MEMOIRS

OF THE

TORREY: BOTANICAL CLUB

VotumeE VIII

Lucien Marcus UNDERWOOD, Editor

Mo. Eot. نا‎ was
15993,
PUBLISHED BY THE CLUB

New York
1899-1902

CONTENTS

I. Francis E. Lrovp. The comparative Embryology of the Rubi-
aceae. Pp. 1-26. //. 7-4. Issued 26 August 1899; pp. 27-111.
Pl. 5-15. Issued 15 February 1902.

2. ALEXANDER W. Evans, The Lejeuneae of the United States and
Canada. Pp. 113-183: p/. 16-22. Issued 15 February 1902.

3- ELIZABETH G. BRITTON and ALEXANDRINA TAYLOR. The Life His-
tory of Vittaria lineata. Pp. 185-211. ۸ 23-31. Issued 30
August 1902.

ERRATA
Page 12, I9th line from top, for mother cell of the embryo sac,
read embryo sac cell.
Page 61, oth line from top, for megaspore, read megaspore mother
cell, ۱ -
. Page 77, 2d line from bottom, for 10: 252, rea 350: 228
Page 102, 6th line from bottom, for and, read i”.

| No. Í, PARTI.

THR

COMPARATIVE EMBRYOLOGY

OF THE

RUBIACEAE

By

FRANCIS ERNEST LLOYD

PART I, ISSUED AUGUST 26, 1899,

PART 2, ISSUED. FEBRUARY 15, 1902.

T MUSS
TUAN

AN A A A DEN TEN

|
|

I. INTRODUCTION

The following study was begun during the summer of 1898 in
the laboratory of the Botanical Institute in München upon the
suggestion and under the direction of Professor K. Goebel. Only
a small beginning was made in München on account of a very
brief visit—a few weeks only—to that city. Since then the work
has been carried on in New York City.

In undertaking a comparative study of the embryology of the
Rubiaceae it is expected to gain a knowledge of the life histories of
a number of types sufficient to lead to a clearer understanding of
the relationships between the genera of a very highly specialized
and polymorphous family. It is intended at the same time to ex-
tend the studies into the supposedly allied families, and thus to
contribute, if possible, to the solution of the more general prob-
lems of relationship among the higher dicotyledons.

In addition to the problems of phylogeny there are those which
arise out of the physiological relations existing between the
sporophyte and the gametophyte. Modern morphology has
acquainted us with the main facts in regard to the process of
antithetic alternation of generation in the higher plants, and the

general physiological relations which are the outgrowth of in-

terdependence of sporophyte and gametophyte. Out of this inter-
dependence, however, a large number of phenomena of both a
structural and physiological character have arisen, offering a field
of investigation hitherto neglected, to a great extent, but which
latterly has been attracting not a little attention. The phenomena
of adaptation in this connection are not a whit less interesting if
more difficult to observe. In the present study these adaptive
phenomena, so far as the writer has succeeded in observing them,
will be set forth. It may, however, very properly be borne in mind
that these phenomena are chiefly those of nutrition, and the small
Memoirs, Torrey Botanical Club, Volume VIII.

[No. 1, Part 1 (pp. 1-26), issued August 26, 1899.]

2 THE CoMPARATIVE EMBRYOLOGY OF THE RUBIACEAE

size of the materials makes any other than morphological evidence
difficult to obtain. For example, it is argued from appearances
that enzymes or at least zymogens are present in the young endo-
sperm, though that such is actually the case is not proven. Since,
however, such substances have been shown to be present and have
been extracted from, e. g., the young date seedling * and since the
action of the sucking organ of that plant on the hard endosperm
is due-to the ability to secrete the zymogens of a cellulose-dissolv-
ing ferment, we have a working basis for such interpretation

The position here assumed that morphology must be regarded
from a physiological point of view is the one emphasized in 1879,
by M. Treub in the introductory part of his ** Notes sur l'embryo-
genie de quelques Orchidées," t where he expressed the hope that

the readers of his paper would agree “que la manière dont les -

embryons absorbent les substances plastiques mérite certainement
d'attirer l attention, et sourtout d’être élucidées lors de recherches
embryogéniques.” That this point of view was not new with him,
Treub took pains, in a detailed historical treatment, to indicate,
though it remained for him to adduce the mass of evidence, em-
bodied in the rich work above referred to, which completely sub-
stantiates the value of his position. To this historical review the
reader is referred for an account of the growth of our knowledge
of the suspensor. The succeeding part of the “Notes” deals

with the origin, structure and function of the suspensor in eleven .

genera, including nineteen species of the Orchidaceae, in which
forms the suspensor is shown to build up for the most part elab-
orate structures of a haustorial character which serve to absorb
nutritive substances to be used by the embryo proper in its devel-
opment. These structures, which attain the highest degree of
specialization in Herminium monorchis, Phalaenopsis grandiflora
and Stanhopea oculata, here take the form of tubes (“boyaux ”)
which penetrate into the surrounding tissues of the ovule.

Two years later L. Guignard} (81) published an account of his
quite thorough studies on the embryology of the Leguminosae, in

* F. C. Newcombe. Cellulose Enzymes, Ann. Bot. 13: 49. 1899.
T Nat. Verh. d. K. Akad. van Wet. و1‎ 1879

t Recherches d’ ne en Mee ay Embryologie des legumineuses.
Ann. sc. nat. Bot. VI. 12: 5-166

i

E
1
7
E

a EE OEN‏ ات زا ان

INTRODUCTION 3

many species of which the suspensor attains a morphological
character quite as striking as that in the Orchidaceae. In the
Viciae especially the suspensor produces a great number of nuclei
sometimes not separated by membranes (Orobus). Here, how-
ever, the endosperm is scanty and the complex. suspensor appears
to take its place. Guignard suggests that the principal function
of the suspensor may be to contribute to the nutrition of the em-
bryo, but admits at the same time that in certain cases the char-
acter of that organ indicates that it serves only as a means of at-
tachment for the embryo.

It is of interest in passing to note that Schleiden and Vogel*
(38) saw the complex suspensor in Lupinus rivularis, but inter-
preted it as a development of the pollen tube in accordance with
their then maintained view of the origin of the embryo,

The next work with a distinct physiological bearing dealing
especially with the function of the suspensor was by S. H. Koor-
derst who, in 1892, investigated the embryology of the teak with
especial reference to the behavior of the embryo in the matter of
nutrition. In this plant a most remarkable condition occurs,
which, while entirely analogous to that already found by Treub in
certain orchids, was shown by Koorders to be morphologically dif-
ferent. The latter investigator found that the suspensor in its
proximal, multicellular part is provided at the close of the “ Ko-
gelstadium ” with “ zuigblasen ” or haustoria, which, however, are
here special endosperm cells which attach themselves to the sus-
pensor and elongate into tubular structures. Basing his inferences
on the distribution of food materials in the ovule and embryo,
which he found to be cuticularized till the cotyledonary phase was
reached, Koorders concluded, also, that the embryo received
nourishment through the agency of the suspensor until the ap-
pearance of the cotyledons, after which these organs, especially,
are concerned in the absorption of foods through their peripheral
membranes.

Touching more especially the rôle of the suspensor there ap-
pears to be no further work. The adoption of a similar point of

* وس موز‎ zur e Re = aa bei Leguminosen. Ver-
handl. der K. . D. Akad. d. Nat 39.

De Nie van Tine e t5 ati). Natuurk. Tijdschr. voor
Ned. Ind. III. 12: 139. ۸7. 7-8. 1892.

4 THE COMPARATIVE EMBRYOLOGY OF THE RUBIACEAE

view by workers on related matters has been productive of some
important publications which we may now with profit glance at
briefly.

The first paper, important in this connection, was published by
M. Westermaier in 1892, entitled “ Zur Embryologie der Phanero-
gamen insbesondere über die sogenannten Antipoden.”* He em-
braced in his observations something like thirty four species of
dicotyledons and monocotyledons, and found in the position and
character of the antipodal cells and in the distribution of food-
stuffs in the ovule, evidence that the view advocated by Vesquet
that the antipodal cells are to be regarded as rudimentary, useless,
and of only morphological worth is entirely insufficient to explain
the facts, and that, on the contrary, we have here to do with a
structure which, while in many cases small and apparently insig-
nificant, are really of very considerable importance in the ultimate
nutrition of the embryo. Latterly A. Osterwaldert (98) by the
study of Aconitum Napellus has established and extended Wes-
termaier’s view which he himself in a short note§ again expressed
as the result of observations on Alstroemeria and other plants.

Mottier’s (93) observations on Senecio aureus || and Chamber-
lain’s (’95) on Aster Novae- Angliae Y extended our knowledge of the
antipodals by showing that in these two plants as in some other
Compositae previously studied by others, a greater number of these
cells occur than is usual. This multiplication of antipodal cells
is of physiological significance, a view which is strengthened by
the observations of D. H. Campbell ** (g9) on Sparganium and
Lysichiton, in which these cells are numerous, reaching in the
former the number of 150 at least. The author himself suggests
that this great development probably indicates an important physi-
ological role.

* Verhandl. der K. L.-C. D. Akad, d. Nat. 56 : 1-39. 1892. (The literature
en published regarding the antipodals is found in this paper. )
** Nouvelles — sur le développement du sac embryonnaire des phanéro-
games angiospermes Ann. sc, nat, Bot. 79.
I rtis zür Emte von Aconitum Napellus L. Flora 85: 254.
2W موی‎ ai Max. ** Historische Bemerkungen zur Lehre von der Bedeutung der
Antipoden-Zelle: Ber. d. D. B. G. 16: 214-216. N. 1898.
|| Bot. Gaz. in 245. Jl. 1893.
Bot. Gaz. 20: 205. My. 1895.
** Bot. Gaz. 27: 153. Mr. 1899.

INTRODUCTION 5

The work of perhaps the most interest which has recently ap-
peared is that of Dr. Gabrielle Balicka-Iwanowska + ('99) who has
studied the behavior of the embryo-sac in the Scrophulariaceae, Ges-
neraceae, Pedalinaceae, Plantaginaceae, Campanulaceae, and Dip-
saceae. All these forms possess a more or less thickened single
integument. Micropylar and chalazal haustoria, which are here
for the first time shown to be of endospermic origin, penetrate
more or less deeply into the integument which, under such condi-
tions, is said to be devoid of any vascular tissue. Those plants
which possess the thickest integument have the most highly de-
veloped haustoria, and these in turn have nuclei whose size and
appearance show them to be active in nutrition. Their limiting
membranes, moreover, are quickly “‘gelatinized” or become
* mucilaginous." The haustoria, in some cases, reach the funicle
(Torenia) and even the placenta (Scoparia). In many species a
special nutritive tissue is found in the chalazal region of the ovule
and the haustorium is for the most part in direct relation with
such tissue. The author does not regard the “ tapetes"' as merely
protective to the embryo-sac, as Hegelmaier held, but holds that
they probably secrete a ferment and exercise a digestive function.
The antipodal cells, when present, are regarded as having only a
less important mission, aud are at best evanescent.

While Dr. Balicka-Iwanowska’s paper was in preparation at
München, Mlle. M. Goldflus* (99), was engaged in the laboratory
of the University of Geneva, in a similar investigation with espe-
cial reference to the Compositae. Her conclusion as to the phys-
iological value of the “epithelium,” the “tapetes” of Balicka-
Iwanowska, agrees with that of the latter, an opinion favored,
though not fully subscribed to, by Schwere. She furthermore at-
tributes to the antipodal cells considerable importance as agents
in nutrition. These cells attain in some of the species studied a
considerable size (Leucanthemum lacustre, Chrysanthemum Leucan-
themum, Helianthus Maximiliani) and are interpreted as a haus-
torium (sucoir) inasmuch as they penetrate the axial part of the
ovule, and stand in relation with a mass of conductive tissue which

t Contribution à l'étude du sac embryonnaire chez certain Gamopetales. Flora,
86: 47

* Sur la structure et les fonctions de l’assise épithéliale et des antipodes chez les
Composées. Jour. de Bot. 12: 374. 1898; 13: 9, 47,87. 1899.

6 THE COMPARATIVE EMBRYOLOGY OF THE RUBIACEAE

in turn is in a similar relation with the vascular bundle. Mlle.
Goldflus bases her opinion not alone on the position of the anti-
podals, for she finds in the cyanophily of these cells evidence that
they represent “ l'intermediaire entre le sac embryonnaire et les
substances digestibles élaborées par l'ovule."

The Rubiaceae in particular have received very little attention.

Schleiden * appears to have been the earliest observer whose obser-

vations may be regarded as pertinent to the present work. He
described the ovule in the Rubiaceae as anatropous (gemmula an-
atropa) and as consisting of a naked nucellus (nucleus nudus), 2. e.,
without any integument (p. 304, /.c.). As Warming points out,
it is easily understood how the single thick integument with the
very delicate micropylar canal would have misled the earlier ob-
servers. These results seem to have remained without modifica-
tion till the appearance of his “ Grundriss," 1 in which, however, it
is further stated (p. 163) that in many Rubiaceae the integument
remains at least as a thin skin which clings to the endosperm, from
which it easily falls off in tatters, as in Coffea.

The work of Hofmeister, § as is well known, coveted a very
large field, and of necessity the treatment of any given group was
relatively brief. The results of his work on the Rubiaceae are
quite brief and faulty. It would, however, be very unfair to
criticise him, on account of the difficulties presented by the
materials, especially in view of the methods then used, methods
which, by the way, in the hands of a master like Hofmeister
gave surprising results. According to this pioneer in embryol-
ogy the antipodals are absent from Galum, Asperula and
Crucianella, in which the embryo-sac is attenuated and swollen
at the micropyle end. The embryo-sac is quickly filled with
endosperm after fertilization has taken place. Within the endo-
sperm, the proembryo develops by forming transverse walls.
The cells of a filamentous structure thus formed produce numer-
ous branches made up of short series of cells of which the outer-

nige Blicke auf die Entwickelungsgeschichte des vegetabilischen Organismus
bei den toas: Wiegmann's Archiv 89, 414.
De Povule. Ann, sc. nat. Bot. MI x: 124 18
Grundriss der Botanik. Leipzig Engelmann, 1850.
& Neure Beobachtungen über Embryobildung der Phanerogamen. Pringsh.
Jahrb, 1:82. 1888.

tcc ETE TS Fr Hr

INTRODUCTION 了

most is rounded and swollen. The cells from the middle of these
branches send out secondary branches here and there, so that the
proembryo as a result of this richness of branching simulates a
bunch of grapes. Out of the lower end of this mass projects the
primary cell row of the proembryo whose terminal cell develops
into the embryo proper. Thus Hofmeister's description of the
suspensor in the genera above named, one in the main correct,
but requiring modification, as will hereafter be seen. He passes
over without criticism and accepts as final the opinion which he
attributes to Schleiden that the endosperm breaks through the
integument.

In Houstonia and Spermacoce the embryo-sac is of less slender
form and somewhat cylindrical, and here, on the other hand, the
antipodal cells are present. The embryo-sac is here also soon
filled with endosperm, although the proembryo remains a very
simple thread of cells from the last of which the embryo proper
arises.

It is unfortunate that no figures of these forms are given so
that it is difficult to estimate exactly the value of the author's
descriptions.

The only other plant of the Rubiaceae. which has received any
attention is the coffee (Coffea arabica), but such work as has been
done does not touch upon the ground of the present paper except
in a remote way. An account of the literature of this subject may
be found in the footnotes of an article by T. F. Hanausek.*

I desire at this point to express my appreciation of the kindness
of Professor Ignatius Urban in putting the collection of growing
Rubiaceae in the Botanical Garden at Berlin at my disposal, and
my gratitude to Professor K. Goebel for his generosity in extend-
ing to me the courtesy of the laboratory and gardens of the
Botanical Institute in Múnchen during six weeks of the summer
of 1898, and in expending much of his own time in discussion and
suggestion when ۳ could ill afford, on account of pressing duties,
to do so.

* Ueber symmetrische und polyembryonische Samen von Coffea arabica L. ۰
D: B. 6.29: 73. 1395.

8 THE CoMPARATIVE EMBRYOLOGY OF THE RUBIACEAE

II. DESCRIPTIVE
Vaillantia hispida

Vaillantia is a small genus of the Ga/ieae containing only two
annual species, indigenous to the Mediterranean region. The plants
are monoecious; the flowers are borne in threes supported bya 。
single broadened spiny stalk. Of the three the middle is pistillate, ۰
the two lateral staminate. The three-flowered peduncles are in |
four vertical rows alternate with the four rows of leaves. The
regularity and perfect radial symmetry of the species studied render
it easy to section a growing tip so as to cut all the ovules of two
opposite rows longitudinally.

The material was obtained at the Botanical Garden at Berlin.

THE ORIGIN AND DEVELOPMENT OF THE NUCELLUS

At the period of the development of the pistillate flower when
the four corolla lobes have met above the hollowed out receptacles
and overarch the stamens, which are as yet merely rounded knobs
at the angles of the sinuses of the corolla, four elevations lying in
the same vertical plane make their appearance near the base of the
ovary. The outer two of these are ridge like, and by their mode
of growth meet later in the transverse plane, fuse, extend upwards
to form the two styles, and downward to form a partition which
divides the originally unilocular ovary into two chambers. The
two others are papilliform and lie on either side of the center of the
floor of the ovary. These are the nucelli. As they develop, the
partition above referred to passes down between them and separates
them. Thelower edge of the partition finally fuses with the tissues
between the funicles of thetwo ovules, making the ovary bilocular.

We may now turn our attention to the nucellus. For a short
time its growth is direetly upwards. (Pl. 1. Fig. 1.) There isa
well marked epidermis, and the cells are quite filled with dense
cytoplasm and large nuclei. More rapid cell division upon the
inner side, however, soon causes the apex of the nucellus to be
directed toward the floor of the ovary. When the bending of the
ovule is complete, and its definitive condition reached, the micropylar

VAILLANTIA HISPIDA 9

end of the ovule lies in a lower horizontal plane than does the base
of the funicle. The micropyle and embryo-sac lie in a curved line,
and the funicle remains short. The ovule is therefore not of the
strictly anatropous type, but is campylotropous, approaching the
anatropous condition. (Pl. 3. Fig. 1.) During the process of in-
version the longitudinal axis comes, of course, to lie in the hori-
zontal plane, and about this time a number of the hypodermal
cells under the apex of the nucellus elongate in a direction parallel
with the longitudinal axis of the nucellus. (Pl. 1. Fig. 2.) These
cells, about 12 in number, constitute the archesporial tissue, and
are recognizable at first only by their size. Their subsequent
changes will be described below. The large number of sporog-
enous cells here found to constitute the archesporium recalls the
condition in Rubus caesius, Geum strictum and Sanguisorba officinalis,
in which, according to A. Fischer* ( 80) numerous sporogenous cells
arise. A similarly large number of these cells has also been found
in Casuarina,t in Loranthus} and in Ranunculus. § This character,
which appears in several widely separated families is probably
therefore not to be regarded as a persistent primitive character in
the Rubiaceae at any rate, but one having a physiological meaning,
as will be shown later. The sporogenous cells now elongate, and
their elongation is accompanied'by periclinal division in the epider-
mis and subadjacent cells of the tissue about the apical part of the
nucellus. (Pl. 1. Fig. 3.) Only in the columnar epidermal cells
immediately beneath which lie the sporogenous cells there occurs
no division, periclinal or otherwise. These, therefore, remain as a
cap of cells, about which arises the integument at first as a low
ridge, but gradually growing over the nucellar cap and forward so
as to form a canal which I shall call the micropylar canal, the
outer end of which forms, in the definitive ovule, the micropyle.
The manner in which the integument has its origin recalls
vividly the figures by Warming || (78) of the nucellus of Thesium,
one of the Santalaceae. In the body of the paper the author asks

—

the question “ Are these ovules in which the nucellus is not cov-

* Jenaisch. Zeit. f. Nat. 14: 122

۱ 18803.

t179. 1875.

10 THE COMPARATIVE EMBRYOLOGY OF THE RUBIACEAE

ered by an integument,” and recounts the forms in which, accord-
ing to earlier students, the nucellus is naked. Some reference to
this point has already been made in the introduction. In endeavor-
ing to contribute to the answer of the question, Warming studied
the development of the nucellus in the type above mentioned, and
found that the epidermal cells at the apex of the nucellus, and
therefore, those which cap the elongated cells which lie in the
axis of the structure, have contents more granular than the re-
maining epidermal cells, and also that the epidermal cells sur-
rounding these capping cells divide tangentially and “il se dé-
veloppe ainsi une assez forte couche de cellules de provenance
épidermique qu'on doit considérer comme un tégument." His
description with the exception of the reference to the more gran-
ular contents of the middle epidermal cells is exact enough to

apply to the Ga/ieae, and one cannot resist the conclusion that his

` interpretation of this collar of tissue as a rudimentary integument

is correct. Guignard, * seven years later, published a fuller ac-
count of the behavior of the nucellus and embryo-sac of Thesium
in which, however, he entirely overlooked Warming’s work and
suggestion. Guignard here makes no reference to the tangential
divisions seen by Warming, nor does he show any such in his
figures. Indeed he takes the position (p. 189, 1. c.) that there is
no integument at all in Thesium, though it must be said that in
making the statement he is referring to the seed, rather than to
the young ovule.

The bottom of the micropylar canal is formed of the few epi-
dermal cells capping the archesporium (Pl. 1, Figs. 5, 6,8). The
walls also are formed by an epidermis continuous with and having
its origin in the outer epidermis of the ovule. The endodermis t

elmaier (Ueber die Keimsack einiger Compositen
und dessen Umhiillung. Bot. Zeit 47: So :

; me-
diately the embryo-sac after its absorptio:

889), comes to surround i

5, 82 m
n of the nucellus is complete. For this layer
1 ion and figured being

Hegelmaier proposed the term endodermis, the t 1
Helianthus annuus, Bidens leucantha and
wickelungsgeschichte der Frucht

J Ns MAA
Zinnia tenuiflora,

cell-layer of the nucellus, ** bei i ö
Nucellus," a conclusion whic

OH CARET TUENDA

Zi

VAILLANTIA HISPIDA 11

thus formed does not, however, have the specialized character
which has been found to be so common in the Gamopetalae. It is
not cutinized, and, as will be seen, breaks down very readily under
the action of the developing embryo-sac.

Meanwhile the archesporial cells are EE considerable
changes which will now be described.

At the time of rapid elongation of the nucellus, the arche-
sporial cells, too, increase in length. This is accompanied by a
good deal of change in the appearance of their cytoplasm and
nuclei. The cytoplasm which at first and for some time is granu-
lar, becomes more and more fibrous in appearance. The fibers run
approximately lengthwise. The granular character is still visible,
but the stringiness becomes more marked as the cells themselves
elongate (Pl. 1, Figs. 3-8). When they reach their maximum
length they are spindle-shaped and appear to run under and above
each other in a most perplexing fashion. Their separating mem-
branes become less and less distinct, until they cannot be made
out, and 1 believe them to be almost, if not entirely, absorbed.
The nuclei, too, rapidly increase in size, while the chromatin,
granular at first, runs through various changes, unnecessary to
describe here, which are preparatory to mitosis.

It should be pointed out here that the nucellar tissue next to
the archesporial tissue on its inner side is made up of cells which
lengthen much more than those of the rest of the nucellus (Pl. 1,
Fig. 5). The innermost of these are, indeed, as a usual thing, as
long as the archesporial cells and appear to partake to a great
degree of their characters. That the nucellar cells next to the
archesporial cells may approach the latter in character has been
noted by Coulter (l.c.) in Ranunculus. They may without doubt
be regarded as underdeveloped sporogenous cells, which in this
case disintegrate, while their substance probably contributes to
the growth of the archesporium.

When the archesporial cells as above mentioned have reached

layer he proposed the term endothelium, a term which is evidently inappropriate.
Schwere's figures as to this detail are very unsatisfactory and unconvincing, while it is
quite evident from Hegelmaier's drawings that the layer to which he gave the name
endodermis is in reality what he thought it to be. This layer is, moreover, an inner
layer, so that epithelium, the term یز‎ by Mlle. Goldfius, seems to me a less appro-
priate term, because more gen

12 THE COMPARATIVE EMBRYOLOGY OF THE RUBIACEAE

their greatest elongation, mitosis takes place. I shall take the
opportunity of describing the cytological details in a separate
paper. For the present purpose it is enough to say that so far as
my own observation goes four megaspores result from the division
of each primitive cell. The appearance of the mass of cells which
results from the division of the archesporial cells is very similar to
that presented by similar cells arising in the same manner in Heli-
anthemum Rhodax as described and figured by A. Fischer* accord-
ing to whom two or three “ mother cells ” are in this form cut off,
which divide each into four or sometimes six daughter cells (meg-
aspores) one only of all these resultant cells becoming the mother
cell of the embryo-sac, the rest suffering disorganization.

Several or all of the archesporial cells may divide; thus may
many megaspores arise,among which there is a considerable discrep-
ency in size (PI. 1, Fig. 9). Of one quartette arising from one mother
cell, the lowest cell may be the largest; in another, one of the
middle ones; then again the uppermost may be the largest. I
have not been able to determine whether there is any constancy in
the selection of a megaspore to be the mother cell of the embryo-
sac. The appearance of such material as I have examined seems
to favor the opinion that any one of them, but more especially
one of the end cells of the quartette, may be chosen.

As a rule only oxe of all the megaspores actually germinates.
Exceptionally in other genera to be hereafter described, I have
found two embryo-sacs completely formed and lying parallel ; one

of these usually lies farther forward than the other, and this one is
functional.

THE DEVELOPMENT OF THE EMBRYO-sAC

At the completion of megaspore formation, the chosen mega-
spore commences a migration along the micropylar canal formed
as above described by the forward growth of the integument. This
migratory movement of the functional megaspore appears to be
unique, for I have not been able thus far to find any reference in the
literature which shows that such migration has hitherto been ob-
served. The cap of epidermal cells which overlies

* Zur Kenntniss der Embry
Zeitschr. 14: 90. 1880

the archesporium

osackentwickelung einiger Angiospermen. Jenaische

۰

as A es‏ زر

"eor.

VAILLANTIA HISPIDA 18

seems to be a barrier to any such migration, for the migrating mega-
spore, so far as 1 have been able to determine, pushes its way to
one side, as shown in Pl. 2, Fig. 1. It is reasonable to conclude
that these capping cells are in a short time destroyed, for in later
stages they are not usually seen. I have seen these cells once in
one species of Galium, when the embryo-sac was mature.

The nucleus of the migrating megaspore, after passing this
barrier, comes to lie in the micropylar canal about midway its
length, while the accompanying cytoplasm lies in a thin layer on
the walls of the canal, and so forms a large vacuole at each
end of the nucleus. The whole megaspore is, roughly speaking,
hour-glass-shaped (Pl. 2, Fig. 1), and its nucleus lies at the con-
striction. The enlarging megaspore encroaches laterally on the
integument, which now for the first time is broken down. The
endodermis gives way first, and is recognized a little later only by
collapsed walls and the nuclei, which are the last elements of the
cells to disintegrate. The persistence of nuclear substance under
the action of digestive ferments suggests that the tissue which be-
haves thus is undergoing digestion. This is the view taken by
Mlle. Goldflus (A c.) in describing the behavior of the integu-
mental tissue about the base of the embryo-sac in the Compositae
the internal part of which is dissolved by “ferments secrétés prob-
ablement par les cellules epithéliales.”

At the constriction in the hour-glass-shaped megaspore the
first division of the megaspore nucleus takes place, after which one
of the daughter nuclei passes forward to the micropylar end of the
cavity. When it reaches this point the second division takes
place in both nuclei (Pl. 2, Fig. 2). There are now four. The
egg apparatus arises in the usual fashion at the micropylar end
(Pl. 2, Fig. 3, 4). In the antipodal region a more unusual state
of affairs occurs.

THE ANTIPODAL APPARATUS

The division of the antipodal nuclei, now two in number, re-
sults in four, one of which is, in the usual manner, concerned in
the formation of the endosperm nucleus. This: passes forward to
coalesce with the corresponding nucleus from the egg region. Of
the three which remain, one enlarges considerably and migrates

14 THE COMPARATIVE EMBRYOLOGY OF THE RUBIACEAE

backwards towards the chalazal region. Its proper cytoplasm be-
comes cut off by a transverse wall from the upper part of the em-
bryo-sac (Pl. 2, Fig. 4). The other two antipodal cells surround
themselves by walls, and take ultimately a lateral position. They
are small and unequal in size and are usually more or less col-
lapsed and apparently of little or no further value; and they over-
stain in the manner of disintegrating cells. Not so, however, with
the odd, basal antipodal, for in these forms (+. e., Vaillantia, Ga-
lium, etc.), it appears to have a distinct and important physio-
logical role.

It has been supposed, since the publication of Hofmeister's
classical researches on the embryology of the Phanerogams (/. ¢.),
that antipodals were absent in those forms of the Rudiaceae studied
by him, excepting Houstonia and Spermacoce. I have confirmed
Hofmeister’s view as to Houstonia, but in addition I am prepared
to show that antipodal cells are present in all Rudiaceae up to this
time studied by me. The smallness of the structures in these
plants, and especially the attenuated character of the basal anti-
podal, together with the similarity of the other two to the adjoin-
ing disintegrating integumental cells will readily account for the
oversight.
The basal antipodal cell is composed at first of cytoplasm which
is tenuous, scanty and difficult to see. The nucleus comes to rest
on one side of the cavity some distance removed from the trans-
verse wall.

Gradually the cytoplasm becomes more pronounced, stains more
deeply and fills the elongating chalazal end. The apex becomes
somewhat expanded into a knob which, as the cell lengthens, is
plunged into the mass of megaspores. These are still evident, but
are now on the way to disintegration as is attested by the fact that
they lose all structure and become black and opaque with haema-
toxylin, often making it very difficult to recognize the apical,
knobbed part of the long antipodal. At the time when the basal
antipodal is first cut off, there is a large mass of proteinaceous
food in the place of the once archesporium, which as just stated,
stains very deeply. As the embryo-sac as a whole reaches ma-
turity, before the flower opens, the cytoplasm about the endosperm
nucleus becomes abundant and dense, and the archesporial food-

VAILLANTIA HISPIDA 15

mass is coördinately reduced, while the basal antipodal shows its
maximum development during this period. The probe-like end,
which, I believe, serves as a haustorium, at this time and some-
what later is filled with cytoplasm which appears finely reticulate,
while the broad end which is adjacent to the upper part of the em-
bryo-sac is occupied by a large vacuole. This large antipodal
must, therefore, stand in an important relation between the food
supply derived from the archesporium and the endosperm cell,
and is probably active in the transportation of the food from the
former to the latter. The megaspore mass thus forms a sort of
nutritive tissue, partly analogous to that described by Dr. Balicka-
Iwanowska (/ c.) as occurring in the region of the integument
near the lower ends of the embryo-sac in the forms studied by
her.

The archesporial tissue is now seen to bear a nutritive rela-
tion to the embryo-sac, which by this time has reached maturity.

THE DEFINITIVE EMBRYO-SAC

We may now pause to sum up the characters of the mature
embryo-sac. E

In shape it is fusiform and much attenuated at the lower end.
The upper part, comprising one third or more of the entire length,
contains the egg apparatus and the endosperm nucleus, and consti-
tutes the expanded part of the embryo-sac. The fusion of the
polar nuclei takes place at about the middle of this part of the
embryo-sac, but at the time of fertilization the resultant nucleus
comes to lie immediately against and partly surrounding the egg
nucleus. The migration of the endosperm nucleus has been no-
ticed by Mme. Balicka-Iwanowska (l. c.) in Digitalis and other
related forms in which the embryo-sac is more or less attenuate.
At the time it reaches the egg, the nucleolus has nearly three
times its original diameter. Its size relative to that of the egg
nucleus is shown in Pl. 2, Fig. 8a. The nucleus itself is larger
than the egg nucleus, but the discrepancy in size is more striking
between the nucleoli. This movement and growth of the endo-
sperm nucleus must be interpreted physiologically ; it is in all
probability in some way connected with the nutrition of the egg.

The antipodals are three in number; two of these are small

16 THE COMPARATIVE EMBRYOLOGY OF THE RUBIACEAE

and relatively inconspicuous; the third and basal, is very long,
being on the average about two thirds the length of the whole em-
bryo-sac, though it may, in exceptional cases which I have noted,
attain a considerably greater length. At its expanded upper end it
is separated from the upper region ofthe embryo-sac by a very thin
transverse wall. Its lower end is somewhat knobbed, and is em-
bedded in a mass of degenerating megaspores. The position of
the nucleus of this antipodal is always lateral and about one fourth
to one half the cell’s length from the upper end (Pl. 2, Figs. 7-11).
It is large and prominent, and evidently active. Its function has
already been discussed. The fate of the antipodals will be discussed
somewhat later.

DEVELOPMENT OF THE EMBRYO

After the egg is fertilized, its vacuole enlarges considerably,
pushing the cytoplasmic end rather more deeply into the endo-
sperm stuff. Immediately the endosperm nucleus commences to
divide so that before the’ first division of the fertilized egg takes
place 20 to 30 endosperm nuclei are present (Pl. 2, Fig. 10). The
proembryo now grows rapidly by the successive occurrence of
approximately transverse divisions. When about five cells have
been formed, the proembryo becomes bent to one side and the
other. This is due to the lateral enlargement of certain of the sus-
pensor cells, as a result of which they bulge out into the endo-
sperm which is now of some bulk. In these enlarging suspensor
cells, the cytoplasm becomes vacuolated to a considerable degree.
These facts and later developments indicate that these cells are
concerned in the absorption of foods in solution from the endo-
sperm, which, however, is not broken down by the enlarging sus-
pensor cells, The cells of the end of the embryo away from the
micropyle are always smaller during the earlier development of
the proembryo, while they get successively larger toward the
micropylar end where the basal cell tapers off acutely, giving the
young proembryo something of a spindle shape (Pl. 4, Figs. 1, 2)-

When the proembryo attains the number of about six cells,
longitudinal divisions occur in some of the suspensor cells. The
divisions sometimes occur in two planes, and the resulting elements
continue to grow and bulge out further into the endosperm and

in a een

VAILLANTIA HISPIDA 17

form haustorial appendages. In some cases when two longitu-
dinal divisions have taken place in a single suspensor cell, a pair
of the resulting quadrants grow out together, forming double
haustoria (Pl. 4, Fig. 5). Again the haustorial cells may divide
transversely ; this occurs only in the region near the embryo proper.

These haustoria, for such I regard them, are at first pro-
nouncedly vacuolated, but as the embryo ages the vacuolation
disappears, and they become densely filled with cytoplasm. After
the embryo proper begins its development, the basal suspensor
cell elongates considerably, increasing its length by a half and be-
comes less vacuolated. Its micropylar end becomes crushed by
the growing endosperm, which now begins to grow into the
micropylar part of the thickening integument, leaving the embryo
in a more central position (Pl. 4, Figs. 9, 10). Later, the basal
cell and even one or two of its neighboring suspensor cells become
disorganized, but the time at which this occurs is not constant, for
while I was not able to find these cells in the preparations from
which Figs. 5 and 6, Pl. 4, were drawn, I did find them in another
specimen of such an age as the one shown in Pl. 4, Fig. ۰

After the embryo proper has differentiated itself from the sus-
pensor, the latter may be seen to be divided clearly into two
regions which differ very markedly in structure. These two
regions may be termed the micropylar and embryonal. The
latter is made up of half a dozen disc-shaped cells which are very
much flattened longitudinally, and show no outgrowths and are
not vacuolated, and they remain so during the entire embryonic
history. The cells of the micropylar region, on the other hand,
become larger and thus increase their absorbing surface. When
they reach their maximum development the suspensor when dis-
sected out resembles a “bunch of grapes." These outgrowths,
to which we have ascribed a haustorial function, are not unlike
the suspensorial haustoria (Keimtrágerblasen, boyaux) described by
` Treub in the Orchidaceae and by Hofmeister“ (49) and Guignard
(2. c.) in Sutherlandia, and are analogous to the “ zuigblasen” oc-
curring in the teak (Tectona grandis) described by Koorders (92).
The upper part of the micropylar region, as shown above, becomes
less important and is finally lost by absorption, so that Ihe emo

* Die Entstehung des Embryo der Phanerogamen. Leipzig, 1849.

