipur (Imphal). 81 Tripura, South Cachar, Mizo, and Lushai. 82 Mikir

AN INTRODUCTION TO FERN GENERA OF THE INDIAN SUBCONTINENT 73

and North Cachar. 83 Khasia-Jaintea Hills. 84 Garo Hills. (For convenience areas 73 to 84 are
referred to as Assam.)

2. THE PLAINS: Pakistan: 85 Lowland Pakistan east of the Indus.
India: 86 West Indian plains. 87 Central Indian plains. 88 Eastern Indian plains.
Bangladesh: 89 North Bangladesh and the delta. 90 Chittagong and hill tracts.

3. SouTH INDIA: India: 91 North-western Ghats, Mahabaleshwar, and Dangs. 92 Central-
western Ghats, Mysore, north and south Kanara, Chikmagalur, and Coorg. 93 Nilgiris, Calicut,
and Coimbatore. 94 Anamalai Hills and Ernakulam. 95 Palni Hills and Madurai. 96 North-
eastern Ghats and Vishakhapatnam. 97 Central-eastern Ghats and Nellore. 98 South-eastern
Ghats, north Arcot, Shevaroy, and Pachaimalai Hills. 99 Southern Ghats, Rajapalaiyam, and
Trivandrum.

4. CeyLon: Sri Lanka: 100 Sri Lanka (Ceylon).

Acknowledgements

The author wishes to thank for their help Dr S. P. Khullar of Chandigarh, Mr A. O. Chater of London, Mr
J.D. A. Stainton of London, and especially Professor R. C. Ching of Peking, who has taken a great deal of
trouble to comment on the manuscript in detail and has corrected numerous small points and has informed
him of additional useful details, including Wallich’s criticism of Don, etc.

86

91

AN INTRODUCTION TO FERN GENERA OF THE INDIAN SUBCONTINENT fs)

Fig. 1 Regions of the Indian subcontinent.

76 CHRISTOPHER R. FRASER-JENKINS

British Museum (Natural History)

3

tig

Ferns of Jamaica

j A guide to the Pteridophytes

:G. R. Proctor

Jamaica. The succinct species descriptions include relevant synonymy and incorpo-
rate distributional data both within and outside Jamaica. Special emphasis is given to
_ the subtle distinctions between closely related species and all genera are illustrated.
_ Keys to the genera and species facilitate a wider use of the flora in the West Indies and
es South America. The author, one time Senior Botanist in charge of the

j This flora records and describes the 579 species and 30 varieties of ferns occurring in
x

_ Herbarium of the Science Museum, Kingston, Jamaica is an outstanding field

botanist and his expertise is reflected in the practicality of the flora and especially in
_ the habitat and ecological information. This volume represents an important addi-
__ tion to our knowledge of the flora of the West Indies.

“1985, 650pp (approx), 135 line illustrations, 22 maps. Hardback.
B69 00895 1 £50.00

1
h
i
|
i
'

Photoset by Rowland Phototypesetting Ltd, Bury St Edmunds, Suffolk — ¥

Printed in Great Britain by Henry Ling Ltd, Dorchester

Bulletin of the
British Museum (Natural History)

A revision of African Sphagnales

Alan Eddy

Botany series Vol 12 No 3 31 January 1985

The Bulletin of the British Museum (Natural History), instituted in 1949, is issued in four
scientific series, Botany, Entomology, Geology (incorporating Mineralogy) and Zoology,
and an Historical series.

Papers in the Bulletin are primarily the results of research carried out on the unique and
ever-growing collections of the Museum, both by the scientific staff of the Museum and by
specialists from elsewhere who make use of the Museum’s resources. Many of the papers are
works of reference that will remain indispensable for years to come.

Parts are published at irregular intervals as they become ready, each is complete in itself,
available separately, and individuaily priced. Volumes contain about 300 pages and several
volumes may appear within a calendar year. Subscriptions may be placed for one or more of
the series on either an Annual or Per Volume basis. Prices vary according to the contents of
the individual parts. Orders and enquiries should be sent to:

Publications Sales,
British Museum (Natural aye
Cromwell Road,
London SW7 SBD,
England. ~

World List abbreviation: Bull. Br. Mus. nat. Hist. (Bot.)

© Trustees of the British Museum (Natural History), 1985

The Botany series is edited in the Museum’s Department of Botany

Keeper of Botany: Mr J. F. M. Cannon

Editor of Bulletin: Mr J. R. Laundon

Assistant Editor: Dr A. J. Harrington

ISBN 0 565 08002 4 Botany series

ISSN 0068-2292 Vol 12 No 3 pp 77-162
British Museum (Natural History)

Cromwell Road Issued 31 January 1985
London SW7 5BD

ek.

oe crs

en al dg

as be

Vee yi
Nes Et

A revision of African Sphagnales

Alan Eddy cece
Department of Botany, British Museum (Natural History), Cromwell Road, London SW7 SBD

FERED SIDE a
;
‘

Contents

ERP eh aca on eran eo aero RO RLE tad Seencne a cedar dae wen dectabOwenderedataoansens oven oi
ARIS OUG AL AGK ROUND Gites fete ratesteaete ets one cersciate dnt aaiee citcdlagainaspln datdeeiadesiesaisiiesmentes 1
ae IML eer aera ce ce dass de CORN eae eine 6s Pasa Pace ac dOroke nWete onthe dd eadaedeosewers 78
Maxononiuc treatment andevolutionary MIStOLY......-.t60.ccecssossseceeeresceececcecoensctoncees 78
PMTEN OMCs IDE CUTIEN Sint a-ciso sc doom ioe se care ttee sake elactoanccdohunsewee aetgcehitiecce seclatieotelns 80
arama ease eS ccd ee es dele < SUN Te eae oo cle nw ostoablcoaicbias Mnieigl Seda weies dues annieas 80
a aN PUR Se Pfs it Sign AOS AES wins Sepa Sand « ake As Awiddw now slow canoes 80

PAINT DATMUUSDS POUCA VILE Beer acai cteacap orcas ofolet ols ates ania eles oie vew'e Nels oaiaala- Saws ¢8Gc vane aed bas 83
NS NNR Fe Sone oaate aces acedua silanes de spsennils | Cay soaet taamusieSanhletans ses siue 90
es UNPRANEMINE RIES Se ek ee cere etc eada tens Vasa oacte aetna retdcdcercdenaasinaer 90
ae A ERIE TEIEPLED 5 25 20a ce child ag Tove Sak gs NG Me tose vahee pose ee vase Hap ik ps nedablgacdiabensiees ee 99
SUSE CHIOMS UD SCCHIIA Ue Nes eerie cP ttetanactbaetetdema. dancicacesa sence asta cnsase te 99

SUSE CHOMMACLOSUIIARIIIIM metered acces qaaect Adiasa sree Ganache nde oeciasieeaaelones« 116

‘SECTIOM (CHS AIG IG Von scresganodcaduncat. So aeOrRoEscue ne eC eb E HEC OCME CH CER EE Ree Ree ceT te ae neene 137
NP IIR A TERR ENG SNe aba a aa e ae Ga aa so ey sont es Pan te Sy coals Srles Laaaeen staves Sade 155
Pea OVU Le OE INC TILSE RANE trench Bie teeis tacts w isa acer dr aicenaleNiciered ae wiatcinssoucwiat geiceore sibuis sires 159
Fens AMeT COS SEM OnPE TG Pe oy PPPER ET Cod Rane Se EUR OAR a einatlicls «ldcle ned aed ocendns osbebaoubaaedavienss iS)
dl ease ee a (ran cttrr ache Re San dacusue wMeniabneiddsdnch(elndciens Scinatabadeadah acta cswenaeates 160

Synopsis

A revision of the Sphagnales of Africa and the East African islands is presented. Twenty species are
recognized, and the treatment includes a key, descriptions, synonymy, figures of the taxa, and brief
discussions of their status and interrelationships. Sphagnum slooveri A. Eddy sp. nov. and S. tumidulum
var. confusum A. Eddy var. nov. are described, and Acrosphagnum C. Mill is regarded as a subsection in
section Subsecunda. The Asiatic species S. cuspidatulum C. Mill. and S. ceylonicum Warnst. are recorded
for the first time in Africa. The following new combinations are made: S. magellanicum Brid. subsp.
grandirete (Warnst.) A. Eddy, S. truncatum Hornsch. var. bordasii (Bescher.) A. Eddy, S. planifolium C.
Mill. var. angustilimbatum (Warnst.) A. Eddy, and S. planifolium var. rugegense (Warnst.) A. Eddy.

Historical background

The history of Sphagnum study in Africa follows familiar lines. It began slowly and spasmodical-
ly with Bridel’s (1826) descriptions of S. ericetorum and two other uncertain (and now
untraceable) taxa from the East African islands, and with Hornschuch’s (1841) treatment of S.
truncatum and S. capense from South Africa. Little progress followed until the latter half of the
century when much of the botanically unexplored regions were becoming accessible. The results
of botanical exploration contained fairly numerous specimens of Sphagnum, especially from the
East African islands and the Cape region. Predictably, repeated collections of what would later
prove to be the same taxon were no hindrance to their being described as new species. By 1897
Cardot, in his Répertoire sphagnologique, was already able to list 45 ‘species’ and numerous
varieties. Names continued to be added to the South African flora, largely due to Warnstorf
(1911, etc.), by which time (1908 onwards) various collections were becoming available from
some of the high mountain regions of East Africa. New species from that area were being
described as recently as 1955.

Characteristically, Warnstorf figures prominently among the authors of new species, but his
output (1890-1911) seems to have been somewhat reserved in comparison with his treatment of
South American material during the same period, presumably indicating that the genus figures

Bull. Br. Mus. nat. Hist. (Bot.) 12 (3): 77-162 Issued 31 January 1985

78 ALAN EDDY

less prominently in its African habitats than it does in, for example, Brazil. Cardot’s Répertoire
(1897) is basically an overview of the genus and contains few systematic evaluations of the
species listed therein. No more recent publication has taken account of the African flora as a
whole, although Garside (1949), in his account of Sphagnum in southern Africa, went some way
towards rationalizing the position with regard to that region. This treatise, and the earlier (1926)
simplistic treatment by Sim, have been recently superseded in Magill’s (1981) coverage.

Ecology

Temperate zone populations of Sphagnum, especially where the genus is an important element
of peatland vegetation, have been subjected to exhaustive ecological investigations. No such
detailed work has yet been carried out on African populations nor is there any evidence to
suggest that the broad requirements of Sphagnum in Africa differ from those of its holarctic
species. Therefore a markedly regional distribution is to be expected on the continent, its
pattern reflecting a similar distribution of relatively cool, moist climatic régimes. At European
latitudes a mean annual rainfall of about 1500 mm is the minimum amount that will support
healthy Sphagnum growth on ground with unimpeded drainage, but this must be part of a
seasonal cycle that does not have a marked dry period. The basic rainfall requirement may be
reduced in areas of extreme oceanicity, particularly where mists are frequent and evaporation
impeded, but is likely to be higher within the tropics where exposed populations are subjected to
periods of relatively intense radiation. The genus is not adapted to withstand seasonal dry
periods, although reasonable survival of occasional droughts has been observed in Britain. High
rainfall alone does not account for the observed distribution of the African species, for
Sphagnum is absent from vast areas of equatorial rain-forest where annual precipitation is more
than adequate. On the whole, sphagna are only moderately shade tolerant, and very few species
have developed affinities for even mildly base-rich soils (and the latter are so far only known at
much higher latitudes). Vigorous phanerogamic growth, high rates of leaf-litter deposition, and
varying degrees of edaphic improvement have proved to be effective barriers to incursion of the
genus.

A number of species either require, or are tolerant of, sub-aquatic or riparian habitats. Such
species are therefore to some degree independent of precipitation provided that groundwater
sources are maintained and they are oligotrophic in character (pH values below 4-5). The
subgeneric divisions Cuspidata and Subsecunda are noted for the aquatic predilections of many
of their species, S. planifolium and S. truncatum providing good African examples. Colonies of
these plants are seldom large in extent, probably because of the often temporary nature of their
habitats (water level changes, stream-bank erosion, etc.) but are able to survive, apparently for
long periods in some cases, in localities which would otherwise be climatologically unsuitable. In
general terms, however, Sphagnum forms a significant part of the bryoflora only in those regions
that have a climatic regime that more or less closely resembles that found, for example, in
western Europe. Broadly speaking, the regions correspond to the ‘Afromontane’ zone (sensu
e.g. White, 1978). Local topography, prevailing winds, and proximity to the oceans have
profound effects on the altitude of the zone but, in the broadest terms, it comes in at around the
2000 m contour, being depressed in the Cape region but somewhat higher in Ruwenzori and
western Usambara.

Taxonomic treatment and evolutionary history

In dealing with the taxonomy of African sphagna, we are faced with a number of problems, the
results of several interacting variables, some of which may be either insoluble or require
chemo-systematic and/or biosystematic investigations that are currently impossible to pursue.
Clearly, we are working with a number of essentially island populations, with varying degrees of
divergence or commonality in regard to their Sphagnum floras. There is more or less universal
agreement among geophysicists that these ‘islands’ are the end products of phases of tectonic
and orogenic movements that had their onset about mid to late Cretaceous. The major events

A REVISION OF AFRICAN SPHAGNALES 79

affecting the phytogeography of Sphagnum included: 1. the early separation of the African
plate from the South American—Australasian—Antarctic mass; 2. the northwards drift (mid-
Cretaceous to early Tertiary) of the African continent until it met Laurasia (closing the Tethys in
mid-Miocene); 3. warping and uplift, concentrated along eastern Africa during the Miocene; 4.
the drastic climatic fluctuations of the Pliocene and Pleistocene (Axelrod & Raven, 1978).

Paleoclimatologists visualise the continent as a whole, when it was 15°-18° further south, as
having an overall cooler and moister climate with more widely distributed and probably more or
less continuous zones of vegetation. There is little doubt that Sphagnum was a widespread genus
during this period, but became progressively restricted during the northwards drift as tempera-
tures rose and arid zones widened. Extinction might have been more widespread were it not for
the refugia provided by the Miocene uplifting and vulcanism in the east.

Speciation rates of bryophytes in general are low, and that of Sphagnum, at least within tropic
latitudes, is probably near the bottom end of a depressed scale. Evolutionary forces are
operating almost entirely on the haploid generation of its life cycle. Furthermore, the outbreed-
ing potential of most Sphagnum species (the majority being dioecious), which is almost
negligible among present-day populations, is likely never to have been high. Therefore, the
episodes of rapid speciation that occurred among flowering plants, particularly during the
Miocene and Pliocene—Pleistocene periods, had muted and possibly sometimes undetectable
effects in Sphagnum. Genetic changes that have produced markedly distinct species or even
genera of angiosperms are paralleled by such minor modifications in Sphagnum that it is difficult
to draw the line between genuine evolutionary steps and simple local and spasmodic variants. At
any rate, endemism at species level is very low in Africa generally, although many of the regional
variants, treated below as varieties or subspecies, might be raised again to species rank by some
authors. The present author prefers the more conservative approach (admittedly somewhat
arbitrarily applied at times) in which the distinction of species rests on more than one character
and is more or less clearly defined throughout the range of the taxon. Regional variants,
sometimes resting on a single character, are mainly regarded as subspecies, while distinctive but
apparently more or less random variants are accorded varietal rank. This, admittedly arbitrary,
treatment has nevertheless shown up patterns of distribution that are significant from the
standpoints of evolutionary history and phytogeography.

Madagascar presents something of a special case for, while its Sphagnum flora is clearly
derived from that occupying the Ruwenzori/Usambara afromontane region and has many
species in common with it, it has also an exceptional number of endemic species. It has only
recently been established beyond reasonable doubt that (a) the original position of Madagascar
was adjacent to Kenya/Somalia and not further south, and (b) initial separation of the
Madagascar sub-plane was an early event, i.e. before mid-Cretaceous (Smith, 1976). While it is
at present impossible to estimate at what point interchange between the island and mainland
floras was terminated, it seems to have been sufficiently early to allow the evolution of endemic
forms (e.g. the almost stenotypic S. tumidulum). It is significant, perhaps, that of those
‘continental’ species which appear to be unaltered in Madagascar, the majority are exceptionally
fertile members of the genus (e.g. S. strictum subsp. pappeanum) so that long-range dispersal
might have been possible. At this juncture it should also be pointed out, however, that some
disjunct taxa, e.g. S. ceylonicum, occur in East Africa and Madagascar in more or less identical
facies but are very rare, or unknown as fruiting plants. Clearly, such taxa are plaeo-disjunct
relics, and in the case of the species mentioned, its occurrence in virtually the same form also in
Sri Lanka is significant.

The diversification into the major infra-generic groups (i.e. subgenera and sections now
widely accepted) of Sphagnum is almost certainly pre-Gondwanan, and those represented in
Africa correspond closely with the other continents (i.e. subgenera Rigida and Sphagnum;
sections Acutifolia, Cuspidata and Subsecunda). Species and species-aggregates, however, apart
from the previously documented pan-tropical taxa S. perichaetiale and S. strictum (Eddy, 1977),
are largely endemic to Africa. Links with Asia rest on the occurrence of S. ceylonicum (referred
to above) and the recently discovered S. cuspidatulum in Madagascar. Tenuous links with the
tropical South American flora are mainly at supra-specific level (e.g. S. gracilescens is thought to

80 ALAN EDDY

have common ancestry with S. capensis agg.; S. cucullatum is similarly related to S. truncatum).
There are no apparent links with the Australasian flora to suggest any significant post-Jurassic
interchanges. Perhaps the most problematical links, if such they are, between the Eurasian and
African floras are represented by the occurrence of $. cuspidatum, a widespread holarctic
species, in Madagascar and Ruwenzori/Usambara. Possibly the Ethiopian region, at a time
when it enjoyed a more equable climate, formed a migratory bridgehead. Phytogeographical
considerations are developed somewhat further in the notes appended to the species descrip-
tions below.

The subgeneric divisions of the genus Sphagnum, as interpreted by the present author, are the
same as those that have been already described and discussed in relation to the tropical Asiatic
flora (Eddy, 1977), with the addition of a subordinate group (i.e. subsection) that is almost
endemic to Africa. Thus, subgenera Sphagnum, Rigida and Isocladus all occur, the latter
subgenus represented by sections Subsecunda, Cuspidata and Acutifolia. Of the groups listed
above, section Subsecunda is perhaps the most ancient in origin, with a consequently wide
geographical range and morphological diversity. A more or less well defined, African element,
previously categorised (Eddy, 1979) as the ‘Sphagnum capense’ group, is worthy of distinction.
C. Miller (1887) recognized the group, using the descriptive term ‘Sphagna mucronata’ but
placed it in the subgeneric group ‘Acrosphagnum’. A description and short discussion is
presented at the appropriate place in the text (see p. 116).

Citation of specimens

As already stated above (p. 78) most of the specimens examined (all cited specimens have been
seen by the author) originate from the relatively few regions of Africa where the degree of
oceanicity is enhanced by high altitude and/or proximity to coasts. Specimens are therefore
grouped into the five broad categories: 1. West Africa — relatively few records of scattered
occurrence from Sierra Leone south to western Zaire and Angola; 2. Southern Africa — mainly
from Cape Province but extending eastwards to Transvaal and Natal; 3. Mlanje/Chimanimani—
the high ridges and plateaux of Zimbabwe, Malawi and northern Mozambique; 4. Ruwenzori/
Usambara — the high altitude ridges of eastern Zaire, Rwanda, Uganda, Kenya and northern
Tanzania; 5. East African Islands — principally Madagascar, Mauritius and Réunion.

SPHAGNUM L.

Sphagnum L., Sp. pl. 2: 1106 (1753); Gen. pl. Sth ed.: 487 (1754).
Lectotype: Sphagnum palustre L., cf. Britton, Fl. Bermuda: 431 (1918).

Sphagnum, as interpreted by modern authors, is the only genus in the class Sphagnopsida. Its
characters are sufficiently numerous, conspicuous and unique, and have been so thoroughly
documented that a full generic description is unnecessary here. Good definitions are presented
in most bryological handbooks, dating back as far as the classic works of Schimper (1857) and
Braithwaite (1880). Generic subdivisions, which are more or less coincident with those
presented by Isoviita (1966), are the same as those presented in Eddy (1977). Therefore
synonymies of the higher taxonomic divisions are presented less fully in this paper, and include
only additional and recently used names.

Key

la_ Branch leaves large, more than 1-6 mm long, concave with markedly blunt, hooded apices.
Stem leaves lingulate to spathulate, more than 1-4 mm long. Stem cortex highly developed,
3 or more layered. Cortical cells of branches more or less uniform, few to many with large
pores, normally furnished with distinct spiral fibrils, at least on the inner walls (subgen.
Spa griesTa)y 5s we hae Seat eee ae kee Cea es One Reine ae OE ae cee eee Cee ee 2
1b Branch leaves various: if large and hooded then stem cortex not highly developed (i.e. 1-2
layered) or stem leaves small, under 1-3 mm long. Cortical cells of branches dimorphic or, if
uniform, then without fibrils :2.7.5..4:22 cotta nececeetedet anera Reena nee ne een ee eee eee 3

2a

2b

3a

3b

4a
4b

Sa

5b

6a
6b

7a

7b

8a

8b

9a
9b

10a
10b

lla

11b

12a

A REVISION OF AFRICAN SPHAGNALES 81

Branch-leaf chlorocysts urceolate in section, always distinctly exposed on the adaxial leaf

surface and normally reaching the abaxial surface. Cortical cells of stem uniformly 1-porose

(very rarely 2-porose in an occasional cell adjacent to the insertion of a fascicle or stem
E01 ..stea, seca: bee neem ie Ba egietaame die aie dle amet ee 1. S. perichaetiale (p. 83)

Branch-leaf chlorocysts mainly oval in section and completely immersed (only here and there

reaching to the leaf surface). Cortical cells of stem often 2— or more porose ...................44.
2. S. magellanicum subsp. grandirete (p. 87)

Plants very robust, with markedly dimorphic branches. Branch leaves more than 2-0 mm long,
usually subsquarrose, but stem leaves very small, under 1-3 mm long. Hyaline cells of
branch leaves short and wide (over 30 um), with pseudolacunae on the adaxial side. Cortical
cells of branches not dimorphic, the majority having a large apical pore.......................0055

20. S. strictum subsp. pappeanum (p. 155)

Plants small to medium sized or, if robust, then branches not markedly dimorphic and stem
leaves more than 1-4 mm long and branch-leaf leucocysts long and narrow, without adaxial
pseudolacunae. Cortical cells of branches, at least distally, dimorphic with distinct retort

SNR aoe tether oiischs ete ee SGM Seleee + ecle'siawieiuceje'saje Velemawiaaaviagelse eelelnaslemanacdenutesilstiesienselet 4
Stem leaves lingulate to spathulate, very broad and fimbriate at apex and devoid of fibrils....... 5
Stem leaves various, sometimes truncate and dentate or eroded above but very rarely broadly

fi ALC ANC LALELY CCVOIG OUMODUS ncccteannare ct senses n<cnsaceaenseaaadecieecvincscnsaceheoseheetaseues 6

Stem leaves strongly widened above and fimbriate around the whole upper portion. Chloro-
cysts of branch leaves with wider exposure on the adaxial leaf face. Branch leaves never in
es OAT OWS ee ca actes ciara! co RR ES aa eee ete Maes AMC ee Madara deade 5. S. fimbriatum (p. 97)
Stem leaves not conspicuously widened above. Chlorocysts of branch leaves with wide abaxial
exposure, often immersed on the adaxial side. Branch leaves 5-ranked ...................0.000005
19. S. cuspidatulum (p. 153)

Middle and upper branch leaves narrowly lanceolate to linear, more than 2-5 times as long as

wide. Chlorocysts of branch leaves always more widely exposed abaxially (sect. Cuspidata) 7
Upper branch leaves lanceolate to ovate, usually less than 2-5 times as long as wide (if

narrower, then chlorocysts more widely exposed adaxially) ..............ccccccceceeeeeeeeeeeeewenes 8
Stem leaves typically fibrillose only in the upper half, border distinctly to markedly widened

below. Branch leaves often strongly involute and appearing setaCeOUS.................c0ceee eens
16. S. cuspidatum (p. 137)

Stem leaves typically fibrillose throughout; border not or only imperceptibly widened below.

branenleaves mone Onless plane to weakly INVOICE 2. 2.. cecsen dceeeneerace-cecerenceencseesenereses
17b. S. planifolium var. angustilimbatum (p. 148)

All or the majority of branch leaves proportionately broad, ovate, usually less than 1-5 times

longer than wide (if narrower, then large, more than 1-8 mm long).....................6c00eeeeee es 9
Branch leaves ovate-lanceolate to lanceolate, 1-5—2-5 times longer than wide ...................... 22
lants robust; branch leaves 1-6—2-5 mm OrmMorelONngs.......o...cnsccse0coscencceoscecwcteeceeedoseectece 10
Plants small to medium sized; leaves less than 1-Smmlong....................0cceceeceeeeceeeeeeeeeeees 16
Apices of branch leaves rounded and eroded, not obviously toothed. Stem cortex

IGE) haere die, shake nmennam ae cm cee ins cn mtecece mae coat toeaaemee os caecaeme scott 9. S. africanum (p. 113)
Apices of branch leaves either more or less mucronate or narrowly to broadly truncate with

conspicuous teeth, if rounded and eroded, then stem cortex 2—3-layered ..................0.00005 11

Abaxial (sometimes also the adaxial) surface of branch-leaf leucocysts furnished with several
to very numerous small, ringed pores along the commissures. Chlorocysts in section with
oval lumina and thickened end walls, with more or less equal exposure on both leaf surfaces 12
Abaxial surface of leucocysts without or with very few pores (sometimes with pseudopores),
adaxial surface with unringed or thinly ringed, often very inconspicuous pores mainly
adjacent to the cell angles. Chlorocysts in section triangular to trapezoid with conspicuously
OUSIDE STC BENT OSI EN semen ag sele opcet pcepetinde: HOSE -Oe Hed eee aR EERE: Pt 4ae EE een ea ecnEe ean eeneeney Eames 15

Stem cortex basically 2—3-layered (spasmodically 1-layered in weak specimens). Plants pale or
with weakly developed secondary pigments (endemic to E. African islands) ......................
8. S. rutenbergii (p. 108)

82
12b

13a

13b

14a

14b

15a

15b

loa

16b

17a
17b

18a

18b

19a

19b

19c

20a
20b

21a
21b

21c

ALAN EDDY

Stem cortex 1-layered (sometimes up to 30 per cent 2-layered in aquatic forms). Plants often
with highly developed secondary pigments, then orange-brown to purplish brown ............

Branch leaf leucocysts with small pores in series along the commissures on the abaxial side;
adaxial side with few or no pores. Chlorocysts weakly biased towards the abaxial leaf face
(not Southof the Salata) ii. tavaue don ewan ert were eee ee 6. S. auriculatum(p.

Branch leaf leucocysts very variable, commonly with several to numerous pores on the adaxial
surface. Chlorocysts median or weakly biased towards the adaxial leaf face (not north of the
Sanaa) ergata Se ee eee eee Re ce OREN: Boner ORC (S. truncatum)

Branch leaves often only shallowly concave, with almost flat, broad and conspicuously
dentate, truncate apices. Pores very variable, often scattered and non-serial on the abaxial
leaf face. Stem leaves often as large as branch leaves, fibrillose to base or almost so;
leucocysts not conspicuously septate ..................... 7a. S. truncatum var. truncatum (p.

Branch leaves very concave, with more or less involute margins above. Pores usually very
numerous and more or less serial on the abaxial leaf face. Stem leaves shorter than branch
leaves, sometimes efibrose in the lower half but often with very numerous septa................

7b. S. truncatum var. bordasii(p.

Branch-leaf leucocysts with few or no pores on the abaxial side; adaxial side with rather faint
pores more or less confined to the cell angles. Stem leaves normally fibrillose to insertion ....

17a. S. planifolium var. planifolium (p.

Branch-leaf leucocysts with 0—several pores and/or pseudopores on the abaxial side; adaxial
side with numerous, sometimes serial and occasionally distinctly ringed pores scattered
along the commissures. Stem leaves commonly efibrose below...................4cceceeeeeeeeeeeees

17c. S. planifolium var. rugegense (p.

Branch-leaf pores minute, less than 3-Oum diameter, few, mainly adjacent to some of the cell
angles on the adaxial side of leaf. Chlorocysts in section oval, deeply immersed below both
IERIE UT RYCES wasn ons -scaeonsbanatoneecaionagebasdcdasteabonbe 10a. S. tumidulum var. tumidulum (p.

Branch-leaf pores small to medium sized, usually more than 3-0 um (up to 10-0 um), usually
numerous on either or both surfaces. Chlorocysts usually narrowly exposed on one or both
eat SUTfa CES Siete cca iscttertaauhs sebsnnte cia tee ne cea cemit cee ats eae Me tebioe tanec Pee nner eee eee

Branch-leaf pores confined to, or much more numerous on the abaxial leaf surface ...............
Branch-leaf pores more abundant on the adaxial leaf surface .............. 0.2... ccceeeeeeeeeeeeeen ects

Chlorocysts of branch leaves more widely exposed on the abaxial side, rarely completely
WMIMET SOG 2 ofan ence thea diahanee wie sriebe neta chee estate eee Se a eas elas aeciesnccc ee eater
Chlorocysts either median and exposed equally on both sides, or with widest exposure on the
AG axtal'S1 de ies 5..% soaeecwcnemed sc eeee ectek ae emcee eeceO cee sO ae Se eia> teeit dace eiaeire Saeed gee eee eee

Plants relatively small and delicate, never with more than 4 branches per fascicle. Branch
leaves variable, often blunt or truncate-dentate at apex. Stem leaves typically lingulate to
spathulate and rounded-obtuse at apex, rarely pseudomucronate, usually fibrillose almost
OR QUIGE TOIDASE é.c.08Sanepemncnigeniesiin aloes ene eee eee ca EIRE RL ache Sas oe ee ie eee 12. S. capense(p.

Plants rather robust, rigid, with 5—6 branches per fascicle. Branch leaves abruptly inrolled at
apices and appearing acute or pseudomucronate. Stem leaves typically oval-triangular and
pseudomucronate, often fibrillose only in the upper half. (Chlorocysts narrowly exposed
abaxiallly))i.c setae. caacence aster cae tenet ee eee cee ace eee 13. S. pycnocladulum (p.

Plants rather weak, with 3—5 branches per fascicle. Branch leaves pseudomucronate. Chloro-
cysts more or less completely immersed........ S. pycnocladulum, ‘beyrichianum’ forms (p.

Chlorocysts of branch leaves median, with equal exposure on both leaf surfaces...................
11. S. davidii (p.

Chlorocysts of branch leaves with distinctly to markedly wider exposure on the adaxial leaf
SUPPAGE SS sc AER Se eRe sae Seat Se Re eee SRE Roe Re ee ee ee eae 15. S. ceylonicum (p.

Chlorocysts of branch leaves usually completely immersed. Stem leaves with border distinctly
widened below? cn. 5-eiss cs eee eer ee neece eee 10b. S. tumidulum var. confusum (p.
Chlorocysts exposed at least on the abaxial side of leaf. Stem-leaf border not obviously

99)

14

105)

106)

146)

148)

117)

17
18
ya)

19

20

124)

130)

131)

121)
135)

120)

expanded! below 222.2402 AGATE gi aettcene sn name heen oe eee eee ote eee 14. S. ericetorum(p. 132)

Chlorocysts median and exposed on both leaf surfaces, Stem leaves large, lingulate ..............
11. S. davidii (p.

121)

A REVISION OF AFRICAN SPHAGNALES 83

22a Branch-leaf apices widely truncate and conspicuously 8—-12-dentate. Stem cortex basically

SE MR EC Mie Wer OOEOSE 25d as Se tmrenvainedstecinn deans aiouenaqeebusacturs 7. 8. truncatum (forms) (p. 101)
22b Branch-leaf apices involute or rounded-truncate, 5—7-dentate. Stem cortex 2-4-layered, often

Rema E CNR AONNE PIOLCS 50 50.0 Jacl Meceree enon seg ¢e acs teronsnaycneicinndenweetiy en dddtniasandestdgeeees 23
23a Chlorocysts of branch leaves with much wider exposure on the abaxial side. Stem cortex never

PICMEOTS oe -ENs 3 aera rs Ratha ore fess cee woieteiseetststore jaunt alsaiaelo Gaelic acviewnlst winged ea ssiebigsreanenscie ZF
23b Chlorocysts of branch leaves with equal or wider adaxial exposure. Stem cortex usually with

METRE ONES SE ALIN). CE erase ben cin deacxcanuniebateds Joan nusracesduher<tapaves send Cletdaniesaess 24

24a Branch-leaf leucocysts (20—)25—35(-40) wm wide in mid leaf; pores large (10-0-20-0 um).
Leucocysts in lower-lateral parts of leaf much larger and more porose than the upper and

AVEDA GL 0 MPT ease eras chee ae a em ths ood ea se Seo Sanu ate Galena he Se stenled ys cae dow ciasinaie oe a ae 2S
24b Branch-leaf leucocysts (15S—)20(—30) um wide, with numerous small to medium (c. 4-0—10-0
um) pores. Lower-lateral leucocysts not markedly enlarged ....................:eeceeeeeeeeeeeee eens 26

25a _ Plants usually tinged with crimson pigments (which turn bluish in alkaline solution). Branch-
FEAeCUCOCYS(S mMONMal’, 1.6. MOCSE Plate aa. .a--cnncaderess-beetec.scceusc cece 3. S. violascens (p. 90)
25b Plants always pale and flaccid. Branch-leaf leucocysts often 1-4-septate... 4. S. reichardtii (p. 96)

26a Chlorocysts of branch leaves median, with equal exposure on both sides of leaf....................
11. S. davidii (p. 121)
26b Chlorocysts of branch leaves with wider adaxial exposure ................... 15. S. ceylonicum (p. 135)

27a Stem leaves broadly truncate and fimbriate at apex, fibrillose only in the extreme apical region
18. S. slooveri(p. 151)

27b Stem leaves blunt and often cucullate at apex, fibrillose in at least the upper third and often
Re TEOR ITEM CO IAS leg te au eticcinenin'ss gnu Stwalvadcienle<Uouceedcevnectigd ecisadas (S. planifolium) 28

28a Adaxial surface of branch-leaf leucocysts with scattered unringed pores. Plants often robust
with branch leaves over 1-8 mmlong.................. 17a. S. planifolium var. planifolium (p. 146)
28b Adaxial surface of branch-leaf leucocysts with numerous, minute, heavily ringed pores. Plants
usually small to medium sized with branch leaves under 1-6mm ....................2c0c0eeeee eee ees
17c. S. planifolium var. rugegense (p. 148)

Subgenus SPHAGNUM

Sphagnum L. subgen. Sphagnum

Sphagnum sect. Cymbifolia Schimper, Syn. musc. eur. 2nd ed.: 847 (1876).
Sphagnum [sect.] 1. Inophloea Russow in Schr. NaturfGes. Univ. Jurjeff 3: 27 (1887).

Type: Sphagnum palustre L.

More or less robust, turgid plants with cucullate branch leaves and markedly dimorphic
branches. Stem hyalodermis highly developed, in 3—4 layers and with internal as well as external
pores. Branch hyalodermis uniform, without distinct retort cells but few to the majority of cells
foraminate. Cells of hyalodermis of branches and stems typically furnished with spiral fibrils.
Outer margins of branch-leaf borders resorbed to form a resorption furrow. Leucocysts wide,
with pseudolacunae on the abaxial side, those near the leaf apex typically much resorbed in the
apical half on the abaxial side, the remainder characteristically projecting and scale-like (hence
leaf tip appearing minutely scabrid).

1. Sphagnum perichaetiale Hampe
(Figs 1 & 2)

in Linnaea 20: 66 (1847). Type: Brazil, Beyrich s.n. (BM — holotype).

S. balfourianum Warnst. in Hedwigia 30: 153, tab. 17 fig. 21, tab. 22 fig. z (1891).

S. arbogastii Ren. & Cardot in Bull. soc. r. Bot. Belg. 32 (2): 8 (1893).

S. marlothii Warnst., Sphagnol. univ.: 471, fig. 79A (1911).

S. drouhardii Cardot apud Ren. & Cardot in A. & G. Grandidier, Hist. phys. Madagascar 39: 42 (1915).

Note: possibly the earliest name for this taxon is S. patens Brid. (1806), also given as S. cymbifolium var.
bourbonense P. Beauv. and, later, S. cymbifolium var. patens (Brid.) Brid. (1826). No specimen of S.

84 ALAN EDDY

Fig. 1 Sphagnum perichaetiale Hampe. A, branch leaf (left); B, stem leaves; C, branch fascicle; D, stem
cortex; E, branch cortex; F, transverse section of stem (B, left fig., and E, left fig. drawn from the type of
S. balfourianum; the remainder from the type of S. arbogastii).

A REVISION OF AFRICAN SPHAGNALES 85

_ Fig. 2 Sphagnum perichaetiale Hampe. A, adaxial (left) and abaxial (right) surface of branch-leaf
leucocysts; B, adaxial surface of branch-leaf leucocyst; C, abaxial surface of branch-leaf leucocyst; D,
transverse sections of branch leaves (A drawn from the type of S. arbogastii; B, C drawn from the type of
S. balfourianum; D from various specimens in BM).

86 ALAN EDDY

patens could be traced in Bridel’s herbarium, and the description, which is inadequate for positive
identification, could equally well apply to S. magellanicum or even some unrelated species. S. patens Brid.
and its synonyms are dismissed here as nomina ambigua.

Plants rather robust, yellowish to brown, with or without additional dull red-brown or purplish
pigmentation, compact to rather lax. Fascicles composed of 4-5 branches: 2 spreading branches
of somewhat tumid appearance, blunt or tapering distally, 8-O—18-0 (—25-0) mm long; 2-3
pendent branches strongly deflexed, pale and relatively thin and attenuated (i.e. branches rather
strongly dimorphic). Stems 0-8-1-2 mm diameter; cortex very well developed, of 3(-4) layers of
large leucocysts which usually have sparse to numerous spiral fibrils; internal walls, end walls
and external walls all furnished with large pores, the exposed outer walls each with a single pore.
Internal cylinder strong, dark brown. Branch cortex uniform, of a single layer of large leucocysts
which are usually fibrillose (especially in distal parts of branches), some or all of them with a
single, non-protuberant pore. Stem leaves very variable, sometimes (strong, heterophyllous
plants) short-lingulate and almost, rarely completely, efibrose, sometimes (weaker plants)
strongly fibrillose and resembling branch leaves in morphology, usually more or less spathulate
and rather weakly fibrillose in the upper half. Apex typically broadly rounded with a thin,
ephemeral border, appearing variously eroded or more or less fimbriate. Border not expanded,
much resorbed above. Leucocysts strongly resorbed on the adaxial side. Branch leaves erect-
spreading, large and very concave, broadly ovate, 1-4~2-4 mm long, 1-0-1-5 mm broad; apices
strongly hooded and usually scabrid dorsally due to projecting, partially resorbed hyaline cells;
border thin, of a single cell row, with a resorption furrow. Leucocysts comparatively broad and
strongly convex, (25—)30—40(—50) um wide, 80-200 um long. Abaxial surface, in mid leaf, with
pseudolacunae or ‘triple pores’ at the conjunction of basal and lateral angles of adjacent cells
but with no, or very few, additional ringed pores in the other cell angles or against the com-
missures; near leaf apex with a large resorption gap in the apical angle; pseudopores usually
present in some lateral angles. Adaxial surface usually lacking pores in mid-leaf but 2-4 series of
leucocysts in the lower marginal zones of leaf with few to several large, circular pores (c.
15-0—-20-0 um) on both surfaces. T.S. leaf. Leucocysts plane to moderately convex adaxially,
very strongly convex abaxially. Photosynthetic cells relatively narrow, urceolate to more or less
oval trapezoid with oval lumina, abaxial wall strongly thickened, rather narrowly (2-0-6-0 um)
exposed on the adaxial side, just reaching, or much more narrowly exposed on the abaxial side
(there sometimes appearing immersed due to the degree of convexity of adjacent leucocysts).
Dioecious. Male plants either absent or of very rare occurrence in the African range of the
species. Female plants detected among some gatherings but always in an unfertilized condition
(see Eddy, 1977).

Distribution. S. perichaetiale is a pan-tropical species which is widespread in Central and South
America, Malaysia, and southern North America. It is much more local in its African
occurrence generally, but is common in the East African islands. No plausible explanation has
been found to explain its absence from East Africa in view of its presence both in South Africa
and Angola, there being no shortage of edaphically and climatically suitable locations for the
species in, for example, the Ruwenzori range.

West Africa. ANGOLA: Huila, Gossweiler 2804 (BM).

Southern Africa. CAPE PROVINCE: Table Mountain, 1020 m alt., Marloth s.n. (BM — isotype of S.
marlothii); Taylor 1105 (BM). Steenberg, ‘rocks in stream. . .’, Esterhuysen 15839 (BM, PC; female but
unfertilized plant); mountains above Hermanus, Vogelpoel CH 12796 (PRE); Groot Kop, Twelve
Apostles, on marshy ground by stream, 800 m alt., Pillans 4280 (PRE); between Bot River and
Kleinmoud, near a stream below Highlands Forest Dept. road, Middlemann 24977 (PRE).

East African Islands; MADAGASCAR: Ankafiafé, Arbogast s.n. (PC — type collection of S. arbogas-
tii); Fianarantsoa, Bessons.n. (PC); *. . . forét de Manongiarivo, . . .. Drouhards.n. (PC-— type collection
of S. drouhardii; Androranga, Mount Anjenabe, 600-800 m alt., Humbert & Capuron s.n. (PC); ‘Marais
pres d’Ambohimasoa’, 1100 m alt., coll.? (PC); Mt. Itrafanomby, Ankazondrano, Humbert s.n. (PC);
Andovoranto, Humbert 565 (BM, PC); ‘Mountain west of Itremo’, Humbert s.n. (9th voyage) (PC);
Plateau d’Isalo, 500-1000 m alt., Humbert s.n. (PC); Massif du Tsaratanana, Humbert s.n. (Sth voyage)
(PC); Fort Dauphin, Decary s.n. (PC); Cremers 2963 (NAM); Marojejy, 2050 m alt., Guillaumet 4068,

A REVISION OF AFRICAN SPHAGNALES 87

4069a, 4211 (NAM, BM); Fort Dauphin, ‘on sand between Ericas’, 100 m alt., Onraedt 71.M.5506,
71.M.5535, 71.M.5540, 71.M.5555 (NAM, BM); Massif d’Isalo, Canon des Singes, Onraedt 71.M.5732;
Tananarive, Lac du Mantasoa, Onraedt 74.M.2106, 74.M.2109 (NAM, BM). MAURITIUS: Balfour s.n.
(BM, herb. Mitten — type collection of S. balfourianum); in bog, north of Mt. Cocotte, 600 m alt. ,Onraedt
71.Ma.045, 71.Ma.0234, 71.Ma.0236 (NAM, BM). REUNION: Bosser 8275pp (BM, PC mixed with S.
pallidum); Plaine des Cafres, in bog, 1600 m alt., Onraedt 69.R.0476, 69.R.0477 (NAM).

Subgenus Sphagnum is comparatively easy to recognize, even superficially, with its strongly
hooded leaf tips and dimorphic branches. There are, however, no superficial characters by
which S. perichaetiale can be distinguished from S. magellanicum. Under the microscope, the
uniformly 1-porose stem cortex and the adaxially exposed chlorocysts are the more important
features of S. perichaetiale. S. magellanicum usually has more abundant branch-leaf pores, and
at least a proportion of stem-cortical cells with two or more pores in the external walls.

S. perichaetiale is a species of great antiquity, and is the most ‘primitive’ member of its
subgenus. Many of its populations are of a relict nature, and it has without doubt disappeared
from many of its former sites. Nevertheless, it is difficult to account for its apparent absence from
continental East Africa in view of its relatively frequent occurrences in the East African islands.

2. Sphagnum magellanicum Brid. subsp. grandirete (Warnst.) A. Eddy, comb. nov.
(Figs 3 & 4)

S. grandirete Warnst. in Magy. Bot. Lap. 1: 43 (1902). Type: Madagascar, Androrangabe, Andovoranto
‘ab indig. collect., 1901’ (PC).

Plants relatively large, lax or rather compact, brown to purple-brown with indistinct to distinct
crimson tints. Fascicles composed of 3-5 strongly dimorphic branches: 2 spreading branches
more or less tumid, 10—20 mm long, blunt or slightly tapering with age; 2-3 pendent branches
of varying length but weak, relatively pale, more or less appressed to stem. Stems 0-8-1-1 mm
diameter. Cortex highly developed, of 3-4 layers of leucocysts which are usually furnished with
thin, more or less spiral fibrils and pores; exposed faces of outer leucocysts with 1-2(-4) pores,
rarely all with just a single pore; inner radial walls also with 1-3 pores. Internal cylinder strong,
dark red-brown. Branch cortex not dimorphic, normally fibrillose (at least in the distal parts of
branches), some or a majority of leucocysts with a large pore near the upper end. Stem leaves
variable, lingulate to spathulate, 1-2-1-8 mm long, 0-9-1-2 mm broad; border narrow, largely
resorbed, forming a thin, ephemeral band 2-3 cells wide around the whole apical margin; not
expanded below. Leucocysts in upper leaf region variable, typically short and wide and lacking
fibrils or more elongated and weakly fibrillose; adaxial surfaces largely resorbed. Branch leaves
large, (1-4-)1-6-2-5 mm long, very concave, broadly ovate to more or less orbicular or
cochleate; apices strongly cucullate and more or less scabrid dorsally (due to projecting dorsal
leucocyst walls); border thin, of 1(—2) cell series, with a well marked resorption furrow.
Pendent-branch leaves ovate to ovate lanceolate, the distal often unbordered. Leucocysts wide,
20-0-35-0(—40-0) x 90-0-170-0 um in upper mid leaf. Abaxial surface with well marked ‘triple
pores’ in the conjunction of the basal and lateral angles of adjacent leucocysts, with or without
additional rather large, ringed pores elsewhere; commonly with a large resorption gap in the
apical angle, especially towards leaf apex. Adaxial surface with or without 1-3 large, circular,
unringed pores, about 20 um diameter; 3-4 series of leucocysts in lateral leaf regions more or less
equally porose on both sides. 7.S. leaf. Leucocysts shallowly and more or less equally convex on
both sides or rather more strongly convex abaxially. Chlorocysts small, narrowly oval, the
majority thin-walled and completely immersed, only here and there reaching the adaxial surface
via a thickened anterior wall (possibly derived from adjacent leucocysts rather than the
chlorocyst itself). Dioecious. Fertile material rare in the species and unrecorded in subspecies
grandirete.

Distribution. The type subspecies of S. magellanicum is circumpolar in the northern hemisphere
and has a bipolar range in as much as it is also common locally in South America and on some of
the sub-antarctic islands. In Africa, the species appears to be represented only by subspecies
grandirete, which is apparently rare and local, being confined to the East African islands.

ALAN EDDY

ction of stem;

rtex; F, transverse se

icum Brid. subsp. grandirete (Warnst.) A. Eddy. A, branch fascicle; B, stem
ch co

ves; D, upper margin of stem leaf; E, bran

: C, branch lea
G, stem cortex (all drawn from the type collection of S. grandirete).

Fig. 3 Sphagnum magellan

leaves

A REVISION OF AFRICAN SPHAGNALES 89

Fig.4 Sphagnum magellanicum Brid. subsp. grandirete (Warnst.) A. Eddy. A, adaxial surface of branch
leaf; B, abaxial surface of branch leaf; C, transverse section of branch leaf (all drawn from the type).

East African Islands. MADAGASCAR: Androrangabe, Andovoranto, ‘ab indig. collect., 1901’ (PC,
herb. Renauld and herb. Paris — isotypes of S. grandirete).

S. magellanicum subsp. grandirete is superficially indistinguishable from S. perichaetiale and is
clearly rather closely related to the latter species. In addition to the immersed chlorocysts, which
provide the safest and most consistent differentiating character, the stem hyaloderm cells of S.
magellanicum frequently have two or more pores in the external wall (two-pored cells are not
unknown in S. perichaetiale, but they are exceedingly rare and confined to an occasional
leucocyst immediately adjacent to the insertion of a stem leaf or branch fascicle).

Typical S. magellanicum, where it occurs in temperate latitudes, is a clearly defined taxon
having considerable morphological stability. This stability tends to break down with tropical
latitudes, but probably not to such a degree that permits ready separation into definable
infraspecific taxa in most regions. Nevertheless, there are problems yet to be solved regarding
the status and relationships of the African plants of subsp. grandirete and other tropical
expressions of the species. Subspecies grandirete clearly links S. magellanicum with the more
primitive S. perichaetiale, and the Madagascan populations may be relict forms of an early
divergence of the two species.

90 ALAN EDDY
Subgenus ISOCLADUS

Sphagnum subgen. Isocladus (Lindb.) Braithw. in Mon. microsc. J. 14: 48 (1875).

Isocladus Lindb. in Ofvers. K. VetenskAkad. Foérh. Stockh. 19: 133 (1862). Type: Sphagnum macrophyl-
lum Brid.

Sphagnum subgen. Litophloea (Russow) A. Andrews in Bryologist 14: 73 (1911).

Very variable but cortical cells of stems and branches not fibrillose and very rarely endowed with
more than a single, unringed pore in the exposed walls; internal walls without large pores.
Branch hyaloderm dimorphic, typically with 1-4 much enlarged and often strongly protuberant
‘retort cells’ with the remaining cells unmodified and eforaminate. Branch leaves rarely with a
resorption furrow. Pseudolacunae, if present, poorly developed (only in some species of section
Acutifolia).

Section ACUTIFOLIA

Sphagnum sect. Acutifolia Wilson, Bryol. brit.: 20 (1855), excl. parte. — Schlieph. in Verh. zool.-bot. Ges.
Wien 15: 413 (1865). Type: Sphagnum acutifolium Schrader = S. capillifolium (Ehrh.) Hedw.
Sphagnum sect. Pycnosphagnum C. Mill. in Flora, Jena 70: 404 (1887).

Plants usually small to medium-sized; secondary pigments often pronounced, crimson or less
commonly orange or brown. Branches markedly dimorphic. Stem hyaloderm well developed,
3—4-layered, the outermost layer often foraminate. Retort cells of branch cortex commonly
solitary. Stem leaves usually strongly modified. Branch-leaf leucocysts strongly inflated, mainly
dorsiporose in mid-leaf but amphiporose in lower-lateral leaf areas; pores typically large and
distinctly ringed. Chlorocysts of branch leaves typically trapezoid in section, always displaced
towards, and with wider exposure on the adaxial leaf face.

Absent from Australasia, otherwise with a distribution more or less coincident with that of the
genus.

Although this section is one of the most diverse and ecologically important groups in the
temperate to subarctic northern hemisphere, it is represented by few tropical taxa, and in Africa
appears to be represented almost solely by a single, very variable but widespread species. In this
respect, the situation is analogous to that found in the Malaysian flora. However, whereas the
Malaysian S. junghuhnianum Dozy & Molkenb. is apparently related to S. subnitens and its
allies, the African S. violascens C. Mill. is much closer to S. russowii Warnst. and the American
species S. meridense (Hampe) C. Mill.

3. Sphagnum violascens C. Mill.
(Figs 5-8)

in Flora, Jena 70: 422 (1887). Type: Mozambique, M. B. de Carvalho (COI).

S. ericetorum Bescher., Fl. bryol. Réunion: 181 (1881), hom. illeg. (non S. ericetorum Brid., 1806).

S. obtusiusculum Lindb. ex Warnst. in Hedwigia 29: 196, tab. 4 fig. 8, tab. 7 fig. 13 (1890). — S.
obtusiusculum var. purpurascens forma brachy-dasyclada Warnst. in Hedwigia 29: 197 (1890). — S.
obtusiusculum var. pallescens Warnst., Sphagnol. univ.: 92 (1911).

S. acutifolium Schrader var. borbonicum Ren. & Cardot apud Warnst. in Hedwigia 29: 196 (1890), nom.
syn.

S. purpureum Schimper ex Warnst. in Hedwigia 29: 197, tab. 4 fig. 9, tab. 7 fig. 12 (1890).

S. borbonicum Warnst., Sammlung europdischen Torfmoose: 385 (1894); in Ren., Musci mascareno-
madagascarienses exs. 50 (1892), nom. nud.; Sphagnol. univ.: 90 (1911), nom. syn.

S. cordemoyi Warnst. in Hedwigia 36: 150 (1897).

S. rodriguezii Ren. & Cardot apud Warnst., Sphagnol. univ.: 90 (1911), nom. syn.

S. scotiae Cardot in Trans. R. Soc. Edinb. 48: 70, plate 1 fig. 1 (1911).

S. laceratum sensu Garside in JIS. Afr. Bot. 15: 71 (1949) (non S. laceratum C. Mill. & Warnst., 1897).

Plants very variable in-superficial appearance, small and compact to medium-sized and rather
lax, sometimes pale yellowish green but usually with some tint of crimson and, in open habitats,
commonly the whole plant deep wine-red. Fascicles usually rather closely set but, especially in
shade forms, sometimes rather distant, composed of (3—)4—S branches which are distinctly to

91

A REVISION OF AFRICAN SPHAGNALES

l
(above) and stem leaf

of upper part of branch leaf; B. adaxial su

of stem; D, branch leaves

C. Mull. A, adaxial surface

lower-median part of branch leaf; C, transverse sec

Fig.5 Sphagnum violascens

92 ALAN EDDY

——
ea“
SSS

Fig. 6 Sphagnum violascens C. Mill. A, branch fascicle; B, stem leaves; C, branch leaves; D, branch
cortex; E, abaxial (left) and adaxial (right) surface of stem leaf in weakly fibrillose region; F, apical
region of stem leaf; G, transverse sections of branch leaf (A, B, (two at left), D (left), E and G drawn
from the type of S. cordemoyi; the remainder from various sources).

A REVISION OF AFRICAN SPHAGNALES 93

yA

Fig. 7 Sphagnum violascens C. Mill. A, lower lateral area of stem leaf; B, branch leaves (left) and stem
leaf (right); C, adaxial surface of branch leaf (right) and abaxial surface (left) from sub-apical leaf region;
D, abaxial surface of branch leaf from lower-lateral leaf region (B drawn from the type of S. scotiae; the
remainder from Guillaumet 4200).

strongly dimorphic. Spreading branches short or much attenuated, 6-0—20-0 mm long; pendent
branches paler, narrow and caudiform, finely tapering, 10-0-25-0 mm. Stems 0-6—0-8 mm
diameter. Cortex very well developed of (2—)3—4 layers of thin-walled, highly inflated leuco-
cysts, the outer series commonly with solitary pores or thinnings; internal cylinder reddish
brown to violet-brown. Branch cortex dimorphic, retort cells solitary or in linear pairs (rarely
threes). Stem leaves lingulate, variable in attitude but mainly erect or erect-spreading 1-3-1-8
mm long, 0-6—0-8 mm wide: apex abruptly narrowed to a broadly truncate, dentate or more or
less erose-fimbriate tip, c. 150-300 um wide; border 2-4 cells wide above, distinctly to strongly
expanded below. Leucocysts variable; in upper part of leaf varying from strongly to weakly
fibrillose, or quite devoid of fibrils, strongly fibrillose tissues more or less identical to those of
branch leaves, otherwise largely resorbed adaxially and with few abaxial pores. Septa few to
numerous, especially in non-fibrillose states. Branch leaves suberect to erect-spreading, unrank-
ed or partially to wholly 5-ranked, numerous (60-80 per branch on average) the lower

94 ALAN EDDY

|

ff \
OURAN
AAI

Fig. 8 Sphagnum violascens C. Mull. A, adaxial leaf surface from subapical leaf region; B, adaxial leaf
surface, mid-leaf; C, transverse section of stem; D, abaxial (left) and adaxial (right) leaf surface, lower
mid-leaf; E, adaxial leaf surface, mid-leaf (all drawn from the type of S. obtusiusculum).

A REVISION OF AFRICAN SPHAGNALES 95

commonly with sub-tubular apices and appearing more or less acute, the more distal shallowly
concave and obviously rounded-truncate, (1-1—)1-3—1-6(-2-0) mm long, 0-6-0-9 mm wide.
Border narrow, 1-2 cells wide, without a resorption furrow. Lower pendent-branch leaves
similar to distal spreading-branch leaves, becoming narrowly lanceolate and delicate. Leuco-
cysts very variable in size, in apical part of leaf about 12-0-20-0 x 70-0-100 um, becoming
progressively larger and up to 35-0 x 200 um in lower-lateral parts of leaf. Abaxial surface in
mid-leaf with 1-5 medium-sized, ringed pores against the commissures, c. 8-0-15-0 um
diameter, which tend to form more or less well-defined ‘triplets’ at the basal angles; apical angle
frequently with a large resorption gap. Adaxial surface with or without 1-3 large, median,
circular, unringed or thin-ringed pores c. 26-0 um diameter. Leucocysts in lower and lateral parts
of leaf with more numerous, large pores on both sides. Leucocysts of pendent-branch leaves
with rather more frequent pores but otherwise identical to those of ordinary leaves. T.S. leaf.
Leucocysts shallowly convex on the adaxial side, much inflated abaxially. Chlorocysts in section
trapezoid with much wider exposure on the adaxial side. Dioecious? No fertile material seen or
reported.

Distribution. Uncommon to rare and of local occurrence in widely scattered localities down the
eastern side of the African continent from Mozambique to the Cape region; rather frequent to
locally common in the mountains of Madagascar and Réunion. The species is intolerant of long
periods of immersion and is to be found on wet, acid soils in open or lightly shaded heathland or
scrub vegetation.

Southern Africa. CAPE PROVINCE: Houtbay, Rehmann exs. 9b (H); Riversdale, Garcias Pass,
Thorne, 1926 (PRE); Sonderende, Wilman 527 (PRE); River Sonderende Mountains, 4000 ft (1220 m)
alt., Stokoe 9261 (PRE); Caledon Division, 1000 ft alt., ‘by stream, much shaded’, Adamson s.n. (PRE,
herb. BOL no. 24436); mountains above Hermanus, Vogelpoel 12794, 12795 (PRE); Fernkloof, above
Hermanus, ‘in bright red, rounded cushions by stream in fynbos’, Brenan M. 2783 (PRE); Cape, Zeyher
s.n. (BM).

Mlanje/Chimanimani. MALAWI: Mt. Mlanje, 2000 m alt., Wilman 1369/52 (BM). MOZAMBIQUE:
M.B. de Carvalho s.n. (COI — holotype of S. violascens).

East African Islands. MADAGASCAR: Pollen & Van Dam s.n. (BM, H — syntype collections of S.
obtusiusculum); Plateau d’Ikongo, Besson s.n. (PC); Ambohimatrara, Betsileo, Berthieu s.n. (PC);
Fianarantsoa, forest of Andrambovato, ‘on granite in sun’, 1000 m alt., abundant,.Descoings s.n. (BM);
Ambatoloana, 1350 m alt., Bosser 16197 (BM); Mt. Tsaratauana, 1200-2400 m alt., P. de la Bathie 187
(PC); North of Marojejy, Guillaumet 4065, 4186, 4200 (BM, NAM). MAURITIUS: Bory St. Vincent s.n.
(BM); Rodriguez s.n. (BM). REUNION: Cordemoy s.n. (PC, in herb. Bescherelle — type collection of S.
cordemoyi), Richard 683; Plaine de Mafate, Rodriguez s.n. (BM, PC — syntypes of S. obtusiusculum),
Bosser 11137 (BM); Plaine de Palmista, 1400 m alt., Bory s.n. (BM); Plaine des Cafres ‘thick moss cushion
on open grasplain (sic) common’, Schlieben 877 (PRE); Plaine des Cafres, Mare a Bone, 1584 m alt.,
Onraedt 69.R.0475; North of Fournaise, 1800 m alt., Gimalac 71.R.6535; Plaine des Palmistes, 1020 m alt.,
Gimalac 70.R.3507 p.p. (BM, NAM); Plaine des Chicots, 1400 alt., Onraedt 71.R.9251 (BM, NAM), De
Sloover 17256 (BM); Savane Mare a Bone, 1570 m alt., Onraedt 73.R.8057 (BM).

S. violascens is in many respects the African analogue of the boreal S. capillifolium (Ehrh.)
Hedw., to which it bears a strong superficial resemblance. It is equally variable in that it may be a
compact, densely branched plant or a lax one, and vary from an intense wine-red colour to pale,
straw-coloured or green, with little or none of the secondary crimson pigmentation in evidence.
Branch-leaf areolation resembles that of subgenus Sphagnum, but the resorption furrow is
lacking and the small leaves with truncate-dentate apices are more akin to those of the other
sections of subgenus /socladus. The wide leucocysts and large, ringed commissural pores are
distinctive, being larger than those of the S. capense group (up to 20 um or more), while the more
widely exposed chlorocysts, trapezoid rather than urceolate in section, prevent confusion with
S. ceylonicum. The stem leaves are very variable, especially in regard to the presence or absence
of fibrils. There does not seem to be any correlation between the degree of stem-leaf fibrillation
with other morphological features such as intensity of pigmentation or degree of branch leaf
ranking, so that there is no logical way in which the species can be meaningfully subdivided into
subspecies or varieties.

S. violascens seems to have no close relatives in Africa apart from S. reichardtii (see below)

96 ALAN EDDY

which is probably a vicariad, derivative taxon. On the other hand, S. capillifolium and S.
russowii Warnst. are obviously rather closely related northern species which would be difficult to
separate, at times, from the present plant if they were to occur together. S. violascens has similar
cortical pores and/or thinnings to those of S. russowii in the stem hyalodermis, and shows some
approach to the latter in the frequent presence of large adaxial branch-leaf pores in mid leaf. Its
stem leaves are much less stable in form, and are much narrower across the tips than those of S.
russowil. The American species, S. meridense (Hampe) C. Miill., is also rather close to both S.
violascens and S. russowii, differing from the former most obviously in having proportionately
wider leaves furnished with more numerous and rather smaller pores.

The present author’s view, that the phylogenetic derivation of Sphagnum subgenus Sphag-
num, is one with direct connections with Sphagnum section Acutifolia, has been stated already
(Eddy, 1977). In addition to certain structural modifications (e.g. the multistratose hyaloderm
and highly inflated leucocysts) there are marked similarities in the ecological preferences of the
two groups. Neither group contains species that occupy aquatic habitats (with the exception of
some boreal, probably derivative species in section Acutifolia) but instead contain most of the
characteristic ‘hummock formers’ familiar to temperate-zone ecologists. Perhaps therein lies the
explanation why two of the more widespread Sphagnum species, one from each group (i.e. S.
perichaetiale and S. violascens), are both absent from the East African afromontane regions,
although common in the East African islands, and occur in the Cape region. Assuming former
distributions to be more continuous, it is presumed that more ecologically sensitive species have
been eliminated from areas in which other taxa managed to survive. Aquatic modes of survival
are manifest among taxa of sections Subsecunda and Cuspidata, both of which are recorded from
regions where S. violascens and S. perichaetiale are unknown.

4. Sphagnum reichardtii Hampe ex Warnst.
(Fig. 9)
in Hedwigia 29: 206, tab. 5 fig. 17, tab. 6 figs 1-3 (1890).
? S. reichardtii Hampe in Reichardt in Fenzl, Reise Novara, Bot. Theil 1 (3): 166 (1870), nom. nud.

Plants lax and pale, apparently poorly developed and modified by habitat, having many reduced
stems and regrowths. Fascicles potentially with 2 spreading branches and 2 pendent branches up
to 16:0 mm long, weakly dimorphic. Stem pale to brownish; cortex (1—)2-3-layered, lacking
superficial pores. Branch cortex dimorphic with 1-2 relatively distinct and slightly protuberant
retort cells. Stem leaves flaccid, ovate-lingulate, on average 1-7 mm long, 0-85 mm wide, widest
at some distance above insertion. Leaf apex truncate and irregularly dentate. Border intact,
slightly expanded below. Leaf tissue characteristically multiseptate, each leucocyst subdivided
by numerous septa which in many cases are clearly derived from aborted chlorocysts. Branch
leaves rather thin-textured, lanceolate, up to 2:2 mm long, 0-7 mm wide, apices widely
truncate-dentate. Leucocysts variable, sometimes identical to those of S. violascens but many or
the majority aberrant in that they are subdivided by ‘septa’ like those of the stem leaves. T_S.
leaf. Chlorocysts rather thin-walled, trapezoid with distinctly wider exposure on the adaxial
side.

Distribution. Apparently endemic to St Paul Island.
East African Islands. ST PAUL ISLAND: Monte St Paul, 1100 [m alt. ?], G. de I’Isle 17 (Exped.
astronom. aux Iles St Paul et d’ Amsterdam 1874/75), (PC — type collection).

Section Acutifolia is rather poorly represented in the southern hemisphere and in looking for
relationships between S. reichardtii and other species of the section, S. junghuhnianum Dozy &
Molkenb. (Malaysia), S. meridense (Hampe) C. Mill. (S. America), and S. violascens (Africa)
are the reasonable possibilities. From a purely geographical standpoint, there seems to be little
doubt that S. reichardtii derives from the same stock as S. violascens. Its claims to specific status
are, however, rather more doubtful, and it is maintained as such here mainly because of the lack
of more positive evidence of identity with the latter species. Even where S. violascens is
common, in Madagascar for example, parallel modifications to a ‘S. reichardtii’ form have not

A REVISION OF AFRICAN SPHAGNALES 97

Fig.9 Sphagnum reichardtii Hampe ex Warnst. A, adaxial surface of branch leaf; B, branch leaves (left)
and stem leaf (right); C, apical region of stem leaf; D, adaxial surface of stem leaf; E, abaxial surface of
branch leaf, unmodified form; F, transverse section of leaf (all drawn from the type).

been collected. In this connection, however, it is interesting to compare the material of S. scotiae
(= S. violascens) from Ascension. It has much the same overall form and structure as S.
reichardtii (and had, in fact, been subsequently misidentified as S$. cuspidatum) but lacks the
characteristic subdivision of the leucocysts by which the present taxon is distinguished.

Warnstorf’s description (1890) of the species is based principally, if not exclusively, on the
collection of G. de l’Isle, whereas Hampe’s nomen nudum pertained to the earlier collections
made, presumably by Jelinek and Hochstetter, during the Novara expedition, 1857/59. It
appears that the latter material no longer exists.

5. Sphagnum fimbriatum Wilson
(Fig. 10)
in J. D. Hooker, Crypt. bot. antarct. voy.: 92 (1845). Type: Falkland Islands, J. D. Hooker (BM).

Plants elongate, rather small, green or pale straw-coloured; capitula with more or less
conspicuous, projecting stem-buds. Fascicles usually rather remote, composed of 4—5 distinctly
| dimorphic branches: 2(-3) spreading branches elongate, thin, finely tapering, 15-0—20-0
(—30-0) mm long; 2(—3) pendent branches very attenuated, pale, appressed to stems. Stems pale,
| yellowish to very pale brown, 0-7—0-9 mm diameter; cortex well developed, consisting of 2-3

98 ALAN EDDY

Fig. 10 Sphagnum fimbriatum Wilson. A, branch leaves; B, stem leaves; C, transverse section of branch
leaf; D, adaxial surface of branch leaf; E, abaxial surface of branch leaf; F, lower lateral region of branch
leaf, adaxial side; G, stem leaf tissue, upper mid-leaf; H, stem leaf tissue, lower mid-leaf; I, transverse
section of stem (all drawn from Wager 1057).

layers of thin-walled leucocysts, the majority of outer cells each having a large pore; internal
cylinder yellowish to yellow-brown. Branch cortex strongly dimorphic; retort cells solitary or in
linear series of 2(—3), slightly protuberant at apertures. Stem leaves erect and appressed to stems
(therefore virtually invisible in surface view), shortly and broadly spathulate, 0-9-1-4 mm long,
0-9-1-2 mm wide in upper half, delicate in texture, expanded and conspicuously resorbed,
fimbriate around the whole apical region; border strongly expanded below into broad patches of

A REVISION OF AFRICAN SPHAGNALES 99

prosenchymatous tissue, evanescent above mid-leaf; leucocysts greatly widened above, devoid
of fibrils and largely resorbed on both sides; septa numerous. Branch leaves rather small,
numerous, erect to erect-spreading, sometimes sub-squarrose, never 5-ranked, more or less
lanceolate, (1-1—)1-3—1-5(-1-9) mm long, 0-6-0-8 mm wide; apices truncate and 5—7-dentate;
border 2-3 cells wide, without a resorption furrow. Leucocysts very variable in size; in upper
mid-leaf relatively narrow, 15-0—20-0 x 90-0—120-0 um, in lower lateral parts of leaf much larger,
20-0-35-0 x 110-200 um. Abaxial surface with relatively few, medium to large ringed pores
mainly in the lower and lateral cell-angles, commonly with a large resorption gap in the apical
angle. Adaxial surface occasionally porose, normally with (1—)3-4(—7) large, unringed or
thin-ringed pores. Leucocysts of lower-lateral regions of leaf with numerous large circular pores
on both sides. Leucocysts of pendent-branch leaves all multiporose; pores very numerous on
abaxial surface, large (10-0-20-0 um), more or less in series along the commissures. T.S. leaf.
Leucocysts strongly inflated on abaxial side, less so to almost plane on adaxial side. Chlorocysts
more or less trapezoid with widest exposure on adaxial side. Monoecious (autoecious).
Antheridia borne towards distal ends of spreading branches; perigonial bracts yellowish. Inner
perichaetial bracts large and convolute, c. 4-5—6-0 mm long; apices retuse, not resorbed.

Distribution. Widespread in the temperate northern hemisphere and bipolar in that it also
occurs in temperate South America and some subantarctic islands. Known from only a single
locality in South Africa where it has possibly been introduced (Magill, 1981: 26).

South Africa. TRANSVAAL: Belfast, Wager 1057 (BM, PRE).

At one time, a record of this species from South Africa did not seem to be too unusual, since it
has a bipolar distribution. Many antarctic bryophytes that occur in Australasia and temperate
South America have also been found in South Africa, and there did not appear to be any logical
reason why Sphagnum should not share this pattern of distribution. However, no other antarctic
Sphagnum, even the widely distributed S. falcatulum Bescher., has been found in southern
Africa. Furthermore, a species of such fecundity and ecological tolerance as S. fimbriatum had it
survived in the region at all, should surely be more widespread than the solitary record indicates.
The suggestion of Magill (1981) that the species might have been introduced is therefore given
considerable support. Personal experience has shown that S. fimbriatum can, with ease, be
successfully cultivated in Britain, and fertile plants readily produced.

Section SUBSECUNDA

Sphagnum sect. Subsecunda (Lindb.) Schimper, Syn. musc. eur. 2nd ed.: 843 (1876). Type: Sphagnum
subsecundum Nees

A large and heterogeneous group that owes its recognition mainly to an absence of secondary
evolutionary characters. It contains taxa that appear to be relatively primitive, but also some
that connect with section Acutifolia and section Cuspidata.

Subsection SUBSECUNDA

| Variable but mainly rather large, often flaccid plants, with monomorphic or only weakly
dimorphic branches. Stem hyalodermis thin, clearly defined, of 1(—3) layers. Branch
hyalodermis sometimes only weakly dimorphic (especially in lower parts of branches), with only
slightly protuberant retort cells. Stem leaves usually smaller than branch leaves. Branch-leaf
| leucocysts usually long and narrow with very small, usually thick-ringed pores (commonly under
6-0 um diameter) scattered or in series along the commissures (free central pores not common).

_ 6. Sphagnum auriculatum Schimper
(Fig. 11)

_ in Mém. prés. div. Sav. Acad. Sci. Inst. Fr. 15: 80, tab. 24 (1857).
S. contortum Schimper var. algerianum (Cardot) Warnst. in Flora, Jena 67: 601 (1884).

Plants medium-sized, green overlaid with brownish and purplish pigments. Fascicles rather

100 ALAN EDDY

Fig. 11 Sphagnum auriculatum Schimper. A, branch fascicle; B, branch leaf; C, stem leaf; D, transverse
section of branch leaf; E, branch cortex; F, adaxial (left) and abaxial (right) surfaces of branch leaf; G,
transverse section of stem (all drawn from Cosson et al., Tunisia).

|

A REVISION OF AFRICAN SPHAGNALES 101

closely set or well spaced, composed of 3-4 more or less monomorphic branches, the weaker
branches paler and more or less pendent. Branches tumid, sometimes curved and cornute, up to
20:0 mm long. Stems 0-6-0-8 mm diameter; hyaloderm of a single layer of thin-walled,
non-porose leucocysts; internal cylinder pale brown to dark brown. Branch cortex indistinctly
dimorphic or, near bases of stronger branches, without differentiated retort cells; retort cells
otherwise in groups of 2-4, not protuberant at apices. Stem leaves lingulate to spathulate,
fibrillose in the upper half or beyond the middle; upper fibrillose tissues identical to branch-leaf
tissues; septa few, leaf apices concave, erose-dentate; border narrow 2-3 cells wide, not
expanded below. Branch leaves incumbent, weakly and inconspicuously 5-ranked, concave,
broadly ovate; apices truncate and c. 5—8-dentate; border 2 cells wide, without a resorption
furrow. Leucocysts proportionately narrow, 14-0—-18-0(—22-0) um wide, (90—)110—150 um long
in mid leaf. Abaxial surface with numerous to abundant, small, ringed pores c. 2-0-4-0 um
diameter, in series along the commissures; free central pores absent or very rare. Adaxial surface
with few or no pores. T.S. leaf. Chlorocysts more or less barrel-shaped to ovate-rectangular,
with oval lumina and thickened walls, more or less median but with a distinct tendency to be
more widely exposed on the abaxial side. Dioecious. Perigonial leaves typically orange-brown.
Inner perichaetial bracts 4-0-4-5 mm long, convolute, fibrillose in the upper 1/3-1/2. Fruit
unknown in Africa, frequent in Europe. Spores c. 32-34 um , coarsely papillose.

Distribution. A circumboreal species, very widespread and locally abundant in Europe;
confined, in Africa, to mountain ranges in Tunisia and Algeria, unrecorded south of the Sahara.

North Africa. TUNISIA: Kroumiria orientale, Bab-Abtk, Cosson, Barratte & Duvals.n. (BM, H, as S.
rufescens).

The description above is based on the few available collections of S. auriculatum from northern
Africa, and does not take into account the extreme range of forms that occur in its European
populations. As represented in African collections, the plant diverges very little from Schim-
per’s type and fits very well the author’s impression of an ‘average’ sample of the taxon. S.
auriculatum is very close, phylogenetically, to S. truncatum, the main difference lying in the
slight, but reasonably consistent, structural difference in the stem leaves and, in the case of S.
truncatum var. truncatum, the branch leaves. S. truncatum var. oligodon, which is superficially
indistinguishable from S. auriculatum, has stem leaves which are fibrillose to the insertion, anda
high proportion of leucocysts which are septate. S. truncatum var. truncatum has in general
proportionately narrower, more widely truncated and abaxially less porose branch leaves.

Under other circumstances, such small morphological differences might seem to provide
inadequate foundations upon which to maintain separate species. However, the ecological
barrier of the Sahara Desert and associated arid zones has prevented any possibility of
introgression and it is convenient, if not perhaps taxonomically consistent, to maintain S.
auriculatum and S. truncatum at species level.

The earliest valid name for this species has yet to be finally established. For example, S.
lescurii Sulliv. antedates S. auriculatum by a couple of years. However, there is at least one
earlier contender in the field (S. denticulatum Brid., Isoviita, pers. comm.) and this author
adheres to the familiar binomial until the nomenclatural tangle, unimportant as regards African
sphagnology, has been finally resolved.

7. Sphagnum truncatum Hornsch.
(Figs 12-15)

in Linnaea 15: 114 (1841). Type: S. Africa, Dutoitskloofberge, 3000 ft., Drége s.n. (BM — isotype).

S. coronatum C. Mill. in Flora, Jena 70: 412 (1887). —S. coronatum var. cuspidatum Rehmann in Warnst.
in Hedwigia 29: 184 (1890). — S. coronatum var. falcatum C. Mill. ex Warnst., Sphagnol. univ.: 306
(1911). — S. coronatum var. fluctuans Rehmann in Dixon & Gepp in Bull. misc. Inf. R. bot. Gdns, Kew
19: 195 (1923), nom. nud.

S. fluctuans C. Mill. in Flora, Jena 70: 414 (1887).

S. convolutum Warnst. in Hedwigia 29: 220, tab. 8 figs 10-12, tab. 10 fig. 6 (1890).

S. cuspidatum Hoffm. var. latetruncatum Warnst. in Hedwigia 29: 220 (1890), nom. syn.

Gl

Tine.

\ J
—

©)

RS

% hal |

y

Fig. 12 Sphagnum truncatum Hornsch. var. trun
(left) and adaxial (right) surfaces of branch leaf; D
bran

catum. A, branch fascicle; B, branch leaf; C, adaxial
af; D, transverse secti F, apex
of branch leaf (all drawn from Werdermann & Oberdieck 534).

ection of leaf; E, branch cortex; p

A REVISION OF AFRICAN SPHAGNALES 103

Fig. 13. Sphagnum truncatum Hornsch. A-C, branch leaves (above) and stem leaves (below) showing
variation in dimensions and degree of isophylly; D, transverse sections of stems. (A drawn from
Rehmann 9; B from the type collection of S. oligodon; C from Drége s.n.; D (left) from Mauve 5179 and
(right) Eddy s.n.).

S. hypnoides Schimper ex Warnst. in Hedwigia 29: 220 (1890), nom. syn.

S. oxycladum Warnst. in Hedwigia 30: 15, tab. 1 fig. 1, tab. 4 fig. a (1891).

S. rehmannii Warnst. in Hedwigia 30: 16, tab. 1 fig. 2, tab. 4 fig. b (1891).

S. marginatum Schimper ex Warnst. in Hedwigia 30: 28, tab. 2 fig. 20, tab. 4 fig. o (1891). — S. marginatum
var. convolutum (Warnst.) Warnst., Sphagnol. univ.: 309 (1911).

_ S. transvaaliense C. Mill. in Warnst. in Hedwigia 30: 32, tab. 2 fig. 22, tab. 5 fig. q (1891).

Plants extremely variable, flaccid to semi-rigid, medium-sized to rather large, pale green, deep
green to yellow-green, with or without orange, brown or, rarely, purple-brown secondary

104 ALAN EDDY

\. \ée
OMe
Fig. 14 A and D (upper fig.) Sphagnum truncatum Hornsch. var. truncatum. A, abaxial (right) and

adaxial (left) surfaces of branch leaf; D, (above) transverse section of leaf. B—D (lower fig.) S. truncatum

var. bordasii (Bescher.) A. Eddy. B, adaxial (left) and abaxial (right) surface of branch leaf; C, adaxial

(right) and abaxial (left) surface of branch leaf; D, (below) transverse section of leaf (A, D drawn from

Rankin 342; B—D from the type collection of S. mathieut).

pigmentation. Fascicles closely set to remote, composed of 2-4, rarely 5 branches which are only
slightly or not at all dimorphic. Branches abbreviated and tumid to elongated, flaccid and
long-tapering, (6-0—)10-0—20-0(—35-0) mm long. Stems normally brown, at least in part, but pale
in etiolated states, 0-5—O0-9 mm diameter. Cortex distinct, of basically a single layer of leucocysts;
sometimes with irregular duplication but never more than 25 per cent bistratose. Cortical pores
absent or rarely present but very faint. Internal cylinder weak or strong, typically brown.

A REVISION OF AFRICAN SPHAGNALES 105

Branch cortex indistinctly dimorphic. Retort cells often indistinguishable from other cortical
cells, especially towards base of branch, then many or all cortical cells uniporose; when distinct,
retort cells in linear series of 2-4, not or very slightly protuberant at apertures. Stem leaves
medium to large, 1-4-2-8 mm long, erect, spreading or more or less reflexed, lingulate,
spathulate to ovate-spathulate, fibrillose in the upper half, at least, and usually almost, or quite
to insertion. Hyaline cells commonly septate, in extreme forms with several septa per leucocyst,
in the upper half of leaf porose, the number and distribution of pores more or less as in branch
leaves. Apex variable: broadly truncate-dentate or partially resorbed and eroded. Border
narrow, 2-4 cells wide, not expanded below. Branch leaves large, relatively few per branch (i.e.
30-40 fully developed leaves, on average, compared to at least 40 in most species), basically, but
usually not obviously, 5-ranked. Leaves more or less uniform, or distal leaves longer and
narrower than lower leaves; very concave and more or less cochleariform, ovate, ovate-
lanceolate to more or less linear-lanceolate (distal leaves); 1-6-3-0 mm long. Apex broadly
truncate-dentate, sometimes very conspicuously so; border (1—)2 cells wide, without resorption
furrow. Leucocysts long and narrow to moderately wide, 20-0—30-0(—35-0) um wide, 110-160 um
long in upper mid leaf. Abaxial surface with few to very numerous, small, mainly ringed pores,
2:0-4-0(-6:0) um diameter, scattered or in series along the commissures. Adaxial surface
equally variable, with few to abundant pores, sometimes with more abundant adaxial than
abaxial pores. 7.S. leaf. Leucocysts more or less equally biconvex or with perceptibly more
inflated adaxial sides. Chlorocysts thick-walled and more or less barrel-shaped with oval lumina,
varying to rather thin-walled and more or less rectangular with oval-rectangular lumina, 5-0—
8-0 um wide, exposed more or less equally on both sides or with slightly wider abaxial exposure.
Dioecious. Male plants virtually indistinguishable from sterile plants, apparently very rare.
Female plants rarely collected, inner perichaetial leaves very large, 4-8—5-5 um long, convolute;
apices not resorbed, obtuse to minutely retuse; tissues fibrillose above, efibrose towards
insertion. Fruit rarely seen; spores c. 35-0 um, rugulose-papillose.

Distribution. From East Africa and the East African islands southwards to the Cape of Good
Hope. On irrigated, acid, mineral, or organic substrates from medium to high altitudes. See
under varieties below.

7a. Sphagnum truncatum var. truncatum
(Figs 12 & 14)

Plants usually soft or flaccid, rather large unless etiolated. Branch leaves often very widely ovate
and concave in lower parts of branches, but at least the median and upper leaves less concave,
ovate-lanceolate to lanceolate; apices conspicuously truncate and 8—15-dentate. Leucocysts with
few to several but not abundant nor serial pores on the abaxial side and commonly with more
numerous pores on the adaxial side. Stem leaves generally fibrillose to insertion, more or less
isophyllous and generally with few or no septa.

Distribution. Southern Africa: locally frequent from the western Cape to Transvaal and
southern Mozambique; absent further north except for a single record from Zaire/Rwanda.
This taxon, as far as collectors’ annotations indicate, is nearly always found in habitats which are
permanently wet, and at least a third of the specimens examined were partly or completely
submerged in acidic mires or streams.

Southern Africa. CAPE PROVINCE: Dutoits kloof, 3000 ft alt., Drége s.n. (BM -— isotype of S.
truncatum); Tafelberges [Table Mt.], MacOwan s.n. (PC — type collection of S$. convolutum); Montagu
Pass, Rehmann 9 (BM, H-SOL - isotypes of S. coronatum); Mountains above Worcester, Rehmann 10

_ (BM, H - isotypes of S. oxycladum); Sonderend, Breutel s.n. (BM, PC — isotypes of S. marginatum);

Gnadenthal, ‘in aquis fluitantib.’, Breutel 2 (BM — holotype of S. fluctuans); Giftberg, Werdermann &

_ Oberdieck 534 (BM); Rustenberg, Retrifs Kloof, Putherill 11035 (PRE); Table Mt., Kasteels Poort,

Pillans 4282 (BM); S.W. Cape, Groot Winterhoek, Osrien 34 (PRE); Fransch Hoek Pass, Schelpe 3878
(BM); Kirstenbosch, Schelpe 3857 (BM); Sandown Bay ‘open wet slope below 1000 ft’ alt., Levyns s.n.
(PRE-BOL no. 24066); Bain’s Kloof, Wilman 226 (BM); Kasteel’s Ravine, ‘sub-aquatic, 2300 ft alt.’,

106 ALAN EDDY

Esterhuysen 15970 (BM); Hex River Valley, ‘in running water’, Esterhuysen 15867 (BM); Stellenbosch,
Jonkers Hoek, Garside 74 (BM); Table Mt., Wager 11, 533 (BM); several collections from Table Mt. and
adjacent regions, including Rehmann 72, 294, Sim 9299. NATAL/KWAZULU: Umpumula, Ellen s.n.
(BM); Port St John, Wager 1091 (BM). TRANSVAAL: Lydenberg, Wilms s.n. (BM — isotype of S.
transvaaliense); Pure Krans, Mauve 5179 (PRE); Drakensberg, Graskop, submerged in water, Rankin
120, 342, 343 (BM); Pietersberg, Codd & Dyer 9147 (PRE); Waterberg, Germishuizen 943 (PRE); Belfast,
Wager 866 (BM). S. MOZAMBIQUE: Lourengo Marques, Balsinhas 1341 (PRE).

7b. Sphagnum truncatum var. bordasii (Bescher.) A. Eddy, comb. nov.
(Fig. 15)

S. bordasii Bescher. in Annls Sci. nat. (Bot.) VI, 10: 330 (1880). Type: Mauritius, Bordas (1879), (PC —
holotype, BM — isotype).

S. oligodon Rehmann ex C. Miill. in Flora, Jena 70: 412 (1887). — S. oligodon var. beyrichii Warnst.,
Sphagnol. univ.: 364 (1911). —S. oligodon var. bachmannii Warnst., Sphagnol. univ. : 363 (1911).

S. mathieui Warnst. in Bull. Soc. r. bot. Belg. 41 (1): 7 (1905).

S. mauritianum Warnst. sensu Warnst., Sphagnol. univ.: 306 (1911), nom. illeg., non S$. mauritianum
Warnst. (1891). — S. oxycladum Warnst. var. mauritianum (Warnst.) Warnst., Sphagnol. univ.: 306
(1911).

S. eschowense Warnst., Sphagnol. univ.: 328, fig. 36d, fig. 52D (1911).

Plants robust and often rather compact with tumid branches. Branch leaves all broad, very
concave, broadly ovate; apices truncate but less conspicuously so (mainly because of the
incurved upper leaf margins), 7—11-dentate. Leucocysts with numerous, more or less serial pores
along the commissures on the abaxial side; adaxial side variable, typically with few pores,
occasionally several, rarely with adaxial pores outnumbering the abaxial ones. Stem leaves
rather short, lingulate to ovate-lingulate, rarely triangular-lingulate, usually abundantly and
conspicuously septate. Apices dentate or rather cucullate and eroded.

Distribution. Southern Africa, southern Ruwenzori (Zaire/Rwanda), and the East African
islands. Generally more northern and eastern in range than the type variety, but overlapping the
range of the latter in southern Africa. Most records indicate a preference for acid mires, but the
plant is rarely, if ever, completely submerged.

Southern Africa. NATAL/KWAZULU: Eschowe, 600m alt., Hankon fils s.n. (H, PC — isotypes of S.
eschowense); Inanda, Rehmann 14 (BM - isotype of S. oligodon); Wager 580 (BM). PONDOLAND (=
TRANSKEI): Bachmann 5 (BM — H, type collections of S. oligodon var. bachmannii); Beyrich 25 (H —
isotype of S. oligodon var. beyrichii). TRANSVAAL: ‘in monte Kwatlamba . . .’, MacLea s.n. (H —
isotype of S. reamannii); Kaapsche Hoop, Wager 319 (BM).

Ruwenzori/Usambara. BURUNDI: ‘tourbiére d’Ijenda’, 1950 m alt., De Sloover 19/179 (BM, NAM).

East African Islands)s MADAGASCAR: Moroansitra, Mathieu s.n. (PC — type collection of S.
mathieui). MAURITIUS: Bordas s.n. (BM, PC - type collections of S. bordasii); between Savondronina
and Ramamasana, Besson (PC). REUNION: Bosser 9492 (PC).

S. truncatum is very closely related to S$. auriculatum Schimper and, indeed, some forms of the
former species could pass unremarked among collections of the latter if the two species ever
occurred together. Fortunately, the Sahara and surrounding arid zones have proven to be an
effective barrier to the transmigration of either species so that, although morphological
differences are slight, identification is relatively easy. S. truncatum var. truncatum is the more
divergent form, and has more widely truncated branch leaves than S. auriculatum; S. truncatum
var. bordasii has more extensively fibrillose and septate stem-leaf leucocysts than otherwise
similar states of the latter species. S. rutenbergii, endemic to the East African islands, also comes
very close to S. truncatum and may be difficult, at times, to separate from that species. It differs
in its basically 2-layered stem cortex (never less than 60 per cent 2—3-layered in fully developed
plants, in contrast to never more than 20 per cent 2-layered in S. truncatum), narrower
branch-leaf chlorocysts (c. 3-0-4-0 um, against 4-0-6-0 um or more in S. truncatum), and very
weakly developed secondary pigments. S. africanum differs in its eroded, rounded branch-leaf
apices and, typically, adaxially immersed chlorocysts. Free midline pores are common in typical
S. africanum, but absent or very rare in the leaf leucocysts of S$. truncatum.

A REVISION OF AFRICAN SPHAGNALES 107

Fig. 15 Sphagnum truncatum Hornsch. var. bordasii (Bescher.) A. Eddy. A, branch fascicle; B, stem
leaf; C, branch leaf; D, branch cortex; E, transverse section of stem; F, transverse section of branch leaf
(drawn from the type of S. bordasii).

The rather large number of synonyms of S. truncatum reflects, not only the relatively wide
range and frequency of the taxon, but also its unusually high degree of polymorphism. In this the
species parallels the protean qualities of its closest relative, the boreal S. auriculatum. The
situation is well exemplified by the extremes treated here as the varieties S. truncatum var.
truncatum and S. truncatum var. bordasii, which may be considered to be analogous to the

108 ALAN EDDY

‘crassicladum’ and ‘auriculatum’ expressions of S. auriculatum sens. lat. in Europe. It is
emphasized that the two varieties described above are connected by a range of forms that is
virtually a cline, and that the division of the species into two varieties is an arbitrary measure that
yields a simplistic picture. While recognition of the major variants is felt to be desirable, further
subdivision would have to take into account a great number of apparently random and
inconstant morphological details.

In ecological behaviour as well as in morphological details, S. truncatum is extremely close to
S. auriculatum. Under different circumstances, it would be tempting to reduce the species to
subordinate rank, i.e. at subspecies, to the latter species. However, there are certain drawbacks
to this approach. S. auriculatum has not been recorded from any other part of the southern (i.e.
Gondwanan) hemisphere, which means that its occurrence in Africa, if not a relic of a
Gondwanan element, would have to be a more recent incursion from the Laurasian flora. If the
latter were true, bearing in mind that modern geologists are agreed that no land connection
existed between Africa and Laurasia until the closure of the Tethys around about the
mid-Miocene, then S. truncatum could surely only have attained its present range and poly-
morphism after many generations of vigorous outbreeding and successful colonisation. If this
were the case, then it is difficult to reconcile such theoretical past performance with its present,
apparently almost universal sterility. Many of the variants appear to be very local in occurrence,
consistent with mainly vegetative propagation. Whether or not such variants are given varietal
names is largely a matter of choice (in some ways, they might be equated with the ‘microspecies’
of apomictic phanerogams), but it is clear that S. truncatum, as a species, exhibits the
characteristics of an ancient relic rather than an expanding colonist.

Some tropical American taxa are undoubtedly closely related to S. truncatum, but their
precise relationships are not yet resolved. For example, the Brazilian species S. cucullatum
Warnst. seems to be scarcely distinct, and other described taxa no doubt exist. Final resolution
of the systematic positions of S$. truncatum, S. auriculatum, and S. rutenbergii in relation to one
another, and to the American plants, may have to depend ultimately on biosystematic and
chemotaxonomic investigations.

The almost endemic southern African range of S. truncatum var. truncatum, in contrast to the
overlapping but basically northern and eastern range of S. truncatum var. bordasii, could
support their separation at a level above ‘variety’. It would be comparatively simple to
enumerate characters which easily separate East African island from aquatic southern African
forms. However, variation in the various characters is virtually random among southern
populations, and ecological data, where available, indicate that ‘truncatum’ characters, while
not necessarily produced entirely by, are nevertheless much exaggerated by, an aquatic mode of
existence. Similarly, subaerial habitats with high atmospheric humidity seem to encourage the
development of the ‘bordasii’ facies. It is admitted, however, that cultivation experiments on a
collection of S$. truncatum var. truncatum (Rankin 342) were only marginally successful, and to
date no unequivocal ‘truncatum’ forms have been found among Madagascan collections.

8. Sphagnum rutenbergii C. Mull.
(Figs 16-19)

in Abh. naturw. Ver. Bremen 7: 203 (1881). Type: Madagascar, Forest of Ambatondrazaka, Rutenberg
(1877) (B, destroyed). Neotype: Madagascar, Imerina, mixed with S. hildebrandtii and S. madegassum,
Hildebrandt (1880) (BM, PC).

S. aequifolium Warnst. in Hedwigia 30: 22, tab. 1 fig. 9, tab. 4 fig. 1 (1891).

S. mauritianum Warnst. in Hedwigia 30: 17, tab. 1 fig. 3, tab. 4 fig. c (1891), non S. mauritianum Warnst.
sensu Warnst., Sphagnol. univ.: 203 (1911).

S. obovatum Warnst. in Hedwigia 30: 18, tab. 1. fig. 4, tab. 4 fig. d (1891).

S. pallidum Warnst. in Hedwigia 30: 170, tab. 19 fig. 30, tab. 24 fig. n (1891).

S. salvanii Warnst. in Hedwigia 47: 119 (1908).

S. flavirameum C. Mill. ex Warnst., Sphagnol. univ.: 381 (1911), nom. syn.

Plants medium-sized (resembling pale forms of S. truncatum) whitish green, with or without
tinges of pale brown or yellowish brown. Fascicles distant or closely packed, composed of

A REVISION OF AFRICAN SPHAGNALES 109

Fig. 16 Sphagnum rutenbergii C. Mill. A, branch fascicle; B—C, branch leaves; D, stem leaves; E,
transverse sections of branch leaves; F, transverse section of stem; G, branch cortex (C drawn from a
| submerged plant; the remainder from the type collection).

(2—)4-5 branches which are not, or slightly dimorphic. Branches rather tumid or elongated and
tapering, arched or somewhat laterally curved and cornute, with 40-60 or more fully developed
leaves on stronger branches, 12-0—25-0 mm long. Stems rather strong, 0-7-1-0 mm diameter;
cortex well developed, composed of (1—)2-3 layers of thin-walled hyaline cells, the outer series

110 ALAN EDDY

It)

ee.
Ne ee
ih

\ st

ta
iV

eal
To
& h
ened
—
ro}

aS
Wi
W\

(cGAR
CI

ern
aad

REL
ane

=
8
ote

\Ls
es

a, |
Fig. 17 Sphagnum rutenbergii C. Mull. A, branch fascicle; B, branch leaves; C, stem leaf; D, stem-leaf
areolation; E, transverse section of leaf; F, branch cortex; G, adaxial (left) and abaxial (right) surfaces of

branch leaf; H, transverse section of stem (all drawn from the type collection of S. obovatum).

frequently with faint pores or thinnings; internal cylinder rather well developed, pale except for
a narrow band adjacent to the hyaloderm which is usually a rich red-brown. Branch cortex
weakly to moderately dimorphic, retort cells in linear groups of 2-4, sometimes not differenti-
ated in the lower parts of branches. Stem leaves ovate-lingulate to ovate-spathulate, large,
1-8—2-8 mm long, 0-8—1-2 mm wide; apex concave, more or less cucullate; border narrow, 2-4

a

A REVISION OF AFRICAN SPHAGNALES 111

Fig. 18 Sphagnum rutenbergii C. Mill. A, stem leaf; B, branch leaf; C, adaxial (left) and abaxial (right)
surfaces of branch leaf; D, transverse section of stem; E, branch fascicle; F, low-magnification figures of
stem sections (A—E drawn from the type collection of S. pallidum; F from a variety of sources).

ole ALAN EDDY

Fig. 19 Sphagnum rutenbergii C. Mull. A-D, adaxial (left figures) and abaxial (right figures) surfaces of
branch leaves (A drawn from Bosser 8275; B from Onraedt 71.R.1039; C from the type collection of S.
aequifolium; and D from the type collection of S. salvanii).

cells wide, not expanded below; tissues fibrillose throughout and more or less identical above to
branch-leaf tissues. Branch leaves medium-sized to large, incumbent, basically 5-ranked, ovate,
(1-4-)1-6-2-0(—2-8) mm long, 0-9-1-2 mm wide; apices truncate and 6—-9-dentate, relatively
narrow in proportion to the leaf dimensions; border narrow, 2 cells wide, without a resorption
furrow. Leucocysts rather wide, 20-30 um X 110-170 um in mid leaf. Abaxial surface with
numerous to abundant small, ringed pores c. 3-0—5-0 um diameter scattered or in series along the

A REVISION OF AFRICAN SPHAGNALES 113

commisures, occasionally with 1-4 additional pores near the cell midline. Adaxial surface
without pores or with a few scattered pores mainly in, or near, the apical and upper-lateral cell
angles. T.S. leaf. Leucocysts more or less equally convex on both sides. Chlorocysts very
narrow, mainly under 5-0 um, narrowly oval-rectangular with oval lumina and incrassate abaxial
and adaxial walls, more or less median and symmetrical with narrow (c. 2:0 um) exposure on
both sides of leaf. Dioecious ? No male plants seen. Female plants seen but only in an
unfertilized condition.

Distribution. Growing in pure or mixed low mats or hummocks in marshy ground between 1500
and 2500 m alt. Endemic to the East African islands.

East African Islands. MADAGASCAR: Imerina, Hildebrandt 2106p.p. (BM, PC — neotype collec-
tions, in mixture with S. tumidulum and S. planifolium); same region, Hildebrandt s.n. (BM, PC — type
collections of S. aequifolium); Amboritro, Amkatokapaitra, 1300-1400 m alt., Salvan s.n. (PC — type
collections of S. salvanii); ‘. . . ohne naheren Standort . . .’ (BM, herb. Mitten — type collection of S.
obovatum); Massif d’Andringitra, 1600-2500 m alt., P. de la Bathie 44 (PC); Tampoketsa, 1600 m alt.,
Humbert s.n. (BM); Massif Andohahelo, Humbert s.n. (BM, PC); Ambositra, Mt. Antety, 1400-1850 m
alt., Humbert & Swingle s.n. (BM); Ambatotoana, 1400 m alt., Bosser 9901, 15888 (BM, PC);
Ambatolaona, Benoist 328 (PC); Ankafana, Cowan 23 (BM); Tananarive, Lake Manasoa, 1400 m alt.,
Onraedt 74.M.2105 (NAM); Ombadirafia, Guillaumet 2202 (BM, NAM). MAURITIUS: Dr Ayres s.n.
(BM - type collection of S. mauritianum). REUNION: Rodriguez 10 (PC-—type collection of S. pallidum);
Plaine des Chicots, De Poli 23 (PC); Salazie, Rodriguez s.n. (PC); Maryakandriana, edge of swamp,
1300 m alt., Bosser 8275 (BM); Plaine des Palmistes, Onraedt 73.R.8089, 73.R.8195 (BM, NAM); Plaine
des Cafres, 1584 m alt., Onraedt 69.R.0702 (BM); Cirque de Mafate, 1600 m alt., Onraedt 71.R.1039
(NAM, BM); Chauvet s.n. (PC).

The secondary pigments present are never intense. This feature, combined with the exception-
ally narrow chlorocysts, gives the plant a rather ‘washed out’ appearance. Well developed
specimens have long branches carrying many more leaves than the branches of S. truncatum
sens. lat., which S. rutenbergii otherwise resembles. Weaker, presumably semi-aquatic phases of
S. rutenbergii, however, may be virtually superficially indistinguishable from pale states of S.
truncatum. The safest character for separation rests on the stem section. The cortex is rather
irregular in development but clearly 2-3 layered for the most part, and never more than 40 per
cent 1-layered.

S. mauritianum was described by Warnstorf in the same work in which he described S.
pallidum, basing the former on a specimen in Mitten’s herbarium which had been collected by
Dr Ayres in Mauritius. When he reduced S. mauritianum to a variety of S. oxycladum Warnst.
(= S. truncatum Hornsch. var. bordasii), Warnstorf did so with reference to a quite different
plant that had been collected in Madagascar by Besson. The latter collection is possibly
conspecific with S. truncatum, but the plant from Mauritius is without doubt the same as S.
pallidum (= S. rutenbergii).

Although Miller’s original specimen was destroyed, and no duplicates can be traced, there is
little uncertainty about the species he described. The selected neotype collection, represented in
BM, PC and possibly other herbaria, is part of a mixed gathering examined by Miller and
comprising three species: S$. madegassum C. Mill. (= S. planifolium), S. hildebrandtii C. Mull.
(= S. tumidulum), and a plant which he recognized as his S. rutenbergii. The last named is clearly
the same taxon as the later described S. pallidum and S. mauritianum of Warnstorf.

9. Sphagnum africanum Welw. & Duby
(Figs 20 & 21)

in Duby in Mém. Soc. Phys. Hist. nat. Genéve 21: 216, tab. 1 fig. 1 (1870). Type: ‘Ad rivulorum latera in
summis provinciae Huilla regni Angolensis ad 5300-5500 pedes alt.’, Welwitsch (BM).
S. angolense Warnst. in Bull. Herb. Boissier II, 1: 1086 (1901).

Medium-sized, rather lax plants varying from dull green tinged with brown to more or less
completely orange-brown. Fascicles commonly composed of only 1-2, never more than three
branches. Branches tumid not contorted, up to 13-0 mm long but usually shorter than 10-0 mm,

ALAN EDDY

Fig. 20 Sphagnum africanum Welw. & Duby. A, branch cortex; B, adaxial (left) and abaxial (right)
surfaces of branch leaf; C, branch leaves; D, transverse section of stem; E, transverse section of leaf (all
drawn from the type).

not at all dimorphic. Stems generally rather weak, 0-3-0-6 mm diameter. Cortex mainly
1-layered but with a marked tendency towards local duplication and up to 50 per cent 2-layered,
exposed cell walls without pores. Internal cylinder varying from pale yellow-brown to deep
red-brown. Branch cortex with only slight differentiation of retort cells, most of the ordinary
cortical cells having an apical pore. Stem leaves large to very large, 1-9-2:8 mm long, 1:2-1-3 mm
wide, always larger than branch leaves, ovate to ovate-lingulate, widest at some distance (c. 1/4 to
1/3) above insertion, fibrillose throughout and virtually identical to branch leaves in structure.
Border narrow, not widened below; apex concave and eroded. Branch leaves (1-2-)1-4—
1-9(-2-1) mm, widely ovate, concave, symmetrical, erect. Border narrow, without a resorption
furrow. Apex broadly rounded, eroded, sub-cucullate, never truncate-dentate. Leucocysts
relatively wide (except in poorly developed plants), about 18-30 x 110-130 um in upper
mid-leaf. Abaxial surface with abundant small, ringed pores (occasionally partly replaced by
pseudopores) in series along the commissures, sometimes with additional pores (ringed or
unringed) in the cell midline; pores (i.e. apertures) 2-0-4-0 um diameter on average. Adaxial

A REVISION OF AFRICAN SPHAGNALES 115

Fig.21 Sphagnum africanum Welw. & Duby. A, transverse section of stem; B, branch cortex; C, adaxial
(left) and abaxial (right) surfaces of branch leaf; D, branch fascicle; E, transverse section of leaf; F,
branch-leaf apex (D drawn from the type collection; the remainder from Townsend 75/261).

surface variable, generally with a few scattered small pores but few to numerous pseudopores,
especially in the upper half of the leaf. T.S. leaf. Leucocysts biconvex, usually more inflated on
the abaxial side. Chlorocysts variable; in weak specimens more or less oval-rectangular and
exposed subequally on both sides of leaf, typically distinctly displaced towards the adaxial face,
with very narrow abaxial exposure and here and there appearing immersed. Fertile material not
seen.

