TV

6 m 5

BULLETIN OF

BOTANY

Vol. 4
1965-1973

BRITISH MUSEUM (NATURAL HISTORY)
LONDON: 1975

DATES OF PUBLICATION OF THE PARTS

No. i ..... November 1965

No. 2 ...... May 1967

No. 3 ...... June 1968

No. 4 ...... May 1969

No. 5 ...... June 1969

No. 6 ..... August 1970

No. 7 ..... August 1971

No. 8 ..... December 1973

ISSN 0068-2292

Printed in Great Britain by John Wright and Sons Ltd. at The Stonebridge Press, Bristol BS4 5NU

CONTENTS

BOTANY VOLUME 4

No. i. Cuticular studies as an aid to plant taxonomy. By C. A. STAGE

No. 2. The genus Elaphoglossum in the Indian Peninsula and Ceylon. By
W. A. SLEDGE ..........

No. 3. Fungi of recent Nepal expeditions. By F. L. BALFOUR-BROWNE
No. 4. A synopsis of Jamaican Myrsinaceae. By W. T. STEARN

PAGE

I

79

97

143

No. 5. The Jamaican species of Columnea and Alloplectus (Gesneriaceae) .

By W. T. STEARN ......... 179

No. 6. New or little known Himalayan species of Swertia and Veratrilla

(Gentianaceae) . By H. SMITH ....... 237

No. 7. A survey of the tropical genera Oplonia and Psilanthele (Acantha-

ceae). By W. T. STEARN ........ 259

No. 8. Angiosperms of the islands of the Gulf of Guinea (Fernando Po,

Principe, S. Tome, and Annobon). By A. W. EXELL . . . 325

Index to Volume 4

413

CUTICULAR STUDIES AS

AN AID TO
PLANT TAXONOMY

CLIVE A. STAGE

BULLETIN OF
THE BRITISH MUSEUM (NATURAL HISTORY)

BOTANY VoL * Na l

LONDON: 1965

CUTICULAR STUDIES AS AN AID
TO PLANT TAXONOMY

BY

CLIVE A. STAGE

(University of Manchester)

Pp. 1-78 ; 10 Text-figures ; Plates 1-5

BULLETIN OF

THE BRITISH MUSEUM (NATURAL HISTORY)

BOTANY Vol. 4 No. i

LONDON: 1965

THE BULLETIN OF THE BRITISH MUSEUM

(NATURAL HISTORY), instituted in 1949, is
issued in five series corresponding to the Departments
of the Museum, and an Historical series.

Parts will appear at irregular intervals as they become
ready. Volumes will contain about three or four
hundred pages, and will not necessarily be completed
within one calendar year.

This paper is Vol. 4, No. i of the Botany series.

Trustees of the British Museum (Natural History), 1965

TRUSTEES OF
THE BRITISH MUSEUM (NATURAL HISTORY)

Issued November, 1965 Price One Pound Fifteen Shillings

CUTICULAR STUDIES AS AN AID
TO PLANT TAXONOMY

By CLIVE A. STAGE

THIS review is an account of the systematic value of the epidermal characters of
angiosperm leaves as they are seen on preparations of the cuticular membrane, and is
based on a detailed study of about 250 species belonging to three families. The
introductory sections outline the history of the taxonomic use of cuticular and epider-
mal characters, the anatomical background of the features observed on cuticular
membranes, the methods employed in the isolation and mounting of the latter,
and a suggested routine for their formal description. The main part of the work
comprises a somewhat detailed account of the cuticular characters which have been
found to be useful in identification and taxonomy, and of the variability that some
of these exhibit in various conditions. The precautions necessary in the use of such
characters are stressed. In the final section the place of cuticular characters in
modern taxonomic work is discussed.

It is hoped that these comments will help to place the systematic use of the
epidermis upon a firmer and more scientific basis than has hitherto been the case.

This paper is a modified portion of a thesis submitted to the University of London
for the Ph.D. degree. The work was supported by grants from the Department of
Scientific and Industrial Research.

I am indebted to the Keeper of Botany, British Museum (Natural History), for
providing laboratory, library and herbarium facilities for the pursuit of this work,
and to Dr. A. W. Exell for his very considerable help and valuable discussion at all
stages. I am also grateful to those members of the Departments of Botany and
Palaeontology, British Museum (Natural History), who have helped me with several
problems, and to Mr. G. Grange, Department of Botany, University of Manchester,
for preparing the photomicrographs.

References (by author and date) given in the text are to the list of Special Literature
at the end of the paper.

4 CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY

i. PREFACE

Plant taxonomy has undergone a great many changes since it was first treated as a
scientific discipline in the seventeenth century, one of the major contributing factors
being the discovery of the use of new fields of investigation to supplement the old.
The ultimate aims of the majority of taxonomists have naturally also changed, the
original hope of cataloguing every species of plant now embracing the idea that their
evolutionary and genetic interrelationships should also be expressed. I accept that
taxonomists should aim at devising " natural " classifications which attempt to
indicate the phylogenetic position of each basic unit, although it seems to me certain
that this goal will be in sight only when the genetical constitution of each taxon is
known and can be expressed with something approaching the precision of a chemical
formula, towards which the present binomial system of nomenclature is no more than
a preliminary step.

Prior to the latter part of the nineteenth century taxonomists used only morpholo-
gical (i.e. macroscopic) characters in their studies, and apart from the inclusion of the
broadest vegetative features they confined themselves to the reproductive organs.
It was not until about 1865 that anatomical (i.e. microscopic) characters were first
used, and considerably later that this became usual practice. The volumes of
Engler & Prantl's Die naturlichen Pflanzenfamilien (1887-1915) included a relatively
large amount of anatomical and vegetative morphological details, and since then
taxonomic monographs have been considered incomplete without them. Modern
workers are constantly employing new diagnostic characters, and work of this kind
will need to continue until every structural, physiological and biochemical feature
has been utilized.

Parts of the plant which have been most successfully investigated from an
anatomical rather than morphological viewpoint include the wood, primary vascular
system, laticifer systems, crystals, epidermis (including stomata and trichomes),
floral vascular supply, sclerenchyma, pollen-grains and chromosomes. A further
group of characters, such as colour, chromosomal behaviour during meiosis, antigen-
antibody reactions and chromatographic analyses of various chemical constituents,
in fact represent plant structure at a third (i.e. sub-microscopic) level : that of the
molecule. Only a synthesis of ah 1 this and other evidence will provide anything
approaching a complete picture of each species, and thus an ideal classification.
Even today there is a great reluctance by many taxonomists to make real use of
any but macroscopic characters, and there are a large number who believe that
vegetative features are to be used only as subsidiaries to evidence from floral studies.

Of all the types of somatic anatomy studied it seems clear that the less superficial
layers of the plant have provided the best evidence for taxonomists, the wood,
where present, perhaps being the most useful of all. The leaf epidermis, or its
cuticle, has received relatively little attention from neobotanical taxonomists and
most of the studies undertaken have been concerned with somewhat specialized
groups of plants, e.g. Gramineae. Palaeobotanists, however, have used epidermal
characters in classification since the beginning of this century, and today their use is

CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 5

practised as a routine procedure, although here the groups of plants concerned have
usually been non-angiosperms. In recent years much more attention has been paid
by both neobotanists and palaeobotanists to their use in angiosperms, but most of
the studies are far from adequate and the subject is still in its infancy.

Except in fresh material of a relatively small proportion of species the leaf
epidermis of modern angiosperms is extremely difficult or impossible to remove
in toto. In fossil species, moreover, it is frequently not or only fragmentarily
preserved, and most of the early work on fossil leaf forms was carried out solely with
the aid of leaf impressions. The discovery that the cuticle, which covers the whole
of the leaf epidermis, bears an imprint of the epidermal features was therefore of
great importance, especially as it often remains on fossil leaves in a well-preserved
state long after all cellular layers have disintegrated. Moreover, except in plants
with a very thin cuticular covering, the cuticle can be removed from the fossil or
modern leaf (whether fresh or as a herbarium specimen) in a continuous sheet, to
which end several methods have been employed. Whilst some epidermal features
are lost in cuticular preparations the cuticle does provide some extra characters,
and the study of the cuticle rather than the epidermis places the work on a consistent
and precisely definable basis which is comparable with palaeobotanical investigations.
The present survey is concerned with the cuticle itself, although epidermal studies
are of course highly relevant.

The level of knowledge now reached with regard to the systematic anatomy of
the epidermis or cuticle is quite impressive in some groups, such as Gramineae or
gymnosperms. In the dicotyledons and other monocotyledons, however, little
extensive work has been attempted. Most investigations have been aimed at
separating small groups of species or genera with little or no attempt at studying
the degree of variation found within the taxa which are being separated. Much
of the literature in the dicotyledons consists of short sections subsidiary to much
wider studies of a more general nature. Work on fossil or recent plants by palaeo-
botanists has been solely for the purpose of identifying fossils by means of a compari-
son with living species rather than to investigate the intrinsic features of either
group, and scarcely any mention is found of the possibilities of the presence of
environmental or other variation. Many of the better surveys, even excluding
the gymnosperms or Gramineae, appear to have been undertaken on plants with
leathery, somewhat xeromorphic leaves (e.g. Lauraceae, Magnoliales). Such leaves
have a thick cuticle which shows the epidermal cell pattern clearly, and for some
reason these xeromorphs usually have a less variable pattern than mesomorphic
species. Studies of the latter types have mostly been of a very limited scope, and
in many cases a quite inadequate amount of material has been used as representative
of each species.

Thus there is a considerable gap in the literature with regard to both the variation
of cuticular patterns within a species and the types of characters and their degree
of taxonomic usefulness which are to be found in " normal " mesomorphic di-
cotyledons. The series of investigations upon which this article is based was aimed

6 CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY

directly at helping to fill this gap. The survey was intended to be intensive rather
than extensive, for, whilst this narrows the range of experimental material, it was
just this close study of a limited group of plants which was needed. Extensive
investigations have been undertaken in the past by a few workers, mainly palaeo-
botanists, but they have invariably been too superficial for general use. The
families of plants chosen for this programme, Combretaceae, Rhizophoraceae and
Avicenniaceae, are almost wholly tropical. They contain about 550, 120 and n
species in twenty, sixteen and one genera respectively, and although all are woody
in habit the species show a great variety of growth forms, including trees, shrubs
and lianes of all degrees of specialization occurring in all types of tropical habitat.
All three families, of which the first two are regarded as very closely related but
the third as very remotely so, contain mangroves. It is suggested that the results
gained from the intensive study of these families have enabled a good number of
general conclusions to be drawn which are applicable to all or most groups of
dicotyledons.

A soundly based theory of cuticular patterns can be put to a good number of
uses besides the obvious applications in identification, taxonomic research and
phylogenetic investigations. These include peat stratigraphy, pharmacognostical
analyses and animal foodstuff research. Only after thorough examination of all
aspects of epidermal and cuticular variation can systematic cuticular evidence be
legitimately used in these fields, and it is hoped that the following sections will
go some way towards realizing that aim.

2. HISTORICAL INTRODUCTION

A brief historical survey of the study of epidermal and cuticular anatomy throws
much light on the situation at the present day, and is a considerable help in under-
standing the areas in which work is now most needed.

The earliest reference to fossil cuticles is apparently the note by Brodie (1842)
" When the sandstone is freshly broken the epidermis of the fossil frequently peels
off . . .", but this observation was not followed up. Three years later Goeppert &
Berendt (1845) figured cuticular fragments of one species of conifer and two species
of angiosperm which they found preserved in amber. These figures appear in a
large volume primarily concerned, amongst plants, with twigs and leaf impressions,
and this is the case with the later cuticular diagrams of Schleiden (1846), Unger
(1853) and Wessel & Weber (1856). In all four cases the figures are crude and
highly diagrammatic, and except where the material has been available for subse-
quent study the true identity of the plants is obscure. For example the leaves
described by Unger as a Potamogeton have been found to belong to a member of
the Loranthaceae, whilst those figured by Schleiden are not identifiable at all
(Edwards, 1935). Wessel & Weber described and figured a number of dicotyledons
which they referred, in most instances, to living species, whilst Goeppert & Berendt
more advisedly used generic names, such as Alnites, which were not applicable to
living plants.

CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 7

The first really important work on the systematic treatment of cuticles was that
of Bornemann (1856), describing fossil cuticles of cycads from the neighbourhood
of Thiiringen. He was the first person to realize the exact nature of the cuticles
which he studied, and went to great pains to explain the relation between them and
the leaves which bore them. He wrote : " The outline of the cell walls of the
epidermis is almost always shown on the (cuticular) membrane by a network of
dark brown lines, which apparently represent the cell walls, but which are much
thinner than these would have been. These brown lines are to be regarded as
parts of a homogeneous cuticle ..." (transl.).

From then onwards a large number of works, from short notes to large volumes,
began to appear on the subject of fossil cuticles. Even so, systematic studies on
the surface features of angiosperms generally made use of the whole epidermis,
and such investigations in the latter half of the nineteenth century were very
numerous. One of the earliest of these was the investigation by Prillieux (1856)
into the types of peltate trichomes found in various members of the Oleaceae, and
he was actually able to separate all 23 species examined on the structure of these
trichomes alone.

The first taxonomist to make constant use of anatomical characters for his
diagnoses was apparently Bureau (1864) in his revision of the Bignoniaceae, although
these characters naturally formed but a very minor part of the descriptions. Soon
after this anatomical characters began to appear more widely in taxonomic treat-
ments, reaching a climax in Engler & Prantl's Die natiirlichen Pflanzenfamilien
(1887-1915). Conversely anatomists were becoming more aware of the systematic
value of many of their studies, and a number of extremely useful surveys appeared
at that time, several of which are unknown to many present-day workers. Weiss
(1865) produced a monumental account of the sizes, numbers and distribution of
stomata in a large number of dicotyledons ; Bokorny (1882) surveyed the distri-
bution and occurrence of pellucid spots in dicotyledon leaves, classifying them
according to their causes ; Bachmann (1886) gave a beautifully illustrated account
of all the types of peltate hairs known, in systematic arrangement ; Grob (1896)
provided a very useful early account of grass-leaf epidermides ; and von Minden
(1899) thoroughly covered the different types of water-secreting organs (e.g. water
stomata, hydathodes) found in dicotyledons. It was perhaps this productive
period which led Fritsch (1903) to summarize the most important systematic
anatomical characters to which taxonomists might pay more attention, and in many
instances his comments still apply today. Solereder's encyclopaedic volumes on
dicotyledon systematic anatomy, first published in 1898-99, represent the epitome
of anatomical works at that time, and together with a supplement were translated
into English by Boodle & Fritsch (1908). They remain a standard reference today,
one of their most valuable aspects being the detailed survey of the principal families
in which a wide range of anatomical characters are to be found.

Although works on fossil cuticles from Bornemann's (1856) time onwards were
equally numerous very few of great value were produced before the present century,

8 CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY

and scarcely any of these were concerned with angiosperms. Schenk (1869-71)
included some drawings of cuticles in his studies of Wealden plants, but of course
none of these was an angiosperm. Zeiller (1882) wrote what must be regarded as
the first review on cuticular studies, and his article brings together much useful
information with many supplementary original facts on cuticles of fossil ferns and
conifers. Perhaps the most important of the early cuticular studies, however, were
those of Nathorst (1907-12). His Paldobotanische Mitteilungen represents the first
work which reported a long series of cuticular studies on a large number of species,
and should be considered as the consolidation of this subject as a truly scientific
discipline.

One of the most important features of Nathorst's work is the improvement in
various techniques, and these were quickly assimilated by palaeobotanists elsewhere
in Europe and in America. This resulted in a considerable number of studies
entirely devoted to cuticles, rather than studies using cuticular characters as sup-
plementary evidence as had previously been usual. Thompson (1912), for instance,
studying Cretaceous conifers, re-examined the genus Frenelopsis and concluded that
Zeiller (1882) was wrong in believing that its stomata had the unusual number of
four guard-cells : in fact these four cells are the subsidiary cells, the guard-cells
being quite normal.

The first noteworthy cuticular investigations by British workers were those of
Thomas & Bancroft (1913) on cycads. They compared cuticles of fossil and recent
species with an emphasis on stomatal characters, and thus uncovered a considerable
amount of detail regarding the relationships of the fossil types. Some of the first
good photomicrographs of cuticles were included in this publication, and others
were produced by Wills (1914) who reported on fossils from British coal-measures.
Thomas (1930) later produced an article supplementing the knowledge of Mesozoic
cycadean fronds set out in the earlier work. To exemplify the spread of cuticular
studies from continental Europe Holden's (1915) paper may be cited, being one of
the first works on Asian cuticles. From her study of Indian fossil conifers she
concluded that epidermal leaf characters may often be of great value in specific
determination, but that they seem of little use as phylogenetic evidence.

By far the most important cuticular investigations are those of Florin, who for
the past forty years has produced a large number of excellent accounts of recent and
fossil gymnosperm cuticles, together with first-class photographs and drawings.
One of his earliest important papers (Florin, 1920) was concerned with a comparison
of fossil conifer cuticles with those of modern conifers, and most of his subsequent
work has been developed along these lines. Florin's most important work on
modern conifers (1931) includes descriptions of the cuticles of almost every known
species, with particular reference to stomata, and a section is also included on the
stoma types of the major gymnospermous fossil and recent groups for comparative
purposes. Later (1938-45) he produced an even larger work on Upper Carboniferous
and Lower Permian conifers, including full details of their cuticles, but this work
also embraces the reproductive structures and other vegetative organs than the

CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 9

leaves. Perhaps his most important work on non-coniferous cuticles was the account
of the stomata of Mesozoic cycads (Bornemann's subject in 1856) and Bennettitales,
attempting to shed some light on the relationships of these groups (Florin, 1933).
More recently Florin has devoted most of his time to studying conifer cone-structures,
and it is fortunate that he continued to investigate conifer relationships rather than
turn his attention to angiosperm cuticles as Edwards (1935) once suggested.

Bandulska (1923) investigated the cuticular anatomy of dicotyledon and conifer
leaves found in Eocene deposits at Bournemouth, and was able to place one of the
latter into an already known fossil species. The dicotyledons, which were un-
identified at the time, were placed into a new fossil form-genus, Dicotylophyllum.
In her following four papers Bandulska described a series of cuticles of modern
dicotyledons and in several genera she placed species which she had previously
described under Dicotylophyllum. These genera were Nothofagus and Fagus in the
Fagaceae (1924) ; Aniba, Lindera, Litsea and Neolitsea in the Lauraceae (1926) ;
Cinnamomum also in the Lauraceae (1928) ; and Tristania and Rhodomyrtus in the
Myrtaceae (1931). This work was the first important investigation of modern
dicotyledon cuticles to be attempted, and it clearly showed that with a great deal
of patience conclusive and positive results can be obtained from this approach.

Odell (1932) published a highly controversial paper which not only condemned
the identification of fossil leaves by their gross form (shape and venation), but also
severely attacked the use of cuticular characters for this purpose. She apparently
originally attempted to identify the older Eocene fossils of the Bournemouth
deposits, where Bandulska had first investigated the younger Eocene leaves, but
concluded that the variation shown in all characters which were normally used to
this end made it an impossible task. The bulk of her paper consists of a series of
conclusions drawn from the study of 170 species of modern angiosperms, but un-
fortunately she does not state the extent of her sample of each species. Her method
of argument is to take each character used in cuticular differentiation separately,
discuss the evidence for the fact that this character is known to vary under environ-
mental or other conditions, or that completely unrelated species are identical (or,
conversely, closely related species quite different) in that character, and conclude
that it is taxonomically worthless. One could, of course, equally well do this with
any other character that has ever been used taxonomically, but because (for example)
Ranunculus ficaria L. possesses about seven to twelve petals one cannot conclude
that petal number is of no diagnostic importance, either in this or in any other group.

It is perhaps worth mentioning two of the twelve or so characters which Odell
considered, as more or less random examples. With regard to the presence and
types of subsidiary cells around the stomata she states that the three main types
recognized by Solereder are not found especially in any taxonomic groups, but are
distributed variously in different genera and families. Moreover in some species
only the stomata on the stems are associated with subsidiary cells. Thus " the
presence or absence of subsidiary cells bordering the stoma can be of little systematic
value ". In the case of non-glandular hairs Odell mentions that the arrangements

TO CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY

and numbers per unit area on a leaf vary tremendously, and are affected by environ-
mental factors such as sunlight, humidity, wind and altitude. Their abundance also
varies on different parts of the same plant or even of the same leaf. Types of hair
often differ within a family, and there is often more than one type even on one leaf,
so that the types of hair are not connected with taxonomic categories. She therefore
decided that " clothing hairs cannot be used in the identification of the vegetative
parts of Angiosperms ", and that their " structure has been found of no diagnostic
value ". Her final conclusions were that " any feature of the epidermis of the
vegetative parts of living Angiosperms is unsatisfactory for diagnostic work ", and
that " the modern method of naming fossil Angiosperms from a combination of the
form, venation, and epidermal structure of their vegetative organs is quite inadequate
for specific or even generic diagnosis".

In almost all instances Odell's statement of fact must be regarded as correct, but
it is in some of her extraordinary conclusions that she is obviously in error. If she
was attempting to show that it is in general impossible to place a vegetative portion
of a plant into its correct family or order on a single epidermal character (or even on
a combination of such characters) then one would readily agree with her, but to
assert that each of the twelve or so characters is "of no diagnostic value ", or
" cannot be used in determinative work ", is obviously quite illogical. Odell
appears to have overlooked two important basic points in drawing her conclusions.
First of these is that the compilation of diagnoses and hence the identification of
species should not be undertaken with a single character, but with a combination
of as many characters as possible. Although two families, genera or species may
not be separable by a single character a combination of several characters will often
enable separation to be effected. Secondly a range of form is often more important
taxonomically than is a fixed type. For example the possession of a variable number
of petals by Ranunculus ficaria is a notable taxonomic feature of this species, not a
reason to doubt the systematic value of petal number. Further criticism of Odell's
remarks is unnecessary. In general one would agree with her that cuticular characters
are usually not useful at the family level or above, but it is unfortunate that she does
not differentiate between those cases where the characters are useful in phylogenetic
classification and those where they are useful for identification.

Belying Odell's claim that stomatal structure " cannot be regarded as of systematic
value " just four of several publications may be cited. Carolin (1954), working
with Dianthus, found that contrary to the usual findings (see below) neither stomatal
size nor frequency could be used to indicate the level of ploidy. However, within
the diploid and within the hexaploid groups most investigated species could be
separated by a series of size, frequency and morphological characters. Stebbins &
Khush (1961) surveyed the " stomatal complex " of most families of monocotyledons.
They found that characters such as the number and type of subsidiary cells gave
valuable clues to the affinities of the plants, so that each family had a predominant
type or arrangement. Watson (1962) found that in the Epacridaceae the stomatal
distribution and orientation on the perianth and leaves, and also the broad structural
types of stomata, were mostly differentiated at the generic or a higher level, and in

CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY n

the Ericaceae '(Watson, 1965) the stomata provided equally important taxonomic
evidence although in several instances there was little or no correlation with the
usually accepted classification.

Pharmacognosists have frequent need to identify leaves or fragments of leaves
used as herbs, and, as the number of species is relatively small, epidermal features
have often made identification possible. Timmerman (1927) produced an early
analysis of one of the characters used in pharmacognosy, namely stomatal frequency.
Using Datura she found that stomata were more frequent at the leaf apex than in
the middle, whilst the ratio on the two epidermides also altered in different parts of
the leaf. She therefore concluded that stomatal frequency was generally of no use for
identifying leaf fragments in this genus, although she was able to separate one species
from the rest of those examined by this character. In the same year, working from
a different angle, Salisbury (1927) also investigated the constancy of stomatal
frequency, adding environmental effects to the positional factors considered by
Timmerman. Salisbury found much the same variation as Timmerman, except that
the proportions on the two epidermides did not appear to alter. He found that if
the proportion of stomata to epidermal cells (the stomatal index, i.e. lOoS/E + S)
was measured instead of the stomatal frequency the variation produced by environ-
ment and position could be completely cancelled, apart from the effect of humidity.
He was able, in fact, to measure the environmental humidity by means of the
stomatal index. This very simple information was of extreme importance to
taxonomists, and the stomatal index is now used as a routine procedure in palaeo-
botany (cf. Harris, 1944). Salisbury's use of the stomatal index had been fore-
shadowed, however, by Loftfield (1921), who noticed that the ratio of stomata to
epidermal cells often remained constant when the stomatal frequency varied.

A less reliable method of overcoming the variation in stomatal frequency was also
used by Baranov (1924), who found that the stomatal frequency of the middle
portion of leaves half-way up the stem was equal to the average frequency of all
parts of all leaves. This does not take into account the fact that most fossils are
only fragmentary, nor the effects of environment.

Working much later than the above, Gupta (1961) also found a considerable
variation in stomatal frequency on leaves from different parts of the plant, and
demonstrated a negative correlation between this and the lamina size. He thus
used an Absolute Stomatal Number, the product of the stomatal frequency and the
lamina size, as a constant, this being comparable to the stomatal index.

Sax & Sax (1937) and Sax (1938) reported on some very interesting and now almost
classic studies on the effect of polyploidy on stomatal size and frequency. They
found a positive correlation between each of these two variables and polyploidy, and
could usually separate the diploids and tetraploids in any given series of plants.
Modern Floras now often contain characters such as stomatal size and stomatal index
in the separation of closely related species where one is a polyploid (e.g. Rorippa
nasturtium-aquaticum (L.) Hayek and R. microphylla (Boenn.) Hyland.).

12 CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY

The great mass of anatomical work which marked the end of the last century
seems to have largely ceased with the publication of Solereder's work and supplement
(1908). Rehfous (1917), however, provided a very useful survey of work on
stomata. In the I93o's systematic anatomy, together, in fact, with taxonomic
studies in general, began to regain popularity, and from then on the number of
systematic studies has increased quite steadily.

Meyer (1932, I932a) investigated the Alismataceae, firstly for taxonomic and se-
condly for phylogenetic reasons. Most of the genera and species Meyer examined
could be separated by anatomical features, but he found little evidence suggesting
close relation between the Alismataceae and Ranunculaceae ; the stomata, for
instance, are quite different. This work was connected with the commencement of
a Systematische Anatomic der Monokotyledonen by Solereder & Meyer (1928-33), of
which only a very small proportion appeared.

Prat (1932) published an extremely important work surveying the systematic value
of grass epidermides, this family without doubt being the most amenable amongst
the angiosperms to epidermal classification. Several important taxonomic changes
have been foreshadowed in the Gramineae by epidermal considerations, such as the
exclusion of Nardus from the Hordeeae or of Eragrostis from the Festuceae, a situation
only elsewhere paralleled in some gymnosperms. Many more recent works on this
subject have appeared, but the exceptional nature of the Gramineae precludes the
necessity of their being considered fully here. The most important work is that of
Metcalfe (1960), which is a complete survey of almost all genera and of all the litera-
ture. Some recent papers which show the amazing systematic value that can
confidently be placed on minute features of grass epidermides are those of Tateoka
et al. (1959) and Borrill (1961). In all cases, however, students of grass epidermal
anatomy have used the whole epidermis in preference to the cuticle.

In the last twenty years a great many systematic anatomical studies have been
undertaken, and the resulting publications provide sections on epidermal and
cuticular characters varying in length with their taxonomic importance in the group
concerned. Mention may be made of the work of Bailey & Nast (1944, 1948) in the
Winteraceae, Illiciaceae and Schisandraceae, Heintzelman & Howard (1948) in the
Icacinaceae, Morley (1953) in the Melastomataceae, Tomlinson (1959) in the Musaceae,
and Paterson (1961) in the Epacridaceae. The degree of use which can be obtained
from the epidermis varies considerably in these groups, as would be expected. Thus
Tomlinson was able to use a number of general and special epidermal characters to
great effect in delimiting at the generic level, but Bailey & Nast, who were two of
the few anatomists to isolate the cuticle and study it as such, found that apart from
its very distinctive appearance throughout the Winteraceae it did not furnish good
taxonomic characters.

Over forty years after Solereder's volumes (1908) Metcalfe & Chalk (1950) published
along similar lines a modern survey of dicotyledon systematic anatomy. The
literature survey is essentially concerned with work done since Solereder's time, and
many details of structure mentioned by the latter are omitted where no further

CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 13

examinations have been undertaken. The greatest improvements and additions are
to be found in the sections on secondary-xylem structure, upon which emphasis is
justifiably placed. With regard to the epidermis and cuticle Solereder's work is
therefore frequently the more informative. In reviewing characters of taxonomic
importance Metcalfe & Chalk obviously consider wood and associated tissues to be
of the highest importance, and epidermal characters are dismissed in a few paragraphs.
With regard to stomata they quite correctly state that the number per unit area is
too variable to be of any great importance, but it is strange that they do not mention
the usefulness of the stomatal index. More recently Metcalfe has commenced a
comparative Anatomy of the Monocotyledons, of which to this date volumes have
appeared on the Gramineae (Metcalfe, 1960) and the Palmae (Tomlinson, 1961).
Without such an important basis as Solereder's dicotyledon volumes the compilation
of a systematic anatomy of the monocotyledons is a considerably greater task, and
when completed will probably prove even more valuable.

In 1954 Metcalfe published a paper which may be compared with one by Fritsch
(1903) half a century before. This set out to indicate the main lines along which
systematic anatomy could be developed, and in the next year (1955) he summarized
recent work on this subject in the monocotyledons, especially the grasses. More
recently Metcalfe (1963) has elaborated his opinions on the desirable trends in
modern comparative plant anatomy.

From about the time of Bandulska's work (1923-31) onwards a considerable
number of articles on fossil dicotyledons began to appear in the literature, in several
cases obviously directly inspired by her publications. Stockmans (1932) reported
on two species discovered in Belgium : a Litsea and a Dewalquea. The first genus,
a member of the Lauraceae, had been discovered in England by Bandulska (1926),
whilst the extinct genus Dewalquea had been previously found in Ireland by Johnson
& Gilmore (1921). Discussions by Stockmans and Johnson & Gilmore, in both cases
well illustrated, on the affinities of this genus produced little concrete theory, several
diverse families being mentioned. Since not all of the species of Dewalquea possessed
peltate scales on the leaf it was suggested that they were perhaps representatives
of two different genera. Whether or not this is true it must be remembered that other
genera known today (e.g. Combretum and Rhododendron) possess scaly and scale-less
species.

Hofmann produced a valuable series of papers from 1926 to 1932 (cf. Hofmann,
1932) on plant remains in early Tertiary formations. In her earlier works she
attempted to give names of living genera or species to the cuticles, and in 1926 she
described fossil plants in ten living genera from these deposits. In 1932, however,
she changed to the more advisable system of describing these old leaves in a fossil
form-genus. Her choice of name for this was Folium, which is nomenclaturally
superfluous owing to Bandulska's early name Dicotylophyllum for the same purpose.
Edwards (1935) justifiably comments that her descriptions of new fossil leaves are
somewhat inadequate, but she gives some good photographs.

Kubart (1927) found great difficulty in distinguishing between the fossil leaves of

14 CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY

Fraxinus (Oleaceae) and Umbellaria (Lauraceae) which he discovered in North
America. He pointed out, however, that had the cuticles been preserved there
would have been no difficulty in discrimination as the epidermal structure of the
two genera is quite different. In support of this he figured a series of recent species
of each genus, showing the great divergence between them. Also working on Ter-
tiary plants, Straus (1930) was able to identify a number of species with modern
genera, and in the cases of the two better-preserved genera (Juglans and Populus)
he gave some good high-powered photographs. Straus pointed out the need for
microscopic examination of the cuticle in cases where two fossil leaves appeared to
be closely similar.

In recent years the great majority of works on fossil cuticles has been concerned
with pre-Tertiary plants, mostly gymnosperms. Perhaps the most important
publications have been those of Harris (i942->), who has covered a wide range of
topics on the Jurassic flora of Yorkshire in several scores of articles, many of these
including cuticular studies which have often proved of considerable diagnostic value.

Cookson (1953) reported on a relatively recent (early Tertiary) cycad from Australia
which, on the basis of its cuticular characters, could be referred to the genus
Macrozamia, which still occurs in Australia today. Cuticles of this fossil and of the
living M. hopei W. Hill ex F. M. Bailey are shown, and due to the great similarity
between them the former is named M. hopeites Cookson.

Apart from Europe and North America some of the most important cuticular
studies have emanated from India, and as examples two of the papers on the Indian
Glossopteris flora by Srivastava and his colleagues may be cited. Srivastava (1956)
produced a most thorough account of all the species of Glossopteris, Gangomopteris
and Palaeovittaria, with very clear descriptions, drawings and photogiaphs. All
species could quite easily be separated on cuticular characters alone, a whole series
of features, concerning epidermal cells, stomata and venation, being employed.
Surange & Srivastava (1956) concurrently considered the generic limits in this group
of leaves. They concluded that the three genera, which were delimited on frond
shape, could be better divided into six genera using the cuticular characters. These
six genera did not correspond in any way to the three established genera, and
pending further information no new generic names were proposed.

Apparently the only general review of cuticular studies in angiosperms is that of
Edwards (1935). He gives an extremely useful survey of the literature to that date,
which has obviated the necessity of very detailed accounts of the earlier work in
this introduction. The account treats recent and fossil monocotyledons and dicoty-
ledons in four separate sections, those on monocotyledons being mostly concerned
with grasses. Many of his citations are annotated with some sort of assessment on
their importance or validity, but in the case of Odell's (1932) paper he quite justifi-
ably records strong protest with a series of objections, mostly in the form of counter-
evidence. One would concur with almost all of his opinions on this and other papers,
but as mentioned previously I prefer to attack Odell's deductions and conclusions
rather than her statement of fact. In several instances it is obvious that Edwards

CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 15

held almost as extreme ideas as Odell, though in a completely opposite direction.
He was clearly convinced of the tremendous and widespread usefulness of cuticular
characters in the angiosperms, which is perhaps to be expected of a palaeobotanist
whose cuticular studies had mostly been concerned with gymnosperms. Had he
made extensive investigations in the angiosperms he would have met, in many
instances, with a much wider range of variation than is found in any gymnosperm,
and than he seemed to admit as existing in the angiosperms. For instance, in
criticizing Odell's statement that the leaves of Campanula latiloba A. DC. and Inula
salicina var. denticulata Borbas are identical, he found that there were in fact good
microscopic differences between them, and he also suggested that Odell's " own
drawings do not suggest identity of epidermal structure ". Odell's figures differ
only in the thickness and degree of undulation of the epidermal cell walls, and had
Edwards been familiar with the range of variation of the cuticles of many angio-
sperms he would have realized that most species show a wider range of variation
than Odell's two drawings. Thus, although in fact the two species do apparently
differ in cuticular structure (fide Edwards), Odell's figures certainly do not suggest it.
Apart from a few criticisms such as the above, Edwards's paper is a valuable and
critical review. His conclusion may be quoted : " given careful and critical work
on well-preserved material, together with a detailed comparison with a wide range
of living forms, results obtained from a study of fossil angiosperm cuticles will be as
valuable as those derived from any other fossil remains, and certainly far more
reliable than those founded on leaf impressions alone ".

Edwards did not mention any articles concerned solely with modern dicotyledon
cuticles, and the first that the writer has traced is the account by Rao (1939) of the
order Magnoliales in the sense of Hutchinson (1926), a very heterogeneous assembl-
age of families about which Rao concludes: " From the point of view of epidermal
studies, the Magnoliales do not represent a co-sanguinary group, but indicate a
convergence ". Rao's paper presents a detailed description of the cuticles of over
50 species representing more than half the genera of the order, but since he admitted
that in most cases only a single leaf of each species was examined the results must
be treated with extreme caution. Objection must be made to some of Rao's con-
clusions, although a detailed criticism will not be attempted here. Rao implied,
for instance, that the cuticular striations found in some species of Magnoliales are
phylogenetically related to those of the Cycadales. However, representatives of
many dicotyledon families possess cuticular striations which are identical with those
of the Magnoliales. Even if cuticular striations in the two orders were phylogeneti-
cally related, and this seems unlikely and in any case could never be proved, no such
inference could be drawn merely from their presence in both groups. Rao strongly
stressed further evidence of a connexion between the Magnoliales and gymnosperms
from his stomatal studies. Using Florin's (1931) two categories of stomatal de-
velopment Rao claimed that two genera of Magnoliales, Euptelea and Cercidiphyllum,
possessed haplocheile stomata, which according to Rao were otherwise virtually
unknown in the dicotyledons yet were typical of all gymnosperms except the
Bennettitales, W elwitschiales and Gnetales. Later work has shown many of Rao's

i6 CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY

conclusions regarding stomatal types to be erroneous, and a modern stomatal survey
of the Magnoliales is badly needed. Jalan (1962), for example, has found that the
stomata of Schisandra are haplocheilic (as, in fact, they are in a great many
dicotyledons), and the same is true of Gnetum (Maheshwari & Vasil, 1961). In
Magnolia and Michelia species the stomata on the leaves are syndetocheilic, but
those on the flowers haplocheilic (Paliwal & Bhandari, 1962). The need for develop-
mental studies rather than observations of mature structure cannot be over-
emphasized, and the approach of several Indian workers in this field is encouraging
(cf. Pant, 1965). Further reference to this is made in the section on stomata.

Ahmad (1964, 1964^ 1964^ has investigated the cuticular and epidermal anatomy
of 17 genera of Solanaceae, and found that a number of characters concerning the
stomata, trichomes, epidermal anticlinal cell walls and cuticular striations are very
useful in identification. The same characters enabled 28 species of Solanum and
23 of Cestrum to be differentiated.

Sinclair & Dunn (1961), on the other hand, working in the monocotyledons,
appear to have made conclusions regarding the usefulness of cuticular characters
at the family and generic levels which are not justified on present information.

It is perhaps largely by chance that the Lauraceae have been studied more
frequently than any other dicotyledon family with regard to their cuticular charac-
ters. Studies already mentioned have been those of Bandulska (1926, 1928), Kubart
(1927) and Stockmans (1932), and there have been other less important ones besides.
Kostermans (1957), in his taxonomic studies of the Lauraceae, mentioned the useful-
ness of cuticular characters, but pointed out the care necessary in the interpretation
of the results of these investigations. He was referring mostly to Bandulska's work.
Marlier-Spirlet (1945) investigated the epidermal anatomy of Cinnamomum more
thoroughly than did Bandulska, and produced many figures. He also showed
vertical sections of the leaf to demonstrate the derivation of various features of the
cuticular preparations which were not readily interpretable from surface views
alone. It was found to be possible to separate many of the eighteen species in-
vestigated, and three broad groups were recognized, containing four, three and
eleven species respectively. The most useful taxonomic criteria appear to be the
shape of the upper epidermal cells and the types of stomata. A continuation of
this study would be of considerable interest and value. Dilcher (1963) undertook
a " cuticular analysis " of an Eocene species of Ocotea, and after a study of its
variation and of the living taxa concluded that it should be considered a distinct
species.

Martin (1955) used cuticular anatomy to a most novel and useful end to which
there appear to be no subsequent references. Previously Parkinson & Fielding
(1930 ; fide Martin, 1955) had compiled notes of some epidermal characters of the
plants used as cattle food, using the solid leaf fragments. In order to study the
diet that sheep choose in various pastures Martin examined the rumen contents
and faeces of these animals to ascertain quantitatively the species of plants re-
presented by their cuticles. In the upland pastures in which the work was pursued

CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 17

the number of species was rather limited and few species were closely related ; thus
Martin was able to separate all forty likely plants by cuticular anatomy and deter-
mine them quantitatively in the faeces. He gives brief descriptions and photographs
of twelve representative species. Of these forty species only sixteen were recognized
in the stomach contents or faeces, including four grasses, two sedges, three rushes,
two mosses and five dicotyledons. Except in the case of two species of Juncus
each genus was represented by only a single species, although other genera were
represented by two or more species among the originally investigated group of
forty plants. It seems likely, however, that some or many mesomorphic herbs
would never be represented due to their extremely thin cuticle, although the only
known cutinase enzyme has been isolated for certain solely from Penicillium
spinulosum Thorn (Heinen, 1960, 1961).

This work is a further indication of the economic use to which cuticular studies
may be put, and emphasizes the need to discover the basic principles underlying
the possible application of this method in various directions.

3. ANATOMICAL INTRODUCTION

Some idea of the precise relationship between the cuticle and the underlying
epidermal cells is essential to a full understanding of the nature of the patterns seen
on acellular cuticular preparations and their interpretation in terms of a cellular
layer.

Although Bornemann (1856) obviously understood the cuticular-epidermal
relationship fully the first good general anatomical accounts are those of de Bary
(1871, 1884), and Hohnel (1878) appears to be the first to have described the
chemistry of the cuticle. During the past thirty years our knowledge of the cuticle
and of its essential constituent, cutin, has steadily grown, accelerated by two major
technical advances. The first of these, polarizing microscopy, became available in
the I920 r s and its use is shown in the work of several Continental writers. The
second, electron microscopy, a post-war development, has been utilized in cuticular
studies mainly by Roelofsen and a number of American workers.

In all vascular plants the basic chemical component of all the cell walls is cellulose,
whose long-chain molecules exist as long compound microfibrils. In the inter-
fibrillar spaces are a number of substances, including other polysaccharides, fatty
materials, water and inorganic and organic solutes. Separating two such walls of
adjacent cells is a cellulose-free layer known as the middle lamella, consisting
basically of calcium and magnesium pectates. The outer face of an epidermal cell,
however, is unusual in being adjacent to the environment rather than to another cell,
and it is on this face that the cuticle is found. The precise structure of the latter
was first ascertained by means of polarizing microscopy (e.g. Meyer, 1938 ; Roelof-
sen, 1952) and was confirmed later by electron microscopy (Roelofsen, 1959).

External to the normal cellulose cell wall of the outer face of each epidermal cell
is found a thin layer of pectic material which is presumably continuous with the
middle lamellae of the anticlinal walls of the underlying epidermal cells (Fig. i).

i8

CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY

Outside this is a usually two-layered wall, which according to the nomenclature of
Roelofsen (1952, 1959) may be termed the cuticular membrane. The inner portion
of this, the cuticular layer, is composed essentially of a cellulosic framework between
the microfibrils of which are encrusted large amounts of cutin. The outer part is a
usually thinner layer, the cuticle proper, which lacks cellulose and is composed
mainly of cutin, which is adcrusted on to the cuticular layer. Esau (1953) refers to
the process of encrustation as cutinization, and that of adcrustation as cuticular i-
zation. The above terminology is adopted hereinafter. In a number of plants,
usually those with leathery leaves, the four layers exterior to each epidermal cell
lumen are simplified insofar as a cellulose-free pectic layer is absent, the cutin-free
innermost cellulose wall gradually merging into the outermost cellulose-free cutin
layer. Rarely are all four layers as sharply defined as suggested in Fig. i, but in
general it can be said that the innermost layers lack cutin and the outermost layers
lack cellulose. Chemically cutin consists of a number of highly polymerized long-
chain hydroxy-fatty acids.

cuticle

increasing
cellulose

[increasing
cutin

middle lamella

? possible boundary

between cutin-cellulose

and cutin-free layers

cell outline

observed on

cuticular preparations

extent of original
cell wall

middle lamella

FIG. i. The organization and interpretation of cuticular membranes and preparations :
sections of epidermides with normal and abnormally extensive cutinization are shown,
together with diagrams showing the significance of the patterns observed on the
respective cuticular preparations.

CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 19

Where the outer walls of two epidermal cells meet, the pectic layer, if present,
may form a T-shaped juncture with the middle lamella of the vertical walls of the
cells. In most cases, however, the cuticular layer projects down between the two
cellulose cell walls for some way, forming the three-dimensional network of so-called
cuticular flanges (Fig. i). The length of the cuticular flanges varies from their
virtual absence to their reaching all the way to the inner wall of the epidermis. In
some leaves a greater proportion of the epidermis may be cutinized, in extreme
cases the cell walls on all faces of the epidermal cells and even the outer walls of
the subepidermal layer being encrusted with a certain amount of cutin. This
internal cutinization is a feature of many xeromorphic plants.

Cuticular flanges are formed by the cuticular layer alone, and the cuticle remains a
thin uninterrupted membrane. Indeed, the cuticular membrane as a whole is a
continuous and virtually extra-cellular sheet, and should be thought of as something
additional to the cell wall. It completely covers the leaf, usually including any
trichomes which may be present, being interrupted only by the stomata. In a
herbaceous plant it appears to surround the whole of the aerial shoot, including
the apical meristems and flowers, although some workers have stated that certain
trichomes lack a cuticular covering. With regard to the root system there is less
certainty, opinions differing whether it is only the root tip and root hairs or nearly
all the root that lacks a cuticular membrane (Lee & Priestley, 1924 ; Priestley,
1943 ; Esau, 1953). At each stoma the cuticular membrane follows the contours of
the guard-cells, and often ends on their inner walls. In some cases, however, it is
more extensive, covering also the adjacent mesophyll cells. Cutinized cell walls
have also been found in more internal regions of the plant (cf. Arzt, 1934), and
Lamarliere (1906) found a " cuticule interne " lining the aerenchyma of water plants.

Right from the time that the leaf is a minute primordium in the apical meristem
it is covered by the rudiments of the cuticular membrane in the form of a very thin
and pliable yet continuous fatty layer. These fatty substances obviously migrate
through the cell wall from the cytoplasm, although there are no visible pores in the
former for the passage of the cutin. On exposure to air the cutin hardens to a
varnish-like cuticle by polymerization. A reasonably hardened cuticular membrane
is present long before the epidermal cells have attained full size, and the problem is
thus one of increasing the surface area of a relatively non-stretchable substance.
Some workers suggest that the epidermal cells may grow in surface area only at
their edges, so that the already hardened cuticular membrane remains undisturbed
(e.g. Schieferstein & Loomis, 1959).

Not only is wax present within cell walls and cuticular membranes but it is also
found in a fairly pure state on the surface of the cuticle of many plants. In recent
years these deposits, which are the cause of the " bloom " on the surfaces of many
leaves, have been studied in considerable detail, since they are important to the
understanding of water repellency. Fairly comprehensive accounts are given by
Mueller et al. (1954), Schieferstein & Loomis (1956, 1959), Scott et al. (1958), Juniper
& Bradley (1958), Roelofsen (1959) and Juniper (1959, i959a, 1960). It is

20 CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY

interesting to note that the type of deposit (frequently variously shaped rods and
plates) appears to be largely characteristic of each plant species. The deposits,
when washed, dissolved or brushed off growing leaves, re-form quite rapidly and in
exactly the same form as previously, and although their abundance varies with the
environment and other factors their non-varying form is obviously in some way
genetically regulated. What determines the form which the wax assumes is of
course unknown, and moreover different parts of the plant (e.g. stem and leaf
epidermides) may have totally different patterns. This subject is obviously of some
taxonomic interest though doubtfully of any widespread practicability, and it is not
considered further in this review.

In studying the cuticular membrane various methods may be employed to isolate
it from the rest of the leaf. All make use of one or both of two important structural
features of the epidermis which will have become apparent from the preceding
paragraphs : firstly, the cuticular membrane is in most plants separated from the
cellulose cell wall within by a thin layer of pectic material ; and secondly, the
cuticle and the bulk of the mature cuticular layer are essentially composed of a
substance (cutin) of very different chemical structure from the main constituent of
the cell wall (cellulose) . Isolation of the cuticular membrane may be effected with
either of two types of substance : an enzymic preparation, either a pectinase which
digests the pectic layers and so frees the cuticular membrane from the underlying
cells and these from each other (Orgell, 1955), or a cruder hemicellulase-cellulase-
pectinase mixture which additionally digests the cell walls (Skoss, 1955) ; or a
macerating solution, composed usually of strong acids, which completely dissolves
away all polysaccharide leaving only cutin unaffected. Very much later the acid
also dissolves the cutin. The enormous resistance of cutin to the effects of age,
micro-organisms and chemicals is, in fact, its greatest attribute with regard to the
present study. Only one organism (Penicillium spinulosum) is known to be able to
break down cutin (Heinen, 1960, 1961). The advantage of the above enzymic method
is that the cutin is always completely unchanged, and in biochemical studies of the
cuticular membrane this method is the only one to be recommended. On the other
hand the macerating solutions are very much quicker and simpler to use, and the
speed of reaction may be adjusted more easily. The cuticular membranes are also
better for microscopic examination since they are considerably cleaner.

The upper surface of such preparations is smooth, or minutely striate or papillose,
whereas the lower surface is protracted into a reticulate series of flanges which
represent the position of the epidermal cell walls. When viewed from above or
below the cuticular membrane is much thicker at the position of these flanges, which
therefore show up, especially if the cutin is stained, as a series of narrow darker
lines representing the shapes of the original epidermal cells (Fig. i). It is important
to realize that these lines only show the position of the central region of each double
cell wall, and that the original walls were present for a considerable thickness on
either side of them. In those cases where there are no cuticular flanges the cuticular
preparations will of course lack the cell outlines, but it is apparently rare in woody
plants for at least the faintest and most fragmentary of outlines not to be present.

CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 21

As mentioned previously some plants have a heavily cutinized epidermis, the whole
of the outer, vertical and inner walls and even the top of the vertical walls of the
subepidermal layers being encrusted with cutin. In these cases the preparations
have cuticular flanges which are very long and join up at their bases, and in surface
view a second fainter outline, that of the subepidermal layer, may be visible below
the epidermal outline. There is some evidence that the cuticular membrane is more
or less completely lacking in cell outlines, or is not even able to be isolated in a
continuous sheet, in many though by no means all herbaceous dicotyledons. In
these cases the whole epidermis, or plastic moulds of it, must be studied.

4. MATERIALS AND METHODS

For the purposes of obtaining cuticular preparations leaves may be used in any
state (fresh, spirit-preserved or dried) and it is fortunate that herbarium specimens
even over 100 years old produce at least as satisfactory preparations of the cuticular
membrane as fresh leaves, exactly the same techniques being used. It is important,
if studies of cuticular membranes are to be used as routine taxonomic procedure, that
the methods involved in their isolation should be simple, quick and reliable.

With regard to the choice of material it is essential to ensure that a representative
sample of each species is examined, and this involves a considerable amount of
preliminary investigation into the variability of the taxa concerned. Thus the
amount of material which can be considered truly representative will vary greatly,
but in the present study I found it necessary to examine only one leaf of each
specimen, except in cases of obviously different-aged leaves. The number of
specimens of each species which was investigated varied according to the amount of
material available, the degree of macroscopic variation apparent, and the ecological
and geographical range of the species concerned. Three to five specimens were
usually found to be a sufficient sample, but care was taken to select those with the
widest morphological and ecological variation. In many cases, however, fewer
specimens were available, while on the other hand special circumstances sometimes
necessitated the examination of up to twenty.

The first task in investigating the taxonomic use of a new character is obviously
to check it with the other characters already used, and for this correctly named
material is absolutely essential. In the present series of investigations the families
concerned are rather well covered in the literature, although the lack of knowledge
concerning two genera necessitated a complete revision of all the species (Exell &
Stace, 1963), and in some families this problem is so marked that it would become the
limiting factor. Other genera of the three families studied still badly require
revision. Whatever the state of knowledge of the classification of the taxon
concerned differences of opinion may exist and occur, and a careful record of the
actual specimens examined is essential. All the permanent slides prepared by
me are marked with the collector, the number, year and country of collection, and
the herbarium from which the material was obtained. This enables the identifica-
tion to be checked whenever required.

CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY

Many different macerating techniques have been employed to isolate the cuticular
membrane. Enzymic maceration is not suitable for taxonomic work as the condi-
tions (temperature, pH, concentration) have to be carefully controlled and the method
is very slow, and in fact in many thick leaves a clean preparation cannot be obtained.
Of the milder chemical macerating solutions Eau de Javelle (CaQ 2 and K 2 CO 3
solutions) is the most used (Nathorst, 1907 ; Bather, 1908 ; Bandulska, 1923),
but for present purposes strong acids are preferable. Schulze's Solution (Schulze,
1855), concentrated HNO 3 with a few KC1O 3 crystals added, is the most popular,
and several improvements and variations have been suggested (e.g. Bather,
1908 ; Thomas, 1912 ; Thomas & Bancroft, 1913 ; Gothan, 1915 ; Harris, 1926).
I found, however, that a slight modification of Jeffrey's Solution (Jeffrey, 1917),
equal parts of 10% chromic and bench concentrated nitric acids, gave the best
preparations in the shortest time, and it has been used without exception. This
mixture was variously diluted to conveniently regulate the time taken for macera-
tion to about sixteen hours, i.e. overnight, and in all 250 or so species examined the
method was entirely successful. Care must be taken not to allow maceration to
continue for too long since over-macerated cuticular membranes are unsuitable for
future manipulation and examination.

Parts of leaf macerated

Cuticular preparations
obtained

margin

u = cuticular membrane from upper epidermis
/ = cuticular membrane from lower epidermis

FIG. 2. Diagram illustrating the origin of cuticular preparations.

For purposes of conformity identical regions of each leaf were macerated (Fig. 2),
so that three pieces of cuticular membrane were made into separate permanent
mounts.

CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 23

In some cases it is necessary to examine the cuticular pattern of species not avail-
able for maceration, or of species whose cuticular membrane cannot be removed as
a continuous sheet. For this purpose the preparation of a plastic epidermal cell
mould can prove of value, the plastic, usually cellulose acetate or nitrate, being
either painted on to the leaf as a solution or pressed on as a thin sheet and then
partly and temporarily redissolved (e.g. Nathorst, 1907, igoya ; Bather, 1907 ;
Thomas, 1912 ; Long & Clements, 1934 ; Sinclair & Dunn, 1961), making use of
the fact that the outer surface of the cuticle usually follows the contour of the
epidermal cells to some extent. The version of this method considered most
suitable is that of North (1956), who used the formula 2 gms. of cellulose acetate
and 0-7 gms. of gentian violet in 100 mis. of acetone, the stain improving the contrast.
Herbarium specimens, whether dry or boiled, rarely give such good results as fresh
leaves, but they are usually quite usable. The method is useless for most thick-
cuticled leaves, which do not show the lines of the epidermal cells on the outer
cuticular surface, and for pubescent leaves, although in the latter case the hairs
can sometimes be removed. Occasionally transparent cell moulds are found already
prepared on herbarium specimens in the form of flakes of old dried herbarium gum,
which may show the epidermal cell outline very clearly.

Frequently sections of leaves must be prepared in order to interpret the observed
cuticular patterns. Fresh or spirit-preserved leaves are naturally easier to section,
but herbarium leaves of great antiquity, after boiling in water for a while, usually
provide good sections when cut by hand or with a freezing-microtome. It is often
more expedient to isolate the whole epidermis rather than the cuticular membrane
alone, this method being particularly applicable to Gramineae. Lactic acid is
generally used to separate the epidermis from the subepidermal layers, and the
epidermis may then be stained with any general botanical stain (Clarke, 1960) .

The technique employed to prepare permanent mounts was again designed to
meet the three requisites of speed, simplicity and reliability. The best stain was
found to be Sudan IV (i% solution in 70% alcohol) which has the advantage of
being progressive, so that overstaining does not occur, but the disadvantage that
dehydration and thus mounting in canada-balsam or similar media is impossible
since the stain is re-dissolved from stained preparations by absolute alcohol. Thus
glycerin-jelly was used as the mountant, and although this is not completely per-
manent most preparations are still perfect after at least thirty years and can at any
time be remounted. Since the cuticular preparations are acellular sheets rather
than pieces of tissue the number of stages from water to alcohol, alcohol to water,
and water to glycerol can be reduced to a minimum, and each stage to a few seconds.
The procedure adopted for preparing permanent mounts may be summarized :
leaf fragments macerated in suitably diluted Jeffrey's Solution in an open watch-
glass ; cuticular membranes washed in water, re-macerated in undiluted solution
and re-washed in water if some cellular tissue remains, and transferred to a slide
in 70% alcohol ; 70% alcohol replaced by Sudan IV solution for fifteen to thirty
minutes ; slide irrigated with water, with 50% glycerol, and then with 100%
glycerol ; and cuticular membranes finally mounted in glycerin-jelly.

24 CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY

Cuticular preparations can be kept indefinitely, before staining and mounting,
in 70% alcohol ; placed in water they become excellent cultures of bacteria, fungi
and protozoa (which feed on the adhering polysaccharides, not on the cutin) in a
day or two.

5. SURVEY OF CUTICULAR CHARACTERS

The present section is a catalogue of the main characters which are present in
cuticular preparations, together with some idea of the extent of their variation from
species to species. In general the ranges of variation discussed indicate those
normally encountered without examining plants of an extreme facies or with a very
specialized adaptation, so that no mention is made of unusual extremes of structure
or of features of very restricted occurrence. Any combination and number of these
characters may be of diagnostic value in a particular group of plants. The first
task is thus to survey the features found on cuticular membranes, and then to
investigate the degree of variation each exhibits. After analysing the causes of
variation the systematic value can be ascertained. Most of the comments made
have been drawn from my experience in investigations of the three families
mentioned previously, and, whenever relevant, examples are usually given from
these groups.

Such a general and, it is hoped, purely objective survey of cuticular characters
has not been previously compiled, and thus the different terms which have been
introduced into the literature are quite uncorrelated. In many cases, for instance,
several descriptive terms exist for a single situation (e.g. cuticular flanges, teeth,
pegs or ribs). When an acceptable and non-ambiguous descriptive term is already
available this has been adopted, but in some cases terms had to be coined, or existing
unsuitable ones replaced. All accepted terms are printed in italics, and it is hoped
that this will encourage nomenclatural stability.

Of the characters observed during microscopic examination of cuticular prepara-
tions some are quite easily observed on whole leaves with an ordinary hand-lens, or
with the naked eye. In these cases microscopic examination usually gives a more
quantitative or precise assessment of the character, although some points (such as
the colour or " texture " of the indumentum, e.g. silky, bristly or velvety) may be
lost with this medium. Other characters are wholly microscopic in nature, and it is
convenient here to treat the two groups separately.

A. MACROSCOPIC CHARACTERS.

The four features discussed here are treated only from the microscopic viewpoint,
although preliminary study of a leaf with a hand-lens or low-power binocular
microscope should always be undertaken.

(i) Leaf venation.

The venation of a leaf, second to its general shape, is frequently the most obvious
character, and descriptions of all new plants contain (or should contain) a note of its
organization. The type of venation is often typical of broad taxonomic groups. In

CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 25

general, dicotyledons have leaves (or leaflets) with a single central main vein
which is a continuation of the petiole, if this is present. This is known as a midrib,
costa or main vein. In some leaves two or more main veins may be present,
although that situation is more characteristic of monocotyledons. Branching from
the midrib are a number of lesser veins or nerves of successively smaller order.
Those branching directly from the midrib may be termed lateral veins, and they are
usually very numerous. However, a limited number, from two or three to a score
or more, are generally much more conspicuous than the rest and are here described
as major lateral veins, as opposed to the less distinct minor lateral veins. In most
descriptions " lateral veins " usually refers to the former only. The number of
major lateral veins, their spacing, whether they arise in pairs or independently on
either side of the midrib, the angle at which they arise, the curvature along their
length, and their method and position of ending are all taxonomic features.

midrib

minor lateral vein
major lateral vein

veinlet termination
venule

secondary vein
tertiary vein

FIG. 3. Terminology of leaf venation.

Such a venation system is pinnate, but palmate systems also exist. In the latter
case there are a number of main veins all arising from a single point and diverging
outwards, and each usually has an accessory venous system resembling that of a
pinnately veined leaf.

Generally, arising from the lateral veins is a series of secondary veins, from these
tertiary veins, and so on. Unlike the branching of a tree, however, the smallest
veins, the venules or veinlets, in most cases do not possess terminations, but usually
join up to form an extensive venule reticulum. The number of orders of veins between
the midrib and the venules is usually one or two only, but it may be more. The
term venule is reserved for the smallest veins with no directional tendency. What-

26 CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY

ever the number of orders of veins, however, there is in these types of leaves usually
a system of venules which emanates directly from the midrib, and almost always
one from all other orders of veins. The venation system (Fig. 3) may thus be
represented :

Secondary Tertiary

Midrib ^Lateral veins ^ veins ^veins, etc.

Venule reticulum

The above systems of venation, terminated by a venule reticulum, are termed
closed, but open systems, where a reticulum is not present, are found in some
dicotyledons.

The spaces in the venule reticulum are known as areolae, veinlet-islands, vein-
islets or inter-reticular spaces. Their shape and size vary considerably and, although
difficult to measure, are often the cause of the different appearance of various leaves.
Levin (1929) found that the average " vein islet area " is of considerable taxonomic
importance.

Although most, or all, venules join up to form the venule reticulum there are often
a number which end blindly, projecting into areolae, these being termed veinlet
terminations (venule terminations). Hall & Melville (1951) calculated the " veinlet
termination number " (number of veinlet terminations per square millimetre of
leaf surface) and found it of diagnostic importance. It was stated to be uncorrelated
with the number of areolae per square millimetre, and less variable over different
parts of the same leaf and hence of greater systematic value. Later work by the
same authors (1954) confirmed that the veinlet termination number was more
constant than the number of areolae, but found that the two showed a slight positive
correlation, and both a negative correlation with leaf size. Gupta (1961) improved
Hall & Melville's methods by devising the " absolute vein-islet number " and
" absolute veinlet termination number ", the product of the respective numbers
and the leaf area. This, of course, cancels the variation due to leaf size and, although
more laborious to calculate, should always be used when whole leaves are available.
In certain disciplines, such as pharmacognosy and palaeobotany, however, whole
leaves are often not available ; in these cases the non-absolute numbers should be
used with the greatest caution.

Leaf venation is treated from a different aspect below, when its position as a
cuticular character will become clear.

(2) Distribution of trichomes.

The very different degrees of pubescence shown by plants have given rise to a great
variety of descriptive terms (e.g. hirsute, pubescent, pilose, ciliate, puberulent,
etc.) which unfortunately have no universally recognized meaning. Using cuticular

CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 27

preparations the precise distribution and frequency of hairs and other trichomes
can be measured, i.e. the individual trichomes can be counted and their positions
noted. Taking into account the enormous variation that is encountered, only
differences of a considerable degree are likely to be of taxonomic value, and the
extremes of a range must be noted.

Precise data may thus sometimes replace such statements as " almost glabrous "
or " sparsely pubescent ", but in general their compilation will not be worth the
labour involved. Besides variation apparently uncorrelated with other factors the
degree of pubescence is dependent more than most characters upon the environ-
mental conditions and the age of the leaf.

Of much more value than simple trichome frequency is the relative frequency of
trichomes on different parts of the same leaf. As is well known, hairs are usually
more abundant on the midrib and veins of a leaf, and these regions are usually the
last to become glabrous as the leaf ages. However, this is by no means always the
situation. In the Combretum psidioides Welw. aggregate, for example, there are
three taxa (treated by Exell (1961) as subspecies of C. psidioides) which differ only
in their distribution of indumentum on the lower leaf surface : one taxon has a
densely pubescent venule reticulum and sparsely pubescent to glabrous areolae
(subsp. psidioides) ; the second a glabrous venule reticulum and tomentose areolae
(subsp. kwinkiti (De Wild.) Exell) ; and the third a tomentose venule reticulum
and areolae (subsp. dinteri (Schinz) Exell). Moreover, a further species in this
aggregate, C. grandifolium F. Hoffm., differs not only in its larger leaves but also
in having glabrous areolae and a sparsely pubescent to glabrous venule reticulum.
This taxon, demonstrating the fourth combination, appears to be a further sub-
species of C. psidioides. Dr. Exell and I have found that all four taxa belonging to
this group have apparently no other constant differences, either macroscopic or
microscopic, than the distribution of hairs. In many plant groups the distribution
of hairs on the leaf margin is frequently also a diagnostic character.

The broad type of trichome, e.g. sessile glands, stalked glands, branched hairs,
peltate hairs, long straight hairs or short curly hairs, etc., is also a character often
used without resorting to the use of a microscope, although this is not to be encour-
aged. Small differences, often vital, may be not apparent with a hand-lens or
binocular microscope, and very distinct taxa can easily be classed together. The
key to the genera of Sapotaceae by Hutchinson & Dalziel (1931), for instance, is
completely misleading due to a mis-statement on the type of hair present. I have
found that the hairs described as " simple (not medifixed) " (as opposed to medifixed,
stalked and thus " T-shaped " hairs) are in fact medifixed but sessile. Such errors
would have been avoided if a microscope had been employed. There have been
many erroneous statements concerning the trichomes of the Combretaceae. A
hand-lens should only be relied upon when two or more very different types of
trichomes are being searched for (e.g. in Epilobium), or special distinctive trichomes
are sought (e.g. in Urtica).