18 THE COMPARATIVE EMBRYOLOGY OF THE RUBIACEAE

with its suspensorial haustoria, to which it is attached by the short
row of disc-shaped cells, comes to lie entirely within the now large
mass of endosperm (Pl. 4, Figs. 9, 10). The structures above
described are probably to be regarded as an adaptation for the
ready absorption of food for the rapid growth of the embryo.
The shortness of the cells of the embryonal region of the sus-
pensor favors the passage of food tothe embryo. Whether or not
the embryo is capable of absorbing food through its own surface is
a question which I am unable to answer at this time. Koorders
(92), working on Tectona, found that the embryo till the com-
pletion of the ** Kogelstadium ” possessed cuticularized walls, thus
preventing the absorption of food. If this be true of Vaillantia
and its close relatives, we have a complete explanation of the sig-
nificance of the haustoria. But even if the embryo is not cuticular-
ized, it is not improbable that the amount of absorbing surface on
the embryo proper alone is insufficient for its rapid development,
and it may be that in this lies the explanation of many features in
other forms not yet cleared up. For example, I have found that
the basal cell in Capsella, figured years ago by Hanstein, and
copied in almost every text-book, actually pushes its way into the
adjacent tissue of the inner integument in the same manner that
the embryo-sac itself, by its growth, in many cases encroaches
upon the surrounding tissue. This fact indicates that the sus-
pensor, in forms in which it reaches so pronounced a development,
as in the Cruciferae, Tropaeolum and others, is of importance in
absorbing nourishment, an opinion which we must credit to Meyen
(39).* I believe this action of the suspensor in the surrounding
tissue to be correlated with the rapid growth of the embryo in
Capsella. Before drawing any conclusions of this kind, however,
it will be necessary to obtain many more data.

There can be little doubt that the rapid development of seeds
is an important adaptive feature, and the study of structures which
are correlated with this ability will be a fertile field for further
study. Such haustorial structures as those found in the Orchida-
ceae may be correlated with the meagre integument which means
a meagre store of food immediately at hand.

The further development of the embryo, which brings the

* Neues system der Pflanzen-Physiologie 3: 331. 18 39.

|
|
|
|

VAILLANTIA HISPIDA 19

cotyledons into view, is accompanied by the dissolution of a zone

of endosperm surrounding the embryo. The cells become mu-
cilaginous, and the embryo is bathed in nutrient substance. The
relative increase of absorbing surface by the development of
the cotyledons is accompanied by the gradual loss of the haus-
toria of the suspensor, of which there may be found, at the time
of the ripening of the fruit, only a few of the disc-shaped cells.
The last of these plays a small part in the formation of the radicle.

THE INTEGUMENT

The earlier development of the integument has already been
described. Some time before complete maturity of the embryo-
sac is reached the cells which form a zone (d, Fig. 1, Pl. 3)
about the attachment of the funicle to the ovule become large and
rounded (Pl. 3, Fig. 1, ۵ ( and their contents become strongly vac-
uolated. These cells contain a good deal of starch at quite an
early condition of the ovule, and when the accumulation of starch
in the integument commences, the process has its beginning here.
At the end of the vascular bundle when it reaches this zone of
enlarged cells, is found a mass of cells which gives the bundle a
club-shaped appearance in longitudinal section. These cells,
like those of the leptome of the bundle, are rich in contents, and
form a center for the distribution of food, which is received by the
large-celled zone above referred to, the function of which appears
to be the preparation of food and the regulation and distribution

of the supply to the meristematic zone. The latter is made up of

smaller cells (Figs. 12, 1c, Pl. 3), in which, after fertilization
multiplication begins. Their dividing walls are periclinal, and
there results a rapid thickening of the integument, with a corre-
sponding increase of the ovary wall, accompanied by secondary
changes leading to the formation of a schizocarp. Within the
meristematic zone the integumental tissue shows signs of disin-
tegration, and this process appears to be as rapid and complete in
the tissue surrounding the basal antipodal as in that about the
endosperm region. We may notice in this connection that the
mass of disintegrated megaspores has after fertilization and before
the first division of the proembryo almost disappeared (Pl. 2,
Fig. 9; Pl. 3, Fig. 3) and the antipodal cells are difficult to

20 THE COMPARATIVE EMBRYOLOGY OF THE RUBIACEAE

see. In some genera they persist for some time, and vestiges may
. be seen after a large mass of endosperm has been formed. After
the megaspore mass is absorbed, the basal antipodal appears to
have no further function.

When the embryo starts to develop a small amount of starch is —

found in the integument, principally in those cells above referred
to which form a zone between the funicle and the meristem of the
integument. A very few granules are also to be found in the cells
immediately about the embryo-sac. As the seed approaches ma-
turity the starch accumulates first in the peripheral cells until all
those cells of the integument which remain are replete.

THE GROWTH OF THE ENDOSPERM

During the first division of the endosperm nucleus the endo-
sperm enlarges, becoming rounded, and encroaches upon the ad-
jacent tissue of the integument, the cells of which become flattened
and give way. At first, separating walls between the cells are not
formed. The nuclei are not parietal, nor do they ever become
so. -The endosperm mass is always solid. A little later the
parietal cells become much more densely filled with cytoplasm
and stain more deeply (Pl. 3, Figs. 20, 2c). This is especially true
of those cells on the side of the endosperm mass facing the longi-
tudinal ‘axis of the ovary, which are the earliest to become dense.
It is on this side that the thickness of the integument is the greatest
and where the food is most abundant. After the endosperm ap-
proaches its limit of growth, these are the last of the peripheral
cells to lose their dense character. The inner cells of the endo-
sperm are strongly vacuolated, the vacuolation becoming more
pronounced towards the center, where the cells ultimately break
down and leave a cavity within which lies the now rapidly matur-
ing embryo. The peripheral cells are digestive, I believe, and ab-
sorptive, for the integuments break down in advance of the growth

۱
3
1
|

of the endosperm, and the cell walls of the disorganized cells be- _

come thickened. These thickened walls stain readily and deeply
with safranin, while the walls of the normal cells show much less
affinity for that stain.

As the endosperm ages it becomes concave on its inner

face, and dipping into the concavity is a mass of integument which

VAILLANTIA HISPIDA 21

remains undigested at the maturing of the fruit. (In Pl. 3, Fig. 2.)
This mass is continuous with a thin layer of integument two cells
in thickness, which surrounds the endosperm. The latter does not,
contrary to the opinion of Hofmeister, break through the integu-
ment, so that the envelopes of the mature fruit consist of the wall
of the ovary lined by an integument (Pl. 3, Figs. 2, 24) two cells
thick and loaded with starch. The endosperm cells up to a time
when the embryo is large and approaching maturity, have thin
walls. Starch accumulates more and more till the cells are gorged.
Just before the complete maturity, however, the walls thicken so
that in the mature seeds the principal foods stored in the endo-
sperm are reserve cellulose* and starch.

The central part of the endosperm is, up to the completion of
the sphere stage of the embryo, continuous, so that the embryo
lies in direct contact with the endosperm cells. When, however,
the cotyledons begin to appear, the endosperm immediately sur-
rounding the embryo breaks down, so that the embryo comes to
lie freely in a cavity which is filled with fluid materials. The be-
havior of the endosperm in this regard, as also the integument dur-
ing the advance of the endosperm upon it, indicates the presence of
digestive agent secreted by the embryo on one hand and the
endosperm on the other.

* The term “reserve cellulose’ is used here provisionally. The exact nature
of the substance in the Ga/iene i is, 1 believe, not yet determined.

Mem. Torr. Bor. CLUB, 8.

Explanation of Plate 1

Fig. 1. The fundament of the nucellus, 670

Fig. 2. The archesporial cells, of which only two are seen in the section, have be-
gun to be differentiated. They apparently underlie only one epidermal capping ce ell.

Fig. 3. The archesporium more strongly developed. The p ericlinal divisions
seg the capping epidermal cells mark the beginning of the ی‎ of the integu-

ie. 4. Transverse section through archesporium ; it is composed in this instance of
twelve cells, X 670.
ig. 5. Integument is well marked; maximal elongation of archespori
disintegration of some of the inner cells has begun, X 670. |
Fi Somewhat older condition of the ovule in which the archesporial cells are
preparing for division ; micropylar canal is begun, X ۰
Fig. 7. First division of the archesporial cells. The arrow is directed toward the
position of the micropyle, X 670.
Fig. 8. A still more ارت روت‎ condition of the ovule in which the micropylar pi
is. > The سس‎ division is taking place in some of the archesporial cells, X ۰
. 9. Definitive megaspore. On the right one sees imperfectly formed and de-
97 megaspores, X 670,

al cells;

(23)

Explanation of Plate 2

. The embryo-sac mother cell has mi egies down the micropylar canal, hav-
ing aban to one side of the nucellar cells which cap the archesporium. The remain
ing pago are beginning to disintegrate, Xx 670.

re . Embryo-sac with four nuclei, X 670,

. Embryo-sac with eight nuclei. One of the antipodals is cut off from the

seven nuclei by a transverse wall, 670.‏ بت
ig. 4. Egg apparatus and antipodals definite. Approach of polar nuclei, 472.‏
Fig. 5. Fusion of polar nuclei completed. The position of endosperm nucleus is‏
near the middle of the sac, 472,‏

. 6 7. Approach of endosperm nucleus toward the egg, and their con-
cómitant increase in size, X 472.
ig. 8. Mature embryo-sac.
X 670.

Fig. 8a. Egg and e ndosperm nuclei of the same, 1350.

Fig. 9. After fertilization ; embryo is still sais called y several endosperm nuclei
are now present; the disintegrated megaspores are now duty absorbed ; the endo-
sperm سم‎ to encroach on the integument,

Fig.

The endosperm. nucleus is now applied to the egg,

o. Embryo two-celled. The دی میج‎ in which walls begin to appear,
rts further upon the integument, X 315.
Fig

. II. Three antipodal cells, X 670.

(24)

Mem. Torr. Bor. CLUB, 8.

cy. ao

-

Ed.

1 Z aa, EE.

e WA
۵ CESS a

zd مس‎
we 4 = :

VAILLANTIA HISPIDA

=

nn
OS

- LA پت‎
Oo

VAILLANTIA HISPIDA

-ssy

STA

-
=

—L Lem)
— سوه‎
— A:

مر

Mem. Torr. Bor. CLUB, 8.

Explanation of Plate 3

Fig. 1. Ovule a after fertilization showing areas made up of cells of different
character. (2) Zone of cells dividing mostly by periclinal walls (see Fig. 12). (4, c)
one of rapidly یت‎ cells (see Fi = 14, Ic) forming a puree completely sur-
rounding the internal mass including the embryo-sac of cells lying be-
tween the funicle and the meristem, made up of TUN cells "en Fig. 14) containing
ROE cytoplasın and starch grains, X 75.
igs. Ia, Û, can 670
ig. 2. Longitudinal section en mature fruit, in. integument, en. endosperm,
em. embryo. The pericarp is seen as a thin layer without the integument.
Fig. 24. Envelopes of mature fruit in detail. /. pe pericarp, 272. integument, ev.
endosperm. The cells of the integument are still supplied with much starch, 220.

EV
Fig. 3. After fertilization. The basal cid has its haustorial end embedded in
a mass of ۳ substance which results from the disintegrated megaspores. Em-
bryo wn one-celled, X 670.
4. Young embryo lying in the endosperm. The latter by its Be pushes
past = nes antipodal into the integumental tissue as toward e, Fig. 1

(25)

Explanation of Plate 4

Fig. ۰ Proembryo of 5 cells, X 670.

Fig. 2. |
3. 上 X 472
5 Successiveiy older stages of the embryo.
3 X 5
a
$. X 28

ka pd r4. Embryo, drawn on larger scale in Fig. y A lying in the endosperm.
3, ER

Br of these last two figures shows the position taken di: the proembryo ]
as a result of the growth of the endosperm at the micropylar region

(26)

PL. de

Mem. Torr. Bor. CLUB, 8.

VAILLANTIA HISPIDA

Memoirs oí the Torrey Botanical Club
The MEMOIRS are A ab d ei intervals and are sold at a uniform price
of THREE DOLLARS a Mos the artieles can be furnished separately

at the prices indicated below. me price a pos
Volume

No. 1.—Bailey, L. H. Studies of the Types of various — of the Genus Carex.

Pp. 85. 1889. Ss be furnished except with full volume.
No. Martindale, I. C. Marine Algae of the New aged Coast and adjacent
Waters Ws Staten Inland. een 87-109. 1889. Price, 50 e
No. 3.—Spru An Enumeration of the ópticas e by Dr. H. H.
Rusby in South ای مج‎ with Description of many new Species. ۰ 113-140. 1890. .
Price, 75 cen
No. 4.—Sturtevant, E. L. On — Fruit. Pp. 141-187. 1890. Price, 75 cents. -

e 2.

No. 1.—Halsted, B. D. On reserv “Food — in Buds and surrounding |
Pp. PL. 1 1, 2. 1890. Price, > 50 ce
No. 2.—V: A. M. and Hol lick, A. وتو‎ to the Botany of vp ;
Pp. dee ns 8, 7 1890. Price, 75 9
No. 3.— Holm, T . Contributions © js Su go of the Germination of 0
N merican Plants. Pp. 57-108. Pl. 5-19. 1891. (Cannot be furnished ise
with dp volume. ER
— Whee A Monograph of edt North American Species of the
SM hie P M9 12. 1891. Price, 75 cen
olume 3
No. 1.—Small, J. K. and Heller, A. A. = the u e Western North Carolina
and ine Territory. . 1-39. 1892. ce, 50 ce
: o. 2.—Morong, T. A Revision a the ی‎ A por aliado with Illustra-
tions ‘of all the Species. Pp. 1-65. Pi. 20-74. 1893. Price, $2.
„No. 3.—Rusby, H. H. An Enu ea of the Plants collected in Bolivia by
Miguel Bang. Pp. 1-67. 1893, us ids

No. 1.—Underwood, L. M. M Sie OPES Part 1, Bibliography. P 1-91.
1893. Price, 75 cen = نی‎ E
No. 2.—Vail, E M. and Small, J. K. Report on the Botanical Exploration of = B
ginia during the Season 1892. Pp. 92-202. PI. 75-82. 1893-1894. Pri 5 50 ۰
No. 3.— Rusby, H.H. AnE Boli by Miguel
Bang—If. Pp. 203-274. 1895. Price, 50
= 4.—Pettit, A. S. Arachis ی‎ Pp. 275-996. Pl. 83-85. 1895. Price

No. 5 .—Rydberg, P. A. Mon h of Ph Pp. 291-
374. 1896. Price, 75 cents. ograp ysalis and related Genera. ۰

noc Volume 5.

ist Pictor der ze (€ phyta growing without Cultivation in Northy 1
orth Ameri pared by — of the Botanical Club, Ameria i
Association for the nas of Scien Pp. 1-377. 1893-1894. Price, $3. وا‎
ooo 6. 3

o. 1.—Rusby, H. H. An Enumeration of the Plants collected in Bolivia by .
Miguel Bang—III. Pp. 1-130. 1896. Price E

No. 2. rdum A.J. A Revision of the “North American Isotheciaceae and Brachy- 、

` $heeia. in 1897, Ses 50 cents. 1
oF = Ay History - Eu gag toed enn ( Haematococcus E
pia). 87 (colored). Price, 50 cents. d

- Nos. 4 and 5 do the volume will be ند‎ during 1899

olume 7.
Howe, M. A. The Hepaticae and = n.o ]
122. 1899. Price, $3.00 ee and Anthocerotes of California. Pp. 1-208.

Volume 8.
Pp. at = T 1899. he s Comparative Embryology of the Rubiaceae. Part I.
re Sa Club also publishes the Bulletin of the Tos Botanica cal Club,
monthly, $2.00 per year. Vol. 25 for 1898 conta tained 650 pages and 29 id
AA L. M. UNDERWOOD, porte
Col nmbia University, New York mr

FURTHER DESCRIPTION OF GENERA

Callipeltis cucullaria
The genus Callipeltis embraces three species indigenous to the
Mediterranean region. In general habit they resemble closely the
plants of the genus Ga/zum, from which, however, they differ in
their smaller size. Callipeltis cucullaria, the species studied, was
collected in growing condition in the Botanical Garden at Berlin.

THE NUCELLI

The nucelli, two in number, arise as in Vaz//antia, on either side
the center of the floor of the ovary. Nor is their course of de-
velopment different from that already described for that genus,
excepting in matters of only minor moment, and these are to be
found in the total number of cells in the nucellus, the greater rela-
tive amount of archesporial tissue, and in the shape of the
definitive ovule.

The archesporium, the cells of which are more numerous than
in Vaillantia, is evident when the nucellus is quite small ( ۸/۰ 5, fig.
7). The epidermal capping cells remain without any further
growth, while immediately around the area occupied by them, the
integument grows out, as is made evident by the numerous peri-
clinal divisions to be seen at this stage (pl. 5, figs. 2 and 3).
By subsequent growth, the integument is developed so as to
shut in a long micropylar canal (p/. 5, fig. 4). As the ovule
attains its definitive form, the absence of the large special cells
seen in the ovule of Vaillantia is noticed, and the sporogenous
cells of the archesporium divide. The division into four meg-
` aspores is by no means complete for each cell. Some of
them, the more centrally placed, divide more rapidly and com-
pletely, while the more peripheral lag behind and sometimes fail of
division, or perhaps divide but once, and so a heterogeneous
mass results, out of which the megaspore which develops into
the embryo-sac makes its way into the micropylar canal. In such
Memoirs Torrey Botanical Club, Volume VIII.

[No. 1, Part 2 (pp. 27-112), issued Ei February, 1902.]

28 COMPARATIVE EMBRYOLOGY OF THE RUBIACEAE

a mixture of cells it is not easy to distinguish this cell, especially
as it often happens that several pass through the initial stages of
migration, and the result is very confusing. I have, however, with
reasonable certainty, been able to pick out the most active cell, as
in figs. 4 and 5. The embryo-sac cell, so far as determined, may
arise out of any of the megaspores, so that we may regard these
for the greater part as physiologically equivalent. The appear-
ances presented by the archesporium at this time match closely
those seen in Alchemilla as described by Murbeck.* Tapetal
cells are not formed.

No starch is seen in the integument until the migration of the
megaspore, when a slight accumulation occurs, evenly distributed
throughout the whole. During the maturing of the embryo-sac,
the amount of starch increases in a zone surrounding it. As the
endosperm starts to grow, zonal accumulation disappears and a
general distribution is again seen which becomes gradually more
pronounced in the peripheral parts.

THE DEVELOPMENT OF THE EMBRYO-SAC

One of the megaspores which have developed as described
above commences to enlarge and to migrate along the micropylar
canal. This takes place when the ovule is well developed and has
taken on its mature form. 应 5, fig. 4, shows these conditions
just previous to the migration of the megaspores. It not infre-
quently happens in this species that two or even three megaspores
commence to migrate at the same time and take paths which are
nearly parallel. In such cases a good deal of confusion arises,
which makes it difficult to interpret properly the preparations
showing this condition. The remaining megaspores retain their
character for some time, some degenerating more rapidly than
others, and the resulting mass adds not a little to the difficulty of
making out clearly what one sees. Occasionally two of the mi-
grating megaspores germinate, although in no cases which I have
observed has development passed beyond the four-celled stage
(Pl. 5, fig. 8) in any but the more advanced and normal one,
which is set apart as the fundament of the functional embryo-sac.

* Murbeck, Sv. Parthenogenetische ها زیون‎ in der Gattung Alchemilla.
Acta Reg. Soc. Physiogr. Lund, 11 : 1- -47. 1901.

CALLIPELTIS CUCULLARIA 29

More than one embryo-sac may sometimes reach perfect devel-
opment, thus supplying the conditions for false polyembryony.
This, however, has not been observed.

The movement of the megaspore down the micropylar canal is
accompanied by histolysis of the endodermis, the cells of which
degenerate and collapse. The contents of the cells are disorgan-
ized and become deeply staining. It is to be noted that the nu-
cleus is preceded in its course by a part of the cytoplasm, while
the nucleus itself travels unusually near, or along the side of the
cavity formed and some little distance back from its apex )۸ 5,
jig. 6). The amount of cytoplasm is now not great, nor does it
increase markedly in amount until the complement of embryo-sac
nuclei is reached. The cells of the egg apparatus, as soon as they
are cut off, become rapidly different in character from the rest in
attaining greater density. Their final character is reached at
about the time of fusion of the polar nuclei or somewhat later.
The synergids then are vacuolated at their free ends, while the
egg-cell has a vacuole at its narrow basal end. The fusion of the
polar nuclei occurs near the middle of the endosperm cavity (pl.
5, fig. 8) and is followed by a movement of the resultant nucleus
toward the egg-cell with which it lies in contact till its first di-
vision. Meanwhile the cytoplasm loses its vacuolated character
and becomes more dense and gorged with food material, consist-
ing, in large part, of starch. The whole endosperm-cell becomes
more rounded also.

THE ANTIPODAL CELLS

When the fusion of the polar nuclei is taking place the anti-
podal nuclei have arranged themselves in their definitive positions,
two of them close together, each with its proper mass of cytoplasm
which is thin and vacuolated, and the third some distance away in
the narrowing end of the somewhat irregular cavity formed by the
megaspore in its migration (A. 5, fig. 9). The cytoplasm of this
latter antipodal cell is at first extremely scanty, so much so that it
cannot be detected with a high power except near the nucleus. The
two small antipodals secrete cell walls and remain in a normal
condition for some time during the development of the embryo,
Their cytoplasm becomes more dense from the time they are first

30 CoMPARATIVE EMBRYOLOGY OF THE RUBIACEAE

cut off, and numerous starch grains may then be seen. The long
antipodal which occupies about three-fourths of the length of the
entire embryo-sac takes on a definite club shape at its free ex-
tremity which reaches quite to the end of the megaspore cavity,
and is finely granular, though not so densely filled with protoplasm
as are the others. A large vacuole appears constantly at the thick
end which abuts against the small antipodals. The vacuole ex-
_tends as far as the nucleus. The interior of the narrow upper
part is also occupied by a vacuole, though this is less evident in
sections on account of the slender character of this region of the
cell. A normal appearance is presented by this cell as by the
other two until the proembryo has reached an eight- or.ten-celled
stage or even older (p/. 6, fig. 5), and an application of the iodine
test demonstrates that at such a time the small antipodals are
gorged with starch while the other shows a similar starch content
at its thick end. The haustorial end stains deep yellow and is
free from starch. No trace of the deeply staining products of dis-
integrated megaspores may be seen after the second or third divi-
sion of the endosperm nuclei, though this is, of course, a somewhat
variable matter. Disintegration of the antipodals sets in with the
rapid extension of the endosperm, and they are finally lost to view.

ENDOSPERM AND INTEGUMENT
Soon after fertilization the endosperm nucleus divides. This
division is followed by several, others, probably five or six, before
any walls are laid down. Its starch content is drawn upon during

these divisions, so that at the time the cell walls are laid down, `

only a very little may be found or none at all. The proembryo
remains unicellular till about the third or fourth division of the
endosperm nuclei occurs. The growth of the endosperm is ac-
companied by destruction of the integument. The peripheral endo-
sperm cells become less vacuolated, the greatest density of con-
tents being found in those cells of the endosperm which lie towards
the raphe, 7. ^, the peripheral cells are most active where the
largest amount of food is available. The accumulation of starch
is gradual at first and is localized in the chalazal end of the endo-
sperm. Later all the cells become gorged with starch, so that at
the time the cotyledons appear, all the cells are well supplied with

-

CALLIPELTIS CUCULLARIA 31

the exception of those which occupy the middle portion of the
endosperm and the layer of absorbing cells next the raphe. The
endosperm continues however to grow, and ceases only when there
is left of the integument only a single layer of cells separating the
endosperm from the pericarp. The limit of growth in extent being
reached, the cells commence to secrete reserve cellulose. Their
lumina, by reason of the accumulating cellulose, become more
contracted and rounded in form. The increase of cellulose is ac-
companied by an evident reduction of the starch content, though
a plentiful, supply is present at the maturity of the seed.

The immediate supply of food necessary for the endosperm is to
be found in the integument, of which the starch content is gradually
increased as it reaches its maximum size. When the seed is ripe,
a single layer of integumental cells only is left, excepting an
island of tissue underlying the termination of the vascular tissue of
the funicle. The endosperm is thus a concavo-convex mass, sur-
rounded by an integumental membrane, this, in turn, being closely
invested by the pericarp four cells in thickness (øl. 6, fig. 72).
A considerable amount of starch is found in all the cells of the
integument, particularly in the chalazal island.

DEVELOPMENT OF THE EMBRYO

As already pointed out, the embryo remains unicellular till the
endosperm nuclei have reached the number of èight to sixteen,
when it begins to develop rapidly, one transverse division follow-
ing another until the embryo is composed of a cell row of ten or
a dozen cells (p/. 6, fig. 6). The largest of these are in the sus-
pensor. The terminal cell, the fundament of the embryo proper,
is the smallest. The suspensor cells bulge out more or less. A
portion of the suspensor may be composed of two rows of cells
)۸ 6, fig. 9), though, whether this is due to 6 division
or to displacement by longitudinal pressure, I cannot say. Prob-
ably, however, the latter. The arrangement and number of cells
in the suspensor is by no means constant, so that it is a matter of
physiological significance alone. By further growth, the bulging
suspensor cells extend themselves, so as to lie between the endo-
sperm cells, but do not cause any destruction. They are filled
with cytoplasm, and have large active nuclei, The development

7

92 COMPARATIVE EMBRYOLOGY OF THE RUBIACEAE

of haustorial cells by the embryo in Callipeltis is not so great as
in Vaillantia, either in point of number or size. With the de-
velopment of the embryo proper and of the endosperm, the basal
cell elongates, allowing the embryo to take a more central posi-
tion in the endosperm. The haustorial portion is then near to the
embryo. The basal cell is finally crushed between the cells of
the endosperm, the haustorial portion becomes absorbed, so that
only a single cell is left as a portion of the root cap of the
embryo.

The development of the embryo is briefly as follows: After
transverse division has ceased, the terminal-cell divides longitudi-
nally once, then a second time. The second cell from the end fol-
lows suit, then the third and finally the fourth. Meanwhile the first
three tiers of cells are giving rise to dermatogen, then to plerome
in the fashion described by many authors for numerous dicotyle-
dons. The fourth tier of cells which apparently never exceed
four in number, remain so and may be recognized in the definitive
embryo as a small protuberance at the top of the root cap.

A great deal of variation takes place, but the plan just set
forth is followed in general and the ultimate product is the
same,

Sherardia arvensis
This plant, the only species of the genus, is a small upright
annual having the whorled foliage of the Galieae with small heads of
several flowers with rose colored tubular corollas. It is distrib-

uted in Western Asia, Europe and North Africa. The material |

for study was obtained in cultivation in Germany and in New
York.

The history and character of the nucelli, integument and mega-
spore mother-cells call for no special treatment. No evidence
has been obtained pointing to differences in the division of the
megaspore mother-cells, or in the selection of a given megaspore
to become the embryo-sac. It thus appears that the megaspores
are of potentially equivalent value. During migration down the

micropylar canal the three serial divisions of ‘the embryo-sac .

nuclei take place and finally a definitive embryo-sac is formed
which differs somewhat from those hitherto described in form and

i
XE

DUE AA

Poe, OSEE: NA ea

See‏ اه

A EA itu SY AA ee و‎

SHERARDIA ARVENSIS 33

proportion (pl. 6, fig. 15). The two small antipodals are con-
cavo-convex, and are set upon the endosperm cell as a cap. They
are densely filled with food materials. The third antipodal is
longer relatively than in either Vaillantia or Callipeltis, and is
scantily supplied with cytoplasm. The distal end is formed into
a club-shaped swelling, and is buried in the mass of disintegrated
megaspores. The swollen end is filled with a tenuous and finally
granular cytoplasm.

The endosperm nucleus lies in the definitive embryo-sac
against the egg, into which position it has moved from a distant .
position where fusion of the polar nuclei takes place (7. 6, figs.
IS; 10, t7).

The development of the embryo proceeds by the formation of
transverse walls, until about a dozen cells in linear arrangement
have been laid down. All but the terminal cell form then the
suspensor in which two distinct regions are to be recognized. The
micropylar region is composed of larger vacuolated cells, which
grow out laterally into the endosperm forcing their way between
the cells. The nucleus lies in the concavity of the outgrowing
haustoria just as is known to occur in young root hairs )۸/ 6,
fig. r9). The proximal portion of the suspensor is composed of
very short disc-shaped cells, which, while the embryo is young,
retain that form. As the embryo becomes older, the more distal
of these elongate, and some grow out into haustoria. "When the
endosperm has reached its maximum size the laying down of re-
serve cellulose begins in the peripheral cells first, and these cells
become unfavorable for the absorption of food by the suspensorial
haustoria. That portion of the suspensor, therefore, which comes
to lie in the region of cellulose formation degenerates par: ۷
with the thickening of the cellulose layer. The function of ab-
sorption thus given up by the more distal part of the suspensor is
then taken up by the more newly formed haustoria. The suspen-
sor becomes loaded with food which is gradually passed on to the
embryo (X. 6, fig. 21).

In the development of the embryo proper there is correspon-
dence with the process as described for Callipeltis. One feature
appears somewhat different, namely, that the longitudinal division
of the terminal cell is followed by transverse division of its quad-

34 COMPARATIVE EMBRYOLOGY OF THE RUBIACEAE

rants, so that it appears as if the terminal five cells had gone to
form the embryo proper instead of four.

The mature ovule consists of a mass of endosperm with cellu- |
lose as the chief reserve food, and one layer of integumental cells
still intact. The whole is closely invested by the pericarp.

The endosperm absorbs the integument till but one or two
cell layers are left, these forming a membranous lining of the
pericarp.

Galium
p (PLATE 7)

The species of Galium which have been studied are the follow-
ing: G. Aparine, recurvum, pilosum, Mollugo, verum, triflorum,
tinctorum, and Parisiense. As the species of this genus are very
similar with regard to the characters which we are considering, it
is not deemed necessary to treat them separately.

The origin of the nucelli is in all respects as in the forms already
described, and the same may be said of the number of cells in the
` archesporium (figs. 1, 2 and 3), their division into megaspores by
two successive divisions (fig. 4), and the subsequent enlargement of
one of these to act as the embryo-sac cell. In occasional instances
two such cells derived from different megaspores may enlarge and
when germination takes place, two more or less completely formed
_ embryo-sacs result ( fig. 1I) In one case three such were found,

although here the third embryo-sac was very abnormal in appear-
ance. The nucellar epidermal capping cells remain intact until
the megaspores are fully formed, but suffer histolysis at about the
time the migration of the embryo-sac mother-cell commences (fig.
4). The histolysis of the endodermis then follows, as the en-
larged megaspore moves forward along the micropylar canal.

The condition of the definitive embryo-sac is reached in the
usual manner. A two-celled condition is shown in jig. 6. The
chief interest in this connection attaches to the form the embryo-sac
takes in the various species of the genus under consideration. The
various degrees of curvature seen in different species is, no doubt,
due to the form of the ovule. The more strictly anatropous the
ovule, the straighter the embryo-sac. In Galium tinctorum in
which the ovule approaches the campylotrous condition, the em-
bryo-sac is bent into an arc of ninety degrees ( fig. ro). In other

GALIUM |. 85

respects save in three particulars the characters of the embryo-sac
are uniform namely, in the form of the endosperm cell, the shape
and disposition of the antipodals and the amount of the food con-
tent. In certain species, e. g., G. Mollugo, and G. Parisiense, the
endosperm cellis almost spherical while in others it is oval. In
G. tinctorum this cell is markedly constricted about midway its
length and this appears to be the normal condition (ig. 70). In
this plant also, the food content of the endosperm cell is evidently
greater, and is composed for the most part of starch. The starch
grains are indeed so crowded that the nucleus is very much dis-
torted by compression ( fig. 70). Finally the antipodal cells may
take two different arrangements. They may be placed tandem
fashion as in G. triflorum ( fig. 7) and G. Mollugo ( fig. 11), or the
two small antipodals may lie in the same transverse plane as occurs
in G. Aparine ( fig. 8) and G. Parisiense ( fig. 12), and some others.
A modification of this latter mode of arrangement is to be found
in G. Zinctorum ( fig. 10), in which plant these cells become rounded
and are placed obliquely with reference to the longitudinal axis of
the embryo-sac. These cells are at the same time densely filled
with starch, a good deal of which material is also to be found in
the long antipodal cell.

The haustorial antipodal cell reaches, in this genus, its great-
est relative length, as is illustrated by our jig. 9 (G. triflorum),
which shows an ovule in longitudinal section, with an embryo-sac
just before fertilization.

The development of the embryo and endosperm in their earlier
stages presents no differences of special note. Considerable varia-
tion is shown in the development of the suspensor in the different
species. In G. Parisiense, for example, there is but a small haus-
torial apparatus formed, while the suspensor of this plant is re-
markable for the unusual elongation of the basal cell (fig. 13). Of
the species studied, the greatest development of the suspensor is
reached in G. Mollugo (fig. 15). ۱

The endosperm arises by division of the endosperm nucleus in
all directions, at first without the formation of cell walls. Soon,
however, these commence to form before, indeed, the suspensor
produces haustoria. The histolytic changes in the integument,
which take place relatively slowly at first, proceed a little later

36 COMPARATIVE EMBRYOLOGY OF THE RUBIACEAE

very rapidly, and, as indicated in fig. 74, the histolysis is not con-
fined to the immediate vicinity of the endosperm. In the same
figure we may note the antipodal cells to be still present, but they
are in a dead condition, and it would not appear that they are in
any way active agents in the action described, but rather that the
enzyme which we may believe is secreted by the endosperm more
readily passes along the course of the antipodal apparatus and is
thus distributed to a distance from the endosperm.

That the behavior of the integument as the endosperm grows
is indeed the result of the action of an enzyme, we have evidence
in the case shown in fg. 15, where the cell-walls of the cells
lining the micropyle remain unaffected, although the surrounding
tissue has been totally destroyed.

It is to be noted also that the endosperm nucleus takes its
position against the egg cell, although this does not appear to be:
the case in G. finctorum. This may be due to the interference of
the large and numerous starch grains which crowd the cell and
would very naturally interfere with the movements of the endo-
sperm nucleus (fig. 70).

The food mass derived from the disintegrating megaspores
becomes exhausted at the maturity of the embryo-sac or soon after
and no trace of this may be seen soon after fertilization. At this
time, too, the fertilized egg loses its vacuole by the increase of
the cytoplasm, and the synergids at the same time disintegrate.

The course of development of the embryo is, in the early stages,
like that in the others previously described.

Asperula
(PLATES 8 AND 9)

Of this genus the following five species have been studied: A.
azurea, galindes, montana, setosa, and tinctoria.

The “development of the ovule and the appearances presented
by the archesporium in the genus Asperula do not differ in general
from the other Ga/ieae. They present, nevertheless, some striking
points of contrast, especially when compared with Crucianella.