Distribution. Scattered occurrences in montane regions in a zone across southern-central Africa
from Huila Province (Angola) to Malawi.

West Africa. ANGOLA: Huila, Welwitsch, Iter Angolense 11 (BM, L, PC—type collections); Lubango,
Borges 249 (BM).

Mlanje/Chimanimani. MALAWI: Mlungasi Falls, 1700 m alt., Hornby 6 (BM). ZAMBIA: Shiwa
Ngandu, ‘moorland drainage in peat bog, common’, 1500 m alt., Greenway 5501 (BM, PRE); Mwinilunga,
1350 m alt., Townsend 75/193 (BM). ZIMBABWE: Mt. Nuza, ‘damp hollows between grass clumps’, 1700
m alt., Gilliland 509B (BM).

116 ALAN EDDY

Ruwenzori/Usambara. ZAIRE: Katanga, Overlaedt, Bryotheca R. Naveau B2, B3, B4 (H); Shaba
Province, 1700 m alt., Townsend 75/261, 75/267 (BM).

The small, serially disposed pores and basically single-layered stem cortex of this taxon suggest
a relationship with S. truncatum. However, there is little risk of confusion between the two
species, the morphology of the leaf apices being quite different and sufficiently stable to
facilitate discrimination between them. Smaller examples of S. africanum can superficially
resemble species of the S. capense group in colour and external features, but lack the highly
developed cortex and branch-retort cells of the latter group.

Warnstorf had not seen Welwitsch’s gathering when he described S. angolense, and seems to
have been influenced by Duby’s misguiding allusion to similarities between S. africanum and S.
pylaesii.

Those plants with comparatively small (c. 1-6-1-8 mm) branch leaves and distinctly adaxially
biased chlorocysts should be considered to represent the normal expression of the species.
Specimens with more or less symmetrically disposed chlorocysts with subequal adaxial and
abaxial exposure may be expressions of the species in response to immersion or shading. In such
widely dispersed populations, however, a degree of infraspecific, genetically based variation is
to be expected in this, as in many other African taxa.

A broader interpretation of S. africanum is given by Magill (1981). A number of names are
drawn into synonymy with this species which, in the present work, are considered to be aligned
more appropriately with S. truncatum. Thus S. rehmannii Warnst., S. transvaaliense C. Mill., S.
oligodon, var. bachmanii Warnst., and S. oligodon var. beyrichii Warnst. are transferred from
synonymy with S. africanum to that of S. truncatum Hornsch. sens. lat.

Subsection ACROSPHAGNUM

Sphagnum sect. Subsecunda subsect. Acrosphagnum (C. Mill.) A. Eddy, stat. nov. — Sphagnum 6:
Acrosphagnum C. Mill. in Flora, Jena 70: 405 (1887).

Sphagnum [sect.] Mucronata Warnst. in Hedwigia 30: 127 (1891). Lectotype: S. mucronatum C. Mill. = S.
tumidulum Bescher.

Plants small to medium-sized, lax or dense, yellow-green to brownish, sometimes with dull
purple or red undertones. Stems typically brownish; hyaloderm very well developed, of 2-3
layers of thin-walled leucocysts, the outer much larger than the inner and usually with an
unringed large pore in the outer wall. Branch cortical cells usually markedly dimorphic, retort
cells in linear groups of 2-3, usually strongly protuberant at apex. Branch leaves usually very
regular in size, all except a few basal ones ovate, concave, distinctly 5-ranked, abruptly
narrowed to mucronate, pseudomucronate or eroded apices. Stem leaves very variable but
typically as large as or larger than branch leaves; fibrillose in the apical region, commonly
throughout, seldom efibrose.

Branch-leaf leucocysts normally relatively short and wide, 25-0-35-0 x 70-0—-100 um long,
pores variable in size and distribution. Chlorocysts with rounded lumina in cross section,
variable in disposition relative to the leaf surfaces.

This group is principally African, with one species occurring also in Sri Lanka. It grades into
subsection Subsecunda, and some non-African taxa could be construed as intermediate between
the two groups (e.g. S. ovatum Hampe of N.E. India, Thailand, and Taiwan; S. gracilescens
Hampe ex C. Mill. of tropical South America). There are no obvious intermediates in the
African flora, although S. africanum, in its small-leaved expressions, exhibits features that could
point to a common origin (see pp. 113-116).

This subsection contains plants that are rather uniform in external appearance but exhibit a
wide diversity in the morphology of their branch-leaf cells. The type species, S. tumidulum
Bescher., is the most divergent, and were it not for the existence of a number of connecting
forms (i.e. S. pycnocladulum var. ericetorum, S. tumidulum var. confusum and S. tumidulum
var. tumidulum) might well be placed in a section of its own. The small, wide, uniform and
numerous branch leaves, often markedly 5-ranked, give rise to plants with a characteristically
neat appearance, reminiscent of the boreal S. tenellum (Brid.) Brid. The hyaloderm of the stem,

A REVISION OF AFRICAN SPHAGNALES IAL,

coupled with the large stem leaves, are distinctive features, while branch-leaf areolation can
approach that of section Acutifolia with regard to width and pore size. However, the latter
features can vary widely, even in a single gathering, and repressed forms (e.g. from unfavour-
able habitats) and less evolved species (e.g. S. davidii) can have narrow, small-pored leucocysts
that are virtually identical to those of subsection Subsecunda. Variability in chlorocyst position,
which is an unusual feature within a section, led to widely differing systematic interpretations by
other authors. This has been discussed at length elsewhere (Eddy, 1979).

Within the subsection, S. davidii is the African species which shows least divergence from
subsection Subsecunda, while the existence of the extreme form, S. tumidulum, suggests that it is
a taxon of considerable age. This view can be reinforced if the morphological features of S.
ceylonicum (p. 135) are accepted to indicate common ancestry with section Acutifolia. S.
tumidulum itself may have had a pre-mid-Cretaceous origin (see above, p. 79) but section
Acutifolia is probably even older. If the suggested phylogenetic interpretation is valid, then
subsect. Acrosphagnum must be a ‘pre-Drift’ taxon.

Where Miller (1887, loc. cit.) briefly described the infrageneric group ‘6. Acrosphagnum’, he
appended the term ‘Sphagna mucronata’. Clearly the epithet ‘mucronata’ was intended to be
descriptive, not nomenclatural, indicating that the author considered S. mucronatum (= S.
tumidulum) to be morphologically representative.

Warnstorf (1891, 1900, 1911) gave no explanation for his selection of the name ‘Mucronata’,
which he ascribed erroneously to Miller, in place of “Acrosphagnum’. Van der Wijk et al (1967)
also seem to be in error where they indicate that Muller himself (‘l.c.’) reduced [sect.]
Acrosphagnum to synonymy.

Since “Acrosphagnum’ is validly described and unambiguously typified, I can find no good
reason for its rejection in favour of Warnstorf’s choice of name.

10. Sphagnum tumidulum Bescher.
(Figs 22 & 23)

in Annis Sci. nat. (Bot.) VI, 10: 329 (1880). Lectotype: Réunion, Richard s.n. (PC).

10a. Sphagnum tumidulum var. tumidulum
(Fig. 22)
S. tumidulum var. macrophyllum Warnst. in Hedwigia 30: 128 (1891).
S. tumidulum var. microphyllum Warnst. in Hedwigia 30: 128 (1891).
S. aculeatum Warnst. in Bot. Zbl. 9: 97 (1882).
S. hildebrandtii C. Mill. in Flora, Jena 70: 420 (1887).
S. mucronatum C. Mill. in Flora, Jena 70: 421 (1887).
S. imbricatum Schimper ex Warnst. in Hedwigia 30: 128 (1891), nom. syn.
S. submucronatum C, Mill. ex Warnst., Sphagnol. univ.: 172 (1911), nom. syn.
S. pugionatum C. Mill. ex Warnst., Sphagnol. univ.: 172 (1911), nom. syn.

Plants small to medium-sized, compact or occasionally rather lax; light yellow-brown or
yellowish-ochre, rarely green. Fascicles closely set, rarely somewhat distant, composed of
(4-)5-6 monomorphic branches (if 6, then one branch usually very weak). Branches very
uniform in structure and form but variable in length, (4-0—)6-0-15-0(—20-0) mm long, of uniform
diameter, not tapering. Stems normally relatively strong, 0-6-1-0 mm diameter, pale to mid
brown. Cortex well developed, basically 2-layered with an outer, highly inflated series of
leucocysts much larger than the inner; most outer cortical cells with a large pore in the exposed
wall. Internal cylinder well developed but individual cells not grossly thickened; walls of outer
layers yellow-brown to deep brown. Median parenchyma yellowish or colourless, but sometimes
with a group of cells in the middle having dark brown walls, resembling a false central strand.
Branch cortex dimorphic: retort cells highly inflated, mainly in linear pairs, slightly to moderate-
ly protuberant at apertures. Stem leaves relatively large, 1-4-2-0 mm long, 0-7—0-8 mm wide,
lingulate to triangular-lingulate, suddenly contracted above to mucronate tips. Border strong,
4-6 cells wide, not or very indistinctly widened below. Hyaline cells always fibrillose in the upper
half of leaf but fibrils often incomplete, rarely fibrillose to insertion; fibrillose tissue more or less

ALAN EDDY

sé

Sk ie

Vel
S-

o™

oS Yee
ZR k
A “>

Fig.22 Sphagnum tumidulum Bescher. var. tumidulum. A, branch fascicle; B, transverse section of stem;
C, branch leaves; D, stem leaves; E, branch cortex; F, transverse sections of leaf; G, adaxial (left) and

abaxial (right) surfaces of branch leaf (all drawn from Richard s.n.).

A REVISION OF AFRICAN SPHAGNALES 119

Y
SARK
LEO
SOREE RS
SRR
RR

ron
LO
ES
rar
OO
OLD

2%,
OO
SOQ
xz

Fig. 23. Sphagnum tumidulum Bescher. var. confusum A. Eddy. A, branch cortex; B, branch leaves; C,
stem leaves; D-E, adaxial (right) and abaxial (left) surfaces of branch leaves; F, transverse section of
stem; G, transverse sections of leaves (C, right fig, and E drawn from Bosser 11369; the remainder from
Bory).

120 ALAN EDDY

identical to branch-leaf tissue. Branch leaves very uniform in size and direction, strictly 5-ranked
in straight rows so that branches appear prismatic in form, small, 1-0-1-6 mm long, ovate, very
concave; apices acute or more or less sharply mucronate with strongly inflexed or ‘pinched in’
margins, apical resorption absent or very weak; border 2-3 cells wide. Leucocysts short and
relatively wide, 70-90 um long, 25-35 um wide; commissural walls of adjacent leucocysts
fused throughout, the common walls generally thin. Abaxial surface with thin fibrils and
scattered to serial pseudopores, with nodular thickenings at the anastomoses so that the cell
surfaces appear reticulate rather than fibrillose; pores absent or few, minute, ringed, circular,
remote from commissures, c. 2-0 um diameter. Adaxial surface similar to abaxial surface but
usually with less numerous pseudopores; pores 1-6 per cell, minute, ringed, remote from
commissures, c. 2:0 um diameter. T.S. leaf. Leucocysts almost plane on both sides, frequently
with thin internal, longitudinal partitions in addition to the poorly developed transverse, partial
septa. Chlorocysts thin-walled, broadly oval to almost circular, deeply immersed and remote
from both leaf surfaces; internal commissural walls smooth.

Distribution. Endemic to the East African islands, in a variety of terrestrial habitats, mainly at
altitudes above 1000 m.

East African Islands. MADAGASCAR: Imerina, Hildebrandt s.n. (BM, PC, mixed with S. pallidum,
source material of S. aequifolium and S. hildebrandtii); between Vinanentelo and Ikongo, Bessons.n. (PC);
Massif Tsaratanana, siliceous soil in humid forest, 2000 m alt., Humbert s.n. (PC); Massif d’Andringitra,
bog hollows at 2000 m alt., Humbert 51,52 (PC); Ankaratra, below Manjakatampo, 2200 m alt., at edge of
stream, Bosser 15048 (PC); Ivohibe, siliceous rocks at 1500-2000 m alt., Humbert 60 (BM); Massif
Beampingaratra, summit of Bekoho, 1500 m alt., Humbert s.n. (BM); Fianarantsoa, Bosser 17.801 (BM,
NAM). MAURITIUS: ‘Tourbiére a Philippia’, 600 m alt., Onraedt 71.Ma.237 (NAM). REUNION:
Lépervanche (PC - lectotype of S. tumidulum); Richard (PC — isosyntype); Rodriquez s.n. (BM); Bory de
St. Vincent s.n. (BM); Plaine des Cafres, 1550 m alt., Onraedt 69.R.103, 69.R.0473 (BM, NAM);
Commune de St Philippe, bog at 1900 m alt., Gimalac s.n. (BM); Cirque de Cilaos, bog at 2300 m alt.,
Gimalac s.n. (BM); river valley north of the Fournaise, bog at 1800 m alt., Gimalacs.n. (BM, NAM, herb.
Onraedt no. 71.R.6536), also from more or less the same region, several other specimens collected by
Gimalac, in NAM and BM under Onraedt herbarium numbers, e.g. 71.R.8505, 71.R.8531; Plateau des
Chicots, 1200 m alt., Onraedt 71.R.9252 (BM, NAM).

The typical, and by far the commonest form of S. tumidulum, is a plant that can be identified
almost at a glance. The numerous, closely set, very uniform, mucronate and strictly 5-ranked,
rather patent and concave branch leaves result in a facies that is highly characteristic and
matched only by some robust states of a few other species of subsection Acrosphagnum which
are rare or absent in the East African islands. The peculiar, reticulate appearance of the
branch-leaf leucocysts may be encountered in some states of S. pycnocladulum but these differ
in the numerous, usually serial, pores on the abaxial leaf surfaces. S. ericetorum has adaxially
seriate pores, but exposed chlorocysts.

S. tumidulum has been recorded from time to time from mainland Africa. With one
exception, all such records have been based on misidentifications (e.g. of S. pycnocladulum and
its allies). The exception is a specimen purporting to originate from ‘Cote orientale d’ Afrique,
recolté en 1913 par le P. Gattang sur les pentes du Mkomboko (2500 m) dans l’Ourongourow’
[Uronguru ?]. The data are supplied at secondhand by R. P. Sacleuse, probably in error since
Gattang appears to have made collections from both East Africa and the East African islands in
the same year.

10b. Sphagnum tumidulum var. confusum A. Eddy, var. nov.

(Fig. 23)
S. tumidulo typico adspectu simillimum sed foliis caulinis longioribus et cellulis hyalinis ad eorum apices
parce vel valde fibrosis. Cellulae hyalinae foliorum ramorum poris majoribus (4-0-6-0 um) manifeste
seriatis in paginis ventralibus instructae.

(Typus: Réunion, Plaine de Palmista, Bory, growing in a mat of a pale form of S. violascens C. Mill.
(BM, herb. Schimper — holotypus. )

A REVISION OF AFRICAN SPHAGNALES IAL

Plants closely resembling the typical variety but stem leaves rather more elongated and distinctly
to extensively fibrillose in the sub-apical region. Hyaline cells of branch leaves furnished with
numerous, clearly seriate, rather large (4-0—6-0 um), ringed pores along the commissures on the
adaxial side.

Distribution of the typical form, but very infrequent or overlooked.
East African Islands. REUNION: Plaine de Palmista, Bory (BM — holotype); Briléde St. Denis, 1100 m
alt., Bosser 11369 (PC).

There is a tendency in section Subsecunda for normally dorsiporose species-complexes to
produce ventriporose variants. Sometimes these are inconsequential from a taxonomic view-
point, being clinally interconnected with ‘normal’ plants. At other times, intermediate states are
rare or unknown and it is then convenient to recognize them by name, usually at an infraspecific
level. In S. tumidulum var. confusum, the mere presence of ‘Acrosphagnum’-type pores is a
character sufficient to separate the taxon from S. tumidulum var. tumidulum. At the same time,
the nature of the pores demonstrates the positive link between S. tumidulum and the ‘S. capense
group’ in subsection Acrosphagnum, while their distribution strongly suggests phylogenetic
proximity to S. ericetorum. It is possible, in fact, to arrange species in a kind of stepped cline,
thus: S. davidii-S. pycnocladulum-S. ericetorum-S. tumidulum var. confusum—S. tumidulum
var. tumidulum, suggesting a monophyletic sequence of evolutionary events.

The variety could very easily be confused with S. ericetorum since, even under the micro-
scope, the superficial resemblance of their branch leaves is striking. However, the stem leaves of
S. tumidulum var. confusum maintain the basically triangular form of the type variety, with
involute, more or less mucronate apices and strong, distinctly expanded borders. In S.
ericetorum, the chlorocysts are narrowly exposed on the abaxial side, not immersed as in the
present taxon.

11. Sphagnum davidii Warnst.
(Figs 24 & 25)

in Allg. bot. Z. 11: 99 (1905). Type: East Africa: ‘Hochmoore in 2240 m. MeereshGhe auf dem Berge
Runsoro 1904 von Dr. J. David gesammelt.’ (BM — isotype). —S. davidii var. brachydasycladum Warnst.
in Mildbr. in Wiss. Ergebn. dt. ZentAfr. Exped. 2: 135 (1910). —S. davidii var. viride Warnst., Sphagnol.
univ.: 383 (1911). —S. davidii var. flavo-fuscescens Warnst., Sphagnol. univ.: 383 (1911).

S. chevalieri Warnst., Sphagnol. univ.: 311 (1911).

S. afro-crassicladum Dixon & Sherrin in Dixon in J. Bot., Lond. 76: 218 (1938).

Plants variable in stature, typically medium-sized to rather robust, rarely as small and delicate as
S. capense, green tinged with orange varying to rich orange-brown. Fascicles usually rather
closely set, composed of at least 5, frequently 6-7 branches; branches varying in vigour and
attitude but not dimorphic, the strongest up to 22-0 mm long but often much shorter. Stems
0-6-1:0 mm diameter; cortex well developed, of 2(—3) layers of thin-walled leucocysts, the
outermost largest and commonly with a large pore in the exposed wall. Internal cylinder strong,
yellow-brown to dark red-brown. Branch cortex dimorphic: retort cells distinct, mainly in linear
pairs; neighbouring cortical cells also sometimes perforate. Stem leaves large, lingulate to more
or less spathulate, 1-6-2-6 mm long, 0-9-1-3 mm wide; fibrillose in the upper half or almost to
insertion, fibrillose portions anatomically similar to branch-leaf tissues; border narrow, 2-4 cells
wide, not expanded below; apex eroded, rounded-obtuse, concave. Branch leaves 5-ranked,
mainly ovate but some basal leaves on strongest branches may be longer and narrower and tend
to be slightly homomallous and asymmetric; border narrow, 1(—2) cells wide, sometimes partially
resorbed in the upper half; apex narrowly truncate-dentate, sometimes appearing subacute with
tightly inrolled upper margins. Leucocysts rather variable; in narrower leaves (when present)
leucocysts narrow, 18-0-20-0 x 90-150 um, in normal leaves shorter and wider, 20-0-30-0 x
80-130 um in upper mid-leaf. Abaxial surface with small, ringed pores, 4-0-6-0 wm diameter, in
series along the commissures; with or without 1-6 additional pores near the cell mid-line.
Adaxial surface variable; typically without or with few pores and/or pseudopores but occasional-
ly multiporose. T.S. leaf. Chlorocysts narrow with oval to narrowly oval rectangular lumina,

122 ALAN EDDY

ON

1*O mm

Ai) € a

Fig. 24 Sphagnum davidii Warnst. A, branch leaves (above and lower left) and stem leaves; B, branch
fascicle; C, adaxial (left) and abaxial (right) surfaces of branch leaf; D, branch cortex; E, transverse
section of stem; F, transverse sections of branch leaf (all drawn from the type collection).

more or less median and narrowly and more or less equally exposed on both sides of leaf; abaxial
and adaxial walls usually strongly thickened. Dioecious. Male plants not seen. Female plants
relatively frequent, but no fertilized material available for study.

Distribution. Apparently a rare and local but widespread species. Its main occurrence is in the
high mountains of Ruwenzori and Runsoro, extending southwards to Mlanje, and in the East
African islands. Also scattered in West Africa.

West Africa. SIERRA LEONE: Tingi Hills, c. 1500 m alt., Morton s.n. (BM). IVORY COAST:

Fi

A REVISION OF AFRICAN SPHAGNALES 123

g.25 Sphagnum davidii Warnst. A, abaxial (left) and adaxial surface of branch leaf (S. chevalieri form);
B, branch fascicle; C, transverse section of leaf; D, branch cortex; E, branch leaves (above) and stem
leaf (below); F, adaxial (above) and abaxial (below) surfaces of branch leaf; G, transverse section of
stem; H, adaxial (left) and abaxial (right) surfaces of branch leaf (A—E drawn from the type of S.

| chevalieri; F from Morton s.n.; and H from Lupton 217).

124 ALAN EDDY

Guineko, Chevalier 5 (PC — type collection of S. chevalieri). NIGERIA: Gongola State, Gangirwal,
c. 7500 ft [2290 m] alt., Chapman 3431 (BM, FHO). CAMEROON: Babadjou Santa, Jacques-Felix 2811
(PC):

Mlanje/Chimanimani. MALAWI: Mlanje Mountains, 6000 ft [1750 m] alt., Lupton 217, 218 (BM).

Ruwenzori/Usambara. ZAIRE: Runsoro, ‘Hochmoore in 2400 m. . .’ David s.n. (BM — isotype of S.
davidii); Ruwenzori, Buamba, 3400-3800 m alt., Alluaud 291, 292 (PC; 291 is the type collection of S.
davidii var. flavo-fuscescens); Eggeling 1341 (BM); Kivu, various localities between 2500 and 3500 m alt.,
Humbert 7844, 7853, 7858, 7859 (BM, PC); Kivu, Hendricks 3350, 3709 (PRE); Kivu, piste du Kahuzi,
2750 m alt., De Sloover 12697, 12776, 12954 (BM, NAM); Rungwe Mt., Leedal 4352 (BM). RWANDA:
Rugege, 2020 m alt., wet rocks with Breutelia, De Sloover 12539 (BM, NAM); Birunga, Karisimbi crater,
3900 m alt., De Sloover 13218, 13264 (BM, NAM). UGANDA: Manwamba Valley, 2800 m alt., Taylor
2886 (BM); Toro District, in heath forest at 4200 m alt., Osmaston 3892 (BM); Ruwenzori, Bigo, 3500 m
alt., Osmaston 3719 (BM); Ruwenzori, in ‘ericetum’ by stream, 3800 m alt., Haeman 841 (BM). KENYA:
Aberdare Mountains, summit of Kinangop, 3750 m alt., Galpin 7926; (PRE); Aberdare Mountains,
Kinangop, Balbo 184 (BM); Taylor 1615 (BM — type collection of S. afro-crassicladum). TANZANIA:
Morogoro, Kitundu, 1500 m alt., Bruce M.13 (BM).

East African Islands. REUNION: Piton des Neiges, 2300 m alt., Gimalac 70.R.5005 (BM, NAM; a
rather ambiguous plant showing some approach to S. ceylonicum in structure); ‘Cirque de Cilaos, Sentier
du Piton des Neiges’, 2000 m alt., De Sloover 17.874 (BM, NAM).

It will be clear from the description that S. davidii is closely related to S. pycnocladulum on the
one hand and to S. ceylonicum (see below) on the other. Comparatively few specimens have
been collected, but some of these are from widely dispersed localities, and they demonstrate a
fairly wide range of variation in habit and porosity. The median, more or less symmetrically
disposed, chlorocysts, which invariably reach both leaf surfaces, offer the safest character by
which S. davidii is distinguished from the other species mentioned.

The plant described under the name S. chevalieri is a variant with markedly ventriporose
branch leaves. In the other West African collections cited above this feature is much less in
evidence, and these plants are matched by similar forms from East Africa. In the southern part
of its East African range plants tend to be smaller, without the narrower basal branch leaves, and
are superficially similar to S. capense.

Some allusion has been made elsewhere (Eddy, 1979) concerning the systematic position of S.
davidii. To reiterate in brief, the narrower basal branch leaves (when present), strongly
isoareolate stem leaves, and median chlorocysts indicate a relatively ‘primitive’ condition within
subsection Acrosphagnum. This is supported to a degree by its wide distribution since it is the
only trans-African species of the subsection. Evolutionary divergence is suggested towards S.
ceylonicum (thence to section Acutifolia?) on the one hand, with adaxially displaced chloro-
cysts, and towards S. pycnocladulum and S. capense (and, ultimately, to S. tumidulum) on the
other, with abaxially displaced or enclosed chlorocysts.

12. Sphagnum capense Hornsch.
(Figs 26-29)

in Linnaea 15: 113 (1841). Lectotype: ‘In der Kluft nach dem Gipfel des Tafelberges’, Ecklon (BM, PC).
Paratypes: ‘Dutoitskloof am Wasserfalle 1600 f. . . .. Drége (BM, PC); ‘An einer kleiner Quelle. . .’,
Ecklon (H). — S. capense var. multiporosum Warnst., Sphagnol. univ.: 428 (1911). — S. capense var.
mollissimum (C. Mill.) Warnst., Sphagnol. univ.: 428 (1911). — S. capense var. austro-molle (C. Mill.)
Warnst., Sphagnol. univ.: 428 (1911).

mollissimum C. Mill. in Flora, Jena 70: 418 (1887). — S. mollissimum var. elongatum Rehmann in
Warnst., Sphagnol. univ.: 428 (1911), nom. syn.

panduraefolium C. Mill. in Flora, Jena 70: 418 (1887).

austro-molle C. Mill. in Flora, Jena 70: 419 (1887).

islei Warnst. in Hedwigia 30: 14, 19, tab. 1 fig. 6, tab. 4 fig. f (1891).

subrotundifolium C. Miull., Gen. musc. frond.: 102 (1900), nom. nud.

wenckei Rollin Ost. bot. Z. 57: 7 (1907).

pappeani Breutel in Warnst., Sphagnol. univ.: 427 (1911), nom. syn.

humbertii Cardot in Bull. Mus. natn. Hist. nat. Paris 22: 342 (1916).

125

A REVISION OF AFRICAN SPHAGNALES

WES
ORS
ROKR

O <
es

(right, cross-hatched figs); D, branch cortex; E, transverse sections of stems (B drawn from Ecklon’s

Fig. 26 Sphagnum capense Hornsch. A, B, branch fascicles; C, branch leaves (left figs) and stem leaves
type collection; the remainder from various South African localities).

126 ALAN EDDY

Fig. 27 Sphagnum capense Hornsch. A-C, adaxial (left figs) and adaxial (right figs) surfaces of branch
leaves; D, adaxial surface of stem leaf; E, abaxial surface of stem leaf; F, transverse sections of leaves (A
drawn from Rehmann 17c; B from the type collection; C from De Sloover 17. 625; D, E from the type
collection and F from various sources).

A REVISION OF AFRICAN SPHAGNALES 127

Fig. 28 Sphagnum capense Hornsch. A, abaxial (left) and adaxial (right) surfaces of branch leaf,
amphiporose form; B, adaxial (left) and abaxial (right) surfaces of branch leaf, amphiporose form; D, E,
abaxial surface of branch leaves, exceptionally multiporose forms; E, adaxial (left) and abaxial (right)
surfaces of branch leaf, average form (A drawn from Rehmann 433; B from Lacey 65; C, from P. de la
Bathie 12127; D from De Sloover 17.660; and E from Mitchell 375).

Plants small and delicate, green in shade but usually tinged with orange-brown and commonly
the whole plant orange-brown, sometimes suffused with reddish-purple. Fascicles dense or
remote, with 2-4 branches which vary in vigour and intensity of pigmentation but are not at all
dimorphic. Stronger branches 4-0—10-0(—13-0) mm long. Stems 0-4—0-8 mm diameter. Cortex
well developed, of 2—3(—4) layers of leucocysts, the outermost up to 45 u wide and usually with a
large pore in the outer wall. Internal cylinder well developed, normally brown. Branch anatomy.
Cortical cells markedly dimorphic; retort cells usually in linear pairs, highly inflated and with the
pores at ends of conspicuous protuberances. Internal cylinder yellow to deep brown. Stem leaves
relatively large, larger than branch leaves, ovate-lingulate, 1-2-1-9 mm long, 0-7—-0-9 mm wide,
fibrillose in the upper half, sometimes almost or quite to insertion. Upper fibrillose tissue
identical to that of branch leaves, lower tissues variable, usually the leucocysts with 1-2
resorption gaps or enlarged pores. Border narrow, not expanded below. Leaf apex variable,
from narrowly truncate and dentate, apparently mucronate due to ‘pinched in’ margins, to long,

128 ALAN EDDY

Fig. 29 Sphagnum capense Hornsch. A, branch fascicle; B, branch leaves (above) and stem leaves
(below); C, adaxial (left) and abaxial (right) surfaces of branch leaves; D, transverse section of stem; E,
abaxial surface of non-resorbed branch leaf cells; F, branch cortex; G, transverse sections of branch
leaves (all drawn from the type of S. is/ei).

wide-ovate, very concave; apices truncate-dentate, less commonly narrow and more or less
mucronate, occasionally eroded; border narrow, typically of a single series of cells, with or
without resorption of the outer wall near the leaf apex. Leucocysts relatively short and wide,
18-0—25-0 x 70-0-110-0 um in upper mid-leaf; sometimes with a marked tendency to have partial
loss of fibrils; pores very variable; typically, abaxial surface with very numerous, ringed pores in
series along the commissures, commonly with one to several additional free pores in, or near,
the cell midline; adaxial surface sometimes devoid of pores (usually at least some pseudopores)
but usually with 1-few pores near the commissures, especially at the cell angles, rarely with very

A REVISION OF AFRICAN SPHAGNALES 129

numerous adaxial pores and only exceptionally with as many or more adaxial than abaxial pores;
pores 3-0-6-0 um diameter. 7.5. leaf. Leucocysts convex on both sides, sometimes more
strongly so adaxially. Chlorocysts narrow, oval to urceolate, mainly enclosed on the adaxial side
but very narrowly exposed abaxially via a thick, refractive wall: less frequently chlorocysts
reaching adaxial surface, or chlorocysts completely enclosed both abaxially and adaxially, but
then always more deeply so adaxially. Dioecious. Antheridial bracts more or less indis-
tinguishable from branch leaves but fibrils usually lost in lower half of bract. Female bracts large
and convolute, ovate to ovate-lingulate, very concave, apices obtuse to sub-acute, not resorbed;
tissue fibrillose almost to insertion and resembling that of branch leaves. Fruit rather rare;
spores yellowish, rugulose, 34-38 um diameter.

Distribution. Growing in low mats on acid, flushed mineral or, rarely, organic substrates at
medium to high altitudes, rarely below 500 m alt. Local to locally frequent in southern Africa
from Cape Province to Mozambique, becoming rarer in the more northerly regions of East
Africa; several records from the East African islands.

Southern Africa. CAPE PROVINCE: ‘In der Kluft nach dem Gipfel des Tafelberges’, Ecklon s.n. (BM,
H, PC -lectotype collections of S. capense); ‘An einer kleiner Quelle’, Ecklon s.n. (H, PC); ‘Dutoitskloof
am Wasserfalle’, 1600 ft. alt., Drége s.n. (BM, PC); Clarkson, Breutel s.n. (BM); Table Mountain,
Rehmann exs. 15 (BM, PRE-collections of S. panduraefolium), exs. 16, 16b, 16c (= Rehmann 433, 433b,
433c) (BM, H, PC, PRE - syntype collections of S$. austro-molle), exs. 17c (BM, etc.); Stinkwater,
Rehmann exs. 18 (H, fertile); ‘Prom. bon Spei’, Berguis s.n. (H, as ‘S. patens Brid.’ in herb. S. O.
Lindberg). Also many other more recent collections from the Table Mountain—Devil’s Peak—Montagu
Pass region, where the species and collectors are frequent, including: Almborn 5241a, b (PRE); Eddy &
Sims s.n. (BM); Esterhuysen 15362, 15427, 15431, 15560 (BM); Garside 74 (PRE); Goldblatt 1357c (PRE);
Howes 8 (PRE); Humbert s.n. (PC); Magill & Schelpe 3072 (H, PRE); Marshall & Crosby 8128 (PRE);
Pillans 3545 (BM); Rodin 3235 (BM); Schelpe 3811, 3865, 3869, 3882 (BM); Schyff 5826 (PRE); Sim 9120,
9139, 9276, 9417, 9465, 9486 (PRE); Spielhaus, 1875 (BM); Townsend 82/196; Van Zinderen Bakker
Bif. -U. 17458, 17458A (fertile and fruiting), 17745 (PRE); Vorster 450, 461, 664, 942, 943, (BM); Wager
88, 311, 511, 591 (BM, PRE); Webster 353 (BM); Weintraub 10301 (BM, fertile); Wilman 535 (PRE);
Wilms 2603, 2629 (BM). TRANSVAAL: Drakensberg, Mt. Anderson, 2200 m alt., Humbert s.n. (PC);
Mt. Kroatlamba, MacMac, Maclea s.n. (H).

Mlanje/Chimanimani. MALAWI: Mlanje Mt., 2000-2200 m alt., Humbert s.n. (BM, PC); Lacey 65
(BM); Lupton 193 (BM); Wilman 1369/52A (PRE). ZIMBABWE: Chimanimani Mts., 6000 ft [2000 m]
alt., Mitchell 375 (BM); Nyumkombe, in cascade over ledge of rock, Gilliland 893 (BM).

East African Islands. MADAGASCAR: ‘sous-bois de la région centrale’, 2000 m alt., P. de la Bathie
12127 (BM); Vakinankaratra, near summit of Triafajarona, 2575 m alt., Viguier & Humbert 1582 (PC),
Viguier & Humbert 1588 (PC -type collection of S. humbertii). REUNION: Cirque de Cilaos, 1950 m alt.,
in bryophyte carpet on wet talus, De Sloover 17.625, 17.660, 17.875 (BM, NAM). AMSTERDAM
ISLAND: G. de I’Isle 42 (BM, PC — type collections of S. is/ei).

This delicate and rather neat species belongs to a group of section Subsecunda which is almost
exclusively African in its occurrence. The group is characterised by the possession of a strongly
developed, 2-4-layered cortex and ovate leaves with wide, multi-porose leucocysts. Associated
features are the generally short, numerous, markedly 5-ranked branch leaves, large stem leaves,
and highly developed retort cells. S. capense is closely related to S. pycnocladulum and it may
not always be possible to distinguish between the former species and underdeveloped states of
the latter. Garside (1949) was rather ambiguous in his treatment of S. pycnocladulum, implying
at One point that it is simply a robust expression of S. capense, nevertheless listing both as
distinct species in his text. S. capense is, on average, a distinctly more delicate plant than S.
pycnocladulum, never has more than 4 fully developed branches per fascicle, and only
exceptionally has pseudomucronate branch leaves. S. davidii, another African species of the
group, is usually more robust, has usually 5 or more branches per fascicle, and has chlorocysts
narrowly and equally exposed on both sides of branch leaves. S. ceylonicum has adaxially
displaced chlorocysts and usually proportionately narrower leaves.

Variation within the species is mainly in the number and distribution of the pores in the leaves.
The pores are, on average, larger and more regular in form than those of other section
Subsecunda, averaging 4-6 wm diameter, occasionally up to 10 wm. The tendency to have

130 ALAN EDDY

additional free pores in the cell midline is a characteristic of S. capense and its relatives, but has
been noted also in S. africanum (distinguished by its almost uniseriate cortex). The degree of
chlorocyst exposure is also variable, from those plants in which there is narrow adaxial, but
wider abaxial exposure to forms with adaxially immersed chlorocysts.

S. panduraefolium C. Mill. is a juvenile or etiolated plant with poorly developed branches
and large stem leaves. Populations from the East African islands include exceptionally small
plants with excessively perforate leaves, but these are matched by some of the collections from
Mlanje. Such plants are not considered to merit taxonomic distinction.

13. Sphagnum pycnocladulum C, Mill.
(Figs 30-32)

in Flora, Jena 70: 420 (1887). Type: S. Africa, ‘Montagu Pass ad cataractam’, Rehmann (1875); Musci
austro-africani 13 (1878-1879) (BM). — S. pycnocladulum var. viride Warnst., Sphagnol. univ.: 176
(1911). —S. pycnocladulum var. fuscescens Warnst., Sphagnol. univ.: 176 (1911).

S. beyrichianum Warnst. in Hedwigia 36: 157 (1897).

S. goetgeanum Warnst. in Hedwigia 47: 98 (1907).

Plants medium-sized to rather robust, rarely small and delicate, green, yellowish to brown.
Fascicles usually spaced at 24 mm on stems, rarely closely packed, composed of at least 5,
commonly 6 or even 7 branches. Branches 8-0-18-0 mm long, uniformly foliate, varying in
length but not dimorphic, although weaker 24 branches more or less pendent. Stems rather
rigid, 0-6—0-9 mm diameter; cortex well developed, of 2-3 layers of thin-walled leucocysts, the
outer layer usually with single pores or thinnings; internal cylinder rather thick, yellowish to
dark brown. Branch cortex dimorphic; retort cells in linear pairs or occasionally threes, highly
inflated and moderately protuberant at apertures. Stem leaves varying in attitude, usually more
or less spreading, larger than branch leaves, 1-0—-1-8 mm long, 0-7-0-9 mm wide, varying from
triangular-ovate (heterophyllous forms) to ovate-lingulate; apex variable, from pseudo-
mucronate (hemiisophyllous forms) to rounded-obtuse and eroded; border 2-4 cells wide, not
expanded towards leaf base. Leaf tissue variable in extent of fibrillation: typically fibrillose only
in the upper half, not rarely almost or quite to insertion, exceptionally devoid of fibrils. Upper
fibrillose tissue more or less identical to that of the branch leaves, lower tissues with one to
several resorption gaps, mainly towards cell ends. Branch leaves very regular in size and
disposition, markedly 5-ranked, ovate and very concave, 0-8-1-4 mm long, 0-8—1-0 mm wide;
apices narrow, more or less abruptly tapered to acute, or narrowly truncated, more or less
dentate tips, always appearing more or less mucronate because of the strongly involute upper
margins (virtually impossible to flatten under a cover glass). Border narrow, without resorption
furrow. Leucocysts 18-0—25-0 um wide, 80-130 um long in upper mid leaf; fibrils on abaxial side
often very irregular and sometimes more or less suppressed in the middle part of cells. Abaxial
surface with small to medium-sized pores (3-0—5-0 um diameter) more or less in series along the
commissures, with or without 1-5 additional pores in, or near, the cell midline. Adaxial surface
without pores, or pores few; cell often appearing reticulate with ramifying fibrils and more or
less pseudoporose. T.S. leaf. Chlorocysts narrow, with oval lumina, very narrowly exposed on
abaxial leaf surface, usually shallowly immersed or just reaching adaxial side. Dioecious.
Antheridial branches virtually indistinguishable from sterile branches. Inner perichaetial leaves
large, c. 4-5 X 2-5 mm, ovate and more or less acute, fibrillose to below the middle. Fruit very
rare; spores 34-36 um diameter, more or less rugulose-papillose.

Distribution. In and around acid springs and mires; rare and of scattered occurrence from
western Cape Province to Ruwenzori.

Southern Africa. CAPE PROVINCE: Montagu Pass, Rehmann, Musci austro-africani 13 (BM — type
collection of S. pycnocladulum). TRANSKEI: Pondoland, ‘mooringe Land auf Sandstein’, Beyrich s.n.
(BM, H-type collections of S. beyrichianum).

Mlanje/Chimanimani. MALAWI: Usafua, 2200 m alt., Goetz 1301 (BM — type collection of S.
goetzeanum); Mlanje, Likabula Valley, Lacey 52 (BM); Mt. Mlanje, 6300 ft [c. 1900 m] alt., Davy 2088
(BM).

A REVISION OF AFRICAN SPHAGNALES 131

OVO
AN
y~N SEIS Y

leaves (right figs); D, adaxial (left) and abaxial (right) surfaces of branch leaf; E, transverse section of
stem; F, transverse section of branch leaf (all drawn from the type collection).

Ruwenzori/Usambara. RWANDA: Rugege, between Mt. Yahahi and Mt. Bigugu, 2450 m alt., De
Sloover 13786 (BM, NAM), 19093 (BM).

The type collection consists of a robust, rigid, very distinctive plant which superficially bears a
strong resemblance to robust forms of S. tumidulum. On the other hand, the weaker expres-
sions, including ‘S. beyrichianum’ forms, have a facies much more like that of S. capense.
Garside (1949) was ambivalent in his treatment of the taxon, but held the opinion that S.
pycnocladulum (which he interpreted more widely than the present author) was the mature,
female phase of S. capense. As a result, most fruiting plants of S. capense were identified by him

132 ALAN EDDY

Fig. 31. Sphagnum pycnocladulum C. Mill. A, branch cortex; B, adaxial (left) and abaxial (right)
surfaces of stem leaf in supapical leaf region; C, adaxial (left) and abaxial (right) surfaces of branch leaf;
D, branch leaves (above) and stem leaves (below) (drawn mainly from the type collection of S.
goetzeanum).

as S. pycnocladulum. While the two taxa are indisputably closely related, and allowing that S.
pycnocladulum is a not altogether satisfactory species, the latter is usually easily identified by its
more or less mucronate leaves and at least some fascicles bearing 5 or more branches.
Furthermore, while unequivocal plants of S. capense have been collected in fruit on a number of

occasions, the well marked examples of S. pycnocladulum available for study are apparently all
barren.

14. Sphagnum ericetorum Brid.
(Fig. 33)
Sp. musc. 1: 19 (1806). Type: Réunion, Plaine de Chicot, Bory de Saint-Vincent (B).

Plants identical in habit to the less robust forms of S. pycnocladulum but differing in the

A REVISION OF AFRICAN SPHAGNALES

Fig.32 Sphagnum pycnocladulum C. Mill. A, branch cortex; B, adaxial surface of branch leaf; C, (2 figs)
abaxial surfaces of branch leaves; D, abaxial (left) and adaxial (right) surface of branch leaf; E,
transverse section of stem; F, branch leaves; G, stem leaf; H, transverse section of leaf (drawn mainly
from the type collection of S. beyrichianum).

134 ALAN EDDY

Fig. 33 Sphagnum ericetorum Brid. A, branch fascicle; B, branch leaves; C, stem leaf; D, branch cortex;
E-F, adaxial (left) and abaxial (right) surfaces of branch leaf; G, transverse section of stem; H,
transverse sections of leaf (A, D and E drawn from Bosser 9501; the remainder from the holotype).

A REVISION OF AFRICAN SPHAGNALES 135

following points: stem leaves usually proportionately larger and more isophyllous in character,
ovate-lingulate, 1-8—2-5 mm long, fibrillose almost or quite to insertion. Leucocysts of branch
leaves with few or no true pores on the abaxial side (but often with abundant pseudopores);
adaxial surface with abundant, medium-sized, ringed pores, 6-0—-10-0 um diameter, along the
commissures, with or without a few additional free pores in or near the cell mid-line. Chlorocysts
occasionally enclosed abaxially as well as adaxially.

Distribution. Endemic to the East African islands, where it is apparently rare. Scanty data
indicate that the plant occurs in wet heathlands at high altitudes.

East African Islands. MADAGASCAR: Massif de l Andringitra, 2000-2500 m alt., Humbert 98 (PC,
also a specimen numbered ‘28’ in herb. Cardot); Massif du Tsaratanana, 2300-2875 m alt., Humbert s.n.
(PC); Vakinankaratra, near summit of Triafajarona, 2575 m alt., Viguier & Humbert 1588 p.p. (PC).
REUNION: Philippia, 2000 m alt., Bosser 9501 (BM, PC).

It is with some hesitation that S. ericetorum is treated as a species rather than a subspecies or
variety of S. pycnocladulum. The characters that discriminate the two taxa, while making
identification a relatively simple operation, are themselves so variable in the other taxa of the S.
capense complex that it is inconsistent to give them greater emphasis here. Although there do
not appear to be intermediate facies between S. ericetorum and S. pycnocladulum, the situation
is in many respects analogous to that existing between S. chevalieri and S. davidii (which are
treated here as simply synonyms). However, if united below the level of species, S. ericetorum
would take precedence over S. pycnocladulum. Bridel’s (loc. cit.) original diagnosis is quite
inadequate to circumscribe a plant which, at that time, was apparently represented solely by the
type specimen in his own herbarium (B). Subsequently, and not surprisingly, many and diverse
plants were erroneously identified as S. ericetorum, and the situation remained confused until
the present time. On the other hand, while the specific status of S$. pycnocladulum may be
regarded as equivocal by some, its identity is reasonably well established, and since both taxa are
rare, it is considered practical here to maintain the nomenclatural status quo by retaining S.
ericetorum at species level.

Ventriporose forms of S. davidii are distinct in their median chlorocysts having equal
exposure on both branch-leaf surfaces, while S. tumidulum var. confusum is separable by its
immersed chlorocysts and pseudomucronate stem leaves with distinctly expanded borders.

15. Sphagnum ceylonicum Mitten ex Warnst.
(Fig. 34)
in Hedwigia 29: 195, tab. 4 fig. 7, tab. 7 fig. 14 (1890). Type: Sri Lanka (Ceylon), Thwaites 262 (BM).
S. kerstenii Hampe in Warnst., Sphagnol. univ.: 135, fig. 30C (1911).
S. keniae Dixon in Smithson. misc. Collns. 69 (2): 7, pl. 1 fig. 1 (1918).
S. vandenbroeckii Nav. in Natuurw. Gijdschr. 4: 144 (1922).
S. ugandense J. Taylor & A. Thompson in Kew Bull. 9: 517 (1955).

Plants medium-sized or rather small, lax or dense, green variegated with brown and orange or
whole plant orange-brown. Fascicles distant or closely set, with S—7 branches; branches rather
short to elongated and tapering, the strongest 9-0-18-0 mm long, monomorphic. Stems 0-6—
0-8 mm diameter, yellowish to deep brown; cortex well developed, of 2(—3) layers of thin-walled
leucocysts, the outer layer usually, at least in part, uniforaminate. Branch cortex dimorphic;
retort cells in linear pairs slightly protuberant at apertures; sometimes, here and there, other
cortical leucocysts also porose. Internal cylinder pale to dark brown. Stem leaves relatively large,
lingulate to oval-lingulate, 1-8—2-6 mm long, 0-9-1-2 mm wide, fibrillose to below mid-leaf,
often to leaf base; border 2-4 cells wide, evanescent at the eroded, rounded-obtuse leaf apex,
not expanded below. Upper fibrillose tissue identical to branch-leaf tissue. Branch leaves
basically 5-ranked, often markedly so, varying from short and widely ovate to longer and
ovate-lanceolate, 1-0-2-1 mm long, 0-7-0-8 mm wide; apices narrowly truncate-dentate but
usually appearing acute or pseudomucronate; border narrow, not resorbed. Leucocysts rather
narrow to relatively short and wide, mainly 20-30 x 90-120 um; abaxial surface multiporose;
pores small to medium-sized, circular, ringed, 5-O-8-0 um diameter on average, in series along

136 ALAN EDDY

50 pam

Fig. 34 Sphagnum ceylonicum Mitten ex Warnst. A, B, abaxial (left) and adaxial (right) surfaces of
branch leaves; C, D, branch leaves (upper figs) and stem leaves (lower figs) and one pendent-branch leaf
(lowermost fig.); E, transverse section of stem (low magnification); F, branch cortex; G, transverse
section of branch leaves, H, transverse section of stem (A, C, G drawn from the type of S. keniae; the
remainder from the type of S. ugandense).

A REVISION OF AFRICAN SPHAGNALES 3i7/

the commissures and usually 1-6 or more free pores near the cell mid-line. Adaxial surface
without pores or pores relatively few. T.S. leaf. Leucocysts more strongly inflated adaxially.
Chlorocysts varying from narrowly trapezoid to narrowly urceolate, thick-walled with more or
less oval lumina, with widest exposure on the adaxial side, varying from narrowly exposed to
shallowly enclosed on the abaxial side. Dioecious (presumably). No male plants seen. Female
plants not rare, but only unfertilized specimens so far available.

Distribution. A local and apparently rare species, more or less confined to high altitudes, in East
Africa, the East African islands, and Sri Lanka.

Ruwenzori/Usambara. UGANDA: Kigezi, Magahinga, in crater swamp at 3650 m alt., Lind 12 (BM —
type collection of S. ugandense). KENYA: Mount Kenya, 4000 m alt., Rune s.n. (H); Mount Kenya,
3630 m alt., Mearns 1560 (BM~—holotype of S. keniae); Teleki Valley, 3000-4000 m alt., Stwarts.n.|R.A.F.
Lyneham Exped., 1980] (BM).

East African Islands. MADAGASCAR: Andohahelo, rocks at summit, 1900 m alt., Humberts.n. (PC).
REUNION: Kersten s.n. (BM, herb. Hampe — type collection of S. kerstenii); ‘Enclos de la Fournaise,
Piton du Crac’, bog in sunlight, 1400 m alt., Gimalac 71.R.8501 (BM, NAM); ‘Sentier du Tremblet au sud
de la Fournaise’, 2000 m alt., Gimalac 71.R.8004 (BM, NAM).

S. ceylonicum closely resembles S. davidii but with a tendency to have slightly longer leaves
drawn to finer apices. It is, however, very variable in this respect, and the only safe discriminat-
ing character is the position of the chlorocyst with distinctly adaxial exposure.

The branch-leaf pores of this, and other, species of the S. capense group are larger on average
than those usually met with in section Subsecunda. In addition, the multi-layered hyaloderm,
and in the case of S$. ceylonicum the adaxially biased chlorocysts, suggest an approach towards
section Acutifolia. In fact, Warnstorf (1911) and later Taylor & Thompson (1955) placed S.
ugandense (= S. ceylonicum) in the latter section. Clearly there could be no question of
including either S. pycnocladulum or S. capense in section Acutifolia, two species very closely
related to S. ceylonicum. Juvenile states of all species of the S. capense group show very clearly
their affinities with section Subsecunda.

Section CUSPIDATA
Sphagnum sect. Cuspidata (Lindb.) Schimper, Syn. musc. eur. 2nd ed.: 829 (1876).

Plants very variable but usually pale and commonly very flaccid; secondary pigments, if
developed, yellow to brown, never red. Stem hyaloderm sometimes rather indistinct, never
foraminate. Fascicles commonly not, or only weakly dimorphic. Retort cells of branch
hyaloderm usually well developed, typically in linear groups of 2-3, rarely solitary. Stem leaves
very variable, isophyllous to markedly heterophyllous. Branch leaves lanceolate to linear.
Branch-leaf leucocysts typically narrow and pauciporose on the abaxial side, adaxial pores
usually unringed and more or less confined to the cell angles, rarely numerous and ringed.
Chlorocysts in section more or less triangular to trapezoid, always displaced towards, and with
only or widest exposure on the abaxial leaf face.

Distribution as in the genus.

This is a large and diverse section, well represented in the tropics as well as in the temperate
zones. The African flora appears to comprise relict, endemic, and apparently primitive forms
(e.g. S. planifolium agg.), in addition to more evolved plants that include incursions from other
continental elements (e.g. S. cuspidatum from Laurasia and S. cuspidatulum from Malaysia), in
addition to obvious neoendemics such as S. slooveri.

16. Sphagnum cuspidatum Ehrh. ex Hoffm.
(Figs 35-37)

Deutschl. Fl. 2: 22 (1796).

S. bernieri Bescher. ex Ren. & Cardot in Grandidier, Hist. phys. Madagascar 39 (atlas, 4): plate 142 (1901).
S. ikongense Warnst. in Magy. bot. Lap 1: 46 (1902).

S. aloysii-sabaudiae Negri in Annali Bot. 7: 161 (1908).

S. gabonense Bescher. apud Warnst., Sphagnol. univ.: 269 (1911).

138 ALAN EDDY

5°O mm p

aR LA ae
AN SAS

\ ‘ (ee
a ‘i \ A
HIN | :
HAAN

Ff

a

C0
ae

Fig. 35 Sphagnum cuspidatum Ehrh. ex Hoffm. A, branch fascicle; B, stem leaves; C, branch leaves
(lower, middle and upper branch leaves, from left to right); D, apex of stem leaf; E, fascicle with female
branch and empty capsule; F, transverse section of stem; G, branch cortex (E drawn from P. de la Bathie
7073; the remainder from the type collection of S. ikongense).

A REVISION OF AFRICAN SPHAGNALES 139

Fig. 36 Sphagnum cuspidatum Ehth. ex Hoffm. A, branch fascicle; B, stem leaves; C, branch leaves; D,
apex of stem leaf; E, transverse sections of leaves; F, transverse section of stem (drawn from the type
collection of S. bernieri).

ALAN EDDY

140

branch leaves; D, branch cortex; E, adaxial surface of apical region of stem leaf (A drawn from the type

Fig. 37 Sphagnum cuspidatum Ehrh. ex Hoffm. A-C, abaxial (left figs) and adaxial (right figs) surfaces of
of S. bernieri; B from the type of S. gabonense; C-E from Humbert 7855).

A REVISION OF AFRICAN SPHAGNALES 141

Plants small to medium-sized, rarely robust, flaccid but sometimes with brittle stems, green or
almost colourless, with or without brownish or orange tints in young branch tips (especially in
male plants). Fascicles variously spaced on stems, seldom closely packed, monomorphic but
varying in vigour, 3—5 branches per fascicle, branches up to 26-0 mm but usually 12-0-20-0 mm.
Stems pale, more or less lacking secondary pigments or, uncommonly, here and there pale
brown (at least in herbarium specimens), 0-4-0-8 mm diameter. Cortex distinct, (1—)2-3
layered, the outer cells thin-walled or slightly incrassate, never with pores or thinnings in the
external walls. Internal cylinder quite well developed, rather abruptly delimited internally,
yellowish. Branch cortex dimorphic; retort cells distinct, usually in linear pairs, only slightly
protuberant at apertures. Internal cylinder pale or pink or light brown (often only one branch in
a fascicle has a pigmented cylinder). Stem leaves erect, spreading or reflexed, triangular-ovate,
always longer than wide, (1-1—)1-2—1-6(-1-8) mm long (occasionally longer in juvenile or
aberrant plants); apices rounded-obtuse to more or less truncate, dentate or eroded, sometimes
slightly hooded or apparently mucronate due to inflexed upper margins. Border strong above,
slightly to very strongly expanded below into patches of prosenchymatous tissue. Branch leaves
conspicuously long and proportionately narrow, straight or somewhat curved and homomal-
lous, basically 5-ranked but often only obscurely so; lowest leaves short and more or less ovate,
the great majority narrowly lanceolate to linear, frequently appearing almost subulate due to
inrolled margins; (1-3—)1-4~-2-5(-—3-0) mm long, 0-5—0-8 mm wide; apices truncate and 5—9-
dentate; border strong, 2-4 cells wide. Hyaline cells long and narrow, 11-0—15-0(—20-0) um long
in mid-leaf; abaxial surface without pores or with a small pore in an upper lateral angle, with or
without an inconspicuous resorption gap in the apical angle; pseudopores absent or frequent;
adaxial surface normally porose but pores often only discernible after staining; pores 2-S(—6) per
cell, adjacent to the cell angles, ringed or unringed, 2-0-4-0 um diameter. T.S. leaf. Hyaline cells
almost plane or shallowly convex abaxially, shallowly to strongly convex adaxially. Photo-
synthetic cells exposed on both sides of leaf, trapezoid in section, more or less thin-walled and
with widest exposure on the abaxial side (sometimes almost as wide as hyaline cells). Dioecious.
Male branches usually tinged orange or brown; perigonial leaves shorter and more concave but
not essentially differing from sterile leaves. Inner perichaetial bracts very large, convolute, up to
5:0 mm long, c. 2:1 mm wide, usually fibrillose in a zone below apices and sometimes,
erratically, elsewhere; apex truncate or retuse, prosenchymatous, not resorbed. Fruit very rare
in African material; spores brown, coarsely papillose, c. 32-0 um diameter.

Distribution. A widespread, circumboreal species occurring from subarctic to equatorial
latitudes, most abundant in oceanic regions of Europe and eastern North America where the
species occurs in terrestrial or aquatic states. In Africa S. cuspidatum occurs exclusively in a
broad zone from Gabon in the west to Ruwenzori in the east, and on the larger islands in the
Indian Ocean, mainly above 2000 m alt., except in the more oceanic parts of Malagasy where it
descends almost to sea-level. The species is frequently fertile in temperate latitudes, but very
rarely so in Africa.

West Africa. GABON: Fernan Vaz, Walker 77 (BM); Duparquet s.n. (PC — type collection of S.
gabonense); ‘Savanes dans la région de bas Ogooué’, Binda, Le Testu 1597 (BM). CONGO REPUBLIC:
Brazzaville, Koechlin 2694 (BM); South of Djambala, Nicklés s.n. (BM). ZAIRE: ‘Lac Leopold II, al’est
de Bankaie’, Gilbert 14767 (H).

Ruwenzori/Usambara. ZAIRE: Kivu, montane forest at 2200 m alt., Humbert 7855 (PC). RWANDA:
Massif des Birunga (= Virunga), Karisimbi, in water in crater, 3850 m alt., De Sloover 13257 (BM, NAM);
Gahinga, at summit crater, 3440 m alt., De Sloover 13542 (BM, NAM). UGANDA: Kigezi, swamp in
crater, 3250 m alt., Thomas 2444 (BM).

East African Islands. MADAGASCAR: North of the island, Bernier s.n. (PC — isotype of S. bernieri);
swamp in forest, east coast below Natitana, P. de la Bathie 7073 (PC, fruiting); Andovoranto, Viguiér &
Humbert s.n. (PC); Port Dauphin, Decary s.n. (PC, fruiting); Midongy, Farafangana, ‘bog in sunlight’,
Decary s.n. (BM); Tananarive, Mount Bity [Ibity?], 2000 m alt., Cremers 1972 (BM); Andondabe,
submerged in bog, Cremers 2306 (BM); 8 km from Santa Suce, Onraedt 70.M.2000, 70.M.2001 (BM,
NAM); Plateau d’Ikongo, Besson 313, 317 (PC-—syntype collections of S. ikongense); Marais de Stampika,
Ambongo, P. de la Bathie 7140 (PC); Antongil, marsh near the sea, P. de la Bathie 7113 (PC); Mt. Ibity,
2000 m alt., P. de la Bathie 7037 (PC).

142 ALAN EDDY

In spite of considerable variation in vigour, degree of stem-leaf fibrillation, branch-leaf
dimensions, and other morphological details, S. cuspidatum will rarely present any difficulty in
identification. Branch-leaf shape alone will be sufficient in most cases to determine the species,
only one other African taxon, S. planifolium var. augustilimbatum, approaching it in leaf shape.
The distinctly expanded stem-leaf border of S. cuspidatum is a feature that is absent in all forms
of S. planifolium.

From a genetic diversity point of view, vicariad taxa may have to travel a considerable distance
before they can be confidently identified using purely morphological characters. Full expression
of derived characters may be obscured, furthermore, in juvenile or ecologically repressed
phases of growth. Taxonomic issues are complicated further by the propensity of Sphagnum to
follow avenues of parallel or convergent evolution. S. cuspidatum, with regard to its tropical
occurrences, exemplifies the kind of problem that may be beyond solution by traditional
herbarium-taxonomical approaches. African populations, under such names as S. bernieri, S.
gabonense, and S. ikongense, seem to fit perfectly well within the traditional interpretation of
S. cuspidatum, even taking into account local tendencies towards minor modifications (e.g.
larger, more extensively fibrillose stem leaves; less widened prosenchymatous patches of cells,
etc.). The present author has been unable to find any morphological characters by which
populations from Madagascar, Zaire, or Ruwenzori can be separated from boreal S. cuspida-
tum. If the suggested conspecificity is valid, then S. cuspidatum features distributional patterns
that differ from all other sphagna, being the only holarctic species to be found also in tropical
Africa south of the Sahara. Recalling that S. cuspidatum occurs also in tropical Asia (extending
as far as New Guinea [Eddy, 1977]), then the species is either a) an old taxon (possibly pre-mid
Cretaceous) that had already become widely distributed before the break-up of Pangea, or b) a
very successful migratory species (it has been found in fruit in Madagascar, and is certainly
commonly very fertile in its temperate range) that managed to infiltrate Africa after the closure
of the Tethys, or c) the African material belongs to a cryptic ‘doppel-ganger’ and is in reality an
endemic taxon superficially indistinguishable from the holarctic plant. The interpretation
favoured by this author is that option a) pertains, S. cuspidatum being an old and genetically
heterogeneous species, representing an ancient divergence from, e.g. S. falcatulum Bescher.
(temperate S. hemisphere).

17. Sphagnum planifolium C. Mill.
(Figs 38-43)

in Flora, Jena 70: 415 (1887). Type: ‘Africa aequinoctialis occidentalis, prope flumen Gabun in Arthing-
ton-cataracta’, Biittner 23 (B, destroyed).

Plants extremely variable, from small, pale and flaccid (resembling S. cuspidatum) to robust and
tumid; secondary pigmentation absent or confined to youngest and male branches which may be
tinged yellowish to brown. Fascicles rather distant to closely set, stem hidden or not; (3—)4—5
branches per fascicle of which 2-3 are pendent but not or only slightly dimorphic; branches
15-0-30-0(-40-0) mm long, finely tapering distally. Stems green, yellowish or pale brown,
0-3-1-1 mm diameter. Cortex distinct, of 2-3 layers of moderately inflated, thin-walled or
commonly somewhat incrassate leucocysts; never with pores in the outer walls. Internal
cylinder weak to well developed, yellowish to pale brown. Branch cortex dimorphic; retort cells
distinct, in linear series of 2-4, only slightly protuberant at apertures. Stem leaves usually
reflexed, triangular-ovate to triangular-lingulate, (1-1—)1-3—2:0 mm long, 0-7—1-2 mm wide at
insertion, fibrillose to at least one third, commonly throughout; upper fibrillose tissue more or
less identical to that of the branch leaves, lower tissues usually with numerous septa and
enlarged, adaxial resorption gaps; apices rounded-obtuse, dentate or eroded, sometimes
slightly cucullate; border not expanded into prosenchymatous patches. Branch leaves typically
broadly lanceolate, becoming narrower distally, in weak plants all leaves more or less narrowly
lanceolate to linear; spiral or partially to completely 5-ranked, incumbent or erect-spreading,
not curved or homomallous, symmetrical; typically leaves range from 1-4-2-8 mm long, 0-6—1-2
mm wide, in weak plants 1-2—2-4 mm long, 0-4—0-8 mm wide. Border narrow, of 2 series of cells;

A

A REVISION OF AFRICAN SPHAGNALES 143

Fig. 38 Sphagnum planifolium C. Mill. var. planifolium. A, branch cortex; B, branch leaves; C, upper
branch leaf; D, stem leaf; E, branch fascicle; F, transverse section of stem (all drawn from the type
collection of S. albicans).

ALAN EDDY

a eues

Enon

S\ererd 7

Beads)
50°52 Oe

Fig. 39 Sphagnum planifolium C. Mill. var. planifolium. A, transverse section of leaf; B, branch fascicle;
C, abaxial (right) and adaxial (left) surfaces of branch leaf; D, stem leaf; E, branch leaves; F, branch
cortex; G, transverse section of stem (drawn from Staudt 876).

A REVISION OF AFRICAN SPHAGNALES

» S.

ae

Re a a

~ ie E is
vovee. i:

Fig. 40 Sphagnum planifolium C. Mill. var. planifolium. A, abaxial (left) and adaxial (right) surfaces of
branch leaf; B, adaxial surface of branch leaf from mid leaf (upper fig.) and near apex of leaf (lower fig. );
C, abaxial surface of stem leaf, upper fibrillose part; D, abaxial surface of branch leaf, lower-lateral leaf
region. E, transverse sections of leaves from apical region of leaf (upper fig.) and mid leaf (drawn from

various sources).

146 ALAN EDDY

apex truncate-dentate, rather wide but often appearing narrow due to inrolled upper leaf
margins. Leucocysts, in broader leaves 20—25(—35) um wide, 100-180 um long, in narrow leaves
12-20 x 90-180 um. Abaxial surface sometimes without pores but with few to numerous
pseudopores in the cell angles and scattered along the commissures, frequently 1— few small
pores present, usually confined to the cell angles, and an inconspicuous resorption gap in the
apical angle. Adaxial surface typically with a small to medium-sized, unringed or thin-ringed
pore (c. 4-0-6-0 um diameter) near (but not against) the commissure in most cell angles, rarely
with additional pores scattered along the commissures. T.S. leaf. Leucocysts shallowly convex
on the abaxial side, strongly convex adaxially. Chlorocysts triangular to trapezoid, always with
much wider abaxial exposure and, in robust plants, frequently distinctly immersed on the
adaxial side; cell walls more or less thin to distinctly incrassate, lumina oval-triangular.
Dioecious. Male plants apparently very rare; perigonial leaves similar to normal leaves. Female
plants apparently less rare, but no fertilized material seen, therefore no fully developed
perichaetia were available for description; fruit and spores unknown.

Distribution. This species favours habitats with a high water-table, the most robust plants having
been collected where these habitats are subject to high degrees of oceanicity. The main
occurrences, judged on frequency of gatherings, are in the afromontane areas of Ruwenzori and
Madagascar, with outlying stations in West Africa (Cameroon). There are no records of the
species from southern Africa. See under varieties.

17a. Sphagnum planifolium var. planifolium
(Figs 38-40)

S. madegassum C. Mill. in Flora, Jena 70: 415 (1887).

S. albicans Warnst. in Hedwigia 32: 3, tab. 1 fig. 2 (1893).

S. bessonii Warnst. in Hedwigia 32: 4, tab. 1 fig. 3 (1893).

S. cardotii Warnst. in Hedwigia 32: 5, tab. 1 fig. 4, tab. 2 fig. 4 (1893).

S. pulchricoma Warnst. in Bot. Zbl. 82: 44 (1900), hom. illeg., non S. pulchricoma C. Mill. (1848).

Robust plants, with broadly lanceolate branch leaves and more or less adaxially immersed
chlorocysts, accord well with Miller’s taxon. Linear-lanceolate to linear branch leaves, in this
form, are confined, when present at all, to the distal parts of branches. Stem leaves in the var.
planifolium may be fibrillose throughout, but not infrequently have the basal portions efibril-
lose. Pores and pseudopores were not always distinguished by earlier authors, the latter (but not
the former) sometimes being numerous on the convex leaf surfaces.

West Africa. 'VORY COAST: 15 km east of Abidjan, Guillaumet 1887 (BM). CAMEROON: ‘Johann
Albrechts Hohe’, Staudt 876 (BM). GABON: ‘Wasserfall des Mougoungoulou auf dem Berg Iboundgi’,
Le Testu s.n. (BM — type collection of S. potieri, fragment only). ZAIRE: between Leopoldville and
Labuka, Laurent 34; Laurent s.n., 1905 (BM, H, PC); marais de Boko, Pynaerts.n. (PC).

Southern Africa. CAPE PROVINCE: Little Winterhoek, Lefson 3312 (BM).

Mlanje/Chimanimani. ZAMBIA: Mwinilunga, on peaty soil among sedges, 1400 m alt., Townsend
75/146; Mpika, swamp on mountain slope, 1725 m alt., Van Zinderen Bakker 973 (BM, form approaching
var. rugegense); Shiwa Ngandu, in swamp by river, 1400 m alt., Van Zinderen Bakker 890 (BM).
ZIMBABWE: Melsetter District, source of Bundi River, 1800 m alt., Chase 6898 (BM).

Ruwenzori/Usambara. ZAIRE: Kivu, crater lake, Tchikesi, Humbert s.n. (PC); Shaba Province
(Katanga), Kundelungu Plateau, 1700 m alt., Townsend 75/268, 75/283 (BM). RWANDA: ‘Marais de la
riviére Hikiberanya’, Rugege, 2450 m alt., De Sloover 13789 (NAM); ‘dans un ruisseau traversant un
marais’, Massif des Birunga (= Virunga), 2400 m alt., De Sloover 13703 (BM, NAM). UGANDA:
Muhavura, Gahinga Pass, c. 9000 ft [2740 m] alt., in bog in open moorland, Burtt 2852 (BM); Kigezi,
locally dominant in swamp, 2250 m alt. , Purseglove 3737 (PRE); same locality, 1800 m alt., Lind 1A (BM);
Kashambya Swamp, Lind s.n. (BM); Burbaria Swamp, Lind 2 (BM); Marius Bog, Lind 6 (BM); Lake
Nabugabo, ‘abundant in waterlogged cavities’, Wood 1068 (BM); Papyrus swamp, Eggeling 794 (BM).
KENYA: South Nyeri District, Aberdare Mountains, in boggy ground, Townsend 75/892. (BM).
TANZANIA: Bukoba, Stuhlmann 1062 (BM, H, PC - type collections of S. albicans), 1690 (BM); ‘.. .
M’pala am Westufer des Tanganjikasees’, Guillemé 220 (H, as S. pulchricoma sensu Warnst.); Bukoba,
1120 m alt., Haarer 2083 (BM); Geilinger 537 (BM, mixed with S. africanum).

A REVISION OF AFRICAN SPHAGNALES 147
\

a) De ye

A :
4)
%

—<—=

Fig.41 Sphagnum planifolium C. Mill. var. angustilimbatum (Warnst.) A. Eddy. A, transverse sections
of leaves; B, stem leaf; C, branch leaves; D, abaxial (left fig.) and adaxial (right figs) surfaces of branch
leaves; E, stem leaf (upper fig.) and branch leaves (E drawn from the type of S. congoanum; the
remainder from the type of S. angusti-limbatum).

East African Islands. MADAGASCAR: between Vinanintelo and Ikongo, Besson s.n. (PC — isotype of
S. bessonii); Imerina, Hildebrandts.n. (BM —type collection of S. madegassum, mixed with S. tumidulum);
Ankazondandy, Boiteau 116 (PC); Tananarive, Mantasoa, in marsh, 1400 m alt., Onraedt 74.M.2105,
74.M.2108 (BM, with S. pallidum); Omalabe, north of Tananarive, 1300-1400 m alt., Bosser 12309 (BM,
NAM).

148 ALAN EDDY

17b. Sphagnum planifolium var. angustilimbatum (Warnst.) A. Eddy, comb. nov.
(Fig. 41)

S. angusti-limbatum Warnst. in Allg. bot. Z. 1: 135 (1895). — S. albicans Warnst. var. angusti-limbatum
(Warnst.) Warnst., Sphagnol. univ.: 263 (1911). Type: ‘Centralafrikan Seengebiet’, Stuhlmann 3927 (H,
herb. Brotherus).

S. stuhlmannii Warnst. in Allg. bot. Z. 1: 172 (1895).

S. congoanum Warnst. in Ren. & Cardot in Bull. Soc. r. bot. Belg. 41 (1): 8 (1905). — S. planifolium var.
congoanum (Warnst.) Warnst., Sphagnol. univ.: 223 (1911).

S. potieri Paul in Revue bryol. lichén. 4: 72, fig. 7, pl. 2 (1931). Note: duplicates of the type collection, with
the locality data ‘Bukoba’ (in Tanzania) are in BM.

Flaccid plants with the majority of branch leaves narrowly lanceolate to linear, with correspon-
dingly narrow leucocysts (i.e. under 20 wm) and chlorocysts mainly exposed on both leaf
surfaces. Such plants may almost lack dorsal pores and/or pseudopores, and the resemblance to
S. cuspidatum is striking. Sometimes the only reliable character by which this taxon may be
distinguished from S. cuspidatum is the non-expanded border of the stem leaf. Forms
approaching this variety seem to be proportionately more frequent in the western than in the
eastern parts of the range of the species.

West Africa. ZAIRE: Kisantu, 550 m alt., Gillets.n. (PC); Kwango, Jormain 2695 (PC); Ndembo, Boko,
Vanderyst s.n. (several gatherings, PC); ‘Entre Dembo (sic) et Kisantu’, Gillet 1535 (BM — type collection
of S. congoanum).

Mlanje/Chimanimani. ZIMBABWE: Chimanimani, 1750 m alt., Wild 47953 (PRE), Lacey 33, 34, 35
(BM).

Ruwenzori/Usambara. UGANDA: Liebenburg (BM, no data). TANZANIA: Bukoba, Stuhlmann
3927 (BM, H - type collections of S. angusti-limbatum), Stuhlmann 4139 (BM).

East African Islands. MADAGASCAR: ‘Forét littorale Tampolo, Fenerive’, Guillaumet 2365 (BM,
NAM).

17c. Sphagnum planifolium var. rugegense (Warnst.) A. Eddy, comb. nov.
(Figs 42 & 43).

S. rugegense Warnst. in Mildbr. in Wiss. Ergebn. dt. ZentAfr. Exped. 2: 135 (1910). Syntypes: East Africa,
Rugege, 1900 m alt., Mildbraed 817, 984 (B, destroyed, BM).

S. recurvatum Warnst. in Mildbr. in Wiss. Ergebn. dt. ZentAfr. Exped. 2: 135 (1910).

? S. ruwenzorense Negri in Annali Bot. 7: 161 (1908).

S. macromolluscum Dixon in J. Bot., Lond. 76: 218 (1938).

Plants varying from rather small and compact to more or less robust but usually smaller and
more compact than in var. planifolium, commonly tinged brown with pale brown stems. Stem
leaves variable, mainly as in the type variety but shorter on average and less extensively
fibrillose. Branch leaves all of the short form, ovate-lanceolate to broadly lanceolate. Hyaline
cells of branch leaves typically with numerous small pores and/or pseudopores on the abaxial
side; adaxial side with abundant, often almost seriate, rather small (4-0-7-0 um), sometimes
distinctly ringed pores.

Distribution. Forms of S. planifolium that can be assigned unequivocally to the variety
rugegense are recorded only from high altitudes in the Ruwenzori/Usambara ranges. Scant
field data indicate that the plants occur in open, flushed situations but not submerged. Plants
having a morphology approaching typical S. planifolium are more widespread, extending
further southwards in range, and also occurring in the East African islands.

Mlanje/Chimanimani. ZAMBIA: Mpika, swamp on mountain slope, 1725 m alt., Van Zinderen Bakker
973 (BM, PRE, intermediate form between var. rugegense and var. planifolium).

Ruwenzori/Usambara. RWANDA: Rugege, 1900 m alt., Mildbraed 817, 984 (BM -— syntypes of S.
rugegense). UGANDA: in bog on shore of Lake Nabugabo, 1150 malt., Eggeling 571 (BM-—holotype of S.
macromolluscum). KENYA: Mount Kenya, 3000 m alt., Mearns 1727 (BM). TANZANIA: ‘. . . M’pala
am Westufer des Tanganjikasees’, Guillemé 220 (H).

Just as the number of synonyms (above) reflects to some degree the variability of S. planifolium,

A REVISION OF AFRICAN SPHAGNALES 149

Fig. 42 Sphagnum planifolium C. Mill. var. rugegense (Warnst.) A. Eddy. A, branch fascicle; B, stem
leaf (left) and branch leaves (right); C, abaxial (left) and adaxial (right) surfaces of stem leaf near leaf
apex; D, abaxial (left) and adaxial (right) surfaces of branch leaf; E, transverse section of stem; F,
transverse sections of branch leaf (all drawn from Mildbraed 817).

150 ALAN EDDY

faraeer
te et in eae al
100 ip
Fig. 43 Sphagnum planifolium C. Mill. var. rugegense (Warnst.) A. Eddy. A, branch leaves; B, stem

leaves; C, abaxial (left) and adaxial (right) surfaces of branch leaves; D, branch cortex; E, transverse
section of stem (all drawn from the holotype of S. macromolluscum).

so the changes in status of many, particularly in Warnstorf’s treatments, indicate the difficulties
encountered when trying to interpret the taxonomic value of the various expressions of the
species. At one extreme, plants are broad-leaved, with a facies similar to the most robust forms
of the European S. riparium Aongstr.; at the other extreme we have flaccid, weak, narrow-
leaved plants superficially almost identical to S. cuspidatum. In relation to S. planifolium the
situation is clearly comparable to that pertaining to S. auriculatum in Europe, and the equally
polymorphic S. truncatum Hornsch. in Africa. The indeterminate genetic heterogeneity, which
is to be expected in a taxon that occurs in widely scattered localities, is so confused with
ecotypical plasticity that, at present, it is almost impossible to extricate significant and stable
variants from the matrix of forms within the species. Presenting the synonymy of S. planifolium
under separate varietal names implies a confidence not wholeheartedly possessed by the author.

The suppression of useful diagnostic characters by immersion, reduced light intensity, and
other factors is well known in Sphagnum, and the associated elongation of leaves, reduction of
pores, and tendency towards isophylly are familiar phenomena among European taxa. Con-
versely, exposure under conditions of high insolation tends to produce the opposite effects. It is
quite possible therefore that the varieties (above) of S. planifolium are produced predominantly
by ecological pressures and it would be instructive to determine to what degree their essential
characters are reversible.

In addition to the varieties of S. planifolium described above, there are plants which
superficially resemble S. cuspidatulum or S. recurvum P. Beauv., in that the branch leaves

A REVISION OF AFRICAN SPHAGNALES ily!

are smaller than average, uniform in size, and 5-ranked. In many of these plants such
resemblance is further enhanced by the tendency for the leaves to become recurved in the dry
condition. There are, however, no anatomical features to suggest that such forms are anything
more than local expressions of the species or of its variety augustilimbatum. S. slooveri (below)
shares some of the superficial characters mentioned above, but is readily distinguished by its
reflexed truncate-fimbriate, almost efibrose stem leaves.

S. planifolium is clearly an archaic taxon in the sense that it possesses a number of features that
can be construed as primitive. Branch-leaf structure in unremarkable, being fairly average for
section Cuspidata while at the same time showing the kind of flexible response to environmental
conditions that seems to be particularly marked in ‘primitive’ taxa rather than those whose
derived characters are more fixed. The stem leaves are unequivocally primitive, both in the
isoareolate nature of their upper tissues and in the unexpanded border. Relationships between
S. planifolium and other species of the section are unclear, although within the matrix of forms
that constitute the species (as interpreted in this work) there can be detected tendencies towards
more advanced forms (e.g. S. recurvum P. Beauv.), and S. planifolium var. rugegense has been
compared with S. mendocinum Sulliv. (Warnstorf, 1910); at the other end of the scale, S.
planifolium var. angustilimbatum is morphologically very close to the Malaysian S. cuspidatum
subsp. subrecurvum (Warnst.) A. Eddy and, perhaps, to the austral species, S. falcatulum
Bescher.

18. Sphagnum slooveri A. Eddy, sp. nov.
(Fig. 44)

Plantae aliquantum parvae, compactae, flavidae. Epidermis caulina stratis 2(—3), a cylindro crasso pallido
plerumque plane distincta. Fasciculi ramorum 4-5 compositi; rami non diversi. Epidermis tamen eorum
diversa; cellulae perforatae plerumque binatim in longitudinem dispositae, non rostratae. Folia caulina
deflexa, triangulata, circa 1-0 mm longa, basi 0-9 mm lata, apicibus late truncatis, circa 0-3 mm latis,
eroso-fimbriatis; limbus superne 3-5 cellulas latus, deorsum distincte dilatatus; cellulae hyalinae in
apicibus fibrosae et interiore superficie poris magnis obsitae; ceterae efibrosae pauciporosae vel eporosae.
Folia ramorum dense imbricata, semper quinquefaria, uniformia, lanceolata, circa 1-3-1-4 mm longa,
0-5 mm lata, apicibus angustis, 3-4 dentatis, involutis, ut videtur acutis vel mucronatis; limbus angustissi-
mus, 1-2 cellulas latus. Cellulae hyalinae ad 20-0 um latae, 100-140 um longae, exteriore superficie
eporosae vel poris singulis in angulis apicalibus obsitae sed pseudoporis multis saepe seriatis ad commis-
suras instructae; interiore superficie multiporosae, poris minutis plerumque incrassato-marginatis, 2-0-4-0
um diametro, prope commissurae nec confinibus dispositis. Cellulae chlorophyllosae sectione transversali
triangulae in interiore folii superficie inclusae (cellulae hyalinae ubi se contingunt parietibus coalitis).
Typus: Africa, Burundi, ‘tourbiére d’Ijenda, 1950 m alt., sur le sol tourbeux’, Jean Louis De Sloover
19.181 (BM — holotypus; NAM — isotypus).

Plants small to medium-sized, rather dense, resembling compact forms of S. cuspidatulum,
yellowish. Stems relatively thick, about 0-9 mm diameter, pale yellowish to brownish yellow;
cortical cells in mainly 2 layers, relatively distinct, imperforate; internal cylinder pale. Fascicles
rather closely set, composed of (4-)5 branches which are rather short, up to 10-0 mm long;
pendent branches not distinguishable. Branch cortex dimorphic but not conspicuously so; retort
cells mainly in linear pairs, relatively short and scarcely rostrate; walls of all epidermal cells
relatively firm. Stem leaves reflexed, about 1-0 mm long and about 0-9 mm across the insertion,
triangular to triangular-lingulate, narrowed from the insertion to broad, more or less truncate,
distinctly fimbriate apices about 300 um across; fibrillose only in a small zone just below apex;
border strong, distinctly widened in the lower half. Adaxial surface of fibrillose leucocysts with
large, circular or irregular resorption gaps on the adaxial side; abaxial side mainly imperforate.
Branch leaves rather small, regular in size and arrangement and closely and regularly quinque-
farious, lanceolate, 1-3-1-4 mm long, about 0:5 mm wide, involute above and tapering to

_ Narrow, tubular, 3—4-dentate tips (which appear to be acute under low magnification). Border

narrow, 1-2 cells wide, without a resorption furrow. Hyaline cells of branch leaves rather
uniform, 20-0 um wide on average, 100-140 um long; abaxial surface without or with only 1-2
pores (of which one is typically in the apical cell angle) but with abundant, often almost seriate

152 ALAN EDDY

DY GCE

AY)

SSS DSSS

Aes
i. eS

— ACT oo

@o2:2; .

oot eccees
ee ae

Fig. 44 Sphagnum slooveri A. Eddy. A, branch fascicle; B, stem leaves; C, branch leaves; D, branch
cortex; E, adaxial (left) and abaxial (right) surfaces of branch leaf; F, apex of stem leaf; G, adaxial
surface of fibrillose region of stem leaf; H, transverse section of stem; I, transverse sections of leaf (all
drawn from the holotype).

A REVISION OF AFRICAN SPHAGNALES 153

pseudopores; adaxial surface with numerous, very small, unringed or frequently heavily ringed
pores set at some distance from the commissures. T.S. leaf. Hyaline cells more or less plane on
the abaxial side, shallowly convex on the adaxial side. Chlorocysts in section oval-triangular,
broadly exposed on the abaxial side but rather deeply immersed below the adaxial side.
Dioecious? Only the female, but unfertilized plant present in the type collection.

Distribution. At present known only from the type locality, in boggy ground.
Ruwenzori/ Usambara. BURUNDI: ‘tourbiére d’Ijenda’, 1950 m alt., De Sloover 19.181 (BM -—
holotype, NAM - isotypes).

This species clearly represents the end-point of an evolutionary line that connects it with S.
planifolium via S. planifolium var. rugegense. At present, the discontinuity manifested in
stem-leaf morphology and, to a lesser extent, the size and arrangement of the branch leaves has
been deemed sufficient to support the recognition of S. slooveri as a species distinct from S.
planifolium var. rugegense. This course also has the advantage of avoiding the addition of yet
another infraspecific epithet to S. planifolium itself. Future collections, however, may reveal
plants of intermediate character, at which time the whole position regarding S. planifolium and
its allies may have to be reviewed.

The species is named in honour of Dr Jean Louis De Sloover, whose numerous collections of
Sphagnum from eastern Africa and other regions have been made available for study by the
author, and have helped to solve many of the systematic problems of the genus in Africa.

19. Sphagnum cuspidatulum C. Mill.
(Fig. 45)
in Linnaea 38: 549 (1874). Type: India, Khasia, J. D. Hooker & T. Thomson 1284 (BM - isotype).

Plants elongate, of regular appearance with well marked capitula, pure green to yellow-green.
Fascicles uniform in appearance and regularly disposed along stems, composed of (4-)5
branches which are distinctly dimorphic: 2—3 spreading branches elongate, (12-0—)15-0—-25-0 mm
long, relatively narrow and multifoliate (more than 80 fully developed leaves per branch); 2-3
pendent branches more or less appressed to stems, of variable length, up to 30-0 mm. Stem
rather strong and rigid, 0-7—-0-9(-1-1) mm diameter; cortex more or less distinct but tending to
intergrade with the internal cylinder, in 2-3 layers of only moderately enlarged leucocysts, never
with pores in the exposed wall and often somewhat incrassate; internal cylinder yellowish, rarely
pale brown. Branch cortex dimorphic; retort cells distinct, in linear pairs, sometimes slightly
protuberant at apertures. Stem leaves lingulate or more or less trapezoid, reflexed, 1-2-1-5 mm
long, 0-7-0-9 mm wide at insertion; apices rounded to broadly truncate and conspicuously
erose-fimbriate across the whole apical region; border evanescent above, slightly to moderately
expanded below but rarely very conspicuously so. Leucocysts narrow at mid leaf and below,
expanded and broadly rhomboid in apical region and largely resorbed; septa numerous. Branch
leaves uniform in size and attitude, very numerous (more than 80 per branch), strictly 5-ranked;
leaves lanceolate, apical parts sub-tubular with strongly incurved margins, 1-4-1-8 mm long,
0-7-0-8 mm wide, truncate-dentate; border 1-2 cells wide, without resorption furrows. Pendent-
branch leaves mainly ovate-lanceolate, concave, 0-9-1-4 mm (longer and narrower on distal,
tapering parts of branches). Leucocysts in apical leaf area narrow, especially on abaxial side,
11-0-18-0 x 70-120 um on abaxial side, 13-0—20-0 um wide adaxially in upper mid leaf, up to
30-0 um wide in basal-lateral area of leaf; abaxial surface usually without normal pores but
commonly with a few rather small pseudopores about the lateral angles and frequently with an
inconspicuous resorption gap in the apical angle; adaxial surface with 3-6 medium-sized,
circular, unringed or faintly ringed pores, c. 6-0-10-0 um diameter, mainly adjacent to the cell
angles (these pores often very faint and only visible after staining). Leucocysts of pendent-
branch leaves not significantly different from those of spreading branches anatomically. T.S.
leaf. Leucocysts more or less plane to shallowly convex abaxially, strongly convex adaxially.
Chlorocysts isosceles-triangular, with rather wide (3-0-6-0 um) exposure on abaxial side, mainly

154 ALAN EDDY

KG pe Yr ty
B55 SS = \ za 0
bo Cie
x ee } f
x ie
Y < SA A <<
4 } A J .

TI
‘é 0 mm_j

{h
at

ca
mi ia ny

Fig. 45 Sphagnum cuspidatulum C. Mill. A, transverse section of stem; B, branch fascicle (with male
branch); C, stem leaves; D, abaxial (left) and adaxial (right) surfaces of branch leaf; E, abaxial surface of
apical region of branch leaf; F, branch leaves; G, apex of stem leaf; H, transverse sections of branch leaf
(all drawn from Bosser 13076).

A REVISION OF AFRICAN SPHAGNALES 55

rather deeply immersed adaxially. Dioecious. Male branches more or less identical to sterile
branches (at least in African material). Female plants unknown and fruit unrecorded.

Distribution. A terrestrial species of tropical oceanic regions, widespread in Indo—Malaya from
western New Guinea to the Himalaya. In Africa known only from single localities in Madagascar
and Rwanda. Possibly this, or a very closely related taxon, also occurs in tropical America (as S.
pulchricoma P. Beauv. ?).

RWANDA: Bysoke, 3600 m alt., De Sloover 13441 (NAM).

MADAGASCAR: Niagarakety, Mt. Moramanga, Bosser 13076 (PC).

This is a well defined species which has no close relative among the other African taxa of section
Cuspidata. Small-leaved states of S. planifolium have very different, never completely efibrose,
stem leaves. In fact, the stem leaves of S. cuspidatulum, as far as African species are concerned,
are more reminiscent of those of S. fimbriatum.

There is a reference to a putative African mainland specimen of S. pulchricoma P. Beauv.
(Warnstorf, 1911: 189) but no trace of the species could be traced in the herbarium cited
(Brotherus). Probably this record is in error. However, the similarities between S. pulchricoma,
as presently and probably imperfectly understood by the author, and S. cuspidatulum are
very close, so much so that it may prove impracticable to maintain them as separate taxa in
future. The discovery of an unequivocal specimen of S$. cuspidatulum among unidentified
material from Madagascar raised the possibility that this taxon could occur in East Africa and
that it might be the plant to which Warnstorf referred.

Surprising though it may seem, that such a characteristic and by no means inconspicuous
species has never been recorded before or since the collection made of it by Bosser in 1959, there
is no reason to doubt the provenance of the specimen cited above. The more recent collection by
De Sloover in Rwanda extends the African range of the species considerably, and indicates a
more widespread occurrence in former times.

Subgenus RIGIDA

Sphagnum subgen. Rigida (Lindb.) A. Eddy in Bull. Br. Mus. nat. Hist. (Bot.) 5 (7): 431 (1977).
Sphagnum 4. Rigida (Lindb.) Schlieph. in Verh. zool.-bot. Ges. Wien 15: 413 (1865), excl. parte. — Lindb.,
Musci scand.: 11 (1879). Type: Sphagnum rigidum (Nees & Hornsch.) Schimper = S. compactum DC.

Plants large-leaved and normally robust, typically brownish or straw-coloured. Stem hyaloderm
well developed, 2—4-layered and foraminate. Branch hyaloderm not dimorphic, all or the
majority of leucocysts slightly to strongly protuberant and with a large pore. Branch fascicles
strongly dimorphic. Stem leaves small, often more or less vestigial. Branch leaves very large,
often exceeding 2-5 mm, commonly squarrose. Branch-leaf leucocysts uniform, rather short and
wide, not highly inflated; adaxial face pauciporose but always with pseudolacunae at the
conjunction of the basal and lateral convergence of three adjacent leucocysts; adaxial face
typically with few to numerous ringed pores. Antheridia typically borne on pendent, not
spreading branches.

This is a well defined, monosectional subgenus containing few species (probably only three in
all) with many features pointing to an evolutionary history paralleling that of subgenus Sphag-
num. The group is represented in Africa by a single taxon, S. strictum subsp. pappeanum which,
although reaching its maximum diversity in that continent, is probably a relict from a formerly
more widespread tropical and subtropical range that included Malaysia and central America.

20. Sphagnum strictum Sulliv. subsp. pappeanum (C. Mill.) A. Eddy
(Figs 46 & 47)

in Bull. Br. Mus. nat. Hist. (Bot.) 5 (7): 433 (1977). — S. pappeanum C. Mill., Syn. musc. frond. 1: 101
(1848). Type: Cape of Good Hope near Zwellendam, Pappe s.n., 1838 (BM, H).

S. lacteolum Bescher. in C. r. hebd. Séanc. Acad. Sci., Paris 81: 725 (1875).

S. patens Bescher. in Ann. Sci. nat. (Bot.) VI, 10: 329 (1880), hom. illeg., non S. patens Brid. (1806).

156

So,

mSulliv. subsp. pappeanum (C. Miill.) A. Eddy. A, branch fascicle; B, stem leaf
C, adaxial (left) and abaxial (right) surfa f branch leaf; D, upper margi
ion of branch leaf (all drawn from Rehmann 12).

Fig.46 Sphagnum strictu
e) and branch leaves; C, adaxial (left) and abaxia
of stem leaf; E, transverse sect

(centr

A REVISION OF AFRICAN SPHAGNALES LS i7

500 ~m

Fig.47 Sphagnumstrictum Sulliv. subsp. pappeanum (C. Mill.) A. Eddy. A, branch leaf; B, stem leaf; C,
adaxial (left) and abaxial (right) surfaces of branch leaf; D, abaxial leaf surface, lower lateral leaf region;
E, branch cortex; F, transverse section of stem (all drawn from the type collection of S. mildbraedii).

S. bescherellei Warnst. in Hedwigia 29: 240, tab. 11 fig. 12, tab. 14 fig. g (1890).

S. sparsifolium Warnst. in Hedwigia 33: 320 (1894). — S. pappeanum var. sparsifolium (Warnst.) Warnst.,
Sphagnol. univ.: 151 (1911).

S. mildbraedii Warnst. in Wiss. Ergebn. dt. ZentAfr. Exped. 2: 134 (1910).

Plants very variable in stature, short and compact to elongate and very robust, pale yellowish or
greenish varying to pale brown. Capitula sometimes concealed among upwardly directed
sub-comal branches. Fascicles closely packed or spaced at short intervals on stems, internodes
concealed by pendent branches, always very markedly dimorphic: spreading branches usually 2,
horizontal or ascending, ending abruptly or more or less suddenly contracted distally and
tapering, (15-0—)20-0—30-0(—40-0) mm long; pendent branches 2-4, strongly deflexed, pale, thin
and tapering, very variable in length from vestigial up to more than 40 mm. Stems pale, yellowish
to yellow-brown, 0-7—1-0 mm diameter; cortex well developed, composed of (2—)3—-4 layers of
hyaline cells, the outermost cells variable in diameter but most having a single large pore in the
external wall; internal cylinder well developed but cells less incrassate than in other sphagna of

158 ALAN EDDY

comparable vigour, yellowish to pale brown. Branch cortex monomorphic, composed of large
leucocysts, several or all of which have a large pore. Stem leaves, for the size of the plant, very
small, (0-7—)0-9-1-2(-1-45) mm long, c. 0-6-0-9 mm across insertion, triangular-ovate with
rounded, more or less eroded apices; border not, or only very slightly and indistinctly expanded
below; tissue normally fibrillose below apex, rarely to much beyond two-thirds, very occasional-
ly completely devoid of fibrils; leucocysts more or less eporose on abaxial surface but with
resorption gaps on the adaxial side. Branch leaves very large, (2-:0—)2-5—4-0 mm long, more or
less ovate to ovate-lanceolate but usually appearing to be abruptly contracted at about the
middle to a narrower, sub-tubular spreading to sub-squarrose limb, widest at about 1/4 to 1/3
above insertion (c. 1.5—-2:-5 mm); apices broad, truncate and strongly 7-10-dentate; border
narrow, of a single cell series, with a resorption furrow. Pendent-branch leaves delicate, the
lower more or less ovate, becoming narrowly lanceolate distally, often lacking a defined border.
Leucocysts usually very regular in size and arrangement in mid-leaf, broad, 30-40 um wide,
90-150 um long; abaxial surface very variable, typically with 1-4 rather large, thin-ringed pores
adjacent to the upper and upper-lateral angles, pores c. 12-0-20-0 um diameter, with or without
pseudopores, seldom lacking pores altogether; adaxial surface always with pseudolacunae
(‘triple pores’) at the confluence of basal and lateral angles of adjacent leucocysts, the external
common apertures frequently very narrow and tri-radiate, with or without additional simple
pores on the cell angles or elsewhere, with or without pseudopores. Leucocysts towards the
lower lateral leaf areas becoming progressively more porose on both sides, pores sometimes
becoming more or less serial. Hyaline cells of pendent-branch leaves all highly porose (resemb-
ling some species of section Subsecunda). Internal commissural walls occasionally smooth but
usually distinctly to rather coarsely papillose. T.S. leaf. Leucocysts very shallowly convex
abaxially, rather more inflated adaxially; partial septa (= fibrils) narrow. Chlorocysts narrow
with narrowly oval lumina, rarely entirely enclosed on both sides of leaf but frequently so on the
adaxial side; abaxial wall, at least, strongly thickened and narrowly exposed. Autoecious.
Antheridia borne on pendent branches, the perigonial leaves scarcely distinct. Lowest perichae-
tial leaves approaching stem leaves in morphology but median and upper very large and strongly
fibrillose; innermost bracts convolute, 4-5—6-5 mm long, elongate-cymbiform; uppermost tissue
more or less identical to that of branch leaves; lowermost resembling that of stem leaves, lacking
fibrils and with adaxial resorption-gaps. Fruit frequent (in favourable habitats); spores
32-35(—40) um diameter, light brown, with very thick walls and varying from almost smooth to
rather coarsely rugulose-papillose.

Distribution. Pan-tropical with wide disjunctions. Confined to regions of high oceanicity at high
altitudes or on islands: Sulawesi, New Guinea, Central America and the West Indies, Africa and
the East African islands. In Africa its main occurrence is in Cape Province in the south, the
Ruwenzori Range in the east, and on Réunion and Madagascar in the Indian Ocean.

Southern Africa. CAPE PROVINCE: ‘Prom. bon. spei prope Zwellendam’, Pappe s.n. (BM, H— type
collections of S. pappeanum); Montagu Pass, Rehmann exs. 12 (H); McOwans.n. (PRE). TRANSVAAL:
Marieskop, Schyff 5647 (PRE); Marieskop Summit, Vorster 1061 (PRE).

Ruwenzori/Usambara. ZAIRE: Karisimbi volcano, moor at 3400 m alt., Mildbraed 2074 (BM — type
collection of S. mildbraedii); same locality, Humbert 7841, 7842, 7843, 7845 (PC); river valley east of
Bahungu, 3180 m alt., Demaret 5277 (H); Runssoro, 2800-3000 m alt., Stuhlmann 2385 (BM) (same
gathering in H labelled ‘Ru-Nssoro 3100 m alt.’); Kivu, several collections from 3000-3300 m alt., De
Sloover 12788, 12789, 12799 (BM, NAM), Troupin 2573 (BM). RWANDA: Bysoke, 3600 m alt., De
Sloover 13442 (NAM); Gahinga, 3440 m alt., De Sloover 13541 (NAM); Birunga, Karisimbi, 3000-3900 m
alt., De Sloover 13029, 13214 (NAM). UGANDA: Mgahinga Nigezi, in swamp in crater, 3300 m alt.,
Thomas 2446 (BM), Lind 13, 14 (BM); Toro District, Ruwenzori at 3500 m alt., Osmaston 3769 (BM);
Ruwenzori, 5000 m alt [?], Grinrod s.n. (BM, no other data); Ruwenzori, 12000 ft [3500 m] alt., Hancock
113 (BM). TANZANIA: Lupanga, mist forest at 2000 m alt., Lovetts.n. (BM).

East African Islands. MADAGASCAR: Mt Tsaratanana, 2400 m alt., P. de la Bathie 16228 (PC,
fertile); Mt Papanga, near Befotaka, on irrigated siliceous rocks, Humbert s.n. (PC). REUNION:
Lépervanche 17 (BM, scraps in herb. Hampe; PC - type collection of S. bescherellei); Mafaté, Rodriguez
s.n. (H); Forestry post of Bébour, edge of stream at 1600 m alt., Onraedt 69.R.0602 (BM); Cirque de
Cilaos, 1950 m alt., De Sloover 17611, 17616 (NAM); Cirque de Cilaos, Coteau Kerreguen, 2250 m alt.,

A REVISION OF AFRICAN SPHAGNALES 159

Gimalac 70.R.3507 (BM, NAM). AMSTERDAM ISLAND: G. de I’Isle s.n. (PC — isotype of S.
lacteolum).

Examination of the considerable number of available gatherings of this species has demon-
strated that, in spite of considerable variation in size, there are no other morphological features
upon which taxonomic subdivisions can be sensibly made. The variety S. pappeanum var.
sparsifolium of Warnstorf is simply the more robust expression of the species from favoured
habitats. S. mildbraedii Warnst. on the other hand, is a rather starved or repressed state which
turns up from time to time among material from particularly high and exposed situations on
Ruwenzori. In these the leaves are not obviously squarrose, the branch-leaf cells are narrower,
and the pores are sometimes more numerous and smaller. However, intermediate forms are not
rare, and even within a gathering the features tend to be inconstant, indicating that the genetic
element in the variation is less significant than the ecological element.

There should be no problem in identifying this species, even at a casual glance. The
sub-squarrose, large branch leaves, extremely dimorphic branches, and minute stem leaves is a
combination of features not found in any other African taxon. Species of subgenus Sphagnum,
even when they lack cortical fibrils, have cucullate, not truncate-dentate leaves and larger stem
leaves. Occasional squarrose-leaved states of the larger-leaved members of section Subsecunda,
e.g. S. truncatum, have large stem leaves, and monomorphic branches.

S. strictum appears to be a species that has rather narrowly circumscribed ecological
requirements, demanding a pronouncedly oceanic environment and seemingly intolerant of
sustained high or low temperatures. This applies equally or possibly even more strongly to S.
strictum subsp. pappeanum than it does to subsp. strictum. In East Africa, S. strictum subsp.
pappeanum occupies a rather narrow altitudinal zone, being more or less confined to 2500-
3500 m alt. , but descends to much lower altitudes on mountain ranges facing onshore winds in the
East African islands and the Cape region of southern Africa. In common with the other species
of subgenus Rigida, this taxon is intolerant of immersion and is therefore unable to extend its
habitat range by adopting a semi-aquatic mode of existence. On the other hand, it is a
monoecious and frequently fruiting plant, so that regeneration from spores may to some extent
compensate for the barriers to survival imposed by its special needs.

Acknowledgements

I am indebted to the staff of the herbaria B, COI, FHO, H, NAM, PC, and PRE who made available the
collections essential to this revision, with special thanks to Dr P. L. Isoviita, Dr J. L. De Sloover, and Dr
W. Schultze-Motel. Fr. M. Onraedt very generously loaned his extensive collections from the East African
islands, held in his private herbarium in Malonne, without which many problems would have remained
unsolved. Thanks are due to M. R. Baudoin for his guidance during a visit to the Laboratoire de
Cryptogamie, Paris. During a brief sojourn in South Africa, fieldwork benefited from the guidance and
hospitality extended by Dr R. E. Magill, Mr C. Everson (South African Institute of Forestry), and
Professor E. A. C. L. E. Schelpe (University of Cape Town). Finally, I wish to express my appreciation for
the unstinted editorial efforts of Dr A. J. Harrington.

References

Axelrod, D. I. & Raven, P. H. 1978. Late Cretaceous and Tertiary vegetation history of Africa. InM. J. A.
Werger (Ed.), Biogeography and ecology of southern Africa 1: 77-130. Den Haag. (Monographiae biol.
31).

Braithwaite, R. 1880. The Sphagnaceae or peat-mosses of Europe and North America. London.

Bridel, S. E. de 1806. Muscologiae recentiorum supplementum seu species muscorum 1. Gotha.

—— 1826. Bryologia universa 1. Leipzig.

Cardot, J. 1897. Répertoire sphagnologique. Bull. Soc. Hist. nat. Autun 10 (1): 235-432.

Eddy, A. 1977. Sphagnales of tropical Asia. Bull. Br. Mus. nat. Hist. (Bot.) 5: 357-445.

—— 1979. Taxonomy and evolution of Sphagnum. In G. C. S. Clarke & J. G. Duckett (Eds), Bryophyte
systematics: 109-121. London.

Garside, S. 1949. Sphagnum in South Africa. JIS. Afr. Bot. 15: 59-78.

160

Hornschuch, C. F. 1841. Muscorum frondosorum novorum, quos in Africa australiori collegerunt Ecklon,
Drége, Mundt et Maire, descriptiones. Linnaea 15: 113-157.

Isoviita, P. 1966. Studies on Sphagnum L. 1. Nomenclatural revision of the European taxa. Annis bot.
fenn. 3: 199-264.

Magill, R. E. 1981. Flora of southern Africa. Bryophyta part 1, fasc. 1. Pretoria.

Miller, C. 1887. Sphagnorum novorum descriptio. Flora, Jena 70: 403-422.

Schimper, W. P. 1857 [‘1858’]. Mémoire pour servir a l’histoire naturelle des sphaignes (Sphagnum L.).
Mem. prés. div. Sav. Acad. Sci. Inst. Fr. 15: 1-97, plates 1-24.

Sim, T. R. 1926. The Bryophyta of South Africa. Trans. R. Soc. S. Afr. 15: i-i1i, 1-475.