As previously mentioned, characters of the texture or overall appearance (includ-
ing colour) of the indumentum are lost in cuticular preparations.

28 CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY

(3) Cork-warts and similar structures.

Very frequently there are to be found on the cuticular preparations of leaves
structures which are usually referable to wounds of one sort or another, but which,
in some cases, represent cork-warts or related structures. Cork-warts are not
abundant in nature, and they are usually constant in occurrence, so they are frequently
useful in identification. In the mangrove genera of the Rhizophoraceae, for instance,
they are present in all species of Rhizophora but in none of the other three genera.
In West Africa, moreover, they are larger in R. mangle L. than in the other two species
of that genus found there, and are said to be diagnostic in the field (Keay, 1953).
They appear as small convex warts of corky tissue up to about 2 mm. across.

Microscopically the position of a cork-wart on the cuticular membrane is marked
by a circular or less regular hole surrounded by modified epidermal cells (Plate i A) .
The latter are usually, though not always, thicker-walled and much smaller than the
normal epidermal cells, and they are arranged in strict rows radiating for some
distance from the hole. It appears as if the radiating cells are meristematic, being
produced to repair the hole . These areas are thus very characteristic and conspicuous .

Unfortunately three types of wounds are extremely similar to the cork- warts.
Purely mechanical accidents are marked by variously shaped, but usually long and
scar-like, areas, the surrounding cells being identical to those of the cork-warts.
In many cases, presumably in the case of more ancient wounds, or perhaps when
only part of the outer wall of the epidermis has been removed .or damaged, a very
thin cuticular membrane, usually scarcely stainable, has grown over the wound,
this being either amorphous or with very thin cell walls. The cells on these areas,
if present, are usually normal-sized, although small cells would have been expected
if they were meristematic. Insect punctures are also frequently found on leaves,
these more often resembling cork-warts than accidental injuries, and probably not
being distinguishable from the former. They are of the same order of size as a small
cork-wart, and usually smoothly circular in outline. They were also reported by
Orgell (1955), who observed an identical structure. Moreover he found that they
frequently became occupied by corky or gummy plugs, showing a very close analogy
to cork-warts. The third type of wound which I have observed is a fungal attack,
found in the form of hemispherical shining excrescences on both leaf surfaces of one
specimen of Buchenavia kleinii Exell. The hole left by these fruiting bodies after
maceration is surrounded by the radiating rows of small cells typical of cork- warts
(Plate i B).

Although cork is more or less resistant to the macerating solutions used in the
present study, owing to its chemical similarity to cutin, it is not preserved on the
cuticular preparations as the wart or plug is not connected to the cuticular membrane,
but falls away during maceration. Where a wound has healed over, and occasionally
a cork-wart has dropped out, however, the tissues may become covered with a thin
cuticular membrane, as observed above.

Thus, although wounds have no intrinsic taxonomic value, they are important
as they closely resemble cork-warts, and thus render the latter considerably less

CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 29

useful in systematics. This equally applies to macroscopic and microscopic studies.

Other types of structures found on leaves, although often not macroscopic
characters, are sufficiently similar to cork-warts to be mentioned here. In many
cases they may be referred to excretory glands (e.g. salt glands, chalk glands, sessile
hydathodes, etc.), which may be macroscopically very conspicuous. The actual
gland is frequently only very thinly covered with a cuticular membrane, and the
adjacent cells may be radially arranged or otherwise modified. Holes may also be
found in the cuticular membrane due to the presence of modified epidermal cells
(e.g. calcified or silicified cells) which have no cuticular covering, but these are usually
surrounded by normal epidermal cells, as are the terminations of long idioblasts.
Cells containing crystals frequently break through to the surface of the leaf, and
these are often surrounded by radially arranged cells, as in wounds. In this
connexion the minute pimples and translucent spots common on some leaves and
often used by taxonomists may be mentioned. They are not epidermal characters :
the former are due to abnormally large epidermal or subepidermal cells, usually
containing crystals ; the latter to similar cells which interrupt the mesophyll.
In fact in most cases the translucent spots correspond with small pimples, although
either one may be present in the absence of the other, or rarely both exist together
on the same leaf but are unconnected.

Wound-like structures are frequently present on cuticular preparations in such a
regular manner that they are almost certainly not wounds, but no macroscopic
equivalent is visible and no microscopic subepidermal structures seem to be
connected with them. In Terminalia catappa L., for example, all specimens
examined showed ragged holes along the leaf margin (Plate i c), and T. chebula
Retz. possesses similar areas there and on the lower epidermis (but not the upper).
In T. bursarina F. Muell. similar areas are present in the major lateral vein axils
on the upper epidermis, corresponding to the presence of domatia on the lower
epidermis. They are caused by the growing of the domatia (see below) right
through the leaf to the upper surface.

Besides hydathodes, other water-secreting organs are the water-stomata, which
are usually distinctly different from the normal stomata in some way (e.g. by their
smaller or larger size). They are, of course, not macroscopic structures, although
they are in some families united into groups which appear as small surface spots
(Solereder, 1908). In the two Combretaceous mangrove genera, Lumnitzera and
Laguncularia, large stomata which may be water-stomata are found scattered over
the leaf surface. Usually they are surrounded by enormous subsidiary cells, and
around the latter the epidermal cells may be radially arranged (Plate ID). In
extreme cases the radially arranged cells form an extensive area resembling those
around a wound or cork-wart, and indeed in some instances the large stoma has
disappeared, leaving a hole. Solereder (1908), in fact, states that stomata may
become underlaid with a development of cork which ultimately erupts to give rise
to a cork- wart. Cork- warts have not been recorded from the Combretaceae,
however, and any connexion between the observed structures and cork is conjectural.

30 CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY

In addition Lumnitzera possesses wound-like areas along the leaf margin (cf.
Terminalia catappa] which appear to be distinct from those developing around the
large stomata, and which are also unconnected with the shallow marginal pits
thought to be domatia.

Finally, the presence in many instances of thin-walled areas, often only a few cells
in size, on the leaf epidermis may be noted. In some cases the component cells are
much more or much less undulate-walled than the rest of the epidermal cells.
Their origin is unknown.

It may thus be seen that the presence of " spots ", " pimples " or " warts " on
leaves may be due to any one of a whole range of epidermal (or subepidermal)
structures, which only a microscopic examination will determine. Their study is,
however, hampered by the presence of various types of wounds and punctures of
external cause which may extremely closely resemble the structural features and
thus limit the determinative value of the latter.

(4) Domatia.

The term domatium is used to denote a small cavity of very varying form, found on
almost any vegetative part of the plant, which is inhabited, or thought to be
inhabited, by some type of animal. Domatia have been almost completely neglected
from a taxonomic viewpoint in the past, and so limited is our knowledge of them
that they are treated rather fully here. It will be seen that on occasions they may
be of great diagnostic significance.

Domatia vary from cavernous passages in rhizomes, stems or petioles to minute
pits or hair-tufts on the under surfaces of leaves. The former are usually occupied
by ants, and the literature concerning these so-called myrmecodomatia is fairly
extensive (see Uphof, 1942). The smaller domatia, which may be up to about
4 mm. in diameter, are often said to be occupied by mites or similar arthropods,
although good evidence for this is usually lacking. They are, however, called
acarodomatia, a term suggested by Lundstrom (1887), and the plants on which
they occur acarophytes or acarophilous plants. Elliot (1911) stated that the mites
could be seen scurrying away from the appropriate regions of the leaves of many
common English trees when the former were disturbed with a needle, and Mani (1964)
reported that acarodomatia may be inhabited by mites such as species of Tarsonemus.

The literature relevant to acarodomatia is remarkably scanty, the only important
papers being a biological study by Lundstrom (1887) and a review by Penzig &
Chiabrera (1903). The latter account is excellent from many aspects, and thus any
sort of systematic review is unnecessary here. The authors reported acarodomatia
from about 425 species and 175 genera in 44 families, all woody dicotyledons. They
are most abundant in the Rubiaceae, in which Penzig & Chiabrera recorded 54
acarophilous genera. As these authors predicted, since 1903 many additional
genera have been shown to possess acarodomatia, but few are of very special note.

CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 31

An exception is Dioscorea, which is the only monocotyledonous genus and also the
only herbaceous one known to possess acarodomatia.

Apart from the fact that they are more or less confined to woody dicotyledons
there is no connexion whatsoever between the type of leaf and the presence or type
of domatia. They do perhaps tend to be rarer on leaves with inconspicuous
venation, but they are not absent from these and any generalization would have a
number of exceptions.

The commonest types of acarodomatia are small pockets or pits in the axils of
the main veins with the midrib on the lower leaf surface. Since the whole domatium,
whatever its shape, is lined with epidermis throughout, cuticular preparations
provide excellent transparent models of the domatia and afford the best method of
studying their anatomy. The preparations can be studied both with a high-power
and a binocular microscope. The domatia are very conspicuous on the cuticular
preparations since, where there is an extended piece of leaf tissue overarching the
main bulk, the cuticular membrane overlaps and so is three layers in thickness.

There is in all probability no system of classification that will accommodate all
types and grades of domatium. The usefulness of some sort of classification is,
however, obvious, and the one which I use to cover leaf domatia has proved
relatively satisfactory. Two main types of domatium are recognized (Fig. 4) :

Marsupiform (pocket-shaped] domatium (domatium marsupiforme] (Plate 2 A, B) .

All grades, from an extremely highly developed " pocket " to a situation which is
scarcely able to be called a domatium exist. They are found in the axils of the
major lateral veins with the midrib, or more rarely of secondary veins with major
lateral veins. In a leaf with very conspicuously raised veins a cuticular preparation
will always show folding of the epidermis of the veins over that of the areolae, and
this effect is of course doubled at a vein branch, especially the branchings from the
midrib. Leaves of Buchenavia reticulata Eichl., for instance, show this to a marked
degree. In some leaves this folding is only evident in the major lateral vein axils,
and here a V-shaped suggestion of a domatium is to be seen, as, for example, in
Terminalia mollis Laws. Penzig & Chiabrera (1903) classed this species as
domatium-bearing, but the case is debatable. In the genus Terminalia, as probably
in many others, a whole series of types from that of T. mollis to the most complex
pocket-shaped domatium is to be seen. This gradation can be visualized as the
gradual filling-in of the notch of the V to form a V-shaped triangular pocket where
all traces of the " arms " of the V have disappeared. These latter types bear no
relation at all to the conspicuousness of the veins, often being found on leaves with
an almost planar venation. The basal point of the triangle points towards the vein
axil, and the mouth of the domatium, forming one side of the triangle, away from it.
Thus the depth of the domatial cavity, which lies between the leaf surface and the
triangular area of tissue, is parallel to the leaf surface. The mouth of the pocket-
shaped domatium is therefore at its broadest and most distal point.

CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY

FIG. 4. Types of domatia : i, primary-axillary lebetiform. 2, primary-axillary
marsupiform. 3, 4, vertical sections of i and 2 respectively, at angle bisecting the
vein axil. 5, primary-axillary hair-tuft. 6, primary-axillary pubescent-marsupiform.
7, marginal lebetiform, in section (Lumnitzera) . 8, baso-laminar revolute (Strephonema) .
9, baso-laminar revolute (Dioscorea). 10, twinned supernumerary lebetiform (Terminalia
catappa) .

CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 33

Lebetiform (bowl-shaped] domatium (domatium lebetiforme] (Plate 2 c, D).

As with the preceding type all grades of bowl-shaped domatia are to be found, the
simplest of which might be visualized as a shallow depression in the leaf surface.
Such shallow depressions are found in Lumnitzera, on the leaf margin, and have been
referred to as domatia by Backer (fide van Steenis, in litt.}. In most cases, however,
bowl-shaped domatia are found in the axils of veins. In Terminalia plagata Merr.
well-developed domatia are present in the major lateral vein axils, but in the
secondary vein axils reduced forms, the shallow depressions visualized above, are
to be found. A series may again be traced where the opening of the hollow becomes
smaller, and/or the hollow itself becomes deeper and wider, the genus Terminalia
showing all types of intermediates. As noted by Penzig & Chiabrera (1903), in
some cases (e.g. T. microcarpa Decne.) the pore of the domatium may be at the level
of the rest of the leaf surface, the pit being entirely internal, whereas in other cases
(e.g. T. catappa) the pit is only partially sunken into the leaf tissue, the pore being at
the apex of a small dome. In the best-developed bowl-shaped domatia the pore is
extremely small in proportion to the large pit beneath (e.g. T. microcarpa}. In
the present type the depth of the domatium is thus at right-angles to the leaf surface.
The pore is usually in the centre of the domatium as viewed from above, but it may
be placed elsewhere. In some abnormal forms of domatium in T. catappa there are
two pores. The pore may be circular, elliptic or slit-like : in the last case it is most
frequently found on elongated domatia, the elongation of both domatium and pore
being parallel to the lateral vein.

The above two types of domatium are nearly always distinct, and it is suggested
that they arose independently, from the overarching veins and the shallow hollow
respectively. Intermediates, although rare, are apparently present, but they are
thought more likely to represent the points of convergence than of divergence. As
shown above, increasing stages in complexity of both types can be traced which are
difficult to interrelate. In apparently intermediate types (e.g. Terminalia moluccana
Lam.) the bowl-shaped domatium has a large pore which is distal to the vein axil, and
the cuticular membrane is often not trebled at its distal edge. The trebled membrane
thus forms a crescent rather than a circle, and moreover the depth of the domatium
appears to be parallel to the leaf surface, although it may not be so. This type
obviously closely resembles the pocket-shaped domatium. The latter may also
approach the intermediate situation by the narrowing of the opening. The
domatium of Buchenavia fanshawei Exell & Maguire, which is classed as the bowl-
shaped type, may possibly be a modified pocket-shaped type as all other species of
Buchenavia which possess domatia have the latter type.

Two other types of domatium may be recognized : the first often very well
defined ; the second of a highly dubious nature.

Re-volute (y oiled-mar giri) domatium (domatium revolutum).

In this type the leaf margin is inrolled, or folded under like the flap of an envelope.
Inrolled domatia are found in such plants as Hevea brasiliensis (Kunth) Muell. Arg.

34 CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY

(Penzig & Chiabrera, 1903) and Strephonema spp. (De Wildeman, 1923 ; confirmed
by my own examination). The inrolling is present at the base of the lamina, just
above the petiole on either side. The degree of inrolling varies considerably, and
the status of the domatium is thus dubious. In the flap-like domatia a specialized
portion of the lamina is folded under ; if this is folded back it forms a regularly
shaped projection from the leaf margin. It was found in dicotyledons (e.g. Ilex sp.)
by Lundstrom (1887) and later in Dioscorea by De Wildeman (1904). The latter
author described the structure in D. smilacifolia De Wild., in a new species
(D. acarophyta De Wild.), and subsequently, from 1912 to 1914, in ten other new
taxa. Burkill (1939) showed that these twelve represented three species only :
D. smilacifolia, D. minutiflora Engler and D. pynaertii De Wild., but he discovered
domatia in a fourth closely related species, D. praehensilis Benth.

Hair-tuft domatium (domatium fasciculatum) .

The vein axils of the otherwise sparsely pubescent or glabrous leaves of a large
number of species of trees may be densely pubescent or pilose, and these tufts of
hairs have been considered to be domatia by Lundstrom (1887) and others. If
mites do take refuge in them this may be accepted. Amongst British trees Tilia,
Alnus and Corylus are examples. In this type there is no modification of leaf tissue
other than the development of the hairs. Although many trees have leaves which
are at first densely pubescent and later scarcely so, and the last areas to lose the
indumentum are the veins of the lower leaf surface, the hair-tuft domatia are often
well defined, even if their association is dubious. They are frequently found in
connexion with the slightest development of a pocket-shaped domatium.

Lundstrom (1887) recognized more or less the same four types of domatia as I do,
although a much smaller range of forms was considered. Penzig & Chiabrera
(1903) closely followed Lundstrom, but described six types. The " hollow "-shaped
(fossette) and " pocket "-shaped (tasche or borsette) domatia were each divided into
two categories according to the presence or absence of hairs. It is not, however,
advisable to distinguish between glabrous and pubescent types because the difference
is apparently of no consequence and every type of intermediate is to be found.
Penzig & Chiabrera appear to separate the " hollows " from the " pockets " in that
the former are sunken into the leaf tissue with their openings on the level of the leaf
surface, whilst the latter (which may be either bowl-shaped or pocket-shaped in
my system) are raised above the leaf surface. Intermediates are, however, as
common as either extreme, and the classification is considered inferior to that used
in the present study. It so happens that Lundstrom's scheme does not commit
itself to agreement with either Penzig & Chiabrera's system or mine, since all his
" hollows " were in fact wholly sunken into the leaf, no raised hollows (as in
Terminalia catappa) being noted.

The latinized terms have been given for the convenience of concise description.
The position of the domatia can also be described by the following terms : marginal,
for those at the leaf margin (e.g. Lumnitzera) ; primary -axillary, for those in the

CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 35

major lateral vein axils (the most common type) ; secondary-axillary, for those in
the secondary or lesser vein axils (e.g. Tilia spp., Terminalia oreadum Diels and
T. plagata) ; baso-laminar , for those on either side of the lamina base (e.g. Quercus
robur L., Strephonema spp. and Dioscorea spp.) ; extra-axillary, for those just below
each major lateral vein axil (e.g. Rudgea lanceolata Benth.) ; and supernumerary,
for those not associated with any particular part of the lamina (e.g. Terminalia
catappa, where they are found scattered up the side of the midrib apart from the
vein axils ; and T. grandiflora Benth. and Conocarpus spp., where they are found
very sparsely scattered over the lower leaf surface). They are occasionally also
found elsewhere : for example in Schinus terebinthifolia Raddi domatia revoluta are
found on leafy wings to the rhachis of the pinnate leaves ; and in one or two genera
they have been reported from young twigs and petioles.

In most cases the epidermis lining and around the domatia is little or not modified.
In the axils of major lateral veins there is usually a patch of small straight- walled
isodiametric cells, and when a primary-axillary domatium is present this area is
merely enlarged, and this type of epidermis lines the domatium. In some cases,
however, the cells do differ in some way from those surrounding the domatium,
usually being smaller. In Terminalia catappa the adjacent cells are often arranged
in radial files for a short distance. In Conocarpus erectus L. var. erectus the epidermis
is absolutely identical in, near and away from the domatia, even to the extent of
the possession of stomata. In C. erectus var. sericeus Forsstr. ex DC. the epidermis
is very densely pubescent, but that lining the domatia is more or less glabrous,
as in var. erectus.

In many cases the domatia are associated with an abnormal amount of hair
development in, on or around the actual structure. In some species, e.g. Thiloa
gracilis (Schott) Eichl., the hairs are of a peculiar type not found elsewhere on the
leaf. In a few species gland-like structures are to be found in the domatium. These
were noted in Anacardium by Lundstrom (1887), and in the Combretaceae they have
been found in Buchenavia parvifolia Ducke and Terminalia archboldiana Exell,
though not very frequently. The nature and function of their secretion, if any,
are unknown. It is just possible that the deeply sunken glands found on the leaves of
Laguncularia are homologous with domatia, but this is not likely as they are found
on both epidermides and have an extremely narrow slit-like opening at the surface.

The examples cited above clearly show that domatia may have a considerable
value in taxonomy, although they have rarely been used. An extreme case is per-
haps that of Buchenavia fanshawei which has well-developed domatia lebetiformia.
Of the other 22 species examined in this genus 13 have domatia marsupiformia and
nine lack domatia altogether. Thus much use of this fact was made in constructing
a key to the genus (Exell & Stace, 1963). In the Combretaceae domatia revoluta
are found in all six species of Strephonema, which is usually treated as a separate
subfamily from the other genera, all of which lack these domatia. The remaining
genera belong to three groups, of which two contain eight genera each. In one of
these groups (Terminalieae) four genera possess domatia (three lebetiform and three

36 CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY

marsupiform) ; whilst in the other (Combreteae) all eight genera possess marsupi-
form domatia only. The third group (Laguncularieae) consists of three genera,
of which one possesses peculiar marginal lebetiform domatia. Thus in the
Combretaceae the occurrence of the various types of domatia to some extent agrees
with the generally accepted subdivision of the family. This is not the usual situa-
tion, however. Penzig & Chiabrera (1903) recorded all six of their types from the
Rubiaceae, and four from several genera, and in general domatia are useful in
identification at the specific level only.

There is no reason not to believe that all specimens of a domatium-bearing species
do possess domatia, although these may be sparser in some cases and may be lacking
altogether on some parts of the plant such as the sucker shoots. Bloembergen
(fide van Steenis, in litt.} found that Alangium chinense (Lour.) Harms and A. kurzii
Craib possess domatia in most of Asia, but not in Malaysia, but, as this situation
has never been reported elsewhere, the absence of domatia in Malaysia or the identity
of the species in the two areas needs careful checking. In the genus Terminalia,
of which I investigated 160 species, no such geographical variation was noted. For
example T. catappa, a native of Asia and Australasia, possesses very typical domatia
whether in its native countries or planted in Africa or America ; and T. arjuna
(Roxb.) Wight & Arn., a native of India, lacks them there or when planted elsewhere
in Asia, Africa or America. A different type of geographical variation, however,
was discovered in Terminalia. Of the 160 species studied 50 (31%) possess domatia,
of which 16 (32%) have domatia marsupiformia and 34 (68%) domatia lebetiformia.
However, in Australasia 21 (62%) of the 34 examined species possess domatia, and
the figures for other continents are : Malaysia 19 (58%) out of 33 ; Asia 4 (13%)
out of 34 ; Africa i (3%) out of 39 ; and America 7 (23%) out of 30. Thus as one
travels east or west from the islands between Australia and Malaysia the percentage
of domatium-bearing species drops sharply. The proportion of the two types of
domatia does not significantly differ from continent to continent.

Extremely little is known about the formation or function of acarodomatia, but
despite suggestions to the contrary it is obvious that mites play no part in their
formation. It is more likely that the mites inhabiting the domatia (if this is a regular
occurrence) are in a purely chance association. It is possible that they may enlarge
the domatia somewhat once they have taken possession of them, and this may be
the cause of the deep domatia in Terminalia bursarina which sometimes reach the
upper epidermis. There is no necessity, however, to consider that this is not a natural
plant process. Lundstrom (1887) claimed that the domatia of Psychotria, Coprosma
and Rhamnus developed in the absence of mites, but, although he was doubtless
correct, his experiments have never been repeated under the exacting conditions
necessary to exclude all mites. Mani (1964) also stated that the mites did not
participate in domatium formation. Bailey (1924) showed that American myrmeco-
domatia developed in the absence of ants. Whilst it is true that seedlings (and often
suckers) lack domatia, these developing only at a later stage, seedling and sucker
leaves often differ in many ways from the other leaves, and van Steenis (1953) states
that in Nothofagus domatia develop by the young sapling stage.

CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 37

Good evidence for the non-participation of mites in domatium formation is that
domatia are definitely formed in predisposed places: e.g. at the bases of the laminae
in Dioscorea, or just below each major lateral vein axil in Rudgea. If the domatia
were formed by mites they would not be so distributed, and neither would they
be so constant in their presence (on all individuals in all geographical areas) and
structure. In the case of primary-axillary domatia these are not always found in
all major lateral vein axils. They are always better-developed towards the lamina
base, but in some species, e.g. Terminalia bursarina, T. grandiflora and Nothofagus
menziesii Oerst., they are found only in the lowest axil on each side of the midrib,
or in the latter two sometimes also in the next axil. Furthermore, only certain
species in any one genus may possess domatia, e.g. only two species out of about
40 in Nothofagus (van Steenis, 1953) and four out of several hundred in Dioscorea
(Burkill, 1939). The domatia are often very hairy, and since many mite-produced
galls also possess this character, definitely as a direct result of the animals' activities,
it has been suggested that pubescent domatia are also formed in this way. There
is no reason to believe this, however, and in most Combretaceae, at least, the pubescent
domatia (often the only pubescent parts of the leaf) are relics of a wholly pubescent
juvenile state, and the hairs may wear off even here at senescence.

The very young leaf develops domatia marsupiformia by an extension of the
pocket tissue over the vein axil, but, if domatia lebetiformia evolved in the manner
previously suggested, ontogeny does not repeat phylogeny as the pits do not appear
to develop by an extension of tissue encroaching over a hollow. In the cases
examined (especially Conocarpus erectus) the pit forms internally, and at first is
only marked by a primary axillary pimple. Later a pore develops in the top of
the pimple, and this enlarges to a varying degree. The hollow appears to develop
internally. Different stages may be seen on one leaf, development being acropetal.
This is an important argument against any participation by mites in domatium
formation. Holtermann (1893) , however, seems to suggest that the pits of Conocarpus
do develop by means of an epidermal intucking, and not as I have observed.

The situation regarding domatia has been discussed at some length, although
numerous other points could have been included, since little is known of them in
general, and they have been largely neglected by taxonomists. With regard to
their position and structure they may be very useful in identification, but their
biological significance is highly obscure. They are doubtless often used as a shelter
by mites and other similar minute arthropods, but in the great majority of cases
no animals have been connected with them. This is in marked contrast to the case
regarding myrmecodomatia. The prefix " acaro- " is thus best rejected and replaced
by one suggestive of the organ on which the domatia are found, e.g. phyllodomatia.

B. MICROSCOPIC CHARACTERS.

(i) Epidermal cells.

There are a large number of directions of variation to be found with respect to the
epidermal cells, even omitting the several characters which are not apparent from
a purely surface examination, but they have rarely been used by taxonomists.

38 CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY

In certain regions the epidermal cells are modified, notably over the veins and leaf
margin and around the stomata and trichome-bases. This section deals only with
the areas not so modified.

Although the shape, size and other characters of the epidermal cells exhibit a wide
range of genotypic variation, which in many cases has definite taxonomic applica-
tion, the cells certainly show an extreme degree of phenotypic variation which is
elsewhere only paralleled by characters such as pubescence and petal colour, etc.

One of the most obvious features of the epidermal cells is their shape ; this is
often quite different on the two epidermides. This fact is often of systematic
importance, although Odell (1932) decided that it indicated that cell shape was
" of no diagnostic value ". The cells are usually isodiametric, i.e. more or less the
same in length and width, with any scattered elongated cells randomly orientated.
In some dicotyledon leaves, however, as well as in many groups of monocotyledons,
the epidermal cells are predominantly elongated. This feature usually occurs in
narrow leaves, the cells lying parallel to, or less often at right-angles to, the leaf axis.
In cases of isodiametric-celled epidermides some species have a characteristically
low (e.g. Guiera senegalensis J. F. Gmel., Plate 4 D) or high number of walls to each
cell, this factor markedly altering the appearance of the epidermis.

In some species the epidermal cells are arranged in small groups separated by
thicker than normal cell walls, where it appears that an originally single cell has
retained its identity after many subsequent divisions. Species of Avicennia show
this feature well. A related phenomenon is the anticlinal division of the epidermal
cells, where cells are divided by one or more thin straight walls which cut a normal-
sized cell into two or more parts. Such secondary division (i.e. occurring after
normal epidermal cell division has ceased, and not being accompanied by an increase
in size) is often taxonomically important. In the genus Laguncularia, for example,
it is absolutely constant, and it is not well developed elsewhere in the family
Combretaceae.

The term " wall " is here used as an abbreviation for the apparent cell wall
produced by the cuticular flanges, and it should always be borne in mind that the
actual cell wall was present for some distance on either side of these. For this
reason some workers prefer the term " cell outline ".

The anticlinal walls of the epidermal cells are either straight, curved or variously
undulate or sinuate (cf . Plate 3 A, B) . This is undoubtedly of taxonomic use in some
instances : it has been used as a major character in Lemna and is useful in many
groups of Combretum. For some unknown reason straight-walled epidermal cells
are commoner in xeromorphic plants than in mesomorphic ones, which typically
have undulate cell walls. The relative advantages and disadvantages of undulate
walls are unknown, but Wylie (1943), who attributes a conductive function to the
epidermis, suggests that their advantage is to increase the area of contact between
adjacent cells. If this is so it is surprising that undulate walls are not commoner
in xerophytes. The type of undulation is very various ; the " waves " may be

CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY

39

most easily measured in exactly the same terms as light waves, i.e. in the three
variables frequency, amplitude and wave-length. In the present study the frequency
was usually expressed in terms of wave-lengths per wall, but in some cases it seems
that the number of peaks (i.e. half wave-lengths) per cell may be a less variable
value. There may be considerable or very little variation in the degree of undulation
on one leaf, but the undulation often markedly decreases towards the leaf margin
and veins. It will be seen from Fig. 5 that the three variables mentioned above
aptly express all the characters of the undulations but one, i.e. the shape of the waves.
These may usually be described as U-shaped, V-shaped or Q-shaped. In some
species peculiar shapes of undulation are found with ornamentations, in the form of
knobs, ridges or T-shaped thickenings, on the outer side of the crest of each wave.
Curved epidermal cell walls may be described as having half a wave-length.

Frequency

1-0

ro

2-0

2-0

2-5

2-5

Wave-length

Amplitude

Vl

X /2

2x /s

2y

2y

2y

0-5 2x /

FIG. 5. Types of cell-wall undulation.

The exact causes of epidermal cell-wall undulation are unknown, the two main
theories involving the suggestions that it is due to the tensions set up between the
mesophyll and the epidermis, and that it is caused by the method of hardening
of the differentiating cuticular membrane (Watson, 1942). It has been found
that whereas many leaves have the epidermal cells undulate-walled throughout

40 CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY

the height of the cells, some have only the outer part of the wall undulate, and thus
the second hypothesis is by far the more likely. Sifton (1963) has provided further
evidence supporting Watson's theory. Amongst species examined in the present
study the situation where only the outer part of the wall is undulate is very con-
spicuous in Ramatuella. Watson, however, was able to produce either condition in
mature ivy leaves by varying the illumination during development, so that the
character is probably rarely of taxonomic use. In some species with straight walls
the isolated cuticular membranes show the tips of the cuticular flanges to be irregu-
larly undulate, but this is probably due to the effect of removing the epidermal
cells. It is especially noticeable in Conocarpus erectus.

As mentioned above, the cell shape is often different on the two epidermides, the
chief varying factor being the degree of undulation. In almost all undulate- walled
leaves the amount of undulation is greater on the lower than on the upper epidermis.
In many cases the former is markedly undulate-walled and the latter straight-walled.
Odell's (1932) remark that a few leaves of Acer pseudoplatanus L. may show the
reverse of this situation, which the bulk of the leaves exhibit, needs to be checked.
vSome species, however, do possess a markedly more undulate-walled upper epidermis
than lower, e.g. Terminalia microcarpa, T. plagata and others.

Epidermal cells vary tremendously in size, average widths being about 20-30 \i.
Even on typically large-celled epidermides, however, small cells are usually to be
found scattered. Some epidermides, in fact, are characterized by their extremely
variable-sized cells. That cell size might be of taxonomic value is indicated by the
knowledge that polyploids usually have larger cells than diploids, although this is
better shown by the stomata than other epidermal cells. Solereder (1908) mentions
the fact that certain species are characterized by a particularly small- or large-celled
epidermis, and this has been noted in the present study. Avicennia marina var.
rumphiana (Hallier f.) Bakh., for example, has a notably smaller-celled epidermis
than other varieties of this species. Lumnitzera littorea (Jack) J. O. Voigt has a
smaller average cell size than has L. racemosa Willd., although some overlap occurs
and this character, like all others, fails to separate the two species on surface cuticular
characters in all cases. The cell size is frequently different on the two epidermides, as
is cell shape. Solereder (1908) states that in general the upper epidermis is composed
of larger cells than the lower, but the opposite was found in the present study.
Characters such as the comparative sizes of parts of the same leaf are extremely
useful as the different parts are all equally affected by external conditions, e.g.
the presence of upper epidermal cells twice as large as the lower epidermal cells is a
better diagnostic character than the presence of upper epidermal cells of an average
length of 40 [X. The cell size usually decreases towards the leaf margin and apex.

Although the boundary between the cuticular membrane and the cellulose cell
wall is usually a smooth line, interrupted only by the cuticular flanges, in some species
regular cellulose pegs protrude outwards into the cuticular membrane, which becomes
very thin in these regions. This may or may not be accompanied by an extension
outwards of the cell lumen in the corresponding position. In either case small

CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 41

elliptic or circular areas are seen on the cuticular membranes where the thin
membrane is less deeply stained. Such areas are very well developed in Combretum
obanense (Baker f.) Hutch. & Dalziel and certain of its relatives, but in all other
species of this genus they are unknown. The thin areas are frequently present in
the bulge of each undulation, and in C. obanense they are so conspicuous that the
normal differentiation between cuticular flanges and non-flange areas is entirely
masked, except on the veins. The epidermis in this case appears to be composed
of minute cells about 5-10 \i across, with extremely thick walls. Other situations
have been described where the outer epidermal cells contain large pores, the cell
contents having been lost. The origin of the pores in the cuticular preparations of
Cinnamomum obtained by Marlier-Spirlet (1945) is obscure.

The surface of each epidermal cell may be flat or convex, or bear a small conical
process. The latter is known as a papilla, and the epidermides or cells which bear
papillae are said to be papillate, but unless the papillae exceed a certain length they
are not seen on surface preparations of cuticular membranes. They may, however,
be viewed on folded-over pieces of membrane. All gradations from simple conical
papillae to long structures better described as hairs exist, and if they are long enough
to become folded over on cuticular preparations they are visible in surface views.
They are often considered to be taxonomically important. The papillae are usually
centrally placed on the epidermal cell, but they may be found on one corner of
each cell, as in Combretum zenkeri Engler & Diels. Elsewhere in this large genus
papillae have been recorded in less than half a dozen species, e.g. C. lanceolatum
Pohl ex Eichl., and in all cases are of the central type.

(2) Leaf venation.

The epidermal cells above and below the veins are usually variously modified, so
that leaf venation is a cuticular character. Characters of the venation which are
microscopic alone, as opposed to the type of venation (which is a macroscopic
character), are the extent and nature of the modification of the epidermal cells.

The midrib and the lateral, secondary, etc. veins are usually enclosed or capped by
sclerenchyma, which may or may not reach either epidermis. In any case the
epidermis above and below these veins is usually markedly modified, in the majority
of cases the cells being elongated parallel to the veins. These elongated cells are
usually rectangular, and, even if the epidermal cells of the areolae have undulate
walls, straight- walled. In general the more xeromorphic a leaf the less conspicuous
are the veins on the epidermis. Usually the degree of modification of the epidermis
is proportional to the prominence of the veins, hence there tends to be a much
greater modification on the lower than on the upper epidermis, although this is not
always so. In most cases a more modified epidermis has more narrowed and
elongated cells over the veins. The ends of the cells may be truncate, oblique or
almost pointed and interlocking (prosenchymatous) ; and the cells arranged end to
end in numerous parallel rows, or less regularly placed. In many cases, especially
when the midrib is well developed on the epidermis but the cells are not very narrow,
the cells are secondarily subdivided at right-angles to the midrib, and in these cases

42 CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY

the cells may be squarish or even broader than long. Some leaves, most frequently
rather xeromorphic leaves with a prominent midrib but rather inconspicuous veins
and venules, have the epidermis modified over the former into strict rows of broader-
than-long cells, the primary and secondary cell divisions being indistinguishable
because all the cuticular flanges are equally thick. Examples found in the Combre-
taceae are Strephonema spp., Combretum demeusei De Wild., Terminalia laxiflora
Engler and many others. This effect is almost always more greatly developed on
the upper epidermis, and furthermore the cells of the veins of lesser prominence
show it to a progressively lesser degree.

The lesser veins and venules showing no directional tendencies have a different
anatomical structure. Above and below, or below only, many of them are bounded
by a dorsiventral extension, the vein-extension or vein-rib, which consists of living
cells elongated parallel to the venules and abutting on to the epidermis below and
often also above. Wylie (1943) considers that these, as well as the epidermis,
constitute a conductive system supplementary to the venous system. When a
venule is connected to the epidermis in this way the latter is modified. Thus it
may be seen that, as with the major veins and the midrib, the upper epidermis is
modified to an equal or lesser extent than the lower epidermis over the lesser veins
and venules. The modification of the cells over the venules (Plate 5 B) is similar
to but much less developed than that over the midrib, etc. The cells are generally
elongated, to varying extents, although secondary division of them is rare or absent.
When the epidermis is mostly composed of undulate-walled cells the venule cells
are usually straight-walled, but sometimes they are slightly, or even equally,
undulate- walled. In the latter case only their elongation and, on the lower epider-
mis, the lack of stomata make them distinguishable.

Thus every intermediate between the very long and narrow midrib cells to the
scarcely modified venule cells exists. The cells in the centre of each vein are mostly
greatly modified, the edges of the veins gradually or abruptly merging into the
normal epidermal cells. The most modified midribs are several hundreds of cells
wide, the smallest venules only one or two. In an average mesomorphic leaf such as
those of many species of Combretaceae the following degree of modification is seen :
midrib composed of about 50 parallel rows of narrow elongated straight-walled cells ;
major lateral veins about half as wide, the cells shorter and wider ; secondary veins
about 10-15 cells wide, the cells only about twice as long as broad ; lesser veins
about 5-10 cells wide ; venules forming venule reticulum about 2-3 cells wide,
the cells little longer than broad, many with slightly undulate walls. Variation
in one extreme is shown by strongly xeromorphic leaves, e.g. of Laguncularia,
where only the midrib is modified on the epidermis, and that not conspicuously ;
and in the other extreme by species with strongly reticulate leaves, e.g. Buchenavia
reticulata, where the venules forming the reticulum are 5 or more cells wide, the cells
being straight-walled and somewhat narrow and elongated. In all cases the lower
epidermis is more greatly modified than the upper, so that the venation of the upper
epidermis of a leaf would resemble that of the lower epidermis of a more xeromorphic
leaf, and so on.

CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 43

From the preceding paragraphs it can be seen that the three variables mentioned
(i.e. the venation pattern ; the degree to which the venation system is recognizable
on the epidermis ; and the type of modification of the cells) are completely indepen-
dent of each other. Because of this, caution must be exercised in interpreting the
extent of the venation system seen on the epidermis as that to be found in the leaf :
it is frequently not so. The smallest veins represented on an epidermis, whether
they are in fact the smallest venules of the leaf or of a much larger order, tend to be
of the 2-3-wide slightly elongated-cell type. Thus a venule reticulum with areolae
3000 [A across seen on an epidermis might represent this structure in the leaf, or
smaller venules may be present but not differentiated on the epidermis. According
to the excellent figures of Wylie (1943), and to my own observation, the ultimate
veinlet terminations rarely have vein-extensions, and are thus rarely visible on an
epidermis. Apparent veinlet terminations seen on the epidermis are due to venules
which are not represented in full.

Although the vein-like structures seen on the epidermis are of course only modified
epidermal cells, they are for convenience referred to as midribs, veins and venules,
etc. Thus the " upper epidermal midrib " or " lower epidermal venules " may be
referred to in descriptions. The unmodified epidermal islands within the venule
reticulum may likewise be referred to as the areolae. In cases where the venules
are not well represented on the epidermis, and the venation system is thus apparently
open rather than closed, the unmodified epidermal cells may be referred to as the
non-venous areas. In general the differentiation of the veins on the epidermis is
subject to rather more phenotypic variation than is the organization of the actual
veins themselves, calling for a greater measure of caution in their taxonomic use.

(3) Leaf margin.

The epidermis is continuous over the leaf margin, connecting the upper and lower
epidermides, but its cells are modified at these positions, providing useful taxonomic
characters. Towards the margin the cells of both epidermides are frequently smaller
and thicker-walled, and if the normal cells are undulate-walled the undulations
become less pronounced. Thus the marginal cells either abruptly or gradually
merge into the normal epidermal cells.

The width of the margin, of course, is partly dependent upon the thickness of
the leaf, and may be of diagnostic value. In most mesomorphic leaves the margin
is not very wide, often not exceeding 5 or 6 cells across. The cells are in most cases
rectangular, straight-walled and variously elongated (Plate 3 c) . Secondary division
of these cells may be apparent, giving rise to squarish or broader-than-long cells.
The margin is thus frequently very similar to a venule, and may be indistinguishable
but for its continuous straight appearance and the absence of other venules branching
from it.

In certain cases, especially in xeromorphic leaves, other types of leaf margin are
to be found. One of the most common of these is composed of small isodiametric
cells, frequently with very thick walls and more or less circular lumina. Sometimes

44

CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY

these are arranged into parallel rows as is normally the case, but in other leaves
they may be quite irregularly arranged. All intermediates are to be found, but in
some cases the margin provides valuable systematic characters. The two Com-
bretaceous mangrove genera, Lumnitzera and Laguncularia, for example, are closely
similar in cuticular characters, but one of the three main differences lies in the leaf
margin. In Laguncularia (Plate 3 D) the cells are much smaller than those elsewhere,
have thick walls and circular lumina and are arranged completely at random. In
Lumnitzera, however, the marginal cells are scarcely smaller than the other epidermal
cells, have little-thicker walls, have angular lumina, are rectangular and slightly
to conspicuously longer than broad, and are always arranged end to end in parallel

rows.

stomatal aperture

outer stomatal ledge

T-piece at
stomatal pole

walls of
guard-cells

radial ) walls of
^subsidiary
tangential j cells

FIG. 6. Terminology of stoma as seen in surface view.

(4) Stomata.

The term stoma is here taken to include the pair of guard-cells with the stomatal
aperture or pore between them. In all vascular plants there are normally two guard-
cells to each stoma, these arising from a single stomatal mother-cell at an early
stage of leaf development, although abnormalities with one or three or more guard-cells
do rarely occur (Dehnel, 1961; Ahmad, ig64a; Pant, 1965) and may even be of
diagnostic value. Other cases have been described of bistratose guard-cells which
have divided by a periclinal wall (Rehfous, 1914). The guard-cells are surrounded
by normal or modified epidermal cells, which in the latter case are termed subsidiary
cells or accessory cells. The whole system is known as the stomatal complex, or
stomatal apparatus. In surface view the guard-cells, in the great majority of
vascular plants, and in almost all dicotyledons, are vaguely reniform. The anticlinal
wall on the side towards the pore is here described as the poral wall, and that towards
the rest of the epidermis the epidermal wall. If subsidiary cells are present their
anticlinal walls abutting on the guard-cells may be known as radial walls, and those
adjacent to the rest of the epidermis as tangential walls (Fig. 6).

CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY

45

In vertical transverse section the guard-cells show an extremely various construc-
tion which is surprising in view of their relative constancy in surface view (Fig. 7).
The guard-cells vary in three main directions : their position relative to the
subsidiary or other adjacent cells ; the relative thickenings of the various walls of
the guard-cells ; and the pattern of cuticular ornamentation of the guard-cells.
All these are important taxonomic characters. In most mesomorphic plants the
guard-cells are on a level with or very slightly sunken below or raised above the rest
of the epidermis, but in some xeromorphic plants they are often sunken to a
considerable degree. In many conifers, as well as in a number of angiosperms, the
guard-cells are very deeply sunken and completely overarched by the subsidiary

back cavity

sub-stomatal
cavity

inner stomatal ledge

poral wall
of guard-cell

inner

epidermal

/ walls of
^guard-cell

subsidiary cell

front cavity

outer wall of guard-cell
outer stomatal ledge

stomatal aperture
FIG. 7. Terminology of stoma as seen in transverse vertical section.

cells, so that in a surface view the former are almost invisible, their position being
marked by a ring of subsidiary cells (known as a poral ring) around a nearly circular
pore. In other plants the subsidiary cells are also sunken. The thickenings of the
guard-cell walls are very uneven in distribution, being connected with the physiology
of the stomatal movements. In general, however, the poral and epidermal walls
are rarely thickened, the former never, whilst the other two walls (i.e. the outer
wall, adjacent to the environment, and the inner wall, adjacent to the inner regions
of the leaf) are variously thickened. In some species, mostly xeromorphic, the
thickenings are very considerable, and the cell lumen may be reduced to an extremely

46 CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY

small proportion of the cell volume. The amount of cuticular ornamentation of
the guard-cells is frequently proportional to the thickening of the cell walls, and
hence to the degree of xeromorphism. The cuticular membrane which covers the
epidermis is continuous over the subsidiary and guard-cells, and either ends where
the inner and epidermal walls meet or also covers a varying amount of the
subepidermal cells which line the space which usually exists below the stoma (the
sub-stomatal cavity] . In most cases the cuticular membrane is very thin on the poral
walls, but thick on the outer and inner walls, especially where each of these meets the
poral wall. Here the cuticular membrane may be protracted into a pair of cuticular
or stomatal ledges, ridges, horns, rims or hooks, which may be termed outer and inner
respectively. This is found in many xeromorphic plants, but more frequently only
the outer stomatal ledge is present. When the stoma is closed the poral walls of
the two guard-cells meet, and above and below this point the outer and inner
stomatal ledges also, if present, meet, or nearly so. Thus two extensions of the
stomatal pore are delimited : a. front cavity and a back cavity (Fig. 7). In some plants,
e.g. Bruguiera spp. (Areschoug, 1902) and Ceriops spp. (Stace, 1963), there are
actually two pairs of outer stomatal ledges subdividing the front cavity into two
portions.

In cuticular preparations the stomata are usually well preserved, and they provide
valuable systematic characters (cf. Plate 3 A, B). Because much stomatal variation
is in the third dimension (depth), however, a number of characters are not usually
observable on cuticular preparations. It is often very difficult to ascertain the
exact relative positions of the guard-cells and their neighbouring cells and the
precise structure beneath the outer stomatal ledge. The cuticular membrane below
the latter, since it curves back over the inner stomatal wall, is conspicuous on many
preparations as a pair of stomatal flaps. A very good idea of the degree of cuticulariz-
ation and sunkenness of the guard-cells and subsidiary cells may often be obtained
by examining cuticular preparations with the membranes folded over on themselves
inside outermost (Plate 5 c) . Also a surprising amount of structure can be observed
once the many lines and ridges on the preparation have been interpreted by means of
vertical sections. This approach was useful in the Rhizophoraceous mangroves,
which have many diagnostic details of stomatal structure. The single known
epidermal difference between Lumnitzera littorea and L. racemosa, the shape of the
guard-cells and subsidiary cells in section, however, was not visible from surface
views of either cuticular preparations or whole epidermides.

However, in surface view the most obvious cuticular characters are present on the
outer walls of the guard-cells. The shape of the guard-cells as a whole is a useful
criterion which is usually constant within species. The commonest shapes are
circular, elliptic and oblong, but some stomata may even be broader than long, and
sometimes angular. The stomatal poles where the two guard-cells meet may be
obtuse, truncate, rounded or retuse, the last apparently being the case in many
xeromorphic plants. The appearance of various thickenings on the outer guard-cell
walls is also of importance. The outer stomatal ledge is usually visible from surface
views, and its thickness and relative distance between the epidermal and poral walls

CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 47

are characters of significance. When this ledge is directly over the poral wall of the
guard-cells it forms a thick rim to the stomatal aperture which Bandulska (1923,
etc.) called the poral rim. At the stomatal poles there may be developed a T-shaped
thickening or T-piece, much used by Bandulska (1924) as a character in the Fagaceae.
Sometimes only the upright or the cross-piece of this is present. In some groups
there is a small thin area at the polar end of each guard-cell, being circular, elliptic
or comma-shaped. Solereder (1908) mentioned several examples and Rao (1939)
found them commonly in the Magnoliales. A few genera are known to possess
peculiar lobe-like extensions of the guard-cell epidermal walls, and these were
noted in Strephonema.

Perhaps of more significance than any of the above characters, however, are the
presence and types of subsidiary cells around the stomata. In his various studies
Florin (1931, 1933) distinguished two types of stomatal development in the gymno-
sperms, based on whether or not the guard-cells and subsidiary cells originate from
the same mother-cell (syndetocheile and haplocheile respectively), and these types
can usually be seen in the mature state because the former has two adjacent sub-
sidiary cells and the latter a variable greater number. Although it was frequently
necessary for Florin to draw conclusions concerning ontogeny by examining the
mature anatomy, since many of his plants were fossils, it is unfortunate that it was
so because the two terms are now more often used for describing the mature anatomy
than the developmental sequence, and it has been shown by many workers that
the presence of subsidiary cells does not necessarily signify a syndetocheile ontogeny,
nor does their absence always indicate a haplocheile one. Thus the terms syndeto-
cheile and haplocheile are better abandoned, and the cells neighbouring a stoma
(whether they be subsidiary cells, i.e. different from the other epidermal cells, or not)
described as mesogenous or perigenous according to whether or not they arise from
the same mother cell as the stoma. These terms were first used by Florin (1933).
In addition Pant (1965) proposed the term mesoperigenous for those situations where
one neighbouring cell is mesogenous and the rest perigenous. This terminology has
the advantages that it is equally applicable to all groups of plants rather than
primarily to the gymnosperms, and that it allows for the fact that mesogenous
subsidiary cells may become indistinguishable from other epidermal cells, or that
perigenous cells may become specialized as subsidiary cells. The last term has no
ontogenetic implications.

The earliest system of subsidiary cell classification was adopted by Solereder (1908)
from the work of Prantl (1872) and Vesque (1889). Six ontogenetic categories were
recognized, three of which were perigenous and three mesogenous. Due to convergence
during maturation only four types were recognizable in the mature state, being
termed Ranunculaceous, Rubiaceous, Caryophyllaceous and Cruciferous. Metcalfe &
Chalk (1950) completely ignored the modes of development and renamed these four
categories as follows : anomocytic, with no distinctive subsidiary cells (Plate 4 A) ;
paracytic, with 2 subsidiary cells lying parallel to the guard-cells (Plate 4 c) ; diacytic,
with 2 subsidiary cells lying transversely to the guard-cells; and anisocytic, with
3 subsidiary cells of which one is distinctly larger or smaller than the other two.

48 CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY

These groups are often of considerable taxonomic significance, in many cases one
type being characteristic of a whole genus or family. In the Combretaceae, for
instance, Strephonema is the only genus with paracytic stomata, and this agrees
with the usual treatment of Strephonema as a separate subfamily. Apart from the
four types of situation mentioned above there are other arrangements of subsidiary
cells for which no terms are generally available, although for descriptive purposes
they are obviously desirable. The term actinocytic was proposed by Metcalfe & Chalk
(1950) for cases where the subsidiary cells, usually four or more in number, are
elongated radially to each stoma, and cyclocytic by Stace (1963), where a similar
number of cells forms a narrow ring round each stoma. Both terms have subsequently
been used by other workers, although they probably refer to mere modifications of
the anomocytic type. Of the 19 genera of Combretaceae without paracytic subsidiary
cells, in fact, two (Lumnitzera and Laguncularia] are cyclocytic (Plate 4 B) and the
rest anomocytic. Bandulska (1931) also found cyclocytic subsidiary cells (" girdle
cells ") in various genera of Myrtaceae. More usually, however, differentiation by
subsidiary cell types is only at the specific level, or is even here absent. In Anopyxis
klaineana (Pierre) Engler (A. ealaensis (De Wild.) Sprague), for example, paracytic,
anisocytic and anomocytic stomata are said to occur on the same leaf (Metcalfe &
Chalk, 1950), and other members of the Rhizophoraceae also show varying
arrangements. It seems probable that only one basic mode of development is
involved, but that the degree of subdivision of the subsidiary cells finally differs,
producing the different arrangements.

A reliable ontogenetic classification is highly desirable, and that proposed by
Pant (1965) seems to be satisfactory. Ten categories are recognized, of which one
is perigenous, three mesoperigenous, and six mesogenous. This nomenclature is
quite separate from one based upon mature topography. The latter scheme would
involve a larger number of categories (perhaps about 15), several of which could
be derived from several different modes of development. Apart from the six terms
mentioned previously one other has so far been named: tetracytic (Metcalfe, 1963),
for cases where there are four subsidiary cells, two polar and two lateral, as in many
monocotyledons. Terms such as paracytic, tetracytic, etc., could conveniently be
prefixed by meso-, peri-, etc., once the mode of development was known. Members
of the Rubiaceae are meso-paracytic (Pant, 1965), for instance, and of the Gramineae
peri-paracytic (Stebbins & Jain, 1960; Stebbins & Shah, 1960; Stebbins & Khush,
1961). All investigated monocotyledons, in fact, have been found to be perigenous.
It seems clear that the developmental studies at present being carried out mainly
by Pant and his compatriots (e.g. Pant, 1965; Pant & Mehra, 1963; Pant & Verma,
1963) will provide valuable evidence for a classification of stomata which will prove
to be of considerable taxonomic and phylogenetic value.

In some plants the subsidiary cells are furnished with distinct papillae, which may
be absent elsewhere on the leaf. These may be quite long and overarch the stomatal
aperture (Solereder, 1908).

Stomata vary considerably in size and this is often a character of some importance.
The relation between the level of ploidy and stomatal size is well known, but stomatal

CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 49

size does not always indicate the former (Carolin, 1954) . In many species the stoma-
tal size also varies on a single leaf, but this may not prevent it from being an
important taxonomic feature. On many leaves abnormally large stomata are to be
found, and these are often, if not almost always, water -stomata, which are supposed
to secrete drops of liquid water. Their presence is noteworthy. Water-stomata, or
water-pores if the surrounding cells are not obviously guard-cells, may also be smaller
than the normal stomata, and sometimes occur in groups (Solereder, 1908).

The distribution and frequency of stomata are their most conspicuous characters,
and are often of considerable systematic value. They are very often, of course,
connected with the ecology of the species, and thus not likely to be of great phylo-
genetic importance. In many hydrophytes, for example, stomata are either absent
or infrequent and vestigial. In some terrestrial plants also some stomata may be
vestigial, and Solereder (1908) describes situations where they may be plugged with
a resinous mass. In plants with floating leaves they are usually on the upper
epidermis only, and the same is true of other species with variously modified leaves
which for some reason have a more concealed lower than upper epidermis, or even
with apparently normal leaves (Solereder, 1908). In most dicotyledons, however,
the stomata are more or less confined to the lower epidermis, although it appears
to be usual to find a few scattered along the vicinity of the upper epidermal midrib
of mesomorphic leaves. Solereder also noted this, and wisely pointed out that it
could have " no great systematic value ". They are present there in most species
of Combretaceae, and are frequently abnormally elongated. They are sometimes
present upon the midrib itself, rather than beside it, and cannot here be functional.
All intermediate situations from the complete absence of stomata on the upper
epidermis to their presence as abundantly as on the lower epidermis are to be found,
even within the Combretaceae alone. Lumnitzera spp., in fact, have stomata more
abundant on the upper than on the lower epidermis. Species showing a slight
increase from normal in the abundance on the upper epidermis have stomata
scattered quite frequently alongside the midrib, more sparsely alongside the veins,
and very sparsely in the rest of the areolae, and the further from the midrib they are
situated the less abnormal is their anatomy. On the lower epidermis stomata are
usually confined to the areolae, or non-venous areas, but sometimes they are to be
found scattered on the midrib. They have seldom been recorded from the leaf
margin of a dorsiventral leaf, or from the venules and veins of the lower epidermis.

Thus stomata are typically found in the areolae of the lower epidermis, and
usually only very sparsely anywhere else. In leaves showing various xeromorphic
characters, especially where a closed system of venules is not apparent on the
epidermides, they may be otherwise distributed. In the latter case they are frequently
found all over the epidermis, except for the veins and midrib, but in some species
they may be present only in special areas, e.g. longitudinal grooves, or depressions.
These latter situations are not found in the Combretaceae, although here there are
some species with very prominent reticulations and thus sunken areolae. The
position and types of sunken areas, the stomatal crypts or stomatal pits, may be very

50 CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY

useful in taxonomy (cf. Morley, 1953). They are quite independent of the individu-
ally sunken stomata found in a number of xeromorphic leaves.

In leaves with stomata confined to areolae or to stomatal crypts the stomata are
not regularly orientated, their long axes being quite randomly directed. In some
leaves, however, especially narrow leaves or those with stomata present in longi-
tudinal grooves, the stomata are all orientated parallel to the leaf axis. In a still
smaller group of species the stomata are orientated at right-angles to the leaf midrib,
e.g. Laguncularia racemosa (L.) Gaertn. f. These characters are usually of absolute
constancy and of sufficient scarcity to be valuable in identification.

The frequency of stomata, either in the areolae or in the non-venous areas, is a
much used character, and frequently misused. A large number of workers have
used the stomatal frequency in order to distinguish between species, with varying
degrees of success. Timmerman (1927), for example, found the character of little
use in Datura since the frequency varied considerably on different portions of the
leaf. Environmental and other factors also cause variation in frequency. Valuable
taxonomic evidence, however, may often be obtained if the stomatal index (iooS/
E + S), based on the proportion of stomata to epidermal cells, is used instead of the
frequency per unit area (Salisbury, 1927), as has been shown by various workers in
modern and fossil plant taxonomy. In Lumnitzera, however, the two species only
differ on stomatal frequency since the stomatal indices of the two species are
identical, the stomatal frequency being greater in one due to the smaller size of all
the cells.

(5) Types of trichomes and trichome-bases.

Despite some attempts by Odell (1932) to prove otherwise, trichome anatomy
provides a very important group of taxonomic characters. The definition of trichomes
as outgrowths of the epidermis is not absolute as there is every grade between these
and the so-called emergences which differ in also involving some subepidermal
layers in the outgrowth, but in the present context both types are included.

The extent to which trichomes are cuticular characters depends upon the degree
to which they are cuticularized. In the case of most emergences and of trichomes
with multiseriate basal parts a high proportion of the outgrowth is preserved on
cuticular preparations, and the structure can be observed quite easily. Thus the
whole or most of such outgrowths is cuticularized, only the more distal regions being
sometimes not so covered. In the case of trichomes with pauci- or uni-seriate basal
parts a similar proportion of the outgrowth may be cuticularized, but quite frequently
only the very basal parts or even none at all possess a cuticular membrane. There
are, of course, many exceptions to these generalizations ; for example the very long
unicellular hairs of Macropteranthes leichhardtii F. Muell. are well cuticularized along
their whole length. Where cuticularization is extensive cuticular preparations
may provide the best means of studying the trichome anatomy.

Outgrowths with large multiseriate bases are not readily broken off at the base,
but when this happens an irregular hole appears in the cuticular preparation.
Trichomes with pauci- or uni-seriate bases are much more easily removed in toto

CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY

from the epidermis, the appearance of the remaining trichome-base (i.e. the adjacent
modified region of the epidermis, exclusive of any parts of the trichome itself)
depending upon the original organization of the basal parts of the trichome, of which
there are two main situations (Fig. 8) . Firstly the trichome may rest upon a normal
or modified epidermal cell or group of cells, or, in the case of some unicellular
trichomes such as root-hairs, the trichome may in fact not be separated from the
epidermal cell by a cell wall. In these types the former position of a lost trichome
will be indicated on the cuticular membrane by a modified epidermal cell or cells,
by a hole in an epidermal cell, or will not be visible at all. Secondly the base of the
trichome may be inserted between the epidermal cells, or may replace an epidermal
cell, so that a lost trichome leaves a pore in the cuticular preparation.

scarcely modified
foot

multicellular
foot

FIG. 8. Four types of trichome-base as seen in vertical section.

Trichome-bases (hair-bases, gland-bases, etc.) of the above types are common
features of cuticular preparations, especially as the trichomes are usually covered
with a thinner cuticular membrane than the rest of the epidermis so that slight
over-maceration removes the trichome completely. Their organization provides
useful diagnostic characters, and it has been found necessary to create a terminology
for the parts of the trichome-base (Fig. 9). The trichome-base cells may or may not
surround a pore. In the former case the trichome-base cell walls may be termed
poral, radial and epidermal, the first of these often being considerably thickened and
then known as the poral rim.

CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY

A trichome occupying the pore of a trichome-base therefore consists of two parts :
the foot, which is inserted into the pore ; and the body, which extends above the
epidermis. The foot may thus be a single cell, a small group of cells, or a peg-like
basal portion of a cell. The parts of the body of the trichome have been given very
different names by various authors according to the trichome structure (e.g. stalk,
head, branches, arms, disk, fringe, rim, etc.) and generalization is impossible and in
act undesirable.

base of poral rim

hair-base cells

poral rim

pore

scarcely
modified foot-cell

epidermal )

poral Walls of hair-base cells

radial )
FIG. 9. Terminology of two main types of trichome-base.

There have been many attempts to classify trichomes (cf. Bachmann, 1886 ;
Solereder, 1908 ; Foster, 1949 ; Seithe, 1960, 1962 ; etc.) but none has been
generally successful. As mentioned previously, not only does each trichome type
grade into the next but there is no real distinction between papillae, trichomes and
emergences. The most usual division, between glandular and non-glandular types,
suffers from the same disadvantages as other systems in that there is no sharp
distinction between the groups, and in any case the division often separates obviously
very closely related types. Furthermore, within each group or subgroup the direc-
tions of variation are often quite different in various plant taxa, so that the only
way to avoid the situation where a single trichome could belong to more than one
group on different points of anatomy would be to create a system with many groups
each with numerous parallel subgroups. It is thus quite clear that no useful general
classification of trichomes can be constructed, but that instead it is advisable to
create systems for each particular plant group investigated as it becomes necessary,

CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 53

and to employ special descriptive terms in each case without attempting to correlate
the terms from taxon to taxon. This has the advantage that newly discovered
types can easily be fitted into the system at any time.

Good accounts of the ranges of variation and of the details of anatomy and func-
tioning of trichomes are given by Solereder (1908) and Netolitzky (1932) respectively,
and so no sort of review is given here. Instead an indication of the types of problems
and the degree of taxonomic usefulness which may be involved in trichome studies
is given by a brief statement of two major points of interest in the Combretaceae. In
all twenty genera of this family peculiar unicellular hairs with distinctive basal
components (known as " Combretaceous hairs ") occur, elsewhere being known only
in a few genera of the Myrtaceae and of the Cistaceae. These are of great interest
for they virtually delimit the Combretaceae from all other families except the other
two mentioned above, and thus indicate a monophyletic origin of this family. The
Myrtaceae, moreover, are generally considered to be closely related to the Combreta-
ceae, and the hairs thus support this view. The presence of the hairs in the Cistaceae,
however, is difficult to understand since this family has never been considered at all
closely related to the Combretaceae. The identity of the Combretaceous and
Cistaceous hairs thus needs to be investigated. In Strephonema, a Combretaceous
genus usually placed in a separate subfamily from all the other genera, typical
Combretaceous hairs are present in one of the species, thus confirming its inclusion
in this family. In the other four species, however, they are absent and are replaced
by hairs unique in the Combretaceae, so that the hairs also support the treatment of
Strephonema as a separate subfamily.

The second point concerns the presence of " scales " (Plate 5 A), which are short-
stalked trichomes with a disk-shaped multicellular head found in a wide range of
plant families (Bachmann, 1886). These trichomes are often extremely distinctive,
but in some families they intergrade with stalked globular glands, with stellate
hairs, or with cup-shaped hairs. In the Combretaceae scales are found in all species
of the small genera Thiloa, Calycopteris and Guiera, and in over half of the large genus
Combretum. These four genera all belong to the same tribe (Combreteae], and the
scales do seem to indicate, therefore, that this classification is well founded. In the
Combretaceae the disk of the scales is always one cell thick, but varies in the degree
and manner of subdivision, giving rise to a very wide range of types. In Rhododend-
ron, a genus of the very distant family Ericaceae, however, the main direction of
variation is in the overall organization of the scale rather than in its division in one
plane, producing an equally varied range of forms (Cowan, 1950 ; Seithe, 1960).
Furthermore, in Rhododendron there are intermediates between all of the numerous
(25 according to Cowan ; 43 to Seithe) types of trichomes, whereas in the Combreta-
ceae there are three main types (Combretaceous hairs, scales and stalked glands)
which are always quite distinct (Stace, 1965). Thus Cowan concluded that all the
trichomes of Rhododendron are phylogenetically directly related, especially as they
all appear to pass through identical early stages of ontogeny. Although this may
be correct in Rhododendron I do not consider it to be the case in Combretum, where it
is not possible to construct a " phylogenetic tree " of trichome types as did Cowan.

54 CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY

The stalked glands of the Combretaceae (Plate 5 B) are found in all the scale-less
genera of Combreteae except Pteleopsis, in all the scale-less species of Combretum
itself, and also in Conocarpus, a genus in another tribe. Their presence in the last
genus and absence from Pteleopsis is puzzling, since the two tribes are otherwise
well denned on a number of characters. The presence and types of stalked glands
or scales are thus very useful taxonomic criteria in Combretum and in other genera
of the Combreteae (Stace, 1961, 1963, 1965 ; Exell, 1963) . In both Combretum and Rhodo-
dendron the major taxonomic positions of trichome differentiation are at the sub-
generic and sectional level, and where a section contains species with unexpectedly
divergent trichome types there is good reason to doubt the homogeneity of that group.