The early stages of development of the ovule may be passed
over without further remark. When the definitive condition of
the archesporium has been attained, the further development of

ASPERULA 37

the ovule, especially of the integument, is such that the upper part
of the sporogenous tissue becomes compressed so that the indivi-
dual cells become displaced and distorted to such a degree that
the whole appears almost abnormal. This appearance is accen-
tuated during and after the division of the mother-cells, which, it
would seem, is by no means as regular as in Crucianclla. This
irregularity is most striking in Asperula montana, in the ovule of
which plant the compression appears to be greatest on the micro-
pylar part of the archesporium. This condition is obviously the
cause also of certain curious abnormalities in the behavior of the
embryo-sac to be described shortly. The irregularities in the be-
havior of the archesporial tissue consist in the differences in rapid-
ity with which tetrad formation takes place, in the partial or total
suppression of tetrad formation in a greater proportion of the mega-
spores, in the early destruction, by pressure, of some of the cells,
and finally are such that the accurate observation of the behavior
of individual cells is rendered almost impossible.

That tetrad formation takes place normally, however, is quite
sure, from the fact that, in spite of the confusion, rows of four cells
each are repeatedly found. The megaspores, sister to the func-
tional embryo-sac cell, persist in apparently normal condition for a
period at least equal to that occupied by the development of the
embryo-sac, and occupy, forasmuch as they do not secrete cell
walls, the same cavity as the embryo-sac. It sometimes happens,
too, that they have quite the appearance of the antipodal cells, so
that it becomes difficult to distinguish between them.

After the functional embryo-sac cell has started to migrate
(pl. 8, fig. 2), some of the other megaspores persist, while others
degenerate. Some, further, secrete cell walls, and undergo changes
which indicate that they are passing through early stages of em-
bryo-sac development. One such cell is shown in the upper right-
hand part of the figure.

It seems equally certain that the tendency is for the upper of
the tetrads to become the embryo-sac. The pressure already
referred to, however, modifies the ordinary mode of procedure, so
that not only may some other of the megaspores rather than the
upper become the embryo-sac, but it may also be forced to develop
in quite the wrong direction. This is the case in Asperula mon-

38 COMPARATIVE EMBRYOLOGY OF THE RUBIACEAE

tana, in which a completely developed embryo-sac is frequently to
be found in the chalazal part of the ovule, or even in the funicle.
Text figs. 1 and 2 illustrate two such instances. In fg. z two

Fic. t, Ovule with a second embryo-sac FiG, 2. Ovule with three embryo-sacs,
lying in the funicle. two in the funicle and one in the usual
Pm The dotted line indicates the fre-

0-Sac
Sa!

Y P g positi 1

embryo-sacs are seen; one in the normal position, 7. ¢., in the
micropylar canal, and one lying parallel to it, but in the funicle.
Both are completely developed. In text fig. 2 three embryo-
sacs may be seen ; the one normal in position, and the two others
lying against eh other, in the funicle, and considerably dis-
torted. The egg poles of these two embryo-sacs are placed
against the pericarp ; that is to say, the embryo-sacs have grown
to the surface of the ovule. In text fig. 3 the egg apparatus
and polar nuclei of one of these abnormally placed embryo-sacs
may be seen to be quite normal except for a small amount of dis-
tortion. Whether the egg of such abnormally placed embryo-
sacs is ever fertilized is doubtful, and no case has been observed.
This plant has flowers which show a peculiar form of dimorphism
caused by the abortion of the ovules in a large percentage of cases.

ASPERULA 39

Only relatively very few ovaries with normal ovules therefore are
to be found, and from these only few seeds were produced under

the cultural conditions at
Bonn, where the material
was collected. In view of
these facts, therefore, the
failure to observe fertiliza-
tion in the abnormally
placed embryo-sacs is not
surprising. On the other
hand, there are grounds for
expectation that the beha-
vior of the pollen tubes is

Fic, 3,
such as to favor such an (gi. ui 1

occurrence.

Part of embryo-sac occurring in the
wing egg apparatus,

Further peculiarities occur in the behavior of the normally
placed embryo-sac in Asperula montana, in that its growth con-
tinues often until the egg end projects beyond the mouth of the

G.
nucleus.'

Protruding embryo-sac with ¡two nuclei in mitosis, Ani., Antipodal

Fic. 5. Protruding embryo-sac with at least seven antipodal nuclei. Polar nuclei

not yet fused.

40 COMPARATIVE EMBRYOLOGY OF THE RUBIACEAE

micropyle and comes to lie between the integument and the peri-
carp. In text fig. 2 the dotted line indicates in a schematic way
the position taken by the embryo-sac under these circumstances,
and text figs. 4 and 5 show two such embryo-sacs in detail. It
is to be noted that the egg apparatus under these circumstances is
always directed away from the funicle. In cases in which fertiliza-
tion takes place, the egg elongates before
division, and the embryo thus comes to lie
within the ovule (text fig. 6). Fig. 7 serves
as an example, although here the egg ap-
paratus is not as far away from the micro-
pyle as in many other cases. The habit
of the embryo of producing a long slender
stalk before division is constant, and may
be seen even where the embryo-sac does
not protrude from the micropyle, where,
namely, the embryo-sac takes the same

Fic. 6. One-celled embryo position as in the other Galicae (text fig. 7).
vo ina protruding em- What the behavior of the pollen tube in

FIG. 7. One-celled دی‎ this plant may be cannot be said, as no
in embryo:sac occurring in the favorable material has so far been found.
sore UAE posto A study of this point cannot fail to be of

interest.

The migration of supernumerary embryo-sacs into the funicle as
above described is of special interest as it bears on the question
of the guidance of the embryo-sac in its movements.

Under ordinary circumstances the embryo-sac cell migrates
down the micropylar canal, breaking through the thin cap o
nucellar tissue. Where, however, the pressure of the integument
is great enough the normal movements of the embryo-sac nucleus
are prohibited, and the embryo-sac makes an irregular growth in
any direction, and reaches no definite point. We may, there ore,
conclude that the guidance of the embryo-sac into the micropylar
canal is mechanical.

The other species of Asperula studied show in general the
same disturbances in the archesporium due to pressure as in
Aspe rula montana, though not to the same degree. This is indi-
cated by the fact that no exception to the normal position of the

ASPERULA 41

embryo-sac has been seen, although similar appearances in the
archesporial tissue are there.

The only other point in connection with the embryo-sac calling
for special mention is the behavior of the antipodal cells. These
are by no means constant in number, and may vary even in the
same species, as far as observed, from three to seven. The most
typical condition is shown in pl. 8, fig. 3, for Asperula azurea,
and in pl. 9, figs. 1 and 2, for. Asperula montana. The long anti-
podal has the usual appearance corresponding as it does to the
homologous cell in the other Galicae. The other two, however,
are here very much enlarged and each consists of dense cytoplasm
with a large active nucleus, and a very large central vacuole. In
the growth of these cells the adjacent tissues present histolytic ap-
pearances similar to those seen in the same tissue when the endo- -
sperm enlarges. There is, therefore, so much evidence that the
antipodal cells are able to secrete substance which acts in the way
indicated upon the living cells.

When more than three antipodal cells occur they are arranged
in pairs forming an irregular double row, recalling the condition in
Aster (Chamberlain) and in Diodia (pl. 9, fig. 5). In some
species the behavior of these cells is very irregular ; one frequently
finds several nuclei in the same cell, and the cells are sometimes
much more slender in proportion than as represented in the figure
(pl. 9, fig. 5). In all cases, however, the same observations
already made with reference to the persistance of the antipodals
and their nutritive relations holds good in Asperula.

THE EMBRYO

Of all the Galicae studied, the genus Asperula is most remark-
able for the extreme development of the suspensor. In 4. setosa
(pl. 8, fig. 4) the haustorial cells attain a large size, and extend
to a considerable distance from the suspensor into the endosperm.
The amount of cytoplasm is not great, and is collected mainly at
the distal ends of the haustoria, where the nucleus is to be found.
At this point the growth of the organ takes place, so that we have
here a very close analogy with the root hair in which, as Haber-
landt * has shown, the nucleus and cytoplasm behave similarly.

* Ueber die Beziehung zwischen Function und Lage des Zellkernes bei den

Pflanzen, Jena, 1887.

42 > COMPARATIVE EMBRYOLOGY OF THE RUBIACEAE

In this form, on account of the large size attained by the haustoria
it is quite easy to see that, as they penetrate the endosperm, the
process is accompanied by mechanical readjustment of the cells of
the latter, without any destruction. 2%. 8, fig. 5, shows a later
stage in the same or a closely related species, from which it will
be seen that as the embryo grows older the haustoria becomes
filled with cytoplasm. Meanwhile their proximal extremities be-
come compressed and drawn out as a result of the pressure exerted
on them by the growing endosperm. The suspensor, exclusive of
the haustoria, then presents a fibrous appearance, with little cyto-
plasm, and is little more than an irregular tube joining the bases .
of the haustoria and the embryo. The haustoria in some instances
are very narrow, and extend to a considerable distance, passing
beyond the embryo proper. Such narrow ones remind one of
fungal hyphae, which pursue an intercellular growth, or of the
pollen tube with its analogous behavior.

In the other species, for example A. azurea and A. galioides,
the suspensor develops haustoria which, instead of penetrating to
a distance into the endosperm, confine themselves to rather narrow
limits, while the immediately surrounding endosperm suffers his-
tolysis. All the preparations of plants in which this behavior
takes place show a perfectly normal structure in the haustoria, in
which the cytoplasm is very dense and finely granular as in A.
galioides, while in the material in which they are embedded no
trace of structure is visible, nothing, in short, but the mucilaginous
irregularly staining products of histolysis.

As the embryo becomes older the haustoria become separated
from each other but still keep a normal appearance. The study
of a complete series, of which p/. 8, fig. 6, represents a component
of three sections, shows no union between several of the haustoria, .
although this fact cannot well be shown in a drawing.

The behavior of the suspensor in the various species of Asper-
ula is therefore so different, as well as the relation between it and
the endosperm, that we are prevented from interpreting the facts
in the same way. In some cases, as described, the relation in- ，
volves histolysis of the endosperm, which indicates that, in such
cases, the suspensor secretes an enzyme.

The earlier cell divisions of the embryo proper are as described

RUBIA TINCTORIA 43

in Callipeltis. It is to be observed, however, that the longitudinal
divisions which are generally confined to the last four cells of the
young embryo here take place also in a half dozen or more of the
cells of the suspensor nearest the embryo (pl. 8, figs. 4 and 5).

Rubia tinctoria
(PLATE 9, FIGURES 1-6)

The nucelli of Rubia are recognizable by their form which in
the young condition, before the integument has commenced to
grow, is characterized by a sharp bend on the inner side of the
funicle. The megaspore mother-cells are fewer in number and
are of different proportions, being shorter and thicker than the
same structures in the rest of the Ga/zeae studied. Their nuclei
are relatively larger (fig. 7). The tapetal cells fail here as
elsewhere, and the division of the spore mother-cells is reg-
ular (fig. 2) giving rise to four equivalent spores. It is a com-
mon occurrence for several of the megaspores to divide with the
result that a confusion is caused. The migration of the proper
embryo-sac cell along the micropylar canal is accompanied by the
usual divisions and the ultimate formation of an embryo-sac ( fig.
5). The haustorial antipodal cell is broader and shorter than
those of the foregoing forms. The other antipodal cells may be
very small, and placed laterally or in some cases are quite large.
In the former case they are filled with cytoplasm, in the latter case
they have a cytoplasmic lining and the interior is filled by a large
vacuole, :

It is interesting to note that the embryo-sac nuclei do not cause
the disintegration of the endodermis so rapidly as in other Gazae,
and appear to force their way along the micropylar canal, without
the immediate collapse of the impinging cells (jig. 3).

In many respects the features of the archesporium and of the
embryo-sac present the same peculiarities as have already been
described in Asperula. There is the same tendency towards the
formation of supernumerary embryo-sacs, and the same characters
are shown by the persisting megaspores (fig. 4), the nuclei of
which are supplied with a scanty amount of cytoplasm, and are
found lying freely in the irregular cavities left by the megaspore
mother-cells.

44 COMPARATIVE EMBRYOLOGY OF THE RUBIACEAE

After fertilization the endosperm grows very rapidly, and gets
a large bulk. Although the growth of the embryo is immediate,
and as rapid at least as in the other plants so far described, it is
proportionately slower, so that in a fruit so large that one would
naturally expect to find a large embryo, one finds, on the con-
trary, a young embryo without cotyledons. The endosperm has
large cells and contains a very small proportion of plasma. For
this reason the endosperm plasmolyzes very easily.

The suspensor of this form is very fully developed (fig. 6).
The cell divisions are, at first, solely transverse. Soon, however,
they become irregular and often strictly longitudinal, while all of
the suspensor cells excepting a few disc-shaped cells near the
embryo proper, and still fewer at the micropylar end of the sus-
pensor develop into haustoria. The relatively greatest develop-
ment is reached by the suspensor when the embryo proper has
about eight to sixteen cells. The pressure exerted by the endo-
sperm upon the suspensor, together with the direction of growth
of the same causes the suspensor to be drawn out, so that the
embryo proper appears to be anchored to the endosperm, which
in this way is enabled to exert a tension upon the suspensor.

Crucianella
(PLATE 10, FIGURES 7-22)

Of the genus Crucianella, three species, namely, ۰ gilanica, E
macrostachya, and C. herbacea have been studied. These present
certain anomalies which are of not a little interest in view of the
close relationship between Crucianella and the rest of the Galicae.

The archesporium consists of 12 to 15 very large embryo-sac-
mother-cells, which, however, vary between themselves in size.
The largest occupy the more central position, and these are also
distinguished by the completeness of their nuclear divisions ; the
other cells divide less regularly or not at all, and the more quickly
show signs of degeneration. The cell walls between the mother-
cells are very delicate, but are present and easily distinguishable.

The division of the megaspore mother-cells each into four

£ PREND
BOREAS i = 3
IPG AV A ne O A

grand-daughter cells is remarkably regular, and results in the E

formation of four megaspores which are alike in size, but not

separated by cell walls. This feature, to which only very occa- ，

CRUCIANELLA 45

sional exceptions may be found, may be regarded from the
phylogenetic point of view as a step further removed from the
condition in Angiosperms in general in which the tendency to form
cell walls may be regarded as an inheritance from forms in which
the megaspores are in maturity free bodies.

Similar failure on the part of the megaspores to form cell walls
has been reported to occur in Eichhor nia * and in Avena and is
probably of much wider occurrence than at present supposed.

Certain authors—Juel, Murbeck and Koernicke especially —
have recently concentrated attention anew to the constancy and
significance of these divisions—the tetrad division. In addition to
the evidence adduced from the character of the nuclear mitoses,
to which reference is made in another part of this paper, the facts
which speak for the morphological equivalency of the cells which
justifies the use of the term megaspores as applied to them have
been set forth at length by Koernicke.t

Remarkable as is the regularity with which the megaspore
mother-cell divisions take place, still more so is the further di-
vision, in very many instances, of a// the megaspores simultaneously.
If we regard these as the first embryo-sac divisions, which we are
justified in doing from a comparative standpoint, we are forced to
regard the eight resultant nuclei derived from a single megaspore
mother-cell as representing four embryo-sacs lying tandem. It
has been determined by actual count that some ovules may thus
be supplied with a dozen young embryo-sacs. Of these two or
three of those derived from the megaspores lying adjacent to the
micropylar canal may commence to develop into functional em-
bryo-sacs. Only one, however, normally attains full development.
The very evident physiological as well as morphological equiva-
lency of the megaspores in this plant irresistibly compels the com-
parison with the conditions seen in certain heterosporous vascular
cryptogams, especially with Se/aginel/a, in which the majority of
the megaspores serve as nutriment for the functional ones.

* Smith, W. R. A Contribution to the Life History of the Pontederiaceae. Bot.
98

Gaz. 15: 1808.
T Cannon, W. A. HEP ageing Study of the Flower ge rq of the Wild
cad

Oat, Avena fatua L. Proc, Cal . Botany, 111. 1: 32
$ Studien an Embryosack- ی ی تون و‎ d Natur. und Heilk.

Sens 1901.

46 COMPARATIVE EMBRYOLOGY OF THE RUBIACEAE

The occurrence of such conditions in widely separated families
is, however, a fact derived from comparative study, which warns
us from inferring phylogenetic continuity. We are dealing here
with recurrent morphological structures, called out in response to
physiological necessities.

Of the four megaspores formed, although as just pointed out
their activity is indicated by their tendency to further division, only
the upper one, namely, the one furthest removed from the chalaza,
develops into a functional embryo-sac. No exception to this rule
has been observed. Instead, also, of enlarging directly at the ex-
pense of the other three by destroying and absorbing them, it
moves into the micropylar canal, meanwhile growing at the ex-
pense of the adjacent tissue of the ovule (jigs. 78, 79, 27, 22). In
so doing the nucellar capping cells are destroyed. The fate of the
remaining megaspores of the tetrad is indicated in fig. 19, 1
which they may be seen to have undergone partial voii e.
and to have been drawn forward toward the embryo-sac as if by
suction. This appearance is quite constant, and is difficult to in-
terpret in any other way. That the materials derived from them
by their gradual disintegration and that of the other megaspores
are used by the embryo-sac is shown by their gradual disappear-
ance inversely to the growth of the embryo-sac.

For some time during the development of the embryo-sac,
however, the tetrad masses maintain their normal appearance, and
indeed elongate considerably, but generally show signs of degen-
eration before it is completely developed.

At the time the embryo-sac cell commences its migration
down the micropylar canal certain changes take place in inner lay-
ers of the adjacent integument These consist in the rapid growth
of the cells to many times their original volume, and the pro-
nounced enlargement of the nuclei (jig. 73). These cells appear
to be very active, and in some cases they fail to secrete cell walls,
so that we find occasional cells provided with several nuclei. This
behavior of cells, namely, their failure to form cell walls immedi-
ately, appears to be connected with a highly nutritive function,
and recalls the conditions in young endosperm, in the suspensor
of some Leguminosae (Guignard, /. c.), and in the megaspore
masses described above. Examination shows that this cell mass

CRUCIANELLA 47

is a specialized portion of the ovule which is especially concerned
in secreting starch, of which a very large amount is present, while
it is at the same time absent from the rest of the ovule.

At the base of the archesporium the cells of the chalazal tissue
appear, in a very early stage, just as the remainder of the cells of
the ovule. As the integument develops, however, and therefore,
as the archesporial cells increase in size, these chalazal cells take
on a different character (c.c., figs. 7, 9 and 13). From being
small cells with a rather scant supply of cytoplasm, and relatively
small nuclei, they become larger, their nuclei grow, and they be-
come densely filled with cytoplasm. This appearance they main-
tain during the periods of growth of the archesporial cells, of their
division into megaspores, and of the subsequent elongation of the
tetrad masses. At the close of the activity of the latter the chal-
azal cells lose their supply of cytoplasm and return to the condi-
tion of the rest of the cells of the ovule.

We may therefore interpret them as physiologically active in
the nutrition of the archesporium. Similar specialized chalazal
cells have been described in many forms, more recently, however,
in Polypompholyx * and in. Stylidium.t

The embryo-sac cell moves rapidly along the micropylar canal,
and as it moves forward the integumental tissue closes behind it.
The course taken may readily be traced by the more deeply stain-
ing cell walls of those cells lying next to the canal. The division
of the embryo-sac cell follows the usual order, as heretofore de-
scribed.

In this genus, however, the development of the embryo-sac is
marked by the very sudden disappearance of the antipodal cells.
Of all the preparations made, in not one have all three of the anti-
podals been seen, although they are developed, as indicated in
fig. 21, ant. The polar nuclei have been seen, and several
cases have been found where one antipodal nucleus appeared
normal. All the evidence indicates, therefore, that the sudden
death of these nuclei is a character of the genus. It must however
be noted that the antipodal end of the embryo-sac extends some

* Lan . K. Untersuchungen über gp" Anatomie, und Samenent-
wicklung v von een und Biblis gigant Flora, 88: 149-206. 1901.
f s, G. P. Beiträge zur Kenntniss de eg Flora, 87: 313-354.
1900. A

48 COMPARATIVE EMBRYOLOGY OF THE RUBIACEAE

distance backwards toward the archesporial tissue, and is in con-
nection with the same by means of the micropylar canal.

The embryo and endosperm present no points of striking differ-
ence from the same structures already described in other Galzeae.

Diodia Virginiana
(PLATE 12, FIGURES 1-12)
I am indebted to Professor Frank S. Earle for collecting and
preserving material of this and the following species described.

NucELLUS. STROPHIOLE AND INTEGUMENT

From the account already given of the origin in the GaZzeae of
the nuclei and basal partition, we have seen that the former rise
independently from the floor of the ovary, and that the latter even
at maturity of the ovules reaches over to one-fourth the distance
from the floor to the roof of the ovarial chamber.

In the forms now to be considered, however, a somewhat dif-
ferent condition is to be found. Here the basal partition originates
at the same time with the nucelli, so that in an appropriate longi-
tudinal section one sees a plate of tissue rising from the floor
of the ovary, from the sides of which the nucelli, seen as hemi-
spherical masses, take their origin (fig. 7). With the further de-
velopment of the ovary, the basal partition and that formed by the
upper parts of the carpels, grow with equal rapidity, so that at the
time the embryo-sac is mature the floor and roof moities of the
ovarial partition fuse at a point half way from the floor to the
roof of the ovary. At this point, therefore, are inserted the funicles
of the ovules, whose growth is such as to give rise to the amphi-
tropous form (figs. 2, 8, and 9).

Early in the development of the ovule at the time when, in as-
suming the amphitropous position, its axis is passing through a
transverse plane, a growth which superficially suggests an integu-
ment makes its appearance first,on the morphologically lower sides
of the funicle. Coincident with the origin of this organ which for
our present purposes we call a strophiole, the single integument
commences to develop as a ring-shaped fundament about the epi-
dermal capping cells of the archesporium. The growth of the in-
tegument is such as to form a long micropylar canal, the mouth

DIODIA VIRGINIANA 49

of which is ultimately covered by the strophiole. Certain features
of special interest attach to this organ which will now be described.

In the first place, the supply of vascular tissue to the ovule is
confined to the strophiole. A single strand of the same passes
through the funicle and upon reaching the strophiole divides,
sending one branch, the chief one, to the chalazal region of the
strophiole, the other toward the micropylar region (a and a’, figs.
2, 8,9). The chalazal branch (a) presents at its extremity a sharp
turn towards the spot at which lies the archesporium.

Secondly the limit of the strophiolar tissue is sharply defined
by the presence of a plate of brown-walled cells (sf. in the figures),
having the same cytological characters as the epidermal cells of
the integuments, which also become brown. In the young ovule
the strophiolar plate is not to be seen, and has its origin at about
the time the embryo-sac reaches its definitive condition, in scat-
tered cells which marked the position of the plate, and which in-
crease in number until the sheet of tissue is complete. When
this takes place there are no fenestrations in the plate excepting
a sometimes smaller one at the extremity of the chalazal branch
of the vascular supply. It ultimately becomes continuous with
the epidermis (epi, fig. 9) and thus the ovule exclusive of the stroph-
iole comes to be completely surrounded by a sheath of cells with
scanty plasmatic content, and deep brown walls.

In the third place the strophiole is characterized by the occur-
rence of a large number of excretory cells which become filled
with raphides. When the seed is mature the strophiole carries a
heavy load of calcium oxalate. As many as two hundred of the
raphide cells have been counted. The remainder of the tissue is
composed of cells which do not differ materially from the paren-
chyma of the integument. From the foregoing facts, it must be
evident that the strophiole is the seat of special metabolic activity,
during the growth of the embryo and endosperm, and that the
strophiolar plate is of no special hindrance to the passage of proper
food materials used by the embryo and endosperm.

ARCHESPORIUM AND EMBRYO-SAC
The archesporium is laid down in certain hypodermal cells of
the nucellus of any early age. No tapetal cells are cut off, nor is

Mo. Bot. Garzea

E KS NANA er NA REN NO

50 COMPARATIVE EMBRYOLOGY OF THE RUBIACEAE

there any further growth of nucellar tissue above the archesporium.
The conditions then, are precisely as in the Galzeae in which the
archesporium is surmounted only by a group of epidermal capping
cells. In the form, here being described, the archesporium, how-
ever, contains only one functional megaspore mother-cell ( £g. 3)
surrounded by a number—about a half dozen—of cells of similar
character, but reduced in size, and as their later history shows,
possessing usually no ability to divide or to act as embryo-sac
fundaments. No instance has been found in which two embryo-
sacs have been present in a single ovule, a witness to the reduced
ability of these cells as compared with their homologues in the
Gaheae.

The functional megaspore, however, grows apace and is easily
to be recognized by its size and appearance. Its cytoplasm,
especially at its outer end, is of a loose reticular structure
which, in this form in particular, is very marked. Whether by
division this cell divides to form megaspores or not, I can not at
present say. In the figure shown (fig. 3) the synapsis stage
there represented indicates that such divisions do occur since the
first division of the embryo-sac nucleus takes place, in a closely
related plant, at a later stage. To say that a direct development
of the megaspore mother-cell into the embryo-sac takes place would,
however, in the absence of positive evidence, be gratuitous. We
therefore assume that such is not the case until a more favorable
opportunity is presented for more careful investigation of the
point,

The embryo-sac cell, when the integument is well grown, is
ready to commence its development. It is to be noticed that in
the case before us the subsidiary archesporial cells do not suffer
disintegration during the enlargement of the functional embryo-sac
cell. The latter when fully developed pushes forward, causing dis-
integration of the epidermal capping cells and enters the micro-
pylar canal. Down this it moves, dividing meanwhile so as to
give rise to the usual complement of embryo-sac cells. The re-
maining undivided megaspore mother-cells follow the moving
embryo-sac cell, and arrange themselves lengthwise in the arche-
sporial cavity and the adjacent portion of the micropylar canal in
such a manner as to form a continuous strand of elongated cells

[121011۸ VIRGINIANA 51

connecting the embryo-sac with the archesporial cavity ( figs. 4,
5, 12). Near this point, as above pointed out, ends the chalazal
branch of the vascular tissue of the strophiole. It must be further
pointed out that the chalazal tissue cells are elongated between the
archesporium and the vascular tissue, completing the path along
which we may believe passes a supply of food materials from the
parent plant to the developing embryo-sac. The persistence of
the subsidiary archesporial cells with normal appearance until the
endosperm is well developed sustains this view.

The embryo-sac in form is cylindric and presents, so far as the
egg apparatus and endosperm nucleus are concerned, no point of
special note. The fusion of the polar nuclei takes place quite near
to the egg. The synergidae have well developed beaks which
have been described for many forms and which I have referred to
in another part of this paper.

THE ANTIPODAL CELLS

The feature of particular note is, however, to be found in the
antipodal cells, which possess the power to divide, and form a tis-
sue of considerable extent. The number of cells has been counted
in several instances, and varies from four to ten. Frequently the
several component cells are not separated by a cell wall, in which
cases two or more nuclei may be found ina common mass of cyto-
plasm. The last antipodal is tapering in form, and overlaps the
adjacent subsidary archesporial cells ( figs. 6, 72), so that there lies,
between the endosperm and the chalazal end of the vascular tis-
sue, a continuous strand of transporting cells. There can be no
doubt of the activity of the antipodal cells in this regard because
of the following facts: The antipodal cells exhibit a degree of
active growth, they are at all times richly supplied with starch, a
substance which is abundant in the embryo-sac during its whole
time of development, and they do not disintegrate, but, on the con-
trary, maintain their form and normal appearance for a long time
after fertilization, while the endosperm is in its earlier most rapid
growth. In appearance they are apparently similiar to the anti-
podal tissue described by Chamberlain (Z c.) in Aster. The coin-
cident persistence of the archesporial cells is also noteworthy in
this connection.

52 COMPARATIVE EMBRYOLOGY OF THE RUBIACEAE

THE ENDOSPERM AND EMBRYO

The endosperm is at first parietal, and arises by free cell for-
mation. At a somewhat early stage the central cavity becomes
obliterated by the growth of the endosperm cells toward the center.
Walls are then secreted and the tissue then presents in the main
the same appearance as the endosperm in the 6۵/26۵6, The pe-
ripheral cells, which are more densely filled with cytoplasm than
the rest, appear to play the roles of digestive and absorptive cells
before which the integument breaks down as the endosperm grows.
The integument breaks down with the usual appearances accom-
panying histolysis, although the cells of the micropyle resist the
digestive action rather more than the rest of the tissue. The cells
of the middle part of the endosperm, as the latter reaches a con-
siderable size, present, on the other hand, a quite different char-
acter. They are larger and have contents which stain in such a
way as to suggest that they also are suffering histolysis. A col-
umn of such cells extends from the embryo two-thirds of the
longitudinal distance through the endosperm. As the embryo
grows in length it comes to occupy this space, displacing the cells.

The embryo is of remarkably slow growth in the earlier stages
of its development. The early growth takes place in the usual
way, forming a suspensor of a few cells which are totally devoid
of such outgrowths as have been described for the Galicae. The
spherical form of the embryo proper is reached only slowly, and
meanwhile the endosperm grows so as to reach nearly its final
size. The later development of the embryo during which the coty-
ledons are formed is more rapid.

Diodia teres

(PLATE 12, FIGURES 13-17; PLATE 13, FIGURES 1-10)

THE NUCELLUS AND STROPHIOLE

With a trifling difference in the configuration of the nucelli, due -
apparently to the somewhat different mechanical relations in the
ovary of this species, their behavior and that of the basal partition
is the same as that in Diodia Virginiana (pl. 12, fig. 13). The in-
sertion of the funicle, however, is not so far removed from the base
of the ovary on account of the somewhat less rapid development

DIODIA TERES 53

of the basal partition. Quite soon the strophiole appears on the
concave side of.the funicle )۸ 72, fig. 74), and the ridge grad-
ually extends to the opposite side. With further development the
micropyle becomes entirely closed by the strophiole, which agrees
with its homologue in Diodia Virginiana in being possessed of
special excretory cells containing raphides in large numbers, and in
being the organ in which the vascular tissue of the ovule is dis-
tributed. The cells adjacent to the vascular tissue are small and
are more densely filled with plasma, and increase in size in ap-
proaching the integument. The latter is not delimited from the
strophiole by means of the brown cells which have been seen in
D. Virginiana, and which are absent from the ovule entirely. The
epidermal covering of the funicle and part of the strophiole is
strikingly developed, and has been described more fully in another
part of this paper. It is sufficient at present to point out the fact
that they are deep columnar cells with thickened walls, and form
a collar of tissue about the funicle, but having no relation to the
nutrition of the ovule (pl. 73, figs. I, 7, 8).

THE ARCHESPORIUM AND EMBRYO-SAC

The archesporium is in all respects similar to that of Diodia
Virginiana (pl. 12, fig. 15). Only one functional megaspore
mother-cell is present which divides to form four megaspores (fl.
12, figs. 16,17). One of these, probably the basal, judging from
appearances, such as are seen, for example, in p/. 73, fig. 2, is the
embryo-sac cell. This grows rapidly, becomes loaded with
starch, and commences its migration along the micropylar canal.
As the nucleus passes forward, a vacuole is formed behind it.
This vacuole is.also present in Diodia Virginiana (pl. 12, figs. 2,
4, and 5).

The embryo-sac when mature is comparatively short, and only
three antipodals are present, and we notice that, although one of
these is much larger than the others, it is nevertheless much re-
duced in length as compared with that cell in the forms previously
described. The persistence of these cells for a considerable
period, together with their heavy starch contents indicate that they
have a considerable amount of activity. The functional activity
of the antipodal cells even in this form where, to all appearances,

54 COMPARATIVE EMBRYOLOGY OF THE RUBIACEAE

they are of less importance than in the other plants studied, is in-
dicated by the distribution of starch. In the unopened flower the
embryo-sac is crowded with this food material, which is, however,
absent from the rest of the ovule. As the flower opens, and at
the time of fertilization, the starch disappears from the embryo-sac,
first from the egg pole, so that in an old flower starch is entirely
absent from the embryo-sac, but has commenced to accumulate
at the chalazal end of the ovule. After fertilization, starch reap-
pears in the antipodal cells, and then in the growing endosperm,
so that when the latter has considerably increased (p/. 13, fig. 6),
the former are very rich in starch content. In this connection it
is to be noticed, also, that the subsidiary archesporial cells soon
disappear, and take no further part in the development of the
ovule. They are, in fact, quite lost to view when the embryo-sac
contains but two cells. The endosperm nucleus lies appressed
against the egg cell for some time before division. The endosperm
is of the parietal type )۸ 13, fig. 6), and its behavior is similar
to that of the same structure in Diodia Virginiana, as is also true
of the integument. E

The shape of the ovary is such that the chalazal end of the in-
tegument becomes folded over the strophiole, the whole produc-
ing the curious effect shown in p/. 13, fig. 8.

The embryo is slow in developing, and no haustorial out-
growths occur. The cells of the suspensor are at first short, and
disc-shaped and later lengthen, bringing the embryo deeper in
the endosperm )۸ 13, fig. 9). That the suspensor is not an
organ of only mechanical significance may be inferred from the
occurrence in it of a great deal of starch up to the time, at least,
when the cotyledons are well started.

Richardsonia pilosa
(PLATE 13, FIGURES 11-18)

Material in cultivation at the New York Botanical Garden.

In the main, this plant agrees with the two species of Diodia
as to the development of the ovule and archesporium, and of the
embryo. The following important points of difference are to
be noted.

The functional megaspore enlarges and. develops into the

RICHARDSONIA PILOSA 55

embryo-sac zz situ. In lengthening, however, it grows forward,
and destroys the nucellar cap, so that these cannot be seen after
the embryo-sac nucleus divides. No migration, strictly speaking,
takes place, so that the antipodal cells appear in the position pre-
viously occupied by the archesporium. The few reduced sub-
sidiary megaspore mother-cells are soon absorbed and are lost
to view.

By the usual number of divisions the eight embryo-sac nuclei
are formed, and the egg apparatus, polar nuclei and antipodal
cells arrange themselves in an oval cavity. The antipodals are
constantly three in number, and are in this plant all nearly of a
size (figs. 14, 15). For some time they possess a large, deeply stain-
ing nucleus, finely granular dense cytoplasm, and a vacuole at
their free ends, and present the appearance of certain glandular
cells, especially the stigmatic cells in some plants, and of syner-
gidae. This fact is important, as it indicates a considerable degree
of physiological activity on the part of the cells.

After fertilization the embryo develops slowly, forming first by
transverse cell divisions a row of disc-formed cells. The suspensor
remains always a short organ, with no special adaptations. The
endosperm is at first parietal, but early fills the cavity by equal
centripetal growth on all sides. At such time as this occurs, there
appear at the apex of the embryo certain very much enlarged.cells
of endospermic origin, with nuclei considerably greater than those
of the rest of the endosperm cells ( £g. 78). Beyond their size, as
just indicated, these cells do not present any peculiarities, and it is
therefore difficult to any whether they possess any special physio-
logical peculiarities, although this seems probable. They corre-
spond in general to similarly placed cells in other types described,
as having a centripetally developing endosperm, differing from
them chiefly in size.

As the ovule enlarges after fertilization the limiting cells of the
strophiole take on the brown coloration already noted in Diodia
Virginiana. This change spreads to the cells surrounding the end
of the vascular bundle. The peculiar function of these cells, if
such they possess, remains problematical. It is certain, however,
that they exert no unfavorable influence on the passage of food
solutions, whether'they have any positive value or not.