Smith, P. J. 1976. So Madagascar was to the north. Nature, Lond. 263: 729-730.

Taylor, J. & Thompson, A. 1955 [‘1954’]. Notes on Sphagna from Uganda. Kew Bull. 9: 517-521.

Warnstorf, C. 1890. Beitrage zur Kenntniss exotischer Sphagna. Hedwigia 29: 179-211, tabs 4-7; 213-258,
tabs 8-14.

—— 1891. Beitrage zur Kenntniss exotischer Sphagna. Hedwigia 30: 12-46, tabs 1-5; 127-178, tabs 14-24.

—— 1892. Einige neue exotische Sphagna. Hedwigia 31: 174-182, tabs 16-17.

—— 1893. Beitrage zur Kenntniss exotischer Sphagna. Hedwigia 32: 1-17, tabs 1-4.

—— 1895. Beitrage zur Kenntnis exotischer Sphagna. Allg. bot. Z. 1895: 92-95, 115-117, 134-136,
172-174, 187-189, 203-206. 227-230.

—— 1897. Beitrage zur Kenntniss exotischer Sphagna. Hedwigia 36: 145-176.

—— 1900. Weitere Beitrage zur Kenntniss der Torfmoose. Bot. Zbl. 82: 7-14, 39-45, 65-76.

— 1901. Sphagnaceae. Jn K. Fritsch, Beitrag zur Flora von Angola. Bull. Herb. Boissier Il, 1:
1086-1087.

— 1902. Vier neue exotische Sphagna. Magy. bot. Lap. 1: 43-46.

—— 1905. Vier neue exotische Sphagna. Allg. bot. Z. 1905: 97-101.

—— 1907. Neue europdische und aussereuropaische Torfmoose. Hedwigia 47: 76-124.

— 1910. Sphagnales. Wiss. Ergebn. dt. ZentAfr. Exped. 2: 134-136.

—— 1911. Sphagnales — Sphagnaceae (Sphagnologia universalis). In A. Engler (Ed.), Das Pflanzenreich
51. Leipzig.

White, F. 1978. The Afromontane region. In M. J. A. Werger (Ed.), Biogeography and ecology of
southern Africa 1: 463-513. Den Haag. (Monographiae biol. 31).

Wijk, R. van der, Margadant, W. D. & Florschiitz, P. A. 1967. Index Muscorum 4. Utrecht. (Regnum veg.

ALAN EDDY

48).

Index

Accepted names are in roman and synonyms in italic; new names are in bold, as are principal references.

An asterisk (*) denotes a figure.

Acrosphagnum (C. Mill.) A. Eddy,
subsect. 77, 80, 116
aculeatum Warnst. 117
Acutifolia Wilson, sect., 79, 80, 90,
96, 116, 117, 137
acutifolium Schrader 90
var. borbonicum Ren. & Cardot
90
aequifolium Warnst. 108, 112*, 113
africanum Welw. & Duby 81, 106,
113, 114*, 115*, 130
afro-crassicladum Dixon & Sherrin
121, 124
albicans Warnst. 143*, 146
var. angusti-limbatum (Warnst.)
Warnst. 148
aloysti-sabaudiae Negri 137
angolense Warnst. 113, 116
angusti-limbatum Warnst. 147*, 148

arbogastii Ren. & Cardot 83, 84*,
85*, 86

auriculatum Schimper 82, 99, 100*,
106, 107, 108

austro-molle C. Mill. 124, 129

balfourianum Warnst. 83, 84*, 85*

bernieri Bescher. 137, 139*; 140*,
141

bescherellei Warnst. 157, 158

bessonii Warnst. 146

beyrichianum Warnst. 130, 133*

borbonicum Warnst. 90

bordasii Bescher. 106, 107*

capense Hornsch. 77, 82, 95, 124,
521267 IQs 28a asle
1B21837)

var. austro-molle (C. Mill.)

Warnst. 124
var. mollissimum (C. Mill.)
Warnst. 124
var. multiporosum Warnst. 124
capillifolium (Ehrh.) Hedw. 90, 96
cardotii Warnst. 146
ceylonicum Mitten ex Warnst. 77,
79, 82, 83, 95, 117, 124, 129,
135, 136*
chevalieri Warnst. 121, 123*, 124
compactum DC. 155
congoanum Warnst. 147*, 148
contortum Schimper
var. algerianum (Cardot)
Warnst. 99
convolutum Warnst. 101, 105
cordemoyi Warnst. 90, 92*, 95
coronatum C. Mill. 101, 105
var. cuspidatum Rehmann 101

var. falcatum C. Mill. ex
Warnst. 101
var. fluctuans Rehmann 101
cucullatum Warnst. 80, 108
Cuspidata (Lindb.) Schimper., sect.
78, 79, 80, 96, 137
cuspidatulum C. Mill. 77, 79, 81,
150, 153, 154*
cuspidatum Ehrh. ex Hoffm. 80, 81,
97, 137, 138*, 139*, 140*, 148
subsp. subrecurvum (Warnst.)
A. Eddy 151
var. latetruncatum Warnst. 101
Cymbifolia Schimper, sect. 83
cymbifolium var, bourbonense
P. Beauv. 83
var. patens (Brid.) Brid. 83

davidii Warnst. 82, 83, 116, 121,
21123", 129. 135
var. brachydasycladum Warnst.
121
var. flavo-fuscescens Warnst.
121, 124
var. viride Warnst. 121
denticulatum Brid. 101
drouhardii Cardot 83, 86

ericetorum Bescher. 90

ericetorum Brid. 77, 82, 116, 120,
132, 134*

eschowense Warnst. 106

falcatulum Bescher. 99, 142, 151
fimbriatum Wilson 81, 97, 98*
flavirameum C. Mill. ex Warnst. 108
fluctuans C. Mill. 101, 105

gabonense Bescher. 137, 140*, 141
goetzeanum Warnst. 130, 132*
gracilescens C. Mill. 79
grandirete Warnst. 87, 88*, 89*

hildebrandtii C. Mull. 113, 117
humbertii Cardot 124, 129
hypnoides Schimper ex. Warnst. 103

ikongense Warnst. 137, 138*, 141

imbricatum Schimper ex. Warnst.
117

Inophloea Russow, [sect.] 83

islei Warnst. 124, 128*, 129

Isocladus (Lindb.) Braithw., subgen.
80, 90

junghuhnianum Dozy & Molkenb.
90, 96

keniae Dixon 135, 136*, 137
kerstenii Hampe 135, 137

laceratum C. Mill. & Warnst. 90

A REVISION OF AFRICAN SPHAGNALES

lacteolum Bescher. 155, 159

lescurii Sulliv. 101

Litophloea (Russow) A. Andrews,
subgen. 90

macromolluscum Dixon 148, 150*
macrophyllum Brid. 90
madegassum C. Mill. 113, 146
magellanicum Brid. 87
subsp. grandirete (Warnst.) A.
Eddy 77, 81, 87, 88*, 89*
marginatum Schimper ex Warnst.
103, 105
var. convolutum (Warnst.)
Warnst. 103
marlothii Warnst. 83, 86
mathieui Warnst. 104*, 106
mauritianum Warnst. 106, 108, 113
mendocinum Sull IV. 151
meridense (Hampe) C. Mill. 90, 96
mildbraedii Warnst. 157*, 158, 159
mollissimum C. Mill. 124
var. elongatum Rehmann 124
Mucronata Warnst., sect. 116
mucronatum C. Mill. 116, 117

obovatum Warnst. 108, 110*, 113
obtusiusculum Lindb. ex Warnst. 90,
94*, 95
var. pallescens Warnst. 90
var. purpurascens Warnst. 90
oligodon C. Mill. 103*, 106
var. bachmannii Warnst. 106,
116
var. beyrichii Warnst. 106, 116
ovatum Hampe 116
oxycladum Warnst. 103, 105
var. mauritianum (Warnst.)
Warnst. 106

pallidum Warnst. 108, 111*, 113
palustre L. 80, 83
panduraefolium C. Mill. 124, 129,
130
pappeani Breutel 124
pappeanum C. Mill. 155, 158
var. sparsifolium (Warnst.)
Warnst. 157
patens Brid. 83, 155
patens Bescher. 155
perichaetiale Hampe 79, 81, 83, 84*,
85*, 89, 96
planifolium C. Mull. 78, 83, 137,
142, 153
var. angustilimbatum (Warnst. )
A. Eddy 77, 81, 142, 147*,
148
var. congoanum (Warnst.)
Warnst. 148
var. planifolium 82, 83, 143*,
144*, 145*, 146
var. rugegense (Warnst.) A.

161

Eddy 77, 82, 83, 148, 149*,
150% 153
potieri Paul 148
pugionatum C. Mill. ex Warnst. 117
pulchricoma Warnst. 146, 155
purpureum Schimper ex Warnst. 90
pycnocladulum C. Mill. 82, 120,
124, 129, 130, 131*, 132*,
133,135, 137
var. fuscescens Warnst. 130
var. viride Warnst. 130
Pycnosphagnum C. Mill., sect. 90
pylaesii Brid. 116

recurvatum Warnst. 148

recurvum P. Beauv. 150

rehmannii Warnst. 103, 116

reichardtii Hampe ex Warnst. 83, 95,
96, 97*

Rigida (Lindb.) A. Eddy, subgen.
79, 80, 155

rigidum (Nees & Hornsch.)
Schimper 155

rodriguezii Ren. & Cardot 90

rufescens (Nees & Hornsch.)
Warnst. 101

rugegense Warnst. 148

russowli Warnst. 96

rutenbergii C. Mill. 81, 106, 108,
LOS Ors di * a2"

ruwenzorense Negri 148

salvanii Warnst. 108, 112*, 113
scotiae Cardot 90, 93*, 97
slooveri A. Eddy 77, 83, 151, 152*
sparsifolium Warnst. 157
Sphagnum L., subgen. 79, 80, 83, 96
strictum Sulliv. 79, 155
subsp. pappeanum (C. Mill.)
A. Eddy 79, 81, 155, 156*,
157
stuhlmannii Warnst. 148
submucronatum C. Mill. ex Warnst.
117
subnitens Russow & Warnst. 90
subrotundifolium C. Mill. 124
Subsecunda (Lindb.) Schlieph., sect.
78, 79, 80, 96, 99, 116, 121
subsecundum Nees 99

tenellum (Brid.) Brid. 116
transvaaliense C. Mill. 103, 106, 116
truncatum Hornsch. 77, 78, 82, 83,
101, 103*, 116
var. bordasii (Bescher.) A.
Eddy 77, 82, 104*, 106, 107*
var. truncatum 82, 102*, 104*,
105
tumidulum Bescher. 79, 116, 117,
124, 131
var. confusum A. Eddy 77, 82,
116, 119*, 120, 135

162

var. macrophyllum Warnst. 117

var. microphyllum Warnst. 117

var. tumidulum 82, 116, 117,
118*

INDEX

ugandense J. Taylor & A. Thompson
135, 136*, 137

vandenbroeckii Nav. 135

violascens C. Mill. 83, 90, 91*, 92*,
93*, 94*

wenckei Roll 124

>

British Museum (Natural History)

Ferns of Jamaica
A guide to the Pteridophytes

G. R. Proctor

This flora records and describes the 579 species and 30 varieties of ferns occurring in
Jamaica. The succinct species descriptions include relevant synonymy and incorpo-
rate distributional data both within and outside Jamaica. Special emphasis is given to
the subtle distinctions between closely related species and all genera are illustrated.
Keys to the genera and species facilitate a wider use of the flora in the West Indies and
northern South America. The author, one time Senior Botanist in charge of the
Herbarium of the Science Museum, Kingston, Jamaica is an outstanding field
botanist and his expertise is reflected in the practicality of the flora and especially in
the habitat and ecological information. This volume represents an important addi-
tion to our knowledge of the flora of the West Indies.

1985, 650pp (approx), 135 line illustrations, 22 maps. Hardback.
0 565 00895 1 £50.00

Photoset by Rowland Phototypesetting Ltd, Bury St Edmunds, Suffolk
Printed in Great Britain by Henry Ling Ltd, Dorchester T.

Bulletin of the
British Museum (Natural History)

Studies in the genus Hypericum L.
(Guttiferae)

3. Sections 1. Campylosporus to
6a. Umbraculoides

Norman K. B. Robson

Botany series Vol 12 No 4 28 February 1985

The Bulletin of the British Museum (Natural History), instituted in 1949, is issued in four ‘
scientific series, Botany, Entomology, Geology (incorporating Mineralogy) and Zoology, a a:
and an Historical series. ?

Papers in the Bulletin are primarily the results of research carried out on the unique and
ever-growing collections of the Museum, both by the scientific staff of the Museum and by
specialists from elsewhere who make use of the Museum’s resources. Many of the papers are
works of reference that will remain indispensable for years to come.

Parts are published at irregular intervals as they become ready, each is complete in itself,
available separately, and individually priced. Volumes contain about 300 pages and several
volumes may appear within a calendar year. Subscriptions may be placed for one or more of
the series on either an Annual or Per Volume basis. Prices vary according to the contents of
the individual parts. Orders and enquiries should be sent to:

Publications Sales,
British Museum (Natural History),
Cromwell Road,
London SW7 5BD,
England.

World List abbreviation: Bull. Br. Mus. nat. Hist. (Bot.)

© Trustees of the British Museum (Natural History), 1985

The Botany series is edited in the Museum’s Department of Botany

Keeper of Botany: Mr J. F. M. Cannon
Editor of Bulletin: Mr J. R. Laundon
Assistant Editor: Dr A. J. Harrington

ISBN 0 565 08003 2
ISSN 0068-2292 Botany series tae

Vol 12 No 4 pp 163-325
British Museum (Natural History)
Cromwell Road a
London SW7 5BD Issued 28 February 1985

Studies in the genus Hypericum L. (Guttiferae) ~~
3. Sections 1. Campylosporus to 6a. Umbraculoides

Norman K. B. Robson
Department of Botany, British Museum (Natural History), Cromwell Road, London SW7 SBD

Contents

(08 DSS ce ccinen Sacer a: BAe EE ae ae e ene v oee r 163
il TIRERO CDI jo agepacceadepioacesbUSuL pede’ donee avGcdL coc ee GRee ee coc cee Cenc e eae nCe ee ae aaa ane 164
Ra EREER IN ree Ro. Per rin data baeg apo doa. sooswal «doen he «ann anhnwianedr idveteesasdue anne 164
DR PSB Spain abit te Bin «Pains Ae Binb oan civitmp emia Go eine odw'ge abe dened mares 165
DCS CTA TEL ELON Peer Ree etree ree aera RCM Nera eidcseencacsitecuionches desis sieweaehaeente 165

(a) Seer. LaCGMmpYIOSDOFUS ANG 2. PSOTOPAYIUM 2.02 .s0<ccsscanececesssnisaccsssenees 165

WE lee Se UL OUE fen sasha seman de macinnn tne wawae gn anndnaeAvaconteveesan'stivrassteewe ana 169
STEM TINS BOWIE fost tcn. acy tinea ares «2 neucacabceptovacdvantssnseseinate nad lciewssaces 174

Ree er eA TEDY OSATIIUIO cs osha. Ceetees Moeticatec seco stteuierae cdbettheadschcensatenton 175

KO) MSCCUNG MIMO COT ater rad crt settee tecet acta cm sree moe iensdiccitcledanedecasesavet ade 176

Ci RSecieGastiabracwlondays sect MOVs ..00600.00. aikeks danelawicves deainneedseveseeens 176
eae re RE AROIOME 7 cep sa. can wcn esd sceeeiausiedesieatansnensisnkGs vndetodeie cavtavevenevnenasenaee 177
LE NOT IIDIES Soeesce pone Segre On INES a SNPC REC EE TE PC OUCH CORT RADE CE CURE ACE ERECT ECOSOC rE eee CEC eREEe 17

Reh eC GI IVIOSPOLUS SDACD), Ie CUCL, sncnciowsnonenasidvacnnpdancinmbednsriane dave <ssenae 178
Et. 22 7 SOLOPVIETA (ADAG) NVIDIA 0 conic sinonnsinnsnidiene cnani naninnsasancnenss ese dcannse 202

GE MMPS CAN.) 2 © LOIS Veneer ete cena nce nates ai sigriiisnes ase dover namesinstesidecee se seinieeins 206
Rete mou Y . RAEIIEA IN. FKODSOM ..0...-.0:00-sccecneercessecesnescnneseonsce 288
NEC ATI ALOSACHIIT ID UNAMEL) GOGTON! .o. cce-crsacecassecacecencestecssesaecesneates 297

SEG OMm OU ONd SKC leper anemone Saeer ceases seeae tne cee te teceueccacadadessteessecedeasess 314
DECt OA OPIDKAGHIONAESINARODSOMM ..a-.scsccncehe cece rire seckecsesansscncetevassevcseotesss 317

EMP NC ALON ECL ZCUIEMIUS eee eremetaes aa naten ates estate ni tek scammnadchtiet bene vausisanwisswacecaua ves 320
MEINE LCTEN COS ameneme enna sone tit. haat on wicnidssuisbisivinate tamed cedettantia tedumaduauesoaGelvlcume de 320
SRN ANTE RENCE? Sater geared Sos WE We Sees vent okld os Wee Gens desig nine wulllllivignva esses cnmetion 322

Synopsis

A systematic account is given of the 64 Old World shrubby species of Hypericum L. (sects 1-6). A new
shrubby species from Mexico (H. umbraculoides N. Robson sp. nov.) is included which belongs to a new
section (sect. 6a. Umbraculoides N. Robson sect. nov.) and is related to the Chinese H. monogynum L. A
discussion of the morphology, chromosome numbers, and distribution of members of the sections
concerned is incorporated.

In addition to the new section and species, the following new taxa are described: in sect. 1: H. balfourii
(Socotra) and H. socotranum subsp. smithii (Socotra); in sect. 3: H. cohaerens (China: Guizhou), H.
subsessile (China: Yunnan, Sichuan), H. siamense (Thailand), H. lagarocladum (China: south central), H.
addingtonii (China: N.W. Yunnan), H. x cyathiflorum (H. addingtonii x hookerianum), H. lacei (E.
Burma), H. henryi subsp. hancockii (China: Yunnan, Burma, Thailand, Sumatra), H. maclarenii (China:
Sichuan), H. bellum subsp. latisepalum (China: W. Yunnan, N. Burma), H. lancasteri (China: N. Yunnan,
S. Sichuan), H. curvisepalum (China: Yunnan, Sichuan, Guizhou), and H. x dummeri (H. forrestii x
calycinum).

The following changes of rank (stat. nov.) are also made: H. longistylum subsp. giraldii (R. Keller) N.
Robson, H. henryi subsp. uraloides (Rehder) N. Robson, H. geminiflorum subsp. simplicistylum (Hayata)
N. Robson, H. hircinum subsp. majus (Aiton). N. Robson, and H. hircinum subsp. albimontanum
(Greuter) N. Robson. In addition, H. x ‘Hidcote’ and H. x ‘Eastleigh Gold’ are recognised as hybrids, and
H. hookerianum ‘Rogersii’ is treated as a cultivar, ‘Charles Rogers’.

Bull. Br. Mus. nat. Hist. (Bot.) 12 (4): 163-325 Issued 28 February 1985

164 NORMAN K. B. ROBSON
1. Introduction

Part 3* of this monograph of Hypericum L. begins the systematic account of the genus. It
includes descriptions of 64 species in seven sections, i.e. sects 1-6 as treated in Part 1 and an
additional, monotypic section, sect. 6a. Umbraculoides, which contains a new species with
interesting phytogeographical relationships. It has also proved necessary to subdivide sect. 7.
Roscyna and to divide sect. 9. Hypericum further. The new sections thus produced are indicated
in Figs 4-6 as sects 7a and 8a but they will not be described formally until sects 7—9 sens. lat. are
treated in a future part.

The species treated here therefore include those of (i) the basic (primitive) sect. 1. Campylo-
sporus (Africa and adjacent islands), (ii) the closely related sect. 2. Psorophytum (Balearic Is.)
and (iii) the immediately derived sect. 3. Ascyreia (south and east Asia and Pontic Turkey),
along with its shrubby derivatives sects 46a (Taiwan and Luzon, Caucasus to Macaronesia,
south Mexico). The other derivations of sect. Ascyreia are dwarf shrubs or wiry shrublets or
herbs (sects 7-19).

Morphology

The species in sects 1—-6a are thus all shrubs from c. 0-3 to c. 10 m in height. Except for some
species in sect. 1, they entirely lack dark glands; but H. xylosteifolium (sect. 6) sometimes has
yellowish glands fringing sepals and bracts. All have marginal punctiform glands, as well as
laminar glands which are primitively linear but have become dissected and eventually punc-
tiform. In addition, many species in sect. 3 have resinous punctiform glands on the lower surface
of the leaf (ventral glands); but these are variable in occurrence and distribution and thus less
useful taxonomically than are the other types (see Robson (1981: 80-82)).

The dissection and shortening of the laminar glands is correlated with changes from parallel
(basically dichotomous) to pinnate and open to closed secondary venation, as well as with
increasingly densely reticulate tertiary venation (see Part 2: 76-80).

The inflorescence in sects 1-6 is almost always 1-flowered or dichasial to monochasial; only
very exceptionally does the axillary flowering shoot comprise more than one node, making the
inflorescence pseudo-dichotomous (Part 2: 85).

In these sections the perianth is normally pentamerous. The sepals are mostly erect in flower
and fruit and the petals spreading to deflexed (‘flowers stellate’); but in sect. 3 there are trends
towards spreading to reflexed sepals and, in sects 3 and 4, to incurved petals (‘flowers
cyathiform’). The five stamen fascicles are always pentamerous and free, except in sect. 6a,
where they are ‘trimerous’ (i.e. united 2 + 2 + 1) (Part 2: 100-102). Sects 2 and 3 differ from
most species in sect. 1 in having petals and stamen fascicles that are deciduous after flowering.
This character difference breaks down partially, however, in H. quartinianum and H. synstylum,
in which the petals (and also the stamens in the former) are tardily deciduous and the styles are
usually completely coherent, foreshadowing the deciduous petals and stamens and completely
coherent styles in the related sect. 20. Myriandra (Figs 1, 3). In H. balfourii and H. socotranum
too, the petals are completely deciduous and the stamen fascicles tardily so; but in these species
the styles are only partially coherent and may even become free in fruit in H. balfourii, a
combination of characters that foreshadows the deciduous petals and free styles of the primitive
members of the related sect. 3 (Fig. 1). The petals and stamens are also deciduous in sect. 5, but
in all other sections derived from sect. 3 they have become persistent again (Fig. 3).

The number of styles and ovary placentae is normally 5 in sects 1-4 and 3 in sects 5—6a, the
styles being appressed and more or less coherent in sect. 1 and free in most of the other sections.
In part of sect. 3, however, there is a reversal of this trend, whereby the styles have become
almost united in H. monogynum and its relatives, H. prattii and H. longistylum; in sect. 4 they
are completely united. This union, though persisting in fruit in these species, nevertheless is not
at all ‘fundamental’; in the other derivatives of H. monogynum (viz. H. cohaerens (sect. 3) and
sects 6 and 6a, as well as the herbaceous species, in sects 8—9a) the styles have become almost or
completely free again.

* For part 1 see Robson (1977a); for part 2 see Robson (1981).

STUDIES IN THE GENUS HYPERICUM L. 165

The ovary and capsule vary from ovoid-pyramidal to globose, and vittae are visible only in H.
xylosteifolium (sect. 6); the capsule, however, is unknown in H. umbraculoides (sect. 6a). In H.
reptans (sect. 3), as well as H. androsaemum and H. x inodorum (sect. 5), the capsule wall
becomes fleshy; only in H. androsaemum does the fleshiness persist until maturity. The seeds
vary from linear-reticulate to linear-foveolate or rarely scalariform-reticulate and are often
carinate or winged (laterally and sometimes apically).

Cytology

In the monotypic sect. 2, as well as in the only taxon counted so far in sect. 1 (H. revolutum
subsp. keniense), the chromosome number is 2n = 24. This is thought to be the primitive number
for the whole genus (Robson, 1981: 150). From this there are several descending series of diploid
numbers, in two sections (13, 30) extending to 2n = 14, with numerous polyploids (tetraploid or
rarely hexaploid or secondary polyploids) on most numbers (Fig. 1). The number 2n = 12 in
one species of sect. 30 is anomalous (Robson, 1981: 166).

In sect. 3 these descending series extend from 2n = 24 to 20, 22 having been counted so far
only in H. oblongifolium (Fig. 4). Tetraploids (2n = 48, 46, 44) also occur in that species, and the
2n = 42 in the closely related H. monogynum may be a continuation of a similar series. 2n = 20 is
the most frequently recorded number in the section, and tetraploids based on it (2n = 40) occur
in sects 5 and 6. Within sect. 3, however, 2n = 40 has been recorded only in H. kouytchense,
which also yielded a count of 2n = 36. The occurrence of 2n = 38 and 36 in the closely related H.
beanii and H. forrestii indicates that x = 9 in this group has arisen at the tetraploid level. The only
other report of this basic number in sect. 3 is an isolated one approximately at the hexaploid level
(H. angustinii), and counts for related species are necessary in order to interpret it. It seems
likely, however, that x = 9 at the diploid level is attained only in derivatives of sect. 3 (Fig. 4).

Among hybrids in sect. 3, however, diploid, triploid and pentaploid (or approximately
pentaploid) numbers have been counted. All are on the base x = 10, if one assumes that the
report of n = 18 for H. x moserianum (Part 2: 152) was based on a misidentification.

The data on which the above analysis is based are all cited in Part 2: Table 7, except for the
count of n = 23 for H. oblongifolium (Singhal, Gill & Bir, 1980). Subsequent efforts by Dr
Gibby to add to the list of counts in sect. 3 have not yet succeeded, owing to the difficulty of
producing satisfactory preparations.

Distribution

In Part 2 it was shown (i) how the species of the basic sect. 1 have their nearest relatives in a
semi-circle of surrounding continents and islands (pp. 182-187, figs 60-61); and (ii) how this
distribution is consistent with the Continental Drift hypothesis (pp. 208-216, fig. 73).

(a) Sects 1. Campylosporus and 2. Psorophytum

The suggested relationships of species in sect. 1 and their derivatives are shown in Fig. 1, and
their distributions and those of the derivative sections are summarised in Fig. 2.* From these
figures it can be seen that the most primitive taxa (1. H. bequaertii and 2a. H. revolutum subsp.
keniense) are confined to the East African mountains, but that 3a. H. lanceolatum subsp.
angustifolium (which is morphologically very similar to H. revolutum subsp. keniense) is widely
disjunct in the island of Réunion. In an earlier publication (Robson, 1979) it was shown how H.
revolutum and H. lanceolatum appear to have evolved parallel subspecies (Fig. 2(i), (ii)), 2b. H.

* These diagrams differ slightly from that in Part 2 (Robson, 1981: fig. 61). The relationships of H. revolutum subsp.
keniense have been resolved cladistically (as far as possible), and sects 27. Adenosepalum and 28. Elodes have been
moved from the H. quartinianum line (v) to the H. roeperianum one (vi) after their relationship to that species had
become apparent (see p. 169). In these and similar diagrams, although no strict cladistic analysis has been attempted,
the branching has been represented in a cladistic manner except where the characters of a taxon are apparently wholly
advanced (apomorphic) in comparison with those of its relative. Such a taxon is shown as directly derived from its
relative. This practice violates one of the Hennigian tenets (i.e. that the ancestral taxon is changed when the clade
branches), but it may often be considered a more accurate representation of the true relationships.

166 NORMAN K. B. ROBSON

revolutum subsp. revolutum being widespread in the mountains of tropical Africa and Arabia
whereas 3b. H. lanceolatum subsp. lanceolatum is confined to Réunion and the Comoro Islands
(Grande Comore). The relatively early separation of Madagascar and the Mascarene region
from Africa must have isolated H. lanceolatum, so that it has given rise only to H. madagascar-
iense (Madagascar) and sect. 26. Humifusoideum, the most primitive species of which is now in
New Guinea but probably reached there via Australia (Robson, 1981: 208-209, fig. 73).

Sect. 1. Campylosporus

Relationships and chromosome numbers (2n)

sai ET 20,18 4 Ht

18,16
gnidiifolium He ee
socotranum
socotranum
bsp.
1 0 subsp. socotranum subsp. smithii

8 synstylum

5

balfourii

9

roeperianum

99 3-19
24,22,20,18,16,14
A | as

48,46,44
24,28

ie quartinianum

40,38,36

5 29] madagascariense
24 2b revolutum 4
highs subsp. en
30 2a revolutum 3b lanceolatum
subsp. keniense subsp. lanceolatum
24

3a

24, 18,16,14

: 1

12 bequaertii

lanceolatum
subsp. angustifolium

24,16

Santomasia

Fig. 1 Sect. 1. Campylosporus: Relationships and chromosome numbers (2n) of the 10 species, the other
sections of Hypericum (in boxes) and the most closely related genus, Santomasia.

STUDIES IN THE GENUS HYPERICUM L. 167

The remaining evolutionary lines in sect. 1 stem from H. revolutum subsp. keniense, and it
does not seem possible to resolve them further cladistically, as each displays a different
morphological advance from H. revolutum subsp. keniense. The line through subsp. revolutum
(Fig. 2(i)), in which the inflorescence remains 1-flowered, shows a disjunction between
north-east tropical Africa and the Balearic Islands, where sect. 2 Psorophytum (H. balearicum)
is endemic. Another line (Fig. 2(iii)), which shows several correlated characters e.g. parietal

Sect. 1. Campylosporus

Distributic Socotra, Levant, S.Turkey

zes
6b

NE.& C.America, ae GSco, 6
Macaronesia, 2 Gr.Antilles, editerranean a
Africa, Europe, 59 21-22 Bahamas, Bermuda
S.W.Asia hidesronakid, te: W.Socotra E.Socotra
Levant, Ethiopia
Turkey 8 N. Somalia
10 Ethiopia

C.Socotra 5

3 trop.Africa

7

S.W.Arabia S.Asia to
E.Africa Bulgaria,
N.Eurasia, N.America to
Guatemala
Balearic Is.
C.America, 4
Gr.Antilles, (i) S.W.Arabia to N&C
Ria Gage Prov. .& C. Madagascar
Cameroun Mtn., F.Poo bye
2a
S.America to E.African,mountains
N.America, \ he. Gr. Comore
Africa, :
S.E.Asia,
Australasia 3
a Réunion

Ruwenzori 1

New Guinea
W.Indonesia,
Africa, Madagascar

Central America

==} Trans-oceanic ee
— Other disjunctions

Fig.2 Sect. 1. Campylosporus: Distribution of the 10 species, the other sections of Hypericum (in boxes)

and the most closely related genus, Santomasia (Central America).

168 NORMAN K. B. ROBSON

placentation and absence of dark glands, gives rise immediately to two trans-Atlantic disjunc-
tions: sect. 29 Brathys with primitive species in northern S. America, Belize, and Cuba, and sect.
30. Spachium with primitive species in southern Brazil; the remaining line involves other species
ofisectl:

The eastern main branch of this line (Fig. 2(iv)) includes H. mysurense (sect. 3, in Sri Lanka
and south India), which, as has been mentioned above, is morphologically close to H. balfourii
of Socotra. These populations, along with H. socotranum (another Socotran endemic), have

Sect. 1.Campylosporus

aeer
auaUa Rain sleletalaln es eecan acta oCe TRAN oe care,
paar terse eeeesee rs eseresserersesccesccsens svete?
aD

completely

Styles
Dark glands S
absent 8 present
Dark glands F
present

ce absent ig ‘

Petals and stamens

deciduous

partly coherent 1

Fig.3 Sect. 1. Campylosporus: Limits of certain characters within the section and in relation to the other
sections. Note the progressions: styles partly coherent > free — completely united; petals and stamens
persistent — deciduous — persistent; dark glands present — absent — present.

STUDIES IN THE GENUS HYPERICUM L. 169

indeed usually been treated as conspecific (see p. 192); but even when separated, as they are
here, they still represent a north-eastward disjunction that seems to be associated with the
separation from Africa and north-eastward movement of the Indian subcontinent. Each
subspecies of H. socotranum is related to a different section or sections, both containing low
shrubby species with ‘3’ stamen-fascicles and 3 styles. Subsp. smithii has its nearest relatives
(belonging to sect. 23. Triadenioides) in Socotra, but other species in sect. 23 are in the Levant
and south Turkey. Subsp. socotranum, on the other hand, has its nearest relatives belonging to
sect. 25. Adenotrias in western Morocco, the more reduced or specialised forms and species of
this section being scattered eastward through the Mediterranean to Syria.

The remaining western branch, which includes those species of sect. 1 that have retained at
least some dark glands and have styles completely coherent or almost so, divides into two. In one
of the sub-branches (Fig. 2(v)) the main lateral leaf veins are unbranched, the tertiary venation,
if evident, is confined to areas between the midrib branches, and the pellucid glands are
therefore mostly elongate (7. H. quartinianum and 8. H. synstylum). These species are confined
to east and north-east tropical Africa and south-west Arabia; but they are closely related to sect.
20. Myriandra, centred in south-eastern N. America, in which the styles remain appressed but
dark glands are absent and, as has been mentioned above, the petals and stamens are deciduous.
In addition, in the more primitive species of this section the tertiary venation is conspicuously
reticulate over the whole leaf.

The other sub-branch of the western branch (Fig. 2(vi)) includes species in which all the leaf
venation has become conspicuously reticulate; the pellucid glands, in consequence, are mostly
punctiform and are confined to the areoles of the reticulum (Part 2: 81, fig. 11d). Although the
present distribution of the primitive species (9. H. roeperianum) extends farther west and south
in tropical Africa than does that of species in branch (v), this species appears to have
differentiated in the north-eastern part of tropical Africa. Three facts support this hypothesis: 1)
the most primitive form occurs there, ii) a variant that has sometimes been recognised as a
separate species (H. schimperi Hochst. ex A. Rich.) or subspecies also occurs there, and iii) a
closely related species that is advanced relative to H. roeperianum is nearly all characters (10. H.
gnidiifolium) has a relict distribution in Ethiopia. H. roeperianum appears to have given rise to
two distinct lines, one (sects 21. Webbia and 22. Arthrophyllum) related to ‘H. schimperi’ and
with a Macaronesia/E. Mediterranean distribution, the other (sects 27. Adenosepalum and 28.
Elodes) related to the typical form and spreading from its Macaronesian ‘centre’ in three
directions: i) northward to north-west Africa and western Europe, ii) eastward to north and east
Africa and thence to the Himalayas and western Eurasia, and iii) southward to tropical and
southern Africa. In both these lines the petals and stamens remain persistent but the styles
become free and reduced to three (Fig. 3).

(b) Sect. 3. Ascyreia

It seems very reasonable to assume, from the morphological and distributional relationships of
1. H. mysurense, that its progenitors drifted north on the Indian subcontinent. This species is
now confined to Sri Lanka and the southern mountains of India, apparently occurring as far
north in the Western Ghats as Konkan in the Bombay region. Its leaves are concolorous but
variably glaucous, and it gives rise to two lines (Fig. 5(A) and (B)), one initially with concolorous
glaucous leaves (Spp. 2-11 and sects 7 and 9-19) and the other with green or discolorous leaves
(Spp. 12-42 and sects 46a and 7a-8a).

Line (A) mostly comprises species that have relict distributions in the mountains on the north
and east of the Indian subcontinent from Nepal eastward, with 6. H. pachyphyllum and 7. H.
augustinii confined to small areas of central Burma and southern Yunnan respectively (branch
(ii)). 5. H. reptans, although occurring along the Himalayan range from Nepal to north-western
Yunnan, has its most primitive form in Burma and is absent from Bhutan, the country in which
its closest relative, 4. H. sherriffii, is a localised endemic.

From this peripheral distribution alone, it is not clear whether these species spread and
diversified before or after the initiation of the Himalayan orogeny; but consideration of the
remaining two species of sect. 3 in this line provides a clue to this problem. 2. H. cordifolium and

170 NORMAN K. B. ROBSON

3. H. podocarpoides are closely related local Nepalese endemics, both directly derivable from
H. mysurense but not one from the other. Both appear to be ancestral to groups of sections that
are centred on the other side of the Himalayan massif, H. cordifolium westward to Mediterra-
nean/European sections radiating respectively from eastern Turkey (sects 17. Hirtella — 19.
Coridium) and western Turkey (sect 10. Olympia — 16. Crossophyllum) (branch (i)) and H.
podocarpoides northward to sect. 7. Roscyna (Altai Mts to Japan, eastern N. America) and its

Sect.3. Ascyreia

Relationships and chromosome numbers (2n)

H x‘Rowallane’ (22x 28) 20 39 curvisepalum
H x cyathiflorum‘Gold Cup’ (prob.27x28) 30
H x‘Hidcote (prob.(27x28)x 14) c.50 (c.48,¢.54,55) 38 35 bellum
H x‘Eastleigh Gold’(28 ?x 36) lancasteri
H x moserianum (31x 14) 50 bone
Hx dummeri (42x 14) peaudehenrul 40 38 37 stellatum 34
41 choisianum
forrestii 42
38,36 36 ead 33
williamsii 99 7a 8 8a ese:
8 9 tenuicaule

18 prattii

ae a cohaerens 37 patulum
augustinii lobbii 17
7 c.54
17 gracilipes 32
uralum
20
ee
henryi
preva gay
pachyphyllum as RS
P .
9 6a 28
16 monogynum (i) hookerianum
20?
griffithii
ae 15 25
5 reptans wilsonii
calycinum
14-20 27 addingtonii
20?
4 lagarocladum
blonaifoli lacei 24
sherriffii 69 7,9,9a 13 0 ey 5 ium a5
18, 16 a
‘32 +48 48,46,44 acmosepalum
40 ” 23
leschenaultii
66 10-16 é
odocarpoides os
18, 16, 14 e P gaitii 20? 929
. 3 12
32 271 siamense
coraionm subsessile
2
§6 17-19 20

mysurense
20,18

™ 24,28,

Fig. 4 Sect. 3. Ascyreia: Relationships and chromosome numbers (2n) of the 42 species and related
sections, and a list of intrasectional hybrids, with parentage. See p. 325.

STUDIES IN THE GENUS HYPERICUM L. 7A

derivative sects 9. Hypericum and 9a. Concinna (north temperate zone) (branch (iii)). It would
seem, therefore, that the Himalayan orogeny was instrumental in cutting off the ancestors of
these sections from those of H. cordifolium/podocarpoides, which no doubt spread on to Eurasia
after the Indian Plate had made contact with it. The species of line (ii) (4. H. sherriffii— 11. H.
gracilipes), which are also derived from the H. cordifolium/podocarpoides affinity , can therefore
be seen to be a third, eastward, development from that group. It probably achieved its initial

Sect.3. Ascyreia

: Yunnan, Sichuan,

eee 39 Guizhou N.W.Yunnan to
35 Xizang &
W.Sichuan, Sichuan, 38 S.Sichuan Arun.Prad.
N.W.Yunnan,N.Burma yy w-Yunnan Riscaiian se
E.Yunnan -Yunnan to
42 =a sad lon) 37 34 Pakistan

(xix) si N.E.Sichuan

Guizhou 36

9 Nepal, Sikkim

a 33 W.Sichuan
-Cnina, Himalaya,
ee S,India, W.Transcaucasia'

41 E.lndian mts ., S.Central China
W.Bengal 31
17 Burma
Guizhou 19 to Pakistan
N. Central China 32
W.Transcaucasia
§6| Taiwan, Luzon, W.Hima’
S.Yunnan D6 Ryukyu Is. ss
7 S.Mexico 30
S.W.China to
Thailand, Sumatra

HedtoraneanSWAraba. 7 ~ 16 E.China , Taiwan?

ol Macaronesia, W.Europe yo 25
6 2B
/ Yunnan to 1,
Assam eke 3 P Hs
to Yunnan, =e See / Nepal , S.India huan
Nepal,Sikkim
5
!

ry alee ~. SMadhes 13 w aie
Koa ulgaria i

S.Central
China

| :
\ N.Temp-Eurasia, W.Indonesia
Se N.America

22

Nepal,Kumaun

21
EuropeMediterr.,

Macaronesia !

Yunnan, Sichuan
1

Sri Lanka,

S.india
A yi?

——} Trans-oceanic disjunctions

to CAsia Africa, etc. — Himalayan disjunctions

Fig. 5 Sect. 3. Ascyreia: Distribution of the 42 species and related sections. The single bars indicate
trans-Himalayan disjunctions, the double bars trans-oceanic disjunctions, and the area enclosed by the
broken line represents the original, cis-Himalayan distribution of the section. See also p. 325.

172 NORMAN K. B. ROBSON

dispersal while the Himalayan range was relatively low; and its distribution would have been
dissected as the mountains became more elevated. Thus the species from the highest altitudes
(H. reptans) has the least dissected or relict distribution.

The other evolutionary line that stems from H. mysurense (Fig. 5(B)) has two main branches,
of which one (iv) has a western sub-branch (vi) with Indian and Himalayan species (Spp. 12, 13,

Sect.3. Ascyreia

Characters

EE SETS
one

=

q
substellate § cyathiform

6 17-19 <
; pelt —— subsessile Leaf
=== _ sessile reticulate
venation

L}
stellate erect | spreading 4

Sepals Corolla

wees & = | -=.]

Corolla Leaves

Fig. 6 Sect. 3. Ascyreia: Limits of certain characters within the section and in related sections. Note the
progression: corolla stellate > subcyathiform — cyathiform. See also p. 325.

STUDIES IN THE GENUS HYPERICUM L. 173

15) and then disjunctions to beyond the Himalayan Plateau (14. H. calycinum and sect. 5.
Androsaemum westward to Turkey and the Mediterranean region). The eastern sub-branch
(vii) has Spp. 16-19 in central China with derivative sections radiating from there, sect. 4.
Takasagoya to Taiwan and Luzon, sects 7a and 8a (as yet unnamed) through China and
south-east Asia to Pakistan, south India, and Thailand, sects 6. Inodora and 8. Bupleuroides to
western Transcaucasia, and sect. 6a. Umbraculoides to southern Mexico. The occurrence of H.
xylosteifolium (sect. 6) and H. bupleuroides (sect. 8) in Georgia and adjacent Turkey implies
that H. monogynum had diversified (in China?) before the elevation of the Tibetan plateau.

In sub-branch (iv) the flowers are always stellate and the early species all have leaves with
densely reticulate venation and free styles, the reticulation becoming less dense in 19. H.
longistylum after the styles have become partially united (in 16. H. monogynum) (Fig. 6). As we
have seen above (p. 164), this trend towards union becomes complete in sect. 4 Takasagoya but
is reversed in 17. H. cohaerens and sects 6, 8 and 8a.

In the second main branch of the second line from H. mysurense (v), the leaves are green or
discolorous (not uniformly glaucous) with at most laxly reticulate venation. The styles remain
free, but the flowers evolve from stellate to cyathiform along several lineages (Fig. 6). The
distribution of this group (Spp. 20-42) ranges from central China westward along the Himalaya
to Pakistan, southward to south India (28. H. hookerianum), and through Indonesia to Lombok
and western Sulawese (22. H. leschenaultii). Unlike the other lines that have been discussed, this
one remains in sect. 3. It has two main branches, which are difficult to differentiate morphologi-
cally as whole branches, although the relationships of the individual species in each are clear
enough. *

Lineage (v) as a whole is centred in south-west China (Yunnan, Sichuan, Guizhou) and, being
directly related to H. mysurense, thus displays another trans-Himalayan disjunction in distri-
bution.

The more northern sub-branch (ix) has a western division (xvi) (33. H. maclarenii — 35. H.
bellum; Sichuan to north-western Yunnan, southern Xizang (Tibet) and along the Himalaya to
Pakistan) and an eastern one (xvii) (36. H. kouytchense — 42. H. forrestii; Guizhou and Sichuan
to northern Burma), the latter also dividing into a mainly north-eastern branch (xviii) (Spp.
37-39) and a mainly south-western one (xix) (Spp. 40-42) linked by 36. H. kouytchense. Spp.
36-39 stand out from the others in line (ix) by the combination of narrow acute spreading sepals
and an acute petal apiculus; Spp. 33-35 have rather bluish-green leaves that are either rather
thin and sub-bullate or undulate-margined, whereas in Spp. 40-42 they are mid-green and
neither bullate nor undulate.

The more southern sub-branch (viii) also has two main divisions of which one (x) represents
an early diversification southwards from Yunnan/Sichuan (20. H. subsessile) through Thailand
(21. H. siamense) to western Indonesia (22. H. leschenaultii). It shows morphological trends
from elliptic to overate or lanceolate leaves and ovate or broadly elliptic to narrowly elliptic or
oblanceolate sepals. H. leschenaultii (x) has converged morphologically with 34. H. choisianum
(xvi), which has evolved along a similar line except for the closed rather than open leaf venation
(See p. 273).

Sub-branch (x1) comprises species that tend to have not only the leaves in one plane (a fairly
common state in species with arching stems), but also the lateral branches. In one line (xiii),
where the stems are mostly erect to arching, this results in a fern-like (frondose) habit (e.g. in 32.
H. uralum); in the other (xii) the extreme species can be prostrate or pendulous (35. H. wilsonii,
26. H. dyeri). Here, again, the primary division of line (ix) is geographical, line (xii) starting to
the north-east (23. H. acmosepalum in north-eastern Yunnan, Guizhou and Hunan) and line
(xiii) to the south-west (29. H. lacei, 30a. H. henryi subsp. hancockii in eastern Burma, southern
Yunnan, northern Vietnam, and Sumatra).

Finally, line (xii) itself has two branches that are at first geographically separate. The
north-eastern (xiv) includes 24. H. lagarocladum and 25. H. wilsonii (north-eastern Yunnan

* See couplet 27, p. 208. In using the key, it may be necessary to try both alternatives at this point; but subsequent
answers should indicate clearly which one is correct for a given specimen.

174 NORMAN K. B. ROBSON

north and east to northern Sichuan and Hunan), as well as the western Himalayan 26. H. dyeri-—
another trans-Himalayan disjunction; the south-western (xv) comprises 27. H. addingtonii
(north-western Yunnan) and the widespread 28. H. hookerianum, which has its most primitive
forms in western Yunnan, Burma, and Thailand.

(c) Sect. 4. Takasagoya

Sect. Takasagoya is clearly related to and derived from H. monogynum, in particular from form
(1) ‘salicifolium’, which, as has been explained, is native to Sichuan (Fig. 7). If, as seems
possible, H. monogynum is truly native in Taiwan, then the geographical disjunction between
sects 3 and 4 disappears. At any rate, sect. Takasagoya divides into two clear groups; 4. H.
subalatum (north-eastern Taiwan) has strongly 4-lined stems and leaves that are narrow and
acute like those of H. monogynum ‘salicifolium’, but smaller. In the other species the stem
becomes 2-lined or terete, and the leaves are broader and obtuse to rounded. Of these, the

Sect.4. Takasagoya
(a) Relationships
3 senkakuinsulare xy
geminiflorum
2 nakamurai ssp. Surg)
4a 4b geminiflorum
subsp. Simplicistylum
1

5 subalatum

H. monogynum (i)
(b) Distribution |

3 Ryukyu Is.

(Senkaku) S. Taiwan , Luzon

jie aren W. & N. Taiwan
1 N. Taiwan

5 NE. Taiwan

4 E. Taiwan

Sichuan [Ns E. Taiwan |

|
Fig.7 Sect. 4. Takasagoya: (a) Relationships of the 5 species; (b) distribution of the 5 species.

STUDIES IN THE GENUS HYPERICUM L. 175

long-sepalled species are in northern and eastern Taiwan (1. H. formosanum, 2. H. nakamurai)
and the Ryukyu Is. (3. H. senkakuinsulare), whilst the last species (5. H. geminiflorum) is in
south Taiwan and the Philippines (Luzon) with a derivate subspecies in the north. Thus here
again the branching of the cladogram has a geographical basis.

(d) Sect. 5. Androsaemum

The nearest species in sect. Ascyreia to those of the basically Mediterranean and Macaronesian
sect. Androsaemum is H. griffithii, endemic to Bhutan and adjacent India (Arunachal Pradesh).
Thus here again we have two taxa with disjunct distributions on either side of the Himalayan
massif; and again, as in the case of H. sherriffii and H. reptans, the nearest ancestral species (H.
oblongifolium) occupies an intermediate geographical position (western Himalaya). The sect.
Androsaemum diagram (Fig. 8) branches first to separate a basically eastern group with
deciduous sepals (4. H. hircinum) from a basically western group with persistent sepals

Sect.5. Androsaemum

(a) Relationships and chromosome numbers (2n)

H. x inodorum (3x4) 40

3 aadeosacriin hircinum
: eda : Ab subsp. cambessedesii
40
foliosum 2
hircinum 4d hircinum:
1 subsp. majus ees subsp. hircinum
40

4a

grandifolium
——————
40
4e hircinum
subsp. albimontanum
griffithii 4c hircinum
subsp. metroi

oblongifolium

| 24,22

¥
48,46,44
(b) Distribution
3 W.Europe, 4b Balearic Is.
Mediterranean to Iran
Azores 2

Ad Sardinia ,Corsica
Mediterranean
S.W.Arabia 4 ae
a a Peloponnisos,
4e Crete,S.Aegean,

Canary Is., Madeira Cyprus

4c Morocco
Bhutan, E.India

Pakistan to Nepal

Fig.8 Sect. 5. Androsaemum: (a) Relationships of the 4 species; (b) distribution of the 4 species.

176 NORMAN K. B. ROBSON

(Spp. 1-3). In the latter, 1. H. grandifolium and 2. H. foliosum are confined to the Atlantic
Islands, whilst 3. H. androsaemum has a dissected W. Europe—Mediterranean distribution that
extends eastwards to Iran and possibly Turkmenistan. H. hircinum, on the other hand, has its
most primitive form in the central and eastern Mediterranean (S. Italy, S. Turkey, and the
Levant) and south-west Arabia (Asir), with the advanced forms on Mediterranean islands
(Balearics, Corsica, Crete, Andhros, Samos, Cyprus) and in the Peloponnisos and Morocco.

(e) Sect. 6. Inodora

The separation of H. xylosteifolium (Georgia and adjacent Turkey) from its nearest relative in
central China (H. monogynum form (ii)) is paralleled in other genera (e.g. Pterocarya in the
Juglandaceae) and appears to be related to the uplifting of the Himalayan massif and the
subsequent increased aridity of central Asia. It is now confined to the wettest (tea-growing)
areas in Transcaucasia.

From the above geographical considerations it is possible to draw the general conclusion that,
in the relationship diagrams (Figs 1-8), the initial forking of a branch nearly always involves
vicariance, i.e. the nearest relatives (‘sister groups’) rarely occur in the same area. Where they
appear to do so (e.g. 8. H. lobbii/9. H. gracilipes or 38. H. lancasteri/39. H. curvisepalum in
sect. 3), then the separation may be ecological (including altitudinal) or cyto-genetical — or
possibly only imaginary; every such pair should be examined especially carefully on account of
the predominantly vicariance-related speciation in Hypericum.

(f) Sect. 6a. Umbraculoides, sect. nov.

When looking through some Hypericum material borrowed from the University of Kébenhavn
(C), I recently recognised that two specimens from southern Mexico (Oaxaca) not only
represented a new species, but also belonged to no hitherto-recognised American section of the
genus. Like the two other monotypic sections its relationships are clearly with H. monogynum
sens. lat.; unlike them (sects 6 and 8), however, the disjunction in distribution is trans-Pacific,
not trans-Himalayan.

Sect. 6a. UMBRACULOIDES N. Robson, sect. nov.

Sect. 3. Ascyreia affinis, sed staminorum fasciculis ‘3’ (i.e. 2 + 2 + 1) post anthesin persistentibus, stylis 3,
differt.

Fruitex glaber, haud nigro-glandulosus, ramificatione laterali. Folia opposita decussata sessilia
libera decidua glanduloso-punctata. [nflorescentia dichasialis vel monochasialis, e nodo supre-
mo orta. Flores stellati homostyli. Sepala 5, libera persistentia integra. Petala 5, post anthesin
decidua, glandulis linearibus vel interruptis instructa. Staminorum fasciculi 5, quorum 4 binatim
coaliti et 1 liber, post anthesin persistentia, filamentis breviter coniunctis. Ovarium trimerum
stylis demum liberis. Capsula adhuc non visa.

Typus: Hypericum umbraculoides N. Robson (p. 318).
DISTRIBUTION: Mexico (Oaxaca).

The Mexican species has oblong to elliptic, obtuse to rounded, cordate-amplexicaul leaves with
densely reticulate venation, like some examples of H. monogynum form (ii) ‘obtusifolium’; but
the inflorescence is corymbose, like a rather condensed version of that in form (1) ‘salicifolium’
of the same species, and has prompted the choice of epithet umbraculoides. It differs from
species in sects 3 and 6 in having ‘3’ (i.e. 2 + 2 + 1) persistent stamen fascicles and from sect. 3 in
its 3 styles, which are free but appear to be proximally appressed when young. The petals are
deciduous, unlike those of H. xylosteifolium (sect. 6); and the inflorescence, sepals and
leaf-shape are also quite unlike those of that species.

The new species (Hypericum umbraculoides N. Robson) cannot therefore be included in any
existing sections. It represents an eastward development from H. monogynum comparable with

STUDIES IN THE GENUS HYPERICUM L. 17

the westward one of H. xylosteifolium; but in this case the disjunction is transoceanic. It is
similar in latitude to the disjunction in the bambusoid genus Yushania Keng fil. (Sichuan/
Yunnan to Mexico and Central America) and north of that in Hydrangea sect. Cornidia
(Taiwan/Philippines; Mexico to Argentina) (Stern, 1978). Within Hypericum, it is south of the
trans-Pacific disjunctions in H. ascyron (sect. 7) (E. Siberia to NE. America), H. concinnum
(sect. 9a, related to H. ascyron sens. lat.) (Japan/Sakhalin to California) and sect. 9. Hypericum
(Japan to S.E. United States and N. Mexico) (see Robson, 1981: 187-192, which requires slight
modification owing to the division of sects 7 and 9). One or two trans-Pacific disjunctions are also
involved in the Guttiferae — Cratoxyleae, in Triadenum and possibly Thornea (see Robson,
1981: 182).

There is no reason to doubt that all the above trans-Pacific disjunctions are the result of
gradual rather than long-distance dispersal, unlike those in Hypericum sect. 30. Spachium
(Robson, 1981: 214). The more northern ones (especially those of H. ascyron and Triadenum)
may just possibly be the result of dispersal over the Bering Bridge, but it is unlikely that the
others are due to such a northern crossing of the Pacific. It is even more unlikely that they
crossed via the Antarctic/South America route. Melville’s ‘Pacifica’ (Melville, 1981) would have
provided a convenient ‘vehicle’, as it is thought to form the present eastern part of Asia and
western part of North America. However, as I pointed out earlier (Robson, 1981: 212),
derivatives of sect. Ascyreia migrating northeastward from India could not have reached the
western half of ‘Pacifica’ until it made contact with the rest of Eurasia—by which time the eastern
(American) part would be unreachable overland. Perhaps, after all, Styer & Stern (1979) are
correct in suggesting that even some of the China—Mexico Pacific disjuncts crossed by the Bering
Bridge at a time when the climate in that region was far more temperate than it is today.

2. Systematic treatment
HYPERICUM L.
Sp. pl.: 783 (1753); Gen. pl. Sth ed.: 341 (1754). Type species: H. perforatum L.

Trees (up to c. 12 m tall), shrubs or perennial to annual herbs, glabrous or with simple uniseriate
hairs, with glandular canals or lacunae containing resins (amber), essential oils (pellucid or
‘pale’) and often hypericin and pseudo-hypericin (red to blackish or ‘dark’). Stems green to
yellow-brown or red and with 2-4(-6) raised lines along each internode when young, those lines
decurrent from the midrib of the leaf above usually most prominent (sometimes expanded to
form narrow wings), eventually usually terete, or wholly terete in some herbs; glabrous or with
indumentum; eglandular or with pale to dark glands (sometimes + prominent) or glandiferous,
simple or rarely branched emergences; bark smooth, red-brown to purple-brown or silvery,
smooth, thin or rarely corky, exfoliating in sheets, patches or irregular strips, not exuding
resiniferous sap or latex. Leaves opposite, decussate or sometimes in alternating whorls of 3-4,
exstipulate, sessile to shortly and gradually petiolate, sometimes with basal articulation, free or
+ united, deciduous at or above the articulation or persistent; lamina entire or occasionally with
gland-fringed auricles or base, rarely wholly glandular-denticulate or -fimbriate, venation
parallel-dichotomous to pinnate or 1-nerved, open or closed, the tertiary absent to densely
reticulate; glands linear to punctiform, pale and/or dark, marginal to laminar; indumentum
absent or present. Inflorescence terminal, cymose, 1—~-flowered, elaborated acrotonally by
dichasium/monochasium-formation or pseudo-dichotomy or a mixture of both methods (‘in-
florescence mixed’), basitonally by axillary flowers or flowering branches, sometimes becoming
thyrsoid or occasionally racemose by suppression of the terminal bud; transition from leaves to
sepals sudden or gradual; bracts and bracteoles often present, usually more similar to sepals,
persistent as long as leaves or occasionally (sects 3, 4) caducous. Flowers bisexual, actinomor-
phic, stellate to campanulate or rarely pseudo-tubular, homostylous or rarely dimorphically
heterostylous. Sepals 4-5 (or abnormally 6 or 3), quincuncial or opposite and decussate (rarely
almost open in sect. 20), equal or + unequal, sometimes foliaceous, free or up to 0-7 united,
persistent or occasionally deciduous, with margin entire or glandular-denticulate to -fimbriate or
eglandular-fimbriate; veins 1—c.11, parallel or divergent, dichotomising or pinnately branched;

178 NORMAN K. B. ROBSON

glands marginal to laminar, linear to punctiform, pale and/or dark; indumentum absent or on
dorsal surface only. Petals lemon to golden yellow or orange or rarely cream or white, often
tinged or veined red dorsally where visible in bud or very rarely wholly carmine-red, 4-5 (or
abnormally 6 or 3), contorted, equal, free, persistent or deciduous, asymmetrical (except in
reduced forms), entire or with sessile marginal glands or glandular-ciliate, usually with +
evident projection at apical point of margin in bud (‘apiculus’), rarely with entire and cucullate
or trifid and flat ligule; veins numerous to few, sometimes dichotomising or forming loops,
reaching margin or not; glands laminar and often marginal, punctiform to linear, pale and/or
dark; indumentum absent. Stamen fascicles 4-5, antipetalous, free or variously united (2+ 1+ 1
+ 1,2+ 2+ 1, (5), (4)) and then with double fascicles antisepalous, glabrous, persistent or
deciduous, each with 1-c. 60 stamens; filaments yellow to orange or rarely cream or white or
crimson, slender, united towards the base only or apparently free or united to above the middle
in a few ‘3’-fascicled species; anthers yellow to orange or reddish, oblong to elliptic, almost
isodiametric, bithecal, with amber to red or blackish gland on connective, dorsifixed (or rarely
apparently basifixed, sect. 4), dehiscing introrsely by longitudinal slits; pollen tricolporate,
spheroidal to prolate, with exine microreticulate to reticulate or a tectum perforatum. Stami-
node fascicles absent or rarely 3, alternating with 1 + 2 + 2 stamen fascicles, scale-like, entire or
bilobed, functioning like grass lodicules. Ovary 2-5-merous, glabrous, with placentation
incompletely to completely axile or + parietal; styles 2-5, elongate, free or partially or
completely united, with stigmas distinct, minute to capitate; ovules 2—» on each placenta, erect
to horizontal or pendulous. Fruit capsular, 2—S-valved, dehiscing septicidally from the apex,
with valves somewhat woody or coriaceous to papyraceous, persistent or rarely deciduous,
rarely tardily dehiscent or indehiscent with valves + fleshy and coloured red to blackish, valves
1—~%-seeded; vittae often + prominent, linear or punctiform (‘vesicles’), amber or rarely
blackish; styles wholly or partially persistent. Seeds small (0-3—1-5 mm long), narrowly cylindric
to ovoid-cylindric or ellipsoid, slightly curved or usually straight, without or with a prominent
unilateral carina or thin and papyraceous wing, without or with an apical and sometimes also
basal expansion, wing-like or thick, or rarely an apical whitish caruncle (sect. 25); testa
yellow-brown to red-brown or purple-brown (‘blackish’) with sculpturing + prominently
reticulate or linear-reticulate to foveolate or scalariform or papillose; endosperm absent;
embryo slender, straight, with cotyledons equal, free, plano-convex, shorter than hypocotyl.

BASIC CHROMOSOME NUMBERS (x): 12, 11, 10, 9, 8, 7, 6; ploidy 2-6.

Hapsirat: Dry, semi-desert areas (e.g. southern Turkey and adjacent Syria and Iraq) to shallow
water, in cold temperate to warm temperate climates.

DISTRIBUTION: World-wide except for (1) southern Chile, Argentina and the Antarctic regions,
(2) southern Oceanic islands, (3) tropical lowlands (excluding H. japonicum, which occurs there
as a rice-field weed), (4) Arctic, subarctic, and alpine regions. Apparently absent from the
following territories: Alaska, Falkland Is., Antarctica, Greenland, Iceland, Svalbard, Lesser
Antilles, Trinidad, Tobago, Dutch W. Indies, Surinam, Cayenne, Cape Verde Is., S. Atlantic
Is., W. Sahara—Mali-Niger-Dahomey westward (except Guinea), Central African Republic,
Gabon, Congo-Brazzaville, S. Tomé, Principe, Annobon, Equatorial Guinea (mainland),
Aldabra, Seychelles, Mauritius, Rodriguez, Indian Ocean islands, Kuwait, Arab Emirates,
Oman, Bahrain, Qatar, S. Yemen (mainland), Mongolia, Pacific Ocean islands (except New
Caledonia, Hawaii, Galapagos, and Revilla Gigedos Islands).

Key to sections
See Robson (1977a: 342-344).

Sect. 1. CAMPYLOSPORUS (Spach) R. Keller
in Engler & Prantl, Nat. Pflanzenfam. 3(6): 209 (1893).

Trees or shrubs up to 12 m tall, evergreen, glabrous, with or without dark glands; branching
lateral. Stems 4-lined and + compressed when young, becoming terete, eglandular; cortex

STUDIES IN THE GENUS HYPERICUM L. 179

exfoliating in longitudinal strips; bark fissured, scaly. Leaves opposite, decussate, sessile or
rarely very shortly petiolate, free, deciduous at basal articulation; lamina entire, with venation
parallel-dichotomous to pinnate, open or closed, the tertiary absent or + densely reticulate;
laminar glands linear to punctiform, pale; marginal gland dots dark and/or pale; vental glands
absent. Inflorescence 1-9(—~)-flowered, branching dichasial/monochasial from 1—2(-4) nodes,
sometimes with subsidiary flowering branches from lower nodes; bracts and bracteoles foliar or
transitional. Flowers stellate to cyathiform or rarely campanulate, homostylous. Sepals 5, free,
persistent, spreading to reflexed in fruit, with margin entire to denticulate or ciliolate; veins
numerous; laminar glands pale, linear or rarely interrupted to punctiform; submarginal dark
glands present or absent; marginal or inframarginal glands absent or dark (or rarely reddish).
Petals 5, persistent or sometimes tardily deciduous, with apiculus present, lateral, rounded, +
prominent or rarely almost absent, margin entire or + minutely glandular-ciliate; marginal
glands absent or amber to dark; laminar glands + numerous, pale, linear and often interrupted
distally, very rarely also dark dots or short streaks. Stamen fascicles 5, free, persistent or rarely
tardily deciduous, each with 20—-45(—80) stamens; filaments united very shortly or for up to 5
mm; anthers yellow, gland amber; pollen type I. Ovary with 5 incompletely axile placentae
(united at the base, free but deeply intrusive above), each -ovulate; styles 5, partly or wholly
coherent or very rarely almost free; stigma subglobose (or rarely bilobed) to small and rounded.
Capsule 5-valved, coriaceous, not vittate. Seeds + narrowly cylindric to fusiform, sometimes
shallowly carinate, rarely with apical wing expansion; testa linear-reticulate or rarely linear-
foveolate.

BASIC CHROMOSOME NUMBER (x): 12; ploidy 2.

HasiraT: open montane or mid-altitude forest, deciduous forest, fringing forest, heathland,
grassland, rocky places, 210-4000 m.

DisTRIBUTION: SW. Arabia (Asir, Yemen) and Socotra; tropical and subtropical Africa from
Guinée, Nigeria, Sudan Republic and Ethiopia south to northern Cape Province; Fernando
Poo; Comoro Is. (Grand Comore), Madagascar, and Réunion.

10 species (+ 3 subspecies).

Key to sect. 1. Campylosporus

At Petals erect at anthesis; stamen filaments united for 2-5 mm; leaf venation wholly dichoto-
mous or with one pair of basal midrib branches ................0..ecceeeeeeee ees 1. bequaertii(p. 180)

Petals spreading or reflexed at anthesis; stamen filaments united for 1 mm or less; leaf
venation usually at least partly pinnately branched or reticulate .....................ccceeeeeee ees Z

2(1) Flowers all solitary; petals and stamens persistent; main leaf-venation either open or closed
bycrOss-Velms to Lorm(complete MtramManroial VEU <2... ....cedseceeoceceecesesncerscavesecuncecne 3

Flowers (at least some) in 2—~-flowered inflorescences or, if all solitary (Sp. 6a), then petals

and sometimes stamens tardily deciduous; main leaf-venation either open or closed by

branches from midrib to form incomplete intramarginal vein ...................c0cceeceeeeeeee ees v
3(2) Styles outcurved above; sepals outcurved in bud; stamens 3040 per fascicle (2. revolutum) 4
Styles ascending above or wholly appressed; sepals erect in bud; stamens 20-25 per fascicle 5

4(3) Leaves without or with 1-2 pairs of cross-veins not interrupting parallel veins; pedicel 5-8
mm long; stamens c. 40 per fascicle, equalling or exceeding styles ................00.ceeceeeeeee ee

2a. revolutum subsp. keniense(p. 184)
Leaves with 3-8 pairs of cross-veins interrupting parallel veins; pedicel 1-5 mm long;
stamens 30-35 per fascicle, shorter than or equalling styles ...................cceeeeeeeee eee eeeees

2b. revolutum subsp. revolutum (p. 185)

5(3) Styles 0-9 coherent or less; sepals ovate to triangular-ovate (3. lanceolatum) ..................... 6
Styles completely coherent; sepals elliptic or lanceolate to narrowly oblong.....................-
RR ene Sore sacl eh f Pee Linea eR Me de AIRES ME dau ntcloe eves Dae 4. madagascariense (p. 191)

180 NORMAN K. B. ROBSON

6(5) Leaves without or with 1-2 pairs of cross-veins not interrupting parallel veins; leaf lamina
concolorous, margin plane; styles c. 1-2-1-5 X ovary, 0-6-0-7 coherent.......................0.

3a. Janceolatum subsp. angustifolium(p. 190)
Leaves with 3-10 pairs of cross-veins interrupting parallel venation; leaf lamina + discol-
orous, margin usually + revolute; styles c. 1-8-3 x ovary, 0-8-0-9 coherent....................

3b. lanceolatum subsp. lanceolatum(p. 190)

7(2) Styles 1-1-2 x ovary, separating + completely in fruit; petals and sometimes stamens tardily
deciduous; dark glands absent; leaves with midrib branches often not reaching main
laterals (6=6: socotranum'sens. lat :)) j -cdac.. senretteore cetbo act ee oe wonipe dequeue ance aaceeee oe cte ne ee 8
Styles 1-2-3 X ovary, remaining united in fruit; petals and stamens persistent (or tardily
deciduous in Sp. 7); dark glands p-esent (except in Sp. 8); leaves with midrib branches
reaching main laterals wens. de cicheas ghee Saw ove Seaman aout oe oo eee ee OEROReTe CERO 10

8(7) Leaves + narrowly elliptic (1: b = 3-6), acute; sepals ovate to ovate-oblong (3-2-5-5 mm
wide), acute to obtuse, entire to denticulate; ovary ovoid; growing on granite ................
5. balfourii(p. 191)
Leaves elliptic to oblanceolate or subcircular (1: b = 1-3-2), acute or apiculate to rounded;
sepals oblong (2-5-3 mm wide), apiculate to rounded, entire; ovary ovoid to ovoid-conic;
growing on limestome'(6. Socotra) 0.00.20. .c.o.ecognee ace eee hoes ac naensee-teeeacas eee eee 9

9(8) Pedicels 1-6 mm long; leaves elliptic to oblanceolate (S-10 mm wide), acute to obtuse;
SHLOMO) TET Ab Manner are Receammacinecadedaqdeacnakacendood: 6a. socotranum subsp. socotranum(p. 193)

Pedicels 5—11 mm long; leaves broadly elliptic to subcircular (8-22 mm wide), apiculate to
rounded: flowers 157) sted tees een ee Roetome aneeee. 6b. socotranum subsp. smithii(p. 194)

10(7) Leaves without or with very laxly reticulate venation; petals tardily deciduous................... 11
Leaves with densely reticulate venation; petals persistent ..................ceceececeeceeecneeseneeees 12

11(10) Leaves sessile, 16-90 mm long; sepals ovate-lanceolate to triangular-lanceolate, 2-5 mm

wide; leaves, sepals and petals with dark glands ......................0005- 7. quartinianum(p. 194)
Leaves petiolate, 10-30 mm long; sepals linear-lanceolate, 1-7-2 mm wide; leaves, sepals
and petals with only pale or reddish glands .......................0eceeeeeeeeeees 8. synstylum(p. 197)

12(10) Leaves with + punctiform laminar glands; styles wholly united or almost so...................04-
9. roeperianum(p. 198)
Leaves with linear to striiform laminar glands; styles c. 0-7 united......... 10. gnidiifolium (p. 201)

1. Hypericum bequaertii De Wild.

in Rev. zool. afr. 8, suppl. bot.: 4 (1920) (‘Bequaerti’); Pl. bequaert. 1: 241 (1922);
GoodinJ. Bot., Lond. 65: 333, t. 582 f. 14 (1927); Hauman in Bull. Acad. r. Belg.,
Cl. Sci. V, 19: 705 (1933); Milne-Redh. in Fl. trop. E. Afr. Hyperic.: 5 (1953);
Hedberg in Symb. bot. upsal. 15(1): 130 (1957); Robson in Kew Bull. 12: 444
(1958); Spirlet, Contr. fl. Congo. Guttif.: 10 (1966); Bamps in Fl. Congo, etc.
Guttif.: 10, t, 1, f. 1D (1970), in Bull. Jard. bot. natn. Belg. 41: 441 (1971), Distr. pl.
afr. 3: map 74 (1971); Robson in Bamps, Robson & Verdcourt, Fl. trop. E. Afr.
Guttif.: 28 (1978), in Kew Bull. 33: 581 (1979). Type: Zaire, Ruwenzori, Butahu
[Butagu] Valley, 3800-4200 m, 7.iv.1914, Bequaert 3757 (BR, holotype; K!).

H. keniense sensu Mildbr., Deutsch. Zentr.-Afr.-Exped. 1907-1908 2: 560 (1913);
Staner in Bull. Jard. bot. Etat Brux. 13: 72 (1934); Lebrun in Bull. Agric. Congo
belge 25: 423 (1934); Robyns, Fl. Parc Nat. Albert 1: 624 (1948); Dale, Indig. trees
Uganda: 157 (1952); pro parte omnes, quoad syn. H. bequaertii.

Icones: Bamps in Fl. Congo, etc. Guttif.: t. 1, f. 1D (1970); Fig. 9.
Shrub or tree to 10(—12) m tall, often rather bushy, with branches ascending. Stems

red when young, soon terete; internodes 2-6 mm long, much shorter than leaves;
bark grey-brown. Leaves* sessile; lamina (10—)18-55(-60) x 4-15 mm, narrowly

* The leaves and petals in the text figures are all three-quarters of natural size unless otherwise indicated.

STUDIES IN THE GENUS HYPERICUM L. 181

M. Tebbs

Fig. 9 H. bequaertii: (a) habit: (b) leaf section; (c) sepal; (d) petal; (e) stamen fascicle; (f) anthers; (g)
ovary; (h) capsule (a X 1;h x 2; c-e, g x 4; b X 12, f x 20). All Taylor 3023.

182

\

NORMAN K. B. ROBSON

lanceolate to narrowly elliptic, acute, margin plane, base cuneate, reflexed-
auriculate, concolorous, often glaucous, subcoriaceous, lower ones soon deciduous;
venation: (3)5 basal or near-basal veins, unbranched, alternating with numerous
parallel dichotomising (non-functional ?) secondary veins, without midrib branches
or cross-veins; laminar glands in secondary venous system, uninterrupted, alternat-
ing with + irregular series of small dots, and with a few scattered large resin glands;
marginal glands pale, sparse. Inflorescence 1-flowered; pedicel c. 10 mm long; bracts
foliar but smaller and broader. Flowers 40-75 mm in diam., cyathiform to campanu-
late; buds ovoid, acute. Sepals 14-20 x 8-12 mm, imbricate, + unequal, soon
recurved in bud, broadly to narrowly ovate, acute to subacute, margin entire to
irregularly denticulate or ciliolate, midrib undifferentiated; laminar glands linear,
numerous; submarginal glands absent or rarely few; marginal glands absent or dark.
Petals* orange (? to bright yellow), flushed red outside, persistent, erect, 27-35 x
16-25 mm, 1-7-2 X sepals, obovate, with apiculus small, rounded; margin entire,
eglandular; laminar glands linear, interrupted distally. Stamen fascicles persistent,
each with c. 30 stamens, longest 20-28 mm long, c. 0-7 x petals, with filaments united
for 2-5 mm. Ovary c. 10 X 5 mm, ovoid-cylindric; styles c. 12 mm long, c. 1:2 x
ovary, 0-75 coherent, suberect distally; stigmas subglobose, sometimes bilobed.
Capsule 20-22 x 13-15 mm, ovoid-cylindric. Seeds dark reddish-brown, c. 1-5 mm
long, cylindric, not carinate, linear-reticulate. 2n = ?

‘In moist or protected localities, co-dominant in open forest or bush, occurring in the
upper part of the ericaceous belt and the lower part of the alpine belt’ (Hedberg,
1957: 131); (3150)3350-4350 m.

+Zaire (Ruwenzori), Uganda (Ruwenzori). Map 1.

ZAIRE. Kivu: Terr. Beni, Ruwenzori, Parc National Albert, viii. 1948, Robyns
3316 (K); Upper Ruanoli Valley, 3720 m, i.viii.1952, Ross 737 (BM).

UGANDA. Western: Toro distr., Ruwenzori; Namwamba Valley, Kilembe, 3600
m, 9.i.1936, Taylor 3023 (BM); Kitandara, 3600 m, viii.1953, Osmaston 3761 (K).

H. bequaertii has the greatest number of primitive characters of any species of
Hypericum. These include the parallel-dichotomous leaf venation, the large, solitary
flowers with massive parts, and the virtual absence of dark glands. The cyathiform
corolla and the relatively long union of the stamen filaments could be either primitive
or specialisations associated with high-altitude conditions.

H. bequaertii is closely related to H. revolutum subsp. keniense, which also occurs
on Ruwenzori but has a lower (but overlapping) altitudinal range and appears to
grow in somewhat drier habitats. H. bequaertii may be distinguished from it by the
persistently cyathiform to campanulate, largely red-tinged corolla. Both these taxa
are distinguished from the remaining mainland species of sect. Campylosporus by
the presence of a papillose epidermis (cf. Spirlet (1967: 13-19)).

2. Hypericum revolutum Vahl

Symb. Bot. 1: 66 (1790); Christensen in Dansk bot. Ark. 4: 39 (1922); Robson in Kew
Bull. 14: 251 (1960), in Fl. Zamb. 1: 381 (1961), in Bamps, Robson & Verdcourt,
FI. trop. E. Afr. Guttif.: 28 (1978), in Kew Bull. 33: 581 (1979); Paiva in Mems Jta
Invest. Ultramar (Bot.) I, No. 38: 76 (1963); Spirlet, Contr. fi. Congo Guttif.: 7, f.
1A-C (1966), in Bull. Inst. fr. Afr. noire A, 29 (1): 13, t. 1 f. 12 (1967); Moggi &
Pisacchi in Webbia 22: 236, f. 1, map 1 (1967); Bamps, Fl. Congo. etc. Guttif. 8, f.
1A, B (1970), in Bull. Jard. bot. natn. Belg. 41: 438 (1971), in Distr. Pl. Afr. 3: map
72 (1971) pro parte, excl. loc. Comores et Réunion; Lisowski, Malaisse &
Symoens in Bolm Soc. broteriana I, 44: 232 (1970); Agnew, Upl. Kenya wild fis:
186 (1974); Killick & Robson in FI. sthn Africa 22: 15 (1976); Troupin, Fi.
Rwanda, Sperm. 1: 299, f. 63 (1) (1978), Fl. Pl. lign. Rwanda: 215, f. 72, 1 (1982).
Types: Yemen, Arabia felix, 1763, Forsskal 796 (C!, lectotype — mihi); ibid.,
Forsskal 792 (C!, syntype).

+ All specimens cited in geographical lists have been seen by the author unless otherwise indicated. The lists are not
necessarily comprehensive, those for the commoner species having been compiled to illustrate distribution.

STUDIES IN THE GENUS HYPERICUM L. 183

Map1 Sect. 1. Campylosporus: 1, H. bequaertii A, 2a. H. revolutum subsp. keniense @.

Shrub or tree (0-3—)1—10(—12) m tall, bushy or slender, with branches ascending to
spreading. Stems red to yellow-brown when young, soon 2-lined, eventually terete;
internodes 2-6 mm long, much shorter than leaves; bark dark brown to grey. Leaves
sessile; lamina (11—)15—45(-60) x 2-5-12 mm, + narrowly elliptic to narrowly oblong
or very rarely oblanceolate, acute, margin plane to recurved, base cuneate, reflexed-
auriculate, paler beneath, sometimes glaucous, subcoriaceous to chartaceous, lower
ones + soon deciduous; venation: (1)3—5 basal or near-basal veins, laterals bran-
ching dichotomously (mostly near the base), with numerous parallel or interrupted
and + vestigial secondary veins, without or with 1-8 cross-veins and a loose or +
dense tertiary reticulum; laminar glands in secondary venous system either (i) not
interrupted and alternating with + irregular series of small dots and short streaks or
(ii) interrupted and wholly streaks and lines or with a few dots; marginal glands pale
to dark, sparse to dense. Inflorescence 1-flowered; pedicel 1-8 mm long; bracts foliar
but smaller and broader. Flowers (35)40-80 mm in diam., stellate; buds ovoid to
subglobose, acute to rounded. Sepals 5-18 x 3-9 mm, imbricate, + unequal, soon
recurved in bud, broadly to narrowly ovate or subcircular, acute to subrounded,
margin entire or usually + irregularly glandular-denticulate or -ciliolate, midrib
undifferentiated or obscure; laminar glands linear, numerous; submarginal glands
absent or dark; marginal glands absent or dark. Petals orange-yellow to golden-
yellow, flushed red to orange outside, persistent, spreading, 15S—40(-43) x 5-26 mm,
2-3-3 X sepals, obovate-oblanceolate to oblanceolate, with apiculus small, rounded;
margin entire or with a few dark sessile glands or glandular cilia; laminar glands
linear, interrupted distally. Stamen fascicles persistent, each with 30-40 stamens,
longest 11-20 mm long, c. 0-5 X petals, with filaments united for c. 1 mm. Ovary
3-5-9 x 3-5—-5-5 mm, ovoid to ovoid-globose; styles 4-12 mm long, 0-8-1-6 X ovary,
c. 0-8 coherent, spreading distally; stigmas subglobose to narrowly capitate. Capsule
9-15 X 7-11 mm, ovoid to subglobose. Seeds reddish-brown, 1:2—1-5 mm long, +

184

NORMAN K. B. ROBSON

narrowly cylindric, sometimes narrowly carinate and/or with distal expansion, finely
linear-reticulate. 2n = 24 (both subspecies).

In open forest, forest margins, savannas, and grassland; 900-4000 m.

From SW. Arabia and Ethiopia to the Transvaal; also in Cameroon and Fernando
Poo. Maps 1, 2.

H. revolutum comprises two subspecies (Robson, 1979), one widespread (subsp.
revolutum), the other confined to the East African mountains (subsp. keniense).

2a. H. revolutum subsp. keniense (Schweinf.) N. Robson

in Bamps, Robson & Verdcourt, Fl. trop. E. Afr. Guttif.: 28 (1978), in Kew Bull. 33:
582 (1979). Type: Kenya, Mt. Kenya, western slope at 1150 m, 1887, von Hoehnel
(Bt, holotype; BM [fragment]!)

H. keniense Schweinf. in von Hoehnel, Reise Rudolfsee: 15, 868 (1892); Engler in
Phys. Math. Abh. K. Akad. Wiss. Berlin 1891: 308 (1892), Pflanzenw. Ost-Afr. C:
274 (1895) pro parte, quoad spec. ex Kenia; Engler & Diels, Veg. der Erde 3 (2):
499 (1921) pro parte, quoad spec. ex Kenia; T. C. E. Fries in Notizbl. bot. Gart.
Mus. Berlin 8: 565 (1923); R. Keller in Engler & Prantl, Nat. Pflanzenfam. 2nd ed.
21: 176 (1925); Good in J. Bot., Lond. 65: 333, t. 582 ff. 12, 13 (1927); Dale, Indig.
trees Uganda: 157 (1952); Milne-Redh. in Kew Bull. 8: 434 (1953), in Fl. trop. E.
Afr. Hyperic.: 5 (1953); Hedberg in Symb. bot. upsal. 15 (1): 130 (1957), Feat.
Afroalpine pl. ecol.: 89 (1964); Dale & Greenway, Kenya trees & shrubs: 235
(1961); Spirlet, Contr. fl. Congo Guttif.: 10, f. 1D (1966), in Bull. Inst. fr. Afr.
noire A, 29 (1): 16, t. 1 ff. 14-15 (1967); Bamps in Fl. Congo etc. Guttif.: 10, f. 1C
(1970), in Bull. Jard. bot. natn. Belg. 41: 440 (1971), in Distr. pl. Afr. 3: t. 73
(1971); Agnew, Upl. Kenya wild fls.: 186 (1974); Troupin, Fl. Rwanda, Sperm. 1:
299, f. 63 (2) (1978).

H. ruwenzoriense De Wild. in Rev. zool. afr. 8 suppl. bot.: 5 (1920); Pl. Bequaert. 1:
242 (1922); Goodin J. Bot., Lond. 65: 333, t. 582 (1927); Staner in Bull. Jard. bot.
Etat Brux. 13: 71 (1934); Lebrun in Bull. Agric. Cong. belg. 25: 423 (1934);
Robyns, Fl. Parc Nat. Albert 1: 623 (1948); Brenan, Checkl. Tang. Terr. 2: 250
(1949); Dale, Indig. trees Uganda: 159 (1952); Spirlet, Contr. fl. Congo Guttif.: 8,
f. 1E-F (1966), in Bull. Inst. fr. Afr. noire A, 29 (1): 16, t. 1 f. 13 (1967). Type:
Zaire, Ruwenzori, Butahu [Butagu] Valley, Bequaert 3705 (BR).

Icon: Bamps in Fl. Congo. etc. Guttif.: f. 1 C (1970).

Stems red when young. Leaves with 3-S basal or near basal veins, without or with 1-2
cross-veins not interrupting the numerous parallel secondary veins or changing their
direction. Pedicel S—8 mm long. Flowers 45-80 mm in diam.; buds acute to obtuse.
Sepals narrowly to broadly ovate. Petals flushed red outside. Stamen fascicles each
with c. 40 stamens equalling or exceeding styles. Styles 7-12 mm long. Capsule
(10)12-15 mm long.

‘In open forest or bush on comparatively dry ground, occurring from the montane
forest belt into the lower part of the alpine belt’ (Hedberg, 1957: 130); (1850—)2400-
3750(—4000) m.

Zaire (Ruwenzori, Mt. Muhi, Mt. Kahuzi), Uganda (Ruwenzori, Elgon), Kenya
(Elgon, Aberdares, Mt. Kenya), Tanzania (Kilimanjaro, Meru, Hanang, Loolmala-
sin). Map 1.

ZAIRE. Kivu: Ruwenzori, Butahu valley, Mutsora—Kalonge path, 2000 m,
13.vii.1954, Osmaston 3301 (BM); Terr. Kalehe, Mont Kahuzi, flanc est, 3240 m,
29.xii.1971, Bamps 2894 (BR).

UGANDA. Western: Toro distr., Ruwenzori, Bujuku valley, near Bigo camp,
3450'm, 1.iv.1948, Hedberg 624 (K); Mbale distr., Mt. Elgon, above Butandiga, 2100
m, vili.1934, Synge 864 (BM).

KENYA. Rift Valley: Trans-Nzoia distr., Mt. Elgon (north-east), 3360 m,
iii.1960, Tweedie 1985 (K); Naivasha distr., Aberdare Range, Kinangop, 3600 m,

STUDIES IN THE GENUS HYPERICUM L.

17.vii.1948, Hedberg 1629 (K). Central: N. Nyeri distr., Mt. Kenya, Sirimon track,
3000 m, 30.xii.1974, Williams 70 (K).

TANZANIA. Northern: Moshi distr., Kilimanjaro, Marangu, 2850 m,
17.xii.1960, F. G. Smith 1401 (K); Arusha distr., Mt. Meru, eastern slope, 2500-3150
m, 6.1.1971, Vanska (H); Mbulu distr. , Loolmalassin Mtn, 3300-3570 m, 17.ix.1932,
Burtt 4211 (K).

H. revolutum subsp. keniense is the nodal taxon in sect. Campylosporus (and hence
in Hypericum as a whole). Only H. bequaertii has more primitive characters, and
even its claim to overall primacy in this respect is open to doubt. Subsp. keniense is
very closely related to H. lanceolatum subsp. angustifolium (q.v.) from Réunion, but
can be distinguished from it inter alia by the outcurving sepals and styles.

The distinctions between subsp. keniense and subsp. revolutum are less clear,
despite the length of the differential descriptions. Intermediate specimens are not
infrequent at altitudes between 1800 and 3600 m, especially in the east and south of
the distribution of subsp. kKeniense. Thus, whereas pure forms of the subspecies are
common on Ruwenzori and Mt. Elgon, they are rarer in the Aberdares and on Mt.
Kenya; on the Tanzanian volcanoes they are much less common than the intermedi-
ates. In an earlier paper (Robson, 1979), I have suggested that the incomplete
evolution of subsp. revolutum from subsp. keniense is likely to have occurred before
the elevation of the East African volcanoes.

‘2b. H. revolutum subsp. revolutum

H. kalmii Forssk., Fl. aegypt.-arab.: cxviii (1775), nomen.

H. kalmianum Vahl, Symb. bot. 1: 66 (1790), in synon., non L. (1753).

H. leucoptychodes Steudel [in sched. Pl. Schimper. Abyss. 834, 1177; Delile in
Rochet d’Héricourt, Sec. voy. Choa: 339 (1846)] ex A. Rich., Tent. fl. abyss. 1: 96
(1847); Good in J. Bot., Lond. 65: 330, t. 582 ff. 3-5 (1927); Norlindh in Bot.
Notiser 1934: 100 (1934); Brendell in Bothalia 3: 580, map (1939); Phillips in
Flower. pl. S. Afr. 20: t. 787 (1940); Brenan, Checkl. Tang. Terr. 2: 249 (1949);
Andrews, Fi. pl. Anglo-Egyptian Sudan 1: 213 (1950); Dale, Indig. trees Uganda:
157, t. 33 (1952); Pardy in Rhodesia Agric. J. 53: 514, tt. 515-6 (1956). Types:
Ethiopia, Ametscha prope Dschenausa, 1.i1.1840, Schimper 834 (P, lectotype
—mihi; FI, G, K!, L, LE, M, MO, PAL, S, Z); in latere boreali montis Bachit,
6.vi. 1838, Schimper 1177 (FI, G, K!, LE, P).

H. lanuriense De Wild., Pl. bequaert. 5: 403 (1932); Robyns, Fl. Parc. Nat. Albert 1:
622 (1948). Types: Zaire, Ruwenzori, Ruanoli [Lanuri] Valley, Bequaert 4460
(BR, lectotype; K!, photo); Zaire, Ruwenzori, vallée du Batagu vers 300 m,
Bequaert 3716 (BR).

H. lanceolatum sensu Choisy, Prod. monogr. fam. Hypéric.: 41 (1821), in DC.,
Prodr. syst. nat. regni veg. 1: 545 (1824); Oliver, Fl. trop. Afr. 1: 156 (1868) pro
parte, excl. typum; R. Keller in Engler & Prantl, Nat. Pflanzenfam. 3 (6): 209, f.
100 T, U (1895) pro parte, 2nd ed. 21: 176 (1925) pro parte; Engler, Pflanzenw.
Ost-Afr. C: 274 (1895), in von Gétzen, Durch Afr. von. O. nach W.: 376 (1895), in
Mildbraed, Deutsch. Zentr.-Afr.-Exped. 1907-1908 2: 560 (1913), in Engler &
Diels, Veg. der Erde 3 (2): 497, t. 229 (1921); Th. & H. Durand, Syl. pl. Afr.: 43
(1909); Baker f. inJ. Linn. Soc. (Bot.) 40: 26 (1911); Mildbraed in H. Meyer, Mitt.
Deutsch, Schutzgeb. Erganzungsch. 6: 96 (1913); R. E. Fries, Wiss. Ergebn.
Schwed. Rhod. Kongo Exp. 1911-1912 1: 150 (1914); Eyles in Trans. R. Soc. S.
Afr. 5: 420 (1916); De Wild., Pl. bequaert. 1: 241 (1922); T. C. E. Fries in Notizbl.
bot. Gart. Mus. Berlin 8: 565 (1923); Staner in Bull. Jard. bot. Etat Brux. 13: 74
(1934); Robyns, Fl. Parc Nat. Albert 1: 620, t. 62 (1948); Milne-Redh. in Fi. trop.
E. Afr. Hyperic.: 4 (1951), in Mem. N.Y. bot. Gdn 8 (3): 221 (1953); Keay &
Milne-Redh. in Fl. W. trop. Afr. 2nd ed. 1 (1): 287, t. 109 (1954); N. Robson in
Kew Bull. 12: 444 (1958) pro parte, quoad loc. Afric.; Pellegrin in Bull. Soc. bot.
Fr. 106: 217 (1959); Cufod. in Bull. Jard. bot. Etat. Brux. 29 suppi.: 588 (1959);
Dale & Greenway, Kenya trees & shrubs: 235, f. 46 (1961); Mooney in Proc. Linn.
Soc. Lond. 174: 147, t. 2a (1963); Gilli in Annin naturh. Mus. Wien 74: 425 (1970);
pro parte omnes, non Lam. (1797).

185

186

NORMAN K. B. ROBSON
Icon: Moggi & Pisacchi in Webbia 22: 239, f. 1 (1967).

Stems yellow-brown when young. Leaves with 3 basal veins, with 3-8 cross-veins
interrupting the + numerous secondary veins and often changing their direction.
Pedicel 1-5 mm long. Flowers 35-55 mm in diam.; buds obtuse to rounded. Sepals
broadly ovate to subcircular. Petals flushed orange outside. Stamen fascicles each
with 30-35 stamens shorter than or equalling styles. Styles 4-8 mm long. Capsule
9-12(—14) mm long.

Montane forest margins, fringing forest, secondary forest, scrub, savannahs and
grassland on dry or marshy ground; 900-4000 m.

Saudi Arabia (Asir), Yemen, Equatorial Guinea (Fernando Poo), Cameroun,
Nigeria, Ethiopia, Sudan Republic, Kenya, Uganda, Zaire, Burundi, Rwanda,
Tanzania, Malawi, Zambia, Zimbabwe, Mozambique, Swaziland, South Africa
(Transvaal, Natal, Cape Prov.). Map 2.

SAUDI ARABIA. Asir: Suda [Al Sudah], 3000 m, i.vii.1946, Vesey-Fitzgerald
16086/1 (BM).

YEMEN. Taiz, Jebel Sabir, 2400-2700 m, 21.x.1975, Hepper 5948 (K).

EQUATORIAL GUINEA. Fernando Poo: Moka region, 1350-1500 m,
30.1.1933, Exell 798 (BM).

NIGERIA. Eastern: Chappal Waddi (Gangirwal), 1500-2400 m, 19.xi.1969,
Jackson, Magaji & Tuley 2084 (K).

CAMEROON. West; Cameroon Mtn, Mann’s Spring, slopes of Vefondi, 2890—
2050 m, 1.iv.1948, Brenan, Jones & Richards 9556 (K).

SUDAN REPUBLIC. Equatoria [El Istwaya]: Didinga Mts, Mt. Lotuke, 2700 m,
i1.1939, Macdonald 100 (BM).

ETHIOPIA. Eritrea: Scimenzana, sul monte Guna guna, 21.1x.1902, Pappi 225
(BM, K). Tigray: Urahut, 2400 m, 28.1.1862, Schimper 906 (BM, E). Gonder:
Wolkofit Pass, 3000 m, 25.v.1953, Greathead 95 (BM, EA). Gojjam: Choké Mts, W.
side of upper Godeb gorge, c. 2700 m, 30.viii.1957, Flenley & Hillier 266 (K). Shoa:
Entotto Mtn, 2850 m, 29.i.1965, Perdue 6435 (K). Keffa: Jimma area, Red Mtn, Aba
dima Farm, 10.1.1962, Siegenthaler 1624 (EA, K). Gamo-Gofa: Gamo, Borodda to
Ciencia, 5.11.1938, Vatova 1698 (FI). Sidamo: Amaro Mts, E. slope of Mt. Delo,
2400-3000 m, Gillett 14912 (K). Arssi: Mt. Boroluccu, c. 50 km SE. of Asella, de
Wilde 10060 (K). Bale: Stella Waghe near Gurio, 3600 m, 4.iv.1958, Mooney 7114
(K). Harerghe: Kondudo Mtn, 2400-2700 m, 4.xi.1961, Burger 1221 (K).

KENYA. Northern Frontier: Mt. Nyiru, 2700 m, 5.1.1959, Newbould 3438 (K).
Rift Valley: S. Kinangop, Chenia stream, 2490 m, 22.xi.1953, Verdcourt 1034 (K).
Central: Mt. Kenya (N. sector), Timberless Gap, 2720 m, 29.vii.1949, Schelpe 2456
(BM). Masai: Narok, Enesambulai Valley, 2640 m, 12.1x.1970, Greenway & Kanuri
14571 (K).

UGANDA. Northern: Karamoja, Morongole Mtn, 2580 m, 1i-xi.1939, Thomas
3318 (K). Western: Kigezi, Mt. Muhavura, 2100-2400 m, 17.x.1929, Snowden 1514
(BM, K). Eastern: Mt. Elgon, Butandinga, 2250 m, Dale U.18 (K).

ZAIRE. Haute Zaire: Terr. Djugu, Mt. Oboro (L. Albert), 2450 m, 26.11.1958,
Bamps 149 (K). Kivu: Terr. Mwenga, Mt. Muhi, 3045 m, 31.viii.1955, Kinet 94 (K).
Shaba: environs de Kasiki, R. Lunangwa, 2050 m, 10.iv.1970, Liswoski, Malaisse &
Symoens 12000 (K).

RWANDA. Kibuye, Gisovu, forét de Nyungwe, 2250-2400 m, 17.vi.1978, Raynal
20620 (BM).

BURUNDI. Route Usumbura—Astrida a 50 km d’Usumburu, 3.1.1948, Taton 803
(K).
TANZANIA. Northern: Arusha, Mbulu, Nou Forest, 2250 m, 22.ix.1966, Car-
michael 1241 (K). Tanga: W. Usambara Mt., near Lushoto, 1260-1650 m, 29.1.1932,
St. Clair-Thompson 259 (K). Western: Kungwe—Mahali Peninsula, S. of Kungwe,
above Kasogi, c. 1500 m, 12.ix.1959, Harley 9612 (K). Eastern: Uluguru Mts,
Banduki, 1200 m, 26.xii.1938, Vaughan 2631 (BM). Southern Highlands: Mbeya,
Chunya Escarpment, 2340 m, 14.xii.1962, Richards 17053 (K).

MALAWI. Northern: Nyika Plateau Valley, c. 4km SW. from Rest House, 2100

STUDIES IN THE GENUS HYPERICUM L. 187

iF
i}
i
1
'
\
)
i}
i}

Map2_ Sect. 1. Campylosporus: 2b. H. revolutum subsp. revolutum.

m, 22.xi.1958, Robson & Angus 246 (BM, K). Central: Dedza, Chongoni Mtn, c.
2100 m, 29.v.1960, Chapman 722 (BM), 722a (K). Southern: Zomba Plateau, by
Chingwe’s Hole, 1880 m, 11.11.1970, Brummitt & Banda 8514 (K).

ZAMBIA. Eastern: Nyika Plateau, 1-6kmN. of Rest House, 2100 m, 27.xi.1955,
Lees 95 (K). [Central: Broken Hill, xi.1928, van Hogsen 1231 (PRE), probably
wrong location].

ZIMBABWE. Eastern: Umtali, rim of Umkarara Valley, Stapleford Forest
Reserve, 1680 m, 16.vii.1955, Chase 5676 (BM, K, SRGH).

MOZAMBIQUE. Manica & Sofala: Gorongoza, planalto da serra, 1800 m,
26.ix.1943. Torre 5937 (BM).

SWAZILAND. Pigg’s Peak, near Havelock, c. 1350 m, 16.ix.1960, Compton
30123 (K).

188

NORMAN K. B. ROBSON

[LESOTHO. Occurrence not confirmed. See Jacot-Guillarmod, Fl. Lesotho: 212
(1971).]

SOUTH AFRICA. Transvaal: Zoutpansberg, c. 6-5 km W. of Lake Fundust,
14.vii.1935, Galpin 14917 (K, LU). Natal: Zululand, Nkandhla Forest, Pateni
Estates, 1200 m, 27.viii.1967, Strey 7636 (K). Cape Province: Kokstad, Mt. Malowe,
1350 m, Tyson 3044 (K).

CULTIVATED. Specimens seen from England, Scotland, Ireland, and U.S.A.
(California), dated 1935-1967, and living material up to 1984.

Except in the areas where it intergrades with subsp. keniense (q.v.), subsp. revolu-
tum is mostly fairly constant in morphology, varying mainly in leaf venation and style
length. In two separate regions, however, relatively broad-leaved forms have
evolved. In the Uluguru Mts, eastern Tanzania (e.g. Bruce 73 — BM, K), the
subspecies tends to have long, elliptic to oblanceolate leaves (23-40 x 6-10 mm; 1: b
= c. 4) with the apex subacute, instead of the usual 10-25 x 2-5 mm (1: b = c. 4-5)
with the apex acute (on mature shoots). In Eritrea, Ethiopia (e.g. Pappi 225 — BM,
K.), there is a form in which the leaves are short and relatively broad (10-15 x 5-6; 1:
b = 2-2-5) with the apex apiculate to obtuse.

3. Hypericum lanceolatum Lam.

Encycl. méth. (Bot.) 4: 145 (1797); Choisy, Prodr. monogr. Hypér.: 41 (1821), in
DC., Prodr. syst. nat. regni veg. 1: 545 (1824); Oliver, Fl. trop. Afr. 1: 156 (1868)
pro parte, quoad typum; Engler in Phys. Math. Abh. K. Akad. Wiss. Berlin 1891:
306 (1892) pro parte, quoad typum; R. Keller in Engler & Prantl, Nat. Pflanzen-
fam. 3 (6): 209 (1893), 2nd ed. 21: 176 (1925) pro parte, quoad pl. ex Réunion;
Cordemoy, Fl. Réunion: 334 (1895); Good in J. Bot., Lond. 65: 330, t. 582 f. 1
(1927); Perrier in Fl. Madag. et Comores 35: 3 (1951) pro parte, quoad typum et
syn. Campylosporus angustifolius et pl. ex Comoros et Réunion; Robson in Kew
Bull. 12: 444 (1958) pro parte quoad typum, in Kew Bull. 33: 582, map 2 (1979), in
Robson & Stevens, Fl. Mascareignes 49. Guttif.: 3 (1980). Type: Réunion, La
Réunion, Ambaville, 1771, Commerson (P-LA!, holotype; LINN-SM!).

Shrub or small tree (0-2—)1—7(-10) m tall, slender, much branched, with branches
drooping then ascending. Stems yellow-brown when young, soon 2-angled, eventual-
ly terete; internodes 2—20(-35) mm long, usually much shorter than leaves; bark
red-brown to grey. Leaves sessile; lamina 17-50 x 3-10 mm, narrowly lanceolate to
+ narrowly elliptic or more rarely elliptic or oblanceolate, acute to rounded, margin
plane to + recurved, base narrowly to broadly cuneate, reflexed-auriculate, paler
beneath or concolorous, sometimes + glaucous, subcoriaceous, lower ones even-
tually deciduous; venation: (1)3—5 basal or near-basal veins, the laterals branching
dichotomously near the base, with numerous parallel or interrupted and + vestigial
secondary veins, without or with 1-10 cross-veins and a loose or + dense tertiary
reticulum; laminar glands in secondary venous system either (i) not interrupted and
alternating with + irregular series of short streaks and lines or (ii) interrupted and
wholly streaks or dots; marginal glands pale to dark, dense. Inflorescence 1-flowered;
pedicel 5-15 mm long; bracts foliar but smaller. Flowers (35—)40-70 mm in diam.,
cyathiform to stellate; buds ovoid, acute. Sepals 7-11 xX 4-7 mm, imbricate,
subequal, erect in bud, ovate to triangular-ovate, acute to rounded, margin entire or
+ minutely glandular-ciliolate (especially towards the apex), midrib undifferenti-
ated or obscure; laminar glands linear, numerous; submarginal glands absent or
dark; marginal glands absent or dark. Petals golden yellow, not flushed red or orange
outside, persistent, spreading, 25-38 x 15-25 mm, c. 3-5 X sepals, broadly to
narrowly obovate, with apiculus small, rounded; margin entire or with minute dark
sessile glands or glandular cilia, rarely with a few dark intramarginal dots distal to
apiculus; laminar glands linear, scarcely interrupted distally. Stamen fascicles persis-
tent, each with 20-25 stamens, longest 15—22(—28) mm long, c. 0-6-0-8 X petals, with
filaments united for c. 1 mm. Ovary 5-10 X 3-5-7 mm, ovoid-conic; styles 8-18 mm
long, 1-2-3 x ovary, 0-6-0-9 coherent, ascending distally; stigmas subglobose to
narrowly capitate. Capsule 12-14 x 10-12 mm, ovoid-pyramidal. Seeds reddish-

STUDIES IN THE GENUS HYPERICUM L. 189

brown, 1-0—1-6 mm long, narrowly cylindric to fusiform, not carinate or winged or
with distal expansion, finely linear-reticulate, 2n = ?

In mid-altitude forest, forest clearings and margins, dwarf forest patches, heathland,
savannahs, and grassland; (550—)1100—2400 m.

Réunion and the Comoros (Grand Comore). Map 3.

H. lanceolatum comprises two subspecies (Robson, 1979), one (subsp. angusti-
folium) confined to Réunion, the other (subsp. /anceolatum) also on Grande Comore.

Map 3 Sect. 1. Campylosporus: 3a. H. lanceolatum subsp. angustifolium A, H. lanceolatum subsp.
lanceolatum @, H. madagascariense @.

190

NORMAN K. B. ROBSON
3a. H. lanceolatum subsp. angustifolium (Lam.) N. Robson

in Kew Bull. 33: 582 (1979), in Robson & Stevens, Fl. Mascareignes 49. Guttif.: 5t. 1,
ff. 4-7 (1980). Type: Réunion, La Réunion, 1771, Commerson (P-LA!, holotype;
K!, P!, isotypes).