Compared with trichome structure trichome size is relatively unimportant, but
nevertheless at times very useful. For example the Combretaceous hairs of the
Combretaceae are relatively very uniform in structure, but those of Macropteranthes
leichhardtii are very distinct in their large size, and those of Guiera senegalensis
extremely so in their very small size. The length of hairs generally is, of course, a
frequently used taxonomic character.

The trichome-base frequently provides characters of importance, most notable of
which is the shape (i.e. the degree of modification) of the trichome-base cells. In
some cases they are scarcely modified from normal epidermal cells, but usually they
are elongated radially to the pore, when this is present. If the normal epidermal
cells are undulate-walled the radial walls of the trichome-base cells are very often
straighter or straight, and frequently their epidermal walls also. Although the
number of trichome-base cells varies considerably within one species, in many cases
there is a typical range, of which 5-8 is the commonest, which may aid in the
diagnosis of the plant. Some species, for example, have a characteristically low or
high number of trichome-base cells. In some species the trichomes are so numerous
that almost every epidermal cell is a trichome-base cell, and radially elongated. In
Guiera (Plate 4 D) many upper epidermal cells are adjacent to two or three hair-bases,
and there are no unmodified upper epidermal cells. The latter are thus typically
elongated in two or three directions, and the cuticular preparations of this genus
are quite distinct in the whole family. In many Combretum species (e.g. C. glutinosum
Perrott. ex DC. and several related species) almost all of the lower epidermal cells
are scale-base or hair-base cells.

Quite frequently, especially in xeromorphic species, the trichome-base cells are
not radially elongated, but have become tangentially divided so that there are
several rings of modified cells, the inner of which is composed of cells longer in the
tangential than in the radial plane. The cuticular flanges of these walls are usually
thick and straight. Sometimes the two different types of trichome-bases mentioned
above form valuable diagnostic characters between species, although in other cases
this is not so and too much emphasis should not be placed on them. In Terminalia
glabrescens Mart., for example, both extremes and all degrees of intermediates may
be found.

CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 55

The size of the pore, where present, varies considerably, but in some cases may
characterize species. In Anogeissus acuminatus (Roxb. ex DC.) Wall, ex Bedd., for
example, the pores are very large (often over 50^ across) and are a quite distinctive
character of this species. The thickness of the poral rim varies in two dimensions,
i.e. towards the centre of the pore (Plate 5 A), and towards the centre of the leaf. In
the former case it may result in a very small pore of only 1-2^ in diameter. The thick
rims are frequently, though not always, accompanied by a considerable thickening
of the radial walls of the trichome-base cells, so that a stellate pattern of thickening
is obtained. In extreme cases, where the trichome-base cells are of the tangentially
subdivided type, all the walls of the trichome-base are conspicuously thicker than
the rest of the epidermal cell walls. The poral rim often becomes deeper by a
development of cuticular membrane around the foot of the trichome, often up to the
inner wall of the epidermis. The diameter of the pore is nearly always narrower at
the base than at the apex, so that the sides and base of the pore can be seen in
surface view.

Where there is more than one type of trichome on a leaf the types of trichome-bases
bearing them are often different (Plate 50). In Combretum, for example, where
hairs are always accompanied by either stalked glands or scales, the poral rims of the
latter are much thicker and deeper than those of the hair-bases, and the two can be
readily distinguished. Moreover the distribution of the scales is different from that
of the stalked glands, the latter being commoner on the veins, the former in the
areolae. Thus the abundance and distribution of each type of trichome can be
ascertained even if the trichomes have been lost. In Conocarpus erectus there are two
types of bases : those with radially and those with tangentially elongated cells.
Stalked glands are found on all of the latter types, on intermediate types and on a
few of the former types ; whilst Combretaceous hairs are confined to the former
types of bases. Thus in this genus the difference is not so well marked.

(6) The cuticle.

Besides providing a means of studying epidermal anatomy in the absence of any
cellular layers the cuticle often possesses characters of its own on its outer surface,
and the cuticular membrane also varies in the degree of its development. All these
characters are worthy of study as potential taxonomic criteria.

The most obvious feature is the degree of development of the cuticular membrane,
as has been mentioned in several preceding sections, notably the cuticular orna-
mentation of the guard-cells and trichome-bases. The thickness of the cuticular
membrane over the normal epidermal cells is also a character of diagnostic importance.
In most cases thicker membranes are characteristic of plants of drier habitats, and
are thus a measure of xeromorphy, although this is not always so. In Lumnitzera
racemosa and Macropteranthes kekwickii F. Muell., for instance, the wall between the
lumen of the epidermal cells and the exterior is quite thick, as might be expected in
these two xeromorphs. However, in the former the cuticular membrane is thick
and has long stout cuticular flanges, whilst in the latter the cuticular membrane is

56 CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY

thin with scarcely any flanges, the non-cutinized epidermal wall accounting for the
greater part of the thickness of the outer compound wall. Because of its relation
to the environment, however, the degree of development of the cuticular membrane
is liable to considerable phenotypic variation, and great care must be exercised in
its use in taxonomy. The cuticular flanges show the same degrees of genetic and
phenotypic variation and may also prove of diagnostic value in some cases, besides
determining the over-all appearance of the cuticular preparation and whether or not
the epidermal cell outlines are visible (Plate 50). In Combretum psidioides, for
example, the cellular organization of the scales is always more or less completely
indiscernible, whilst in all other closely related groups it is clearly visible on cuticular
preparations. Again, xeromorphic species usually have well-developed flanges, but
some species have thick cuticular membranes yet very short and narrow flanges.

The other character of the cuticular membrane is the type of minute ornamentation
of the outer surface of the cuticle, this having been used by a number of past workers.
The vast majority of plants have a smooth or minutely granulated cuticle when
viewed with a light microscope, and no use can be made of this fact beyond the
absence of any definite markings. The latter, as seen in many mesomorphic plants,
take the form of a series of minute grooves and ridges which are commonly termed
striations from their appearance in surface view. Their presence and degree of
development do not always seem to be related to xeromorphy, although some
species of Combretaceae have been found in which the degree of striation is not
constant. In their less-developed state the striations are usually found running
parallel to the midrib and veins, and radially to the trichome-bases and stomata, and
the presence of such a pattern is probably of little importance. In general the
striations bear no relation to the cell outlines, each marking passing over a number
of cells, but exceptions have been found where each striation is always contained
within a single cell outline. When the striations are better developed they appear
on the non-venous areas unconnected with the trichomes or stomata, and their
orientation, arrangement and degree of development may be very various. Where
they are found in areolae they are usually contained within one areola, i.e. they do
not continue across the venules. Where the veins are less extensive on the epidermis
the striations may be much longer. They are extremely conspicuous, for example,
in Macropteranthes kekwickii and M. montana (F. Muell.) F. Muell., where they are
found on all parts of the epidermis. In section the surface of the cuticle appears
papillate. In extreme cases, several of which have been found in Combretum, the
whole cuticle is covered with striations, those radial to the trichomes and stomata
and those parallel to the veins being continuous over other areas of the epidermis
(Plate 5 D). Sometimes the striations are so strong, and the cuticular flanges so
weak, that the former obscure all signs of the cell outlines.

In Combretum zenkeri the long papillae have a conspicuously echinate cuticle.
The series of raised cuticular anastomosing ridges reported by Solereder (1908) in
several families appear to be rare.

CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 57

6. DESCRIPTION OF CUTICULAR CHARACTERS

Just as the formal description of a new plant species follows a well-defined and
well-known pattern, so should the characterization of a cuticular membrane, although
hitherto this has not been the case. Two types of description are required for various
purposes : a relatively full account of the cuticular membrane characters ; and an
abbreviated account or diagnosis. A suggested form for the former, after many
adjustments, is as follows :

1. Distribution of species ; habitat.

2. Morphology of the leaf :

a. general morphology as seen with the naked eye.

b. special morphology venation pattern and prominence, trichome types
and distribution, clear spots, pimples, etc.

3. Account of the cuticular membrane :

a. general features thickness, ease and quality of preparation.

b. shape, arrangement and size of upper epidermal cells of :
i. venation system ;

ii. non-venous areas.

c. shape, arrangement and size of lower epidermal cells of :
i. venation system ;

ii. non-venous areas.

d. shape, arrangement and size of margin cells.

e. distribution, frequency, orientation, size and structure of stomata, in-
cluding the subsidiary cells.

f. distribution, frequency, size and structure of trichomes and trichome-
bases.

g. presence of any other features, e.g. cork-warts, hydathodes, water-stomata,
domatia.

h. presence of markings, e.g. striations, on the cuticle.

4. If necessary : interpretation of surface cuticular characters, primarily of the
stomata, by means of leaf sections. A brief mention of the organization of the
hypodermal and inner layers of the leaf might not be out of place as a general
background to the cuticular characters.

Such a description is an obvious prerequisite for the use of cuticular characters in
taxonomy and identification, and full allowance must be made for environmental and
other variation. Characters which are thought to be likely to vary beyond the
limits observed should be indicated.

For the purposes of a diagnosis the third section above, i.e. the description of the
cuticular characters themselves, should alone be used, and this in an abbreviated
though essentially similar form to that above. To this end certain abbreviations
may be employed, comparable to Ivs. , fls. and fr. in other descriptions. The following
are suggested : epid. (epiderm-is, -ides or -al) ; 1. and u. (lower and upper) ; cutic.
(cuticle (s), cuticular or cuticular membrane (s), unless these need to be distinguished) ;

58 CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY

stom. (stom-a, -ata or -atal); ven. (venule(s)); ret. (reticulum) ; w.l./wall (wave-
lengths per wall) ; ampl. (amplitude) ; isod. (isodiametric) ; 2ry (secondary) ; lat.
(lateral(ly)) ; longit. (longitudinal (ly)) ; trans, (transverse (ly)) ; tang, (tangential
(ly) ; rad. (radial(ly)).

In some cases these procedures may need to be modified, just as with normal
taxonomic descriptions, but in general their use would lead to a greater uniformity
and thus ease of reference and comparison. A diagnosis of a typically meso-
morphic species is given here :

Diagnosis of Combretum fruticosum (Loefl.) Stuntz.

Cutic. medium, thickness ; features clear. U. epid. : midrib broad and conspicuous, of many
longit. rows of narrow elongated cells ; lat. veins conspicuous but narrower ; minor veins and
some ven. distinguishable but ven. ret. absent or fragmentary ; cells of non-venous areas
averaging c. 30 (j. across, isod., with straight walls, or less often with undulate walls with up to
i^ w.l. /wall and 5 \L ampl. L. epid. : midrib and lat. veins broader and more conspicuous than
on u. epid. but minor veins and ven. not so much more conspicuous and ven. ret. still ill-defined ;
cells of areolae averaging c. 20-25 V- across, isod., with faint undulate walls with up to 2-J- w.l. /wall
and 8 pi ampl. Margin of a very few rows of slightly elongated rectangular, straight-walled
cells. Stom. very frequent in non- venous areas of 1. epid., sparse alongside u. epid. midrib,
randomly orientated on 1. epid., c. 22-30 x 15 jx ; epid. walls very faint ; poral walls quite
conspicuous ; stom. ledges absent ; poles mostly rounded, sometimes retuse or obtuse ;
adjacent epid. cells 4-7, unmodified. Combretaceous hairs frequent on midribs and lat. veins
of both epid., but sparse elsewhere ; hair-bases with pore c. 10-25 ^ across ; poral rim some-
what thickened ; hair-base cells mostly 5-9 and slightly rad. elongated, with straight, slightly
thickened rad. walls ; internal compartments long and pointed. Scales bowl- shaped with
raised convex cutic. membrane, frequent in 1. epid. areolae, sparse on 1. epid. venous areas and
on u. epid., there frequently commoner on or near midrib ; scale-bases with pore c. 10-25 V-
across ; poral rim considerably thickened ; scale-base cells c. 8-13, considerably rad. elongated
with thin straight rad. walls ; scales c. 100-200 (JL across, circular in surface view, scalloped at
margin, divided into c. 35-70 cells by rad. walls alone, few of the cells reaching the scale centre,
all reaching the scale margin ; stalk of scale uniseriate.

7. VARIATION OF CUTICULAR CHARACTERS

The main disadvantage of many taxonomic characters is that they will vary
independently of the genotype, and it is thus important to discover to what extent
and under which conditions this variation will occur. In all groups certain characters
vary so much that they are useless in taxonomy, but since these characters are not
the same ones in every plant group every character is worthy of study. With regard
to cuticular characters variation is due to three major causes : the age of the leaf,
or rather its degree of maturity ; the environment in which the plant is situated ;
and the position of the leaf upon the plant, that is to say the internal and external
environment of the leaf. In general the characters affected by these three factors
are those of size, frequency and degree rather than the actual anatomy or organi-
zation of the particular structures. Thus the majority of the characters previously
discussed show relatively little of this type of variation. Much of the variation
shown within species is of course genetically controlled ; this must be equally
carefully studied, and, if possible, distinguished from the environmental and develop-
mental variation.

CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 59

In the case of variation due to age it is obvious that in extremely young leaves
every cuticular character would be quite unlike that character on a mature leaf, and
such immature leaves are usually of no taxonomic use with regard to either macro-
scopic or microscopic features. The present section therefore only deals with leaves
that are likely to be chosen for cuticular examination, i.e. leaves which are over
half-grown.

The relevant literature shows a generalized sequence of the development of the
various cellular layers of a typical simple leaf to be established (Avery, 1933 ; Esau,
1953 ; Slade, 1957 ; etc.). The leaf primordium arises as a bud-like outgrowth of
the apical meristem of the stem, and develops by its own apical meristem into a
finger-like projection which corresponds to the petiole and midrib. The lamina
later grows out from this as a pair of lateral wings, this activity of a marginal meri-
stem usually commencing before the cessation of that of the apical meristem, in
most cases before the leaf is a millimetre long. When apical growth has ceased the
leaf elongates by intercalary growth. The thickness of the lamina in terms of cell
number is determined at a very early stage in development, when the leaf is a few
millimetres long, and subsequent increase in cell number is solely by anticlinal
divisions. Very soon after the lamina has begun to develop the major lateral veins
appear in its central portion, branching from the midrib and keeping pace with the
lamina extension by intercalary growth. The lesser veins and venules develop
similarly at progressively later stages. Recent work (Pray, 1963) has shown that
the veinlet terminations also develop in the same manner, and not by dissociation
of parts of : the venule reticulum in the expanding leaf as was formerly thought to
be the case.

Before the leaf approaches its mature size cell division ceases altogether, the
remaining increase in leaf size being by cell expansion alone, and, although some of
the layers are distinguishable before cell division ceases, most of them acquire their
characteristic form after this point. The first parts of the leaf to end cell division
are the two epidermides, followed by the spongy mesophyll and lastly the palisade
mesophyll. Not all regions of the leaf stop producing more cells at the same time,
moreover, the leaf apex reaching maturity first and the base last, and superimposed
on this sequence the leaf margin continues cell division longer than the midrib region.
The tip of the leaf is thus the most mature, and the marginal parts the least. Since
the epidermis is the first layer to show a cessation of cell division, cell enlargement
shows its greatest development there. Stomata develop from epidermal initials
soon after the number of cells in the lamina thickness has been established, and their
mature shape is attained in most cases before that of the other epidermal cells.
The times and rates of trichome development usually appear to run parallel to those
of the stomata, and in almost all the cases investigated trichomes and stomata had
reached their mature structure before the leaf was its mature size. Thus one of the
latest features of the epidermis to appear is the mature epidermal cell shape and
size.

Examination of data from a wide range of taxa has led to the conclusion that there
are six cuticular characters which are regularly to be expected to show sufficient

6o CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY

environmental and developmental variation to pose serious taxonomic difficulties.
These are discussed separately.

A. EPIDERMAL CELL SIZE.

A number of workers (e.g. Yapp, 1912 ; Salisbury, 1927 ; Watson, 1942) have
found that the epidermal cells are larger on leaves in more humid or more shaded
situations, and Odell (1932) cites workers who also discovered a decreasing cell size
with excess carbon dioxide, drier air, drier soil and greater altitude. In general
these size differences are proportional to a change in leaf area. The cell size also
decreases with an increased height of insertion upon the axis of many herbs (Yapp,
1912 ; Stober, 1917 ; Odell, 1932 ; Turrell, 1942 ; Ashby & Wangermann, 1950 ;
etc.). Although this might seem to reflect the more shaded and humid environment
of lower leaves Ashby & Wangermann (1950) concluded that the difference in cell
size was not dependent upon the climate when the leaf unfolds, even though in
ipomoea at least the later the time of seed sowing the larger were the epidermal cells,
but found (i95oa) that the factors affecting this character operate mainly by in-
creasing the period of cell division in upper leaves. Ashby (1948) discovered that
the decrease in cell size was accompanied by a decrease in leaf area and cell number,
and although the latter were affected by water supply the former was not, since
plants in dry and wet conditions produced the same-sized epidermal cells. This is
contrary to the results of most of the above workers.

Some investigations into the relation between cell size and leaf maturity were
made by me on a number of species, using North's cellulose acetate film technique
(North, 1956). In the two main species investigated, Quercus robur and Carpinus
betulus L., the leaves grow by cell division and enlargement until they are about
one-third their mature size, when cell division ceases and subsequent leaf enlarge-
ment is entirely due to cell enlargement. In species with undulate epidermal cell
walls, e.g. C. betulus, the cells appear for a time to be enlarging more rapidly than
the leaves. This apparent anomaly is due to the fact that cells can increase in
length and breadth without any increase in area while the wall undulations are
developing (cf. Fig. 10).

Secondly, the nature of the abnormally small leaves found at the base of most
sucker shoots was investigated in Quercus robur, Buchenavia capitata (Vahl) Eichl.
and Platanus X hybrida Brot., cells from fully mature leaves of as wide a size range
on the same sucker as possible being examined. No significant cell-size difference
whatsoever was found, an increase in mature leaf size of a factor of three producing
no increase at all in the cell size in all three species. Representative figures obtained
from Quercus and Buchenavia are :

Quercus Buchenavia

Lf. size Av. cell Igth. Lf. size Av. cell Igth.

9-0 X5-O cm. 30-0 [x 3-5 X2-o cm. 42-5 [A

6-0 X3-o cm. 28-0 [i 2-5 Xi-3 cm. 41-5 y.

4-5 X3-0 cm. 31-0 (A i-o xo-6 cm. 40-5 y.

3-3 X2-o cm. 28-5 (a.

CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 61

This is in direct contrast to the results obtained by Turrell (1942) and others working
with herbs showing typical heteroblastic development (i.e. varying leaf types up the
stem) where the upper and lower leaves differed in a number of size characters
besides the overall leaf size, and thus indicates that the small leaves at the base of
suckers do not represent normal heteroblastic growth. Since different-sized leaves
on a single sucker differ in cell number but not cell size the factors regulating the
mature leaf size must take effect early on by determining the point at which cell
division stops. The same results were obtained from abnormal suckers on coppiced
trees of Quercus which possessed leaves over twice as long and twice as broad as
those on normal suckers, yet which had epidermal cells of the same size.

Finally, the average epidermal cell sizes of different parts of a single leaf were
measured, when in every case the cells increased in size with the distance from the
leaf apex and from the leaf margin. Representative measurements for Quercus
robur are : extreme leaf apex, 29 X22 (x ; extreme lamina base, 36 xsi [J. ; extreme
margin, half-way from base to apex, 30 X20 (x ; half-way from base to apex and
from margin to midrib, 34 X27 \JL.

These observations merely serve to emphasize the variability of epidermal cell
size not only with age and minor genetic variation but with the environment,
position of the leaf on the shoot and the position of the cells in the leaf. However,
there does appear to be less variation than might be expected with respect to the
cells of leaves on a single shoot which attain different sizes at maturity, and this is
encouraging from the angle of systematic anatomy. Should the character appear
to be of taxonomic use great care must be taken to use strictly comparable mature
material, and to ascertain the degree of variation to be found. Thus with herbarium
material this feature should not be used unless very well marked.

An additional complication, the secondarily subdivided epidermis, is unlikely to
cause any confusion as the nature of the extra walls is usually obvious (see 2 above) .
They presumably develop long after normal cell division has stopped, and represent
a renewal of meristematic activity.

A very similar situation to the above exists regarding stomatal size, although in
this case the leaf size is not wholly and directly dependent upon the character.
Measurements indicate, in fact, that the stomata may become mature in size
and anatomy before the epidermal cells have fully expanded, and so the use of
absolutely mature leaves may not be so essential. Various workers, especially
those previously mentioned, have found that stomata become longer nearer the
base of a herbaceous plant, in shade, in moist air and in moist soil. These results
are not always capable of repetition, however, and some workers (e.g. Odell, 1932)
found no difference in stomatal size on sun and shade leaves. Thus the variation
of this character in the particular group concerned should be fully ascertained
before it is taxonomically employed, but the evidence does suggest that the average
stomatal size is less subject to phenotypic variation than that of the epidermal cells.

62 CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY

B. STOMATAL FREQUENCY.

Since the number of stomata that appear on a leaf is determined at the end of the
period of cell division, long before a leaf reaches its full size, the stomatal frequency
expressed as the number of stomata per unit area clearly decreases as the leaf
expands. Again, therefore, only mature leaves are suitable for comparative
measurements of stomatal frequency. Moreover the stomatal frequency often
varies considerably on different parts of the same leaf, and on different leaves of the
plant. Salisbury (1927) found that the lowermost leaves of a plant, and the basal
and the midrib regions of the leaf, had a lower stomatal frequency, which was
reflected to an equal extent on both epidermides when the upper epidermis possessed
stomata. The increased frequency of stomata on higher (smaller) leaves of herbs has
also been noted by Yapp (1912), Stober (1917), Rea (1921) and Turrell (1942), but
Odell (1932) and others have found that the frequency increases towards the base
and midrib of each leaf.

Environmental factors also regulate the stomatal frequency, the latter being
lowered by humid air, wet ground, shade from sun and wind, lower altitude, and a
lesser concentration of carbon dioxide (fide Odell, 1932).

Fortunately, however, the above variation can apparently be almost entirely
cancelled by recording stomatal frequency in terms of the proportion of stomata to
epidermal cells :

No. of stomata

TT 7 : 7 r^ -, , -. 17- xioo =Stomatal Index.

No. of stomata + No. of epidermal cells

Since the variation in stomatal frequency caused by the above factors is apparently
due to the increase or decrease in epidermal cell (and to a lesser extent stomatal)
size, but not to cell number, the stomatal index is not affected. Salisbury (1927),
who was the first to use this value, found that one factor did, however, alter the
stomatal index the humidity. It is not known how widespread this phenomenon
is, but it seems that the diverse variation shown by stomatal frequency can often
be more or less overcome. More work is needed, however, to establish that the
stomatal index is always as invariable as has been claimed : Yapp's (1912) data,
for example, would suggest that it is not. Gupta (1961) also used a value which
cancelled the effect of the position and environment the absolute stomatal number.
This is obtained by calculating the product of the stomatal frequency per unit area
and the leaf area, and can therefore be used as an alternative to the stomatal index.
Reyenga & Karstens (1964) described an unusual situation in the sepals of
Hydrangea, which are small and pink during anthesis, but which afterwards turn
green and enlarge considerably. This period of growth includes not only the
formation of new stomata but also the disintegration and disappearance of some of
the original ones, the stomatal index rising due to the predominance of the former
process. Moreover, the original stomata have actinocytic subsidiary cells with
radial cuticular striations, whilst those of the later stomata are anomocytic and
not striated. Furthermore, Dehnel (1960) reported in Begonia the disintegration of

CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 63

stomata which have been wounded. Phenomena of this kind are probably
exceptional, however, and need rarely enter into consideration.

Rea (1921) found that in Campanula rotundifolia L. the hydathodes showed the
same positional and environmental variation as the stomata.

C. TRICHOME FREQUENCY.

The general situation regarding trichome density is similar to that of stomatal
frequency in that usually the number of organs is determined when the leaf is very
young and the frequency thus drops as the leaf expands. Thus a densely pubescent
young leaf may develop into a sparsely pubescent mature leaf. Yapp (1912),
however, found that in Filipendula ulmaria (L.) Maxim, there was an additional
development of hairs after the leaf began to unfold, the original pubescence being
developed in the leaf bud. As in the case of stomata the pubescence often varies
from leaf to leaf, from one part of a leaf to another, and under a variety of environ-
mental conditions. The variation in trichome frequency on different parts of one
leaf is a well-known and important taxonomic character. Yapp (1912) and Stober
(1917) both found that the pubescence on plants generally rose with increasing
distance up the aerial shoot : in many plants, in fact, the radical leaves are almost
glabrous and the upper cauline leaves conspicuously pubescent. The inner radical
leaves are also usually more pubescent than the outer.

The same workers, in general, as those who have found a variation in other
characters, have also documented the variation of trichomes in different environ-
mental conditions (fide Odell, 1932). Hairs have been found to be more abundant
in greater sunlight, greater wind exposure, drier air, drier soil and greater altitude.
McDougall (1927), however, reported that far more hairs developed in reduced
sunlight in Lactuca biennis (Moench) Fernald (L. spicata auct.). It is possible that
in some of these cases the proportion of hairs to cells does not vary, but at least in
many instances it does. In general there is no possibility, therefore, of the use of a
trichome index comparable to the stomatal index ; trichome density is a much more
variable character than stomatal frequency, and its taxonomic use is thus more
restricted. Only differences of a considerable degree or differences in the relative
distribution of trichomes are reliable. The one advantage of a measurement of hair
density in terms of hairs per cell rather than of hairs per unit area is that the former
cancels the effect of age in any one leaf, except as noted below. As in the case of
stomata the number of hairs per epidermal cell may remain constant as the leaf
expands, and it is unfortunate that this value does not also remain constant in
different environmental conditions and positions on the plant.

A further complicating factor concerned with trichome frequency is the
phenomenon of glabrescence : the gradual loss of hairs with age. As explained
previously most hairs may be localized on cuticular preparations by the presence
of a hair-base, the centre of this often being occupied by a pore. It was thought
that the number of hair-bases per epidermal cell might remain constant throughout
the life of a leaf, even if most of the hairs themselves dropped off. Unfortunately

CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY

in most cases this was found to be untrue. Observations were made chiefly on a
number of species of Buchenavia, Conocarpus and Combretum ranging from
extremely small young leaves, from which cuticular membranes were difficult to
prepare, to senescent leaves that had begun to fall from the trees. In the species of
Buchenavia examined, notably B. capitata and B. kleinii, the very young leaves are
uniformly densely pubescent or sericeous, whilst the very old leaves are mostly
glabrous with a few hairs still remaining on the midribs and major lateral veins.
Microscopic examination of the former showed that almost every epidermal cell
was adjacent to one or more hair-bases, but as the leaf expanded and the hairs began
to fall off the hair-bases gradually became occluded, and in the mature state the
majority were not visible at all. The more actively growing (i.e. the more immature)
a leaf the more quickly occluded are unoccupied hair-bases. Occlusion is effected
by the hair-base cells encroaching inwards until no pore remains. When loss of
the trichomes is delayed until after leaf growth has ceased, however, the vacated
hair-bases often remain distinct thereafter, as is the case with the diagnostically
important scales of Combretum spp., although this is not true of the hairs of
Conocarpus erectus var. sericeus.

Thus the use of trichome frequency as a taxonomic character must be preceded
by an exhaustive series of observations on a wide range of material in order to
discover the cuticular changes accompanying growth and senescence, as well as the
degrees of phenotypic variation.

D. EPIDERMAL CELL SHAPE.

The epidermal cells gain their characteristic shape gradually from the time that
cell division ceases onwards, and in general this is completed before the process of
cell expansion has ended. Certain types of cell shape, however, may be finally
assumed after the leaf has reached full size. The only important example of this is
the undulation of the lateral epidermal cell walls. Since the undulations occur
equally on either side of an imaginary mid-line, i.e. into each of the two adjacent
cells, the apparent length of the cell can increase (by twice the value of half the
amplitude) without any increase in volume (Fig. 10) .

original length = I
original breadth - b
original area = Ib

original cell wall

final cell wall

amplitude = x
final length = / +x
final breadth b + x
final area = Ib

FIG. 10. Diagram illustrating growth of cell-wall undulations.

In most cases examined, however, the development of undulations takes place
whilst the leaf is expanding, and can give rise to the apparent phenomenon that the

CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 65

cells are growing faster than the leaf. The final amplitude of the undulations is
usually reached as the cells (and the leaf) attain full size, although the maximum
growth of the undulations ceases some time before this. In Buchenavia capitata,
however, the leaves at flowering time, which are fully grown but considerably
thinner and more pubescent than when mature, have conspicuously less undulate
walls than those at fruiting time, and in this case at least the undulations do increase
in amplitude without any increase in cell area. This is most apparent on the upper
epidermis. According to Watson (1942) it is the stresses set up by the hardening
cuticular membrane which cause the plastic cell walls to become undulate, and when
the cells have become mature their walls are too rigid to allow further increase of
undulation. It is thus clear that absolutely mature leaves must be examined in
order to ascertain the degree of undulation of the cell walls.

Much variation in this undulation has been noted in various environmental
conditions, strongly undulate to straight walls often being observed within a single
species. Brenner (1900), amongst the earlier workers, found that considerable
differences could be found between plants in " normal " and damp conditions : in
two species of Crassulaceae the normally straight-walled cells were converted to
markedly undulate-walled cells, and in a third species the exactly opposite result was
obtained. A considerable number of workers have noted that the amplitude of the
undulations increases with increased shade, humidity and soil dampness (Odell,
1932). Watson (1942) found that epidermal cells of Hedera in sunny situations had
6-3 crests per cell, whilst those in shade had 8-7. Moreover the latter extended
throughout the height of the cells, whereas the former were confined to their outer
edge. This was explained on the basis that in the sun the cells become mature
and rigid more quickly, so that the hardening cuticular membrane does not have an
effect to such a great depth as in the shade.

Yapp (1912), Stober (1917) and other workers (fide Odell, 1932) have found that
the undulation of the cells of many species decreases the higher the point of insertion
on the stem, in some cases radical leaves having undulate-walled cells and cauline
leaves straight-walled cells.

With regard to other features of cell shape Brenner (1900) claimed that Crassula
portulacea Lam. has strongly papillose cells in damp conditions, but flat cells in
" normal " situations.

Few other characters of the epidermis show as much variation as the cell-wall
undulation, and although this has been used taxonomically in a number of groups
its use requires the most extreme caution.

E. LEAF VENATION.

Variation in the actual venation of leaves is apparently not very frequent, and
when it does occur not drastic. The characters found to vary, if any, are exemplified
by the situation found in Medicago saliva L. by Turrell (1942), where the larger leaves,
which had larger epidermal cells and stomata than normal, possessed larger areolae,

66 CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY

i.e. the number of veins and venules was not different, but the spaces between them
were larger. Levin (1929) and Hall & Melville (1951) claimed that the areola size
was diagnostically important, but the latter authors (1954) and Gupta (1961) later
found that it in fact varied with leaf size.

The most variable venation character with regard to the present study is the
degree to which the various orders of veins are represented on the epidermis. This
feature appears as the epidermal cells are enlarging, but its exact time of develop-
ment probably varies from group to group. In Buchenavia capitata and B. kleinii,
for instance, it develops earlier on the lower epidermis, where it is quite well developed
well before the cells are full-sized, than on the upper epidermis, where it has not
fully developed when the leaves are fully grown : in the latter case both the smaller
venules and undulations of the cells develop as the full-sized leaf matures. In fact
it is the non-appearance of undulations on the venule cells which causes them to be
distinguished so easily.

No comments concerning the phenotypic variation of the epidermal veins seem
to have been made in the literature, but considerable variation was in fact found in
a number of species. Since the species in which this was noted were all of tropical
origin the reasons could not be definitely explained, butj there is little doubt that
they are partly environmental and positional in origin. The differences were
concerned with the size of the smallest venules distinguishable on the epidermis,
and thus with the sizes of the areolae and the degrees to which the venous system
appeared to be open or closed. Usually the range of variation within a species
gives extremes with different-sized areolae only, but in some cases a species shows
open and closed venous systems of all degrees.

It appears that the environmental conditions producing a leaf with fewer veins
discernible on the epidermis in fact act primarily by producing a thicker leaf due to
the larger cells and inter-cellular spaces. This in some way seems to permit fewer
of the vein extensions to extend from the venules as far as the epidermis, so that
fewer venules produce epidermal cell modifications. Greater shade and humidity
are the commonest conditions which cause these effects.

F. DEVELOPMENT OF CUTICULAR MEMBRANE.

The mature thickness of the cuticular membrane is not attained until a fairly late
stage, when cell enlargement is completed. The membrane is, however, thick
enough to be isolated in toto even before cell division has stopped, and the cuticular
flanges are then well enough developed to show quite clearly the outlines of the cells.
Nevertheless absolutely mature leaves are essential for comparative taxonomic
purposes.

Furthermore environmental conditions are known to affect the cuticular membrane
thickness and cuticular-flange length. Skoss (1955) found that the " cuticle " was
thicker in leaves exposed to the sun, and the same was discovered by a number of
other workers who also correlated a thicker cuticular membrane with a drier soil,
drier air, more exposure to wind, a higher altitude and other factors (fide Odell, 1932).

CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 67

Also the cuticular membrane is usually thicker on the upper stem leaves than on
the lower stem or radical leaves (Yapp, 1912) but Stober (1917) reported that in
some plants it was thickest on the lower stem and upper radical leaves. This
cuticular thickness is accompanied by a thickening of the outer epidermal cellulose
cell wall.

Variation may also be caused by other, unknown, conditions, since Stevens (1932)
noted a fluctuation in the cuticular thickness on cranberry fruits (V actinium
macrocarpon Aiton) from year to year which was not correlated with fruit size or
climate, and which was exactly repeated in 33 different varieties of the plant. In
Lumnitzera racemosa, a mangrove with a normally very thick cuticular membrane,
a single specimen was encountered with very abnormally thin leaves due to the
sparse development of the usually abundant water-storage tissue. This specimen
had a very much thinner cuticular membrane than usual, but the epidermal anatomy
was otherwise quite typical, and the habitat gave no clue to the reason for the
peculiarity.

Conspicuous cuticular striations are absent from most leaves, but when they are
present they do not develop until a relatively late stage, and their usefulness is
restricted to mature leaves. They are, however, of great taxonomic value in some
groups, being highly characteristic of three of the four species of Macropteranthes,
for example. No very young leaves of markedly striate species of the Combretaceae
were available, but Martens (1934) has followed the development of the cuticular
striations on certain petals. There is, however, a considerable variation in the
degree of development of the cuticular striations in some species, and full account
of this must be taken. In Combretum molle R. Br. ex G. Don, for example, the only
known gathering from South West Africa is most distinctive in its extremely
conspicuously striated cuticle, the striations being so strong as to obscure the cell
outlines (Stace, 1961). In tropical Africa, however, where this species is very
abundant, all degrees of striation are found from the previous situation to the
presence of only a few striations placed radially to each trichome-base, a situation
present in many or even most species of the genus. Since the cell outlines in the
latter cases are very clear, the two extremes have an entirely different aspect. The
cause of this variation is, of course, unknown, but would seem likely to be environ-
mental as the climate of the South West African habitat is relatively very dry.

G. SUMMARY.

It is clear that a great deal of variation is to be encountered within one species in
certain cuticular characters, although little or none may be found in others. Varia-
tion may be attributed to minor genetic (genotypic) differences or to certain
measurable influences, both genotypic and phenotypic, of three main types
developmental, positional and environmental. Obviously variation due to all of
these factors must be fully considered before the characters are utilized taxonomically.

In many or most plants the lower leaves are in a different environment from the
upper, and their structure often reflects this. In Filipendula ulmaria, for example,
the lower leaves are in a much more shaded and humid situation, and the type of

68 CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY

variation from the upper leaves which they show is similar to that which is shown
by upper leaves of other plants which are in a wholly damper environment, although
this is not to say that the characteristics of the lower leaves are actually caused by
their environment, as are those of upper leaves of plants in damp, shaded situations.
The characters of these leaves are nevertheless to be regarded as relatively meso-
morphic, and those of leaves in drier conditions as xeromorphic. The commonest
mesomorphic cuticular characters are : a thinner cuticular membrane and shorter
flanges ; more undulate epidermal cell walls ; fewer hairs ; less densely situated
stomata ; larger epidermal cells and stomata ; and larger areolae.

A third level of differentiation of mesomorphic and xeromorphic characters,
beyond various parts of one plant and plants of one taxon, is between different
species of one larger taxon. In this case the characters are genetically regulated,
and they differ to some extent from those mentioned above, mainly in that they are
concerned with structure to a greater extent than size and frequency, but they do
show a number of similarities. The xeromorphic characters as exemplified by the
Combretaceae are : a thicker cuticular membrane and longer cuticular flanges ;
straighter cell walls ; more hairs (in a group in which hairs are of frequent occurrence
the more xeromorphic species are usually the most pubescent, but in other groups
xeromorphs are frequently extremely glabrous, e.g. mangroves) ; lesser development
of veins and venules on the epidermis ; presence of stomata on both leaf surfaces
(in those groups in which there is a tendency for stomata to appear on the upper
epidermis the character is usually most developed in the most xeromorphic species,
but in other groups the stomata of xeromorphs are frequently more limited in
distribution on the lower epidermis alone, e.g. they may be confined to grooves or
crypts) ; more sunken guard-cells ; and, within the anomocytic stomatal group,
the tendency to develop distinctive epidermal cells adjacent to the guard-cells.

It is the task of the taxonomist to distinguish between phenotypic and genotypic
variation. In the former case the problem is to correlate it with specific environ-
mental or other factors and to ascertain the degree of the variation, and in the latter
to decide at what taxonomic levels it should be recognized, if at all.

8. GENERAL CONCLUSIONS

It cannot be too heavily emphasized that there are no guiding principles enabling
assessment of the taxonomic usefulness of a given character in a taxon in which it
has not been previously utilized ; neither are there any means of predicting the
characters most likely to prove of greatest diagnostic importance in an uninvestigated
group. It is thus unwise and unjustifiable to draw on experience in one taxon
and to apply it to another, even if the two plant groups are very closely related.
In the Combretaceae, for instance, the flowers and trichomes provide the most
useful systematic characters in the tribe Combreteae, the fruits being relatively
uniform in structure ; whereas in the closely related Terminalieae the fruits and the
epidermis are the only organs showing a large number of known taxonomic criteria.
The diagnostic value of pollen-grains, chromosomes and wood anatomy in this

CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 69

family has as yet been little studied. Furthermore, in general, characters tend to
vary and assume a greater or lesser taxonomic importance quite independently of
one another, and there is no single character which universally or even usually
overrides all others in importance. Even in the case of the chromosomes, for which
the latter has most frequently been claimed, there are very many taxa in which
characters other than the number and morphology of these organelles are of greatest
significance, indicating that the genetic information in the chromosomes is no more
manifested in their over-all shape than in that of the flowers, fruits, leaves or
other organs.

There is, therefore, no reason to doubt the validity of epidermal features merely
because in some groups they do vary greatly or do not agree with differentiation by
other means. If it is possible for two taxa to differ solely on the shape of the stamens
or leaves then it is possible for two others to be distinguished only on the form of
their hairs or epidermal cells. It is very doubtful, however, whether a difference in a
single feature of one organ (whatever that organ) will differentiate between taxa at
a higher level than " minor genetic variants ", and almost certainly not between
" species " as normally understood. In reaching taxonomic conclusions evidence
from as many sources as possible should be incorporated into the catalogue of
information which will lead to the proposed classification. No facet of the plant,
morphological, physiological, biochemical or genetic, should be omitted. This
becomes abundantly clear when the case of taxa whose systematic position varies
with the character selectively utilized (e.g. Pteleopsis and Calycopteris in the Com-
bretaceae] is considered.

Undoubtedly minor exceptions to these generalizations exist, though by no means
disproving the rule. There does appear, for instance, to be some degree of cor-
relation between the diagnostic value of the epidermal cells and stomata and the
degree of xeromorphy of the taxon. Most groups which have been found to possess
taxonomically valuable epidermal characters are xeromorphic to some extent, e.g.
gymnosperms, Lauraceae, Magnoliaceae, Ericaceae, Epacridaceae, various mangroves,
and others. This is perhaps an expression of the idea that plants which are more
highly specialized tend to be less variable, potentially as well as actually. It seems
obvious that the greater the complexity of an organ the more features there are
to be used by taxonomists, and thus the greater the diagnostic value of that organ.
There is, in addition, the commonly held belief that characters of apparently little
or no differential survival value (e.g. pollen-grain sculpturing, delimitation of
cells of certain trichomes) are those with the greatest systematic and phylogenetic
significance.

Attempts have often been made to generalize on the taxonomic level at which
various diagnostic characters differentiate. Fritsch (1903), for example, concluded
that gland anatomy was more indicative above the generic level, and the presence
or absence of papillate epidermal cells at the specific level. Edwards (1935), how-
ever, pointed out that in different groups stomatal frequency can be a family,
generic or specific criterion, or of no systematic importance at all. This is true of

70 CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY

almost every character as the many exceptions to Fritsch's and others' claims
testify. Thus, apart from noting that in a particular family or other group certain
characters are more important than others at the various taxonomic levels little or
no valid generalization can be made. A good example illustrating this point is
the possession of a very densely or very sparsely pubescent leaf epidermis : in
Ramatuella it serves to distinguish between R. argentea Kunth and R. virens Spruce
ex Eichl. ; in Hibiscus vitifolius L. between subsp. vitifolius and subsp. vulgaris
Brenan & Exell ; in Conocarpus erectus between var. sericeus and var. erectus ; and
in many species of Epilobium between no more than various phenotypic modifications.
In these cases, therefore, the level at which the character differentiates is presumably
dependent upon the other characters with which it is correlated, and with the causes
of the differences. Similarly, the presence of stomata on the upper epidermis is
important at the family level within the mangrove habit, but amongst the non-
mangrove members of the Combretaceae it is of little significance, varying from
species to species or often even infraspecifically. Stomatal size is frequently
directly the result of the level of ploidy, and indicates a second reason for the fact
that characters differentiate at varying taxonomic levels : namely that taxonomists
have often failed to accord similar ranks to groups of taxa showing exactly the same
differences. Thus Rorippa microphylla and R. nasturtium-aquaticum, Galium
palustre L. subsp. palustre and subsp. tetraploideum Clapham, and Ranunculus
ficaria var. ficaria and var. bulbifer Marsden- Jones are three pairs of taxa of which
the first is diploid and the second tetraploid, although the resultant stomatal size
differences apparently show differentiation at three distinct levels, if the taxonomic
ranks assigned above are accepted 1 .

The need to investigate exhaustively all aspects of the genotypic and phenotypic
variation of a " new " character has been previously stressed. In cases where
characters have been used before this preliminary survey has been undertaken the
identifications attempted have often proved faulty, and have cut across decisions
based on other characters. When the potentiality of the character is understood,
however, and accurate identification is possible, the taxonomy of the group may on
occasions be improved in the light of the discoveries. The present cuticular study
has resulted in several such proposals in the classification of the Combretaceae. It is
not to be expected that the new evidence will cause a drastic re-classification to
become desirable, but rather that the new character might so shift the balance of
the evidence from all sources that in some places taxonomic adjustments seem
advisable. Entirely new classifications based on one type of evidence are usually
shorter-lived than the original.

Epidermal characters, then, are likely to provide not only a means of identification
(especially of sterile material) but also as valuable taxonomic and phylogenetic clues
as most other characters, and they exhibit much the same characteristics with
respect to their pattern of variability. Edwards (1935) commented that the

1 Rorippa microphylla was in fact first described as a variety of R. nasturtium-aquaticum, and Ranunculus
ficaria var. bulbifer has been raised to subspecific rank as R. ficaria subsp. bulbifer (Marsden-Jones)
Lawalrec.

CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 71

cuticular differences between closely related species or infraspecific taxa are usually
slight or absent. This is by no means always so. When all the differences between
two closely allied taxa are analysed it is usually found that a few characters are
conspicuously different, many slightly or scarcely so, and others identical. Thus,
although it is most likely in such a case that the cuticular characters will also be
only slightly or not different, the same is true of any other single feature. Occasion-
ally, however, it is just that particular character being studied that shows the
greatest difference between the two taxa. Several examples where this is true of
the cuticular characters have been discovered within the Combretaceae. It seems
that a knowledge of the relative usefulness and an understanding of the level of
differentiation of cuticular characters in the angiosperms as a whole will only be
objectively obtained when cuticular studies have been absorbed into the general
practice of plant taxonomy.

I do not claim that cuticular characters are of any outstandingly fundamental
or all-important significance, as did some of the early exponents of the taxonomic
use of pollen-grain or chromosome features. They are, however, to be regarded as
characters of undoubted importance as further pieces of the jigsaw of complete
systematic evidence, and at times they are in fact of greater value in identification
and taxonomy than are any other characters of which we know.

SPECIAL LITERATURE

AHMAD, K. J. (1964). Cuticular studies in Solanaceae. Canad. Journ. Bot. 42 : 793-803.

(i964a). On stomatal abnormalities in Solanaceae. Science &> Culture 30 : 349-351.

(19640). Epidermal studies in Solanum. Lloydia 27 : 243-250.

(i964c). Cuticular studies with special reference to abnormal stomatal cells in Cestrum.
Journ. Ind. Bot. Soc. 43 : 165-177.

ARESCHOUG, F. W. C. (1902). Untersuchungen iiber den Blattbau der Mangrove-Pflanzen.

Biblioth. Bot. 56.
ARZT, T. (1934). Untersuchungen iiber das Vorkommen einer Kuticula in den Blattern

dikotyler Pflanzen. Ber. Deutsch. Bot. Ges. 51 : 470-500.
ASHBY, E. (1948). Studies in the morphogenesis of leaves. II. The area, cell size and cell

number of leaves of Ipomoea in relation to their position on the shoot. New Phytol. 47 :

I77-I95-

- & WANGERMANN, E. (1950). Studies in the morphogenesis of leaves. IV. Further
observations on area, cell size and cell number of leaves of Ipomoea in relation to their
position on the shoot. New Phytol. 49 : 23-35.

- & (i95oa). Studies in the morphogenesis of leaves. V. A note on the origin of

differences in cell size among leaves at different levels of insertion on the stem. New
Phytol. 49 : 189-192.

AVERY, G. S. (1933). Structure and development of the Tobacco leaf. Amev. Journ. Bot.

20 : 565-592.
BACHMANN, O. (1886). Untersuchungen iiber die systematische Bedeutung der Schildhaare.

Flora 69 : 387-400, 403-415, 428-448.
BAILEY, I. W. (1924). Notes on neotropical ant-plants. III. Cordia nodosa Lam. Bot. Gaz.

77 : 32-49.

72 CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY

BAILEY, I. W., & NAST, C. G. (1944). The comparative morphology of the Winteraceae.
V. Foliar epidermis and sclerenchyma. Journ. Arnold Arb. 25 : 342-348.

- & - - (1948). Morphology and relationships of Illicium, Schisandra and Kadsura. I.
Stem and leaf. Journ. Arnold Arb. 29 : 77-89.

BANDULSKA, H. (1923). A preliminary paper on the cuticular structure of certain dicotyle-
donous and coniferous leaves from the Middle Eocene flora of Bournemouth. Journ. Linn.
Soc. Land., Bot. 46 : 241-269.

- (1924). On the cuticles of some recent and fossil Fagaceae. Journ. Linn. Soc. Lond.,
Bot. 46 : 427-441.

- (1926). On the cuticles of some fossil and recent Lauraceae. Journ. Linn. Soc. Lond.,
Bot. 47 : 383-425-

- (1928). A Cinnamon from the Bournemouth Eocene. Journ. Linn. Soc. Lond., Bot.

48 : I39-I47-

- (1931). On the cuticles of some recent and fossil Myrtaceae. Journ. Linn. Soc. Lond.,
Bot. 48 : 657-671.

BARANOV, P. (1924). De la methode quantitative-anatomique de 1'etude des plantes. I.

Distribution des stomates. Bull. Univ. As. Centr. 7 : 30-35.
BARY, A. DE (1871). Ueber die Wachsuberziige der Epidermis. Bot. Zeit. 29 : 129-139,

145-154, 161-176, 566-571, 573-585, 589-600, 605-619.
(1884). Comparative anatomy of the vegetative organs of the phanerogams and ferns. Transl.

F. O. Bower & D. H. Scott. Oxford.
BATHER, F. A. (1907). Nathorst's use of collodion imprints in the study of fossil plants.

Geol. Mag., Dec. 5, 4 : 437-440.
(1908). Nathorst's methods of studying cutinised portions of fossil plants. Geol. Mag.,

Dec. 5, 5 : 454~459-
BOKORNY, T. (1882). Ueber die " durchsichtigen Punkte " in den Blattern. Flora 65 :

339-350, 355-368, 371-381. 387-397, 4II-4I7-

BORNEMANN, J. G. (1856). Uber organische Reste der Lettenkohlengruppe Thiiringens. Leipzig.
BORRILL, M. (1961). Epidermal characteristics in the diploid subspecies of Dactylis glomerata

L. Journ. Linn. Soc. Lond., Bot. 56 : 453-458.

BRENNER, W. (1900). Untersuchungen an einigen Fettpflanzen. Flora 87 : 387-439.
BRODIE, P. B. (1842). Notice on the occurrence of plants in the Plastic Clay of the Hamp-
shire coast. Proc. Geol. Soc. Lond. 3 : 592.
BUREAU, E. (1864). Monographie des Bignoniacees. Paris.
BURKILL, I. H. (1939). Two notes on Dioscoreas in the Congo : (i) The acarodomatia of

D. minutiflora Engl. and D. smilacifolia De Wild. Proc. Linn. Soc. Lond. 151 : 57-60.
CAROLIN, R. C. (1954). Stomatal size, density and morphology in the genus Dianthus.

Kew Bull. 9 : 251-258.
CLARKE, J. (1960). Preparation of leaf epidermis for topographic study. Stain Technol. 35 :

35-39-
COOKSON, I. C. (1953). On Macrozamia hopeites an Early Tertiary cycad from Australia.

Phytomorphology 3 : 306312.

COWAN, J. M. (1950). The Rhododendron leaf '. Edinburgh & London.
DEHNEL, G. S. (1960). Response of stomata to wounding. Bot. Gaz. 122 : 124-130.

- (1961). Abnormal stomatal development in foliage leaves of Begonia aridicaulis. Amer.
Journ. Bot. 48 : 129-133.

DILCHER, D. L. (1963). Cuticular analysis of Eocene leaves of Ocotea obtusifolia. Amer.

Journ. Bot. 50 : 1-8.
EDWARDS, W. N. (1935). The systematic value of cuticular characters in recent and fossiJ

angiosperms. Biol.Rev. & Biol. Proc. Cambr. Phil. Soc. 10 : 442-459.
ELLIOT, G. F. S. (1911). Botany of to-day. Ed. 2. London.
ENGLER, A., & PRANTL, K. (1887-1915). Die natiirlichen Pflanzenfamilien. Leipzig.

CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 73

ESAU, K. (1953)- Plant anatomy. New York & London.

EXELL, A. W. (1961). New species and subspecies of Combretum from southern tropical
Africa. Mitteil. Bot. Staatssamml. Munch. 4 : 3-7.

- (1963). Plants of the 1957-58 West African Expedition : IV. A new species of Combretum
from Northern Nigeria. Kew Bull. 17 : 233-236.

- & STACK, C. A. (1963). A revision of the genera Buchenavia and Ramatuella. Bull.
Brit. Mus. (Nat. Hist.}, Bot. 3 : 1-46.

FLORIN, R. (1920). Uber Cuticularstrukturen der Blatter bei einigen rezenten und fossilen
Coniferen. Ark. for Bot. 16 (6).

- (1931). Untersuchengen zur Stammesgeschichte der Coniferales und Cordaitales. Erster
Teil : Morphologic und Epidermisstruktur der Assimilationsorgane bei den rezenten
Koniferen. K. Svensk. Vetenskapsakad. Handl., Ser. 3, 10 (i).

- ( I 933)- Studien liber die Cycadales des Mesozoikums nebst Erorterungen iiber die
Spaltoffnungsapparate der Bennettitales. K. Svensk. Vetenskapsakad. Handl., Ser. 3,
12 (5).

- (1938-45). Die Koniferen des Oberkarbons und des Unteren Perms. Palaeontographica

85 (B).

FOSTER, A. S. (1949). Practical plant anatomy. Ed. 2. Toronto, New York & London.
FRITSCH, F. E. (1903). The use of anatomical characters for systematic purposes. New

Phytol. 2 : 177-184.
GOEPPERT, H. R., & BERENDT, G. C. (1845). Der Bernstein und die in ihm befmdlichen

Pflanzenreste der Vorwelt. Berendt, Bernst. Org. Reste Vorwelt i (i).
GOTHAN, W. (1915). t)ber die Methoden und neue Erfolge bei der Untersuchung kohlig

erhaltener Pflanzenreste. Sitzungsber. Ges. Naturforsch. Freunde Berl. 1915 : 43-48.
GROB, A. (1896). Beitrage zur Anatomie der Epidermis der Gramineenblatter. Biblioth.

Bot. 36.
GUPTA, B. (1961). Correlation of tissues in leaves, i. Absolute vein-islet numbers and

absolute veinlet termination numbers. Ann. of Bot., New Ser.. 25 : 65-70.
HALL, J. P., & MELVILLE, C. (1951). Veinlet termination number. A new character for the

differentiation of leaves. Journ. Pharm. &> Pharmacol. 3 : 934-941.

- & - - (1954). Veinlet termination number some further observations. Journ.
Pharm. &> Pharmacol. 6 : 129-133.

HARRIS, T. M. (1926). Note on a new method for the investigation of fossil plants. New
Phytol. 25 : 58-60.

- (1942-*-). Notes on the Jurassic flora of Yorkshire. i->. Ann. &> Mag. Nat. Hist.,
Ser. n, 9 : 568-*-.

- (1944). Notes on the Jurassic flora of Yorkshire. 15. Ann. & Mag. Nat. Hist., Ser.
ii, ii : 685-690.

HEINEN, W. (1960). tJber den enzymatischen Cutin-Abbau. I. Mitteilung : Nachweis
eines " Cutinase "-Systems. Act. Bot. Neerl. 9 : 167-190.

- (1961). Uber den enzymatischen Cutin-Abbau. II. Mitteilung : Eigenschaften eines
cutinolytischen Enzyms aus Penicillium spinulosum Thorn. Act. Bot. Neerl. 10 : 171-189.

HEINTZELMAN, C. E., & HOWARD, R. A. (1948). The comparative morphology of the Icacina-

ceae. V. The pubescence and the crystals. Amer. Journ. Bot. 35 : 42-52.
HOFMANN, E. (1932). Epidermisreste und Blattabdriicke aus den Braunkohlenlagern des

Geiseltales. Nov. Act. Leopold., Neue Folge, i : 59-74.
HOHNEL, F. VON (1878). Einige Bemerkungen iiber die Cuticula. Oesterr. Bot. Zeitschr.

28 : 81-87, Ir 5- 121 -

HOLDEN, R. (1915). On the cuticles of some Indian conifers. Bot. Gaz. 60 : 215-227.
HOLTERMANN, C. (1893). Beitrage zur Anatomie der Combretaceen. Forhandl. Vidensk.-Selsk.

Christiania 1893 (12).
HUTCHINSON, J. (1926). The families of flowering plants. I. Dicotyledons. London.

- & DALZIEL, J. M. (1931). Flora of west tropical Africa. 2 (i). London.

74 CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY

JALAN, S. (1962). The ontogeny of stomata in Schisandra grandiflora Hook. f. & Thorns.

Phytomorphology 12 : 239-242.

JEFFREY, E. C. (1917). The anatomy of woody plants. Chicago.
JOHNSON, T., & GILMORE, J. G. (1921). The occurrence of Dewalquea in the coal-bore at

Washing Bay. Sci. Proc. R. Dublin Soc., New Ser., 16 : 323-333.
JUNIPER, B. E. (1959). The surfaces of plants. Endeavour 18 : 20-25.

- ( I 959 a )- The effect of pre-emergent treatment of Peas with trichloracetic acid on the
submicroscopic structure of the leaf surface. New Phytol. 58 : 1-4.

- (1960). Growth, development and effect of the environment on the ultra-structure of
plant surfaces. Journ. Linn. Soc. Lond., Bot. 56 : 413-419.

& BRADLEY, D. E. (1958). The carbon replica technique in study of the ultrastructure of

leaf surfaces. Journ. Ultrastruct. Res. 2 : 16-27.

KEAY, R. W. J. (1953). Rhizophora in West Africa. Kew Bull. 8 : 121-127.
KOSTERMANS, A. J. G. H. (i957)- Lauraceae. Reinwardtia 4 : 193-256.

KUBART, B. (1927). Einige Bemerkungen zur Frage des Nachweises der Lauraceen-Gatiung
Umbellularia in den Tertiaren " Redwood "-Bestanden Nordamerikas. Biol. Gen. 3 :

585-594-
LAMARLIERE, L. G. DE (1906). Sur les membranes cutinisees des plantes aquatiques. Rev.

Gin. Bot. 1 8 : 289-295.
LEE, B., & PRIESTLEY, J. H. (1924). The plant cuticle. I. Its structure, distribution, and

function. Ann. of Bot. 38 : 525-545.
LEVIN, F. A. (1929). The taxonomic value of vein islet areas. Quart. Journ. Pharm. (&

Pharmacol. 2 : 17-43.

LOFTFIELD, J. V. G. (1921). The behaviour of stomata. Carneg. Inst. Wash. Publ. 314.
LONG, F. L., & CLEMENTS, F. E. (1934). The method of collodion films for stomata. Amer.

Journ. Bot. 21 : 7-17.
LUNDSTROM, A. N. (1887). Pflanzenbiologische Studien. II. Die Anpassungen der Pflanzen

an Thiere. Nov. Act. R. Soc. Sci. Upsal., Ser. 3, 13 (10).
MCDOUGALL, W. B. (1927). Plant ecology. London.
MAHESHWARI, P., & VASIL, V. (1961). The stomata of Gnetum. Ann. of Bot., New Ser., 25 :

313-319-

MANI, M. S. (1964). Ecology of plant galls. Monogr. Biol. 12.

MARLIER-SPIRLET, M. L. (1945). Sur quelques epidermes de feuilles de Cinnamomum L.

Bull. Jard. Bot. Et. Brux. 17 : 266-305.
MARTENS, P. (1934). Recherches sur la cuticle. III. Structure, origine et signification du

relief cuticulaire. Protoplasma 20 : 483-515.
MARTIN, D. J. (1955). Features on plant cuticle. An aid to the analysis of the natural diet

of grazing animals, with especial reference to Scottish hill sheep. Trans. & Proc. Bot.

Soc. Edin. 36 : 278-288.
METCALFE, C. R. (1954). An anatomist's view on angiosperm classification. Kew Bull.

9 : 427-440.

- ( J 955)- Recent work on the systematic anatomy of the monocotyledons. Kew Bull.
9 : 523-532.

- (1960). Anatomy of the monocotvledons. I.Gramineae. Oxford.

- (1963). Comparative anatomy as a modern botanical discipline. Advanc. Bot. Res.
i : 101-147.

- & CHALK, L. (1950). Anatomy of the dicotyledons. Oxford.

MEYER, F. J. (1932). Beitrage zur Anatomic der Alismataceen. Beik. Bot. Centralbl. 49 (i)
309-368.

- (i932a). Die Verwandtschaftsbeziehungen der Alismataceen zu den Ranales im Lichte
der Anatomie. Engler, Bot. Jahrb. 65 : 53-59.

MEYER, M. (1938). Die submikroskopische Struktur der kutinisierten Zellmembranen .
Protoplasma 29 : 552-586.

CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY 75

MINDEN, M. VON (1899). Beitrage zur anatomischen und physiologischen Kenntnis wasser-

secernierender Organe. Biblioth. Bot. 46.
MORLEY, T. (1953). The genus Mouriri (Melastomaceae) . A sectional revision based on

anatomy and morphology. Univ. Calif. Publ. Bot. 26 : 223-299.
MUELLER, L. E. ( CARR, P. H., & LOOMIS, W. E. (1954). The submicroscopic structure of

plant surfaces. Amer. Journ. Bot. 41 : 593-600.
NATHORST, A. G. (1907-12). Palaobotanische Mitteilungen. i-n. K. Svensk. Vetenskap-

sakad. Handl., Ny Foljd, 42 (5)~48 (2).

- (igoya). t)ber die Anwendung von Kollodiumabdrucken bei der Untersuchung fossiler
Pflanzen. Ark. for Bot. 7 (4).

NETOLITZKY, F. (1932). Die Pflanzenhaare. Linsbauer, Handb. Pflanzenanat. 4 (29).
NORTH, C. (1956). A technique for measuring structural features of plant epidermis using

cellulose acetate films. Nature 178 : 1186-1187.
ODELL, M. E. (1932). The determination of fossil angiosperms by the characteristics of their

vegetative organs. Ann. of Bot. 46 : 941-963.
ORGELL, W. H. (1955)- The isolation of plant cuticle with pectic enzymes. PI. Physiol. 30 :

78-80.
PALIWAL, G. S., & BHANDARI, N. N. (1962). Stomatal development in some Magnoliaceae.

Phytomorphology 12 : 409-412.
PANT, D. D. (1965). On the ontogeny of stomata and other homologous structures. PI. Sci.

Ser. Allahabad i : 1-24.

- & MEHRA, B. (1963). Development of caryophyllaceous stomata in Asteracantha longifolia
Nees. Ann. of Bot., New Ser., 27 : 647-652.

- & VERMA, B. K. (1963). Development of stomata in leaves of Notonia grandiflora DC.
Journ. Ind. Bot. Soc. 42 : 384-391.

PATERSON, B. R. (1961). Systematic studies of the anatomy of the genus Acrotriche R. Br.

I. The leaf. Austral. Journ. Bot. 9 : 197-208.
PENZIG, O., & CHIABRERA, C. (1903). Contribute alia conoscenza delle piante acarofile.

Malpighia 17 : 429-487.
PRANTL, K. (1872). Die Ergebnisse der neueren Untersuchungen iiber die Spaltoffnungen.

Flora 55 : 305-312, 321-328, 337~34 6 - 3 6 9-3 82 ,-
PRAT, H. (1932). L'epiderme des Graminees. Etude anatomique et systematique. Ann.

Sci. Nat., Ser. 10, Bot. 14 : 117-325.
PRAY, T. R. (1963). Origin of vein endings in angiosperm leaves. Phytomorphology 13 :

60-8 i.

PRIESTLEY, J. H. (1943). The cuticle in angiosperms. Bot. Rev. 9 : 593-616.
PRILLIEUX, E. (1856). De la structure des poils des Oleacees et des Jasminees. Ann. Sci.

Nat., Ser. 4, Bot. 5 : 5-14.
RAO, H. S. (1939). Cuticular studies of Magnoliales. Proc. Ind. Acad. Sci., Sect. B, 9 :

99-116.
REA, M. W. (1921). Stomata and hydathodes in Campanula rotundifolia L., and their relation

to environment. New Phytol. 20 : 56-72.
REHFOUS, L. (1914). Les stomates des Celastracees. Bull. Soc. Bot. Geneve, Ser. 2, 6 : 13-18.

- (1917). Etude sur les stomates. These No. 605, Universite de Geneve.

REYENGA, T. W., & KARSTENS, W. K. H. (1964). Changes in the epidermal pattern of the
sepals of Hydrangea macrophylla (Thunb.) D.C., " Otaksa " in their successive growth
phases. Formation of new stomata, disintegration of old stomata. Act. Bot. Neerl. 13 :
340-351.

ROELOFSEN, P. A. (1952). On the submicroscopic structure of cuticular cell walls. Act.
Bot. Neerl. i : 99-114.

- (1959). The plant cell-wall. Linsbauer, Handb. Pflanzenanat., ed. 2, 3 (4).
SALISBURY, E. J. (1927). On the causes and ecological significance of stomatal frequency,

with special reference to the woodland flora. Phil. Trans. R. Soc. Lond., Ser. B, 216 : 1-65.

76 CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY

SAX, H. J. (1938). The relation between stomata counts and chromosome number. Journ.

Arnold Arb. 19 : 437-441.
SAX, K., & SAX, H. J. (1937). Stomata size and distribution in diploid and polyploid plants.

Journ. Arnold Arb. 18 : 164-172.
SCHENK, A. (1869-71). Beitrage zur Flora der Vorwelt. Ill, IV. Palaeontographica 19 :

1 34, 203-276.