56 COMPARATIVE EMBRYOLOGY OF THE RUBIACEAE

Houstonia
(PLATE 14)

The two species of the genus Houstonia which have been
studied are Houstonia coerulea and H. longifolia, both of which
have been found to be in essential agreement.

The style, which in Houstonia is single, is provided with two
stigmas, and arises by the concrescence of two ridge-like fun-
daments which develop, as in the GaZeae, laterally and oppo-
site to each other, in the inside of the hollow torus (s, figs. 7, 2
and 3). During concrescence a transverse ridge (figs. 1, 2 and 3)
is formed on the inside of the ovary roof, which, by downward
growth, ultimately divides the originally single ovarial chamber
into two locules. This ridge, which may be described as hanging
from the roof of the ovary, unites along its free lower border
with a parallel, low broader ridge developed in the floor of the
ovary. This floor ridge (b, figs. 2 and 3), which must be regarded
as representing two fused carpellary margins, is not evident until
the two rounded fundaments of the placentae (5, fig. r) have
reached considerable size. These two structures have the same
position and appearance as the fundaments of the ovules in the
Gaticae. In the latter forms, however, the two carpellary margins
produce each a single ovule ; in Houstonia a knob-shaped placenta
is formed in each locule occupying the same topographic relations
as the ovules in the Ga/reae.

Each placenta soon begins to show a number of protuberant
growths, which may early be recognized as the primordia of
ovules. The development of these is of peculiar interest.

Up till the time when the archesporium is to be distinguished
the course of development of the ovule is like that in the ۰
A single row of epidermal cells cap the archesporium, but these
cells, which in the GaZeae remain without further division, in
Houstonia suffer periclinal divisions (Ag. 4), as do also the
adjacent epidermal cells. This process is continued until the
archesporium comes to lie in the middle of the nucellar mass. Vo
integument at all is developed ; indeed, the regular periclinal cell
divisions which characterize the early stage in the development of
the integument in the Ga/zeae are absent. The suppression of the
integument accounts for the absence of the micropyle.

HOUSTONIA 67

By growth in all directions the mass of the ovule is increased,
and finally a bowl-shaped form is attained, with the concave sur-
face turned toward the placenta. This change in form is accom-
panied by lengthening of the funicle, which has histologically the
appearance of being a part of the placenta, but which in reality is
derived by the regular division of the cells (figs. 8 and ro, ff),
which constitute a very short stalk in the younger condition,

It is interesting to note in passing that the “zwcellus nudus”
of Schleiden, which that botanist incorrectly thought to be found
in certain Rubiaceae, is realized in Houstonia.

It must be evident that generalizations made heretofore with
reference to the occurrence of ovules without integument in de-
generate plants must fall to the ground.

THE ARCHESPORIUM

The archesporium consists of a single large embryo-sac-mother-
cell, surrounded by a single layer of usually about six small
slender suppressed ones which show no tendency to divide. They,
however, do not all degenerate, but some at least take on special
cytological characters and take up their position in the cavity
vacated by the functional embryo-sac-mother-cell in its develop-
ment ( fig. 9). In this figure one may see two such mother-cells
which have the appearance of glandular cells, resembling in this
regard the antipodal cells. From the same figure it is also evident
that some of the mother-cells persist in this character for a con-
siderable period, until, at least, the completion of the embryo-sac.

When the growth of the nucellar cap has proceeded until the
archesporium comes to lie in the middle of the naked nucellus,
the division of the megaspore mother-cell takes place. Four
megaspores arise by two divisions, of which, contrary to the
general rule among the Rubiaceae studied, the basal megaspore
becomes the embryo-sac cell. In fig. 6, in which the arrow indi-
cates the egg pole, are shown the four cells, of which the basal
is the largest. Its sister cell is small, while the remaining pair of
cells are evidently destined also to be of small dimensions.

The embryo-sac cell develops into the embryo-sac 7m situ.
jig. 7 shows the two-celled, and fig. 8 the four-celled, condition.
The definitive condition is shown in fig. 9. In its growth, the

58 COMPARATIVE EMBRYOLOGY OF THE RUBIACEAE

embryo-sac destroys the surrounding nucellar tissue ( £g. 8), but
more rapidly in the direction of the egg pole, and leaving the tis-
sue about the archesporium intact. The suppressed megaspores,
which are represented on either side the functional one in Sig. 5;
meanwhile enlarge somewhat, and a vacuole appears in the cyto-
plasm. In this condition they may be seen, when the embryo-sac
is mature, occupying the space previously mapped out as the
archesporium, behind the antipodal cells, to which they are very
similar in cytological characters, differing only in size. The anti-
podals are all of a size, densely filled with finely granular cyto-
plasm, and having a large vacuole in the end adjacent to the en-
dosperm cells. Soon after fertilization both the megaspores, which
persist up till this time, and the antipodal cells degenerate, and are
seen as irregular, deeply staining masses.

Before the endosperm nucleus commences to divide it is seen
lying against the young embryo. The growth of the endosperm
is at first parietal, remaining so until two or three irregular layers
of cells have been formed. As the result of the rapid increase in
amount of vacuolation, however, the endosperm cells increase in
size so as to obliterate the enclosed cavity, thus passing from the
parietal condition to that found in the Galieae, but by a different
mode of development (figs. ro, II, and 12).

The development of the embryo is exceedingly slow. From
the condition, in which it has a rather characteristic cylindrical
form, it becomes multicellular by transverse divisions which occur
slowly. No haustoria are produced, nor is there any point of par-
ticular interest in the later development.

When mature the seed consists of a seed coat of a few layers
of very much compressed cells derived from the outer zone of the
integument, enveloping the endosperm, which, in turn, surrounds
the embryo.

The cells of the endosperm are rich in proteids with a few
scattered starch grains. The cell walls are moderately thick,
while those of the outer layer of cells are considerably thickened.

The separation of the seed from the placenta takes place at
the plane of the insertion of the funicle in the seed, along the plane,
therefore, in which the growth of the funicular cells by division
takes place (ff, fig. 10).

SUMMARY AND CONCLUSIONS 59

SUMMARY AND CONCLUSIONS

The results presented in the foregoing pages have been derived
from a study of 23 species representing 9 genera of the Rubiaceae.

In view of the great size and complexity of the family as at
present understood, any conclusions based upon so small a pro-
portion of the whole segregate may be applied with probable
truth only to the genera represented by the species studied.

Of the lot, the Galieae have been the most thoroughly
- studied, and this group has been found to present a high degree
of homogeneity, and the characters ascertained, with one exception
to be noted below under 5 may be received as applying with only
um differences to all the plants of Galicaz.

. In all the plants considered, excepting in those of the genus
a two ovules are present. Their primordia arise in the
floor of the ovary on either side the floral axis, and develop in
such a manner as to produce approximately anatropous ovules
provided with a single integument and with a greatly reduced
nucellus which is not, to be distinguished except at an early age,
when it takes the form of a cap consisting of a single layer of cells
crowning the archesporium. Inthe Compositae this layer forms a
complete investment of an unicellular archesporium. It would
appear, then, that the nucellus in these Rubiaceae is suppressed, a
view however which is hardly in accord with the facts ; for the ap-
parent lack of a nucellus such as we find in the Compositae, may
be regarded as a result of the thickening of the tissue of the pri-
mordium about the sides of the archesporium and the origin of the
much thickened integument nearer the morphological apex of the
primordium. According to this interpretation the nucellus و‎
marked by the more abundant growth of its sides and its intimate
relation to the integument.

This circumstance, if, as I hold, a fact, leads toa a n
delimiting the nucellus as ordinarily defined, a difficulty however
no greater than that encountered in defining any organ when
. viewed in the light of modern morphological conceptions.

2. In the Spermacoceae there is, in addition to the integument,
a second outgrowth, derived from the funiculus and here called a
strophiole. This structure contains the vascular supply of the

*

60 COMPARATIVE EMBRYOLOGY OF THE RUBIACEAE

ovules and possesses also in large numbers special excretory cells
which become loaded with raphides. |

Similar cells are found in some @alieae in the immediate vicinity
of the embryo-sac for some time before and after fertilization, and
the róle which the contents play in the physiological economy of
the ovule is by no means clear.

It may turn out that in both instances the calcium oxalate is
of some positive value, and the term “excretory ” as used above
must be regarded as provisional.

3. In Houstonia each locule of the ovary is provided with a
club-shaped placenta bearing each a number of ovules. The to-
pographic relations of a placenta in Houstonia and an ovule in the
Galieae are the same, and no differences have been observed in
their origin.

The ovules in Houstonia are relatively very small and of a very
simple type in that an zztegument is wholly absent. There is ac-
cordingly no micropyle and the archesporium becomes deeply :
buried in the nucellar tissue by the growth of the capping cells and
of the tissue at the sides of the archesporium toward the apex of
primordium. The archesporium further consists of but one func-
tional embryo-sac-mother-cell.

Two questions arise concerning the above comparison ; first,
whether we may regard the ovule of the Galieae and the placenta
with its ovules in Houstonia as morphological equivalents, and
secondly, whether we have here a case of correlation.

To the first we can give no positive answer at present since a
more careful study of the origin of the primordia is needed. If
such should prove to be the case we should have a sort of branch-
ing or compound sporangium, a most interesting condition if true.
: To the second question a more decided answer in the affirma-
tive may be given in view of the fact that where the placenta is
formed the ovules are arrested both as regards the integument and
archesporium and in total relative size. That the archesporium is
really suppressed appears from the fact that commonly about seven
embryo-sac-mother-cells are laid down, all but one of which fail to
reach the normal size.

It must be added that the belief that a simple naked ovule,
that is, one without any integument, is correlated with the para-

Li

,

*

SUMMARY AND CONCLUSIONS 61

sitic habit of growth, must now be thrown aside. This conclusion
was suggested inferentially by Goebel* in connection with his ob-
servation on the ovules of Crinum Asiaticum of which he remarked :
“Sie zeichnen sich durch eine, sonst nur bei Schmarotzerpflanzen
bekannte Eigenthümlichkeit aus, sie sind nackt, es fehlen die
sonst in Ein- oder Zweizahl vorhandenen Integumente.”

4. The archesporium throughout contains a number (7-15) of
megaspore mother-cells, the largest number occurring in the spe-
cies of the Galicae. Each megaspore—barring the arrested ones
at the side of the archesporium—divides twice to form four mega-
spores, which, except very infrequently, are not separated by
walls. This condition is comparable to that described as occur-
ring with some frequency in Zichhornia} and Avena.} All the
megaspores are both morphologically and physiologically equiva-
lent. It is, in consequence, impossible to pick out with the eye
the functional megaspore, which appears usually to be the outer.
As Koernicke § has recently emphasized, the embryo-sac cell in
the great majority of Phanerogams is derived from the last mega-
spore, 2 e., the one at the chalazal pole, because usually the divi-
sions which give rise to the megaspores are such as to give the
largest proportion of plasma to that cell, a phenomenon which he
holds to be connected with its more favorable position with refer-
ence to the source of nutriment. This view finds support in the
facts here summarized in so far that the larger number of mega-

spores and undivided megaspore mother-cells form a nutritive tis-
sue surrounding more or less completely the embryo-sac cell
which arises near the longitudinal axis of the mass. That the
functional megaspore, 7. ¢., the embryo-sac cell is found near the
middle, and not in the sides of the archesporium is due, I hold, to
the mechanical relation between the cells brought about by the
growth of the ovule which is such as to produce no unequal pres-

sure on the archesporium.

Vi e eeg Schilderungen, t1: p.
1 Smith, W. R. A Contdibition to de Life History of the Pontederiaceae. Bot.
Gaz. 15: m

189 8.
iCannon, W. A. A morphological Study of the Flower on Pes of the Wild
Oat, Arena fatua L. Proc. Cal. Acad. Botany, III. 1: 329
Studien an Euibeyötóck- -Matterzellen, Sitzungsber. de Néeaéchen. Gesellsch.
Natur- und Heilkunde zu Bonn.

62 COMPARATIVE EMBRYOLOGY OF THE RUBIACEAE

That, in forms in which a pluricellular megasporangium is pres-
ent, any or all of the megaspores have the ability to develop into
embryo-sacs proves in a most convincing manner the morpholog-
ical equality of the megaspores, and the regular division of each
of the megaspore mother-cells into four megaspores must be
regarded as a true tetrad division.

At this point it is of interest to compare this result with those
reached by Murbeck * who has studied A/chemilla, a representa-
tive of the Rosaceae. The presence ofa multicellular archesporium
in this family was first made known by Strasburger. + According
to Murbeck twelve to sixteen mother cells are present, of which
the peripheral and usually one of the centrally placed, fail to
divide, and moreover never give rise to embryo-sacs. The re-
mainder, six to nine in number, divide at least once and of the
daughter-cells some divide a second time, and the granddaughter-
cells thus formed are regarded, properly, as megaspores, It may
be noted in passing that here there is no reduced number of chro-
mosomes, reproduction being parthenogenetic. Of the megaspores
thus produced, two or more may develop into young embryo-sacs,
and what is of immediate and special interest in this connection,
in the same Sporogenous cell-row, tivo or more may commence their
development into embryo-sacs. It thus appears that Alchemilla and
Crucianella are precisely comparable in this regard; while the evi-
dence derived from the latter is still more final because of the
actual division of all the megaspores derived from a simple mother-
cell. Further cytological evidence that these are to be regarded
as true tetrad divisions is presented on another page.

It is scarcely necessary to add that these facts accord an extra-
ordinary strength to the view that the ovule is a sporangium.

In the Spermacoceae and Oldenlandeae the archesporium con-
tains only one megaspore mother-cell of special size, from which
the embryo-sac cell is derived. Several suppressed megaspores
are present, surrounding the large megaspore mother-cell, which
divides to form four megaspores, the basal of which forms the
embryo-sac cell.

ee Embryobildung in der Gattung Alchemilla. Acta. Reg. Soc.
۱ , 11: No. y. I. Fora more extended examination of the relevant
literature see this valuable paper.
Angi

f Die ospermen und Gymnospermen,

¥

SUMMARY AND CONCLUSIONS 63

The appearance of a pluricellular archesporium may by no
means be considered as primitive. It has been shown by several
workers to occur in widely separated families,* and with certainty
may be said to have no phylogenetic significance. On the con-
trary, the meaning is purely physiological, and such conditions
have arisen in various forms in response to similar conditions.
When present they offer favorable nutritive conditions for the
chosen embryo-sac. This view comports at least with the case of
Alchemilla as interpreted by Murbeck (/. c.) The archesporial
tissue not directly concerned in the formation of the embryo-sac
takes activity or passively a nutritive rôle. If their rôle is passive
they gradually disintegrate, and are absorbed. This appears to
be the case in the Galieae, Diodia teres and Richardsonia. In the
Diodia Virginica they grow chiefly in length and arrange them-
selves asa transporting tissue. As such they connect the embryo-
sac with the vascular supply of the ovule, and persist for a
considerable period after fertilization. In Houstonia the subsid-
lary megaspore mother-cells appear to persist till about the time
the embryo commences to develop, and undergo regular and con-
stant changes in size and form. These changes indicate their
physiological importance.

5. The embryo-sac cell behaves in two ways. In Houstonia
and Richardsonia it develops in situ. Inthe Ga/zeae and in Diodia
it moves forward out of its original position, breaking through the
nucellar capping cells, and passing along the micropylar canal. Its
passage is intercellular, and its action is so far analogous to the
movement of the pollen tube, as described by Murbeck and myself.
As it moves it derives nutriment from the disintegrating cells
adjacent to it. The disintegration is such as to suggest very
strongly the secretion of an enzyme by the embyro-sac.

The embryo-sac presents some remarkable features as regards
the antipodal cells. These, in the GaZeae, with the possible ex-
ception of one species of Asperula, are three in number, one of
which is very much elongated. The free end of this cell is plunged
into the mass of disintegrating megaspores and absorbs the same.
The physiological importance of all the antipodals is indicated by

*For full citation of the literature see the papers of Murbeck and Koernicke

D

64 COMPARATIVE EMBRYOLOGY OF THE RUBIACEAE

their structure, persistence, food content, and the behavior of the
surrounding tissues.

The genus Crucianella offers an exception to the general con-
dition in the Galieae. Although antipodals are developed no evi-
dence has been obtained that they take on any special cytological
characters. Indeed their tenure of life is so short that the three
antipodal cells have not been seen at the same time. Their func-
tion is probably, therefore, at best but passive, and is carried out by
the mere giving up of their substance in death to the embryo-sac.

In Diodia teres, also, the number of antipodal cells is three, but
present a difference in that the long one is relatively much shorter,
a feature which is undoubtedly correlated with the smaller amount
of archesporial tissue. They may, however, not be considered as
physiologically inactive since they secrete a large amount of starch,
and persist for a long time.

In Diodia Virginiana the number of antipodals is greater, the
number varying, so far as observed, between four and ten. These
are arranged in a long series, physiologically equivalent to the
single long antipodal in the GaZeae. The subsidiary megaspore
mother-cells, as above pointed out, persist and form a continua-
tion of the series. The antipodal cells retain their normal char-
acter for a considerable period covering the earlier period of endo-
sperm development. :

In Richardsonia and //oustonia the antipodal cells are equal in

all regards, and appear to have fess permanent value. They
acquire the cytological characters of glandular cells such as are
found in certain nectarial tissues and stigmas, and are probably
active in some kind of secretion. "Their relative small importance
as judged by their size is connected with the behavior of the
embryo-sac cell and the much suppressed archesporium.
That, in the words of Westermaier (4. c., Part 1), we have to do,
in the antipodals, with “an anatomical-physiological apparatus,
and not with a useless rudimentary structure which may be under-
stood only from the viewpoint of comparative morphology,” * is
evident from the facts as here summarized :

= is of great interest in this connection to recall the remark of “Hofmeister made
Mim fty years ago : “ Die am Chalaza-Ende entstandenen, durch ihre betráchtliche
rösse oft sehr ausgezeichneten Zellen scheinen keine andere Bestimmung zu haben,
als die der Verarbeitung des Nahrungsstoffes für den werdenden Embryo." Die Ent-
stehung des Embryo der Phanerogamen, p. 59. 1849.

SUMMARY AND CONCLUSIONS 65

(a) The form and cytological structure .of these cells, and the
changes of their food content.

(4) Their tendency in some plants to multiply and form a
special tissue, of the nature of conductive or nutritive tissue.

c) The behavior of the surrounding tissues notably the arche-
sporial cells, which are absorbed.

We may not overlook the case of Cructanella, for which evi-
dence is lacking that the antipodals play any but a very brief,
passive role.

7. The embryo in the GaZeae develops rapidly, especially in
its early stage, and possesses a remarkable feature in the presence
of a highly developed embryonic organ, the suspensor, the
cells of which elongate laterally, and penetrate the surrounding
endosperm in an intercellular manner without causing any disinte-
gration of the endosperm. These outgrowths are termed haustoria.
Analogous organs are known in a number of forms, and probably
occur in many more in one form or another. The function of the
suspensor in these forms is therefore not alone to bring the embryo
into favorable position with relation to the food supply in a me-
chanical sense, but to act as a temporary embryonic root. This
function is secondary, the former being regarded as primary, as for
example in the case of Selaginella * in which it appears to have no
further significance.

A strikingly analogous condition appears in certain species of
Tropaeolum, according to Dickson, in which the embryo pro-
duces “ branches from the base of the suspensor ” which “ serve
as organs of nutrition for the developing embryo— foetal roots, in
fact" These organs penetrate into the pericarp.

Similar organs have been described by Treub f in the Orchida-
ceae and by Hofmeister,$ followed by Guignard || in the Legu-
minosae.

Pfeffer, W. Die Entwickelung des Keimes der Gattung Se/agınella. Hansteins,
bot. er 1: No
fOn the تسس‎ of Tropaeolum peregrinum (L.) and T. speciosum (Endl.
rt

and Paepp.). Trans. Roy. Soc. 17 : Part 2.

t Notes sur l'embryogenie ye quelques Orchidees. Naturk. Verh. K. Akad, 19:
1879.

2 Die Entstehung des Embryo der Phanerogamen, 1849.

l| Ann. sci. nat, VI. 12: 5. 1881.

COMPARATIVE EMBRYOLOGY OF THE RUBIACEAE

It is, further, not improbable that the peculiar features of the
suspensor in some Cruciferae, 0 and many other plants are
connected with similar physiological conditions.

The embryo in the Spermacoceae and in Houstonia develops
very slowly, especially in its early stages. We - have to note here
the complete absence of the adaptive characters in the suspensor
seen in the 6۵/2۵۵۵ ; the suspensor is of a very simple structure,
and upon anatomical evidence alone we might attribute to it only
the mechanical functions above indicated. In view of the distri-
bution of food stuffs as already described, however, we must
recognize the suspensor in these forms as a less special haustorial
organ than in the Ga/ieae, but still an organ of that nature.

It may further be noticed that the production of haustoria
from the suspensor of the embryo, and, so far as known, from no
other part, is probably an expression of polarity in the embryo.

8. The endosperm is of two kinds, according to its method of
development, corresponding to two of Hegelmaier's * types.

In the Galieae, the endosperm during its whole course of de-
velopment fills completely the endosperm cavity (Hegelmaier's first
type) The peripheral cells on the basis of their cytological char-
acter and the behavior of the adjacent tissues may be regarded as
concerned in the secretion of an enzyme, and in the absorption of
the food derived by digestion of the reserve food substances in the
integument.

In the Spermacoceae and Houstonia the endosperm is at first
parietal, later c coming to fill completely the enclosed cavity (Hegel-
maier's second type). In addition to the peripheral cells, which
appear much as in the Gadieae, the centrally placed cells in Diodia
and A/chardsonia take on a more special character with a prob-
lematical significance.

THE TETRAD AND EMBRYO-SAC MITOSES
(PLATES 9 AND 11) ۱
The following account of the mitoses of the archesporium and
of the embryo-sac is based upon a study of Asperula montana,
Crucianella macrostachya and C. gilanica. Of the two genera,

* rc über die ee des Dicotyledonen—Endosperms. Nova
885.

Acta Akad. Naturforscher., 49 : No.

TETRAD AND EMBRYO-SAC MITOSES 67

Crucianella offered by far the more favorable material both on ac-
count of the size of the cells and of the absence of unequal pressure
due to peculiarities in the development such as occur in Asperula.
The former has therefore received more exhaustive study, although
in regard to certain points Asperula has given important results.

When the ovule is young the embryo-sac-mother-cells are to
be distinguished only by their greater size. During the early de-
velopment of the integument, however, or even earlier, their cyto-
plasm becomes denser. At about the same time a large number
of fibers appear in the cytoplasm traversing the cells longitudinally
but in somewhat irregular fashion. -The fibers are more or less
curved and taper off at their ends.

Inasmuch as the embryo-sac-mother-cells attain a strikingly
great length before the first division takes place (pl. 10, fig. 16)
and have their cytoplasm crowded with these fibers, they present
a very unique appearance. In some cases the fibers traverse the
cells from end to end, bending aside to avoid the nucleus. The
condition here described apparently corresponds to that found by
Mottier * and Dixon + in LZzuz, and by Duggar in Symplocarpus,
during the earlier development of the embryo-sac mother-cell, ex-
cepting in certain matters of detail and interpretation.

From the figures of Mottier and Duggar one would judge
that the fibrillar structures result from a rearrangement of the
cytoplasmic reticulum. That the large thick fibers found by me
in the developing embryo-sac-mother-cell are formed in the same
way seems to me not to be the case, for, in addition to their great
thickness, I have not been able to see a pulling out or distortion of
the adjacent cytoplasm. Reference will be made to a possible in-
terpretation of these structures beyond.

Again, according to Mottier, the orientation of the fibers in
Lilium is not constant in different cells of the same stage of growth.
They may run in various directions or they may sometimes forma
dense felt-like zone about the nucleus. Further, this cytoplasmic
differentiation disappears before the commencement of mitosis, and
+ Ueber das Verhalten der Kerne bei = an des Embryosackes und
die Vorgänge bie der Befruchtung. Jahrb. 3 - 97.

On the chromosomes of Lilium longi mm Proc. Irish Acad. 1 + 946,

+ Studies in the en of the € grain in Symplocarpus ۳ and
Peltandra undulata, Bot. Gaz. 29: 81.

68 COMPARATIVE EMBRYOLOGY OF THE RUBIACEAE

by inference, has nothing to do with the formation of the mitotic
figure. At the beginning of mitosis the cytoplasm shows again a
uniform reticular structure, and the spindle of the first division
takes its origin in radial or more or less irregularly placed fibers
which penetrate the nuclear cavity, and form a multipolar spindle
primordium.

With regard to the disappearance of the earlier formed thick
fibers I am not able to establish a parallel in the Rubiaceae, for, in
the embryo-sac-mother-cell in the plants studied they persist until
the earlier stages of mitosis, at least through diakinesis. That
they are in some way connected with the kinoplasmic figure dur-
ing mitosis would appear from the circumstance that their ends on
reaching the nucleus bend and run as figured in Mottier's ۸ 2,
fig. 17, of the paper referred to. This condition, as seen in Cru-
ctanella macrostachya is shown in pl. 11, fig. 37. From this and
the adjoining figures it will be seen that the origin of the spindle
is multipolar, and, with the possible exception of the peculiar be-
havior above described, corresponds to the descriptions of Oster-
hout * and others. The definitive spindle of the first division is
strictly bipolar, and differs in the three plants in one or two details
of form only.

In Crucianella gilanica the spindle is of very regular, broad,
oval form, with sharply-pointed ends ; in C._macrostachya it has a
looser, more slender form, with attenuated ends, while in Asperula

this character is still more accentuated. Such differences may in’

the main be regarded as inconstant and of no general importance,
and may in many cases also be appearances referable to the plane
in which the material happens to be cut. Insufficient observation
of this point may very easily lead to false impressions, as Stras-
burger t has hinted, as with reference to the character of
the spindle poles.

Generally the ends of the spindle poles lie freely in the cyto-
plasm, inasmuch as the spindle axis usually lies parallel to the
longo axis of ms cell, which, as oe shown, is er per

* Ueber E
"s Pg TETE don d a Spindel bei سر‎ Jah wiss.

Ueber Red
am e Spindelbildung, Centrosomen und Cilienbildner im

*

TETRAD AND EMBRYO-sac MITOSES 69

Occasionally, however, the spindle is found to lie obliquely, and
one or both ends may become connected with the ectoplasm.*

In no case is a centrosome or centrosphere to be found. I
have been able to follow finely-pointed spindles to their extreme
end without finding any trace of the body which can without.doubt
be regarded as such. Ina recent paper by Meves and von Korff
on the Myriopoda, ۲ the ‘‘centralkérper” are described as lying
very near to the periphery of the cell, and are apparently not di-
rectly in connection with the spindle. The latter, however, is not
pointed, while its fibers nevertheless are directed toward the kinetic
centers,

Thus is formed a condition in which the centrosomes are con-
siderably removed but still in control of the spindle ends.

In spite of the obvious difference in the character of the spindle
itself, I was led to examine carefully every part of the cytoplasm
lying in or near the spindle axis, but with a negative result.

The multipolar origin of the spindle in certain plant cells has
' been for some time regarded by the majority of botanists as speak-
ing against the existence of the centrosome in plants in which the
spindle has such an origin. Although a contrary view has been
held by Guignard f it would still appear that the former position
is more consonant with the facts as they are at present understood.

For we would expect in such cases either as many centrosomes
as spindle ends, or two whose influence would be exerted over the
whole kinoplasmic figure (archoplasm). The former is shown to
be the case in the development of the curious worm-shaped sper-
matozoon in Paludina, an account of which has lately appeared. $
According to Meves the centrosphere of the primary spermatocyte
contains a number of central granules which act as a unit. In
the second mitosis the granules (Centralkórperkórner) separate
and take a position in the periphery of the cell, and from these as
centers are formed a number of half or partial spindles. The re-
sult is a picture which resembles in its later development to some

* Ectoplasm is here used to signify the ** Hautschicht ”of the Germans.

t Meves, Fr , and von Korff, K. ** Zur Kenntniss der Zelltheilung bei ENSEM
Archiv mik. Anat. und Entwickelungsgeschichte, 57 : 481-486. fl. 27. 1901

I Centrosomes in Plants. Bot. Gaz. 15 : 15 1999.

¿Meves, Fr. Ueber die sog. wurmfórmigen Samenfäden von Paludina und über
ihre Eo Viele: Mitth. Verein Schles.-Holst. Aerzte. 10: No. 1

70 COMPARATIVE EMBRYOLOGY OF THE RUBIACEAE

degree the multipolar spindle primordium in plants, to which the
author unhesitatingly compares it, with the inference that the
plant cytologists have still overlooked the structures in question.
'But the comparison is without warrant, for an examination of the
facts shows that we are dealing with two different modes of action.

The origin of the spindle is in the two cases quite different.
In Paludina it is in the periphery of the cell, while in the higher
plants it is at the nucleus. The development of the spindle is in
the former therefore centripetal; in the latter centrifugal. We
fail therefore to see any warrant for the conclusion that there is
any parallel between the phenomena described by Meves, and
those seen in the formation of the plant spindle of multipolar origin.

The parallel fails also in the following regard. In Paludina
the spindle fibers develop from the centers. In plants, even ad-
mitting the existence of centers, the fibers are developed to a very
considerable degree before the centrosomes, or centrosome-like
bodies, when such are recognized, appear.

For the present, therefore, we must adhere to the view that
the centrosome is absent from the higher plants.

We do not attempt to meet in detail the statements in certain
recent papers to the contrary. * It is quite impossible to do this
except by working with the same materials and methods, and this
has not been done by me.

Nor, as Strasburger (/ c.) has shown, is positive evidence
against the occurrence of the centrosome in higher plants want-
ing. Repeated studies on pollen development have been carried
out, showing that the ends of the spindle find an insertion in the
ectoplasm and the results of my own observations tend strongly to
confirm the view that the ectoplasm serves in many instances as a
place of connection in those cells in which the spindle is long
enough to bridge the distance from side to side. One instance
(Asperula) was seen in which a pluripolar diarch spindle had its
four apices (appearing in a single section) were all inserted in the
bct Los 9, fig. 16).

— ME E M M

Bernard, Ch. Recherches sur les sphere attractions chez Lilium candidum,
Helosis ری‎ etc. Jour. de Bot. 14: 118, 177, 206. 1900,

amanuochi, S. Einige وس ور‎ uber die Centrosomen in den Pollen-
mueren von Lilium longiflorum. Bot. Centralbl. Beihefi fte, 10: 301. 1901.

. H. ۸ Contribution to the Life Histo
ry and cytology of Erythronium.
Bot. Gaz. i DN June, 1901. ytoiogy 7

TETRAD AND EMBRYO-sac MITOSES 71

When, however, the true apex of the spindle does not insert
itself immediately in the ectoplasm, one may usually discover a fine
strand of kinoplasm extending from the apex of the spindle to the
ectoplasm (/. 9, fig. ro).

In further support of the view that the spindle often stands in
some intimate relation with the ectoplasm I may cite the instance
from which fig. 4 of Pl. و‎ was taken. The axis is here oblique to
the optical plane, and one pole is seen to end distinctly at a point
from which extend fibers in a somewhat radial fashion. Careful
examination revealed the fact that the radially placed fibers lay in
the peripheral layer of the cytoplasm, that is in the ectoplasm, and
the latter therefore appears to bear a distinct mechanical relation
to the spindle.

In many cases in which the dimensions of the cell are much
greater relatively to the size of the spindle such attachments have
not been found. This lack is especially obvious in the large
embryo-sac-mother-cell of Crucianella, in which, however, as in
the pollen spindles, extra-nuclear “supporting fibers " may be ob-
served to run out from the spindle apices to the ectoplasm in the
direction of the equator (pl. rz, figs. 16, 77). Similar conditions
prevail in the pollen mother-cell in Crucianella macrostachya
(pl. 11, fig. 35), where but a single row of sporogenous cells
occurs in each locule of the anther, The cells are relatively large
and the mitotic figure in both the first and second divisions lies
freely in the cytoplasm. In this instance neither the insertion of
the spindle poles, nor of the extra-nuclear supporting fibers in the
ectoplasm could be seen with certainty. Indeed the orientation
of the spindle fibers during the second division is often such as to
preclude the possibility of the insertion of one of the poles at least,
as will be seen in p/. rz, fig. 35, which shows that the axes of
the two spindles lay in the same plane at right angles to each
other; unless, indeed, we may expect to find the pole of one
spindle passing through the body of the other. _

In the developing pollen in many plants a distinct zone of
very granular cytoplasm has been described by Juel,* Lawson, f
2. Bie Kaenthellungen | in den Pollenmutterzellen von Ferien diti fulva und die
bei denselben auftretenden Unregemässigkeiten. Jahrb. wiss. Bot. 30: 204. 189

Some observations on the Development of the Karyokinetic — in the
Pollen mother-cells of Cobaca scandens. Proc. Cal. Acad. Botany, III. 1 : 899.

72 FE" COMPARATIVE EMBRYOLOGY OF THE RUBIACEAE

Byxbee, * Williams, + and others, and it has been suggested by
Strasburger that, in those spindles of which the ends do not reach
the ectoplasm, or in the absence of other mechanical support, the
pericaryoplasm (of Lawson) may serve that purpose.

I have observed this zone in the plants under consideration,
but in instances the poles of the mitotic figure reach the ectoplasm
)۸ 11, fig. 2). In this event the zone thins out as it approaches
the spindle pole. In the pollen mother-cell of Crucianella macro-
stachya this zone has not been observed and the spindle appears to
lie quite freely and without any special mechanical relations.

The amount of variation in the observations of the topographic
relations of the spindle ends leads to the suspicion that we are yet
unable to form a satisfactory explanation of them. The most far
reaching interpretation among the botanists is that of the Stras-
burger school, and is based upon the mechanical conception that
the spindle is made up of contractile fibers (Zug-fasern) which, in
order to pull apart the chromosomes, must have a more or less
rigid mechanical support. This has been found in the ectoplasm,
in the pericaryoplasm, in the vacuoles,f or the spindle ends may
be buttressed by extra-nuclear supporting fibers.

That such mechanical relations as described by Strasburger
and others really exist we may not doubt. All the appearances
lead irresistibly to that conclusion. But I am by no means con-
vinced that we may infer from these appearances anything as to
the intimate workings of the mechanism, and the contractile fiber
explanation may have to undergo perhaps a considerable degree
. of modification before it may be accepted as satisfactory.

Reference has already been made in another part of this paper
to the absence of cell walls between the megaspores. It will be
seen from pl. Ir, fig. 20, and pl. 9, fig. 23, that the connecting
fibers are formed, and the thickenings indicative of the cell plate
are present. In Crucianella the presence of these fibers determines
the appearance of the cytoplasm for some time (pl. 10, fig. 9).
At the end of the third division in the same plant the nuclei are

*'The Development of the eg si in the Pollen Mother- cells of
Lavatera. Proc. Cal. Acad. Botany, III.

The Origin of the LOUPE inb in e coerulea Linn. Proc. Cal.
Acad. Botany, III. 1 : 1899.
+ Strasburger, 1. c.

TETRAD AND Empryo-sac MITOSES 73

arranged sometimes in two rows, and the connecting fibers run
from one nucleus to the other in such a manner as to recall
Mottier's* figure of the four nuclei in the upper end of the embryo-
sac in Helleborus foetidus, and the appearances which one usually
meets with during the formation of cell walls in endosperms (pl.
10, fig. 18).

The prophase of the first division in both pollen and embryo-
sac-mother-cells is characterized by the stage known as diakinesis
(of Hácker), in which the chromosome pairs become condensed
into short, thick rods which lie parallel or become partially twisted
about or crossed upon each other. The members of each pair
may fuse at one or both ends, and as the result of more or less
separation in the middle, may take also the form of rings either
completely closed, or open at one point. Between these more
definite forms the chromosome pairs may assume all sorts of inter-
mediate conditions (Pl. zz, figs. 1, 14, 32, 37, 38).