H. angustifolium Lam., Encycl. Méth. (Bot.) 4: 145 (1797); Choisy, Prodr. monogr.
fam. Hypéric.: 41 (1821), in DC., Prodr. syst. nat. regni veg. 1: 545 (1824);
one Fl. Réunion: 334 (1895); Good in J. Bot., Lond. 65: 330, t. 582 f. 2

1927).

Campylosporus angustifolius (Lam.) Spach, Hist. Nat. Vég. Phan. 5: 425 (1836), in
Ann. Sci. Nat. (Bot.) II, 5: 363 (1836).

Norysca angustifolia (Lam.) Blume, Mus. bot. Lugd. Bat. 2: 22 (1856).

H. lanceolatum forma angustifolium (Lam.) H. Perrier in Fl. Madag. et Comores
135: 3 (1951).

Icon: Robson & Stevens, Fl. Mascareignes 49. Guttif.: t. 1, ff. 4-7 (1980).

Stem internodes 2-6 mm long. Leaves 17-32 x 3-6 mm, narrowly lanceolate-elliptic
to narrowly lanceolate, acute, margin plane, concolorous or somewhat paler be-
neath, with 3-5 basal or near-basal veins, without or with 1-2 cross-veins not
interrupting the numerous parallel secondary veins or changing their direction.
Pedicel c. 5mm long. Stamen fascicles 15-18 mm long, c. 0-7 X petals. Styles 8-10 mm
long, c. 1-1-1-5 X ovary, 0-6-0-7 coherent. Seeds 1:2-1-6 mm long.

In dwarf forest patches, heathland, and rough pasture; 1900-2400 m.

Réunion. Map 3.
REUNION. Plaine des Ramparts, c. 1900 m, 10.v.1976, Coode 5175 (K); Plaine
des Salazes, 2400 m, iv.1957, Léo (P).

H. lanceolatum subsp. angustifolium is very similar to H. revolutum subsp. keniense,
differing only in the sepals (which are erect in bud) and the smaller number of
stamens per fascicle. It intergrades with subsp. /anceolatum at altitudes between 1500
and 1900 m, in a similar way to that in which the two subspecies of H. revolutum
intergrade. For a discussion of the distribution and variation in these species, see
Robson (1979).

3b. H. lanceolatum subsp. lanceolatum

H. penticosia Commerson ex Lam., Encycl. méth. (Bot.) 4: 145 (1797), in synon.

Campylosporus reticulatus Spach, Hist. nat. vég. Phan. 5: 424 (1836), in Ann. Sci.
nat. (Bot.) II, 5: 363 (1836), nom. illegit.

Norysca lanceolata (Lam.) Blume, Mus. bot. Lugd. Bat. 2: 22 (1856).

H. revolutum sensu Robson in Kew Bull. 14: 251 (1960); Bamps in Fl. Congo etc.
Guttif.: 8 (1970), in Distr. pl. Afr. 3: map 72 (1971) pro parte omnes, quoad pl. ex
Comores et Réunion.

Icon: Robson & Stevens, Fl. Mascareign. 49. Guttif.: t. 1 ff. 1-3 (1980).

Stem internodes 2—20(-35) mm long. Leaves 20-50 x 3-10 mm, narrowly lanceolate
or narrowly elliptic to elliptic or oblanceolate, acute to rounded, margin usually +
revolute, discolorous, with (1)3 basal veins, and with 3-10 cross-veins interrupting
and often deflecting the secondary veins to form numerous streaks or dots. Pedicel
5-15 mm long. Stamen fascicles 16—22(—28) mm long, c. 0-75 X petals. Styles 9-18 mm
long, c. 1-8-3 X ovary, c. 0-8-0-9 coherent. Seeds 1-1-2 mm long.

In mid-altitude forest and clearings, and in heathland; 550-1800 m.

Comoros (Grande Comore), Réunion. Map 3.

COMOROS. Grande Comore, 1886, Humblot 1581 (BM, P) (N.B. Printed labels
of this collection wrongly state ‘Angouan’).

REUNION. Centier de la Roche Ecrite, above St Denis, c. 1200 m, 25.xi.1973,
Coode & Badré 4198 (K); route du Maido, 1500 m, 16.xi.1970, Friedmann 567 (P).

H. lanceolatum subsp. lanceolatum is more variable than subsp. angustifolium, but

STUDIES IN THE GENUS HYPERICUM L.

can usually be distinguished from it by the large, reticulate-veined leaves with
revolute margins, and the longer, more coherent styles.

4. Hypericum madagascariense (Spach) Steudel

Nomencl. Bot. 2nd ed, 1: 788 (1840); Robson in Kew Bull. 33: 583 (1979). Type:
Madagascar, Prov. Emerina [Imerina] in montibus meridiem versus Hilsenberg &
Bojer (P!, holotype; BM!).

Campylosporus madagascariensis Spach, Hist. nat. vég. Phan. 5: 426 (1836), in Ann.
Sci. nat. (Bot.) I, 5: 363 (1836).

H. lanceolatum sensu R. Keller in Engler & Prantl, Nat. Pflanzenfam. 3 (6) 209
(1893), 2nd ed. 21: 176 (1925) pro parte, quoad pl. Madagasc.; Perrier, F/. Madag.
et Comores 135: 3 (1951) pro parte, excl. typum et syn. Campylosporus angustifo-
lius et pl. ex Comoros et Réunion.

Icon: Perrier, F/. Madag. et Comores 135: f. I, 1-6.

Shrub 1-3 m tall, ‘ericoid’, with branches ascending or spreading. Stems reddish-
brown when young, soon 2-angled, eventually terete; internodes 1-5—10 mm long,
shorter than leaves; bark red-brown. Leaves sessile; lamina 8-18 x 2-9 mm,
narrowly oblong or narrowly elliptic-oblong to broadly elliptic or obovate, obtuse to
rounded-apiculate, margin plane or recurved, base cuneate, reflexed-auriculate,
paler beneath, not glaucous, subcoriaceous, lower ones eventually deciduous;
venation: | basal vein, the laterals interrupted by 6-10 cross-veins and reduced to +
short streaks and dots in the areoles of the + dense tertiary reticulum; laminar glands
in secondary venous system, numerous, streaks or dots; marginal glands pale to
dark, dense to sparse, or absent. Inflorescence 1-flowered; pedicel 2-4 mm long;
bracts intermediate in form. Flowers 25-35 mm in diam., stellate; buds ovoid,
subacute. Sepals 7-11 x 2-3 mm, imbricate, subequal, erect in bud but eventually
more or less spreading, ovate to triangular-ovate, obtuse to rounded, margin entire
or with a few usually glandular cilia towards the apex, midrib + distinct; laminar
glands linear, numerous; submarginal glands dark; marginal glands absent or dark.
Petals golden yellow, not flushed red or orange outside ?, persistent, spreading or
reflexed, 12-20 x 6-9 mm, c. 2 X sepals, obovate-oblanceolate, with apiculus small,
rounded; margin entire or with minute dark sessile or subsessile glands, with dark
intramarginal dots at outer margin; laminar glands linear, scarcely interrupted
distally, and occasionally also dark, punctiform. Stamen fascicles persistent, each
with c. 25 stamens, longest 7-10 mm long, c. 0-5 X petals, with filaments united at the
base. Ovary 3-4 x 2-5-3 mm, broadly ovoid; styles S—6 mm long, c. 1-5 X ovary,
completely coherent; stigmatic mass capitate. Capsule 7-8 x 6-7 mm, broadly ovoid.
Seeds yellowish-brown, c. 1-5 mm long, narrowly cylindric, not carinate or winged or
with distal expansion, finely linear-reticulate. 2n = ?

In clearings and damp depressions in montane forest; (1400—)1700—2700 m.

Madagascar (Tsaratanana massif (northern) and Imerina massif (central). Map 3.

MADAGASCAR. Centre (Centre-Nord): Massif du Tsaratanana, plateaux
supérieurs et hauts sommets de l’ Amboabory a |’ Antsianongatalata, 2600-2700 m,
xi-xii.1937, Humbert 18376 (P). Centre (Centre-Moyen): Prov. Imerina, Massif
d’Ankaratra ?, [1822-1823], Hilsenberg & Bojer (BM, P).

H. madagascariense was not recognised as distinct from H. lanceolatum sensu lato by
either Keller or Perrier de la Bathie, although its completely coherent styles and
narrower sepals, smaller flowers and smaller, relatively broader leaves make it easily
distinguishable. It appears to be derived from H. lanceolatum subsp. lanceolatum.

All the recent collections that I have seen have come from the north of the island,
and so the species may now be extinct in the centre.

5. Hypericum balfourii N. Robson, sp. nov.

H. mysurensi Wight et Arnott affinis, sed petalis laterale apiculatis apiculo semper
rotundato, staminorum fasciculis tarde deciduis, stylis brevioribus (5-8 mm

191

192

D eG

NORMAN K. B. ROBSON
longis), ad 1/2 vel 5/6 coalitis, capsulis longioribus (13-15 mm longis), differt.

BM!, E!).

H. mysorense sensu Balf. f. in Trans. R. Soc. Edinb. 31: 26 (1888); Hutchinson, Fam.
Fl. Pl.: 186 (1926), 2nd ed. 2: 297 (1959), 3rd ed.: 336 (1973); Good in J. Bot.,
Lond. 65: 334, t. 582 f. 15 (1927); Robson in Kew Bull. 12: 444 (1958) pro parte
omnes, quoad pl. socotranum.

H. mysurense sensu Moggi & Pisacchiin Webbia 22: 260, f. 8, map 5 (1967) pro parte,
excl. typum.

H. socotranum sensu Moggi & Pisacchi in Webbia 22: 257 (1967) pro parte, quoad
spec. Forbes.

Icon: Moggi & Pisacchi in Webbia 22: 263, f. 8 (1967).

Shrub or tree 0-9-2-1 m tall, with branches spreading or subpendulous (always ?),
Stems red when young, soon terete; internodes 4-11 mm long, shorter than leaves;
bark grey. Leaves sessile; lamina 17-40 x 4-12 mm, + narrowly elliptic or rarely
oblanceolate, acute, margin plane, base narrowly cuneate, not or scarcely reflexed-
auriculate, concolorous, + glaucous, subcoriaceous, lower ones soon deciduous;
venation: (3)5(7) basal or lower lateral unbranched veins (i.e. midrib + 1-3 pairs),
the midrib branched pinnately above, with numerous secondary veins and some-
times a lax tertiary reticulum; laminar glands in secondary venous system, streaks
and lines; marginal glands pale, dense. Inflorescence 1-3(-10)-flowered, from 1-2
nodes, if from 2 nodes then internode not condensed; pedicels 2-6 mm long; bracts
intermediate. Flowers 40-75 mm in diam., stellate; buds ovoid-conic, acute to
subacute. Sepals 6-8 x 3-2-5-5 mm, imbricate, unequal, ovate to oblong, acute to
broadly obtuse, margin entire to distally eglandular-denticulate, midrib obscure;
laminar glands linear, numerous; submarginal and marginal glands absent. Petals
golden yellow, unmarked or (?) flushed orange outside, deciduous, spreading, 20-35
x 10-20 mm, broadly to narrowly obovate, with apiculus broad, rounded; margin
entire, eglandular; laminar glands linear, interrupted distally. Stamen fascicles
tardily deciduous, each with 45—70(—80) stamens, longest 8-17 mm long, c. 0:3 x
petals, with filaments united for c. 0-5 mm. Ovary 5-6-5 x 3-54 mm, ovoid; styles
5-8 mm long, 1-1-2 x ovary, 0-5—0-9 coherent (at least in flower), erect or spreading
distally; stigmas capitate or subcapitate. Capsule 13-15 — 7-9 mm, ovoid to narrowly
ovoid-pyramidal. Seeds yellowish-brown, c. 1-5 mm long, narrowly cylindric, not
winged, slightly carinate, finely linear-foveolate. 2n = ?

On granite slopes and rock crevices; 450-900 m.

Socotra, central. Map 4.
SOCOTRA. Haggiher Mts, Adho Dhemalu, 810 m, 19.iv.1967, Smith & Lavra-
nos 435 (K); Reiged, 450 m, 4.iv.1953, Popov SO/359 (BM, EA).

Balfour (1888) included these Socotran plants in H. mysurense Wight & Arnott, a
species from Sri Lanka and south India and the most primitive member of sect. 3.
Ascyreia. The relationship between H. balfourii and H. mysurense is close, and the
former is abnormal in sect. Campylosporus in having deciduous petals and stamens.
Nevertheless, H. balfourii can always be distinguished by the shorter, + united
styles, the lateral (not subapical), rounded petal apiculus, the tardily deciduous
stamens and the absence of a condensed inflorescence internode, and usually by the
more obtuse sepals. The apiculus position and union of the styles clearly indicate that
H. balfourii should be included in sect. Campylosporus, not sect. Ascyreia, although
the deciduous petals and stamens indicate its intermediate position between these
sections.

6. Hypericum socotranum Good

in J. Bot., Lond. 65: 334, t. 582 f. 16 (1927); Robson in Kew Bull. 12: 444 (1958);
Moggi & Pisacchi in Webbia 22: 257, f. 6, map 4 (1967) pro parte, excl. spec.

B. Balfour 246 (BM!, holotype; E!, K!).

|

:
|

STUDIES IN THE GENUS HYPERICUM L.

193

Map 4 Sect. 1. Campylosporus: 5. H. balfourii @, 6a. H. socotranum subsp. socotranum A, 6b. H.
socotranum subsp. smithii @.

Shrub or tree 0-9-4-5 m tall, bushy, with branches + ascending. Stems red when
young, soon terete; internodes 440 mm long, shorter than to exceeding leaves; bark
grey. Leaves sessile; lamina 12-30 x 5—22 mm, elliptic to oblanceolate or subcircular,
acute to rounded, margin plane, base narrowly to broadly cuneate or subangustate,
not reflexed—auriculate, concolorous, + glaucous, subcoriaceous, lower ones soon
deciduous; venation: 1—3(5) basal or lower lateral veins (i.e. midrib + 1-2 pairs), the
laterals unbranched or occasionally dichotomising, or the outer ones pinnately
branched, the midrib pinnately branched, with vestigial secondary venation and
obscure tertiary reticulum; laminar glands in secondary venous system, streaks and
lines or dots; marginal glands pale, dense. Inflorescence 1-S-flowered, from one
node; pedicel 1-11 mm long; bracts foliar but smaller. Flowers 40-75 mm in diam.,
stellate; buds ovoid-conic to ovoid, subacute to obtuse. Sepals 4-5-6:5 x 2-5-3 mm,
imbricate, unequal, ovate-lanceolate to oblong, apiculate or obtuse to rounded,
margin entire, midrib obscure; laminar glands linear numerous; submarginal and
marginal glands absent. Petals golden yellow, not tinged ?, deciduous, spreading,
15-40 x 9-24 mm, c. 3-6 X sepals, broadly obovate, with apiculus broad, rounded;
margin entire, eglandular; laminar glands linear, interrupted distally. Stamen fasci-
cles tardily deciduous, each with 45-70 stamens, longest 7-12 mm long, 0-3-0.5 x
petals, with filaments united for c. 0-5 mm. Ovary 5-6-5 X 3-5-4 mm, ovoid to
ovoid-conic; styles S—8 mm long, 1-1-2 x ovary, 0-3-0-9 coherent, spreading distally;
stigmas capitate or subcapitate. Capsule 12-13 xX 6-7 mm, ovoid to narrowly
ovoid-pyramidal. Seeds yellowish-brown, c. 1-5 mm long, narrowly cylindric, not
winged, slightly carinate, finely linear-foveolate. 2n = ?

On limestone escarpments and rock crevices; 210-600 m.
Socotra, east and west. Map 4.

It has been very difficult to decide on the status of the three taxa comprising the

| aggregate H. socotranum, as there are good reasons for treating them all respectively

as species or as subspecies. In the end, the primitive taxon of granite areas in the
centre of Socotra has been given specific rank as H. balfourii, whereas the other two,

_ growing on limestone at either end of the island, have been treated as subspecies.

_ 6a. H. socotranum subsp. socotranum

H. lanceolatum sensu Balf. f. in Trans. R. Soc. Edinb. 31: 27 (1888).
Icon: Moggi & Pisacchi in Webbia 22: 259, f. 6 (1967).

194

wl LQ vos

NORMAN K. B. ROBSON

Stem internodes 4-16 mm long, shorter than leaves. Leaves 12-24 x 5-10 mm,
elliptic to oblanceolate, acute or apiculate to obtuse, with 1-3(5) + basal veins, and
gland dots visible on dorsal surface. Inflorescence 1-flowered; pedicel 1-6 mm long.
Sepals 5-6-5 x 2-5-3 mm, apiculate to obtuse. Petals 15-30 x 9-15 mm. Stamen
fascicles 8-12 mm long. Ovary ovoid-conic; styles 0-3-0-6 coherent. Capsule 12-13 x
6-7 mm.

Socotra, western (Qallansiyah region); 210 m. Map. 4.
SOCOTRA. Qallansiyah, Jebel Ma’ali, above Qaysuh, 210 m, 27.1ii.1967, Smith
& Lavranos 133 (BM, K).

6b. H. socotranum subsp. smithii N. Robson, subsp. nov.

H. socotrano Good affinis, sed pedicellis et plerumque caulis internodis longioribus,
foliis latioribus, haud dorsaliter glanduloso-punctata, floribus plerumque numer-
osioribus, stylis longiore connatis, differt. Type: Socotra. Hammaderoh, Majhah
escarpment, 510 m, Smith & Lavranos 272 (BM!, holotype; K!).

Icon: —.

Stem internodes 6-40 mm long, often longer than the leaves. Leaves 15-30 x 8-22
mm, broadly elliptic to subcircular, apiculate to rounded, with 5(7) + basal veins,
without gland dots visible on dorsal surface. Inflorescence 1-5-flowered; pedicel S—11
mm long. Sepals 4-5-6 x 2-5-3 mm, obtuse to rounded. Petals 18-40 x 10-24 mm.
Stamen fascicles 7-11 mm long. Ovary ovoid to ovoid-conic; styles 0-5—0-9 coherent.
Capsule 12 X 6 mm (immature ?).

Socotra, eastern (Hammerdoh); 390-600 m. Map 4.

SOCOTRA. Hammerdoh, Majhah escarpment, 390 m, 8.iv.1967, Smith & Lavra-
nos 319 (BM, K); on limestone plateau [Grunhin Plateau] near the upper reaches of
the Wadi Zirik, 600 m, 16.v.1967, Smith & Lavranos 729 (BM, K).

Smith & Lavranos 729 is described as ‘a tree about 15’ [4-5 m] high; much lopped for
camel fodder’.

7. Hypericum quartinianum A. Rich.

Tent. fl. abyss. 1: 97 et Icones: t. 21 (1847); Oliver, Fl. trop. Afr. 1: 156 (1868) pro
parte, excl. syn. Schimper.; Engler in Phys. Math. Abh. K. Akad. Wiss. Berlin
1891: 307 (1892); R. Keller in Engler & Prantl, Nat. Pflanzenfam. 3 (6): 209
(1893), 2nd ed. 21: 176 (1925); T. C. E. Fries in Notizbl. bot. Gart. Mus. Berlin 8:
565 (1923); Good in J. Bot., Lond. 65: 332, t.582 f.10 (1927); Dale, Indig. trees
Uganda: 157 (1952); Milne-Redh. in Fl. trop. E. Afr. Hyperic.: 3 (1953); Robson
in Kew Bull. 12: 444 (1958), in Fl. Zamb. 1: 380 (1961); Cufod. in Bull. Jard. bot.
Etat Brux. 29, suppl.: 589 (1959); Dale & Greenway, Kenya trees & shrubs: 237
(1961); Moggi & Pisacchi in Webbia 22: 252, f. 4, map 3 (1967); Bamps, Fl. Congo,
etc. Guttif.: 13 (1970), in Bull. Jard. bot. natn. Belg. 41: 443 (1971), in Distr. pl.
Afr. 3: map 76 (1971); Agnew, Upl. Kenya wild fis.: 186 (1974); Robson in Bamps,
Robson & Verdcourt, Fl. trop. E. Afr. Guttif.: 29 (1978). Type: Ethiopia, Choa,
Quartin-Dillon & Petit s.n. (P, holotype; K!).

H. affine Steudel ex Oliver, Fl. trop. Afr. 1: 156 (1868), in synon.

H. ulugurense Engler in Bot. Jb. 28: 434 (1900); R. Keller in Engler & Prantl, Nat.
Pflanzenfam. 2nd ed. 21: 176 (1925); Good in J. Bot. Lond. 65: 332, t. 582 f. 11
(1927); Dale, Indig. trees Uganda: 157 (1952). Type: Tanzania, Uluguru Mts,
Stuhlmann 9247 (Bt).

Icones: Moggi & Pisacchi in Webbia 22: 255, f. 5 (1967); Fig. 10.

Shrub or tree 0-5-6 m tall, loosely branched, with branches spreading or ascending.
Stems red when young, soon terete; internodes 10-26 mm long, shorter than leaves;
bark grey. Leaves sessile or very shortly petiolate (to 1 mm); lamina (16—)20-90 x
5-27 mm, ovate-lanceolate to oblong-elliptic or oblanceolate, acute to subacute
or subapiculate-obtuse, margin plane to recurved, base cuneate to cordate-

STUDIES IN THE GENUS HYPERICUM L. 195

Fig. 10 H. quartinianum: (a) habit; (b) leaf section; (c) flowering branch; (d) sepal; (e) petal; (f) stamen
fascicle; (g) anthers; (h) capsule (a, c x 1; d-f,h x 4; b x 8; g x 20). (a) Schlieben 4230; (b-e) Mendonca
720; (f-h) Greathead 92.

amplexicaul, sometimes subauriculate, paler beneath, not or scarcely glaucous,
chartaceous, lower ones eventually deciduous; venation: 5—7(9) basal and lower
lateral veins (i.e. midrib + 2-4 pairs), the laterals unbranched, the midrib pinnately
branched distally, usually linked by cross-veins, with numerous secondary veins
continuous or + interrupted by + dense tertiary reticulum; laminar glands in
secondary venous system, lines alternating with + irregular series of small dots and
short streaks; marginal glands dark, dense. Inflorescence (1)3-9-flowered, from 1(2)

196

NORMAN K. B. ROBSON

nodes, often with lateral flowering branches; pedicels 10-25 mm long; bracts
intermediate. Flowers 35-70 mm in diam., stellate; buds ovoid, acute. Sepals 5-15 x
2-5 mm, imbricate to basally open, unequal, reflexed in fruit, ovate-lanceolate to
triangular-lanceolate, acute, margin glandular-ciliate or -denticulate, midrib un-
differentiated; laminar glands linear, numerous; submarginal and marginal glands
dark. Petals golden yellow to lemon-yellow, not tinged, tardily deciduous, spreading
to reflexed, (17—)20-41 x 10-25 mm, 2-3 x sepals, obovate, with apiculus obtuse;
external margin with few sessile or intramarginal dark reddish glands, internal
margin dark-glandular-ciliate; laminar glands linear, not interrupted. Stamen fasci-
cles tardily deciduous, each with c. 30 stamens, longest 15—22(-25) mm long,
0-5-0-75 X petals, with filaments very shortly united. Ovary 5-8 x 3-5 mm, rather
broadly to narrowly ovoid; styles 8-13 mm long, 1:2-1-6 X ovary, completely or
almost completely coherent; stigma mass capitate or subcapitate. Capsule 10-16 x
5—10 mm, broadly to narrowly ovoid. Seeds yellow-brown, c. 1 mm long, cylindric,
not winged, slightly carinate, shallowly linear-reticulate. 2n = ?

In rocky places, gulleys, and river banks in upland grassland or deciduous woodland;
1150-3300 m.

Yemen, Ethiopia, Sudan Republic (south), Kenya (Mt. Elgon), Uganda (Mt. Elgon,
Karamoja), Tanzania (southern), Zaire (Shaba), Malawi (Misuku, Nyika, Vipya),
Zambia (Mbala, Serenje), Mozambique (Lago). Map 5.

YEMEN. Jebel Raymah above Hadia, on NW. of Jebel Thallamlam, near
Na’ama, c. 1800 m, i.vi.1979, Wood 2817 (BM).

SUDAN REPUBLIC. Equatoria [El Istwaya]: Torit, near Moimoi Dongotonas,
1800 m, 21.1.1950, Jackson 1096 (BM).

ETHIOPIA. Tigray: Djenda, Demben, 30.iii1.1872, Steudner 1439 (K). Gonder:
near Gondar 3000 m, 25.5.1958, Greathead 92 (BM, EA, K). Gojam: Chamoga
valley, S.S.E. of Debre Marcos, 2500 m, 31.1.1960, Kuls 144 (FR). Shoa: Chellamo
forest, 72 km W. of Addis Ababa, 2700 m, 8.xii.1953, Mooney 5087 (K). Wellega:
Hupé, 1950 m, 22.11.1957, Mooney 6785 (K.). Keffa: 8 km S. of Jimma on Bonga
road, 1760 m, 23.xii.1961, Meyer 7794 (K). Gamo Gofa: Uva range, 19 km E. of
Bulki, W. slope, 1500 m ?, 13.iv.1949, Thesiger 1809 (BM). Sidamo: hill to SW. of
Agheremariam (also to E.), 2070 m, 3.xii.1952, Gillett 14601 (K). Arssi: W. shore of
L. Zouai, 1.1899, Welby (K). Harerghe: between Herna and Asbe Tefari, + 2100 m,
22.xii.1962, Burger 2455 (K).

KENYA. Rift Valley: Trans-Nzoia, SW. Elgon, 2010 m, 11.ii1.1958, Symes 301
(K).

UGANDA. Northern: Karamoja, Napak, 2100 m, 11.1938, Sangster 418 (K).
Eastern: Mt. Elgon, Sipi, 1950 m, ix.1934, Synge 1078 (BM).

ZAIRE. Shaba: Kasenia, falls of the Mulamba, ix.1962, Schmitz 7838 (K).

TANZANIA. Western: Ufipa, above Milepa, L. Rukwa north, 1800-2100 m,
3.x.1936, Michelmore 742 (K). Eastern: Morogoro, Bundiki, c. 1350 m, 17.viii.1951,
Greenway & Eggeling 8596 (K). Southern Highlands; Mbeya, foot of Wenzel
Hekmann Crater [L. Ngozi], 1650 m, 28.viii.1936, Burtt 6233 (BM, K). Southern:
Songea, Matengo Hills, c. 8km N. of Miyau, 1570 m, 21.v. 1956, Milne-Redhead &
Taylor 10406 (K).

MALAWI. Northern: Chitipa, Misuku Hills, Walindi Forest, 2000 m, 12.xi.1958,
Robson 583 (K); Vipya Plateau, near Mzuzu, 1200 m, 25.ix.1966, Pawek 109
(MAL).

ZAMBIA. Northern: Serenje, Kundalila Falls, 13.x.1963, Robinson 5697 (K).

MOZAMBIQUE. Niassa: Lago, Maniamba, rio Lualeze, 1200 m, 10.x.1942,
Mendonca 720 (BM, LISC).

Plants of H. quartinianum from the northern part of its range (Yemen, Ethiopia,
Kenya, Uganda) tend to have smaller, narrower, cuneate-based leaves with poorly
developed reticulate venation, and small flowers, whereas those from the southern
part (Tanzania, Mozambique, Malawi, Zambia, Zaire) tend to have larger, broader,
cordate-based leaves with relatively well-developed reticulate venation, and larger
flowers. The correlation of these various trends is incomplete, however, so that the
southern form cannot be distinguished as H. ulugurense.

STUDIES IN THE GENUS HYPERICUM L.

Map5 Sect. 1. Campylosporus: 7. H. quartinianum @, 8. H. synstylum A.

8. Hypericum synstylum N. Robson

in Kew Bull. 12: 433 (1958); Moggi & Pisacchi in Webbia 22: 259, f. 7, map 4 (1967).
Type: Somalia, Gan Libah, Golis Range, Hrialajat, 9.vii.1945, Glover & Gilliland
1168 (K! holotype; BM!, EA).

197

198

NORMAN K. B. ROBSON
Icon: Moggi & Pisacchi in Webbia 22: 261, f. 7 (1967).

Shrub 1-5-3 m tall, diffusely branched, with branches spreading. Stems orange when
young, soon terete; internodes 6-11 mm long, shorter than leaves; bark reddish-
brown. Leaves with petiole 1-2 mm long; lamina 10-30 x 5—12 mm, oblong or
elliptic-oblong to ovate, apiculate-obtuse to rounded, margin plane, subincrassate,
base broadly cuneate to rounded, not auriculate, paler beneath, scarcely glaucous,
chartaceous, lower ones soon deciduous; venation: 7—9 basal and lower lateral veins
(i.e. midrib + 34 pairs), the laterals unbranched, linked by cross-veins, with
secondaries reduced to short streaks in areolae of lax tertiary reticulum; laminar
glands in secondary venous system, short streaks, without distinct alternating series
of dots; marginal glands reddish or pale, irregular. Inflorescence 1—2-flowered;
pedicel up to 2-5 mm long; bracts intermediate, acute. Flowers 25-33 mm in diam.,
stellate; buds ovoid, acute. Sepals 5—7 x 1-7-2-2 mm, scarcely imbricate to basally
open, equal, linear-lanceolate, acute, margin entire, midrib obscure; laminar glands
linear to punctiform, numerous; submarginal glands absent; marginal glands pale or
absent. Petals golden yellow, not tinged, tardily deciduous, spreading, 13-16 x 8-10
mm, c. 2-5 X sepals, obovate, with apiculus very small or absent; margin entire,
eglandular; laminar glands linear, not interrupted. Stamen fascicles persistent, each
with 15-20 stamens, longest 7-10 mm long, 0:5—0-75 x petals, with filaments united
for c. 0-5 mm. Ovary 5-6-5 X 2-3 mm, narrowly ovoid; styles 6-7 mm long, 1-1—1-2 x
ovary, completely coherent; stigma mass narrow, not capitate. Capsule 8-10 x 5-6
mm, narrowly ovoid. Seeds reddish-brown, 1-2-1-5 mm long, cylindric, not winged
or carinate, shallowly linear-reticulate. 2n = ?

In limestone rocks on black soil, c. 2100 m.

Ethiopia (Harerghe [Harar]), Somalia (North). Map S.

ETHIOPIA. Harerghe: Mt. Achim, 2100 m, 14.x.1954, Bally 10057 (K); road to
Combulcha and Harar, W. of Ejersa-Goro (9°28'N, 42°12’E), + 2100 m, 2.xi.1963,
Burger 3320 (K).

SOMALIA. North: Golis Range, Ga’an Libah Forest Reserve, 2.11.1973, Bally &
Melville 16232 (EA), 16233 (K).

H. synstylum 1s clearly intermediate in many respects between H. quartinianum and
the primitive members of the N. American sect. 20. Myriandra. Thus, where it
differs from H. quartinianum (e.g. in the absence of dark glands and the entire sepals
and petals), it approaches H. frondosum. It is apparently confined to two disjunct
areas c. 300 km apart.

9. Hypericum roeperianum W. G. Schimper ex A. Rich.

Tent. fl. Abyss. 1: 96 (1847); R. Keller in Engler & Prantl, Nat. Pflanzenfam. 3 (6): 20
(1893), 2nd ed. 21: 176 (1925); Good in J. Bot., Lond. 65: 331, t. 582 f. 6 (1927);
Exell & Mendonga, Consp. fl. Angol. 1: 119 (1937), 369 (1951); Bredell in Bothalia
3: 582, t. 1 f. 11, map (1939); Brenan, Checkl. Tang. Terr. 2: 250 (1949);
Milne-Redh., Fl. trop. E. Afr. Hyperic.: 3 (1951); Dale, Indig. trees Uganda: 159
(1952); Robson in Kew Bull. 12: 444 (1958), in Fl. Zamb. 1: 380 (1961); Cufod. in
Bull. Jard. bot. Etat Brux. 29,suppl.: 589 (1959); Dale & Greenway, Kenya trees &
shrubs: 237 (1961); Spirlet, Contr. fl. Congo Guttif.: 8 (1966); Moggi & Pisacchi in
Webbia 22: 244, f. 2, map (1967) pro parte, excl. subsp. gnidiifolium; Bamps, Fl.
Congo, etc. Guttif.: 12 (1970), in Bull. Jard. bot. natn. Belg. 41: 441 (1971), in
Distr. pl. Afr. 3: map 75 (1971); Agnew, Upl. Kenya wild fls.: 186 (1974); Killick &.
Robson in Fl. Sthn Africa 22: 16 (1976); Robson in Bamps, Robson & Verdcourt,
Fl. trop. E. Afr. Guttif.: 29 (1978). Types: Ethiopia, [Gonder] prope Dschenausa
ad montem Aber, 4.i.1840, Schimper 866 (P, lectotype — Moggi & Pisacchi (1967);
FI; G; K! LE; MO) (the K specimen is annotated ‘prov. Semien Jul. 1843’ in ms.);
[Tigray] ad montem Kubbi prope Adoam, i., Schimper (n.v.).

H. schimperi Hochst. [in Flora, Jena 24 (1): 29 (1841)] ex A. Rich., Tent. fl. Abyss. 1:
97 (1847); Oliver, Fl. trop. Afr. 1: 150 (1868); Engler in Phys. Math. Abh. K.
Akad. Wiss. Berlin 1891: 307 (1892); R. Keller in Engler & Prantl, Nat. Pflanzen-

STUDIES IN THE GENUS HYPERICUM L.

fam. 2nd ed. 21: 176, f. 73V, W (1925); Good in J. Bot., Lond. 65: 331, t.582 f. 7
(1927); Dale, Trees & shrubs Kenya: 36 (1936); Cufod. in Bull. Jard. bot. Etat
Brux. 29, suppl.: 589 (1959). Types: Ethiopia, [Tigray] Ad rupes . . . montis
Kubbi proxime ad ecclesiam ‘Arba tensa’ (in regione Memsach), 29.xii.1837,
Schimper 132 (P, lectotype;* G; K!; L; LE; M; PAL; W!); Abyssinia, 1842,
Schimper 1160 (FI; G; K!; L; LE; M; P; UPS; W!); [Gonder] in provincia
Ouodgerate in sylvis Goumasso, Q.-Dillon & Petit (K!; P).

H. quartinianum sensu Oliver, Fl. trop. Afr. 1: 156 (1868) pro parte, quoad spec.
Schimper.; De Wild. in Contr. fl. Katanga: 132 (1921); T. C. E. Fries in Notizbl.
Bot. Gart. Mus. Berlin 8: 565 (1923) pro parte, excl. typum.

H. schimperi var. y sensu Oliver, Fl. trop. Afr. 1: 157 (1868).

H. quartinianum var. roeperianum (W. G. Schimper ex A. Rich.) Engler in Phys.
Math. Abh. K. Akad. Wiss. Berlin 1891: 307 (1892); Pichi-Sermolli, Miss. Stud.
Lago Tana (Bot.) 1: 97 (1951).

H. schimperi var. angustisepalum Engler in Phys. Math. Abh. K. Akad. Wiss. Berlin
1891: 307 (1892). Type: Angola, Huilla, Welwitsch (B+, holotype) (same collec-
tion as type of var. Auillense).

H. schimperi var. huillense Hiern, Cat. Afr. pl. Welw. 1: 56 (1896). Type. Angola,
Huilla, Morro de Lopolo, 23.v.1860, Welwitsch 1055 (BM!, holotype; B+; COT;
K!; LISU).

H. riparium A. Chev. in Bull. Soc. Fr. 54 (8): 8 (1907), Expl. Bot. Afr. Occ. Franc. 1:
49 (1920); Good in J. Bot., Lond. 65: 331 (1927); Keay & Milne-Redh. in Fl. W.
trop. Afr. 2nd ed. 1: 287 (1954); Pellegrin in Bull. Soc. bot. Fr. 106: 217 (1959).
Type: guinée, Diagiussa, iv.1905, Chevalier 13460 (P, holotype; K!).

H. conrauanum Engler in Bot. Jb. 40: 555 (1908); Good in J. Bot., Lond. 65: 332, t.
582 f. 9 (1927). Type: Cameroon, Baberong, 1470 m, i.iv.1898, Conrau 28 (Bt,
holotype; BM!; E!).

H. roeperanum var. schimperi (Hochst. ex A. Rich.) Moggi & Pisacchi in Webbia 22:
249 (1967).

Icon: Moggi & Pisacchi in Webbia 22: 248, f. 2 and 251, f. 3 (1967).

Shrub or tree 0-6-5 m tall, bushy, with branches ascending. Stems red when young,
soon 2-lined to terete; internodes 10-30 mm long, shorter than leaves; bark reddish-
brown. Leaves sessile; lamina (25—)30-100(-115) x (S—)10-40 mm, lanceolate or
occasionally ovate to elliptic or oblong-elliptic, acute to obtuse, margin recurved to
revolute, base cuneate, sometimes subauriculate, paler beneath, sometimes
glaucous, chartaceous to subcoriaceous, lower ones + persistent; venation: 7(9)
basal and lower lateral veins (i.e. midrib + 3(4) pairs), the laterals and midrib
branches forming dense tertiary reticulum, with secondaries reduced to dots in
areolae; laminar glands in secondary venous system, dots (sometimes elongate),
without distinct alternating series; marginal glands dark, dense. Inflorescence (1)3-
o-flowered, from 1—2(-4) nodes, often with lateral flowering branches; pedicel 5-20
mm long; bracts foliar or intermediate. Flowers 30-55 mm in diam., stellate; buds
ovoid, acute. Sepals (2:-5—)5-8 x (1-5—)3-5-5 mm, imbricate, unequal or unequal,
ovate-lanceolate to ovate, acute to obtuse, margin subentire to glandular-ciliate or
-denticulate, midrib undifferentiated; laminar glands linear, numerous; submarginal
and marginal glands dark. Petals golden or bright yellow, sometimes tinged reddish,
persistent, spreading to reflexed, (12—)20-35 x (7-)10-20 mm, 3-4 sepals, obovate,
with apiculus small, rounded; margin entire, eglandular, or with submarginal to
marginal dark glands or dark-glandular-ciliate; laminar glands linear, not inter-
rupted, occasionally also a few dark dots or short streaks. Stamen fascicles persistent,
each with c. 35—45 stamens, longest 10-20 mm long, 0-5-0-65 x petals, with filaments
very shortly united. Ovary 4-7 X 3-5-5 mm, ovoid to subglobose; styles 8-12 mm
long, 1-6-2-3(-3) X ovary, completely coherent or rarely almost so; stigma mass
capitate. Capsule 9-14 x 7-10 mm, broadly to narrowly ovoid. Seeds yellow-brown,

* Moggi & Pisacchi chose the Petit specimen as lectotype; but their choice is unacceptable
because Richard validated Hochstetter’s name, thereby indicating one of Schimper’s speci-
mens as the lectotype. As Hochstetter cited only Schimper 132 (the one specimen collected
by 1841), I select this as the new lectotype.

199

200

NORMAN K. B. ROBSON

c. 1 mm long, cylindric to narrowly pyriform, not winged or carinate, shallowly
linear-reticulate. 2n = ?

Dry evergreen forest margins, bushland or grassland, or beside rivers or streams;
840-3300 m.

From Ethiopia to the Transvaal, Zambia, and Angola; also in Guinée, Nigeria, and
Cameroon. Map 6.

GUINEE. [Fouta Djalon] Diagiussa, cascade du Ditinn, 840 m, iv.1905, Chevalier
13460 (K).

NIGERIA. North: Plateau Prov., Jos Distr., Sha, 1350 mi.1960, Wimbush 65 (K).
East: Ogoja Prov., Obudu Div., Ikwette, 1560 m, 29.xii.1948, Savory & Keay FHI
25219 (K); Vogel Peak massif, 1340 m, 13.xii.1957, Hepper 1595 (K).

CAMEROON. West: above Bamenda, 2400 m, 18.1.1928, Migeod 333 (BM, K); 6

\
te ieee
-/ ’ ‘
/ ee ! ‘
oF \ ‘

t
|
I
i
(
(
(
(
\
\
i
\
!
\

/
1
t

Map6_ Sect. 1. Campylosporus: 9. H. roeperianum ®@, 10. H. gnidiifolium A.

STUDIES IN THE GENUS HYPERICUM L.

km N.N.E. Nkambe, prés village Gépka (Ntumbe), 1600 m, xi.1974, Satabie 80 (K).
East: Bamboutos Mts, Djuttitsa, 20.v.1966, Meurillon C.N.A.D. 355 (K).

ETHIOPIA. Tigray: Addi Rasa bei Axum, 2021 m, 5.i.1863, Schimper 782 (BM).
Gonder: Monte Jesus Tabor, 22.iii.1937, Pichi Sermolli 186 (K). Welo: Lalibela,
2840 m, 7.1.1924, Harlan (US n.v.). Gojam: Choké Mts, W. side of upper Godeb
gorge, c. 2700 m, 30.viii.1957, Flenley & Hiller 266a (K). Shoa: Blue Nile road c. 100
kmN. of Addis Ababa, between Fiché and Debre Libanos, + 2000 m, 25.iv.1966, de
Wilde 10875 (K).

KENYA. Northern Frontier: Mt. Nyiru above Tum, 1800-2100 m, vii.1960,
Kerfoot 1934 (K). Turkana: Turkana distr., Karasuk, S.E. ridge of Kachagalau Hill,
2650 m, 22.iv.1959, Osmaston 4457 (K). Rift Valley: Laikipia distr., above Thom-
son’s Falls, 2325 m, Verdcourt 3814 (K). Central: North Nyeri distr., N. of Nanyuki,
8.vii.1938, Pole Evans & Erens 1214 (E, K). Masai: Narok, Usso Nyiro, 1800 m,
23.viii.1960, Paulo 704 (K).

UGANDA. Northern: Karamoja distr., Napak Mts, 2250 m, vi.1950, Eggeling
5912 (K). Eastern: Mbale distr., Mt. Elgon, 2400 m, 1.1918, Dummer 3602 (K).

ZAIRE. Shaba: 42 km W. of Kabiashia, 1090 m, 31.vili.1966, Malaisse 4561 (K).

TANZANIA. Northern: Masai distr., Ngorongoro Conserv. Area, Endulen to
Ngorongoro, S. slopes of Mt. Satiman, 2350-2730 m, 18.ix.1977, Raynal 19131 (BM,
P). Tanga: Lushoto distr., Mlalo, Kilangwi forest, vili.1955, Semsei 2286 (K).
Western: Mpanda distr., Mahali Mts, Kasieha, 960 m, 26.ix.1958, Newbould &
Jefford 2698 (K). Southern Highlands: Rungwe distr., Kyimbila, 1350 m, 1909, Stolz
27A(LU).

ZAMBIA. Western: Mwinilunga distr., R. Matonchi below dam, 21.x.1937,
Milne-Redhead 2880 (K, PRE).

ANGOLA. Benguela: R. Guito near Quipeio, c. 1500 m, 2.v.1937, Exell &
Mendonca 1883 (BM). Huila: Sa da Bandeira, Serra de Chella, 10.vii.1967, Henri-
ques & Brites 1116 (BM).

ZIMBABWE. Central: Makoni, 1440 m, vi.1917, Eyles 795 (BM, K, SRGH).
Eastern: Umtali distr., Stapleford Forest Reserve, heights above Inyamakarara
Valley, 1680 m, 28.xi.1955, Chase 5890 (BM, PRE, SRGH).

MOZAMBIQUE. Manica‘e Sofala: Manica, Macequece, caseata de Penha
Longa, 6.vi.1948, Mendonca 4524 (BM, LISC).

SOUTH AFRICA. Transvaal: Pilgrim’s Rest distr., Mariepskop area, c. 900 m,
18.xi.1958, Killick & Strey 2424 (K).

H. roeperianum is closely related to H. quartinianum; it has leaf-venation, infloresc-
ence and other characters that are similar but more specialised. Of its 7-8 disjunct
areas, two (Guinée and Nigeria/Cameroun) have plants with relatively broad leaves;
but the variation is not disjunct between these and the eastern population, so that H.
riparium cannot be recognised as distinct.

In Ethiopia, Uganda, and Kenya there are two forms, which intergrade mor-
phologically and are not geographically or ecologically distinct. One (H. roeper-
ianum sensu stricto) has leaves elliptic to lanceolate, acute, slightly discolorous, the
venation densely reticulate, not prominent, with small punctiform areolar glands,
and sepals ovate-lanceolate; the other (H. schimperi) has leaves oblong or oblong-
lanceolate, acute to obtuse, markedly discolorous, with laxer venation prominent
beneath and larger slightly elongate areolar glands, and sepals ovate. These forms
have been recognised as varieties by Moggi & Pisacchi, but the intermediate forms
(e.g. Scott Elliot 6569 (BM, K), from Kenya, Gilgit) precluse the recognition of H.
schimperi even as a variety.

10. Hypericum gnidiifolium A. Rich.

Tent. pl. Abyss. 1: 98 (1847) (‘gnidiaefolium’); Oliver, Fl. trop. Afr. 1: 157 (1868);
Engler in Phys. Math. Abh. K. Akad. Wiss. Berlin 1891: 307 (1892); R. Keller in
Engler & Prantl, Nat. Pflanzenfam. 3 (6): 209 (1893), op. cit. 2nd ed. 21: 176
(1925); Good in J. Bot., Lond. 65: 332, t. 582 f. 8 (1927); Robson in Kew Bull. 12:
444 (1958), 33: 582, ff. IC5, 2J, K (1979); Cufod. in Bull. Jard. bot. Etat Brux. 29,
suppl.: 599 (1959). Types: Ethiopia, [Tigray] ‘Ouadgerate, altit. 8500 and 9500

201

202 NORMAN K. B. ROBSON

pedum ad rivulos,’ Quartin-Dillon & Petit s.n. (P, lectotype; K!). Moggi &
Pisacchi (1967) point out that, of the five Q.-Dillon & Petit specimens in Paris (P),
three are labelled ‘Ouadgerate’ or ‘Ouedgerate’. Of these, they selected that cited
above as lectotype.

H. roeperanum subsp. gnidiifolium (A. Rich.) Moggi & Pisacchi in Webbia 22: 250
f. 4 (1967).

Icon: Moggi & Pisacchi in Webbia 22: 253, f. 4 (1967).

Shrub (?) or tree to 4-5 m tall, much branched, with branches ascending. Stems
yellow-brown when young, soon 2-lined to terete; internodes 6-10 mm long, shorter
than leaves; bark reddish-brown. Leaves sessile; lamina 17-28 x 4-10 mm, narrowly
elliptic to oblong-elliptic, acute to subapiculate, margin plane, subincrassate, base
cuneate, not or scarcely auriculate, paler beneath, not glaucous, chartaceous, lower
ones gradually deciduous; venation: 3—-5(7) basal or lower lateral veins (i.e. midrib +
1—2(3) pairs), the laterals and midrib branches forming rather dense tertiary reticu-
lum, with secondaries reduced to short streaks in areolae; laminar glands in
secondary venous system, short streaks, without alternating series of dots; marginal
glands pale and dark, dense. Inflorescence (1)3—S-flowered, from 1—2 nodes, without
lateral flowering branches; pedicels 6-15 mm long; bracts foliar or intermediate.
Flowers c. 35 mm in diam., stellate; buds ovoid, obtuse. Sepals 6-7 x 3-5-5 mm,
imbricate, subequal, broadly elliptic or ovate-elliptic to oblong, acute to obtuse,
margin entire to glandular-ciliolate, midrib undifferentiated; laminar glands linear,
numerous, submarginal glands dark or absent; marginal glands absent or reddish.
Petals golden yellow, tinged red, persistent, spreading to reflexed, 16-20(—25) x 8-9
mm, 2.5-3 x sepals, obovate-oblanceolate, with apiculus almost absent, rounded;
margin internally dark-glandular-ciliate; laminar glands linear, not interrupted.
Stamen fascicles ‘subpersistent’, each with c. 20 stamens, longest 8-14 mm long,
0-5—0-65 X petals, with filaments united for c. 0-5 mm. Ovary 5-6 X 3-5-4 mm,
broadly ovoid; styles 11-13 mm long, c. 2-2 x ovary, c. 0-7 coherent; stigmas small,
rounded. Capsule 7-8 X 6 mm, broadly ovoid-pyramidal. Seeds unknown. 2n = ?

Banks of streams: 2700-2900 m.

Ethiopia (Tigray, Shoa). Map 6.
ETHIOPIA. Tigray: Wogera, Maye-Borhha plateau, 2500-2700 m, Quartin-
Dillon & Petit (K). Shoa: Shoa, Quartin-Dillon & Petit (K).

Like H. roeperianum, H. gnidiifolium has leaves with densely reticulate venation
and a 1—several-flowered inflorescence, and occurs within the same area in Ethiopia.
As the leaves and flowers are relatively small, it thus appears at first glance to be a
reduced form of H. roeperianum. Nevertheless, it has some characters that seem to
be more primitive (elongate laminar leaf-glands, mixed pale and dark marginal
leaf-glands, marginal sepal-glands absent or reddish, styles only partially united), all
of which suggest that H. gnidiifolium is specifically distinct from H. roeperianum. I
have seen no recent collections of this species.

“)

es eee

Sect. 2. PSOROPHYTUM (Spach) Nyman
Consp. Fl. Eur.: 132 (1878).

Shrub up to 1-5 m tall, evergreen, glabrous, without dark glands; branching lateral or rarely
pseudo-dichotomous. Stems 4-lined, not compressed when young or becoming terete, verru-
cose-glandular; cortex exfoliating in short flakes; bark scaly. Leaves opposite, decussate, sessile,
free, deciduous at basal articulation; lamina entire, with venation pinnate, closed, the tertiary
obscure; laminar glands punctiform, pale; marginal gland dots pale; ventral glands absent.
Inflorescence 1-flowered; bracts foliar; bracteoles sepaline, involucral. Flowers stellate, homo-
stylous. Sepals 5, free, persistent, spreading to reflexed in fruit, with margin entire to denticulate;
veins numerous; laminar glands pale, linear, interrupted distally, sometimes also punctiform;
submarginal and marginal glands pale. Petals 5, deciduous, without or with very shallow lateral
apiculus, margin entire; marginal glands absent; laminar glands numerous, pale, linear,

STUDIES IN THE GENUS HYPERICUM L. 203

sometimes interrupted distally. Stamen fascicles 5, free, deciduous, each with 18-25 stamens;
filaments united very shortly; anther gland amber. Ovary with 5(4) incompletely axile placentae
(united at the base, free but deeply intrusive above), each -ovulate; styles 5(4), free; stigma
small. Capsule 5(4)-valved, coriaceous, longitudinally vittate. Seeds cylindric, not carinate,
without apical wing expansion; testa linear-reticulate.

BASIC CHROMOSOME NUMBER (x): 12; ploidy 2.
Hasirat: open dry woods and rocky places, 30-1200 m.
DistTRIBUTION: Balearic Is.

1 species.

1(11). Hypericum balearicum L.

Sp. pl.: 783 (1753); Curtis, Bot. Mag. 4: t. 137 (1791); Choisy, Prodr. monogr. fam.
Hypéric.: 43 (1821), in DC., Prodr. syst. nat. regni veg. 1: 546 (1824); Treviranus,
Hyper. animad.: 8 (1861); Marés & Vigineix, Cat. rais. pl. vasc. Iles Baléares: 61
(1880); R. Keller in Engler & Prantl, Nat. Pflanzenfam. 3 (6): 211 (1893), 2nd ed.
21: 177 (1925); Fiori & Paoletti, Fl. anal. d'Italia 1: 384 (1898), Icon. fi. ital.: t. 1248
(1898); Knoche, Fl. Balearica 2: 184, maps (1922); Fiori & Paoletti, Nuovo fl.
d Italia 1: 520 (1924); Stefanoff in God. Agr.-les. Fak. Univ. Sofiya 11: 143 (1933),
in Pflanzenareale 4: Karte 2a (1933); N. Robson in Fl. europaea 2: 263 (1968);
Bean, Trees & shrubs hardy in Br. Isles 8th ed. 2: 408 (1973); Barcel6, Fl. Mallorca
3: 173, f. 987 (1979). Types: Balearic Is., Majorca, Herb. Linn. 943/1 (LINN!,
lectotype-mihi); also citations from Hort. Cliff., van Royen, Magnol and Clusius.
The Herb. Linn. specimen is chosen in preference to that in Herb. Cliff. on
account of the new diagnosis provided in Sp. pl.: Hypericum floribus pentagynis,
caule fruticoso, foliis ramisque cicatrisatis.

H. verrucosum Salisb., Prod. stirp. horto Chapel Allerton: 369 (1798), nom. superfl.
Type as for H. balearicum L.

Psorophytum undulatum Spach, Hist. nat. veg. Phan. 5: 413 (1836), in Annis Sci. nat.
(Bot.) II, 5: 360 (1836), nom. superfl. Type as for Hypericum balearicum L.

H. balearicum var. ochroleucum R. A. W. Herm in Knoche, Fl. Balearica 2: 184
(1922), nomen.

Psorophytum balearicum (L.) Y. Kimura in Nakai & Honda, Nova fl. jap. 10: 22
(1951).

Icones: Barcel6, Fl. Mallorca 3: 176-177, f. 987 (coloured and b. & w. photos)
(1979); Fig. 11.

Shrub or small tree 0-6-2 m tall, usually forming rounded bushes, with branches erect
to ascending, sometimes branching pseudo-dichotomously. Stems yellow-green
when young, glandular-verrucose, especially along the (4) stem lines, terete after
first season; internodes 2-10 mm long, shorter than leaves; bark reddish-brown.
Leaves sessile; lamina (6-)8-15 x 3-7 mm, ovate to oblong-elliptic or narrowly
oblong, rounded, margin + undulate, base broadly cuneate to rounded, paler
beneath, not glaucous, coriaceous, lower ones gradually deciduous; venation: 1
basal vein (midrib), 1 pair of main laterals (intramarginal vein) and c. 5 subsidiary
laterals (cross veins) joining intramarginal vein, with or apparently without obscure
tertiary reticulum; laminar glands large, vesicular, prominent, amber, often inters-
persed with + numerous small pale dots (beneath); marginal glands dense. Infloresc-
ence 1-flowered; pedicel 3-11 mm long; bracts sepaline, involucral. Flowers 15—40
mm in diam., stellate; buds ovoid-subglobose, obtuse to rounded. Sepals 5-7 x 2-5
mm, imbricate, + unequal, erect in bud, + reflexed in fruit, rounded, margin entire
to apically denticulate, midrib undifferentiated or obscure; laminar glands linear,
interrupted distally, numerous, sometimes also 1-4 large amber vesicular; marginal
glands absent. Petals golden yellow or very rarely pale yellow, slightly red-tinged
outside, deciduous, spreading, 10-20 x 4-9 mm, 2-3 X sepals, narrowly obovate to
oblanceolate, with apiculus very small, rounded or absent; margin entire; laminar
glands linear, sometimes interrupted distally. Stamen fascicles deciduous, each with

Ltt BD

NORMAN K. B. ROBSON

204

EO SB
LZ;

PA 0 yo
D7 Kes
rel \é

Stax
fe

a Yi

TOLER
Cosy iar

SS

Ale

—,

iy

fp

fascicle; (h) anthers; (i) capsule (a X 1;c X 2;d,i x 4; f,g, X 6;e X 8;b x 12;h X 20). (a,b, d,i)J. & M.

Fig.11 H. balearicum: (a) habit; (b) leaf section; (c) flower bud; (d) flower; (e) sepal; (f) petal; (g) stamen
Cannon 3804; (c) Maude s.n.; (e-h) Auquier et al. 7614.

STUDIES IN THE GENUS HYPERICUM L. 205

18-25 stamens, longest 8-18 mm long, c. 0-8 x petals, with filaments very shortly
united. Ovary 3-S x 2-3 mm, narrowly ovoid-conic; styles 6-13 mm long, 2-2-5 x
ovary, suberect or slightly outcurved distally; stigmas small. Capsule 8-12 x S-7 mm,
narrowly ovoid to ovoid-pyramidal. Seeds reddish-brown, c. 1-7 mm long, cylindric,
not carinate or winged, finely linear-reticulate. 2n = 24.

Dry woods and calcareous rocky or stony places; 30-1200 m.

Balearic Islands (Mallorca, Menorca, Ibiza, Cabrera, Dragonera). Map 7.

MALLORCA. Soller distr., Soller side of pass to Lluch, 2.vi.1954, J. & M.
Cannon 2648 (BM); Pollensa distr., Cala San Vicente, track to western headland,
10.iv.1971, J. & M. Cannon 3804 (BM); Palma distr., Genova, prés Palma,
19.iv.1919, Sauvan-Denis in Sennen 3668 (BM, K).

MENORCA. Alayor, Son Blanc, 20.vi.1876, Rodriguez (K).

IBIZA. Cala Talamanca, headland of Grosa y Vatarina between Ibiza harbour
and Bay of Talamanca, 11.iv.1968, Bowden & Sims 1931 (BM).

CABRERA. Fide Knoche, Fl. balear. 2: 186 (1922), 3: 322 (1923).

DRAGONERA. Fide Knoche, Fil. balear. 2: 186 (1922), 3: 321 (1923).

CULTIVATED. Specimens seen from England (1743-1983), France, and Ger-
many (1875).

The glandular-verrucose stems and leaves and the involucre-like bracteoles disting-
uish H. balearicum from all other species of Hypericum. It has such an isolated
position in the genus that its affinities are not immediately obvious. The leaf-
venation and floral characters, however, indicate that H. revolutum subsp. revolu-
tum is its closest relative, in particular the broader-leaved form of that species that
occurs in Ethiopia.

H. balearicum is apparently confined to the Balearic Islands, where it is quite
common on Mallorca but rare on the others on which it occurs. The Italian locality
cited by some authors (Liguria, Appenines above Savona) is the result of introduc-
tion (see Knoche, loc. cit.).

|

ee
| Map7 Sect. 2. Psorophytum: 1. H. balearicum @.

|

206 NORMAN K. B. ROBSON
Sect. 3. ASCYREIA Choisy
Prodr. monogr. fam. Hypéric.: 37, 38 (1821).

Trees (rarely), shrubs or shrublets up to 4-5 m tall, evergreen or deciduous, glabrous, without
dark glands; branching nearly always lateral. Stems 2—4-lined and + compressed when young,
sooner or later becoming terete, eglandular; cortex exfoliating in strips or scales; bark fissured,
smooth. Leaves opposite, decussate, sessile to shortly petiolate, free, deciduous at basal
articulation; lamina entire, with venation pinnate, open or usually closed, the tertiary almost
absent (peripheral) or + densely reticulate; laminar glands striiform to punctiform or rarely
interrupted-linear, pale; marginal gland dots pale, dense; ventral resinous gland dots sometimes
present. Inflorescence 1-25-flowered, branching dichasial/monochasial from 1—2(3) nodes,
sometimes with subsidiary flowering branches from lower nodes; bracts and bracteoles foliar or
reduced, not sepalline. Flowers stellate to cyathiform, homostylous. Sepals 5, free or very rarely
united at the base (Sp. 19 in part), persistent, erect to reflexed in fruit, with margin entire or
minutely denticulate or ciliolate; veins numerous; laminar glands pale, linear or interrupted to
rarely punctiform; submarginal, inframarginal glands absent or pale; marginal glands absent.
Petals 5, deciduous, with apiculus present, lateral to terminal, + prominent or sometimes
obsolete, margin entire or minutely denticulate or glandular-denticulate; marginal glands
absent or amber; laminar glands + numerous, pale, linear. Stamen fascicles 5, free, deciduous,
each with c. 20-100 stamens; filaments united very shortly; anthers yellow to red, gland amber;
pollen types I-IV. Ovary with 5 incompletely (or completely ?) axile placentae, ~-ovulate;
styles 5, free or partly united; stigma capitate to small and rounded. Capsule 5-valved,
coriaceous, rarely fleshy at first (Sp. 5), not vittate. Seeds + narrowly cylindric to fusiform or
cylindric-ellipsoid, often carinate or laterally winged, with apical expansion sometimes wing-
like; testa linear-reticulate to linear-foveolate.

BASIC CHROMOSOME NUMBER (x): 12-9; ploidy 2, 4 (6).

Hasirat: open grassland, stony hillsides or cliffs, streamsides, thickets or deciduous woodland
(Sp. 14); 30-4800 m.

DISTRIBUTION: Pontic Turkey and Bulgaria; Pakistan, India, Sri Lanka, Nepal, Sikkim, Bhutan,
Burma, W. Indonesia (Sumatra, Java, Bali, Lombok, Sulawesi), Thailand, Vietnam, China
(north to southern (Xizang [Tibet], Sichuan, Shaanxi, Hubei, Anhui, Shandong, and Hebei),
Taiwan (?).

42 species (+ 4 subspecies).

Key to sect. 3. Ascyreia

1 Leaves (at least upper) sessile, without visible dense reticulation beneath; styles free .......... 2

Leaves all subsessile or shortly petiolate or, if sessile, then visibly + densely reticulate-
veinedibeneathand/onwithistyles)immite dies. essesteee es ereeeeaeesne eee eee eee eeee eee eee eee eee 7
2(1) Styles 1:8-2-5:x ovary; petal-apiculusiaciite:. sdrtcc-ceeseacpene sora -eces.-tete ae reneeee er seee ee eee 3
Styles 1-1-5 X ovary; petal apiculus obtuse to rounded .....................ccececeeceeceeeeeeeteeeeees 4

3(2) Leaves + elliptic, base cuneate to reflexed-auriculate; petals c. 3 X sepals; capsule 8-13 mm
WAGE ss sce cisrcns Ros alecg cradh arse ae CA ER ee eee 1. mysurense(p. 210)

Leaves + oblong, base shallowly cordate-auriculate to rounded; petals 1-5—2-5 x sepals;
capsule’ 36m wide feria: ae ee ee cae ee eee 2. cordifolium(p. 213)
4(2) Leaves linear to narrow-oblong or -lanceolate (1: b = 3-7), sharply acute to subacuminate.... 5
Leaves oblong to broadly ovate (1: b = 1-7-3 or rarely to 3-5), subacuminate to rounded ...... 6

5(4) Leaves linear to narrowly (oblong-)lanceolate, 25-80 mm long; sepals ovate to lanceolate;
inflorescence|o-Hhowene disse tere ee nee eee eater ere ee eee eeee eee eee 3. podocarpoides (p. 214)

Leaves elliptic to elliptic-oblong, 10-14 mm long; sepals narrowly elliptic; inflorescence
Lflowened |. ost cicclecteneoeale ca astasn ce teeren dhe setae eee ee Ce ee 4. sherriffii(p. 217)

6(4) Sepals lanceolate, acute to obtuse; leaves oblong to oblong-lanceolate 6. pachyphyllum(p. 219)

STUDIES IN THE GENUS HYPERICUM L.

Sepals ovate or oblong to elliptic, obtuse to rounded; leaves oblong-lanceolate to broadly

SEA LC es ee een rae are ete eclectic sisiaiacleiiainigarctne na getter wacbnaajaioemerame eine 7. augustinii(p.

7(1) Stems prostrate to ascending or pendent, forming low mats or clumps; styles + equalling
ER NOONE RD soot in oer rae RAt a ote Cnc? Rite act Per eee nee Mee ete Pam ace ante 5. reptans(p.
Stems erect to straggling or, if creeping and rooting (Sp. 14), then styles 1-5-3 X ovary ........

8(7) Styles straight or gradually outcurving; leaves not densely reticulate-veined .....................
Styles + abruptly curved or flexed below apex or, if straight, then leaves densely reticulate-
Rea AD Ne reat ae tte? Sains tiaolt mamas D oda Miona nina hemioaddvaleiaadalioanntlidinie wniidaindaiveiassiniees

9(8) Sepals erect in bud and fruit, rounded or obscurely apiculate; stamens 0-5—0-6 x petals;
POM EMME US OOLUISE creates Sane rimet ante om taracset sticenne tint at retceeroacemer natin Zane nck acters

Sepals spreading in bud and fruit, + markedly apiculate or acute; stamens 0-25-0-4 x petals;
MemeraA LIENS Lire TO SUDAGUEE Ve N no. hsae tee cddeudennsthoundunies senbwemildd lade osanshsutloastSiantislloe ves

10(9) Sepals 6-9 mm long; petals narrowly obovate, 15—20 mm long; leaves ovate or ovate-oblong

Fovanceolate (I Di—s2—2 -5)) hacwsce ates eemteetia. tronceene-caMennetcesek cmpice saces 8. williamsii(p.

Sepals 4-5 mm long; petals + broadly obovate to subcircular, 10-13 mm long; leaves
lanceolate to narrowly elliptic (1: b = 3-4) 0.0... ee eeeceecneeeeeeeeeeeees 9. tenuicaule(p.

11(9) Sepals ovate-oblong to broadly elliptic, apiculate or rarely rounded; petals 20-26 mm long;
Pe Tamar dl CP EEN EAL Se Ca ciae eed hasan ew sceaxa usrteenianeschstbrats ssevanemaeasenesed 10. lobbii(p.

Sepals narrowly elliptic to linear-oblong, acute; petals 12-18 mm long; plant a ‘low
BUT TS MOTT ete sary ciescerals at acslotial marae isos Ga ade aetarad at emad achesneauacen sec esite 11. gracilipes(p.

12(8) Leaves sessile with base cuneate; styles free or loosely coherent ..................0cceeeeeneceen sense
Leaves subsessile or shortly petiolate or, if sessile, then base rounded to cordate or styles +
THIGIIE. sone SognemashescondCd lore cach b Sooheeces care ROR HC CC nan ecco Pe RE MEET CRE. neec roe eon tec ee eet eee ae see ae

13(12) Leaves concolorous, venation incompletely reticulate ......................0000. 1. mysurense(p.
eaves == (CISCOOLOUS, VENAtlONCOMMleEte WTEMCULALG .ocreciaccencsedseerar seated vesssesseseccrescsnenc

14(13) Sepals erect in fruit, (6-)9-13 mm long; stamens 0-5—0-65 x petals; branches erect to

ASC EMCI area ncte teleqcbina sets secict- eciinate peda eta eestor of asshaiefsaina sigauisinclaelgaiseis sian aohelses 12. gaitii(p.

Sepals ascending in fruit, S—8 mm long; stamens 0-65-0-75 x petals; branches spreading to

Sito Clin OT Sem-SCAMGEMt sa... ceardcstieccrseacecsesuecbencnnceasets s+ 13. oblongifolium(p.

15(12) Leaves densely reticulate-veined beneath or styles + united; styles at least 1-5 X ovary........
Leaves laxly or not visibly reticulate-veined beneath; styles free, less than 1-5 xX ovary
Pee DSH Oe20, 57 ANC NVOLOS OL LA). (va. d.sctcnscrwedccweresuvnceates ent stierseedeaoesomsswaed oane

IGG) es ilesirec or loosely coherent inllowen half x. -<ccceccccsaascseceemvesecwehace sence snetdaccuclencenas
LE SIMICe CAO SUUOlAP OR ee saaarct cei telae aeto aes cee dene re ls aati santsbissnisit dew Siva d dreds cee weldubiwstewe

17(16) Leaf base cordate, lamina ovate-oblong to ovate, with reticulate venation + markedly

HLOMUMe Mt De Me ables seem esene sete cose a-imerteccamnes cect ees ecteneeae ee tae 15. griffithii(p.

Leaf base cuneate to rounded, lamina elliptic or oblong to ovate-oblong, with reticulate
Ven AatloOnin Of OnscakCely, PLOmmMene DEMEAU a raaaceeemee se casaaeeeiaece ccseeetsieeictnssscaeeceesea ss

Peay esiviesixee; Sepals broadly elliptic tOSUDEITCULAL <0. ....22..2.c.c0ceceecceccescascecccerceccescssceeneess

Styles coherent in lower half at first; sepals linear......................0.:0c0eee es 17. cohaerens(p.

19(18) Leaves all sessile; anthers yellow; sepals elliptic-oblong to narrowly ovate; stems +

spreading but not creeping Or rooting ...............ee cee eeeeeceee eee eee eens 13. oblongifolium(p.

Leaves (at least lower) shortly petiolate; anthers reddish; sepals broadly elliptic to obovate

or subcircular; stems creeping and rooting .................ccccceceeeeeeeeee snes 14. calycinum(p.

20(16) Leaves densely reticulate-veined beneath, (20—)30-110 mm long; inflorescence 1-30-
MOWELECateLiMinalOmlOngisOOtSmerercesraactcmecotntasctccrtoee ceotncc cee neccmaees os cme tecceeeasee:

Leaves without visible reticulate venation beneath, 10-31 mm long; inflorescence always
1-flowered, terminal on long and on short lateral shoots (19. longistylum) .....................

21(20) Leaves with base cuneate to rounded, or if cordate then apex rounded; leaf lamina usually

broaderator abovemilddle a neeecceste soar sasetecsece ce seceacaccseceateces 16. monogynum(p.

Leaves with base cordate-amplexicaul, apex acute to acuminate; leaf lamina elliptic-ovate to

broaglhovaters-eaccccae cemensie sce ses Ore sucmallian snseee ten aciabsaies eben eseetiean geass 18. prattii(p.

207
219)

218)
8

11

221)
222)

223)

224)
13

15

210)
14

225)

226)
16

23

iL7/
20

230)

18

19
235)

226)
228)

21

22

231)
236)

208 NORMAN K. B. ROBSON

22(20) Ovary and capsule ellipsoid, usually stipitate......... 19a. longistylum subsp. longistylum(p. 238)
Ovary and capsule globose, sessile ....................:0:006 19b. longistylum subsp. giraldii(p. 239)

23(15) Leaves with + distinct but lax reticulate venation beneath; styles 1-1-5 x ovary; sepals
broad; anthers usually orange or reddish (hybrids of 14. calycinum)......................06.006 24

Leaves without visible reticulate venation or, if visibly reticulate, then styles shorter than
ovary or sepals narrowly oblanceolate to linear; anthers pale to deep yellow .................. 27

24(23) Sepals apiculate, broadly elliptic; petal apiculus curved-acute; anthers yellow
LAG Foie See Re TITEL RA AO EE aE 36x. kouytchense x calycinum(p. 276)
Sepals rounded, ovate-oblong to obovate; petal apiculus rounded; anthers orange or

DE AGISH 56 onttlecet esioessnaece tomes ctee sete Be Sa Cnr ane Be nee cia SOG DeR CaN ace OR CeE CHEESE CEE EERE eee 25

25(24) Anthers brick-red; stamens 0-5—0-75 x petals; green leaves usually apiculate; stems
JO Weal Chin gig. feces cbrac sete tecertiscacttacs ue eesaenente mere steer eee eee eeemne 31x. x moserianum (p. 266)

Anthers orange; stamens 0-35-0-5 x petals; green leaves not (or rarely sub-) apiculate;
stems)high-arching.onspreading toascemdine)see.-.re-eee- eens eee see eee teeeenaee see eee ee ene 26

26(25) Leaves triangular-lanceolate, acute to obtuse; petals golden yellow, not red-tinged;
branchesiarchinptoispreadimgie----cet-e eee eeeeeee eeeteen eee eee eee eeeee 27xx. x ‘Hidcote’ (p. 254)

Leaves oblong-ovate, subapiculate to rounded; petals deep golden yellow, sometimes
tinged red outside; branches spreading-ascending .....................2s0000 42x. x dummeri(p. 287)
27(23) Leaves in one plane (distichous) or, if tetrastichous, then broadest at middle..................... 28
Leaves in four rows (tetrastichous), broadest below middle ...................cccc0cceceeceeeeeeeeees 44
28(27) Sepals spreading to reflexed in bud and fruit; styles shorter than (or rarely equalling) ovary 29
Sepals outcurved to erect in bud and fruit or, if spreading (Sp. 26), then styles 1-5—2 x ovary 32
29(28) Leaves elliptic to oblong or ovate-oblong, subsessile; stems persistently 4-lined................. 30
Leaves narrowly lanceolate to broadly ovate, + shortly petiolate; stems soon terete ........... Si

30(29) Leaves all narrowly elliptic, without clear intramarginal vein; stamens. 0-7 X petals; sepals
GVale COCMIOE Hee hc ences echce cre nee eee ee Sete oa ee eae eee 20. subsessile(p. 239)

Leaves oblong or ovate-oblong (lower) to elliptic (upper), usually with clear intramarginal
vein; stamens 0-35-0-5 x petals; sepals triangular-ovate ..................... 21. siamense(p. 240)

31(30) Sepals acute to subacute; ovary and capsule narrowly ellipsoid to narrowly ovoid-conic;
leaves triangular-lanceolate to broadly ovate .................ceeeeeeeeee ees 22. leschenaultii(p. 242)

Sepals apiculate to rounded; ovary and capsule ovoid-ellipsoid to ovoid; leaves oblong-
lanceolate to vate seer scscsc sata nes eae eae tnwtlteon Meee oee earner ce: 22x. x ‘Rowallane’ (p. 244)

32(28) Leaves with distinct, usually continuous, intramarginal vein; sepals acute to acuminate;
stamens07/5—0: 85 <petalstae acer eereceecneee ee eeee steer eee neers 23. acmosepalum(p. 245)

Leaves without intramarginal vein or, if partial one present (Sp. 29), then sepals apiculate-
obtuse to rounded and stamens 0-35—-0-5 X petals ..............cccceesecececeeeeeeeeneneeeeeeeeeenees 33

33(32) Sepals without or with very narrow hyaline margin; stems arching or spreading to pendulous
but not frondose (i.e. with branches in one plane), not or scarcely ancipitous when young 34

Sepals with marked hyaline margin; stems erect to arching or spreading and then often
frondose, markedly ancipitous at least when young....................eceeceeeeeeceeeeeeeeeneeneaes 39

34(33) Styles 0-35-1 X ovary; sepals usually ovate-lanceolate to obovate-spathulate, acute to
TOW said sated inde lee tienlnerieler sine aise be era a eel fa Sg eR ee 35
Styles 1-5-2 X ovary; sepals lanceolate to linear, acute or shortly aristate .................02.020-- 38

35(34) Leaves narrowly elliptic to oblong-elliptic; flower-buds acute to obtuse; stamens 0-6—0-7 x
PO tals a. 25 se -tian com sagidb ra pieimnr ach eectcaete ae aaa ar eset a ue cei eosscetaet 24. lagarocladum (p. 247)
Leaves broader below middle; flower-buds obtuse to rounded; stamens 0-25—0-4 x petals.... 36

36(35) Sepals acute to obtuse or rarely rounded-apiculate, not becoming ribbed; stamens 40-45 per
PASCi chee. «os: xcen opens MMe at cheascsc Cree cree aoe tse Sa ae nN cE 37

Sepals rounded or very rarely rounded-apiculate, often becoming markedly ribbed; stamens
60-80 perfasetcle aiicis. os) Saeeacocbheis Sas seb ate Avera aieta se eee eee aoe 28. hookerianum(p. 255)

37(36) Leaves apiculate-obtuse to rounded, elliptic-oblong to ovate-lanceolate; sepals ovate to
oblong-spathulate; styles 0-7-1 X Ovary ............:eecceceeeeeneeeneeeeeenees 27. addingtonii(p. 251)

STUDIES IN THE GENUS HYPERICUM L.

Leaves acute to apiculate-obtuse, lanceolate; sepals ovate-lanceolate or narrowly elliptic to

oblong-spathulate; styles c. 0-6 X OVATY ..........0.. cece eeeeeeee eee es 27x. x cyathiflorum(p.

38(34) Sepals erect in bud and fruit, ovate-oblong to ovate-lanceolate or elliptic; petals obovate.....

25. wilsonii(p.

Sepals spreading in bud and fruit, narrowly oblong-lanceolate to linear; petals oblanceolate

26. dyeri(p.

39(33) Leaves oblong to elliptic-oblong, with incomplete intramarginal vein.............. 29. lacei(p.

Leaves narrowly elliptic or broader below middle, without intramarginal vein (H. patulum
ean erg NN NES yt ean sa Piss no Shs alas sacs ROIS CI SIO old esate en « GUESS

40(39) Sepals entire, at least outer broadly elliptic or broadly oblong to circular; stems erect, not

BRONMGL OS Cotas eta taciniesvacieisteaisattewanopemoant goaneemeemersntect acct 30a. henryi subsp. hancockii(p.

Sepals eroded-denticulate or, if entire, then elliptic or narrowly oblong to oblanceolate;
SemNeleChTOMprleadiny, SOMEUMES IO NGOSE 1. «noose Sian sniae cles je oie onje ue sernseveeeiasels

41(40) Sepals eroded-denticulate to subentire, usually apiculate, broadly elliptic to broadly ovate;
leatapex usually apicilate-obtuse to rounded... 20.5. -c0ecccsccsccasdevecs cieedhewseedecacaaedevnes

Sepals entire, rarely apiculate, elliptic to narrowly oblong or obovate-spathulate; leaf apex

Bee LOLOUNGEH-APICU Ate seen esoenee natn scce nineties ane apes tare cane nacsiannes tees eaiseeuacena tome

42(41) Stems erect to arching or rarely divaricate, not frondose, + persistently 4-lined; leaf-apex
acute to rounded but rarely apiculate; capsule 10-14 mm long ....................ccceeeeeeee ees

I rn ee che et ar re meena wars Satan tec araccundugenenss 30b. henryi subsp. henryi(p.

Stems spreading, sometimes frondose, soon 2-lined; leaf-apex obtuse to rounded, always

Bpreniare eapsmie Oil mim lOMpee eacesesescc.ccs-teasecacessecsessceefarcecdoecane 31. patulum(p.

43(41) Flower-buds acute to obtuse; sepals subacute to rounded; stems erect or arching, not or only

Riba MORO OSE 225 ci nice de nena eeateneadenpwinencattisareekvnelen aes 30c. henryi subsp. uraloides(p.

Flower-buds obtuse to rounded; sepals rounded; stems arching, often frondose...............

44(27) Leaves with main lateral venation (usually conspicuous) closed and/or with tertiary venation
rather densely reticulate; sepals broadest at middle, entire ....................ecceceeeeeeee eee ees

Leaves with main lateral venation (usually obscure) open, with tertiary venation not
conspicuously reticulate; sepals usually broadest below middle, often denticulate ...........

45(44) Sepals outcurved to recurved in bud and fruit, usually acute; leaves lanceolate to triangular-
ovate, usually subacute to acuminate; flowers stellate to shallowly cyathiform, buds acute

FS) ZUBIN Sere ab goon make Groen Eee bcmOod mE BpeeSTOre OCR COR OBE On Cet: NN Soa nee atone Cane ie eens

Sepals erect in bud and fruit, usually obtuse to rounded; leaves lanceolate-oblong to broadly
rhombic or subcircular, obtuse to emarginate; flowers + deeply cyathiform, buds obtuse

Com Aane IVACHLe) TO LOUNGE (35, DEMIUIM) <.....0.-coserhacesurasanavconssrsasessacnscasvecsecascanssnees

46(45) Petals obovate-oblanceolate, flowers stellate; stamens c. 0-6 X petals; leaves glaucous

beneath pettole 0-S—2mmilonoree-. sce seestseotees tec staSacceseatt oss s.tnoets 33. maclarenii(p.

Petals broadly obovate to obovate-subcircular, flowers shallowly cyathiform; stamens
0-35-0-4 x petals; styles 0-85-1 x ovary; leaves not glaucous beneath, petiole 2-4 mm

IKDIg fied oadronndtedaduacocrtec Beetoneecoco: caccren Sag CoeeRE EE arch nC che nneE amt nice A aaa 34. choisianum(p.

47(45) Sepals acute to obtuse, elliptic, often foliaceous; flowers 40-60 mm in diam.; styles 0-35-0-7
X ovary; leaves narrowly oblong to ovate-oblong (1: b = usually 2 or more) .................5.

RE aes dese ein sie. ote aucin adraaGiniae nt iarastaattpate htacitecile/aiid Aarnasio ee 35a. bellum subsp. latisepalum(p.

Sepals rounded, narrowly elliptic to obovate, not foliaceous; flowers 25-35 mm in diam.:

leaves ovate-oblong to subcircular (1: b = 2 orless) .......... 35b. bellum subsp. bellum(p.

48(44) Sepals lanceolate to narrowly elliptic or oblanceolate, acute or acuminate; petal apiculus
ACILEOODLUSCOMODSOLELE meena eee erent eee Tansee eee rec neeet Meteee sesteet > ec sasnneaceseininttamene
Sepals ovate to circular or obovate, acute to rounded; petal apiculus obtuse to rounded or

EAS GRE set ete emer ee SMP Ree eee RRC ee See ren ire cutis cto vac rit miesacoaneost ager

| 50(49) Styles equalling or longer than ovary; stamens c. 0-6 X petals; inflorescence (1)3-14-

209
253)

248)

249)
258)

40

261)

41

42

43

263)
265)

263)

. 268)

45

48

46

47

270)

271)

274)
275)

210 NORMAN K. B. ROBSON

FO WEN isteis seca holes lok RST eA Saas Ds eae eso ase se maa cree soe 51
Styles shorter than ovary; stamens 0-3-0-4 x petals; inflorescence 1(2-4)-flowered ............ a2

51(50) Ovary 4-5 mm long, styles c. 1-2 x as long, straight or flexuous; inflorescence branches
relatively slender; leaves with dense ventral glands ...................:.:00666- 37. stellatum(p. 278)

Ovary 5-6-5 mm long, styles 1-1-1 x as long, outcurved; inflorescence branches usually
relatively stout; leaves usually with few or without ventral glands ......... 38. Jancasteri(p. 279)

52(50) Petals 30-35 mm long; flowers shallowly cyathiform, leaves elliptic-oblong to oblong-
lanceolate, obtuse to rounded; styles c. 0-75—0-9 X ovary ..... 36xx. x ‘Eastleigh Gold’ (p. 277)

Petals 15-20 mm long; flowers deeply cyathiform; leaves triangular-ovate to triangular-
laniceolate sstylesic.O SOWA cee. tee ser mcciso wa eeaer ware uene are radeees 39. curvisepalum (p. 281)
53(48) Sepals acute to obtuse; stems + persistently 4-lined ................-ceseeeeee eee eeeeeeeeeeeeneeeee enna 54
Sepals subapiculate to rounded; stems soon terete ...............66eeesseeeeeneeees 42. forrestii(p. 286)

54(53) Styles exceeding ovary; stamen fascicles 0-75—0.85 x petals; leaf apex usually rounded .......
41. pseudohenryi(p. 283)

Styles equalling or shorter than ovary; stamen fascicles 0-5—0-7 x petals; leaf apex usually
obtuse oriapicullatio cs rota eh. iu neletlik san GlOOUS TEL Ree rene Oe eee 40. beanii(p. 282)

1(12). Hypericum mysurense Wallich ex Wight & Arnott

Prodr. fl. penins. ind. or. 1: 99 (1834); Thwaites, Enum. pl. zeyl.: 48 (1858)
(‘mysorense’); Dyer in Hook. f., Fl. Brit. Ind. 1: 253 (1874) (‘mysorense’);
Cameron, Cat. pl. Bot. Gard. Bangalore 2nd ed.: 16 (1891); Cooke, Fl. Bombay 1:
74 (1901) (‘mysorense’); Rao, FI. pl. Travancore: 28 (1914) (‘mysorense’); Gam-
ble, Fl. Pres. Madras 1: 70 (1915) (‘mysorense’); Fyson, Fl. Nilgiri & Pulney
hill-tops 1: 37 (1915) (‘mysorense’); ibid. 2: 28, t. (1915); R. Keller in Engler &
Prantl, Nat. Pflanzenfam. 21: 176 (1925)(‘mysorense’); 2nd ed. FI. S. Ind. hill stns
1: 46 (1932) (‘mysorense’) 2: 20, t. (1932); Ramaswamy & Razi, Fl. Bangalore
distr.: 394 (1973) (‘mysorensis’); Saldana & Nicolson, Fl. Hassan distr.: 128
(1976). Types: India. Mysore, Nundydroog, 17.iii.1806, Heyne in Wallich 4808
(K-W!, lectotype—mihi; E!, K!, isolectotype); [Tamil Nadu] Neelgherries, Notan
in Wallich 4808C (K-W! syntype).

H. auritum Moon, Cat. Ind. exot. pl. Ceylon: 56 (1824), nomen.

H. mysurense Wallich, Numer. list No. 4808 (1831), nomen; Wight, Cat. Ind. pl.: 20,
no. 331 (1833), nomen.

Norysca myrtifolia Spach, Hist. nat. vég. Phan. 5: 428 (1836); in Annls. Sci. nat.
(Bot.) II, 5: 364 (1836); Blume, Mus. bot. Lugd. Bat. 2: 22 (1856). Type: India,
‘Dans les montagnes de l’Inde’, Leschenault (P!).

Norysca mysorensis (Wallich ex Wight & Arnott) Wight, Ic. pl. ind. or. 1: t. 56 (1838)
(‘Norisca’); Blume, Mus. bot. Lugd. Bat. 2: 22 (1856); Y. Kimura in Nakai &
Honda, Nova fl. jap. 10: 98 (1951).

H. norysca Steudel, Nomencl. bot. 2nd ed, 1: 789 (1840). Type as for Norysca
myrtifolia.

Icones: Wight, Ic. pl. ind. or.: t. 56 (1838); Fernando, Wildflowers Ceylon: 64. t. 16,
f. 8 (1954); Fig. 12.

Shrub (0-6)1-2-2-4 m tall, bushy, with branches erect to ascending. Stem red, 4-lined
and ancipitous when young, eventually terete; internodes 4-16 mm long, shorter
than leaves; bark grey. Leaves sessile 4-ranked; lamina 7-45 x 3-18 mm, narrowly to
broadly elliptic or oblong-elliptic, acute, margin plane, base cuneate to reflexed-
auriculate, concolorous, not or scarcely glaucous, subcoriaceous to chartaceous,
lower ones soon deciduous; venation: 3(4) pairs of main laterals, not or sparsely
branched, the midrib pinnately branched, with tertiary reticulum lax and inconspi-
cuous or rarely rather dense; laminar glands linear, between main lateral veins, or +
interrupted by reticulum to form streaks and dots; ventral glands absent. /nfloresc-
ence 1-3-flowered or, if flowers more (4-6), then usually from two close nodes, rarely
up to c. 10-flowered from apical node with flowering branches from second node:
pedicel 7-20 mm long; bracts foliar but smaller, persistent. Flowers 40-60 mm in

STUDIES IN THE GENUS HYPERICUM L.

a, ee
mo x 7 Sees

ae
; Ri Ry oh LB yf?
Wy VaE We
==" Vy, UY
Al

Fig.12 H. mysurense: (a) habit: (b) leaf section; (c) flower bud; (d) flower; (e) sepal; (f) petal; (g) stamen
fascicle; (h) anthers; (i) capsule (a x 1; d x 2;c, e-g,i X 4; b x 8;h X 20). (a) Sauliére 7, Clarke 10498A;
(b, c) Vine 10; (d-h) Evershed s.n.; (i) Hohenacker 435.

pAb

212

NORMAN K. B. ROBSON

diam., stellate; buds broadly to narrowly conical, acute. Sepals 6-10(13) x 3-5 mm,
free, imbricate, subequal, erect in bud and fruit, rather broadly to narrowly ovate or
ovate-triangular, sharply acute to subacute, margin entire or very rarely subentire,
midrib sometimes visible, often then incrassate above, other veins not prominent;
laminar glands linear, numerous. Petals rich golden yellow, sometimes tinged red,
spreading, 20-33 x 13-22 mm, c. 3 X sepals, narrowly obovate, with apiculus
subterminal, acute to rounded; margin entire, eglandular. Stamen fascicles each with
c. 40 stamens, longest 10-17 mm long, c. 0-5 x petals; anthers yellow. Ovary (4)5-6
x 34 mn, broadly ovoid to ovoid-pyramidal; styles (8)10-13 mm long, c. 2:5 x
ovary, free or loosely coherent, erect; stigmas small. Capsule (8)10-12(14) x 8-13
mm, broadly ovoid to narrowly ovoid-conic. Seeds dark reddish-brown, 2:2 mm
long, cylindric, not winged, slightly carinate, finely linear-foveolate. 2n = ?

In thickets and on wooded and open hillsides, on poor soil; 1050-2100 m.

Sri Lanka, south India north to Konkan (Tamil Nadu, Kerala, Karnataka, western
Maharashtra). Map 8.

SRI LANKA. Central Prov.: Nuwara Eliya Distr., 5 km S. of Nuwara Eliya, near
Nanu Oya, 1890 m, 10.iii.1969, Grierson 1082 (E); Adam’s Peak, iii.1846, Thwaites
(K); Nuwara Eliya [Neuer Ellia], 1800 m, Gardner 112 (A, BM, K); [without precise
locality], 1853, Thwaites 48 (BM, BO, FR, G).

INDIA. Tamil Nadu: Madurai Distr., Palni [Pulney] Hills, Kodaikanal,
19.v.1897, Bourne (K); Anaimalai [Anamalley] Hills, pre-1885, Beddome 371 (BM);

yn nN
=a}

Map8 Sect. 3. Ascyreia: 1. H. mysurense @, © (unlocalised). H. gaitii @.

STUDIES IN THE GENUS HYPERICUM L.

Nilgiri Hills, above Wellington—Kotogiri Road, Black Bridge Reserve Forest,
21.ii1.1944, Sinclair 3374 (E); in montibus Nilagiri, 1851, Hohenacker 1113 (BM, JE,
K). Kerala: Munnar, Devicolam road, 1500 m, 13.i1i.1934, Erlanson 5554 (A).
Karnataka: prope Mercara, ii.1847, Hohenacker 435 (BM, E, JE, K); Baba Budan
Hills, Santaveri, xi.1908, Meebold 10149 (E). Maharashtra: Konkan, etc., Stocks
(A, BM, K, L); Bombay Presidency, pre-iv.1878, Dalzell (K).

H. mysurense shows clinal variation from Sri Lanka and the Palni Hills of Tamil
Nadu northward to near Bombay. The southern plants are closer to the African H.
revolutum subsp. keniense and H. lanceolatum subsp. angustifolium from Réunion
(sect. 1. Campylosporus) in having narrow leaves without or with little tertiary
reticulation and with linear laminar glands, and long acute buds and sepals. Towards
the northern end of its range, the leaves tend to be relatively broader with marked
tertiary venation and dissected laminar glands, and the buds and sepals shorter and
subacute. It does not seem possible, however, to recognise infraspecific taxa.

The Socotran plants sometimes included in H. mysurense belong to sect. 1.
Campylosporus (5. H. balfourii).

2(13). Hypericum cordifolium Choisy

in DC., Prodr. syst. nat. regni veg. 1: 545 (1824); Dyer in Hook. f., Fl. Brit. Ind. 1:
253 (1874) pro parte excl. syn. H. acutum Wall.; R. Keller in Engler & Prantl, Nat.
Pflanzenfam. 2nd ed. 21: 176 (1925); BanerjiinJ. Bombay Nat. Hist. Soc. 51: 774
(1953); Anon. in Bull. Dep. Med. Pl. Nepal 4: 7 (1973); Murata in Acta Phyt.
Geobot. 25: 110 (1973); Robson in J. Jap. Bot. 52: 276, f. 1 (1977), in Hara &
Williams, Enum. fl. pl. Nepal 2: 61 (1979). Type: Nepal, Sheopore [Sheopuri]
1821, Wallich 4804 (G-DC!, holotype; BM!, E!, F!, K-W!, L!, SING!).

H. bracteatum Buch.-Ham. ex D. Don, Prodr. fl. nepal.: 217 (1825), nom. illegit.
superfl.; Wallich, Numer. list: No. 4804 (1831); Wallich, Pl. asiat. rarior. 3: t. 220
(1832). Types: (all Nepal) ‘ad Thankote [Tancote] in Nepalia propria’, 16.iv.1802,
Hamilton (BM!, lectotype); ‘ad Narainhetty’, 10.x.1802, Hamilton (BM!); Nepa-
lia, 1819-20, Wallich (BM!, SING!); Nepalia [Sheopore], 1821, Wallich 4804
(BM!, E!, FI!, K-W!, L!, SING!); Nepalia [Sheopore], ii.1821, Wallich (BM!, K!)
(all syntypes).

H. lungusum Buch.-Ham. ex D. Don, Prodr. fl. nepal.: 218 (1825), in synon.

Eremanthe cordifolia (Choisy) K. Koch, Hort. dendrol.: 65 (1853).

Norysca cordifolia (Choisy) Blume, Mus. bot. Lugd. Bat. 2: 22 (1856); Y. Kimura in
Nakai & Honda, Nova fl. jap. 10: 98 (1951); in Kihara, Fauna & fl. Nepal Himal.:
179 (1955); in Hara, Fl. eastern Himal. 2: 81 (1971).

Norysca urala var. angustifolia Y. Kimura in Hara, FI. eastern Himal.: 210 (1966) pro
parte quoad specim. Hara, Kanai & Kurosawa. 20.ix.1963.

Icon: Wallich, Pl. asiat. rarior. 3: t. 220 (1832).

Shrub 1-1-3 m tall, with branches erect to arching or pendent. Stems 2- or partially
4-lined and ancipitous when young, soon terete; internodes 7-20 mm long, shorter
than leaves; bark dark grey. Leaves sessile; lamina 18—-40(-62) x 5—16(—23) mm,
elliptic-oblong to oblong, apiculate or shortly acuminate, margin plane, base round-
ed to shallowly cordate-auriculate, paler beneath, + glaucous on both sides,
subcoriaceous; venation: 3 pairs main laterals, with intermediates and midrib
branches distally, without visible tertiary reticulum; laminar glands streaks and a few
dots; ventral glands dense. Inflorescence 1-7-flowered, from apical node, and with
1—3-flowered lateral branches from up to 12 nodes, rounded-pyramidal to cylindric,
sometimes with subsidiary flowering branches below; pedicels 7-14 mm long; bracts
ovate, persistent. Flowers 30-50 mm in diam., stellate; buds narrowly conic to
cylindric, acute. Sepals 8-13 x 3-S mm, free, imbricate, subequal to rather unequal,
erect to ascending in bud and fruit, lanceolate or rarely ovate or ovate-elliptic,
sharply acute to apiculate, with margin entire; midrib obscure or not visible, veins
not prominent; laminar glands numerous, linear. Petals bright yellow, sometimes
tinged red, spreading, (15—)20—22 x 7-12 mm, 1-5—2-5 X sepals, narrowly obovate to
oblanceolate, with apiculus acute, subterminal; margin entire, eglandular. Stamen

213

214 NORMAN K. B. ROBSON

fascicles each with c. 25 stamens, longest c. 15 mm long, 0-65—0-75 x petals; anthers
bright yellow. Ovary 4-6-5 x 2-5-4-5 mm, ellipsoid; styles 7-5—12 mm long, 1-8-2-5
X ovary, free, erect, sometimes outcurved near apex; stigmas capitate. Capsule 8-10
— 5-6 mm, + broadly ovoid. Seeds dark reddish-brown, c. 0-9 mm long, narrowly
cylindric, not carinate, shallowly linear-reticulate. 2n = ?

On steep banks, rocky hillsides and open, dry situations; 900-1900 m.

Nepal (central, in Kathmandu region; also east, one record). Map 9.

~
rc :
: : a oo Cee
sy é AN = Za Ag aa

; N
Map 9 Sect. 3 Ascyreia: 2. H. cordifolium @, 3. H. podocarpoides O.

NEPAL. Central: Sheopuri, N. of Kathmandu, 1875 m, 5.iv.1965, Schilling 302
(K); Katunji Ridge, 1050 m, 8.iv.1953, Gardner 194 (BM, E); Tribeni, 1500 m,
10.iv.1973, Dobremez 1792 (BM); near Kathmandu, Sundarijal Waterfall, 1600 m,
20.1x.1963, Hara, Kanai & Kurosawa (TI); Gurka Himalaya, N. of town of Gurka, +
1050 m, 29.11.1983, Schilling 2656 (BM). East: hills N. of Dharan, 20.x.1980.
Schilling (BM).

CULTIVATED. Specimens seen from India (Calcutta, nineteenth century) and
England (1980).

H. cordifolium is a relict species which occupies a critical position, as regarded both
morphology and distribution, for the understanding of the evolution of the genus.
Ancestrally it is linked (to the south) with H. mysurense; morphologically it is the
nearest species in sect. 3. Ascyreia to sect. 10. Olympia (radiating from south-
western Anatolia) and its derivatives (sects 11-16), and sect. 17. Hirtella (radiating
from eastern Anatolia) and its relatives (sects 18-19) (see p. 325); and it has a near
relative in the Himalaya itself (H. podocarpoides). Hundley & KoKo’s record of H.
cordifolium from Burma (Lace, 1961: 19) should be treated with reserve in the
absence of a confirmatory specimen. It could be an error for 6. H. pachyphyllum.

H. cordifolium is easily distinguished from all other species in sect. Ascyreia by its
combination of elliptic-oblong to oblong, auriculate leaves, ovate, acute to apiculate
sepals, and long styles. For the differences between it and H. podocarpoides, see
below (p. 215) and Robson (1977b: 277).

3(14). Hypericum podocarpoides N. Robson

in J. Jap. Bot. 52: 276, f. 1 (1977); in Hara & Williams, Enum. fi. pl. Nepal 2: 62
(1979). Type: Nepal, from Tinpipli to Nepalthoke, c. 900 m, 7.iv.1948, Banerji 95
(CAL, holotype; Hb. Banerji!, isotype).

H. acutum Wallich, Numer. list: No. 4807 (1831), nomen; Dyer in Hook. f., Fi. Brit.
Ind. 1: 253 (1874) nom. synon., non Moench (1784).

STUDIES IN THE GENUS HYPERICUM L.

H. cordifolium sensu Dyer in Hook. f., Fl. Brit. Ind. 1: 253 (1874) pro parte, quoad
syn. H. acutum Wall., non Choisy (1824).

H. hookerianum var. lineare Banerjiin J. Ind. bot. Soc. 31: 152 (1952); inJ. Bombay
nat. Hist. Soc. 51: 774 (1953); in Candollea 19: 219 (1964); in Rec. bot. Surv. Ind.
19 (2): 27 (1966). Type as for H. podocarpoides.

Icon: Fig. 13.

Shrub 0-45-1-2 m tall, with branches spreading or pendulous. Stems 4-lined and
ancipitous when young, eventually terete; internodes 5-15 mm long, shorter than
leaves; bark grey-brown. Leaves sessile or with very short flat petiole; lamina
(20-)25-80 x 4-20 mm, linear to narrowly lanceolate or narrowly oblong-lanceolate
(broader towards the inflorescence), sharply acute to subacuminate, margin plane to
recurved, sometimes reddish, subincrassate, base auriculate-truncate or subcordate
to rounded, discolorous, densely glaucous beneath, coriaceous; venation: 2-3 pairs
main laterals with numerous midrib branches distally, without visible tertiary
reticulum; laminar glands mostly undulating lines and streaks (crossing the veins)
with some dots; ventral glands sparse or absent. Inflorescence S—13(—20)-flowered,
from 1—2(3) nodes, subcorymbiform to broadly pyramidal, or sometimes pseudum-
bellate owing to the very short apical internode, occasionally with subsidiary
branches below; pedicels 4-25 mm long; bracts oblong or ovate to lanceolate, those
on main stem foliar and usually in a pseudo-whorl, others reduced, deciduous.
Flowers (25—)35—50 mm in diam., stellate; buds ovoid, acute. Sepals 8-14 x 3-6 mm,
free, imbricate, subequal, spreading in bud and fruit, narrowly ovate to lanceolate,
acute to subacute, with margin entire or slightly irregular to eroded-denticulate
towards apex; midrib of outer 1-3 visible, veins not prominent; laminar glands
numerous, linear or interrupted. Petals bright yellow, not tinged red, spreading,
13-28 X 8-12 mm, 1-6-2 X sepals, + narrowly obovate, with apiculus very acute to
almost absent, lateral; margin entire, eglandular. Stamen fascicles each with c. 30
stamens, longest 8-15 mm long, 0-5—0-65 x petals; anthers bright yellow. Ovary
4-5-6 X 2:54 mm, narrowly to rather broadly ellipsoid; styles 3-5—6:5 mm long,
1-1-2 X ovary, free, suberect to divergent, outcurved near apex; stigmas capitate.
Capsule 8-13 X 7-10 mm, ellipsoid to ovoid. Seeds dark reddish-brown, 0-9-1 mm
long, narrowly cylindric, very narrowly carinate, very shallowly linear-reticulate.
2n =?

On grassy slopes, rock faces and in stony river beds, in dry situations; (570)840—
2100(—2400) m.

Nepal (central, west), India (Kumaun). Map 9.

NEPAL. Central: Kali Gandaki, Tatopani, 1350 m, 25.v.1954, Stainton, Sykes &
Williams 5438 (BM); Godavari, S. of Kathmandu, 1950 m, 9.xii.1965, Schilling 713
(K). West: Kali Gad, 1200 m, 25.vii.1953, Tyson 118 (BM).

INDIA. Uttar Pradesh: Kumaun, Ramari, 1350 m, Strachey & Winterbottom 5
(BM, GH, K); Kumaun, between Gini [Ganai] and Munshiari [Mansiari], 1800-2400
m, 17.viii.1884, Duthie 2768 (K).

H. podocarpoides has been confused with H. cordifolium since Dyer made them
synonymous in Fl. Brit. India (Dyer, 1874). They are, however, quite distinct.
Although they approach one another in leaf shape, the respective variations do not
overlap; and they also differ in leaf thickness, colour and glandularity, and in
inflorescence shape, stamen length, ovary size and style length.

Whereas H. cordifolium occurs in a restricted area of central Nepal with an
isolated locality in the east of that country, H. podocarpoides is found mainly further
west (Robson, 19775: fig. 1). The area of overlap in central Nepal, however, is more
apparent than real, as H. podocarpoides is absent from the immediate region where
H. cordifolium occurs, although it is present to the east and south-east of it. There
also appears to be a gap in the distribution of H. podocarpoides in the west of Nepal.
In addition, H. cordifolium flowers in the spring and H. podocarpoides in the autumn
(teste A. Schilling).

It therefore seems that, although H. cordifolium and H. podocarpoides are both
closely related to H. mysurense and to each other, they have diverged from this

215

216 NORMAN K. B. ROBSON

Fig. 13. H. podocarpoides: (a) habit; (b) leaf section; (c) flower bud; (d) sepal; (e) petal; (f) stamen
fascicle; (g) anthers; (h) ovary; (i) capsule (a x 1; c,i X 4; b, d-f, h x 6; g x 40). (a-g) Stainton, Sykes &
Williams 8705; (h) Schilling 2431; (i) Stainton, Sykes & Williams 575.

STUDIES IN THE GENUS HYPERICUM L.

common ancestor and subsequently come to occupy overlapping areas without
hybridising or introgressing.

4(15). Hypericum sherriffii N. Robson

in Notes Roy. bot. Gdn Edinb. 41: 133, ff. 1, 3 (1983). Type: Bhutan, Chungkar,
24.xi.1938, Ludlow & Sherriff 6784 (BM!, holotype).

Icon: Robson & Long in Notes Roy. bot. Gdn Edinb. 41: 134, f. 1 (1983).

Shrub 0-6-1:2 m tall, with branches spreading or pendulous, + frondose ? Stems
4-lined and ancipitous when young, eventually unlined and compressed; internodes
1-3-5 mm long, shorter than leaves; bark greyish-brown. Leaves sessile or subsessile;
lamina 10-14 x 3-5 mm, elliptic to elliptic-oblong, acute to subacute-apiculate,
margin plane, base cuneate, discolorous, glaucous beneath, coriaceous; venation:
2-3 pairs of looping main laterals without other midrib branches or tertiary reticu-
lum; laminar glands short streaks and dots; ventral glands dense. Inflorescence
1-flowered; pedicels S—6 mm long; bracts deciduous. Flowers not seen. Sepals 5-8 x
1-5-2 mm, free, not imbricate ?, subequal, spreading in fruit, linear-elliptic, acute,
with margin entire or minutely denticulate below apex; midrib visible, veins not
prominent; laminar glands 4—6, linear, punctiform and + numerous distally. Petals,
stamen fascicles and ovary not seen; styles (in fruit) 3-5-4 mm long; stigmas small.
Capsule 8-11 x 5-5-5 mm, ellipsoid-ovoid. Seeds dark reddish-brown, c. 0-8 mm
long, narrowly cylindric, very narrowly carinate or not, very shallowly linear-
foveolate. 2n = ?