SCHIEFERSTEIN, R. H., & LooMis, W. E. (1956). Wax deposits on leaf surfaces. PI. Physiol.
31 : 240-247.

& - (1959). Development of the cuticular layers in angiosperm leaves. Amer. Journ.
Bot. 46 : 625-635.

SCHLEIDEN, M. J. (1846). Ueber die fossilen Pflanzenreste des Jenaischen Muschelkalks.

Schmid & Schleid., Geognost. Verhdltn. Saalth. Jena 2.
SCHULZE, F. (1855). Bemerkungen iiber das Vorkommen wohlerhaltener Cellulose in Braun-

kohle und Steinkohle. Ber. Verhandl. K. Preuss. Akad. Wiss. Berl. 1855 : 676-678.
SCOTT, F. M., HAMNER, K. C., BAKER, E., & BOWLER, E. (1958). Electron miscroscope

studies of the epidermis of Allium cepa. Amer. Journ. Bot. 45 : 449-461.
SEITHE, A. (1960). Die Haarformen der Gattung Rhododendron L. und die Moglichkeit ihrer

taxonomischen Verwertung. Engler, Bot. Jahrb. 79 : 297-393.

- (1962). Die Haararten der Gattung Solanum L. und ihre taxonomische Verwertung.
Engler, Bot. Jahrb. 81 : 261-335.

SIFTON, H. B. (1963). On the hairs and cuticle of Labrador Tea leaves. A developmental

study. Canad. Journ. Bot. 41 : 199-207.
SINCLAIR, C. B., & DUNN, D. B. (1961). Surface printing of plant leaves for phylogenetic

studies. Stain Technol. 36 : 299-304.
SKOSS, J. D. (1955). Structure and composition of plant cuticle in relation to environmental

factors and permeability. Bot. Gaz. 117 : 55-72.
SLADE, B. F. (1957). Leaf development in relation to venation, as shown in Cercis siliquastrum

L., Prunus serrulata Lindl. and Acer pseudoplatanus L. New Phytol. 56 : 281-300.
SOLEREDER, H. (1908). Systematic anatomy of the dicotyledons. Transl. L. A. Boodle & F. E.
Fritsch ; revis. D. H. Scott. Oxford.

& MEYER, F. J. (1928-33). Systematische Anatomie der Monokotyledonen. (Incomplete.)
Berlin.

SRIVASTAVA, P. N. (1956). Studies in the Glossopteris flora of India 4. Glossopteris,
Gangomopteris and Palaeovittaria from the Raniganj Coalfield. The Palaeobotanist 5 :

i-45-

STAGE, C. A. (1961). Cuticular characters as an aid to the taxonomy of the South-West
African species of Combretum. Mitteil. Bot. Staatssamml . Munch. 4 : 9-19.

- (1963). Cuticular patterns as an aid to plant taxonomy. Ph. D. thesis (London). Ined.

- (1965). The significance of the leaf epidermis in the taxonomy of the Combretaceae.
I. A general review of tribal, generic and specific characters. Journ. Linn. Soc. Lond.,
Bot. 59 : 229-252.

STEBBINS, G. L., & JAIN, S. K. (1960). Developmental studies of cell differentiation in the
epidermis of monocotyledons. I. Allium, Rhoeo, and Commelina. Developmental Biol.

2 : 409-426.

- & KHUSH, G. S. (1961). Variation in the organization of the stomatal complex in the
leaf epidermis of monocotyledons and its bearing on their phylogeny. Amer. Journ. Bot.

48 : 51-59-

- & SHAH, S. S. (1960). Developmental studies of cell differentiation in the epidermis of
monocotyledons. II. Cytological features of stomatal development in the Gramineae.
Developmental Biol. 2 : 477-500.

STEENIS, C. G. G. J. VAN (1953). Results of the Archbold Expeditions. Papuan Nothofagus.
Journ. Arnold Arb. 34 : 301-374.

CUT1CULAR STUDIES AS AN AID TO PLANT TAXONOMY 77

STEVENS, N. E. (1932). Thickness of cuticle in Cranberry fruits. Amer. Journl Bot. ig :

432-435-
STOBER, J. P. (1917). A comparative study of winter and summer leaves of various herbs.

Bot. Gaz. 63 : 89-109.
STOCKMANS, F. (1932). Sur des epidemics de dicotyledonees (Dewalquea gelindenensis Saporta

et Marion et Litsea elatinervis Saporta et Marion) des Marnes de Gelinden (Paleocene).

Bull. Mus. R. Hist. Nat. Belg. 8 (9).
STRAUS, A. (1930). Dikotyle Pflanzenreste aus dem Oberpliozan von Willershausen (Kreis

Osterode, Harz) i. Jahrb. Preuss. Geol. Landes. Berl. 51 (i) : 302-336.
SURANGE, K. R., & SRIVASTAVA, P. N. (1956). Studies in the Glossopteris flora of India -5.

Generic status of Glossopteris, Gangomopteris and Palaeovittaria. The Palaeobotanist 5 :

46-49.
TATEOKA, T., INOUE, S., & KAWANO, S. (1959). Notes on some grasses. IX. Systematic

significance of bicellular microhairs of leaf epidermis. Bot. Gaz. 121 : 80-91.
THOMAS, H. H. (1912). On some methods in palaeobotany. New Phytol. n : 109-114.
(1930). Further observations on the cuticle structure of Mesozoic cycadean fronds.

Journ. Linn. Soc. Land., Bot. 48 : 389-415.

- & BANCROFT, N. (1913). On the cuticles of some recent and fossil cycadean fronds.
Trans. Linn. Soc. Lond., Ser. 2, Bot. 8 : 155-204.

THOMPSON, W. P. (1912). Structure of the stomata of certain Cretaceous conifers. Bot. Gaz.

54 : 63-67.
TIMMERMAN, H. A. (1927). Stomatal numbers : their value for distinguishing species. Pharm.

Journ. & Pharmacist 118 : 241-243.
TOMLINSON, P. B. (1959). Anatomical approach to the classification of the Musaceae. Journ.

Linn. Soc. Lond., Bot. 55 : 779-809.

- (1961). Anatomy of the monocotyledons. II. Palmae. Oxford.

TURRELL, F. M. (1942). A quantitative morphological analysis of large and small leaves of

Alfalfa, with special reference to internal surface. Amer. Journ. Bot. 29 : 400-415.
UNGER, F. (1853). Iconographia plantarum fossilium. Abbildungen und Beschreibungen

fossiler Pflanzen. Denkschr. Kaiserl. Akad. Wiss. Wien, Math.-nat. Cl. 4 (i) : 73-118.
UPHOF, J. C. T. (1942). Ecological relations of plants with ants and termites. Bot. Rev. 8 :

563-598.
VESQUE, J. (1889). De 1'emploi des caracteres anatomiques dans la classification des vegetaux.

Bull. Soc. Bot. France 36 : XLI-LXXVII.
WATSON, L. (1962). The taxonomic significance of stomatal distribution and morphology in

Epacridaceae. New Phytol. 61 : 36-40.

- (1965). The taxonomic significance of certain anatomical variations among Ericaceae.
Journ. Linn. Soc. Lond., Bot. 59 : 111-125.

WATSON, R. W. (1942). The effect of cuticular hardening on the form of epidermal cells.

New Phytol. 41 : 223-229.
WEISS, A. (1865). Untersuchungen iiber die Zahlen- und Grossenverhaltnisse der Spaltoft'-

nungen. Jahrb. Wiss. Bot. 4 : 125-196.
WESSEL, P., & WEBER, O. (1856). Neuer Beitrag zur Tertiarflora der Niederrheinischen

Braunkohlenformation. Palaeontographica 4 : 111-178.
WILDEMAN, E. DE (1904). Sur 1'acarophytisme chez les monocotyledones. Compt. Rend.

Acad. Sci. Par. 139 : 551-553.

- (1923). Notes sur le Strephonema gilleti De Wild, plante oleagineuse du Congo. Bull.
Jard. Bot. Et. Brux. 8 : 119-124.

WILLS, L. (1914). Plant cuticles from the Coal-measures of Britain. Geol. Mag., Dec. 6,
i : 385-390.

78 CUTICULAR STUDIES AS AN AID TO PLANT TAXONOMY

WYLIE, R. B. (1943). The role of the epidermis in foliar organization and its relations to the

minor venation. Amer. Journ. Bot. 30 : 273-280.
YAPP, R. H. (1912). Spiraea ulmaria, L., and its bearing on the problem of xeromorphy in

marsh plants. Ann. of Bot. 26 : 815-870.
ZEILLER, R. (1882). Observations sur quelques cuticules fossiles. Ann. Sci. Nat., Ser. 6,

Bot. 13 : 217-238.

\

Bull. B.M. (N.H.) Bot. 4, i

PLATE i

A, Rhizophora mucronata Lam. : lower epidermis of leaf, showing cork-wart (X42o).
B, Buchenavia kleinii Exell : upper epidermis of leaf, showing wound caused by fungal
attack (X42o). C, Terminalia catappa L. : leaf margin, showing characteristic hole
( X42o). D, Lumnitzera racemosa Willd. : lower epidermis of leaf, showing water-stoma

(X 4 20).

Bull. B.M. (N.H.) Bot. 4, i

PLATE 2

D

A, Buchenavia grandis Ducke : lower epidermis of leaf, showing marsupiform domatium
(x8o). B, B. capitata (Vahl) Eichl. : lower epidermis of leaf, showing marsupiform
domatium ( x 60). C, Ternrinalia catappa L. : lower epidermis of leaf, showing lebetiform
domatium ( x8o). D, Conocarpus erectus var. sericeits Forsstr. ex DC. : lower epidermis
of leaf, showing lebetiform domatium ( x6o).

Bull. EM. (N.H.) Bot. 4, i

, Rhizophora mangle L. : lower epidermis of leaf, showing strongly cuticularized stomata
and straight epidermal cell walls ( X42o). B, Buchenavia fans/iawei Exell & Maguire :
lower epidermis of leaf, showing thinly cuticularized stomata and undulate epidermal
cell walls ( x 420). C, Combretum nigrescens King : leaf margin ( x 170). D, Laguncularia
racemosa (L.) Gaertn. f. : leaf margin ( xiyo).

PLATE 4

Bull. B.M. (N.H.} Bot. 4, i

A, Ramatuella virens Spruce ex Eichl. : lower epidermis of leaf, showing anomocytic stomata
(X42o). B, Laguncularia racemosa (L.) Gaertn. f. : lower epidermis of leaf, showing
cyclocytic stomata ( X42o). C, Strephonema sericeum Hook. f. : lower epidermis of leaf,
showing paracytic stomata ( X42o). D, Guieva senegalensis ]. F. Gmel. : upper epidermis
of leaf, showing numerous simple hair-bases ( X42o).

\

Bull. B.M. (N.H.) Bot. 4, i

PLATE 5

D

A, Thiloa glaucocarpa (Mart.) Eichl. : lower epidermis of leaf, showing scale with strongly
cuticularized scale-base ( x^2o). B, Combretum obovatum F. Hoffm. : lower epidermis of
leaf, showing stalked glands on the raised venule reticulum ( x 1 70) . C, Ceriops tagal
(Perrott.) C. B. Robinson : lower epidermis of leaf, showing stoma and cuticular flanges
in profile on cuticular membrane folded inside outermost ( X42o). D, Ouisqualis hensii
(Engler & Diels) Exell : upper epidermis of leaf, showing cuticular striations and three
types of trichome-base ( X42o).

\

PRINTED IN GREAT BRITAIN
BY THOMAS DE LA RUE &
COMPANY LIMITED LONDON

THE GENUS ELAPHOGLOSS UM

IN THE INDIAN PENINSULA

AND CEYLON

W. A. SLEDGE

BULLETIN OF
THE BRITISH MUSEUM (NATURAL HISTORY)

BOTANY Vol. 4 No. 2

LONDON: 1967

1 MAY 1967

THE GENUS ELAPHOGLOSSUM IN THE
INDIAN PENINSULA AND CEYLON

BY

W. A. SLEDGE

(University of Leeds)

%

Pp. 79-96

BULLETIN OF

THE BRITISH MUSEUM (NATURAL HISTORY)

BOTANY Vol. 4 No. 2

LONDON : 1967

THE BULLETIN OF THE BRITISH MUSEUM

(NATURAL HISTORY), instituted in 1949, is
issued in five series corresponding to the Departments
of the Museum, and an Historical series.

Parts will appear at irregular intervals as they become
ready. Volumes will contain about three or four
hundred pages, and will not necessarily be completed
within one calendar year.

In 1965 a separate supplementary series of longer
papers was instituted, numbered serially for each
Department.

This paper is Vol. 4, No. 2 of the Botany
series.

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

Trustees of the British Museum (Natural History) 1967

TRUSTEES OF
THE BRITISH MUSEUM (NATURAL HISTORY)

Issued 2 May, 1967 Price Seven Shillings

THE GENUS ELAPHOGLOSSUM IN THE
INDIAN PENINSULA AND CEYLON

By W. A. SLEDGE

IN a comment on the systematic list of ferns in a paper by Manton and myself on the
cytology and taxonomy of the pteridophytes of Ceylon, I expressed the view
(Phil. Trans. Roy. Soc., ser. B. 238 : 158 (1954)), that the names currently employed
for the four species of Elaphoglossum were in need of revision. It subsequently
became apparent that both the taxonomy and the nomenclature of the Indian
species also required re-examination.

The names employed in Beddome's works are mostly those originally given to
African or tropical American plants with which Indian ones were at that time
considered identical. The same names are now used in a more restricted sense and
the discrimination of several species with more limited distributions within the areas
of the aggregate species has left in doubt the correct names to be assigned to Indian
and Ceylon species. So far as I am aware the only paper on Indian species of
Elaphoglossum which has appeared since the time of Beddome is one by Biswas
(Bull. Misc. Inf. Kew 1939 : 237-241 (1939)) in which two new species are described.
Both of these supposed new species however are unnecessary renamings of species
already described by Blume and Fee.

In his Ferns of Southern India, Beddome described under the name Elaphoglossum
conforme a species which appears to be not uncommon in the Nilgiri Hills. Biswas,
who described it as new, was correct in rejecting this identification but appears
not to have questioned the identification of another species growing in the
Nilgiri and adjoining hills which Beddome and later authors called E. stigmatolepis
(Fee) Moore ; indeed he makes no reference anywhere in his paper to this species.
Beddome's E. conforme was in fact based on specimens of the same species as that
first described by Fee as Acrostichum stigmatolepis, and it is therefore Beddome's
E. stigmatolepis which requires a new name. In his Handbook to the Ferns of British
India, Ceylon and the Malay Peninsula (1883), Beddome amplified the distribution
credited to E. conforme to include the Himalayas, Ceylon and Malaya. The plant
from the Himalayas is E. marginatum, which had long been regarded as a synonym of
E. conforme but which Biswas correctly reinstated as a distinct species. Neither
E. stigmatolepis nor E. marginatum occur either in Ceylon or Malaya and these regions
were included through errors of identification and do not cover a third species un-
accounted for amongst those described by Beddome.

The Indian specimens which Beddome originally referred to E. laurifolium (changed
in his Handbook to E. latifolium) are identical with Blume's Javan Acrostichum
angulatum, which name he cited as a synonym of E. laurifolium in his Ferns of
Southern India. This is the other species which Biswas needlessly described as a
new species. E. angulatum is an easily recognized species on account of its long-
creeping rhizome and very distinctive rhizome scales, yet most Ceylon gatherings of

BOX. 4, 2. 6

82 ELAPHOGLOSSUM IN THE INDIAN PENINSULA AND CEYLON

this species are identified in herbaria not as E. laurifolium but as E. conforme.
The true Acrostichum laurifolium of Thouars is confined to Tristan da Cunha and
Gough Island, and another species from Ceylon and southern India with short-
creeping rhizome, closely placed fronds and different rhizome scales, which is usually
labelled E. laurifolium in herbaria, is E. commutatum, a species which was founded by
Mettenius partly on a Ceylon gathering made by Thwaites but which is not illustrated
in Beddome's works.

As regards the ferns that Beddome called E. squamosum, this name was currently
used to cover superficially similar plants ranging from tropical America eastwards
to southern India and Ceylon. African, Madagascan and Mascarene representatives
of the aggregate species have since been split off from Jamaican plants, whence the
type was described. The Nilgiri and Ceylon plants which were united under the
same name by Beddome are certainly distinct from one another and as neither
can be matched by African or Mascarene gatherings, I have described these as two
new species, using names already given to herbarium specimens at Kew by Krajina
but never published by him.

Of the names used by Beddome, only E. spatulatum stands unchanged and even
this very distinct species has recently been incorrectly excluded as a Ceylon species
by Madame Tardieu-Blot (Not. Syst. 15 : 432 (1959)), and the equally distinct Mada-
gascan E. schizolepis has erroneously been substituted for it.

The confusion which has surrounded the application of so many of these names is
discussed in more detail later in this paper under the individual species ; but it may
be useful to list here the illustrations and names given in Beddome's Ferns of Southern
India together with their correct identifications :

F.S.I, t. 196 as E. viscosum = E. stelligerum (Wall, ex Baker)

Alston & Bonner

F.S.I, t. 197 as E. squamosum = E. nilgiricum Krajina ex Sledge
F.S.I, t. 198 as E. conforme = E. stigmatolepis (Fee) T. Moore
F.S.I, t. 199 as E. stigmatolepis = E. beddomei Sledge
F.S.I, t. 200 as E. laurifolium = E. angulatum (Blume) T. Moore
F.S.I, t. 213 as E. spathulatum = E. spatulatum (Bory) T. Moore

E. marginatum (Wall, ex Fee)

T. Moore

No illustration in F.S.I.

E. commutatum (Mett.) Alderw.

van Rosenb.
E. ceylanicum Krajina ex Sledge

At the time of writing, a proposal (Anderson in Regn. Veg. 40 : 18 (1965)) to con-
serve the generic name Elaphoglossum Schott ex J. Smith (Hooker's Journ. Bot.
4 : 148 (Aug. 1841)) against A coniopteris C. Presl (Tent. Pterid. : 236 (1836)) is under
consideration. It has been approved by the Special Committee for Pteridophyta
(Taxon 15 : 333 (1966)) but awaits ratification.

My grateful thanks are tendered to the Directors and Curators of the herbaria at
the British Museum (Natural History) (BM), the Botany School, University of
Cambridge (CGE), the Royal Botanic Garden, Edinburgh (E), the Royal Botanic

ELAPHOGLOSSUM IN THE INDIAN PENINSULA AND CEYLON 83

Gardens, Kew (K) and the Division of Systematic Botany, Peradeniya, Ceylon
(PDA) for access to, or for the loan of, specimens in their charge.

KEY TO THE SPECIES

Fronds coriaceous or at least stiff in texture, margins with a cartilaginous border,

surfaces glabrous or with minute scales.
Rhizome long-creeping, fronds spaced.

Rhizome scales papery, broad, pale brown ; lamina 2'5~4'5 cm. broad

i. E. angulatum
Rhizome scales not papery, dark brown or black ; lamina 1-5-2 -5 cm. broad

2. E. stigmatolepis
Rhizome short-creeping, fronds crowded.

Rhizome scales entire, linear (5-10 times as long as broad), stipes much

shorter than lamina . . . . . 3. E. beddomei

Rhizome scales more or less fimbriate, narrow (4-5 times as long as broad),

stipes up to half as long as lamina.

Rhizome scales dark brown, 5 mm. long, acuminate 4. E. marginatum

Rhizome scales brown, 10 mm. long, attenuate with crisped tips

5. E. commutatum
Fronds thickly herbaceous or soft in texture, margins without a cartilaginous border,

surfaces conspicuously scaly or hairy.

Scales of stipe and lamina stellate with long rays . . 6. E. stelligerum
Scales not stellate.

Scales ovate with long marginal teeth.

Fronds densely paleate, the lower surface wholly obscured by imbricating
scales ; scales of stipe broad, mostly concolorous 7. E. nilgiricum
Fronds paleate but neither surface obscured by scales ; scales of stipe
narrow, mostly with their borders and teeth dark-coloured

8. E. ceylanicum
Scales filiform, hair-like . . . . . . 9. E. spatulatum

i. Elaphoglossum angulatum (Blume) T. Moore, Index Fil. : 5 (1857).

Acrostichum angulatum Blume, Enum. PL Jav. : 101 (1828) ; Fl. Jav., Fil. : 25, t. 6 (1829).
Fee, Mem. Fam. Foug. 2 : 32 (1845). Kunze in Linnaea 24 : 248 (1851). Christ in
Neue Denkschr. Schweiz. Ges. Naturw. 36 (i, 2) : 46, 49 (1899).

Olfersia angulata (Blume) C. Presl, Tent. Pterid. : 234 (1836).

Acrostichum conforme sensu Hook., Sp. Fil. 5 : 198 (1864) pro parte ; non Swartz. Hook. &
Baker, Synops. Fil. : 401 (1868) pro parte.

Acrostichum marginatum sensu Thw., Enum. PL Zeyl. : 380 (1864) ; non Wall, ex Fee.

Elaphoglossum laurifolium sensu Bedd., Ferns S. Ind. : 67, t. 200 (1864) ; non T. Moore.

Elaphoglossum latifolium sensu Bedd., Handb. Ferns Brit. Ind. : 416 (1883) ; non J. Smith.

Elaphoglossum ogatai C. Chr., in Dansk. Bot. Ark. 9 (3) : 67 (1937).

Elaphoglossum krajinae Biswas in Bull. Misc. Inf. Kew 1939 : 240, t. i fig. 2 (1939).

Rhizome long-creeping with stipes 1-3 cm. apart, clothed with pale, rufous-brown,
papery, thin-walled, broadly ovate, acute scales, 5x2 mm., with subentire margins.
Fronds stalked 15-35 (45) cm., stipe of sterile fronds up to 15 cm. or more, often as

84 ELAPHOGLOSSUM IN THE INDIAN PENINSULA AND CEYLON

long as the lamina but sometimes only half or less their length, scaly, the scales like
those of the rhizome but with irregularly fringed margins ; lamina elliptic, 2-5-5 cm -
broad, apex acute, base narrowed and shortly decurrent on the stipe, margins with a
cartilaginous, revolute border, lower surface with scattered, minute, brown, more or
less stellately-laciniated scales, midrib with a few larger irregularly laciniate scales ;
texture coriaceous. Fertile fronds on stalks about as long as or longer than the
sterile ones, 2-3 cm. broad. Spores 27-30 X 18-21 JLI.

SOUTH INDIA: Anamallays, i,8oom., on trees, " F.S.Ind. f. 200 Elaphoglossum
laurifolium" Beddome (BM). Kudiabad, Sispara Rd., 2,400 m., Nov. 1883, Gamble
I 3473 (K). Bear Shola, Kodaikanal, Pulney Hills, 19 May 1898, Bourne 4972 (K).
Kodaikanal Waterfall, Pulney Hills, 30 May 1898, Bourne 4973 (K).

CEYLON : Thwaites C.P. 1311 (BM ; CGE ; K ; PDA). Gardner 1165 (BM ;
CGE, Nuwara Eliya, on trees, 1,500-2,100 m., Sept. 1844 ; E ; K). Gardner 1310
(E ; K). Nuwara Eliya, Thomson (K). Same locality, Freeman 369 A, 3?oB, 37 iC
(BM). Same locality, Day in Herb. Henderson (E.). Same locality, creeping over
tree trunks, 9 May 1906, Matthew (K). Pedrotalagala, 2,100 m., on mossy trees,
26 Dec. 1950, Holttum (SING). Adams Peak, on trees, 2,100 m., 14 Feb. 1908,
Matthew (K). Same locality, 1,950 m., 14 Dec. 1950, Sledge 615 (BM). Namunukula,
Uva Province, 1,950 m., 24 Feb. 1954, Sledge 1208 (BM). 1855, R. W. Rawson 1039
(BM). Mrs. Can ex Herb. T. Moore (K). Wall in Herb. Henderson and in Herb.
Neill Fraser (E).

South India (Nilgiri, Pulney and Anamallay Hills), Ceylon, Sumatra, Java,
Borneo, New Guinea, Philippine Islands, Tonkin, Formosa.

The long-creeping rhizome and broad, membranous, pale brown or rufous-brown,
concolorous scales readily distinguish E. angulatum from all other Asiatic species.
It differs also from all other Indian or Ceylonese species in having the veins united
by a connecting strand at their extremities a character of the Aconiopteris (E.
nervosum (Bory) Christ) group. This structural feature is however invisible in living
and herbarium specimens and can only be seen after treatment with a clearing agent.
Its spores are smaller than those of the other marginate-leaved species occuring in
India and Ceylon. E. stigmatolepis , which it most resembles in growth habit, may be
distinguished by its dark rhizome scales and the narrower fronds appearing punctate
beneath on account of the more peltate scales and the black points of attachment
left where these have been shed.

Ceylon gatherings of this fern are usually named E. conforme and Indian ones
either E. conforme or E. laurifolium (or E. latifolium). They certainly resemble
African E. conforme but they agree far better and are indeed indentical with authentic
specimens of Blume's A. angulatum at Kew and with other gatherings from Java.
Kunze (loc. cit.) long ago reached the same conclusion when he referred a Nilgiri
gathering (Schmid 65) to this species. A. angulatum was later merged in A. conforme
by Hooker in the Species Filicum, and subsequent authors followed Hooker in their
treatment of Ceylon and Indian plants.

Beddome's illustrations of what he names E. conforme and E. laurifolium in his
Ferns of Southern India and the descriptions in his Handbook have been the greatest

ELAPHOGLOSSUM IN THE INDIAN PENINSULA AND CEYLON 85

source of confusion surrounding the taxonomy of Indian Elaphoglossums. In the
former work his illustration (t. 198) and description, purporting to be A. conforme
Swartz, were based on plants on a sheet in the Kew collection annotated by him
" type of tab. 198 F.S.I. ". The specimens on this sheet represent unmistakable
though small and not very characteristic examples of E. stigmatolepis (Fee) Moore.
(Beddome's illustration (t. 199) of what he calls E. stigmatolepis Fee represents another
species.) The two different scales depicted represent examples taken from the
rhizome and, in the case of the shorter and broader one with markedly jagged margins,
from the stipe or midrib of the frond. The enlargement of a portion of the underside
of the frond shows stellate scales quite different from the minute peltate scales with
shortly fimbriate margins which characterize E. stigmatolepis, and this drawing could
not possibly have been made from any of the fronds on the sheet referred to. There
is moreover no reference in the accompanying description or in the later description
in the Handbook to the presence of stellate scales on the fronds. The earlier work
similarly contains no description of the form or colour of the rhizome scales, but in
the Handbook these are referred to as " blackish " in colour, a description which is
applicable to the specimens on the sheet and appropriate for E. stigmatolepis (and
even, though somewhat less accurately, for E. marginatum, which, from the amplified
distribution assigned in the Handbook to " conforme ", he now included in that species)
but which is completely false for E. angulatum or for the true African E. conforme.

It is clear from the evidence provided by Beddome's own specimens at Kew and
the British Museum that he was unfamiliar with E. angulatum for, whereas E.
stigmatolepis (i.e. Beddome's E. conforme) is stated to be " very common " and is
represented by several gatherings of his, he appears to have collected E. angulatum
once only. This single gathering is in the British Museum collection, the sheet being
labelled by Beddome " Anamallays. 6,000 ft., on trees. Elaphoglossum laurifolium
(Thouars) F.S.Ind. f. 200 ". The description of E. laurifolium in the Ferns of
South. India is followed by the citation of " E. angulatum Bl. En. 201 [sic] : Id. Fl.
Jav. 25, t. 6 " in synonomy and the reference to the rhizome as " wide-creeping with
bright chestnut or golden scales " and the other distinctions attributed in the
Handbook to his E. conforme and E. latifolium (the name there substituted for E.
laurifolium of the Ferns of Southern India) become explicable when it is realized that
his E. conforme is based on E. stigmatolepis and his E. laurifolium (or E. latifolium)
is based on E. angulatum.

Christensen states (in Gard. Bull. Str. Settl. 7 : 291 (1934)) a propos of Blume's
figures of his A. conforme (i.e. E. commutatum) and A. angulatum, " I am by no
means sure that the two ' species ' illustrated by Blume are really different ". And
so it might seem from the plates themselves, for whilst the plate of the latter is an
admirable representation of E. angulatum, that of the former is less characteristic
of E. commutatum particularly in the spacing and small size of the fronds. Blume's
drawings and descriptions of the rhizome scales of his two species however indicate
such considerable differences and they agree so very closely with the scales of E.
commutatum and E. angulatum that I have no hesitation in accepting the descriptions
and the plates as illustrations of these two species. Fee's plate of A. laurifolium
does not represent the same species. It is a Mascarene plant which has since been

86 ELAPHOGLOSSUM IN THE INDIAN PENINSULA AND CEYLON

described by Mme. Tardieu-Blot (Not. Syst. 15 : 433 (1959)) as E. alstonii.

The type of E. ogatai is at the British Museum. Of this Christensen (DanskBot.
Ark. 9 (3) : 67 (1937)) says " it is extremely like E. angulatum (Bl.) Moore . . . differs
however in its remarkably thin texture and the distinctly but shortly decurrent
lamina ". Tagawa, who is familiar with the living plant, states (in Mem. Coll. Sci.
Univ. Kyoto, ser. B, 20 : 28 (1951)) that its fronds are thin-coriaceous and that no
distinction can be drawn in the other respects. One of Bourne's Indian plants
(Bear Shola, Pulney Hills) also has quite thin fronds, and the more ovate rather
than elliptic fronds which look distinctive in the type specimen of E. ogatai can be
matched by a specimen of Matthew's from Adam's Peak, Ceylon. I therefore accept
Tagawa's reduction of E. ogatai to a synonym of E. angulatum.

Biswas's E. krajinae is quite clearly no more than a redescription of E. angulatum,
which he evidently never took into consideration when separating Indian and Ceylon
plants from African E. conforme.

2. Elaphoglossum stigmatolepis (Fee) T. Moore, Index Fil. : xvi, 15 (1857),
368 (1862). Christ in Neue Denkschr. Schweiz. Ges. Naturw. 36 (i, 2) : 52
(1899).

Acrostichum stigmatolepis F6e, Mem. Fam. Foug. 2 : 62, t. 24 f. 2 (1845). Hook., Sp. Fil.

5 : 216 (1864).
Elaphoglossum conforme sensu Bedd., Ferns S. Ind. : 67, t. 198 (1864) ; Handb. Ferns

Brit. Ind. : 416, fig. 247 (1883) pro parte ; non J. Smith.
Elaphoglossum ballardianum Biswas in Bull. Misc. Inf. Kew 1939 : 239, t. i fig. i (1939).

Rhizome long-creeping, stipes 5-10 mm. apart, clothed with ovate-lanceolate scales,
3-5 X 1-2 mm., brown at the base and blackish above, with subentire or sparsely
fringed margins and acute tips. Fronds stalked, 15-35 cm - l n g> stipes of sterile
fronds variable in length, sometimes as long as the lamina but commonly much less,
scaly, at least when young, with rather broad scales with jagged margins ; lamina
narrowly elliptic, 1-5-2-5 cm. broad, apex acuminate, base narrowed and decurrent
on the stipe, margins with a narrow, revolute, cartilaginous border, midrib sparsely
scaly beneath with short broad scales like those on the stipe, lower surface clothed with
minute, dark, peltate scales with fimbriate margins, upper surface with scattered scales
of the same form when young, becoming smooth with age ; texture stiffly mem-
branaceous to subcoriaceous. Fertile fronds on stipes normally exceeding those of
the sterile fronds, 1-2 cm. broad. Spores (39) 42-45 (48) X 27-30 fi.

SOUTH INDIA : " Ind. Orient Neilgherries, 159 " (CGE, Pisotype). Nilgiris,
2,100 m., " type of tab 198 F.S.I.", Beddome (K). Same locality, 1860, Beddome 5
(K). Same locality, 2,100 m., Oct. 1889, Gamble 21420 (K). Same locality, 2,100 m.
Nov. 1870, Henderson (E). Same locality, Wight 50 (K). Same locality, on trees and
rocks, 1842, Gough 3234/58 (K). Ootacamund, 2,100 m., Oct. 1885, Gamble 16963
(BM). Same locality and date, Gamble in Herb. Blanford (E). Same locality, 2,100
m., Aug. 1885, Gamble 16661 (K). Same locality, Babcock (K). Same locality,
Cockburn 85 (BM). Dodabetta, Nilgiri District, 2,550 m., June 1883, Gamble
12038 (E). Kudiabad, Sispara Road, 2,400 m., Nov. 1883, Gamble 13474 (K).

ELAPHOGLOSSUM IN THE INDIAN PENINSULA AND CEYLON 87

Moir Point, Pulney Hills, 2,340 m., stream, July 1937 and May 1938, Aroticasamy 10

(BM).

South India (Nilgiri and Pulney Hills).

Elaphoglossum stigmatolepis differs from the other marginate-leaved species in its
narrower, less coriaceous fronds, which are more tapering above and studded beneath
with minute disc-like scales with shortly fimbriate margins. The points of attach-
ment to the lamina appear as black dots when the scales have been shed. These
are referred to by Fee as "petits points qui paraissent tre de nature glanduleuse:
ne serait-ce pas la base persistante des squames? ' ' . Young fronds show more sparsely
distributed scales on the upper surface and in older fronds, where they have been
shed, similar minute black scars are recognizable. Fee does not allude to these in his
description but they are correctly shown in his illustration.

The rhizome scales are more pointed than those of E. angulatum, pale brown at
the base but almost black, and often somewhat glossy, in the upper parts, and hence
through imbrication they appear darker than the uniformly pale, rufous-brown and
thinner, more papery scales of E. angulatum. As in that species, the stipe bears
broad, brown scales with jagged margins and these are continued on to the lower
part of the midrib. The spores are considerably larger than in E. angulatum,
agreeing in size with those of E. marginatum, from which, and from E. commutatum,
it differs in its more strongly creeping rhizome, in its frond form and texture, and in its
scale characters.

Fee's description of E. stigmatolepis was based on a gathering " in herb, de Lessert.,
[sic] sub no. 159 ". The collector's name was evidently not stated on the label but
the provenance is given as "in Indiis orientalibus, Neilgherries ". There is in the
Cambridge University collection a sheet bearing a complete and well-preserved plant
with three sterile and one fertile frond attached to a long-creeping rhizome 15 cm.
long. No collector's name is inscribed on the label but this also reads " Ind. orient
Neilgherries 159 ". The specimen agrees so exactly with Fee's description and plate
that it is almost certainly an isotype. Wight 50 in herb. Kew is also an excellent
match for Fee's illustration and the Cambridge plant.

In Fee's description the fronds are said to be " membranaceis, papyraceis " and
he does not include this species in the coriaceous-leaved group to which A . margina-
tum, A. conforme and A. laurifolium are referred. At the same time the fronds are
described as narrowly re volute, and re volute edges are commonly associated with the
stiff-leaved species. I have seen no living plants of E. stigmatolepis but to judge from
herbarium specimens the fronds are thin but stiff and are distinctly less coriaceous
than those of the other marginate-leaved species.

E. ballardianum Biswas is merely a redescription of E. stigmatolepis and inaccurate
in the statement that the fronds lack a border. One of the gatherings which Biswas
cites is labelled Acrostichum stigmatolepis and another has had the same name pen-
cilled on the sheet, yet Biswas makes no reference to this species anywhere in his
paper, though he could hardly have failed to note its description in the literature
which he cites or that it was founded on plants from the identical area of his own
supposed new species.

88 ELAPHOGLOSSUM IN THE INDIAN PENINSULA AND CEYLON

3. Elaphoglossum beddomei Sledge, sp. nov.

Elaphoglossum stigmatolepis sensu Bedd., Ferns S. Ind. : 67, t. 199 (1864) ; Handb. Ferns

Brit. Ind. : 418 (1883) ; non T. Moore.
Acrostichum stigmatolepis sensu Baker in Hook. & Baker, Synops. Fil., ed. 2 : 521 (1874) ;

non Fee.

Rhizoma repens, paleis linearibus brunneis marginibus integris 6-9 mm. longis basi
i mm. latis vestitum, stipitibus crebrioribus. Frondes stipitatae, 20-35 cm - longae ;
stipites frondium sterilium plerumque breves et 3-6 cm. longi, vel, si longiores,
dimidia parte longitudinis laminae breviores, juventutes paleaci paleis linearibus
fimbriis marginalibus paucis ; lamina coriacea, anguste elliptica, 2-3 cm. lata, in
apicem acutum diminuta, basi gradatim decrescens et in alas angustas aut cristas in
stipite decurrens, marginibus angustis revolutis cartilagineis, infra paleis minutis
stellatis vestita, supra in frondibus junioribus paleis similibus vestita, in vetustioribus
glabrescens. Frondes fertiles eas steriles longitudine tola aequantes sed stipes
quam lamina aequaliter longus ; lamina 1-5-2-5 cm. lata. Sporae (36) 39-42 (45) X
24-30 fi.

SOUTH INDIA: Anamallays, 1,050 m., " F.S.Ind. f. 199 Elaphoglossum stigmatolepis
Kunze ", Beddome (BM, holotype). Palghat Hills, 1,500 m., Beddome (K). Nilgiri
Hills, 2,100 m., Beddome (K). Same locality, Beddome 6 (K). Same locality 1,500
m., 1870, Henderson (E). Sispara Ghat, Nilgiri District, Nov. 1883, Gamble 13418
(K).

South India (Nilgiri, Anamallay and Palghat Hills).

Elaphoglossum beddomei differs from E. stigmatolepis in its longer and narrower,
uniformly brown rhizome scales, the margins of which are entire or at most with an
occasional marginal fimbriation. Its shortly stalked more coriaceous fronds are
decurrent on the stipes, forming narrow wings or ridges often extending to the base,
and the minute scales on the under surfaces of the fronds are stellately laciniated.
The stipes bear linear scales markedly different from those of E. stigmatolepis, in
which they are short and broad and are continued on to the midrib of the lower
surface of the fronds.

Beddome's illustration well portrays the habit of this species. In one of the three
sterile fronds depicted the narrow wing extending downwards from the decurrent
lamina to the base of the stipe is correctly shown. Sometimes the stipes are con-
siderably longer than those illustrated, and then the decurrent wings merge into
ridges on the sides of the stipes. The fringed scale represents one of those clothing
the stipe, as the rhizome scales have margins which are mostly quite entire though a
few fimbriations may sometimes be present.

When Beddome first used the name E. stigmatolepis for this species he added a
query both after the name and following the citation of Fee's figure. Baker's
annotations on Beddome's specimens at Kew and his citation of Beddome's illustra-
tion show that this was the plant which he had in mind when adding Acrostichum
stigmatolepis to the second edition of the Synopsis Filicum and, following Baker, the
query was dropped by Beddome in his Handbook. Two distinct species however occur
in the Nilgiri and adjoining hills in South India and the true E. stigmatolepis is the
plant which was subsequently distinguished by Biswas as E, ballardianum. He

ELAPHOGLOSSUM IN THE INDIAN PENINSULA AND CEYLON 89

evidently accepted the present species, on the evidence of Beddome's illustration and
the named sheets at Kew, as being E. stigmatolepis , though he makes no reference to
this species anywhere in his paper but compares E. ballardianum with E. conforme,
to which its relationship is certainly more remote.

4. Elaphoglossum marginatum (Wall, ex Fee) T. Moore, Index Fil.: 8, u
(1857), 36i (1862).

Acrostichum marginatum Wall., Nunier. List : 2, n. 17 (1829), nom. nud.
Acvostichum marginatum Wall ex Fee, Mem. Fam. Foug. 2:31 (1845), excl. syn. Blume.
Acrostichum conforme sensu Hook., Sp. Fil. 5 : 198 (1864) pro parte. Hook. & Baker,

Synops. Fil. : 401 (1868) pro parte.
Elaphoglossum conforme sensu Bedd., Handb. Ferns Brit. Ind. : 416 (1883) pro parte minore

et excl. fig. 247 ; non J. Smith. Tagawa in Mem. Coll. Sci. Univ. Kyoto, Ser. B, 20 : 29

Elaphoglossum fuscopunctatum Christ in Bull. Herb. Boiss. 6 : 867 (1898) ; in Neue Denkschr.
Schweiz. Ges. Naturw. 36 (i, 2) : 51 (1899).

Rhizome stout, short-creeping, stipes close, clothed with dark brown, thick-walled,
narrow, lanceolate, attenuate, shortly and irregularly fringed scales, 4-5 X r-i'5 mm.
Fronds stalked, 20-50 cm., stipe of sterile frond up to half as long as lamina but
often much less, deciduously scaly ; lamina elliptic, 2-4 (5) cm. broad, apex acute,
base gradually narrowed and decurrent on the stipe, margins with a translucent
cartilaginous border, lower surface with scattered minute, more or less stellately
laciniated, brown scales, upper surface glabrous ; texture coriaceous. Fertile
fronds on longer stalks than the sterile ones, lamina slightly narrower. Spores
42-45 X 27-30 p.

NEPAL : Sheopore, 1829, Wallich 17 (K). Wallich 174 (BM). Sheopuri Lekh,
Katmandu, 1,350 m., 19 Aug. 1954, Stainton, Sykes and Williams 6917 (BM).
Between Torke and Okhaldunga, 1,950 m., 2 Nov. 1954, Zimmermann 1997 (BM).

SIKKIM : Darjeeling, 2,280 m., 9 Aug. 1875, C.B. Clarke 26913 (K). Same locality,
1,800 m., Oct. 1881, Gamble 9884 (K). Kohima, 1,800 m., 21 Oct. 1885, C. B. Clarke
40971 (K). Neebay (?), 2,250 m., 16 Oct. 18 , C. B. Clarke 25337 (BM). Jeylep Rd.,
1,800 m., Nov. 1880, Gamble 9938 (K). Treutler 988 (K). 2,100 m., Sept. 1882,
Levinge (K). Jerdon (K).

BHUTAN : Kancham, Punakha, 1,950 m., 28 Aug. 1914, Cooper 3058 (BM).
Trashiyangsi Valley, Tobrang, on rock faces, dense jungle, 29 Aug. 1934, Ludlow &
Sherriff902 (BM ; E).

ASSAM : Sohra River, 1,500 m., 16 Oct. 1872, C. B. Clarke 19239 (BM ; K). Same
locality, 1,500 m., 28 Nov. 1871, C. B. Clarke 14778 (K). Same locality, 1,350 m.,
15 Oct. 1872, C. B. Clarke 18826 (K). Vale of Rocks, Sept. 1886, C. B. Clarke
45828 (K). Suruseem, Khasia, 1,200-1,800 m., 26 May 1850, Hooker 6- Thomson (K).
Kalapanee, Khasia 1,650 m., 6 Aug. 1850 and 28 Oct. 1850, Hooker & Thomson (K).
Assam, 1870-80, G. Mann (K). Ukhrul, Manipur State, on rocks in the forest,
1,800 m., 12 Aug. 1948, Kingdon Ward 17935 (BM).

Himalayas eastwards from Nepal, Sikkim, Bhutan, Khasi Hills, Manipur, Yunnan,
Formosa.

go ELAPHOGLOSSUM IN THE INDIAN PENINSULA AND CEYLON

Elaphoglossum marginatum differs from the other coriaceous-leaved species in its
thicker, more shortly creeping rhizome with more crowded stipes. E. angulatum is
readily distinguished by its long-creeping habit with well-spaced fronds and much
broader, thinner and light brown rhizome scales. E. stigmatolepis also differs in its
more rampant habit, its narrower fronds, which are more tapering above and below,
and by the different form of the scales on the lower surface. Both Fee and Moore
included Java in the distribution of E. marginatum through inclusion of the plants
later distinguished by Mettenius as Acrostichum commutatum. From this E. mar-
ginatum is best distinguished by its stouter rhizome and more clustered fronds and
its darker coloured rhizome scales, which lack the long-attenuate flexuose and curled
apices and long marginal fimbriations of E. commutatum.

Christ makes no mention of Acrostichum marginatum in his monograph of Elapho-
glossum, but his E. fuscopunctatum described from Yunnan I consider the same as
Wallich's plant. There are specimens at Kew and the British Museum of Henry
9158, on which Christ's species was based, and these are, to me, indistinguishable
from some Himalayan gatherings of E. marginatum (e.g. Hooker < Thomson, from
Khasia in Herb. Kew ; Ludlow & Sherriff 902 from Bhutan in Herb. Edin.). The
scales in several other collections of E. marginatum give a conspicuously fusco-
punctate appearance to the undersides of the fronds. Tagawa (loc. cit.) reduced
both to forms of E. conforme.

5. Elaphoglossum commutatum (Mett.) Alderw. van Rosenb., Malayan Ferns,
Suppl. i : 427 (1917).

Acrostichum conforme sensu Blume, Enum. PI. Jav. 2 : [275] (1828) (errore " aemulum " in
p. 101) ; Fl. Jav., Fil. : 23, t. 5 (1829) ; non Swartz. Hook., Sp. Fil. 5 : 198 (1864)
pro parte. Hook. & Baker, Synops. Fil. : 401 (1868) pro parte.

Acrostichum marginatum Fee, Mem. Fam. Foug. 2 : 31 (1845) pro parte, incl. syn. Blume.

Elaphoglossum laurifolium sensu T. Moore, Index Fil. : 359 (1862) pro parte, quoad specim.
ex India, Ceylon, Java ; non T. Moore, op cit. : xvi (1857).

Acrostichum laurifolium sensu Thw., Enum. PI. Zeyl. : 380 (1864) ; non Thou.

Acrostichum commutatum Mett. apud Kuhn in Ann. Mus. Bot. Lugd.-Bat. 4 : 292 (1869).

Rhizome creeping, stipes rather close, clothed with narrow, brown scales with
fimbriate margins and long-attenuate, flexuose and twisted tips, 10 X 2 mm. Fronds
stalked, 15-30 (50) cm., stipe of sterile fronds variable in length, up to half as long
as the lamina but often much less, scaly when young with spreading, linear, fimbriate,
dark brown scales and with smaller, appressed, stellately laciniated scales, but be-
coming smooth with age ; lamina elliptic, 2-5 (6) cm. broad, apex acute, base
gradually narrowed and decurrent on the stipe, margins with a translucent, cartila-
ginous, revolute border, lower surface with scattered, minute, more or less stellately
laciniated scales, upper surface with similar scales when young, becoming glabrous;
texture coriaceous. Fertile fronds on stalks normally exceeding those of the sterile
ones and sometimes the stipes longer than the sterile fronds, 1-5-3-0 cm. broad.
Spores 33-36 X 24-27 11.

SOUTH INDIA : Bolampatty Hills, 1,500 m., Beddome (K). Nilgiri Hills, 1,200 m.,
Nov. 1870, Henderson (E).

ELAPHOGLOSSUM IN THE INDIAN PENINSULA AND CEYLON 91

CEYLON: Thwaites C.P. 1310 (BM ; CGE ; E; K; PDA). Hoolankande Peak,
Matale, 10 Jan. 1862, Brodie (E). Nuwara Eliya, Freeman 368 A (BM). Kabot's Gap,
Rangala, 8 Sept. 1927, Alston 2375 (PDA). Knuckles, 1881, no collector's name
(PDA). Ramboda Pass, 1,650 m., epiphytic on tree trunk by stream in jungle,
17 Dec. 1950, Sledge 646 (BM). Hoolankande, 1,350 m., 20 Jan. 1954, Sledge 1010
(BM). Jungle above Le Vallon Estate, 1,500 m., 9 Feb. 1954, Sledge 1116 (BM).
Parawella Falls, Kandapola, near Nuwara Eliya, 1,425 m., 19 March 1954, Sledge
JJ29 (BM). Robinson 20 (K). Central Prov., Naylor Beckett (K). 1871, Randall
in Herb. R. W. Rawson 3285 (BM). R. W. Rawson 1027-1802 (BM). Henderson (E).
Anderson (E).

South India (Nilgiri Hills), Ceylon, Java.

Elaphoglossum commutatum is intermediate in habit between E. marginatum and
E. angulatum, with less crowded stipes than the former but with shorter internodes
than the latter, in which the fronds are 1-2 cm. or more apart. It differs from
E. marginatum in its rhizome scales as described under that species, and in its smaller
spores. From E. angulatum, apart from its much less widely creeping rhizome, it
differs in its markedly narrower, darker brown, finely acuminate and strongly
fimbriate rhizome scales. The stipes bear similar linear dark-coloured scales inter-
mixed with small appressed stellate scales like those of the under surface of the
blades, which spread down the stipes, though in old fronds the stipes may be practi-
cally naked. In E. angulatum the broad, pale brown scales which clothe the rhizome
spread up the stipe and are continued on the lower part of the midrib of the lamina,
becoming narrower and more laciniated upwards and gradually merging into the
minute stellate scales of the under surface. Stellate scales are absent on the stipes of
E. angulatum.

Mettenius based his description (loc. cit.) on plants of Blume's from Java and on
Thwaites C.P. 1310 from Ceylon. His references to Thwaites's numbered collection
and to the description and figure of Blume's A. conforme and his own accurate
description of the rhizome scales fix the identity of the species. Van Alder wereldt
van Rosenburg validly transferred the specific epithet to Elaphoglossum though the
plants he had in mind were doubtfully the same as those of Mettenius. The species
has been treated in most works as a synonym of E. laurifolium (Thou.) Moore,
which was described from Tristan da Cunha. Christensen (Res. Norw. Sci. Exped.
Tristan da Cunha, 1937-38, 6 : 19 (1940)) expressed the view that E. laurifolium was
endemic to Tristan da Cunha. Through the kindness of Mr. James Dickson I have
been able to examine good specimens of E. laurifolium collected by him on Tristan da
Cunha in 1962. These and other gatherings from Tristan and from Gough Island
stand well apart from Ceylon and South Indian plants in their wide-creeping habit
with long internodes, their dark brown-black rhizome scales, which lack the fim-
briated edges and crisped apices of E. commutatum, and their larger spores (45-48 X
30-33 /*)

E. commutatum stands between E. laurifolium and E. callifolium, the latter being
a larger plant with bigger rhizome scales (up to 2 cm. long), which have entire
margins. Backer and Posthumus (Varemiora voor Java: 251 (1939)) maintained
E. commutatum (sub. nom, E, laurifolium) and E. callifolium as distinct species,

92 ELAPHOGLOSSUM IN THE INDIAN PENINSULA AND CEYLON

though they express the view that the former may be no more than a form of the
latter. I have no doubt as to Ceylon and South Indian plants being distinct from
E. callifolium but I am much less sure if plants from Java and elsewhere in Indonesia
which have been called E. laurifolium are identical with those from India and Ceylon.
Specimens so named from Java, Sumatra, Borneo and Indo-China have rhizome
scales the margins of which are less irregular and they lack the frizzed tips of those of
E. commutatum. They approach E. callifolium in these respects though the scales
are narrower and shorter.

I suspect E. permutatum Alderw. van Rosenb. (Bull. Jard. Bot. Buitenz., ser. 2,
16 : 13 (1914)) to be synonymous with E. commutatum. There is an isotype (Matthew
696 from Mt. Sago, Sumatra) at the British Museum and this has rhizome scales with
fimbriate margins and crisped apices although the description states that the scales
are " integerrimis ". In its other characters the cotype does not appear to me to be
significantly different from E. commutatum.

6. Elaphoglossum stelligerum (Wall, ex Baker) T. Moore ex Alston & Bonner
in Candollea 15 : 216 (1956).

Acrostichum neriifolium Wall., Numer. List : 2, n. 16 (1829), nom. nud.
Acrostichum stelligerum Wall., op. cit. : 65, n. 2167 (1830), nom. nud.
Elaphoglossum stelligerum T. Moore, Index Fil. : 15 (1857), 368 (1862), nom. nud.
Acrostichum viscosum sensu Hook., Sp. Fil. 5 : 220 (1864) pro parte ; non Swartz.
Elaphoglossum viscosum sensu Bedd., Ferns. S. Ind. : 67, t. 196 (1864) ; Handb. Ferns

Brit. Ind. : 420 (1883) ; non J. Smith.

Acrostichum stelligerum Wall, ex Baker in Hook. & Baker, Synops. Fil, ed. 2 : 521 (1874).
A cvostichum yunnanense Baker in Bull. Misc. Inf. Kew 1898 : 233 (1898).
Elaphoglossum yunnanense (Baker) C. Chr. in Contrib. U.S. Nat. Herb. 26 : 327 (1931).

Rhizome creeping with stipes close together, clothed with linear, attenuate, dark
castaneous, glossy scales, 3-5 X 0-5 mm., with occasional fimbriations in their lower
parts, elsewhere with entire or obscurely dentate margins. Fronds stalked, 20-40
cm. long, stipes of sterile fronds densely scaly with a mixture of stellate scales with
slender rays and lanceolate scales with long-fimbriate margins ; lamina narrowly
elliptic 1-2-5 cm - broad, apex acute, base gradually narrowed and decurrent on the
stipe, lower surface with abundant, loose, reddish-brown, stellate scales with long rays
which diverge from the leaf surface, the midrib with similar scales intermixed, especially
in the lower part, with broader-centred ones with long-fimbriate margins ; upper
surface stellate-pubescent when young, becoming more or less smooth with age, the
scales paler than those on the lower surface ; texture thin but firm. Fertile fronds
on longer stipes than the sterile ones, lamina narrower, to i cm. broad. Spores
42-45 X 30 11.

NEPAL : 1820, Wallich 16 (BM ; E ; K). Wallich ex coll. Hooker f. & Thomson 96
(E). Panapa, 1,546 m., 6 Sept. 1954, Zimmermann 1036 (BM). Between Maneb-
hanjyang and the Sun Kosi, 1,500 m., 5 Nov. 1954, Zimmermann 2058 B (BM).

SIKKIM : 600 m., Beddome (BM). 2,400 m., 1868, Henderson (E). 1871, Levinge
(K). Darjeeling, 1871, Hope (E). Same locality, coll. Levinge, ex Herb. Hope (E).
Same locality, Oct. 1873, C. B, Clarke 21486 (K). Goke, 17 July 1880, Levinge (E).

ELAPHOGLOSSUM IN THE INDIAN PENINSULA AND CEYLON 93

BHUTAN : Towards Kolepani, 3 Nov. 1835, Griffith (K).

ASSAM : Khasia, 1,350 m., Hooker & Thomson (E ; K). Same locality, 1,200 m.,
16 Sept. 1850, Hooker & Thomson 2285 (K). Same locality, 1,200 m., 16 Nov. 1871,
C. B. Clarke 14678 (K). Jowye, Jainteas, 1,200 m., 19 Oct. 1867, C. B. Clarke 5938
(K). Mariao, 1,200 m., 7 Nov. 1871, C. B. Clarke 15369 (BM ; E). Bogapani,
Khasia, 1,050 m., 29 Oct. 1871, C. B. Clarke 16367 (K). Same locality, 1,200 m.,
July 1890, G. Mann (E).

SOUTH INDIA: Pulney Mts., Sept. 1836, Wight 51 (K). Same locality, 1,800 m.,
Beddome (BM ; K). Periya Shola, Pulneys, 21 Apr. 1898, Bourne 4971 (K). Kodai-
kanal, Pulney Hills, 1,960 m., May 1937, Aroticasamy 7 (BM). Nilgiri, 1860,
Beddome 4 (K). Anamallays, 1,350 m., " F.S.Ind. f. 196 ", Beddome (BM). Same
locality, 1864, Henderson (E). Shevaroy Hills 1,200 m., 1898, coll. /. W. Furrell 66
ex Herb. Hope (E). Massif du Shevaroy, 10 Oct. 1939, Faucheux (BM). Honey Rock
stream, Yercaud, Shevaroy Hills, 1,100 m., 25 Oct. 1962, Ghatak Gig^A (BM).

Yunnan, Indo-China.

Elaphoglossum stelligerum is readily distinguished from all other Indian species
by its loose, rufous-brown, stellate scales with long rays which are not appressed to
the surfaces of the stipes or fronds. The linear, dark-coloured, somewhat glossy
rhizome scales are also markedly different from those of other species. It is closely
related to E. petiolatum (Swartz) Urban, described from Jamaica and formerly
credited with a pantropic distribution.

I see no significant difference between E. stelligerum and Baker's Acrostichum
yunnanense. Baker's description states " ad A. stigmatolepidem, Fee, magis accedit;
frondibus linearibus ad marginem paleis parvis ciliatis et paleis facialibus profunde
stellatim fissis differt " ; but the comparison with E. beddomei A. stigmatolepis
sensu Baker being E. beddomei mihi is incomprehensible since the plant Baker had
before him is manifestly far closer to, and is indeed in my opinion specifically identical
with, E. stelligerum. Ching evidently dismissed it as a valid species since there is a
second sheet of Henry 10310, the type gathering of A . yunnanense, in the same folder
at Kew which he has labelled E. petiolatum (Sw.) Urban. Christensen upheld
E. yunnanense (in Contrib. U.S. Nat. Herb. 26 : 327 (1931)) but his comments on the
distinction between it and E. petiolatum are equally applicable to E. stelligerum,
which he did not distinguish. Nor do I agree with Holttum (Fl. Malaya 2 : 455
(1954)) that Malayan specimens agree better with the type of E. yunnanense than
with South Indian specimens. They are more robust than either Henry's Yunnan
collection or Indian gatherings.

E. stelligerum is very closely related to E. blumeanum (Fee) J. Sm. from Malaysia
and E. salicifolium (Willd. ex Kaulf.) Alston from the Mascarene Islands (type from
Reunion) and Africa, all being formerly included in E. petiolatum Swartz, which was
described from Jamaica. Malaysian plants are larger than those from India and
the scales on the two surfaces of the sterile fronds are dissimilar. Indian and
Mascarene plants have the scales on the two surfaces alike and differences between
them are so slight that they probably represent no more than geographical variants.
The names are retained however pending a much-needed revision of the whole
E. petiolatum complex.

94 ELAPHOGLOSSUM IN THE INDIAN PENINSULA AND CEYLON

7. Elaphoglossum nilgiricum Krajina ex Sledge, sp. nov.

Elaphoglossum squamosum sensu Bedd., Ferns S. Ind. : 67, t. 197 (1864) ; Handb. Ferns
Brit. Ind. : 420, fig. 251 (1883) excl. specim. ex Ceylon ; non J. Smith.

Rhizoma breviter repens, paleis anguste lanceolatis vel linearis, pectinato-dentatis,
fusco-castaneis, marginibus nigris, nitidis, vestitum. Frondes caespitosae, breviter
stipitatae, 5-25 (40) cm. longae ; stipites paleis ovatis vel late ovatis, pallido-ferru-
gineis in margine dentibus setiformibus praelongis, densissime obtecti ; lamina
anguste oblonga vel longitudo-ligulata, 1-2 cm. lata, apice obtusa, in basi sensim
attenuata, in marginibus et utraque superficie copiose squamosa, infra paleis brun-
neis, mollibus, lanceolatis ad late ovatis, margine dentibus setiformibus fimbriata,
confertis imbricatis occultata, supra paleis similibus minus confertis, pallidioribus,
aliquibus paleis in ambitu plus-minusque orbiculatis sed in margine radiis tenuibus
longis instructis ; textura crassiter herbacea. Forma frondium fertilium ut in
sterilibus.

SOUTH INDIA : Sisparah Ghat, Nilgiris, 1,500 m., " type of tab 197. FSI.",
Beddome (K, holotype). Same locality, 1,800 m., Nov. 1883, Gamble 13468 (K).
Same locality, Miss Cockburn 81 (BM). Nilgiri, 1860, Beddome 7 (K). Pykara Falls,
Nilgiri; District, 1,800 m., June 1883, Gamble 12009 (K). Nediwattan, Nilgiri Hills,
i, 800 m., May 1866, Gamble ex Herb. Blanford (E). Nilgiri Hills, Henderson (E).

Endemic to Nilgiri Hills.

The type of Elaphoglossum Mrtum (Swartz) C. Chr. was described from Jamaica.
This name, or the synonym E. squamosum (Swartz) Urban, was formerly used in a
collective sense to cover plants ranging from tropical America eastwards to southern
India and Ceylon, which are the only regions in Asia whence plants have been found.
Several species based on collections from Africa, Madagascar, and the Mascarene
Islands have since been given independent rank. As South Indian and Ceylon plants
not only differ from each other but cannot be matched satisfactorily with any
Jamaican, African or Mascarene gatherings which I have examined, I have adopted
the unpublished names affixed to the Kew sheets by Krajina in 1937.

Specimens from the Nilgiri Hills are very uniform and they differ consistently
from Ceylon plants in the scales of the stipes being markedly broader with their
margins and setiform teeth rarely dark-coloured as in E. ceylanicum, save at the base
of the stipe adjacent to the rhizome. They also differ in their much more densely
paleate fronds, the lower surfaces of which are hidden beneath the closely imbricating
pale-brown scales.

Beddome's illustration (Ferns S. Ind. : t. 197 (1864)), is a fair representation of
the species except that the stipes are unaccountably portrayed as smooth and glabrous
whereas his description states that they are " very scaly " and both in his own and
all other gatherings they are invariably densely paleate with patent scales. The
isolated, long, falcate and apparently glabrous frond depicted on the left-hand side
of the plate is meaningless and, not surprisingly, is omitted from the reproduction
of this plate in the Handbook.

ELAPHOGLOSSUM IN THE INDIAN PENINSULA AND CEYLON 95

8. Elaphoglossum ceylanicum Krajina ex Sledge, sp. nov.

Rhizoma breviter repens, paleis linearibus pectinato-dentatis fusco-castaneis vel
nigris vestitum. Frondes caespitosae, breviter stipitatae, 5-25 cm. longae ; stipites
paleis anguste lanceolatis, attenuatis, acutis, ferrugineis, in margine dentibus
setiformis praelongis, plerumque marginibus dentibusque fuscis, obtecti ; lamina
oblongo-elliptica, 1-2 cm. lata, in apice acuta vel acuminata, in basi sensim attenuata,
in marginibus et utraque superficie paleis tenuibus brunneis ; supra pallidioribus
vestita, sed infra, praeter in juventute, non paleis occultata.

CEYLON : Gardner 1164 (BM ; CGE on trees in dark forests at Nuwara Eliya,
Sept. 1844 ; K, holotype). Thwaites, C.P. 3292 (BM ; CGE ; E ; PDA). Central
Province, on trees and rocks, Naylor Beckett 401 (K). Peacock Hill, Pussilawa,
1870, Randall in Herb. R. W.Rawson 3284 (BM). Pundaloya, 1,500 m., Freeman
3J2 A, 3J4A, 3753 (BM). Nuwara Eliya, Freeman 3738 (BM). Walker (K).
Robinson 22 (K). Rocks by stream at Lonach near Norton Bridge, Central Province,
900 m., 13 Dec. 1950, Sledge 59$ (BM). Ramboda, on mossy rock by stream in jungle,
1,650 m., 17 Dec. 1950, Sledge 648 (BM). Kuda Oya, Ramboda, on rocks in shade
by stream through jungle, 1,725 m., 28 Dec. 1950, Sledge 769 (BM).

Endemic.

The distinctions between this species and E. nilgiricum are referred to under the
previous species. The fully developed fronds are acute at the apex and green in
colour since neither surface is covered by a continuous coating of scales as is the lower
surface of the frond in E. nilgiricum. Young fronds however are very scaly. In
both species the lamina scales vary in shape, the body of the scales being mostly
broadly to narrowly ovate in outline but some are more or less orbicular. All have
the margins fringed by long setiform teeth.

E. ceylanicum seems to come closest to E. deckenii var. rufidulum (Willd. ex Kuhn)
Tardieu-Blot (Not. Syst. 15 : 430, t. 4 f. 6-10 (1959) =E. hirtum var. rufidulum
(Willd. ex Kuhn) C. Chr. in Dansk. Bot. Ark. 7 : 170 (1932)) from Madagascar and
Reunion. In that species, however, the fronds are borne on long stipes and the
shape of the rhizome scales and their marginal teeth are different.

Plants from Madeira and the Azores have jet black scales on the stipes and under
surface of the costae, intermixed with ferrugineous ones with and without thickened
and darkened margins. Such scales are not present in E. nilgiricum or E. ceylanicum.
Specimens from Madeira and the Azores are closest to E. splendens (Bory) Brackenr.
from Reunion and Mauritius but are better treated as a separate species. 1

1 Elaphoglossum paleaceum (Hook. & Grev.) Sledge, comb. nov.

Acrostichum paleaceum Hook. & Grev., Ic. Fil. 2 : t. 235, Alph. Index et Syst. Index (1831) (errore

" vestitum " in expl. tab.).