In some instances the members of the chromosome pairs may
be separated from each other for a considerable distance, and the
linin strands which unite their ends, may under such conditions be
seen with the greatest clearness )۸ 9, fig. 27, a). In Galtonia
candicans, according to Schniewind-Thies + the same behavior has
been noted.

The formation of the rings takes place much more completely
and earlier in 4Asperula in which genus also the condensation of
the large loose rings into small thick ones may be followed with
ease (pl. 9, figs. 1 and 2). The appearances so presented recall
the figures of vom Rath + of the chromosome rings in Gryllotalpa
excepting, however, that no phenomena of tetrad formation (in
the sense of the zoologists) may be observed.

During the prophase of this division the number of chromo-
somes may in these forms quite easily be determined and with
a high degree of certainty has been found to be oxe-half the usual
number in the sporophyte. The reduced number, which has been
determined by counts, not only in the prophases, but in the ana-

a das Verhalten der Kerne, etc. Jahrb. wiss. Bot. 1897.
e Reduction, der Chromosome Zahl und die ihr folgenden Kerntheilungen in
den nie der Angiospermen. Jena, 1901.

} Zur Kenntniss der Spermatogenese von Gryllotalpa vulgaris.
1892,

Arch. Mic. Anat.

74 COMPARATIVE EMBRYOLOGY. OF THE RUBIACEAE

phase of the first and second pollen division and in the first, second,
third and fourth embryo-sac divisions, is in Crucianella the hitherto
unobserved and surprising number ten, in Asperula, twelve.

These numbers, however, are believed, on the basis of a large
number of counts at different times, to be inconstant. In Crucia-
nella eight to eleven have been counted, and in Asperula a similar
amount of variation has been observed. Allowing for error in
counting, which, in these plants is not great, both on account of
the smallness of the cells (with the consequent frequency with
which the whole nucleus appears in a single thick section) and of
the number of chromosomes, I believe it must be maintained that
variation in this regard occurs, and that this is a fact which has
heretofore received insufficient recognition.

The condensation of the chromosome pairs proceeds all through
the prophase until the equatorial plate has been completed, when
they take the form of a double Vs, one of which lies above and
one below the equatorial plane, with theis ‘angles directed toward
the axis of the spindle. This arrangement is difficult to recognize
in metaphase on account of the extreme condensation of the chro-
mosome, but in very early anaphase may be seen with sufficient
distinctness.

My observations and conclusions appear to coincide with those
of Duggar* for Peltandra. That author's interpretation of his
figures of the corresponding division in Symplocarpus as indicating
a qualitative division, however, would in my judgment seem not to
be justified. For the chromosomes are here, as represented in
Pl. 1, figs. 7 and 7a of Duggar's paper, of double V-form with the
angle directed toward the center of the spindle. These separate
first at the angle, and the degree of similarity in the two plants is
So great as to render them susceptible of but a single interpreta-
tion, which is opposed to the view that there is a qualitative divi-
Sion, unless it can clearly be shown that the separation of the limbs
of the V is a separation, not at their ends, but across the middle
of the chromosomes.

Attention should further be directed to the above mentioned

observation, namely, that in diakinesis and the later prophase the
Mene ERR nd a

* Studies in the development of the poll in ; d
Peltandra undulata. Bot. Gaz. 19: 81. Fr grain in Symplocarpus foetidus an

TETRAD AND Empryo-sac MITOSES 15

ends of the chromosome pairs do not invariably fuse even when
they take the form of rings. With the approach of metaphase the
two members if separated become laid together. The fixation of
the fibers then takes place somewhere between and not at the ends
of the chromosomes, so that the possibility of the occurrence of a
transverse splitting of the chromosomes at this time, as suggested:
most recently by Schaffner * for Erythronium, seems to me to be
excluded.

If the ends of the chromosome pairs always fused completely
it would probably be impossible to tell whether the breaks during
anaphase occurred at the points of fusion or somewhere between
these, When, however, such a fusion does not take place the
simpler and more obvious interpretation does not involve the idea
of a reducing division.

The separation of the daughter chromosomes takes place with
extreme regularity but curiously enough with quite constant differ-
ences in the two genera. In Asperula the ends of the daughter.
Vs separate simultaneously and the chromosomes take with little
variation the form of a conventional heart. This is to be explained
by the shortness of those bodies and the point at which the fibers
are inserted (pl. 9, figs. 3, 22).

In Crucianella, on the contrary, the ends of the V-formed seg-
ments scarcely ever separate at the same time. The ones which
remain united the longer are attached to each other with so great
a degree of firmness that it is only after the greater mass of each
daughter chromosome reaches about half the distance to the
spindle poles that they finally separate (pl. 17, figs. 2, 3, 16, 17,
18, 39). The forms taken by the separating chromosomes under
these conditions have received special study, and all of them sug-
gest very definitely the behavior of plastic masses under tension,

` Examination of the detailed representations on pl. 77 (fig. 3, a—
f; fig. 18, a-g) will reveal this point quite clearly. „Especially in-
structive are such forms as shown in fig. 78, a, as they bear on
the question of transverse division. Were the chromosome here
breaking transversely and not undergoing separation at united
ends, it is difficult to see why such a bulge at the point of separa-
* A Contribution to the Life History and Cytology of Erythronium. Bot. Gaz.
: 369-387. 1901

76 COMPARATIVE EMBRYOLOGY OF THE RUBIACEAE

tion should occur. On the other hand, it is quite easy to under-
stand such an appearance on the supposition that a plane of ad-
hesion exists at that point.

Whether the unequal separation of the daughter chromosomes
is due to the failure of the mantle fibers to exert an equally dis-
tributed pull upon them, or to the fact that the fusion of the chro-
mosome ends is not equal cannot be determined. It is a further
point worthy of note that the two free ends of an incompletely
separated chromosome pair are always directed away from the
spindle axis, as if the mantle fibers were stretched along a resistant
surface. After the separation of the daughter chromosomes is
complete, a readjustment of tensions takes place so that the heart-
shaped form spoken of in connection with Asperula is approached
(pl. TE, figs. 4, 16, 17, 18,8). =

During late anaphase a second longitudinal splitting of the
chromosomes is seen with a fair degree of clearness (ZL 9, figs.
6-9). On account of the shortness of the resulting chromo-
somes, however, it is not easy to determine the forms which they
take. For the most part they appear as pairs of straight rods,
although some evidence that they are slightly bent into a V is
seen. Itis not at all improbable that a considerable amount of
readjustment in the relations of mantle fibers and chromosomes
takes place and this would explain the change in the shape of the
chromosomes.

The appearances here described have been constantly seen in
a large number of cells and there is very little doubt that the first
division in Asperula is heterotypic in the sense adopted by Stras-
burger (A c., 99).

The corresponding longitudinal splitting of the chromosomes
in Cructanella does not take place so early as in Asperula, and may
be seen with clearness only in the telophase. 77. zr, fig. 5, repre-
sents sucha stage. It appears not possible to resolve the chromo-
some pairs thus formed during the whole of the period extending
from this time till the metaphase of the second division; for as
shown in pl. rr, Fig. 6, the pairing is indistinguishable, although
a completely resting condition, as described for Larix by Juel, ۳ in

* Juel, H. O. Beiträge zur Kenntniss der Tetradtheilung. Jahrb. wiss. Bot. 35‘
626, 1900,

TETRAD AND EMBRYO-sac MITOSES TT

Pellia, by Davis,* in Asc/epzas,t by Strasburger, and in certain
Liliaceae by Schniewind-Thies i has not been observed to
_occur.

From the facts above presented it will be seen that in Crucia-
nella, as in Asperula, the first splitting of the chromosomes is longi-
tudinal. In early anaphase the daughter rods which open out
from a double V-form into rings, separate finally at the ends of the
limbs of the V, the angles being directed toward the pole. Later
a second longitudinal splitting appears, the formation of grand-
daughter rods being completed before the close of the first mitosis.
This division is, therefore, heterotypic and corresponds in essential
details to the heterotypical mitosis in the spermatocytes of the
salamander as described by Flemming.

The final evidence for this conclusion in regard to the embryo-
sac-mother-cell has not been obtained without any doubt, but I
believe that we may conclude from the otherwise complete corre-
spondence in the two cases that the first embryo-sac-mother-cell
division is also heterotypic.

The resting stage occurring between the first and second di-
visions is characterized by its short duration, and by the varying
degree of diffusion of the chromatin. This variation is considerable
not only among related species, but is evident in comparing the
pollen and embryo-sac tetrad divisions in the same plant, as
Strasburger $ has observed to be the case in Lilium, It is notable
that in the Rubiaceae alsothe daughter nuclei of the pollen mother-
cell enter the resting stage less completely than in the corre-
sponding nuclei in the ovule. If this prove to be generally true we
may attribute the behavior to the closer analogy between the tetrad
divisions in pollen and those of cryptogams, and the greater irregu-
larity in the tetrad divisions in the ovule to the changed condi-
tions which have been brought about in that organ.

_ Although the variation in the character of the resting stage is

* Davis, B. M. Nuclear Studies in Pelia. . Bot 15: 147.

T Einige isn zu der Pollenbildung per Asclepias. Ber. deutsch. bot.
Gesellsch. 19: 1901.

D. Ps T. Die Reduction der Chromosomezahl und die ihr folgen-
den Kerntheilungen in den Embryosacmutterzellen der Angiospe permen. Jena, 1901

3 Ueber Cytoplasmastructuren, Kern und Zelltheilung. Jahrb. wiss. Bot. 10: 252.
1897.

78 COMPARATIVE EMBRYOLOGY OF THE RUBIACEAE

real and considerable, all the cases studied are alike in the absence
of a nucleolus, in the indistinctness of the nuclear membrane, which
appears as if surrounded by an ill-defined zone of fine kinoplasmic
fibers running in various directions, and in the loss, at least so far
as direct observation can take us, of the individuality of the grand-
daughter chromosomes, even when, as in Crucianella, the second
splitting takes place in early telophase.

The occurrence of a more or less complete resting stage at the
end of the first division is a phenomenon of wide occurrence among
the phanerogams. Authors are not agreed, however, as to the
degree into which the daughter nuclei enter at this stage. More
recently Schniewind-Thies has found a considerable amount of
variations in a number of forms studied comparatively by her. It
appears, therefore, that we may explain this disagreement on the
ground of the occurrence of a considerable degree of variation in
this regard.

It does not appear to be necessary to be concerned about the
loss of individuality by the chromosomes under the circumstances
at least so far as the question of reduction is involved. A second
longitudinal splitting being demonstrated to occur before this takes
place, we are compelled to believe in the absence of positive evi-
dence to the contrary that the rehabilitation of the chromosomes
is complete. The Weissmannian interpretation once shut out, it
remains for us to explain the variable degree into which the grand-
daughter nuclei enter the resting condition apart from it. I cannot
agree with Davis, nor do I find for him any justification in the
literature of the subject, either on the zoological or botanical side
of the question, when he remarks “ This conception * is intimately
concerned with, and necessarily a part of, the recent studies
of Guignard and Strasburger on the double longitudinal splitting
of the chromosomes.” (L €, po 150%)

Indeed Strasburger ae has recently shown that in Ascle-
pias (l. c.) a complete resting stage, with a nucleolus, is formed
between the first and second pollen divisions, without making any
remark further than to call attention to the fact.

* That the chromosomes maintain their edidi d in the OR nuclei after
the first mitosis of the pollen mother-cell.

TETRAD AND EMBRYO-sAC MITOSES 79

THE SECOND Division

During the prophase of the second division in the embryo-sac-
mother-cell delicate radiating fibers in great numbers are found
surrounding the nucleus and extending from the same in all direc-
tions (pl. zz, fig. 42). In certain cells where the walls come into
contact with the nucleus these fibers then appear in sections as if
there were two distinct systems arising from the opposite sides
of the nucleus. They thus simulate asters. In point of fact, how-
ever, they form a zone about the nucleus lying parallel to the com-
pressing cell walls (p/. zz, fig. 15)..

Such radiations have been figured and described by many
authors, but accounts are, in spite of a multitude of detail, still in
such a condition as not to warrant us in believing that we as yet
understand fully the relation between the nucleus and these radia-
tions.

Their resemblance, however to certain appearances is parthe-
nogenetically developing eggs of Zoxopneustes after treatment with
magnesium chloride has led Wilson * to make the following sug-
gestion which I give in his own words.

“The formation of a perinuclear zone in the magnesium eggs
strongly suggests the accumulation of ‘kinoplasm’ about the
nucleus in the early prophase of cell division in the higher plants,
as described by Osterhout and Mottier, and a number of later
observers, though the later phenomena are in many respects very
different.’

According to Wilson’s direct observations on the living Mg-
eggs, a primary radiation appears surrounding the nucleus and
extending far out to the periphery of the cell. “This radiation is
interpreted as a flow of hyaloplasm which, in part at least, gives
rise to the appearances making up the “aster.” . The gradual dis-
appearance of these radiations is accompanied by the formation of
a clear perinuclear zone, followed by “enlargement of the nucleus
and the disappearance of the nuclear membrane accompanied by a
reduction of the rays almost to the vanishing point." (4 c.)

The comparison is strengthened by the facts described earlier

۶ peine Studies in Cytology, I. A cytological Study of artificial Parthe-
nogenesis in Sea-Urchin Eggs. Archiv Entwickelungsmechanik der Organismen. 12:

80 COMPARATIVE EMBRYOLOGY OF THE RUBIACEAE

in this paper in connection with the development of the embryo-
sac-mother-cell. The numbers of fibers then seen and their ap-
parent behavior with respect to the nucleus, as already indicated,
in many ways harmonize with Wilson’s suggestion. If further
study should show this harmony to be complete, we would then
have strong grounds, in the absence of direct observation of the
living material, for the belief that these fibers are currents of viscid
material (kinoplasm, hyaloplasm) the view first taken by Auerbach,
and later adopted by Fol, Strasburger + and others in modified
form. ۱

It would appear that a further and more particular study of the
behavior of the kinoplasm during the prophase in plants from the
suggestive point of view just indicated would have a distinct bear-
ing upon the vexing problem of the nature of the fibers.

In the prophase of the second division the granddaughter
chromosomes reappear in Crucianella as pairs of rods lying parallel
to or partly twisted about each other and lying on the nuclear.
walls (pl. 11, fig. 21). In Asperula the chromosomes are so
short that they appear as pairs of spheres (A. 9, fig. 16), though
in metaphase they frequently appear as pairs of short rods, one
lying above and one below the equatorial plate. The mantle
fibers are usually attached to the chromosomes at their ends, one
chromosome lying above and one below the equatorial plane
(pl. 11, figs. 8, 22 and 35) Occasionally, however, they may
be attached at some point along their length. During anaphase,
therefore, the chromosomes have the form of short rods with their
longitudinal axes directed toward the spindle poles (pl. zr, fig.
24), or less frequently of bent rods with the loop similarly di-
rected (pl. zr, fig. 25) There are some minor differences between
the form of the chromosomes in the second pollen- and embryo-
sac-mother-cell divisions. In the latter the rods are longer and
more slender ( rz, fig. 24); in the former shorter and thicker
(A^. 11, fig. 9( On account of this fact and the further one that
in Crucianella gilanica the spindles are very small in the pollen
cells, the character of the structures is rather difficult to deter-
mine. In Crucianella macrostachya, in the anther locules of which
but one row of pollen mother-cells is present, the opportunities

T Zellbildung und Zelltheilung. 3 Aufl., 367.

TETRAD AND EMBRYO-saCc MITOSES 81

for observations. are better, and here the chromosomes are seen
distinctly as paired rods, with the fibers for the most part attached
to their end (pl. zz, figs. 35 and 36). :

The contrast offered between the appearances seen during the
first and second divisions are well seen in ۸ zz, figs. 16 or 17,
and 24, all of which have been taken from the same ovule. The
spindles shown in figs. 76 and 24, pl. 11, may indeed be seen in
the same field in adjacent megaspore mother-cells.

According to the interpretation here given of the behavior of
the chromosomes the application of the facts to the Weissmannian
view of reduction is made impossible.

That, however, a real reduction in the quantity of chromatin
occurs as well as a reduction in the number of chromosomes has
been held by Guignard, according to whom the second division
which follows on the first without pause reduces the nuclein which
the grand-daughter nuclei receive by a half, compared with the
amount which the ordinary nuclei of the sporophyte have. Such
reduction, however, results in equivalent cells as regards their he-
reditary qualities.

A reduction in this sense certainly occurs in the plants under
consideration, if we may argue from the relative sizes of the nuclei.
The figures which present the facts, namely, pl. zz, figs. 6 and
73, are drawn to the same scale and show that the nuclei at the
end of the second division are much smaller than those at the end
of the first. True, the cells themselves are also smaller and this
indicates that the cause of such quantitative reduction is to be
found in the rapidity with which the divisions follow each other,
thus preventing an intervening period of growth.

THE THIRD Division

The resting stage at the close of the second division is com-
plete, a nucleolus is formed, and a considerable pause intervenes
before the third division takes place.

It has been already pointed out on another page that in the
genus Crucianella the megaspores are unique in that they are all
functional in so far that they frequently undergo the first embryo-
sac division. All the divisions which take place in this way are
alike in character, that is they are typical mitoses. The further

82 COMPARATIVE EMBRYOLOGY OF THE RUBIACEAE

divisions of the supernumerary megaspores have not been ob-
served in any other genus, and in all probability do not take place.
It would appear from comparative study that the large size of the
megaspores and the absence of compression by the surrounding
tissues offer favorable conditions for the access of food materials
and the equally rapid development of all the daughter and grand-
daughter cells, thus giving them an equal start as embryo-sac
cells. The embryo-sac cell which finds itself in the most favorable
position as regards the micropylar canal then commences a more
rapid development, which takes place at the expense of the remain-
ing megaspores. In the prophases of the third and succeeding
divisions the chromosomes are about twice as long as those in the
second division, the fibers are usually attached at about their
middle point, where they are bent at a slight angle; the angle is
directed toward the center of the spindle (pl. rz, figs. 27, 30).
This division conforms, therefore, to the ordinary vegetative or
typical division, differing only in the possession of the reduced
number of chromosomes, which have been determined to be in
Crucianella ten. It is not necessary for our present purpose to
describe the remaining divisions in detail, more than to say that
they fully correspond in character to the third.
The embryo-sac divisions in Asperula were not studied.

SUMMARY AND CONCLUSIONS

1. During the growth of the embryo-sac-mother-cell before the
first division a large number of coarse fibers are present in the cyto-
plasm which persist through diakinesis of the first division. These
are, in part at least, continuous with the circumnuclear complex of
fibers preceding the formation of the spindle, and do not result from
a rearrangement of the cytoplasmic reticulum. Further it is sug-
gested that they be interpreted as currents of kinoplasm (hyalo-
plasm) corresponding to the radiations described by Wilson as oc-
curring in the sea-urchin egg.

2. The spindle in both pollen- and embryo-sac-mother-cell is
of multipolar origin. The nucleus is at first surrounded by a felt-
work of curved tangentially placed fibers which gradually segre-

gate into conical groups. By fusion of these groups a bipolar
spindle is formed.

TETRAD AND EMBRYO-sAC MITOSES 83

3. Centrosomes or centrosome-like bodies are not present.
The poles of the spindle are finely tapering points with various
topographic relations described in the body of the text. The sug-
gestion made by Meves, prompted by his observations on Paludina,
is not pertinent.

Under certain conditions, as for example in the pollen-mother-
cell, where the proportions of the cell are not too great relatively
to the size of the nucleus, the spindle ends find an insertion in the
ectoplasm. At the point of insertion a small thickening may fre-
quently be seen, and from this point run out radially placed,
branching fibers in the ectoplasm. It is suggested that these are
currents in the ectoplasm and may represent kinoplasm passing
from the spindle into the ectoplasm during anaphase.

4. At the end of the first and second embryo-sac-mother-cell
divisions, and at the end of the third division also when such oc-
curs, as is generally the case in Crucianella, a system of connecting
fibers arises without, however, the formation of cell walls. The
megaspores and their derivatives in Crucianella remain in a syn-
cytial condition.

5. A reduced number of chromosomes appears in the prophase
of the first division of the pollen- and embryo-sac-mother-cells ;
in Crucianella this number is ten; in Asperula twelve. These
numbers are maintained 7 the subsequent ۰

The prophase of this division is characterized by the stage
known as diakinesis. The metaphase offers on the whole the
same characters as have been described by a number of authors for
various plants and differs from them only in points of detail refer-
able to the proportions of the chromosomes. The frequently dif-
ficult observation of the so-called ring form of the chromosome
pairs in metaphase isa case in point. When the chromosomes are
relatively more slender, as in the Liliaceae, the appearances are
relatively more distinct. In many other plants they are less so
on account of the extreme shortness and thickness of the chro-
mosomes.

The second splitting of the chromosomes is longitudinal, and
takes place in Asperula during the late anaphase of the first di-
Vision. In Crucianella a similar splitting of the chromosomes
takes place in the telophase of the first division and is here also

longitudinal.

84 CoMPARATIVE EMBRYOLOGY OF THE RUBIACEAE

The first division in these plants is therefore heterotypic, in the
sense of Flemming and Strasburger, and the second is homotypic,
and results merely in the separation of the granddaughter rods al-
ready formed. The individuality of these bodies is, however, lost,
so far as can be seen, in the resting nucleus preparatory to the
second mitosis by fusion of the rods and by more or less com-
plete diffusion of the chromatin.

It is concluded from this evidence that the divisions described
are true tetrad divisions, and the four resultant cells as spores. It
has on several occasions been urged that the position of these cells
in the nucellus of angiosperms speaks against this interpretation,
but it is not surprising that such an arrangement should occur in
the ovule when we remember that the position of the spindle and
through the spindle often of the cell wall is determined by growth,
and that the growth of the sporogenous tissues is conditioned by
a very different sort of environment from that found in the sporan-
gium of the pteridophyta.

But the evidence is fast accumulating to slıow that. it really
frequently happens that the divisions are not such as to produce a
single row of cells. This evidence has recently been collated by
Max Koernicke * to which may be added the cases recorded by
G. M. Holferty t as occurring in Potamogeton.

Adding to the above the facts derived from a cytological study
of the divisions themselves, I fail to see the cogency of the state-
ment made by Atkinson f that the embryo-sac of the angiosperms
“arises directly from the nucellar (sporangial) tissues or from the
archesporium, without the intervention of spores." That the
spores in many forms are losing or have already lost their physio-
logical identity as spores is indeed true, but it is certainly a most

remarkable fact that cannot be brushed aside that in the most ex-

treme cases Lilium for example, the process of giving rise to these,
cells is identical in all essential respects, to the process in all other
plants, both pteridophyta and spermatophyta, in which this point
has received careful study.

* Studien an Embryo sac Mutterzellen. Er)
T Ovule and Embryo of Potamogeton natans. t. Gaz. 13: 339. ۰
: 1 s ie Homologies and probable Origin of the Embryo-sac. Science, 1. 13:

TETRAD AND EMBRYO-sac MITOSES 85

On similar grounds the failure of the megaspores to form cell
walls might be urged against the interpretation of these cells as
such, but here again this character has dropped out without, it
seems to me, invalidating the view that ‘the cells in question are
spores which are thereby removed one step further from their more
primitive condition in which a cell wall was eminently necessary.

To be sure we are not rid of the conclusion that, in those plants
in which the embryo-sac arises directly from the mother-cell the
embryö-sac is constructed, so to speak, of four spores ; or that,
when the mother-cell suffers one division giving rise to two
daughter-cells of which one only is functional, the embryo-sac is
constructed of two spores. But after all, spores, in the sense here
meant, are equivalent vegetative cells of a somewhat special sort,
but with no necessarily separate existence, teleologically speaking,

except in those plants in which the spores are the functional equiv-
alent of the seeds in the higher plants.

The conclusion is therefore drawn that the cells under discus-
sion are distinguishable by their origin, that they are not cut out
of the cycle of development and that their morphological. persist-
ence is connected with the function of the mitoses; while on the
other hand, the spore “as such * * * is wanting" * because the
particular features of specialization which characterize spores have
been lost in connection with the loss of the more primitive spore
function.

The following question is very naturally raised at this point,
namely, at what point in the life cycle does the gametophyte com-
mence its existence. There can of course be no categorical, or
at least no final answer given under our present understanding,
But we may state as fairly in accord with the facts, that the mor-
phological identity of the gametophyte has been reduced along with
the change in the function of the spores, and the two periods of
development have been merged into each other, to different degrees
in different plants. In the most extreme cases, as Zilum, the
gametophyte appears to commence histologically with the embryo-

* Atkinson, Z. c. The italics are his.

t Coulter and Chamberlain in their recently published volume ** Morphology of the
Spermatophytes ’’ (New York, 1901), take a tentative position. In their ene
**close the history of the sporophyte with the appearance of the spore mother-cell,
Since it seemed to them “* to be the best defined line of demarkation ” (page iv).

6 COMPARATIVE EMBRYOLOGY OF THE RUBIACEAE

‚sac-mother-cell. Against the adoption of such an interpretation it
may be urged that, where each of four spores form the primordium
potentially or actually of an embryo-sac, the gametophyte becomes
spilt up into four individuals. To be sure there are analogies to
come to our aid, such as the constant development of four embryos
from a single proembryo or the much more widely distributed
phenomena of vegetative reproduction.

But in spite of these analogies, it would seem more natural to
regard the gametophyte as an individual by coalescence, having
its origin in four like vegetative cells whose primitive function has
been lost. These four cells become grafted on each other, so to
speak. The chief difficulty seems to be due to our formal concep-
tional idea of an individual, which in nature, outside of our con-
sciousness, does not exist, :

In comparison with the cases at present before us we may re-
call the very important results obtained by Juel * and by Murbeck.t

Juel found that in Antennaria divica a tetrad division takes
place, and his account of the process differs in no essential detail
from mine. [m Antennaria alpina, however, in which partheno-
genesis is now known to obtain, the tetrad division does not occur,
and the embryo-sac-mother-cell develops directly into the embryo-
sac without any reduction in the number of the chromosomes or
the usual chromosome-ring formation. Nor is this process to be
compared with that in some plants in which megaspores are not
delimited in the usual manner, but in which parthenogenesis does
not occur (Ziium). In such plants the first two divisions are the
same in character as the two divisions occurring when four mega-
spores are formed.

In Alchemilla, as shown by Murbeck, parthenogenesis also oc-
curs, and in such event the megaspore mother-cell divides some-
what irregularly, forming at most three cells, because of the failure
of one of the daughter-cells to divide. These divisions are not
different in character from ordinary divisions ; or, in other words,
the heterotypic and homotypic divisions fail, and the chromosome
number remains constant.

* Vergleichende Untersuchungen über typische u und و سس‎ he Fortpflan-
zung bei der Gattung Antennaria. K. Sv. Vet. Akad. Handl. 33: No. 5. 1900.

1 Parthenogenetische Embryobildung in der Gattung pente dd Acta Reg. Soc.
Physiogr. Lund. 11: No. 7. 1901.

TETRAD AND EMBRYO-sAC MITOSES 87

In the light of the discovery of the fact that true tetrad mitoses
do not occur in parthenogenetically reproduced plants * we are
more than ever forced to insist upon the conclusions stated pre-
viously, that, in such plants as certain Liliaceae, the condition
must not be interpreted as a suspension of spore formation, but as
a reduction in the ontogeny of the gametophyte correlated with a
loss of the spore function, without a loss of the morphological |
character of the spores so far as their development, as indicated
by their mitoses, is concerned. The failure of true tetrad divi-
sions in Alchemil'a and Antennaria is a notable exception which
proves the rule. :

The peculiarities characteristic of the tetrad divisions are now
known to occur in so many plants that we may well pause to esti-
mate their morphological significance, apart from the relation they
hold to the maintenance of a constant number of chromosomes.

In general acceptation the term archesporium has signified a
cell or cell aggregate from which the spore mother-cells are de-
veloped, as first defined by Goebel. t .

These cells cannot, however, be distinguished from vegetative
cells except by the fact that they are richer in food and plasmic
content—they have the character of embryonic cells. According
to their period of growth and the number of divisions they undergo
their derivatives, the spore mother-cells (taken collectiveiy the
sporogenous tissue), are few or many. Now the cells of the spor-
ogenous tissue are different from all other cells in the life cycle
in that they always suffer tetrad division, and thereby exhibit very
constant morphological characters which raise them to special
morphological rank. This cell, or tissue, as the case may be, by
virtue of the special character attaching to it, demands special
recognition in terminology, and it would seem both natural and
justifiable to regard it as constituting the archesporium. The use
of the term in this way seems to me simple and unmistakable.

6. Thethird and following divisions are typical, and the reduced
number of chromosomes is preserved.

In Crucianella all of the four megaspores in a given series may

* Juel (7. c.) interprets the phenomenon in Antennaria alpina as a special case
of apogamy.
f Bot. Zeit. 38: 545, 56%. 1880; 39: 681, 697, 713. 1881.

38 COMPARATIVE EMBRYOLOGY OF THE RUBIACEAE

and often do develop further, and undergo a simultaneous division
(the first embryo-sac division), from which the actual physiological
and morphological equivalence is inferred.

THE BEHAVIOR OF THE POLLEN TUBE IN
DIODIA AND RICHARDSONIA

(PLATE 15)

While engaged in the foregoing study of the embryology of
certain Rubiaceae, namely, Diodia teres, Diodia Virginiana and
Richardsonia pilosa, some conditions of rather more than usual
interest attaching to the behavior of the pollen tube in these forms
were met with. In view of the attention which the subject has of
late received in some quarters—attention which has not by any
means been barren of important results—it seems well at this
time to present the outcome of acloser study of the phenomena in
the above-mentioned plants, although the general subject in the
Rubiaceae promises as a reward of future investigation further re-
sults of interest.

The flower in both genera presents the same chief features in
common. The differences are chargeable to the different numer-
ical relations in them, Diodia having a bicarpellary and Richard-
sonia a tricarpellary ovary. This fact having been mentioned,
we shall from this time on ignore it, since it has no bearing in
particular upon the following description and for the sake of sim-
plicity we shall take Diodia as the type.

MORPHOLOGICAL RELATIONS OF THE STYLE

The locules of the inferior ovary are separated by a partition

of two portions, distinct in origin and in anatomical structure.
The lower portion, or basal element, arises as a ridge from the
floor of the ovary, and inserted on its margin are the ovules, one
in each locule. This basal element reaches to somewhat less than
one-half the depth of the ovary. The remaining distance is oc-
cupied by a partition—the roof element—derived from the styles
which growing out laterally from the upper part of the hollowed-
out receptacles join at the middle point and grow upwards and down-
wards to form the style. This consists therefore of two ( Diodia) or
three ( Richardsonia) concrescent elements. The roof and floor

BEHAVIOR OF THE PoLLEN TuBE 89

elements of the ovarial partition.fuse at their point of meeting and
so complete the separation of the originally single ovarial chamber
into locules.

HISTOLOGY

When the parts above described are young, the cells are, of
course, uniformly isodiametric. After all the primordia are laid
down, and the period of rapid growth prior to the assumption of
the definitive form sets in, the relative rapidity of growth of the
several organs is accompanied by changes in the proportions of the
component cells. At about the time of fertilization the histolog-
ical features are as follows. i

The basal element of the ovarial partition consist of thin-walled
parenchyma whose elements are slightly elongated in the direc-
tion of the flower axis and contain the vascular tissue which sup-
plies the ovules. Where the stylar portion of the partition abuts
against the basal elements, the cells of the latter are flattened out
transversely, this being due partly to the pressure and partly to the
mode of growth. The epidermis of the basal element of the par-
tition has, at this point, very much thickened walls. The thicken-
ing commences in the outer walls (fig. 7) and extends at the
period of fertilization to the lateral, but not to the basal walls
(fig. 5). At the same time the regularity of the layer of cells is
lost in the process of fusion which takes place between the basal
and the stylar elements of the partition.

The stylar portion of the partition is composed at the time of
fertilization of three distinct tissues. The innermost of these, or
that lying in the axis of the compound style, and that with which
we are here concerned constitutes a conductive tissue and is com-
posed of elongated narrow cells with transverse or slightly oblique
end walls and no intercellular spaces. These cells may be further
distinguished from the surrounding tissues by their denser cyto-
plasmic content, and thicker, more deeply staining walls. Capus
(2) has noted that the cells of the conductive tissue in the plants
studied by him have thickened walls (collenchyma), but in Glche-
milla arvensis according to Murbeck (10) no such thickening 1s
present in any part of the course of the conducting tissue, a con-
dition which is found in the free portion of the style of the-plants
under consideration. Here the cells are longer than in the stylar

90 COMPARATIVE EMBRYOLOGY OF THE RUBIACEAE

partition, have thin walls with somewhat thinner cytoplasm, but no
intercellular spaces. The conductive tissue is readily recognized
in the style proper by the smaller transverse diameters of its cells.
The change from this thin-walled tissue to the thick-walled is

quite abrupt and takes place at the level of the roof of the

ovary.

At the lower extremity of the style, at the point namely where
it fuses with the basal partition, the conductive tissue suddenly ex-
pands from being long and narrow, to short, irregular and isodia-
metric. There is also an evident increase in the thickness of the
cell-walls. The line of fusion which may be recognized easily in
the young condition as represented in jig. I is now so irregular
as to be of itself unobservable. This irregularity and confusion of
cells appears to be due to the thickening and partial gelatiniza-
tion of the cell-walls or more especially of the middle lamella,
for one may notice in sections that irregular lines of separation oc-
cur between the cells, thus indicating their loose relations. ' Such
changes have been noted, also, by Capus (2) although according
to him such changes are much more pronounced in the style and
stigma So far as the writer's observation goes such pronounced
changes do not occur in the upper portion of the style, although
here the walls are a little thickened, and that they are on the
more and more pronounced the deeper one penetrates the whole
style.

At this point of fusion, as described in the previous paragraph,
the ovules, two in Diodia, three in Richardsonia, are inserted upon
the edge of the basal partition. The epidermis above described
a undergoing a thickening of its walls is continuous with the
epidermis of the ovules, a description of which is necessary at this
point,

The ovule is, broadly speaking, anatropous, and possesses, in
addition to the single integument characteristic of the gamopetalae,
a second outgrowth, the origin of which has been elsewhere de-
scribed (p. 5 3) asa secondary enlargement of the funicle, a sort of
strophiole. A profile view of this organ, as seen in Diodia teres, i$
Xt fig. 1, and as seen in Rickardsonía in jig. 9.
ure of this organ with which we are at present concerned
und modification of the epidermis to form a special con-

striking feat
is the profo

"

BEHAVIOR OF THE POLLEN TUBE 91

ductive tissue for the pollen tubes. When young the cells of this
tissue are of an approximately cubical or columnar form ( fig. 7)
with thin walls. As the ovule develops they change into deep
columnar cells with thick walls excepting at their inner ends ( fig.
5), where the basal and lateral walls for a little distance are thin.
The contents of these cells are more dense than those of the deeper
lying cells and the. cytoplasm is very finely granular. Their nuclei
are large with very finely divided chromatin. These cells appear
to correspond in general with the cells
described by Capus (2) as occurring at
the base of the funicle in Dracaena ele-
gans and by Dalmer (3) in the placenta
of Mahonia aquifolium. From. those in
Dracaena, however, the cells in Diodia
teres and Richardsonta differ in having
thick walls, and from those in Mahonia
in the fact that they do not secrete the
mucilage which, according to Dalmer,
serves as food for the pollen tube. They
are not glandular cells, but, strictly
speaking, cells of a conductive tissue as -
is shown by the behavior of the pollen
tube towards it. With reference to the
strophiole it may be remarked that this
organ is present in other members of the
same family, but that, so far as known,
in no other is such a special tissue pres-
ent, excepting as we shall see beyond |
in Diodia Virginiana. On the upper side Fis. 8. ies ge nd
ofthe funicle these cells have somewhat i pd F e bé whe =
different dimensions from, but otherwise

are exactly like those of the lower side where they have the longest
ongitudinal diameter (tert figs. 8 and 9). Here the strophiole
lies against the ovules so as completely to close the micropyle.
It will be seen that there is thus formed a collar of columnar epi-
dermis which is continuous around both sides of the funicle and,
as will be shown, completes the path of the conductive tissue
from the style to the micropyle.