On cliff faces and steep rocky slopes; 2100 m.

Bhutan (south-east). Map 10.
BHUTAN. Chungkar, 2100 m, 24.xi.1938, Ludlow & Sherriff 6784 (BM).

Despite the absence as yet of flowering specimens, it is possible to see H. sherriffii as
a relative of H. podocarpoides, smaller in all its parts and with cuneate leaf-bases,
narrower sepals and solitary flowers. The small leaves on apparently frondose shoots
with very short internodes remind one of H. uralum, but the spreading habit, narrow
sepals and narrow capsules are quite different.

PANG

+H+t+o <a

lr 29 = <

poets

Map 10 Sect. 3 Ascyreia: 4. H. sherriffiiO, 5. H. reptans ©, 6. H. pachyphyllum &,7. H. augustinii ©, 10.

H. lobbii * , 11. H. gracilipes O.

218

NORMAN K. B. ROBSON

5(16). Hypericum reptans Hook. f. & Thomson ex Dyer

in Hook. f., Fl. Brit. Ind. 1: 255 (1874); R. Keller in Engler & Prantl, Nat.
Pflanzenfam. 2nd ed. 21: 176 (1925); Hand.-Mazz., Symbolae sinicae 7: 401
(1931); N. RobsoninJ. Roy. Hort. Soc. 95: 492 (1970); [Lancaster], Hilliers’ man.
trees & shrubs: 152 (1971); Bean, Trees & shrubs hardy in Br. Isles 8th ed. 2: 422
(1973); N. Robson in Hara & Williams, Enum. fl. pl. Nepal 2: 62 (1979). Types:
Sikkim, Lachen, 3300 m, 10.vii.1849, J. D. Hooker (K!, lectotype—mihi); Lachen,
3300 m, 17.vi.1849, J. D. Hooker (K!, syntype); temperate Sikkim Himalaya,
2100-3300 m, J. D. Hooker (BM!, E!, FI!, GH!, K!, L!, SING!, syntypes).

Icon: Mansfield in Bull. Alpine Gard. Soc. 5: 178 (1937).

Shrublet, prostrate or ascending to 0-3 m tall, forming clumps or mats up to 1 m in
diam., sometimes pendent from rocks, with branching pinnatiform, branches root-
ing. Stems red to orange, 4-lined and ancipitous when young, eventually 4-lined to
subterete; internodes 5—15 mm long, shorter than to exceeding leaves; bark reddish-
brown, flaking. Leaves petiolate, with petiole 0-5—1-5 mm long; lamina 7—16(—22) x
2-9 mm, increasing in size and relative width up stem, elliptic or elliptic-oblong to
oblanceolate or more rarely obovate, obtuse to rounded, margin plane, base
cuneate, paler or + glaucous beneath, chartaceous; venation: 1—2(3) pairs main
laterals, with midrib branched distally and dense, often obscure tertiary reticulum;
laminar glands medium to short streaks and dots; ventral glands absent. Infloresc-
ence 1-flowered, with flowering branches from middle and upper parts of current
stem; pedicels 4-8 mm long; bracts (uppermost leaf-pair) foliar, persistent. Flowers
20-30 mm in diam., + deeply cyathiform; buds ovoid-ellipsoid, rounded. Sepals 6-14
x 2-5-6-:5 mm, free, imbricate, unequal, reflexed in bud, spreading in fruit, oblong to
obovate or oblanceolate, obtuse or rarely apiculate to rounded, with margin entire;
midrib distinct, veins not prominent; laminar glands distal streaks and dots, relative-
ly few. Petals deep golden yellow, sometimes flushed red, + markedly incurved,
11-18 x 7-12 mm, c. 1-3-1-9 X sepals, broadly obovate, with apiculus subterminal,
rounded, almost absent; margin entire, eglandular. Stamen fascicles each with c.
20-30 stamens, longest 4-5-6 mm long, 0-25—0-35 x petals; anthers golden yellow.
Ovary 34 x 3-5 mm globose; styles (2-5) 3-4(4-5) mm long, about equalling ovary,
free, erect, + sharply outcurved near apex; stigmas subcapitate. Capsule 6-10 x
6-10 mm, globose to depressed-globose, indehiscent, + baccate, brick-coloured
when ripe. Seeds dark reddish-brown, 0-6-0-7 mm long, cylindric to cylindric-
ellipsoid, narrowly carinate, very shallowly linear-foveolate. 2n = ?

Dry to moist, open to shaded habitats (gravel or earth slopes, on boulders, hanging
over cliffs, marshy places), (2100)2640—3900(4200) m.

China (NW. Yunnan, SE. Xizang [Tibet]), Burma (north), India (Arunachal
Pradesh), Sikkim, Nepal (east, central). Map 10.

CHINA. Yunnan: Salween-Kiukiang Divide, Haipuh, 3500 m, 2.xi.1938, Yu
22945 (E); Gongshan [Kungshan, Tala], 11.xi.1940, Feng 8361 (KUN). Xizang
[Tibet]: Médog [Meto, Motui], 2700 m, 3.viii.1974, Zhuanzang team 74-4025
(KUN).

BURMA: Kachin: Nam Tamai valley, 3000 m, 8.1x.1937, Kingdon Ward 13183
(BM); Seingkhu Wang valley, 3300 m, 30.vii.1926, Kingdon Ward 7224 (BM).

INDIA. Arunachal Pradesh: Lo La, Pachakshiri, 2850 m, 8.x.1938, Ludlow &
Sherriff 6564 (BM).

SIKKIM. Mempup, 3000 m, 30.vii.1913, Cooper 381 (A, BM, E); Yeumthang,
3900 m, 30.vii.1947, Cave 175 (E, K).

NEPAL. Central: Annapurna Himal, Seti Khola, 3150 m, 5.viii.1954, Stainton,
Sykes & Williams 6688 (BM, E). East: Foketey, 3900-4200 m, 1930, Dhwoj 491
(BM, EB).

CULTIVATED. Specimens have been seen from England (1983), Scotland (c.
1975), Ireland (1968), and the Netherlands (1945).

The flowers of H. reptans are like small versions of those of H. hookerianum with the
sepals more foliose and the ovary globose. The habit, however, is quite different,

STUDIES IN THE GENUS HYPERICUM L.

varying from ascending to prostrate, with flowering branches from any or all nodes
along the rooting stem. In addition, the small, elliptic to obovate leaves and the
indehiscent, + baccate fruits are distinctive. The branching, leaf form, leaf venation
and sepal form all indicate a relationship with H. sherriffii (from Bhutan); and the
distribution along the Himalayan range on either side of that country (but apparently
not in Bhutan itself) is consistent with this relationship.

6(17). Hypericum pachyphyllum Collett & Hemsley

inJ. Linn. Soc. 28: 24, t. 3 (1890). Type: Burma, Shan Hills, 1200 m, Collett 660 (K!,
holotype).

Icon: Collett & Hemsley, tom. cit.: 24, t. 3 (1890).

Shrub or undershrub (height unrecorded), with branches erect, rigid. Stems 4-lined
and ancipitous when young, eventually terete; internodes 8-20 mm long, shorter
than leaves; bark reddish-brown. Leaves sessile or with very short flat petiole; lamina
35-50 x 10-20 mm, oblong to oblong-lanceolate, subapiculate to obtuse, margin
plane, reddish, subincrassate, base auriculate to broadly cuneate, discolorous,
densely glaucous beneath, coriaceous; venation: 3 pairs main laterals, with midrib
branched distally, without visible tertiary reticulum; laminar glands mostly lines and
long streaks; ventral glands sparse. Inflorescence c. 10—25-flowered, from 2-4 nodes,
subcorymbiform, with very short apical internode, sometimes with branches below;
pedicels 9-17 mm long; bracts on main stem broadly ovate, forming a pseudo-whorl,
others reduced, oblong to elliptic, deciduous. Flowers 40—50(—60) mm in diam.,
stellate or shallowly cyathiform; buds broadly to narrowly conic, acute to subacumin-
ate. Sepals 8-14 x 3-5-5-5 mm, free, imbricate, subequal, + spreading in bud and
fruit, lanceolate, acute to subacute, with margin entire; midrib obscure except
sometimes near apex, veins not prominent; laminar glands numerous, linear or long
streaks. Petals bright? yellow, not tinged red, spreading or slightly incurved, 20-25 x
9-13 mm, c. 2-2-5 X sepals, oblanceolate to rather broadly ovate, with apiculus
subterminal, rounded or almost absent; longest 10-14 mm long, c. 0-5 xX petals;
anthers bright yellow. Ovary 5-6 x 3-5 mm, narrowly ellipsoid; styles 7-9 mm long,
1-1-5 X ovary, divergent, outcurved towards apex; stigmas capitate. Capsule 10-13
Xx 6-8 mm, ovoid. Seeds not seen. 2n = ?

On open slopes and in thickets and marshy areas; 600-1200 m.

Burma (Shan States). Map 10.

BURMA. Shan States: between Lashio and Namkhan, 600-900 m, 22.xii.1952,
Kingdom Ward 20307 (BM); Shan Hills, 1200 m, v.1888, Collett 660 (K); [no
locality], pre ii.1906, Brandis (K).

H. pachyphyllum is intermediate, morphologically and geographically, between the
Nepalese H. podocarpoides and H. augustinii (from southern Yunnan). It has a
restricted relict distribution.

7(18). Hypericum augustinii N. Robson

in J. Roy. Hort. Soc. 95: 495 (1970); [Lancaster], Hilliers’ man. trees & shrubs: 150
(1971); Bean, Trees & shrubs hardy in Br. Isles 8th ed. 2: 408 (1973); N. Robson in
J. Jap. Bot. 52: 278, f. 2 (1977). Type: China, Yunnan, Papien R., Shih-Ping,
x.1898, Henry 13242 (K!, holotype).

Icon: Fig. 14.

Shrub c. 1 m tall, with stems erect or arching, without or with a few short lateral
branches. Stems 4-lined and ancipitous when young, soon unlined, eventually terete;
internodes 25—70 mm long, mostly exceeding the leaves; bark reddish-brown. Leaves
all sessile or the lower (or rarely all but uppermost) with flat petiole to 1-5 mm long;
lamina (30—)37-75 x (10—)15—44 mm, oblong-lanceolate or oblong-ovate to broadly
ovate, acute to rounded-apiculate, margin plane, not or scarcely incrassate, base
rounded to subcordate, the upper amplexicaul, paler and + glaucous beneath,

219

220 NORMAN K. B. ROBSON

Fig.14 H. augustinii: (a) habit; (b) leaf section; (c) flower bud; (d) sepal; (e) petal; (f) stamen fascicle; (g)
anthers; (h) capsule (a x 1; b-f, h x 4; g x 40). All Yeos.n.

STUDIES IN THE GENUS HYPERICUM L.

coriaceous; venation: 3 pairs main laterals, with midrib branched distally, without
visible tertiary reticulum; laminar glands long streaks and dots; ventral glands dense.
Inflorescence (1) 3-13-flowered, from 1-2 nodes, subcorymbiform, with short or very
short terminal internode, without branches below; pedicels 6-12 mm long; bracts on
main stem broadly ovate, usually forming a pseudo-whorl, others reduced, broadly
ovate to lanceolate, deciduous. Flowers 40-66 mm in diam., stellate to subcyath-
iform; buds broadly ovoid, obtuse. Sepals (7—)10-15 x 4-9 mm, free, imbricate,
subequal to unequal, erect in bud, + spreading in fruit, + broadly oblong to broadly
elliptic or ovate, subapiculate or obtuse to rounded, with margin entire or slightly
eroded towards apex; midrib distinct in outer sepals, obscure except near apex in
inner ones, veins not prominent; laminar glands numerous, linear to punctiform.
Petals pale to bright golden yellow, not tinged red, spreading or slightly incurved,
20-36 X 14-26 mm, c. 2-5-3 X sepals, obovate, with apiculus subterminal, rounded
or almost absent; margin entire or very minute denticulate, eglandular. Stamen
fascicles each with 60-70 stamens, longest 10-20 mm long, c. 0-5 x petals; anthers
golden yellow. Ovary 5-6 x 4-5 mm, broadly ovoid; styles 6-8 mm long, c. 1:2 x
ovary, erect to gradually divergent, outcurved towards apex; stigmas capitate.
Capsule 10-12 x 9-10 mm, broadly ovoid. Seeds dark reddish-brown, c. 1-5 mm,
long cylindric, not or scarcely carinate, shallowly linear-reticulate. 2n = c. 54.

Collected at 1200-1890 m.

China (southern Yunnan). Map 10.

CHINA. Yunnan: Papien R., Shih-Ping, 1200 m, x.1898, Henry 13242 (K);
Shiping, 1700 m, 12.iii.1959, Wu 868 (KUN); Kunyuan, 1890 m, 10.i.1963, Zhang
100365 (KUN); Kunmingshan, Mekong R., Kengbung [Cheli], 1866 m, 1935,
Wissmann 1172 (W).

CULTIVATED. Specimens seen from England (1923-1983), Eire (1967), and the
Channel Islands (Jersey, 1970).

H. augustinii, a local species from the extreme south of Yunnan, is related to H.
pachyphyllum (from eastern Burma). It is a stout plant with large, coriaceous leaves
and large flowers, characters which may be a result of polyploidy. It has not proved
possible yet to obtain a consistent chromosome number. The numbers so far
counted, which range from 48 to 55, suggest that H. augustinii may be a hexaploid on
the base x = 9 (i.e. 2n = 54). In Europe, the species flowers too late in the season to
allow fruits to develop.

8(19). Hypericum williamsii N. Robson

in J. Jap. Bot. 52: 279, f. 2 (1977); in Hara & Williams, Enum. fi. pl. Nepal 2: 63
(1979). Type: Nepal (West), Poyora, 2850 m, 28.vii.1952, Polunin, Sykes &
Williams 354 (BM!, holotype).

Icon: —.

Shrub 0-7-1-3 m, tall, bushy, with branches ascending, not frondose. Stems orange-
brown, 2-lined when young, soon terete; internodes 10-20 mm long, shorter than to
slightly exceeding leaves; bark grey-brown. Leaves petiolate, with petiole 0-5-2 mm
long; lamina 20-49 x 8-24 mm, lanceolate or ovate-oblong to ovate, obtuse to
rounded, apiculate, margin plane, not incrassate, base cuneate-angustate to round-
ed, glaucous beneath, chartaceous; venation: 34 pairs main laterals, the midrib
branched distally, with scarcely visible lax tertiary reticulum; laminar glands streaks
and dots; ventral glands dense. Inflorescence 1—7(-16)-flowered, from 1(2) node(s),
corymbiform, usually with short apical internode, without flowering branches from
middle of current stem; pedicels 3-6 mm long; bracts lanceolate, acute, deciduous.
Flowers 30-40 mm in diam., shallowly cyathiform; buds broadly ovoid, obtuse.
Sepals 6-9 x 3-5-5 mm, free, imbricate, subequal, ascending in bud and fruit, oblong
to elliptic-oblong, rounded or more rarely apiculate-obtuse, with margin entire or
minutely denticulate, hyaline; midrib of outer sepals + distinct, veins not or slightly
prominent; laminar glands linear, numerous. Petals bright yellow, not tinged red,
slightly incurved, 15-20 x 13-15 mm, c. 2:5 X sepals, narrowly obovate, with

221

0

222. NORMAN K. B. ROBSON

apiculus lateral rounded to obscure, margin entire to subdenticulate, with a row of
inframarginal gland dots. Stamen fascicles each with 45—70 stamens, longest 9-12 mm
long, c. 0-6 X petals; anthers deep yellow. Ovary 6-7 x 4-5-5:5 mm, ovoid to
ellipsoid-ovoid; styles 5-S—7 mm long, 0-9-1 x ovary, erect, outcurved towards apex;
stigmas slightly capitate. Capsule c. 13 x 9-10 mm, + broadly ovoid. Seeds not seen.
2n=?

On rocky or grassy slopes, often in shade; 1800-2850 m.

Sikkim, Nepal (two areas: 82°-83°E and around 85°15’E). Map 11.

SIKKIM. Gangtok, between Orchid Sanctuary and Rumtok, Martam, 1800 m,
27.vi.1969, Hara, Kurosawa & Ohashi T. 701000 (TI).

NEPAL. Central: Leyley Deorali, 1800 m, 12.vii.1967, Manandhar 7106 (BM,
KATH). West: Bheri Valley, near Tarakot, 2610 m, 27.vi.1952, Polunin, Sykes &
Williams 2320 (BM).

Map 11 Sect. 3. Ascyreia: 8. H. williamsii @, 9. H. tenuicaule @.

H. williamsii is related to H. pachyphyllum and H. augustinii, having the same erect
habit and elliptic-oblong to oblong sepals; but the leaves are broader below the
middle and petiolate. In some respects (e.g. longer stamens and usually entire sepals
and petals) it is nearer to H. augustinii than is H. lobbii, but the thinner sepals with a
hyaline margin and the obtuse flower buds are more advanced characters.

9(20). Hypericum tenuicaule Hook. f. & Thomson ex Dyer

in Hook. f., Fl. Brit. Ind. 1: 254 (1874); R. Keller in Engler & Prantl, Nat.
Pflanzenfam. 2nd ed. 21: 176 (1925); N. Robson in Hara & Williams, Enum. fl. pl.
Nepal 2: 62 (1979). Type: Sikkim, Lachen valley, 2400 m, 3.viii.1849, J. D. Hooker
(K!, holotype; BM!, L!, isotypes).

H. patulum subsp. hookerianum var. tenuicaule (Hook. f. & Thomson ex Dyer)
Kuntze, Rev. gen. pl. 1: 60 (1891).

Norysca urala var. angustifolia Y . Kimura in Hara, FI. eastern Himal.: 210 (1966) pro
parte, incl. typum. Type: Nepal, Yamphodin to Ghatte, 2100-1600 m, 18.xi.1963,
Kanai, Murata & Togashi T.6304573 (TI!, holotype).

Icon: —.

STUDIES IN THE GENUS HYPERICUM L.

Shrub c. 1-5 m tall, bushy, with branches ascending to arching, not frondose. Stems
red, 4-lined when young, soon terete; internodes 8-14 mm, shorter than leaves; bark
reddish-brown to grey-brown. Leaves petiolate, with petiole 0-5—2 mm long; lamina
15-58 x 4-19 mm, lanceolate to rarely elliptic, acute to rounded, not or scarcely
apiculate, margin plane or rarely undulate, not incrassate, base cuneate to angustate,
+ glaucous beneath, chartaceous; venation: 2-3 pairs main laterals, with midrib
branched distally, without visible tertiary reticulum; laminar glands streaks and dots;
ventral glands dense. Inflorescence 1-4(—7)-flowered, from 1(2) node(s), corymbi-
form, with short (sometimes very short) apical internode, without flowering bran-
ches from middle of current stem; pedicels 9-15 mm long; bracts squamiform, acute,
deciduous. Flowers 15-30 mm in diam., + deeply cyathiform; buds subglobose,
obtuse to rounded. Sepals 4-S x 1-8-2-3 mm, free, subimbricate, equal, erect in bud
and fruit, oblong to oblong-spathulate, rounded or rarely with minute apiculus, with
margin entire or minutely eroded-denticulate (and then hyaline); midrib obscure or
indistinct, veins not prominent; laminar glands linear, numerous. Petals bright
yellow, sometimes tinged red, + incurved, 10-13 x 8 x 12 mm, c. 2-5 X sepals,
obovate to subcircular, with apiculus subterminal, rounded or almost absent; margin
entire, with or without a few intramarginal or marginal gland dots. Stamen fascicles
each with 30-35 stamens, longest 5—7 mm long, c. 0-5 X petals; anthers bright yellow.
Ovary 4-5 x 2:5—4 mm, ovoid; styles 3-5-6 mm long, 0-9-1-2 x ovary, erect, straight
or outcurved very near apex; stigmas small. Capsule 8-11 x 4-7 mm, + broadly
ovoid. Seeds dark brown, 0-7-0-8 mm, cylindric, not carinate, shallowly linear-
foveolate. 2n = ?

On ‘rocky bank under trees’ (Beer, Lancaster & Morris 9493); 1300-3150 m.

Sikkim, Nepal (east). Map 11.

SIKKIM. Kulhait, 2100 m, 11.x.1870, Clarke 13013B (K).

NEPAL. East: Zongi to Iladanda, 12.xi.1963, Kanai, Murata & Togashi T.
6304580 (BM, TI); between Tudam and Chyamtang, 3150 m, 12.xi.1971, Beer,
Lancaster & Morris 10734 (BM).

CULTIVATED. Specimens seen from England (1975), grown from seed ex Beer,
Lancaster & Morris 238 (Nepal (East): Bagang Khola).

H. tenuicaule resembles H. williamsti, but differs in its more delicate habit with
narrower, not or scarcely apiculate leaves and small flowers on longer pedicels. The
areas of these two species overlap in Sikkim only; H. tenuicaule occupies the ‘gap’ in
the distribution of H. williamsii between Sikkim and the Khatmandu region. It seems
clear that H. tenuicaule is derived from H. williamsii, or they have had a common
ancestor; but even in Sikkim they are apparently always mutually distinguishable.

H. tenuicaule has, however, been confused with H. uralum; these sympatric
species can always be differentiated by habit, stem-lines, and styles: frondose shoots,
4-lined stems, and styles 0.6—0.9 x ovary in H. uralum; divaricately branched shoots,
terete stems, and styles 0.9-1.2 x ovary in H. tenuicaule. The presence of resinous
glands on the underside of the leaf in the latter will immediately distinguish it from H.
hookerianum.

10(21). Hypericum lobbii N. Robson

inJ. Roy. Hort. Soc. 95: 496 (1970); Bean, Trees & shrubs hardy in Br. Isles 8th ed.,
2: 418 (1973); N. Robson in J. Jap. Bot. 52: 278, f. 2 (1977). Type: cultivated in
Hort. Veitch [Exeter and Chelsea] ex India, Khasia, Mufflong, coll. T. Lobb,
Herb. Hooker. (K!, holotype).

H. oblongifolium sensu Hook. f. in Curtis’s bot. Mag. 82: t. 4949 (1856), non Choisy
(1821).

H. hookerianum sensu hott. pro parte, non Wight & Arnott (1834).

Icon: Hook. f. in Curtis’s bot. Mag. 82: t. 4949 (1856).

Shrub 1-2-1-5 m tall, rather compact, with branches erect (or ascending ?), not

frondose. Stems reddish, slightly 4-angled and ancipitous when young, soon terete;
internodes 10-40 mm long, shorter than to exceeding leaves; bark reddish-brown.

223

224

’

+

NORMAN K. B. ROBSON

Leaves petiolate, with petiole 0-5-2 mm, long; lamina 20-45 x 9-25 mm, ovate-
lanceolate to triangular-ovate, obtuse to rounded, apiculate, margin plane, not
incrassate, base + broadly cuneate to shortly angustate, + glaucous beneath,
subcoriaceous; venation: 2(3) pairs main laterals, with midrib branched distally,
without distinct tertiary reticulum; laminar glands mostly dots (short streaks towards
midrib and base); ventral glands dense. Inflorescence 1—16(—c.24)-flowered, from
1-2 nodes, corymbiform, sometimes with short apical internode, without flowering
branches from middle of current stem; pedicels 4-6 mm long; bracts intermediate
(below first flower), narrowly elliptic, deciduous. Flowers 30-50 mm in diam.,
shallowly cyathiform; buds ovoid, subapiculate to obtuse. Sepals 7-11 x 5-9 mm,
free, imbricate, subequal or unequal, + spreading in bud and fruit, ovate-oblong to
broadly elliptic, rounded or usually apiculate, with margin entire or usually finely
eroded-denticulate, + hyaline, midrib of outer ones distinct, veins not prominent;
laminar glands linear, numerous. Petals golden yellow, not tinged red, slightly
incurved, 20-26 x 19-22 mm, 2-5-3 X sepals, broadly obovate, with apiculus,
subterminal, rounded; margin finely glandular-denticulate, with a row of inframar-
ginal gland dots. Stamen fascicles each with c. 50-60 stamens, longest 8-9 mm long, c.
0-35-0-4 x petals; anthers deep yellow. Ovary 5-6 x 44-5 mm, long, + broadly
ovoid; styles 4-5—5 mm long, 0-9 xX ovary, + erect, outcurved towards apex; stigmas
subcapitate or capitate. Capsule 9-12 x 7-9 mm, ovoid. Seeds not seen. 2n = ?

Habitat unknown; 1800-3600 m.

India (Meghalaya). Map 10.

INDIA. Meghalaya: Khasi Hills, Kurz 285 (BO, E).

CULTIVATED. Specimens seen from England (1853-1882), Germany (1886),
and U.S.A. (California) (1915).

H. lobbii appears to be a very local plant of the Khasi Hills. It is well described and
illustrated by J. D. Hooker (see above) from a cultivated plant, but seems to have
almost died out in Europe subsequently. Recent specimens that were thought to
belong to it have nearly all turned out to be H. forrestii. The tendency for the sepals
to spread is not so marked as in the next species, H. gracilipes (see Hooker’s
illustration), which is why confusion can occur with H. forrestii. The latter species,
however, has leaves that are not apiculate, sepals erect in bud and fruit, petals entire,
and stamens 0-4—0-6 xX petals.

H. lobbii is related to H. williamsii, from which it can be distinguished by the
spreading and usually apiculate sepals, the relatively shorter stamens, the broader,
denticulate petals, and the shorter ovary and styles.

11(22). Hypericum gracilipes Stapf ex C. Fischer

in Kew Bull. 1940: 32 (1940) [Parry, Lakhers: 587 (1932), nomen]. Type: India,
Mizoram, Lushai Hills, Neikdawn, 1200 m, vii.1926, Mrs N. E. Parry 74 (K!,
holotype).

Icon: —.

Shrub, c. 0-5 m tall, ‘undershrub’, with branches erect, not frondose. Stems reddish,
slightly 4-angled and ancipitous when young, soon terete; internodes 20-70 mm long,
shorter than to much exceeding leaves; bark grey-brown. Leaves petiolate, with
petiole 0-5-2 mm long; lamina (15) 23-60 x (6)9-25 mm, ovate to lanceolate, acute
to obtuse or rounded-apiculate, margin plane, not incrassate, base narrowly to
broadly cuneate or shortly angustate, + glaucous beneath, chartaceous; venation:
2(3) pairs main laterals, with midrib obscurely branched distally, usually without
distinct tertiary reticulum; laminar glands dots and sometimes a few short streaks,
ventral glands dense. Inflorescence 3—10-flowered, from 1-3 nodes, corymbiform,
with short apical internode, without flowering branches from middle of current stem;
pedicels 5-10 mm long; bracts linear-lanceolate to linear, deciduous. Flowers 25-35
mm in diam., cyathiform; buds + narrowly ovoid, acute or subacuminate. Sepals 5—9
x 1-3(4) mm, free, subimbricate or open, subequal, + spreading in bud and fruit,
narrowly elliptic to linear-oblong, acute to acuminate, with margin entire; midrib

STUDIES IN THE GENUS HYPERICUM L.

usually distinct, veins not prominent; laminar glands linear or interrupted, 4-10.
Petals golden yellow ?, not tinged red, + incurved, 2-18 x 19-15 mm, c. 2-3 x
sepals, broadly obovate, with apiculus subterminal, rounded; margin very finely
glandular-denticulate (or entire ?), with a few inframarginal gland dots. Stamen
fascicles each with c. 50 stamens, longest 5—6 mm long, c. 0:25-0:35 x petals; anthers
deep yellow. Ovary 3-5-5 x 2-3 mm, ovoid to ellipsoid; styles 34 mm long, c. 0-8 x
ovary, suberect, outcurved towards apex; stigmas subcapitate. Capsule 9-13 x 5—7
mm, ellipsoid. Seeds dark reddish-brown, c. 1 mm long, cylindric to ovoid-cylindric,
scarcely carinate, very shallowly linear-reticulate. 2n = ?

Shaded (sandstone) banks and streamsides; 1200-2250 m.

Bangladesh (Chittagong), India (Mizoram, Manipur, Nagaland, west Bengal). Map
10.

BANGLADESH. Chittagong: Chittagong Division, Cowan (E).

INDIA. Mizoram: Lushai Hills, N. Vaulaiphai, 1500 m, xii.1927, Parry 62a (K).
Manipur: Sirhoi, 1800 m, 16.vii.1948, Kingdon Ward 17794 (BM). Nagaland:
Kedima, 1800 m, 20.vii.1935, Bor 6313 (K). West Bengal: Darjeeling, Ghum to
Lopchu, 2100 m, 19.vi.1960, Kenai, Murata & Togashi 2947 (K),; Ghum, 2250 m,
4.vii.1919, Cave (E).

H. gracilipes is like a small, scrubby version of H. /obbii, differing in size of flower, in
the much narrower, acute sepals and in the narrower and usually smaller leaf (Bor
6313 has large leaves — to 60 X 25 mm). The habit becomes scrubbier and the parts
smaller (in general) northwards, the plants from near Darjeeling being rather like H.
uralum apart from the narrow, acute sepals and non-frondose habit. They have been
included in H. gracilipes here, but study of further material may show them to be
taxonomically distinct.

12(23). Hypericum gaitii Haines

in J. & Proc. asiat. Soc. Beng. I, 15: 311, t. 10 (1920). Type: India. Bihar, on the
plateau (‘pats’) of Chota Nagpur, Neterhat, v.1918, Haines 4327 (K!, holotype).

Icon: Haines in J. & Proc. asiat. Soc. Beng. I, 15: 311, t. 10 (1920).

Shrub 0-7-1-5 m tall, bushy, with branches erect to ascending. Stems red to
yellow-brown, 4-lined and ancipitous when young, soon terete; internodes c. 10-40
mm long, shorter than leaves; bark grey-brown. Leaves sessile; lamina 30-90 x 8-30
mm, elliptic or oblong-elliptic to oblanceolate, acute to obtuse-apiculate, margin
plane, base cuneate, paler beneath, + glaucous, subcoriaceous to chartaceous, lower
ones eventually deciduous; venation: 3(4) pairs main laterals, branched, the midrib
pinnately branched, with tertiary reticulum rather dense and visible beneath;
laminar glands very short streaks and small dots; ventral glands absent. Inflorescence
(1)3-7-flowered, from apical node, subcorymbiform; pedicel 4-25 mm long (in
fruit); bracts foliar (persistent) to small, elliptic (deciduous). Flowers (20—)40-55(-
65) mm in diam., stellate; buds ovoid ?, subacute ? Sepals (6-)9-13 x 4-6 mm, free,
imbricate, subequal, erect in bud and fruit, ovate-lanceolate to narrowly elliptic,
acute to rounded, margin entire or irregularly and minutely denticulate, midrib
obscure; other veins not prominent; laminar glands linear to striiform, numerous.
Petals bright yellow, not tinged red, spreading, 25-30 x 15-20 mm, 34 x sepals,
narrowly obovate, with apiculus subterminal, rounded; margin entire, eglandular.
Stamen fascicles each with c. 30 stamens, longest 15-18 mm long, c. 0-5-0-65 x
petals; anthers yellow. Ovary 6-8 x 4-5 mm, ovoid; styles 10-12 mm long, 1-5S—1-7 x
ovary, free, suberect with apex outcurved; stigmas small. Capsule 13-17 x 7-9 mm,
narrowly ovoid-conic. Seeds dark reddish-brown, 10-12 mm long, cylindric, slightly
carinate, shallowly linear-foveolate to almost ribbed-scalariform. 2n = ?

In shade on cliffs and along streams; 900-1560 m.

India (Bihar, Orissa). Map 8.
INDIA. Bihar: Netarhat, 900 m, v.1918, Haines 4327 (K); Surguja State, Nawa-
dih, 1110 m, 10.v.1940, Mooney 1298 (K); Surguja State, Mainpat, 24.v.1938,

225

226

«

NORMAN K. B. ROBSON

Mooney 780 (K). Orissa: Pal Lahara State, Malyagiri Hill, 1140 m, 12.vi.1937,
Mooney 458 (K); Koraput Distr., near Pottangi, Turia Konda, 1410-1560 m,
10.x.1950, Mooney 4077 (K).

H. gaitii, which was described by Haines as intermediate between H. mysurense and
H. cernuum (i.e. oblongifolium), is indeed both morphologically and geographically
intermediate between these species, having a scattered distribution in areas of high
ground in the north-east of peninsular India. It is separated from the areas of H.
mysurense by the dry region of the Deccan and from that of H. oblongifolium by the
Ganges valley. As the high ground in the south of Orissa continues into Andhra
Pradesh, H. gaitii may well occur in that province. It is also related to H. monogynum
(q.v.), which is geographically separated from it, not only by the Ganges Valley, but
also by the Himalayan massif and the Tibetan plateau. H. gaitii differs from H.
mysurense by the stem internodes, which are longer and soon terete, by the larger
leaves with more densely reticulate venation and by the longer narrower sepals,
rounded petal apiculus, relatively shorter styles and larger capsule. From H.
oblongifolium it is distinguished by the narrower leaves, larger sepals erect in bud
and fruit, relatively shorter stamens, relatively shorter styles and narrower capsule.

13(24). Hypericum oblongifolium Choisy

Prodr. monogr. fam. Hypéric.: 42, t. 4 (1821); in DC., Prodr. syst. nat. regni veg. 1:
545 (1824); N. Robson inJ. Roy. Hort. Soc. 95: 489 (1970), in K. H. Rechinger, FI.
iranica 49: 4 (1968); in Nasir & Ali, Fl. W. Pakistan 32: 3, f. 1 I-L (1973), in Hara &
Williams, Enum. Fl. Pl. Nepal 2: 62 (1979). Type: ‘Ex Indiis orientalibus’,
[Hardwicke ?| (G-DC, holotype!; BM!).

Hypericum sp. sensu Hardwicke in Asiatick Researches 6: 369 (1801).

H. cernuum Roxb. [Hort. Bengal.: 59 (1814) nomen] ex D. Don., Prodr. fl. nepal.:
218 (1825); Roxb., Fl. indica 3: 400 (1832); Cambess. in Jacquem., Voy. Inde 4:
30, Atlas 2: t. 33 (1844); Dyer in Hook. f., Fl. Brit. Ind. 1: 253 (1874); Collett, Fl.
simlensis: 555 (1902); Banerjiin J. Bombay Nat. Hist. Soc. 51: 774 (1953), in Rec.
Bot. Surv. Ind. 19(2): 27 (1966). Type: India, cult. ex Kashmir, ‘Between Burdwar
and Shreenagar’, seed ex Hardwicke, 1797, Roxburgh (CAL, holotype).

H. speciosum Wallich, Numer. list: No. 4803 (1831), nomen.

Norysca cernua (Roxb. ex D. Don) J. Voigt, Hort. suburb. Calcutta: 90 (1845); K.
Koch, Hort. dendrol.: 66 (1853); Blume, Mus. bot. Lugd. Bat. 2: 22 (1856).

Norysca oblongifolia (Choisy) Blume, Mus. bot. Lugd. Bat. 2: 22 (1856) pro parte,
excl. syn. H. uralum, H. patulum; Y. Kimura in Nakai & Honda, Nova fl. jap. 10:
102 (1951).

H. patulum var. oblongifolium (Choisy) Koehne, Deutsch. Dendr.: 415 (1893) pro
parte quoad typum.

H. aitchisonii J. R. Drumm. ex R. Keller in Bot. Jahrb. 58: 191 (1923) in synon.;
R. Keller in Engler & Prantl, Nat. Pflanzenfam. 2nd ed. 21: 176 (1925), nomen.

H. cernuum var. typicum R. Keller in Bot. Jahrb. 58: 191 (1923), in synon.

H. cernuum var. lanceolatum R. Keller in Bot. Jahrb. 58: 191 (1923), in synon.

Icon: Choisy, Prodr. monogr. fam. hypéric.: t. 4 (1821).

Shrub 0-45—1-2(—2-4) m tall, much branched, with branches spreading or drooping,
sometimes straggling or semi-scandent. Stems red, 4-lined and ancipitous when
young, soon terete; internodes 15-50 mm long, shorter than or rarely exceeding
leaves; bark grey. Leaves sessile; lamina 30-93 x 10-42 mm, elliptic to oblong or
ovate-oblong, obtuse or apiculate to rounded, margin plane, base cuneate to
rounded, paler beneath, sometimes glaucous, chartaceous, lower ones eventually
deciduous; venation: 3—4(5) pairs main laterals, branched, the midrib pinnately
branched, with tertiary reticulum lax to dense, conspicuous; laminar glands short
lines or streaks or dots in the reticular areolae; ventral glands dense to sparse.
Inflorescence 1-3(-8)-flowered, from apical node, corymbiform, sometimes also on
short lateral shoots from up to c. 16 nodes towards base of current growth, then
cylindric-ellipsoid; pedicels 4-20 mm long; bracts intermediate, persistent to small,

STUDIES IN THE GENUS HYPERICUM L.

lanceolate, deciduous. Flowers 35-65(—75) mm, in diam., stellate; buds ovoid,
obtuse to rounded. Sepals 5-8 x 2-5-5 mm, free, imbricate, subequal to somewhat
unequal, erect in bud, + ascending in fruit, narrowly ovate to elliptic-oblong, acute
to rounded, margin entire or sometimes minutely denticulate near apex; midrib
obscure but apex sometimes incrassate, other veins not prominent; laminar glands
numerous, linear. Petals bright yellow to bright yellow-orange, not tinged red,
spreading, 20-30 x 10-16 mm, c. 4 X sepals, obovate to oblanceolate, with apiculus
lateral, rounded, margin entire, eglandular. Stamen fascicles each with c. 30 stamens,
longest 15—20 mm long, c. 0-7 x petals, anthers yellow. Ovary 4-7 x 3-5-5 mm,
ovoid-pyramidal to broadly ovoid; styles 9-14 mm long, c. 2-3 x ovary, free, erect or
slightly apically outcurved; stigmas small to subcapitate. Capsule 9-17 x 7-11 mm,
ovoid to ovoid-conic. Seeds dark reddish-brown, 1-1-2 mm long, cylindric or
cylindric-ovoid, narrowly carinate, very shallowly linear-foveolate to ribbed-
scalariform. 2n = 24, 44, 46, 48.

In damp or sheltered cliff crevices, in thickets or Quercus forest, or on grassy or stony
hillsides; (600)750—2700 m.

Pakistan (north), India (Kashmir to Kumaun), Nepal (west, central). Map 12.

PAKISTAN. Kurram: Kurram Valley near Kurram, iv.1879, Aitchison 65 (DD,
K). Swat: Swat, iv.1949, Herb. Stewart. 70 (RAW), Hazara: Kagan, Malkandi,
1350 m, 20.v.1897, Duthie (K, RAW). Rawalpindi: Murree, Brewery Estate near
Octrol, 8.v.1975, Dar, Amin & Safraz Khan 494 (BM). Jhelum: Shahpur, Salt
Range, Sakesar, 7.iv.1902, Drummond 13820 (E, K). Azad Kashmir: above Palan-
dri, 1440 m, 13.iv.1969, Ali & Nasir 5481 (RAW).

INDIA. Kashmir: Jelum Valley, Uri, 1200 m, Ludlow & Sherriff 7588 (BM).
Himachal Pradesh: Kangra, Bhadwar, 600 m, 6.v.1933, Koelz 4405(A); Simla,
Ushan Valley, 1260 m, 25.iv.1888, Watt 8886(E). Uttar Pradesh: Mussoorie, Song
Khad, 1920 m, 5.v.1948, Fleming 427(A); Garhwal, Westufer des Naini Tal, 2100 m,
8.v.1961, Krdusel (FR).

NEPAL. West: Tila Valley, Kalikot, 1650 m, 20.iv.1950, Polunin, Sykes &
Williams 1929 (BM, L). Central: Ganesh Himal, Gatlang, 1950 m, 26.iv.1952,
Stainton 3627 (BM, E).

CULTIVATED. Specimens seen from England ex Kashmir, Srinagar (1983).

Despite the recorded variation in chromosome number from diploid (2n = 24) to
tetraploid (2n = 44, 46, 48), morphological variation in H. oblongifolium does not
appear to be great. Stem and leaves tend to be more glaucous and parts somewhat
smaller towards the western (drier) end of its range, but there is no evidence of
morphological discontinuity. It is separated geographically from its nearest relatives

Su ~~ se)

Map 12 Sect. 3. Ascyreia: 13. H. oblongifolium ©, 15. H. griffithii @.

220

228

NORMAN K. B. ROBSON

by disjunctions of varying size, viz. to the south of the Ganges Valley (H. gaitii), to
northern Turkey (H. calycinum) and to Bhutan (H. griffithii).

Although Hardwicke (1801) collected this species from ‘between Dosay and
Bedeyl’ in 1796 and sent seed to Calcutta Botanic Garden in 1797 from ‘between
Hurdwar and Shreenagur [Srinagar]’, Roxburgh named it H. cernuum only in 1814
and did not himself provide a valid description until 1832, by which time David Don
(1825) had validated this name. Meanwhile, Choisy (1821) had described a specimen
sent from Herb. Lambert (BM) to Geneva in 1816 as H. oblongifolium. This
specimen is annotated ‘Nidas’ or ‘Sides’. This could well be ‘Sides’, because
Hardwicke 4 (BM) is labelled ‘Sides of high Mountains’. The BM specimen is thus
likely to be part of the type collection.

14(25). Hypericum calycinum L.

Mant. pl. 1: 106 (1767); Curtis, Bot. Mag. 5: t. 146 (1796); J. E. Sm., Eng. Bot. 29: t.
2017 (1809), in Trans. Linn. Soc. Lond. 10: 266 (1811); Choisy, Prodr. monogr.
fam. Hypéric.: 42 (1821); in DC., Prodr. syst. nat. regni veg. 1: 546 (1824);
Sibthorp & Lindley, Fl. graeca 8: 52, t. 771 (1833); Treviranus, In Hyper. gen.
spec. animady.: 8 (1861); Boissier, Fl. orient. 1: 789 (1867); Britten in J. Bot.,
Lond. 53: 68 (1915); R. Keller in Engler & Prantl, Nat. Pflanzenfam. 2nd ed. 21:
176 (1925); Stefanoff in God. Agr.-les. Fak. Univ. Sofiya 11: 143 (1933), in
Pflanzenareale: Karte 2a (1933); Czeczott in Mitt. Kénigl. Naturwiss. Inst. 10: 59,
f. 14 (map) (1937); Gorshkova in Shishkin & Bobrov, Fl. U.R.S.S. 15: 211 (1949);
Milano in Publs. Inst. bot. agric. Argentina 8 (128): 16, f. 4 (1961); Salisbury in
Watsonia, 5: 368 (1963); Yakar, Renk. Turk. Bitk. Atlasi 2: t. 30 (1965); N.
Robson in Davis, Fl. Turkey 2: 365, map 20 (1967), in Tutin et al., Fl. europaea 2:
263 (1968), in J. Roy. Hort. Soc. 95: 489 (1970); Jordanov & Kozuharov in
Jordanov, Fl. Rep. Pop. Bulg. 4: 229, t. 43 f. 2 (1970); [Lancaster], Hillier’s man.
trees & shrubs: 150 (1971); Bean, Trees & shrubs hardy in Br. Isles 8th ed. 2: 410
(1973). Type: Turkey or cultivated (‘Habitat in America septentrionalis’, lapsus
calami), Herb. Linn. 943/7 (LINN!, lectotype); ex Herb. Cliff., Herb. Linn. 943/6
(LINN!, syntype).

H. ascyron sensu Miller, Gard. dict. 7th ed.: no. 7 (1759) 8th ed.: no. 7 (1768), non L.
(1753).

Ascyron coriaceum Moench, Meth. bot.: 130 (1794), nom. illegit. Type as for H.
calycinum L.

H. venosum Lam., Encycl. méth. (Bot.) 4: 146 (1797). Type: Herb. Juss. (P).

H. grandiflorum Salisb., Prod. stirp. horto Chapel Allerton: 369 (1798), nom. illegit.
Type as for H. calycinum.

H. calycinum var. B sensu Choisy, Prodr. monogr. fam. Hypéric.: 43 (1821).

Ascyrum calycinum (L.) Poiret, Tabl. encycl. méth. 3: 199, t. 642 ff. 1-22 (1823)
(‘calicinum’).

H. calycinum var. acutifolium Choisy in DC., Prodr. syst. nat. regn. veg. 1: 546
(1824). Type: ‘Hyper. florib.’, 1807, Herb. Montbret. ? (G-DC!).

Eremanthe calycina (L.) Spach, Hist. nat. vég. Phan. 5: 422 (1836), in Annls Sci. nat.
(Bot.) II, 5: 363 (1836); K. Koch, Hort. dendrol.: 65 (1853).

Androsaemum calycinum (L.) C. Presl ex Steudel, Nomencl. bot. 2nd ed. 1: 94
(1840); Clemoncet in Bull. Ass. Nat. Vallée Loing Massif Fontainebleau 42: 13
(1966).

Eremanthe venosa (Lam.) K. Koch, Hort. dendrol.: 65 (1853).

Norysca calycina (L.) Blume, Mus. bot. Lugd. Bat. 2: 22 (1856); Y. Kimura in Nakai
& Honda, Nova fl. jap. 10: 98 (1951).

Norysca venosa (Lam.) Blume, Mus. bot. Lugd. Bat. 2: 22 (1856).

Icon: Curtis, Bot. Mag. 5: t. 146 (1796).

Shrub 0-2-0-6 m tall, evergreen, with creeping branching stolons and erect, usually
unbranched stems. Stems brownish-orange to red, 4-lined, ancipitous when young;
internodes 20-60 mm long, shorter than leaves; bark grey. Leaves subsessile or with
petiole up to 2 mm long; lamina 45—95(-—104) x 15-39(-45) mm, oblong to elliptic or

STUDIES IN THE GENUS HYPERICUM L.

+ narrowly ovate, obtuse to apiculate, margin plane, base cuneate to rounded, paler
beneath, not glaucous, coriaceous, lower ones eventually deciduous; venation: 3-4
pairs main laterals, not clearly distinct from midrib branches, with tertiary reticulum
dense, conspicuous especially beneath; laminar glands very short streaks or dots in
the reticular areolae; ventral glands absent. Inflorescence 1(2-3)-flowered; pedicels
absent or 10-12 mm long; bracts small, lanceolate, persistent. Flowers 50-80(—95)
mm in diam., stellate; buds broadly ovoid to globose, rounded. Sepals 10-20 x 7-15
mm, free, imbricate, markedly unequal, erect in bud, + ascending in fruit, broadly
elliptic to subcircular or obovate, + cucullate, margin entire or minutely and
remotely denticulate, midrib not distinct, veins not prominent; laminar glands
linear, distally dissected, numerous. Petals bright yellow, not tinged red, spreading,
25-40 x 13-22 mm, 2-2-5 x sepals, obovate to oblanceolate, with apiculus lateral,
rounded, margin entire or minutely and remotely denticulate, eglandular. Stamen
fascicles each with 90-120 stamens, longest 20-30 mm long, c. 0-75 X petals; anthers
red. Ovary 5-9 x 3-5—6:5 mm, ovoid to ovoid-conic; styles (5 or rarely 4) 12-20 mm
long, 1-5-3 x ovary, free, erect or spreading; stigmas small. Capsule 10-20 x 10-13
mm, ovoid to ovoid-conic. Seeds reddish-brown, 1-5-2 mm long, cylindric, not or
narrowly carinate, shallowly linear-reticulate to -foveolate. 2n = 20.

In shady woods (especially Quercus and Fagus) and on dry or damp shaded banks;
30-1200 m.

Bulgaria (south-east), Turkey (north, along the Black Sea eastward to Trabzon).
Widely cultivated and often naturalised. Records from west Transcaucasia (e.g. by
Ledebour, Boissier, and Gorshkova) are apparently based on introductions. Map
13:

BULGARIA. Burgas: Strandzha Mts, 100 m, (fide Fl. Rep. Pop. Bulg.)

TURKEY. Kirklareli: Strandscha-dagh [Instranca daglari], c. 100-500 m,
5.vii.1927, Mattfeld 3443 (A). Istanbul: (Europe) Sariyer, 21.vi.1953, Baytop 2761
(BM, H); (Asia) Polonesk6y, 7.ix.1939, Post (G-P). Kocaeli: Marmara Sea between
Izmit and Yalova, 14.ix.1957, K. H. Rechinger 15333 (W). Bursa: Mt Olympus [Ulu
dag] supra Brussa [Bursa], 200 m, 21.v.1899, Bornmiiller 4232 (E); Ulu dag,
2.vii.1972, Himmetoglu H8 (BM). Balikesir: Mt. Ida [Kaz dag] circa font. Scaman-
dri, 29.vii.1883, Sintenis 407 (A, BM, E, G, JE, K, S, U). Sakarya: Karasu to

229
fe)

:
:

V

~

Map 13 Sect. 3. Ascyreia: 14. H. calycinum @, O (unlocalised); also in vil. Trabzon (see Map 33).

230

NORMAN K. B. ROBSON

Sogiitlu, 30 m, 7.viii.1962, Davis & Coode D. 39087 (E, K). Bolu: Diizce to
Akcakoca, 320 m, 14.vii.1962, Davis & Coode D. 38517 (E, K). Zonguldak: 9 km S.
Devrek, c. 300 m, 8.vi.1962, Sorger T62/85/1 (E). Kastamonu: between Inebolu and
Bolu, Bernhard (fide Fl. Turkey). Trabzon: Bender Erekli, Handel-Mazzetti (fide Fl.
Turkey).

CULTIVATED. Specimens seen from the British Isles (all countries), France,
Germany, Switzerland, Austria, Sweden, Finland, U.S.A.

NATURALISED. Specimens seen from the British Isles (all countries), France,
Italy, Yugoslavia (Istria), Greece, U.S.S.R. (Krym, Georgia). Recorded from
Argentina.

H. calycinum is a popular species in cultivation because it tolerates dry shady
conditions, although it flowers better in more open situations. Its flower is one of the
largest in the genus. It was first introduced into England from near Istanbul
(probably from the Belgrade Wood) by Sir George Wheeler in 1676 and soon
escaped from cultivation (Britten, 1915).

H. calycinum appears to spread by seed as well as vegetatively in its native area;
but in the British Isles its reproduction is almost wholly vegetative. It is self-
compatible, and Salisbury (1963) suggested that this is due to (i) degrees of failure in
embryo development and (ii) inefficient endosperm nutrition before the resting stage
is reached, both conditions resulting from too low temperatures during pollination
and fertilisation. Likewise, the absence of sun-baking during capsule maturation
would be deleterious.

The isolation of H. calycinum from near relatives precludes natural hybridisation;
but artificial hybrids have been made with H. patulum (H. x moserianum Luquet ex
André, q.v.), H. kouytchense and H. forrestii (see Robson, 1981: 170), and also H. x
cyathiflorum.

15(26). Hypericum griffithii Hook. f. & Thomson ex Dyer

FI. Brit. Ind. 1: 253 (1874); Biswas in Bull. bot. Surv. India 13: 160 + f. (1973); Sahni
& Naithani in Indian Forester 106: 865-868 (1981); Robson & Long in Notes Roy.
bot. Gdn Edinb. 41: 136, f. 3 (1983). Type: Bhutan: descent to Dimree Nuddee, 1920
in 1862-1863, Griffith 833 = EIC 1820 (K!, holotype; BM!, CAL, K!, isotypes).
Icon: Biswas in Bull. bot. Surv. India 13: 160 (1973).

Shrub 0-9-3 m tall, deciduous, with branches erect. Stems dark reddish-brown,
2-lined and subancipitous when young, soon terete; internodes 20-35 mm long,
shorter than leaves; bark reddish-brown. Leaves sessile, amplexicaul; lamina
(40-)45-117 x (27-)37-60 mm, ovate-oblong to ovate, obtuse-apiculate (or rarely
acute) to rounded, margin plane, base cordate, not or scarcely paler beneath, not
glaucous, chartaceous; venation: (6-)7—8(-9) pairs main laterals, branched, the
midrib pinnately branched, with tertiary reticulum very dense, prominent on both
sides or beneath only; laminar glands punctiform, large and small in the reticular
areolae; ventral glands absent. Inflorescence 5—20-flowered, from apical node,
subcorymbiform, sometimes also from second node and on lower short lateral
shoots, the whole then pyramidal; pedicels 8-20 mm long; bracts small, lanceolate,
deciduous. Flowers c. 35 mm in diam., stellate; buds conic, acute. Sepals 2-5-S x
1-5-2 mm, free, imbricate, subequal, erect to outcurved in bud, + spreading in fruit,
ovate to lanceolate or oblong-elliptic, acute, sometimes apiculate, with apex incras-
sate, margin entire, midrib distinct or not, other veins not prominent; laminar glands
few, striiform to punctiform. Petals golden yellow, not tinged red, spreading,
(13-)15-20 x 6-10 mm, 5-6 X sepals, obovate to oblanceolate, with apiculus lateral.
acute; margin entire, eglandular. Stamen fascicles each with 20-35 stamens, longest
(8-)12-15 mm, 0-7-0-8 X petals; anthers yellow. Ovary 4-5 Xx 2:5 mm, narrowly
ovoid to ovoid-ellipsoid; styles 8-10 mm long, 1-4-2 xX ovary, free but sometimes
coherent in lower half, erect, apically sometimes slightly spreading; stigmas subcapi-
tate. Capsule 9-10 x 7mm, narrowly ovoid to ellipsoid. Seeds dark reddish-brown, c.

0-9 mm long, cylindric, narrowly carinate, linear-foveolate to ribbed-scalariform. 2n
=?

STUDIES IN THE GENUS HYPERICUM L.
On dry hillsides and in scrub jungle along river bank; (900?) 1050-1920 m.

Bhutan (south and central), Arunachal Pradesh (west), see Robson & Long
(1983): fig. 3. Map 12.

BHUTAN. Central: Descent to Dimree Nuddee, 1920 m, 1862-63], Griffith 833
(K); Bhutan, [1862-63], Griffith 1820 (BM, K); Kuru Chu Valley, 1500 m,
28.viii.1915, Cooper 4700 (BM); Rungzhung, Gamri Chu, 1050 m, 18.iii.1936,
Ludlow & Sherriff 1198 (BM, E); Ghunkara, Trashigong Chu, 1200 m, 30.v.1947,
Ludlow, Sherriff & Hicks 12576 (A, BM); below Mongar, 1700 m, 16.vi.1979,
Grierson & Long 1990 (BM). South: Samchi district, Torsa valley, 900-1200 m,
iv.1905, G. L. Searight 108 (CAL n.v.).

INDIA. Arunachal Pradesh: Kameng district, near Jamiri, 1200 m, iv. & x.1977,
Sahni & Naithani Ser.II, 546 (DD, n.v.) Bennet & Naithani 3244 (DD n.v.).

H. griffithii is apparently restricted to two disjunct areas (south-western Bhutan and
east-central Bhutan/western Arunachal Pradesh). In Bhutan it was not collected
between 1862-63, when Griffith found it, and 1915. Both the Ludlow & Sherriff and
Grierson & Long expeditions rediscovered it by retracing Griffith’s route.

It is clearly intermediate between the more erect, broad-leaved forms of H.
oblongifolium (of Nepal) and the more primitive species of sect. 5. Androsaemum
(H. grandifolium, H. hircinum), which differ essentially in having a trimerous
gynoecium.

16(27). Hypericum monogynum L.

Spl. pl. 2nd ed.: 1107 (1763); Miller, Gard. dict. [7th ed.: No. 11 (1759) nom. illegit. ]
8th ed.: no. 11 (1768); Thunb., Fl. Jap.: 297 (1784); Curtis, Bot. Mag.: t. 334
(1796); Blume, Bijdr. Fl. Ned. Ind. 1: 141 (1825); N. Robson in Blumea 20: 251
(1973); Bean, Trees & shrubs hardy in Br. Isles 8th ed. 2: 419 cum fig. (1973); N.
Robson in Nasir & Ali, Fl. W. Pakistan 32: 3 (1973); in Fl. malesiana I, 8: (1974).
Type: Miller, Figures plants: 101, t. 151, f. 2 (1760) (BM!, typotype).

H. chinense L., Syst. nat. 10th-ed. 2: 1184 (1759), Diss. exhib. C. N. Hellenius: 6
(1776), Amoen. Acad. 8: 323 (1785); Trew, Pl. rar.: t. 21, fig. II (1784); Lam..,
Encycl. méth. (Bot.) 4: 144 (1797); Willd., Sp. pl. 3 (2): 1449 (1803); Choisy,
Prodr. monogr. fam. Hypéric.: 40 (1820); in DC., Prodr. syst. nat. regni veg. 1: 545
(1824); G. Don, Gen. Syst. 1: 602 (1831); Bunge in Mem. Acad. Sci. St. Petersb. 2:
84 (1833); Choisy in Zollinger, Syst. Verz. Ind. Archipel. 2: 150 (1857), Pl. javan.:
5 (1858); K. Koch, Hort. dendrol. 1: 494 (1869); Hance in J. Bot., Lond. 17: 81
(1879); Ito & Kaku, Koisikawa-Syokubutu-Somoku-Zusetu 2: 17, t. 17 (1884);
Hemsley in J. Linn. Soc. 23: 72 (1886); Anon in Gdnrs’ Chron. II, 1: 705, f. 135
(1887); Koehne, Deutsch. Dendrol: 415 (1893); Diels in Bot. Jahrb. 29: 475 (1900);
R. Keller in Bot. Jahrb. 32: 548 (1904); Matsumura & Hayata, Enum. pl.
Formosa: 60 (1906); Léveillé in Bull. Soc. bot. Fr. 53: 499 (1906), 54: 590 (1908);
Schneider, ///. Handb. Laubholzk. 2: 337, f. 224 (1909); Hayata, Icon. pl. formos.:
78 (1911); Matsumura, /nd. pl. Jap. 2: 365 (1912); Rehder in Sargent, Pl. Wils. 2nd
ed. 2: 404 (1915); Makino & Nemoto, Fl. Jap: 540 (1925); R. Keller in Engler &
Prantl, Nat. Pflanzenfam. 21: 176 (1925); Chen, III. man. chin. trees & shrubs: 841,
t. 735 (1937); Rehder, Man. cult. trees & shrubs 2nd ed.: 639 (1940); Backer, Fl.
Java 1: 382 (1963); N. Robson in J. Roy. Hort. Soc. 95: 489 (1970); [Lancaster],
Hilliers’ man. trees & shrubs: 150 (1971); Knight in J. Roy. Hort. Soc. 96: 38
(1971), non Osbeck (1757) nec. Retz. (1789). Type as for H. monogynum L.

H. aureum Lour., Fl. cochinch.: 472 (1790). Type: China, ‘Habitat incultum prope
Cantonem Sinarum’, Loureiro (no specimen traced).

Norysca chinensis (L.) Spach, Hist. nat. vég. Phan. 5: 427 (1836), in Annls Sci. Nat.
(Bot.) IT, 5: 364 (1836); J. O. Voigt, Hort. suburb. Calcutta: 90 (1845); K. Koch,
Hort. dendrol.: 65 (1853); Blume, Mus. bot. Lugd. Bat. 2: 22 (1856); Miquel, Fi.
Ind. bat. 1 (2): 514 (1859); Y. Kimura in Nakai & Honda, Novafl. jap. 10: 103, f. 42
(1951).

H. chinense var. B sensu Hooker & Arnott, Bot. Beechey’s voy.: 172 (1840).

H. salicifolium Sieb. & Zucc. in Abh. K. Akad. Wiss: Munch. 4 (2): 162 (1843); R.

Zoi

Zoe

NORMAN K. B. ROBSON

Keller in Bot. Jahrb. 33: 548 (1904); H. Léveillé in Bull. Soc. bot. Fr. 53: 499
(1906), R. Kellerin Engler & Prantl, Nat. Pflanzenfam., 2nd ed. 21: 176 (1925); Y.
Kimura in Bot. Mag., Tokyo 54: 88 (1940). Type: Japan, ‘Crescit in Miaco,
Osakka, alibi ob pulchritudinem florum’, Thunberg (UPS).

Norysca salicifolia (Sieb. & Zucc.) K. Koch, Hort. dendrol.: 65 (1853); Blume, Mus.
bot. Lugd. Bat. 2: 23 (1856).

H. chinense var. salicifolium (Sieb. & Zucc.) Choisy in Zoll., Syst. Verz. Ind.
Archipel. 1: 150 (1854) (‘salicifolia’).

Norysca aurea (Lour.) Blume, Mus. bot. Lugd. Bat. 2: 22 (1856).

Norysca punctata Blume, Mus. bot. Lugd. Bat. 2: 23 (1856). Type: ‘Hypericum
chinense (haud Linn.) Chois. in De Cand. Prodr. I: 545 (excl. Syn) Spr. Syst. Veg.
III 341. 5’. (Blume was misled by Choisy’s description of the leaves and calyx as
black-dotted.)

H. monogynum var. salicifolium (Sieb. & Zucc.) [Maxim. in Bull. Acad. Petersb. 27:
428 (1881), in synon.] André in Revue hort. 61: 464 (1889), in adnot..

H. chinense subsp. salicifolium (Sieb. & Zucc.) Kuntze, Rev. gen. pl. 1: 60 (1891).

H. chinense subsp. latifolium Kuntze, Rev. gen. pl. 1: 60 (1891). Type: Taiwan,
‘Formosa’ [Oldham 29 ?] (NY, holotype; BM!, K!).

H. chinense subsp. obtusifolium Kuntze, Rev. gen. pl. 1: 60 (1891). Type: India,
‘Calcutta’ [probably Hort. Bot. Calcutta, Wallich 4819 C (K!, holotype; BM!)].
H. prattii sensu Rehder in Sargent, Pl. Wils. 2: 404 (1915) pro parte quoad Wilson

1604, 2420.

Komana salicifolia (Sieb. & Zucc.) Y. Kimura ex Honda, Nomina pl. jap.: 222
(1939); Hisauchi, Kika-syokubutsu: 180 cum fig. (1950), comb. invalid.

Norysca chinensis var. salicifolia (Sieb. & Zucc.) Y. Kimura in Nakai & Honda,
Nova fl. jap. 10: 107 (1951).

Icones: Curtis, Bot. Mag.: t. 334 (1796); Y. Kimura in Nakai & Honda, Nov. fl. jap.
10: 105, f. 42 (1951); Fig. 15.

Shrub 0-5-1-3 m tall, bushy or usually with branches lax, spreading. Stems red,
2(4)-lined and ancipitous when young, soon terete; internodes 10-50 mm long,
shorter than leaves; bark orange-brown. Leaves sessile or with petiole up to 1-5 mm
long; lamina 20-112 x 10-41 mm, oblanceolate or elliptic to oblong or more rarely
lanceolate to ovate-triangular or ovate, acute to rounded, often minutely apiculate,
margin plane, base cuneate to rounded or in upper ones sometimes truncate to
cordate, paler beneath, not glaucous, chartaceous; venation: 4-6 pairs main laterals,
branched, not always distinct from the midrib branches, with tertiary reticulum
dense, conspicuous; laminar glands very small, punctiform; ventral glands absent.
Inflorescence 1—15(—30)-flowered, from apical node, loosely subcorymbiform, some-
times also from uppermost 1-3 nodes, rarely with 1-2 pairs of subsidiary branches;
pedicels 8-28(50) mm long bracts small, linear-lanceolate, caducous. Flowers 30-65
mm in diam., stellate; buds ovoid, subacute to obtuse. Sepals 4:5-13 X 1-5-6 mm,
imbricate, equal or unequal, + spreading in bud and fruit, broadly or narrowly
elliptic or oblong to lanceolate or oblanceolate, acute to rounded, margin entire;
midrib distinct, other veins not prominent; laminar glands + numerous, basally
linear to striiform, punctiform towards apex. Petals golden yellow to lemon-yellow,
not tinged red, spreading, 20-34 x 12-20 mm, 2-5-4-5 xX sepals, triangular-obovate,
with apiculus iateral, acute to rounded or obsolete; margin entire, eglandular.
Stamen fascicles each with 25-35 stamens, longest 18-32 mm long, almost equalling
petals; anthers yellow to pale orange. Ovary 2-5-5 x 2-5—3 mm, ovoid or ovoid-conic
to subglobose; styles 12-20 mm, long, c. 3-5-5 X ovary, united almost to the apices
then outcurved or very rarely free up to halfway; stigmas small. Capsule 6-10 x
4-7 mm, broadly ovoid or rarely ovoid-conic to subglobose. Seeds dark reddish-
brown, c. 2 mm, long, cylindric, narrowly carinate, shallowly linear-reticulate to
linear-foveolate. 2n = 42.

Cliffs, glens, gorges, limestone hills, roadsides, in dry localities; 0-1500 m.

SE. China (Shaanxi, Sichuan, Guizhou, Guangxi, Hubei, Jiangxi, Guangdong,
Fujian, Zhejiang, Anhui, Jiangsu, Shandong; also recorded from Henan and

STUDIES IN THE GENUS HYPERICUM L. 233

Fig. 15 H. monogynum: (a) habit; (b) leaf section; (c) flowering branch; (d) flower bud; (e) sepal; (f)
petal; (g) stamen fascicle; (h) anthers; (i) ovary; (j) capsule (a, c x 1;b, d-g,i, j X 4;h x 20). (c) Staunton
& Nelson s.n. (form(ii) ); all others Read 1298 (form (i) ).

234 NORMAN K. B. ROBSON

Map 14 Sect. 3. Ascyreia: 16. H. monogynum ®, 17. H. cohaerens O.

Hunan). Probably introduced in the eastern lowlands. Native or possibly naturalised
in Taiwan. Map 14.

CHINA. Shaanxi: Shiquan Xian, 880 m, 15.vi.1959, Xing 1463 (SZ). Sichuan:
Nanchuan Xian, 1050-1350 m, 1.vi.1928, Fang 1241 (E, K); Fengjie, 1200 m,
23.vi.1958, Zhou 26377 (KUN). Guizhou: Dushan, 900 m, 24.vii.1959, Li 1356
(KUN). Guangxi: (locality uncertain), 3.viii.1956, Liang 100012 (IBSC). Hubei:
Ichang, 300-500 m, v.1907, Wilson 2422 (BM, E, K); Zhuxi Xian, 18.vi.1959, Li
9529 (SZ) Jiangxi: Kiukiang, 1874, Forbes 836 (BM). Guangdong: Yeung Shan, S. of
Linchow, vii-ix.1932, Tsui 487 (K); Ying Tak, 10.iv.1914, Tutcher in Hb. Hongkong.
10618 (IBSC, K). Fujian: Amoy Island, 20.v.1923, Chung 1636 (A, E, K). Zhejiang:
S. of Ping Yung, 60-175 m, vi-vii.1924, Ching 1912 (E). Anhui: Li Shan, NW. of
Chemen, 210 m, S.viii.1925, Ching 3137 (E). Jiangsu: Ox Head Mountain near
Nanking, vi.1922, Merrill 11392 (K). Shandong: Tai Ching Kung, Lao Shan,
13.viii.1930, Chiao 2909 (E, K).

TAIWAN. Taipeh: Grass Mountain, 20.v.1959, Liu & Kuo 123 (TAI). Ilan:
Taihoku, Ins. Kizan, 3.vii.1932, Masamune & Suzuki NTU 076878 (TAI). Taitung:
Taitung, Sasaki NTU 076877 (TAI).

CULTIVATED. Specimens seen from China, Hong Kong, Japan, India, Indone-
sia, Australia, Mauritius, Zimbabwe, South Africa, U.S.A., Mexico, Costa Rica,
Cuba, Jamaica, Martinique, Scotland, England, Ireland, France, Germany, Estonia
and (?) Portugal.

NATURALISED. Japan, N. and E. Taiwan.

H. monogynum is a very variable species in which the main clinal variation has been
complicated by another set of variations in part of the range of the species.

STUDIES IN THE GENUS HYPERICUM L.

(i) The form with elliptic to oblanceolate, acute, cuneate-based leaves, a very lax
inflorescence and narrowly elliptic, acute sepals (Kuntze’s subsp. salicifolium) is
nearest to H. gaitii of India. It is a native of upland regions of central China (Shaanxi,
Sichuan, western Hubei), whence it has apparently spread (at least partly by
introduction) into the coastal lowlands from Fujian north to Shandong. From these
lowland regions it was introduced into Japan at an early date and brought to Europe
from there by Oldham in 1861. It is now cultivated widely in Europe, N. America and
other temperate regions, where it is more or less hardy.

(ii) In eastern Hubei and Jiangxi, form (i) is gradually replaced by one with
shorter, oblong leaves, obtuse to rounded at the apex and rounded at the base, a less
lax inflorescence and narrowly oblong, obtuse sepals (Kuntze’s subsp. obtusifolium),
which attains its extreme form in coastal Guangdong. From this area it was
introduced into England by Sir George Staunton and David Nelson, who accom-
panied Lord Macartney on his famous mission to China in 1753. It was grown in the
Earl of Northumberland’s garden at Stanwick and subsequently at Chelsea Physic
Garden by Philip Miller, whose illustration is the type of both H. monogynum and H.
chinense L. This lowland form was introduced to India (Calcutta Botanic Garden)
and is now widely cultivated in the warmer regions of the world. According to Miller,
it can survive outside in a very warm situation in England.

The form native or naturalised in Taiwan (Kuntze’s subsp. /atifolium) is somewhat
intermediate between forms (i) and (ii) but has solitary flowers.

(iii) Throughout the range of form (i) there are plants with a similar lax inflores-
cence but with broader leaves and sepals (the latter sometimes being foliaceous).
The leaves may vary from broadly elliptic, acute and cuneate at the base (e.g. Fang
12366 from Chengdu, Sichuan) to broadly oblong, rounded and cordate at the base
(e.g. Wilson 1094 from Changyang, western Hubei). Plants of the latter extreme
form were identified by Sargent (Pl. Wilsonianae) as H. prattii (q.v.). Specimens of
this form have also been seen from Guizhou and Shandong, perhaps introduced.

(iv) In an extreme form of (iii), occurring in Sichuan, the leaves are short, elliptic
to ovate or triangular-ovate, obtuse to rounded at the apex, broadly cuneate to
rounded at the base (e.g. Xiong & Li 91225 and 91554 (SZ), both from Nanchuan
Xian). The sepals, however, are long and narrow, as in H. cohaerens.

Although these forms differ widely in their extreme manifestations, and indeed
appear to be related to different species and even sections (Fig. 4, p. 170), it does not
seem possible to recognise them taxonomically owing to complete morphological
and geographical intergradation.

Typical examples of the forms may be identified by the following key:

1. Leaves narrowly elliptic to oblanceolate, apex acute to subacute,
base cuneate; inflorescence usually lax (‘salicifolium’) ............... Form (i)

Leaves broadly elliptic to oblong or ovate (or, if oblanceolate, then

apex obtuse to rounded), apex obtuse to rounded or rarely acute,

base cuneate to sub-rounded; inflorescence usually + dense or

TLONMELSISO IMAL Vcore css terse 3 Sette eee eat ee aeRO OR oe eects ea seein cistie ves al 2
2. Leaves narrowly oblong to lanceolate, apex obtuse to rounded, base
rounded or rarely cuneate (‘obtusifolium’) ......................00.0000 Form (ii)
Leaves broadly oblong or broadly elliptic to ovate (or if lanceolate,
then acute), apex acute to rounded, base rounded to cordate ....... 3

3. Leaves broadly oblong to broadly elliptic or lanceolate; sepals
broadly elliptic to broadly oblong or foliaceous, apex obtuse to

TTI ofa (ALTE Oe Te a ee ee Form (iii)
Leaves broadly elliptic-oblong to triangular-ovate or ovate; sepals
MAELO lyelliptic. Apex acute (“OVATUM’). .....c<.-.000 sscendaumonaa-ntoanse Form (iv)

17(28). Hypericum cohaerens N. Robson, sp. nov.

H. monogyno L. affinis, sed foliis petiolo 1-5-2 mm longo, floribus in alabastro
subacuminatis, sepalis anguste lanceolatis vel linearibus, petalis quam sepalis
duplo longioribus apiculo acuto, stylis libris sed proximaliter cohaerentibus,

235

236

<5 + <<

<>

NORMAN K. B. ROBSON

differt. Type: China, Guizhou, Fan-Tsing Shan, 2000 m, 9.vii.1931, S. S. Sin
F.51002 (IBSC!, holotype).

Icon: —.

Shrub, with branches spreading ? Stems red, 4-lined and ancipitous when young,
soon terete; internodes 15-23 mm long, shorter than leaves; bark reddish-brown.
Leaves petiolate, with petiole 1-5-2 mm long; lamina 25-50 x 12-23 mm, oblong to
elliptic, acute to rounded, margin plare, base cuneate to subangustate, paler
beneath, not glaucous ?, chartaceous; venation: 2-3 pairs main laterals, all or the
upper sometimes uniting with midrib branches to form + arched inframarginal vein
(i.e. brochidodromous), with tertiary reticulum very dense, not very conspicuous;
laminar glands very small, punctiform; ventral glands absent. Inflorescence 1—7-
flowered, from apical node, loosely subcorymbiform; pedicels 8-11 mm long; bracts
linear, caducous. Flowers 35-50 mm in diam., stellate; buds ovoid, subacuminate.
Sepals 9-12 x 1-5-3 mm, slightly imbricate or open, equal, spreading in bud and
young fruit, narrowly lanceolate to linear, acute, margin entire; midrib distinct, one
other pair of veins sometimes visible beneath; laminar glands linear. Petals golden
yellow ?, not tinged red, spreading, 16-25 x 8-13 mm, c. 2 X sepals, oblanceolate,
with apiculus subterminal, acute, margin entire, eglandular ? Stamen fascicles each
with c. 30 stamens, longest 16-20 mm, long (or probably longer), c. 0-8-0-95 x
petals; anthers yellow. Ovary 4-5 x 3-4 mm, ovoid to subglobose; styles 10-12 mm
long, c. 2-2-2:5 X ovary, at first up to half coherent but becoming free, straight;
stigmas small. Capsule and seeds unknown. 2n = ?

No habitat cited; 1450-2000 m.
China (Guizhou, Yunnan). Map 14.
CHINA. Yunnan: Daguan, 1450 m, 16.v.1973, Sun 0620 (KUN) (see also type).

Apart from the free or almost free styles (which in this species are certainly a
secondary development), H. cohaerens is morphologically more advanced than H.
monogynum, e.g. in (i) the tendency to form brochidodromous (looped) leaf-
venation, (ii) the longer petioles, (iii) the narrower sepals and (iv) the acute petal
apiculus. In altitude (1450-2000 m) it exceeds that of nearly all H. monogynum
records (0—1500 m). Its leaf shape and venation approach those of H. formosanum
and its relatives in sect. 4. Takasagoya, but this is probably a parallel development.
Its nearest relatives appear to be H. monogynum forms (iii) and (iv) (see Fig. 4).

18(29). Hypericum prattii Hemsley

in J. Linn. Soc. 29: 303 (1892); Diels in Bot. Jahrb. 29: 475 (1900); H. Léveillé in
Bull. Soc. bot. Fr. 54: 590 (1908); Rehder in Sargent, Pl. Wilson. 2: 404 (1915) pro
parte excl. Wilson 1640, 2420. Types: China, Sichuan, chiefly near Tachienlu,
Pratt 381 (K!, lectotype—mihi; BM!, syntype); Min river, Faber 424 (K!, syntype);
sine loc., Henry 8808 (K!, syntype).

Icon: —.

Shrub 0-75-1 m tall, with branches spreading. Stems red, 4-lined and ancipitous when
young, soon terete; internodes 10-55 mm long, shorter than leaves; bark reddish-
brown. Leaves sessile; lamina 40—-110(145) x 20-51(70) mm, broadly ovate to
elliptic-ovate, acute to shortly acuminate, margin plane, base cordate-amplexicaul
or (uppermost) rounded, rather paler beneath, not glaucous, chartaceous; venation:
4-5 pairs main laterals, branched, the midrib pinnately branched, with tertiary
reticulum dense, conspicuous especially beneath; laminar glands punctiform, very
small; ventral glands absent. Inflorescence 3—10(—24)-flowered, from apical node,
loosely subcorymbiform; pedicels 3-10(25) mm long (--60 mm in fruit); bracts small,
linear, caducous. Flowers 35-55 mm in diam., stellate; buds ovoid, obtuse to
rounded. Sepals 9-17(-19) x 3—9(-11) mm, free, imbricate, subequal to unequal,
erect in bud, ? in fruit, lanceolate to broadly ovate-cordate (foliaceous), rounded,
with margin entire; midrib distinct, other veins not prominent; laminar glands c.
9-numerous, basally linear, distally punctiform, Petals golden yellow, not tinged

STUDIES IN THE GENUS HYPERICUM L.

red, spreading, 20-26 x 9-18 mm, 1-5-2-3 x sepals, obovate to oblanceolate, with
apiculus lateral, rounded or obsolete; margin entire, eglandular. Stamen fascicles
each with 30-40 stamens, longest 18-24 mm long, almost equalling petals; anthers
yellow. Ovary 4-6 x 2-5-4-5 mm, ovoid; styles 11-19 mm long, c. 3-3-5 x ovary,
united almost to the apices then spreading; stigmas small. Capsule and seeds not
seen.

‘Glens’ (Wilson 2421a); 300—c. 3000? m.

China (north-eastern Yunnan ?, Sichuan, western Hubei). Map 15.

CHINA. Yunnan: environs de Yunnan-fou [Kunming], v.1908, Alleizette (L) (see
note below). Sichuan: [no precise locality], pre-ii. 1890, Henry 8808 (K); Yangtze-
kiang, Min R., pre-xii. 1887, Faber 424 (K); chiefly near Tachienlu [Kangding]
2700-4050 m, Pratt 381 (BM, K); Emei Shan [Mt. Omei], 1000 m, 22.vii.1957, Yang
55980 (KUN, SZ). Hubei: Ichang, 300 m, v.1907, Wilson 2421a (BM, E, K).

ee €
H. longistylum subsp. giraldii O.

H. prattii is closely related to H. monogynum, but differs in its ovate, often
acuminate, leaves, nearly always with cordate-amplexicaul base. Using these dis-
tinguishing characters, Wilson 1604 and 2420, cited as H. prattii by Rehder in
Sargent, Pl. Wils. 2: 404 (1915), must be transferred to H. monogynum. These
specimens are somewhat intermediate; but from a study of the available material, it
seems better to treat H. monogynum and H. prattii as distinct species rather than to
give the latter the varietal status suggested by Rehder.

The Wilson specimen (from W. Hubei) has larger flowers and broader sepals than
have the original collections from Sichuan, and was collected at a much lower
altitude. The absence of other records from near Kunming (let alone elsewhere in
Yunnan) throws doubt on the provenance of the Alleizette specimen.

237

Map 15 Sect. 3. Ascyreia: 18. H. prattii x, + (doubtful), 19a. H. longistylum subsp. longistylum @, 19b.

238

NORMAN K. B. ROBSON
19(30). Hypericum longistylum Oliver

in Hook., Icon. pl. 16: t. 1534 (1886); Diels in Bot. Jahrb. 29: 476 (1900); H. Léveillé
in Bull. Soc. bot. Fr. 54: 591 (1908); Pampini in Nuovo G. bot. ital. 11, 17: 670
(1910); Rehder in Sargent, Pl. Wilson. 2: 404 (1915); R. Kellerin Engler & Prantl,
Nat. Pflanzenfam., 2nd ed. 21: 176 (1925) Types: China, Hubei, Ichang, Henry
582 (K!, lectotype-mihi), Henry 217 (K!, syntype), Henry 994 (K!, syntype).

Shrub c. 1 m tall, erect, with long branches divaricate and short ones pinnate. Stems
red, 2-4-lined and ancipitous when young, eventually terete; internodes 10-30 mm
long, shorter than to exceeding leaves; bark pale grey. Leaves subsessile or with
petiole up to 1 mm long; lamina 10-31 x 6-16 mm, narrowly oblong to elliptic or
subcircular, rounded to subapiculate, margin plane, base cuneate to shortly angus-
tate, + densely glaucous beneath, chartaceous; venation: c. 3 pairs main laterals
(faint), the midrib branches not or scarcely visible, without or rarely with very faint
tertiary reticulum; laminar glands small to very small dots; ventral glands absent.
Inflorescence 1-flowered, terminal and on short lateral branches; pedicels 8-12 mm
long; bracts foliar, persistent. Flowers 25-45(-50) mm in diam., stellate; buds
narrowly ellipsoid, acute. Sepals 3-6(-10) x 0-5—2(-3) mm, free or united at the
base, imbricate to open, equal or subequal, spreading or recurved in bud and fruit,
linear or rarely elliptic, acute, with margin entire; midrib + conspicuous, other veins
not prominent; laminar glands c. 4, basally linear, punctiform toward apex. Petals
golden yellow to orange, not tinged red, spreading, (11—)15—22(-24) x 4-8(-10) mm,
2-5-3-5 x sepals, oblanceolate, without or almost without apiculus, margin entire,
eglandular. Stamen fascicles each with c. 15-25 stamens, longest 15-25 mm long;
anthers yellow. Ovary 34 x 2-3 mm, ellipsoid to globose, sometimes substipitate;
styles 10-18 mm long, c. 3-5-6 X ovary, united almost to the apices then spreading;
stigmas small. Capsule (4-)6-12 x 4-5 mm, ellipsoid to subglobose, sometimes
substipitate. Seeds not seen. 2n = ?

On cliffs and dry banks; 200-2090 m.

China (Hubei, Shaanxi, Hunan). Also recorded from Gansu, Sichuan, Henan, and
Anhui. Map 15.

H. longistylum is most similar morphologically to form (iv) of H. monogynum from
Sichuan (Fig. 4). Pampanini, in Nuovo G. bot. ital. I1, 17: 670 (1910), described three
varieties of this species: (i) var. longistylum with large flowers and ellipsoid ovary and
fruit, (ii) var. silvestrii with small flowers and ellipsoid ovary and fruit, (iii) var.
giraldii with large flowers and globose ovary and fruit. Vars (i) and (ii) both occur at
lower altitudes (200-1200 m) in north and west Hubei, Sichuan and Hunan (and
probably Ankui and Henan) and cannot be distinguished on flower size; whereas var.
giraldii is found at higher altitudes (1950-2090 m) in central to north Hubei and in
central and north Shaanxi and possibly Gansu. The difference in ovary shape
between vars (i)—(ii) and (iii) seems to be constant, although some specimens of ‘var.
silvestrit’’ have rather broader ovaries than is usual in var. longistylum. It seems
possible, then, to recognise two subspecies, rather than varieties, as follows:

19a(30a). H. longistylum subsp. longistylum

H. longistylum var. silvestrii Pampanini in Nuovo G. bot. ital. I, 17: 670, f. 15b
(1910). Types: China, Hubei, Sce-men, Silvestri 1490 (FI!, lectotype); presso
Siang-yang, Silvestri 1486 pro parte (FI!); sine loco, Silvestri 1489 (FI); Sce-men,
Silvestri 1490a (FI!); Kai-scian, Silvestri 1491 (FI!) (all syntypes).

Norysca longistyla (Oliver) Y. Kimura in Nakai & Honda, Novafl. jap. 10: 98 (1951).

Icon: Oliver in Hook., Icon. pl. 16: t. 1534 (1886).
Sepals 3-6 mm long. Ovary and capsule ellipsoid, usually substipitate.

China (north and east Sichuan, north and west Hubei, north Hunan, recorded from
Henan, Anhui); 200-1200 m. Map 15.
CHINA. Hubei: Hsing-shan Hsien, 600 m, 8.vi.1907, Wilson 2423 (BM, E, K);

STUDIES IN THE GENUS HYPERICUM L.

Ichang, 300 m, iv.1907, Wilson 2424 (BM, K); U-tan-scian [Wu-tang shan], iii.1912,
Silvestri 6197 (BM, FI); Wudang shan, near South Crag Temple, 800 m, 16.v.1983,
Lancaster 1184 (BM); Nan-T’o [Nanto] and mountains to northward, pre ii.1887,
Henry 1999 (K); Jun Xian 11.viii.1959, Xing 11098 (SZ). Hunan: Dayong, 470 m,
23.v.1978, Sheng 1320 (IBSC).

19b(30b). H. longistylum subsp. giraldii (R. Keller) N. Robson, stat. nov.

H. giraldii R. Keller in Bot. Jahrb. 33: 548 (1904); Pavolini in Nuova G. bot. ital. Il,
15: 406 (1908); H. Léveillé in Bull. Soc. bot. Fr. 54: 590 (1908). Type: China,
Shaanxi, ‘in monte Lun-san-huo’, Giraldi 539 (FI!, holotype).

H. longistylum var. giraldii (R. Keller) Pampanini in Nuovo G. bot. ital. I1, 17: 670,
f. 15¢ (1910).

Icon: Pampanini, in Nuovo G. bot. ital. I1, 17: 670, f. 15c (1910).
Sepals 5-7 mm long. Ovary and capsule globose, sessile.

China (north-west Hubei, central and north Shaanxi, recorded from Gansu as H.
longistylum); 1950-2090 m. Map 15.

CHINA. Hubei: Zan-lan-scien [Zan-lan Shan], iv.1912, Silvestri 6196 (FI); Scian-
Kiu, iii.1912, Silvestri 6200 (FI); Ou-tan-scian [Wu-tang Shan], 2090 m, vii.1907,
Silvestri 1488 (FI); Monti Triori, 1950 m, 3.vii.1907, Silvestri 1487 (FI). Shaanxi.
Monti del Lun-san-suo, v.1895, Pio Nesi in Giraldi 539 (FI); N. Central China,
pre-1900, Fr. Hugh (BM) (type collection ?).

20(31). Hypericum subsessile N. Robson, sp. nov.

H. mysurensi Wight & Arnott affinis, sed foliis subsessilibus venatione densiore
reticulati, floribus vade cyathiformibus, sepalis latioribus, stylis brevioribus, differt.
Type: China, Yunnan, Tali Range, vili.1929, Forrest 28133 (BM!, holotype; E!,
isotype).
Icon: —.

Shrub c. 1 m tall. Stems red, 4-angled and ancipitous when young, eventually terete;
internodes 10-30 mm long, shorter than leaves; bark grey-brown. Leaves subsessile,
with very short flat petiole; lamina 35-65 x 7-20 mm, narrowly elliptic, acute,
margin plane, base cuneate, paler to glaucous beneath, subcoriaceous; venation: 3-4
pairs main laterals, branched, the midrib pinnately branched, with tertiary reticulum
rather lax, not very conspicuous; laminar glands small dots and short streaks; ventral
glands sparse to dense. Inflorescence 1-8-flowered, from 1(2) node(s); pedicel 8-20
mm long; bracts small, lanceolate to foliar, deciduous. Flowers 35-45 mm in diam..,
shallowly cyathiform; buds ovoid, acute. Sepals 10-18 x 4-10 mm, free, imbricate,
unequal (foliaceous), outcurved in bud, reflexed in fruit, ovate to elliptic, acute or
acuminate, with margin entire; midrib clearly visible, veins not prominent; laminar
glands c. 8, basally linear, punctiform toward apex. Petals bright yellow, tinged red
dorsally, slightly incurved, 17-20 x 9-11 mm, c. 2 X sepals, oblanceolate-obovate,
with apiculus subterminal, acute; margin entire, eglandular. Stamen fascicles each
with 40-60 stamens, longest 12-15 mm, long, c. 0-7 X petals; anthers bright yellow.
Ovary 6-8 X 5-6 mm, ovoid-conic; styles 5-6 mm long, 0-8-0-9 x ovary, free,
outcurved near apex; stigmas narrowly capitate. Capsule (16—)18-21 x (7—)10-12
mm, narrowly ovoid-conic. Seeds dark reddish-brown, 1-1—1-5 mm long, cylindric,
with terminal and unilateral wing, shallowly linear-foveolate. 2n = ? See also p. 325.

Thickets; 2400-2550 m.

China (Yunnan, Sichuan). Map 17.
CHINA. Yunnan: Tali Range, vili.1929, Forrest 28133 (BM, E). Sichuan:
Hanyuan Hsien, 2400-2550 m, 11.x.1928, Fang 3744 (E, K).

As well as having affinities directly with H. mysurense, H. subsessile is related to both
H. siamense and H. acmosepalum and is morphologically intermediate between
them. It differs from the former by its longer, narrower leaves and larger foliaceous

CC) * <SS

239

240

WW +8 <&

NORMAN K. B. ROBSON

sepals, and from the latter by its subsessile leaves with lax reticulate venation, longer
(often foliaceous) sepals, broader petals and shorter stamens and styles.

H. subsessile is apparently very local in occurrence, having been found in only
two widely separated localities. The Yunnan plant has solitary flowers, whereas the
one from Sichuan has 1-8-flowered branches. Otherwise the collections are similar
and appear to be conspecific.

21(32). Hypericum siamense N. Robson, sp. nov.

H. leschenaultio Choisy affinis, sed foliis ellipticis vel oblongis haud lanceolatis vel
ovatis, apice acutis vel obtuse apiculatis, vena intramarginali atque nervatura
secundaria pinnate valde manifestis instructis, sepalis ovato-triangularibus, differt.
Type: Thailand, Doi Chiengdao, c. 1900 m, 6.xi.1922, Kerr s.n. (BM!, holotype;
ABD!, isotype).

H. sp. sensu Smitinand in Nat. Hist. Bull. Siam Soc. 21: 100 (1966).

H. cf. patulum sensu Smitinand, tom. cit.: 102 (1966).

Icon: Fig. 16.

Shrub 0-3-3 m tall, bushy, with short ascending lateral branches. Stems red, 4-lined
and ancipitous when young, eventually terete; internodes 5—20 mm, long, shorter
than leaves, bark grey-brown. Leaves subsessile or petiolate, with petiole 1-2 mm
long, broad; lamina 25-36 x 7-18 mm, oblong (lower) to elliptic (upper), acute to .
apiculate-obtuse, margin plane, base + broadly cuneate to rounded or subauricu-
late, paler beneath, not glaucous, chartaceous; venation: 1-3 pairs main laterals, the
outer joining upper laterals to become intramarginal vein, all dark and easily visible
(when dry), with lax or scarcely visible tertiary reticulum; laminar glands very short
streaks and dots; ventral glands sparse to dense. Inflorescence 1—S-flowered, from
1-2 nodes, subcorymbiform, pedicels 10 mm, long; bracts narrowly elliptic, persis-
tent. Flowers 40-50 mm in diam., shallowly cyathiform ?, buds ovoid, acute to
subacute. Sepals 8-12 x 3-6 mm, free, imbricate, unequal, spreading to reflexed in
bud and fruit, triangular-ovate, acute, margin entire, midrib visible below, veins
slightly prominent; laminar glands numerous, linear, interrupted towards apex.
Petals deep ? golden yellow, tinged red dorsally, slightly incurved, 20-24 x 12-17
mm, 2-2-5 X sepals, narrowly to broadly obovate, with apiculus obtuse to rounded,
lateral, margin entire, eglandular. Stamen fascicles, each with c. 45-50 stamens,
longest 10-11 mm. long, c. 0.35-0.5 X petals; anthers golden yellow to orange.
Ovary 5-7 X 3-3-5 mm, ovoid to narrowly ovoid-conic; styles 2-5-6 mm long,
0-4-0-8(-1) x ovary, free, outcurved near apex; stigmas small. Capsule c. 15—7 mm,
ovoid-conic. Seeds pale orange-brown (? immature), c. 1-0 mm long, cylindric,
narrowly carinate, shallowly linear-reticulate to linear-foveolate. 2n = ?

Shrub layer of open evergreen hill forest on exposed ridges, on crystalline limestone;
1900-2200 m.

Thailand (Northern: Doi Chiengdao). Map 16.

THAILAND. Northern: all from Chiengmai, Doi Chiengdao: c. 1900 m,
6.xi.1922, Kerr s.n. (ABD, BM); 2100-2200 m, 3.xii.1961, Smitinand & Anderson
7303 (K); 2000-2200 m, 10.xi.1962, Smitinand, Poore & Robbins 7789 (AAU, BKF,
E, L); 1900-2100 m, 17.viii.1963, Smitinand & Sleumer 1055 (L, SING); c. 2000 m,
7.xii.1965, Hennipman 3277 (L); 1900-2175 m, 14.ix.1967, Shimizu, Koyama &
Nalampoon T. 10109 (BKF); 1900-2100 m, 26.ix.1971, Murata et al. T. 15262 (BKF,
K); 1900-2100 m, 27.ix.1971, Murata et al. T. 15214 (AAU).

H. siamense, like H. subsessile and H. apiculatum, is a relict species. It shows some
resemblances to H. subsessile, but its leaf shape and venation, together with its
triangular-ovate sepals, make it unmistakable.

It is clearly a near ancestor of H. leschenaultii, of which the nearest population is in
Sumatra, differing from it inter alia by the oblong to elliptic leaves, with a clear
intramarginal vein, markedly pinnate laterals and a shorter petiole, and usually by
the shorter sepals.

STUDIES IN THE GENUS HYPERICUM L. 241

——

az
ia] 2
tf
Fi
PEL ¢
H Hild] 4
Pale] s
iff
iff
;
f}/ 8/4
j
i i
rai

—

> ——

Ss a ——r
SSS

Fig. 16H. siamense: (a) habit; (b) leaf section; (c) flower bud; (d) sepal; (e) petal; (f) stamen fascicle; (g)
anthers; (h) ovary; (i) capsule (a x 1;c, i x 4; b, d-f, h x 6; g X 20). All Murata et al. T15214.

242 NORMAN K. B. ROBSON

AWE ‘ a
ag yes inn
ae

?

.
ac. Ss

e — 5
= oF 8d a6 OS

Map 16 Sect. 3. Ascyreia: 21. H. siamense @, 22. H. leschenaultii e.

22(33). Hypericum leschenaultii Choisy

in DC., Prodr. syst. regni nat. veg. 1: 545 (1824); Guillemin in Delessert, Ic. sel. pl. 3:
17, t. 27 (1837); Choisy in Zollinger, Syst. Verz. Fl. Ind. Archip. 2: 151 (1854);
Stapf in Curtis’s bot. Mag. 152: t. 9160 (1925); Hochreutiner in Candollea 2: 435
(1925); R. Keller in Engler & Prantl, Nat. Pflanzenfam. 2nd ed. 21: 176 (1925);
Doct. v. Leeuwen in Trop. Natuur. 16: 115, f. 3 (1927), in Verk. K. Akad. Wet
Amsterdam Afd. Natuurk. (2de sect.) 31: 181 (1933); Merrill in Contr. Arn.

STUDIES IN THE GENUS HYPERICUM L.

Arbor. 8: 107 (1934); v. Steenis in Bull. Jard. bot. Buitenz. II, 13: 219 (1934);
Bakker & Bakh. f., Fl. Java 1: 382 (1963); Robson in J. Roy. Hort. Soc. 95: 489
(1970); v. Steenis, Mtn fi. Java: t. 23 f. 6(1972); Bean, Trees & shrubs hardy in Br.
Isles 8th ed. 2: 417 (1973); Robson in Fl. malesiana I, 8: 17, ff. 12, 13 (1974). Type:
Indonesia, Java, Leschenault (P).

H. javanicum Blume, Bijdr. fl. Ned. Ind. 1: 141 (1825); Hasskarl, Cat. pl. hort.
bogor. alter.: 213 (1844); Moritzi, Syst. Verz. Zollinger: 25 (1845-56); Choisy in
Zollinger, Syst Verz. Ind. Archipel. 2: 151 (1954). Type: Java, ‘In altis montium
Javae insulae’. [Mt. Gede] Blume 557 (L!, holotype; BM!).

H. coriaceum Blume, Bijdr. fl. Ned. Ind. 1: 142 (1825). Type: Java, ‘In cacumine
montis Tjerimai Provinciae Cheribon’, Blume (L!, lectotype ;BM!).

H. triflorum Blume, Bijdr. fl. Ned. Ind. 1: 142 (1825); Hasskarl, Cat. Pl. Hort.
Bogor. Alter: 213 (1844); Moritzi, Syst. Verz. Zollinger: 25 (1845-46); B[urbidge]
in The Garden 23: 158 cum fig. (1883); R. Keller in Engler & Prantl, Nat.
Pflanzenfam. 3 (6): 209, 210, f. 100 (1893), 2nd ed. 21: 176, f. 73 (1925). Type:
Java, ‘In cacumine montis Gede’, Blume (L!).

H. triflorum var. angustatum Blume, Bijdr. fl. Ned. Ind. 1: 143 (1825). Type: Java,
‘In vicinitate catarractarum montis Gede’, Blume (L!).

H. nervosum Choisy in Zollinger, Syst. Verz. Ind. Archipel. 1: 151, 152 (1854);
Koorders in Natuurk. Tijdschr. 60: 259 (1900); Ernst in Karsten & Schenk, Veget.
Bild. 7 (2): t. 4-5 (1910), non H. nervosum D. Don (1825), nom. illegit. Type:
Java, ‘In montosis Waliran, supra 5000 ft.’, Zollinger 1565 (n.v.).

Norysca leschenaultii (Choisy) Blume, Mus. bot. Lugd. Bat. 2: 24 (1856); Miquel, FI.
Ind. bat. 1 (2): 515 (1859); Y. Kimura in Nakai & Honda, Nova fl. jap. 10: 98
(1951).

Norysca javanica (Blume) Blume, Mus. bot. Lugd. Bat. 2: 24 (1856); Miquel, FI.
Ind. bat. 1 (2): 514 (1859).

Norysca coriacea (Blume) Blume, Mus. bot. Lugd. Bat. 2: 24 (1856); Miquel, Fi.
Ind. bat. 1 (2): 515 (1859).

Norysca rorida Blume, Mus. bot. Lugd. Bat. 2: 514 (1856); Miquel, Fl. Ind. bat. 1
(2): 514 (1859). Type: Java, Blume (L!).

Norysca nervosa (Choisy) Miquel, Fl. Ind. bat. 1 (2): 515 (1859).

H. hookerianum var. leschenaultii (Choisy) Dyer in Hook. f., Fl. Brit. Ind. 1: 254
(1874) pro parte, quoad typum.; Besant in Gdnrs’ Chron. 74: 135, f. 49 (1923).
H. patulum [subsp.] (a) leschenaultii (Choisy) Kuntze, Rev. gen. pl. 1: 60 (1891)

(‘leschenaultianum’).

H. patulum subsp. variabile Kuntze, Rev. gen. pl. 1: 60 (1891). Type: Java, Dieng,
1600-2300 m, 16.viii.1875, Kuntze (NY, photo!); Sumbing, 3100 m, 13.viii.1875,
Kuntze (NY, photo!).

H. hookerianum sensu Buysmann in Flora, Jena 107: 358 (1914), non Wight &
Arnott (1834).

H. leschenaultii var. typicum Hochr. in Candollea 2: 435 (1925) ‘typica’. Type as for
H. leschenaultii Choisy.

H. leschenaultii var. typicum forma elatius Hochr. in Candollea 2: 435 (1925)
(‘elatior’). Type: Java (central), Mt. Perahoe, c. 2550 m, 5.xi.1904, Hochreutiner
2391 bis (G).

H. leschenaultii var. coriaceum (Blume) Hochr. in Candollea 2: 435 (1925)
(‘coriacea’).

Norysca hookeriana var. leschenaultii (Choisy) Y. Kimura in Hara, Fl. eastern
Himalaya: 210, 2nd ed: 81 (1971) pro parte, quoad typum.

Icon: Stapf in Curtis’s bot. Mag.: t. 9160 (1925).

Shrub or small tree 0-5—2-5 m tall, dense and rounded or + lax and elongate, with
branches + spreading. Stems red to orange, 4-lined and ancipitous when young,
eventually terete; internodes 15-30 mm, shorter than leaves; bark grey-brown,
smooth. Leaves petiolate, with petiole 1-2 mm long; lamina 25-80 x 10-37 mm,
triangular-lanceolate to narrowly or rarely more broadly ovate, acute to apiculate-
rounded, margin plane to recurved, base broadly cuneate to rounded, paler to
glaucous beneath, chartaceous to subcoriaceous; venation: 3-4 pairs main laterals,

243

244

NORMAN K. B. ROBSON

the midrib pinnately branched, with lax tertiary reticulum; laminar glands short
streaks and dots; ventral glands absent or + dense. Inflorescence 1—-3(-10 or very
rarely —18)-flowered, from 1-2 nodes, subcorymbiform; pedicels 5-12 mm long;
bracts linear, persistent. Flowers (35—)40-70 mm in diam., shallowly cyathiform;
buds + broadly ovoid, acute to subacute. Sepals 7-20 x 2-8 mm (to 22 X 8 mm in
fruit), free, imbricate, unequal, spreading to recurved in bud and fruit, narrowly
oblong or + narrowly elliptic or rarely ovate to oblanceolate, acute to subacute, with
margin entire; midrib sometimes visible below, visible and sometimes incrassate
above, veins not prominent; laminar glands numerous, linear, interrupted towards
apex. Petals deep golden yellow, not tinged red, slightly incurved, (20—-)25-45 x
14-30 mm, c. 2-3 X sepals, broadly obovate to subcircular, with apiculus acute to
obtuse, lateral, margin entire, eglandular. Stamen fascicles each with c. 80 stamens,
longest 9-12 mm long, c. 0-35 X petals; anthers golden yellow. Ovary 6-9 x 4-5 mm,
+ narrowly ovoid-conic to narrowly ellipsoid-acuminate; styles 3-5-7 mm long,
0-5-0-9 x ovary, free, + sharply outcurved near apex; stigmas small. Capsule
10-20 x 8-10(-13) mm, narrowly ovoid-conic to narrowly ellipsoid. Seeds dark
orange-brown to reddish-brown, 0-8-1 mm long, cylindric, scarcely carinate, shal-
lowly linear-reticulate to linear-foveolate. 2n = 20 ? (c.f. its presumed hybrid, 22 x.
H. x. ‘Rowallane’).

Grassy slopes, thickets, and open woodland; 1500-3300 m.

Indonesia (Sumatra, Java, Lesser Sunda Islands, south-west Sulawesi). Map 16.

SUMATRA. Atjeh: Gaju & Alas Lands, summit of Mt. Goh Lembach, c. 2900 m,
21-22.ii1.1937, v. Steenis 9112 (L); Westtop G. Losir, 2950-3500 m, 5-6.ii.1937, v.
Steenis 8642 (BO). Barat: G. Kerintji[Kurintji], 3000 m, 4.v.1920, Biinnemeijer 9999
(GH, K, L). Selantan: G. Dempo, summit, 3000 m, viii.1923, Brooks 15917 (BO).

JAVA. Barat: Goenseng, summit of Mt. Pangrango, 3000 m, 29.iii.1950, v.
Oststroom 13355 (CANB, FI, L, PNH). Tengah: Dieng plateau, S.W. helling G.
Pangoran, c. 2150 m, 6.iv.1920, v. Slooten (L.). Timur: Besoeki, Jang Plateau West,
Taman Hidoep, c. 1900 m, 13.vii.1938, v. Steenis 10786 (A, L, SING).

BALI. Bei Jirta auf Weg noch Gug Agoeng, c. 2100 m, 2.vi.1912, Arens 16 (BO,
L).
LOMBOK. G. Rindjani, Segare Anak, 1925-2000 m, 11 v.1909, Elbert 1196 (L,
PNH, SING).

FLORES. Sano Patjo-Gurung, 23.v.1970, Verheijen 3856 (L.).

SULAWESI. Selatan: G. Bonthain, N.W. side, 7.vi.1921, Biinnemeijer 11894 (K,
L); B. Rantemario, 2800-3000 m, vi.1929, Kjellberg 3959 (BO).

CULTIVATED. Specimens seen from England (1882-3) and Ireland (1925).

H. leschenaultii is a variable species, the variation being to some extent clinal from
west to east. Its presence to the east of the Wallace Line (in Lombok, Flores, and
Sulawesi) is no doubt due to migration after the Malaysian and New Guinea regions
made contact for the second time (see v. Steenis (1979), especially pp. 115-119 and
the distribution map of H. leschenaultii, fig. 11).