It would seem that Hooker and Greville realized during the preparation of the Icones Filicum that
Lowe's manuscript name Acrostichum vestitum was a later homonym. Its appearance in the text
accompanying the plate, in spite of the fact that A . paleaceum is used on the plate and in the indexes, is
to be regarded as an error.

g6 ELAPHOGLOSSUM IN THE INDIAN PENINSULA AND CEYLON

9. Elaphoglossum spatulatum (Bory) T. Moore, Index Fil. : 14 (1857).
Bedd., Ferns S. Ind. : 71, t. 209 (err. 213) (1864).

Acrostichum spatulatum Bory, Voy. Mers d'Afriq. i : 363, t. 20 f. i (1804). Fee, Mem. Fam.

Foug. 2 : 51, t. 14 f. 3 (1845).
Acrostichum acutum Fee ex Kuhn in Linnaea 36 : 44 (1869).

Rhizome creeping with stipes rather close together, clothed at the apex with
filiform, rufous scales. Sterile fronds up to 10 cm. long, stipes about equalling the
lamina in length and covered with patent, reddish, setaceous scales ; lamina lanceolate
to rhomboid-lanceolate, 1-1-5 cm - broad, base narrowed into the stipe, apex usually
acute sometimes obtuse, margins and both surfaces covered with hair-like scales ;
texture rather thickly herbaceous. Fertile fronds smaller than the sterile ones and
on much longer stipes, lamina broadly oval or almost orbicular, conduplicately
folded when young and often splitting at the apex when mature.

CEYLON : Thwaites C.P. 989 (BM ; E ; K ; PDA). Adam's Peak, Moon (BM).
Nuwara Eliya, Rawson W. Rawson 1041 (BM). Same locality, Freeman 376 A,
37?B (BM). Sita Eliya, Hakgala, March 1885 (PDA). Hawa Eliya, 25 Aug. 1926,
/. M. de Silva (PDA). Kuda Oya, Ramboda Pass, 1,700 m., in shady places on rocks
by stream in jungle, 28 Dec. 1950, Sledge JJi (BM). 1,500-1,800 m. " Type of tab
209 FSI ". Thwaites in Herb. Beddome (K). 1870, Thwaites in Herb. Henderson (E).
On wet rocks at about 1,800 m., Hutchison (E.). Wall (E). Robinson 23 (K).
Gardner (K).

Africa (N. & S. Rhodesia, Transvaal, Natal, Cape), Madagascar, Reunion.

A very distinct species, much smaller than all other Indian and Ceylon species. '
Closely related plants, which have often been treated as belonging to the same species,
occur in tropical America and Tristan da Cunha ; but Christensen (Res. Norw. Sci.
Exped. Tristan da Cunha, 1937-38, i (6) : 20 (1940)) considered the species on that
island, E. obtusatum (Carm.) C. Chr., to be " certainly quite different from E. spathula-
tum ".

Madame Tardieu-Blot (Not. Syst. 15 : 432 (1959) ; in Humbert, Fl. Madag.,
Polypod. 2 : 49 (1960)) has erroneously transferred the Ceylon records of E. spatulatum
to E. schizolepis (Baker) Christ. The type of the latter from Madagascar is at
Kew and is quite unlike any Ceylon plant.

PRINTED IN GREAT BRITAIN
BY ADLARD & SON LIMITED
BARTHOLOMEW PRESS, DORKING

FUNGI OF RECENT NEPAL
EXPEDITIONS

FRANCES L. BALFOUR-BROWNE

BULLETIN OF

THE BRITISH MUSEUM (NATURAL HISTORY)
BOTANY Vol. 4 No. 3

LONDON: 1968

FUNGI OF RECENT NEPAL EXPEDITIONS

BY

FRANCES L. BALFOUR-BROWNE,

Pp. 97-141 ; 4 Text-figures

BULLETIN OF

THE BRITISH MUSEUM (NATURAL HISTORY)
BOTANY Vol. 4 No. 3

LONDON: 1968

THE BULLETIN OF THE BRITISH MUSEUM

(NATURAL HISTORY), instituted in 1949, is
issued in five series corresponding to the Departments
of the Museum, and an Historical series.

Parts will appear at irregular intervals as they become
ready. Volumes will contain about three or four
hundred pages, and will not necessarily be completed
within one calendar year.

In 1965 a separate supplementary series of longer
papers was instituted, numbered serially for each
Department.

This paper is Vol. 4, No. 3 of the Botany
series.

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

Trustees of the British Museum (Natural History) 1968

TRUSTEES OF
THE BRITISH MUSEUM (NATURAL HISTORY)

Issued 24 June, ig68 Price Nineteen Shillings

FUNGI OF RECENT NEPAL EXPEDITIONS

By FRANCES L. BALFOUR-BROWNE

THE fungi recorded below were collected during four general botanical expeditions
made in the Nepal highlands by (i) J. D. A. Stainton, W. R. Sykes and L. H. J.
Williams in 1954, (2) J. D. A. Stainton in 1956, (3) A. H. Norkett in 1961-62 and (4)
J. D. A. Stainton in 1962. Expeditions (i) and (4) were in Central Nepal while
the other two were mainly in Eastern Nepal.

The collections were made at altitudes between 300 m. and 5,000 m. At the
highest levels were Conifer and Rhododendron forests with Evergreen Oaks coming
in from below, and the lower levels were characterized by rice terraces, bamboos,
ferns and Castanopsis forest. Sheals & Inglis (1965) give details of the local geog-
raphy and the nature of the terrain encountered in Expedition (3).

In this account references have been restricted to the authorities of the names
used and to the more significant or well-known synonyms. For additional references
to most of the species the revised (1960) edition of Butler & Bisby's indispensable
Fungi of India by Vasudeva should be consulted.

In some groups the delimitation of genera is very controversial and consequently
the classification and the names employed are very much a personal matter of
opinion with very little agreement amongst workers. This is particularly so for
the Polyporaceae. During the last twenty years new systems of classification for
this group have been published by Pilat (1936-42, Europe), Cunningham (1947,
1947-50, 1965, New Zealand), Corner (1932, 1953, general), Overholts (1953, America),
Imazeki (1943, Japan), Donk (1933, 1960, general), Singer (1962, general), Bondart-
zev (1953, Russia), Nobles (1958, Canada), Pinto-Lopez (1952, Portugal), Kotlaba
& Pouzar (1957, Europe), Teixeira (1962, Brazil). These systems have been based
on a variety of criteria: general anatomical and morphological structure, hymenial
structure, hyphal structure, biosystematics, hyphal thickening and the presence or
absence of clamps, and physiological characters. Significant as are these characters,
nevertheless owing to the difference in emphasis placed upon them by different
authors, considerable difficulty arises in attempting to derive a stable or consistent
nomenclature. Here, therefore, the well-known, old or mainly Friesian subdivisions
have been used for the Polyporaceae and the less conservative names included in
the synonymy.

Recently suggested, but still tentative, relationships of certain agarics (so hitherto
regarded) with polypores rather than with other agarics, or vice versa, are also not
taken up here. The second edition of R. Singer's The Agaricales in Modern Taxo-
nomy (1962) should be referred to for new ideas on the classification of this group.

As regards the Clavariaceae, these have been named by Dr. E. J. H. Corner of
Cambridge, and each identification for which he is responsible is indicated by his
initials in parentheses.

In mountainous countries such as Nepal with an annual rainfall in some areas
of 500 cm., the difficulty is to get the plants dry. The tendency is therefore to
press them too enthusiastically, with the consequence that some of the agarics and

BOX. 4, 3 7

ioo FUNGI OF RECENT NEPAL EXPEDITIONS

Clavariae were tissue-paper thin and their hyphal structure indiscernible. To off-set
this it would have been valuable to have had some of the soft and fleshy fungi
preserved in fluid, and full notes as to colour, texture and shape when fresh are
always much desired. For the rest, the material was in good condition and together
represents the largest collection so far from this difficult and until recently almost
inaccessible region: 160 species excluding some immature and over-ripe specimens
which so far have resisted identification. Previous records consist of about two
dozen species collected by Sir Joseph D. Hooker over a century ago and described
by Berkeley (1850, 1851, 1852, 1854) an d about 70 species collected by Polunin,
Sykes and Williams in 1952 and reported in a previous number of this journal
(Balfour-Browne, 1955).

All the specimens cited in this paper are in the herbarium of the British Museum
(Natural History).

PHYCOMYCETES

ALBUGINACEAE

ALBUGO BLITI (Biv.) Kuntze, Revis. Gen. PL 2 : 658 (1891). Best. Biga in Sydowia
9 : 347 (1955)-

Uredo bliti Biv., Stirp. Rar. Sicil. 3:11 (1815). '

Cystopus bliti (Biv.) De Bary in Ann. Sci. Nat., Ser. 4, Bot. 20 : 131, tab. 13 figs. 13-15

(1863).

Caeoma amaranthi Schwein. in Trans. Amer. Phil. Soc., New Ser., 4 : 292 (1832).
Cystopus amaranthi (Schwein.) Berk, in Grevillea 3 : 58 (1874).

NEPAL: Bhurungdi Khola, 1,600 m., on Amaranthus lividus L., 2oth May, 1954,
Stainton, Sykes & Williams 5342.
Distribution: Worldwide.
For a recent review of the genus Albugo consult Bestagno Biba (torn. cit. : 339-58).

PERONOSPORACEAE

SCLEROSPORA GRAMiNicoLA (Sacc.) Schroet. in Cohn, Krypt.-Fl. Schles. 3 (i) : 236
(1886).

Protomyces graminicola Sacc. in Nuov. Giorn. Bot. Ital. 8 : 172 (1876).

NEPAL: Bongakhani, 2,130 m., on grass, 22 Aug. 1954, Stainton, Sykes & Williams

3943-

Distribution: America, Europe, Africa, India, Australia.

PYTHIACEAE

PK'/TOPHTHORA INFESTANS (Mont.) De Bary in Journ. Roy. Agric. Soc., Ser. 2, 12
2^0 (1876).

Botrytis infestans Mont, in L'Institut, Sect, i, 13 : 313 (1845).

NEPAL: Murigurja Gad, 2,500 m., on potato, 27 July 1954, Stainton, Sykes
Williams 3654.
Distribution: Worldwide.

PUNCH OF RECENT NEPAL EXPEDITIONS 101

ASCOMYCETES

MORCHELLACEAE

MORCHELLA ELATA Fries, Syst. Mycol. 2 : 8 (1822).

NEPAL: Siklis, north of Pokhara, 3,000 m., on rotten tree trunk, 21 Apr. 1954,
Stainton, Sykes & Williams 4956.

Distribution: America, Europe, India, China, Japan, Australia.

HELVELLACEAE

HELVELLA CRISPA Fries, Syst. Mycol. 2 : 14 (1822).

NEPAL: Lete, 2,600 m., beneath conifers, 27 Aug. 1954, Stainton, Sykes & Williams

7501.

Distribution: Worldwide; previously recorded from Nepal in 1955.

HUMARIACEAE

ALEURIA AURANTIA (Fries) Fuckel in Jahrb. Nass. Ver. Naturk. 23-24 : 325 (1870).

Peziza aurantia Fries, Syst. Mycol. 2 : 49 (1822).

NEPAL: Ghar Khola, 2,130 m., 14 June 1954, Stainton, Sykes & Williams 5759.
Distribution : Worldwide.

GEOGLOSSACEAE

GEOGLOSSUM AFFINE (Durand) Sacc. & Trav. in Sacc., Syll. Fung. 19 : 756 (1910).
Maas Geest. in Persoonia 4 : 23 (1965).

Gloeoglossum affine Durand in Ann. Mycol 6 : 420 (1908).

NEPAL: Annapurna Himal, Seti Khola, 3,830 m., on shady banks, 28 July 1954,
Stainton, Sykes & Williams 6540. South of Gurjakhani, 4,000 m., among dwarf
Rhododendron, 17 Aug. 1954, Stainton, Sykes & Williams 3879.

Distribution : United States, Himalayas.

These specimens were examined and identified by Dr. Maas Geesteranus.

MITRULA ROSEA Lloyd, Mycol. Not. 61 : 885, t. 129 fig. 1529 (1919).

NEPAL: Gosainkund, Malemchi, 2,800 m., on bare earth, 30 May 1962, Stainton

3789-

Distribution: India.

Pale rose coloured ascophores about i cm. tall with smooth or contorted caps.

SCLEROTINIACEAE

RUTSTROEMIA sp. cf. FiRMA (Fries) Karst. in Bidr. Kann. Finl. Natur & Folk 19 :
108 (1871).

Peziza firma Fries, Syst. Mycol. 2 : 117 (1822).

Ciboria firma (Fries) Fuckel, Symb. Mycol. : 312 (1869).

Phialea firma (Fries) Gill., Champ. Fr., Discom, : 101, t. 74 fig. 2 (1883).

FUNGI OF RECENT NEPAL EXPEDITIONS

NEPAL: Dhankuta Province, near Mahe, 1,300 m., on dead twig, 20 Sept. 1961,
Norkett 5/75 B.

Distribution (of R. firma) : Worldwide.

The fungus agrees well with this species except that its spores are somewhat
short, 9-11 x 4-5 /*. The material, however, is barely ripe.

HYALOSCYPHACEAE

Perrotia malemchiensis Balfour-Browne, sp. nov. (Fig. i.)

Apothecia sparsa, superficialia, sessilia, carnosa, uda pallide brunnea, sicca cinna-
barina, extus villosa, 500-1,000 /* diam. ; setae cylindraceae ad basim i cohaerentes
apice liberae et acutae, septatae, minute granulosae, pallide ochraceae ; asci clavati,
recti vel curvuli, octospori, 80-110 X 9-10 /*; sporae clavatae vel fusoideae, hyalinae,
5-7 septatae, 20-30 X 4-5 /i; paraphyses filiformes, septatae, hyalinae.

Ad ligna et cortices arborum frondosarum.

FIG. i. Perrotia malemchiensis Balfour-Browne. A, asci containing ascospores;
B, paraphyses; c, ascospores; D, external hair; E, apothecium. Stainton 3j68.

FUNGI OF RECENT NEPAL EXPEDITIONS

103

NEPAL: Gosainkund, Malemchi, 2,400 on a dead tree, 29 May 1962, Stainton,
3768 (holotype).

This differs from related species not only in colour and measurements but in spore
septation: P. fusca Mull. & Dennis, i-septate; P. lutea (Phill.) Dennis, up to 31-
septate; P. himalayensis Mull. & Dennis, 3-septate.

DERMATEACEAE

Mollisia dhankutae Balfour-Browne, sp. nov. (Fig. 2.)

Apothecia superficialia, sessilia, usque ad 1-5 mm. diam., disco sordide flavido-
albo; excipulo fusco-brunneo, pseudoparenchymatico ; asci cylindraceo-clavati,
octospori, poro jodo tincto, 80-90 X 9-10 ju; ascosporae biseriatae, elongato-

2mm

FIG. 2. Mollisia dhankutae Balfour-Browne. A, asci and paraphyses;
B, ascospores ; c, apothecia. Norkett 7751" A .

104 FUNGI OF RECENT NEPAL EXPEDITIONS

fusoideae, uniseptatae, hyalinae, 16-20 X 3-5-4 /*; paraphyses filiformes, septatae,
2 /*, supra usque 3-5 /JL.
In ramis siccis Bambusae.

NEPAL: Dhankuta Province, Taplejung district, Sanghu, 1,750 m., on dead
bamboo stump, 19 Nov. 1961, Norkett 7757 A (holotype). Sanghu, 2,060 m., 9 Nov.
1961, Norkett 7349.

Although very close to Mollisia caesia var. andina Dennis (Kew Bull. 14 : 441
(1960)) from Venezuela, it differs in the absence of a whitish margin to the receptacle,
in the slightly larger and more acutely pointed spores and in the different host. It
differs also from Cenangella bambusicola Rick (Broteria 5 : 37 (1906)), which was
described from living bamboo in South America, in that the paraphyses do not form
an epithecium and the apothecia are not at first white, but externally very dark
brown with an off-white disc.

OSTROPACEAE

VIBRISSEA TRUNCORUM Fries, Syst. Mycol. 2 : 31 (1822).

NEPAL: Rambrong, Lamjung Himal, 4,500 m., on dead roots of Rhododendron,
10 July 1954, Stainton, Sykes & Williams 6269..

Distribution : America, Europe. Apparently not previously recorded from Nepal
or neighbouring countries, possibly because it is easily overlooked, being small and
generally found on roots, frequently submerged in water, or on debris.

Apothecia scattered or in groups, with orange discs, 3-5 mm. diam., seated on
pale stalks, blackish below, 15 X 2-3 mm. Asci 250-300 X 6-8 fi; ascospores
acicular, hyaline, multiseptate, 180-200 X 1-5 fi. Paraphyses branched, filiform
with spherical heads.

DIATRYPACEAE

DIATRYPE CHLOROSARCA Berk. & Broome in Journ. Linn. Soc. Lond., Bot. 14 : 123

(i873).

NEPAL: Tumlingtar, Arun ravine, on dead twigs, 23 Dec. 1961, Norkett 9060.

Distribution : Ceylon and India.

The Nepal collection agrees with the type material described from Ceylon : spores
7-9 x 2-2-5 jn, hyaline to light brown.

HYPOCREACEAE

cf. BALANSIA ANDROPOGONIS Syd. apud H. & P. Syd. & Butl. in Ann. Mycol. 9 :

395 (1911). Patel, Gokh. & Kulk. in Indian Phytopath. 4 : 65 (1951).

NEPAL: Mayangdi Khola, 1,000 m., on inflorescence of Chrysopogon aciculatus,
4 Sept. 1954, Stainton, Sykes 6- Williams 4137.

Distribution (of B. andropogonis) : India, Philippines.

The inflorescence of the host is so deformed and shrouded by the fungus as to
give the appearance of quite a different genus of grass. Only the Ephelis stage of
the fungus is represented: conidia acicular, 20-24 x 1 '5 /*

FUNGI OF RECENT NEPAL EXPEDITIONS 105

EPICHLOE CINEREA Berk. & Broome in Journ. Linn. Soc. Lond., Bot. 14 : in

(i873).

NEPAL: Taplejung district, Dhankuta Province, Sanghu, 2,000 m., on an indeter-
minable grass haulm, 17 Oct. 1961, Norkett 5696 C.

Distribution: Previously recorded from Ceylon. Sydow and Butler's record from
Mysore (Ann. Mycol. 9 : 394 (1911)) appears to be a different species, as Petch (Ann.
Roy. Bot. Gdns. Peradeniya 7 : 88 (1920)) has noted.

PHYLLACHORACEAE

CATACAUMA REPENS (Corda) Theiss. & Syd. in Ann. Mycol. 13 : 383 (1915).

Sphaeria repens Corda, Icones Fung. 4 : 42, tab. 9 fig. 123 (1840).
Phyllachora repens (Corda) Sacc., Syll. Fung. 2 : 597 (1883).

NEPAL: Phewa Tal, 800 m., on fallen leaves oiFicus religiosa, 6 May 1954, Stainton,
Sykes & Williams 5254.

Distribution: Cuba, Natal, India.

Several species of Catacauma have been described from Ficus spp. but this Nepal
collection is quite typical of C. repens.

XYLARIACEAE

HYPOXYLON MULTIFORME (Fries) Fries, Summa Veg. Scand. : 384 (1849).
Sphaeria multiformis Fries, Syst. Mycol. 2 : 334 (1823).

NEPAL: Taplejung, Mewa Khola, 2,750 m., on moss-covered branch (? birch),
22 Jan. 1962, Norkett 9300. Ganesh Himal, Ankhu Khola, 2,130 m., on rotting
log, 12 May 1962, Stainton 3694.

Distribution: Widespread in Northern hemisphere; previously recorded from
Nepal.

HYPOXYLON TRUNCATUM (Schwein.) J. H. Mill, in Trans. Brit. Mycol. Soc. 17 : 130
(1932); Monogr. World Sp. Hypoxylon : 95 (1961).

Sphaeria truncata Schwein. in Schr. Naturf. Ges. Leipz. i : 44 (1822). Fries, Syst. Mycol.
2 : 442 (1823).

NEPAL: Bakhri Kharka, north of Pokhara, 2,000 m., 25 Apr. 1954, Stainton, Sykes
6- Williams 5065.

Distribution: Tropical and semitropical ; America, Africa, China, Japan.

Miller (loc. cit., 1961) should be consulted for an account of this species and its
considerable synonymy.

XYLOSPHAERA HYPOXYLON subsp. ADSCENDENS (Fries) Dennis in Bull. Jard. Bot.
Bruxelles 31 : 124 (1961).

Sphaeria adscendens Fries in Linnaea 5 : 537 (1830).

Xylosphaera adscendens (Fries) Dennis in Kew Bull. 13 : 102 (1958).

Xylaria hypoxylon f. tropica H. & P. Syd. & Butl. in Ann. Mycol. 9 : 418 (1911).

Xylaria hypoxylon var. tropica (H. & P. Syd. & Butl.) Balf.-Browne in Bull. Brit. Mus.

(Nat. Hist.), Bot. i : 216 (1955).
BOT. 4, 3. 7

io6 FUNGI OF RECENT NEPAL EXPEDITIONS

NEPAL: South of Gurjakhani, 3,500 m., 18 Aug. 1954, Stainton, Sykes & Williams

Distribution: South America, West Indies, Africa, India, Nepal, Indonesia.

This is the fungus previously recorded from Nepal as Xylaria hypoxylon var.
tropica (Balfour-Browne, loc. cit.). Dennis (loc. cit., 1961) considers that it should
be treated as a separate subspecies of Xylosphaera hypoxylon Dumort. Dennis had
not seen the type specimen of Xylaria hypoxylon forma tropica and left open the
question of whether that name was a synonym of Xylosphaera hypoxylon subsp.
adscendens. I too have not seen the type, but I follow Dennis in considering that
this is a separate subspecies.

XYLOSPHAERA MELLISSII (Berk.) Dennis in Kew Bull. 13 : 104 (1958); in Revista
Biol. i : 186 (1958).

Hypoxylon mellissii Berk, in Melliss, St. Helena : 379 (1875).

Xylaria mellissii (Berk.) Cooke in Grevillea n : 85 (1883).

Xylaria arbuscula Sacc. in Michelia i : 249 (1878). J. H. Mill, in Bothalia 4 : 265 (1942).

NEPAL: Dhankuta Province, Chainpur district, Tumlingtar, Sabhaya Khola, 600
m., " growing out of a niche in rock ", 8 Dec. 1961, Norkett 8471 A.

Distribution : Generally common in the tropics and subtropics, but not previously
reported from the Himalayan region.

The Nepal collection consists of branched stromata with long slender stalks and
short cylindrical heads ; ascospores are 14-16 X 4-5 fi.

XYLOSPHAERA POLYMORPHA (St. Amans) Dumort., Comment. Bot. : 92 (1822).
Dennis in Bull. Jard. Bot. Bruxelles 31 : 140 (1961).

Sphaeria polymorpha St. Amans, Fl. Agenaise : 520 (1821).
Xylaria polymorpha (St. Amans) Grev., Fl. Edinensis : 355 (1824).

NEPAL: Arun Valley, Sabhaya Khola, 2,300 m., on tree trunk in forest, 7 Sept.
1956, Stainton 1607.
Distribution : Worldwide.
For a recent account and complete synonymy see Dennis (loc. cit.).

XYLOSPHAERA TELFAIRII (Berk.) Dennis in Kew Bull. 13 : 106 (1958) ; in Bull. Jard.
Bot. Bruxelles 31 : 119 (1961).

Sphaeria telfairii Berk, in Ann. Nat. Hist. 3 : 397 (1839).

Xylaria telfairii (Berk.) Fries in Nov. Acta Reg. Soc. Sci. Upsal., Ser. 3, i : 127 (1851).

NEPAL: Arun Valley, Kasuwa Khola, 2,800 m., on tree trunk in forest, n Sept.
1956, Stainton 1619. Arun Valley, Hatiar, 2,600 m., 21 Aug. 1956, Stainton
14.08.

Distribution: Africa, Ceylon, India, Indonesia, Australia.

For a recent account of the species, Dennis (loc. cit.) should be consulted.

FUNGI OF RECENT NEPAL EXPEDITIONS 107

USTULINA DEUSTA (Fries) Petrak in Ann. Mycol. 19 : 279 (1921).

(Sphaeria deusta Fries, Syst. Mycol. 2 : 345 (1823).
Ustulina vulgaris Tulasne frat., Sel. Fung. Carp. 2 : 23, tab. 3 figs. 1-6, (1863), nom.
super fl.

NEPAL: Bakhri Kharka, north of Pokhara, 2,000 m., 25 April 1954, Stainton,
Sykes & Williams 5056.
Distribution : Worldwide.

CORONOPHORACEAE

CORONOPHORA EPiSTROMA Syd. apud. Syd., Mitter & Tand. in Ann Mycol. 35 : 231

(1937)-

NEPAL: Sanghu, gulley below camp, 2,000 m., on dead twig, 21 Dec. 1961, Norkett

6390.

Distribution: Originally recorded from Allahabad, and not reported since.

This species was originally described as parasitic in the stromata of Haplosporella
phyllanthina and again in the present gathering it is growing within the remains of
a fungal pycnidium, but the identity of the latter could not be determined, nor that
of the host twigs.

PLEOSPORACEAE

FENESTELLA FENESTRATA (Berk. & Broome) Schroet. in Cohn, Krypt.-Fl. Schles.
3(2) : 435 (1897). (Fig. 3)-

Valsa fenestrata Berk. & Broome in Ann. Mag. Nat. Hist., Ser. 3, 3 : 366, t. 10 fig. 14

(i859).
Fenestella princeps Tulasne frat., Sel. Fung. Carp. 2 : 207 (1863), nom. superfl.

NEPAL: Dhankuta district, Chitre, on dead twig, 2,000 m., 20 Sept. 1961, Norkett
5159 A.

Distribution: N. America, Europe.

The fungus occurs on an unidentified dicotyledonous twig. The pseudothecia
are superficial, i stipitate, arising in small groups on an inconspicuous stroma.
Asci cylindrical, 250 X 20 /A, approx. Ascospores 30-50 X 12-14 fi, broadly fusi-
form, dark brown, the end cells being i hyaline ; there are 3 main and several lesser
transverse septa and several longitudinal septa.

Petrak (Sydowia 8 : 165 (1954)) describes Cucurbitaria pakistanica from Choa
Saidan Shah, on Acacia modesta. This resembles the Nepal fungus in general
structure but its spores are considerably smaller, the average size being 13-22 X
8-10 fi.

VENTURIACEAE

REHMIODOTHIS OSBECKIAE (Berk. & Broome) Theiss. & Syd. in Ann. Mycol. 12 : 192
(1914).

Dothidea osbeckiae Berk. & Broome in Journ. Linn. Soc. Lond., Bot. 14 : 134 (1873).

io8

FUNGI OF RECENT NEPAL EXPEDITIONS

2mm

FIG. 3. F ene stella fenestrata (Berk. & Broome) Schroet. A, asci containing
ascospores; B, ascospore; c, stroma with pseudothecia. Norkett 5x59 A.

NEPAL: Dhankuta Province, Taplejung district, Sanghu, 2,000 m., on Osbeckia
sp., 4 Oct. 1961, Norkett 5696 D.

Distribution: Previously recorded only from Ceylon.

POLYSTOMELLACEAE

SCHNEEPIA sp. cf. DISCOIDEA (Racib.) Racib. ex Theiss. & Syd. in Ann. Mycol. 13 :
203 (1915).

Parmularia discoidea Racib., Parasit. Algen & Pilze Java's 2:21 (1900).

NEPAL: Dhankuta Province, Taplejung district, Sanghu, 1,600 m., on fern,
Davallia sp., 7 Jan. 1962, Norkett 8552.

Distribution (of 5. discoidea) : Previously recorded on Poly podium longissimum
from Java.

FUNGI OF RECENT NEPAL EXPEDITIONS 109

The Nepal material is not quite ripe, but appears to represent the above species.
Its rounded black stromata occur on the underside of the fronds and are somewhat
raised in the centre. The asci and the i-septate brownish spores, which are not
fully mature, measure approximately 35 X 8 ju, and 10 X 4 / respectively.

Several black stromatic Ascomycetes with i-septate spores have been described
on ferns. Careful comparison of type and authentic material is needed and would
probably reveal that a number of names are synonyms.

BASIDIOMYCETES

USTILAGINACEAE

CINTRACTIA CARICIS (Pers.) Magnus in Verh. Bot. Ver. Prov. Brandenb. 37 : 79
(1896).

Uredo caricis Pers., Syn. Meth. Fung. : 225 (1801).

NEPAL: Lamjung Himal, 4,500 m., on inflorescence of Cyperaceae, 14 July 1954,
Stainton, Sykes & Williams 6350. East of Chalike Pahar, 4,500 m., 4 Aug. 1954,
Stainton, Sykes 6- Williams 3743.

Distribution: Worldwide.

FARYSIA OLIVACEA (DC.) H. & P. Syd. in Ann. Mycol. 17 : 41 (1919).

Uredo olivacea DC., Fl. Fran. 6 : 78 (1815).

NEPAL: Bhurungdi Khola, 1,600 m., smut balls in inflorescence of Carex cruciata
Wahl., 20 May 1954, Stainton, Sykes & Williams 5336. Near Lumsum, 2,300 m.,
24 Oct. 1954, Stainton, Sykes & Williams 9127.

Distribution: Worldwide.

SPHACELOTHECA HYDROPIPERIS (Schumach.) de Bary, Vergl. Morph. Biol. Pilze :
187 (1884).

Uredo hydropiperis Schumach., Enum. PI. Saell. 2 : 234 (1803).

NEPAL: Southwest of Gurjakhani, 3,300 m., in flower oiPolygonum campanulatum,
20 Oct. 1954, Stainton, Sykes < Williams 9104. Dhankuta Province, Taplejung
district, Sanghu, 2,000 m., 25 Nov. 1961, Norkett 7532 A.

Distribution: Worldwide on Polygonum sp. Previously recorded from Nepal in
1955-

SPHACELOTHECA REILIANA (Kiihn) Clint, in Journ. Mycol. 8 : 141 (1902).
Ustilago reiliana Kiihn in Rabenh., Fung. Eur. : No. 1998 (1875).

NEPAL: Murigurja Gad, 2,500 m., on Zea mays, 27 July 1954, Stainton, Sykes 6
Williams 3653.
Distribution: Recorded in most countries where maize is grown.

no FUNGI OF RECENT NEPAL EXPEDITIONS

USTILAGO BISTORTARUM (DC.) Korn. in Hedwigia 16 : 38 (1877).
Uredo bistortarum DC., Fl. Fran9. 6 : 76 (1815).

NEPAL: East of Chalike Pahar, 4,830 m., on leaves of Polygonum affine, 22 Sept.
1954, Stainton, Sykes & Williams 4541.

Distribution: Worldwide.

Forming in this material elongated pustules over the under surface of the leaf
lamina, and not occurring on the margins. The spores are pale purple, angular to
globose, 10-16 fji, minutely verrucose.

This species, common in most parts of the world, appears not to have been re-
corded previously from India or neighbouring countries.

Ustilago hordei (Pers.) Lagerh. in Mitt. Badischen Bot. Ver. [2] (59) : 70 (1889),
non Ustilago hordei Bref . (li

Uredo segetum var. hordei Pers., Syn. Meth. Fung. : 224 (1801).

NEPAL: Gurjakahni, 2,800 m., on barley inflorescence, I June 1954, Stainton,
Sykes & Williams 2941. Tegar, north of Mustang, 6 Aug. 1954, Stainton, Sykes <
Williams 2252.

Distribution : Worldwide.

The name used for this species is illegitimate, being a later homonym of Ustilago
hordei Bref. (in Nachr. Klub Landwirthe Berl. 221 : 1593 (28 June 1888)), which
was based on the loose smut of barley, as a name for which it antedated U. nuda
(Jensen) Kellerm. & Swingle (see below). Before the correct names of these two
species can be established, it is necessary to fix the application of the name Uredo
segetum Pers. (Syn. Meth. Fung. : 224 (1801)), and also to investigate the status
and application of the names Uredo carbo-tritici and Uredo carbo-hordei quoted by
Kellerman & Swingle (in Annu. Rep. Exp. Stat. Kansas State Agric. Coll. 2 : 262,
269, 278 (1890)) from a work by Philippar not available to me. I therefore adopt
for the present the names in current use.

USTILAGO MORINAE Padw. & Azmat. Khan in Mycol. Pap., Imp. Mycol. Inst. 10 : i

(I944)-

NEPAL: Above Seng Khola, 4,500 m., in flowers of Morina, 25 June 1954, Stainton,
Sykes & Williams 3254.

Distribution : Recorded from Kashmir in 1944 on Morina longifolia.

Ustilago nuda (Jens.) Kellerm. & Swingle in Annu. Rep. Exp. Stat. Kansas State
Agric. Coll. 2 : 277 (1890).

Ustilago segetum var. nuda Jens, in Journ. Roy. Agric. Soc., Ser. 2, 24 : 406 (1888).
Ustilago nuda Rostr. in Tidsskr. Land0kon 8 : 745 (1889), nom. nud.

NEPAL: Chimgoan, north of Tukucha, 3,000 m., on barley in field, 3 June 1954,
Stainton, Sykes & Williams 904. Gurjakhani, 2,830 m., on wheat, I June 1954,

FUNGI OF RECENT NEPAL EXPEDITIONS in

Stainton, Sykes 6- Williams 2940. Tamur valley, Mewa Khola, on wheat, 17 May
1956, Stainton 343.

Distribution: Worldwide.

For the status of this name, see under Ustilago hordei above.

USTILAGO PIPERI Clint, in Proc. Boston Soc. Nat. Hist. 31 : 382 (1904). G. W.
Fisch., Man. N. Amer. Smut Fungi : 291, fig. 123 D (1953).

NEPAL: Near Seng Khola, 4,600 m., on Polygonum rumicifolium, 9 Aug. 1954,
Stainton, Sykes 6- Williams 38og.

Distribution : United States of America.

The Nepal collection agrees well with this species; it forms large pustules on the
underside of the leaves; the spores are pinkish purple, spherical or subspherical,
7-9 fi diam., marked with striae. This species differs from U. bistortarum in the
distinctly smaller spores.

TILLETIACEAE

MELANOTAENIUM SELAGINELLAE Henn. & Nyman in Warb., Monsunia i : 2 (1900).

NEPAL: Dhankuta Province, Taplejung district, Sanghu, on Selaginella sp., 27
Oct. 1961, Norkett 6551 A.

Distribution: Previously recorded from Tjibodas in Java.

In the present collection the smut occurs mostly on the basal portion of the leaves,
but occasionally it covers the whole leaf surface. Spores dark brown, globose,
coarsely verrucose, 15-18 fi diameter.

MELAMPSORACEAE

COLEOSPORIUM BARCLAYENSE Bagchee in Ind. Forest Rec., New Ser., Bot. 4 : 53

(1950).

NEPAL: Annapurna Himal, Mardi Khola, 4,160 m., on Senecio alatus Wall, ex
DC., 20 Sept. 1954, Stainton, Sykes & Williams 8528.

Distribution: India.

The teleutospore stage is represented in this collection. Morphologically, the
Coleosporium species on Senecio are difficult to separate, but, on the basis of inocula-
tion experiments carried out by Bagchee (loc. cit.) on rusts on different species of
Senecio, that on 5. alatus would appear to be his C. bar clay ense.

COLEOSPORIUM CAMPANULAE (Pers.) Kickx, Fl. Crypt. Flandres 2 : 54 (1867).
Gaum, in Beitr. Krypt.-Fl. Schweiz 12 : 113, fig. 99-100 (1959).

Uredo campanulae Pers., Syn. Meth. Fung. : 217 (1801).

NEPAL: Near Jagat, 2,800 m., on Campanula sp., 5 July 1954, Stainton, Sykes &
Williams 3376. Above Lumsum, 3,700 m., on Campanula seedlings, 23 Oct. 1954,
Stainton, Sykes & Williams 9122. Rambrong, Lamjung Himal, 3,000 m., on Lobelia
seguinii var. doniana (Skottsb.) Wimm., 27 Oct. 1954, Stainton, Sykes & Williams

ii2 FUNGI OF RECENT NEPAL EXPEDITIONS

8306. Annapurna Himal, Mardi Khola, 2,600 m., on Lobelia sp., 20 Sept. 1954,
Stainton, Sykes 6- Williams 8531. Baglung, 1000 m., on Wahlenbergia gracilis DC.,
20 Apr. 1954, Stainton, Sykes & Williams 548.

Distribution: Widely represented in the Northern hemisphere and previously
recorded from Nepal.

The uredospores occurred on stems as well on both leaf surfaces. A greater
proportion of the spores on Wahlenbergia were oblong and elongated compared with
those on Campanula. Probably two biological forms are represented. Gaumann
(loc. cit.) should be consulted for an account of at least six such forms which have
been isolated from different genera and species of Campanulaceae.

COLEOSPORIUM INULAE Rabenh. in Bot. Zeit. 9 : 455 (1851). P. & H. Syd., Monogr.
Ured. 3 : 609 (1915). Vasud. in Butl. & Bisby, Fungi of India, Rev. Ed. : 85
(1960).

Uredo inulae Kunze in Klotsch, Herb. Viv. Mycol., No. 589 (1844), nom. nud.

NEPAL: Pasgam, 1,500 m., on Inula cappa DC., 25 June 1954, Stainton, Sykes &
Williams 5938.

Distribution: Europe, North Africa, Asia Minor, India.

The uredo stage is represented and this chiefly on the upper surface of the leaves,
with very few pustules on the lower.

COLEOSPORIUM PEDICULARIDIS Tai in Farlowia 3 : 100 (1947).

NEPAL: Taglung, south of Tukucha, Kali Gandaki, 4,000 m., on leaves of Pedicu-
laris sp., 22 Sept. 1954, Stainton, Sykes & Williams 7987. Panchasi, 2,300 m., on
Pedicularis sp., 15 Oct. 1954, Stainton, Sykes & Williams 8944.

Distribution: China.

There is little doubt that the Nepal fungus represents this species. The uredosori
are chiefly on the under surface of the leaves, light yellow and soon pulverulent.
The uredospores are coarsely and densely echinulate, 18-30 x 12-20 ju,, oblong to
irregularly globose.

COLEOSPORIUM PLECTRANTHI Barcl. in Journ. Asiatic Soc. Bengal 59 (2): 89 (1890).

NEPAL: Sattewati, 2,000 m., on Plectranthus sp., 12 Oct. 1954, Stainton, Sykes &
Williams 8950.

Distribution: India, Japan.

COLEOSPORIUM SENECIONIS (Pers.) Fries, Summa Veg. Scand. : 512 (1849).
Uredo farinosa var. senecionis Pers., Syn. Meth. Fung. : 218 (1801).

NEPAL: Above Lumsum, 3,160 m., on Senecio graciliflorus DC., 10 Sept. 1954.
Stainton, Sykes & Williams 4327. Dhankuta Province, Milke Danda, 3,160 m.,
on Senecio graciliflorus DC., 16 Nov. 1961, Norkett 7121 A.

Distribution : Worldwide.

Uredospore stage. See note under Coleosporium barclayense.

FUNGI OF RECENT NEPAL EXPEDITIONS 113

HYALOPSORA POLYPODII (Pers.) Magn. in Ber. Deutsch. Bot. Ges. 19 : 582 (1902).
Uredo linearis var. polypodii Pers., Syn. Meth. Fung. : 217 (1801).

NEPAL: Annapurna Himal, Seti Khola, 4,160 m., on Polypodium malacodon,
3 Aug. 1954, Stainton, Sykes & Williams 6603.

Distribution: North America, Europe, India, Japan.

MELAMPSORA sp. cf. HIRCULI Lindr. in Acta. Soc. Fauna & Flora Fenn. 22 (3) : 19

(1902).

NEPAL: Annapurna Himal, 4,150 m., on Saxifraga moorcroftiana Wall., 2 Aug.
1954, Stainton, Sykes & Williams 6587.

Distribution (of M. hirculi): Europe (Finland, Russia, Switzerland).

Although this species so far has apparently only been recorded from Europe, the
Nepal collection agrees very closely with the original description of M. hirculi, but
the fungus occurs not only on the under surfaces, but more frequently on the upper
leaf surfaces, which become considerably discoloured and blotchy. Uredospores are
globose, ovate or ellipsoid, minutely verrucose, 16-25 X 14-18/6; paraphyses are
abundant and capitate or clavate, 35-60 X 10-20 ju with a thick wall. The teleuto-
spores are brown, oblong, 35-55 X 8-12 /*.

MILESINA sp. cf. EXIGUA Faull in Journ. Arnold Arb. 12 : 218 (1931).

NEPAL: Dhankuta Province, Taplejung district, 1,950 m., on Diplazium sp.,
9 Jan. 1962, Norkett 864.2 A.

Distribution (of M. exigua) : Poland, Japan and the Siberian coast.

Uredosori only are represented; they occur in brown discoloured areas on both
sides of the fronds. Many of the spores are somewhat irregularly polygonal in
shape. They measure 28 X 15 / (average) and are quite smooth.

PUCCINIACEAE

FROMMEA DUCHESNEAE (Arth.) Arth. in Bull. Torrey Bot. Club 44 : 504 (1917) ; in
N. Amer. Fl. 7 : 731 (1925).

Kuhneola duchesneae Arth. in N. Amer. Fl. 7 : 185 (1912).

Frommea obtusa [var.] duchesneae (Arth.) Arth., Man. Rusts U.S. & Canada : 93 (1934).

Gaum, in Beitr. Krypt.-Fl. Schweiz 12 : 1177 (1959)-
Frommea obtusa-duchesneae Vienn.-Bourg. in Rev. Path. Veg. Entom. Agric. Fr. 33 : 38

(1954), nom. superfl.

NEPAL: Ghar Khola, 1,600 m., on Duchesnea indica, 14 June 1954, Stainton, Sykes
& Williams 5767.

Distribution: Nepal, and North America and France, where the host is naturalized.

It seems that Arthur in 1934 (loc. cit.) had second thoughts about treating the
rust on Duchesnea indica as a full species and, describing it as a " less robust form "
compared with F. obtusa, regarded it merely as a variety of the latter species.

In the Nepal collection the uredo-sori are abundant, the uredospores measure

BOX. 4, 3. 7

n 4 FUNGI OF RECENT NEPAL EXPEDITIONS

9-14 X 18-20 fi, distinctly small for F. obtusa (which occurs on Potentilla sp.) but
in agreement with the spores described on Duchesnea from America by Arthur and
from France by Viennot-Bourgin. No other spore form was present on the Nepal
plants. In the absence of firm evidence to support synonymy with F. obtusa the
name F. duchesneae is retained here. This is the first record of the rust in what is
considered to be the host plant's centre of origin.

GYMNOSPORANGIUM PADMARENSE Balf.-Browne in Bull. Brit. Mus. (Nat. Hist.) i :

205, fig. 2 (1955).

NEPAL: Near Gurjakhani, 3,000 m., on Juniperus wallichiana, 3 June 1954,
Stainton, Sykes 6- Williams 2969.

Distribution: Nepal.

The material under consideration is old, the sori are broken up and the teleuto-
spores have lost their pedicels.

GYMNOSPORANGIUM CUNNINGHAMIANUM Barcl. in Sci. Mem. Med. Off. Army Ind.

5 : 78, t. 1-3 (1890).

NEPAL: Village south of Chakure Lekh, 6 Apr. 1952, Polunin, Sykes < Williams
1870.

Distribution: India, Nepal.

PHRAGMIDIUM INCOMPLETUM Barcl. in Journ. Asiatic Soc. Bengal 59 (2) : 83 (1890).

NEPAL: Sikh's, north of Pokhara, 3,000 m., on Rubus sp., 21 Apr. 1954, Stainton,
Sykes & Williams 4950.

Distribution: India.

The uredospore stage only is present, and it agrees well with the original account,
except that it is chiefly epiphyllous. Uredospores have a thick, 3-4 fi, epispore,
which is warted. There are no paraphyses.

PHRAGMIDIUM sp. cf. NEPALENSE Barcl. in Journ. Asiatic Soc. Bengal 60 (2) : 220

(1891).

NEPAL: Bhuji Khola, 2,800 m., on Potentilla nepalensis, 16 Oct. 1954, Stainton
Sykes & Williams 9058.

Distribution (of P. nepalense) : India.

Uredosori only are represented. In the original description it was not statec
whether the epispore is smooth or echinulate. Padwick & Azmatullah Khan
(Mycol. Papers, Imp. Mycol. Inst. 10 : 4 (1944)) record this species and describe the
uredospores as finely echinulate. This agrees with those on the present collection
and the measurements are similar.

PUCCINIA CARICIS var. HIMALAYENSIS (Barcl.) Padw. & Azmat. Khan in Mycol
Papers, Imp. Mycol. Inst. 10 : 9 (1944).

Aecidium urticae var. himalayense Barcl. in Sci. Mem. Med. Off. Army Ind. 2 : 29, t. 4
5 figs. 8-17 (1887); in Journ. Asiatic Soc. Bengal 56 (2) : 368 (1887).

FUNGI OF RECENT NEPAL EXPEDITIONS 115

NEPAL: Jagat, 2,600 m., on Urtica sp., 5 July 1954, Stainton, Sykes & Williams

Distribution: India, Himalayas.

Cf. Padwick & Azmatullah Khan (loc. cit.) for an account of the somewhat con-
fused synonymy.

PUCCINIA FAGOPYRI Barcl. in Journ. of Bot. 28 : 261 (1890).

NEPAL: Dhankuta Province, Taplejung district, Sanghu, 2,000 m., onFagopyrum,
ii Nov. 1961, Norkett 7173 A.
Distribution: India.

PUCCINIA GENTIANAE (Strauss) Link in L., Sp. PL, Ed. 4, 6 (2) : 73 (1825).
Uredo gentianae Strauss in Ann. Wetter. Ges. 2 : 102, t. n fig. 33 (1811).

NEPAL: South of Gurjakhani, 4,000 m., on Gentiana sp., 8 June 1954, Stainton,
Sykes 6- Williams 3071.

Distribution : Widely distributed in the northern Hemisphere.
Only the aecidial stage is represented in the Nepal material.

PUCCINIA GLUMARUM (J. K. Schmidt) Erikss. & Henn. in Medd. K. Landtbr.-Akad.
Exp. 38 : 141 (1896).

Uredo glumarum J. K. Schmidt in Allgem. Okonom.-Tech. Flora i : 27 (1827).

NEPAL: Gurjakhani, 2,800 m., on Triticum vulgare, I June 1954, Stainton, Sykes
6- Williams 2939.
Distribution: Worldwide.

PUCCINIA GRAMINIS Pers., Syn. Meth. Fung. : 228 (1801).

NEPAL: Ghasa, Kali Gandaki Valley, 2,500 m., on Berberis sp., 31 May 1954,
Stainton, Sykes & Williams 5499', same locality and host, 13 June 1954, Stainton,
Sykes & Williams 5750.

Distribution : Worldwide.

Some of the pustules were a bright pink and considerably swollen.

PUCCINIA LANTANAE Farl. in Proc. Amer. Acad. Arts & Sci. 18 : 83 (1883). G.
Laund. in Mycol. Papers, Commonw. Mycol. Inst. 89 : 43 (1963).

NEPAL: Sanghu, Dhankuta Province, 2,000 m., on Justicia diffusa, 15 Oct. 1961,
Norkett 6124.

Distribution: North and South America, India, Indonesia, Philippines, China,
Japan.

PUCCINIA LEUCOPHAEA H. & P. Syd. & Butl. in Ann. Mycol 10 : 258 (1912).

NEPAL: Ghar Khola, 2,000 m., on Colquhounia coccinea, 14 June 1954, Stainton,
Sykes 6- Williams 5758.

n6 FUNGI OF RECENT NEPAL EXPEDITIONS

Distribution: India.

Aecidial stage only and chiefly epiphyllous; only a few isolated sori on the lower
surface and on the petioles. In the original account the aecia were said to be
hypophyllous.

PUCCINIA POLYGONI-AVICULARIAE Pers., Syn. Meth. Fung. : 227 (1801).

Puccinia polygoni Alb. & Schwein., Consp. Fung. : 132 (1805), nom. superfl. Gaum, in
Beitr. Krypt.-Fl. Schweiz 12 : 775 (1959).

NEPAL: Dhankuta Province, Taplejung district, Sanghu, 2,000 m., on Polygonum
nepalense, 25 Nov. 1961, Norkett 7530 A.

Distribution : Worldwide.

The teleutospore stage is represented. This species and P. polygoni-amphibii are
united by some mycologists but treated as separate by others. The two species
are maintained here not only on biological grounds but also on account of definite
small morphological distinctions, which apply not only to European plants, as
generally stated, but also to American specimens as exemplified in the British
Museum herbarium. In P. polygoni-amphibii, usually occurring on Polygonum
amphibium, P. lapathifolium and related species, the teleutosori remain for long
covered by the epidermis and form small pimply pustules, and the spores are
frequently somewhat bent and easily lose their pedicels. On the other hand, in P.
polygoni-aviculariae, usually on Polygonum dumetorum and P. convolvulus, the
pustules rapidly burst through the epidermis of the host, leaving smooth black
cushions of straight, stalked teleutospores.

In the present instance the host, Polygonum nepalense, resembles in general
appearance and texture Polygonum convolvulus, and the teleutospores and sori of
the fungus agree exactly with those described for P. polygoni-aviculariae.

PUCCINIA PULVERULENTA Grev., Fl. Edin. : 432 (1824). Gaum, in Beitr. Krypt.-Fl.

Schweiz 12 : 929 (1959).

NEPAL: Near Dogadi Khola, 4,300 m., on Epilobium sp., 23 June 1954, Stainton,
Sykes < Williams 3226.

Distribution: Worldwide.

PUCCINIA USTALIS Berk, in Hook., Journ. Bot. 6 : 207 (1854).
Puccinia songarica Jacz. in Hedwigia 39 : (130), fig. i (1900).

NEPAL: Rambrong, Lamjung Himal, 4,000 m., on Ranunculus sp., 29 June 1954,
Stainton, Sykes & Williams 6016.

Distribution: India, Turkestan, Mongolia.

RAVENELIA EMBLICAE Syd. apud H. & P. Syd. & Butl. in Ann. Mycol. 4 : 438 (1906).

P. & H. Syd., Monogr. Ured. 3 : 293 (1914).

NEPAL: Chainpur path, Tumlingtar, 800 m., on Phyllanthus emblica, 21 Dec. 1961,
Norkett 8100.

FUNGI OF RECENT NEPAL EXPEDITIONS 117

Distribution: Previously recorded from India and Burma.

Ravenelia phyllanthi Mundk. & Thirum. (Mycol. Papers, Imp. Mycol. Inst. 16 :
24, fig. 18 (1946)) seems to be synonymous. This was described on Phyllanthus
polyphyllus from Mysore.

UREDINALESForm Genera

AECIDIUM CRINI Kalchbr. in Grevillea n : 26 (1882). Mundk. & Thirum. in Mycol.
Papers, Imp. Mycol. Inst. 16 : 16 (1946).

Aecidium amaryllidis H. & P. Syd. & Butl. in Ann. Mycol. 10 : 274 (1912).

NEPAL: Dana, Kali Gandaki Valley, 1,600 m., on Crinum amoenum Roxb. ex
Ker-Gawl., 13 June 1954, Stainton, Sykes & Williams 5738.
Distribution : India, South Africa.

AECIDIUM INFREQUENS Barcl. in Journ. Asiatic Soc. Bengal, 59 (2) : 105 (1890).

NEPAL: Nr. Dogadi Khola, 4,300 m., on Geranium collinum Steph. ex Willd., on
open slopes, 23 June 1954, Stainton, Sykes & Williams 3227.

Distribution: India, Nepal, Japan.

The aecidia cover considerable areas of the lower side of the leaves, showing as
light brown patches on the upper surface. Each aecidium is 250-300 ^ diam.
The size, ornamentation, colouring of the aecidia and aecidiospores correspond
exactly with those originally described for A. infrequent by Barclay on a Geranium
sp. (? nepalensis] from Simla. The present fungus agrees also with the details
given for A . sanguinolentum on other Geranium spp. in Finland, Russia and America by
Lindroth (Bot. Notiser 1900 : 241), and he suggested this might be a synonym of
Barclay's rust.

Polunin, Sykes 6- Williams 4765, recorded in 1955 in this Journal (Balfour-Browne,
1955) as A. infrequent is microscopically similar to the fungus now reported, but
differs in that the aecidia are grouped in orbicular patches with a small bare spot
in the centre of each patch, i.e., as described by Lindroth for the less heavily in-
fected specimens of his A. sanguinolentum. The host of this earlier collection, which
has now been identified also as G. collinum, is more elegant and slender. Possibly
the heavy rust infection is responsible for the coarser growth of the host specimen
of Stainton, Sykes & Williams 3227.

A . sanguinolentum, in consequence of inoculation experiments from Geranium spp.
of European origin (Lindroth, loc. cit.), has been described as a stage in life history of
Puccinia polygoni-amphibii Pers. However, until experiments are made using
specimens of the Nepal fungus, no certain conclusions can be drawn as to its relation-
ships or alternative host plants.

In the meantime A. infrequent Barcl. is the name preferred on grounds of distri-
bution and would be also on grounds of priority should this species prove to be
identical with A . sanguinolentum.

n8 FUNGI OF RECENT NEPAL EXPEDITIONS

AECIDIUM SCUTELLARIAE Syd. apud H. & P. Syd. & Butl. in Ann. Mycol. 5 : 504
(1907).

NEPAL: Ghar Khola, 1,800 m., on Scutellaria scandens D. Don., 14 June 1954,
Stainton, Sykes & Williams 5763.

Distribution: Himalayas.

A similar fungus, Aecidium scutellariae-indicae Dietel, has been described from
Japan on Scutellaria indica var. japonica. This may be identical with the above
species (P. & H. Syd., Monogr. Ured. 4 : 115 (1923)).

PERIDERMIUM ORIENTALS Cooke in Ind. Forester 3 : 91 (1877) " orientalis ".
Aecidium complanatum Barcl. in Journ. Asiatic Soc. Bengal 59 (2) : 101 (1890).

NEPAL: Dhaibungkot, 1,600 m., on dead pine needles, 31 May 1949, Polunin 041.
Near Beni, 1,300 m., on needles of Pinus longifolia, 23 May 1954, Stainton, Sykes (
Williams 2794.

Distribution: India, Nepal, Bhutan.

UREDO HYPERICI-MYSORENSIS Petch in Ann. R. Bot. Gdns. Peradeniya 6 : 213

NEPAL: Sanghu, Dhankuta Province, 1,400 m., on Hypericum sp., 12 Nov. 1961,
Norkett 7088.

Distribution: Ceylon, India.

AURICULARIACEAE

AURICULARIA DELICATA (Fries) Henn. apud Bresad., Henn. & Magn. in Engl., Bot.
Jahrb. 17 : 492 (1893).

Laschia delicata Fries in Linnaea 5 : 533 (1830).

Laschia tremellosa Fries, Summa Veg. Scand. : 325 (1849).

NEPAL: Ranipauwa, north of Beni, Kali Gandaki, 1,000 m., 3 Sept. 1954, Stainton,
Sykes & Williams 7629.

Distribution: mostly tropical; America, Africa, India, Australia, Pacific.

AURICULARIA MESENTERICA Pers., Mycol. Eur. i : 97 (1822).

NEPAL: Arun Valley, Num, 1,500 m., on tree trunk, 30 Aug. 1956, Stainton 1459.

Distribution: America, Europe, India, Indonesia, Australia.

The fructifications are broadly attached, many more or less disciform, and they
therefore super fici ally resemble A. peltata Lloyd. However the hairs are much
longer, up to 500 /JL.

AURICULARIA POLYTRICHA (Mont.) Sacc. in Atti R. 1st Veneto, Ser. 6, 3 : 722 (1885).

Exidia polytricha Mont, in Belang., Voy. aux Indes-Or. 2 : 154 (1834).
Hirneola polytricha (Mont.) Fries in K. Vet.-Akad. Handl. 1848 (i) : 146 (1849).

NEPAL: Tamrang Khola, 2,300 m., on branch of tree, 21 Nov. 1961, Norkett 7889.
Distribution: Worldwide,

FUNGI OF RECENT NEPAL EXPEDITIONS 119

TREMELLACEAE

GUEPINIA HELVELLOIDES (Fries) Fries, Elenchus Fung. 2 : 31 (1828).

Tremella helvelloides Fries, Syst. Mycol. 2 : 211 (1822).

Phlogiotis helvelloides (Fries) Martin in Amer. Journ. Bot. 23 : 628 (1936).

Tremella rufa Pers., Mycol Eur. i (i) : 103 (1822).

Gyrocephalus rufus (Pers.) Bref. in Unters. Gesammtgeb. Mykol. 7 : 131 (1888).

NEPAL: Taglung, Kali Gandaki, 3,500 m., 22 Sept. 1954, Stainton, Sykes & Williams

7990.
Distribution : North America, Europe, China, India.

TREMELLA MESENTERICA Fries, Syst. Mycol. 2 : 214 (1822).

NEPAL: Dhankuta Province, near Mahe, 1,300 m., on dead tree, 20 Sept. 1961,
Norkett 5175 D.

Distribution : Worldwide.

EXOBASIDIACEAE
EXOBASIDIUM Sp.

NEPAL: Lete, Kali Gandaki Valley, 3,800 m., on Rhododendron campanulatum,
4 June 1954, Stainton, Sykes & Williams 5607. Above Sauwala Khola, 3,800 m.,
on Rhododendron lepidotum, 15 Sept. 1954, Stainton, Sykes & Williams 4430. Near
Lumsum, 2,300 m., on Rhododendron seedlings, 24 Oct. 1954, Stainton, Sykes 6-
Williams 9132.

Specific identification could not be made as the collections in all cases were very
over-ripe. Several species of Exobasidium have been described on Rhododendron.
References to the literature on Exobasidium can be found in Sundstrom (Phytopath.
Zeitschr. 40 : 213-17 (1960)) and in McNabb (Trans. R. Soc. N.Z., Bot. i : 267
(1962)).

AGARICACEAE

ARMILLARIA MELLEA (Fries) Kummer, Fiihr. Pilzk. : 134 (1871).
Agancus melleus Fries, Syst. Mycol i : 30 (1821).

NEPAL: Arun Valley, Kasuwa Khola, on tree trunk in forest, n Sept. 1956,
Stainton 1618.

Distribution : Worldwide.

CLITOCYBE TABESCENS (Fries) Bresad., Fung. Trident. 2 : 84, t. 197 (1900).
Agaricus tabescens Fries, Hymenomyc. Eur., Ed. 2:111 (1874).

NEPAL: Lete, Kali Gandaki Valley, 2,600 m., in leaf mould at base oiPinus chylla,
3 June 1954, Stainton, Sykes 6- Williams 555-Z".
Distribution : Worldwide.

120 FUNGI OF RECENT NEPAL EXPEDITIONS

COPRINUS COMATUS (Fries) Gray, Nat. Arrang. Brit. PL i : 633 (1821).
Agaricus comatus Fries, Syst. Mycol. i : 307 (1821).

NEPAL: Chimgaon (north of Tukucha), Kali Gandaki, 4,500 m., 17 July 1954,
Stainton, Sykes < Williams 184.6.
Distribution : Worldwide.

COPRINUS DISSEMINATUS (Fries) Gray, Nat. Arrang. Brit. PI. i : 634 (1821).

Agaricus disseminatus Fries, Syst. Mycol. i : 305 (1821).

Psathyrella disseminata (Fries) Qu61. in Mem. Soc. fimul. Montbeliard, Ser. 2, 5 : 153
(1872) (reimpr. quam Champ. Jura Vosg. : 123 (1872)).

NEPAL: Arun Valley, Hatiar, 2,300 m., on fallen tree in forest, 20 Aug. 1956,
Stainton 1394.

Distribution : Worldwide.

CREPIDOTUS MOLLIS (Fries) Staude in Festg. Mitgl. XIX Versamml. deutsch. Land-
und Forstwirthe Coburg : 71 (1857) (reimpr. quam Schwamme Mitteldeutschl. :

71 (1858)).

Agaricus mollis Fries, Syst. Mycol. i : 274 (1821).

NEPAL: Arun Valley, Kasuwa Khola, 2,800 m., on tree trunk in forest, n Sept.
1956, Stainton 1617.

Distribution: America, Europe, China, Japan, Australia.

Very badly crushed in pressing but the layers of parallel and gelatinous hyphae
were readily observed; spores 9 X 5 fi, smooth.

GOMPHUS FLOCCOSUS (Schwein.) Sing, in Lloydia 8 : 140 (1945).

Cantharellus floccosus Schwein. in Trans. Amer. Phil. Soc., New Ser. 4 : 153 (1832).

NEPAL: Above Sauwala Khola, 3,300 m., on earth bank in Quercus forest, 13
Sept. 1954, Stainton, Sykes & Williams 4375.

Distribution: Recorded from North America, China, Japan, as well as from Nepal
(1955).

LACCARIA LACCATA (Fries) Cooke in Grevillea 12 : 70 (1884).

Agaricus laccatus Fries, Syst. Mycol. i : 106 (1821).

NEPAL: Arun Valley, Barun Khola, 4,000 m., in short grass, pinkish brown,
15 Sept. 1956, Stainton 1662.
Distribution : Worldwide.

LACTARIUS PUBESCENS (Krombh.) Fries, Epicrisis Syst. Mycol. : 335 (1838).

Agaricus pubescens Krombh., Naturg. Abbild. & Beschreib. Essb., Schadl. & Verdacht
Schwamme 2 : 24, t. 13 figs. 1-14 (1832).

FUNGI OF RECENT NEPAL EXPEDITIONS 121

NEPAL: Arun Valley, Barun Khola, 4,000 m., in short grass, 15 Sept. 1956, Stainton
1660.

Distribution: Apparently worldwide, but it is uncertain how many records under
the name of the coarser L. torminosus (Fries) Gray, from which many mycologists
have not separated this species, refer to it. Very few species of Lactarius have as
yet been recorded from India or any of the neighbouring countries.

LEPIOTA ERMINEA (Fries) Gill, Hymenomycetes : 59 (1874).
Agaricus ermineus Fries, Syst. Mycol. i : 22 (1821).

NEPAL: Mathand, near Pokhara, 1,120 m., on shady bank, " white except top of
cap which is brown " ', 22 June 1954, Stainton, Sykes & Williams 5852.
Distribution: Europe, India, Australia.

MARASMIUS CRINIS-EQUI F. von Muell. ex Kalchbr. in Grevillea 8 : 153 (1880).

Marasmius equicrinis F. von Muell. ex Berk, in Journ. Linn. Soc. Lond., Bot. 18 : 383

(1881), nom. superfl.

NEPAL: Murigurja Gad, 2,500 m., on dead vegetation near ravine track, 27 July
1954, Stainton, Sykes & Williams 3647.

Distribution: America, India, Ceylon, Philippines, Australia.

Panus polychrous (Lev.) Singer ex Balfour-Browne, comb. nov.

Lentinus polychrous Lev. in Ann. Sci. Nat., Set. 3, Bot. 2 : 175 (1844).

Lentinus vellereus Berk. & Curt, in Journ. Linn. Soc. Lond., Bot. 10 : 301 (1868).

Lentinus kurzianus Currey in Trans. Linn. Soc. Lond., Ser. 2, Bot. i : 120, t. 20 fig. u

(1876).
Panus polychrous Sing., Agaricales in Modern Taxonomy, Ed. 2 : 172 (1962), nom. invalid.

NEPAL: Tumlingtar, Sabhaya Khola, 600 m., on dead tree, 20 Dec. 1961, Norkett
8108.

Distribution: Cuba, India, Nepal, Ceylon, Philippines, Australia.
A few additional synonyms are given by Singer (loc. cit.).

PANUS TIGRINUS (Fries) Sing, in Lilloa 22 : 275 (1951).

Agaricus tigrinus Fries, Syst. Mycol. i : 176 (1821).

Lentinus tigrinus (Fries) Fries, Epicrisis Syst. Mycol. : 389 (1838).

NEPAL: Midam Khola, Chisankhu, 650 m., on dead tree trunk, 4 May 1954,
Stainton, Sykes & Williams 5214.
Distribution : Worldwide.

PHOLIOTA SQUARROSA (Fries) Kummer, Fiihr. Pilzk. : 84 (1871).

Agaricus squarrosus Fries, Syst. Mycol. i : 243 (1821).

NEPAL: Chimgaon (north of Tukucha), Kali Gandaki, 3,500 m., in forest, at base
of conifer, 14 Sept. 1954, Stainton, Sykes < Williams
Distribution : North America, Europe, Japan.

122 FUNGI OF RECENT NEPAL EXPEDITIONS

SCHIZOPHYLLUM COMMUNE Fries, Syst. Mycol i : 330 (1821).