92 COMPARATIVE EMBRYOLOGY OF THE RUBIACEAE

THE COURSE OF THE POLLEN TUBE

In the free upper portion of the style the cells of the conduc-
tive tissue are, as above pointed out, long and very narrow.
Through this tissue the pollen tube takes a somewhat devious,
though on the whole direct course ( fig. 2), and as the tube has a
. transverse diameter much greater than that of the neighboring
conductive tissue cells, one would expect that the mechanical
pressure alone would cause not a little disturbance in the conduc-
tive tissue. Such, however is not the case. The pollen tube does
not destroy the cells among which it passes either by chemical in-
fluences, such as digestion, or mechanically. It is true that the
cells are frequently disturbed in position, as is plentifully evidenced
by the form of the nuclei when the pollen tube happens to exert
pressure on a cell in such a way as to affect its nucleus (fig. 3).
Without exception, however, the nuclei and cytoplasm stain per-
fectly normally and evenly. The walls of the pollen tubes are to
a considerable degrees thicker than those of conducting cells,
while the cytoplasm of the former is more coarsely granular than
that of the latter. |

When the pollen tube reaches the stylar partition it meets with
conductive cells with thicker walls and of larger transverse dimen- '
sions. The nuclei are here long-oval in form. . Here the path has
about the same character as in the free part of the style. Occa-
sionally one finds a sharp turn such as is shown in fg. 3, which
represents a small part of the conductive tissue in longitudinal
section showing a pollen tube cut transversely. A study of this
preparation shows clearly that the effect of the pollen tube upon
the tissue in this part of its course through which it passes is
mechanical. Instances are frequently found in which the pollen
tube passes athwart the conductive tissue, either directly or more
or less obliquely. At first glance it is somewhat difficult to at-
tribute such behavior on the part of the pollen tube to mechan-
ical influences. When, however, we reflect upon the extreme
sensitiveness of the protoplasm of the pollen tube to stimuli,
it becomes less difficult. The impingement of the end of the pol-
len tube upon the end wall of a cell is, we believe, enough to turn
it out of its direct path. Upon approaching the point of fusion of
the basal and roof elements of the ovary septum the pollen tube

BEHAVIOR OF THE POLLEN TUBE 93

meets with cells of similar histological characters as those of the
cells through which it has immediately passed, save in dimensions
and in the mucilaginous condition of the inner lamella. Here
they are irregular and isodiametric as above stated, and are derived
in part from each element of the septum. Through this mass of
tissue the pollen tube takes a more irregular course, showing
some tendency to branch, and passes from the stylar tissue into
the upper part of the basal septum, into, therefore, the epidermis
above described as continuous with the columnar epidermis of the

- 9. Ovule and part of the stylar tissue of Richardsonia pilosa, showing a
of E ی‎ tubes.

strophiole. Upon meeting with the ordinary thin-walled paren-
chyma fof this, the tube abruptly turns at right angles to its
former course (text fig. 9). In no case does the tube penetrate
beyond the thick-walled fusion cells further than to basal septum.
The contents of the pollen tube may be seen well in the lower
region of the style. In properly stained preparations the genera-
tive nuclei have been repeatedly found. The vegetative nucleus,
however, has not been seen in this region. The cytoplasm is
. densely and finely granular, entirely filling the tube, while in the
older parts of. the same only a thin granular layer may be dis-
cerned on the wall, or none at all.

*

94 CoMPARATIVE EMBRYOLOGY .OF THE RUBIACEAE

In this part of its course the pollen tubes which are usually to
be found in numbers produce a destructive effect upon the cells,
upon which they impinge. The effect has the appearance of be-
ing due to mechanical pressure owing in part to the number of
tubes present, and in part to their larger diameter. The contents
of the conductive cells thus affected appear homogeneous and
deeply staining and the nuclei lose their normal appearance, while
the cells are crushed out of shape (fig. 4). The effect of the tube
is unequal, destroying cells, now here, now there, without any dis-
solution of the walls, showing at once the difference between a
mechanical operation, and one involving the histolysis of the cir-
cumjacent tissue as by digestion.

The behavior of the pollen tubes when they reach the sub-
epidermal parenchyma of the basal partition is quite uniform, and
is strongly suggestive of a purely mechanical influence. In the
case of Richardsonia there is formed at that point by virtue of the
confluence of the funicles of the three ovules, a basin so to speak,
filled with the isodiametric irregular conductive cells. Should we
plant a seedling in a basin of soil the roots would penetrate till the
walls of the basin had been reached, and they would then glide
along the walls. Such a case appears to be quite analogous to
the one before us, excepting that geotropism plays no part in the
behavior of the pollen tube. In Diodia teres the penetration of
the epidermis of the basal partition does not take place as in
Richardsonia ; but rather the tubes bend upward in the irregular
cells of the stylar conductive tissue, passing at first between the
walls of the style and strophiole, but soon penetrating the columnar
epidermis of the latter ) fig. 8).

Having now reached the thick-walled epidermis, the pollen
tube travels along till it enters the regular thick-walled columnar
cells which constitute the collar of connective tissue about the
funicle and the strophiole. It then passes out of the fusion tissue
and travels in a fairly direct path upwards, that is toward the
chalazal region of the ovule (text figs. 8 and 9). It gradually
toms, however, so as to pass around one side or the other of the
TA us m finds itself in the epidermis of the lower side of
de ae 9 t this time the course is more or less irregular, and

sometimes branches te 0 Occasionally in the

BEHAVIOR OF THE POLLEN TUBE 95.

region of the micropyle it passes beyond the surface of the epi-
thelium (£g. 6). It then glides along the surface a short distance
and then passes between the conductive epidermis and the integu-
ment (text fig. 1). Ultimately it finds the micropyle and travels
along it to the embryo-sac.

Two points of importance are to be noted at this point, the en-
tirely or almost entirely intercellular path of the pollen tube, and
the direction of its path which is not in the direction of least re-
sistance, but at right angles to the longtudinal axis of the cells of
the conducting tissue. |

During this part of its course, that is, as it is passing through
- the fusion tissue and the strophiolar conductive tissue, the pollen
tube as above pointed out not. infrequently branches (fg. 7).
Nawaschin has noted a similar tendency on the part of pollen
tubes in the Birch (11, p. 22), to branch “wo das betreffende
gewebe keine deutliche Anordnung seiner Zellen in longitudinalen
Reihen aufweist.” When in the strophiolar conductive tissue, the
pollen tube shows more or less inability to pursue a path which is
constantly at right angles to the component cells. It may then
slip out to the surface. This happens most frequently in Richard-
sonia in the neighborhood of the micropyle where there is little or
no pressure exerted upon the strophiole by the adjacent structures.
In Diodia teres where the strophiole in the micropylar region is
tightly wedged between the ovule and the wall of the ovary, the
pollen tubes take constantly a deeper course (text fig. 5). These
phenomena lead to the suggestion that the cause which has brought
about the habit of intercellular growth in the pollen tube is a me-
chanical one, and this factor may probably be called upon to ac-
count for the behavior of the pollen tube in many forms. We
have in mind such cases as occur, for example, in the Cannabinaceae
of which Zinger (23) has given an account. According to this
author, the intercellular node of growth is of phylogenetic signifi-
cance, and, by means of such interpretation, brings his view into
line with that of Nawaschin, who has vigorously supported the
contention that the behavior of the pollen tube in the Amentiferae
is inherited from their ancestral forms, and has endeavored, on this
theory, to connect the lower Angiosperms with the Gymnosperms
A discussion of this question is not pertinent here, but it may be

96 COMPARATIVE EMBRYOLOGY OF THE RUBIACEAE

said in passing that recently Murbeck (10) has, we believe, suc-
cessfully, contested the‘view of Nawaschin. I believe that a study
of the facts in a larger number of plants will lead to a more gen-
eral recognition of the principle laid down by Murbeck (10) in the
words, “ Der intercellulare Wachstumsmodus des Pollenschlauches

istal itzubezeichnen, * * e

and of the mechanical conditions as constituting not the least im-
portant factors in leading to such behavior. As a further illustra-
tion of peculiar mechanical conditions 1 would cite the instance
of Cynomorium in which, as shown by Pirotta and Longo (18), the
micropyle is completely closed and there appears in its place a cone
of conductive tissue through which the pollen tube must pene-
trate. Zinger (21) has shown somewhat similar relations in Can-
nabis.

When, as shown above, the pollen tube passes to the surface,
it travels a short distance along the conductive tissue, soon finding
its way between the strophiole and the integument. Under such
circumstances the wall of the pollen tube may well be seen ( fig. 6):

The effect of the pollen tubes upon the cells of the collar of
conductive tissue appears to be purely mechanical, as in the case
of that in the style. Forcing its way between the cells, the pollen
tube causes in them but little or no distortion and, so far as I have
observed, no destruction (figs. 5, 6). It would appear that,
although the cells of the conductive tissue form a true tissue, and
are not crowded papillae, their thickened walls are nevertheless
not bound together firmly but rather are loosely related by a soft,
mucilaginous middle lamella.

Having reached the micropyle, the pollen tube now passes
into the micropylar canal. The cells lining the latter are char-
acterized by somewhat thicker, more deeply staining walls than
those of the other integumental cells. The pollen tube eventually
reaches the synergids. In the plants under discussion these cells
have the “ hyaline, striated tips," as described by Strasburger (21),
extremely well developed. The generally accepted view that the
synergids serve as a conductive apparatus is strengthened by the
analogy which 1 venture to believe exists between their curious
cellulose tips and the thickened and partially gelatinized cell walls
of the conductive tissue.

%

e

BEHAVIOR OF THE POLLEN TUBE 97

4 Diodia Virginiana

In this species the conductive tissue is of a somewhat different
character from that in the two plants already discussed. In the
style both above and below the roof of the ovary the cells are
cylindrical, their transverse walls being transverse, though oc-
casionally oblique in varying degrees. The nuclei are not so much
elongated, and correspondingly the length of the cells is not so
great. The cell walls are thin throughout the whole range of
stylar conductive tissue, excepting near the region of fusion of the
basal and stylar parts of the partition where a slight thickening of
the walls occurs. The conformation of the stylar conductive tis-
sues is the same as in the other species.

In the character of the conductive tissue of the ovule, how-
ever, lies the greatest difference, a difference which, as we shall
see, is connected with the behavior of the pollen tube. The tissue
is disposed on the strophiole in a band around the insertion of the
funicle, and extending downwards towards the micropylar region.
The accompanying text figure 10 shows in a schematic way

the arrangement. The component epithelial cells are approx-

imately cubical in form (fig. 8), are fairly richly supplied with
cytoplasm, and have a large nucleus. The outer walls are thick-
ened, and exactly correspond in all respects to the homologous
tissue in the other species studied in the young condition, before
any further changes take place, 2. e., before the cell elongates into
columnar form. At the time of pollination, a mucilaginous secre-
tion appears so that at this time there lies on the surface of the
conductive tissue a layer of mucilage of a thickness of nearly half
the depth of the cells. The mucilage may be traced between the
cells to some depth.

This conductive tissue corresponds in general to that which
has been described by Dalmer (3) as occurring on the placenta in
certain Liliaceae both in form and in the occurrence of the muci-
laginous secretion through which that author has shown that the
pollen tubes grow, and from which they derive nutrition which
enables them to grow onward toward the micropyle. Apparently
similar cells have been described by Lang (5) in Byblis and Poly-
pompholyx.

98 COMPARATIVE EMBRYOLOGY OF THE RUBIACEAE

The course of the pollen tube through the style needs no re-
mark since what has been said above applies throughout. One
point, however, needs mention, namely, that here the cellulose
plugs described by Treub (22) as occurring in the pollen tubes in
Casuarina, by Nawaschin (12, 16) in the Amentiferae and by
Murbeck (10) in Alchemilla are here very easily and clearly seen-
Two of these plugs are shown in figs. ro and rr. It will be seen
from the outline of the plug and the form of the somewhat

. shrunken pollen tube contents that these fit each other. One finds
sometimes that the tube on one side of the plug is empty ( £g. 70).
In other cases cytoplasm occurs in both sides (fig. rr). Similar
plugs occur lower down in the fusion tissue, but they have not
been observed elsewhere.

: When the pollen tube reaches
the epidermis of the basal partition

(text fig. ro), it stops at that point,

and does not enter between the

cells as occurs in 0۰

They then bend upward and travel

between the two tissue masses until
they come to the collar of con-
ductive epidermis of the strophiole,
in the mucilaginous outer wall of

: which they pursue their way. Af-

FiG. 10. Two views of the ovule ter moving upward, 2 e, toward

of Diodia Virginiana, showing the the chalaza, a little distance, they

conductive epidermis (dotted) and the turn, following the band of se-
path of the pollen tube; و‎ funicle. . & : } :

( Schematic. ) creting epidermis as shown in

text fig. 10, and being com-
pletely buried in the mucilage, as occurs in plants studied by

Dalmer (3). The behavior of the pollen tube toward the walls of

the secreting cells needs particular mention in view of the fact that
the wall of the former coalesces with the outer walls of the latter
in such arway as to be completely indistinguishable. In one pre-
paration the pollen tube was cut transversely at one point of its
course and showed this relation especially well. In fig. 9 this is
well shown, from which it will be seen that the outer part of the
tube wall is thin, thickening markedly on the sides nearer the epi-

A

BEHAVIOR OF THE POLLEN TUBE 99

dermis, and, as above said, coalescing with its outer wall. It is not
clear, however, whether the pollen tube wall as represented, and
as it appears to the eye in the preparation belongs, properly

speaking, entirely to the pollen tube. At any rate the soft char-

acter of the walls and their coalescence makes the determination of
this point difficult.

When the pollen tube reaches the
lower end of the strophiole, it travels, as
before, superficially around and into the
rift between this and the integument. This
rift is filled with the slimy secretion de-
rived entirely from the epidermis of the
strophiole, inasmuch as the epidermis of
the integument is not secretory. In this
way the micropyle is at last reached and
entered.

If now we compare the three types de-
scribed as regards the development of the
strophiolar conductive tissue, and the be-
havior of the pollen tube, we may accord-
ingly arrange them into a series in the
following order and with the characters
indicated. Fic. 11. Diodia Virgini-

I. Diodia Virginiana.—The conduct- ana. Ovule and portion of the
ive tissue consists of epidermal cubical wes E HA peia ase
cells with thickened outer walls covered ” ^ °
with a mucilage secreted by them. The mechanical relations of the
ovary are'such that the pollen tubes may easily lie between the
strophiole and the ovary wall. The pollen tube travels super-
ficially along the funiculus and strophiolar conductive tissue, but
imbedded in the mucilaginous secretion.

2. Diodia teres.—The conductive tissues consists of columnar
epithelial cells with thickened outer and lateral walls. The middle
lamella is mucilaginous. There is no'secretion on the outer sur-
face. The mechanical relations are such that the strophiolar con-
ductive tissue is compressed between the ovule and ovary wall.
The pollen tube enters the said conductive tissue immediately on
commencing its upward movement along the funicle and remains
within till it reaches the micropyle. *

100 COMPARATIVE EMBRYOLOGY OF THE RUBIACEAE

3. Richardsonia pilosa—The conductive tissue as in Diodia
teres, but somewhat different histological relations exist in the fusion
tissue of the ovary septum. The mechanical relations are such
that the strophiolar conductive tissue above the funicle is com- `
pressed between the ovary wall and the ovule. Below, 2. e., toward
the micropyle, the pressure is reduced. The pollen tube enters
the conductive tissue at the base of the stylar portion of the parti-
tion, travels by the intercellular method in the same till it reaches
the micropylar region where it often though not always bends to-
wards the surface.

The above-named facts speak for the mechanical conditions in
the ovary as important factors in determining the course of the
pollen tube, whether intercellular or not.

In review the following results and conclusions ate presented :

I. In Richardsonia pilosa and Diodia teres the course of the
pollen-tube is chiefly intercellular, a phenomenon clearly secon-
dary in view of the fact that in a closely related species, Diodia
Virginiana, the pollen tube pursues its path freely in the “ovary
cavity in that part of its course.

An important factor in determining such behavior of the pollen
tube is found in the mechanical relations (pressure) existing be-
tween the ovule and the ovary wall. This is added as evidence
supporting the contention of Murbeck to the effect that the inter-
cellular mode of growth of the pollen tube is a matter of physio-
logical meaning.

Inasmuch as such a difference in the behavior of the pollen-
tube as has been pointed out may occur between two closely re-
lated species of the same genus, I infer, contrary to the view of
Nawaschin, that the behavior of the pollen tube is of very limited
significance—indeed of no practical significance—from a phylo-
genetic viewpoint. The condition is analogous to that of the
archesporium in the angiosperms, which occurs, so far as at
present known, in eight genera of the Rosaceae and in a dozen
others widely scattered through the whole range of forms from
the Rubiaceae (5, 6) to Casuarina.*

This statement does not preclude the possibility that Nawas-

> AON ee
A A summary of the facts and literature pertaining thereto will be found in paper
y Murbeck and Koernicke already quoted,

BEHAVIOR OF THE POLLEN TUBE 101

chin's contention, that the primitive behavior of the pollen-tube
was entirely intercellular, is true.

2. The conductive tissue of the style consists of: (I) very thin-
walled cylindrical cells in the upper part and (2) thicker walled
cells in the lower part. The latter pass from a cylindrical to an
irregular isodiametric form in the fusion tissue, in which the
middle lamella is evidently soft and yielding.

3. A conductive tissue is here described which is morpholog-
ically a part of the ovule, or, more exactly, a part of a peculiar
secondary outgrowth which I have designated the strophiole. It
takes two histological forms : (1) In Diodia Virginiana it is com-
posed of cubical, mucilage secreting cells, with thickened outer
walls, corresponding to analogous cells found in many plants on
the placenta. (2) In Diodia teres and Richardsonia pilosa the
homologous tissue is composed of columnar, thick-walled cells
(the thickening being absent from their bases only) with a soft
middle lamella. No outer mucilage layer is here found.

In these plants we find a conductive tissue composed of elon-
gate, regular cells, in which their longitudinal axes are placed at
right angles to the intercellular path of the pollen tube for the
direct passage of which this condition is mechanically less favor-
able than in the kind of conductive tissue in which the longitudinal
axes of the component cells are parallel to the path of the pollen
tube.

4. The pollen tube does not in general act unfavorably upon
the cells with which it comes into contact. When however it
does it may be referred to mechanical causes. This conclusion
appears to accord with the observations of others.

5. The pollen tube shows a tendency to branch where the
cells are irregular and isodiametric, or where the long axis of the
conductive tissue cells are placed at right angles to the course
of the pollen tube. This behavior speaks against the view of
Miyoshi: “Im Griffel werden die Pollenschlauche wesentlich me-
chanisch zum Fruchtknoten gelenkt” (8).

6. There remains an undetermined factor in the guidance of
the pollen tube. We cannot tell how the same is able to find its
way to the egg in the definite way in which it does. The pollen
tube is negatively aérotropic, and is chemotropic, as shown by

102 COMPARATIVE EMBRYOLOGY OF THE RUBIACEAE

Molisch (9). These facts explain its behavior toward the stigma,
but the later behavior in the conductive tissue remains unexplained.
According to Molisch again, chemotropism “ muss bei der Wan-
derung des pollenschlauches zur Eizelle, resp. bei der Auffindung
derselben in vielen Fallen eine wichtige Rolle zufallen.” But it
is difficult to see how chemotropism can be an efficient factor
unless there be a differential distribution of the stimulant.
Miyoshi on the other hand, has maintained, in the words
above quoted, that the operation is a purely mechanical one.
Nawaschin (16), to whom we are indebted for a critical exposition
of these two views, has pointed out that both presuppose that “ der
Pollenschlauch Reizbewegungen ungehindert ausführen könne,”
a condition which is found only where the pollen tubes have free
space in which to move. If we assume that such are the conditions
in Diodia Virginiana, in which the pollen tube moves freely in the
ovarial cavity, what may we say of Diodia teres and Richardsonia
Pilosa, in which the pollen tube is not free to move in the ovarial
cavity, but pursues an intercellular path, parallel to that which we
may believe, from the case of Diodia Virginiana, it would take were
the mechanical conditions favorable, and this in spite of the char-
acter of the conductive tissue which is mechanically unfavorable.

I believe that we may be justified in maintaining :

t. With Nawaschin, that the occurrence of a secreting con-
ductive tissue speaks as much for the nutrition of the pollen tube
as for the necessity of mechanical guidance ; that, moreover, the
ability of the pollen tube to secrete a cellulose-dissolving enzyme
indicates the former, as in point of fact the pollen tube does
not penetrate the cells of the conductive tissue ; and that the as- ۰
sumption that in order to respond toa stimulus, the pollen tube
must have unoccupied space in which to move is unnecessary and
probably untrue.

2: With Molisch, that chemotropism is the important factor,
and the guidance of the pollen tube; but that the distribution of
the ed in the conductive tissue is a differential one ; and, as
Nawaschin (16) suggests, “ die Verbreitung der Reizstoffe Kann
zwischen den Papillen des Leitgewebes in F olge der capillarkrafte

leicht zu Stande kommen,” extending his statement, however, to
cells as well as papillae,

BIBLIOGRAPHY 103

. The experiments of Miyoshi (8) by which he showed that
pollen tubes will penetrate the cut end of a style and travel toward
the stigma, and that when a lateral cut is made into the conduc-
tive tissue the pollen tubes will grow out in various directions, we
maintain not only do not support his view that the pollen tubes are
guided in a purely mechanical way, but also that they do not
contradict the suggestion of such a differential distribution of an
appropriate stimulant, inasmuch as a cut in the conductive tissue
would in all probability set up disturbances by which the ordinary
distribution of the stimulant might be reversed.

It may, of course, be that Miyoshi's contention is true for the
style, although I believe that the case is not proven ; but that the
pollen tube is guided in one way in one region of the conductive
tissue, and in another way in another, seems quite improbable.

3. The suggestion is submitted that the synergidae are the
source of the stimulant, and constitufe the center of distribution
of the stimulant. To this idea I am led by the cytological char-
acter of those cells, and by their position as the only effective
center, exclusive of the egg itself. We do not forget the experi-
ment of Strasburger (21) on Torenia, the results of which were,
however, negative. Searching experimentation is still needed to
throw light on the perplexing problem which we have attempted
somewhat further to elucidate.

A A List OF LITERATURE PERTAINING TO THE POLLEN TUBE.

1. Benson, M. Contributions to the Embryology of the Amentiferae,
Part I. Trans. Linn. Soc. Bot. 3: 409. 1894.

Capus, G. Anatomie du tissu conducteur. Ann. des sci. nat. Bot.
VI. 7: 209-291. 1878.

Dalmer, M. Ueber die Leitung der Pollenschlauche bei den Angio-
spermen. Jen. Zeitschr. 14: 1-39.

. Koernicke, M. Studien an Embyrosack-Mutterzéllen. Sitzungsber.
d. Niederrhein. Gesellsch. Natur. und Heilkunde zu Bonn. 4
Mar. 1901.

Lang, F. X. Untersuchungen iiber Morphologie, Anatomie und
Samenentwickelung von Polypompholyx und Byblis gigantea. . Flora,
88: 149-206. 1901.

N

v

>

in

104 COMPARATIVE EMBRYOLOGY OF THE RUBIACEAE

6. Longo, B. La mesogamia nella comune Zucca (Curcurbita pepo
Linn.). Rendiconti della R. Accad. dei Lincei 10: 168-172.
3 Mr. 1901,

7. Miyoshi. Ueber Chemotropismus der Pilze. Bot. Zeitung. 52:
1-28. 1894.

8. Ueber Reizbewegungen der Pollenschläuche. Flora, 78: 76-
93- 1894.

Molisch, H. Zur Physiologie des Pollens, mit besonderer Riick-‏ .و
sicht auf die chemotropischen Bewegungen der Pollenschlauche.‏
Sitzungsber. K. Acad. Wiss. Wien. 102: 423-448. Jl. 1893.‏

10. Murbeck, Sy. Ueber das Verhalten des Pollenschlauches bei
Alchemilla arvensis (L.) Scop. und das Wesen der Chalazogamie.
Acta ۶ Physiogr. Lund. 11: no. 2.2.3907,

11. Nawaschin, S. Zur Embryobildung der Birke. Bull. de 1’ Acad.
imper. des:sci. de St. Pet. 13: 345-348. D. 1892.

Ueber die gemeine Birke (Betula alba L.) und die mor-

phologische Deutung der Chalazogamie. 42: no. 12. 1894.

12.

13. RER Kurzer Bericht meiner fortgesetzten Studien über die Em-
bryologie der Betulineen. Ber. deutsch. bot. Gesellsch. 12: 163-
169. 1894.

14 Neue Ergebnisse über die Embryologie der Hazel ( Corylus

Avellana). Bot. Centralb. 43: 104-106. . 1895.

15. — — Ein neues Beispiel der Chalazogamie. Bot. Centralb. 43:
353-357. 1895. -

euer das Verhalten des Pollenschlauches bei der Ulme.

l'Acad. imper. des sci. de St. Pet. V. 8: 345-358. “1898.

1 Zur Entwickelungsgeschichte der Chalazogamen. Corylus

Avellana L. Bull. de l'Acad. imper, des sci. de St. Pet. 10: 99-
115. 1899.

ie Pirotta R.e Longo, B. Osservazione e Ricerche sulle Cynomo-

nsiderazioni sul percorso del Tubo pollinico

"elk apio ene oa Ann. R Te. Dot di Rome 9: 97-

IIS. 1900.

Basigamia, Mesogamia, Acrogamia. Rend. della R. Acc.

sag 8: 296-298. 1899. (Ref. in Bot. Centralb. 86: 93, 94:

20, Strasburger,
Jena. 1879.

16.

Bull. de
17

19.

E. Die Angiospermen und die Gymnospermen.

EXPLANATION OF PLATES 105

21. ———— Ueber Befruchtung und Zelltheilung.

22. Treub, M. Sur les Casuarinées et leur place dans le zen nat-
urel. Ann du Jard. Buit. 10: 145-231. 1891.

23. Zinger, N. Beiträge zur Kentniss der weiblichen Blüten und In-
florescenzen bei Cannabineen. Flora, 85 : 189-253. 1898. (The
author presumes that this paper includes Zinger’s previously stated re-
sults which are recorded as follows: (1) Beitrag zur Entwickelungsge-
schichte des Hanfs. Trav. Soc. Imper. Nat. St. Pet. Livr. 1. Comptes
Rendus des Sciences. 1896. No. 4. (2) Beitrag zur Morpho-
logie der Fam. Cannabineae. /éid. Comptes Rendus. 1896.
No. 7-8.)

Explanation of Plates

'The letters in the following gen signify as follow
ant., antipodal c cell (or nucleus). n. c., nutritive cells i the chalaza.
ar., archespo û. t., pollen-t
€. efi., Eie epidermis. s». srophiolar den of brown cells.
e. s., embryo-sac. 5. £., strophiol :

PLATE 5. Callipeltis cucullaria
Fic. 1. An early stage in the growth of the nucellus. The men is de-
limited and the periclinal divisions giving rise to the integument are to be se
Fıc. 2. Older conditions of the nucellus in which the archesporial cells are ap-
intend maturity.
Fic, 3. A nucellus cut es showing a larger number of archespores than
wade be seen in a radial s

G. 4. A radial ieu section through a mericarp. The formation of
Sun has been consummated and one especially large one may be seen before it
has started to migrate. The micropylar canal is, therefore, intact.

. A group of sace in which one, the lowest one in the figure, has en-
iet ndi is preparing for migra

G. 6. The migrating megaspore may be seen lying in the micropylar tanal, the
E rans of which has been end by the activity of the moving cell.

Fic he final division of the embryo-sac cells to form the eight cells of the
nearly mature condition of that o

Fic. 8. An instance in which two megaspores have migrated and have given rise
to two nn formed embryo-sacs.

G. 9. An embryo-sac after fertilization has taken place. The embryo is in the
vera state and the endosperm nucleus is in its first division. The material de-
rived from the disintegrating megaspore has completely disappeared.

PLATE 6.
Fics, 1-12. Callipeltis cucullaria.

Fic. 1. Ovule in longitudinal section at the time of the migration of the mese
sac-mother-cell, and before its division. The remains of the megaspore mass may sti
e seen

106 COMPARATIVE EMBRYOLOGY OF THE RUBIACEAE

IG. 2. An example in which several megaspores have began migration. In this
figure and in Fig. 1 may be seen the compressed endodermis
- 3. Embryo-sac what fusion of the polar nuclei. The different cells have
now taken on their definite
Fic. 4. Embryo-sac with one-celled embryo. One synergid remains intact.
Fic. 5. A still later condition with 11-celled embryo. Antipodals still intact.
Fics. 6-9. Various embryos in various stages of development.
Fic. 10. Suspensor, showing و‎ lts most extreme ise ی‎ in this species.
FIG. 11. Ovule with end 1 ity (seed a
The — cells which are ‘still wer ما‎ are dawn on he left s

. Section through the pericarp ی‎ cell layers), integument (one cell "D ;
and ie. outer cell layer of the endosperm

Fics, 13-23. Sherardia arvensis.

Fic, 13. Mass of megaspore mother-cells.

Fic, 14. Embryo-sac with the migrating embryo-sac-mother-cell.

Fic. 15. Definitive embryo-sac just previous to fertilization.

Fics. 16, 17. Egg apparatus and fusing polar nuclei.

Fic. 18. Two-celled embryo with endosperm.

Fic. 19. Older maid in dá two regions in the suspensor are evident. The
basal region ve hausto:

Fic. 20. Shows a somewhat more developed condition of the ی‎ proper than
that seen in a

Fic uspensor of nearly mature embryo. The basal part of the haustorium
which Due ی‎ ap in the intact endosperm has degenerated. The haustoria of the
part of the suspensor which lies in the digested endosperm are active.

Fic. 22. Ovule from which Fig. 21 is taken.

Fic. 23. Peripheral portion of endosperm showing outer absorbing cells.

PLATE 7. Galium
Figs. 1, 2. Galium Mollugo. Two successive stages in the development of the
nucellus’
Fic. 3. Galium recurvum. In this — the archesporium has reached maturity
and the i en has started its = wth.
. Galium recurvum

Fic. A mega: spore-mass just before the migration of the
مت‎ -mother-cell.

The ا‎ cells are not yet destroyed.
Fic. 5. The megaspore mother-cell in migration.

Fic. 6. Galium Parisiense.

G. 7. Galium triflorum.

n. 9 in order to show the

The two-celled stage of the embryo-sac.
-sac. The same embryo-sac is shown in
length ۳ tha long antipodal.
G. 8. Galium Aparine. An embryo-sac in which the endosperm has begun to
form a in which the embryo is composed of six cells,
Galium triflorum. Ovule in longitudinal section showing the proportions

FIG. ۰
the ee. and the relative position of the end of the food-conducting tissue
the raphe.

EXPLANATION OF PLATES 107

Fic. 10. Galium tinctorium. The embryo-sac. The constriction of the endo-
sperm salt and the irregular nucleus are to be seen.
Fic. 11. Galium Mollugo. Two embryo-sacs produced in the same ovule.
; Ovule with endosperm and young embryo. The

Fic. 12. Galium Parisiense. An AYSE: The embryo has begun its

: Fic. 15. The lower half of a mericarp of Galium Mollugo. The undigested walls
of the salle lining the micropyle are to be seen at the end of the suspensor. emb, em-
bryo ; end, endosperm ; /, funicle ; 7, integument; 2, pericarp; 5. e, stylar column.

PLATE 8. Asperula

Fic. 1. Asperula azurea. A nucellus in an incipient stage of development.

Fic. 2. Asperula setosa. The migrating megaspore.

Fic: 3. Asperula azurea. An embryo-sac. Here two antipodals attain a spher-
ical form.

Fic. 4. A young embryo, In this form the suspensorial haustoria reach a high
degree of نی‎

Fic. 5: Asperula sp. An embryo showing the extreme development reached by
the suspensor in this form. The surrounding outline represents the limit of the endo-

Fic. 6. A. azurea. This figure illustrates the persistence of the suspensor till a
late stage of embryonic growth.
Fic. 7. A. galivides—embryo- c.
Fic. 8. A. gulioides. A rather early stage in the development of the embryo. The
haustoria are و رو‎ to erg out into the endosperm
. galioi f an deans which has reached a still

Fic
greater degre ef ver re bel depicted in fig. 6.

PLATE 9. Asperula montana
Fic. 1. Pollen mother-cell nucleus in the stage of diakinesis.
Fic. 2, a, b, c, d, e. Chromosome pairs and rings from prophases of the nucleus
of the ی کر‎ cell.

ollen mother-cell, early anaphase of the first divisio
indle, the end of which is seen ی‎ to the ecto-

FIG. 5, a, û, c. Chromosomes, early anaphase, division (2. m
"IG. 6. Late anaphase ; chromosome pairs show

ing three EEE

Fics. 7-9. Chromosome pairs from anaphase, first division (p. m. c.).

Fic. 10. End of spindle (anaphase, first division) attached " means of kino-
plasm strand to the ectoplasm.

FiG. 11 al first division (f. m. c.).

Fics. 12, 13. d of the first divisiobirealing stage (P. m. ¢.)

Fic. 14, a, ۰ e i pmi telo phase, first division.

Fic. 15, a, 4, c, d. Chromosome pairs during m — of the second division.

Fic, 16. Pluripolar diarch spindle (second division, 7. m.

108 COMPARATIVE EMBRYOLOGY OF THE RUBIACEAE

FIG. 17. pod view of chromosomes (12) metaphase, second division.
Fic, 18. Anaphase, second division
Fic. 19. Polar view anaphase, Mad division.

Fic, 21, a, 6. Chromosome ipie prophase, first embryo-sac-mother-cell division,
Fic. 22. Early anaphase, first embryo-sac-mother-cell division.
Fic. 23. Telophase, first ten embryo-sac-mother-cell.

PLATE Io
Fics. 1-6. Rubia tinctoria

Fic vpn. ovule before the integument has started to grow. Two archesporial
cells are ber

Fic. 2. یب‎ -cells of megaspore mother-cell in division.

Fic. 3. Embrvo-sac with four cells.

Fic. 4. Embryo-sac with two cells. At e s a second embryo-sac is seen contain-
ing but one nucleus. Above, in the archesporial cavity are several megaspore nuclei.

. Embryo-sac, defaitive,

Fic d Embryo. The embryo proper is very young, consisting of 13 cells. The

suspensor is exceedingly well developed as a haustorial organ

Fics. 7-22. Crucianella

Fic. 7. C. gilaniva. Young ovule, showing archesporium of very large cells, and
the nutritive cells at the chalazal end.

Fic. 8 ughter-cells, at close of first division.