In the British Isles, H. leschenaultii is hardy in southern England and in Ireland;
but several hardier species (e.g. H. addingtonii) have been grown under that name.

The Sumatran form is the most primitive and approaches H. siamense in having,
for example, more prominent and more numerous main lateral leaf veins than usual
(4 pairs instead of 3), larger flowers and broader sepals. For differences between H.
leschenaultii and H. siamense see p. 240.

22x(33x). Hypericum x ‘Rowallane’ (Armytage Moore)
(22. H. leschenaultii x 33a. H. hookerianum ‘Charles Rogers’)

in J. Roy. Hort. Soc. 68: 380, xlvii (1943), ‘Rowallane Hybrid’; Comber in Gdnrs’
Chron. 120: 232, f. 106 (1946); J. B. PatoninJ. Roy. Hort. Soc. 81: 477 (1956); H.
‘Rowallane Hybrid’ G. Thomas in Gdnrs’ Chron. III, 147: 254 (1960); N. Robson
inJ. Roy. Hort. Soc. 95: 484, 490 (1970); [Lancaster], Hilliers’ man. trees & shrubs:
152 (1971); Bean, Trees & shrubs hardy in Br. Isles 8th ed. 2: 415, t. 51 (1973).

STUDIES IN THE GENUS HYPERICUM L.

Standard specimen: Scotland, Wigtown, Logan, 22.ix.1977, R.B.G. Edinb. C.
11687 (BM!).

H. hookerianum ‘Rowallane’ F. Schneider in Meded. Inst. Vered. Tuinbouwgew.
Wageningen 252: 21 (1966); in Dendroflora 2: 21 (1966).

Icon: Thomas in Gdnrs’ Chron. III, 157: 255 (1960).

Shrub up to 3 m tall, with branches erect, gradually outcurving. Stems orange, 4-lined
and ancipitous when young, becoming quadrangular, eventually terete; internodes
10-70 mm long, shorter that to exceeding leaves; bark grey-brown. Leaves petiolate,
with petiole 1-2 mm long; lamina 27-67 x 10-33 mm, ovate to oblong-ovate or
oblong-lanceolate, apiculate-obtuse, base cuneate to truncate or subcordate, paler
or + glaucous beneath, chartaceous to subcoriaceous; venation: 34 pairs main
laterals, the midrib pinnately branched, with tertiary reticulum lax and obscure
or apparently absent; laminar glands short streaks and dots; ventral glands
absent. Inflorescence 1-3-flowered, from apical node, subcorymbiform; pedicels
8-10(—15) mm long, bracts lanceolate or elliptic to oblanceolate, persistent. Flowers
50-75 mm in diam., shallowly cyathiform; buds ovoid, acute to obtuse. Sepals 10-14
x 4-7 mm, free, imbricate, unequal, spreading in bud and fruit, oblanceolate to
obovate-spathulate, apiculate or rounded, with margin entire or apically eroded-
denticulate; midrib visible or obscure, veins becoming slightly prominent; laminar
glands linear, numerous. Petals deep golden yellow, not red-tinged, slightly in-
curved, 30-40 x 25-32 mm, c. 3 X sepals, obovate to subcircular, with apiculus
subterminal, + rounded; margin entire, eglandular. Stamen fascicles each with 70-90
stamens, longest 8-10 mm, long, 0-25—0-35 x petals; anthers golden yellow. Ovary
8-11 x 4-6 mm, ovoid to ovoid-ellipsoid; styles 6-7-5 mm long, c. 0-7 X ovary, free,
diverging, outcurved towards apex; stigmas narrowly capitate. Capsule c. 13 x 10
mm, ovoid. Seeds rarely set, hybrid apparently usually sterile.* 2n = 20.

Known only in cultivation.

Cultivated specimens seen from England, Wales, Scotland, Ireland, and the U.S.A.
(California).

H. x ‘Rowallane’ arose spontaneously in 1932 in the garden of Rowallane, Saintfield,
Co. Down, N. Ireland, owned by T. Armytage-Moore. Cuttings were obtained by L.
Slinger and propagated at Slieve Donard Nursery, Newcastle, Co. Down. It was first
offered for sale by that nursery a few years later, and was given an Award of Merit by
the Royal Horticultural Society in 1943.

The parentage of the hybrid is assumed, on grounds of both morphology and the
proximity of suitable parental plants, to be H. leschenaultii x “rogersii” i.e. H.
hookerianum ‘Charles Rogers’.

23(34). Hypericum acmosepalum N. Robson

inJ. Roy. Hort. Soc. 95: 494, f. 238 (1970); [Lancaster], Hilliers’ Man. trees & shrubs:
150 (1971); Bean, Trees & shrubs hardy in Br. Isles 8th ed. 2: 405 (1973). Type:
China, Yunnan, on the Lichiang plain, 2460 m, v.1921, Forrest 19448 (K!,
holotype; E!, isotype).

H. oblongifolium sensu hort. pro parte, non Choisy (1821).

H. patulum var. oblongifolium sensu hort. pro parte, non sensu Koehne (1893) nec
H. oblongifolium Choisy (1821).

H. kouytchense sensu hott. pro parte, non H. Léveillé (1904).

H. patulum var. henryi sensu hort. pro parte, non Veitch ex Bean (1905).

H. henryi sensu hort. pro parte, non H. Léveillé & Vaniot (1908).

Icon: N. Robson in J. Roy. Hort. Soc. 95: f. 238 (1970).
Shrub 0-6-2 m tall, with branches erect to ascending. Stems orange, 4-angled and
ancipitous in first year (or longer), then terete; internodes 10-50 mm long, shorter

* A few apparently ripe seeds were observed in old capsules at Nymans, West Sussex, England
(8. viii. 1982).

245

U* Se

246

CS + <>

NORMAN K. B. ROBSON

than to exceeding leaves; bark grey-brown. Leaves broadly petiolate, with petiole
0-5—1(1-5) mm, long; lamina 18-42(-60) x 6-15(-20) mm, oblong or elliptic-oblong
to narrowly elliptic (sometimes lanceolate towards apex of shoot and oblanceolate
towards base), obtuse (rarely subacute) or apiculate to rounded, margin plane, +
recurved, base cuneate, markedly paler to glaucous beneath, chartaceous to sub-
coriaceous; venation: 1-2 pairs main laterals (the upper forming distinct, often +
straight intramarginal vein), with midrib rather obscurely branched distally, with
rather dense but very obscure or invisible tertiary reticulum; laminar glands + small
dots and sometimes short streaks, ventral glands sparse to rather dense. Inflores-
cence 1—3(-6)-flowered, from apical node, subcorymbiform; pedicels 7-17 mm long;
bracts foliar to lanceolate, persistent. Flowers 30-50 mm in diam., stellate; buds
ovoid, acute to subapiculate. Sepals (S—)6-9(-11) x 3-4(-6) mm, free, imbricate,
subequal, + outcurved in bud and fruit, ovate to narrowly lanceolate, acute or
acuminate, with margin subentire or minutely and + irregularly denticulate (espe-
cially towards apex); midrib + conspicuous, veins not prominent; laminar glands
linear or interrupted, c. 8. Petals deep yellow, sometimes tinged red, spreading or
reflexed, 16-25 x 10-15 mm, 2-5-3 X sepals, obovate, with apiculus lateral, subacute
to obtuse, margin entire or often minutely glandular-denticulate especially around
apiculus. Stamen fascicles each with 40-65 stamens, longest (10)15-18 mm, long,
0:75-0:85 X petals; anthers yellow to orange-yellow. Ovary 5-7 X 3-5—4:5 mm, +
narrowly ovoid-conic; styles (3—)4-6(-8) mm, long, equalling to slightly longer than
ovary, free, suberect, outcurved near apex; stigmas truncate to narrowly subcapi-
tate. Capsule 9-15 xX 8-10 mm, ovoid to narrowly ovoid-conic, turning bright red
during maturation. Seeds dark orange- to reddish-brown, 1-1-1 mm long, narrowly
cylindric, narrowly carinate with terminal expansion, shallowly linear-foveolate.
2n =?

Forest glades, roadside banks, hillsides among scrub, open stream-sides; 900—
3000 m.

China (N. and E. Yunnan, S. Sichuan, SW. Guizhou, Guangxi). Map 17.

Map 17 Sect. 3. Ascyreia: 20. H. subsessile 0, 23. H. acmosepalum @.

STUDIES IN THE GENUS HYPERICUM L.

CHINA. Yunnan: Likiang range, Long Yi Mtn, Mung Hua [Munghwal], vi.1934,
McLaren’s Colls. L94A (BM, E, K); Cangshan, above Yangbi, Shangchang, 2600 m,
9.v.1981, Sino-Brit. Exped. Cangshan 423 (BM, E); near Kunming, Dashao, 2400 m,
26.iv.1981, Sino-Brit. Exped. Cangshan K.052 (BM, E); vicinity of Yunnan-sen
[Kunming], 1906 ?, Maire 136 (BM, E); Ping-pien Hsien, 1500 m, 18.vii.1934, Tsai
60972 (BO). Sichuan: Xichang Xian, 2100 m, 23.v.1978, Zhao 4577 (SZ). Guizhou:
Shui-cheng, outside E. gate, 23.x.1930, Tsiang 9442 (E, K). Guangxi: Lingyun,
1100 m, 19.x.1957, C. C. Chang 10428 (IBSC).

CULTIVATED. Specimens seen from England (1970; live 1984), Ireland (1967),
U.S.A. (1959).

H. acmosepalum appears to be an early offshoot of the line that gave rise to the H.
patulum and H. lagarocladum groups. Its stellate flowers with long stamens and
styles and acute sepals, and its bright red young fruits, are reminiscent of H.
kouytchense; but its erect habit, elliptic to oblong leaves with a marked intramarginal
vein, and deep yellow petals are quite distinct and make it easily distinguishable. Its
nearest (ancestral) relative appears to be H. subsessile, which is more advanced in its
shorter stamens and styles and broader sepals, but has a more primitive leaf.

24(35). Hypericum lagarocladum N. Robson, sp. nov.

H. acmosepalo N. Robson affinis, sed caulibus tenuoribus arcuatis, foliis anguste

ellipticis vel oblongo-ellipticis venis lateralibus venam intramarginalem continuum

haud formantibus, petalis incurvis, staminibus stylisque brevioribus, inter alia

differt. Type: China, Yunnan, Kunming, Qiqu, 8.v.1956, B.-y.Qiu 51837 (KUN!,

holotype; SZ!, isotype).

H. hookerianum sensu R. Keller in Bot. Jahrb. 33: 549 (1904); H. Léveillé in Bull.
Soc. bot. Fr. 54: 590 (1908); Rehder in Sargent, P/. Wils. 2: 403 (1915) pro parte,
quoad spec. Pratt 292; non Wight & Arnott (1834).

Icon: —.

Shrub 0-5—1-5 m tall, with branches arching to spreading, slender, often rather lanky.
Stems orange, 4-lined, not ancipitous, sometimes becoming 2-lined or terete;
internodes 10-33 mm long, shorter than leaves; bark purplish-brown. Leaves
narrowly petiolate, with petiole 1-1-5 mm long; lamina 18-30(-45) x 6-11(-27) mm,
narrowly elliptic or rarely lanceolate-elliptic to + broadly oblong-elliptic, acute to
rounded, margin plane, base cuneate, paler beneath, not glaucous, chartaceous;
venation: 3(4) pairs main laterals, lower free or all looped, with subsidiaries and
intermediates sometimes almost as strong, but without visible tertiary reticulum;
laminar glands dots and short or very short streaks; ventral glands absent or rarely
(Hunan) dense. Inflorescence 1-3-flowered, from apical node; pedicels 2-7 mm long;
bracts reduced, narrowly elliptic to linear, persistent (?). Flowers 30-45 mm, in
diam., substellate to shallowly cyathiform; buds broadly (or rarely narrowly) ovoid,
acute or apiculate to obtuse. Sepals (16-)7-10 x (3—)3-5—5-5 mm, free or almost so,
imbricate, equal or subequal, erect or apically outcurved in bud, erect or suberect in
fruit, ovate to oblong-ovate or lanceolate, acute to obtuse, with margin entire to
minutely and + irregularly denticulate; midrib often obscure, veins not prominent;
laminar glands linear, sometimes distally interrupted, c. 12-14. Petals golden yellow,
+ shallowly incurved, 18-23 x 10-15(—18) mm, 2-5-3 xX sepals, + narrowly to rather
broadly obovate, with apiculus subterminal, rounded or obsolete; entire or inner
margin minutely glandular-denticulate. Stamen fascicles each with 40-45 stamens,
longest 12-18 mm, long, 0-6-0-7 x petals; anthers yellow. Ovary 5-8 x 3-5 mm,
narrowly ovoid-conic to ovoid; styles 4-7 mm long, 0-5—0-85 x ovary, free, rather
slender, suberect to gradually outcurved, more distinctly outcurved distally; stigmas
scarcely capitate. Capsule c. 12 x 8mm, ovoid-conic to ovoid. Seeds purplish-brown,
c. 1-4 mm long, narrowly cylindric, shortly apiculate, shallowly carinate, shallowly
linear-reticulate. 2n = ?

Roadside banks and grassy slopes; 1900-2550 m.
China (NE. Yunnan, W. Sichuan, Guizhou, W. Hunan). Map 18.

247

248 NORMAN K. B. ROBSON

CHINA. Yunnan: Lo Shiueh Mtn, 13.iv.1936, McLaren’s colls. U173 (BM, E);
Kunming to to Xiaguan, 260 km, 2400 m, 27.v.1981, Sino-Brit. Exp. Cangshan K149
(BM, E). Sichuan: Tianquan Xian, 1500 m, 4.vi.1956, He 44213 (SZ); Omei-hsien,
14. vii.1937, Chien 6098 (E). Guizhou: Shiheryung to Lungchang Puting, 14.v.1935,
Teng 0306 (IBSC); Kaili Xian, 19.v.1959, Cao 1440 (KUN). Hunan: Qianyang,
4.viii.1953, Li 199 (IBSC).

H. lagarocladumis related to H. acmosepalum but differs in its more slender, arching
stems, thinner, more elliptic and less discolorous leaves without a marked intramar-
ginal vein, erect or suberect sepals, usually shallowly incurved petals, relatively
shorter stamens and usually relatively shorter styles. The eastern form (from
Guizhou and western Hunan) has an ovoid or ovoid-conic ovary (4-5-6 mm wide)
and usually relatively narrow, acute to obtuse leaves, and is related to H. wilsonii;
the north-western form (from north and east Yunnan and west Sichuan) has a
narrowly ovoid-conic ovary (3-4-5 mm wide) and usually relatively broad, obtuse to
rounded leaves, and is related to H. addingtonii. Apart from the ovary width, which
shows a cline, all character ranges seem to overlap; and so it seems undesirable to
recognise these populations as subspecies.

The eastern form shows a trend eastward in which the habit becomes more
spreading, the sepals narrower, the flowers reduced to one, and the styles shorter.
These plants (from Guizhou and Hunan) have in most extreme form the slender,
spindly shoots which have suggested the specific epithet and approach H. wilsonii
most closely. For possible hybrids between H. lagarocladum and H. pseudohenryi,
see p. 284.

25(36). Hypericum wilsonii N. Robson

in J. Roy. Hort. Soc. 95: 492 (1970); [Lancaster], Hilliers’ man. trees & shrubs: 152
(1971); Bean, Trees & shrubs hardy in Br. Isles 8th ed. 2: 423 (1973). Type: China,
Hubei, Patung Hsien, vii.1907, Wilson 2419 (BM!, holotype; E!, K!, isotypes).

H. sp. sensu Rehder in Sargent, Pl. Wils. 3: 452 (1917).

H. kouytchense sensu Milne-Redh. in Curtis’s bot. Mag. 157: t. 9345 (1934) et hort.,
non H. Léveillé. (1904).

STUDIES IN THE GENUS HYPERICUM L.
Icon: Milne-Redh. in Curtis’s bot. Mag. 157: t. 9345 (1934).

Shrub 0-5-1 m tall, spreading, with branches pendulous to somewhat prostrate.
Stems red, 4-lined and ancipitous when young, becoming 2-lined, eventually terete;
internodes 10-45 mm long, shorter than to equalling leaves; bark grey-brown.
Leaves petiolate, with petiole 0-5—1-5 mm long; lamina 23—55(-60) x (6-)12-29 mm,
elliptic or elliptic-lanceolate to lanceolate or ovate-lanceolate, subacute or apiculate-
obtuse to rounded, margin plane, base broadly cuneate to rounded, paler or +
glaucous beneath, chartaceous; venation: 4-5 pairs main laterals, with midrib
obscurely branched distally, with scarcely visible tertiary reticulum; laminar glands
dots and short to longish streaks, ventral glands sparse or absent. Inflorescence
(1)3-c. 22-flowered, from 1(2) node(s), subcorymbiform; pedicels 8-12 mm long;
bracts foliar to narrowly lanceolate, deciduous? Flowers 40-50(-60) mm in diam.,
stellate; buds narrowly ovoid, acute to subacuminate. Sepals 7-10 x 2-4-5 mm, free,
imbricate, equal, + erect in bud and fruit, lanceolate or narrowly elliptic, acutely
acuminate to shortly aristate, with margin entire; midrib conspicuous, veins not
prominent; laminar glands linear, c. 10. Petals golden yellow, not tinged red,
spreading, 20-25 x 12-20 mm, 2-5-3 X sepals, obovate, with apiculus subterminal,
acute; margin minutely glandular denticulate towards apex or entire. Stamen fas-
cicles each with 30-35 stamens, longest 10-15 mm long, 0-35-0-5 x petals; anthers
golden yellow. Ovary 5-6 x 3-3-5 mm, + broadly ovoid; styles 7-9 mm long, c.
1-5-1-8 X ovary, free, erect, outcurved at apex; stigmas small. Capsule 7-11 x 5-6
mm, ovoid. Seeds dark orange-brown, narrowly cylindric, scarcely carinate, +
shallowly linear-reticulate. 2n = ?

Cliffs and thickets; (600?)1000-1750 m.

China (W. Hubei, E. Sichuan and possibly Jiangsu; also recorded from Hunan).
Map 18.

CHINA. Hubei: S. Patung, 1885-1888, Henry 6099 (BM, K); no precise loc.,
vi.1907, Wilson 2216 (E, K). Sichuan: Wushan Hsien, 1000 m, 1907, Wilson 256
(seed only); Huidong Xian, 1750 m, 3.vii.1978, Zhao 5802 (SZ).

CULTIVATED. Specimens seen from England (1915), Scotland (1977), Ireland
(1930-80), France (1929), U.S.A. (1929).

In H. wilsonii, the spreading habit developed in the eastern form of H. lagarocladum
is more pronounced, so that its stems frequently trail over the ground, and it has (i)
relatively and absolutely shorter stamens and (ii) a smaller ovary, so that the styles
are relatively longer. The typical specimens that I have seen are restricted to a small
region in the Yangtze valley near the Hubei—Sichuan border; but Li (ined.) records it
from Hunan.

All the cultivated material of H. wilsonii would appear to have originated from
Wilson 256 (see above). Anon. 5809, from Jiangsu? (Buyijiao R. area) has narrower
petals, relatively longer styles, a subglobose ovary and oblong-elliptic discolorous
leaves, all characters that tend towards H. dyeri.

26(37). Hypericum dyeri Rehder

in J. Arnold Arbor. 20: 422 (1939); N. Robson in K. H. Rechinger, FI. iranica 49: 4
(1968), inJ. Roy. Hort. Soc. 95: 490 (1970), in Nasir & Ali, Fl. W. Pakistan 32: 4,
f. 1 A-D (1973); Bean, Trees & shrubs hardy in Br. Isles 8th ed. 2: 423 (1973); N.
Robson in Hara & Williams, Enum. fi. pl. Nepal 2: 61 (1979). Types: India, Uttar
Pradesh, east Kumaon, Blinkworth in Wallich 4817A (K-W!, lectotype; BM!).

H. govanianum Wallich, Numer. list.: No. 4815 (1831), nomen.

H. coriaceum Royle, Illustr. bot. Himal.: 131 (1834), nomen, non Blume (1825).

H. lysimachioides Wallich [Numer. list.: No. 4817 (1931) nomen] ex Dyer in Hook. f.,
Fl. Brit. Ind. 1: 254 (1874); Collett, Fl. simlensis: 56 (1902); R. Keller in Engler &
Prantl, Nat. Pflanzenfam. 2nd ed. 21: 176 (1925); Parker, For. fi. Punjab, etc. 2nd
ed.: 28 (1934); non Boiss. & Noé (1853). Types as for H. dyeri.

H. cernuum sensu Dyer in Hook. f., Fl. Brit. Ind. 1: 253 (1874) pro parte, quoad syn.
H. govanianum, non H. cernuum Roxb. ex. D. Don.

249

250 NORMAN K. B. ROBSON

Icon: N. Robson in Nasir & Ali, Fl. W. Pakistan 32:5, f. 1 A-D.

Shrub 0-6-1-2 m tall, with branches + widely spreading. Stems orange to yellow-
brown, 2—4-lined and ancipitous when young, soon terete or persistently 2-lined;
internodes 10-30(-55) mm, shorter than or very rarely exceeding leaves; bark
grey-brown. Leaves petiolate, with petiole 1-2 mm long; lamina 10—50(-60) x
5—24(-35) mm, oblong-lanceolate or lanceolate to ovate, acute to apiculate-obtuse
or rounded, margin plane, base cuneate to rounded, discolorous, markedly glaucous

beneath, chartaceous to papyraceous; venation: 2-3 pairs main laterals, the upper
sometimes forming a + prominent partial inframarginal vein, with midrib pinnately
branched, the branches forming a rudimentary reticulum, tertiary reticulum other-

wise very rarely visible; laminar glands short streaks and dots; ventral glands dense.
Inflorescence 1-31(or more)-flowered, from 1—2-nodes, subcorymbiform to broadly
pyramidal, without or with 1 pair of branches below; pedicels 6-11(—16) mm long;
bracts foliaceous to linear-lanceolate, gradually smaller distally, persistent. Flowers
(15-)20-30(-35) mm in diam., stellate; buds ovoid, subacute to rounded. Sepals
4-12 x 1-2-5(-3) mm, free, imbricate at base, equal, spreading in bud and fruit,
narrowly oblong-lanceolate to linear, acute, with margin entire; midrib visible, veins
not prominent; laminar glands interrupted-linear, c. 6-10. Petals bright yellow, not
tinged red, spreading, 10-18 x 3-7(-9) mm, c. 1-5—2-5 x sepals, oblanceolate, with
apiculus subterminal and rounded or absent; margin entire, eglandular. Stamen
fascicles each with c. 20 stamens, longest 8-12(-15) mm long, c. 0-7-0-85 x petals;
anthers bright yellow. Ovary 3-5-5 x 2-8-4 mm, broadly ovoid to subglobose; styles
4-7 mm long, 1-5-2 X ovary, free, suberect or gradually divergent, spreading at
apex; stigmas capitate. Capsule 7-10 x 5—7 mm, subglobose. Seeds dark reddish-
brown, c. 0-9 mm long, narrowly cylindric, not or narrowly carinate, shallowly
linear-foveolate. 2n = 20.

On dry open slopes, steep banks, and cliffs, sometimes in woodland, rarely in forest
(central Nepal, atypical); 1300-3000 m.

Nepal (Central, West), India (Bihar, Kumaun to Kashmir), Pakistan (Hazara, Swat,
Dir). Map 19.

Map 19 Sect. 3. Ascyreia: 26. H. dyeri @.

STUDIES IN THE GENUS HYPERICUM L.

NEPAL. Central: Kali Gandaki Valley, 1800 m, 31.v.1954, Stainton, Sykes &
Williams 5478 (BM); Lukarban Khola, W. of Beni, 5.v.1954, Stainton, Sykes &
Williams 447 (BM). West: near Maikot, 2550 m, Stainton, Sykes & Williams 3302
(BM, E).

INDIA. Bihar: Suratpur, x.1942, Mundkur 100 (A). Uttar Pradesh: N. Gharwal,
Nandagiri Valley, 2460 m, 11.vi.1919, Osmaston 1061 (A); Mussoorie, Old Brewery
Road, 1950 m, vi.1914, Anderson (E). Himachal Pradesh: Simla, Jakko, 27.vi.1916,
Rich 257 (K); Chamba, Kilor, Bhandal Valley, 2100 m, 30.ix.1919, Parker (K).
Kashmir: Basaoli [Basoli], 1500 m, 26.vii.1876, Clarke 31615 (K); Batol Pass, Jumu
Road, c. 1800 m, vili.1931, Stewart 12991 (A).

PAKISTAN. Hazara: Murree Hills, Changla Gali, c. 2100 m, 27. viii.1918, Stewart
3910 (A, RAW). Swat: Kulali, c. 11 km up Swat R. from Bahrein, c. 1650 m, Rodin
5608 (RAW). Dir: Gujar to Dir, 2100-2400 m, 30.viii. 1962, Stewart, Nasir & Siddiqi
1432 (RAW).

CULTIVATED. Specimens seen from Eire (1922), U.S.A. (1921) and St. Helena
(1832).

Despite their wide disjunction, H. dyeri is clearly related to H. wilsonii (q.v.).
Presumably the disjunction was caused, or at least widened, by the eruption of the
Himalayan massif.

One of the Nepal specimens of H. dyeri (Stainton, Sykes & Williams 447) has large,
thin leaves and relatively long pedicels, and is said to grow ‘in forest’.

27(38). Hypericum addingtonii N. Robson, sp. nov.

H. x cyathifloro N. Robson similis, sed foliis latioribus, ovato-lanceolatis vel
oblongo-lanceolatis vel elliptico-oblongis, apice acutis vel obtusis vel rotundato-
apiculatis, alabastris apice obtusis, sepalis latioribus apice apiculatis vel obtusis vel
rotundato-apiculatis, petalis suborbicularibus, ovario ovoideo breviori, differt.
Type: cultivated, England, W. Sussex, Fittleworth [ex horto cornubiensil], 6.x.1972,
Addington (BM!, holotype).

H. leschenaultii sensu hort. pro parte, non Choisy (1821).

Icon: Fig. 17

Shrub 1-5-2 m tall, spreading to 2:5 m wide, with branches arching to spreading.
Stems yellow-brown, 4-angled but not ancipitous when young, soon terete; inter-
nodes 10-50 mm long, shorter than leaves. Leaves petiolate, with petiole 1-2-5 mm
long; lamina (20—)25-85 x 10-35 mm, elliptic-oblong to ovate-lanceolate or oblong-
lanceolate, apiculate or obtuse to rounded, margin plane, base cuneate, paler
beneath, not glaucous, chartaceous; venation: 3—4(5) pairs main laterals, all or only
distal ones closed, with subsidiaries rarely prominent and tertiary reticulum not
visible; laminar glands dots and short streaks; ventral glands absent or rarely sparse.
Inflorescence 1-—3(—5)-flowered from apical node; pedicels 2-10 mm long; bracts
reduced, lanceolate, persistent. Flowers (30—)50-65 mm in diam., shallowly cyathi-
form; buds ovoid, obtuse. Sepals 7-10 x 4-5—6-2 mm, free, imbricate, sub-equal,
erect in bud and fruit, ovate to oblong-ovate or oblong-spathulate, acute or apiculate
to obtuse or rounded-apicuiate, margin entire or minutely denticulate, sometimes
narrowly hyaline; midrib distinct, veins becoming prominent after flowering; lami-
nar glands linear or interrupted, numerous. Petals golden yellow, shallowly in-
curved, (20—)25-32 x (12-)15-32 mm, 34 x sepals, broadly obovate to subcircular,
with apiculus lateral rounded; margin entire. Stamen fascicles each with 40-45
stamens, longest 12-15 mm long, c. 0-4 xX petals; anthers yellow. Ovary 5-7 xX
3-5 mm, ovoid; styles 4-5—5(—7) mm, 0-7-0-8(-1) X ovary, free, suberect, outcurving
near apex; stigmas scarcely capitate. Capsule c. 20 x 10-12 mm, ovoid to cylindric-
ovoid. Seeds dark reddish-brown, 1-1-2 mm long, cylindric, not or scarcely carinate,
shallowly linear-reticulate. 2n = 20 ?

Habitat unknown; 1800-3100 m.
China (NW. Yunnan). Map 18.

251

O+4 &

252 NORMAN K. B. ROBSON

Fig. 17 H. addingtonii: (a) habit; (b) leaf section; (c) flower bud; (d) sepal; (e) petal; (f) stamen fascicle;
(g) anthers; (h) capsule (a x 1; b-f, h x 4; g x 20). All Wayments.n.

STUDIES IN THE GENUS HYPERICUM L.

CHINA. Yunnan: Gongshan [Kungshan] 9.x.1940, Feng 8297 (KUN); Shu-yang
valley, lat. 25°20'N, 1800 m, 13.v.1913, Forrest 9928 (E, K); headwaters of Red
River or Menghuaho, from Maokai to Tatsang, 2200 m, 8-12.iv.1922, Rock 3030
(EO).

CULTIVATED. England: Hampshire, Ampfield, Hilliers’ Jermyns Lane
Nursery, 29.vi.1970, Lancaster s.n. (BM); Manchester, Fallowfield, Whitworth
Lane, Univ. Bot. Dept. Experimental Gdns, 3.vii.1981, Wayment s.n. (BM). Also
seen elsewhere, e.g. at the Royal Horticultural Society’s Garden, Wisley (1983).

H. addingtonii, which occurs only in north-west Yunnan, was probably introduced to
cultivation by Forrest and appears to be quite widespread in British gardens. It has
been confused (by me as well as others) with H. leschenaultii and H. beanii; but the
chartaceous leaves and sepals, the sepal shape and constantly erect attitude, and the
spreading habit are among the characters distinguishing it from H. leschenaultii,
whilst it differs from H. beanii by the spreading habit, the more slender and soon
terete stems, and the often larger flowers with much shorter stamens.

H. addingtonii is most closely related to H. lagarocladum, in which the stem is
often 4-lined, the leaves always broadest at the middle, acute to obtuse and usually
smaller, and the flower generally more primitive (i.e. with narrower acute sepals,
narrower petals, longer stamens and often narrower ovary and capsule). It shows an
unusually wide variation in leaf size (20-85 mm in length). The Rock specimen is
somewhat atypical, having narrower sepals than usual, reddish stems and a few
ventral leaf-glands. In some of these respects it approaches H. lagarocladum and in
others H. hookerianum.

27x(38x). Hypericum x cyathiflorum N. Robson, hybr. nov.

Hybrida hortensis e Hyperico addingtonii N. Robson et H. hookeriano Wight & Arn.
verosimiliter exorta, longitudine staminorum inter eas species media, ad illud foliis
sepalisque acutis vel apiculatis, ad hoc foliis angustibus accedens. Standard speci-
men: Cultivated, Jackman’s Nursery, Woking, Surrey, England. 13.vii.1970, Jack-
man s.n. (BM!).

H. ‘Lawrence Johnston’ hort.

H. patulum ‘Gold Cup’ Jackman (1952).

H. hookerianum ‘Gold Cup’ Jackman (c. 1960).

H. beanii ‘Gold Cup’ N. Robson in J. Roy. Hort. Soc. 95: 488, in clav. (1970);
[Lancaster], Hilliers’ man. trees & shrubs: 150 (1971); Bean, Trees & shrubs hardy
in Br. Isles 8th ed. 2: 406 (1973).

Icon: N. Robson in J. Roy. Hort. Soc. 95: f. 236 (1970).

Shrub up to c. 1-5 m tall, with branches widely spreading. Stems orange, 4-angled but
not ancipitous when young, soon 2-lined, eventually terete; internodes 10-60 mm
long, shorter than leaves; bark reddish-brown. Leaves rather broadly petiolate, with
petiole 1-2-5 mm long; lamina 30-75 x 8-20 mm, lanceolate, acute to apiculate,
margin plane, base cuneate, paler beneath, not glaucous, chartaceous; venation: 4-5
pairs main laterals, all free or subdistally closed, with subsidiaries not prominent and
obscure tertiary reticulum; laminar glands dots and short streaks; ventral glands
absent. Inflorescence 1-9-flowered, from 1-3 nodes, subcorymbiform; pedicels 5—
10 mm long; bracts reduced, + narrowly elliptic, persistent. Flowers 40-50 mm in
diam., cyathiform; buds narrowly ovoid, acute. Sepals 6-9 x 3-3-4-5 mm, free,
imbricate, subequal, erect in bud and fruit, ovate-lanceolate or elliptic to oblong-
spathulate, acute to apiculate-obtuse or rarely rounded-apiculate, margin entire to
minutely denticulate, hyaline, sometimes reddish; midrib + distinct, veins not
prominent; laminar glands linear, c. 12. Petals golden yellow, + shallowly incurved,
20-30 x 15-19 mm, 2-5-3 xX sepals, broadly oblong-obovate, with apiculus subter-
minal, obtuse; margin entire. Stamen fascicles each with c. 45 stamens, longest 9-13
mm long, 0-3-0-4 x petals; anthers yellow. Ovary 7-8 x 4-5-5 mm, ovoid-conic;
styles 4-5-5 mm long, c. 0-6 X ovary, free, suberect, outcurved near apex; stigmas
truncate. Capsule c. 20 X 10 mm, ovoid-conic. Seeds not seen. 2n = 30.

253

254

NORMAN K. B. ROBSON
Known only in cultivation.

CULTIVATED. England: Wisley, Surrey, Royal Horticultural Society’s Garden,
29.vii.1970, Robson s.n. (BM); Winchester, Hants, Hilliers’ Old Winchester
Nursery, 27.vii.1970, Lancaster s.n. (BM). Scotland: Benmore, Argyll, Younger
Bot. Gard., 4.ix.1977, Herb. Edinb. C. 11670 (BM).

H. x cyathiflorum was apparently introduced to cultivation by Jackman’s of Woking,
Surrey, England, about 1952 as H. patulum ‘Gold Cup’. The name was later (c. 1960)
changed to H. hookerianum ‘Gold Cup’, probably on account of the terete mature
branches; and I (Robson, 1970), on account of its acute to subacute sepals and buds,
placed it under H. beanii.

The source of Jackman’s material has not been located. It may be significant,
however, that this plant is grown at Hidcote Manor, Gloucestershire, under the
name ‘Lawrence Johnston’, after the late owner of Hidcote Manor and maker of its
garden. A chance seedling originating there might well have been thus named. The
chromosome number (2n = 30) suggests that it is a triploid on the base x = 10, and its
relatively poor fruiting would tend to support this hypothesis (see also 27xx. H. x
‘Hidcote’). I have not seen any wild-collected material of this plant, which is
strikingly intermediate between H. addingtonii and H. hookerianum. It is therefore
very likely that it arose in cultivation in western Europe, probably in Britain after the
introduction of H. addingtonii by Forrest (?). As other hybrids between H. adding-
tonii and H. hookerianum may be (or may have been) produced, I propose that the
plant described above to be known as H. x cyathiflorum ‘Gold Cup’.

27xx(38xx). Hypericum x ‘Hidcote’, hybr. nov.
(27x. H. x cyathiflorum ‘Gold Cup’ x 14. H. calycinum)

H. patulum ‘Hidcote Variety’ Synge in J. Roy. Hort. Soc. 75: C (1950). Standard
specimen: cultivated, England, Surrey, Woking, Chobham, T. Hilling & Co’s
nurseries, 1.viii.1950, Anon. (K!).

H. patulum ‘Hidcote’ Plaisted & Lighty in Nat. hort. Mag. 38: 124 (1959) et auct.
hort.

H. ‘Hidcote’ Thomas in Gdnrs’ Chron. III, 147: 227, 254t. (1960); N. Robson in J.
Roy. Hort. Soc. 95: 491, f. 244 (1970); [Lancaster], Hilliers’ man. trees & shrubs:
151 (1971); Bean, Trees & shrubs hardy in Br. Isles 8th ed. 2: 413 (1973); N.
Robson in Bull. Br. Mus. nat. Hist. (Bot.) 8: 170 (1981).

H. hookerianum ‘Hidcote’ I. F. Schneider in Meded. Inst. Vered. Touinbouwgew.
252: 20 (1966), in Dendrofl. 2: 20 (1967).

H. ‘Hidcote Gold’ sensu hort. amer.

Icones: Thomas in Gdnrs’ Chron. III, 147: 227, 254 (1960).

Shrub up to 1-75 m tall, bushy, with branches arching to spreading. Stems reddish to
orange, 4-angled and slightly ancipitous when very young, usually soon terete or
rarely persistently 2-4-lined; internodes 10-40 mm long, usually shorter than leaves;
bark reddish-brown. Leaves petiolate, with petiole 0-5—2 mm long; lamina 30-60 x
10-25 mm, triangular-lanceolate, acute to obtuse or sometimes slightly mucronate,
margin plane, base broadly to rather narrowly cuneate, paler beneath, not glaucous,
chartaceous; venation: 3-5 pairs main laterals, all closed, with subsidiaries and lax
tertiary reticulum + conspicuous; laminar glands dots and short streaks; ventral
glands rather dense to sparse. Inflorescence 1-c. 16-flowered, from 1-2 nodes,
subcorymbiform; pedicels 6-13 mm long; bracts elliptic to narrowly oblong, even-
tually deciduous. Flowers (35—)40-65 mm in diam., cyathiform; buds broadly ovoid,
obtuse to rounded. Sepals 8-11 x 6-9 mm, free, imbricate, equal to subequal, erect
in bud and fruit, ovate-oblong to broadly elliptic, rounded, margin finely ciliolate or
denticulate distally, hyaline; midrib usually obscure, veins not prominent; laminar
glands linear, numerous. Petals golden yellow, shallowly incurved, 15-35 x 15-25
mm, 2-3 X sepals, obovate, with apiculus lateral, rounded; margin entire. Stamen
fascicles each with 40-60 stamens, longest 8-12 mm long, 0-35-0-5 x petals; anthers
orange. Ovary 7-9 x 5-6 mm, ovoid; styles 8-10 mm long, 1-1-5 x ovary, free, erect

STUDIES IN THE GENUS HYPERICUM L.

or suberect, outcurved near apex; stigmas small. Capsule deciduous when still
immature. Seed not developing. 2n = c. 50 (c. 48, c. 54, 55).

Known only in cultivation.

CULTIVATED. England: Kew, Arboretum Nursery, 8.viii.1959, Anon. (K);
Wisley, Roy. Hort. Soc’s Garden, vii.1966, Brickell s.n. (BM). Ireland: Offaly, Birr
Castle, 6.x.1980, Anon. (DBN). New Zealand: Canterbury, Lincoln, D.S.I.R.,
14.1.1980, Sykes 2/80 (BM).

Also seen in Scotland and France. Grown also in the U.S.A. and many other
countries.

No material of H. x ‘Hidcote’ from natural habitats has been seen; and it appears to
be always sterile, the flower dropping after the fruit has begun to swell. Morphologi-
cally it is intermediate between H. x cyathiflorum and H. calycinum, and cytologi-
cally an origin from these two species with chromosome doubling is possible (c.15 +
10 = c.25 X 2 = c.50). Circumstantial evidence indicates that the cross may have
occurred (probably spontaneously) at Hidcote Manor, Gloucestershire, where both
suspected parents were and are growing (Robson, 1981). In an attempt to resynthe-
sise H. x ‘Hidcote’, the cross H. x cyathiflorum ‘Gold Cup’ 9 x H. calycinum CO was
made by Miss Julie Westfold in 1984, but the results are not yet available.

H. x ‘Hidcote’, like most other artificial Hypericum hybrids, shows a tendency to
produce shoots with variegated leaves. These are narrowly lanceolate, narrower
than in H. x moserianum ‘Tricolor’, and the whole shoot looks weak and sickly.
Plants have nevertheless been grown from cuttings of such shoots and sold by
nurseryman as H. ‘Hidcote variegatum’. It seems unnecessary to encourage the
distribution of such a poor plant by formally describing it.

28(39). Hypericum hookerianum Wight & Arnott

{in Wight, Cat. Ind. Pl.: 20, No. 332 (1833) nomen] Prodr. fi. penin. ind. or. 1: 99
(1834); Wight, Icon. pl. ind. or. 3: t. 959 (1845); Dyer in Hook, f., Fl. Brit. Ind. 1:
254 (1874), excl. var. leschenaultii; Gamble, Fl. pres. Madras 1: 70 (1915); Fyson,
Fl. Nilgiri & Pulney hill-tops 1: 37 (1915), 2: 29, t. (1915); R. Keller in Engler &
Prantl, Nat. Pflanzenfam. 2nd ed. 21: 176 (1925); Fl. S. Ind. hillstns 1: 47 (1932), 2:
30 (1932); Y. Kimura in Hara, Fl. eastern Himalaya: 210 (1966); Mathew in Rec.
Bot. Surv. India 20: 45 (1969); N. Robson in J. Roy. Hort. Soc. 95: 490 (1970);
[Lancaster], Hilliers’ man. trees & shrubs: 151 (1971); Bean, Trees & shrubs hardy
in Br. Isles 8th ed. 2: 414 (1973); N. Robson in Hara & Williams, Enum. fi. pl.
Nepal 2: 61 (1979). Type: India, ‘Southern Provinces’, Wight 332 (K!, holotype;
E!, isotype).

H. patulum sensu D. Don, Prodr. fl. nepal.: 218 (1825), non Thunb. ex Murray
(1774).

H. oblongifolium sensu Wallich, Pl. asiat. rarior. 3: t. 244 (1832) et auct. mult., non
Choisy (1821).

Norysca hookeriana (Wight & Arnott) Wight, Illustr. Ind. bot. 1: 110 (1840)
‘Norisca’; Blume, Mus. bot. Lugd. Bat. 2: 22 (1856); Y. Kimura in Nakai & Honda
Nova fl. jap. 10: 98 (1951).

H. patulum subsp. (y) hookerianum (Wight & Arnott) Kuntze, Rev. gen. pl. 1: 60
(1891).

H. patulum var. oblongifolium sensu Koehne, Deutsche Dendrol.: 415 (1893), non
H. oblongifolium Choisy (1821).

H. garrettii Craib in Kew Bull. 1913: 66 (1913). Type: Thailand, Doi Intanon,
2142-2165 m, 12 & 26. x.1910, Garrett 67 (K!, holotype; BM!, E!, isotypes).

Norysca urala var. angustifolia sensu Y. Kimura in Hara, Fl. eastern Himalaya, 2nd
Rep.: 82 (1971) pro parte, quoad spec. ex Bhutan.

Icon: Wallich, Pl. asiat. rarior. 3: t. 244 (1832).

Shrub 0-3-2-1 m tall, bushy, round-topped, with branches erect to spreading. Stems
red to yellowish, 4-lined and ancipitous when young and usually soon terete, or

255

256

) + <=

NORMAN K. B. ROBSON

always terete; internodes 12-60 mm long, shorter than to exceeding leaves; bark
grey-brown. Leaves petiolate, with petiole 14 mm long; lamina (17—)25-78 x
(7—-)10-32 mm, narrowly lanceolate or oblong-lanceolate to broadly ovate, acute or
obtuse to apiculate or rounded, margin plane, base narrowly cuneate to subcordate,
paler or + glaucous beneath, chartaceous; venation: (2)3—4 pairs main laterals, with
midrib pinnately branched above, without visible tertiary reticulum; laminar glands
short to very short streaks and dots; ventral glands absent or + dense. Inflorescence
1-5-flowered, from apical node, subcorymbiform; pedicels 3-16 mm long; bracts
lanceolate or narrowly oblong to obovate-spathulate, deciduous. Flowers 30-60 mm
in diam., + deeply cyathiform; buds broadly ovoid to subglobose, broadly obtuse to
rounded. Sepals 5-10 x 4-8 mm, free, (often broadly) imbricate, subequal, erect in
bud and fruit, obovate or obovate-spathulate to subcircular or elliptic or oblong-
elliptic, rounded or rarely rounded-apiculate, with margin entire or rarely very finely
eroded-denticulate; midrib visible or + obscure, veins often prominent, especially in
fruit; laminar glands linear, sometimes interrupted near apex, numerous. Petals
deep golden to pale yellow, not tinged red, markedly incurved, 15-30 x 15-25 mm,
c. 3 X sepals, broadly obovate to subcircular, with apiculus subterminal, obtuse to
rounded, or absent; margin entire, eglandular. Stamen fascicles each with 60-80
stamens, longest 5—9 mm long, 0-25—0-35 x petals; anthers golden yellow, sometimes
tinged red ? Ovary S—7(-8) x 4-5(—6) mm, broadly ovoid, acute; styles 24(-7) mm
long, 0-35-0-7 (-0-9) xX ovary, free, gradually outcurved towards apex; stigmas
narrowly capitate. Capsule 9-17 x 7-12 mm, ovoid to ovoid-conic. Seeds dark
reddish-brown, 0-7-1 mm long, cylindric, not or scarcely carinate, shallowly linear-
reticulate. 2n = 20?

Dry to moist, open to half-shaded habitats (grassy or rocky slopes or cliffs, thickets,
open forest, forest margins); 1200-3000 m.

China (west Yunnan, south-east Xizang [Tibet]), Thailand (north), Burma (north,
central, and Mt. Victoria), India (Arunachal Pradesh, Bengal, Meghalaya,
Manipur; Mysore, Tamil Nadu), Bhutan, Tibet (south-east), Sikkim, Nepal (east).
Map 20.

CHINA. Yunnan: above Yangbi, Ziyang, 2560 m, 10.v.1981, Sino-Brit. Exped.
Cangshan 469 (BM, E); Xiaguan, Erhai Park, 1950 m, 1.v.1981, Sino-Brit. Exped.
Cangshan 10 (E).

THAILAND. Northern: Chiang Mai, Doi Inthanon, c. 2400 m, 2.v.1921, Kerr
5312 (ABD, BM, K); Chiang Rai, Doi Chong, north part, 1850 m, 19.11.1968,
Hansen & Smitinand 12674 (AAU, C, E, K).

BURMA. Shan: Gussampo Patmung, Zawgyi Reserve, 2100 m, pre 25.vii.1921,
Rogers (E). Kachin: hills NW. of Htawgaw, 2400 m, vii.1925, Forrest 27024 (A, E,
K). Chin: Mt. Victoria, Mindat Ridge, 2250 m, 22.v.1956, Kingdon Ward 22262
(BM).

BANGLADESH. Chittagong: Palutalun, Rangbi, 4.viti.1916, Cowan 289 (E.).

INDIA. Manipur: Sirhoi, on the ridge, 2400 m, 16.vii.1948, Kingdon Ward 17796
(BM). Meghalaya: Khasi Hills, Mawphlang, c. 1800 m, 28.vi.1953, Koelz 33232 (L).
Arunachal Pradesh: Delei Valley, 1500-1800 m, 27.vii.1928, Kingdon Ward 8484
(K); Pachakshiri, Nyug La, 2700 m, 5.vii.1936, Ludlow & Sherriff 1906 (BM);
Balipara Frontier, Phutang, 2100 m, 16.v.1935, Kingdon Ward 11438 (BM). W.
Bengal: Darjeeling, Tonglo, 3000 m, 12.ix.1875, Clarke 27530 (BM, K). Mysore:
Baba Budan Hills, vi.1847, Cleghorn (E). Tamil Nadu: Nilgiri Hills, Ootacamund,
Lovedale, 2100 m, x.1889, Gamble (K); Anaimalai Hills, pre 1885, Metz in Beddome
379 (BM); Palnai Hills [Pullney], 1915, Sauliére 303 (BO).

BHUTAN. North: Shimitang, Tsampa, 3300 m, i.vii.1949, Ludlow & Sherriff
19273 (BM). Central: Punakha, Sinchu La, 3000 m, 4.ix.1967, Bowes-Lyon 5073
(BM). South: Deothang distr., Tshilingor to Riserboo, 8 km, c. 2100 m, 26.vi.1979,
Grierson & Long 2270 (E).

TIBET. Chayul Chu, 6-4 km below Lung, 2400 m, 10. vii. 1936, Ludlow & Sherriff
2337 (BM); Pad Lo, near Chumbi, 17.vii.1878, Dungboo (BM, K).

SIKKIM. Karponang, 2700 m, 25.vi.1945, Bor & Kirat Ram 20604 (L).

NEPAL. East: Barun Khola, 3000 m, 25.xi.1971, Beer 12316 (BM). Central:

STUDIES IN THE GENUS HYPERICUM L. 257

Map 20 Sect. 3. Ascyreia: 28. H. hookerianum @ , 28a. H. hookerianum ‘Charles Rogers’ O.

Khatmandu, Sheopuri Lekh, 1950 m, 19.viii.1954, Stainton, Sykes & Williams 6940
(BM).

CULTIVATED. Specimens have been seen from England, Scotland, Wales, and
Eire.

H. hookerianum is a variable species related to H. addingtonii. From Yunnan, where
the form most similar to that species occurs in the Dali region, it shows clinal
variation in two directions: (i) through central Burma (‘rogersii’), Manipur and
Meghalaya to the Himalayan range at Bhutan and thence west to central Nepal and
east to northern Burma; (ii) through central Burma to northern Thailand (‘garrettii’),
western Burma (Mt. Victoria) and southern India. Both clines show trends from red
4-angled stem, small acute ovate-lanceolate leaves and large flowers with rounded-
apiculate sepals (Yunnan) to orange terete stem, large narrowly lanceolate rounded
leaves and small flowers with obovate-spathulate rounded sepals. In cline (i) the
leaves in the early stages are glaucous beneath with dense ventral glands, whereas the

258

NORMAN K. B. ROBSON

later stages (most Himalayan specimens) resemble plants from Thailand, Mt.
Victoria and southern India in having leaves pale beneath without ventral glands. In
Meghalaya there is a local variation in which the leaves become very broadly ovate
and the plant looks superficially like H. bellum subsp. latisepalum.

These variation trends appear continuous, so that, despite the wide range of H.
hookerianum, no infraspecific taxa can be recognised. Plants have been introduced
into cultivation, however, from two distinct parts of the range, viz. the (i) Himalaya
and Mt. Victoria (showing advanced characters as described above) and (ii) central
Burma. The latter form has been named var. rogersii and merits recognition as a
cultivar (Map 20):

28a(39a). H. hookerianum ‘Charles Rogers’, nom. nov.

H. hookerianum [var.] rogersii hort., nomen; Thomas in Gdnrs’ Chron. III, 147: 254
(1960).

H. rogersii hort., nomen.

H. hookerianum ‘Rogersii’ N. Robson in J. Roy. Hort. Soc. 95: 484 (1970); Bean,
Trees & shrubs hardy in Br. Isles 8th ed. 2: 415 (1973), nom. illegit. in forma latina.
Standard specimen: [cult. ex ?] Burma, Shan State, Gussampo Patmung, Zawgyi
Reserve, C. G. Rogers s.n. (E!).

Icon: —.

Stem stout, more persistently 4-angled than in the type; leaves oblong-ovate to
broadly ovate, acute to obtuse; flowers 1-3, 40-60 mm in diam., on stout pedicels;
sepals subcircular to obovate-spathulate, becoming markedly ribbed.

H. hookerianum has been confused with H. choisianum, as they both occur in similar
habitats in the eastern and central Himalaya. They are, however, nearly always
easily distinguishable morphologically (see p. 273); and H. hookerianum occurs at
lower, though overlapping, altitudes (1200-3300 m as against 2400-4800 m for H.
choisianum in the same area).

29(40). Hypericum lacei N. Robson, sp. nov.

H. henryi H. Léveillé & Vaniot affinis, sed foliis oblongis vel ovato-oblongis,
staminis quam petalis pro ratione brevioribus, staminis et stylis et capsulis longiori-
bus, differt. Type: Burma, Ruby Mines Distr., Kyatpyin, 22.x.1912, Lace 6275 (K!,
holotype; E!, isotype).

Icon: Fig. 18.

Shrub, with stems probably erect, not or weakly (?) frondose. Stems orange, 4-lined
and ancipitous when young, eventually terete; internodes 10-20 mm long, shorter
than leaves; bark reddish-brown. Leaves petiolate, with petiole c. 1 mm long; lamina
25-51 X 8-18 mm, oblong to elliptic-oblong, apiculate-obtuse, margin plane, base
cuneate, densely glaucous beneath, chartaceous; venation: 2-3 pairs main laterals
from near base, innermost joining upper laterals to becoming incomplete and +
undulating intramarginal vein, all rather dark (when dry), without visible tertiary
reticulum; laminar glands very short streaks and dots; ventral glands dense. In-
florescence 1-6-flowered, from one node, with short apical internode, subcorymbi-
form; pedicels 6 mm long; bracts narrowly oblong to lanceolate, deciduous. Flowers
40-50 mm in diam., shallowly cyathiform; buds ovoid, subacute to obtuse. Sepals
5-6 x 3-5 mm, free, imbricate, + unequal, erect in bud and fruit, elliptic to obovate,
obtuse to rounded, usually apiculate, with margin hyaline and minutely eroded-
denticulate near apex; midrib sometimes apparent, veins not prominent; laminar
glands linear or interrupted, numerous. Petals golden yellow, slightly incurved,
20-25 x 15-17 mm, c. 4 X sepals, obovate, with apiculus subterminal, rounded,
margin entire, with a row of inframarginal gland dots. Stamen fascicles each with c. 50
stamens, longest 9-12 mm long, c. 0-35-0-5 x petals; anthers yellow. Ovary 6-7 x
4-5 mm, broadly ovoid to ovoid-conic; styles S—6 mm long, 0-75 X to almost as long
as ovary, free, outcurved towards apex; stigmas small. Capsule 16-20 x 8-10 mm,
ovoid-conic. Seeds unknown. 2n = ?

STUDIES IN THE GENUS HYPERICUM L. 259

Fig. 18 H. lacei: (a) habit; (b) leaf section; (c) flower bud; (d) sepal; (e) petal; (f) stamen fascicle; (g)
anthers; (h) capsule (a x 1; b-f, h x 4; g x 40). All Lace 6275.

260 NORMAN K. B. ROBSON
Habitat unknown.

Burma (Shan State). Map 21.
BURMA. Shan: Ruby Mines Distr., Kyatpyin, 22.x.1912, Lace 6275 (type);
Upper Burma, pre-1897, Burke 346 (E).

H. lacei is an apparently narrowly localised species that resembles H. henryi most
closely in general; but, in its oblong to elliptic-oblong leaves and square (i.e. not
remaining ancipitous) young stems, it is nearer H. lagarocladum and H. acmosepa-
lum. It is thus clearly a relict species.

By ace mes KS \
Map 21 Sect. 3. Ascyreia: 29. H. lacei A, 30a. H. henryi subsp. hancockii * (part; see also Maps 22, 23),
30b. H. henryi subsp. henryi O, 31. H. patulum @.

30(41). Hypericum henryi H. Léveillé & Vaniot

in Bull. Soc. bot. Fr. 54: 591 (1908): Lauener in Notes Roy. bot. Gdn. Edinb. 27: 3
(1966). Type: China, Guizhou, ‘environs de Kouy-Yang, dans la montagne’,
Bodinier 1933 (E!, holotype).

H. patulum sensu N. Robson in J. Roy. Hort. Soc. 95: 491 (1970) pro parte quoad
syn. H. henryi, non Thunb. ex Murray (1784).

Shrub 0-5-3 m tall, bushy, with stems erect to arching, not or weakly (?) frondose.
Stems reddish to yellowish, 4-lined and ancipitous when young, usually eventually
2-lined or terete; internodes 10-20 mm long, usually shorter than leaves; bark
reddish-brown. Leaves subsessile or with petiole to c. 1 mm long; lamina 10-40 x
4-16 mm, narrowly elliptic or elliptic-oblong to lanceolate or ovate, acute to obtuse

STUDIES IN THE GENUS HYPERICUM L.

or more rarely rounded, margin plane, base angustate or cuneate to rounded,
densely glaucous beneath, chartaceous; venation: 2—3(4) pairs main laterals, with
midrib branched distally, without or with scarcely visible lax tertiary reticulum;
laminar glands streaks and dots; ventral glands sparse to dense. Inflorescence
1-S(—7)-flowered, from 1-2 nodes, subcorymbiform, usually with short apical inter-
node, sometimes with 1(2)-flowered branches from middle region of stem; pedicels
4-7 mm long; bracts narrowly oblong to lanceolate, deciduous. Flowers 15-52 mm in
diam., stellate-incurved to cyathiform; buds broadly ovoid to globose, subacute to
rounded. Sepals 4-9 x 2-5-6 mm, free, imbricate, unequal, erect or (?) rarely +
spreading bud and fruit, oblong or elliptic to obovate-spathulate or broadly ovate or
circular, acute (rarely) or apiculate to obtuse or rounded, with margin entire or +
eroded-denticulate, hyaline; midrib distinct or not, veins not or rarely slightly
prominent; laminar glands linear to punctiform, numerous; intramarginal glands
dense. Petals golden to pale yellow, sometimes tinged red, + spreading or incurved,
8-25 x 6-15 mm, 24 X sepals, narrowly to broadly obovate, with apiculus lateral,
rounded to obscure; margin entire, with a row of inframarginal gland dots. Stamen
fascicles each with (30—)40—60 stamens, longest 5-13 mm long, c. 0-5 x petals;
anthers deep yellow. Ovary 3-5-7 x 2-5—5-5 mm, + broadly ovoid; styles 2:5-6 mm
long, 0-7-1-2 ovary, erect, outcurved towards apex; stigmas scarcely capitate.
Capsule 9-12 X 8-10 mm, narrowly ovoid-pyramidal to subglobose. Seeds dark
brown, 1-1-2 mm long, cylindric, not or scarcely carinate, shallowly linear-foveolate.
2n=?

In dry, usually open habitats (slopes, thickets, open forest); 1300-3314 m.

China (Yunnan, west Guizhou, south Sichuan), Vietnam (north), Burma (Shan,
Kachin), Thailand (north), Indonesia (north Sumatra). Maps 21-23.

H. henryi is the basic member of the patulum group (Spp. 33-35), the form of subsp.
hancockii from southern Yunnan being the most closely related to H. lacei and H.
acmosepalum; but it is very variable and not easy to differentiate from H. patulum
and H. uralum. Indeed it would be possible to treat it as one very variable species (H.
patulum), with five subspecies. In view of the nomenclatural history of the group,
however, it seems preferable to recognise three species (H. henryi, H. patulum, and
H. uralum), two of the subspecies of henryi forming intermediate taxa between H.
henryi subsp. hancockii and the other species. Thus subsp. hancockii (southern
Yunnan, adjacent Burma, Vietnam and Thailand, and northern Sumatra) is linked
northeastwards to H. patulum by subsp. henryi and northwestwards to H. uralum by
subsp. uraloides.

30a(41a). Hypericum henryi subsp. hancockii N. Robson, subsp. nov.

a subsp. henryi sepalis integris, acutis vel rotundatis, foliis plerumque anguste
ellipticis vel lanceolatis, acutis vel obtusis, differt. Type: China, Yunnan, Mengtze,
1895, Hancock Kew 116 (K!, holotype and isotype).

H. garrettii Craib in Kew Bull. 1913: 66 (1913) pro parte, quoad specim. Kerr 6300.

H. garrettii var. ovatum Craib, Fl. siam. enum. 1: 111 (1925). Type: Thailand, Payap.
Muang Fang, Doi Pahom Pok, c. 2400 m, 2.iv.1921, Kerr 5188 (K!, holotype;
ABD!, BBK!, BM!, E!, isotypes).

Icon: —.

Stems erect, 4-lined and ancipitous at first, sometimes becoming 2-lined. Leaves
subsessile to shortly petiolate; lamina 15-40 x 5-16 mm, narrowly elliptic or
lanceolate to ovate-oblong or rarely ovate, acute to obtuse, base angustate to
rounded. Inflorescence 1-6-flowered. Flowers 25-50 mm in diam., stellate-incurved
to cyathiform; buds ovoid to globose, subacute to rounded. Sepals (at least the outer)
broadly elliptic or broadly oblong to circular or very rarely obovate-spathulate or
apiculate, entire. Petals 14-25 x 10-15 mm, narrowly to broadly obovate. Ovary
5-6 X 4mm, ovoid; styles 4-6 mm long, 0-8-1-2 x ovary. Capsule 10-14 x 6-9 mm
narrowly ovoid-pyramidal to broadly ovoid.

261

262

NORMAN K. B. ROBSON

Map 22 Sect. 3. Ascyreia: 30a. H. henryi subsp.
hancockii ® (part; see also Maps 21, 23).

China (south Yunnan), Vietnam (north), Burma (Shan), Thailand (north), Indone-
sia (north Sumatra). Maps 21-23.

CHINA. Yunnan: Xichou [Hsichou], 1500-1700 m, 16.ix.1960, Wen 60-0072
(KUN); Chihtsun, 26.vii.—, Henry (K); Jingping [Chinping], 1960 m, 11.v.1956, ?
1158 (KUN); Szemao, 1800 m, 6.x.1899, Wilson in Hb. Hong Kong 2751 (K).

VIETNAM. North: Sa Pa [Cha Pa], 1500 m, 26.x.1963, Przybylski 135 (K).

BURMA. Shan: Taunggyi, 1500 m, 6.x.1899, Dickason 938 (A).

THAILAND. Northern: Chiang Mai, Doi Inthanon, Doi Angka, c. 2000 m,
14.viii.1922, Kerr 6300 (ABD, BM, K); Chiang Rai, Doi Pahom Pok, below summit,
NW. of Phan, 2000-2350 m, 13.ix.1967, Iwatsuki, Fukuoka & Chitayungkun T. 9690
(AAU, K).

SUMATRA. Atjeh: Laut Poepandji, 1900 m, 3-4.ix.1934, v. Steenis 6528 (A, L,
SING). Utara: Mt. Tandikat, N. of Kandang Empat, 1700-2400 m, 24.v.1955,
Meijer 3937 (L.). Barat: Korinchi Peak [Koerintji], 2190 m, iv.1914, Robinson &
Kloss 105 (BM, K, SING, US).

NEW ZEALAND (naturalised). South Island: West Coast near Ross, 21.ii.1947,
Healy H. 252 (BM).

CULTIVATED. Scotland: Edinburgh, Roy. Bot. Gard., ex seed Wang & Liu
85065 [Yunnan, Pingpien, 1300 m, 22.xi.1940], 25.ix.1977, C. 11659 (BM, E). See
note below.

STUDIES IN THE GENUS HYPERICUM L.

Tsai 62615, Mao 3171 and Wang & Liu 85065 (all from the Pingpien area) are
morphologically intermediate between subsp. hancockii and subsp. henryi.

30b(41b). Hypericum henryi subsp. henryi

H. patulum sensu N. Robson in Yb. int. Dendr. Soc. 1980: 141 (1981).
Icon: —.

Stems erect to arching or divaricate, + persistently 4-lined, ancipitous. Leaves
shortly petiolate; lamina 15-30 x 6-17 mm, ovate-lanceolate or rarely elliptic to
broadly ovate, acute (or rarely apiculate) to rounded, base cuneate to rounded.
Inflorescence 1-7-flowered. Flowers 20-35 mm in diam., cyathiform; buds ovoid to
subglobose, obtuse to rounded. Sepals broadly oblong or broadly elliptic to broadly
ovate or circular, apiculate or rounded, subentire to denticulate, erect. Petals
10-20 x 8-14 mm, broadly obovate. Ovary 4-5—5-5 x 3-5—5 mm, broadly ovoid to
subglobose; styles 4-5 mm long, c. 0-9 x ovary. Capsule 10-14 x 8-10 mm, broadly
ovoid.

China (central and east Yunnan, Guizhou). Map 21.

CHINA. Yunnan: Lo Shiueh Mtn, viii.1937, McLaren’s colls U 124 (BM, E);N.
end of Cangshan, road to Huadianba, 2400 m, 17.v.1981, Sino-Brit. Exped. Cang-
shan 264 (E); near Kunming, Hua Hong Dong, 2000 m, 26.iv.1981, Sino-Brit.
Exped. Cangshan K. 037 (E); Luquan [Luchuan], 2600 m, 25.x.1940, Zhang 0324
(KUN). Guizhou: Pin-fa [Pin-pa], 11.viii.1908, Cavalerie 3317 (BM, E, K) (see also
type).

CULTIVATED. Specimens seen from England (1982), Scotland (1977), and the
U.S.A. (California, 1959).

Some of the plants from the Kunming region (e.g. Sino-Brit. Exped. Cangshan
K106), tend towards H. patulum morphologically; but they have tall, arching, 4-lined
stems and ovate-lanceolate rather than oblong-lanceolate leaves.

30c(41c). Hypericum henryi subsp. uraloides (Rehder) N. Robson, stat.
nov.

H. uraloides Rehder in Sargent, Pl. Wilson. 3: 452 (1917). Type: China, Sichuan,
banks of Min River, vi.1903, Wilson 3258 (A, holotype; BM!, E!, isotypes).

Icon: —.

Stems erect to arching or distally frondose, + persistently 4-lined, ancipitous. Leaves
subsessile to shortly petiolate; lamina 10-27 x 3—10(-12) mm, narrowly elliptic or
narrowly lanceolate to ovate-lanceolate, acute to rarely obtuse, base angustate to
cuneate or rarely rounded. Inflorescence 1-3-flowered. Flowers c. 12-40 mm in
diam., stellate-incurved to cyathiform; buds ovoid, subacute or obtuse. Sepals
elliptic to oblanceolate or narrowly oblong, subacute or apiculate-obtuse to rounded,
entire. Petals 10-15 x 5-12 mm, broadly obovate, sometimes pale yellow. Ovary
4-7 x 2:5—5 mm, ovoid-pyramidal to subglobose; styles 2-5 mm long, 0-5—1 X ovary.
Capsule 8-11 X 6-8 mm, broadly ovoid-pyramidal to globose.

China (north Yunnan, south Sichuan, west Guizhou), Burma (Kachin). Map 23.

CHINA. Yunnan: NE. of Tengyueh [Tengchung], 1800 m, vii.1912, Forrest 8626
(E, K); Changning, 2000 m, 11.viii.1960, Tong 6003 (KUN); Cangshan, above
Yangbi, Zhaimen, 2400 m, 6.v.1981, Sino-Brit. Exped. Cangshan 280 (E); Liichuan,
1700 m, 17.ix.1973, Tao 90 (KUN); Lo Shiueh Mtn, viii.1937, McLaren’s colls U124
(BM, E). Sichuan: Miyi Xian, 16.vi.1958, Chen et al. 10204 (SZ); see also type.
Guizhou: Pu’an, 1800 m, 7.ix.1959, An 1329 (KUN).

BURMA. Kachin: Bhamo, 30.ix.1908, Cubitt s.n. (E).

CULTIVATED. England: Hampshire, Chandler’s Ford, seed ex Sino-Brit.
Exped. Cangshan 264 [Yunnan, Cangshan, above Yangbi, Mofanygou, 2400 m,
26.iv.1981], viii.1983, Lancaster s.n. (BM).

The type of subsp. uraloides is smaller in all its parts and more delicate than the other

263

264 NORMAN K. B. ROBSON

Map 23 Sect. 3. Ascyreia: 30a. H. henryi subsp. hancockii * (part; see also Maps 21, 22), 30c. H. henryi
subsp. uraloides 5x, 32. H. uralum @ (part: see also Map 24).

STUDIES IN THE GENUS HYPERICUM L.

specimens, and its locality (though imprecise — Min River) is further north than their
area. There seems to be no reason, however, to doubt that it and they belong to the
same taxon.

31(42). Hypericum patulum Thunb. ex Murray

Syst. veg. 14th ed.: 700 (May-June, 1784); Thunb., Fl. jap.: 295 (Aug. 1784); Icon.
pl. jap. Dec. sec.: t. 7 (1800); Choisy, Prod. monogr. fam. Hypéric.: 41 (1821), in
DC., Prodr. syst. nat. regni veg. 1: 545 (1824) pro parte excl. syn. et var., in
Zollinger, Syst. Verz. Fl. Ind. Archip. 1/2: 151 (1854) pro parte, quoad spec Jap.;
Treviranus, Jn Hyper. gen. sp. animad: 8 (1861); Maxim. in Bull. Acad. Imp. Sci.
St. Petersb. 27: 429 (1882) et Mél. biol. 11: 161 (1882), pro parte excl. syn. Ind.;
Hemsley in J. Linn. Soc. 23: 73 (1886) pro parte excl. syn.; Franchet, Pl. delavayi:
103 (1889); Kuntze, Rev. gen. pl. 1: 60 (1891) pro parte quoad subsp. normale;
Koehne, Deutsche Dendrol.: 415 (1893); R. Keller in Bull. Herb. Boiss. 5: 638
(1897); Diels in Bot. Jahrb. 29: 476 (1900); H. Léveillé in Bull. Soc. bot. Fr. 53: 499
(1906), in op. cit. 54: 591 (1908); R. Keller in Engler & Prantl, Nat. Pflanzenfam.,
2nd ed. 21: 176 (1925) pro parte; Rehder in J. Arn. Arbor 15: 100 (1934); Y.
Kimura in Bot. Mag., Tokyo 54: 88 (1940); Lauener in Notes Roy. bot. Gdn.
Edinb. 27: 4 (1966); N. Robson inJ. Roy. Hort. Soc. 95: 491 (1970) pro parte, excl.
syn. H. henryi; [Lancaster], Hilliers’ man. trees & shrubs: 152 (1971); Bean, Trees
& shrubs hardy in Br. Isles 8th ed. 2: 421 (1973); N. Robson in J. Jap. Bot. 52: 278,
f. 2 (1977). Type: Japan: ‘Crescit in insula Nipon, saepe cultum, alibique’,
Thunberg (UPS, holotype).

Norysca patula (Thunb. ex Murray) J. Voigt, Hort. suburb. Calcutta: 90 (1845);
Blume, Mus. bot. Lugd. Bat. 2: 23 (1856); Y. Kimura in Nakai & Honda, Nova fl.
jap. 10: 99, f. 41 (1951).

Eremanthe patula (Thunb. ex Murray) K. Koch, Hort. dendrol.: 65 (1853).

H. gumbletonii Lavallée, Arb. segrez.: 25 (1877), nomen.

H. uralum sensu Hance inJ. Bot., Lond. 16: 104 (1878), non H. uralum Buch.-Ham.
ex D. Don (1823).

H. pseud-patulum hort. in Gdnrs’ Chron. II, 17: 149 (1882), in synon.

H. patulum subsp. (a) normale Kuntze, Rev. gen. pl. 1: 60 (1891). Type as for H.
patulum Thunb. ex Murray.

H. hookerianum sensu R. Keller in Bot. Jahrb. 33: 549 (1904), non Wight & Arnott
(1834).

H. argyi H. Léveillé & Vaniot in Bull. Soc. Bot. Fr. 54: 591 (1908); H. Léveillé, Fi.
Kouy-Tcheou: 198 (1914), in Mem. R. Acad. Cienc. Artes Barcelona III, 12: 553
(1916). Type: China, Kiang-sou [Jiangsu], d’Argy (E!, holotype).

Komana patula (Thunb. ex Murray) Y. Kimura ex Honda, Nomina pl. jap.: 509
(1939); Hisauchi, Kika-syokubutsu: 179 (1950).

Icon: Y. Kimura in Nakai & Honda, Nova Fl. Jap. 10: 100, f. 41 (1951).

Shrub 0-3-1-5(-3) m tall, bushy, with branches spreading, sometimes weakly fron-
dose. Stems reddish to orange, 4-lined or 4-angled when young, soon 2-lined,
sometimes eventually terete; internodes 840 mm long, shorter than or rarely
exceeding leaves; bark greyish-brown. Leaves petiolate, with petiole 0-5—2 mm long;
lamina 15-60 x 5-30 mm, lanceolate or oblong-lanceolate to ovate or oblong-ovate,
obtuse to rounded, always apiculate, margin plane, not incrassate, base narrowly or
broadly cuneate to shortly angustate, rather glaucous beneath, chartaceous; vena-
tion: 3 pairs main laterals, with midrib branched distally and scarcely visible lax
tertiary reticulum; laminar glands short streaks and dots; ventral glands + dense.
Inflorescence 1-15-flowered, from 1-2 nodes, corymbiform, sometimes with short
apical internode, sometimes with 1-3-flowered branches from middle of stem;
pedicels 2-4(—7) mm long; bracts narrowly elliptic to narrowly oblong, deciduous.
Flowers 25-40 mm in diam., + cyathiform; buds broadly ovoid, obtuse. Sepals
5-10 x 3-5-7 mm, free, + imbricate, subequal or unequal, erect in bud and fruit,
broadly ovate or broadly elliptic or subcircular to oblong-elliptic or obovate-
spathulate, obtuse to rounded or retuse and usually apiculate, with margin finely

265

266

NORMAN K. B. ROBSON

eroded-denticulate to ciliolate, hyaline, often reddish; midrib often distinct, veins
not or slightly prominent; laminar glands linear, numerous. Petals golden yellow, not
tinged red, + incurved, 12-18 x 10-14 mm, 1-5-2-5 x sepals, oblong-obovate to
broadly obovate, with apiculus lateral, + rounded to obsolete; margin entire or
slightly eroded-denticulate, with a row of inframarginal gland dots. Stamen fascicles
each with c. 50-70 stamens, longest 7-12 mm long, 0-40-5 x petals; anthers bright
yellow. Ovary 5-6 X 3-5-4 mm, + broadly ovoid; styles 4-5-5 mm long, 0-75—0-95 x
ovary, + erect, outcurved towards apex; stigmas not or scarcely capitate. Capsule
9-11 x 8-10 mm, broadly ovoid. Seeds dark brown, 1-1-2 mm long, + cylindric, not
or scarcely carinate, shallowly linear-foveolate, 2n = 36? (record is probably based
on a misidentification, see Robson (1981: 150-151)).

In dry, open habitats (thickets, scrubby slopes, cliffs); (300) 450-2400 m.

China (native in north Guizhou, Sichuan; probably introduced in other parts of
China, certainly introduced in Taiwan, Japan, and South Africa (Natal); specimens
seen from Shaanxi, Jiangsu, Anhui, Zhejiang, Fujian, and Guangxi; also recorded
from Hunan, Hubei, and Jiangxi). Map 21.

CHINA. Sichuan: cliffs of R. Yangtze above Chungking, 2.vi.1903, Wilson 3261
(BM, K); Mt. Omei, x.1980, Lancaster 623 (BM); Nanchuan Xian, 1120 m,
19.ix.1957, Xiong & Li 93606 (SZ); Huidong Xian, 1750 m, 3.vii.1978, Zhao 5801
(SZ). Guizhou: Tsunyi Hsien, Liang Feng Yah, 1100 m, 19.vui.1931, Steward, Chiao
& Cheo 5 (BM, E, K, L). Guangxi: Fabao, 1001 m, 5.xii.1958, Chang 13767 (IBSC).
Shaanxi: Hugiaba, 1.1936, Fenzel 1018 (W). Jiangsu: no precise locality or date,
d’Argy (E). Anhui: no precise locality, comm. viii.1929, Chang 1885 (K). Zhejiang:
Chungshan Hsien, 40 li W. of Wenchow, 250-450 m, 8.vi.1924, Ching 1885 (E, K).
Fujian: Nanjing, 17.v.1959, Huang 190370 (IBSC).

TAIWAN (naturalised). Chiayi: A-li-shan, road to Museum, 2200 m, 19. viii.1966,
v. Steenis 20922 (L).

JAPAN (cultivated and naturalised). Kyushu: Fukuoka Pref., Kasuya Co., Hako-
zaki, 2.vi.1965, Ichikawa 188 (A). Honshu (Central), Kinki: Prov. Kii, Hashimoto,
between Yatate and Hosokawa, 400 m, 3.vii.1972, Tamura 24686 (E). Honshu
(Central), Chubu: Sagami, Mt. Hakone, 14.vi.1907, Sakurai s.n. (H). Honshu
(Central), Kanto: environs de Yokohama, c. 1904, Borel (G).

NATAL (naturalised). Mpendhle district, Carters’ Nek on Nottingham—Under-
berg road, c. 1830 m, 24.xii.1982, Hilliard & Burtt 16088 (BM).

CULTIVATED. Specimens seen from China, Japan, Taiwan, England, Scotland,
Ireland, France, Germany, Austria, Italy, Switzerland, Ecuador (also naturalised),
and Colombia.

H. patulum has a more northern distribution than H. henryi subsp. henryi, from
which it can be distinguished by the more spreading habit, 2-lined stems, oblong-
lanceolate to ovate rather than elliptic or ovate-lanceolate to ovate leaves, and
usually by the smaller capsules. It was introduced to Britain from Nagasaki by
Oldham in 1862 (Robson, 1970).

H. patulum is less hardy in Europe than species such as H. pseudohenryi or H.
forrestii. These species grow at higher altitudes than H. patulum and were introduced
to cultivation later, when botanical collectors penetrated further into the moun-
tainous regions of Yunnan and Szechwan. Most of them were at first regarded as
‘improved’, hardier versions of H. patulum, and were named as varieties of that
species. However, they have all proved to be specifically distinct, and not even
closely related to it.

31x(42x). Hypericum x moserianum Luquet ex André
(31. H. patulum x 14. H. calycinum)

Revue hort. 61: 463 (1889) (‘Moserianum’); Plaisted & Lighty in Nat. hort. Mag. 38:
124 (1958); Thomas in Gdnrs’ Chron. III, 147: 227 (1960); Milano in Publs Inst.
Bot. agr. Argentina 8, 128: 18, f. 5 (1961); Fosberg in Occ. Pap. B.P. Bishop Mus.
24: 21 (1969); Robson in J. Roy. Hort. Soc. 95: 491 (1970); [Lancaster], Hilliers’

STUDIES IN THE GENUS HYPERICUM L.

man. trees & shrubs: 151 (1971); Bean, Trees & shrubs hardy in Br. Isles 8th ed. 2:
420 (1973). Type ?: ‘Seedling from H. patulum’, Moser’s nursery, Versailles,
x.1887, Herb. Nicholson (K!, isotype ?).

Icon: Thomas in Gdnrs’ Chron., III, 147: 227 (1960).

Shrub 0-3-0-5(-0-7) m tall, bushy, semi-evergreen, with branches spreading,
arching, not very frondose. Stems red, 4-lined when young, soon 2-lined to terete
except immediately below nodes (4-lined); internodes 8-43 mm, shorter than to
exceeding leaves; bark greyish-brown. Leaves petiolate, with petiole 1-2-5 mm long;
lamina 22-60 x 7-36 mm, oblong-lanceolate or rarely lanceolate to oblong-ovate or
ovate, acute to rounded-apiculate or very rarely rounded, margin plane, not
incrassate, base cuneate to rounded, paler beneath, not glaucous, subcoriaceous;
venation: 4 pairs main laterals, with midrib branched distally, and distinctly visible
lax tertiary reticulum; laminar glands medium to short streaks and dots; ventral
glands + dense. Inflorescence 1-8-flowered, from 1-2 nodes, corymbiform, some-
times with shorter apical internode, without flowering branches from middle part of
stem; pedicels 2-5 mm long; bracts narrowly elliptic to narrowly oblong, deciduous.
Flowers 45-60 mm in diam., stellate to slightly cyathiform; buds broadly ovoid to
subglobose, obtuse. Sepals 7-10 x 6-10 mm, free, imbricate, unequal, broadly
oblong-elliptic to subcircular, rounded, erect in bud, + spreading in fruit, with
margin ciliate or ciliolate, scarious, often with submarginal red zone; midrib obscure
or indistinct, veins not prominent; laminar glands linear, numerous. Petals bright
yellow, not tinged red, spreading or + incurved, 21-30 x 15-20 mm, c. 3 X sepals,
obovate, with apiculus subterminal to lateral, rounded; margin entire or ciliate, with
few inframarginal gland dots. Stamen fascicles each with c. 50-70 stamens, longest
12-18 long, c. 0-5—0-75 x petals; anthers reddish. Ovary 6-5-9 xX 5—6 mm, ovoid to
ovoid-conic; styles 8-11 mm long, 1-1-5 X ovary, erect, outcurved at apex; stigmas
not capitate. Capsule 9-12 x 9-11 mm, broadly ovoid. Seeds dark brown, 0-9-
1-1 mm long, + cylindric, not or scarcely carinate, shallowly scalariform-reticulate.
2n = 36 (Sugiura, 1936), c. 50 (Gibby in Robson, 1981).

Known only in cultivation.

CULTIVATED. Specimens seen from England, Scotland, Ireland, France, the
Netherlands, Germany, U.S.A. (Massachusetts, California).

H. x moserianum is the result of an artificial cross (H. patulum 2 x H. calycinum ©’)
made in c. 1887 at Moser’s nursery at Versailles. It has the general habit of a low H.
patulum (with low spreading branches); but the influence of H. calycinum is apparent
in the subcoriaceous leaves with marked reticulate venation (visible beneath), the
relatively large flowers with more spreading petals, the reddish anthers and the
relatively long stamens and styles. It is partially fertile; the capsules contain a mixture
of fully developed seed and flattened or shrivelled seed. From some developed seeds
my wife raised a varied batch of seedlings, some of which have flowered and seem
healthy. These plants show the re-assortment of characters that one would expect
from F, hybrid seed.

The reported chromosome numbers of H. x moserianum (2n = 36 and c. 50) are
difficult to reconcile with its known parentage. H. patulum (2n = 36) x H. calycinum
(2n = 20) should give 2n = 28. Even if the record of 2n = 36 for H. patulum is based
on a misidentification and that species in fact has 2n = 20 like H. uralum, the number
should be 2n = 20. Only further research will resolve this apparent anomaly. If the
cross was between 2n = 36 (unreduced) and n = 10, giving 3n = 46, that triploid
number would not be consistent with the partial fertility of H. x moserianum.

A variegated form of the hybrid, in which the red sap tinges some of the white parts
pink, arose spontaneously in 1894, a short time after the original cross was made. It
has smaller leaves and flowers than the typical form and, although not vigorous in
growth, is now widely used as a foliage plant.

267

OO 4

268

NORMAN K. B. ROBSON
31xa(42xa). H. x moserianum ‘Tricolor’

Hillier’s man. trees & shrubs: 152 (1971); Bean, trees & Shrubs hardy in Br. Isles
8th ed. 2: 420 (1973).

H. x moserianum var. tricolor Maumené in Le Jardin 8: 186, f. 93 (1894); Plaisted &
Lighty in Nat. Hort. Mag. 38: 125 (1958). Type: cult. France, Orleans, 79 route
d’Olivet, Chenault (specimen ?).

H. moserianum forma tricolor (Maumené) Rehder, Bibl. cult. trees & shrubs: 463
(1949); Thomas in Gdnrs’ Chron. II, 147: 227 (1960).

H. x moserianum nothomorph tricolor (Maumené) N. Robson in Bull. Br. Mus. nat.
Hist. (Bot.) 8: 170 (1981).

Icon: Maumené in Le Jardin 8: 187, f. 93 (1894).

Leaves variegated with pink and white, narrowly lanceolate. Flowers c. 20-30 mm in
diam., deeply cyathiform, sterile.

CULTIVATED. Specimens (live) seen from England (1984).

32(43). Hypericum uralum Buch.-Ham. ex D. Don

in Curtis’s bot. Mag. 50: t. 2375 (1823), Prodr. Fl. Nepal.: 218 (1825); Schilling in J.
Roy. Hort. Soc. 94: 224 (1969); Anon. in Bull. Dept. Med. PI. Nepal 3: 117 (1970);
N. Robson in J. Roy. Hort. Soc. 95: 491, f. 423 (1970); [Lancaster], Hilliers’ man.
trees & shrubs: 152 (1971); Murata in Acta Phytotax. Geobot. 25: 111 (1973);
Bean, Trees & shrubs hardy in Br. Isles 8th ed. 2: 421 (1973); N. Robson in Nasir &
Ali, Fl. W. Pakistan 32: 7 (1973), in Hara & Williams, Enum. fl. pl. Nepal 2: 62
(1979). Type: Nepal, Narainhetty, Hamilton (BM!, lectotype), see p. 000.

H. patulum var. attenuatum Choisy in DC., Prodr. syst. nat. veg. 1: 545 (1824);
Kuntze, Rev. gen. pl. 1: 60 (1891) sub subsp. normale Kuntze; Gagnepain, Fi. gen.
Indo-chine suppl. 1: 249 (1943) pro parte quoad typum. Type: Nepal, Sheopore,
1821, Wallich 4809A (G-DC!, holotype BM!, E!, FI!, K-Wall!, L!, SING!).

H. patulum sensu Wallich, Numer. list: No. 4809 (1831); Dyer in Hook. f., Fl. Brit.
Ind. 1: 254 (1874); Burkill in Rec. botl Surv. Ind. 4: 99 (1910); Parker, For. fl.
Punjab, Hazara & Delhi, ed. 2: 29 (1924); BanerjiinJ. Bombay Nat. Hist. Soc. 51:
774 (1953), 55: 251 (1958), in Rec. botl Surv. Ind. 19 (2): 27 (1966); non Thunb. ex
Murray (1774).

Norysca urala (Buch.-Ham. ex D. Don) K. Koch, Hort. dendrol.: 66 (1853); Y.
Kimura in Nakai & Honda, Nova fl. jap. 10: 102 (1951), in Kihara, Fauna & fi.
Nepal Himal.: 180 (1955), in Hara, Fl. eastern Himal.: 210 (1966), in Bull. Dept.
Med. Pl. Nepal. 2: 11 (1969), in Hara, Fl. eastern Himal. 2: 82 (1971).

H. nepalense K. Koch, Dendrol. 1: 497 (1869), in synon., non H. napaulense Choisy
(1824).

H. ramosissimum K. Koch, Dendrol. 1: 497 (1869), in synon., non Ledeb. (1842).

H. uralense Lavallée, Arb. segrez.: 25 (1877), sphalm.

H. patulum var. uralum (Buch.-Ham. ex D. Don) Koehne, Deutsche Dendrol.: 415
(1893); Rehder, Man. Cult. Trees & Shrubs., ed. 2: 638 (1940).

Norysca urala var. angustifolia Y. Kimura in Hara, FI. eastern Himal.: 210 (1966) pro
parte excl. typum.

H. hookerianum ‘Buttercup’ hort.

Icon: D. Don in Curtis’s bot. Mag. 50: t. 2375 (1823).

Shrub, 0-3-2 m tall, bushy, with stems erect, arching, often frondose. Stems red,
4-lined or 4-angled and strongly ancipitous when young, eventually 2-lined or terete;
internodes 5—20 m long, shorter than leaves; bark reddish-brown. Leaves petiolate,
with flat petiole 0-5—1 mm long; lamina 10-40 x 4-24 mm, all lanceolate or older ones
ovate, acute to rounded-apiculate, margin plane, base narrowly to rarely broadly
cuneate, + densely glaucous beneath, chartaceous; venation: 3 pairs main laterals,
with midrib branches and tertiary reticulum scarcely visible; laminar glands streaks |
(towards midrib) and dots; ventral glands usually + dense. Inflorescence
1-3(-10)-flowered, from 1-2 nodes, subcorymbiform, if from 2 nodes then with short

STUDIES IN THE GENUS HYPERICUM L. 269

apical internode, often with 1—-3-flowered branches from middle of stem; pedicels 3-7
mm long; bracts narrowly oblong, deciduous. Flowers 15-30 mm in diam., + deeply
cyathiform; buds broadly ovoid to globose, obtuse to rounded. Sepals 3-5—6(9) x
(1)2-5(6-5) mm, free, imbricate, subequal or unequal, erect in bud and fruit, oblong
or elliptic to obovate-spathulate, rounded or very rarely obtuse, with margin entire,
narrowly hyaline; midrib not or scarcely distinct, veins not or rarely slightly promin-
ent; laminar glands linear, numerous. Petals golden to deep yellow, not tinged red,

incurved, 9-18 x 5-12 mm, 2-5-3 X sepals, broadly obovate to subcircular, with i)
apiculus lateral to subterminal, rounded to obscure; margin entire, without infra-

marginal gland dots. Stamen fascicles each with 40-60 stamens, longest 4-6(8) mm

long, 0-25—0-5 x petals; anthers golden to deep yellow. Ovary 3-5 Xx 2:5-3 mm,

broadly ovoid to globose; styles 2-5—4-5 mm long, 0-6-0-9(1) x ovary, erect and +

divergent near apex or wholly outcurving; stigmas narrowly capitate. Capsule

7-11(13) x 7-11 mm, subglobose (or more rarely broadly ovoid) to globose. Seeds

dark brown, 0-4—0-6 mm long, cylindric-ellipsoid, scarcely carinate, shallowly linear-

foveolate. 2n = 20.

In dry, open habitats (grassy or rocky slopes, open woodland, pastures, cliff edges),
and sometimes in thickets and by streams; 1500-3600 m.

China (Xizang [Tibet]), Burma (Kachin, Chin), India (Manipur, Mizoram?, Naga-
land, Meghalaya), Bhutan and Himalaya west to Pakistan (Sarghoda). Maps 23, 24.

CHINA. Xizang [Tibet]: Senge Dzong, near Tawang, 2700 m, 14.xi.1913, Bailey
12 (E); Rima, 1500 m, 11.viii.1950, Kingdon Ward 20121 (BM).

BURMA. Kachin: Bhamo, 30.ix.1908, Cubitt s.n., (E); Adung Valley, vicinity of
Tahawndam, ii.xi.1931, Kingdon Ward 10153 (BM). Chin: Mt. Victoria, Old
Kanpelet, 20.v.1924, Cooper 5922A (E, SING).

INDIA. Manipur: Khongui [Imphal], N.E. ranges, 1800 m, Watt 6328 (E, K).
Nagaland: Naga Hills, Khiyakhu, 1800 m, Bor. 5084 (K). Meghalaya: Khasi Hills,
Shillong, 1500-1800 m, 12.vii.1949, Kingdon Ward 18641 (BM). Arunachal Pradesh:
Delei valley, 1500 m, 12.vii.1928, Kingdon Ward 8444 (K). West Bengal: Labah
[Kalimpong], 2100 m, ix.1904, Haines BB 983 (E, K). Uttar Pradesh: Mussoorie,

Map 24 Sect. 3. Ascyreia: 32. H. uralum @ (see also Map 23).

270

NORMAN K. B. ROBSON

Landour, 1950 m, 1.xii.1936, R. R. Stewart 15777 (A). Himachal Pradesh: Simla,
Bushahr, Manjoli, 4.vii.1934. Parmanand 354 (E). Kashmir: Basaoli, 1650 m,
26.ix.1876, Clarke 31592 (K).

PAKISTAN. Sargodha: Shahpur, Mt. Sakesar, 1500 m, 4.iv.1902, Drummond
13806 (K).

NEPAL. West: Dhaulakot, 2100 m, 17.viii.1952, Polunin, Sykes & Williams 464
(BM, E). Central: Annapurna Himal, Sati Khola, 2250 m, 5.viii.1954, Stainton,
Sykes & Williams 6706 (BM, E). East: Batasay to Halnale Bhanjang, 1.xii.1963,
Hara et al. T. 6304584 (BM, TI).

SIKKIM. Toong, 1800 m, 6.vii.1909, Smith & Cave 879 (A).

BHUTAN. Gyasa Dzong, Mo Chu, 3330 m, 3.vii.1949, Ludlow & Sherriff 17400
(BM); Tobrang (Trashiyangse Chu), 2100 m, 6.vii.1949, Ludlow & Sherriff 20493
(BM, E).

CULTIVATED. Specimens seen from England (1834-1983), Scotland (1964),
Eire (1967), and Germany (1886).

H. uralum is closely related to (and apparently derived from) H. henryi subsp.
uraloides; some of the specimens from Khasiya approach the latter in morphology
rather closely. Indeed, it is difficult to describe good characters to separate these
taxa, although they seem quite distinct in cultivation. However, the presence of
frondose branching (i.e. with lateral branches in one plane) in H. uralum, in
combination with the usually broader leaves and sepals, will usually suffice to
distinguish it from H. henryi subsp. uraloides.

In Flora of West Pakistan (Robson, 1973), I cited Curtis’s bot. Mag.: t. 2375 (1823)
as the type of H. uralum, as the cultivated plant from which the illustration was made
was apparently not preserved. It was grown from Nepal seed (probably sent by
Wallich in 1818-21) by Whitby, Brame & Milne at their Fulham Nursery, London,
and flowered in 1822. It is clear, however, that Don obtained at least the epithet
uralum from Hamilton’s notes; and in the introduction to Prodromus florae
nepalensis (Don, 1825) he claims to have seen all Hamilton’s Nepal specimens,
collected in 1802-3. The Hamilton specimen of H. uralum in BM must, therefore,
have some type status. Don’s description in Curtis’s bot. Mag. was probably derived
initially from the Nepal specimen and confirmed or modified after he had seen the
cultivated plant, so that specimen and illustration should be regarded as syntypes. As
a specimen is always to be preferred to an illustration when choosing lectotypes, I
have selected the Hamilton specimen as lectotype of H. uralum.

33(44). Hypericum maclarenii N. Robson, sp. nov.

H. choisiano Wallich ex. N. Robson affinis, sed foliis petiolo breviora lamina
angustiora subtus glauca haud reticulata, alabrastris apiculatis, sepalis haud foliosis,
excurvatis vel patulis, petalis obovato-oblanceolatis patulis, staminibus longioribus,
ovario ovoideo-conico stylis aequantibus vel paulo brevioribus, differt. Type: China,
Sichuan, Tatsienlu [Kangding], 1938, McLaren’s Collectors AD151 (E!, holotype).

H. stellatum sensu N. Robson in J. Roy. Hort. Soc. 95: 493 (1970) pro parte.
Icon: —.

Shrub 0-75-1 mm tall, with branches erect. Stems red, shallowly 4-lined and
sometimes ancipitous when young, soon terete; internodes 10-25 mm long, usually
shorter than leaves; bark reddish-brown. Leaves petiolate, with petiole 0-5—2 mm
long; lamina 25-40 x 7-10 mm, narrowly lanceolate, acute to subacute, margin
plane, base cuneate, + densely glaucous beneath, chartaceous; venation: 34 pairs
main laterals, branched, the lower sometimes free, the midrib pinnately branched,
all branches clearly visible and forming undulating inframarginal vein, with tertiary
reticulum almost invisible; laminar glands long (in lower leaves) to short streaks and
dots; ventral glands + sparse. Inflorescence 1-4-flowered, from terminal node,
subcorymbiform; pedicels 7-10 mm long; bracts reduced, linear-lanceolate, persis-
tent ? Flowers 40-50 mm in diam., stellate; buds narrowly ovoid, apiculate. Sepals
7-11 x 2-5-5 mm, free, open, subequal to unequal, + outcurved in bud, spreading in
fruit, narrowly elliptic, acute to acuminate, with margin entire; midrib clearly visible,

STUDIES IN THE GENUS HYPERICUM L.

veins not prominent; laminar glands linear or + interrupted, c. 12-14. Petals golden
yellow, sometimes tinged red dorsally, spreading, 20-25 x 12-15 mm, 2-5-3 x
sepals, obovate-oblanceolate, with apiculus acute to obtuse, lateral, margin entire,
eglandular. Stamen fascicles each with c. 50 stamens, longest 13-15 mm long, c. 0-6 x
petals; anthers golden yellow. Ovary 7-8 x 3-5—S mm, ovoid-conic; styles 6-8 mm
long, 0-85—1 x ovary, free, outcurved near apex; stigmas narrowly capitate. Capsule
12-15 x 7-8 mm, narrowly ovoid. Seeds dark brown, 1-1-5 mm long, cylindric,
carinate, shallowly linear-reticulate to linear-foveolate. 2n = ?

Steep rocky banks; c. 2000 m.

China (W. Sichuan). Map 26.

CHINA. Sichuan: Tatsienlu [Kangding], 1923. Cunningham 171 (E), 1924, Cun-
ningham 530 (E); Luding—Kangding road, Kangding Ho gorge, c. 2000 m,
12.ix.1981, Lancaster L. 863 (BM).

CULTIVATED. Specimens seen from England (1983-ex Lancaster L. 863).

H. maclarenii, like H. subsessile, is an isolated relict species, having been collected
only in the vicinity of Kangding. It has affinities with three groups of species, being
closest to H. choisianum and H. bellum but also showing similarities to H. subsessile
and even to the H. kouytchense — H. stellatum group. From H. choisianum and H.
bellum sens. lat. it differs by its erect, not foliaceous sepals, narrower, spreading
petals and longer stamens and styles; from H. subsessile it differs by its subterete
stems, petiolate, lanceolate, + glaucous leaves and the floral characters already
mentioned; and from the H. kouytchense group by the lateral petal apiculus, shorter
styles and leaves glaucous beneath with markedly ‘looped’ venation and (in the lower
ones) striiform glands.

34(45). Hypericum choisianum Wallich ex N. Robson

in Nasir & Ali, Fl. W. Pakistan 32: 6, f. 1 E-H (1973); in Hara & Williams, Enum. fi.
pl. Nepal 2: 61 (1979). Type: India, Kumaon, R. Blinkworth in Wallich 4805
(K-Wall.!, holotype).

H. choisianum Wallich, Numer. list: No. 4805 (1831), nomen.

H. oblongifolium sensu Wallich, Numer. list: No. 4810 (1831), non Choisy (1821).

H. hookerianum var. leschenaultii sensu Dyer in Hook. f., Fl. Brit. Ind. 1: 254
(1874); Banerji in J. Bombay nat. Hist. Soc. 51: 774 (1953); non H. leschenaultii
Choisy (1824).

Norysca hookeriana var. leschenaultii sensu Y. Kimura in Hara, Fl. eastern Himal.:
210 (1966), non H. leschenaultii Choisy (1824).

Icon: N. Robson in Nasir & Ali, Fl. W. Pakistan 31: 5, f. LE—-H (1973).

Shrub (0-1)-2 m tall, bushy, with branches erect to ascending. Stems red to orange,
4-lined and ancipitous when young, eventually terete; internodes 15-55 mm long,
usually shorter than leaves; bark grey-brown. Leaves petiolate, with petiole 2-4 mm
long; lamina 25-88 x 10-42 mm, triangular—lanceolate or rarely triangular-ovate to
ovate, acute or acuminate to obtuse or rarely rounded, margin plane, base broadly
cuneate to rounded or subcordate, paler beneath, not glaucous, chartaceous;
venation: 3-5 pairs main laterals, branched, the lower pairs sometimes free, the
midrib branches impressed, with tertiary reticulum almost invisible to marked and
rather dense; laminar glands streaks and dots; ventral glands absent. Inflorescence
1-7-flowered, from terminal node, subcorymbiform; pedicels 3-5-11 mm long;
bracts foliar to narrowly elliptic, persistent. Flowers 40-70 mm in diam., shallowly to
deeply cyathiform; buds ovoid, sharply acute to obtuse. Sepals 7-18 (or longer if
markedly foliaceous) x 2-10 mm, free, imbricate or open, unequal, spreading to
recurved in bud and fruit, narrowly to very broadly elliptic, acute to apiculate or
more rarely obtuse, with margin entire; midrib clearly visible, veins not usually
prominent; laminar glands linear, interrupted towards apex, numerous. Petals deep
golden yellow, sometimes tinged red dorsally, slightly to markedly incurved, 16-30
x 15-22 mm, 1-7-2-2 X sepals, broadly obovate to obovate-circular, with apiculus
subterminal, rounded, margin entire, eglandular. Stamen fascicles each with 60-80

CO -

Zi

Dye,

ee

NORMAN K. B. ROBSON

stamens, longest 6-10 mm long, 0-35—0-4 x petals; anthers golden yellow. Ovary
(5)6-8(9) X 3-5-5-5 mm, + broadly ovoid; styles 3-5 mm long, 0-35-0-7 x ovary,
free, outcurved near apex; stigmas narrowly capitate. Capsule (9-)14-19 x 8-12
mm, ovoid-conic to subglobose. Seeds dark brown, 0-7-1 mm long, cylindric to
cylindric-ellipsoid, carinate or shallowly winged, shallowly linear-reticulate to
linear-foveolate. 2n = ?

Dry to moist, open to half-shaded situations (grassy or rocky slopes or cliffs, in
thickets or in Rhododendron forest); 1600-4800 (Tibet), mainly 2400-4200 m.

China (Yunnan, S. Xizang), Burma (north), Bhutan, Sikkim, Nepal, India (W.
Bengal, Kumaun to Kashmir), Pakistan (Hazara). Map 25.

CHINA. Yunnan: Jing-dong, Feng-Guan-Shan, 1600 m, 29.iv.1940, Li 3458
(IBSC); Taron-Taru Divide, Barcuwhang, 3.ix.1938, Yii 20107 (E). Xizang (Tibet):
Yatung [Yathang], 1897, Hobson (K); Karma valley, 3150 m, 25.vi.1922, Mt. Everest
Exped. (K); Niangdam La, Kharta, 4800 m, 2.vi.1922, Mt. Everest Exped. (K).

BURMA. Kachin: Upper Adung Valley, N. of Tahawndam, 3000-3300 m,
22.x.1931, Kingdon Ward 10133 (BM).

Map 25 Sect. 3. Ascyreia: 34. H. choisianum @, 35a. H. bellum subsp. latisepalum ©, 35b. H. bellum

subsp. bellum *.

BHUTAN. Rathong Chu, 3900 m, 1913, Ribu & Rhomo 6545 (E); Sinchu La,
Timpu, 2700 m, 10.viii.1914, Cooper 3373 pp. (E); Dongla Lingtsi and Trashiyang-
tsi, 3300 m, 10.viii.1915, Cooper 4398 (BM, E); Lao, Me La Chu, 2700 m, 5.vii.1949,
Ludlow, Sherriff & Hicks 20822 (BM); below Sengor, c. 3000 m, 6.vii.1979, Grierson
& Long 2540 (BM, E).

SIKKIM. Tonglo, 3000 m, vii.1882, Gamble 10414 (FI, K); Singalilah pass,
3300 m, 1.1848, Hooker 103 (K).

NEPAL. East: Lumding Khola, 2850 m, 13.vii.1974, McCosh 399 (BM, E).
Central: Lamjung Himal, 3450 m, 16.vii.1954, Stainton, Sykes & Williams 6372
(BM). West: above Ranmagaon, 3300 m, 2.vii.1954, Stainton, Sykes & Williams
3348 (BM).

STUDIES IN THE GENUS HYPERICUM L.

INDIA. W. Bengal: Darjeeling between Ghum and Lopchu, 2100 m, 19.vi.1960,
Kanai, Murata & Togashi (K). Uttar Pradesh: Kumaun, Younrani, 2100 m,
24.vi.1948, Chand 790 (L); N. Gharwal, Nandargiri Valley, 2850 m, 12.vi.19,
Osmaston 1062 (A). Himachal Pradesh: Bashahr, between Khadrala and Baghi,
2700 m, 19.vi.1928, Parker 3057 (K); Kulu, versus transitum Jalauri, 2850 m,
27.vii.188, Drummond 8427 (E, K). Kashmir: Kishtawar Distr., Bangar, 3600 m,
31.vii.1943, Ludlow & Sherriff 9254 (BM).

PAKISTAN. Azad Kashmir: Poonch [Punch] to Bantara Gali, 2400 m,
4.vii.1952, Stewart & Nasir 24097 (BM, RAW). Hazara: Nandi, Kund, Anon.
(RAW).

CULTIVATED. Specimens seen from England (1972-1983) and Ireland (1980).

H. choisianum has sometimes been confused with the Indonesian H. leschenaultii, on
account of the narrow sepals that occur in both species; but it differs from that species
particularly in the ovary and capsule, which are ovoid-conic to subglobose (not
narrowly conic), but also in its relatively longer stamens and styles, broader ovary
and capsule, longer petiole and frequently more densely reticulate leaf venation. H.
choisianum seems to be a derivative of 33. H. maclarenii, whereas 22. H. leschenaul-
tii is derived from 21. H. siamense in a somewhat parallel way.

H. choisianum can also be very similar to 28. H. hookerianum, which grows in a
lower but overlapping zone of the central and eastern Himalaya; but the erect,
rounded, never foliaceous sepals, relatively shorter stamens and styles, and bluish
leaves without reticulate venation of the latter usually serves to distinguish it.