NEPAL: Ranipauwa (north of Beni), Kali Gandaki, 1,000 m., on tree, 12 Sept.
1954, Stainton, Sykes & Williams 7818. Tumlingtar, Sabhaya River, Chainpur
district, 600 m., 12 Dec. 1961, Norkett 8480. Hinwan Khola, Chainpur, 660 m.,
on dead stick, 21 Dec. 1961, Norkett 9024.

Distribution: Worldwide.

HYDNACEAE

HERICIUM ERINACEUS (Fries) Pers., Mycol. Eur. 2 : 153 (1825).
Hydnum erinaceus Fries, Syst. Mycol. i : 407 (1821).

NEPAL: Above Sauwala Khola, 3,300 m., on Quercus in thick forest, 15 Sept.
1954, Stainton, Sykes & Williams 4415. Chimgaon, Kali Gandaki, 3,300 m., 14
Sept. 1954, Stainton, Sykes 6- Williams 7821.

Distribution: America, Europe, India, Japan, China.

HYDNELLUM ZONATUM forma VESPERTILIO (Berk.) Coker & Beers, Stipitate Hydnums
of Eastern U.S. : 80 (1951).

Hydum vespertilio Berk, in Hook., Journ. Bot. 6 : 167 (1854).

NEPAL: Taglung, Kali Gandaki, 3,500 m., 22 Sept. 1954, Stainton, Sykes 6- Williams
7997.

Distribution: America, Europe, India.

POLYPORACEAE

AMAURODERMA RUGOSUM (Blume & Nees) Torrend in Broteria, Ser. Bot. 18 : 127
(1920).

Polyporus rugosus Blume & Nees in Nov. Act. Phys.-Med. Acad. Caes. Leop.-Car. 13 : 21,
t. 7 (1826).

NEPAL: Dhankuta Province, below Sanghu, on roots of bamboo, 1,800 m., 27
Feb. 1962, Norkett 10233.

Distribution: Mostly tropical; Madagascar, India, Ceylon, Java, Philippines.

GANODERMA APPLANATUM (Pers.) Patouill. in Bull. Soc. Mycol Fr. 5 : 67 (1889).
Humphr. & Lewis in Philipp. Journ. Sci. 45 : 514 (1931).

Polyporus fomentarius var. applanatus Pers., Mycol. Eur. 2 : 80 (1825).
Polyporus applanatus (Pers.) Wallr., Fl. Crypt. Germ. 2 : 591 (1833).

NEPAL: Bakhri Kharka, north of Pokhara, 1,800 m., on rotten tree trunk, 24
April 1954, Stainton, Sykes & Williams 5060. Taglung, Kali Gandaki, 3,300 m.,
19 Oct. 1954, Stainton, Sykes < Williams 8198. Dharan Bazar, Terai forest, south
of Gopa Gurkha Camp, 250 m., on old tree, 27 Feb. 1962, Norkett 10234.

Distribution : Worldwide.

FUNGI OF RECENT NEPAL EXPEDITIONS 123

GANODERMA LUCIDUM (Fries) Karst. in Rev. Mycol. 3 (9) : 17 (1881).
Polyporus lucidus Fries, Syst. Mycol. i : 353 (1821).

NEPAL: Bakhri Kharka, north of Pokhara, 2,000 m., on rotten tree trunk, 25
April 1954, Stainton, Sykes & Williams 5075.
Distribution : Worldwide

FOMES PECTINATUS (Klotzsch) Gill., Hymenomycetes : 686 (1874).
Polyporus pectinatus Klotzsch in Linnaea 8 : 485 (1833).

NEPAL: Between Bakhri Kharka and Rambrong, 2,300 m., on rotten tree trunk,
26 Apr. 1954, Stainton, Sykes & Williams 5081.

Distribution: America, Europe, India, Australia, Philippines.

FOMES MARGINATUS (Fries) Gill., Hymenomycetes : 683 (1874).
Polyporus marginatus Fries, Syst. Mycol. i : 372 (1821).

NEPAL: Taglung, Kali Gandaki, 3,300 m., on forest tree, n July 1954, Stainton,
Sykes & Williams 1751. Also at 3,500 m., 22 Sept. 1954, Stainton, Sykes & Williams

7992.

Distribution: America, Europe, India, Nepal, China, Japan.

POLYPORUS ARCULARIUS Fries, Syst. Mycol. i : 342 (1821).

Polyporellus arcularius (Fries) Pilat in Kav. & Pilat, Atlas Champ. 3 : 75, t. 30-31, fig.
18 (1936).

var. ARCULARIUS.

NEPAL: Midam Khola, Chisankhu, 660 m., 4 May 1954, Stainton, Sykes 6- Williams
52 jo. Kabre, Kali Gandaki, 2,000 m., 13 June 1954, Stainton, Sykes & Williams
5742. Dhankuta Province, Chainpur district, Tumlingtar, 600 m., on dead trunk,
13 Dec. 1961, Norkett 8109 B; 8815 A ; and on 14 Dec. 1961, Norkett 8ioj B.

Distribution : Worldwide ; previously recorded from Nepal.

var. STRIGOSUS Bourd. & Galz., Hymenomyc. Fr. : 532 (1928).

NEPAL: Chipli, North of Pokhara, 2,600 m., on rotten tree trunk, 18 Apr. 1954,
Stainton, Sykes & Williams 4.882.
Distribution : Worldwide.
Distinguished by its marginal hairs.

POLYPORUS PARGAMENUS Fries, Epicrisis Syst. Mycol. : 480 (1838). Overh., Polyp.

U.S., Alaska & Canada : 336 (1953).

NEPAL: Sanghu, Milke Danda Forest, 2,900 m., on old dead tree, 16 Nov. 1961,
Norkett 7129.

124 FUNGI OF RECENT NEPAL EXPEDITIONS

Distribution: Widespread in temperate and tropical regions.

The material is in good condition but not sporing. There is much confusion over
the use of this name, P. biformis Klotsch and P. cervinus Fries. It is hoped to make
a more critical study of the problem shortly. Meanwhile Overholt's interpretation
of P. pargamenus is adopted.

POLYPORUS CORRUGATUS Pers. apud Gaud, in Freyc., Voy. aut. Monde Uranie &
Physicienne, Bot. : 172 (1826).

Earliella corrugata (Pers.) Murrill in Bull. Torrey Bot. Cl. 34 : 468 (1907).
Polystictus persoonii Cooke in Grevillea 14 : 85 (1886).
Daedalea sanguinea Klotsch in Linnaea 8 : 481 (1833).

NEPAL: Chainpur district, Tumlingtar, Dhankuta Province, 600 m., on dead tree,
13 Dec. 1961, Norkett 8109 E.

Distribution: West Indies, India, Nepal, East Indies and throughout most of the
tropics.

POLYPORUS PICIPES Fries, Epicrisis Syst. Mycol. : 440 (1838). Overh., Polyp. U.S.,
Alaska & Canada : 262 (1953).

Polyporellus picipes (Fries) Karst. in Bidr. Kann. Finl. Natur. & Folk 37 : 31 (1882).
Pilat in Kav. & Pilat, Atlas Champ. 3 : 99, t. 44 fig. 1-3, t. 46 fig. b, fig. 24, p. 105
fig. B (1937)-

NEPAL: Ghar Khola, 3,100 m., on dead trunk, 15 June 1954, Stainton, Sykcs &
Williams 5769. Arun Valley, Kasuwa Khola, 2,800 m., n Sept. 1956, Stainton
1624.

Distribution: Worldwide.

POLYPORUS SQUAMOSUS Fries, Syst. Mycol. i : 343 (1821).

NEPAL: Rambrong ridge, north of Pokhara, 3,300 m., on rotten tree trunk, 27
Apr. 1954, Stainton, Sykes < Williams 5103.
Distribution: Worldwide.

POLYPORUS SULPHUREUS Fries, Syst. Mycol. i : 357 (1821).

Grifola sulphured (Fries) Pilat in Beih. Bot. Centralbl. 52 (B) : 39 (1934).
Laetiporus sulphureus (Fries) Bondartz. & Sing, in Ann. Mycol. 39 : 51 (1941).

NEPAL: Chimgaon, Kali Gandaki, 3,500 m., 14 Sept. 1954, Stainton, Sykes &
Williams 7828. Arun Valley, Kasuwa Khola, 2,800 m., on tree in forest, n Sept.
1956, Stainton 1621.

Distribution : Worldwide ; previously recorded from Nepal.

POLYPORUS ZONALIS Berk, in Ann. & Mag. Nat. Hist. 10 : 375, t. 10 fig. 5 (1843).

NEPAL: Karelung, Madi Khola, 660 m., on rotten branch, 23 June 1954, Stainton,

Sykes & Williams 59JJ.

FUNGI OF RECENT NEPAL EXPEDITIONS 125

Distribution: tropical and semi-tropical; Central and South America, Cuba,
India, Indonesia, China, Australia.
Resembles the type but is a little thicker. Spores globose and no cystidia.

POLYSTICTUS AFFINIS (Blume & Nees) Fries in Nov. Act. Soc. Sci. Upsal., ser. 3,
i : 75 (1851).

Polyporus affinis Blume & Nees in Nov. Act. Phys.-Med. Acad. Caes. Leop.-Car. 13 : 18,

t. 4 (1826).
Microporus affinis (Nees) Kuntze, Revis. Gen. PI. 3 (2) : 495 (1898).

NEPAL: Arun Valley, Sashaya Khola, 660 m., on tree trunk in forest, 4 Sept.
1956, Stainton 1578. Dharan Bazar, Terai Forest, south of Gopa Gurkha Camp,
250 m., 27 Feb. 1962, Norkett 10224.

Distribution : widespread in tropical and subtropical regions ; previously recorded
from Nepal.

POLYSTICTUS CINNAMOMEUS (Gray) Sacc., Syll. Fung. 6 : 210 (1888).

Strilia cinnamomea Gray, Nat. Arrang. Brit. PI. i : 645 (1821).

Polyporus cinnamomeus (Gray) Fries, Epicrisis Syst. Mycol. : 468 (1838). Overh., Polyp.

U.S., Alaska & Canada : 386 (1953).
Coltricia cinnamomea (Gray) Murrill in Bull. Torrey Bot. Cl. 31: 343 (1904).

NEPAL: Taglung, Kali Gandaki, 3,500 m., in wood, 22 Sept. 1954, Stainton, Sykes
& Williams 799 1.

Distribution : Worldwide ; previously recorded from Nepal.
Very close to P. perennis but distinguished by its more uniform and silkier cap.

POLYSTICTUS HIRSUTUS (Fries) Fries in Nov. Act. Soc. Sci. Upsal., ser. 3, i : 86
(1851)-

Polyporus hirsutus Fries, Syst. Mycol. i : 367 (1821).

Coriolus hirsutus (Fries) Quel., Enchir. Fung. : 175 (1886). Bourd. & Galz., Hymenomyc.
Fr. : 561 (1928).

NEPAL: Dhankuta Province, Milke Danda Forest, 260 m., 29 Nov. 1961, Norkett
8307 A.

Distribution : Worldwide.

POLYSTICTUS PERULA (Fries) Fries in Nov. Act. Soc. Sci. Upsal., ser. 3, i : 73 (1851).

Polyporus perula Fries, Syst. Mycol. i : 349 (1821).

Polyporus xanthopus Fries, Syst. Mycol. i : 350 (1821).

Polystictus xanthopus (Fries) Fries in Nov. Act. Soc. Sci. Upsal., ser. 3, i : 74 (1851).

Microporus perula (Fries) Hariot in Bull. Soc. Mycol. Fr. 7 : 206 (1891).

NEPAL: Rupakot Tal, 800 m., on rotten branch, 5 May 1954, Stainton, Sykes &
Williams 5233. Arun Valley, Hinwan Khola, 800 m., on rotten log, 4 Sept. 1956,

126 FUNGI OF RECENT NEPAL EXPEDITIONS

Stainton 1533. Sanghu, 3,000 m., on tree stump, 2 Oct. 1961, Norkett 5540. Chain-
pur district, Tumlingtar, 600 m., 13 Dec. 1961, Norkett 8109 C.

Distribution: Widespread in tropical and sub-tropical areas; previously recorded
from Nepal.

POLYSTICTUS SANGUINEUS (Fries) Fries in Nov. Act. Soc. Sci. Upsal., ser. 3, i : 75

(1851)-

Polyporus sanguineus Fries, Syst. Mycol. i : 371 (1821).

Pycnoporus sanguineus (Fries) Murrill in Bull. Torrey Bot. Cl. 31 : 421 (1904).

NEPAL: Midam Khola, Chisankhu, 660 m., 4 May 1954, Stainton, Sykes & Williams
5218. Midam Khola, Karelung, 600 m., 23 June 1954, Stainton, Sykes & Williams
59x0. Kusma, 660 m., 2 Nov. 1954, Stainton, Sykes & Williams 9270. Sanghu,
3,000 m., 6 Oct. 1961, Norkett 5709.

Distribution : Mostly tropical and sub-tropical.

The Nepal material is thin and smooth and conforms with P. sanguineus. P.
cinnabarinus Fries, at one time considered to be a synonym, has been shown to be
distinct on the basis of cultural interfertility tests (McKay in Mycologia, 51 : 465-

73 (I959))'

POLYSTICTUS SUBAFFINIS Lloyd, Mycol. Not. 40 : 550, fig. 755 (1916).

NEPAL: Surauti Khola, 660 m., on dead bamboo, 12 Aug. 1954, Stainton, Sykes <
Williams 6869.

Distribution: Japan, Java, Madagascar.

The present collection appears to agree completely with Lloyd's species as he
figured it from Umemura's Japanese specimen, but it is doubtful whether this species
is distinct from Polystictus affinis (Fries) Fries.

POLYSTICTUS TABACINUS (Mont.) Sacc., Syll. Fung 6 : 280 (1888).

Polyporus tabacinus Mont, in Ann. Sci. Nat., Ser. 2, Bot. 3 : 349 (1835).

NEPAL: Siklis, north of Pokhara, 2,100 m., on rotten tree trunk, 22 Apr. 1954,
Stainton, Sykes < Williams 4974. Gurjagaon, 3,000 m., on dead tree, 25 Sept.
1961, Norkett 5481. Milke Danda Forest, 3,000 m., on old dead tree, 16 Nov. 1961,
Norkett 7127.

Distribution: South America, Africa, India, East Indies, China, Australasia.

9-10 pores per mm., setae subulate, dark brown. This species differs from
P. iodinus in having smaller pores.

POLYSTICTUS TEPHROLEUCUS (Berk.) Sacc., Syll. Fung. 6 : 275 (i

Trametes tephroleuca Berk, in Hook., Journ. Bot. 6 : 165 (1854).

Coriolus tephroleucus (Berk.) Bondartz., Trutov. Ghrib. Evr. Chasti S.S.S.R. & Kavk. :

492, fig. 126 (I953)-

Coriolus favoliporus Pilat in Bull. Soc. Mycol. Fr. 52 : 313, t. 3 figs. 3-4 (1937).
Trametes favolipora (Pilat) Pilat in Kav. & Pilat, Atlas Champ. 3 : 267, t. 182, fig. 105

(1939).

FUNGI OF RECENT NEPAL EXPEDITIONS 127

NEPAL: Lulo Khola, 16 Sept. 1952, Polunin, Sykes ( Williams 3448. Near
Lumsum, 2,300 m., on dead tree stump, 24 Oct. 1954, Stainton, Sykes & Williams

Distribution: Asia: Kazakstan, India and East Nepal.
Fine specimens but the pores in Stainton, Sykes & Williams 9137 are mostly
discoloured owing to a mycelial growth over the hymenium.

POLYSTICTUS VERSATILIS (Berk.) Fries in Nov. Act. Soc. Sci. Upsal., Ser. 3, i : 92
(1851).

Trametes versatilis Berk, in Lond. Journ. Bot. i : 150 (1842).

Polyporus versatilis (Berk.) Romell in Bih. K. Svensk. Vet. -Akad. Handl. 26, (3, 16) : 35

(1901). Lloyd, Mycol. Not. 50 : 703, figs. 1049-50 (1917). Overh., Polyp. U.S.,

Alaska & Canada: 325 (1953)-

NEPAL: Ghar Khola, 2,600 m., 3 May 1954, Stainton, Sykes & Williams 5444.
Maikot, 2,600 m., on stump, 4 July 1954, Stainton, Sykes & Williams 3363. Tum-
lingtar, Chainpur, 600 m., on old tree in ravine, 9 Dec. 1961, Norkett 8693. Dharan
Bazar, Terai forest, south of Gopa Gurkha Camp, 230 m., 27 Feb. 1962, Norkett
10232.

Distribution: America, Madagascar, India, Malaya, Indonesia, Japan, China.

POLYSTICTUS VERSICOLOR (Fries) Fries in Nov. Act. Soc. Sci. Upsal., Ser. 3, i : 86
(1851).

Polyporus versicolor Fries, Syst. Mycol. i : 368 (1821).
Coriolus versicolor (Fries) Quel., Enchir. Fung. : 175 (1886).

NEPAL: Ranipauwa, Kali Gandaki, 1,000 m., 3 Sept. 1954, Stainton, Sykes &
Williams 7637. Lete, Kali Gandaki, 2,800 m., 17 Sept. 1954, Stainton, Sykes &
Williams 7893. Bakhri Kharka, Pokhara, 2,000 m., 25 Apr. 1954, Stainton, Sykes
6- Williams 5075. Arun Valley, Hatiar, 2,600 m., 21 Aug. 1956, Stainton 1412.
Taplejung, above Sanghu, Milke Danda Forest, 3,000 m., 16 Nov. 1961, Norkett
7124 A. Mewa Khola, 2,700 m., 23 Jan. 1962, Norkett 9345 and 9345 A.

Distribution: Worldwide; previously reported from Nepal.

DAEDALEA UNICOLOR Fries, Syst. Mycol. i : 336 (1821). Overh., Polyp. U.S.,
Alaska & Canada : 125 (1953).

Coriolus unicolor (Fries) Patouill., Ess. Tax. Fam. & Genr. Hym6nomyc. : 94 (1900).

NEPAL: Sanghu, 3,000 m., i Nov. 1961, Norkett 6728.
Distribution : America, Europe, North Africa, China, Australia.

TRAMETES CERVINA (Schwein.) Bresad. in Ann. Mycol. i : 81 (1903).

Boletus cervinus Schwein. in Schrift. Naturf. Ges. Leipz. i : 96 (1822).

Coriolus cervinus (Schwein.) Bondartz., Trutov. Ghrib. Evr. Chasti S.S.S.R. & Kavk. :

493, fig. 127 (1953)-
Polyporus biformis sensu Berk, in Ann. & Mag. Nat. Hist. 3 : 392 (1839), non Klotzsch.

128 FUNGI OF RECENT NEPAL EXPEDITIONS

NEPAL: Ulleri, north of Kusma, Kali Gandaki, 2,600 m., on tree in wood, i Nov.
1954, Stainton, Sykes 6- Williams 8275.

Distribution: Europe, Russia, India, Ceylon, China, Australia.

For nomenclature of this fungus see note under Polypoms pargamenus (p. 123).

TRAMETES GIBBOSA (Fries) Fries, Epicrisis Syst. Mycol. : 492 (1838).

Daedalea gibbosa Fries, Syst. Mycol. i : 338 (1821).

Pseudotrametes gibbosa (Fries) Bondartz. & Sing, in Ann. Mycol. 39 : 60 (1941).

NEPAL: Siklis, north of Pokhara, 21 Apr. 1954, Stainton, Sykes 6- Williams 4957.
Distribution: Europe, Africa, India, China.

LENZITES BETULINA (Fries) Fries, Epicrisis Syst. Mycol. : 405 (1838).

Daedalea betulina Fries, Syst. Mycol. i : 333 (1821).

Trametes betulina (Fries) Pilat in Kav. & Pilat, Atlas Champ. Fur. 3 : 327, t. 220, fig. 142
(1940).

NEPAL: Arun Valley, Hatiar, 2,600 m., on tree trunk, 21 Aug. 1956, Stainton 1412.
Siklis, north of Pokhara, 2,500 m., on rotten tree trunk, 19 Apr. 1954, Stainton,
Sykes & Williams 4926.

Distribution: Worldwide.

LENZITES PALISOTII (Fries) Fries, Epicrisis Syst. Mycol. : 404 (1838).

Daedalea palisotii Fries, Syst. Mycol. i : 335 (1821) " Palisoti ".

Daedalea applanata Klotzsch in Linnaea 8 : 481 (1833).

Lenzites repanda Fries, Epicrisis Syst. Mycol. : 404 (1838).

Lenzites applanata (Klotzsch) Fries, Epicrisis Syst. Mycol. : 404 (1838).

NEPAL: Arun Valley, Sabhaya Khola, 800 m., on tree trunk, 3 Sept. 1956, Stainton
15^6. Dhankuta Province, Chainpur district, Tumlingtar, 600 m., on dead branch,
13 Dec. 1961, Norkett 8109 A.

Distribution: Widespread, especially in the southern hemisphere.

LENZITES SUBFERRUGINEA Berk, in Hook., Journ. Bot. 6 : 134 (1854).

Gloeophyllum subferrugineum (Berk.) Bondartz., Trutov. Ghrib. Evr. Chasti S.S.S.R. &
Kavk. : 50 (1953)-

NEPAL: Lete, Kali Gandaki Valley, 2,600 m., on rotten tree trunk, 12 June 1954,
Stainton, Sykes & Williams 5730; same locality, 8 July 1954, Stainton, Sykes &
Williams 1639. Taglung, Kali Gandaki, 27 Aug. 1954, Stainton, Sykes < Williams

7494-

Distribution: India, Nepal, Philippines, Japan.

No spores were found in any of the gatherings. Stainton, Sykes & Williams 5730
has a grey cap, 1639 has grey cap with wide brown margin, and in 7494 the cap is
entirely brown.

FUNGI OF RECENT NEPAL EXPEDITIONS 129

THELEPHORACEAE

CORTICIUM CAERULEUM (Pers.) Fries, Epicrisis Syst. Mycol. : 562 (1838).

Thelephora caerulea Pers., Mycol. Eur. i : 147 (1822). Fries, Elench. Fung, i : 202

(1828).

NEPAL: Ganesh Himal, Ankhu Khola, 2,800 m., on bark in broad-leaved forest,
17 May 1962, Stainton 3731.

Distribution: America, Europe, India, Australia, Japan.

HYMENOCHAETE MOUGEOTII (Fries) Cooke in Grevillea 8 : 147 (1880).

Thelephora mougeotii Fries, Elench. Fung, i : 188 (1828).

NEPAL: Annapurna Himal, on branches of Rhododendron campanulatum, 30 Aug.
1954, Stainton, Sykes & Williams 6641. Rambrong, Lamjung Himal, on branch of
Betula utilis, 7 July 1954, Stainton, Sykes & Williams 6202. Mewa Khola, 1,300 m.,
on dead wood, i Feb. 1962, Norkett 9182. Ganesh Himal, Ankhu Khola, 2,800 m.,
on bark in broad-leaved forest, 17 May 1962, Stainton 3732. Ganesh Himal,
Mailung Khola, 4,000 m., 20 May 1962, Stainton 3744.

Distribution: Europe, India, Nepal, Australia, New Zealand, China.

HYMENOCHAETE RHEICOLOR (Mont.) Lev. in Ann. Sci. Nat., Ser. 3, Bot. 5 : 151
(1846).

Stereum rheicolor Mont, in Ann. Sci. Nat., Ser. 2, Bot. 18 : 23 (1842).
Stereum tenuissimum Berk, in Hook., Lond. Journ. Bot. 6 : 510 (1847).
Hymenochaete sallei Berk. & Curt, in Journ. Linn. Soc. Lond., Bot. 10 : 333 (1868).
Hymenochaete tenuissima Berk, apud Berk. & Curt, in Journ. Linn. Soc. Lond., Bot. 10 :

333 (1868) nom. nud.
Hymenochaete tenuissima (Berk.) Berk. & Broome in Journ. Linn. Soc. Lond., Bot. 14 :

67 (1873).
Stereum elegantissimum Spegazz. in An. Soc. Cient. Argent. 17 : 78 (1884).

NEPAL: Dhankuta Province, Taplejung district, Sanghu, 2,000 m., 17 Oct. 1961,
Norkett 5696 B; same locality, 15 Nov. 1961, Norkett 7112 C.

Distribution : North and South America, Africa, West Indies, India.

HYMENOCHAETE RUBIGINOSA (Fries) Lev. in Ann. Sci. Nat., Ser. 3, Bot. 5 : 151

(1846).

Thelephora rubiginosa Fries, Syst. Mycol. I : 436 (1821).

NEPAL: Chainpur district, Tumlingtar, 600 m., on dead tree, 13 Dec. 1961, Norkett
8109 D; same locality, 16 Dec. 1961, Norkett #567.
Distribution : Worldwide.

130 FUNGI OF RECENT NEPAL EXPEDITIONS

HYMENOCHAETE TABACINA (Fries) Lev. in Ann. Sci. Nat., Ser. 3, Bot. 5 : 152 (1846).

Thelephora tabacina Fries, Syst. Mycol. i : 437 (1821).

Stereum tabacinum (Fries) Fries, Epicrisis Syst. Mycol. : 550 (1838).

NEPAL: Annapurna Himal, on branches of Rhododendron campanulatum, 3 Aug.
1954, Stainton, Sykes & Williams 6642. Near Dogadi Khola, 4,300 m., on dead
shrubs, ii Aug. 1954, Stainton, Sykes & Williams 3818.

Distribution: Worldwide.

LOPHARIA CRASSA (Lev.) Boidin in Bull. Soc. Mycol. Fr. 74 : 479 (1958).

Thelephora crassa Lev. in Ann. Sci. Nat., Ser. 3, Bot. 2 : 209 (1844).

Stereum umbrinum Berk. & Curt, apud Berk, in Grevillea i : 164 (1873).

Hymenochaete vinosa Cooke in Grevillea 8 : 149 (1880).

Laxitextum crassum (Lev.) Lentz in U.S. Dept. Agric., Agric. Monogr. 24 : 20 (1955).

NEPAL: Dhankuta Province, Dhankuta, near Mahe, 1,300 m., 20 Sept. 1961,
Norkett 5^57 C. Sombu, 1,600 m., 23 Sept. 1961, Norkett 5289. Dhankuta Pro-
vince, Taplejung district, Sanghu, 3,000 m., 3 Oct. 1961, Norkett 5639 A ; same
locality, 2,000 m., 15 Nov. 1961, Norkett 7112 C.

Distribution: America, Europe, Africa, India, Australia, New Zealand.

STEREUM OSTREA (Fries) Fries, Epicrisis Syst. Mycol. : 547 (1838).

Thelephora fasciata Schwein. in Schrift. Naturf. Ges. Leipz. i : 106 (1822).

Thelephora ostrea Fries, Elench. Fung, i : 175 (1828).

Thelephora versicolor var. fasciata (Schwein.) Fries, loc. cit.

Stereum fasciatum (Schwein.) Fries, Epicrisis Syst. Mycol. : 546 (1838).

NEPAL: Arun Valley, Khandbari, 2,300 m., on rotting log, 31 Aug. 1956, Stainton
1472.

Distribution: widespread; previously recorded from the Himalayas.

Since Fries did not treat Thelephora fasciata Schwein. as a separate species in his
Elenchus, which is part of the starting-point for the Fungi caeteri, the epithet of his
T. ostrea must be adopted when the two names are regarded as synonyms.

STEREUM HIRSUTUM (Fries) Gray, Nat. Arrang. Brit. PI. i : 653 (1821).
Thelephora hirsuta Fries, Syst. Mycol. i : 439 (1821).

NEPAL: Chipli, north of Pokhara, 3,600 m., 18 Apr. 1954, Stainton, Sykes
Williams 4885. Arun Valley, Hatiar, 2,600 m., on rotting log, 21 Aug. 1956,
Stainton 1410. Arun Valley, Sibrung, 27 Aug. 1956, Stainton 1447. Kasuwa Khola,
3,300 m., 12 Sept. 1956, Stainton 1637. Milke Danda Forest, 3,000 m., 16 Nov.
1961, Norkett 7125. Sanghu, 2,300 m., 23 Nov. 1961, Norkett 8196.

Distribution : Worldwide.

FUNGI OF RECENT NEPAL EXPEDITIONS 131

STEREUM ROSEO-CARNEUM (Schwein.) Fries, Acta Soc. Sci. Upsala, Ser. 3, i : 112
(1851).

Thelephora roseo-carnea Schwein. in Schrift. Naturf. Ges. Leipz. i : 107 (1822).
Laxitextum roseo-carne^im (Schwein.) Lentz in U.S. Dept. Agric., Agric. Monogr. 24 : 22
(1955)-

NEPAL: Dhankuta Province, Taplejung district, Sanghu 2,060 m., 15 Nov. 1961,
Norkett Jii2 B.

Distribution: North and South America, Japan, China.

The fructifications are pinkish buff, resupinate on twigs; the paraphyses have
branching tips; spores 8 X 4-5 fi.

STEREUM SANGUINOLENTUM (Fries) Fries, Epicrisis Syst. Mycol. : 549 (1838).
Thelephora sanguinolenta Fries, Syst. Mycol. i : 440 (1821).

NEPAL: Gurjakhani, 3,160 m. on small branches, 30 July 1954, Stainton, Sykes <
Williams 3678.
Distribution : North America, Europe, South Africa, Australia, New Zealand.

STEREUM SUBPILEATUM Berk. & Curt, apud Berk, in Hook., Journ. Bot. i : 238
(1849). Lentz in U.S. Dept. Agric., Agric. Monogr. 24 : 36 (1955).

Stereum insigne Bresad. in Nuovo Giorn. Bot. Ital. 23 : 158 (1891).

Xylobolus subpileatus (Berk. & Curt.) Boidin in Rev. de Mycol. 23 : 336 (1958).- Lentz
in Sydowia 14 : 118 (1960).

NEPAL: Bakhri Kharka, north of Pokhara, on rotten tree trunk, 24 Apr. 1954,
Stainton, Sykes & Williams 5059. Milke Danda Forest, Dhankuta Province, 2,800
m., 2 Nov. 1961, Norkett 6806 A ; same area, 3,000 m., 16 Nov. 1961, Norkett 7124.

Distribution: America, Europe, India, Indonesia, China, Japan.

Lentz and Boidin should be consulted for modern interpretations of this aggregate
species.

STEREUM sp. cf. SULCATUM Burt apud Peck in New York State Mus. Annu. Rep.
54, i, App. i : 154 (1901).

NEPAL: South of Gurjakhani, 3,600 m., on tree in Abies forest, 16 Aug. 1954,
Stainton, Sykes & Williams 3868.

Large robust sporophores superficially like those of a large S. princeps (Jungh.)
Lev. but having a tomentose sulcate pileus, coarse, more or less parallel skeletal
hyphae intermingled with generative hyphae, and no acanthophyses as such, but
merely some slightly granular cystidial hyphae. The spores, irregularly globose,
smooth or very faintly punctate and amyloid, 4-5-6 / diameter, resemble those of
S. sulcatum Burt and those of 5. taxodii Lentz & McKay (Mycologia 52 : 262 (1960)),
two species recently transferred by H. L. Gross to his genus Echinodontium (Myco-
path. & Mycol. Appl. 24 : 8, u (1964)). The Nepal fungus however differs in the

132 FUNGI OF RECENT NEPAL EXPEDITIONS

absence of large encrusted cystidia and in the possession of large flabelliform reflexed
pilei. In view of the large and conspicuous fructifications it would seem unlikely
that this fungus has not been recorded previously. I therefore defer describing it
as new to science.

THELEPHORA CARYOPHYLLAEA Fries, Syst. Mycol. i : 430 (1821).

Phylacteria caryophyllea Patouill., Hym. Eur. : 154 (1887), nom. nud.
Phylacteria caryophyllea (Fries) Patouill. ex Bourd. & Maire in Bull. Soc. Mycol. Fr. 36 :
76 (1920).

NEPAL: Taplejung district, above Sanghu, 2,000 m., amongst moss on earth,
12 Oct. 1961, Norkett 5927 A.
Distribution : Worldwide.

VARARIA RHODOSPORA (Wakef.) G. H. Cunn. in Proc. Linn. Soc. N.S.W. 77 : 291
(1953); in N.Z. Dept. Sci. Industr. Res. Bull. 145 : 100 (1963).

Stereum duriusculum sensu Bresad. in Ann. Mycol. 6 : 43 (1908), non Berk. & Broome.
Asterostromella rhodospora Wakef. in Kew Bull. 1915 : 372 (1915). Banergee in Journ.

Ind. Bot. Soc. 14 : 45 (1935).
Asterostromella dura Bourd. & Galz. apud Bourd. & Maire in Bull. Soc. Mycol. Fr. 36 :

74 (1920).
Dichostereum durum (Bourd. & Galz.) Pilat in Ann. Mycol. 24 : 223 (1926).

NEPAL: Sanghu, 820 m., 9 Nov. 1961, Norkett 7324.

Distribution: America, Europe, Africa, India, Japan, Australia, New Zealand.

The species was recorded from India by Banergee. The Nepal collection con-
sists of tough, resupinate thalli covering earth beneath tree roots, about 2-3 mm.
thick, mid-brown, of stratose context. Only a few basidia were observed embedded
in the dichophysoid paraphyses ; spores globose, straw-coloured, echinulate, 5-6 ju,
diameter; context hyphae brown, dendroidly and dichotomously branched, most
markedly and densely in the hymenial layer, where they form the brown dendro-
physes and dichophyses; intermingled with them are finer, readily stained hyphae.
Rogers & Jackson (Farlowia I : 309 (1943)), treat Dichostereum durum as a synonym
of Vararia pallescens (Schwein.) Rog. & Jacks. Type material of Thelephora pal-
lescens Schwein. in the B.M. Herbarium differs in several particulars, notably its
finer context, and would appear to be not merely a different growth-form, as these
authors suggested, but a distinct species.

CLAVARIACEAE

CLAVULINA MUSSOORIENSIS Corner, Thind & Dev in Trans. Brit. Mycol. Soc. 41
204 t. 8 fig. 3, text-fig, i (1958).

NEPAL: Near Gurjakhani, 2,800 m., among grass on open slope, 28 July 1954,
Stainton, Sykes & Williams 3670.
Distribution: India.

(E.J.H.C.)

FUNGI OF RECENT NEPAL EXPEDITIONS 133

Clavulina alta Corner, sp. nov.

Receptacula ad n cm. alta, alba, sicco luride flava; stipite 2-6 cm. x 3-8 mm.,
bene evoluto; ramulis inferioribus polychotomis v. applanato-rrmltifidis, superior-
ibus i mm. latis dichotomis v. cristatis, axillis inferioribus 3-6 mm. latis. Sporae
9-5-14 x 7-5-9 /*, subglobosae, lacrymiformes v. pyriformes, apiculo I /i longo.
Basidia 6-5-7-5 /JL lata, bispora. Hymenium incrassatum ; cystidiis nullis ; hyphis
subhymenialibus 6-17 /i latis, fere pseudoparenchymaticis. Hyphae 3-12 /^ latae,
fibulatae, tenue tunicatae, cellulis potius brevibus.

NEPAL: Chimgaon, north of Tukucha, Kali Gandaki, 3,500 m., on ground beneath
conifers, 14 Sept. 1954, Stainton, Sykes & Williams 7825. (Herb. Mus. Brit, holotype).
This resembles C. rugosa (Fries) Schroet. in the large spores and wide sub-
hymenial hyphae, and C. cristata var. coralloides Corner in the form of the fruit-body.
I have not seen such a distinct form before, and the spores are constantly rather
narrow for their length.

(E.J.H.C.)

Lentaria macrospora Corner, sp. nov.

Receptacula ad 10 cm. alta, gregaria v. caespitosa, carneoflavida ; stipite ad
25 X 2-4 mm., axillas inferiores polychotomas versus plus minus dilatato; ramulis
superioribus teretibus dichotomis strictis ascendentibus, 1-2 mm. latis. Sporae
20-30 X 3'7-5'5 /*, hyalinae, cylindricae, v. subclavatae, saepe curvatae v. sig-
moideae, et allantiformes, tenue tunicatae, haud amyloideae. Basidia ad 45 x
9-10-5 fji; sterigmatibus 2-4, 7-8 fJb longis. Hymenium incrassatum; cystidiis nullis.
Hyphae 2-5-7 /* latae, fibulatae, tunicis ad 0-5 {JL vix incrassatis; in mycelio 2-5-4 /*
latae, fibulatae, monomiticae, tunicis tenacibus sed vix incrassatis, passim partibus
ampulliformibus ad 15 /* latis inflatae, crystallis sphaeroideis 2-7 /* latis inter hyphas
numerosis.

NEPAL: Tamur Valley, Ghunsa, east of Walungchung Gola, 4,300 m., on ground
under conifers, 27 July 1956, Stainton 1145. (Herb. Mus. Brit, holotype).

This resembles in shape and colour the common tropical L. surculus (Berk.)
Corner but the spores are much longer, the hyphal walls are scarcely thickened, and
the habitat seems to be humicolous. Many basidia had 1-3 long spiculiform sterig-
mata, but they may have been abnormal and formed after collection.

(E.J.H.C.)

RAMARIA aff. BOTRYTOIDES (Peck) Corner, Monogr. Clavaria, Ann. of Bot., Mem.

i : 562 (1950).

Clavaria botrytoides Peck in N.Y. State Mus., Mus. Bull. 94 : 49, t. 93 figs. 5-7 (1905).

NEPAL: Taglung, south of Tukucha, Kali Gandaki, 3,300 m., on ground beneath
trees, n July 1954, Stainton, Sykes & Williams i6gi.

Distribution (of Ramaria botrytoides} : America, southern parts of Australia and
Tasmania, Japan ; when the species is interpreted in a wide sense.

(E.J.H.C.)

134 FUNGI OF RECENT NEPAL EXPEDITIONS

RAMARIA SUECICA (Fries) Donk in Med. Bot. Mus. Herb. Rijks Univ. Utrecht 9 :
105 (1933). Corner, Monogr. Clavaria, Ann. of Bot., Mem. i : 629 (1950).

Clavaria suecica Fries, Syst. Mycol. i : 469 (1821).

Clavariella suecica (Fries) Karst. in Bidr. Kann. Finl. Natur. & Folk 37 : 187 (1882).

NEPAL: South of Gurjakhani, 3,300 m., on damp shady forest bank, 18 Aug.
1954, Stainton, Sykes < Williams 3902.
Distribution: Europe, China, Canada, Northern U.S.A.

(E.J.H.C.)

RAMARIA FLACCIDA (Fries) Ricken, Vadem. Pilzfr. : 254 (1918).

Clavaria flaccida Fries, Syst. Mycol. i : 471 (1821).

NEPAL: Taglung, south of Tukucha, Kali Gandaki, 3,300 m., in pine wood, 22
Sept. 1954, Stainton, Sykes & Williams 7971.

Distribution: America, Europe, South Africa, Australia, China, Japan.

(E.J.H.C.)

RAMARIA OBTUSISSIMA (Peck) Corner, Monogr. Clavaria, Ann. of Bot., Mem. i : 609

Clavaria obtusissima Peck in N.Y. State Mus., Mus. Bull. 167 : 39 (1913).

NEPAL: Taglung, south of Tukucha, Kali Gandkai, 3,000 m., beneath conifers,
12 July 1954, Stainton, Sykes & Williams 1790.
Distribution: U.S.A. and Canada.
The pink form is represented.

(E.J.H.C.)

RAMARIA SUBAURANTIACA Corner apud Ball-Browne in Bull. Brit. Mus. (Nat. Hist.),

Bot. i : 200 (1955).

NEPAL: North of Barse, 4,000 m., on Abies stump, 14 Aug. 1954, Stainton, Sykes
6- Williams 3851.

Distribution: Tibet.

Spores 10-5-15 X 5-6 fi, rather coarsely subverrucose. No clamps.

(E.J.H.C.)

CYPHELLACEAE

Chromocyphella bryophyticola Balfour-Browne, sp. nov.

Fungus cupulatus, cupulis sessilis, levis, griseo-albis, 0-5 mm. diam. ; contextis
tenuis, mollis, 15 fi latis ex hyphis elongatis efformatis ; hymenio levo, brunneo ;
basidiis cylindricis, 15-16 X 4-5 fi', sporis globosis, 5-7 ^ diam., brunneis, punctatis.

NEPAL: Sanghu, 1,800 m., on moss, Pterobryopsis, and on an intermingled liver-
wort, on shady rock, 8 Nov. 1961, Norkett 7292. (Herb. Mus. Brit, holotype).

FUNGI OF RECENT NEPAL EXPEDITIONS 135

This fungus is not unlike Cyphella chromospora Patouill. (Tab. Anal. Fung, i : 19,
r. 32 (1883)), but the spores are larger. It differs from Chromocyphella burtii
ridge Cooke (Sydowia, Beiheft 4 : 137 (1961)), in its smaller basidia and its definitely
punctate spores.

This appears to be the first record of a " Cyphella " in the Himalayan region.

SCLERODERMATEACEAE

SCLERODERMA AURANTIUM Pers., Syn. Meth. Fung. : 153 (1801).

NEPAL: Mathand, near Pokhara, 1,160 m., on shady bank, 22 June 1954, Stainton,
Sykes & Williams 5851.
Distribution : Worldwide.
Spores 8-10 /*, reticulated.

LYCOPERDACEAE

BOVISTA sp. cf . BOVISTOIDES (Cooke & Massee) Ahmad, Gasteromycetes W. Pakistan,
Publ. Dept. Bot. Univ. Panjab n : 16 (1952).

Mycenastrum bovistoides Cooke & Massee in Grevillea 16 : 26 (1887).

NEPAL: Above Dogadi Khola, amongst short grass on exposed slope, 21 June
1954, Stainton, Sykes & Williams 3211.

Distribution (of Bovista bovistoides) : India.

The Nepal collection appears to be very close to this species but has very slightly
warted spores, whereas Ahmad described them as smooth. The capillitium threads
agree in being chestnut brown and unpitted.

BOVISTA ECHINELLA Patouill. in Bull. Soc. Mycol. Fr. 7 : 165 (1891).

Bovistella echinella (Patouill.) Lloyd, Mycol. Notes 23 : 286, t. 89 figs. 1-2 (1906).
Lycoperdon echinella (Patouill.) Ahmad in Journ. Ind. Bot. Soc. 20 : 138 (1941).

NEPAL: Chainpur district, Tumlingtar, 600 m., on earth near base of cliff of
Sabhaya River, 9 Dec. 1961, Norkett 8682 A.

Distribution : North and South America, Jamaica, Europe, Pakistan.

The spores are smooth, not echinulate as described by Patouillard, but show
" lines " beneath the outer membrane which at a certain focus appear like spines;
the pedicels are mostly 6 /i long.

GEASTRUM FIMBRIATUM (Fries) E. Fisch. in Engl. & Prantl, Nat. Pflanzenfam., 2
Aufl., 7a : 73 (1933).

Geaster fimbriatus Fries, Syst. Mycol. 3 : 16 (1829).

NEPAL: Tukucha, Kali Gandaki, 3,600 m., 12 Oct. 1954, Stainton, Sykes &
Williams 8212.
Distribution: America, Europe, Africa, India, Australia.

136 FUNGI OF RECENT NEPAL EXPEDITIONS

GEASTRUM HARIOTII (Lloyd) E. Fisch. in Engl. & Prantl, Nat. Pflanzenfam., 2 Aufl.,
7a : 73 (1933).

Geaster hariotii Lloyd, Mycol. Not. 25 : 311, t. 99 figs. 7-9 (1907).

NEPAL: Tumlingtar, Sabhaya Khola, 600 m., on earth beneath bamboo, 20 Dec.
1961, Norkett 9002.

Distribution: South America, West Indies, Europe, Ceylon, Australia (according
to Cunningham, v. below).

The eight specimens from Nepal resemble the descriptions of Geastrum hariotii
very closely in being non-hygroscopic, in having a sessile endoperidium, dark,
sulcate peristome and small, 3-3-5 /* diam., minutely verrucose spores. But the
surface of the endoperidium is furfuraceous or granular rather than pitted as des-
cribed by some authors (Cunningham, Gasteromycetes Austral, and N.Z. : 165
(1942)).

LYCOPERDON PYRIFORME Pers., Syn. Meth. Fung. : 148 (1801).

NEPAL: Tukucha, Kali Gandaki, 3,500 m., 12 Sept. 1954, Stainton, Sykes &
Williams 7806.

Distribution : Worldwide.

FUNGI IMPERFECTI

SPHAEROPSIDACEAE

CONIOTHYRINA AGAVES (Durieu & Mont.) Petr. & Syd. in Beih. Rep. Spec. Nov.
Regn. Veg. 42 : 322 (1927).

Phoma agaves Durieu & Mont, in Mont., Syll. PL Crypt.: 271 (1856).
Coniothyrium agaves (Durieu & Mont.) Sacc., Syll. Fung. 3 : 318 (1884).

NEPAL: Sanghu, 2,000 m., on Agave, 17 Nov. 1961, Norkett 7154', Sombu, 1,600
m., on Agave, 23 Sept. 1961, Norkett 5313.

Distribution: America, South Europe, Africa, India.

NECTRIOIDACEAE

ASCHERSONIA sp. cf. viRiDANS (Berk. & Curt.) Patouill. in Bull. Soc. Mycol. Fr
7 : 48 (1891).

Hypocrea viridans Berk. & Curt, in Journ. Linn. Soc. Lond., Bot. 10 : 376 (1868).
Aschersonia disciformis Patouill. in Bull. Soc. Mycol. Fr. 8 : 136 (1892).
Aschersonia viridula Sacc. in Ann. Mycol. n : 547 (1913).

NEPAL: Phewa Tal, 800 m., on leaves of Castanopsis sp., 8 May 1954, Stainton
Sykes < Williams 5272 (a).

Distribution (of A . viridans} : Central America (Trinidad, Vera Cruz, Cuba, Ecua
dor), Brazil, Mexico.

The Nepal fungus was growing on white fly. It differs from previous accounts in
the greater number of pycnidia to each stromatic cushion, which is completely dotted
over with the greenish ostioles; conidia 12-16 X 1-5-2/4.

FUNGI OF RECENT NEPAL EXPEDITIONS 137

Another Aschersonia also was collected on Castanopsis leaves; this appears to be
close to A. caespiticia Syd. (in Engl., Bot. Jahrb. 54 : 260 (1916)), but differs in
the rough surface of the tubercles and its small basal cushions. Tilhar, 3 Nov.
1954, Stainton, Sykes & Williams 9251 (a).

LEPTOSTROMATACEAE

MELASMIA SALICINA Tulasne frat., Sel. Fung. Carp. 3 : 119, t. 15 figs. 15-17 (1865).

NEPAL: East of Chalike Pahar, 4,160 m., on Salix sp., 25 Sept. 1954, Stainton,
Sykes < Williams 4587.

Distribution: not previously collected in this part of the world but its perfect
state, Rhytisma salicinum Fries, has been recorded from the Punjab.

MELANCONIACEAE

Mastigonetron americanum (Mont.) Balfour-Browne, comb. nov.

Pestalotia americana Mont, in Gay, Hist. Chile, Bot. 7 .-481 (1850). Guba, Monogr.

Monochaetia & Pestalotia : 268 (1961).

Seiridium liquidambaris Berk. & Curt, apud Berk, in Grevillea 2 : 154 (1874).
Mastigonetron fuscum Klebahn in Myc. Centralbl. 4 : 18, fig. 37-38 (1914).
Monochaetia liquidambaris (Berk. & Curt.) Guba, loc. cit. (1961), nom. synon.

NEPAL: Taplejung district, Sanghu, Tamrang Khola, 2,000 m., on dead twigs,
19 Oct. 1961, Norkett 6319 A.

Distribution: North and South America.

The acervuli, about 0-5 mm. in diameter, are scattered over the twigs and resemble
lenticels in appearance. The conidia, 20-27 X 9- IO- 5 /*> are dark brown, unicellular,
ovoid or ellipsoid, each with a hyaline apical appendage 30-40 x 2 fi, and a pedicel
12 X 2 fi, approximately.

The fungus agrees exactly with Klebahn's species and apparently also with the
type of P. americana (in spite of Montague's description of the conidia as biseptate)
since Guba, loc. cit., states that " Montague's drawings of the fungus and my study
show dark colored i-celled ellipsoid conidia ". The drawing mentioned is apparently
unpublished.

In view of its unicellular conidia the species cannot be included in either Mono-
chaetia or Pestalotia. As for Seiridium Nees & Henry, (Syst. Pilze : 18, t. 3 (1837)),
arguments can be brought forward for retaining it as an earlier name for Mastigone-
tron if one regards as accurate the elder Nees's (Syst. Pilze & Schwamme : 22 (1816),
t. i fig. 19 (1817)) description and drawing of unicellular appendaged conidia. This
non-septate condition of the conidia was accepted by Berkeley & Curtis when they
selected this genus for the fungus on Liquidambar. Alternatively, Seiridium can
be regarded as an older name for Monochaetia if Fries's (Syst. Mycol. 3 : 473 (1832))
statement that his examination of Nees's material showed multiseptate conidia is
taken as correct. The original collection appears to be lost but other collections
since then and reputed to be the same species, i.e. S. marginatum, are invariably

138 FUNGI OF RECENT NEPAL EXPEDITIONS

described as having a brown septate conidium with a transparent apical seta. The
probable explanation for this discrepancy is that Nees chanced to examine and
illustrate the unripe fungus, i.e. before the conidia became septate, while they were
still unicellular, spindle-shaped and contained grey granular protoplasm. Fries, on
the other hand, and all subsequent workers have described the mature fungus, which
represents what is now generally regarded as typical Monochaetia. Taking this latter
view, or better still treating Seiridium as a nomen confusum, this generic name cannot
be used in place of Mastigonetron and therefore the correct name for the Nepal
fungus becomes Mastigonetron americanum, as cited above.

STILBACEAE

ARTHOBOTRYUM NILGIRENSE Subram. in Proc. Ind. Acad. Sci., Sect. B, 46 : 324,
fig. i (1957).
NEPAL: Chainpur, Tumlingtar, 660 m., on old bamboos, 15 Dec. 1961, Norkett

Distribution: Previous record and original description from bamboo, Sim's Park.
Coonoor, Madras, 1956.

Podosporium himalensis Balfour-Browne, sp. nov. (Fig. 4).

Synnemata atrobrunnea numerosa et effusa ex hyphis parallelibus juxtapositis
efformata. Conidia atrobrunnea, 1-12 septata, obclavata, leniter curvata et irregu-
lariter disposita, 9-15 x 16-100 /*.

NEPAL: Lamjung Himal, 4,500 m., black woolly growth on branches of Rhododen-
dron campanulatum, 14 July 1954, Stainton, Sykes 6- Williams 6341. (Herb. Mus.
Brit., holotype).

The general form of the synnemata is similar to that described by Subramanian
(Journ. Ind. Bot. Soc. 35 : 73 (1956)) for his Prathoda saparva; the conidia however
are different and the conidiophores hardly distinct from the hyphae and not geni-
culate. The Nepal material resembles an extremely luxuriant form of Podosporium
rigidum Schwein., originally described from Carolina. The hyphae are hormiscium-
like and very probably any portion breaking away can regenerate fresh growth,
independently of conidial reproduction.

STILBUM CINNABARRINUM Mont, in Ann. Sci. Nat., Ser. 2, Bot. 8 : 360 (1837).

Stilbum lateritium Berk, in Ann. Nat. Hist. 4 : 291, pi. 8 fig. 2 (1840).

NEPAL: Tumlingtar, by shore of Sabhaya Khola, 600 m., n Dec. 1961, Norket
8107 C.

Distribution: North and South America, Cuba, Dominica, Africa, Nepal, India,
Ceylon, Australia.

Pleonectria pseudotrichia (Schwein.) Wollenw. is its perfect form. The present
collection occurs on unnamed bark and is very sparing.

FUNGI OF RECENT NEPAL EXPEDITIONS

139

FIG. 4. Podosporium himalensis Balfour-Browne. A, general aspect; B, conidia and
conidiophorous hyphae. Stainton, Sykes &> Williams 6341.

STILBUM INCONSPICUUM Currey in Trans. Linn. Soc. Lond., Ser. 2, Bot. i : 129

(1876).

Stilbum kurzianum Cooke in Grevillea 16 : 71 (1888).

NEPAL: Sanghu, 2,000 m., 17 Nov. 1961, Norkett 7153.
Distribution: India.

On dead twigs of Rosa sp. Synnemata 3-4 mm. tall, pale orange soon becoming
cinereous; conidia rod-shaped 7-9 x 3 fi.

140 FUNGI OF RECENT NEPAL EXPEDITIONS

TUBERCULARIACEAE

EPICOCCUM ANDROPOGONIS (Rabenh.) Schol-Schwarz in Trans. Brit. Mycol. Soc.
42 : 171, t. 9% ii (1959).

Cerebella andropogonis Rabenh. in Bot. Zeit. 9 : 669 (1851).

NEPAL: Bhadauri, near Pokhara, 2,000 m., on inflorescence of a grass, i Nov.
1954, Stainton, Sykes 6- Williams 8326.

Distribution : Worldwide but chiefly in the tropics and subtropics.

This genus has been revised recently by Langdon (Mycol. Commonw. Mycol. Inst.
Papers, 61 : 1-18 (1955) under the name Cerebella), and by Schol-Schwarz (torn. cit. :

EPICOCCUM PURPURASCENS Schlechtend., Fl. Berol. 2 : 136 (1824). Link in L., Sp.
PL, Ed. 4, 6 (2) : 108 (1825).

Epicoccum nigrum Link, loc. cit. Schol-Schwarz in Trans. Brit. Mycol. Soc. 42 : 170
(1959).

NEPAL: Argam, near Pokhara, 830 m., on leaves and stems of Acrocepkalus indicus
(Burm.) O. Kuntze, 10 Sept. 1954, Stainton, Sykes & Williams 7146.
Distribution : Worldwide.

SPECIAL LITERATURE

BALFOUR-BROWNE, F. L. (1955). Some Himalayan Fungi. Bull. Brit. Mus. (Nat. Hist.),

Bot. i : 189-218.
BERKELEY, M. J. (1850). Decades of Fungi. Hook., Journ. Bot. 2 : 42-51, 76-88, 106-112.

(1851). Decades of Fungi. Hook., Journ. Bot. 3 : 39-49, 77-84, 167-172, 200-206.

(1852). Decades of Fungi. Hook., Journ. Bot. 4 : 97-107, 130-142.

(1854). Decades of Fungi. Hook., Journ. Bot. 6 : 129-143, 161-174, 204-212, 225-235.
BONDARTZEV, A. S. (1953). Trutov'ie grib'i Europel skol chasti S.S.S.R. i Kavkaza. [Poly-

poraceae of the European part of the U.S.S.R. and the Caucasus.] Leningrad.
BUTLER, E. J. & BISBY, G. R. (1960). Fungi of India. 2nd ed., revised by R. S. VASUDEVA.

New Delhi.

CORNER, E. J. H. (1932). The fruit-body of Polystictus xanthopus. Ann. of Bot. 46 : 71-111.
~~ ( I 953)- The construction of Polypores. I. Introduction. Phytomorphology 3 : 152-167.
CUNNINGHAM, G. H. (1947). Notes on the classification of the Polyporaceae. N.Z. Jourt

Sci. &> Technol. 28, A : 238-251.

(1947-1950). New Zealand Polyporaceae. N.Z. Dept. Sci. Industr. Res. Bull. 72-83.

(1965). Polyporaceae of New Zealand. N.Z. Dept. Sci. Industr. Res. Bull. 164.

DONK, M. A. (1933). Revision der niederlandischen Homobasidiomycetae Aphyllophora-
ceae. II. Med. Bot. Mus. Herb. Rijks Univ. Utrecht 9 : 1278.

(1960). The generic names proposed for Polyporaceae. Persoonia i : 173-302.
IMAZEKI, R. (1943). [Genera of Polyporaceae of Nippon.] In Japanese. Bull. Tokyo Sci.

Mus. 6, i-ni.
KOTLABA, F. & POUZAR, Z. (iQ57). Poznamky k tfideni evropskych chorosu. [Notes or

the classification of European pore Fungi.] Ceskd Mykol. n : 152-170.
NOBLES, M. (1958). Cultural characters as a guide to the taxonomy and phylogeny of the

Polyporaceae. Canad. Journ. Bot. 36 : 883-926.

FUNGI OF RECENT NEPAL EXPEDITIONS

141

OVEKHOLTS, L. O. (1953)- The Polyporaceae of the United States, Alaska and Canada. Ann

Arbor, Mich.
PILAT, A. (1936-1942). Polyporaceae. In KAVINA, C. & PILAT, A., Atlas des Champignons

de I' Europe, 3. Praha.
PINTO-LOPEZ, J. (1952). " Polyporaceae " Contribuiao para a sua biotaxonomia. Mem. Soc.

Broter. 8 : 5-191.
SHEALS, J. G. & INGLIS, \V. G. (1965). The British Museum (Natural History) expedition to

east Nepal 1961-62. Introduction and lists of localities. Bull. Brit. Mus. (Nat. Hist.},

Zool. 12 : 95-114.

SINGER, R. (1962). The Agaricales in modern taxonomy. 2nd ed. Weinheini.
TEXIEIRA, A. R. (1962). The taxonomy of the Polyporaceae. Biol. Rev. Cambr. Phil. Soc.

37 : 5i-8i.

\

PRINTED IN GREAT BRITAIN
BY ADLARD & SON LIMITED
BARTHOLOMEW PRESS, DORKING

A SYNOPSIS OF
JAMAICAN MYRSINACEAE

W. T. STEARN

BULLETIN OF

THE BRITISH MUSEUM (NATURAL HISTORY)
BOTANY Vol. 4 No. 4

LONDON: 1969

A SYNOPSIS OF
JAMAICAN MYRSINACEAE

BY

WILLIAM THOMAS STEARN

Pp. 143-173; 25 Text-figures; Plates 6-13

BULLETIN OF

THE BRITISH MUSEUM (NATURAL HISTORY)
BOTANY Vol. 4 No. 4

LONDON: 1969

THE BULLETIN OF THE BRITISH MUSEUM
(NATURAL HISTORY), instituted in 1949, is
issued in five series corresponding to the Departments
of the Museum, and an Historical series.

Parts will appear at irregular intervals as they become
ready. Volumes will contain about three or four
hundred pages, and will not necessarily be completed
within one calendar year.

In 1965 a separate supplementary series of longer
papers was instituted, numbered serially for each
Department.

This paper is Vol. 4, No. 4 of the Botany series.

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

Trustees of the British Museum (Natural History) 1969

TRUSTEES OF
THE BRITISH MUSEUM (NATURAL HISTORY)

Issued 30 May, 1969 Price i 45.

A SYNOPSIS OF JAMAICAN MYRSINACEAE

By WILLIAM T. STEARN

SUMMARY

The family Myrsinaceae is represented in Jamaica by three genera, Ardisia, Wallenia and
Myrsine (including Rapanea) . Recognition of the species presents especial difficulty in Wallenia,
which is dioecious ; some taxa are known only from specimens with male flowers or female flowers
or fruits, i.e. in only one of three states needed for adequate discrimination ; few are known in all
three; 15 species, two occurring also in Cuba, the others endemic, are listed here. Ardisia is
represented by eight species, one (A. solanacea) introduced and naturalized; Myrsine by two,
which Mez placed in Rapanea. Keys, synonymy and statements of distribution are given.
Ardisia brittonii and A. byrsonimae are described as new; A. rosea Urban (1908), non A. vosea
King & Gamble (1905), is renamed A. urbanii.

INTRODUCTION

STUDENTS of the family Myrsinaceae tend to be censorious of the work of Carl Mez
(1864-1944) on this difficult and to most botanists unattractive family, wherein, as
C. L. Lundell (19660;) has remarked, " most generic characters are weak and those
separating species are often weaker." Nevertheless, they must needs be grateful
to Mez for his comprehensive survey of its West Indian members published in Urban's
Symbolae Antillanae 2 : 389-433 (1901) and for his world monograph in Engler's
Das Pflanzenreich IV. 236 (1902), in which he provided a helpful basic classification,
co-ordinated the extensive literature, and enumerated and described the material
then available. The subsequent great increase of specimens in herbaria and a more
critical attention to matters of nomenclature and typification than prevailed then
have, however, made necessary many alterations and some drastic revisions of Mez's
treatment; moreover, some characters emphasized by him in his keys have been
found impracticable. E. Y. Hosaka, for example, in his account of the Hawaiian
species of Myrsine sensu lato (in Occ. Pap. B.P. Bishop Mus. 16 : 25-76 (1940))
concluded that Rapanea, Suttonia and Myrsine, kept as separate genera by Mez,
overlap so much in their characters that they should be combined in one genus,
Myrsine, as indeed they were by Bentham & Hooker in 1873 ; H. C. D. de Wit (in
Bull. Jard. Bot. Bruxelles 27 : 233-242 (1957)) has united Heberdenia with Ardisia.
On the other hand, Lundell (in Wrightia 3 : 88 (1963)) has raised Mez's Ardisia sect.
Synardisia to generic rank as Synardisia. A few examples may suffice to indicate
the nomenclatural rectifications needed. The name Ardisia coriacea Swartz (1788),
based on Jamaican material, was misapplied by Mez to a species which does not
occur in Jamaica, i.e. A. obovata Ham. (A. guadalupensis Griseb.). The name
Ardisia sanguinolenta Blume (1825) is the correct name for A. zollingeri A. DC.
(1844) : A . wallichii A. DC is the correct name for A . sanguinolenta Wall, ex Mez
(1902). Merrill has called attention (in Lingnan Sci. Journ. n : 50 (1932) and in
Contrib. Arnold Arb. 8 : 131 (1934)) to Mez's misinterpretation of A. humilis Vahl
(1794), which, according to Merrill, is the same as A. hainanensis Mez (1902), while

BOT. 4, 4 8

146 A SYNOPSIS OF JAMAICAN MYRSINACEAE

A. humilis sensu Mez is A. elliptica Thunb. (1795) (syn. A. obovata Blume (Dec.
i825~March 1826), non A. obovata Ham. (Nov. 1825)). As regards the East Asiatic
Myrsinaceae, a much needed nomenclatural and taxonomic overhaul has been made
by E. H. Walker (in Philipp. Journ. Sci. 73 : 1-258 (1940), with supplementary
notes in Journ. Arnold Arb. 23 : 344-355 (1942), in Bot. Mag. Tokyo 67 : 105-111,
155-162, 203-213, 247-255 (1954) and in Quart. Journ. Taiwan Mus. 12 : 161-194
(1959)). Unfortunately, there exists no comprehensive survey of the American
Myrsinaceae later than Mez's of 1902, although many new species have been published
since by Urban, Standley and C. L. Lundell, who has also revised the genus Parathesis
(Lundell, 19666) and the Guatemalan members of the family (Lundell, I966a). The
author of a local Flora is accordingly obliged to make a study more or less independent
of Mez's work and to extend his enquiries beyond his own geographical area. The
following synopsis of Jamaican Myrsinaceae is a by-product of preparing a more
detailed account for the Flora of Jamaica ; descriptions and particulars of localities
to be given there in detail are either summarized below or omitted.

Mez in 1901 based his account of the Myrsinaceae occurring in Jamaica on 14 wild
and two cultivated gatherings of Ardisia, 15 of Wallenia, 13 of Rapanea (i.e. Myrsine),
i.e. 42 in all. For the Flora of Jamaica I have examined at least 120 Jamaican
gatherings of Ardisia, 130 of Wallenia, and 50 of Myrsine, i.e. in all more than 300
gatherings. This increase in material has made evident the distribution of the
Myrsinaceae in Jamaica, but has not greatly simplified the problems of their classifica-
tion. Thus, despite much field work, the female flowers of Wallenia clusioides,
W. crassifolia, W. discolor, W. erythrocarpa and W. xylosteoides and the male flowers
of W. discolor, W. elliptica, W. erythrocarpa, W. punctulata and W. sylvestris have
apparently missed collection; they have not been observed in the material studied,
which, in addition to that of the British Museum (Natural History), London (cited
below as BM), has included the specimens of the Systematisch-Geobotanisches
Institut, Universitat Gottingen (GOET), the Institute of Jamaica, Kingston,
Jamaica (IJ), the Royal Botanic Gardens, Kew (K), the New York Botanical Garden
(NY) and the University of the West Indies, Mona, Jamaica (UCWI) and the lecto-
type of Wallenia grisebachii at the Botanische Staatssammlung, Munich. Thanks are
here expressed to the Directors of these Institutions for their invaluable co-operation
in lending or making available for study this extensive material. Nevertheless, the
following account as regards the status and distinctive characters of some taxa is
necessarily pro visional. Text-figures 21-25 have been drawn by Joanna C. Webb
(Mrs. D. Erasmus).