Fic. 9. Megaspores formed by two divisions of the d cells. The end
cell has become the embryo-sac cell and has begun to develop as

3 Fic, 10. C. herbacea, Ovule, in outline, سر ین‎ its typograp Chalazal nu-

tritive tissue and megaspore masses above and an embr yo-sac An both partially
surrounded by the specialized nutritive tissue derived bon ۳ integument.

Fic. 11. Megaspore masses (seen in fig. ro), enlar,

Fic. 12. een (seen in fig. zo), enlarged, re antipodal nucleus only.
Polar nuclei not yet fus

Fic. 13. Nutritive eset tissue from another ovule; archesporium and
chalazal yep cells above; emb below.

IG. I4. A trinucleate cell from توت‎ nutritive tissue.
Fic, 1 » Embryo showing, in particular, the suspensor.

Fics. 16-22. Crucianella macrostachya

Fic. 16. Two embryo-sac-mother-c ells. The one on the right with a E

e first division; on the left two spindles of the second division.
Fe 1 Anaphase and telophase of the second divi sion.
i 19,

in mitosis,

+ 22, do embryo-sac ( with two nuclei) and four sets of megaspores. m.

of one set are in mitosis,

EXPLANATION OF PLATES 109

PLATE 11
Fics, 1-31. Crucianella gilanica
Fic. 1-13. Pollen tetrad divisions.
Fic. 1, Chromosome pairs seen during prophase of the first division.
Fic. 2. Anaphase of the first division; the chromosomes are in various stages of

separation.
Fic. 3. The forms assumed by es daughter chromosomes.
Fic. 4. Daughter chromosomes in later anaphase

Fic. 5. Chromosome Yes in ias of the first division,

Fic. 6. End of first division

Fic. 7. az pairs as seen in A of the second division.

in 8. Metaphase, second divisi 10

G. 9. Late anaphase, second division.

dem IO. Anaphase, second deinen, spindle showing its connection with the
ectoplasm.

Fic. 11. Equatorial plate, polar view ; ten chromosomes.

Fic, 12. Equatorial plate, polar view; eleven chromosomes.

Fic. 13. End of second division. Drawn to the same scale as fig. 6, for com-
parison of the size of nuclei

Fics. 14-25. Embryo: -sac tetrad divisions.

Fic. 14. Diak

Fic. 15. دم‎ of rings, the aster-like radiations correspond to those seen
in Fig. 42.
Fics. 16, 17. Anaphases, first division.

Fic. 18, a-g. Forms assumed by the chromosomes during anaphase of the first
division.

Fic. 19. Telophase, first division.

IG. 20. Telophase, first division.
Fic, 21. Prophase, second division
22. Mb qt second division.

Fic. 23. on some pairs an second division.
rats, second pe visi
Fic. 25 Er mosomes, a ine Be division.

Fic. 26. Chromosomes, sph, first division.

Fics. 27-30 Embryo-sac divisio

Fic. 27. First embryo-sac aividoas in metaphase.

Fic. 28. Second embryo-sac division, polar view.

Fic. 29. Chromosome pairs , second embryo-sac division.
son.

Fic, 3t. Mitotic figure from the nucellus.
Fics. 32-44. Crucianella macrostachya.
Fics. سوم‎ Pollen tetrad divisions.
Fic. 32. Prophase, first division.
Fic. 33. Metaphase, first division: ; ten chromosomes.
Fic. 34. Very early anaphase, first divisio:
Fic. 35. even ene a rude one vie in lateral and Zone in polar
view; ten chromoso:
G. 36. ee pair, metaphase, second division.

110 COMPARATIVE EMBRYOLOGY OF THE RUBIACEAE

Fics. 37-42. Embryo-sac tetrad division.
Fics. 37, 38. Prophase, first division.
Fics. 39, 40. Early anaphase, first division
Fic. 41, Separating chromosomes, anaphase, first division.
Fic. 42. Prophase, second division. Perinuclear radiations.
Fic. 43. First embryo-sac division.
Fic. 44. Separating chromosomes of the first embryo-sac division.
PLATE 12
Fics. 1-12. Diodia Virginiana
Fic. 1. Basal partition with nucelli.

IG. 2. Ovule with one-celled embryo-sac. 2, a’, branches of vascular bundle,
Fic, 3. Part of young ovule, the integument commencing, showing archesporium.
Fic. 4. One-celled embryo-sac in migration. The remaining megaspores seen

above following the embryo-sac

IG. 5. A similar condition in another ovule.

Fic. 8. Ovule with definitive embryo-sac, Horen the relation of parts. 2,4”,
branches of vascular tissue. The isolated cells have appeared which later form a con-
tinuous ری‎ as seen in fig. 9, sp

Fic. 9. Older ovule.

Fics. fia 11, I2. The antipodal end of three embryo-sacs showing the number
and arrangement of antipodal cells in each.

FIGS. 13-1 é Diodia teres

Fic. 13. Basal partition and young nucelli.

Fic, 14. Somewhat older condition ar the strophiole (s¢) beginning to de-
m

15. Archesporium. One large functional megaspore mother-cell flanked by

a nave one.

Fic. 16. Young ovule with archesporial cell in ‘its second division ; strophiole
well marked.
Fic. 17. Second division of the megaspore mother-cell.

PLATE 13
Fics, 1-10. Diodia teres
Fic. 1. Young ovule, with one-celled embryo-sac. The strophiole is seen to close
the mieropyle,
Archesporium showing the large, functional megaspore, and the tie
sister مش تست بو‎ to separate. The arrow indicates position of micropyle.
Fic. 3. Embryo-sac cell وج‎ ; at the base are seen two nutritive cells.
en 4. Two-celled embryo-sa
G. 5. Embryo-sac.

IG. 6. Embryo-sac after fertilization containing a one-celled embryo. The anti _
normal.

podals are still

f

Fic. 7, Hi. ovule in longitudinal section containing the embryo-sac shown inde 。

tail in Fig.
Fic, 8. on older condition with more advanced endosperm and embryo.
Ri detail,

10, More advanced embryo, showing the suspensor after stretching.

EXPLANATION OF PLATES 111

Fics. 11-18. Richardsonia pilosa

FiG. 11. Nucel
Fic, 12. Ovule in TA sac cell. The strophiole is seen.
Fic. 13. Embryo-sac
Fic, 14. Embryo-sac eid small antipoda
Fic. 15. An antipodal cell as it appears aem the cell wall is secreted.
Fic. 16. Ovule with endosperm and embryo.
E 17. Embryo of Fig. 16 in detail.

G. 18. Older embryo and the special cells of the endosperm.

PLATE 14. Houstonia

Lettering : 5, stylar primordia; 7, placental primordia; 7, roof element of ovarial
partition; 4, basal element of ovarial partition; ff, plane of cell division adding to the
unicle; 2, antipodal cells

Fic. 1. Æ. coerulea. Longitudinal section through a young flower.

Fic. 2. 24. longifolia. Longitudinal section through an older flower.

Fic. 3. Æ. longifolia. A similar section after the fusion of the basal and roof
partitions.

Fic. 4. A nucellus with a single functional megaspore mother-cell,

Fic. 5. Archesporium, showing two smaller megaspore mother-cells which do not
divide. The arrow indicates the direction of the egg pole.
. 6. Two megaspores formed, and the second daughter-cell in division.
IG. 7. Embryo-sac with tw
IG. 8. Ovule in which the tires contains four nuclei in telophase

Fic. 9. A. coerulea. Definitive embryo-sac. At the antipodal pole there are two
megaspore mother-cells still persistent

Fic. 10. Z. coerulea. Ovule with three-celled embryo, and endosperm.

Fic. ۰ One esili: embryo.

Fic, 12. Five-celled embryo.

PLATE 15
. Conductive epidermis ( Diodia Virginiana) in young ee in the re-
gion x fasion pisc: the roof and basal elements of the ovarial

Fr A portion of the stylar conductive tissue penetrated pis T ten tubes.
Portods d three such tubes are shown.

Fic, 3. A portion of the stylar tissue showing a pollen tube taking a brief course
xui a large E with the longitudinal axes of the cell (Richardsonia pilosa).

G. 4. Pollen tubes as seen in the base of the stylar tissue, turning to pass into
the - Priser tissue ( Æ. pilosa
Fic. 5. Mouth of micropyle and adiacent conductive tissue, with a pollen tube
(R: pilosa).

Fics. 6, 7. Ovular conductive tissue of the kind seen in fig. 5, cut transversely to
the er axes of the cells. Parts of pollen tubes are seen passing between the
cells er

"P Conii epidermis of Diodia Virginiana.

Fie 9. A small portion of the surface of e conductive epidermis of Diodia
Virginiana showing the pollen tube on the su

Fics, 10, 11. Parts of pollen tube pos س‎ plugs (Diodia Virginiana).

TABLE OF CONTENTS

PART 1
Introduction 3... by wn a Ue eo ¿ib e i
Descriptive: Vaillantia hispida موم موم م۰۰۰۰‎ 8
Explanation of plates......... codice ماخ‎ eee eee 23
PART 2
Further description of genera.
CA 27
WA 32
O E ETT ULAIL. .. c E 34
DIEN. qnx eene re ee ecc i ee 36
ais AC c ks RA 43 |
CHE تپ‎ Ces A ws ru vu von i E 44
AROS VOTAN. ee ence keene Her et tie 48
IMIR ll bods وی‎ UU VR des n AN 52
a PUG sorrise ee- r ourer oru e 54
LAS ei d ME 56
Summary and CONCLUSIONS 4... err nr ma 59
The tetrad and embryo-sac itoses rer et! 66
Summary and. conclusions... وان وی‎ seo rH S 82
The behavior of the pollen tube
Diodia teres and Richardsonia ۵6/0506. + '"" 88
A A nmt 97
Késults asd conclusions مر‎ O 100
DO cr ÓN 103
Explanation of plates 105

۰
..
* » ^
"oO RIN K RR M B4 MM b.e * Bo.» o * b ue "o t "T

(112)

PLATE 6.

2 mE mer dn با‎
» ELS EY EX c x.
3 = y E
E ۳

CALLIPELTIS CUCULLARIA.

13- 23. SHERARDIA ARVENSIS.
THE MELIOTYPE PRINTING CO., BOSTON.

1-12.

Mem. Torr. Bor. CLUB, 8.

7.

P

Torr. Bor. Crus, 8.

Mem.

v 5
DY

c7:
(4

Mem. Torr. Bor. Crus, 8 ‚ : Pr.8

Bor. CLUB, 8

TORR.

MEM.

19

F, E. L. del.

ASPERULA MONTANA

=
2
VÉ
x »
N
a
aN e

Mem. Torr. Bor. CLUB，8

i

5
V

Be NP
ERA 35

A

lint uu

s A 7 ۱
2 e | N
(A

| N Wi E

1-31-CRUCIANELLA GILANICA

32-44 CRUCIANELLA MACROSTACHYA

ga

4

i7

due p

—

تحت نی PDA‏

EE
ER
TO In
GITE. 2

Ao

:
3
3

: 8
BEHAV

Bor

Torr.

Mem.

Memoirs of the Torrey Botanical Club

The MEMOIRS are published at irregular intervals and are sold at a uniform
‘price of THREE DOLLARS a volume. Most of the articles can be furnished
separately at the prices indicated below. The price includes postage

Volume 1.

No. 1.—Bailey, L. H. Studies of the Types of various Species of the Genus
Carex. Pp. 1-85. s

No. 2, —Martindale, I. C. Marine gom of fà ge iori Coast and adjacent
Waters of Staten Island. Pp. 87-109.

No. 3.—Spruce, R. An ری‎ a E the Heic o ecole by Dr. H. H.
Rusby in South America. Pp. 113-140. 1890. Price
No. 4 turtevant, E. L. On seedless Fruit. E pron 1890. Price,
75 cents.

Volume 2.
No. 1.—Halsted, B. D. On reserve Food nn in Buds and surrounding
Parts. Sant 1890. Price, 50 cen
No. 2.—Vail, A. M. and Hollick, A. ‘Contributions to the Botany of Virginia.
Pp. 27-56. Pl. 8, 4. 1890. Price, 75 ce
No. 3,—Holm, T. 0 ios ون‎ ‘Knowledge of SFE Germination of some
North American Plants. pa 57-108. Pl. 5-19. 1891. (Notfurnished separately. )
o. 4.—Wheelock, W. A Monograph of the ryp American Species of
the Bis Polygala. Pp. ae 52. 1891. Price, 75 cents.

Volume 3.
No. 1.—Small, J. K. and Heller, A. A. On the Flora of Western North Caro-
lina and contiguous Territory. Pp. 1-39. 1892. Price, 50 cents
No. 2.—Morong, T. A Revision of the North American Keane with Illus-
trations of all the papa de 1-65. Pl. ehe :1893. Price, $2.00.
No. 3.—Rusby, H. H. An Enumeration of the en collected in Bolivia by -
Miguel Bang. Pp. 1-67. D Price, 50 cents.

Volume 4.

No. 1 Underwood, L. M. Index Hepaticarum. Part 1, Bibliography. Pp. a x

1-91. 1893. Price, 75 cents.
No. 2.— Vall, A. M. and Small, J. K. Report on the e Botanical Exploration of
_ Virginia during the Season 1892- Pp. 92-202. Pl. 75-82. 1893-1994. Price, 50-

Miguel Bang—II. Pp. 203-274. 1895. Price, 50 cents.

No. 3.—Rusby, H. H. An Enumeration of the Plants collected in Bolivia by mu

۳ No. 4.—Pettit, A. S. Arachis hypogaea L PR E 1895. Price > ;
cents.

No. 5 一 Byaberg P. A. Monograph ot Papa and related Genera. Pp. وه‎ =

: w 1896. Price, 75
: Address, TORREY BOTANICAL CLUB,

E E | Columbia ی‎ New York.

| Memoirs of the Torrey Botanical Club :

as MEMOIRS are E. at irregular intervals and are sold at a uniform -
ce of THREE DOL volume. Most of the articles can be furnished
ی یی‎ at the prices مر‎ below. The price includes postage.

Volume 5.

List of Pteridophyta and Spermatophyta growing without Cultivation in North-
eastern North America. Pp. 1-377. 1893-1894. Price, $3.00

Volume 6.

No. 1.—Rusby, H.H. An Enumeration of a en collected in
Miguel Bang—III. Pp. dou 18 6

y LJ. Reyi

: pluvialis). Pp. 211-244. PI. es 87 (colored).
Ps No. 4.—Underw vi M
= to 1832. Price, 50 cen
No. 5.— Fink, B. شوه‎ on the Lichen Distribution in the Upper Mi
. Valley. Price, 25 cents. |
Volume 7.
Howe, M. A. The Hepaticae and Anthocerotes of California. Pp. 1-908. 7
88-122. Sag Price, $3.00.- ume
Volume 8.
No. 1.—Lloyd, F. E. The comparative Embryology of the Rubiaceae.
. Pp. 1-26. Pl. 1-4. 1899. Price, 50 cents. Part II. Pp. 27-111. Pl. 5-15. 1
ce, er

۰

No, 2—Bvans, A. W. The Lejeuneae of the United States and Canada. Pp.
d 113-183, Pl. 16-22. 1902, Price, $1.00. A
ER o. 3.—Britton, E. G. & es A. The Life History of Vittaria E
s p. 23-31. 1902, TM 75 cen

7 Volume 9. : 2
. Salmon, E. S. A Monograph of the Erysiphaceae. Pp. 1-292. Pl.
1900. Pres $3.00. — = |

| Volume 10.
| Burgess, E E. s. ‘Studies i in the Genus Aster.

Part I. In Press. —

A ی را‎ BASE Re

The Lejeuneae of the United States and Canada
By ALEXANDER W. Evans
(PLATES 16-22)

The subtribe Lejeuneae, as defined by Schiffner, is made up of
the large and comprehensive genus Lejeunea as understood by |
Spruce. The latter author divided this genus into thirty-seven
natural divisions, which he looked upon as subgenera, while Schiff-
ner raised nearly all of these subgenera to generic rank. Each of
these authors has his own adherents. Writers on the hepaticae
of Europe, where the Lejeuneae are poorly represented, tend to
follow Spruce, recognizing but a single genus, while writers on the
hepaticae of tropical regions, where the Lejeuneae attain a most
luxuriant and varied development, tend to follow Schiffner. It is
probable that an intermediate course would be more nearly cor-
rect, that some of Schiffner's genera are worthy of generic rank,
but that others ought to be united. Until, however, the many
species are more clearly defined and more accurately understood
than they are at present, it seems most practicable, and perhaps
wisest, to recognize Schiffner's genera as such, at least in a tenta-
_ tive way—a course which is adopted in the present paper

For descriptions of the various genera, the reader is referred to
the writings of Spruce * and of Schiffner.* Reference may also be
made to a recent paper by the writer,f where the division of the
group into genera is more fully discussed.

HISTORICAL
The earliest work devoted to North American hepaticae is
that of Schweinitz, published in 1821. In this páper the genus
Jungermannia i is still used in its old sense and includes seg ee

* Hep. Amaz. et And. 63-308. 1884.
T Engler & Prantl, Nat. Pflanzenfam 117-131. 1893.
j The Hawaiian Hepaticae of the hie thie Trans. Conn. Acad. 10:
387-462. . 1900.
Me

| ey Botanical Club, Volume VIII.
[No. 2 issued 15 February, 1902.]
113

Sl ce ux

114 LEJEUNEAE OF THE UNITED STATES AND CANADA

manniaceae and the Metzgeriaceae of recent American authors.
Among the species referred to this genus are three which would
now be included in the Lejeuneae. These are /. serpyllifolia, J.
transversalis and J. clypeata. The first of these is a species which
we share with Europe. It is the type-species of the genus
Lejeunea, as originally described, and is now known as Lejeunea
serpyllifolia Lib. or preferably as L. cavifolia (Ehrh.) Lindb.
Apparently Schweinitz did not really know this species from
America, the specimens so named in his herbarium being partly
Kadula tenax Lindb. and partly Frullania Virginica Gottsche.
The second species is not the true West Indian J. transversalis
Swartz, as Schweinitz supposed, but is the same as his J. clypeata,
which he correctly distinguished as new. This third species is now
referred to the genus Archilejeunea. Apparently in ignorance of
Schweinitz's paper, Lehmann, in 18 38, described the same species
as new under the name Lejeunea Dorotheae.

In 1841 and in 1845 two important exsiccatae of North Ameri-
can bryophytes appeared in which several Lejeuneae were included.
In the first of these, Drummond's Mosses of the Southern States,
the specimens were determined by Wilson and Hooker and, in the
case of new species, were accompanied by printed descriptions.
The Lejeuneae of this collection are labeled Jungermannia serpyl-
folia, J. auriculata and J. parvula, the last two being described as
new. The first of these is not correctly determined and is in fact
a mixture of two distinct species, Euosmolejeunea duriuscula (Nees)
and Lejeunea Americana (Lindb.), proposed as a species in the
Present paper. |. auriculata is the type of Mastigolejeunea auricu-
lata Schiffn., a species now known to have a wide distribution in
tropical America. J. parvula, however, is apparently the same as
the European Cololejeunea minutissima ( Sm.) Schiffn., which has
subsequently been collected many times in our Southern States.
In the second exsiccata, Sullivant’s Musci Alleghanienses, we find
six Lejeuneae—Phragmicoma clypeata, Lejeunea serpyllifolia, L.
serpyllifolia var., L. cucullata, L. calcarea and L. minutissima.
The first and last of these are correctly determined. Of the re-
mainder, Z. serpyllifolia is really Z. Americana; L. serpyllifolia
var. is Euosmolejeunea duriuscula; L. cucullata, although com-
pared by Sullivant with a specimen communicated by Montagne,

HISTORICAL 115

is not the same as what we now call Microlejeunea cucullata (R.
Bl. & N.) Jack & Steph., of Java, but is apparently to be referred
to Mierolejeunea lucens (Tayl.), a species at that time undescribed ;
while Z. calcarea is not precisely the same as the European Colole-
Jeunea calcarea (Lib.) Schiffn. but agrees better with the North
American representative of this species, C. Biddlecomiae (Aust.),
which was not published until 1890.

The next notes we find on our Lejeuneae are by European
writers. From 1844 to 1847 Gottsche, Lindenberg and Nees von
Esenbeck issued their classical Synopsis Hepaticarum in five parts.
The Lejeuneae are treated in the second and third parts, which
appeared in 1845, and in the fifth or supplementary part, which
was published two years later. In the second and third parts the
following five Lejeuneae are recorded from North America: Phrag-
micoma versicolor, Lejeunea transversalis, L. catenulata, L. Doro-
theae and L. serpyllifolia. The first of these species is a composite ;
so far as American material is concerned, it is the same as Junger-
manna auriculata Wils. & Hook., which is unjustly reduced to it
as a synonym. Lejeunea transversalis is admitted solely on the au-
thority of Schweinitz. Z. catenulata, under -which Jungermannia
transversalis Schwein. strangely appears as a synonym, is Neuro-
lejeunea catenulata (Nees) Schiffn. It is a species of tropical
America, but there is no adequate evidence that it occurs within
the United States. The last two species require no comment. In
1846 Taylor described a large number of new hepaticae from
various parts of the world and among them six Lejeuneae, pre-
sumably from North America. Many years later it was proved
that four of these species, viz.: Phragmicoma testudinea, P. cyclo-
stipa, Lejeunea polyphylla and L. longiffora,* did not come from Ohio
at all, as Taylor supposed, but were really collected near Para in
Brazil. A fifth species, Z. calyculata, is a synonym of Archilejeunea
clypeata. It is probable that the sixth species, Z. /ucens, was also
originally collected at Para. This plant, however, has a wide

* AIL four of these species are reduced to synonymy by Stephani (Hedwigia,
29: 1890). Phragmicom ı testudinea is Zyenolsjeunsa macroloba (Mont.) Schiffn.,

116 LEJEUNEAE OF THE UNITED STATES AND CANADA

range, extending into the United States and is apparently the same
as Sullivant's Z. cucullata. In the supplementary part of the
Synopsis Hepaticarum these six species of Taylor are quoted with
little comment, and L. minutissima, L. calcarea and L. cucullata
are recorded from North America on the authority of Sullivant.

`. The first and second editions of Gray's Manual contain descrip-
tions of the hepaticae by Sullivant. In the first edition, published
in 1848, only four Lejeuneae are included. These are L. serpylh-
folia, L. clypeata, L. calcarea and L. cucullata. - In the second edi-
tion, published in 1856, species from the Southern States are added,
and we find eleven Lejeuneae. The list includes the four species
quoted above, Taylor's six new species, as well as L. minutissima
and L. auriculata. L. lucens Tayl., however, is quoted merely as
a synonym of Z. cucullata.

In Austin’s papers on hepaticae, published from 1869 to 1879,
we find seven species of Lejeuneae described as new and three species
recorded for the first time from the United States. Two of the new
species, Z. Sullivantiae ('72) and L. Mohrü (75) are now reduced
to synonymy ; the first being referable to Euosmolejeunea duriuscula
and the second to E. opaca (Gottsche) Steph. A third species, L5
biseriata ('69) proved to be a moss, as Austin himself afterwards
decided. Three other species, L. Ravenelü, L. Caroliniana and L.
laete-fusca, all published in 1876, are so incompletely described
that we know little about them. Even the type specimens, where
accessible, do not aid us; they are very fragmentary and are either
sterile or in poor condition. Under the circumstances it seems ad-
visable to allow these three names to disappear from our literature.
Austin's seventh species, the only one here retained, is £L. 0
(75), which belongs to the genus Cololejeunea. Of the three spe-
cies which Austin first accredited to the United States, Phragmi-
coma xanthocarpa does not agree with authentic specimens of this
widely distributed tropical species but should be referred to Archi-
lejeunea Sellowiana Steph.; the specimens referred to the European
Lejeunea ovata, now Harpalejeunea ovata (Hook.) Schiffn., are prob-
ably correctly named ; while those referred to Z. /aete-virens Mont.,
a West Indian species, are in too poor a condition to be definitely
determined.

In 1875 the Finnish botanist, Lindberg, in a lengthy paper on

HISTORICAL 117

Irish hepaticae, includes interesting notes on some of the Lejeu-
neae in American exsiccatae. The most important of these are
the following: Zejeunea ulicina, of Europe, is accredited to North
America; Z. parvula (the Jungermannia parvula of Drummond's
collection) is recognized as a species, and Sullivant's L. minutissima
is reduced to it as a synonym; the American specimens of L. ser-
Pyllifolia are all referred to the new var. Americana Lindb.; L.
Sullivantiae Aust. is given the new name Z. Austini Lindb., on
account of the previously described Z. Sullivantii Gottsche. The
evidence that Microlejeunea ulicina (Tayl.) occurs in North Amer-
ica does not seem to be conclusive. Lejeunea parvula and L. Sul-
livantiae have already been commented upon. Lindberg’s var.
Americana is a composite, and, while it includes the Z. Americana
of the present paper, it includes also typical Z. cavıfolia.

In 1884 Spruce, in his important work on South American
hepaticae, accredits the rare European Z. diversiloba Spruce to
North America. He apparently refers to this species the speci-
mens which Sullivant called ۲۰ cucullata, although this is not defi-
nitely stated. Three years afterwards Spruce referred Sullivant's
specimens somewhat doubtfully to Z. /ucens Tayl.

The numerous papers on hepaticae by Underwood contain fre-
quent references to the Lejeuneae. In his preliminary list of
North American species, published in 1882, fourteen species of
Lejeuneae are noted. In his Descriptive Catalogue of 1884, de-
scriptions are given of these fourteen species and of four others
omitted from the earlier list. All of these species had been pre-
viously recorded. In the sixth edition of Gray's Manual, pub-
lished in 1890, Underwood described the hepaticae, and it is inter-
esting to note that the four Lejeuneae which he there recognizes
are the same ones which Sullivant described in the first edition of
the Manual, forty-two years before. In the same year Underwood
published a description of Leyeunea Macount Spruce, and Pearson,
a plate of L. Biddlecomiae Aust., accompanied by descriptive notes.
Both of these species belong to the genus Cololejeunea. Two years
later Underwood recorded another Cololejeunea, the European Le
Rossettiana Massal., but there seems to be no very adequate evidence
that this species has been collected in America.

The last published list of North American Lejeuneae is that of

118 LEJEUNEAE OF THE UNITED STATES AND CANADA

Stephani, which appeared in 1892. Twenty species are recog-
nized ; of these, Zejeunea (Eu-Lejeunea) Underwoodi and L. (Micro-
Lejeunea) Cardoti are described as new, while L. (Euosmo-Lejeunea)
trifaria Nees and L. (Colo-Lejeunea) Wrightii Gottsche are re-
corded for the first time from the United States. The first of the
new species, judging from authentic specimens, is identical with
Lejeunea Sullivantiae Aust. and is therefore a form of Zuosmo-
lejeunea duriuscula; the second, however, seems to be distinct.
The specimens referred to Euosmolejeunea trifaria are dioicous
and agree with Z. opaca, while those referred to the undescribed
Lejeunea Wrighti, from Cuba, are apparently an abnormally devel-
oped form of Cololejeunea minutissima. In 1895 Stephani ac-
credited Lejeunea fava Swartz to Florida. The occurrence of this
widely distributed species is certainly to be expected within our
limits, but a specimen from Florida collected by Rau and deter-
mined as L. fava by Stephani himself seems to be referable to
Spruce's var. albida of this species, a plant which the writer is
unable to distinguish from L. Americana.

In the Hepaticae Americanae of Underwood and Cook, the
first decade of which appeared in 1887, and which is still unfin- "
ished, a number of Lejeuneae have been distributed. Those from
the United States or Canada will be noted in connection with the
description of species.

It will be seen from the preceding account that the opinions of
writers with regard to several of our species have been both varied
and uncertain. This fact becomes even more evident when we
consult the Lejeuneae in herbaria or in published exsiccatae. We
are almost certain to find many false determinations among them.
In some cases a single species will appear under several different
names; in other cases several distinct species will appear under
the same name. This confusion is partly because the Lejeuneae
often grow mixed together, partly because certain of the species
are extremely variable, and partly because the published descrip-
tions, especially those of the older writers, are often so general and
so incomplete as to be quite valueless. It becomes often necessary
therefore to examine type-specimens, in order to learn upon what
particular form a given species was originally founded. It be-
comes equally necessary, in the case of a variable species of wide

DISTRIBUTION OF SPECIES 119

range, to examine a large series of specimens from as many dif-
ferent localities as possible. It is only by this latter method that
one can acquire a sufficient knowledge to distinguish between
specific characters and those peculiarities which are simply due to
variability. It should also be remembered in this connection that
plants which grow under abnormal conditions tend to develop
abnormally or imperfectly, and this is most strikingly true of the
Lejeuneae. It may even happen that a specimen bearing peri-
anths and capsules may utterly fail to show the foliar characters
of the species, on account of the crowding together of the fruiting
branches. Such abnormal specimens, although interesting from a
morphological point of view, cannot always be determined with
certainty, and yet it is unfortunately true that many species of
Lejeuneae have been founded on just such forms. It is sometimes
a question as to what should be done with species whose type-
specimens are of this character, but probably in doubtful cases the
best course is to discard them altogether. `

DISTRIBUTION OF SPECIES

As a result of his own studies the writer is able to distinguish
twenty-three species for the United States and Canada. Three of
these are proposed as new, one is raised from varietal rank and
six are recorded for the first time. The remaining species have
all been previously noted. Four of our species occur in Europe,
thirteen are more or less widely distributed in tropical America and
six appear to be endemic. It should be noted, however, that
two of these six species have near relatives in Europe and that
the remaining four are closely allied to species of tropical America.
It is very probable also that the ranges of several of these appar-
ently endemic species will be extended into tropical regions by
future explorations. Three of our species, Lejeunea cavifolia, L.
patens and Cololejeunea Macounii are distinctly northern in their
distribution. Archilejeunea clypeata, A. Sellowiana and Colole-
Jeunea Biddlecomiae all have wide ranges extending from eastern
Canada or New England to or beyond the Gulf States. Aarpale-
Jeunea ovata has been found from Virginia to northern Georgia,
while Microlejeunea Ruthii has been collected only in Tennessee.
Our other species are characteristically southern, although Zeyeunea

120 LEJEUNEAE OF THE UNITED STATES AND CANADA

Americana and Cololejeunea Jooriana have been found as far north
as North Carolina, and Microlejeunea lucens apparently occurs in
Virginia. Six of these southern species have not been collected -
outside of Florida. With the exception of Cololejeunea Macounit,
of British Columbia, no species are known from the Pacific Coast
region ; no species moreover have been reported from the Rocky
Mountains nor from the Great Plains east of the Rocky Mountains,
their absence or scarcity apparently being due to the dry atmo-
spheric conditions. With regard to the number of species in indi-
vidual States, Florida leads with fifteen, Louisiana ranks second
with nine and Alabama third with eight. Few of the remaining
States have more than three species apiece, and even east of the
Mississippi River there are several States which have had no Le-
jeuneae whatever reported from them.

The brief notes which have just been given are mainly of tem- |
porary value. The Lejeuneae have been so largely neglected by
collectors that our knowledge of their geographical distribution is
at best very incomplete. The rich harvest of species which
Underwood gathered in Florida shows how little we had previ-
ously known about the Lejeuneae of that state and indicates that
much is still to be learned. The collections of Langlois in Louisi-
ana and of Mohr in Alabama point to similar conclusions. Even
in the north, the recent discovery of Lejeunea patens in Newfound-
land and in Cape Breton shows how fragmentary our knowledge
really is, It should be noted also that specimens exist in herba-
ria, too incomplete for identification and yet undo ubtedly distinct
from any recorded species. Some of these are probably identical
with described species from tropical America, others are probably
undescribed ; but for the present they must be left in doubt. It
is clear, therefore, that much remains to be done to complete our
knowledge of the Lejeuneae, and it is to be hoped that the pres-
ent paper may incite collectors to pay more attention to these in-
conspicuous but beautiful plants.

The revision which follows is based largely on the collections
of Lejeuneae in the Underwood herbarium, in the United States
National Herbarium and in the herbaria at Yale University. The
curators of other herbaria have also been most generous in allow-

A ee ee o تن‎ E Sima a= =. TT w€—

DISTRIBUTION OF SPECIES 121

ing the writer to examine type specimens in their charge, and
other correspondents have kindly furnished material for study.

Analytical Key to the Species

1. Underleaves present, undivided. 2.
Underleaves present, bifid. 6.
Underleaves absent. 19.

2. Plants pale, perianth five-keeled. 3
Plants more or less tinged with brown or purple. 5.

3. Lobule bluntly pointed at the apex or apículate. I. Archilejeunea clypeata.
Lobule with a TM and slender apical tooth. 4.

4. Lobe widely slightly convex or plane, averaging 1.2 X 0.95 mm., postical

margin usually fadi ng a distinct angle with keel. 2. Archilejeunea Sellowiana.
Lobe obliquely spreading, strongly convex, averaging 0.7 X 0.45 mm., postica al
margin usually forming a continuous line with keel. 3. A rchilejeunea eonchifoha.

5. Lobule entire on margin or with a Aen apical e, — leaf-cells averaging

21 X 12 a inflorescence usuall g , perianth trigonous
throughou 4. Mastigljeune a auriculata.
Lobule ae d on margin with about four teeth, median leaf-cells averaging 23
H, 9 inflorescence usually innovating on both sides, perianth six- to eight-

plicate i in upper part. s. Brachiolejeunea corticalis,

6. Lobe acute, at least on well developed leaves, underleaves with rounded division,

and broad, shallow s . Harpalejeunea ovata.
Lobe rounded or rarely تن‎ underleaves with obtuse to acute divisions and nar-
row sinus. 7

7. Plants firm in texture, yellowish- or brownish-green, inflorescence dioicous. 8.
Plants delicate in texture, whitish or pale green, varying to dark green. 10.

8. Underleaves small, distant, narrowed toward base. ۰ Euosmolejeunea duriuscula.

se :

Underleaves large, contiguous to imbricated, rounded or cordate at
9. Leaves loosely imbricated, the lobe broadly ovate, obliquely spreading.
8. Euosmolejeunea opaca.
Leaves densely imbricated, the lobe orbicular, widely spreading.
9. Cheilolejeunea polyantha.
10. Branches sometimes flagelliform and leafless but with persistent underleaves, €
inflated, perianth (so far as known) with a plane antical face. I.
Branches never flagelliform.
II. Lobes without basal ocelli, median leaf-cells vasis 21 u in diameter, inflores-
ence autoicous. joe ejeunca phyllobola.
Lobes with basal ocelli, median leaf-cells averaging IO u in diameter, inflorescence
ioicous 11. Cheilolejeunen versifolia.

12.

12. Lobule u not inflated, inflorescence autoicous, perianth plane or nearly so on
antica 12. Cheilolejeunea pililoba.
13.

Lobule vowel perianth with a distinct antical keel.

m
دیا‎

. Lobe widely spreading, inflorescence auto ico 14.
Lobe obliquely spreading to suberect, ehe dioicous.

16.

122 LEJEUNEAE OF THE UNITED STATES AND CANADA

14. Lobes and underl distinctly crenulate on the margins, underleaves smaller than
13. Lejeunea patens.
15.

lobules.
Lobes and underleaves entire.or nearly so.

15. Plants pale to dark green, underleaves as large as lobules or larger, not rounded at
base, perianth abruptly narrowed at base. 14. Lejeune ia.
Plants pale, underleaves larger than lobules, rounded at base, perianth gradually
narrowed toward base. 15. Lejeunea Americana,

16. Median leaf-cells 18-20 u in diameter. I7:
Median leaf-cells 12-13 u in diameter. 18.

17. Underleaves longer than broad, deeply bifid, often unidentate on sides.
Microlejeunea lucens.
Underleaves as broad as long, bifid to middle, never unidentate on sides.
17. Microlejeunea Ruthit.
18, Lobule much smaller than lobe, often obsolete, lobe indistinctly ocellate at base,

averaging 0.25 mm. in length. 18. Microlejeunea Cardoti.
Lobule nearly as large as lobe, lobe not ocellate at base, averaging 0.17 mm. in
length. I9. Microlejeunea bullata.