In Nepal, H. choisianum takes on a rather distinct facies in which the typical form
(leaves with venation + obscurely reticulate, flowers solitary or rarely 2-3, sepals
ovate to lanceolate) is replaced by one in which the leaves always have markedly
reticulate venation and the flowers are usually in cymes of 3~7 with narrowly
lanceolate to narrowly elliptic sepals. Although this form is easily recognisable, it is
connected by too many intermediates with the typical one to allow it taxonomic
recognition.

The Chinese and Burmese plants, from isolated localities divorced from the main
Himalayan range of the species, are nearest to H. bellum; and the one from furthest
east (Li 3458) is unusual in having almost elliptic leaves with a rounded apex.

35(46). Hypericum bellum H. L. Li

in J. Arnold Arbor. 25: 308 (1944); Robson in J. Roy. Hort. Soc. 95: 491, f. 239
(1970); [Lancaster], Hilliers’ man. trees & shrubs: 150 (1971); Bean, Trees &
shrubs hardy in Br. Isles 8th ed. 2: 408 cum fig. (1973). Type: China, Yunnan,
oe oh Valley (Taron), Mt. Chingtinglaka, 28.vii.1938, Yii 19497 (A!, holo-
type; E!).

Shrub 0-3-1:5 m tall, often forming low thickets, with branches dense, erect or
arching. Stem red to orange, 4-angled and slightly ancipitous at first, very soon
terete; internodes 10-80 mm long, usually equalling or exceeding leaves; bark
grey-brown. Leaves petiolate, with petiole 0-5-3 mm long; lamina 15-65(-—78) x
7-46 mm, lanceolate-oblong or ovate-oblong to broadly rhombic-ovate or subcircu-
lar, obtuse to rounded or emarginate, often apiculate, margin plane or undulate,
base + broadly cuneate or rounded to truncate or subcordate, paler or glaucous
beneath, chartaceous; venation: 3-4 pairs main laterals (the upper not distinctly
intramarginal), with midrib obscurely branched distally with or apparently without
+ clear lax tertiary reticulum; laminar glands dots and short streaks, ventral glands
absent or + dense. Inflorescence 1—7-flowered, from apical node, subcorymbiform,
rarely with flowering branches from lower nodes; pedicels 3-14 mm long (—30 mm in
fruit); bracts foliar to narrowly elliptic, persistent to deciduous. Flowers (25—)30-
50(-60) mm in diam., cyathiform; buds broadly ovoid, obtuse to rounded. Sepals
3-12 x 2-5-8 mm, sometimes enlarging in fruit, free, imbricate, subequal, erect in
bud and fruit, broadly elliptic or subcircular to broadly or narrowly oblong, acute or
obtuse to apiculate or rounded, with margin entire or finely eroded-denticulate
towards apex and often scarious; midrib rarely conspicuous, veins not prominent;

273

274

NORMAN K. B. ROBSON

laminar glands linear, c. 12. Petals golden yellow to butter-yellow or rarely pale
yellow, not red-tinged, incurved, 15-30 x 11-25 mm, 3-5 x sepals, broadly to
narrowly obovate, with apiculus subterminal, rounded, margin entire. Stamen
fascicles each with 25-65 stamens, longest 6-15 mm long, 0-35-0-6 x petals; anthers
deep golden yellow. Ovary 4-8 x 3-7 mm, narrowly ovoid to subglobose; styles 3-8
mm long, 0-6-1 x ovary, free, suberect to divergent, outcurved near apex; stigmas
small. Capsule 10-15 x 6-15 mm, narrowly to broadly ovoid, often puckered. Seeds
dark reddish-brown, 0-8-1 mm long, narrowly cylindric, + carinate, shallowly
scalariform-reticulate. 2n = ?

Scrub or forest margin, open dry grassy slopes, sometimes by streams or in cultivated
ground; (1440)1800-3900 m.

China (W. and NW. Yunnan, S.W. Sichuan, S.E. Xizang [Tibet]), N. Burma, India
(Arunachal Pradesh). Map 25.

H. bellum, like H. choisianum, is related to the relict H. maclarenii and differs from it
in its cyathiform flowers with broader petals and broader leaves. From H.
choisianum it can be distinguished by its oblong to ovate leaves that are usually
broader and never acute; and the forms of these species that approach one another in
leaf shape differ in style and stamen length and in sepal aspect.

The more primitive subspecies of H. bellum (subsp. latisepalum, which
approaches H. choisianum) has longer, broadly elliptic sepals, which are often
foliose but nearly always acute or obtuse to apiculate, the styles and stamens are
longer, and the leaves vary from narrowly oblong to broadly ovate. It has a dissected
distribution and shows a morphological trend westwards from W. Yunnan to N.
Burma, where its variation and range almost merge with those of the type subspecies.

Subsp. bellum occurs in NE. India and on the north side of the Himalayan range
from southern Xizang to SW. Sichuan and NW. Yunnan. Its sepals are subcircular to
narrowly oblong but always rounded, the styles and stamens are usually shorter, and
the leaves vary from oblong-ovate through broadly ovate to broadly rhombic or
subcircular. Some specimens from the east of its range approach the Chinese rather
than the Burmese form of subsp. /atisepalum, which suggests that subsp. bellum may
prove to be diphyletic. In that case, the epithet would apply to the Chinese form and
the Tibetan one would be rendered nameless. The latter has an undulate leaf margin
and is probably the only form of subsp. bellum cultivated now in Europe.

35a(46a). Hypericum bellum subsp. latisepalum N. Robson, subsp. nov.

a subsp. bellum foliis pro ratione angustioribus margine semper planis, sepalis
longioribus latioribus, acutis vel obtusis vel apiculatis, petalis ovariisque maioribus,
staminibus stylisque longioribus, differt. Type: China, Yunnan, Tali [Dali], top of
Yin Yo Mtn, viii.1932?, McLaren’s Colls, 214C (BM!, holotype; E!, isotype).

Icon: —.

Leaves with petiole 1-5-3 mm long; lamina (30—-)37—-87 x (13)1646 mm, narrowly
oblong or lanceolate-oblong to ovate-oblong (1: b usually = 2 or more). Flowers
40-60 mm in diam., buds acute to obtuse. Sepals 8-13 x (5)6-8 mm, broadly elliptic,
acute to obtuse or apiculate, entire. Petals 23-37 x 18-30 mm. Stamens fascicles with
longest stamen (10)14-21 mm long, (0-5)0-6-7 x petals. Ovary 7-8 X 6-7 mm,
broadly ovoid; styles 6-7 mm long; 0-75-1-1 x ovary. Capsule 12-15 x 10-15 mm,
broadly ovoid.

China (western Yunnan, northern Burma). Map 25.

CHINA. Yunnan: Cangshan, above Yangbi, Shanchang, 2700 m, 9.v.1981,
Sino-Brit. Exp. Cangshan 421 (BM, E); Salwin-Kiukiang Divide, Panbahlung,
2500 m, 22.x.1938, Yii 20827 (KUN).

BURMA. Kachin: Nam Tamai valley (Adung Wang to Gamlang Wang), 3000—
3300 m, 2.x.1937, Kingdon-Ward 13350 (BM).

CULTIVATED. Specimens of Sino-British Expedition to Cangshan (421, 550,
693) seen from England (1983).

STUDIES IN THE GENUS HYPERICUM L.

35b(46b). Hypericum bellum subsp. bellum
Icon: Robson in J. Roy. Hort. Soc. 95: f. 239 (1970).

Leaves with petiole 0-5—2-5 mm, long; lamina 15-65 x 7-43 mm, ovate-oblong or
broadly rhombic to subcircular (1: b = 2 or less). Flowers 25-35 mm in diam.; buds
obtuse to rounded. Sepals 3-9 x 2-5-6 mm, narrowly elliptic to obovate, rounded or
rarely subapiculate, entire or finely eroded-denticulate. Petals 15—25(-30) x
11-21(-21) mm. Stamen fascicles with longest stamen 6-10(—11) mm long, 0-35-0-4(-
0-6) X petals. Ovary 4-6 x 3-3-5 mm, broadly to narrowly ovoid; styles 3-6 mm long,
0-6-1 X ovary. Capsule 10-15 x 6-10 mm, broadly to narrowly ovoid.

China (NW. Yunnan, SW. Sichuan, SE. Xizang), India (Arunachal Pradesh). Map
2,

CHINA. Yunnan: Atuntze [Tehtsin], Dokerla, 2800-3000 m, 1.xi.1937, Yi 7823
(BM, E); Chungtien, Haba, 2600 m, 22.xi.1937, Yiu 14939 (BM, E). Sichuan:
Pao-hsing-hsien [Mupin], 22.vii.1936, Chu 3213 (BM, E, K); Mao Hsien, vii.1930,
Wang 21846 (KUN). Xizang [Tibet]: Rong T6 valley, above Migu, Zayul, 3000—-
3300 m, 28.xi.1933, Kingdon Ward 11024 (BM); Kongbo, Chema Dzong, Pasum
Chi, 3300 m, 17.ix.1947, Ludlow, Sherriff & Elliott L. & S. 15737 (BM, E); Qagan,
2650 m, 2.vii.1973, Qing 73-54 (KUN).

INDIA. Arunachal Pradesh; Walong, 2400-2700 m, 28.ix.1950, Kingdon Ward
22018 (BM).

CULTIVATED. Specimens seen from England (via France, probably ex Monbeig
(1912-1919), and ex Ludlow, Sherriff & Elliot 15737 (1952-1983)).

H. bellum subsp. bellum was first introduced into Europe from Yunnan by the
French Jesuit missionary Pierre Monbeig, and Otto Stapf of Kew named it H.
monbeigii in the late 1920s. This name, however, was never published. Recent
European cultivated material originated from later introductions from Tibet by
Ludlow & Sherriff.

36(47). Hypericum kouytchense H. Léveillé

in Bull. Soc. Agric. Sci. Arts Sarthe 39: 322 (1904), in Bull. Soc. bot. Fr. 54: 592
(1908), in Reprium nov. Spec. Regni veg. 6: 375 (1909) [reprint of 1904 paper];
Lauener in Notes Roy. Bot. Gard. Edinb. 27: 3 (1966); N. Robson inJ. Roy. Hort.
Soc. 95: 490, ff. 240, 241 (1970); [Lancaster], Hilliers’ man. trees & shrubs: 151
(1971); Bean, Trees & shrubs hardy in Br. Isles 8th ed., 2: 417 (1973). Type: China,
Guizhou, mont de Lou-tsang-koan [Lou shan Guan], 31.v.1897, Bodinier 1603
(E!, holotype; A!, isotype).

H. patulum sensu H. Léveillé in Bull. Soc. bot. Fr. 54: 591 (1908) pro parte, quoad
specim, Bodinier 2662, non Thunb. ex Murray (1784).

H. ascyron sensu H. Léveillé, Fl. Kouy-Tchéou: 198 (1914), non L. (1753).

H. longifolium sensu H. Léveillé, Fl. Kouy-Tchéou: 199 (1914) pro parte, quoad
Bodinier 1774, non H. Léveillé. (1904).

Norysca kouytchensis (H. Léveillé) Y. Kimura in Nakai & Honda, Nova fl. jap. 10:
98 (1951) (‘koutchense’).

H. pendulifiorum hort. — Plaisted & Lighty in Nat. Hort. Mag. 38: 125, 128 (t.)
(1959); Thomas in Gdnrs’ Chron. III, 147: 255 cum fig. (1960); nomen.

H. patulum var. grandiflorum hort. — Thomas in Gdnrs’ Chron. III, 147: 226 (1960),
nomen.

H. grandiflorum sensu hott. pro parte, non Salisb. (1798).

H. patulum ‘Sungold’, hort. amer.

H. patulum ‘Laplace’ (Boom c. 1934).

H. patulum ‘Summergold’ (Boom c. 1956).

Icon: N. Robson in J. Roy. Hort. Soc. 95: ff. 240, 241 (1970).
Shrub 1-1-8 m tall, with branches arched or pendulous. Stems red, 4-lined when

young, becoming 2-lined, eventually terete; internodes 10-40 mm long, shorter than
to exceeding leaves; bark reddish-brown. Leaves petiolate, with petiole 0-5—1-5 mm

275

se
’
¢

276

NORMAN K. B. ROBSON

long; lamina 20-58 x 6-30 mm, elliptic or lanceolate to ovate or triangular-ovate,
acute to obtuse or rarely rounded-apiculate, margin plane, base cuneate or sub-
angustate to rounded, paler beneath, not to scarcely glaucous, chartaceous; vena-
tion: 3-4(5) pairs main laterals, with midrib branched distally, without or with
obscure tertiary reticulum; laminar glands dots and short streaks, ventral glands +
dense. Inflorescence 1-7(-11)-flowered, from 1(2) node(s), subcorymbiform;
pedicels 5-10 mm long; bracts intermediate to narrowly lanceolate, deciduous.
Flowers 40-65 mm in diam., stellate; buds narrowly ovoid, acute to subacuminate.
Sepals 7-15 X 2-5-7 mm, free, imbricate, equal, + spreading in bud and fruit +
narrowly ovate to lanceolate, acute to acutely acuminate, with margin entire; midrib
conspicuous, veins not prominent; laminar glands linear, c. 10-11. Petals bright
golden yellow, not tinged red, spreading or recurved, 24-40 x 16-25 mm, c. 3 +
sepals, obovate-oblong to obovate, with apiculus subterminal, acute; margin
minutely glandular-denticulate towards apex. Stamen fascicles each with 35-50
stamens, longest 18-29 mm long, 0-65—-0-8 x petals; anthers golden yellow. Ovary
6-8 x 4-6 mm, ovoid-pyramidal to narrowly ovoid; styles 8-10 mm long, c. 1-2-
1-35 x ovary, free, erect, slightly outcurved at apex; stigmas small. Capsule 17-20 x
8-10 mm, + narrowly ovoid-pyramidal to ovoid, red when immature. Seeds dark
purplish-brown, 2-3-2 mm long, narrowly cylindric, narrowly winged, almost
smooth. 2n = 36, 40 (also 20?, see below).

Pastures, hillsides, streamsides, among rocks; c. 1500-2000 m.

China (Guizhou). Map 26.

CHINA. Guizhou: Kouy-yang [Guiyang], mont a Kien Lin chan Hem,
20.vii.1899, Bodinier 2662 (E); Tafig-tchafig (Houang tsa6 pa), vi.1907, Esquirol
1461 (E).

CULTIVATED. Specimens seen from China, England, Scotland, France, the
Netherlands, Germany, Finland, U.S.A. (east, central, west, and Hawaii). It is
naturalised in New Zealand (S. Island, south-western distr., near Lake Paringa), fide
W. R. Sykes (in litt. et photo, 1983).

H. kouytchense has a rather restricted distribution in north-western Guizhou. In
some respects it seems to be one of the most primitive species in sect. Ascyreia, e.g. in
its long styles and stamens, acute leaves and sepals, large stellate flowers and conical
ovary and capsule. Yet its arching to pendulous habit and petiolate leaves with
punctiform laminar glands are relatively advanced characters; and it is clearly related
to H. beanii and H. maclarenii. On the other hand, its acute petal apiculus and long
stamens and styles indicate a close relationship with the H. stellatum group.

The type specimen of H. kouytchense and, probably, Bodinier 1774 (which I have
not seen) differ from the other material studied (both wild and cultivated) in having
more delicate stems, narrowly elliptic to narrowly lanceolate leaves that are smaller
and thinner, and smaller flowers with narrower sepals, in all giving the plant a more
delicate appearance. These differences suggest that the type is diploid (2n = 20), like
H. uralum, H. dyeri and (by inference from the occurrence of that number in H. x
‘Rowallane’) H. hookerianum and H. leschenaultii. On the other hand, the culti-
vated material, as well as Bodinier 2662 and Esquirol 1461, has stouter stems,
thicker, lanceolate to ovate leaves, and larger flowers with broader sepals, all
characters that would be expected in the tetraploid that it has proved to be (2n = 40,
36). These numbers support its close relationship with H. beanii (2n = 36) and H.
forrestii (2n = 38, 36); and H. pseudohenryi will doubtless also be shown to be
tetraploid. When more material has been studied, both morphologically and cyto-
logically, it may be best to treat H. kouytchense as two subspecies, diploid and
tetraploid. At present, however, I shall maintain it undivided.

36x(47x). Hypericum kouytchense x calycinum

This hybrid was made by Mr D. Walker of North Mymms, Hatfield, Herts; but as I do
not know if it is still in cultivation and have a specimen of only one flower and two
pairs of leaves (coll. 18.ix.1976), I have refrained from naming it or describing it in
detail. It is intermediate between the parents, having lanceolate to elliptic, apiculate

STUDIES IN THE GENUS HYPERICUM L. PLT

Map 26 Sect. 3. Ascyreia: 33. H. maclarenii A, 36. H. kouytchense ©, 37. H. stellatum yx, 38. H.
lancasteri @ , 39. H. curvisepalum O.

leaves with rather dense but obscure tertiary reticulation beneath; a flower 80 mm in
diam.; sepals 10-12 x 6-7 mm, broadly elliptic and rounded-apiculate; petals 40 x
28 mm, intermediate in form with uncinate to obtuse subterminal apiculus; stamens
16-17 mm long, 0-4 x petals, with yellow anthers; and ovary ovoid, 10 X 6mm, with
styles 13 mm (1-3 X ovary), gradually outcurving.

36xx(47xx). Hypericum x ‘Eastleigh Gold’

Hilliers’ man. trees & shrubs: 150 (1971). Standard specimen: cultivated at
Ampfield, Hampshire, Hilliers’ Jermyns Lane Nursery, 24.vii.70, Lancaster s.n.
(BM!).

H. beanii ‘Eastleigh Gold’ N. Robson in J. Roy. Hort. Soc. 95: 488 [491 in synon.]
(1970).

Shrub to c. 1 m tall, loose, with spreading drooping branches. Stems reddish, 4-lined
and ancipitous when young, soon 2-lined, eventually terete, stout; internodes 20—
60 mm long, shorter than to exceeding leaves; bark reddish-brown. Leaves petiolate,
with petiole c. 1 mm long; lamina 25-51 x 10-25 mm, elliptic-oblong to oblong-
lanceolate, obtuse to rounded, margin plane, base + broadly cuneate to rounded,
paler or somewhat glaucous beneath, chartaceous; venation: c. 4 pairs main laterals,
with scarcely visible subsidiaries, without visible tertiary reticulum; laminar glands
dots and short streaks; ventral glands dense. Inflorescence 14-flowered, from apical

278

NORMAN K. B. ROBSON

node, rather lax; bracts foliaceous or reduced and lanceolate, deciduous. Flowers
50-65 mm in diam., subcyathiform; buds narrowly ovoid, acute to subapiculate.
Sepals 10-12 x 2:5-4-5 mm, free, not imbricate, equal to subequal, outcurved in bud
and fruit, narrowly oblong or narrowly oblong-lanceolate, acute to acuminate, with
margin entire; midrib visible, veins not prominent; laminar glands linear, c. 10.
Petals golden yellow, not tinged red, slightly incurved, 30-35 x 20-25 mm, c. 3 X
sepals, oblong-obovate, with apiculus subterminal, obtuse; inner margin minutely
denticulate. Stamen fascicles each with 30-50 stamens, longest 10-12 mm long,
0-3-0-4 x petals; anthers deep yellow. Ovary 7-5-8 x 4mm, narrowly ovoid-conic;
styles 5-5—6-5 mm long, 0-75—-0-9 x ovary, free, erect, outcurved towards apex;
stigmas small. Capsule c. 12 X 6 mm (or larger ?), narrowly conic, often puckered.
Seeds not seen. 2n = ?

Hybrid (?) seedling in cultivation.

CULTIVATED. Specimens seen from England (The Hillier Arboretum,
Ampfield, Hampshire (see type); Chelsea Physic Garden, London; Royal Botanic
Gardens, Kew; Wakehurst Place, W. Sussex).

Hypericum x ‘Eastleigh Gold’ arose c. 1964 as a seedling in the Eastleigh Nursery of
Messrs Hillier & Sons and was selected by R. Alford (nursery foreman). It was first
included by me provisionally under H. beanii as a cultivar (1970) but later treated as a
cultivar unassigned to species (1971). It is almost certainly of hybrid origin and
therefore justifies the above use of the hybrid sign; but its parentage is not clear. H.
kouytchense appears to be involved (shape of sepals and ovary), and the oblong,
rather glaucous leaves suggest H. acmosepalum as the other parent; but the short
stamens would seem to rule it out. H. hookerianum would provide the short stamens
and even possibly the leaf shape, but not the sepal shape. Perhaps cytological studies
would provide the answer to this problem.

37(48). Hypericum stellatum N. Robson

inJ. Roy. Hort. Soc. 95: 493, f. 237 (1970); [Lancaster], Hilliers’ Man. trees & shrubs:
152 (1971); Bean, Trees & shrubs hardy in Br. Isles 8th ed. 2: 422 (1973). Type:
cultivated in Hort. Kew. ex France, Les Barres (de Vilmorin), 14.vii.1904, Bean
(K!, holotype).

H. lysimachioides sensu hort. non Wallich (1831, nomen) nec Boiss. & Noé (1867).

H. dyeri sensu hort.—Plaisted & Lighty in Nat. hort. Mag. 38: 125 (1959); Thomas in
Gdnrs’ Chron. III, 47: 226 (1960); non Rehder (1939).

Icones: N. Robson in J. Roy. Hort. Soc. 95: f. 237 (1970); Fig. 19B.

Shrub c. 1-2-5 m tall, with branches spreading to subpendulous. Stems orange-red,
4-lined and ancipitous when young, soon 2-lined, sometimes becoming terete;
internodes 10-31 mm long, shorter than leaves; bark reddish-brown. Leaves
petiolate, with petiole 1-2 mm long; lamina 20-55 x 10-22 mm, oblong-lanceolate or
lanceolate to narrowly ovate, acute to obtuse or rounded-apiculate, margin plane,
base cuneate to rounded, paler or sometimes + densely glaucous beneath, char-
taceous; venation: 4(5) pairs main laterals, with midrib branching + obscurely
distally, without visible tertiary reticulum; laminar glands dots and short streaks,
ventral glands dense. Inflorescence 1—14-flowered, lax, with branches slender, from
terminal node; pedicels 10-15 mm long; bracts narrowly lanceolate, deciduous.
Flowers 25-40 mm in diam., stellate; buds + broadly ovoid, apiculate to shortly
acuminate. Sepals 8-13 x 2-5 mm, free, imbricate or open at base, equal, widely
spreading to subrecurved in bud and fruit, + narrowly lanceolate, acute, with margin
entire (reddish) midrib conspicuous, veins not prominent; laminar glands linear c.
10. Petals golden yellow, sometimes tinged red, spreading to shallowly incurved,
12-20 x 8-14mm, c. 1-5 X sepals, obovate, with apiculus subterminal, acute, margin
entire or minutely denticulate towards apex. Stamen fascicles each with 30-55
stamens, longest 10-13 mm long, c. 0-6 X petals; anthers golden yellow. Ovary 4-6 X
3-4 mm, ovoid-conic to ovoid; styles 6-9-5 mm long, 1:2-1-5 Xx ovary, usually

STUDIES IN THE GENUS HYPERICUM L.

flexuous and + twisted: stigmas scarcely capitate. Capsule 10-15 x 6-8 mm, ovoid.
Seeds c. 1-1 mm long, dark red-brown, narrowly cylindric, not carinate, shallowly
and loosely reticulate. 2n = ?

Thickets and slopes; 800-1350 m.

China (NE. Sichuan). Map 26.

CHINA: Sichuan: Chenkou Xian, 1350 m, 17.vi.1958, Dai 100888 (SZ); ibid.,
800 m, 27.viii.1958, Dai 102054 (SZ).

CULTIVATED. Specimens seen from England (1982), France (1904), Germany
(1924), Poland (1912), and U.S.A. (1962).

Hypericum stellatum can be distinguished from H. dyeri, with which it has been
confused, by the relatively shorter styles (c. 1-2-1-5 x ovary, not 1-5-2 x), larger
flowers, broader petals and larger leaves without markedly reticulate venation.

The original concept of H. stellatum was based mainly on cultivated plants of
unknown origin, but it included some herbarium material of which some has
subsequently proved to belong to H. choisianum and two hitherto undescribed
species, H. lancasteri and H. curvisepalum. The status of the cultivated specimens,
however, has been clarified by recent collections made in Yunnan and from the study
of specimens in Chengdu herbarium (SZ). It appears that there are two populations,
respectively in northern Yunnan and north-eastern Sichuan, which can be separated
by the size of the ovary, the length and form of the styles, the stoutness of the
inflorescence, and usually by the density of ventral leaf glands.

The northern population is H. stellatum and the southern one, which merits
specific status (taking into account the above differences and the geographical
disjunction), I have named H. lancasteri.

38(49). Hypericum lancasteri N. Robson, sp. nov.

H. stellato N. Robson affinis, sed foliis lanceolatis vel oblongo-lanceolatis vel
triangulari-lanceolatis glandulis superficiebus plerumque haud vel parce instructis,
inflorescentiae ramificationibus pro ratione crassioribus, ovario majori stylis quam
ovario pro ratione longioribus curvatis haud flexuoso-torsivis, differt. Type: China,
Yunnan, above Dali, Qingbixu, 2000 m, 23.v.1981, Sino-Brit. Exped. Cangshan
1096 (E!, holotype).

Icon: Fig. 19A.

Shrub 0-3-1 mm tall, with branches suberect to spreading. Stems purplish-red,
4-lined but scarcely ancipitous when young, soon 2-lined, eventually terete; inter-
nodes 10—40(-60) mm long, shorter than to exceeding leaves; bark reddish-brown.
Leaves petiolate, with petiole 1-1-5 mm long; lamina 30-60 x 9-30 mm, oblong-
lanceolate or lanceolate to triangular-lanceolate, acute to rounded, margin plane,
base cuneate to rounded, paler or sometimes + densely glaucous beneath, char-
taceous; venation: 3-4 pairs main laterals, with midrib branching distally, without
visible tertiary reticulum; laminar glands dots and short streaks, ventral glands
usually sparse or absent. Inflorescence 1—11-flowered, lax, with branches relatively
stout, from 1-3 nodes; pedicels 13-30 mm long; bracts foliaceous, gradually reduced
in successive nodes, deciduous. Flowers 30-55 mm in diam., + stellate to subcyathi-
form; buds narrowly to broadly ovoid, acute to apiculate or subacuminate. Sepals
8-11 x 34 mm, free, imbricate, unequal or equal, outcurved to widely spreading in
bud, widely spreading to recurved in fruit, lanceolate to ovate or oblong-ovate, acute
to subacuminate, with margin entire (reddish); midrib conspicuous, veins not
prominent; laminar glands c. 10. Petals golden yellow, not tinged red, spreading to
shallowly incurved, 17-28 x 13-18 mm, c. 2-2-5 X sepals, oblong-obovate, with
apiculus subterminal to lateral, acute to obtuse, margin entire. Stamen fascicles each
with c. 45-50 stamens, longest 11-16 mm long, c. 0-6 x petals; anthers golden yellow.
Ovary 5-6-5 X 3-5-5 mm, ovoid; styles S—7 mm long, 1-1-2 X ovary, outcurved
distally, not twisted; stigmas scarcely capitate. Capsule 13-17 x 8-10 mm, ovoid.
Seeds dark red-brown, 1-1-3 mm long, narrowly cylindric, not or incompletely
carinate, shallowly and loosely reticulate. 2n = ?

279

280 NORMAN K. B. ROBSON

Fig. 19 A. H. lancasteri: (a) habit; (b) leaf section; (c) ovary. B. H. stellatum: (d) habit; (e) leaf section;
(f) ovary. C. H. curvisepalum: (g) habit; (h) leaf section; (i) ovary (a, d, g X 1; b, c, e, f, h,i x 4). A.
S.B.E.C. K036; B. Robson s.n.; C. Schneider 1550.

STUDIES IN THE GENUS HYPERICUM L.
Dry grassy banks and slopes; 1750-2550 m.

China (N. Yunnan, S. Sichuan). Map 26.

CHINA. Yunnan: Kunming, Hua Hong Dong, 2000 m, 26.iv.1981, Sino-Brit.
Exp. Cangshan K001 (E), K992 (E), K036 (BM, E); K039 (E); Kunming, Western
Hills, c. 2500 m, 17.x.1980, Lancaster L. 750 (BM), Rushforth KR 327 (E); Tali
[Dali] Range, 1906, Forrest 4300 B (BM); Cangshan, above Yangbi, Xieshanhe,
1850 m, v.1981, Sino-Brit. Exp. Cangshan 34 (E); Tong-tchouan [Hweitseh], 2550
m, vi.1913, Maire s.n. (E). Sichuan: Huidong Xian, 1750 m, 3.vii.1978, Zhao 5802
(SZ).

CULTIVATED. Specimens seen from England (1984) and Scotland (1982), ex
Lancaster L. 749, L. 750 and Rushforth KR327.

For a comparison between H. lancasteri and H. stellatum see after description of the
latter (p. 279).

39(50). Hypericum curvisepalum N. Robson, sp. nov.

H. lancasteri N. Robson affinis, sed floribus minoribus profunde cyathiformibus,
staminibus petalis circa triplo brevioribus, stylis ovario circa duplo brevioribus,
capsulis valvis crassioribus, differt. Type: China, Sichuan, ‘inter Woloho et Humati’,
C. Schneider 1550 (E!, holotype; K!, isotype).

Icon: Fig. 19C.

Shrub 0-3-1-2 m tall, with branches spreading to pendulous. Stems yellow-brown
(young parts purplish at first), 4-lined when young, soon terete, slender, internodes
10-25 mm long, shorter than leaves; bark grey-brown. Leaves petiolate, with petiole
0-5-1 mm long; lamina 20-40 x 8-20 mm, triangular- lanceolate to triangular-ovate,
acute or more rarely obtuse to rounded, margin plane, base rounded to shallowly
cordate, + glaucous beneath, chartaceous; venation: 3-4 pairs main laterals,
branched, the midrib pinnately branched, with tertiary reticulum obscure or in-
visible; laminar glands streaks (occasionally elongate) and dots. Inflorescence 1(3)-
flowered, from apical node; pedicels 6-10 mm long; bracts linear or usually foliar.
Flowers 20-40 mm in diam., deeply cyathiform; buds ovoid, acute to apiculate.
Sepals 8-14 x 3-5 mm, free, imbricate, unequal, outcurved or spreading and
purplish in bud and fruit, ovate to lanceolate or narrowly elliptic, subacute to
acuminate or rarely apiculate-obtuse, margin entire, midrib narrow sometimes
indistinct; laminar glands linear, interrupted distally, 8. Petals deep yellow, marked-
ly incurved, 12-22 x 8-17 mm, 1-1—1-5(—2) x sepals, broadly obovate to subcircular,
with apiculus subterminal, obtuse to rounded; margin entire, eglandular. Stamen
fascicles each with c. 60 stamens, longest 10-12 mm long, 0-35-0-7 x petals; anthers
deep (?) yellow. Ovary 6-8 x 4-5-6 mm, + broadly ovoid; styles 34 mm long, c. 0-5
X ovary, free, outcurved near apex; stigmas narrowly or scarcely capitate. Capsule
(12)14-17 x 8-10 mm, ovoid-conic to broadly ovoid, with thickly coriaceous valves.
Seeds dark reddish-brown, 0-8-1 mm long, cylindric, not winged, not or scarcely
carinate, shallowly linear-foveolate. 2n = ?

Dry or rocky hillsides and open woodland; 1800-3000 m.

China (north Yunnan, south Sichuan, west Guizhou). Map 26.

CHINA. Yunnan: c. 15 km W. of Kunming, Western Hills, around Tai Hua
Temple, 2300 m, 24.iv.1981, Sino-Brit. Exp. Cangshan K.118 (E); Tali Range, 1906,
Forrest 4300A (BM); above Dali, Qingbixu, 2300 m, 23.v.1981, Sino-Brit. Exp.
Cangshan 1081 (E); above Weishan, Santaipo, S. end of Cangshan, 2350 m,
21.v.1981, Sino-Brit. Exp. Cangshan 1198 (BM); above Yangbi, Zhaimen, 2400 m,
6.v.1981, Sino-Brit. Exp. Cangshan 0283 (E); Tong-Tchouan, 2500-2600 m, v.1913,
Maire 443/1913 (E), 444/1913 (BM, E); Tehching, Wuyen, 2400 m, 2.vi.1937, Yu
8401 (BM). Sichuan; between Woloho and Humati, 2800-3000 m, 14.vi. 1914,
Schneider 1550 (E, K). Guizhou: Pu’an Xian, 1800 m, 7.ix.1959, Anchuan Team
1176 (KUN).

CULTIVATED. Specimens seen from England (1983).

281

282

CS SS

NORMAN K. B. ROBSON

H. curvisepalum is clearly derived from H. lancasteri, which has shorter styles than
H. stellatum and sometimes 1-flowered inflorescences. In general, however, H.
curvisepalum has smaller, more deeply cup-shaped, nodding flowers with relatively
shorter stamens and styles; and it is usually more slender and delicate except for the
fruits which, although smaller, tend to be rather massive. Its leaves tend to be shorter
and broader and have a shorter petiole and longer translucent glandular streaks.

I was reluctant, at first, to describe two closely related species with such overlap-
ping distributions; but the above differences seem to be constant. It seems possible,
therefore, that H. lancasteri gave rise to H. curvisepalum by ‘trend variation’ and
subsequent isolation, and that the present syntopy of these taxa is secondary.

40(51). Hypericum beanii N. Robson

in J. Roy. Hort. Soc. 95: 490, excl, f. 235 (1970); [Lancaster], Hilliers’ man. trees &
shrubs: 150 (1971); Bean, Trees & shrubs hardy in Br. Isles ed. 8, 2: 406 (1973).
Type: Cultivated, R.B.G. Kew ex China, Yunnan (Henry 179/1898), 4.x.1899 (K!,
holotype).

H. patulum var. henryi Veitch [in Gdnrs’ Chron. III, 36: 229 (1904), nomen] ex Bean
in Gdnrs’ Chron. II, 38: 179 (1905); Plaisted & Lighty in Nat. Hort. Mag. 38: 124
(1959); Thomas in Gdnrs’ Chron. III, 147: 226 (1960). Type as for H. beanii.

H. pseudohenryi sensu N. Robson in Yb. int. Dendr. Soc. 1980: 141 (1981).

Icon: Lancaster in Yb. int. Dendr. Soc. 1980: 124, t. 54 (1981) as H. pseudohenryi.

Shrub 0-6-2 m tall, bushy, with branches erect or arching. Stems red to orange,
4-angled and ancipitous at first, eventually terete; internodes S40(—-50) mm long,
shorter than or exceeding leaves; bark reddish brown. Leaves petiolate, with petiole
1-2-5 mm long; lamina 25-65 x 10-35 mm, narrowly elliptic or oblong-lanceolate to
lanceolate or ovate-lanceolate, acute or apiculate to obtuse or sometimes rounded,
margin plane, base cuneate to rounded, paler or glaucous beneath, chartaceous to
subcoriaceous; venation: (2)3-5 pairs of main laterals, all free or the upper forming a
partial undulating intramarginal vein, the midrib branching distally, with lax obscure
tertiary reticulum; laminar glands dots and short to + long streaks, ventral glands
dense, sometimes only near midrib, or absent. Inflorescence 1—-14-flowered, from
apical node, subcorymbiform, often with lateral flowering branches below; pedicels
3-20 mm long; bracts foliar to narrowly lanceolate, persistent. Flowers 30-45 mm in
diam., stellate to cyathiform; buds ovoid-conic to broadly ovoid, acute to bluntly
apiculate. Sepals 6-11(—14) x 3-6-5(—10) mm, free, imbricate (sometimes markedly
so), equal or subequal, erect to spreading in bud and fruit, ovate to oblong-ovate or
broadly elliptic, acute or apiculate to obtuse with margin hyaline, entire or distally
minutely denticulate; midrib conspicuous, + prominent, veins rarely conspicuous;
laminar glands linear, + interrupted distally, c. 10-14. Petals golden yellow, not
red-tinged, spreading to rather deeply incurved, 15-33 x 10-30 mm, c. 2-4-5 x
sepals, oblong-obovate to subcircular, with apiculus lateral to subterminal, obtuse to
rounded; margin entire to irregularly eroded-denticulate, eglandular. Stamen fasci-
cles each with 40-55 stamens, longest 10-15 mm long, 0-5—0-7 x petals; anthers
golden yellow. Ovary 6-9 x 4-5 mm, ovoid-pyramidal to narrowly ovoid-cylindric;
styles 4-9 mm long, 0-65—1-1 x ovary, free, suberect, outcurved near apex; stigmas
narrowly capitate to truncate. Capsule 15-20 x 8-11 mm, narrowly ovoid-conic to
ovoid. Seeds dark reddish- to purplish-brown, 1-1-5 mm long, narrowly cylindric,
broadly carinate, shallowly linear-reticulate. 2n = 36.

In open forest or thickets or by streams, and on grassy or stony slopes; 1500-2400 m.

China (E. Yunnan, W. Guizhou). Map 27.

CHINA. Yunnan: Kunming, Hua Hong Dong, 2100 m, 18 x 1980, Lancaster 752
(BM); vicinity of Yunnan-sen [Kunming], [rec. xi.1906], Maire 291 (E); near
Kunming, Dashao, 2400 m, 26.iv.1981, Sino-Brit. Exped. Cangshan K. 047 (BM, E);
Little Stone Forest, c. 120 km SE. of Kunming, 1750 m, 19.x.1980, Lancaster L.711
(BM); Mengtze, south mountains, 1500 m, 1897?, Henry 9986 (K). Guizhou:
Tcheou-fong, viii.1904, Esquirol 187 (E).

STUDIES IN THE GENUS HYPERICUM L. 283

Map 27 Sect. 3. Ascyreia: 40. H. beanii O, 41. H. pseudohenryi @, 42. H. forrestii O.

CULTIVATED. England: Kew, Royal Botanic Gardens, Arboretum,
19.ix.1905, Henry 179/1898 (K); Hampshire, Ampfield, Jermyns Lane Nursery,
3.viil.1970, Lancaster s.n. (BM). Specimens of Lancaster L.711 (Little Stone
Forest), L.751 and 752 (Hua Hong Dong) and L.653 (Western Hills), all from near
Kunming, flowered in England and Scotland in 1982 or 1983.

H. beanii is a variable species that shows affinities on the one hand with H.
kouytchense and on the other with H. pseudohenryi and H. forrestii. It shows a north
to south cline from (i) a form with leaves elliptic, acute, with long glandular streaks,
flowers stellate or shallowly cyathiform and stamens and styles relatively long to (ii) a
form with leaves ovate, obtuse to rounded, with very short glandular streaks or dots,
flowers deeply cyathiform and stamens and styles relatively short.

The type of H. beanii is a cultivated specimen grown at Kew from seed sent by Dr
Henry in 1898 from Yunnan, probably from the Mengtze region.

41(52). Hypericum pseudohenryi N. Robson

in J. Roy. Hort. Soc. 95: 493 (1970); [Lancaster], Hilliers’ man. trees & shrubs: 152
(1971); Bean, Trees & shrubs hardy in Br. Isles 8th ed.2: 406 (1973). Type: China,
Yunnan, Chung-tien district, north of Chung-tien in Tonwa Territory, 3000 m,
iv—v.1932, Rock 24673 (BM!, holotype; A!, E!, K!, isotypes).

H. patulum sensu R. Keller in Bot. Jahrb. 44: 48 (1909), non Thunb. ex Murray
(1784).

284

-O

ees

aaa
?

NORMAN K. B. ROBSON

H. patulum var. henryi sensu Rehder in Sargent, Pl. Wils. 2: 403 (1915) et hort. pro
parte, non Veitch ex Bean (1905).
H. henryi sensu hort. pro parte, non H. Léveillé & Vaniot (1908).

Icon: Fig. 20.

Shrub 0-7-1-7 m tall, + spreading, with branches erect to arching. Stems red,
4-angled and ancipitous in first year, then terete, stout; internodes 8-60 mm long,
shorter than to exceeding leaves; bark grey-brown. Leaves petiolate, with petiole
0-5-1 mm long; lamina 20-66(—80) x 5-35 mm, ovate or ovate-oblong to lanceolate
or lanceolate-oblong, rounded or rarely apiculate-obtuse, margin plane, base nar-
rowly to + broadly cuneate, paler or somewhat glaucous beneath, chartaceous;
venation: 2-3 pairs of main laterals (the upper forming distinct undulating intramar-
ginal vein), the midrib branched distally, with lax obscure tertiary reticulum; laminar
glands dots and short streaks; ventral glands densest or present only near midrib.
Inflorescence 1-7(-c. 25)-flowered, subcorymbiform, from apical node; pedicels
4-11 mm long; bracts foliar to narrowly lanceolate, persistent. Flowers 30-55 mm in
diam., stellate to subcyathiform, buds ovoid-pyramidal, subacute. Sepals 6-9(-13) x
3-7 mm, free, imbricate, subequal, erect to outcurved in bud and fruit, broadly to
narrowly ovate-oblong, acute or subacuminate to obtuse, with margin narrowly
hyaline, entire or minutely denticulate towards apex; midrib conspicuous, veins not
prominent; laminar glands linear, 8-10. Petals golden yellow, not red-tinged,
spreading to reflexed, 16-30 x 10-17 mm, 2-5-3 X sepals, obovate, with apiculus
subterminal, obtuse; margin entire to irregularly eroded-denticulate, eglandular.
Stamen fascicles each with c. 40 stamens, longest 14-20 mm long, 0:75-0:85 xX petals;
anthers golden-yellow. Ovary 5-9 x 3-5-6 mm, + broadly ovoid; styles 5-S—11 mm
long, somewhat longer than ovary, free, suberect to divergent, outcurved near apex;
stigmas truncate. Capsule 12-17 x 10-14 mm, ovoid-conic to ovoid. Seeds dark
orange-brown, 15-20 mm long, narrowly cylindric, narrowly carinate, shallowly
linear-reticulate. 2n = ?

Pine forest, thickets, dry grassy or stony slopes; 1400-3800 m.

China (north Yunnan, central Sichuan). Map 27.

CHINA. Yunnan: Dongchuan Shi [Hweitseh], 2740 m, 12.viii.1964, Dian Dong-
bei group 468 (KUN); Lichiang Snow Range, 2800 m, 29. viii.1937, Yii 15503 (BM,
E); in regione Yungning-Yungpeh inter Pij et Paoto, c. 2600 m, 25.vi.1914,
Schneider 3054 (E). Sichuan: Muli, Wachin, Yatsa, 3100 m, 5.xi.1937, Yui 14729
(BM, E); Tachien-lu [Kangding], 1500-2400 m, xi.1908, Wilson 1355. (BM, K); west
and near Wénchuan Hsien, 1200-1800 m, vii.1908, Wilson 2418 (BM, E, K);
Huidong Xian, 2550 m, 20.vii.1960, Hu 60-1390 (SZ); Tianquan Xian, 1700 m,
7.vi.1953, Tsiang 34335 (SZ).

IRELAND (naturalised). Co. Cork, Glengariff, Glengariff wood, 3.viii.1983,
Stern s.n. (BM).

CULTIVATED. Specimens seen from England (1919-1983), Ireland (1973—
1980), France (1926), the Netherlands (1933-1936), Finland (1933), and the U.S.A.
(1911-1959).

H. pseudohenryi is related to H. beanii but has a distribution (in north-western
Yunnan and Sichuan) wholly to the north-west of that species. It differs from it in its
oblong-lanceolate to oblong-ovate leaves, usually with a rounded apex and a quite
marked partial intramarginal vein, and its ovoid (not pyramidal) capsules. H.
pseudohenryi appears to be one of the commonest species of sect. Ascyreia in
Sichuan, along with H. monogynum and H. patulum.

Two specimens appear to be intermediate between H. pseudohenryi and H.
lagarocladum: Sichuan: Tianquan Xian, 21.v.1956, He 44087 (SZ); Yanyuan,
1900 m, 23.v.1978, Zhao et al. 4586 (SZ). Since both localities are in the region where
the distributional areas of these species overlap, it seems likely that these specimens
are hybrids.

STUDIES IN THE GENUS HYPERICUM L. 285

Ave
\ i
EEO Se a
ny fe
5

Fig.20 H. pseudohenryi: (a) habit; (b) leaf section; (c) flower bud; (d) sepal; (e) petal; (f) stamen fascicle;
(g) anthers; (h) inflorescence in young fruit; (i) capsule (a, h X 1; b,c, i x 4; d-f x 6; g x 20). All Forrest
5885.

286

or:

NORMAN K. B. ROBSON

42(53). Hypericum forrestii (Chittenden) N. Robson

inJ. Roy. Hort. Soc. 95: 491, f. 242 (1970); [Lancaster], Hilliers’ man. trees & shrubs:
151 (1971); Bean, Trees & shrubs hardy in Br. Isles 8th ed. 2: 412 (1973). Type:
cultivated, R.B.G. Kew ex China, Yunnan (Forrest A.669), 11.vii.1922, Hort.
Kew. H. 1061 (K!, holotype).

H. hookerianum sensu Rehder in Sargent, Pl. Wils. 2: 403 (1915) pro parte excl.
spec. Pratt 292; Hand.-Mazz., Symb. sin. 7: 401 (1931).

H. patulum ‘Rothschild’s Form’ Bean in Gdnrs’ Chron. III, 72: 235, f. 93 (1922),
nomen.

H. patulum var. forrestii Chittenden in J. Roy. Hort. Soc. 48: 234 (1923); Plaisted &
Lighty in Nat. Hort. Mag. 38: 124 (1959); Thomas in Gdnrs’ Chron. III, 147: 226
cum fig. (1960); Schneider in Meded. Inst. Vered. Touinbougew, Wageningen 252:
21 (1966), in Dendrofl. 2: 21 (1967) pro parte excl. fig. et syn. H. patulum
grandiflorum, H. patulum ‘Sungold’.

H. patulum forma forrestii (Chittenden) Rehder, Bibl. cult. trees & shrubs: 463
(1949).

H. calcaratum hort., nomen.

H. patulum var. henryi sensu hort. pro parte, non Bean (1904).

Icones: Thomas in Gdnrs’ Chron. II, 147: 227 (1960); also Bean (1922) and Robson
(1970) supra cit.

Shrub 0-3-1-5 m tall, bushy, with branches + erect. Stems red to orange, 4-angled
and slightly ancipitous when young, soon terete; internodes 10-45(-60) mm long,
shorter than or rarely exceeding leaves; bark grey-brown, smooth, fissuring. Leaves
petiolate, with petiole 0-5-2 mm long, + broad; lamina 20-53(—60) x 9-32(-35) mm,
lanceolate or triangular-ovate to + broadly ovate, obtuse to rounded or slightly
retuse, margin plane, base broadly cuneate to rounded, paler beneath, chartaceous;
venation: 4—S pairs main laterals, forming undulating intramarginal vein with midrib
branches, with obscure or invisible tertiary reticulum; laminar glands short streaks
and dots; ventral glands dense, especially near midrib. Inflorescence 1-c. 20-
flowered, from 1 or occasionally 2 nodes, subcorymbiform, pedicels 4-10 mm long;
bracts lanceolate to + foliaceous, persistent. Flowers (25—)35—60 mm in diam., +
deeply cyathiform; buds broadly ovoid, obtuse to rounded. Sepals 6-9 x 3-8 mm,
free, imbricate, subequal to equal, erect in bud and fruit, ovate or + broadly elliptic
to subcircular, rounded or rarely subapiculate, with margin entire or finely eroded-
denticulate towards apex and often + hyaline; midrib distinct, veins not prominent;
laminar glands linear, + interrupted distally, 12 or more. Petals golden yellow, not
tinged red, markedly incurved, 18-30 x 11-25 mm, 3-3-5 x sepals, broadly obovate,
with apiculus subterminal, rounded; margin entire (?) or remotely glandular-
subdenticulate. Stamen fascicles each with 40-65 stamens, longest 10-15 mm long,
0-4-0-6 X petals; anthers golden yellow. Ovary (4-5—)6-8 x 44-5 mm, broadly
ovoid; styles 4-7 mm long, 0-7-0-9(-1) x ovary, free, outcurved near apex; stigmas
small. Capsule 12-18 x 8-14 mm, + broadly ovoid. Seeds dark reddish-brown,
1—2-1-7 mm long, narrowly cylindric, distally slightly carinate or winged, very
shallowly scalariform-reticulate. 2n = 36, 38.

Open stony situations on hillsides, sometimes by streams or at Pinus forest margins;
1500-3300(—4000 ?) m.

China (N.W. Yunnan, S.W. Sichuan), N.E. Burma. Map 27.

CHINA. Sichuan: Tachien-lu [Kangding], 1500-2600 m, vii.1908, Wilson 1355a
(BM, EK); ibid., 2100-2500 m, x.1910, Wilson 4338 (K); Gongshan [Kungshan],
22.xi.1959, Feng 24188 (KUN); Tianquan Xian, 23.vii.1953, Tsiang 34993 (SZ).
Yunnan: Tsekou, c. 1904, Monbeig (K); N. end of Lichiang valley, 2700-4150 m,
vi.1906, Forrest 2425 (E, K); E. flank of Tali [Dali] Range, 25°40’'N, 2100-3300 m,
vi.ix.1906, Forrest 4300 (BM, E, K); hills around Tengyueh, 25°N, 1500-2100 m,
viii.1912, Forrest 7743 (BM, E, K); Salwin valley, Yuragan, 1700 m, 22.vii.1938, Yu
23076 (E).

BURMA. Kachin: Ridge of Naung chaung/Nmai Divide, 1200-1800 m,

STUDIES IN THE GENUS HYPERICUM L.

20.vii.1914, Kingdon Ward 1843 (E); Myitkina, 4-8 km from Kangfang, 16.vi.1938,
Naw Mu Pa 17415 (K).

CULTIVATED. Specimens seen from England, Scotland, Ireland, France, Ger-
many, Finland, and the U.S.A.

H. forrestii is closely related to H. beanii, differing from it in having a terete stem,
more ovate, rounded leaves, a more deeply cyathiform flower with rounded sepals,
somewhat relatively shorter stamens and usually relatively shorter styles, and a
broadly ovoid rather than ovoid to ovoid-conic capsule. H. forrestii usually occurs at
much higher altitudes than does H. beanii, and, since Forrest’s plants that were the
original source of cultivated material were collected at c. 3000-4000 m, it has proved
to be quite hardy in northern Europe. H. forrestii therefore replaced H. beanii in
popularity as a garden plant in the period from c. 1920-1950; but it, in turn, was
replaced by H. x ‘Hidcote’.

42x(53x). Hypericum x dummeri N. Robson, hybr. nov.
42. H. forrestii x 14. H. calycinum

Hybrida hortensis e Hyperico forrestii N. Robson et H. calycino L. genita, foliis
oblongo-ovatis ad 50 mm longis subtus pallidis manifeste sed laxe reticulato-venatis,
floribus leviter cyathiformibus in alabastro cinnabarinis coloratis, staminibus quam
petalis 2-5 plo brevioribus antheris aurantiacis, stylis ovario circa 1-2 plo longioribus,
et aliis characteribus inter parentes media. Type: cultivated at Limpsfield, Surrey,
viii. 1982, Robson 1901 (BM!, holotype).

Shrub up to c. 0-7 m tall, lower and with branches more spreading-ascending than in
H. forrestii. Stems red, 4-angled but not ancipitous when young, eventually terete;
internodes 25-35 mm long, shorter than leaves; bark reddish-brown. Leaves
petiolate, with petiole c. 2 mm long; lamina 35-48 x 18-22 mm, oblong-ovate,
rounded to subapiculate, base cuneate, markedly paler or somewhat glaucous
beneath, subcoriaceous; venation: c. 4 pairs of main laterals, with marked but lax
tertiary reticulum; laminar glands dots and short streaks, obscure; ventral glands
dense to sparse. Inflorescence 1-4-flowered, from 1 node, subcorymbiform; pedicels
3-6 mm long; bracts reduced, elliptic. Flowers c. 55 mm in diam., shallowly
cyathiform; buds broadly ovoid, rounded. Sepals 6-9 x 6-8 mm, broadly oblong to
obovate or subcircular, rounded, with margin eroded-denticulate, hyaline and
submarginal zone red; midrib obscure, veins not prominent; laminar glands linear, c.
13. Petals deep golden yellow, tinged orange-red outside, c. 30 x 23 mm, c. 4-5 X
sepals, narrowly obovate, with apiculus lateral, rounded to obsolete; margin minute-
ly eroded-denticulate towards apex. Stamen fascicles each with 25-27 stamens,
12 mm long, 0-4 x petals; anthers deep orange. Ovary 7 X 6mm, ovoid; styles 8 mm
long, c. 1:2 x ovary, free, erect, gradually outcurved; stigmas small. Capsule dark
red when immature. Seeds not seen.

CULTIVATED. England, Herts, North Mymms, 22.vii.1973, D. Walker s.n.
(BM).

H. x dummeri is named after Mr Peter Dummer of Hilliers’ Nurseries (now The
Hillier Arboretum), Ampfield, Hants, who, at my suggestion, made the cross that
resulted in the above hybrid. The cross (Q forrestii x C’ calycinum) was made (c.
1975) in an attempt to re-create H. ‘Hidcote’; but the resultant hybrid is clearly
different from the latter (q.v.). It is nevertheless a striking plant, with the deep green
of the upper surface of the leaf contrasting with the paler reticulate-veined lower
surface; and the deep reddish flower buds and young fruits, together with the orange
anthers and lowish, spreading habit, make it quite distinct from other members of
sect. Ascyreia. Somewhat earlier (1971) another hybridiser, Mr D. Walker of North
Mymms, Hatfield, Herts, made the same cross with similar results. His plants were
completely sterile. A seedling from the Dummer cross growing at Chelsea Physic
Garden has no red tinges on the petals. H. x dummeri thus appears to be variable,
and so I propose to treat the clone described above as a cultivar: c.v. ‘Peter
Dummer’. According to Mr Dummer (in litt. 20.x.1977), the hybrid occasionally
produces variegated shoots (c.f. H. x moserianum and H. x ‘Hidcote’).

287

288 NORMAN K. B. ROBSON
Sect. 4. TAKASAGOYA (Y. Kimura) N. Robson
in Blumea 20: 252 (1973).

Shrubs or shrublets up to c. 1-5 m tall, deciduous, glabrous, without dark glands; branching
lateral. Stems 4-lined and + ancipitous when young, often eventually terete, eglandular; cortex
exfoliating in long scales; bark smooth, scaly. Leaves opposite, decussate, sessile or subsessile,
free, deciduous at basal articulation; lamina entire, with venation pinnate, usually partially or
completely closed, the tertiary reticulum almost invisible or obscure; laminar glands punc-
tiform, pale; marginal gland dots pale; ventral glands absent. Inflorescence 1-3-flowered,
branching dichasial/monochasial from up to 14 nodes sometimes with subsidiary branches;
bracts foliar, bracteoles + reduced. Flowers stellate to cyathiform, homostylous. Sepals 5, free
or united at the base, persistent, with margin entire; veins 5—9; laminar glands + pale, linear to
punctiform; inframarginal and marginal glands absent. Petals 5, deciduous, with apiculus
present, subterminal, sometimes much reduced, or absent, and margin entire; marginal glands
absent; laminar glands + numerous, pale, linear to punctiform. Stamen fascicles 5, free,
deciduous, each with 6-c. 40 stamens; filaments united very shortly; anthers yellow, with gland
amber; pollen type III. Ovary with 5 incompletely or completely axile placentae, »-ovulate;
styles 5, completely united; stigmas completely united, forming capitate to ellipsoid or subglo-
bose mass. Capsule 5-valved, coriaceous, not vittate. Seeds cylindric to clavate, narrowly
carinate to laterally winged, with apical expansions sometimes wing-like; testa linear-reticulate
to scalariform-reticulate.

BASIC CHROMOSOME NUMBER (x): ?; ploidy ?.

Hasirat: stony or rocky slopes or clefts; 0-2400 m.

DIsTRIBUTION: Ryukyu Is. (Uotori I.), Taiwan, Philippines (Luzon).
5 species (+ 1 subspecies).

Sect. Takasagoya appears to be derived from sect. Ascyreia, H. subalatum being most closely
related to H. monogynum form (i) ‘salicifolium’ or the form of that species native (?) in Taiwan.
(See p. 235 and Fig. 4, p. 170.)

Key to sect. 4. Takasagoya

1 Sepals 3-5-10 mm long, mostly lanceolate or narrowly elliptic to linear, acute to subacute (or
rarely rounded); often with distinct miidmtb. (2220.0. iaessaceet cere de aneente speech oeane ee eee 2

Sepals 1-3 mm long, mostly broadly ovate to oblong, obtuse to rounded, without distinct

MAIGIAD: ... 24 « cag dd tee tees eb soeaiet bese Uloare achtde we sides tags oe SHEN a ak Gera sae a so eee ee

2(1) Stems terete when mature; leaves ovate to elliptic or obovate, subacute to rounded...............
Stems 4-angled to 4-winged when mature; leaves usually narrowly elliptic to lanceolate, acute

fo Subacute. 2255, css. ccees enc caceter ene: seer a oace cee sean eee Rare eee eee 5. subalatum(p. 296)

3(2) Sepals obovate to lanceolate, 7-10 mm long, ascending in fruit; petals 10-20 mm long; leaves
OVATE to CLS CC a ie aiaetsn bse oi ieee eet eis ee casa nee eee 1. formosanum(p. 289)

Sepals linear-lanceolate, 3-5-8 mm long, reflexed in fruit; petals 20-28 mm long; leaves elliptic
toobovate ..! kas ssc. eaeemereetee oes aee see ena aRER Ree ieee eee eee 2. nakamurai(p. 289)

4(1) Flowers from up to 14 nodes below apex, substellate; capsule narrowly cylindric or cylindric-
ellipsoid to cylindricscomiciziis/. 228 FI es eee Ee 5
Flowers terminal, solitary, cyathiform; capsule narrowly ovoid .......... 3. senkakuinsulare(p. 292)

5(4) Styles 1-3-2 x as long as ovary; sepals 1-2-5 mm long; branches often spreading and
pendulous). CPt cy wuere eee tee 4a. H. geminiflorum subsp. geminiflorum(p. 294)
Styles about as long as ovary; sepals 2-5—3 mm long; branches erect or ascending.................
Ree er io soon oe tune oa bacwseaueuceodooddede 4b. H. geminiflorum subsp. simplicistylum(p. 295)

STUDIES IN THE GENUS HYPERICUM L.

1(54). Hypericum formosanum Maxim.

in Bull. Acad. Sci. St. Petersb. 27: 428 (1881), in Mél. Biol. 11: 160 (1882); Forbes &
Hemsley in J. Linn. Soc. 23: 73 (1886); H. Léveillé in Bull. Soc. bot. Fr. 53: 499
(1906) pro parte, 54: 590 (1908) pro parte, excl. spec. Faurie.; Matsumura &
Hayata, Enum. pl. Formosa: 41 (1906) pro parte, excl. spec. Faurie. anno 1903;
Hayata, Icon. pl. formos. 1:79 (1911); Matsumura, Index pl. jap. 2: 366 (1912) pro
parte; R. Keller in Engler & Prantl, Nat. Pflanzenfam. 2nd ed. 21: 176 (1925)
(‘formosum’ ); Sasaki, List. pl. Formos.: 295 (1928) pro parte excl. syn.; Makino &
Nemoto, Fi. Jap. 2nd ed.: 748 (1931); Susuki in Masamune, Short fl. Formosa: 141
(1936) (‘formosana’); N. Robson in Blumea 20: 253 (1973), in Liet al., Fl. Taiwan
2: 630 (1976) excl. t. 429. Type: Taiwan. Taipei, ‘prope Tamsuy’ [Tansui], 1864,
Oldham 31 (LE, holotype; BM!, K!).

H. patulum sensu Matsumura & HayatainJ. Coll. Sci. Univ. Tokyo 22: 41 (1906) pro
parte, quoad spec. cit.; Hayata, Icon. Pl. Formos. 1: 79 (1911) pro parte, quoad
loc. Taiwan. cit.; Matsumura, Index Pl. jap. 2: 368 (1912) pro parte, quoad loc.
Taiwan cit.

Takasagoya formosana (Maxim.) Y. Kimura in Bot. Mag. Tokyo 50: 499, ff. 1, 3a
(1936), in Nakai & Honda, Nova fl. jap. 10: 89, ff. 37(1), 38(1), 39(1951).

Icon: Y. Kimura in Nakai & Honda, Nova fl. jap. 10: f. 39 (1951); Fig. 21.

Shrub c. 22 cm tall, with branches spreading, arching. Stem 4-lined and ancipitous
when young, soon 2-lined to terete; internodes 1-5-6 mm long, shorter than leaves;
bark grey-brown. Leaves sessile; lamina 20-60 x 11-29 mm, ovate or elliptic to
oblong-elliptic, subacute to rounded, margin plane, base cuneate to rounded, paler
beneath, not glaucous, subcoriaceous, lower ones tardily deciduous; venation: 1-2
pairs main laterals (the upper forming intramarginal vein) and 7 pinnate midrib
branches, with tertiary reticulum very faint or invisible; laminar gland dots promin-
ent; intramarginal glands dense. Inflorescence 1-3-flowered, terminal, and 1-2-
flowered on subsidiary branches from up to 7(—10) nodes below; pedicel 5-10 mm
long; bracteoles narrowly elliptic, subpersistent. Flowers 25-35 mm in diam., stellate
to very shallowly cyathiform; buds + broadly ovoid, subacute to obtuse. Sepals
7-10 X 1-5-6 mm, very slightly connate, imbricate, subequal to unequal, erect in
bud, ascending in fruit, lanceolate or narrowly elliptic to oblanceolate or elliptic-
ovate, acute to subacute, 5—7-veined with veins reticulately branched, midrib distinct
but narrow; laminar glands linear and punctiform; inframarginal glands scattered,
indistinct. Petals golden yellow, not veined or tinged red, spreading with incurved
tip, (10—-)13-17(—20) x 8-12 mm, c. 2 X sepals, obovate, with apiculus short, Stamen
fascicles each with c. 25-40 stamens, longest 8-12 mm long, c. 0-7 petals. Ovary
3-45 x 3-3-5 mm, ovoid to subglobose; styles (6-S—)7—8(—8-5) mm long, 1:2-2-5 x
ovary, slender; stigmatic mass subglobose. Capsule 8-9 X 6 mm, broadly ovoid.
Seeds 0-1-0-2 mm long, ? linear-reticulate. 2n = ?

On well-drained banks or in stony areas; 0—c. 500 m.
Taiwan (Taipei). Map 28.

TAIWAN. Taipei: Kelung, 13.v.1903, Faurie 114 (BM, E, W); Duan-duan,
18.v.1944, Masamune 4560 (TAI).

The long sepals easily distinguish H. formosanum from all other species in sect.
Takasagoya except H. nakamurai, which has larger flowers and elliptic to obovate
leaves. Its nearest relative in sect. 3. Ascyreia appears to be H. monogynum forma
(iv) ‘ovatum’, from central China.

2(55). Hypericum nakamurai (Masamune) N..Robson

in Blumea 20: 253 (1973), in Li et al., Fl. Taiwan 2: 628 (1976). Type: Taiwan.
Hualien, Karen-gun, Karenky6-ty6, Mt. Seisuizan, c. 2000 m, 9.ix.1939, T.
Nakamura 3734 (TAI!, holotype).

Takasagoya nakamurai Masamune in Trans. Nat. Hist. Soc. Formosa 30: 410 cum
fig. (1940); Y. Kimura in Nakai & Honda, Nova fl. jap. 10: 91 (1951).

Se

289

290 NORMAN K. B. ROBSON

Fig. 21 H. formosanum: (a) habit; (b) leaf section; (c) flower bud; (d) sepal; (e) petal; (f) stamen fascicle;
(g) anthers; (h) capsule (a x 1; b x 4; c-f, h x 6; g x 20). All Oldham s.n.

STUDIES IN THE GENUS HYPERICUM L.

Map 28 Sect. 4. Takasagoya: 1. H. formosanum @, 2. H. nakamurai ©.

Icon: —.

Shrub over 50 cm tall, with branches + spreading. Stem 4-lined and ancipitous when
young, soon terete; internodes 6-35 mm long, shorter than to exceeding leaves; bark
grey-brown. Leaves sessile or subsessile; lamina 10—27(-30) x 5—10(-13) mm, oblong
to elliptic or obovate, acute to rounded, margin plane, base cuneate to angustate,
paler beneath, not glaucous, subcoriaceous, lower ones tardily ? deciduous; vena-
tion: 1 pair main laterals (intramarginal), with c. 6 midrib branches, the tertiary
reticulum almost invisible; laminar gland dots prominent; intramarginal glands
dense. Inflorescence 1-3-flowered, terminal, sometimes with 1-flowered short sub-
sidiary branches from up to 3 nodes below; pedicels c. 10 mm long; bracteoles
narrowly elliptic or elliptic-lanceolate, persistent. Flowers 40-60 mm in diam.,
stellate ?; buds not seen. Sepals 3-5—S(-8) x 0-7—-1-:2(—-2) mm, very slightly connate,
scarcely imbricate, equal, spreading ? in flower, spreading or reflexed in fruit,
linear-lanceolate to linear, subacute to rounded, 5-veined with veins reticulately

291

292

NORMAN K. B. ROBSON

branched, midrib indistinct; laminar glands striiform and punctiform; inframarginal
glands indistinct or absent. Petals bright yellow, not veined or tinged red, spreading
?, 20-28 x 10-20 mm, c. 3-5-6 X sepals, obovate, with apiculus short. Stamen
fascicles each with c. 15 stamens, longest 12-16 mm long, c. 0-6 X petals. Ovary
2-5—3(-5) X c. 1:5 mm, narrowly ovoid-ellipsoid; styles 10-5—12(-15) mm long, c. 4 x
ovary, slender; stigmatic mass capitate. Capsule 8-12 x 4-6 mm, cylindric-ellipsoid.
Seeds dark yellowish-brown, 1-2—1-4 mm long, narrowly cylindric, narrowly carinate
with long apical expansion narrow or sometimes narrowly winged, linear-reticulate.
2n=?

In open limestone rock crevices; 1400-2400 m.

Taiwan (Hualien). Map 28.

TAIWAN. Hualien: Mt. Luanshan to Mt. Patolushan, N.W. of Hualien, 2000-
2200 m, 4.viii.1963, Tamura, Shimizu & Kao 21707 (E); around Mt. Chingshui,
1800-2400 m, i.v.1961, Shimizu 12543 (TAI).

H. nakamuraz is closely related to H. formosanum, but usually differs in having
narrower leaves, smaller, narrower sepals that are reflexed in fruit, larger petals (?
always), relatively longer styles and narrower fruits, as well as a distinct geographical
and altitudinal distribution. However, a specimen of H. formosanum from Mt. Ta
Tung, Taipei Co. (Sasaki s.n., NTU 077118) is somewhat intermediate.

3(56). Hypericum senkakuinsulare Hatusima

inJ. Geobot. 21: 2, cum tab. (1973); Y. Kimura in Asahi hyakka, sekaino shokobutsu
[Weekly Asahi Encycl., Plants of the World] (No. 64): 1511 cum tab. (1977). Type:
Ryukyu Is., Senkaku Group, Uotori I., 360 m, S. Tamaki (URO, holotype).

Icon: Y. Kimura in Asahi hyakki, sekai no shokobutsu (No. 64): 1511 (1977).

Shrub c. 0.6 m tall, with branches spreading. Stem soon terete; internodes 10-20 mm
long, shorter than leaves; bark reddish-brown. Leaves sessile; lamina 23-35 x 10-18
mm, elliptic or oblong-elliptic to obovate, rounded, margin plane, base cuneate to
cuneate-angustate, not glaucous, chartaceous, deciduous; venation: 1 pair main
laterals (intramarginal) and c. S—6 pinnate midrib branches, without visible tertiary
reticulum; laminar gland dots numerous; intramarginal glands. Inflorescence 1-
flowered, terminal, without subsidiary branches; pedicels c. 10 mm long; bracteoles
elliptic. Flowers c. 40 mm in diam., cyathiform; buds narrowly ovoid, subacute.
Sepals c. 3 xX 2mm, very slightly connate ?, not or scarcely imbricate, subequal, erect
in bud, broadly oblong, obtuse; laminar glands linear. Petals bright yellow, not
veined or tinged red, incurving, c. 25 x 18 mm, c. 6 X sepals, obovate, very short,
apiculus rounded. Stamen fascicles each with c. 15 stamens, longest c. 15 mm long, c.
0-7 X petals. Ovary with styles 9 mm long, slender. Capsule c. 10 mm long, narrowly
ovoid. Seeds not recorded. 2n = ?

Near rocky summit of mountain; 360 m.

Ryukyu Islands (Senkaku Group). Map 30.
RYUKYUIS. Senkaku Group: Uotori I. (not seen).

H. senkakuinsulare is known from only the type locality. Itis apparently related to H.
formosanum, the geographically nearest species, but differs in its consistently
rounded leaves, its solitary cyathiform flowers with smaller sepals, and (?) its
narrower capsules.

4(57). Hypericum geminiflorum Hemsley

in Ann. Bot. 9: 44 (1895); Henry in Trans. Asiat. Soc. Japan 24, suppl. 1: 19 (1896);
Matsumura & Hayata, Enum. pl. Formosa: 41 (1906); H. Léveillé in Bull. Soc.
bot. Fr. 54: 590 (1908); Hayata, Icon. pl. formos. 1: 76 (1911), 3: 41 (1913);
Matsumura, Index Pl. Jap. 2: 366 (1912); Makino & Nemoto, FI. Japan ed. 2: 748
(1931); N. Robson in Blumea 20: 253 (1973), in Fl. Malesiana I, 8: 19 (1974), in Li

STUDIES IN THE GENUS HYPERICUM L.

et al., Fl. Taiwan 2: 630 (1976). Type: Taiwan, Prov. Takao [Kaohsiung], Apes’
Hill, Henry 1155 (K!, holotype).

Shrub 0-3-1-5 m tall, lax, with branches erect to ascending or pendulous. Stem
4-lined and ancipitous when young, eventually 2-lined to terete; internodes 9-35 mm
long, shorter than leaves. Leaves subsessile; lamina 18-45 x 6-22 mm, oblong or
oblong-lanceolate to elliptic or ovate, acute to obtuse or rounded-apiculate, margin
incrassate, base broadly to narrowly cuneate, paler beneath, not glaucous, char-
taceous, lower ones tardily deciduous; venation: 1-2 pairs main laterals (the upper
sometimes forming distinct intramarginal vein) and c. 4-9 pinnate midrib branches,
with tertiary reticulum faint or invisible; laminar gland dots sometimes prominent
above; intramarginal glands dense. Inflorescence 1-(2-3) flowered, terminal and on
solitary or paired short subsidiary branches from up to 14 nodes below: pedicels 3—
4 mm long; bracteoles reduced, deciduous. Flowers 20-30 mm in diam., stellate;
buds ovoid, acute to obtuse. Sepals (1-)1-5-3-5 x 1-2-5 mm, free or slightly connate,

293

Map 29 Sect. 4. Takasagoya: 4a. H. geminiflorum subsp. geminiflorum @ (part; see also Map 30), 4b. H.

geminiflorum subsp. simplicistylum @, 5. H. subalatum *.

294

NORMAN K. B. ROBSON

imbricate, equal or subequal, broadly ovate to oblong or subcircular, subacute to
rounded, 1—5-veined with veins unbranched, midrib not distinct; laminar glands
mostly linear. Petals bright yellow or rarely white, not veined or tinged red,
spreading, 9-15 x 5-7 mm, 4-9 x sepals, obovate, apiculus obsolete or absent.
Stamen fascicles each with 5-11 stamens, longest 6-10 mm long, c. 0-65 x petals.
Ovary 2-5-5 X 1-5—2-5 mm, + narrowly ellipsoid; styles 3-5—6(—7) mm long, 1-2 x
ovary, + slender; stigmatic mass capitate to ellipsoid or cylindric. Capsule 5-11 x
3-5-5 mm, narrowly cylindric to cylindric-ellipsoid or narrowly cylindric-conic. Seeds
dark reddish-brown, 0-6—1-:5 mm long, narrowly cylindric or narrowly ellipsoid to
fusiform, narrowly carinate with + long apical expansions or unilaterally + deeply
winged, very shallowly linear-reticulate. 2n = ?

Open or exposed stony ground; 1000-1800 m.
Taiwan, Philippines (Luzon). Maps 29, 30.

H. geminiflorum comprises two subspecies (treated as varieties by Robson, 1973,
supra cit.), one found at relatively low altitudes in south-east Taiwan and Luzon
(subsp. geminiflorum), the other at higher altitudes in central and north-west Taiwan
(subsp. simplicistylum).

4a(57a). H. geminiflorum subsp. geminiflorum

H. trinervium Hemsley in Ann. Bot. 9: 144 (1895); Henry in Trans. Asiat. Soc. Japan
24: suppl. 1: 19 (1896); Matsumura & Hayata, Enum. pl. Formosa: 43 (1906); H.

Map 30 Sect. 4. Takasagoya: 3. H. senkakuinsulare ©, 4a. H. geminiflorum subsp.
geminiflorum @ (see also Map 29).

STUDIES IN THE GENUS HYPERICUM L.

Léveillé in Bull. Soc. bot. Fr. 54: 590 (1908); Hayata, Icon. pl. formos. 1: 79
(1911); Matsumura, Index pl. jap. 2: 369 (1912); Sasaki, List pl. Formosa: 296
(1928). Type: Taiwan, Prov. Takao [Kaohsiung], South Cape, Schmiirer in Henry
906 (K!, lectotype—mihi), 906A (K!, syntype).

H. loheri Merrill in Philipp. J. Sci. C, 4: 294 (1909). Type: Philippines, Luzon.
Benguet Prov., Ambuklao [Ambuelo (US)], Loher 66 (K!, lectotype—mihi; US!);
Benguet Prov. Mt. Ugo, xii.1908, Ramos in PBS 5716 (US! syntype); Zambales
[Pampanga] Prov., Mt. Pinatubo, iv.1907, Foxworthy in PBS 2562 (BO!, US!,
syntypes).

H. acutisepalum Hayata in J. Coll. Sci. Tokyo 30 (1): 38 (1911), Icon. pl. formos. 1:
77, t. 15 (1911); Ito, Ill. formos. pls: 363 (1927); Makino & Nemoto, Fl. Japan 2nd
ed.: 747 (1931); Liu, U/l. nat. introd. lign. pl. Taiwan 1: 306, f. 254 (1960); Iconogr.
cormophyt. sin. 2: 874, f. 3478 (1972). Type: Taiwan: Prov. Taitya [Taitung],
Nant6, vii.1907, Kawakami 3245 (TI!, holotype).

H. formosanum sensu Hayata, Suppl. icon. pl. formos. 6: 7 (1916); N. Robson in Li
et al., Fl. Taiwan 2: 63D (1976) pro parte, quoad t. 429.

2H. pustulosum R. Keller in Engler & Prantl, Nat. Pflanzenfam. 2nd ed. 21: 176
(1925), nomen.

H. lackeyi Elmer in Leafl. Philipp. bot. 9: 3190 (1934). Type: Philippines, Luzon.
Pampanga Prov., Zambales Mts, Mt. Pinatubo, Camp Stotsenburg, Elmer 21989
(PNH!, holotype; BM!, BO!, GH!, K!, L!, PNH!, SING!, Z!).

Takasagoya acutisepala (Hayata) Y. Kimura in Bot. Mag., Tokyo 50: 501, f. 3j-o
(1936), in Nakai & Honda, Nova fl. jap. 10: 92, ff. 37-2, 38-3, (1951).

T. geminiflora (Hemsley) Y. Kimura in Bot. Mag. Tokyo 50: 501 ff. 2, 4 j—n (1936), in
Nakai & Honda, Nova fl. jap. 10: 92, ff. 37-5, 38-4, 40 (1951).

T. trinervia (Hemsley) Y. Kimura in Bot. Mag., Tokyo 50: 503, (1936), in Nakai &
Honda, Nova fl. jap. 10: 95, f. 38-6 (1951).

Icones: N. Robson in Lietal., Fl. Taiwan 2: 631, t. 429 (1976); Ho, Trop. pls. Taiwan
in colour 3: 114, t. 1 (1982).

Shrub 0-5—1-5 m tall, with branches often spreading and pendulous. Sepals 1-2:5 mm
long, broadly ovate or triangular-ovate or subcircular to oblong-lanceolate. Ovary
2-5-3-5 mm long; styles 4-6(—7) mm long, 1-3-2 x ovary. Capsule narrowly cylindric
to narrowly cylindric-conic.

Open stony ground; 300-1200 m (Taiwan), 1000-1500 m (Luzon).

Taiwan (Kaohsiung, Pintung, Taitung, Hualien, Ilan), Philippines (Luzon). Maps
29, 30.

TAIWAN. Kaohsiung: Apes’ Hill, pre v.1894, Henry 1155 (K). Pintung:
Kaoshihfo [Kusukusu], 28.vi.1912, Hayata & Sasaki (TI). Taitung: Taichung,
Sung-mao, 15.x.1970, Kuoh, Chuma & Chang 8761 (H). Hualien: from Patakang to
Tienhsiang, 28.xii.1960, Shimizu & Kao 10644 (TAI). Ilan; vicinity of Nan-shan,
1200 m, 26.ix.1964, Koyama 24056 (E).

PHILIPPINES. Luzon. Cagayan Prov.: Penablanca, iti-v.1929, Ramos PBS
76892 (K.) Benguet Prov.: Mangayan to Baguio, x.1921, Ramos & Edano PBS 40501
(A, BO, K, US). Pampanga Prov.: Zambales Mts, Mt. Pinatubo, Camp Stotsen-
burg, v.1927, Elmer 21989 (BM, BO, GH, K, L, PNH, SING, Z).

H. trinervium, at first glance, seems rather distinct from H. geminiflorum in its
stouter shoots and pedicels and relatively broad leaves with a strong submarginal
vein. These characters, however, occur sporadically in other individuals of H.
geminiflorum, and so H. trinervium appears to be no more than a local race from the
extreme south of Taiwan.

Forms with white flowers have been recorded in both Taiwan (Kawakami,
Shimada & Ito 28.iv.1917) and the Philippines (cf. description of H. loheri).

4b(57b). H. geminiflorum subsp. simplicistylum (Hayata) N. Robson, stat.
nov.

H. simplicistylum Hayata in J. Coll. Sci. Tokyo 30 (1): 40 (1911), Icon. pl. formos. 1:

SHER eB

295

296

QO HES <>

NORMAN K. B. ROBSON

79, t. 16 (1911); Makino & Nemoto, Fl. Japan 2nd ed.: 752 (1931); Suzuki in
Masamune, Short fi. Formosa: 141 (1936). Type: Taiwan. N6k6ézan, ad 1800 m
alt., vi.1908, Kawakami & Mori 4507 (TI!, holotype).

H. formosanum sensu Suzuki in Ann. Rep. bot. Taihoku Gdns 1: 158 (1931) pro
parte, quoad spec. Suzuki 6375.

Takasagoya simplicistyla (Hayata) Y. Kimura in Bot. Mag., Tokyo 50: 502, ff. 3 r-s,
4 fH, in Nakai & Honda, Nova fl. jap. 10: 93, ff. 37-4, 38-5 (1951).

H. geminiflorum var. simplicistylum (Hayata) N. Robson in Blumea 20: 254 (1973),
in Liet al., Fl. Taiwan 2: 632 (1976).

Icon: Hayata, Icon. pl. formos. 1:79, t. 16 (1911).

Shrub c. 0-3-0-4 m tall, with branches erect or ascending. Sepals 2-5-3 mm long,
broadly ovate to oblong. Ovary 44-3 mm long; styles 3-5—5 mm long, 1-1-3(-1-5) x
ovary. Capsule cylindric to cylindric-ellipsoid.

Exposed stony ground (mountain slopes, roadsides); 1500-1800 m.

Taiwan (Taipei, Ilan, Taichung, Nantou, Chiayi). Map 29.

TAIWAN. Taipei: Raisya, 1.vii.1916, Matuda 1333 (TAI). Ilan: Inter Kinyan et
Piyahau, 24.ix.1930, Suzuki 6375 (TAI). Taichung: Lakulaku, 14.xii.1923, Sasaki
s.n. (TAI). Nantou: Nengkaoshan, Tung-yen to Yin-hai, 11.viii.1971, Huang, Hsien
& Kao 5677 (TAI). Chiayi: Nitakayama. viii.1935, Sasakis.n. (TAI).

Subsp. simplicistylum is apparently an upland derivative of subsp. geminiflorum.
Although it tends to be smaller in all parts and more erect, there is a certain overlap
in character variation between it and the more lowland taxon that prevents their
recognition as separate species. The relative style length and fruit shape, however,
are usually diagnostic. This variation, when considered along with the distinct
geographical and altitudinal distributions of the two taxa, suggests that their
appropriate rank is subspecies rather than variety.

5(58). Hypericum subalatum Hayata

in J. Coll. Sci. Univ. Tokyo 30: 41 (1911), Icon. pl. formos. 1: 77 (1911); Makino &
Nemoto, Fl. Jap. 2nd ed.: 752 (1931); Suzukiin Masamune, Shortfi. Formosa: 141
(1936); N. Robson in Blumea 20: 253 (1973), in Liet al., Fl. Taiwan 2: 643 (1976).
Type: Taiwan, Prov. Taihoku [Taipei], Kussyaku, secus rivulos, 8.vi.1903, Faurie
115 (TI!, holotype; BM!, E!, W!).

H. formosanum sensu H. Léveillé in Bull. Soc. bot. Fr. 53: 499 (1906) pro parte;
Matsumura & Hayata, Enum. pl. Formosa: 41 (1906) pro parte; H. Léveillé, op.
cit. 54: 590 (1908) pro parte, omnes quoad Faurie 115.

H. kushakuense R. Keller in Bot. Jahrb. 58: 191 (1923). Type: Taiwan, Kushaku,
Faurie 115 (B+, holotype; BM!, E!, TI!, W!).

Takasagoya subalata (Hayata) Y. Kimura in Bot. Mag., Tokyo 50: 500, f. 3 p-q
(1936), in Nakai & Honda, Nova fl. jap. 10: 91, ff. 37 (6), 38 (2) (1951).

Icon: —.

Shrub over 50 cm tall, with branches + ascending. Stem persistently 4-angled to
narrowly 4-winged; internodes 10-25 mm long, shorter than leaves; bark reddish-
brown. Leaves sessile; lamina 19-70 x 5-16 mm, narrowly elliptic to lanceolate or
narrowly oblong-elliptic, acute to subacute or rounded-apiculate, margin plane, base
cuneate, paler beneath, not glaucous, chartaceous, lower ones tardily deciduous;
venation: 2-3 pairs main laterals (sometimes forming distinct intramarginal vein)
and 6-7 pinnate midrib branches, with tertiary reticulum faint or almost invisible;
laminar gland dots prominent; intramarginal glands dense. Inflorescence 1-flowered,
terminal and on short subsidiary branches from up to c. 6 nodes below; pedicel 6—
12 mm long; bracteoles foliar to subulate, deciduous. Flowers c. 25 mm in diam.,
substellate; buds ovoid, acute. Sepals 5-8 x 1-2(-3) mm, free or almost so,
imbricate, subequal to unequal, erect in bud and fruit, oblong-linear to elliptic or
oblanceolate, subacute to acute or shortly acuminate, 7—9-veined with veins little
branched, midrib + obscure; laminar glands mostly linear. Petals bright yellow,

STUDIES IN THE GENUS HYPERICUM L. 297

tinged red, + spreading, c. 10-12 x 6 mm, 1-5-2 X sepals, obovate, with apiculus
obsolete. Stamen fascicles each with c. 15 stamens, longest c. 9-10 mm long, c. 0-8 x
petals. Ovary 2-4 x c. 0-7-1-3 mm, + narrowly ovoid; styles (3-3—)4~7 mm long, c.
2-2-5 x ovary, slender; stigmatic mass capitate. Capsule 7-9 x 3-4 mm, narrowly
ovoid to cylindric. Seeds dark yellowish-brown, 0-9-1-1 mm long, narrowly cylindric
or narrowly ovoid-cylindric, scarcely carinate with long narrow apical expansion,
very shallowly linear-reticulate. 2n = ?.

In open limestone rock crevices; 400-900 m.

Taiwan (Taipei, Ilan, Hualien). Map 29.

TAIWAN. Taipei: Hsintien, i.1914, Faurie 1067 (BM, G). Ilan: Heirinbi district,
Kueishan I., 28.iii.1912, Price 294 (K). Hualien: Shuiyan to Luanshan, NW. of
Hualien, 700-900 m, 21.viii.1963, Tamura, Shimizu & Kao 21508 (E).

The 4-angled mature stem distinguishes H. subalatum from all other species in sect.
Takasagoya.

Sect. 5. ANDROSAEMUM (Duhamel) Godron
in Grenier & Godron, Fl. France 1: 320 (1847).

Shrubs 0-3-2 m tall, deciduous, glabrous, without dark glands; branching lateral. Stems 2—4-lined
(subfoliar lines less prominent or absent) and + ancipitous when young, eventually terete, eglandular;
cortex exfoliating in strips; bark smooth, scaling or fissured. Leaves opposite, decussate, sessile to shortly
petiolate, free, deciduous at basal articulation; lamina entire, with venation pinnate, partially (or rarely
completely ?) closed, the tertiary densely reticulate; laminar glands punctiform, pale; marginal gland dots
pale; ventral glands absent. Inflorescence (1—)3-c. 20-flowered, branching dichasial to monochasial from
1-2 nodes, sometimes with subsidiary branches; bracts and bracteoles foliar (persistent) or reduced
(deciduous), not sepaline. Flowers stellate, homostylous. Sepals 5, free, persistent or deciduous, with
margin entire; veins 5—11; laminar glands pale, linear to punctiform; inframarginal glands pale. Petals 5,
deciduous, with apiculus present, subterminal, small, or absent, and margin entire; marginal glands
absent; laminar glands + numerous, pale, linear, distally interrupted. Stamen fascicles 5, free, deciduous,
each with 20-40 stamens; filaments united very shortly; anther gland amber; pollen types IIJ-IV. Ovary
with 3 (abnormally 5) incompletely axile placentae (parietal in upper c. 2/3), »-ovulate; styles 3
(abnormally 5), free; stigma subcapitate to narrowly capitate. Capsule 3 (abnormally 5)-valved, coriaceous
to chartaceous and completely to partially dehiscent or + thin-walled, fleshy and tardily dehiscent or
indehiscent, narrowly vittate. Seeds cylindric, broadly carinate to partially or completely unilaterally
winged, with apical expansions sometimes wing-like; testa linear-reticulate.

BASIC CHROMOSOME NUMBER (x): 10; ploidy 4.
Hasirat: evergreen forest and dry open areas (Sp. 1); damp, shaded places (other species); c. 200-2100 m.

DISTRIBUTION: Macaronesia, NW. Africa, W. & S. Europe, W. Mediterranean islands, Crete, S. Aegean
islands, Turkey, Transcaucasia, N. Iran, Turkmenistan, Cyprus, Levant south to N. Israel, SW. Saudi
Arabia.

4 species (+ 1 hybrid).

Key to sect. 5. Androsaemum

1__ Sepals shrivelling and deciduous before fruit ripens; styles 3-5 x ovary; leaves (when crushed)
ISH ALLY SOE E OF SOAGS (4c MIL CHBVEN) soa o< wien 500 weno Woe we oslewisouivcuveddeusided use ooseveenseencaeceess 2

Sepals persistent at least until fruit ripens; styles 0-5-3 x ovary; leaves (when crushed) very

Ett (ie SAM ALL )ISMICMU SOL POA A. ue. Jathngdkas. capeneicksaataeaesbapeasedisccsseoeeuaeeeenecs

2(1) Styles 10-13 mm long; flowers 20-30 mm in diam. (petals 10-15 mm long); leaves lanceolate to

thianoular-lanceolatesae.-ecereee ei aee ssc cse ete ce recess: 4b. hircinum subsp. cambessedesii(p. 311)
Styles 14-24 mm long or, if 12-13 mm long, then flowers 30-40 mm in diam. (petals 15-21 mm
[OT) I SAVES\V ARO US seee mn eene ere ter er en enn ch ose Re oI Oe ENS acd ce asad wha beteowa S4ciator 3

3(2) Capsule (8)9-14 mm long; leaves narrowly ovate to narrowly lanceolate or triangular-
FANICE OLAGe MIOSHN ACUI CO G@OLUSE! fas ends. cte suecdosesos-ndetatnunadendecbecauduedceenessecesesansonsaee
Capsule 6—8(9) mm long; leaves broadly ovate to ovate-lanceolate, mostly obtuse to rounded.. 5

298 NORMAN K. B. ROBSON

4(3) Branches erect; leaves goat-scented when crushed................. 4a. hircinum subsp. majus(p.
Branches spreading or pendent; leaves not goat-scented ........ 4d. hircinum subsp. metroi(p.

5(3) Petals 15-18 mm long; leaves with apex obtuse to rounded, base rounded to truncate, margin
|0) C1 at Sean nemrca hie AMA SEasBenAore Suamtanes ance ee se oedecaeyae te 4c. hircinum subsp. hircinum(p.

Petals 18-20 mm long; leaves with apex apiculate or rounded, base cordate to rounded or
shortly angustate, margin usually + crisped-undulate 4e. hircinum subsp. albimontanum(p.

6(1) Styles c. 0-5 x ovary; petals equalling or shorter than sepals; ripe fruit fleshy, dark-brown to
black; lucent: indehiscentinec..-- seotascerenaerereerer ee name sceeh sercrr eer. 3. androsaemum(p.
Styles 1-3 x as long as ovary; petals 1-5—4 x sepals; ripe fruit dry, not usually lucent, dehiscent

7(6) Petals c. 1-5-2 X sepals; fruit coloured bright red during maturation, dehiscing tardily and
incompletely; outensepals = broadly OVatC ho .c.c.c- ocean senna ee te 3x. x inodorum(p.

Petals 34 X sepals; fruit not coloured bright red during maturation, dehiscing completely;

outer sepals + narrowly oblong or = narrowly elliptic... >. cn0..2seutsesd- secede aes eee eee

8(7) Inflorescence branches widely spreading; capsule coriaceous, ovoid-ellipsoid to ellipsoid;
leaves broadly triangular-ovate to oblong-ovate; petals (16)18-22 mm long 1. grandifolium(p.

Inflorescence branches + narrowly ascending; capsule chartaceous, + broadly cylindric-
ellipsoid; leaves + narrowly ovate to triangular-lanceolate; petals 10-18mm_ 2. foliosum(p.

1(59). Hypericum grandifolium Choisy

Prodr. monogr. fam. Hypéric.: 38, t. 3 (1821), in DC., Prodr. syst. nat. regni veg. 1:
544 (1824); Cham. & Schlecht. in Linnaea 3: 120 (1828); Lowe, Fl. Madeira: 75
(1868); Menezes, FI. archip. Madeira: 28 (1914); R. Keller in Engler & Prantl,
Nat. Pflanzenfam. 2nd ed. 21: 177 (1925); Stefanoff in God. Agr.-les. Fak. Univ.
Soflya 11: 144 (1933), 12: 82 (1934); in Pflanzenreale 4: Karte 2a (1933); Schaeffer,
PI. Canary Islands: 144 + photo (1963), Pfl. Kanar. Inseln, 2nd ed.: 148 + photo
(1967); Kunkel, G., Endemismos Canarios: 295 (1977); Kunkel, G. & M. A., Fi.
Gran Canaria 4: 42, t. 164 (1979). Type: Canary Islands. ‘In insula Tenerife’, 1806,
Chr. Smith (G-DC!, lectotype—mihi); ‘in insula Tenerife’, 1807, Broussonet

(G-DC!, syntype).

H. erectum Solander ex R. Br. [in Buch] in Abhandl. Physik. Kl. K6nigl. Preuss.
Akad. Wiss. Berlin 1816-17: 75 (1819) in synon., non Thunb. ex Murray (1784).
H. canariense Brouss. ex Webb & Berth., Phyt. Canar. 1: 51 (1836) in synon, non L.

(1753).

Androsaemum webbianum Spach, Hist. nat. vég. Phan. 5: 418 (1836), in Ann. Sct.

nat. (Bot.) II, 5: 362 (1836); Webb & Berth., Phyt. Canar. 1: 51 (1836), ibid.,

t.4E

(1841); Pitard & Proust, Les Iles Canaries: 134 (1909). Type as for Hypericum

grandifolium Choisy.

H. grandiflorum sensu Ceballos & Ortuna, Veg. Fl. for. Canar. occid.: 388 t. 158

(1951) et auct., non Salisb. (1798) nec. hort.

H. elatum sensu Plaisted & Lighty in Nat. Hort. Mag. 38: 131 (1959); Lid in Skr.

Norske Videns. -Akad. Oslo, Math. -Nat. Kl. I1, 23: 120 (1967) et auct. plur.
Icon: Kunkel, G. & M. A., Fl. Gran Canaria 4: t. 164 (1979).

Shrub 0-5—1-8(—2?) m tall, bushy, with branches erect or ascending. Stems 4-lined
and ancipitous when young, soon 2-lined, eventually terete; internodes 24 mm long,
shorter than leaves; bark scaling. Leaves sessile, + amplexicaul; lamina (30—)40—
90 x (20-)25—45 mm, broadly triangular-ovate to oblong-ovate, obtuse (or rarely
acute) to rounded, margin plane, base cordate to rounded, not paler beneath, not
glaucous, papyraceous; venation: 4-5 pairs main laterals, with tertiary reticulum
prominent above; laminar glands small; intramarginal glands dense. Inflorescence
1-13-flowered, from 1—2 nodes with internode not condensed, broadly pyramidal to
subcorymbiform with branches widely spreading, occasionally with accessory flow-

Note: The petals in text figures appearing on pp. 298-309 are approximately natural size and

the leaves half natural size.

STUDIES IN THE GENUS HYPERICUM L.

ers, sometimes with flowering branches from up to 4 nodes below, the whole then
cylindric; pedicels S—11 mm long; bracts foliar or reduced and linear to linear-
subulate. Flowers (25—)30-45 mm in diam.; buds ellipsoid-subglobose, obtuse to
rounded. Sepals 6-8 x 2-3 mm, imbricate, unequal to subequal, spreading to
reflexed and enlarging after anthesis, erect and persistent in fruit, lanceolate to
narrowly oblong, acute to subacute, 7—9-veined from base with veins + branched and
reticulate, midrib distinct; laminar glands linear and punctiform; inframarginal
glands dense. Petals golden yellow, not tinged red, spreading, (16—)18-22 x (4-)6-
8 mm, 3-4 X sepals, narrowly oblong-lanceolate, with apiculus obtuse or absent.
Stamen fascicles each with (25—)30-40 stamens, longest 15-20 mm long, c. 0-9 x
petals. Ovary 4-S(-6) X (2-5—)3-3-5 mm, narrowly ovoid to ovoid-ellipsoid, acute;
styles (rarely 5 in number) (7-5—)9-S—13(-17) mm, 2-3 x ovary, erect, narrowly
divergent towards apex, slender; stigma narrowly capitate. Capsule 8-13 x 6-8 mm,
ovoid-ellipsoid to ellipsoid, acute, coriaceous, completely dehiscent. Seeds yellow-
ish-brown, c. 1-2 mm long, broadly carinate or with narrow unilateral wing, almost
without terminal appendages. 2n = 40.

In evergreen Laurus forest and on dry stony hillsides and cliffs; 200-1500 m
(Canaries), 400-500 m (Madeira).

Canary Islands (all islands), Madeira. Map 31.

CANARY IS. La Palma: Brena Alta to El Paso road, E. end of Cumbre tunnel,
1000 m, 21.iv.1977, Jarvis, Gibby & Humphries 411 (BM). Gomera: Monte del
Cedre, top of Barranco del Hermingua, 900 m, 8.v.1977, Jarvis 605 (BM). Tenerife:
Monte de la Mercedes, Mirador, 4km W. of Pico del Ingles, 800 m, 29.iv.1977, Jarvis
451 (BM). Gran Canaria: Los Lagunetas, 26.vi.1931, Frey & Stord s.n. (H). Also in

299

Map 31 Sect. 5. Androsaemum: 1. H. grandifolium yx, * (records), 4a. H. hircinum subsp. majus @, O
(naturalised), 4b. H. hircinum subsp. cambessedesii Wi, 4c. H. hircinum subsp. hircinum A, 4d. H.
hircinum subsp. metroi #, 4e. H. hircinum subsp. albimontanum D.

300

NORMAN K. B. ROBSON

Hierro, Fuerteventura and Lanzarote, fide Hansen & Sunding, Fl. Macaron.
Checkl.: 51 (1979).

MADEIRA. Sao Vicente valley, 28.vii.1954, Moore s.n. (BM); near Curral das
Freiras, 12.vii.1974, McClintock s.n. (BM); Queimadas, 880 m, 26.vi.1957, Malberg
sm(H):

H. grandifolium does not show any marked geographical variation. Among the
species of sect. Androsaemum with persistent sepals, it can be distinguished by its
relatively lax inflorescence, relatively large flowers and completely dehiscent cap-
sules. It has frequently been confused with H. elatum Aiton (= 3x. H. x inodorum
Miller, q.v.).

2(60). Hypericum foliosum Aiton

Hort. kew. 3: 104 (1789); Choisy, Prodr. monogr. fam. Hypéric.: 46 (1821), in DC.,
Prodr. syst. nat. veg. 1: 544 (1824); Watson in Godman, Nat. hist. Azores or W.
Islands: 142 (1870), in Trelease in Anns. Rep. Mo. botl. Gdn 8: 99 (1897); R.
Keller in Engler & Prantl, Nat. Pflanzenfam. 2nd ed. 21: 177, f. 73M, N (1925);
Stefanoff in God. Agr.-les. Fak. Univ. Sofiya 11: 144 (1935), 12: 82 (1934), in
Pflanzenareale 4; karte 2a (1933); Palhinha, Cat. pl. vasc. Agores: 74 (1966); N.
Robson in Flora europaea 2: 263 (1968); non H. foliosum sensu Jacq., Hort.
Schoenbrun. 3: t. 299 (1798) (= H. prolificum L.) nec Brouss. ex Webb & Berth.,
Phyt. Canar. 1: 45 (1836) in synon. (= H. reflexum L.f.). Type: Azores. San
Miguel, 1777, Masson (BM!, holotype).

Icon: Trelease in Ann. Rep. Mo. botl Gdn 8: t. 19 (1897).

Shrub 0-5-1 m tall (or taller ?), bushy (?) with branches erect or + spreading. Stems
4-lined and ancipitous when young, soon 2-lined, eventually terete; internodes 1—
3 mm long, shorter than leaves; bark fissuring. Leaves sessile or subsessile, some-
times amplexicaul; lamina 35-60 x 10-32 mm, + narrowly ovate to triangular-
lanceolate, obtuse (or rarely acute) to rounded, margin plane, base rounded or more
rarely cordate, slightly paler beneath, not glaucous, papyraceous; venation: 4-5
pairs ascending main lateral veins, with tertiary reticulum prominent on both sides;
laminar glands small; intramarginal glands dense. Inflorescence 1—-9-flowered, from
1—2 nodes with condensed internode, subcorymbiform to pseudo-umbellate with
branches + narrowly ascending, without accessory flowers, sometimes with 1-3-
flowered branches from node below; pedicels 7-12 mm long; bracts reduced,
linear-lanceolate to linear-subulate. Flowers (20—)25—30(-35) mm in diam.; buds
ellipsoid-subglobose, rounded. Sepals (2—)3-6(-7) x 1-5—2-5(-4) mm, imbricate,
unequal to subequal, spreading to subdeflexed and enlarging after anthesis, deflexed
and persistent in fruit, triangular-lanceolate or lanceolate to oblong-elliptic, acute to
obtuse, 5—7-veined from base with veins sometimes + branched and reticulate;
laminar glands linear and punctiform; inframarginal glands dense. Petals golden
yellow, not red-tinged, spreading, (10-)12-18 x 5-8 mm, 34 x sepals, oblan-
ceolate, with apiculus rounded or obsolete. Stamen fascicles each with 20-30
stamens, longest 12-18 mm, equalling or slightly exceeding petals. Ovary 4-5(-6) x
3-4-5(-5) mm, broadly ovoid to subglobose, acute to obtuse; styles 5—10 mm long,
1-5-2-5 X ovary, erect to suberect, narrowly divergent near apex, slender; stigmas
narrowly capitate. Capsule 8-13 x 7-10 mm, + broadly cylindric-ellipsoid, acute to
obtuse, chartaceous, somewhat fleshy at first, soon drying, tardily and sometimes
incompletely dehiscent. Seeds yellowish-brown, 1-2-1-5 mm long, unilaterally
winged with terminal winged appendages. 2n = ?

In damp shaded placed in the mountains, 220-800 m.

Azores (all islands). Map 32.

AZORES. Santa Maria: Feiteiras, 220 m, 1.v.1972, Goncalves 3909 (BM). Sao
Miguel: no precise locality, v.1956, Pickering 126 (BM). Terceira: Santa Barbara,
Caminho Flora, 900 m, 12.vii.1971, Goncalves 3236 (BM). Sao Jorge: Santo
Antonio, 520 m, 4.ix.1971, Goncalves 3472 (BM). Pico: Landroal, 800 m,
27.vii.1971, Goncalves 3304 (BM). Graciosa: No precise locality, 1868 ?, Drouet

STUDIES IN THE GENUS HYPERICUM L.

(BM). Faial: no precise locality, vii.1842, Watson s.n. (E, K). Flores: Ribeira
Barqueiros, 5.vi.1938, Stord s.n. (H). Corvo: Estrada do Farol, 250 m, 9.vi.1971,
Goncalves 2645 (BM).

H. foliosum is intermediate in most respects between H. grandifolium and H.
androsaemum except for the leaves, which are frequently narrower than those of
both these species. It differs from H. grandifolium in its more condensed inflores-
cence with smaller flowers, usually shorter styles and thinner capsule valves.

3 (61). Hypericum androsaemum L.

Sp. pl.: 784 (1753); Reichenb., Ic. fl. german. 6: t. 352 (1844), Syme, Eng. bot. 3rd
ed.: t. 264 (1865); Boiss., Fl. orient. 1: 788 (1867); Willk. & Lange, Prodr. fi. hisp.
3: 588 (1878); Fiori, Nuova FI. ital. 1: 520 (1923); Hayek, Prodr. fl. pen. balc. 1:
530 (1925); R. Keller in Engler & Prantl, Nat. Pflanzenfam. 2nd ed. 21: 177 (1925);
Stefanoff in God. Agr.-les. Fak. Univ. Sofiya, 10: r.1f.2,t.2f. 4, t. 3 f. 4 (1932),
11: 141 (1933), 12: 81 (1934), in Pflanzenareale 4: karte 1b (1933); Guinea Lopez,
Viscaya y su Paisaje vegetale: 215 X map (1949); Gorshkova in Fl. U.R.S.S. 15:
214, t. 9 f. 4 (1949); Ross-Craig, Drawings Br. Pls 6: t. 6 (1952); Grossheim, FI.
Kavkasa 2nd ed. 6: 166, t. 30 f. 2, karte 181 (1962); N. Robson in Fl. Turkey 2: 366,
ff. 11, 13, map 21 (1967), in Fl. europaea 2: 263 (1968), in Fl. iranica 49: 5 (1968), in
Green, Plants wild & cult.: 80 (1973); Jordanov & Kozhukharov in Jordanov, FI.
R.P. Bulg. 4: 230, t. 43 f. 1 (1970); Franco, Fl. Portugal: 449 (1971); Stjepanovikh-
Veselichikh in Josifovic, Fl. Srbije 3: 107, t. 31 f. 1 (1972); Bean, Trees & shrubs
hardy in Br. Isles 8th ed. 2: 407 (1973); Zangeri, Fl. italica 1: 394 (1976); Fournier,
Quartres flores de France 2nd ed.: 456 (1977); Bouchard, FI. pratique de Corse, ed.
3: 176 (1977); van Rompaey & Delvosalle, Atlas fi. belg. lux. ed. 2: t. 417 (1979).
Type: cultivated, Hort. cliff. 380 (BM!, lectotype) — see discussion below (p. 000).