HISTORY

The first members of the Myrsinaceae to be recorded from Jamaica, indeed from
all America, were Ardisia tinifolia and Wallenia laurifolia gathered by Hans Sloane
in 1688 or 1689. He collected the Ardisia " on the Mountains near Mr. Elletson's
Plantation in Liguanee ", i.e. in the parish of St. Andrew, Jamaica, and publishec
it in his Catalogus Plantarum: 169 (1696) and Voyage to the Islands Madera etc
2 : 98, t. 205 f. 2 (1725) as " Jasminum forte, arboreum, foliislaurinisobtusislatioribus

A SYNOPSIS OF JAMAICAN MYRSINACEAE 147

atrovirentibus, flore pentapetalo racemose purpureo reflexo ". Concerning this,
Sloane wrote: " This Tree riseth to about thirty Foot high, having a Clay or Ash-
colour'd pretty smooth Bark; its Twigs are set about with Leaves which are very
smooth, of a dark green Colour, having a quarter of an Inch long Footstalks, being
four Inches long and two broad in the Middle, where broadest, having an eminent
middle Rib. The Flowers stand on several little Branches, being pentapetalous,
purplish, bow'd back, having yellow Stamina in their Middle." Olof Swartz found
the same species " in aridis submontosis Jamaicae " during his stay in the West
Indies from 1784 to 1786, and gave it the binomial Ardisia tinifolia when founding
the genus Ardisia in his Nova Genera et Species Plantarum sen Prodromus : 48 (1788).
Swartz referred five species to Ardisia: I. A. tinifolia from Jamaica; 2. A. coriacea
(now regarded as conspecific with A. tinifolia}, its provenance vaguely recorded by
him as " India occidentalis "; 3. A. serrulata (now Parathesis serrulata (Swartz) Mez)
from Hispaniola, these three being of his own collecting ; 4. A . laterifolia (now Stylo-
gyne laterifolia (Swartz) Mez), vaguely recorded as from "India occidentalis";
5. A. parasitica (now Grammadenia parasitica (Swartz) Griseb.) from " Montserrat "
(later corrected to Dominica) ; these last two were known to him only from specimens
in the herbarium of Sir Joseph Banks, now at the British Museum (Natural History),
as the asterisk against the diagnosis indicates. Later authors have accepted A.
tinifolia as the type-species of Ardisia. Swartz's description of the species in his
Flora Indiae Occidentalis i : 468 (1797) is fairly detailed and accompanied by a
reference to his " Icon. fasc. 3 ". Unfortunately, this third fascicle of his Icones
Plantarum incognitarum (fasc. I, 1794; fasc. 2, 1800) was never published, although
Swartz sent the illustrations and text to Schreber in Erlangen for publication.
Schreber died in 1810, Swartz in 1818; the Palmische Verlagsbuchhandlung in
Erlangen published a note in Flora (Regensburg) 3 : 144 (March 1820) offering these
illustrations to whoever would undertake their publication, but no one responded.
Their subsequent history is unknown until about 1890 when the Berlin botanist
Ignaz Urban acquired a volume containing presumably all that had survived of
Swartz's 200 drawings of West Indian plants. They are cited in Urban's Symbolae
Antillanae (cf. i : 164 (1898)) as " Sw. Ic. ined ", and in Fawcett & Rendle's Flora of
Jamaica (i : xviii, etc. (1910)) as " Sw. Icon, ined: ". They could not be traced
at the Botanisches Museum, Berlin-Dahlem after the 1939-45 war and hence were
believed to have been destroyed by bombing in March 1943, as stated in Sitwell and
others, Great Flower Books : 77 (1956). Luckily, however, Urban had sold this
precious volume in 1922 (cf. Journ. Bot. (Lond.) 60 : 361 (1922)) toC. A. Lindman
in Stockholm and it is now preserved in the library of the Royal Swedish Academy
of Sciences; Plates 6 and 7 are reproduced here from these drawings by permission
of the Royal Swedish Academy of Sciences, to which grateful acknowledgement is
made. Sets of photographs of Swartz's unpublished drawings have been deposited
in the libraries of the British Museum (Natural History), London, the Institute of
Jamaica, Kingston, Jamaica, and the Hunt Botanical Library, Pittsburgh, U.S.A.
Swartz's coloured drawings representing Ardisia tinifolia and A. coriacea, together
with A . serrulata and A . parasitica, leave no doubt as to the application of the names.
Swartz applied the name A. coriacea to a Jamaican plant (without evident veining

148 A SYNOPSIS OF JAMAICAN MYRSINACEAE

of the leaves but otherwise very similar to his A. tinifolia) which was described by
Mez in 1901 as A. harrisiana; hence, as pointed out by Urban (Symb. Antill. 3 : 330
(1902); op. cit. 8 : 519 (1921)), Mez used the name A. coriacea wrongly for a species
of the Lesser Antilles, i.e. A. obovata (A. guadalupensis) .

According to Sloane's account (Voy. Nat. Hist. Jam. I : 234 (1707)) of the
Jamaican plant now known as Wallenia laurifolia, but recorded by him as " Bryonia
nigra fruticosa, foliis laurinis, floribus, racemosis, speciosis ", he " gathered it, if
I rightly remember, in St. Maries, near Cabeca del oro in the North-side of this Island
amongst the Woods " ; he described the flowers simply as " at the top of the Branch . . .
standing in a bunch together, being many very beautiful and small ". On this
species Swartz in 1788 (Nov. Gen. & Sp. PL : 2, 31) founded the genus Wallenia,
naming it in honour of Matthew Wallen, from whom Swartz received much hospitality
while in Jamaica. Meanwhile, Jacquin had published in his Selectarum Stirpium
Americanamm Historia : 17 (1763) the provisional designation Petesioides laurifolium
referring to incomplete material collected by him in Santo Domingo (Haiti) between
1755 and 1759, and Petesioides was accordingly cited in the 1952, 1956, and 1961
editions of the International Code of Botanical Nomenclature, under No. 6304, as
a nomen genericum rejiciendum necessitating the conservation of Swartz's Wallenia.
Linnaeus had, however, ruled in Critica botanica : 34 (1737) that " nomina generica
in oides desinentia, e foro Botanico releganda sunt " (generic names ending in oides
are to be banished from the botanical forum). Hence, when Linnaeus himself used
designations such as Baccharioides, Bannister oides, Cornutioides, Euonymoides,
Hibiscoides, Oxycoccoides, Senecioides etc. in his Flora Zeylanica (1747) for obscure
plants and when loyal followers of Linnaean method and nomenclature such as
Jacquin, Loefling and Rottboll used designations such as Celosioides, Diodiodes,
Malpighioides, Sideroxyloides , Staehelinoides and Viscoides, they did not intend them
to be permanent names for the genera concerned but merely as provisional or token
names indicating resemblance; they intended that these -oides designations should
later be replaced by correctly formed names. Jacquin's description of his Petesioides
indicates that he possessed only a specimen with functionally female flowers; he
believed that it represented a new genus which could not be properly established
until fruits became available; accordingly, in his own words, " ego interea ab habitu
nomen fluxum imposui " (I meanwhile, from its habit of growth, have given it a
perishable name). Such token designations are in a different category from the
pre-Linnaean generic names ending in -oides, such as Capnoides, Ficoides, Nymphoides,
brought into post-Linnaean use as definite generic names by Miller, Hill, Adanson,
Medicus and others. In a like spirit Rottboll in 1772 described Schoenoides and
Scirpoides as new genera to which names would be given later ; in 1773 he named the
first Kyllinga, the second Fuirena. Swartz thus acted rightly according to the
procedure of his times in replacing Petesioides by Wallenia and Jacquin evidently
approved Swartz's action, for he himself adopted the name Wallenia in his Plantarum
rariorum Horti Caesarei Schoenbrunnensis Descriptions I : 13, t. 30 (1797). Valid
publication of " Petesiodes Jacq." as a real generic name dates from Otto Kuntze's
adoption of it in 1891 (Revis. Gen. PI. 2 : 402).

In 1788 Swartz named, as Samara coriacea (Nov. Gen. & Sp. PI. : 32), a third

A SYNOPSIS OF JAMAICAN MYRSINACEAE 149

Jamaican member of the family, referring it to the Linnaean genus Samara, which is
now included in Embelia Burman f. This species, found " in sylvis montium
Jamaicae australis ", is congeneric with Rapanea guianensis Aublet, as Swartz noted,
but Rapanea itself is not well separated from My r sine L.

The collections including Myrsinaceae made in Jamaica during the first half of the
nineteenth century by Macfadyen, Alexander, McNab, Nathaniel Wilson, March, and
Purdie and used by Grisebach in preparing his Flora of the British West Indian
Islands : 392-397 (1861) provided the material for publication there of Wallenia
clusiifolia, W. venosa, Ardisia xylosteoides , and A. clusioides, and the record of A.
humilis (actually A. solanacea) as naturalized in Jamaica. Grisebach's book (cf.
Stearn, 1965) seems to have given the impression for many years that no further
collecting need be done in the West Indies. William Harris's collecting from 1894
onwards brought about a new era in the botanical exploration of Jamaica, revealing
the existence there of a multitude of new species which had escaped the notice of
earlier collectors (cf. Urban, Symb. Antill. 6 : 70-86 (1909)). In 1895 Krug & Urban,
working out Harris's collections from the Blue Mountains, Jamaica, published two
further species, Myrsine acrantha and Ardisia densiflora. Mez in 1901 recorded from
Jamaica four species of Ardisia, A. harrisiana being proposed as new, eight species
of Wallenia, and four of Rapanea. At the same time he defined more satisfactorily
the genera, by transferring to Wallenia some species which Grisebach had included
in Ardisia, although he separated Rapanea from Myrsine, reducing the latter to four
African and Asiatic species typified by M. africana L.

The acquisition of new Jamaican material, mostly collected by Harris and by N. L.
Britton of the New York Botanical Garden, and the re-study of the older material
seen by Mez led Urban between 1908 and 1915 to describe as new Ardisia dictyoneura,
A. troyana, Wallenia calyptrata, W. corymbosa, W. discolor, W. elliptica, W. erythro-
carpa, W. punctulata and W. sylvestris and Britton to describe Petesioides sub-
verticillata (now Wallenia subverticillata] . Urban's descriptions are wonderfully
accurate, clear and detailed and are accompanied by helpful critical notes in which
he made the most of inadequate material, but as usual he never provided a key to
aid their recognition. The construction of such a key has indeed proved very
difficult, partly because some taxa are only known in one of their three possible
states thus of the 16 species of Wallenia distinguished here, only six have been
collected in their staminate, pistillate and fruiting states but also because of the
variation among these plants, which makes every gathering seem a little different
from other gatherings. This is discussed further below under Wallenia.

GEOGRAPHICAL DISTRIBUTION

Four, if not five, of the Jamaican members of the Myrsinaceae occur outside the
island. These are Ardisia compressa, which extends from Mexico to northern South
America, Myrsine coriacea, which extends from Cuba over Jamaica, Hispaniola,
Puerto Rico and the Lesser Antilles, and from Mexico over Central America, to
Bolivia and Argentina, Wallenia laurifolia, which occurs also in Cuba and Hispaniola,
and W. subverticillata, which also occurs in Cuba. Ardisia densiflora is closely allied

150 A SYNOPSIS OF JAMAICAN MYRSINACEAE

to A. spicigera of Mexico and A. gentlei of British Honduras; indeed, according to
Lundell (19660) the three are conspecific. The ornamental Asiatic species A.
solanacea (Text-fig. 8) has become naturalized in several places. All the others appear
to be endemic to Jamaica, although Myrsine acrantha may probably be found on other
islands; specimens of it are likely to be referred to Myrsine (Rapanea) guianensis.
Members of the family may be found in Jamaica at any altitude from sea-level to the
top of Blue Mountain Peak (7,400 ft., 2,250 m.).

Although it would be premature to define types of distribution within Jamaica on
the basis of present material, this is adequate to indicate relative ranges. Thus
Myrsine acrantha (Text-fig. 20), M. coriacea (Text-fig. 19), Ardisia tinifolia (Text-
figs. 1-4), Wallenia laurifolia (Text-fig. 17), and W. venosa (Text-fig. 14) have
comparatively wide ranges extending both east and west. A few are localized in the
east, i.e. Ardisia brittonii (Text-fig. 5), Wallenia sylvestris (Text-fig. 18), and W.
subverticillata (Text-fig. 9). Some others are restricted to the higher parts of the
Blue Mountains, i.e. Ardisia densiflora (Text-fig. 6), Wallenia calyptrata (Text-fig. 13),
W. crassifolia (Text-fig. 10), and W. fawcettii (Text-fig. 16). As far as is known,
Ardisia byrsonimae (Text-fig. 5), A. urbanii (Text-fig. 5), Wallenia corymbosa (Text-
fig, n), W. elliptica (Text-fig. 9), and W. punctulata (Text-fig. 13) are restricted to the
middle of the island. Ardisia dictyoneura (Text-fig. 6), Wallenia grisebachii (Text-
fig. 15), and W. purdieana (Text-fig. 16) occur both in the middle and west, Ardisia
compressa (Text-fig. 7), Wallenia clusioides (Text-fig. 9), and W. erythrocarpa (Text-
fig. 12) only in the west. As there has obviously been no bias in making the collec-
tions here analysed, the species such as Ardisia brittonii (Text-fig. 5), A. byrsonimae
(Text-fig. 5), A. urbanii (Text-fig. 5), Wallenia discolor, W. elliptica (Text-fig. 9),
W. erythrocarpa (Text-fig. 12), and W. punctulata (Text-fig. 13), which are known
only from single gatherings or from gatherings made at the same place, must be of
very limited range.

Specimens are cited below under Jamaican parishes, proceeding usually from
west to east or from north to south within the parish.

FAMILY CHARACTERS

The family Myrsinaceae as represented in the West Indies consists of small trees
or shrubs, with the leaves usually alternate, rarely whorled, and without stipules.
The flowers are regular, bisexual or unisexual, with parts in fours or fives, and are
grouped in axillary or terminal racemes or panicles sometimes condensed into
fascicles or umbels. The calyx is persistent, with segments free or basally connate,
imbricate or valvate. The corolla is usually gamopetalous but in Myrsine acrantha
parted to the base. The stamens are inserted on the corolla, opposite to and the
same number as the lobes of the corolla, with the anthers opening introrsely by
lengthwise slits or apical pores. The ovary, superior in West Indian genera, is
one-chambered with a free central placenta, into which the few or many ovules are
sunk, appearing as slight or distinct bulges on its surface ; the style may be short or
long, with the stigma undivided or lobed or fimbriate. The fruit in West Indian
genera is a more or less globose drupe. The family as a whole includes about 900

o

"

3

o

f

s

156 A SYNOPSIS OF JAMAICAN MYRSINACEAE

species and occurs in the tropics and sub-tropics of the Old and the New World.
Of its 30 or so genera, nine are represented in the West Indies as a whole but only
three in Jamaica. Those absent from Jamaica are Parathesis (Cuba, Hispaniola,
Puerto Rico), Stylogyne (Lesser Antilles), Conomorpha (Lesser Antilles), Weigeltia
(Lesser Antilles), Cybianthus (Trinidad) and Grammadenia (Puerto Rico, Lesser
Antilles) .

KEY TO JAMAICAN GENERA

Flowers in usually terminal (sometimes axillary and then long-pedunculate) racemes

or panicles; style slender; stigma inconspicuous.

Flowers bisexual, with stamens and style both well developed in the same flower;
corolla c. 6-12 mm. long; ovules numerous, scattered in several series over
the placenta . . . . . . . . . i. Ardisia

Flowers unisexual, dioecious, the male flowers with well developed exserted
stamens, the female flowers with abortive included stamens but exserted
styles; corolla 1-5-6 mm. long; ovules 3-4 in a single horizontal row on the
placenta . . . . . . . . . .2. Wallenia

Flowers in umbellate clusters on very short, scaly axillary spurs ; style short or none
(in American species) ; stigma conspicuous . . . -3- Myrsine

i. ARDISIA Swartz

ARDISIA Swartz, Nov. Gen. & Sp. PL : 3, 48 (1788) ; Fl. Ind. Occ. i : 467 (1797) ;

nomen conservandum No. 6285.
Lectotype: A. tinifolia Swartz.

KEY TO JAMAICAN SPECIES

Inflorescences all axillary; pedicels 1-4 cm. long, clustered almost umbellately at

the end of the peduncle; anthers 5-6 mm. long . . 8. A. solanacea

Inflorescences terminal (sometimes with a few axillary inflorescences below) ; pedicels

mostly 1-7 mm. long, sometimes to 12 mm. long but then racemosely arranged

along the lateral peduncle; anthers 1-5-5 mm. long:

Lateral peduncles compound, i.e. bearing 3 to 4 secondary peduncles each
umbellately 2-6-flowered; leaves 7-16 cm. long, the blades narrowly obovate
to oblanceolate ; filaments of stamens 0-5 mm. long . . 7. A. compressa
Lateral peduncles simple, i.e. directly bearing 3 to 28 pedicelled flowers but no
secondary peduncles; leaves mostly less than 12 cm. long, the blades various;
filaments of stamens 1-5 mm. long:

Leaves closely clustered at the tips of shoots, less than 5-5 cm. long; inflorescence
contracted, less than 4 cm. long and broad, the branches 3-6-flowered, the
pedicels 1-4 mm. long; central Jamaica . . . 6. A. byrsonimae

Leaves more widely spaced, some or all more than 5-5 cm. long; inflorescence
more than 4 cm. long and broad, the branches usually many-flowered, the
pedicels 2-15 mm. long:

A SYNOPSIS OF JAMAICAN MYRSINACEAE 157

Inflorescence somewhat fastigiate, the densely many-flowered branches

ascending at c. 40; corolla c. 5 mm. long; anthers c. 1-5 mm. long, much

shorter than the filaments; Blue Mountains . . 5. A. densiflora

Inflorescence looser, the branches spreading at c. 50-60 ; corolla 7-12 mm.

long; anthers 2-5-5 rnm. long, equalling or longer than the filaments:
Pedicels 1-4 mm. (rarely to 5mm.) long; leaf -blades with well-marked
reticulate venation . . . . . 4. A. dictyoneura

Pedicels 4-15 mm. long; leaf-blades not markedly reticulate:
Sepals 2-5-4 mm. broad:

Anthers 4-5-5 mm. long, apically dehiscent; sepals almost or quite

glabrous; eastern Jamaica . . . . 2. A. brittonii

Anthers 3-3-5 mm. long, laterally dehiscent almost to the base; sepals

ciliate; western Jamaica. . . . . 3. A. urbanii

Sepals 1-2 mm. broad:

Sepals ovate or broadly ovate :

Anthers 3-3-5 mm. long . . . i. A. tinifolia var. i

Anthers 4-5 mm. long .... A. tinifolia var. 2

Sepals mostly narrowly ovate or oblong-lanceolate :

Leaf -blades mostly less than 6 cm. long; sepals 2-2-5 mm - l n g

A . tinifolia var. 3
Leaf-blades 7-12 cm. long; sepals 2-5-3 mm. long A. tinifolia var. 4

Subgenus i. ARDISIA
ARDISIA subgenus ARDISIA; subgenus typicum generis.

Ardisia Swartz, Nov. Gen. & Sp. PI. : 3, 48 (1788), sensu stricto.
Pickeringia Nutt. in Journ. Acad. Sci. Philad. 7 : 95 (1834).
Ardisia sect. Euardisia Endl., Gen. PI. : 736 (1839).

Ardisia subgenus Pickeringia (Nutt.) Mez in Urban, Symb. Antill. 2 : 396 (1901); in Engler,
Pflanzenr. IV. 236 : 79 (1902).

Lectotype: A. tinifolia Swartz.

Inflorescence always terminal; branches simply racemose. Stamens with con-
spicuous filaments. Species all American.

i. Ardisia tinifolia Swartz, Nov. Gen. & Sp. PI. : 48 (1788); Fl. Ind. Occ. i : 468
(1797). Griseb., Fl. Brit. W. Ind. Is. : 396 (1861). Mez in Urban, Symb.
Antill. 2 : 400 (1901); in Engler, Pflanzenr. IV. 236 : 83, fig. n F-K (1902).
(Text-fig. 23G, H; pi. 6, 7.)

Ardisia coriacea Swartz, Nov. Gen. & Sp. PI. : 48 (1788); Fl. Ind. Occ. i : 470 (1797).
Ardisia harrisiana Mez in Urban, torn. cit. : 401 (1901); in Engler, torn. cit. : 83 (1902).
Ardisia troy ana Urban, Symb. Antill. 5 : 455 (1908).
" Jasminum forte, arboreum, foliis laurinis obtusis latioribus atrovirentibus, flore pentapetalo

racemoso purpureo reflexo " Sloane, Cat. PI. Jam.: 169 (1696); Voy, Jam. Nat. Hist. 2:

98, t. 205 fig. 2 (1725).

The species Ardisia tinifolia as here accepted includes plants diverging in form and
size of leaf, in number of flowers, shape of sepals and length of anthers. Selected
specimens, e.g. Harris 8741 (type-collection of A. troy ana] from near Troy, Central

158 A SYNOPSIS OF JAMAICAN MYRSINACEAE

Jamaica, with very small leaves, Proctor 16621 from Orange River Valley, St. James,
with comparatively elongated sepals, and Loveless 1198 from the Long Mountain,
near Kingston, with long anthers, could reasonably be regarded as exemplifying
independent species and were indeed treated by the writer as such until the difficulty
of finding stable combinations of characters and of allocating specimens intermediate
in one character or another to these supposed taxa made it more practical to place
them all in one species, within which four major populations (Text-figs. 1-4) are
more or less distinguishable though not easily defined. They are simply given
numbers below. Swartz's original diagnosis of his A. tinifolia in 1788 is brief but
the more detailed description in his Flora Indiae Orientalis, and his unpublished
illustration (PL 6, see p. 147) and authentic specimens (BM) clearly establish the
identity of the plant he had in mind. A representative modern gathering is Harris
12087, collected between Constant Spring and Bardowie, St. Andrew, with veining
on the lower surface of the leaf plainly evident. It typifies the population (var.
i below) found in the southern parishes of Jamaica, with an altitudinal range of
50 ft. (15 m.) to 3,500 ft. (1,070 m.). Here belongs Swartz's A. coriacea with veining
on the lower leaf-surface faint or not evident, likewise represented by an unpublished
drawing (PL 7). Diverging from typical A . tinifolia principally by having somewhat
longer anthers is the population (var. 2 below) characteristic of the Long Mountain
near Kingston, at 500 ft. (150 m.) to 1,000 ft. (300 m.). Small short-petioled leaves
evidently veined characterize plants of central Jamaica (var. 3 below) named A.
troy ana by Urban, which also differ from typical A . tinifolia in having shorter branches
of the inflorescence. These occur at 700 ft. (210 m.) to 2,500 ft. (750 m.). Their
range slightly overlaps that of a population (var. 4 below) with larger longer-petioled
leaves, looser many-flowered inflorescences and slightly shorter filaments, found in
the northern parishes of Jamaica from almost sea-level to 700 ft. (210 m.). They
occur in the same area at Sherwood Content, Trelawny, and merit particular study
there in the field.

Together the plants here placed in Ardisia tinifolia thus give the impression of
a population within which morphologically and geographically differentiated taxa
have been formed but have not achieved complete independence.

Var. 1.

A. tinifolia Swartz, loc. cit. (1788).
A. coriacea Swartz, loc. cit. (1788).
A. harrisiana Mez, loc. cit. (1901).

Leaf-blades mostly elliptic, sometimes narrowly obovate, the apex obtuse or
acute, 5-12 cm. long, 1-5-7 cm - broad. Sepals imbricate, ovate or broadly ovate,
rounded, c. 1-5-2-5 mm. long, 1-2-2 mm. broad. Anthers 3-3-5 mm. long.

ST. ELIZABETH: Proctor 13419 (BM; IJ), Proctor 11344 (BM; IJ).

CLARENDON: Proctor 22739 (BM; IJ), Proctor 9712 (IJ), Proctor 10233 (U) Stearn
1044 (BM), Adams 8424 (UCWI).

ST. ANN: Alexander (K).

ST. CATHERINE: Proctor 9310 (BM; IJ), Proctor 8373 (IJ), Proctor 22739 (IJ).

ST. ANDREW: Britton 3461 (NY), Steam 59 (BM), Proctor 11428 (IJ), Harris 5970

A SYNOPSIS OF JAMAICAN MYRSINACEAE 159

(BM; NY), Harris 11137 (NY; UCWI), Newill 3449 (IJ), Harris 5926 (NY), Harris
12081 (BM; NY; UCWI), Harris 9033 (BM; NY; UCWI), Powell 850 (IJ), Proctor
23694 (IJ), Harris 6600 (BM; NY; UCWI : type-collection of Ardisia harrisiana Mez),
Britton 2987 (NY), Proctor 24511 (BM), Proctor 23982 (BM).
ST. THOMAS: Adams 12579 (UCWI).

Var. 2.

Leaf-blades elliptic or broadly elliptic, a few broadly obovate, the apex obtuse or
rounded, 4- 9 cm. long, 2-6 cm. broad. Sepals imbricate, ovate or broadly ovate,
c. 2-2-5 mm. long, 1-5-2 mm. broad. Anthers 4-5 mm. long.

KINGSTON and ST. ANDREW: Britton 377 (NY), Robbins 1496 (UCWI), Proctor 7386
(IJ), Webster 4986 (BM), Loveless 1198 (UCWI), Yuncker 17315 (BM), Anderson &
others 626 (UCWI).

Var. 3.

Ardisia troy ana Urban, loc. cit. (1908).

Leaf-blades elliptic, the apex obtuse, c. 3-10 cm. long, 1-5-6 cm. broad (usually
less than 6cm. long, 3-5 cm. broad). Sepals narrowly ovate, 2-2-5 mm. l n g>
i- 1-5 mm. broad. Anthers 3-3-5 mm. long.

TRELAWNY: Harris 8741 (BM; NY; UCWI: type-collection of A. troy ana Urban),
Britton 605 (NY), Harris 9473 (NY; UCWI), Howard, Proctor & Steam 14677 (BM),
Howard, Proctor 6- Steam 14679 (BM; IJ), Proctor 11060 (BM; IJ), Proctor 22543
(BM; IJ), Proctor 24458 (BM; IJ), Proctor 25676 (IJ).

CLARENDON: Harris 12769 (NY; UCWI), Proctor 9772 (IJ), Proctor 10233 (IJ).

MANCHESTER: Proctor 25625 (BM; IJ).

Var. 4.

Leaf -blades narrowly obovate or elliptic, the apex acute or shortly acuminate,
5-12 cm. long, 2-6 cm. broad. Sepals scarcely or not imbricate, ovate to obovate-
lanceolate, obtuse, c. 1-5-2-5 mm. long, 1-1-5 mm - broad. Anthers c. 3 mm. long.

HANOVER : Proctor 11280 (BM ; I J) .

WESTMORELAND: Harris 10220 (BM; K; NY; UCWI).

ST. JAMES: Harris 10318 (BM; NY; UCWI), Britton & Rollick 2355 (NY) .Proctor
16621 (BM; IJ), Proctor 24292 (BM; IJ).

TRELAWNY : Proctor 11062 (BM ; I J) .

ST. MARY: Loveless 2437 (UCWI), Proctor 7546 (IJ), Yuncker 17838 (BM).

2. Ardisia brittonii Steam, sp. nov. (Text-fig. 21; PI. 8a)

Arbor parva ad 4-5 mm. alta, rawM^'shornotinisprimo f errugineis pulverulentisque ,
annotinis nigrescentibus glabrisque. Folia dissita, c. 4-8 mm. inter se distantia,
magna, patentia, longipetiolata ; lamina oblongo-elliptica, apice obtusa, margine
integra, basi cuneata in petiolum attenuata, 8-17 cm. longa, 3-8 cm. lata, glabra,
utrinque punctata, coriacea, venis numerosis utrinque prominulis marginem versus
reticulate anastomosantibus ; petiolus 1-2-2 cm. longus. Inflorescentiae terminales
multiflorae laxae, c. 15 cm. longae, 10 cm. latae, supra glabrae, inferne minute

BOT. 4, 4 9

i6o

A SYNOPSIS OF JAMAICAN MYRSINACEAE

pubescentes, ramulis patentibus ad gem. longis racemose 8-i2-floris; pedicelli
inferiores ad 1-5 cm. longi. Calyx c. 4-5-5 mm. longus; sepala imbricata parum
inaequalia, ovata vel late oblonga, rotundata, c. 4-4-5 mm. longa, 2-5-3 mm. lata,
punctata, margine glabra. Corolla rosea, c. 1-3 cm. longa; petala in 1/4 altitudinis
coalita, apice obtusa, parte libera oblonga c. 4 mm. lata, per anthesin revoluta,
nigro-punctata et lineolata. Stamina quam petala breviora, c. 1-5 mm. supra
corollae basin inserta; antherae anguste triangulares, c. 4-5-5 mm. longae, apice
obtusa rimis brevibus dehiscentes, basi cordatae; filamenta 3 mm. longa. Ovarium
late ovoideum; stylus c. 9 mm. longus, stigmate minuto.

FIG. 21. Dissected flower of Ardisia brittonii Stearn (Harris & Britton 10558), X 3.

ST. THOMAS : Bath to Cuna Pass, Sept., 1908, Harris 6- Britton 10558 (NY, holotype:
BM; K;UCWI).

This species of eastern Jamaica has been named in honour of Nathaniel Lord
Britton (1859-1934), founder of the New York Botanical Garden and author of
many botanical works on the eastern United States and the West Indies; see
National Acad. Sci. U.S.A. Biogr. Mem. 19 : 147-202 (1938) for a biography by
E. D. Merrill and a bibliography by J. H. Barnhart.

Ardisia brittonii differs from the widespread A. tinifolia principally in its broader
sepals and in having somewhat larger flowers and longer anthers than is usual in
A . tinifolia. It agrees more closely with A . urbanii of central Jamaica but may be
distinguished from that by its glabrous calyx and its longer anthers, which dehisce
only at the apices.

3. Ardisia urbanii Stearn, nom. nov. (Text-fig. 22.)

Ardisia rosea Urban, Symb. Antill. 5 : 456 (1908); non A. rosea King & Gamble (1905).

TRELAWNY: Harris 9419 (BM; NY; type-collection), Harris 9418 (BM; NY;
UCWI), Britton 659 (NY).

Ardisia urbanii is known only from collections made at Troy on the southern side
of the Cockpit Country by Harris & Britton in 1906 at about 1,600 ft. (530 m.). The

A SYNOPSIS OF JAMAICAN MYRSINACEAE

161

size of the calyx (with sepals 3-5 mm. long, 2-5-4 mm - broad) distinguishes it from
A. tinifolia; the ciliated sepals and the anthers dehiscent to the base distinguish it
from A. brittonii of eastern Jamaica. According to Harris, it forms a small tree to
2 5 ft. (7-5 m.) high, with rosy flowers. It is named in honour of Ignaz Urban (1848-
1936), the author of many important works on the West Indian flora; see Ber.
Deutsch. Bot. Ges. 48 : (205)-(225) (1931) for a biography and bibliography by
Th. Loesener.

FIG. 22. Dissected flower of Ardisia urbanii Stearn (Britton 659), x 4.

4. Ardisia dictyoneura Urban, Symb. Antill. 6 : 28 (1909).

HANOVER: Harris 10345 (BM; NY; UCWI; type-collection), Proctor 11316 (BM;
IJ), Britton 2277 (NY).

WESTMORELAND : Proctor 7321 (BM ; I J) .

ST. JAMES : Proctor 22979 (BM ; I J) .

TRELAWNY: Proctor 10610 (BM; IJ).

MANCHESTER: Proctor 11028 (BM; IJ), Proctor 11036 (BM; IJ), Proctor i82gg (IJ),
Stearn 351 (BM).

ST. ANN: Pwn^'e (K), Proctor 22835 ( BM i U). Proctor 26724 (BM), Stearn 567 (BM),
Proctor 11896 (IJ), Howard & Proctor 15039 (IJ), Proctor 7494 (IJ).

ST. CATHERINE: Howard & Proctor 13620 (IJ; NY).

Ardisia dictyoneura is a species of western and central Jamaica at 1,000-3,000 ft.
(300-900 m.). The type collection was made by -Harris in 1908 at Fray Woods in
Hanover. It has often been collected in fruit, only twice in flower, and is notable
for its profusely and finely reticulate-veined coriaceous mostly obovate or narrowly
obovate leaves and its comparatively large fruits, 6-5-8 mm. in diameter when
mature, borne on pedicels only 1-4 mm. long.

BOT. 4, 4 9

162 A SYNOPSIS OF JAMAICAN MYRSINACEAE

5. Ardisia densiflora Krug & Urban in Notizbl. Bot. Gart. & Mus. Berlin i : 79
(1895). Urban, Symb. Antill. I : 385 (1899); op. cit. 6 : 28 (1909). Mez in
Urban, Symb. Antill. 2 : 398 (1901); in Engler, Pflanzenr. IV. 236 : 82 (1902).
C. L. Lundell in Fieldiana, Bot. 24 (8) : 141 (1966). (Text-fig. 23F.)

Ardisia spicigera Donn. Smith in Bot. Gaz. 27 : 434 (1899), fide Lundell, loc. cit. (1966).
Ardisia gentlei C. L. Lundell in Field & Lab. 13 : n (1945), fide Lundell, loc. cit. (1966).

ST. ANDREW: Harris 5227 (K; UCWI), Harris 5431 (BM; K; type collection),
Harris 7657 (BM ; K ; NY ; UCWI), Harris 6578 (BM ; UCWI), Harris 10029 (BM ; NY ;
UCWI), Harris 6077 (NY; UCWI), D. Watt 8225 (NY; UCWI), Maxon < Kittip 1343
(NY), Proctor 24511 (BM; IJ), Proctor 25610 (BM; IJ).

The discovery in 1894 of Ardisia densiflora in the Blue Mountains of Jamaica, on
which it grows at about 3,000-4,000 ft. (900-1,200 m.), was one of the early results
of the renewed collecting in Jamaica at the end of the nineteenth century. It is a
well-marked species with a densely many-flowered inflorescence, the branches of
which are ascending, almost fastigiate, and which bear up to 25 small white flowers
on pedicels 1-6 mm. long. The filaments of the stamens are much longer than the
longitudinally dehiscent anthers. According to Lundell, A. spicigera from Mexico
and A . gentlei from British Honduras are both conspecific with A . densiflora, which
thus has a fairly wide but sparse Central American distribution comparable with
that of Achimenes erecta (Lam.) H. P. Fuchs, a gesneriad restricted in Jamaica to the
Blue Mountains.

6. Ardisia byrsonimae Stearn, sp. nov. (Text-fig. 23A-E; PI. 8b).

Arbor parva ad 4-5 m. alta, ramulis hornotinis ferrugineis pulverulentisque,
annotinis cinereis glabrisque. Folia pro genere parva, ad apices ramulorum conferta,
1-5 mm. inter se distantia, ascendentia, brevipetiolata ; lamina obovata, apice
rotundata interdum emarginata, margine integra, basi cuneata, c. 3-5-5 cm. longa,
I '3~3'5 cm - lata, glabra, utrinque punctata, coriacea, venis vix prominulis; petiolus
25 mm. longus. Inflorescentiae terminales fere sessiles multiflorae densae, c. 3 cm.
longae et latae, primo pulverulentae, ramulis ascendentibus ad 3 cm. longis racemose
3-6-floris; pedicelli 1-4 mm. longi. Calyx c. 3 mm. longus; sepala imbricata, ovata,
rotundata, c. 2-2-5 mm. longa, 1-5-2 mm. lata, margine scariosa ciliataque. Corolla
rosea, c. 7-8 mm. longa; petala in 1/4 altitudinis coalita, apice acuta, parte libera
lanceolata, 2-2-5 mm. lata, per anthesin recurvata, punctata. Stamina quam petala
breviora, c. 1-5-2 mm. supra corollae basin inserta; antherae anguste triangulares,
c. 3-5-5 mm. longae, apice mucronatae et rimis brevibus quasi poratim dehiscentes,
basi cordatae; filamenta c. 2 mm. longa. Ovarium late ovoideum; stylus c. 8-9 mm.
longus, stigmate minuto. Fructus transverse late ellipsoideus, c. 4-5 mm. altus,
5-6 mm. latus, glanduloso-punctatus, styli basi c. I mm. longa coronatus.

CLARENDON: Peckham Woods, 760 m., Dec. 1917, Harris .12799 (BM; NY; UCWI),
same locality, Aug. 1954, Webster & Proctor 5428 (BM), same locality, Jan. 1955,
Proctor gj43 (BM; IJ), same locality, Dec. 1955, Stearn 9 (BM, holotype; K), Proctor
11408 (IJ).

A SYNOPSIS OF JAMAICAN MYRSINACEAE

163

J.C.W

FIG. 23. Ardisia byrsonimae Stearn: A, fruiting specimen (Steam 9), x ; B, buds (Proctor
9743), X 2; C, dissected flower (Stearn 9), x 3; D, anther (Stearn 9), x 5; E, fruit
(Stearn g], x 4. Ardisia densiflora Krug & Urban: F, dissected flower (Harris 10029),
X 4. Ardisia tinifolia Swartz; G, dissected flower (Harris 12081], x 3; H, placenta
with ovules (Harris 12081], x 20. Ardisia solanacea Roxb. ; I, dissected flower (Harris
11637), calyx and ovary x 1-5, corolla and stamens x 2.

Ardisia byrsonimae is a slow-growing tree endemic to wooded limestone hill-tops
in upper Clarendon, Jamaica. It differs from other Jamaican species in its compara-
tively small leaves crowded at the ends of the shoots and its contracted inflorescences
with very short pedicels. From A. tinifolia var. 3 (A. troy ana Urban), likewise
inhabiting the western interior of Jamaica, it may also be distinguished by its

164 A SYNOPSIS OF JAMAICAN MYRSINACEAE

markedly imbricate and more rounded sepals. The epithet byrsonimae was given
to this species by N. L. Britton under Icacorea but never published by him. It
refers to a resemblance in habit between it and certain West Indian species of
Byrsonima (Malpighiaceae), notably B. lucida (Mill.) Rich., with small obovate
leaves usually clustered at the ends of the shoots.

\
Subgenus 2. ICACOREA (Aublet) Pax

ARDISIA subgen. ICACOREA (Aublet) Pax in Engler & Prantl, Nat. Pflanzenfam.
4 (i): 94(1889). Mez in Urban, Symb. Antill. 2 : 391(1901).

Icacorea Aublet, Hist. PI. Guiane Franc. 2, Suppl. : i (1775).

Ardisia sect. Icacorea (Aublet) Griseb., Fl. Brit. W. Ind. Is. : 395 (1861).

Type (by monotypy): Ardisia guianensis (Aublet) Mez (Icacorea guianensis Aublet).
Inflorescences terminal or both axillary and terminal; branches compoundly
racemose, bearing several umbellate or corymbose clusters. Stamens with short
filaments. Species all American.

7. Ardisia compressa Kunth, Nov. Gen. & Sp. PI. 3 : 245 (1819). Mez in Urban,
Symb. Antill. 2 : 393 (1901); in Engler, Pflanzenr. IV. 236 : 89 (1902). C. L.
Lundell in Fieldiana, Bot. 24 (8) : 139 (1966).

Ardisia decipiens sensu Griseb., Fl. Brit. W. Ind. Is. : 395 (1861); non A. DC.
Icacorea compressa (Kunth) Standley in Contrib. U.S. Nat. Herb. 23 : mo (1924).

HANOVER: Proctor 10013 (BM; IJ), Harris 10253 (BM; K; NY), Britton & Rollick
2129 (NY), Britton 6- Rollick 2164 (NY).

WESTMORELAND : Proctor & Huttings 22060 (BM ; I J) .

ST. JAMES: Proctor 16166 (BM; IJ).

Ardisia compressa ranges from Mexico to northern South America, including
Trinidad, but is not recorded from any other West Indian island. Humboldt and
Bonpland collected the type near Caripe, Venezuela. Most continental specimens
have the apex of the leaf more gradually tapered than have Jamaican specimens,
which approximate most to some from British Honduras. The Jamaican specimens
are all from the west of the island. The British Museum (Natural History) possesses
an unlocalized specimen from Jamaica collected by William Wright who was in
Jamaica in 1765-76 and 1782-85, living most of the time at Hampden, in Trelawny,
and Mez suggests that this was of cultivated origin, which, however, seems unlikely.
It is easily distinguished from other Jamaican species by its more intricately branched
inflorescence, the lateral peduncles of which bear three or four secondary peduncles
each with two to six small, whitish, brown-flecked flowers; the filaments of the
stamens are about 0-5 mm. long, and much shorter than the anthers.

Subgenus 3. TINUS Mez

ARDISIA subgen. TINUS Mez in Engler, Pflanzenr. IV. 236 : 124 (1902).
Lectotype: Ardisia elliptica Thunb. (A. humilis sensu Mez, pro parte; non Vahl).

A SYNOPSIS OF JAMAICAN MYRSINACEAE 165

Inflorescences axillary. Stamens with very short filaments. Species Asiatic and
Australian, one often naturalized elsewhere.

8. Ardisia solanacea Roxb., PL Coast Coromand. I : 27, t. 27 (1795). Sims in
Curt. Bot. Mag. 40 : t. 1677 (1814). Mez in Urban, Symb. Antill. 2 : 402 (1901);
in Engler, Pflanzenr. IV. 236 : 132 (1902). E. H. Walker in Philipp. Journ.
Sci. 73 : 58 (1940). (Text-fig. 23 I.)

Ardisia humilis sensu A.DC. in DC., Prodr. 8 : 129 (1844); non Vahl. Wight, Ic. PL Ind.

Br. 4 : t. 1212 (1848). Griseb., Fl. Brit. W. Ind. Is. : 396 (1861). C. B. Clarke in Hook.

f., Fl. Brit. Ind. 3 : 529 (1882). De Wild., Ic. Select. Hort. Thenen. i : 79, t. 19 (1900).
Icacorea solanacea (Roxb.) Britton, Fl. Berm. : 284 (1918).

ST. JAMES : Proctor 16461 (BM ; I J) .

ST. MARY: Harris 11637 (NY; UCWI), Proctor 20615 (BM; IJ), Adams 12076
(UCWI).

ST. ANDREW: Grabham (NY), Adams 10094 (BM).

PORTLAND: Powell 717 (BM; IJ; NY), Proctor 16546 (BM; IJ; NY), Proctor 19723
(BM; IJ; NY), Marble 852 (NY), Yuncker 18826 (BM), Adams 11482 (BM; UCWI).

ST. THOMAS: Nichols 191 (NY), Powell 854 (IJ), Harris 6- Britton 10585 (NY;
UCWI), Proctor 9217 (BM; IJ), Webster & Wilson 5218 (BM; IJ), Howard, Proctor &
Steam 14809 (BM; IJ), Proctor 16450 (BM; IJ), Yuncker 17523 (BM), Yuncker
18826 (BM; UCWI), Harris 6040 (NY), Britton 3555 (NY), Adams 10471 (UCWI).

A native of India, Malaya and China, long cultivated in tropical gardens, this has
become naturalized here and there in Jamaica, usually along streams or on shaded
road banks, and is sometimes called " Blackberry " or " Craingcraing ". It is
easily recognized by its axillary few-flowered inflorescences with comparatively
large pale rose flowers clustered almost umbellately at the end of the long peduncle.

2. WALLENIA Swartz

WALLENIA Swartz, Nov. Gen. & Sp. PL : 2, 31 (1788); Fl. Ind. Occ. i : 247 (1797);
nomen conservandum No. 6304.

Petesioides Jacq. ex Kuntze, Revis. Gen. PL 2 : 402 (1891), nom. superfl.

Type (by monotypy) : W. laurifolia Swartz (Petesioides laurifolium Jacq.)

Wallenia belongs to the tribe Myrsineae characterized by having the ovules few
and arranged in a single series around the placenta. Both it and Myrsine can be
readily distinguished by their small flowers, with the corolla 1-5-6 mm. long, from
Ardisia of the tribe Ardisieae, characterized by having numerous ovules arranged in
several series or apparently scattered over the placenta. The loose many-flowered
inflorescence of Wallenia readily distinguishes it from Myrsine (Rapanea) which has
flowers crowded in axillary clusters. The Jamaican species all belong to the sub-
genus Wallenia (subgenus Euwallenia Mez) which is confined to Jamaica, Cuba and
Hispaniola and has an erect terminal paniculate inflorescence; the other subgenus,
Homowallenia Mez, with pendulous axillary racemose inflorescences, occurs in Cuba,
Hispaniola, Puerto Rico, and the Lesser Antilles but not in Jamaica.

166 A SYNOPSIS OF JAMAICAN MYRSINACEAE

The dioecious nature of Wallenia possibly militates against the formation of well-
defined taxa within the genus ; it certainly makes difficult the correlation of characters
in herbarium material. The corolla of male flowers is much longer than the calyx;
the anthers are borne on long exserted filaments. The corolla of female flowers is
shorter or little longer than the calyx; the filaments are very short and the anthers
aborted. Consequently female flowers are much less conspicuous than male flowers
and are liable to be passed over by collectors as not being fully developed. Thus
there is rarely available, from the same locality, male and female flowering material
and mature fruiting material obviously belonging to the same species, which would
provide the means of correlating other material. Instead some members of the
genus are only known in one of these three possible states. The variability of the
genus, presumably maintained by " out-breeding ", is such that two gatherings
rarely match completely. Mez and Urban accordingly tended to describe each new
gathering as representing a new species distinguished from its congeners by a com-
bination of subtle rather than well-marked characters, the latter being few within
the subgenus Wallenia (Euwallenia Mez) . The lack of comparable material makes it
difficult to assess the worth of such characters. In preparing the present account it
has not been possible to devise a key making full use of floral characters and unfortu-
nately the vegetative characters used tend either to overlap or to be difficult to
express without appearing too definite and exclusive or too vague and hence useless.
Thus while the key should lead to the identification of typical material of the species,
it would be too optimistic to expect conclusive results with all specimens. Wallenia
laurifolia occurs in Cuba, Hispaniola and Jamaica, W. subverticillata in Cuba and
Jamaica; the other species accepted here are apparently endemic to Jamaica.

KEY TO JAMAICAN SPECIES

Leaves with base broadly rounded and slightly cordate, paired or in whorls of three
together, fairly large (3-17 cm. broad), the petiole short (less than 6 mm. long)
or lacking :
Leaves rounded at the apex, 7-17 cm. broad:

Leaves sessile . . . . . . . . . i. W. clusioides

Leaves with 3-6 mm. long petiole . . . . . 2. W. elliptica

Leaves narrowed towards the acute or obtuse apex, 3-6 cm. broad

3. W. subverticillata

Leaves with base cuneate or attenuate, rarely rounded (the petiole then I cm. or
more long), usually alternate, only occasionally paired or whorled, the petiole
short or long :
Leaves reticulate-veined beneath, with slightly raised cross-veins between the main

lateral veins :
Petiole short, c. 1-8 mm. long; blades under n cm. long:

Leaves mostly broadest at the middle and gradually narrowed from there to

the apex:

Petiole very short, 1-4 mm. long; blade inconspicuously punctate beneath

4. W.fawcettii

A SYNOPSIS OF JAMAICAN MYRSINACEAE 167

Petiole longer (5-8 mm. or more long) ; blade conspicuously punctate

beneath 10. W. venosa

Leaves mostly broadest above the middle with the apex rounded or obtuse:
Leaf-blades 2-8 cm. long, less than 3-5 cm. broad; crest of Blue Mountains

5. W. crassifolia
Leaf -blades larger (5-11 cm. long) :

Leaves coriaceous, firm, greyish or glaucescent; female calyx c. 1-4 mm.
long . . . . . . . . 6. W. corymbosa

Leaves chartaceous, thinner:

Leaves drying light green; female calyx c. 1-5 mm. long

IT. W. purdieana
Leaves drying brownish; female calyx I mm. long

7. W. xylosteoides

Petiole longer (c. 8-30 mm. long) ; blades mostly more than n cm. long:
Leaf -blades nearly three times as long as broad, mostly narrowly elliptic :
Base of blade narrowly cuneate or attenuate . . 8. W. erythrocarpa
Base of blade rounded or abruptly contracted . . 12. W. grisebachii
Leaf -blades about twice as long as broad, mostly elliptic or narrowly obovate :
Lateral veins spreading outwards from the midrib at an angle of about 40

9. W. punctulata
Lateral veins spreading outwards from the midrib at an angle of about

50-70:

Male flower with calyx 1-5-2-5 mm. long, corolla 2-5-4 mm - l n g persis-
tent fruiting calyx not more than 1-5 mm. long:

Leaves usually densely blackish-punctate and somewhat mottled

beneath, strongly veined, the base cuneate . 10. W. venosa

Leaves inconspicuously punctate and not mottled beneath, lightly

veined, the base cuneate or rounded or abruptly contracted:
Leaves drying light green ; inflorescence glabrous, loose ; male flower
with corolla yellowish, not punctate; style about 0-8 mm. long

ii. W. purdieana

Leaves drying brownish; inflorescence minutely pubescent, compact;
male flower with corolla reddish, distinctly punctate; style
about 1-5 mm. long . . . . 12. W. grisebachii

Male flower with calyx c. 3mm. long, corolla 5-6 mm. long; persistent

fruiting calyx about 2-5 mm. long . . 13. W. calyptrata

Leaves pinnately veined beneath, i.e. without raised cross-veins between the main

lateral veins :

Blades of some or all leaves abruptly contracted at base into the petiole;

inflorescence usually minutely pubescent . . . 12. W. grisebachii

Blades of all leaves gradually narrowed into the petiole ; inflorescence glabrous :

Blades of leaves mostly less than 12 cm. long, 6 cm. broad; petiole not more

than i cm. long 14. W. laurifolia

Blades of leaves up to 25 cm. long, 11 cm. broad, mostly more than 12 cm.
long, 6 cm. broad; petiole 1-2 cm. long:

168 A SYNOPSIS OF JAMAICAN MYRSINACEAE

Fruits vertically ribbed . . . . . 15. W. sylvestris

Fruits not ribbed . . . . . . . 16. W. discolor

1. Wallenia clusioides (Griseb.) Mez in Urban, Symb. Antill. 2 : 411 (1901); in
Engler, Pflanzenr. IV. 236 : 245 (1902).

Ardisia clusioides Griseb., Fl. Brit. W. Ind. Is. : 396 (1861) excl. specim. Alexander.

HANOVER: Harris 10333 (BM; NY), Proctor 10028 (IJ), Proctor 26590 (IJ).

WESTMORELAND: Pur die (K, lectotype), Harris 10210 (BM; NY).

ST. JAMES: Harris 9x77 (BM; NY).

ST. ELIZABETH: Harris 976 j (NY), Britton mo (NY), Proctor 26530 (IJ).

W. clusioides is a species of western Jamaica notable for its very large sessile and
subcordate leaves 15-31 cm. long, 7-17 cm. broad. It forms a small tree or tall
shrub 6-20 ft. (2-6 m.) high. The female flowers are unknown. The greenish-
yellow male flowers have the calyx 1-5-2 mm. long, the corolla about 2-5 mm. long.

2. Wallenia elliptica Urban, Symb. Antill. 6 : 30 (1909). (PI. ga.)

Ardisia clusioides Griseb., Fl. Br. W. Ind. Is. : 396 (1861) pro parte, quoad specim.
Alexander.

ST. ANN: Alexander 569 (K; NY; type-collection), Britton & Rollick 2776 (NY).

An obscure species described by Urban on a specimen collected by Alexander at
Grierfield near Moneague, St. Ann, W. elliptica is provisionally kept apart from
W. clusioides on account of the shortly petioled leaves. The male flowers are
unknown. The female flowers have the calyx and corolla about 1-5 mm. long.

3. Wallenia subverticillata (Britton) Ekman ex Urban, Symb. Antill. 9 : 412
(1925). Alain in Leon & Alain, Fl. Cuba 4 : in (1957). (Text-fig. 24K.)

Petesioides subverticillatum Britton in Bull. Torrey Bot. Club 37 : 355 (1910).

PORTLAND: Proctor 22102 (IJ).

ST. THOMAS: Britton 3937 (NY), Harris & Britton 10694 (NY, holotype; UCWI),
Harris & Britton 10701 (K; NY), Britton 4049 (NY), Proctor 28681 (BM).

Wallenia subverticillata, a species of mountain woods at 1,000-2,000 ft. (300-
600 m.), is distinguished by having its leaves grouped in pairs or threes; they are
almost sessile and are much narrower and more attenuate towards the apex than
those of W. clusioides but are nevertheless extremely variable in shape and size,
e.g. 5-20 cm. long, 3-5 cm. broad on the same plant. The type was collected by
Harris and Britton on a slope below Big Level, eastern St. Thomas. Rather
surprisingly, specimens of what appear to be the same species have been collected
in Las Villas, Cuba, by Ekman. The calyx of male flowers is 2-2-5 mm - l n g f
female flowers about i mm. long; the corolla of male flowers is about 2-5 mm. long,
of female flowers about i mm. long, slightly longer than the calyx.

4. Wallenia fawcettii Mez in Urban, Symb. Antill. 2 : 408 (1901); in Engler,
Pflanzenr. IV. 236 : 244 (1902). (PI. gb.)

A SYNOPSIS OF JAMAICAN MYRSINACEAE 169

PORTLAND: Maxon & Killip 714 (NY).

PORTLAND or ST. THOMAS: Harris 5422 (NY, type collection).

ST. THOMAS: Proctor 9684 (IJ).

From other relatively small-leaved and short-petioled species, Wallenia fawcettii
may be recognized by its more attenuate leaves, which are gradually narrowed above
the middle to a usually acute apex. It is known only from the crest of the Blue
Mountains, at about 5,000-5,500 ft. (1,500-1,700 m.) ; the type was collected by
Harris near Portland Gap but whether on the Portland or St. Thomas side is un-
certain. The calyx of male flowers is about 1-5 mm. long, of female flowers about
I mm. long; the corolla of male flowers is about 2-5 mm. long, of female flowers
unknown.

5. Wallenia crassifolia Mez in Urban, Symb. Antill. 2 : 409 (1901); in Engler,
Pflanzenr. IV. 236 : 244 (1902). (PL roa.)

Wallenia venosa Griseb., Fl. Brit. W. Ind. Is. : 394 (1861) pro parte, excl. specim. Wilson.

ST. ANDREW: Proctor 9890 (IJ; NY), Proctor 9524 (BM; IJ), Maxon 6- Killip 1026
(NY), Philipson 932 (BM), E. G. Britton 3866 (NY), Purdie (K, holotype).

ST. THOMAS: Webster & Wilson 5445 (IJ), Proctor 9207 (IJ), Steam 101 (BM),
Proctor 11458 (IJ).

Wallenia crassifolia inhabits mountain rain forest along the crest of the Blue
Mountains between 4,750 and 6,500 ft. (1,400-2,000 m.). That the smallness of its
elliptic or narrowly elliptic leaves is not simply a reduction occasioned by its high
mountain environment is shown by the occurrence of the large-leaved and
undoubtedly distinct W. calyptrata in the same habitat. The type-locality of W.
crassifolia is Morces Gap, where it was first collected by William Purdie in 1843.
The calyx of male flowers is about 2 mm. long, of female flowers about I mm. long,
the corolla of male flowers about 4 mm. long, of female flowers unknown.

6. Wallenia corymbosa Urban, Symb. Antill. 5 : 457(1908). (Text-fig. 24A-J.)
TRELAWNY: Harris 8720 (BM; NY; UCWI; type-collection), Harris 9095 (NY;

UCWI).

CLARENDON: Harris 10864 (NY), Harris 10996 (NY), Harris 12781 (BM; NY;
UCWI), Harris 12803 (NY; UCWI).

ST. ANN: Howard & Proctor 15037 (IJ), Howard & Proctor 15170 (BM; IJ), Proctor
26739 (BM).

Typical Wallenia corymbosa inhabits the central hills of Jamaica at about 2,000-
2,500 ft. (600-760 m.), probably extending well into the Cockpit Country, and is
notable for its thick greyish or glaucescent narrowly obovate leaves.

The type-locality is on Crown Lands near Troy, where William Harris collected it
at 2,500 ft. (760 m.) in 1904. Subsequent collections in the same area have provided
an unusually adequate representation of it in male and female flower and in fruit.
The flowers are described as " pale yellow with red spots " or " cream with red

170

A SYNOPSIS OF JAMAICAN MYRSINACEAE

stripes ". The calyx of male flowers is 1-5-2 mm. long, of female flowers about
i '4 mm. long, the corolla of male flowers 3-4 mm. long, of female flowers about
2-5 mm. long. The fruits are about 3-5 mm. broad.

JCW

FIG. 24. Wallenia corymbosa Urban: A, female flowering specimen (Harris 12803), x ^;
B, female flower (Harris 12803), X 6; C, dissected calyx and gynoecium of female flower
(Harris 12803), x 8; D, dissected corolla of female flower with aborted stamens (Harris
12803), x 8; E, placenta of female flower (Harris 12803), x 20; F, male flower (Harris
12819), x 8; G, dissected corolla of male flower with aborted gynoecium (Harris 12819),
x 8; H, anther of male flower (Harris 12819), x 20; I, fruit (Harris 12781), X 5; J, leaf
(Harris 12781), x . Wallenia subverticillata (Britton) Ekman: K, leaves (Harris d>-
Britton 4049), x J. Wallenia grisebachii Mez : L, leaf-base (Proctor 8247), x \.

A SYNOPSIS OF JAMAICAN MYRSINACEAE 171

7. Wallenia xylosteoides (Griseb.) Mez in Urban, Symb. Antill. 2 : 409 (1901);
in Engler, Pflanzenr. IV. 236 : 244 (1902). (PI. lob.)

Ardisia xylosteoides Griseb., Fl. Br. W. Ind. Is. : 395 (1861) tantum quoad pi. Jamaic.

ST. ANN: Alexander 568 (K, lectotype).

Grisebach reported his Ardisia xylosteoides from " Haiti! ; Mexico! ; New Granada! "
as well as Jamaica; these records cannot apply all to one species. Mez in 1901
restricted the epithet xylosteoides to Jamaican material collected by Alexander on
St. Ann's Road, Moneague. The same species is represented by Jamaican Plants
1118 (UCWI) and 1141 (K) without locality. The calyx of female flowers is about
i mm. long, the corolla scarcely longer. Male flowers and fruit are unknown. The
epithet indicates a resemblance in leaf-form to Lonicera xylosteum L., the Xylosteum
of pre-Linnaean authors.

8. Wallenia erythrocarpa Urban, Symb. Antill. 6 : 29 (1909). (PL na.)

HANOVER: Harris 10343 (BM; NY; UCWI; type-collection).

WESTMORELAND: Harris 7080 (UCWI).

This species is still only known from material collected by Harris, the type from
Fray Woods at 1,650-1,800 ft. (500-550 m.) in 1908; he gathered other material at
Grandvale at 500 ft. (150 m.) in 1898. It is notable for its large but relatively narrow
long-petioled leaves with prominent reticulate veining, their narrowly elliptic-oblong
shape and their size (11-21 cm. long, 4-5-7-5 cm. broad) distinguishing it from W.
venosa. The calyx in fruit is about i mm. long. Male and female flowers are un-
known.

9. Wallenia punctulata Urban, Symb. Antill. 7 : 322 (1912). (PL lib.)

MANCHESTER: Britton 3281 (NY; type collection).

Wallenia punctulata is another little-known species, known only from the type-
collection gathered by N. L. Britton in 1908 between Brown's Town and Porus, but
distinguished from other species with prominently reticulate-veined leaves by its
more oblique lateral veins which pass outwards from the midrib at an angle of about
40. The calyx in fruit is about 2 mm. long, the corolla of female flowers according
to Urban about 1-7 mm. long.

10. Wallenia venosa Griseb., FL Brit. W. Ind. Is. : 394 (1861) quoad specim.

I Wilson. Mez in Urban, Symb. Antill. 2 : 410 (1901); in Engler, Pflanzenr. IV.
236 : 245 (1902). Urban, Symb. Antill. 7 : 322 (1912). (PL I3a.)
HANOVER: Britton 2425 (NY), Britton 2292 (NY), Britton 2326 (NY).
WESTMORELAND : Harris 10248 (BM ; NY) .
TRELAWNY: Harris 8768 (BM; NY; UCWI).
ST. ELIZABETH: Harris 9925 (BM; K; NY; UCWI).
MANCHESTER: Britton 3712 (NY).
CLARENDON: Proctor 18400 (BM; IJ), Proctor 26514 (BM; IJ), Proctor 27637 (BM;

U).

ST. ANN : Proctor 26514 (BM) .

172 A SYNOPSIS OF JAMAICAN MYRSINACEAE

ST. ANDREW: Harris 5633 (BM; NY; UCWI), Harris 6126 (BM), Harris 16032
(BM; NY; UCWI), Adams 11845 (UCWI).

ST. THOMAS: Harris 6- Britton 10745 (NY), Britton 3963 (NY), Britton 3988 (NY),
Britton 4145 (NY) .

The lectotype of this widespread species is Wilson 388 (GOET, K), collected in
1858 without specified locality, which was designated by Mez in 1901. Wallenia
venosa has prominently reticulate-veined leaves with black dots beneath sometimes
so numerous as to give it a somewhat mottled appearance. The calyx of male
flowers is 1-5-2 mm. long, of female flowers 0-8 mm. long, the corolla of male flowers
about 3 mm. long, of female flowers about I mm. long. The Alexander specimen
(GOET) from the Blue Mountains cited by Grisebach under W. venosa belongs to
W. crassi folia.

n. Wallenia purdieana Mez in Urban, Symb. Antill. 2 : 408 (1901); in Engler,
Pflanzenr. IV. 236 : 243 (1902). (PI. I2a.)

HANOVER: Harris 10305 (BM; UCWI), Proctor 11301 (IJ).

ST. JAMES: Proctor 22992 (BM; IJ).

TRELAWNY: Harris 9096 (UCWI), Britton 510 (NY), Proctor 15663 (BM; IJ),
Howard, Proctor ( Steam 14663 (BM; IJ).

MANCHESTER: Purdie (K, type), Proctor 16118 (BM; IJ), Howard, Proctor 6-
Wagenknecht 20520 (NY), Harris & Britton 10601 (BM; NY; UCWI), Robertson
5416 (UCWI).

Wallenia purdieana is a species of western and central Jamaica between 1,000 ft.
(300 m.) and 3,000 ft. (900 m.). The type was collected by William Purdie in
Manchester, but without specified locality. The leaves on drying retain their green
colour unlike those of other species which become brownish or grey. The calyx of
male flowers is about 2 mm. long, of female flowers about 1-5 mm. long, the corolla
of male flowers about 3 mm. long, of female flowers about 1-5-2 mm. long.

12. Wallenia grisebachii Mez in Urban, Symb. Antill. 2 : 411 (1901) quoad specim.
Wullschlagel 1338; in Engler, Pflanzenr. IV. 236 : 245 (1902). Urban, Symb.
Antill. 6 : 31 (1909). (Text-fig. 24!.; PL I2b.)
Wallenia laurifolia sensu Griseb., Fl. Brit. W. Ind. Is. : 394 (1861) pro parte; non Swartz.

ST. JAMES: Proctor 22154 (BM; IJ).

ST. ELIZABETH: Howard & Proctor 13745 (IJ; NY), Howard 6- Proctor 13755 (IJ).

TRELAWNY: Perkins 1385 (K).

MANCHESTER: Wullschlagel 1338 (Munich, lectotype), S. Brown 300 (NY), Robertson
5470 (UCWI), Adams 10116 (UCWI), Adams 11766 (UCWI).

CLARENDON : Proctor 8247 (I J) .

ST. ANN: Britton 6- Rollick 2776 (NY; UCWI).

The species W. grisebachii as accepted here consists of plants with fairly large
coriaceous leaves having the rounded base contracted abruptly into the petiole and
becoming brownish on drying. Mez separated W. grisebachii from among material
included by Grisebach in W. laurifolia. Urban further divided this, restricting the
name W. grisebachii to the taxon represented by Wullschlagel 1338 collected at

A SYNOPSIS OF JAMAICAN MYRSINACEAE 173

Fairfield, Springfield, in Manchester parish, between 1847 an d 1849. The inflores-
cence is compact; the flowers are rusty brown. The calyx of male flowers is 2 mm.
long, of female flowers 1-5-2 mm. long, the corolla of male flowers about 4 mm. long,
of female flowers about 1-5 mm. long.

13. Wallenia calyptrata Urban, Symb. Antill. 5 : 458 (1908); in Fedde, Repert.
Sp. Nov. 13 : 469 (1915).

ST. ANDREW: Shreve s.n. (NY), /. A. Harris 6- Lawrence 15490 (NY), Adams 7440
(UCWI).

PORTLAND: Proctor 26717 (BM; IJ).
ST. THOMAS: Proctor 9638 (IJ), Proctor 9683 (IJ; NY), Stearn 104 (BM), Proctor

H457 (IJ).

Although the type-gathering of W. calyptrata made by Rehder in 1903 at Morces

Gap has not been seen, the agreement of Urban's description with the other material
collected along the crest of the Blue Mountains leaves no doubt as to its identity.
Its range here seems to be between 5,000 ft. (1,500 m.) and 7,000 ft. (2,100 m.) in
mountain mist-forest. It has prominently reticulate-veined leaves 8-14 cm. long,
3-5-6 cm. broad, and larger male flowers than other species. The calyx of male
flowers is 3-3-5 mm. long, of female flowers 2-5 mm. long, the corolla of male flowers
5-6 mm. long, of female flowers 3 mm. long.

14. Wallenia laurifolia Swartz, Nov. Gen. & Sp. PI. : 31 (1788); Fl. Ind. Occ.
1:248 (1797). Mez in Urban, Symb. Antill. 2 : 407 (1901); in Engler,
Pflanzenr. IV. 236 : 243 (1902). Urban, Symb. Antill. 8 : 521 (1921). Alain
in Leon & Alain, Fl. Cuba 4 : no, fig. 42 (1957).

Petesioides laurifolium Jacq., Select. Stirp. Amer. Hist. : 17 (1763), nom. invalid.
Wallenia clusiifolia Griseb., Fl. Brit. W. Ind. Is. : 394 (1861).

" Bryonia nigra fruticosa, foliis laurinis, floribus racemosis speciosis " Sloane, Cat. PI. Jam. :
106 (1696); Voy. Jam. Nat. Hist, i : 234, t. 145 f. 2 (1707).

HANOVER : Proctor 10035 (BM ; I J) .

MANCHESTER: S. Brown 138 (NY), Proctor 24175 (BM; IJ), Crawford 775 (NY),
Farr or Sanderson 22334 (IJ), Harris or Britton 10613 (NY; UCWI), Howard or
Proctor 14964 (IJ), Howard, Proctor or Wagenknecht 20520 (BM; IJ), Harris 12862
(BM; NY), Howard, Proctor or Stearn 14708 (BM; IJ).

CLARENDON: Harris 11188 (NY; UCWI).

rST. ANN: Stearn 603 (BM), Proctor 11907 (IJ), Proctor 8632 (IJ), Proctor 16142 (IJ).
ST. CATHERINE: Britton 2616 (NY).
ST. MARY: Sloane, Herb. IV. 494 (BM), Proctor 7554 (NY; IJ).
ST. ANDREW: Harris 6880 (BM; NY; UCWI).
PORTLAND: Stearn 538 (BM), Proctor 11841 (IJ), Wight 195 (NY), Maxon or
Kittip 830 (NY), Harris or Britton 10767 (K; NY), Proctor 11347 ( BM i U). Proctor
11352 (IJ), Proctor 16259 (IJ), Howard, Proctor or Stearn 14758 (BM; IJ), Stearn 226
(BM; UCWI), Britton 4121 (NY).
ST. THOMAS: Britton 4109 (NY), Powell 1647 (BM; IJ).

I 7 4 A SYNOPSIS OF JAMAICAN MYRSINACEAE

This widespread species, first described from Hispaniola and also known from
Cuba, is mainly distinguished by its relatively inconspicuous veining with the areas
between the main lateral veins flat, not reticulate with prominent veinlets. The
material from the upper part of the John Crow Mountains exemplified by Proctor 9822
gathered at 1,500-2,500 ft. (450-750 m.), with very small leaves (to 6 cm. long) and
reduced inflorescence, differs much in appearance from that of the coast, exemplified
by Steam 226, but these differences are assumed to result from the diversity of
habitat, as small leaves approaching these are to be found on some specimens from
elsewhere referred to this species, e.g. Howard, Proctor & Steam 14708 from
Manchester, Proctor 8632 from St. Ann, and Proctor 10035 from Hanover, and inter-
mediates occur moreover on the John Crow Mountains. The calyx of male flowers
is 1-5-2-5 mm. long, of female flowers about 1-5 mm. long, the corolla of male flowers
2-5-3 mm. long, of female flowers about 1-5 mm. long.

15. Wallenia sylvestris Urban in Fedde, Repert. Sp. Nov. 13 : 468 (1915).
(PI. I3b.)

PORTLAND: Proctor 10477 (Ui NY), Proctor 16258 (IJ), Proctor 5266 (BM; IJ),
Proctor 5244 (IJ).

PORTLAND or ST. THOMAS: Harris 6- Britton 10720 (NY, type-collection), Harris &
Britton 10690 (K; NY; UCWI), Harris 6- Britton 10776 (NY; UCWI).

This species is apparently confined to mountain woods on the John Crow Mountains
at 1,500-2,500 ft. (450-750 m.). The type-locality is the southern end of the John
Crow Mountains near the boundary of the parishes of Portland and St. Thomas. It
is notable for its large long-petioled leaves with elliptic or narrowly elliptic (rarely
obovate or narrowly obovate) blades up to 22 cm. long, 9 cm. broad. The calyx and
corolla of female flowers are about 2 mm. long. The male flowers are unknown.
The subglobose fruits are slightly punctate and vertically ribbed.

16. Wallenia discolor Urban, Symb. Antill. 6 : 29 (1909).

ST. ANN: Alexander (fide Urban).

Described by Urban from fruiting material collected by Alexander, probably in
St. Ann but without recorded locality, this appears to be distinguished from W.
laurifolia by its larger leaves (18-25 cm - l n g)> shorter pedicels (0-5-3 mm. long) and
coarsely punctate but not ribbed fruits. Male and female flowers are unknown.

3. MYRSINE L.

MYRSINE L., Sp. PI. i: 196(1753), Gen. PI. ed. 5: 90(1754).
Type (by monotypy) : M. africana L.

The genus Myrsine as accepted here includes the groups kept apart by Mez as
Myrsine, Rapanea and Suttonia.

KEY TO JAMAICAN SPECIES

Young shoots and petioles minutely pubescent ; leaf-blades mostly narrowly elliptic
and mostly less than 2-5 cm. broad; corolla lobes joined for the lower 0-5 mm.
or more . . I. M. coriacea

A SYNOPSIS OF JAMAICAN MYRSINACEAE 175

Young shoots and petioles glabrous ; leaf -blades mostly obovate or narrowly obovate
and mostly more than 2-5 cm. broad; corolla lobes free to the base

2. M. acrantha

1. Myrsine coriacea (Swartz) R. Br. ex Roem. & Schult. in L., Syst. Veg., ed.
nov. 4 : 511 (1819). A. DC. in DC., Prodr. 8 : 99 (1844). (Text-fig. 25E-H.)

Samara coriacea Swartz, Nov. Gen. & Sp. PI. : 32 (1788); Fl. Ind. Occ. i : 261 (1797).

Caballeria ferruginea Ruiz & Pa von, Syst. Veg. Fl. Peruv. & Chil. i : 280 (1798).

Myrsine ferruginea (Ruiz & Pavon) Spreng. in L., Syst. Veg., ed. 16, i : 664 (1825).

Myrsine laeta sensu Griseb., Fl. Brit. W. Ind. Is. : 392 (1861); non A.DC.

Rapanea coriacea (Swartz) Mez in Urban, Symb. Antill. 2 : 428 (1901) ; in Engler, Pflanzenr.
IV. 236 : 380 (1902).

Rapanea ferruginea (Ruiz & Pavon) Mez in Urban, Symb. Antill. 2 : 429 (1901); in Engler,
Pflanzenr. IV. 236 : 381 (1902). Urban, Symb. Antill. 8 : 522 (1921). Britton & Wilson
in Sci. Surv. Puerto Rico 6 : 61 (1925). Alain in Le6n & Alain, Fl. Cuba 4 : 112, fig. 43
(1957). J. F. Macbr. in Publ. Field Mus. Nat. Hist., Bot. 13 (5) : 168 (1959). Little
& Wadsworth, Common Trees of Puerto Rico : 432, fig. 203 (1964).

ST. JAMES : Norman 50 (BM) .

TRELAWNY: Webster, Proctor & Powell 5355 (BM; IJ), Britton 486 (NY), Adams
12413 (UCWI).

MANCHESTER: Proctor 21920 (BM; IJ), Britton 3216 (NY), Adams 8466 (UCWI),
Proctor 23118 (BM; IJ).

CLARENDON: Proctor 9723 (IJ), Proctor 8427 (IJ), Proctor 15905 (BM; IJ; UCWI).

ST. ANN : Proctor < Howard 14992 (I J) , Howard, Proctor & Steam 14607 (BM ; I J) ,
Proctor 74? i (BM; IJ), Skelding 3016 (UCWI), Adams 12698 (UCWI).

ST. CATHERINE: Howard & Proctor 13599 (U)-

ST. ANDREW: Adams 12675 (UCWI), Britton 147 (NY).

PORTLAND: Philipson 933 (BM), Proctor 10110 (IJ; NY), Howard, Proctor &
Steam 14767 (BM; IJ).

ST. THOMAS: Webster & Wilson 5454 (IJ), Steam in (BM), Proctor 11463 (IJ),
Maxon 9442 (US), Proctor 1108 (IJ), Steam 505 (BM).

A widely ranging species, recorded in the West Indies from Cuba, Hispaniola,
Jamaica, Puerto Rico and the Lesser Antilles and on the American mainland from
Mexico to Peru, Chile, Bolivia and Argentina, this is commonly known as Rapanea
ferruginea, a name based on Peruvian material; Swartz's Samara coriacea, based on
Jamaican material being, in my opinion, conspecific, his epithet is adopted here.
It is sometimes called " colic wood " in Jamaica, according to Proctor.

2. Myrsine acrantha Krug & Urban in Notizbl. K. Bot. Gart. Mus. Berlin i : 79
(1895); in Urban, Symb. Antill. i : 380 (1899). (Text-fig. 25A-D.)

Myrsine coriacea sensu Griseb., Fl. Brit. W. Ind. Is. : 392 (1861); non (Swartz) R. Br.
Rapanea guianensis Mez in Urban, Symb. Antill. 2 : 431 (1901); in Engler, Pflanzenr. IV.

236 : 392 (1902) pro parte, quoad pi. Jamaic. ; non Aubl.
Rapanea acrantha (Krug & Urban) Mez in Urban, Symb. Antill. 2 : 433 (1901); in Engler,

Pflanzenr. IV. 236 : 376 (1902).

176

A SYNOPSIS OF JAMAICAN MYRSINACEAE

JCfff

FIG. 25. Myrsine acrantha Krug & Urban: A, fruiting specimen (Robertson 5372), x ;
B, flower (Harris 8295), x 10; C, dissected flower with aborted gynoecium (Harris 8295),
X 10; D, fruit (Harris 7061), x 5. Myrsine coriacea Swartz: E, leaf (Steam ill), X \',
F, leaf (Steam 505), x ; G, dissected corolla with stamens (Steam in), X 10; H,
dissected calyx with gynoecium (Steam in), X 10.

WESTMORELAND : Proctor 6- Mullings 22061 (BM ; I J) .

ST. ELIZABETH: Britton 1320 (NY), Proctor 7754 (BM; IJ), Howard & Proctor
13682 (IJ).

MANCHESTER: Howard, Proctor 6- Wagenknecht 20521 (BM; NY; UCWI), Robertson
5372 (BM; K; UCWI), Harris 8295 (BM; NY; UCWI), Proctor 21908 (BM; IJ),
Proctor 11619 (BM; IJ), S. Brown 159 (NY), Adams 10114 (UCWI), Proctor 23130

(IJ).

A SYNOPSIS OF JAMAICAN MYRSINACEAE 177

ST. ANDREW: Harris 6072 (NY), Harris 3398 (K; NY; UCWI; type-collection of
Myrsine acranthd), Harris 5398 bis (BM; UCWI), A. von der Porten 2035 (U)> Harris
.5528 (NY; UCWI), Loveless 1502 (IJ; UCWI).

PORTLAND: Proctor 19737 (BM; NY; UCWI).

ST. THOMAS: Proctor 9580 (BM; IJ).

Myrsine acrantha has hitherto been regarded as a rare and obscure species, known
only from the type and topotype collections gathered by Harris in 1894 and 1895
at Strawberry Hill Cottage high in the Blue Mountains of St. Andrew. Mez distin-
guished it by its free petals from the two other Jamaican species accepted by him
under the names Rapanea ferruginea and R. guianensis. Thus according to his own
generic key in the Pflanzenreich IV. 236 : 15 (1902) it should be referred to the
otherwise Pacific Ocean group Suttonial In fact all the Jamaican specimens hitherto
referred to Myrsine (or Rapanea} guianensis agree with M. acrantha in having obovate
or narrowly obovate leaves, flowers borne on very short bract-covered lateral shoots,
and free petals. Myrsine guianensis (Aubl.) Kuntze does not occur in Jamaica; it
was described by Aublet in 1775 (Hist. PI. Guiane Fran9. i : 121 ; 3 : t. 46) as Rapanea
guianensis from French Guiana and is accepted by Mez as ranging from Florida over
the West Indies to the Guianas and over Mexico and Central America to Bolivia.
Typical M. guianensis as represented by specimens from French Guiana (Aublet!
Sagot 913! Martin!} is notable for its large obovate leaves up to 10 cm. long, 7 cm.
broad, with a glossy almost varnished upper surface and distinctly punctate lower
surface; plants of similar character occur in Bolivia, Colombia, British Guiana, and
Surinam. Very similar to this but distinguishable by the leaves being marked with
short lines beneath is M. trinitatis A. DC. occurring in Trinidad, Tobago, Grenada,
Martinique and Puerto Rico. Plants from Hispaniola, Cuba, Jamaica, the Bahamas
and Florida stand apart from these in their smaller leaves, rarely more than 8 cm.
long, 2-5 cm. broad, less glossy above, punctate but not lined beneath. Whether
more than one taxon is represented in this mostly sterile material is uncertain.
Some specimens from Cuba, e.g. Ekman 15603 (Oriente province), and from Haiti,
e.g. E. C. Leonard 8456 (Departement du Nord) in their leaves and short spurs
unfortunately lacking flowers, closely resemble the Jamaican M. acrantha. In the
Bahamas and Florida the plants of this alliance as represented by A . Curtiss 44 from
New Providence, Bahamas, and R. A. Howard 8047 and Tracy 7454 from Florida
have their five corolla-lobes fused at the base for about 0-4-0-5 mm., as figured by
Small (Man. Southeast Fl. : 1029 (1933)). They represent a species distinct from
M . acrantha for which the correct name appears to be Myrsine punctata.*

* Myrsine punctata (Lam.) Steam, comb. nov.

Sideroxylum punctatum Lam., Tabl. Encyc. & Meth., Bot. 2: 42 (1794) [from Carolina, Michaux].

Bumelia punctata (Lam.) Roem. & Schult. in L., Syst. Veg., ed. nov. 4 : 498 (1819).

Myrsine floridana A. DC. in Trans. Linn. Soc. London 17 : 107 (1834); in DC., Prodr. 8 : 98 (1844)

[from Florida].
Rapanea guyanensis sensu Small, Man. Southeast Fl. : 1029, fig. (1933); non Aublet.

SPECIAL LITERATURE

GRISEBACH, A. H. R. 1859-64. Flora of the British West Indian Islands. London.
HOSAKA, E. Y. 1940. A revision of the Hawaiian Species of Myrsine (Suttonia, Rapanea),
Myrsinaceae. Occ. Pap. Bernice P. Bishop Mus., Honoluhi. 16 : 25-79.

178 A SYNOPSIS OF JAMAICAN MYRSINACEAE

LUNDELL, C. L. 19660,. Flora of Guatemala: Myrsinaceae. Fieldiana, Bot. 24 (8) : 135-200.
19666. The genus Parathesis of the Mvrsinaceae. Contrib. Texas Res. Found. Bot. Studies,

5. Renner, Texas.
MEZ, C. 1901. Myrsinaceae. Urban, Symb. Antill. 2 : 389-433.

1902. Myrsinaceae. Engler, Pflanzenreich IV. 236 (Myrsinaceae).

STEARN, W. T. 1965. Grisebach's Flora of the British West Indian Islands: a biographical

and bibliographical introduction. Journ. Arnold Arb. 46 : 243-285.
SWARTZ, O. 1797. Flora Indiae occidentalis i : 247-249, 468-469.
URBAN, I. 1898. Bibliographia Indiae occidentalis bo tanica. Urban, Symb. Antill. i : 3-195.

1909. Incrementa siphonogamarum florae jamaicensis. Urban, Symb. Antill. 6 : 70-131.

WALKER, E. H. 1940. A revision of Eastern Asiatic Myrsinaceae. Philipp. Journ. Sci.

73 : 1-258.

IQ54- Concerning the Myrsinaceae (" Ardisiaceae ") of Japan. Bot. Mag. Tokyo 67 :

105-111, 155-162, 203-213, 247-255.

Dr. WILLIAM T. STEARN, D.Sc., Sc.D.

Dept. of Botany

BRITISH MUSEUM (NATURAL HISTORY)

CROMWELL ROAD, LONDON, S.W.7

Bull. Br. Mm. nat. Hist. (Bot.) 4, 4

PLATE 6

ARDISlATlnifotia

Ardisia tinifolia Swartz; original drawing by Swartz.

BOT. 4, 4

Bull. Br. Mus. nat. Hist. (Bot.) 4, 4

PLATE 7

Ardisia tinifolia Swartz; original drawing by Swartz representing A. coriacea Swartz.

Bull. Br. Mus. nat. Hist. (Bot.) 4, 4

PLATE 8

a 5,

\

Bull. Br. Mus. nat. Hist. (Bot.) 4, 4

PLATE 9

J

\

Bull. Br. Mus. nat. Hist. (Bot.) 4, 4

PLATE 10

18
8

8

~

_~

-~

e 03

\

Bull. Br. Mus. nat. Hist. (Bot.) 4, 4

PLATE ii

\

Bull. Br. Mus. nat. Hist. (Bot.) 4, 4

PLATE

-*
12

<L>

\

Bull. Br. A/MS. nat. Hist. (Bot.) 4, 4

PLATE 13

*t

u-

II,

O

^

r^

.

X

*

i i *

!I1 ^41

i'$ t\*

141

i j

^

'or uo,

*t\A il %

_?

**

. ("V ,

r~ *

i;

<

s? *

05

1

- -i ^\

*' >

<a

,

;}

I

4->

C

O

\

BY ADLARD & SON LIMITED
BARTHOLOMEW PRESS, DORKING

THE JAMAICAN SPECIES OF

COLUMNEA AND ALLOPLECTUS

(GESNERIACEAE)

WILLIAM T. STEARN

BULLETIN OF
THE BRITISH MUSEUM (NATURAL HISTORY)

BOTANY Vol. 4 No. 5

LONDON : 1969

THE JAMAICAN SPECIES OF COLUMNEA AND
ALLOPLECTUS (GESNERIACEAE)

BY

WILLIAM THOMAS STEARN

. \ y .

British Museum (Natural History)

Pp. 179-236; 29 Text-figures, Plates 14-21

BULLETIN OF

THE BRITISH MUSEUM (NATURAL HISTORY)
BOTANY Vol. 4 No. 5

LONDON 11969

THE BULLETIN OF THE BRITISH MUSEUM

(NATURAL HISTORY), instituted in 1949, is
issued in five series corresponding to the Departments
of the Museum, and an Historical series.

Parts will appear at irregular intervals as they become
ready. Volumes will contain about three or four
hundred pages, and will not necessarily be completed
within one calendar year.

In 1965 a separate supplementary series of longer
papers was instituted, numbered serially for each
Department.

This paper is Vol. 4, No. 5 of the Botany series.

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

Trustees of the British Museum (Natural History) 1969

TRUSTEES OF
THE BRITISH MUSEUM (NATURAL HISTORY)

Issued 10 June, 1969 Price i 8s.

THE JAMAICAN SPECIES OF COLUMNEA AND
ALLOPLECTUS (GESNERIACEAE)

By WILLIAM T. STEARN

With Notes on CYTOLOGY and CHEMOTAXONOMY
By BRIAN D. MORLEY

SUMMARY

Eleven species of Columnea L., together with hybrids between two species (C. rutilans and
C. urbanii), and two species of Alloplectus Mart, occur in Jamaica. All are endemic and have
distinctive ranges within the island. Columnea proctorii and C. urbanii, here described as new
species, have previously been misidentified.

The present paper deals with their history and morphological characteristics, the typification
of the names Columnea, Achimenes and Alloplectus, the definition of the genera Columnea and
Alloplectus, and the effect of pollination by humming-birds on their evolution; it also provides
notes on cytology and chemotaxonomy as well as a key, descriptions and enumerations of speci-
mens. Diagnostic features of pubescence and calyx are illustrated. A survey by methods of
numerical taxonomy (taxonometrics) using both a wide range of characters and a reduced range
gave a classification of the species in general agreement with one reached by traditional methods.

INTRODUCTION

THE following paper on the Jamaican species of Columnea and Alloplectus is the out-
come of the preparation of the text on the Gesneriaceae for Fawcett and Rendle's
Flora of Jamaica, volume 6. A study of all the available herbarium material before
my visit to Jamaica in 1955-56 (cf. Steam, 1959) indicated the need for further
collecting and study of the species, despite the excellence of Morton's careful and
critical revisions of 1944, and I gathered specimens of them whenever I came across
them. Thanks to the co-operation of Mr. George R. Proctor of the Institute of
Jamaica I was able to see most of them in their type localities or nearby. Subsequent
collecting by Proctor has further increased our knowledge of their distribution. In
1965 Dr. Brian Morley visited Jamaica to study the cytology of the genus Columnea,
using wild material. His field work revealed the existence of hybridization between
C. rutilans and C. urbanii and he was able to introduce most of the species into
cultivation at the Royal Botanic Gardens, Kew. Study of the abundant material
now available made evident the need also to consider the definition of the genera
Columnea and Alloplectus on a wider basis. It provided data which appeared suitable
for testing taxonometric analysis as a means of classifying the species. These
matters are dealt with below.

THE TYPE OF COLUMNEA

The genus Columnea L. (sensu stricto) comprises more than 100 species, all native
to tropical America. It attains its greatest morphological diversity and breaks up

BOX. 4, 5. II

182 THE JAMAICAN SPECIES OF COL UMNEA ANDALLOPLECTUS

into its greatest number of species in northern South America and Central America :
17 species, for example, are recorded from Costa Rica alone. In the West Indies as
a whole it is represented by some 16 species, one of which, C. scandens L., is the
type-species by monotypy. Charles Plumier collected this between 1689 and 1697
" in locis insulae martinicanae ", i.e. in Martinique, Lesser Antilles, and founded a
new genus upon it dedicated to the Italian botanist Fabius Columna (i.e. Fabio
Colonna, 1567-1650). He published a brief description, together with an engraving
showing floral details, in his Nova Plantarum Americanarum Genera : 28, t. 33 (1703)
as follows:

" Columnea est plantae genus flore monopetalo, personate, cujus labium superius
non nihil fornicatum & excavatum, inferius vero tripartitum. Ex calyce autem
surgit pistillum, posticae floris parti, ad instar clavi infixum, quod deinde abit in
fructum globosum, mollem, seminibusque plenum exiguis, & oblongis.

Columneae species sunt.

Columnea scandens, phaeniceo flore, fructu albo,
Columnea scandens, flore lutescente, fructu albo."

Plumier had, however, drawn up a much more detailed description and had made
a coloured drawing depicting the flowers as red, the original of which is at the
Museum National d'Histoire Naturelle, Paris, but of which there are copies in the
University Library, Groningen, the Herbarium Library of the Royal Botanic
Gardens, Kew (Plum. MSS. 2 : 117), and the Department of Botany, British Museum
(Natural History), London (Plum. MSS. 2 : 104). That now at Groningen forms part
of the set of copies of Plumier's drawings originally made for Boerhaave, which
passed after his death to Johannes Burman, Linnaeus's friend, at Amsterdam. In
1756 Burman published an engraving made from it, excellently portraying the habit
of the whole plant, in Plantarum Americanarum Fasciculus quartus: t. 89.
Lamarck's account in the Encyclopedic methodique, Botanique 2 : 66 (1786), which is
fuller than any previously published, was based on Plumier's manuscript material
in Paris. Plumier's plant with yellowish flowers is assumed to be a rare colour
form.

Linnaeus himself possessed no specimens of Columnea and he based his description
of the genus in his Genera Plantarum: 373 no. 921 (1737), 5th ed. : 286 no. 710 (1754)
and likewise his protologue of C. scandens in the Species Plantarum 2 : 638 (1753)
primarily on Plumier's published text and figures, although it is evident that between
the preparation of the first edition (1737) and the second edition (1742) of the Genera
Plantarum he saw herbarium material of a Columnea which enabled him to amend
his 1737 statement " STAM : Filamenta duo (pingit, an quatuor) ", based on Plumier's
figure which portrays only two stamens, to " STAM. Filamenta quatuor, quorum duo
longiora, sub labio superiore recondita. Anther ae simplices " in 1742. He may well
have seen material during his visit to Paris in 1738. In any event he stated " Habitat
in Gallia aequinoctiali ", meaning the French possessions in tropical America. The
traditional use of the name Columnea scandens L. for the red-flowered species
occurring in Martinique is thus well-founded. This ranges in the Lesser Antilles
from Montserrat, by way of Martinique, St. Vincent and Grenada, to Trinidad and
Tobago, but does not occur in the Greater Antilles.

THE JAMAICAN SPECIES OFCOLUMNEA AND A LLOPLECT US 183

SLOANE'S JAMAICAN MATERIAL

The first record of any species now included in the genus Columnea was made not
by Plumier but by Hans Sloane and resulted from his stay in Jamaica from December
1687 to March 1689. He found " in montosis Insulae Jamaicae sylvis " a species
which he listed as " Rapunculus fruticosus foliis oblongis integris villosis ex adverse
sitis flore purpureo villoso " in his Catalogus Plantar urn quae in Insula Jamaica
sponte proveniunt : 58 (1696). This entry would be completely cryptic but for the
description and the rather crude engraving in his later Voyage to the Islands Madera . . .
and Jamaica i : 157, t. 100 f. i (1707) and the specimen in Herb. Sloane 3 : fol. 21
at the British Museum (Natural History) on which these were based. Sloane's artist
Everard Kickius unfortunately failed to distinguish the calyx from the corolla,
both being very hairy, but examination of Sloane's specimen shows his plant to be
Columnea fawcettii (Urban) Morton, which was not distinguished as a species until
1944. Sloane probably collected it on Mount Diablo.

THE TYPE OF ACHIMENES

Patrick Browne, failing to connect Linnaeus's genus Columnea with any Jamaican
plant, established a new genus A chimenes in his Civil and Natural History of Jamaica :
270, t. 30 (1756). He included two species. The first of these, his " Achimenes
major, herbacea, subhirsuta, oblique assurgens " to which he referred Sloane's
" Rapunculus fruticosus . . ." is undoubtedly a Columnea, probably C. fawcettii,
since he mentions it as " most commonly met with in the woods of New Liguanea
and St. Ann's " and describes the " divisions of the cup ", i.e. calyx-segments, as
" pinnated at the sides, somewhat like those of the garden rose ". Browne's generic
description was based on this species, as is evident, for example, from his account of
the irregular corolla: " limbus erectus, in quatuor lacinias inaequales, inaequaliter
sectus; lacinia superior recta, latiuscula, profunde crenata, sive bifida; later ales
oblongae & a superiori oblique decendentes; infima anguste patula, ad medietatem
floris incisa, longissima." Hence Achimenes P. Browne (1756) as to type is congeneric
with Columnea L. Browne's second species " Achimenes minor, erecta, simplex ..."
is the plant commonly known as Achimenes coccinea (Scop.) Pers. or A. pulchella
(L'Herit.) Hitchc., but whose correct name is A. erecta (Lam.) H. P. Fuchs. Browne
regarded it as agreeing " in the most essential parts [of the flower] with the fore-
going ", although it has an almost regular corolla. When Persoon dealt with these
two Brownean species in his Synopsis Plantarum 2 : 164 (1806) he referred Browne's
first species to Columnea, as Burman had already done in 1756 and Swartz in 1788,
but he retained the name Achimenes for the second species, at the same time providing
a new generic description referring to this alone: " Cal. superus 5-phyllus. Cor.
inf undibulif ormis ; limbus planus, 5-lobus, subaequalis. Rudimentum filamenti 5-ti.
Anther ae connexae. Caps, bilocularis." Under the name Achimenes Persoon in
fact established a new genus distinct from Achimenes P. Browne (sensu stricto) and
the name Achimenes has been consistently used for it. This now includes about 25
species, of which 13 are in cultivation, together with many cultivars, some of hybrid

184 THE JAMAICAN SPECIES OF COLUMNEA AND A LLOPLECT US

origin. Being a later homonym of Achimenes P. Browne, it needs conservation; the
type-species of Achimenes Pers. is A. coccinea (Scop.) Pers., i.e. A. erecta (Lam.)
H. P. Fuchs.

THE TYPE OF A LLOPLECT US

The genus Alloplectus was founded by Martius, Nov. Gen. Sp. PL Brasil. 3 : 53
(1829), with the following differential character: " Calyx liber, coloratus, penta-
phyllus, foliolis imbricatis, plus minus connatis, duobus interioribus. Corolla infera,
tubulosa vel claviformis, rectiuscula, limbo brevi quinquelobo aut quinquedentato.
Stamina quatuor didynama, cum quinti postici rudimento minimo, e basi tubi.
Annulus hypogynus in glandulam posticam tumens. Capsula baccans, ovata,
coriacea, unilocularis, bivalvis, seminibus numerosis oblongis." He included two
Brazilian species, both described and illustrated in detail, A. sparsiflorus Mart, and
A. circinatus Mart.

In the list of conserved generic names in the International Code of Botanical
Nomenclature the name Alloplectus Mart. (1829) is conserved under No. 7860 against
Crantzia Scop. (1777) and Vireya Raf. (1814) and its lectotype given as A. sparsiflorus
Mart. (= A. hirtellus (Schott) Preston), although A. circinatus would have been a
better choice since Martius saw no mature fruits of A. sparsiflorus but had both
flowering and fruiting material of A . circinatus and described the fruits and seeds of
Alloplectus from this species. His two Brazilian species are, however, closely allied
and differ in several characters from many species now referred to the genus Allop-
lectus. If, however, a broad concept of the limits of Alloplectus is accepted, as in
the present paper, Dalbergaria Tussac (Fl. Antilles i : 141 (1811-13) must be added
to its synonomy as a nomen rejiciendum, since its type, D. phaenicea Tussac, loc. cit.
(1811-13) is the same Hispaniola species (Alloplectus sanguineus (Pers.) DC., Prodr.
7 : 546 (1839))* as Rafmesque's Vireya sanguinolenta, i.e. the type of Vireya Raf.
already listed as a nomen rejiciendum. The short-lived Palermo Specchio delle
Scienze o Giornale enciclopedico di Sicilia,^ wherein (i : 194 (June, 1814)) Rafmesque
published Vireya, is an extremely scarce periodical and his account is accordingly
reprinted here :

" VIREYA. Cal. 5partitus, lac. equal, laciniatis, Corolla peripetala tubulosa
5dentata, basi gibbosa; Stam.4. didynamica inclus. Ov. liberum oblong, basi
monadeno, stigma 2lamellat. Capsula i.locul. 4valve polysperma, semina recept.
4. longitudinal, affixa. Planta, fol. oppos. flor. axillaris. Oss. Mi. fu comunicato
questo nuovo genere del Sign. Turpin col. barbaro nome di Caonabo, al quale ho
sostituito il nome dell' egregrio moderno naturalista e filosofo Virey. Si appartiene
alia mia seconda Classe Mesogynia, 5. Ordine Epidia, Famiglia Didynamia: ha delle
affinita colli generi Browallia, e Lindernia.

Vireya sanguinolenta. Caule erecto simplex, foliis oppositis, alternis multo major,
oblongis integris acutis sub-villosis subtus sanguinolentis, florib. axillarib. solitaris
sessilis. Oss. Nasce nellTsola di S. Domingo offerisce le singolari particolarita di

* Martius, Nov. Gen. & Sp. 3 : 57 (1829) suggested that Besleria sanguinea Pers. belonged to Alloplectus
but he did not make the combination A. sanguineus attributed to him by Jackson (Index Kew. i : 84
(1893)), Urban (Symb. Antill. 2 : 358 (1901)), and Morton (Contr. U.S. Nat. Herb. 29 : 4 (1944)).

f The British Museum (Natural History) now possesses a photostat copy.

THE JAMAICAN SPECIES OFCOLUMNEA AND A LLOPLECT US 185

avere alternativamente le foglie dieci volte piii grande delle opposte, e coperte al di
sotto di una larga macchia sanguinosa: ha le corolle d'un giallo livido."
Vireya Raf. is thus a synonym of Dalbergaria.

RECORDS OF COLUMNEA AND ALLOPLECTUS IN JAMAICA

The first post-Linnaean author to deal with the Jamaican species of Columnea
was Olof Swartz, who travelled extensively in Jamaica between 1784 and 1786 and
succeeded in finding an extraordinary number of rare plants. His journeys took
him into the parishes of Trelawny, Hanover, St. Elizabeth, St. Andrew, St. Catherine
and St. Thomas. His collections included three new species of Columnea, i.e.
C. hirsuta, C. hispida and C. rutilans, of which he published diagnoses in his Nova
Genera et Species Plantarum seu Prodromus : 94 (1788) and detailed descriptions in his
Flora Indiae Occidentalis 2 : 1080-1086 (1800). Under his C. hirsuta he cited as
synonyms Sloane's " Rapunculus fruticosus . . ." and Browne's " Achimenes major
herbacea . . .", but from the significant note in his description, '' Hirsuties pulchre
articulata," and his specimen in the British Museum (Natural History) it is clear that
he based his C. hirsuta on material of the species growing high in the Blue Mountains
which has conspicuous multicellular hairs on the leaves. To this species the name
C. hirsuta has always been attached. Swartz himself ascended Blue Mountain Peak
in 1785 and, while a guest of Matthew Wallen at Coldspring, St. Andrew, in the Blue
Mountains, he had other opportunities of collecting the species. In the west of
Jamaica he collected C. rutilans, conspicuous for its leaves red beneath. Swartz's
third species, C. hispida, was long enigmatic. He himself collected it only in fruit
" in montibus altis Jamaicae occidentalis, ad rupes umbrosas ". Its flowers in fact
remained unknown until February 1957; the flowering material collected by
Wullschlagel in Manchester which Grisebach in 1862 (Fl. Brit. W. Ind. Is. : 465)
referred to C. hispida belongs to another species, C. urbanii. Swartz's own descrip-
tion and his specimens at the British Museum (Natural History) and the Rijksmuseet,
Stockholm, show C. hispida to be a very distinct species and justify Morton's com-
ment in 1944: " From description I am not able to identify this with any of the
species known from Jamaica. It is perhaps a valid species, not collected again."
Swartz, as was unfortunately his custom, gave no precise locality. In 1784 he
travelled across western Jamaica from Montego Bay on the north coast to Savanna-
la-Mar on the south coast and it would seem probable that he then came across this
very rare and local species, for White Rock Hill, St. James, where Proctor and I had
the good fortune to rediscover it in March 1956, lies near a route between these towns.

Grisebach's account of Columnea in his Flora of the British West Indian Islands :
464-465 (1862) added a further species, C. argentea Griseb., notable for its sericeous
leaves and other parts, which the missionary Heinrich Rudolf Wullschlagel (1805-64)
of the Moravian Brothers had collected in 1849 at Nazareth, Manchester. Evidently
unaware of Grisebach's publication, Hanstein described it anew in 1865 as C.
wullschlaegeliana. This again is a very local species. In 1956 I was overjoyed to
find it growing among rocks near the Moravian Chapel at Nazareth, which must have
been the very place where Wullschlagel gathered it. As noted above, Grisebach
included under C. hispida Swartz a species here named C. urbanii.

186 THE JAMAICAN SPECIES OF COLUMNEA ANDALLOPLECTUS

Grisebach placed three Jamaican species in the genus Pterygoloma, which was
founded by Hanstein in 1854 (Linnaea 26 : 211) on a species described and illustrated
in Bot. Mag. 72 : t. 4250 (1846) as Attoplectus repens Hook, and there stated to have
been collected by William Purdie " in the ascent of the Sierra Nivada, Santa Martha ",
Colombia. The first of these, which Grisebach erroneously identified with Ptery-
goloma repens (Hook.) Hanstein and enumerated as such, is a dwarf creeping species,
with flowers not so markedly zygomorphic as in typical species of Columnea; it had
been collected in Jamaica both by Wullschlagel and Purdie. In 1901 Urban distin-
guished it as Columnea jamaicensis. Grisebach's two other species, P. pubescens
from western Jamaica and P. cristatum from central Jamaica, based on collections
by Wilson, Purdie, Alexander and Wullschlagel, agreed with the first in their low
creeping habit but differed in having almost cylindrical or clavate corollas with very
short almost equal rounded lobes. They are easily distinguished from one another
by the calyx-segments, entire in P. pubescens, pinnatifid with spreading lobes in
P. cristatum. Fawcett listed all three under Attoplectus (see p. 234) in 1893. The
association of the three accords with the results of a taxonometric analysis of the
group (see p. 213).

Nothing more was published on the Jamaican species of Columnea until 1901,
when Urban included them in his " Enumeratio Gesneriacearum " (Symb. Antill.
2 : 344-388 (1901)). He revised nomenclature, listed synonyms and specimens,
and amended previous descriptions in the careful and scholarly manner which makes
his contributions to West Indian botany so valuable as quarries of information, but
as usual he failed to provide a key to the plants concerned, an omission which detracts
so much from the practical utility of his work. He recorded five species as Jamaican,
expressing his opinion about their distinctness by the comment " Fortasse omnes
Columneae species in Jamaica obviae formis intermediis inter sese conjunctae sunt,
C. jamaicensi excepta ". There are indeed some specimens intermediate between
recognized species, but these would appear to result from hybridization. All the
species accepted by Urban have definite characters associated with distinctive
geographical ranges which justify their separation. C. hirsuta he regarded as a very
variable species, within which, despite the limited material then available, he
distinguished seven taxa. On the basis of the much more abundant and adequate
material now at hand, it seems more reasonable to refer Urban's varieties and forms
of C. hirsuta to four species as follows:

var. genuina (p. 361) = C. hirsuta Swartz

var. pallescens (p. 362) = C. hirsuta Swartz

var. concolor (p. 362) = C. hirsuta Swartz

var. subintegra (p. 362)

forma wullschlaegelii (p. 362) = C. urbanii Stearn

forma hansenii (p. 363) = ? C. harrisii (Urban) Britton

ex Morton
forma harrisii (p. 363) = C. harrisii (Urban) Britton

ex Morton
var. fawcettii (p. 363) = C. fawcettii (Urban) Morton

THE JAMAICAN SPECIES OF COLUMNEA AND A LLOPLECT US 187

In 1909 Urban added another species, Columnea brevipila, collected by Harris on
Bluefields Mountains, Westmoreland.

Morton in 1944 (a, b) revised the West Indian species of Columnea and Alloplectus.
With much more material for study than Urban had, most of it collected by Harris
and by Britton, he described seven Jamaican species of Columnea in detail, raising
two of Urban's varieties of C. hirsuta to specific rank as C. harrisii and C. fawcettii
and adding a new species, C. subcordata, from Windsor, Trelawny. Unlike Urban,
he provided a key to all the West Indian species of both Columnea and Alloplectus.
Examination of living material during my visit to Jamaica in 1956 (cf. Stearn, 1959)
confirmed the taxonomic soundness of Morton's work, already evident from a prior
survey of herbarium material.

DEFINITION OF COLUMNEA AND ALLOPLECTUS

The difficulties of defining the American genera of Gesneriaceae, among them
Columnea and Alloplectus, are reflected in the different views of authors on their
generic limits. As stated by Leeuwenberg (1958 : 293), " there are several cases of
close inter-relationship of species usually placed in separate genera, rendering the
boundaries between those genera vague or even untenable. Judging from these
cases alone one would be compelled to unite large groups of genera or even complete
tribes under one genus; this is particularly true for the Columneinae. The typical
species of the genera are however so different that such a procedure would be quite
undesirable. But part of those generic limits are based rather on convenient charac-
ters than on conspicuous discontinuities of features. A consequence of this state of
affairs is the presence of what may be called ' borderline species ', i.e. species that
have some characters of one and some of an other genus." Such a species in Jamaica
is Columnea jamaicensis, referred by Grisebach to Pterygoloma, by Fawcett to
Alloplectus, by Urban to Columnea. The two Brazilian species (see above) on which
Martius founded the genus Alloplectus in his Nova Genera et Species Plantar um
Brasiliensium 3 : 53, t. 223 (1829) are markedly different from Columnea in their
narrowly tubular rather ventricose corolla with the limb almost regular and very
shortly lobed, i.e. shallowly 5-toothed or 5-lobed, the lobes almost equal, in their
free anthers and their rather fleshy 2-valved capsule, but these distinctions have not
remained clear-cut as more and more species of the group have been discovered.
A few examples will serve to indicate the lack of correlated differences. Thus a
species such as Columnea aureonitens Hook. (Bot. Mag. 73 : t. 4294 (1847)) has a
narrowly tubular corolla with very short almost equal lobes, but four of the anthers
cohere as in typical zygomorphic species of Columnea, while C. kalbreyeri Hook. f.
(Bot. Mag. 108 : t. 6633 (1882)) has a zygomorphic corolla but the four anthers are
completely free from one another. C. guianensis Morton, illustrated by Leeuwenberg
( X 958 : 387, fig. 16), agrees with many typical species of Columnea in having the
leaves of a pair markedly unequal in size, but its corolla is narrowly cylindrical with
a small almost regular limb of five short erect lobes, the two dorsal ones however
connate, and the anthers are free from one another as in a typical Alloplectus. The
same is true of C. sanguinea (Pers.) Hanstein (Dalbergaria phaenicea Tussac) with

i88 THE JAMAICAN SPECIES OF COLUMNEA AND A LLOPLECT US

markedly unequal leaves but a regular cylindrical corolla and anthers cohering only
in pairs. Alloplectus savannarum Morton (illustrated by Leeuwenberg, 1958 : 365,
fig. 9) has unequal leaves and cohering anthers but a cylindrical corolla and a two-
valved capsule as in typical Alloplectus. A possible distinction based on the form of
the corolla tube, i.e. gradually widened and not contracted at the mouth in Columnea,
ventricose with a contracted mouth in Alloplectus, is obscured by such species as
Columnea fdipes Oliver (Hook., Ic. PI. 25 : t. 2428 (1896)) with ventricose tube but
connate anthers. Characters of the fruit, an almost globose berry in typical
Columnea, a capsule separating into two fleshy or thick valves in typical Alloplectus,
might provide means of separation, but A. domingensis with a regular corolla has a
berry-like fruit like a typical Columnea and the fruits of many species placed currently
in Columnea or Alloplectus are unknown; the correlation of carpological, habit and
floral characters in the group as a whole is accordingly doubtful. Attempts to
separate the two genera by referring those with a coloured calyx and free anthers to
Alloplectus, the type of which, A. hirtellus, has a dark red calyx, and those with a
green calyx and connate anthers to Columnea are frustrated by such a species as
C. scheideana Schlecht. (illustrated in Bot. Mag. 70 : t. 4045 (1843)) with a large
purplish-red calyx as in Alloplectus but a markedly zygomorphic 2-lipped corolla
and connate anthers typical of Columnea. Anisophylly is likewise unavailing, since
the type-species of both Columnea and Alloplectus have the two leaves of a pair equal.
In short there is no apparent discontinuity between Columnea and Alloplectus. O.
Kuntze in 1891 (Rev. Gen. PL 2 : 470) united the two, together with Hypocyrta and
Nematanthus, into one genus, Columnea (sensu lato), which would now include about
220 species, defined as follows: " Ovarium omnino superum. Disci glandula postica
magna ceterae o vel parvae. Stamina 4. Antherarum loculi paralleli. Filamenta
basi dilatata unilateraliter connata cum corolla breviter connata. Calyx 5-partitus
vel alte 5-fidus. Corolla apice variabilis exappendiculata. Fructus subbaccatus
vel demum bivalvis." This course certainly has simplicity to recommend it as
regards nomenclature but obscures the special character of Columnea (sensu stricto)
as a highly developed group. The prevailing gamopetalous type of corolla found
in Gesneriaceae and allied families consists of a fairly long tube with a spreading
limb which is shorter than the tube and slightly or distinctly two-lipped; the two
upper lobes form an upper lip while the lower lobe and the two lateral lobes together
form a lower lip which serves as a landing platform for insects seeking nectar at the
base of the tube. The stamens are usually included within the corolla-tube and self-
pollination is often possible; it occurs, for example, in Alloplectus species under
cultivation. The fruit is usually a capsule. Alloplectus (sensu stricto) represents
this general unspecialized type. The corolla of Columnea (sensu stricto) represents
a reversal of the trend towards development of an expanded lower lip able to support
the weight of a bee. In Columnea the upper side of the corolla has become prolonged
forward, the two uppermost lobes being fused and enlarged into a hood sheltering
the exserted anthers and stigma; the two lateral lobes are reduced and they spread
either sideways or upwards; the lower lip is also reduced and turned down or back.
By these features, together with the horizontal poise of the flowers and their usually
conspicuous red or yellow colouring, the flower of Columnea (sensu stricto) has

THE JAMAICAN SPECIES OF COLU M N E A AND A LLOPLECT US 189

become perfectly adapted to pollination by nectar-seeking humming-birds on the
wing (Text-fig. 5). The reduced lower lip and lateral lobes offer no obstacle to the
bird as it flies under the lengthened upper lip and thrusts its beak and tongue down
to the abundant nectar at the base of the tube, while its wings beat unimpeded on
both sides of the lip. The anthers and stigma, being directly under the upper lip,
are so placed that they touch the top of the head of the hovering bird. The coherence
of the anthers ensures that they deposit the maximum of pollen in one area on the
bird's head where the stigma of another flower will first come into contact with it.
The corolla as a whole is firm enough in texture to sustain vibration by the wing-beats
of a humming-bird (mostly 20-25 beats per second, but ranging from 8 to 80 according
to the size of the bird; cf. Greenewalt, 1960). Such a flower well exemplifies " the
syndrome of ornithophily " as characterized by Faegri & van der Fiji (1966 : 109).
This has been achieved by the successful association of several divergences from the
almost regular flower typical of Alloplectus, which divergences have evidently also
occurred separately and independently, as the anomalous intermediate species
mentioned above indicate.

The proposal is accordingly made that Columnea should be defined more narrowly
than is customary, by including within it only species of this ornithophilous habit,
characterized by having a distinctly bilabiate red or yellow corolla, with the four
upper (posterior) lobes fused and extended into a galea, the lower (anterior) lobe
narrow and bent downwards, and having exserted stamens with the four anthers
initially connate. Those species not satisfying this definition should be transferred
to Alloplectus* or possibly other genera. Thus Alloplectus might likewise be more
narrowly defined and restricted to species with a regular corolla and a capsular fruit
akin to the type-species. To accommodate the species not fitting into this and
Columnea (sensu stricto), Dalbergaria and possibly also Pterygoloma might then be
restored.

A situation of this kind, in which a well-marked specialized offshoot, such as
Columnea (sensu stricto), is still connected to a generalized stock, such as Alloplectus

* Acceptance of this viewpoint makes it necessary to transfer to A lloplectus the following West Indian,
Guianaian and Venezuelan species allocated to Columnea by Morton and by Leeuwenberg:

Alloplectus afflnis (Morton) Stearn, comb. nov.

Columnea affinis Morton in Fieldiana 28 : 529 (1953).

Alloplectus aureonitens (Hook.) Stearn, comb. nov.

Columnea aureo-nitens Hook, in Bot. Mag. 73 : t. 4294 (1847). Leeuwenberg in Act. Bot. Neerl.
7 = 383 (1958).

Alloplectus calotrichus (Donn. Smith) Stearn, comb. nov.

Columnea calotricha Donn. Smith in Bot. Gaz. 40 : 9 (1905). Leeuwenberg, torn. cit. : 385 (1958).

Alloplectus cubensis (Urban) Stearn, comb, nov

Columnea sanguinea var. cubensis Urban, Symb. Antill. 2 : 359 (1901).

Columnea cubensis (Urban) Britton in Torreya 5 : 215 (1905). Morton in Contr. U.S. Nat. Herb.

2 9 ' 5 ( I 944)- -Alain in Leon & Alain, Fl. Cuba 4 : 472 (1957).

Alloplectus guianensis (Morton) Stearn, comb. nov.

Columnea guianensis Morton in Bull. Torrey Bot. Club 75 : 564 (1948). Leeuwenberg, torn. cit. :
386 (1958).

i go THE JAMAICAN SPECIES OF COLUMNEA AND A LLOPLECT US

(sensu lato), by a few species showing intermediate stages of development, does not
permit of a solution both logical and convenient and hence proves disconcerting to
a tidy-minded taxonomist intent on emphasis of discontinuity, but it is all the more
interesting and worthy of study as indicating possible lines of evolution within the
group. The most practical procedure would seem to be to recognize as genera both
the large heterogeneous groups manifesting a wide range of characters occurring
in many different combinations without consistent correlations and also at the same
time the allied apparently monophyletic groups separated from them by possessing
consistent correlations of characters whereby these form biologically important or
conveniently distinguishable entities, the few intermediate species being referred to
the more heterogeneous groups. The case of Alloplectus and Columnea parallels
in some ways that of Lilium and Nomocharis (cf. Sealy, 1950) where a solution
according with this viewpoint has proved satisfactory.

The Jamaican species, apart from Columnea jamaicensis, fit readily into Columnea
with a strongly bilabiate corolla and Alloplectus with a regular cylindrical corolla.
C. jamaicensis is like A. pubescens and A. grisebachianus in its low creeping habit
and has the bilabiate character of the corolla less marked than in typical Columnea,
but the tube expands gradually, the upper lip consists of two fused lobes with two
spreading lateral lobes, and the four anthers are connate. These associated floral
characters justify the allocation of C. jamaicensis, and hence of Pterygoloma, to
Columnea, as defined here, even though on the sum of its other characters, as is shown
by taxonometric analysis, it has more in common with the Jamaican species of
Alloplectus.

CHARACTERS DISTINGUISHING THE SPECIES

The Jamaican species of Columnea and Alloplectus are soft- wooded plants of
creeping or sprawling habit, rather sparingly branched, with stems which may even
reach a length of several metres in some robust specimens, the leaves then being
clustered near the growing tip and the branchlets marked with the conspicuous scars
of the fallen leaves. Columnea jamaicensis (PI. 14), Alloplectus pubescens and
A. grisebachianus have slender creeping branchlets only 1-3 mm. thick, whereas the
others have much stouter branchlets about 5-10 mm. thick, initially erect but
drooping downwards with age when the plants grow as epiphytes on trees. The
leaves are paired, those of a pair being almost equal in size and shape, as in C. urbanii
(PL 16), to markedly unequal, as in C. proctorii (PL 17) and C. rutilans (PL 21), the
smaller leaf of the pair having a shorter petiole as well as a shorter blade and often
falling earlier than the larger leaf. This anisophylly, although used by Leeuwenberg
(1958 : 295) in his study of Guiana Gesneriaceae to separate Alloplectus sect. Allo-
plectus with the leaves of a pair almost equal from Columnea sect. Collandra with
leaves of a pair markedly unequal, varies in the Jamaican species, being apparently
constant in some, such as C. hispida, C. subcordata and C. harrisii, but variable in
others such as C. fawcettii and C. hirsuta, which may have the leaves of a pair almost
equal or markedly unequal. The difference between the large and small leaves is
particularly evident in C, rutilans, C. proctorii and C. subcordata.

THE JAMAICAN SPECIES OF COLUMNEA ANVALLOPLECTUS 191

The proportions as well as the actual dimension of the leaf-blade provide characters
of some diagnostic value, as Morton demonstrated by measuring 50 representative
leaves of each species when he had enough material available. Thus in Columnea
jamaicensis, C. hirsuta and C. subcordata, the blades average about two times as long
as broad, in C. fawcettii, C. brevipila and C. harrisii, between 2-4 and 2-7 times as
long as broad, and in C. rutilans and C. argentea, about 3-4 to 3-7 times as long as
broad. The leaf -base is remarkably asymmetric in C. harrisii and C. hispida
C. rutilans stands apart from the others in having the underside of the leaf either
completely red or else conspicuously red-veined, and this red colouring is evident
in some of the hybrids between C. rutilans and C. urbanii.

Diversity of hair-covering provides features of diagnostic value, notably the hair-
covering on the upper surface of the leaf-blade, where four main types of hairs may
be distinguished (Text-fig, i, p. 192) :

1. Hairs erect, S-g-celled, red, to 3-5 mm. long, e.g. Columnea hispida (Text-fig.
i A), in which also occur minute erect colourless, 3-celled hairs.

2. Hairs erect, 5-io-celled, colourless, 0-8-2-5 mm - l n g> e -g- C. hirsuta (Text-fig.
iC), C. proctorii, C. subcordata, C. urbanii (Text-fig. iB); in C. hirsuta these are
associated with minute almost appressed 2-celled hairs.

3. Hairs bent forward and somewhat or almost appressed, 4-6-celled, colourless,
scarcely i mm. long, e.g. C. argentea (Text-fig. iF), C. harrisii (Text-fig. lE)
and C. rutilans (Text-fig. iD), in all of which they are associated with minute
almost appressed 2-celled hairs.

4. Hairs bent forward, almost appressed, 2-celled, colourless, 0-2-0-4 mm - l n g.
exclusively present in C. brevipila (Text-fig. iD) and C. fawcettii (Text-fig. iH),
associated with longer hairs in C. argentea (Text-fig. iF), C. hirsuta (Text-fig.
iC), C. harrisii (Text-fig. lE), and C. rutilans (Text-fig. iG).

Suppression of the long hairs in C. hirsuta would produce leaves like those of
C. fawcettii.

The flowers are axillary, with one to five in an axil, on pedicels 1-3 cm. long, varying
from species to species in the type of hair-covering.

The most useful characters for specific distinction are provided by the calyx (Text-
fig. 2) . The segments may increase a little in size after flowering as the fruit matures
but retain their shape, being, for example, about 12 mm. long in flower but 16 mm.
long in fruit in Columnea jamaicensis, about 2-5 cm. in flower but 3-5 cm. in fruit in
C. rutilans. They are mostly between 2 and 3 cm. long. C. subcordata (Text-fig.
2G) stands apart from the other species, in which the five segments are free almost
to the base, by having four segments united for about the lower third of their length
and the fifth almost free.

The segments are linear and entire in C. hispida (Text-fig. 2A) and lanceolate or
narrowly lanceolate and entire or scarcely toothed in C. argentea (Text-fig. 2C),
C. brevipila, C. harrisii, C. proctorii (Text-fig. 2B), C. urbanii (Text-fig. 2H) and
A. pubescens (Text-fig. 2QG). In C. hirsuta and C. fawcettii (Text-fig. 2D; Text-fig.
28B) the segments bear one to three remote but usually well-marked teeth on each
margin. In C. rutilans (Text-fig. 2F) and A. grisebachianus (Text-fig. 296) they are

192

THE JAMAICAN SPECIES OF COLUMNEA AND ALLOPLECTUS

laciniate in the lower part, with 3-5 prominent teeth on each margin. These
differences in shape, together with differences in hair-covering, enable most species to
be recognized by the calyx alone. The hairs of the calyx are red in C. hispida, C.
fawcettii, C. hirsuta, C. hispida and C. rutilans, colourless in the others.

The corolla as mentioned above (p. 186) is small, narrowly tubular and regular in
Alloplectus grisebachianus (Text-fig. 2gA, B) and A. pubescens and slightly 2-lipped

brevipila

urbanii

D

harrisii

rutilans
hirsuta \ H fawcettii

hispida

FIG. i. Hairs on upper leaf-surface of: A, Columnea hispida Swartz; B, C. urbanii Stearn;
C, C. hirsuta Swartz; D, C. brevipila Urban; E, C. harrisii (Urban) Morton; F, C. argentea
Griseb.; G, C. rutilans Swartz; H, C. fawcettii (Urban) Morton. (All drawn from living
plants cultivated at Kew.)

THE JAMAICAN SPECIES OF COLUMNEA ANDALLOPLECTUS 193

in Columnea jamaicensis (Text-fig. 28H), but larger and strongly 2-lipped in the other
species. It is completely yellow in C. argentea (Text-fig. 5), C. brevipila, C. hispida
C. subcordata and C. urbanii (Text-fig. 28G) but longitudinally striped with red in
the others, including many of the hybrids between C. rutilans and C. urbanii.

The filaments of the stamens are glabrous in Columnea argentea and C. brevipila
but pubescent in most other species.

jamaicensis

F rutilans

subcordata

H

urbanii

FIG. 2. Calyx of: A, Columnea hispida Swartz (Steam 449); B, C. proctorii Steam, with
apex of pedicel (Steam 451) ; C. argentea (Steam 359) ; D, C. fawcettii (Steam 157) ; E, C.
jamaicensis Urban, a (Wullschldgel I2go), b (Steam 928); F, C. rutilans Swartz (Steam
453): G, C. subcordata Morton (Steam 472); H, C. urbanii Stearn (Steam 361).

194

THE JAMAICAN SPECIES OF COLUMNEA ANDALLOPLECTUS

Certain species can thus be easily recognized by distinctive and unusual characters.
For example, Columnea jamaicensis has small leaves about 3-5 cm. long, a short
calyx and a small corolla not strongly bilabiate. C. rutilans stands apart from
other species in having its leaves red below and its calyx-segments partly laciniate.
C. subcordata has a bilabiate calyx with four segments fused for their lower third.
C. fawcettii with toothed calyx-segments and C. brevipila with entire calyx-segments
differ from other species in having only very short appressed 2-celled hairs on the
upper surface of the leaf. C. argentea has relatively narrow leaves silvery-grey with
long somewhat appressed sericeous hairs. C. hispida is notable for the long erect
reddish hairs covering the upper side of the leaf as well as the calyx and for the
relatively small corolla.

CYTOLOGY
By BRIAN D. MORLEY

The chromosome numbers of species of Columnea have been reported by Rogers
(1954), Eberle (1956), Fussell (1958), Lee (19620,, b, 1964, 1966, 1967), Morton (1963),
Lee & Grear (1963) and Morley (1967). Investigation of the chromosome numbers
of ten Jamaican species and four natural hybrids selected from a hybrid swarm
between C. rutilans and C. urbanii confirms the reports for C. brevipila, C. fawcettii
and C. hispida by Lee (1964, 1966) and shows that all the Jamaican species have a
diploid number of 2n = 18 and that their basic number is the same as that of other
Columnea species, x = 9. The plants studied were as follows:

Species Locality

C. hirsuta Hardwar Gap, St. Andrew-Portland.

C. fawcettii Mount Diablo, St. Ann.

C. rutilans Ramgoat Cave, Trelawny.

C. rutilans Near Catadupa, St. James.

C. argentea Shooters Hill, Manchester.

C. urbanii Top Hill, Manchester.

C. urbanii Near Newport, Manchester.

C. urbanii Rose Hill Village, Manchester.

C. proctorii Near Troy, Trelawny.

C. proctorii? Ramgoat Cave, Trelawny.

C. hispida Near Catadupa, St. James.

C. harrisii Seven Rivers, St. James.

C. harrisii Near Leamington, Westmoreland.

C. brevipila Cho Cho Gulley, Westmoreland.

C. jamaicensis Near Leamington, Westmoreland.

Plant
reference no.

Morley 33

Morley 2

(coll. E. Lodge)

Morley 3
Morley 35
Morley 20
Morley 36
Morley 14
Morley 442
Morley S.6
Morley 8.7
Morley 8.17

(coll. C. D. Adams)
Morley 5
Morley 40
Morley 23
Morley S.n
Morley 24
Morley 27

(coll. C. D. Adams)

18
18

18
18
18
18
18
18
18
18
18

18
18
18
18
18
18

THE JAMAICAN SPECIES OF COLUMNEA ANDALLOPLECTUS

195

C. jamaicensis Near Leamington, Westmoreland.

C. rutilans X Top Hill, Manchester.

urbanii hybrid 4
C. rutilans X Top Hill, Manchester.

urbanii hybrid 6
C. rutilans x Top Hill, Manchester.

urbanii hybrid 8
C. rutilans x Top Hill, Manchester.

urbanii hybrid n
A. grisebachianus Mount Diablo, St. Ann.

Morley S.io 18
Morley 16, 44, 45 18

Morley 47 18

Morley 15, 46 18

Morley 48 18

Morley s.n. 18
(coll. E. Lodge)

Although the chromosomes are small, they display sufficient morphological
diversity to allow recognition of the following three karyotype series. A metacentric
position is defined as an approximate position of 50 % along the length of a chromo-
some and submetacentric as 60 %~70 % along a chromosome.

Series i. Karyotypes with large chromosomes, i.e. longer and thicker, where the
smallest chromosome is always longer than broad. Two pairs of chromo-
somes with submetacentric centromere position, seven pairs of chromo-
somes with metacentric centromere position; see Text-fig. 3.

[ Species: C. hirsuta, C. fawcettii, C. urbanii (from an area north of

Mandeville) .

Series 2. Karyotypes with large chromosomes, i.e. longer and thicker, where the
smallest chromosome is always longer than broad. Three pairs of
chromosomes with submetacentric centromere position, six pairs of
chromosomes with metacentric centromere position; see Text-fig. 3.
Species: C. urbanii (from an area south of Mandeville), C. brevipila, C.
rutilans, C. jamaicensis.

Series 3. Karyotypes with small chromosomes, i.e. shorter and thinner, where the
smallest chromosome is always as long as broad. Three pairs of chromo-
somes with submetacentric centromere position, six pairs of chromosomes
with metacentric centromere position ; see Text-fig. 4.
Species: C. argentea, C. hispida, C. harrisii, C. proctorii.

Series I karyotypes characterize the species Columnea hirsuta and C. fawcettii,
which are morphologically and biochemically distinct from other Jamaican species
and are only found in eastern Jamaica. Most C. urbanii karyotypes belong to
Series 2, with the exception of one plant which had a Series i karyotype.

Although Columnea jamaicensis is included in Series 2, it is unlike any other
Jamaican species, having a different phenotype and a karyotype with two pairs of
satellites. The metacentric satellite pair of C. urbanii (66.1460) requires further
analysis.

EOT. 4, 5 12

ig6 THE JAMAICAN SPECIES OFCOLUMNEA AND A LLOPLECT US

Karyotypes of Jamaican species
of Columnea. x25OO

66-991 hirsuta.

66-1627 fawcettii.

" a " i* if** *%-<*

fv II <t 9m n P 1 I* * %

661042 urbanii.

*M

* Mift3ttl**ail

66-1460

t; 18 as

66-960

urbanii.

ss **

brevipila.

it *s IS u :i as )

66-1084 rutilans.

66-1A68

jamaicensis.

CD

FIG. 3. Karyotypes of some Jamaican species of Columnea.

The hybrids between Columnea rutilans and C. urbanii have karyotypes inte
mediate in morphological details between those of the parents; see Text-fig. 4.

The Jamaican species with smaller and more asymmetric karyotypes, i.e. Columnea
argentea, C. hispida and C. harrisii in Series 3, have the more limited distributions in
Jamaica.

While, on evidence so far available, no changes in chromosome number have
accompanied the evolution of Columnea, changes in chromosome structure have
occurred, as is evident from gross morphological studies of Jamaican and Central
American Columnea karyotypes.

THE JAMAICAN SPECIES OF COL UMNEA A.NDALLOPLECTUS 197

Karyotypes continued.

66-1119 argent ea.

2 S3 ft* ? S* IS > I*

66-1180 hispida.

83 fid is if ISli :t KIT

66-1465 harrisii.

66-1008

subcordata.

A ;; ft ;*. r. v. nX*l**f*l*S

hybrid 3.

H 2 i f * ..*

<% M I I A I **%%* A %

66-1257 hybrid A.

*S :A nft iX S&SS 5s % tr

66-1291 hybrid 5.

; ; s s c s U - ; : t , t - t i

66-1152 hybrid 6.

5S nr, (A S* SX tt *s ss ?A

FIG. 4. Karyot