H. bacciferum Lam.., Fl. frang. 3: 151 (1778); St. Lager in Cariot, Etudes des fleures
8th ed. 2: 140 (1889); nom. illegit. Type as for H. androsaemum.

Androsaemum officinale All., Fl. pedem. 3: 147 (1785); Choisy, Prodr. monogr. fam.
Hypéric.: 37 (1821), in DC., Prodr. syst. nat. regni veg. 1: 543 (1824); Spach, Hist.
nat. vég. Phan. 5: 415 (1836), in Annls Sci. nat. (Bot.) I, 5: 361 (1836); Rouy &
Foucaud, Fl. France 3: 348 (1896); Coste, Fl. France 1: 255 cum fig. (1901); Y.
Kimura in Nakai & Honda, Nova fl. jap. 10: 21 (1951). Type: Italy, Pedemontana
(various localities (TO)).

Androsaemum vulgare Gaertner, De fructibus 1: 282, t. 59 f. 2 (1788), nom. illegit.
Type as for H. androsaemum.

H. floridum Salisb., Prodr. stirp. hort. Chapel Allerton: 369 (1796), nom. illegit.
Type as for H. androsaemum.

H. androsaemum var. aureum hort. ex Nicholson, Handlist trees & shrubs Kew
Gardens 1: 37 (1894), nomen.

H. bacciforme Bubani, Fl. pyren. 3: 343 (1901), nom. illegit. Type as for H.
androsaemum.

Icon: Ross-Craig, Drawings Br. pls 6: t. 6 (1952).

Shrub 0-3-0-7 m tall, bushy with numerous stems from branching but not rooting
base, with branches erect. Stems 2-lined when young, eventually terete; internodes
3-9 mm long, shorter than to exceeding leaves; bark fissured or scaling. Leaves
sessile, sometimes amplexicaul, without goat-like smell when crushed; lamina
(25—)40-120(-150) x 20-80 mm, oblong-ovate or triangular-ovate to broadly ovate,
rounded or rarely obtuse, margin plane, base truncate to cordate, paler beneath, not
glaucous, papyraceous; venation: 4-5 pairs of ascending main lateral veins, with
tertiary reticulum + prominent on both sides; laminar glands small; intramarginal
glands dense. Inflorescence 1-11-flowered, from 1-2 nodes with + condensed
internode, corymbose to pseudo-umbellate with branches + narrowly ascending,
without accessory flowers, sometimes with 1—3-flowered branches from node below;
pedicels 8-14 mm long; bracts reduced, linear-lanceolate to subulate. Flowers 15—
25 mm in diam.; buds globose, rounded. Sepals 6-12(-15) x 3-7 mm, imbricate,
markedly unequal, spreading to deflexed and enlarging after anthesis, deflexed and

301

302

NORMAN K. B. ROBSON

persistent in fruit, oblong-ovate to broadly ovate, rounded, 7—9(—11)-veined from
base with veins + densely reticulate; laminar glands punctiform or occasionally
striiform; intramarginal glands rather dense. Petals golden yellow, not tinged red,
incurved, 6-10(-12) x 3-7 mm, equalling or slightly shorter than sepals, obovate,
without apiculus. Stamen fascicles each with 20—-25(—30) stamens, longest 7-11 mm
long, c. 0-9-1:1 x petals. Ovary 4-5 x 3-5-4-5(-5S) mm, ellipsoid-subglobose,
rounded to truncate; styles 2-2-5 mm long, c. 0-5 X ovary, erect, the upper half
sharply outcurved; stigmas subcapitate. ‘Capsule’ 7-12 x 6-8 mm, broadly cylindric-
ovoid or cylindric-ellipsoid to globose, rounded to retuse, thin-walled and fleshy,
maturing from reddish to dark reddish-brown or purplish-black, lucent, indehiscent
(but sometimes splitting in three if compressed laterally when dry). Seeds reddish-
brown, c. 1 mm long, partially or rarely completely ‘unilaterally winged, without
terminal winged appendages. 2n = 40.

In damp or shaded places, lowland to 1800 m (in Iran).

British Isles (widespread, but absent from NE. [except for naturalised plants],
extreme N., Outer Hebrides, and northern islands), Belgium (Ardennes), France
(W., also scattered in E. and S. — Oise, Cher, Is¢re, Aveyron, etc.), Spain (N.),
Portugal (N., south to Cintra), Morocco, Algeria, Tunisia, Sicily (W.), Sardinia,
Corsica (extinct ?), Italy (NW. south to Toscana, Elba ?, SE. — Ischia, Napoli ?,
Basilicata, also elsewhere but possibly naturalised). Yugoslavia (N. — Pula, Croatia,
Bosnia, Serbia), Bulgaria (SE.), Turkey (Thrace, Pontus, Amanus), U.S.S.R.
(Georgia, Azerbaijan, Daghestan, Turkmenistan), Iran (N. — Azerbaijan, Gilan,
Mazanderan, Gorgan). Introduced into Australia, New Zealand (North, South,
Stewart, and Campbell Islands), and Chile. Map 32.

GREAT BRITAIN. Somerset: Frome, Orchardleigh, 27.viii.1958, Wycherley 267
(BM). Sussex: Hastings, Pett, viii.1914, Redgrove s.n. (BM). Norfolk: Happisburg,
viii.1928 Meinertzhagen s.n. (BM). Hereford: Wye gorge, Dripping Well,
18.vii.1914, Adamson s.n. (BM). Caernarvon: Bardsey, vii? 1933, Butlers.n. (BM).
Westmorland: Grasmere, 22.vii.1939, Frankland s.n. (LIV). Dumfries: Brockle-
hurst, Collin, 5.viii.1952, Balfour-Browne 12 (BM). Argyll: Morven, Savory Bay
near Lochaline, 13.vii.1966, Cannon & Kendrick 221 (BM). Inverness: Isle of
Rhum, N. side of Loch Scresort, 5.viii.1959, Jermy 338 (BM). W. Ross: near Kyle of
Lochalsh, 21.vii.1909, Marshall s.n. (BM).

IRELAND. Kerry: Derrynane, Castle Cove, 26.vi.1930, Foggitts.n. (BM). South
Tipperary: outside Cappamurra House, 26.vii.1975, Pankhurst 192 (BM). Wicklow:
Glendalough, 25.viii.1951, Klingstadt s.n. (H). Galway: Connemara, near Clifden,
Faul, 16.vi.1953, Whittaker s.n. (BM). Donegal: Drummonaghan Wood,
26.vii. 1886, Kinahan s.n. (BM).

BELGIUM. Ardennes (fide van Rompaey & Delvosalle, Atlas Fl. belge lux., ed.
2: carte 417 (1979) ).

FRANCE. Calvados: Port Usembourg, vii.1868, Housnot s.n. (BM). Loire
Atlant.: forét de Gavre, 18.vi.1878, Gadeceau s.n. (BM). Cher: Forét d’Allogny,
7.vii.1851, Héribaud-Joseph 817 (BM). Landes: a l’ouest de Bernéde, 4 km SSE. de
Aire sur l’Adour, 2.vii.1977, Bamps 4734 (BM). Pyrenées Atlant.: Cambo, 1880,
Wilmott s.n. (K); Pyrenées Orient.: Vernet-les-Bains, 22.vii.1926, Ronniger s.n.
(W). Aveyron: Villefranche, bois de La Baume, vi.1849, ? (H). Isére: Cult. ex
Saint-Savin, prés de Bourgoin-Jallieu, bois du Laccas, 250 m, vii—vili.1877, Barbezat
163 (H). Alpes Maritimes: Fontan, 18.viii.1886, Reverchon 176 (BM).

SPAIN. Catalufia: Gerona, Villa de Arrozos, vi.1932, Krisogono 73 (BM).
Navarra: Alsasua, 26.vi.1952, Rovainen s.n. (H). Vascongadas: Viscaya, Miraval-
les, 29. vii.1914, Elias s.n. (BM). Galicia: La Coruna, near Coruna, viii.1888, Guendi
s.n. (BM).

PORTUGAL. Minho: Serra do Gerez [Geréz], 15.vi.1887, Murray s.n. (BM).
Aveiro: Oliveira d’Azemeis. Coimbra: Coimbra, 1849-50, Welwitsch 921 (BM).
15.v.1888, Murray s.n. (BM).

ALGERIA. Soummam: Zigalon, near El Kseur, 24.iv.1937, Alston & Simpson
37569 (BM).

TUNISIA. Jendouba: Ain-Draham, 1883, Miss. Bot. Tun. 1883 s.n. (K).

STUDIES IN THE GENUS HYPERICUM L. 303

Map 32 Sect. 5. Androsaemum: 2. H. foliosum A, 3. H. androsaemum ®@, © (records).

CORSICA. Haute Corse; Prope Orezza, 1830, Salis s.n. (K); Corti to Ajaccio,
Vivazione, 500-700 m, 3.vii.1970, Verdcourt 4781A (K).

SARDINIA. No detailed record or specimen seen (fide 5: Pignatti, Fl. d’Italia 1:
345 (1982)).

SICILY. Messina: Mistretta, Boschi di Amidda, vii.1887, Miojacono, PI. Ital.
Sel., Cent. IV 370 (BM).

ITALY. Basilicata: Pignola, Acqua fredda, 1200-1450 m, 6.vii.1923, Gavioli in
Fiori & Beguinot, Fl. Ital. Exs., ser. II, 2862 (BM, K). San Marino: Canepa,
26.vi.1912, Pampanini s.n. (K). Toscana: Gricigliano presso Firenze, vii.1890,
Martelli s.n. (H). Liguria: Apuania, Massa Montignosa, 14.vi.1951, Ranhala s.n.
(H). Piemonte: Valdensa, vi.1852, Rostan s.n. (BM). Lombardia: Lago Maggiore,
zwischen Stresa und Belgirate, 22.vi.1928, Ronniger s.n. (W). Veneto: Verona, Le
Ferrazze, 13.vi.1913, Rigo s.n. (BM).

SWITZERLAND. Ticino: Brissaga, 1865, Murets.n. (K).

YUGOSLAVIA. Bosnia: Banjaluka, Wald ober dem Tragspitzen, 7.vii.1929,
Ronniger s.n. (W).

BULGARIA. Strandzha Mts. (fide Jordanov & Kozhukharov in Jordanov, Fi.
Rep. Pop. Bulg. 4: 230 (1970)).

TURKEY. Kirklareli: Istranca dag. between Demirk6y and Velikakéy,
24.vi.1960, Kayacik 54 (E). Bursa: in Olympo Bythino, Noé (K). Bolu: Bolu dag,
2.vii.1957, Kuhne 1065 (E). Kastamonu: Kire to Inebolu, 1000 m, Czeczott 567 (E).
Sinop: Cangal dag above Ayancik, 700 m, 26.vii.1962, Davis 38162 (E, K). Samsun:
Kara dag, 10 km E. of Samsun, 200 m, 15.vi.1963, Tobey 319 (K). Giresun: Tamdere
to Yavuzkemal, near Karinca, 1500 m, 13.viii.1952, Davis 20761 (BM, E, K). Rize:
Pontic Range above Rize, 450 m, 14.v.1960, Furse & Synge 147 (K). Gaziantep:
Kurd dag, ix.1891, Post (BM).

U.S.S.R. (Transcaucasia). Georgia: Chakvi near Batumi, 30.vil.1979, Lancaster

304

NORMAN K. B. ROBSON

s.n. (BM). Azerbaijan: Baku, Kuba, supra pag. Cziczi inferiore versus, M. Clit dag,
900 m, 9.vii.1919, Alexcenko 1630 (S). Daghestan: Lenkoran, 1878, Haussknecht
(JE).

IRAN. Azerbaijan: Astara R. gorge, 20 km W. of Astara, 500 m, 21.vii.1964,
Grant 271 (W). Gilan: Navrud (Asalem) to Herowabad (Khalkhal), 1000-1300 m,
15.iii.1971, Lamond 2965 (E). Mazanderan: Elburz Mts, Golban Forest, 1200 m,
31.viii.1967, Walton 269 (E). Asterabad: Gorgan, entre Naudeh et Sharud, 1000—
2000 m, 11.vi.1956, Schmid 6068 (W).

U.S.S.R. Turkmenistan: Turcomania, Hb. Hooker. (K).

CULTIVATED. Specimens seen from England, Scotland, France.

NATURALISED. Australia: Tasmania, Port Cygnet, pre-1914, Kitson (BM).
New Zealand: N. Island, Wellington, Wainui-o-Mata, v.1929, Meebold 4622 (M);
Stewart Island, near Half Moon Bay, ii.1910, Chase & Leland 318 (BM). Chile:
Concepcion, San Juan, 1893-6, Neger (M); Valdivia, La Union, 10.ii.1964, Kausel
4718 (H).

H. androsaemum can be distinguished from both H. foliosum and H. x inodorum (H.
androsaemum x hircinum) by the short styles, the black or dark red-brown baccate
fruit, and the petals, which are equal to or shorter than the sepals. The shiny berry is
easily detachable from the receptacle and is eaten (and the seeds are dispersed) by
birds. If not eaten, after about a month it dries and withers on the plant; but it can be
split into three valves by slight pressure, and may do so spontaneously after
over-wintering.

The larger-flowered forms of H. androsaemum tend to have red-tinged leaves and
sepals and larger, cylindric-ovoid fruits that become brownish-red early and ripen to
dark reddish-brown; whereas those with smaller flowers tend to have green leaves
and sepals and smaller, cylindric-ellipsoid to globose fruits that ripen later to a bright
red before becoming shiny black. These forms, however, appear to be linked by a
continuous series of intermediates. In addition, a form with yellow-green leaves (var.
aureum hott.) is sometimes seen in cultivation; and a form with variegated leaves has
been found in two gardens in Ireland, having been collected originally at Lisnavagh
in Co. Carlow (D. McClintock, pers. comm.; cult. spec. in BM). This form also
occurred as a seedling at the Hillier Nursery, Winchester, Hampshire.

The selection of a lectotype of H. androsaemum L. is not as simple as it appeared
to me to be earlier, when I cited ‘Cult. in horto Upsaliensis’ in Fl. Iranica, Guttiferae:
5 (1968). Linnaeus first cites Hortus upsaliensis, in which the phrase-name has been
changed from that in Hort. Cliff., “caule fruticoso ancipiti’ having been substituted
for the earlier ‘foliis ovatis pedunculo longioribus.’ He could have observed the
ancipitous stems on the Herb. Cliff. specimen also, but no doubt did not happen to
have noted this character at the time. Linnaeus also cites Bauhin’s Pinax and
Dodoens’ Stirp. Hist. Pempt., finishing with ‘Habitat in Angliae, Ilvae sepibus’ and
the sign for ‘tree or shrub’.

The geographical data were derived respectively from Ray and Burser (?); but the
first two phrase-names are based on plants seen by Linnaeus himself. No specimen of
the plant that he grew in his ‘frigidarium’ at Uppsala is known to exist (in London,
Stockholm, Uppsala, Helsinki, or Moscow); but there is a specimen in Herb. Cliff.
(BM). The one is Herb. Linn. (LINN 943.12), of which the provenance is unknown,
is not H. androsaemum but H. x inodorum Miller. I therefore select the Herb. Cliff.
specimen (p. 380, No. 4), an authentic specimen seen by Linnaeus, as the lectotype
of H. androsaemum.

3x(61x). Hypericum x inodorum Miller

Gard. Dict. 8th ed.: no. 6 (1768); N. Robson in Flora europaea 2: 263 (1968), in
Green (ed.), Plants, wild & cult.: 81 (1973), in Fl. Gr. Brit. & Ireland 1: 000 (1985);
[Lancaster], Hillier’s man. trees & shrubs: 151 (1971); Bean, Trees & shrubs hardy
in Br. Isles 8th ed.: 415 (1973). Type: specimen cultivated in Chelsea Physic
Garden, Miller (BM!, holotype).

H. elatum Aiton, Hort. kew 3: 104 (1789); Desroux in Lam., Encycl. méth. (Bot.) 4:
156 (1797); A. L. de Jussieu in Ann. Mus. natn. Hist. nat. 3: 162, t. 17 (1804);

STUDIES IN THE GENUS HYPERICUM L. 305

Choisy, Prodr. monogr. fam. Hypéric: 38 (1821), in DC., Prodr. syst. nat. regni
veg. 1: 544 (1824); P. W. Watson, Dendr. Brit. 2: 85 (1825); Syme, Engl. bot. 3rd
ed.: 265 cum tab. (1865); R. Kellerin Engler & Prantl, Nat. Pflanzenfam. 2nd ed.,
21: 177 (1925) pro parte excl. loc. cit.; Stefanoff in God. Agr.-les. Fak. Univ.
Sofiya 11: 144 (1933) pro parte excl. loc. cit.; Turrill in Curtis’s bot. Mag. 173:
t. 376 (1962) pro parte quoad typum et syn. H. anglicum, A. pyramidale, A.
parviflorum. Type: cult. Hort. Kew, 1781, Hb. Goodenough (K!, holotype).

H. androsaemum sensu Sowerby & Sm., Eng. bot.: t. 1225 (1804) et auct.

Androsaemum pyramidale Spach, Hist. nat. vég. Phan. 5: 417 (1836), in Annis Sci.
nat. (Bot.) II, 5: 362 (1836); K. Koch, Hort. dendr.: 65 (1853); nom. illegit.
superfl. Type as for H. elatum Aiton.

A. pyramidale forma grandifolia Spach, Hist. nat. vég. Phan. 5: 417 (1836). Types:
P. W. Watson, Dendr. brit. 2: t. 85 (1825) and Jussieu in Ann. Mus. natn Hist. nat.
3: 162, t. 17 (1804), syntypes.

A. parviflorum Spach in Annls Sci. nat. (Bot.) II, 5: 361 (1836); K. Koch, Hort.
Dendr.: 65 (1853). Types as for H. pyramidale forma grandifolia [sic] Spach.

Hypericum anglicum Bertol., Fl. ital. 8: 310 (1853); Babington in Ann. Mag. nat.
Hist. II, 15: 92 (1855), in Trans. bot. Soc. Edinb. 5: 42 (1958); Walker-Arnott in
Ann. Mag. nat. Hist. Il, 6: 362 (1860). Types: Italy, Venato, R. Brenta at S.
Fortunato. Bertoloni (BOLO, syntype); woods at Angarano, Montini (BOLO,
syntype).

H. multiflorum sensu hort. ex Dippel, Laubholzkunde 3: 39 (1893) in synon., non
Kunth (1922).

H. x urberuagae P. & S. Dupont in Monde Plantes 325: 3 (1958), nom. illegit. altern.
Type: Spain, Viscaya, Urberuaga de Ubilla, P. & S. Dupont (Herb. Dupont. ?).

Androsaemum x urberuagae P. & S. Dupont in Monde Plantes 325: 3 (1858), nom.
illegit. altern. Type as for Hypericum x urberuagae.

Hypericum persistens 1. F. Schneider in Meded. Inst. Vered. Touinbougew.,
Wageningen 252: 22 (1966), in Dendrofl. 2: 22 (1967). Type: cult. in Hort.
Wageningen (WAHO).

H. persistens ‘Elatum’ I. F. Schneider in Meded. Inst. Vered. Touinbougew.,
Wageningen 252: 22 (1966), in Dendrofl. 2: 22 (1967). Type: as for H. x persistens
I. F. Schneider.

Icon: Syme, Engl. Bot. 3rd ed.: t. 265 (1865).

Shrub 0-6-2 m tall, bushy with numerous stems from branching but not rooting base,
with branches erect. Stems 2-lined when young, eventually terete; internodes
2-8-5 mm long, shorter than to exceeding leaves; bark scaling. Leaves sessile to
subsessile, sometimes amplexicaul, rarely with goat-like smell when crushed; lamina
35-110 x 20-60 mm, oblong-lanceolate to broadly ovate, subacute to rounded,
margin plane, base rounded to cordate, somewhat paler beneath, not glaucous,
papyraceous; venation: 4-5 pairs of ascending main lateral veins, with tertiary
reticulum + prominent on both sides; laminar glands small; intramarginal glands
dense. Inflorescence (1—)3-c. 23-flowered, from 1—2 nodes without condensed inter-
node, corymbose to subcorymbose, with branches + narrowly ascending, without
accessory flowers, often with 1-3-flowered branches from node below; pedicels 6-15
mm long; bracts reduced, lanceolate to subulate. Flowers 15-30 mm in diam.; buds
globose, rounded. Sepals 5-9-5 x 2-5 mm, imbricate, markedly unequal, deflexed
and enlarging after anthesis, persistent at least until fruit ripens, oblong or oblong-
lanceolate to broadly ovate, subacute to rounded, 5—9-veined from base with veins +
branched and reticulate; laminar glands mostly punctiform; inframarginal glands
dense. Petals golden yellow, not tinged red, spreading to incurved, 8-15 x 4-5-7 mm,
c. 1-5-2 x sepals, oblanceolate to narrowly obovate, without or almost without
apiculus. Stamen fascicles each with 20-30 stamens, longest 10-20 mm long, c.
1-2-1-3 x petals. Ovary 3-5-6-5 x 2-5 mm, ellipsoid to ovoid-ellipsoid, subacute to
subrounded; styles (6-)8-16 mm long, c. 1-2-5 X ovary, erect, obtusely divergent in
the upper 1/3-1/4; stigmas narrowly capitate. Capsule 6-13 x 4-8 mm, ellipsoid or
ovoid-ellipsoid to cylindric-ellipsoid, obtuse to rounded, thin-walled and + fleshy,
maturing from cerise or bright red to dark brown, sublucent, sometimes dehiscent at

306

NORMAN K. B. ROBSON

the apex. Seeds reddish-brown, c. 1-2-1-5 mm long, partially or rarely completely
unilaterally winged, sometimes with terminal winged appendages. 2n = 40.

In damp or shaded places, lowland.

Possibly natural distribution (where areas of parents overlap): SW. France (Pyr.
Atl., Landes), NW. Spain (La Coruna, Guipuzcoa ?), Corsica (Bastia), Italy
(Liguria). Introduced into and naturalised (or originated de novo) in the British Isles
(all countries, mainly S. and W. England and the Channel Islands), France (scat-
tered), Portugal (Porto), Switzerland (Valais), Italy (Veneto), Madeira, New Zea-
land, Mexico, and Chile.

GREAT BRITAIN. Cornwall: near Menabilly, 1931, Medlin s.n. (K). Somerset:
Worle, 28.vii.1899, White s.n. (BM, K). Surrey: Camberwell, Dulwich Woods,
5.x.1958, Lousley s.n. (BM). Essex: Thorpe-le-Soken, vii—viii.1900, Standen s.n.
(BM, K). Herefordshire: Goodrich, viii.1861, Watkins s.n. (BM). Caernarvon:
Llanfairfechan, 5.viii.1938, Last s.n. (K). Yorkshire: Cayton Bay, near Scar-
borough, 22.vii.1938, Evans s.n. (BM). Cumberland: Sebergham, ix.1886, Lomax
s.n. (BM). Wigtown: near Portpatrick, H.B. Balfour (K). Buteshire: Arran,
vii.1840, Maclagan s.n. (BM).

IRELAND. Sligo: Lough Gill, 17.viii.1913, Barton s.n. (BM). Down: Donard
Lodge, ix.1860, Arnott s.n. (K).

CHANNEL ISLANDS. Sark: east coast, fl. 19.vii.1953, Sowerby 305 (BM).

FRANCE. Maine et Loire: La Baumette, 21.vii.1853, Genevier s.n. (BM).
Rhéne: Arnas, 6.vii.1875, Gandoger 808 (BM, JE). Landes: Boucou, prés Bayonne,
vi.1869, Bordére s.n. (K). Pyren. Atlant.: Bayonne, vi.1870, Bordére (K).

SPAIN. Galicia: La Coruna, Betanzos, vii.1901, Peyils.n. (BM).

PORTUGAL. Porto: Juxta Oporto, pre-1867, de Paivas.n. (K).

CORSICA. Bastia, 4.vii.1875, Gandoger s.n. (FR, JE).

ITALY. Fide Pignatti, Fl. d'Italia 1: 345 (1982).

SWITZERLAND. Valais: Sion, vi.1872, Wolf s.n. (K).

MADEIRA. ‘Ubique in sepibus et rupestribus’, 300-1000 m, vii.—x.1865, Mandon
34 (BM, JE, K).

MEXICO. Morelia: Rericon prés Morelia, 26.ix.1911, Arséne (L).

CHILE. Valdivia: Lake Llanquihue, 1912, Calvert s.n. (BM). Chiloe: Concep-
cién, Fundo Piruquina, 28.i.1933, Belus.n. (M).

CULTIVATED. Specimens seen from England, France, Belgium, the Nether-
lands, Germany, Denmark, Sweden, U.S.A.

H. x inodorum appears to have originated spontaneously, both in cultivation
(England prior to 1759) and in natural habitats, where the parental species grow
together. The hybrids are at least partially fertile (chromosomes of all species in sect.
Androsaemum so far counted are 2n = 40), and segregation as well as presumed
back-crossing has resulted in a range of intermediate forms. The larger-flowered tall
forms (H. elatum) have H. hircinum subsp. majus as a parent, whereas in the dwarfer
smaller-flowered forms (H. x inodorum sensu stricto, H. multiflorum sensu hort.) the
hircinum parent is apparently subsp. cambessedesii or possibly sometimes subsp.
hircinum. As there is a continuous range of variation in H. x inodorum between these
extremes, it is not possible to recognise two nothovarieties. On the other hand,
segregation and selection have resulted in the establishment of two groups of
cultivars (see below, p. 307).

H. x inodorum has been the subject of misunderstanding in more than one way:

(i) Aiton (Hort. Kew. 3: 104 (1789)) gave the origin of H. elatum as North
America, and it was not until the publication of Coulter’s account of North American
Hypericum (in Bot. Gaz. 11: 82 (1886)) that it was firmly recognised as an Old World
species.

(ii) Several authors, including Rehder (Man. Cult. Trees & Shrubs ed. 2: 641
(1940)), have confused it with H. grandifolium and given its natural distribution as
the Canary Islands and Madeira, one of the latest being Turrill (Curtis’s bot. Mag.
173: t. 376 (1962)). Schneider, who was also misled by Rehder’s statement into
thinking that H. elatum was a Canary Island plant, gave the plant that he knew in

STUDIES IN THE GENUS HYPERICUM L.

cultivation a new name — H. x persistens (see synonymy). The label of the plant from
Madeira cited above (Mandon 34) is almost certainly another result of this confusion;
Mandon, having apparently thought that his plant was the same as the one that is
widespread on the island (H. grandifolium), described it as occurring everywhere:
‘ubique’.

Miller chose the epithet inodorum to distinguish his species from the well known
H. hircinum, which always (in the forms then in cultivation) had a goat-like smell.
See also Willdenow (1802: 1449). This distinction is usually valid; but, not sur-
prisingly, there are some forms of H. x inodorum in which the smell characteristic of
hircinum is quite apparent.

As was mentioned above, two form-groups of H. x inodorum have been recog-
nised as cultivars.

3xa(61xa). H. x inodorum ‘Elstead’

Hilliers’ man. trees & shrubs: 151 (1971); N. Robson in Green (ed.), Plants, wild &
cult.: 81 (1973). Standard specimen: England, Surrey, Kew, Royal Botanic
Gardens, Arboretum, 21.ciii.1960, Bot. Mag. II, 173: t. 376 (K!).

H. elatum ‘Elstead Variety’ (Ladhams 1933); Chittenden in J. Roy. Hort. Soc. 59:
305, cxxix (1934); Thomas in Gdnrs’ Chron. 147: 254 (1960).

H. elatum sensu Turrill in Curtis’s bot. Mag. 173: t. 376 (1962) pro parte, quoad tab.
et descr. pro parte.

H. persistens ‘Elstead’ I. F. Schneider in Meded. Inst. Vered. Touinbougew.,
Wageningen 252: 22 (1966), in Dendrofl., 2: 22 (1967).

Icon: Turrill in Curtis’s bot. Mag. 173: t. 376 (1962).

Flowers 25-30 mm in diam. Styles c. 6 mm long, about equalling ovary. Capsule
16-17 x 8mm, becoming bright pinkish-red (not cerise) during maturation.

ENGLAND. Surrey: Kew, Royal Botanic Gardens, 14.viii.1960, Turrill (K).

Named cultivars also include ‘Summergold’ (golden yellow leaves), ‘Ysella’,
‘Goudelsje’, ‘Hysan’ and ‘Beatty’s Variety’. These all show different forms of
variegation and all occasionally appear in catalogues or lists.

4(62). Hypericum hircinum L.

Sp. pl.: 784 (1753); Choisy, Prodr. monogr. fam. Hypéric.: 39 (1821), in DC., Prodr.
syst. nat. regni veg. 1: 544 (1824); Treviranus, Hyper. gen. sp. animad.: 7 (1840):
R. Keller in Engler & Prantl, Nat. Pflanzenfam. 2nd ed. 21: 177 (1925); Stefanoff
in God. Agr.-les. Fak. Univ. Sofiya 10: t. 3 f. 7 (1932), 11: 143 (1933), 12: 82 (1934),
in Pflanzenareale 4: karte 2a (1933); Sauvage in Bull. Soc. Sci. nat. phys. Maroc 38:
123 (1958); N. Robson in Fl. europaea 2: 263 (1968); Plaisted & Lighty in Nat.
Hort. Mag. 38: 130 (1959); Thomas in Gdnrs’ Chron. 147: 255 (1960); [Lancaster],
Hilliers’ man. trees & shrubs: 151 (1971); Bean, Trees & shrubs hardy in Br. Isles
8th ed. 2: 414 (1973). Types: Hort. cliff. 381 [sphalm. 331 in Sp. pl.], no. 9; Herb.
Linn. 943. 16 (LINN!, lectotype—mihi; BM-Herb. cliff.!, syntype).

Androsaemum hircinum (L.) Spach, Hist. nat. vég. Phan. 5: 419 (1836), in Ann. Sci.
nat. (Bot.) II 5: 362 (1836); K. Koch, Hort. Dendrol.: 65 (1853); Nyman, Consp. fl.
europ. 1: 131 (1878); Rouy & Foucaud, FI. France 3: 349 (1896); Coste, Fl. France
1: 256 cum fig. (1901).

Icon: Fig. 22.

Shrub (0-2—)0-5—1-8(—3) m tall, bushy with numerous stems from branching but not
rooting base, with branches erect to spreading or pendulous. Stems 4-lined and
ancipitous when young, eventually 2-lined or terete; internodes 2-9 mm long,
shorter than to exceeding leaves; bark grey-brown, fissured. Leaves sessile to
subsessile, sometimes amplexicaul, when crushed usually smelling of goats (caproic
acid); lamina 20-65(-75) Xx (8-)12-27(-35) mm, broadly ovate to triangular-
lanceolate, acute to rounded, margin plane or crisped-undulate, base cordate or

307

308 NORMAN K. B. ROBSON

Fig. 22. H. hircinum. A. subsp. majus: (a) habit; (b) leaf; (c) leaf section; (d) flower bud; (e) sepal; (f)
petal; (g) stamen fascicle; (h) anthers; (i) ovary; (j) capsule. B. subsp. cambessedesii: (k) leaf; (1)
capsule. C. subsp. hircinum: (m) leaf; (n) capsule. D. subsp. albimontanum: (0) leaf; (p) capsule (a, b, k,
m, o X 1; c-g,i,j,l,n xX 4;h xX 16).

STUDIES IN THE GENUS HYPERICUM L.

rounded to cuneate or shortly angustate, somewhat paler beneath, very rarely
glaucous, chartaceous to papyraceous; venation: 2-4(—S) pairs of ascending main
lateral veins, with tertiary reticulum + prominent on both sides; laminar glands
small; intramarginal glands dense. Inflorescence (1—)3-c.20-flowered, from 1-2
nodes without condensed internode, broadly pyramidal, often with 3-flowered
branches from node below and flowering branches from up to 4 nodes below that, the
whole then narrowly pyramidal to subcylindric; pedicels 4-10 mm long; bracts rarely
foliar, usually reduced, lanceolate to subulate. Flowers (20—)2540 mm in diam.;
buds + globose, rounded. Sepals (2—)4-8 x 1-3 mm, imbricate or open, somewhat
unequal, decurving and sometimes slightly enlarging after anthesis, deciduous
before fruit ripens, lanceolate to narrowly ovate, acute to shortly acuminate,
5—7-veined from base with veins + branched and reticulate; laminar glands linear to
striiform; inframarginal glands + dense. Petals golden yellow, not tinged red,
spreading, 11-21 x 4-9 mm, 3-4 x sepals, oblanceolate or oblong-oblanceolate to
narrowly obovate, with apiculus subterminal and obsolete or absent. Stamen fas-
cicles each with c. 20 stamens, longest 12-22 mm long, 1-1-2 x petals. Ovary 3-5 x
2-3-5 mm, ellipsoid, acute; styles 10-24 mm, 3-5 x ovary, erect, narrowly divergent
in the upper 1/3-1/4; stigmas narrowly capitate. Capsule 5—14 x 4-7 mm, ellipsoid or
ovoid-ellipsoid to subcylindric, coriaceous, maturing from green to mid-brown, dull,
incompletely dehiscent. Seeds orange- to reddish-brown, c. 1-:2-1-5 mm long,
completely unilaterally winged, with terminal winged appendages. 2n = 40.

In damp or shaded places, often beside streams, (30—)300—1200 m (to 2100 m in Saudi
Arabia).

France (possibly native in SW.; naturalised in W., north to Calvados and Manche,
and S.-central), Spain (N. from Catalufa to Santander; naturalised in extreme S. and
in Gibraltar), Balearic Islands (Mallorca), Italy (Liguria, Lazio, Campania, Basilica-
ta, Calabria, Elba, Ischia, Lipari Islands), Sicily, Sardinia, Corsica, Morocco,
Greece (Peloponnesus), Crete. Andhros, Samos, Rhodes, Turkey (S.), Cyprus,
Syria, Lebanon, Israel (N.), Saudi Arabia (Asir). Naturalised also in Portugal,
British Isles (all countries), and Yugoslavia (Pula). Map 31.

The natural distribution of H. hircinum on the western European mainland is
difficult to establish because of the frequency with which it has escaped from
cultivation. The map published by Sauvage (1958) includes northern Portugal, the
Landes, Pyrenees, and Cataluna, and a continuous strip of land from there down the
western side of Italy to Basilicata, but doubtfully excludes Sicily and the extreme
south of Calabria. In Flora Europaea (Robson, 1968b), on the other hand, the
regional advisers took the view that the species is not native anywhere in Portugal,
Spain, or France, but is so in Sicily. From a study of variation, specimen labels and
floras, it would appear that H. hircinum grows in natural habitats away from
habitation in at least some localities in Spain, France, and Sicily, and that the
endemic subspecies of the Balearic Islands (subsp. cambessedesii) resembles most
closely specimens from Cataluna. It seems likely, therefore, that this species is a
native component of the floras of France, Spain, and Sicily. On the Italian mainland,
there seems to be a gap in its distribution between (i) Liguria and (ii) Lazio to
southern Calabria.

H. hircinum is very variable; but the variation, though clearly geographical, is not
sufficiently discontinuous to allow recognition of more than one species. It is
possible, however, to distinguish no less than five subspecies (Fig. 8). The most
primitive form (of subsp. majus), which is the most similar one to H. grandifolium,
occurs in southern Italy, Sicily, and (rarely) Crete, and rather similar plants are
found in Rhodes, S. Turkey, the Levant, and Asir. From the two Mediterranean foci
trends may be discerned leading to pendulous branches (Morocco — subsp. metroi),
to narrower, acute leaves and smaller parts (Liguria, Pyrenees etc., ending in the
Balearic Is. — subsp. cambessedesii), to smaller, broader, rounded leaves with plane
margin and smaller fruits (Sardinia, Corsica — subsp. hircinum), and to smaller,
broader, apiculate to rounded leaves with crisped-undulate margin and smaller fruits
(Cyprus, Samos, Andros, Greece, Crete — subsp. albimontanum).

The description of H. hircinum L. in Species plantarum is clearly based primarily

309

310

NORMAN K. B. ROBSON

on that in Hortus Cliffortianus, the phrase name being taken directly from this work.
It does not indicate any particular subspecies. There is one specimen of H. hircinum
in Herb. Cliff. (BM) and a duplicate of it in the Linnaean Herbarium (LINN). One of
these is clearly the indicated lectotype, and I have selected the one in Herb. Linn.
(943.16). Both these specimens, however, are clearly of the subspecies from Corsica
and Sardinia (subsp. obtusifolium (Choisy) Sauvage), which must therefore become
subsp. hircinum. Linnaeus appears to have based his concept in 1753 of H. hircinum
largely on this subspecies; whereas his Hort. Cliff. synonymy (1738) included
‘Hypericum frutescens maj. & minus. Dill. elth. 182. t. 151, f. 181, 182’, in Species
plantarum he omitted ‘majus’ and ‘f. 181’. The other citations in Species plantarum
(to ‘Bauh. pin.’ and ‘Clus. hist.’), as well as the distribution (Habitat in Sicilia,
Calabria, Creta’), refer in part to other subspecies; but this fact cannot be taken as an
indication that another element should be chosen as lectotype. The subspecies called
subsp. hircinum by Sauvage therefore requires another name, and it is appropriate to
use the Dillenian epithet majus from Hortus elthamensis.

4a(62a). H. hircinum subsp. majus (Aiton) N. Robson, stat. nov.

H. hircinum [var.] a majus Aiton, Hort. kew. 3: 105 (1789). Type: Dillenius, Hort.
elthamensis: t. 151, f. 182 (1732).

H. hircinum sensu Choisy in DC., Prodr. syst. nat. veg. 1: 544 (1824) pro parte, excl.
6 obtusifolium; P. W. Watson, Dendrol. brit. 2: t. 86 (1825); Syme, Eng. bot. 3rd
ed.: t. 266 (1865); Boissier, Fl. orient. 1: 788 (1867) pro parte, excl. spec. graec. et
cret.; Willk. & Lange, Prodr. fl. hisp. 3: 589 (1878); Fiori & Paoletti, Nuovo fl.
analit. Italia 1: 520 (1924); Post, Fl. Syria, Pal. & Sinai 2nd ed. 1: 228 (1932); H.
Lindberg in Acta Soc. sci. fenn. I B, 2 (7): 23 (1946); Gismondi, Prosp. fi. ligust.
336 (1950); K. H. Rechinger in Arkiv. for Bot. II, 5: 291 (1960); Zohary, Fi.
Palaest. 1: 221, t. 324 (1966); N. Robson in Davis, Fl. Turkey 2: 366, f. 13(4), map
20 (1967); Mouterde, Nouvelle Fl. Lib. Syrie 2: 520, t. 225 f. 1 (1970);
Zangheri, Fl. Italica 1: 394 (1976); Fournier, Quatre Flores de France 2nd ed.:
456 (1977).

H. hircinum subsp. hircinum sensu Sauvage in Bull. Soc. Sci. nat. phys. Maroc 38:
127 (1958).

Icones: Zohary, Fl. Palaest. 1: t. 324 (1966); Fig. 22A.

Plant (0-5-)1-1-5 m tall, with stems erect or ascending. Leaves 30-75 x (10-)
14-32 mm, narrowly ovate to lanceolate or triangular-lanceolate, acute to obtuse or
more rarely rounded, margin plane, base subcordate to rounded, goat-scented.
Flowers 25-40 mm in diam. Petals (13)15-21 mm long. Styles (13)15—24 mm long.
Capsule 8-14 x 5-7 mm.

N. and S. Spain, SW. France, NW. and SW. Italy and adjacent islands, N. Sicily,
Crete, Rhodes, S. Turkey, Syria, Lebanon, N. Israel, Saudi Arabia (Asir). Natural-
ised elsewhere (see p. 309). Map 31.

GREAT BRITAIN (naturalised). Cornwall: near Par Harbour, 15.ix.1948,
Alston s.n. (BM). Somerset: Clevedon, Court Hill, 16.vii.1884, White s.n. (BM).
Sussex: Surrey Hatch, Fernhurst, 22.viii.1918, Barton s.n. (BM). Middlesex:
London, Westminster, wall of The Mall by Marlborough House, x.1968, McClintock
s.n. (BM). Norfolk: Holkham, 18.viii.1882, Linton s.n. (BM). Pembroke: Hundle-
ton, near Pembroke, vii.1916, Wallis s.n. (BM). Caernarvon: Llanfairfechan,
26.ix.1884, Bailey s.n. (BM). Anglesey: Garth Ferry, 8.viii.1936, Redgrove s.n.
(BM). Lancashire: Southport, Crossens, 13.viii.1949, Holder 4944 (LIV). York-
shire: Settle, 19.ix.1868, Joads.n. (K). Wigtownshire: Sorbie, 1897, McAndrew s.n.
(BM). Inverness: North Ballahulish, 11.viii.1939, Wilmott s.n. (BM).

IRELAND (naturalised). Cork: near Cork, Glanmire, by T11, 2.viii.1975, Pank-
hurst 273 (BM). Mayo: Bongor Erris, viii.1957, McClintock s.n. (BM).

FRANCE (naturalised and ? native). Calvados: Vierville, vii.1925, Bédel 401
(BM). Indre-et-Loire: Locher, 29.vii.1857, Genevier s.n. (BM). Loire Atl.: vis a vis
le Bois Cillac, viii.1891, Le Pellerin s.n. (BM). Char. Mar.: Dompierre, viii.1887,
Teneron s.n. (BM). Vendée: Noirmoutier, vii. 1869, Viaud-Graud-Marais & Gobert

STUDIES IN THE GENUS HYPERICUM L.

s.n. (BM). Landes: Castillon-sur-Tarnas, 27.vi.1881, Blanchet s.n. (BM, E). Pyr.
Atl.: St. Esprit prés Bayonne, 19.vii.1880, Bonnet s.n. (K). Rhéne: Denivé,
7.vii.1875, Gandoger s.n. (FR). Cote d’Or: Pouillenay, 2.x.1929, Desplantes in
Duffour 5856 (BM).

SPAIN (native ?). Catalufia: Figueras, fossé en face de Moli de la Torre,
11.vii.1912, Joseph in Sennen 1378 (BM, E, JE). Asturias: Santander, 10.x.1851,
Lange (K).

GIBRALTAR (naturalised). Gibraltar, 23.xii.1840, Lemann (K).

PORTUGAL (naturalised). (Fide Franco, Nova Fl. Port. 1: 448 (1971).)

ITALY. Liguria: near Genoa (fide Parlatore, Fl. Ital. 5: 507 (1872)). Toscana:
Elba, 18.vii.1837, Marciana (E). Lazio (fide Pignatti, Fl. d’Italie 1: 345 (1982)).
Campania: Positano (near Amalfi), 13.vii.1953, Davis 17431A (E); Isola d’Ischia,
18.viii.1891, Martelli (E). Puglia, Basilicata (fide Pignatti, Joc. cit.). Calabria: 4 km
north of Sinopoli, 510 m, 30.viii.1966, Lambert 351 (BM). Lipari Is. (fide Parlatore
tom. cit.: 508).

SICILY. Palermo: Palermo, vi.1902, Ross 319 (BM, E, JE). Messina: c. 12 km
NW. of Floresta towards Tortorici, c. 1300 m, 2.vi.1979, Davis & Sutton 64337 (BM).
Catania: c. 8 km W. of Zafferana Etnea, slopes of Etna, c. 1000 m, 5.vi.1979, Davis
& Sutton 64520 (BM).

YUGOSLAVIA (naturalised). Pula: Molara fra Barana e Campagnano,
22.vi.1859, Haeckel s.n. (JE).

GREECE. Dhodhekanisos: Rodhos, Monte Profeta prope Salakos, c. 300 m,
2.vii.1935, K. H. Rechinger 8497 (BM, K).

CRETE. Lasithi: Mt. Aphendhi Kavutsi, 23.vi.1937, Lempberg 548 (K).

TURKEY. Antalya: Gebiz (Pisidia). Bozburum dag at Penargozu yayla, 600 m,
23.vii.1949, Davis 15394 (E, K). Konya: Near Ermanek, 1872, Péronin 153 (K, S).
Icel: Silifke to Anamur, 50 m, 9.Ix.1962, Kayacik 2927 (E). Seyhan: Amanus
Gebirge zwischen Osmaniye und Harpuz, c. 900 m, 13.vii.1978, Ehrendorfer, Sorger
et al. 787-35-6 (WU). Antakya: Iskenderun, between Sogukaluk and Karliktefe,
18.vii.1971, G. & E. Sezik 165 (BM).

SYRIA. Halep: Kurd Dagh, ix.1891, Posts.n. (K). Latakia: Monts Nusairy [Jebel
el Ansariye], Ain-Halakim, 750-900 m, vi.1910, Haradjian 3474 (K). Damascus:
NW. of Damascus, W. of Souk Wadi Barada, Jabal Hafiel, 21.vi.1963, Barkoudah
652 (E).

LEBANON. Becharre, 18.viii.1945, Davis 10088 (E); ad Brummana, 700-800 m,
7.vii.1897, Bornmiiller 237 (E, JE, K); Antilebanon, Jdeide village, 750 m,
31.vii.1945, Harding, Hardy & Whitehorn (BM).

ISRAEL. Dan: Reserve near source of the Jordan. 18.11.1967, Hepper (K).

SAUDI ARABIA. Asir: Bani Shahr, 2250 m, 8.vi.1946, Thesiger s.n. (BM);
Taif-Abha road, 12 km S. of An Numas, Taunoumah, 1800 m, 15.iv.1979, Collenette
1396 (K).

CULTIVATED. Specimens seen from England, Germany, and U.S.A.

4b(62b). H. hircinum subsp. cambessedesii (Nyman) Sauvage

in Bull. Soc. Sci. nat. phys. Maroc 38: 127 (1958). Type: (as for H. canariense sensu
Cambessédes) Majorca, ‘In insulae Majoris torrente dicto Malluch prope Lluch’,
1824, Cambessédes (K!, holotype).

H. hircinum [var.] 8 minus Aiton, Hort. kew. 3: 105 (1789) pro parte, quoad specim.
‘Kew 1777’ (BM): Bean, Trees & shrubs hardy in Br. Isles 8th ed. 2: 414 (1973).

H. canariense sensu Cambessédes, Enum. pl. ins. Balear.: 54 (1827).

Androsaemum cambessedesii Nyman, Consp. fl. europ.: 131 (1878). Type as for H.
canariense sensu Cambessédes (1827).

Hypericum cambessedesii (Nyman) Cosson ex Marés & Vigin., Cat. rais. pl. Baléar.:
330, t. 3 (1880); Knoche, Fl. Balear. 2: 183 cum carta (1922); Stefanoff in God.
Agr.-les. Fak. Univ. Sofiya 10: tt. 2. f. 6, 3 f. 8 (1932), ibid. 11: 143 (1933), 12: 82
(1934), in Pflanzenareale 4: karte 2a (1933).

H. minus sensu hort. ex Nicholson, Handlist trees & shrubs Kew Gardens: 39 (1894),
in synon.

311

312

NORMAN K. B. ROBSON

H. hircinum var. cambessedesii (Nyman) Ramos in Trab. Dep. Botanica Ciudad
Univ., Madrid 12: 49 (1983).

Icones: Marés & Vigin, Cat. rais. Pl. Balear.: t. 3 (1880); Fig. 22b.

Plant 0-2-1 m tall, with stems erect. Leaves 2248 x 7-20 mm, lanceolate to
triangular-lanceolate, subacute to obtuse, margin plane, base rounded to cuneate,
goat-scented. Flowers 20-30 mm in diam. Petals 10-15 mm long. Styles 10-13 mm
long. Capsule 8-14 xX 5-7 mm.

Balearic Islands (Mallorca). Map 31.

BALEARIC ISLANDS. Mallorca: In montibus |. Prats de Massonelles, pr. Lluc,
1000 m, 27.vi.1935, Gros in Fl. Iber. Sel. , Cent. II, 147 (BM, JE, K); c. 1:5 km inland
from mouth of Torrente de Pareis towards monastery of Lluch, c. 60 m, 13.11.1971,
L. F. & I. K. Ferguson 2945 (BM).

CULTIVATED. Specimens seen from England (from earlier than 1777).

Although the epithet minus was originally applied by Dillenius to the taxon that has
to be known as subsp. hircinum (i.e. subsp. obtusifolium (Choisy) Sauvage), Aiton’s
confusion of this with subsp. cambessedesii has resulted in the latter’s currently being
called ‘var. minus’ in horticultural circles. The confusion between this subspecies and
H. canariense, which was originated by earlier authors than Cambessédes (1827)*,
has persisted down to the present day, so that H. cambessedesii has been recorded in
error for Macaronesia by several authors.

Subsp. cambessedesii is very similar to the form of subsp. majus in Catalufa,
differing essentially from it only in size of parts.

4c(62c). H. hircinum subsp. metroi (Maire & Sauvage) Sauvage

in Bull. Soc. Sci. nat. phys. Maroc 38: 127 (1958). Type: Morocco, Mont Tazzeka,
vers 1200 m, Métro & Sauvage s.n. (MPU, holotype).

H. metroi Maire & Sauvage in Bull. Soc. Hist. nat. Afr. Nord 31: 13 (1940); Sauvage
in Trav. Inst. sci. cherif. (Bot.) 8: 67-72, f. 3 (1956). Type as above.

Icon: Sauvage in Trav. Inst. sci. cherif. (Bot.) 8: 69, f. 3 (1956).

Plant 1-3 m tall, with stems spreading to pendent. Leaves c. 50-70 X 16 mm,
narrowly ovate or lanceolate to triangular-lanceolate, subacute to obtuse, margin
plane, base rounded, not goat-scented. Flowers 30-40 mm in diam. Petals 16-20 mm
long. Styles 12-17 mm long. Capsule 10-12 xX 7 mm.

Morocco (Moyen Atlas — massif du Tazzeka). Map 31.
MOROCCO. Moyen Atlas: Taza distr., Oued Rmila, 1000 m, 15.xi.1970, Sgnder-
gaard s.n. (E).

Subsp. metroi is very similar to the large-flowered form of subsp. majus from
southern Italy, differing essentially from it only in habit and in the absence of scent.

4d(62d). H. hircinum subsp. hircinum

H. hircinum [var.] B minus Aiton, Hort. kew. 3: 105 (1789) pro parte, quoad syn.
Dillen.

H. hircinum [var.] 6 obtusifolium Choisy in DC., Prodr. syst. nat. veg. 1: 544 (1824).
Type: Corsica, ‘In petrosis humidis montium Corsicorum, 1821,’ Salzmann (G —
DC!, holotype).

H. hircinum var. pumilum P. W. Watson, Dendrol. brit. 2: t. 87 (1825). Type:
England (cultivated), London ‘Mr Knight’s, King’s Road, Chelsea, 13.ix.1821’.
No specimen traced.

H. hircinum sensu Moris, Fl. Sardoa 1: 316 (1837); Briquet, Prodr. fl. Corse 2 (2):
143 (1936) et auct. cors. et sard.

* Two Miller specimens at BM, one in the general collection and the other in Herb. Sloane
t. 288 f. 74 (both of which are H. hircinum subsp. cambessedesii), have ‘Hypericum frutescens
canariense multiflorum. Hort. Amstel. 2. 135’ on their label, although this phrase name
ref-rs to H. canariense L.

STUDIES IN THE GENUS HYPERICUM L.

H. hircinum subsp. obtusifolium (Choisy) Sauvage in Bull. Soc. Sci. nat. phys. Maroc
38: 127 (1958).

Icones: P. W. Watson, Dendrol. brit. 2: t. 87 (1825); Fig. 22C.

Plant c. 1 m tall, with stems erect. Leaves 25-45 x 12-27 mm, + broadly ovate,
obtuse to rounded, margin plane, base truncate to rounded, rarely goat-scented.
Flowers 25-40 mm in diam. Petals 15-18 mm long. Styles 14-20 mm long. Capsule
(6-)8-9 x 4-6 mm.

Sardinia, Corsica. Map 31.

SARDINIA. Gallara: secundum Rio Mannu (Rio di Liscia), loco dicto il Vignale
prope Luogosanto, 150 m, vii.1905, Vaccari in Fiori & Béguinot, Fl. Ital. Exsice.,
ser. II, 816 (BM, E, K); Arrondissment de Tempio, Monte Limbardo, 14.vii.1882,
Reverchon 261 (E, JE, K).

CORSICA. Bastia, 27.v.1900, Kugler s.n. (JE); Evisa, 840 m, 3.viii.1932, Aellen
30 (K); Bastélica, 5.vi.1878, Reverchon 5 (E, Fr, K); Montagne de Cagna, Ansteig
zum Monaco-Pass, nahe von Gianuccio, 550 m, 19.vii.1973, Bocquet 15507 (BM).

CULTIVATED. Specimens seen from England (1974) and Holland (Herb. cliff).

H. hircinum subsp. hircinum was obviously well known to 18th century botanists and
gardeners, but it seems to be due merely to chance that Linnaeus was more familiar
with it than with the more widespread subsp. majus. Although it is quite distinct from
subsp. majus in Corsica, the differences in Sardinia are not always so clear.

In gardens, P.-W. Watson’s var. pumilum (which is clearly subsp. hircinum from
his illustration) has been confused with var. minus Aiton (= subsp. cambessedesii)
because Aiton’s variety (q.v.) includes elements of both subspecies. The epithet
minus, however, originated with Dillenius, Hort. Elthamensis: t. 151 f. 182 (1732);
and his illustration is clearly of subsp. hircinum.

4e(62e). H. hircinum subsp. albimontanum (Greuter) N. Robson, stat.
nov.

H. hircinum L., Sp. pl.: 784 (1753) pro parte, quoad loc. Creta cit.

H. hircinum sensu Sibth. & Sm., Fl. Graeca 8: 53, t. 773 (1833); Hayek, Prodr. fi. pen.
Balc. 4 (4): 531 (1925); K. H. Rechinger, Fl. Aegaea: 262 (1943) pro parte, excl. pl.
Rhodenses; N. Robson in Meikle, F/. Cyprus 1: 294 (1977); et auct. Graec.

H. hircinum var. albimontanum Greuter, Prem. coll. OPTIMA, Crete 1975, guide
excurs.: 25 (1977). Type: Crete, Ep. Sfakia, b.Aj. Rumeli, 100 m, 10.vii.1961,
Greuter 3837 (Herb. Greuter., holotype; G, W, Z, isotypes).

Icones: Sibth. & Sm., Fl. graeca 8: t. 773 (1833); Fig. 22D.

Plant c. 0-5—1 m tall, with stems erect. Leaves 30-45 x 12-35 mm, broadly ovate to
ovate-lanceolate, apiculate or obtuse to rounded, margin plane or usually + marked-
ly crisped-undulate, base cordate to cuneate or shortly angustate, goat-scented.
Flowers 28-40 mm in diam. Petals 18-20 x 10 mm. Styles 14-21 mm long. Capsule
5-8 x 5-6 mm.

Greece (Peloponnisos: Mt. Malevo), Crete, Kikladhes (Andhros), Dhodhekanisos
(Samos), Cyprus. Map 31.

GREECE. Peloponnisos: M. Malevo Laconiae prope Platanos, 1050 m, 4—
16.vii.1856, Orphanides 820 (E, JE).

CRETE. Kissamos: Helos (Enneachoria), 400 m, vi.1932, Atchley 1371 (K).
Selinos: in umbrosis vallis pluvii Pelekanotikos prope Vutas, c. 300 m, 2.vi.1942,
K. H. Rechinger f. 13553 (Herb. Huber-Morath). Khania: Ardiini, vii.1943, K. H.
Rechinger f. 15070 (BM); Perivoglia, 1820, Sieber s.n. (E, FR, G-DC, H, JE, K).
Sphakia: Levka Ori, inter Samaria et casam pastoralem Potamos, c. 900 m,
15.vi.1942, K. H. Rechinger f. 13808 (BM, K). Rethymno: Bene, 31.v.1915,
Gandoger 12460 (K). Lassithi: Kera below Lassithi-Plain, 800 m, 12.viii.1971, Jermy
& Brownsey 9031 (BM). Hierapetra: Mt. Aphendhi Kavutsi, 23.vi.1937, Lempberg
548 (K).

KIKLADHES. Andhros: prope Palaeochori, vi.1930, Guiol s.n. (BM).

313

314 NORMAN K. B. ROBSON

DHODHEKANISOS. Samos: Infra pagum Nenedes, c. 100 m, 6.iv.1934, K. H.
Rechinger f. 3783 (BM, K).

CYPRUS. Lapithos, 240 m, 6.v.1955, Merton 2235 (K).

CULTIVATED. Living plants seen in England 1981-1983.

The relatively broad leaves with crisped-undulate margin of typical H. hircinum
subsp. albimontanum appear quite distinct from those of subsp. majus. There are
specimens, however (e.g. Reverchon 7.vii.1883, from Malaxa, Crete), in which these
characters are less pronounced; and true subsp. majus occurs in eastern Crete and in
Rhodes.

Sect. 6. INODORA Stef.
in God. Agr. -les. Fak. Univ. Sofiya 11: 147 (1933) (‘Inodorum’).

Shrub up to 1-5 m tall, deciduous, glabrous, without dark glands; branching lateral. Stems
4-lined, + compressed when young, becoming 2-lined. Leaves opposite, decussate, shortly
petiolate, free, deciduous at basal articulation; lamina entire, with venation pinnate, closed, the
tertiary, densely reticulate; laminar glands punctiform, pale; marginal gland dots pale; ventral
glands absent. Inflorescence 1-c. 30-flowered, branching dichasial/monochasial from 1-3 nodes,
sometimes with subsidiary branches; bracts and bracteoles foliar or reduced, sometimes
sepalline. Flowers stellate, homostylous. Sepals 5, free, persistent, with margin entire or
glandular: veins (3)5-7, reticulately branched; laminar glands linear to punctiform, pale;
marginal glands absent or sessile to shortly stipitate, yellowish. Petals 5, persistent, without
apiculus, margin glandular; laminar glands numerous, pale, linear. Stamen fascicles 5, free,
persistent, each with c. 10 stamens; filaments united very shortly; anthers yellow, gland amber;
pollen type III. Ovary with 3-4 loosely axile placentae, -ovulate; styles 3-4, free; stigma small.
Capsule 3—4-valved, subcoriaceous, longitudinally vittate. Seeds cylindric, not carinate or
winged, without winglike apical expansion; testa linear-reticulate.

BASIC CHROMOSOME NUMBER (x): 10; ploidy 4.
Hasirat: forest and shaded places, 0-670 m.
DistRIBUTION: Turkey, U.S.S.R. (Georgia).

1 species.

1(63). Hypericum xylosteifolium (Spach) N. Robson

in Notes Roy. bot. Gard. Edinb. 27: 185 (1967), in Davis, Fl. Turkey 2: 366, f. 11,
map 20 (1967); [Lancaster], Hilliers’ man. trees & shrubs: 152 (1971); Bean, Trees
& shrubs hardy in Br. Isles 8th ed. 2: 423 (1973). Type: Turkey, ‘Hypericum
cappadocicum frutescens, foetido simile sed inodorum’, Tournefort[Cor. Inst. Rei
Herb.: 18 (1703)] (P, holotype; BM!).

H. inodorum Willd., Sp. pl. 3: 1449 (1802); Choisy, Prodr. monogr. fam. Hypéric.:
39 (1821), in DC., Prodr. syst. nat. veg. 1: 544 (1824); Boissier, Fl. orient. 1: 789
(1867); Dippel, Handb. Laubholzk. 3: 41, f. 24 (1893); Lipsky, Fl. Kavk.: 258
(1899); Woronow in Kusn., N. Busch & Fomin, Fl. Cauc. Crit. 3 (9): 12 (1906); R.
Keller in Engler & Prantl, Nat. Pflanzenfam. 2nd ed. 21: 177 (1925); Stefanoff in
God. Agr.-les. Fak. Univ. Sofiya 10: tt. 2 f. 10, 3f. 11 (1932), ibid. 11: 148 (1933),
ibid. 12: 82 (1934), in Pflanzenareale 4: Karte 2b (1933); Gorshkova in Shishkin &
Bobrov, Fl. U.R.S.S. 15: 217, t. 9f. 3 (1949); Plaisted & Lighty in Nat. Hort. Mag.
38: 130 (1959); Kem.-Nat. in Fl. Gruzii 6: 217 (1960); Grossheim, FI. Kavk., ed. 2,
6: 168, karte 182 (1962). Type: Turkey, ‘Hypericum cappadocicum frutescens,
foetido simile sed inodorum’, Gundelsheimer (B-WILLD, holotype).

Androsaemum xylosteifolium Spach, Hist. nat. vég. Phan. 5: 420 (1836); Jaub. &
Spach, III. pl. orient. 1: 74, t. 38 (1842).

Hypericum ramosissimum Ledeb., FI. ross. 1: 449 (1842). Type: U.S.S.R., Georgia,
‘Guriel in depressis’, Nordmann 396 (H!, lectotype—mihi).

H. ramosissimum [var.] a subuniflorum Ledeb., loc. cit. (1842). Type: U.S.S.R.,

STUDIES IN THE GENUS HYPERICUM L.

Georgia, ‘In umbrosis sylvis per Imereti, Mingreli et Guriel frequens’, vi.1830,
Nordmann (H!, holotype).

H. ramosissimum [var.] B intermedium Ledeb., loc. cit. (1842). Type: U.S.S.R.,
Georgia, Imeretia, Nordmann 131 (H!, holotype).

H. ramosissimum [var.] y multiflorum Ledeb., loc. cit. (1842). Type: U.S.S.R.,
Georgia, ‘Guriel in depressis’, Nordmann 396 (H!, holotype).

H. rariflorum Steven ex Rupr., Fl. Cauc.: 298 (1869), in synon.

H. inodorum var. subuniflorum (Ledeb.) R. Keller ex Somm. & Lev. in Acta Horti
Petrop. 16: 93 (1900).

H. inodorum var. intermedium (Ledeb.) R. Keller ex Somm. & Lev., loc. cit. (1900).

H. inodorum var. multiflorum (Ledeb.) R. Keller ex Somm. & Lev., loc. cit. (1900).

H. inodorum var. sommieri R. Keller ex Somm. & Lev., loc. cit. (1900). Type:
U.S.S.R., Georgia, Namachvani, 23.vii.1890, Sommier & Levier s.n. (FI, holo-
type).

H. inodorum var. glandulosum Stef. in Kew Bull. 1931: 29 (1931). Type: U.S.S.R.,
Georgia (?), ‘Armenia’, Szovitz s.n. (K!, lectotype—mihi; BM!, UPS!); U.S.S.R.,
Georgia, ‘Caucasus’, 1838, Ledebour s.n. (K!, syntype).

H. inodorum var. integrisepalum Stef., loc. cit. (1931). Type: Turkey, Rize, ‘Falaises
entre Rhizé et Mapavré (Lazistan)’, Balansa 1377 (K!, holotype; BM!, G!, JE!).

Icones: Jaub. & Spach, IIl. pl. orient. 1: t. 38 (1842); Fig. 23.

Shrub up to 1-5 m tall, deciduous, with branches erect to spreading. Stems green to
pinkish, 4-lined and ancipitous when young, soon 2-lined, eventually terete; inter-
nodes 10-40 mm long, shorter than leaves; bark fissuring. Leaves petiolate, with
petiole 4-10 mm long; lamina (15—)20—73 x 8-26 mm, oblong or elliptic to lanceolate
or oblong-ovate, rounded or truncate to apiculate or obtuse, margin plane, base
rounded to broadly cuneate, paler beneath, not glaucous, chartaceous; venation:
(3)4-6 pairs of ascending main lateral veins, not always distinct from the midrib
branches, with tertiary reticulation dense, conspicuous; laminar glands small; intra-
marginal glands dense. Inflorescence 1—-7(-c. 25)-flowered, terminal, broadly pyra-
midal to subcorymbiform, from 1—2(3) nodes, sometimes with 1-2 pairs of subsidiary
branches; pedicels 4~-7(-10) mm long; bracts foliar or reduced or linear, and then
sometimes with sessile yellowish marginal glands, deciduous. Flowers 15-30 mm in
diam., stellate; buds ellipsoid, rounded. Sepals 4-5-12 x 1-3-2-5 mm, open (not
imbricate), subequal to unequal, very narrowly oblong or narrowly elliptic to
oblanceolate-spathulate, acute, margin entire or with sessile to shortly stipitate
yellowish flat-topped glands; (3)5—7-veined reticulately branched, midrib distinct;
laminar glands linear to punctiform. Petals golden yellow, not tinged red, spreading,
8-15 x 1-5-4-5 mm, c. 1-3-2 X sepals, narrowly oblanceolate to narrowly obovate,
without apiculus, margin with spaced sessile dark red glands. Stamen fascicles each
with 10-11 stamens, longest (8—)12-15 mm long, about equalling petals; anthers pale
orange. Ovary 2:5-3:5 x 2-3 mm, ovoid-conic to subglobose; styles 7-9 mm long,
2-5-3 X ovary, free, erect to suberect, narrowly divergent in upper 1/3. Capsule
5-7 x 5-7 mm, subglobose to globose. Seeds pale brown, c. 0-7 mm long, cylindric,
not carinate, shallowly linear-foveolate. 2n = 40.

Deciduous forest, shady banks and cliffs, 0-670 m.

Turkey (Giresun, Trabzon, Rize, Coruh), U.S.S.R. (Georgia). Map 33.

TURKEY. Giresun: 25 km S. of Giresun, 500 m, 7.vii.1969, Sorger 69-23-28 (BM,
Hb. Sorger). Trabzon: Of district, Siirmene to Of, coast, 11.vii.1958, Huber-Morath
14973 (Hb. Hub.-Mor.). Rize, 100 m, 30.vi.1931, Gérz 683 (BM). Coruh: Hopa,
near sea level, 22.vi.1957, Davis 29871 (BM, E, K).

GEORGIA. Adzharia: Chakvi near Batumi, 30.vii.1979, Lancaster s.n. (BM).
Imeretia: In sylva Sagori prope Kutais, vi.1931, Woronow IX 432 (K). Abkhasia:
Sukhumi, 20 m, 11.vi.1959, Davis 33655 (E, K).

ENGLAND (naturalised). North Yorkshire: Ripon, near Monkton Moor, 1979,
Houseman s.n. (BM). Lancashire: Silverdale, Elmslack Well, Eaves Wood,
viii. 1978, Jepson s.n. (BM).

CULTIVATED. Specimens seen from England, Scotland, and Sweden.

ca
+
'

315

316 NORMAN K. B. ROBSON

Fig. 23H. xylosteifolium: (a) habit; (b) leaf section; (c) flower bud (sepals entire); (d) flower bud (sepal
margin glandular); (e) sepal; (f) petal; (g) stamen fascicle; (h) anthers; (i) ovary; (j) capsule (a x 1; b x
4; c,d, j X 6;e-g,i X 10;h Xx 40). (a, b, d-i) Gérz 690: (c) Davis & Hedge D. 29871; (j) Lancaster s.n.

STUDIES IN THE GENUS HYPERICUM L.

Map 33 Sect. '6. Inodora: il H. ahicarendhs e,O i Kavkasa, edn: additional records); Sect. 3.
Ascyreia: 14. H. calycinum O (part; see also Map 13).

H. xylosteifolium has the appearance of a small-flowered H. monogynum, but taller
and with persistent petals and stamens and fewer, free styles. It forms small thickets
with spreading leafy branches and relatively inconspicuous flowers. There is some
evidence to suggest that it flowers more freely in the (wetter) west of Britain than in
the (drier) east.

None of the described varieties of H. xylosteifolium appear to be distinct. Those
of Ledebour merely reflect variation in luxuriance of growth. On the other hand, the
forms with (respectively) larger entire sepals (and bracts) and smaller gland-fringed
ones represent the extremes of a west—east cline. All the Georgian plants that I have
seen have gland-fringed sepals and bracts, whereas in most of the Turkish specimens
they are entire. There is, however, an area in eastern Turkey where there occur
transitional forms that prevent the recognition of two varieties. The Tournefort
specimen (type collection) at BM is of this intermediate form, having sepals entire or
with one or two marginal glands.

Sect. 6a. UMBRACULOIDES N. Robson

Shrub, deciduous, glabrous, without dark glands; branching lateral. Stems 2-lined (?) and
ancipitous when young, soon terete. Leaves opposite, decussate, sessile, free, deciduous at basal
articulation; lamina entire, with venation pinnate, closed, the tertiary densely reticulate;
laminar glands punctiform, pale; marginal gland dots pale; ventral glands dense. Inflorescence
3—c.16-flowered, branching dichasial/monochasial from uppermost node, without subsidiary
branches; bracts and bracteoles reduced ?, caducous. Flowers stellate, homostylous. Sepals 5,
free, persistent, with margin entire; veins c. 13-15, unbranched; laminar glands linear, pale;
marginal glands absent. Petals 5, deciduous, apiculus subrounded, margin entire; laminar glands
numerous, pale, linear to interrupted. Stamen fascicles 5, united 2 + 2 + 1, persistent, with total
of 45-50 stamens; filaments united very shortly; anthers yellow, gland amber; pollen type ?

318

NORMAN K. B. ROBSON

Ovary with 3 axile (?) placentae, ? ovulate; styles 3, free (or partially coherent at first ?); stigma
small. Capsule not seen.

BASIC CHROMOSOME NUMBER (x): ?; ploidy ?

HasiratT: unknown.

DISTRIBUTION: Mexico (Oaxaca).

1 species.

fl

+

1(64). Hypericum umbraculoides N. Robson, sp. nov.

H. monogyno affinis, sed sepalis latiuscule oblongo-ellipticis, staminorum fasciculis
2 + 2 + 1 coadunatis post anthesin persistens staminibus in quoque fasciculo
paucioribus, stylis tribus liberis vel basin versus breviter coalitis, differt. Type:
Mexico, Oaxaca, San Pedro Alto, x.1842, Liebmann 3037 (C!, holotype &
isotype).

Icon: Fig. 24.

Shrub, deciduous. Stems reddish, 2-lined (?) and ancipitous when young, soon
terete; internodes 10-20 mm long, shorter than leaves; bark grey, cortex fissuring.
Leaves sessile; lamina 30-58 x 17-30 mm, oblong to elliptic-oblong, subapiculate to
rounded, margin plane, base cordate-amplexicaul, paler beneath, not glaucous,
thinly chartaceous; venation: 7-9 pairs main laterals and midrib branches (not
distinct), with tertiary reticulation dense, conspicuous, + prominent beneath;
laminar glands varying in size; intramarginal glands dense. Inflorescence 3-c.16-
flowered, terminal, corymbiform to pseudumbellate, from apical node only; pedicels
6-17 mm long; bracts not seen, caducous. Flowers c. 22-30 mm in diam., stellate;
buds narrowly cylindric-ellipsoid, rounded. Sepals 3-4-5 x 2-5-3-5 mm, imbricate,
unequal, broadly oblong to ovate-oblong, rounded, margin entire; c. 13-15-veined,
with veins unbranched, midrib prominent; laminar glands linear. Petals golden ?
yellow, not tinged red, spreading to reflexed, 14-17 x 7-8 mm, 4-5 sepals,
oblanceolate, with apiculus lateral, subrounded, almost lobiform, margin entire,
eglandular. Stamen fascicles totalling 45-50 stamens, longest 10-12 mm long, c. 0-7 x
petals; anthers yellow. Ovary 4-5 x 1-5-2 mm, narrowly pyramidal; styles 9-10 mm
long, c. 2 X ovary, possibly shortly coherent at the base, eventually (or always ?)
free, gradually outcurved. Capsule and seeds unknown. 2n = ?

Habitat unknown.

Mexico (Oaxaca). Map 34.
MEXICO. Oaxaca: San Pedro Alto, x.1842, Liebmann 3037 (= Hypericaceae 35)

(C).

Map 34 Sect. 6a. Umbraculoides: 1. H. umbraculoides @.

STUDIES IN THE GENUS HYPERICUM L. 319

f

Fig. 24 H. umbraculoides: (a) habit; (b) leaf section; (c) flower bud; (d) sepal; (e) petal; (f) stamen
fasicle; (g) anthers; (h) ovary (a x 1; b x 4;c x 6;d,e, f, h X 8; g x 20). All Liebmann 3037.

320 NORMAN K. B. ROBSON

H. umbraculoides which has as yet been collected only once and appears to be
restricted to a small area on the south side of the Cordillera in Oaxaca State, is clearly
an ancient relict, its nearest relative (H. monogynum, sect. 3) being in China. Its
leaf-shape and glandularity, pseudo-umbellate inflorescence and general floral
morphology all indicate this relationship, although it is more advanced in having ‘3’
persistent stamen fascicles and 3 styles that are not more than shortly coherent at the
base.

3. Acknowledgements

Iam particularly grateful to Roy Lancaster, who has encouraged and helped me by collecting plants, seeds
and specimens in Nepal and China, by photographing plants and herbarium specimens in China, and by
providing information regarding his own and other collections cultivated in the British Isles and the
Netherlands. Tony Schilling has also helped me greatly by collecting specimens and seed in Nepal, and by
assembling and growing a wide range of Hypericum species and hybrids at Wakehurst Place, Ardingly,
West Sussex. For looking after my own study collection at the Chelsea Physic Garden I must thank Allen
Paterson and Miss Virginia Nightingale. For hybridising members of sect. Ascyreia for me, I am grateful to
Julie Westfold (Chelsea Physic Garden), Peter Dummer (Hilliers’ Nurseries), and Donald Walker. I am
also grateful to others who provided information, specimens, or living material, especially Prof. Li Xiwen
(Kunming), Dr Tseng-Chieng Huang (Taipei), David Long (Edinburgh) , David McClintock (Platt, Kent),
Dr Bassett Maguire (New York), Peter Nicholas (Hidcote Manor, Gloucestershire), and Keith Rushforth
(The Hillier Arboretum). I am indebted for the loan of photographs to Geoffrey Herklots, Sir Colville
Barclay, and Mrs Frances Lesueur, as well as Roy Lancaster; and for the loan of specimens to the directors
of the following herbaria: Arnold Arboretum, Harvard University (A); Herbarium Jutlandicum, Aarhus
(AAU); Jardin botanique national de Belgique, Meise (BR); Royal Botanic Garden, Edinburgh (E);
Botanic Museum, University of Helsinki (H); South China Institute of Botany, Guangzhou (IBSC); Royal
Botanic Gardens, Kew (K); Kunming Institute of Botany, Kunming (KUN); Museum National d’Histoire
Naturelle, Paris (P); Department of Botany, Sichuan University, Chengdu (SZ). I must also thank Dr
Mary Gibby for making numerous chromosome counts on the material at Chelsea, Miss Ann Lum for
translating data on Chinese herbarium labels, and Miss Jane Bellamy, Dr Margaret Collinson, and Miss
Marian Short for plotting distributions.

With regard to the production of this paper, I am deeply indebted to Mrs Margaret Tebbs for drawing the
plates and figures, and for making some of my sketches into text figures. I am also grateful to Miss Sylvia
Gould and my wife, Mrs Eve Robson, for improving the rest of my illustrations, to Miss Rachel Hampshire
and Bob Press for drawing maps, to Mrs Ronaq Khan for printing data for the figures, and to Miss Loveday
Hosking for typing most of the manuscript and Mrs Shirley Burton and Miss Maria Duda for additional
typing. Finally, I would thank various colleagues for helpful discussions and suggestions, in particular Mr
Arthur Chater for reading the typescript.

4. References

Aiton, W. 1789. Hortus kewensis 3. London.

Balfour, I. B. 1888. Botany of Socotra. Trans. R. Soc. Edinb. 31: i-Ixxv, 1-446.

Britten, J. 1915. Note on Hypericum calycinum L. J. Bot., Lond. 53: 68.

Cambessédes, J. 1827. Enumeratio plantarum quas in insulis Balearibus collegit. Paris.

Choisy, J. D. 1821. Prodromus d’une monographie de la famille des Hypéricacées. Geneva & Paris.

Coulter, J. M. 1886. Revision of North America Hypericaceae. —I. Bot. Gaz. 11: 78-88.

Dillenius, J. J. 1732. Hortus elthamensis. London.

Don, D. 1825. Prodromus florae nepalensis. London.

Dyer, W. T. T. 1874. Hypericineae. In J. D. Hooke1, Flora of British India 1: 252-258.

Hardwicke, J. 1801. Enumeration of plants noticed in the preceding tour between Hurdwar and Sirinagur,
in the months of April and May, 1796. Asiatick Researches 6: 369.

Hedberg, O. 1957. Afroalpine vascular plants. A taxonomic revision. Symb. bot. Upsal. 15 (1): 1409.

Lace, H. J. 1961. List of trees, shrubs, herbs and principal climbers, etc. recorded from Burma, 3rd ed.
Rangoon.

Melville, R. 1981. Vicarious plant distributions and the palaeogeography of the Pacific region. In G. Nelson
& D.E. Rosen (Eds), Vicariance biogeography: a critique. New York.

Moggi, G. & Pisacchi, A. 1967. Adumbratio florae aethiopicae. 14. Hypericaceae. Webbia 22: 233-289.

Rehder, A. 1940. Manual of cultivated trees and shrubs. 2nd ed. New York.

STUDIES IN THE GENUS HYPERICUM L. 321

Robson, N. K. B. 1968a. No. 49. Guttiferae. Jn K. H. Rechinger (Ed.), Flora iranica. Graz.

— 1968b. Guttiferae (Clusiaceae). In T. G. Tutin et al. (Eds), Flora europaea 2: 261-269. Cambridge.

— 1970. Shrubby Asiatic Hypericum species in cultivation. J. R. Hort. Soc. 95: 482-497.

— 1973. No. 32. Guttiferae. In E. Nasir & S. I. Ali (Eds), Flora of West Pakistan. Rawalpindi.

— 1977a. Studies in the genus Hypericum L. (Guttiferae). 1. Infrageneric classification. Bull. Br. Mus.
nat. Hist. (Bot.) 5: 291-355.

— 1977b. Notes on some Nepalese and Indian Hypericum. J. Jap. Bot. 52: 276-288.

— 1979. Parallel evolution in Africa and Mascarene Hypericum. Kew Bull. 33: 571-584.

— 1981. Studies in the genus Hypericum L. (Guttiferae). 2. Characters of the genus. Bull. Br. Mus. nat.
Hist. (Bot.) 8: 55-226.

— & Long, D. G. 1983. Notes relating to the flora of Bhutan: VII. Notes on Hypericum L. Notes R. bot.
Gdn Edinb. 41: 133-139.

Salisbury, E. J. 1963. Fertile seed production and self-incompatibility of Hypericum calycinum in England.
Watsonia 5: 368-376.

Sauvage, C. 1958. Hypericum hircinum et les espéces affines dans le basin méditerraneén occidental. Bull.
Soc. Sci. nat. phys. Maroc 38: 123-130.

Singhal, V. K., Gill, B.S. & Bir, S.S. 1980. Hypericaceae. Jn A. Love (Ed.), Chromosome number reports
LXIX. Taxon 29: 703-730.

Spirlet, M. 1967. Etude taxinomique des épidermes foliaires des Hypéricaceés et des Guttiferaceés du
basin du fleuve congo. Bull. Inst. fr. Afr. noire 29 (1), A.: 5-91.

Steenis, C. G. G. J. van 1979. Plant-geography of east Malesia. Bot. J. Linn. Soc. 79: 97-178.

Stern, W. L. 1978. Comparative anatomy and systematics of woody Saxifragaceae. Hydrangea. Bot. J.
Linn. Soc. 76: 83-113.

Styer, C. H. & Stern, W. L. 1979. Comparative anatomy and systematics of woody Saxifragaceae. Deutzia.
Bot. J. Linn. Soc. 79: 291-319.

Sugiura, J. 1936. A list of chromosome numbers in Angiospermous plants II. Proc. imp. Acad. Japan 12:
144-146.

Turrill, W. B. 1962. Hypericum elatum. Curtis’s bot. Mag. 173: t. 376.

Willdenow, C. L. 1802. Species plantarum 3. Berlin.

Systematic Index

Accepted names are in roman and synonyms in italic; new names are in bold, as are principal references.

An asterisk (*) denotes a figure.

Androsaemum calycinum (L.) C.
Pres] ex Steudel (=25) 228
cambessedesii Nyman (=62b)
311
hircinum (L.) Spach (+62) 307
officinale All. (=61) 301
parviflorum Spach (=61x) 305
pyramidale forma grandifolia
Spach (=61x) 305
pyramidale Spach (=61x) 305
x urberuagae P. & S. Dupont
(=61x) 305
vulgare Gaertner (=61) 301
webbianum Spach (=59) 298
xylosteifolium Spach (=63)
314
Ascyron coriaceum Moench (=25)
228
Ascyrum calycinum (L.) Poiret
(=25) 228

Campylosporus angustifolius (Lam.)
Spach (=3a) 190

madagascariensis Spach (=4)
191
reticulatus Spach (=3b) 190

Eremanthe calycina (L.) Spach

(=25) 228

cordifolia (Choisy) K. Koch
(=13) 213

patula (Thunb. ex Murray) K.
Koch (=42) 265

venosa (Lam.) K. Koch (=25)
228

Hypericum Sect. 5. Androsaemum

(Duhamel) Godron 175*, 297

Sect. 3. Ascyreia Choisy 170*,
171*, 172*, 206

Sect. 1. Campylosporus (Spach)
R. Keller, 166*, 167*, 168*,
178

Sect. 6. Inodora Stef. 314

Sect. 2. Psorophytum (Spach)
Nyman 202

Sect. 4. Takasagoya
(Y. Kimura) N. Robson 174*,
288

Sect. 6a. Umbraculoides N.
Robson 176, 317

acmosepalum N. Robson (34)
245, 246*

acutisepalum Hayata (=57a)
295

acutum Wallich (=14) 214

addingtonii N. Robson (38)
248*, 251, 252*

affine Steudel ex Oliver (=7)
194

aitchisonii J. R. Drumm. ex R.
Keller (=24) 226

androsaemum L. (61) 301, 303*

androsaemum sensu auct.
(=61x) 305

androsaemum var. aureum hott.
ex Nicholson (=61) 301

anglicum Bertol. (=61x) 305

angustifolium Lam. (=3a) 190

Hypericum — cont.

argyi H. Léveillé & Vaniot
(=42) 265

ascyron sensu H. Léveillé (=47)
275

ascyron sensu Miller (=25) 228

augustinii N. Robson (18) 217*,
219, 220*

aureum Lour. (=27) 231

auritum Moon (=12) 210

bacciferum Lam. (=61) 301

bacciforme Bubani (=61) 301

balearicum L. (11) 203, 204*,
205*

balearicum var. ochroleucum
R.A.W. Herm. (=11) 203

balfourii N. Robson (5) 191,
193*

beanii ‘Eastleigh Gold’ (=47xx)
277

beanii ‘Gold Cup’ (=38x) 253

beanii N. Robson (51) 282, 283*

bellum H. L. Li (46) 273

bellum subsp. bellum (46b)
2D ans)

bellum subsp. latisepalum N.
Robson (46a) 272*, 274

bequaertii De Wild (1) 180,
181*, 183*

bracteatum Buch.-Ham. ex D.
Don (+13) 213

calcaratum hott. (=53) 286

calycinum L. (25) 228, 229*,
S17

calycinum vat. acutifolium
Choisy (=25) 228

calycinum var. § sensu Choisy
(=25) 228

cambessedesii (Nyman) Cosson
ex Marés & Vigin. (=62b)
311

canariense Brouss. ex Webb &
Berth. (=59) 298

canariense sensu Cambessédes
(=62b) 311

cernuum Roxb. ex D. Don
(=24) 226

cernuum sensu Dyer pro parte
(=37) 249

cernuum var. lanceolatum R.
Keller (=24) 226

cernuum var. typicum R. Keller
(=24) 226

chinense L. (=27) 231

chinense subsp. latifolium
Kuntze (=27) 232

chinense subsp. obtusifolium
Kuntze (=27) 232

chinense subsp. salicifolium
(Sieb. & Zucc.) Kuntze (=27)
232

chinense var. salicifolium (Sieb.
& Zucc.) Choisy (=27) 232

chinense var. B sensu Hooker &
Arnott (=27) 231

choisianum Wallich ex N.

NORMAN K. B. ROBSON

Robson (45) 271, 272*

cohaerens N. Robson (28) 234*,
235

conrauanum Engler (=9) 199

cordifolium Choisy (13) 213,
214*

cordifolium sensu Dyer, pro
parte (=14) 215

coriaceum Blume (=33) 243

coriaceum Royle (=37) 249

curvisepalum N. Robson (50)
277* , 280*, 281

x cyathiflorum ‘Gold Cup’ (38x,
C.V.) 254

x cyathiflorum N. Robson (38x :
38 x 39?) 253

x dummeri N. Robson (53x :
53 X 25) 287

dyeri Rehder (37) 249, 250*

dyeri sensu hort. (=48) 278

x ‘Eastleigh Gold’ (47xx :
47 X 2?) 277

elatum Aiton (=61x) 304

elatum ‘Elstead Variety’
(=61xa) 307

elatum sensu auct. (=59) 298

elatum sensu Turrill pro parte
(=61xa) 307

erectum Solander ex R. Br.
(=59) 298

floridum Salisb. (=61) 301

foliosum Aiton (60) 300, 303*

formosanum Maxim. (54) 289,
290*, 291*

formosanum sensu auct. pro
parte (=58) 296

formosanum sensu Hayata
(=57a) 295

formosanum sensu Susuki
(=57b) 296

forrestii (Chittenden) N.
Robson (53) 283*, 286

gaitii Haines (23) 212*, 225

garrettii Craib (=39) 255

garrettii sensu Craib pro parte
(=41a) 261

garrettii var. ovatum Craib
(=41a) 261

geminiflorum Hemsley (57) 292

geminiflorum subsp.
geminiflorum (57a) 293*, 294*

geminiflorum subsp.
simplicistylum (Hayata) N.
Robson (57b) 293*, 295

geminiflorum var.
simplicistylum (Hayata) N.
Robson (=57b) 296

giraldii R. Keller (=30b) 239

gnidiifolium A. Rich. (10) 200*,
201

govanianum Wallich (=37) 249

gracilipes Stapf ex C. Fischer
(22) 217*, 224

grandifolium Choisy (59) 298,
299*

grandiflorum Salisb. (=25) 228

grandiflorum sensu auct. (=59)

grandiflorum sensu hort. pro
parte (=47) 275

griffithii Hook. f. Thomson ex
Dyer (26) 227*, 230

gumbletonii Lavallée (=42) 265

henryi H. Léveillé & Vaniot
(41) 260

henryi sensu hott. pro parte
(=34) 245

henryi sensu hort. pro parte
(=52) 284

henryi subsp. hancockii N.
Robson (41a) 260*, 261,
262*, 264*

henryi subsp. henryi (41b) 260*,
263

henryi subsp. uraloides
(Rehder) N. Robson (41c)
263, 264*

‘Hidcote’ (=38xx) 254

‘Hidcote Gold’ (=38xx) 254

x ‘Hidcote’ (38xx : 38 Xx 25)
254

hircinum L. (62) 307

hircinum L. pro parte (=62e)
313

hircinum sensu auct. (=62a) 310

hircinum sensu auct. (=62d) 312

hircinum sensu auct. (=62e) 313

hircinum subsp. albimontanum
(Greuter) N. Robson, (62e)
299* , 308*, 313

hircinum subsp. cambessedesii
(Nyman) Sauvage (62b) 299*,
308*, 311

hircinum subsp. hircinum (62d)
299* , 308*, 312

hircinum subsp. hircinum sensu
Sauvage (=62a) 310

hircinum subsp. majus (Aiton)
N. Robson (62a) 299*, 308*,
310

hircinum subsp. metroi (Maire
& Sauvage) Sauvage (62c)
299*, 312

hircinum subsp. obtusifolium
(Choisy) Sauvage (=62d) 313

hircinum var. albimontanum
Greuter (=62e) 313

hircinum var. cambessedesii
(Nyman) Ramos (=62b) 312

hircinum var. majus Aiton
(=62a) 310

hircinum var. minus Aiton pro
parte (=62b) 311

hircinum var. minus Aiton pro
parte (=62d) 312

hircinum var. obtusifolium
Choisy (=62d) 312

hircinum var. pumilum P. W.
Watson (=62d) 312

hookerianum ‘Buttercup’ hort.
(=43) 268

hookerianum “Charles Rogers’

STUDIES IN THE GENUS HYPERICUM L.

(39a) 257*, 258

hookerianum ‘Gold Cup’
(=38x) 253

hookerianum ‘Hidcote’
(=38xx) 254

hookerianum ‘Rogersii’ (=39a)
258

hookerianum ‘Rowallane’
(=22a) 245

hookerianum sensu Buysmann
(=33) 243

hookerianum sensu hott. pro
parte (=21) 223

hookerianum sensu Rehder
(=53) 286

hookerianum sensu R. Keller
(=42) 265

hookerianum sensu R. Keller
pro parte (=35) 247

hookerianum var. leschenaultii
(Choisy) Dyer (=33) 243

hookerianum var. leschenaultii
sensu Dyer (=45) 271

hookerianum var. lineare
Banerji (=14) 215

hookerianum var. rogersii hort.
(=39a) 258

hookerianum Wight & Arnott
(39) 255, 257*

x inodorum ‘Elstead’ (61xa) 307

x inodorum Miller (61x : 61 x
62) 304

inodorum var. glandulosum
Stef. (=63) 315

inodorum vat. integrisepalum
Stef. (=63) 315

inodorum var. intermedium
(Ledeb.) R. Keller ex Somm.
& Lev. (=63) 315

inodorum var. multiflorum
(Ledeb.) R. Keller ex Somm.
& Lev. (=63) 315

inodorum var. sommieri R.
Keller ex Somm. & Lev.

(=63) 315

inodorum var. subuniflorum
(Ledeb.) R. Keller ex Somm.
& Lev. (=63) 315

inodorum Willd. (=63) 314

javanicum Blume (=33) 243

kalmianum Vahl (=2b) 185

kalmii Forssk. (=2b) 185

keniense Schweinf. (=2a) 184

keniense sensu Mildbr. (=1) 180

kouytchense H. Léveillé (47)
215,201

kouytchense sensu hort. pro
parte (=34) 245

kouytchense sensu Milne-Redh.
(=36) 248

kouytchense x calycinum (47x:
47 x 25) 276

kushakuense R. Keller (=58) 296

lacei N. Robson (40) 258, 259*,
260*

lackeyi Elmer (=57a) 295

lagarocladum N. Robson (35)
247, 248*

lancasteri N. Robson (49) 277*,
279, 280*

lanceolatum forma
angustifolium (Lam.) H.
Perrier (=3a) 190

lanceolatum Lam. (3) 188

lanceolatum sensu auct. (=2b)
185

lanceolatum sensu auct. (=4) 191

lanceolatum sensu Balf. f. (=6a)
193

lanceolatum subsp.
angustifolium (Lam.) N.
Robson (3a) 189*, 190

lanceolatum subsp. lanceolatum
(3b) 189*, 190

lanuriense De Wild. (=2b) 185

‘Lawrence Johnston’ (=38x)
253

leschenaultii Choisy (33) 242*

leschenaultii sensu hort. pro
parte (=38) 251

leschenaultii var. coriaceum
(Blume) Hochr. (=33) 243

leschenaultii var. typicum forma
elatius Hochr. (=33) 243

leschenaultii var. typicum
Hochr. (=33) 243

leucoptychodes Steudel ex A.
Rich. (=2b) 185

lobbii N. Robson (21) 217*, 223

loheri Merrill (=57a) 295

longifolium sensu H. Léveillé
(=47) 275

longistylum Oliver (30) 238

longistylum subsp. giraldii (R.
Keller) N. Robson (30b)
237-239

longistylum subsp. longistylum
(30a) 237*, 238

longistylum var. giraldii (R.
Keller) Pampanini (=30b) 239

longistylum var. silvestrii
Pampanini (=30a) 238

lungusum Buch.-Ham. ex D.
Don (=13) 213

lysimachioides sensu hott.
(=48) 278

lysimachioides Wallich ex Dyer
(=37) 249

maclarenii N. Robson (44) 270,
Paifife

madagascariense (Spach)
Steudel (4) 189*, 191

metroi Maire & Sauvage (=62c)
312

minus sensu hort. (=62b) 311

monogynum L. (27) 231, 233*,
234*

monogynum vat. salicifolium
(Sieb. & Zucc.) André (=27)
232

X moserianum Luquet ex André
(42x : 42 x 25) 266

323

X moserianum forma tricolor
(Maumené) Rehder (42xa)
268

X moserianum n.m. tricolor:
(Maumené) N. Robson
(42xa) 268

X moserianum ‘Tricolor’ (42xa)
268

X moserianum vat. tricolor
Maumené (=42xa) 268

multifiorum sensu hort. (=61x)
305

mysorense sensu auct. (=12)
210

mysorense sensu Balf. f. (=5)
192

mysurense sensu Moggi &
Pisacchi (=5) 192

mysurense Wallich ex Wight &
Arnott (12) 210, 211*, 212*

nakamurai (Masamune) N.
Robson (55) 289, 291*

nepalense K. Koch (=43) 268

nervosum Choisy (=33) 243

norysca Steudel (=12) 210

oblongifolium Choisy (24) 226,
227*

oblongifolium sensu Hook. f.
(=21) 223

oblongifolium sensu hort. pro
parte (=34) 245

oblongifolium sensu Wallich
(=39) 255

oblongifolium sensu Wallich pro
parte (=45) 271

pachyphyllum Collett &
Hemsley (17) 217*, 219

patulum forma forrestii
(Chittenden) Rehder (=53)
286

patulum ‘Gold Cup’ (=38x) 253

patulum ‘Hidcote’ (=38xx) 254

patulum ‘Hidcote Variety’
(=38xx) 254

patulum ‘Laplace’ (=47) 275

patulum ‘Rothschild’s Form’
(=53) 286

patulum sensu D. Don (=39)
255

patulum sensu H. Léveillé pro
parte (=47) 275

patulum sensu Matsumura &
Hayata (=54) 289

patulum sensu N. Robson pro
parte (=41) 260

patulum sensu N. Robson pro
parte (=41b) 263

patulum sensu R. Keller (=52)
283

patulum sensu Wallich (=43)
268

patulum subsp. hookerianum
(Wight & Arnott) Kuntze
(=39) 255

patulum subsp. hookerianum
var. tenuicaule (Hook. f. &

Hypericum — cont.

Thomson ex Dyer) Kuntze
(=20) 222

patulum subsp. leschenaultii
(Choisy) Kuntze (=33) 243

patulum subsp. normale Kuntze
(=42) 265

patulum subsp. variabile Kuntze
(=33) 243

patulum ‘Summergold’ (=47) 275

patulum ‘Sungold’ (=47) 275

patulum Thunb. ex Murray (42)
260*, 265

patulum var. attenuatum Choisy
(=43) 268

patulum var. forrestii Chittenden
(=53) 286

patulum vat. grandiflorum hott.
(=47) 275

patulum var. henryi Veitch ex
Bean (=51) 282

patulum var. henryi sensu hott.
pro parte (=34) 245

patulum var. henryi sensu hott.
pro parte (=53) 284

patulum var. henryi sensu
Rehder (=52) 284

patulum var. oblongifolium
(Choisy) Koehne (=24) 226

patulum var. oblongifolium
sensu hort. pro parte (=34)
245

patulum var. oblongifolium
sensu Koehne (=39) 255

patulum var. uralum (Buch.
-Ham. ex D. Don) Koehne
(=43) 268

penduliflorum hort. (=47) 275

penticosia Commerson ex Lam.
(=3b) 190

persistens ‘Elatum’ (=61x) 305

persistens ‘Elstead’ (=61xa) 307

persistens 1. F. Schneider (=61x)
305

podocarpoides N. Robson (14)
214*, 216*

prattii Hemsley (29) 236, 237*

prattii sensu Rehder pro parte
(=27) 232

pseudohenryi N. Robson (52)
283* , 285*

pseudohenryi sensu N. Robson
(=51) 282

pseudo-patulum sensu hott.
(=42) 265

pustulosum R. Keller (=57a?)
29

quartinianum A. Rich. (7) 194,
195*, 197*

quartinianum sensu auct. (=9)
199

quartinianum vat. roeperianum
(W. G. Schimper ex A.
Rich.) Engler (=9) 199

ramosissimum K. Koch (=43)
268

NORMAN K. B. ROBSON

ramosissimum Ledeb. (=63)
314

ramosissimum vat. intermedium
Ledeb. (=63) 315

ramosissimum var. multiflorum
Ledeb. (=63) 315

ramosissimum Vat.
subuniflorum Ledeb. (=63)
314

rariflorum Steven ex Rupr.
(=63) 315

reptans Hook. f. & Thomson ex
Dyer (16) 217*, 218

revolutum sensu auct. (=3b) 190

revolutum subsp. keniense
(Schweinf.) N. Robson (2a)
183*, 184

revolutum subsp. revolutum
(2b) 185, 187*

revolutum Vahl (2) 182

riparium A. Chev. (=9) 199

roeperianum subsp. gnidiifolium
(A. Rich.) Moggi & Pisacchi
(=10) 202

roeperianum vat. schimperi
(Hochst. ex A. Rich.) Moggi
& Pisacchi (=9) 199

roeperianum W. G. Schimper
ex A. Rich. (9) 198, 200*

rogersii sensu hort. (=39a) 258

x ‘Rowallane’ (33x : 33 x 39a)
244

ruwenzoriense De Wild. (2a) 184

salicifolium Sieb. & Zucc.
(=27) 231

schimperi Hochst. ex A. Rich.
(=9) 198

schimperi var. angustisepalum
Engler (=9) 199

schimperi var. huillense Hiern
(=9) 199

schimperi var. y sensu Oliver
(=9) 199

senkakuinsulare Hatusima (56)
292, 294*

sherriffii N. Robson (15) 217*

siamense N. Robson (32) 240,
241*, 242*

simplicistylum Hayata (=57b)
295

socotranum Good (6) 192

socotranum sensu Moggi &
Pisacchi (=5) 192

socotranum subsp. smithii N.
Robson (6b) 193*, 194

socotranum subsp. socotranum
(6a) 193*

speciosum Wallich (=24) 226

sp. sensu Hardwicke (=24) 226

sp. sensu Rehder (=36) 248

sp. sensu Smitinand (=32) 240

stellatum N. Robson (48) 277*,
278

stellatum sensu N. Robson pro
parte (=44) 270

subalatum Hayata (58) 293*, 296

subsessile N. Robson (31) 239,
246*

synstylum N. Robson (8) 197*

tenuicaule Hook. f. & Thomson
ex Dyer (20) 222*

triflorum Blume (=33) 243

triflorum var. angustatum
Blume (=33) 243

trinervium Hemsley (=57a) 294

ulugurense Engler (7) 194

umbraculoides N. Robson (64)
176, 318*, 319*

uraloides Rehder (=41c) 263

uralense Lavallée (=43) 268

uralum Buch.-Ham. ex D. Don
(43) 264*, 268, 269*

uralum sensu Hance (=42) 265

x urberuagae P. & S. Dupont
(=61x) 305

venosum Lam. (=25) 228

verrucosum Salisb. (=11) 203

williamsii N. Robson (19) 221,
DDE

wilsonii N. Robson (36) 248*

xylosteifolium (Spach) N.
Robson (63) 314, 316*, 317*

Komana patula (Thunb. ex Murray)

Y. Kimura ex Honda (=42)
265

salicifolia (Sieb. & Zucc.) Y.
Kimura ex Honda (=27) 232

Norysca angustifolia (Lam.) Blume

(=3a) 190

aurea (Lour.) Blume (=27) 232

calycina (L.) Blume (=25) 228

cernua (Roxb. ex D. Don) J.
Voigt (=24) 226

chinensis (L.) Spach (=27) 231

chinensis var. salicifolia (Sieb.
& Zucc.) Y. Kimura (=27)
232

cordifolia (Choisy) Blume
(=13) 213

coriacea (Blume) Blume (=33)
243

hookeriana (Wight & Arnott)
Wight (=39) 255

hookeriana var. leschenaultii
(Choisy) Y. Kimura (=33)
243

hookeriana var. leschenaultii
sensu Y. Kimura (=45) 271

javanica (Blume) Blume (=33)
243

kouytchensis (H. Léveillé) Y.
Kimura (=47) 275

lanceolata (Lam.) Blume (=3b)
190

leschenaultii (Choisy) Blume
(=33) 243

longistyla (Oliver) Y. Kimura
(=30a) 238

myrtifolia Spach (=12) 210

mysorensis (Wallich ex Wight &

Arnott) Wight (=12) 210
nervosa (Choisy) Miquel (=3)
243
oblongifolia (Choisy) Blume
(=24) 226

patula (Thunb. ex Murray) J. O.

Voigt (=42) 265
punctata Blume (=27) 232
rorida Blume (=33) 243
salicifolia (Sieb. & Zucc.) K.
Koch (=27) 232
urala (Buch.-Ham. ex D. Don)
K. Koch (=43) 268

STUDIES IN THE GENUS HYPERICUM L.

Kimura pro parte (=13) 213
urala var. angustifolia Y.
Kimura pro parte (=20) 222
urala var. angustifolia sensu Y.
Kimura pro parte (=39) 255
urala var. angustifolia sensu Y.
Kimura pro parte (=43) 268
venosa (Lam.) Blume (=25) 228

Psorophytum balearicum (L.) Y.

Kimura (=11) 203
undulatum Spach (=11) 203

325

Kimura (=57a) 295

formosana (Maxim.) Y. Kimura
(=54) 289

geminiflora (Hemsley) Y.
Kimura (=57a) 295

nakamurai Masamune (=55)
289

simplicistylum (Hayata) Y.
Kimura (=57b) 296

subalata (Hayata) Y. Kimura
(=58) 296

trinervia (Hemsley) Y. Kimura
(=57a) 295

urala var. angustifolia Y. Takasagoya acutisepala (Hayata) Y.

Notes added in proof

Pp. 170, 171, 214, and figs 4-6. It now seems more likely that the Hirtella group (Sects 17-19) is most closely
related to the variable H. oblongifolium. This alternative view is morphologically more fitting, and the 2n
= 24 chromosome numbers in sect. 17. Hirtella can then be regarded as primitive in the section.

P. 239. Cultivated material of H. subsessile (S.B.E.C. 551, originally identified as H. acmosepalum) shows
more variation than the two cited specimens, the lower leaves being oblong with a rounded-apiculate apex.

oh

2 jay ay : x aN . , " a
" rate pean pei Hage
Cis : 4 ; 38 es tata hee Candi quer’
nn. et ea
1G a Bjigi 1) ear?
< @ . he = 4 es vane A. hate gia rf

= ‘ ns +
x ae did ft pra
bs oe ‘ mm Of «a
tte oa
ae nisy a 2s
= a. babhaaatoy _
* i oe = >
: \ , ry Maes 7 Pate
or > 3 = i J ’ oJ ie “ rs (ke
Ly : & ¢ *
9 The BAW, son oi
coxcsahn 1 i Sieetl -
=
‘ erry B's % 23 me
Ka f My
‘
(p
‘
~~
a
= ——

British Museum (Natural History)

Ferns of Jamaica
A guide to the Pteridophytes

G. R. Proctor

This flora records and describes the 579 species and 30 varieties of ferns occurring in
Jamaica. The succinct species descriptions include relevant synonymy and incorpo-
rate distributional data both within and outside Jamaica. Special emphasis is given to
the subtle distinctions between closely related species and all genera are illustrated.
Keys to the genera and species facilitate a wider use of the flora in the West Indies and
northern South America. The author, one time Senior Botanist in charge of the
Herbarium of the Science Museum, Kingston, Jamaica is an outstanding field
botanist and his expertise is reflected in the practicality of the flora and especially
in the habitat and ecological information. This volume represents an important
addition to our knowledge of the flora of the West Indies.

1985, 650pp (approx), 135 line illustrations, 22 maps. Hardback.
0 565 00895 1 £50.00

Photoset by Rowland Phototypesetti
Printed in Great Britain by F

Ling L

: ,

@ a 7
St = ae i

i =

i

ix

Pets
nae

ine

ce