. Lobule three fourths the length of lobe or longer, lobe rounded at the apex and
without hyaline cells, crenulate on the margin, stylus reduced to a single papilla,

-
o

inflorescence autoicous 20. Cololejeunea minutissima. i

Lobule about half as long as lobe. :
Outer surface of lobe rough from projecting cells or wart-like thickenings or both,
hyaline cells absent, inflorescence dioicous or autoicous, 21.
Outer surface of lobe almost smooth, the cells scarcely if at all convex, hyaline
cells often present at the apex and along antical margin, stylus usually two cells
long, inflorescence synoicous or paroicous, 23. Cololejeunea Jovriana.

Y
9

to
-

: Lobe narrowed toward the apex, with strongly convex or conical cells, the project-
ing wall sometimes thickened, stylus usually two to ten cells long.
Cololejeunea Biddlecomiae.

a.
Lobe broad at the apex, with slightly convex cells, each bearing in the middle a

large globoid thickening, stylus usually two cells long.
22. Cololejeunea ۵

DESCRIPTION OF SPECIES
1. ARCHILEJEUNEA CLYPEATA (Schwein.) Schiffn.
PLATE 16, Fics. 1-11
Jungermannia clypeata Schwein. Spec. Fl. Amer. Sept. Crypt.
Hep. 12: 1831:
Jungermannia transversalis Schwein. L c. (not Swartz).
Phragmicoma clypeata Nees, Naturgesch. Eur. Leberm. 3:
248. 1838.
Lejeunea Dorotheae Lehm. Pugillus, 7:15. 1838. GER
N. Syn. Hep. 332. 1845.

ARCH ILEJEUNEA CLYPEATA

Lejeunea Carolinensis Mont.; G. L. & N. /. c. (as synonym).

Lejeunea calyculata Tayl. Lond. Jour. Bot. 5: 388. 1846.
G L & N: Syn. Hep. 752. 1847.

Lejeunea clypeata Sull.; Gray, Manual, Ed. I., 685. 1848.

Symbiezidium calyculatum Trevis. Mem. r. ıst. Lomb. III. 4:
403. 1877.

Lejeunea (Archi-Lejeunea) clypeata Spruce, Hep. Amaz. et
And. 90. 1884.

Archilejeunea clypeata Schiffn.; Engler & Prantl, Nat. Pflanzen-
fam. 1%: 130. 1893.

Pale glaucous-green or whitish, growing in broad depressed
mats: stems prostrate, 0.085 mm. in diameter, closely appressed
to matrix, irregularly branched: rhizoids scanty, sim le or
branched : leaves imbricated, the lobe obliquely spreading, slightly
falcate, broadly oblong to suborbicular, convex, averaging 0.6 mm.
long and 0.45 mm. wide, slightly revolute at the rounded apex,
margin crenulate from projecting cells, antical margin rounded at
base and arching across or just beyond axis; lobule inflated, ovate-
triangular, 0.2 mm. long, 0.12 mm. wide, keel straight or slightly
curved, forming a very obtuse angle with postical margin of lobe
and sometimes almost continuous with it, free margin slightly
curved, involute to beyond apex, then obliquely truncate to end of
keel, apex blunt or subacute, consisting of a single projecting cell
or rarely of two cells; cells of lobe convex with rather thick pro-
jecting walls and thin vertical walls, trigones distinct though some-
times small, acuminate, indeterminate thickenings absent except
sometimes near base of lobe, cells at edge of lobe 15 in diameter,
in the middle and at the base 19 £: underleaves distant, orbicu-
lar, 0.2 mm. long, rounded at the apex, abruptly narrowed and
neither rounded nor cordate at base, very short decurrent and at-
tached bya slightly curved line of insertion : inflorescence autoi-
cous or dioicous: 9 inflorescence usually borne on a leading
branch, rarely on a short branch, innovating on one or rarely on
both sides, the innovations usually simple and sterile ; bracts com-
plicate, unequally bifid, scarcely or not at all winged on keel, the
lobe obovate, slightly falcate, rounded at the apex, 0.75 mm. long,
0.45 mm. wide, crenulate on the margins, lobule oblong, rounded
at the apex, 0.4 mm. long, 0.12 mm. wide ; bracteole ovate-oblong,
0.6 mm. long, 0.4 mm. wide, slightly narrowed toward base,
truncate or rarely retuse at the broad apex; perianth obovoid,
0.85 mm. long, 0.6 mm. wide, cuneate toward base, truncate
above and with a short broad beak, lateral keels sharp, antical
keel short and low, postical keel long, broad and sharply two-

124 LEJEUNEAE OF THE UNITED STATES AND CANADA

angled when young, the keels roughened from projecting cells or
sometimes with narrow rudimentary wings: & spikes occupying
short branches, oblong ; bracts closely imbricated in two to four
pairs, strongly inflated, slightly and subequally bifid with rounded
lobes and strongly arched keel; bracteoles similar to the under-
leaves but smaller, limited to base of spike: antheridia in pairs.

Type-locality, Salem, North Carolina (Schweinitz).

On rocks and trees. Connecticut (Hall, Evans); New York
(Peck); Pennsylvania (James, Lea); Delaware (James, Commons);
District of Columbia (Coville, Holzinger) ; Virginia (Mrs. Britton
and Miss Vail); West Virginia (Millspaugh); North Carolina
(Schweinitz) ; South Carolina (Green); Tennessee (Underwood);
Georgia (Underwood, Small); Alabama (Mohr); Mississippi
(Lloyd and Tracy); Missouri (Russell); Arkansas (Coville);
Louisiana (Langlois). |

Exsic.: Musc. Alleg. 262 (as Phragmicoma clypeata). Hep.
Bor.-Amer. 95 (as Phragmicoma clypeata) Hep. Amer. 50 ۰
(as Lejeunea (Archi-Lej.) clypeata). :

Archilej.unea clypeata and the two following species are very
closely related. They differ from typical members of the genus in
their pale color, less robust habit and short antheridial spikes. In
these peculiarities they show an approach to the genus Cheilole-

Jeunea, from which their undivided underleaves should probably E

exclude them. Recent writers on the hepaticae, however, tend to
ascribe less value to this character than was originally done by
Spruce, and species with undivided underleaves have been de-
scribed in the genera Euosmolejeunea, Cheilolejeunea and Pycnole-
Jeunea. Under these circumstances the generic position of A. c/y-

Peata and its immediate allies cannot be regarded as thoroughly | |

established.

In studying 4. clypeata, the writer has been able to examine
the type of Jungermannia clypeata from the Schweinitz herbarium,

the type of Zejeunea Dorotheae from the Lindenberg herbarium and _

a portion of the type of Z. calyculata from the Taylor herbarium.

The specimens in the Schweinitz herbarium labeled Jungermannia
transversalis have also been examined, as well as the plants which — |
Schweinitz sent to Torrey under this name. All of these various —

specimens evidently belong to the same species. In Z. Dorotheae
and in the plants labeled J. transversalis, the apex of the lobule

aM

ARCHILEJEUNEA SELLOWIANA 125

is sometimes composed of two superposed cells, instead of a single
cell as is usual in the species. They show, however, even on
robust stems, no other approach to either of the following species.

2. ARCHILEJEUNEA SELLOWIANA Steph.
PLATE 16, FIGS, 12-20

Phragmicoma xanthocarpa Aust. Hep. Bor.-Amer. 95b. ۰
Not Jungermannia xanthocarpa Lehm. & Lindenb.; Lehmann,
Pugillus, 5: 8. 1832 (Lejunea xanthocarpa Lehm. € Lindenb. ;
G. L. € N. Syn. Hep. 330. 1845).

Archilejeunea Sellowiana Steph. Hedwigia, 34: 62. 1895.

Lejeunea velata Gottsche ; Stephani, /. c. (as synonym).

Pale glaucous-green or whitish, growing in broad depressed
mats: stems prostrate, often radiating from a center, 0.17 mm. in
diameter, closely appressed to matrix, irregularly branched :
rhizoids abundant: leaves closely imbricated, the lobe widely
spreading almost at right angles with axis, ovate-oblong, some-
times slightly falcate, plane or slightly convex, 1.2 mm. long and
0.95 mm. wide on robust individuals, slightly revolute at the
rounded apex, margin entire or subcrenulate from projecting cells,
antical margin rounded at base and arching across or just beyond
axis; lobule somewhat inflated toward base, oblong, 0.6 mm.
long, 0.25 mm. wide, keel slightly arched, forming a distinct but
very obtuse angle with postical margin, free margin curved,
slightly inflexed near base, otherwise plane, apex acuminate,
sometimes hamately curved, the apical tooth mostly three to cight
cells long and two to four cells wide at base, ending in a single
row of two or three cells, sinus beyond apex deep and rounded
then passing very obliquely into postical margin, free margin of
lobule, except for apical tooth usually entire but, in rare Cases,

poorly developed ; cells of lobe convex with rather thick project-
ing walls and thin vertical walls, trigones large, acuminate to
obtuse, intermediate thickenings not infrequent in middle of lobe
and abundant toward base, mostly circular in outline, cells at edge
of lobe 12 p in diameter, in the middle and at the base 18 p: un-
derleaves contiguous to imbricated, broadly orbicular to reniform,
0.6 mm. long, 0.7 mm. wide, rounded, truncate or retuse at the
apex, rounded or subcordate at base, attached by a slightly curved
line of insertion : inflorescence autoicous: 9 inflorescence usually
borne on a short branch, more rarely on a leading branch, inno-

126 LEJEUNEAE OF THE UNITED STATES AND CANADA

vating one side, the innovation simple and sterile ; bracts compli-
cate, unequally bifid, sometimes with a narrow and rudimentary
wing along keel, the lobe oblong-ovate to obovate, suberect to
widely spreading, rounded at the apex, 0.9 mm. long, 0.6 mm.
wide, lobule oblong, rounded to subacute, 0.5 mm. long, 0.2 mm.
wide; bracteole oblong to obovate, 0.5 mm. long, 0.45 mm. wide,
narrowed toward base, truncate at the apex; perianth obovoid,
about half exserted, 0.9 mm. long, 0.7 mm. wide, cuneate toward
base, broad and truncate above and with a short, narrow beak,
lateral keels sharp, antical keel short and low, postical keel long,
broad and prominent, sharply two-angled when young and some-
times showing a short supplementary fold between the angles near
apex, keels slightly roughened from projecting cells: 4 spikes
occupying short branches, similar to those of A. clypeata : spores
brownish-green, angular, variable in length, with thickened walls
marked by minute verruculae and small orbicular patches of low,
radiating ridges, averaging 23 » in short diameter.

Type-locality, Brazil (Sellow).

On trees and rocks: Rhode Island (Bennett); Delaware
Commons); Virginia (Seymour, Mears); North Carolina (John-
son); South Carolina (Miss DuBois); Tennessee (Ruth, Scoville);
Georgia (Underwood); Florida (Austin, Farlow, Lighthipe, Un-
derwood, Straub, Nash); Alabama (Mohr, Earle and Baker,
Maxon and Pollard); Mississippi (Hall, Lloyd and Tracy); Louis-
iana (Langlois); Texas (Mohr).

Exsic.: Hep. Bor.-Amer. 95 b (as Phragmicoma xanthocarpa).

A. Sellowiana was first distinguished from A. clypeata by
Austin, who considered it identical with Lejennea xanthocarpa
Lehm. & Lindenb., a widely distributed tropical and subtropical
species occurring in both America and Africa. Austin distributed
the plant as ۸ xanthocarpa and it has since been
known in America under this name. Through the kindness
of Stephani, the writer has been enabled to study a specimen
of A. Sellowiana collected at Petropolis, Brazil, by Rudolph.
This specimen, which is quoted immediately after the type in the
original description, agrees very closely with the specimens from
the United States. It is probable, therefore, that the range of the
species extends through Mexico and Central America into Brazil.
It is worthy of note that the sterile type-specimens, collected by
Sellow, were first referred doubtfully by Stephani to 4. clypeata.*

* Hedwigia, 29: 20. 1890.

ARCHILEJEUNEA SELLOWIANA 127

Austin incorrectly referred to his “ Phragmicoma xanthocarpa,”
as synonyms, Jungermannia transversalis Schweinitz and Lejeunea
catenulata Nees.* The first of these has already been commented
upon. The second is a distinct species of the genus Veurolejeunea,t
and its presence in the United States has not been definitely estab-
lished. According to the synopsis it was found in “ Hymenophyllo
ciliato et Neckera abietina Hook. Americae septentrionalis." Un-
fortunately the genus Hymenophyllum is not known from the United
States, Æ. ciliatum itself being found in tropical America. Weck-
era abietina (Alsia abietina Sull.), on the other hand, is found in
California and northward along the Pacific coast, a region from
which no Lejeuneae have been recorded except the recently de-
scribed Cololejeunea Macounii of British Columbia. Under the cir-
cumstances we must exclude Veurolejennea catenulata from our list.

The most important differences between A. Sellowiana and A.
clypeata are drawn from their vegetative organs. A. Sellowiana is
the more robust of the two species; the lobes of its leaves spread
more widely and are at the same time more falcate ; the lobule is
oblong instead of ovate-triangular in outline, and its apex is acu-
minate instead of apiculate; the cells of the lobe have more con-
spicuous trigones and more numerous intermediate thickenings,
the latter being practically absent in A. clypeata, its much
broader underleaves finally are more crowded together and are
usually subcordate at the base. The perichaetial bracts and
perianths are so variable in both species that the slight differences
between them are of little diagnostic value. In A. Sellowiana,
however, the female flower is usually borne on a short branch,
while in A. elypeata it is usually borne on a leading branch. Un-
fortunately even this difference is inconstant. The perigonial
spikes are essentially the same in both species.

In the true 4. ranthocarpa (Lehm. & Lindenb.) Steph., the lobe
is much more strongly convex than in A. Sellowiana, and the mar-
gin is revolute both at the apex and along the entire postical side.
In the lobule also the free margin is strongly involute for the
greater part of its length. The leaf-cells are similar in the two

* Syn. Hep. 323. 1845.
+ Spruce, Hep. Amaz. et And. 84. 1884; Schiffner ; Engler & Prantl, Nat. Pflan-
zenfam. 13: 13%; 1893.

128 _LEJEUNEAE OF THE UNITED STATES AND CANADA

species, but intermediate thickenings in A. xanthocarpa are very |
infrequent.
3. Archilejeunea conchifolia sp. nov.
PLATE 17, FIGS. 1-9

Pale yellowish-green or whitish, growing in depressed mats: _
stems prostrate, 0.1 mm. in diameter, closely appressed to matrix,
irregularly branched: rhizoids abundant: leaves densely imbri-
cated, the lobe obliquely spreading, not falcate, ovate, very con-
vex, 0.7 mm, long, 0.45 mm. wide, strongly revolute at the
rounded apex and along the postical margin, antical margin slightly
rounded at base, arching across or just beyond axis, margin entire
or very vaguely and sparingly crenulate ; lobule inflated, oblong,
0.45 mm. long, 0.17 mm. wide, keel more or less arched, usually
forming a continuous line with the revolute postical margin but
sometimes slightly indented at junction with postical margin, free
margin curved, involute to or beyond apex, sinus beyond apex
deep and rounded then passing very obliquely into the revolute
postical margin, apex acuminate, the apical tooth mostly three to
six cells long and three or four cells wide at base, ending in a
single row of two or three cells; cells of lobe slightly papillose,
with thickened free walls and thin vertical walls, trigones small,
intermediate thickenings very scanty near base of lobe, otherwise
wanting, cells at edge of lobe 14 A in diameter, in the middle and
at the base 20 p: underleaves contiguous or imbricated, broadly

obliquely or widely spreading, rounded at the apex, 7
mm. long, 0.45 mm. wide, lobule narrowly ovate to lingulate, ob-
tuse to acute at the apex, 0.45 mm. long, 0.17 mm. wide ; bracteole _
obovate, 0.5 mm. long, 0.35 mm. wide, narrowed toward base,

eels sharp, antical keel distinct, about half as long as perianth,

postical keel long, broad and sharply two-angled when young,
keels smooth or nearly so: 4 spikes occupying short branches, |
similar to those of 4. clypeata.

Type-locality, Orange Bend, Florida (Underwood, 60).

MASTIGOLEJEUNEA AURICULATA 129

On trees. South Carolina (Ravenel, Miss Du Bois); Alabama
(Baker); Florida (Underwood).

Exsic.: Hep. Amer. 50 f. p. (as Lejeunea (Archi- Lej.) clypeata).

A. conchifolia is of about the same size as A. clypeata, and
agrees with it in the position of its female flowers. It differs in its
obliquely spreading, convex lobes with revolute apex and margin,
in its oblong lobule with acuminate apex and in its imbricated un-
derleaves, usually broader than long. From 4. Sellowiana it dif-
fers mainly in its smaller size and in the characters of its convex
leaf-lobes, which are not at all falcate.

A. conchifolia is perhaps closer to A. xanthocarpa than is the
preceding species, agreeing with it in its convex leaf-lobes with
revolute postical margins. It is, however, much less robust, its
lobes spread more obliquely, its lobules are relatively larger and
their apices can usually be made out without dissection.

4. MASTIGOLEJEUNEA AURICULATA (Wils. & Hook.) Schiffn.
PLATE 17, FIGS 10-19

1 auriculata ۰ Hook.; Drummond, Musc.
Amer. St. Merid. 170. 1841.

Phragmicoma versicolor Lehm. & Lindenb. ۸۰ p.; G. d & N.
Syn. Hep. 297. 1845.*

Phragmicoma teretiuscula Lindenb. & Gottsche, ۵۰ p., l. c. 745.
1847.

Lejeunea auriculata Sull.; Gray, Manual, Ed. II., 699. 1856.

Ptycocoleus auriculatus Trevis. Mem. r. Ist. Lomb. III. 4: 405.
1877.

A Phy agmicoma ‘versicolor of the si A re col-

lected at Madras by Wight and on Drummond’s Louisiana specimens, the type of
In ye supplement to this work, the authors
d

ep. 301) and the Mexican specimens to P. versicolor.
Notes that thé name ‘‘ versicolor,’ although finally restricted by the authors of the

ata." According to the strict rules of priority,
replace the name **repleta" of Taylor, but apparently there is enough reason for
giving it up altogether. Of course the name ۴ teretiuscula’’ has no claims whatever.

130 LEJEUNEAE OF THE UNITED STATES AND CANADA

Lejeunea (Mastigo-Lejeunea) auriculata Spruce, Hep. Amaz. et
And. 102. 1884.

Mastigolejeunea auriculata Schiffn.; Engler € Prantl, Nat.
Pflanzenfam. 1°: 129. 1893.

Glaucous green, varying to brownish or purplish in older parts,
growing in depressed tufts : stems prostrate or pendent, 0.15 mm.
in diameter, arising from a creeping caudex, irregularly pinnate or
on 2 plants rarely dichotomous, branches widely spreading, some-
times small-leaved and flagelliform : rhizoids scanty : leaves densely
imbricated, closely appressed when dry, widely spreading and sub-
squarrose when moist, the lobe ovate-oblong, more or less falcate,

mm. long, 0.5 mm. wide, apex usually obtuse, sometimes
rounded or subacute, margin entire or subrepand, antical margin

and shallow, passing very obliquely into postical margin of lobe,
lobule on robust stems and branches often appearing simply as a

mediate thickenings occasionally present toward base of lobe, cells
at edge of lobe 8» in diameter, in the middle 21 x 12 p, at the base
23 X 144: underleaves imbricated, broadly orbicular, o.5 mm.
long, 0.55 mm. wide, with a broad, median ridge, truncate of
slightly retuse at apex, cuneate toward base, indistinctly decurrent
or minutely auriculate at base and attached by a slightly curved
line of insertion, margin irregularly sinuate, more or less revolute
at apex and near base : inflorescence autoicous or dioicous: 9 in-

complicate, unequally bifid, the lobe ovate, suberect to widely
spreading, 1.2 mm. long, 0.6 mm. wide, the apex usually subacute,
but varying from rounded to acuminate, margin irregularly sinuous,
lobule oblong to obovate, bluntly pointed, truncate or retuse at
apex, 0.7 mm. long, 0.35 mm. wide; bracteole 0.8 mm. long, 0.7
mm. wide, obovate from a slightly narrowed base, at the apex
rounded, truncate

line, about one third exserted, 1.5 mm. long, 0.6 mm. wide, rounded
at base and at apex, beak short and broad, strongly compressed and

y

BRACHIOLEJEUNEA CORTICALIS 131

with a high, narrow, postical keel, antical surface plane or nearly so,
keels not winged: 3 inflorescence borne on 2 plants, occupying
ordinary branches or innovations, usually proliferating and some-
times repeatedly floriferous ; bracts in six to twelve pairs, similar `
to the leaves, but with smaller, less widely spreading lobes, more
strongly arched keels and larger, more inflated lobules, truncate
to retuse at apex; bracteoles similar to underleaves, found
throughout spike ; antheridia borne singly : spores greenish, thick-
walled, minutely verruculose, averaging 34, in short diameter.

Type-locality, New Orleans and Louisiana (Drummond).

On trees. Florida (J. D. Smith, Underwood, Straub) ; Ala-
bama (Mohr) ; Louisiana (Drummond, Langlois). The species is
widely distributed in tropical America.

Exsic.: Musc. Amer. St. Merid. 170 (as Jungermannia auri-
culata). Hep. Amer. 133 (as Zejeunea (Mastigo-Lej.) auriculata).

Mastigolejeunea auriculata belongs to a tropical and subtrop-
ical genus which has no other known representatives in the United
States. The species bears, however, a certain superficial resemb-
lance to Brachiolejeunea corticalis, with which it has been confused.
The latter species is readily distinguished by its crenulate lobules,
larger leaf-cells, usually double innovation and pluriplicate perianth.

According to Spruce,* M. auriculata can scarcely be distin-
guished from the Javan M. humilis (Gottsche) Schiffn., to which
he would also reduce M. repleta (Tayl.). Stephani f states that
all three species are amply distinct. The writer has examined a
specimen of M. humilis and finds that this species, although very
close to M. auriculata, differs in its smaller size, in its much less
variable lobule, which ends in a long outwardly curved tooth
consisting of three or eight cells in a single row, in its more
emarginate underleaves and in the broader lobes and lobules of its
perichaetial bracts. According to Stephani M. repleta is distin-
guished by its apiculate leaves.

5. BRACHIOLEJEUNEA CORTICALIS (Lehm. € Lindenb.) Schiffn.
PLATE 18, Fics. 1-11
Jungermannia corticalis Lehm. & Lindenb. ; Lehmann, Pugil-
lus, 4: 50. 1832.

* Hep. Amaz. et And. 103. 1884.
T Hedwigia, 29: 9. 1890.

132 LEJEUNEAE OF THE UNITED STATES AND CANADA

Frullania Leprieuri Nees & Mont. ; Mont. Ann. des Sc. Nat.
IE 14:333. M. 20. را‎ 2.1840;

Phragmicoma corticalis Lehm. & Lindenb. ; G. L. & N. Syn.
Hep. 297. 1845. jd

Phragmicoma melanophloea Mont. & Nees, 7 c. 297 (as .—
synonym). :

Lejeunea linguaefolia Tayl. Lond. Jour. Bot. 5: 390. 1846.
G. L. & N. Syn. Hep. 754. 1847 (fide Stephani).

Phragmicoma Leprieurii Mont. Sylloge, 86. 1856.

Ptychocoleus corticalis 'Trevis. Mem. r. Ist. Lomb. III. 4: 405.
1877.
Lejeunca (Brachio-Lejeunea) corticalis Steph. H edwigia, 29: 8.
1890. ur
Brachiolejeunea corticalis Schiffn. Hedwigia, 33: 180. 1894 7
(in obs.).

: , in entire, antical margin curved, à
Ing across axis and rounded or subcordate at base, postical margin |
of keel; lobule strongly inflated ，

cells at edge of lobe 1 3 # in diameter, in the middle 2 3 x 16 A, at
the base 28 x 23 u: underleaves slightly imbricated, orbicular-quad- _
rate to reniform, 0.45 mm.- ong, 0.45 mm. wide, truncate or retuse -
at the broad apex, sides parallel or converging toward base : inflor-
escence dioicous: 9 inflorescence borne on a leading branch, usually
innovating on both sides, more rarely on one side, the innovations

,

usually once or twice floriferous ; bracts complicate, unequally bifid,

HARPALEJEUNEA OVATA (8

keel broadly winged in lower half, the wing entire, semicircular to
semiovate in outline, lobe ovate, obliquely spreading, rounded at
apex, subcordate at base, irregularly sinuous on margin, 0.95 mm.

wide, rounded at apex and with a short, broad beak, somewhat
compressed in upper half with sharp lateral keels, postical keel
broad, sharply two-angled and with a short, supplementary fold
near apex, antical keels one to three, low and short: ¢ spikes
spikes borne on leading branches, usually proliferating and repeat-
edly floriferous, bracts in about six pairs, less spreading than the
leaves, the lobe not falcate, lobule nearly as long as lobe but nar-
rower, inflated, apex blunt to subacute, margin eñtire; bracteoles
similar to underleaves, found throughout spike.

Type-locality, Jamaica (Herb. Lehm. & Lindenb.).

Ontrees. Lake Worth, Florida (Underwood). The species
has a rather wide distribution in tropical America.

Brachiolejeunea corticalis has not before been recorded from the
United States. The Florida specimens, however, agree closely
with the type and also with the specimens from Cuba, distributed
in Hep. Amer. 144. There is no danger of confusing the species
with any of our other Lejeuneae except possibly Mastigolejeunea
auriculata, under which the most important differences have already
been enumerated.

The genus Brachiolejeunea has several representatives in trop-
ical America, and among them 2. bicolor (Nees) Schiffn. is perhaps
most closely related to B. corticalis. B. bicolor is a more robust
plant, the lobes of its leaves tend to be more sharply pointed, its
underleaves are more or less auriculate at base and its perianth is
regularly pluriplicate, usually with ten distinct folds in the upper
part.

6. HARPALEJEUNEA OVATA (Hook.) Schiffn.

Jungermannia. serpyllifolia $ ovata Hook. Brit. Jung. 1816
(not figured).

Lejeunca ovata Tayl.; G. L. & N. Syn. Hep. 376. 1845.

Lejeunea (Harpa-Lejeunea) ovata Spruce, Hep. Amaz. et And.
165. 1884,

134 . LEJEUNEAE OF THE UNITED STATES AND CANADA

Lejeunea Molleri Steph. Hedwigia, 26: 3. pl. 1.f. 2. 1887.
Harpalejeunca ovata Schiffn.; Engler & Prantl, Nat. Pflanzen-
fam. i: 729: 1807.

Pale to dark green, loosely depressed-caespitose or scattered i
among other hepatics: stems prostrate, 0.05 mm. in diameter, -

closely appressed to substratum, irregularly branched : rhizoids

ew: leaves contiguous or somewhat imbricated, the lobe con- f

vex, obliquely to widely spreading, falcate-ovate, 0.3 mm. long,

5 mm. wide, gradually narrowed beyond middle, apex often
reflexed, mostly acute or short-acuminate, rarely obtuse, margin
subentire or irregularly sinuate, antical margin curved, arching

unequally bilobed, antical lobe oval, obtuse, postical lobe similar

but smaller. i
Type-locality, Ireland (Taylor).

PIG A ee ee
sc

EUOSMOLEJEUNEA DURIUSCULA 135

On trees and fallen logs. Southern States (Sullivant, fide
Austin); Virginia (Mrs. Britton and Miss Vail); Tennessee
(Ruth); Georgia (Underwood). The species also occurs in
western Europe but is rare. Pearson * states that it grows ón
shaded rocks but according to Camus + it is also found on bark.

Harpalejeunea ovata has been illustrated three times within
recent years. Stephani's figure, already noted in the synonymy,
was drawn from specimens collected in Portugal. The writer has °
illustrated the specimens from Virginia, ? while Pearson has pub-
lished figures drawn from British material.$ The perianth, known
only from the brief description in the Synopsis Hepaticarum, does
not appear in any of these illustrations.

The American specimens agree closely with the type-material
in the Taylor herbarium but show a greater variability in the
shape of the bracteole than European writers describe. There
are no male specimens among them, and the description given
above of the perigonial spike is taken wholly from Pearson.

There is no other species from the United States with which
H. ovata ought to be confused. The acute lobes of its leaves and
its emarginate-bifid underleaves with rounded divisions are unique
among our Lejeuneae. The species, however, finds several close
allies in tropical regions, where the genus /arpalejeunea is largely
represented.

7. Euosmolejeunea duriuscula (Nees)
PLATE 18, Fics. 12-13

Jungermannia serpyllifolia Wils. & Hook. p. p. Drummond,
Musc. Amer: St. Merid. 171. 1841.

Lejeunea duriuscula Nees; G. L. & N. Syn. Hep. 364. 1845.

Lejeunea flexuosa Lindenb. /. c. 385. 1845 (fide Stephani).

Lejeunea serpyllifolia var. Sull. Musc. Alleg. 273. 1845.

Lejeunea Sullivantiae Aust. Bull. Torrey Club, 3: 15. 1872.

Lejunea Austini Lindb. Acta Soc. Sci. Fenn. 10: 489. 1875.

Lejeunea (Cheilo-Lejeunea) duriuscula Spruce, Hep. Amaz. et
And. 259. 1884.

* Hep. Brit.(Isles, 43. 9۰
f Bull. Soc. Bot. de France, 47: 201. 1900.
T ito : Vail, Mem. Torrey Club, 4: ۸. 87. 1894.
àL. E pl, 8.

136 LEJEUNEAE OF THE UNITED STATES AND CANADA

Lejeunea (Euosmo-Lejeunea) duriuscula Steph. Hedwigia, 29:

80. 1890.
Lejeunea (Eu-Lejeunea) Underwoodii Steph. Bot. Gazette, 17:
17 h 4802,

Cheilolejeunea duriuscula Schiffn.; Englers Bot. Jahrb. 23:
592. 1897 (in obs.).

Yellowish-green, sometimes tinged with brownish, depressed-
caespitose : stems prostrate, loosely appressed to substratum,
0.08 mm. in diameter, sparingly and irregularly branched : rhizoids
scanty : leaves slightly imbricated, the lobe convex, obliquely to
widely spreading, falcate-ovate, 0.4 mm. long, 0.35 mm. wide,
rounded or very obtuse at the apex, margin slightly and irregu-
larly angular-sinuate, antical margin arching across or just beyond
axis, rounded at base, postical margin forming an obtuse angle
with keel; lobule strongly inflated, triangular-ovate, measuring
when inflated 0.1 mm. long, 0.08 mm. wide, keel slightly arched
or almost straight, somewhat decurrent, free margin strongly in-
volute to beyond the obtusely pointed or apiculate apex, then
obliquely truncate to end of keel ; cells of lobe plane or slightly
convex, with somewhat thickened walls and large, triangular
trigones, averaging 13 ۸ in diameter at edge of lobe, 16 م‎ in the
middle and 21 y at base: underleaves distant, ovate to orbicular,
0.17 mm. long, 0.15 mm. wide, bifid about one half with acute,
slightly spreading lobes and narrow usually acute sinus, cuneate
and slightly decurrent at base, sometimes very minutely auricu-
late : inflorescence dioicous: 9 inflorescence borne on a leading
branch, innovating on one or sometimes on both sides, the inno-
vations long, often floriferous ; bracts complicate, unequally bifid,
the lobe suberect to widely spreading, ovate to falcate-ovate,
rounded to obtuse at the apex, margin as in leaves, 0.5 mm. long,
0.3 mm. wide, lobule ovate to lanceolate, rounded to acuminate
at the apex, 0.17

,

pairs, very concave and strongly inflated, shortly and subequally
bifid with a strongly arched keel and rounded or obtuse lobes,
slightly crenulate at apex ; bracteoles similar to underleaves but
smaller, present only in lower part of spike.

EuosMOLEJEUNEA DURIUSCULA 137

Type-locality, Surinam (Curie).

On trees or rotten logs, often growing with mosses or other
hepatics. Florida (J. D. Smith, Farlow, Underwood) ; Alabama
(Sullivant, Mohr); Mississippi (Langlois, Lloyd and Tracy); Louisi-
ana (Drummond, Featherman, Langlois, Faxon). The species has
a wide distribution in tropical America.

Exsic.: Musc. Amer. St. Merid. 171 f. p. (as Jungermannia `
serpyllifolia). Musc. Alleg. 273 (as Lejeunea serpyllifolia var.).
Hep. Bor.-Amer. 96 (as Lejeunea Sullivantiae). Hep. Amer. 135
(as Lejeunea (Eu-Lej.) Underwoodit).

Lejeunea Sullivantiae Aust. has been more or less of a puzzle
to hepaticologists, and yet the original description, although brief,
is not misleading. The specimens distributed by Austin and by
Sullivant, moreover, are both abundant and characteristic. Un-
fortunately Austin makes no note in his description of the female
flowers, although they are fairly abundant in the material which he
studied. The probability is that he overlooked them because they
had developed no perianths.

A few years after Z. Sullivantiae was published, Lindberg
studied the specimens in Musc. Alleg. 273 and Hep. Bor.-Amer.
96 and proved that they were identical with the dark tufts in No.
171 of Drummond's exsiccata. He also stated that they could
not be distinguished from Z. ¢hymifolia Nees, ? y of the Synopsis
Hepaticarum,* a plant which he knew from male specimens col-
lected in French Guiana by Leprieur. Z. thymifolia has since
been reduced to Z. fava Swartz,t and the questionable variety 7
is referred by Stephani to L. duriuscula, a species with which
Lindberg was apparently unfamiliar Objecting to the name “ Z.
Sullivantiae”” on account of the older Z. Sullivantii Gottsche,$
Lindberg gave to Austin's species the name Z. Austini, and pre-
sumably referred to it the var. 7, to which allusion has just been
made.

M LL Hep. Amaz. et. And. 271; Schiffner, Conspect. Hep. Arch. Ind. 249.
1898
f Hedwiga, 29: 83. 1890. It should perhaps be noted that Stephani's speci-
mens of var. y were collected by Beyrich in Brazil and are the plants mentioned in the
Synopsis. Gottsche (Mex. Leverm. 218. 1863) refers these same speci imens to Z.
Seren: Lehm. & Lindenb. = Hep. 382).
x. Leverm. 196. 6

138 LEJEUNEAE OF THE UNITED STATES AND CANADA

The validity of Z. Austini, so far as the specimens from the a

United States are concerned, is recognized both by Underwood and
by Stephani. The latter author first referred it to the subgenus

Eu-Lejeunea, but at present considers it a Chellolejeunea. With it —

he compares his Zu-Lejeunea Underwoodii, from Florida, a plant
which he is now also inclined to place in Cheilolejeunea. It is evi-

dent from the above statements that Z. duriuscula, L. Austini and.

L. Underwoodu are all considered distinct species by Stephani.
The writer has examined all the specimens which Lindberg re-
ferred to L. Austini with the exception of those from Guiana. He

has also examined the specimens of Z. Underwoodii distributed by .

Underwood and Cook and those of Euosmolejeunea duriuscula in
Hepaticae Spruceanae, collected near Pará, in Brazil. All of these
plants show female flowers without perianths, and there are no
male plants among them. Asa result of this study it would ap-
pear that all of these specimens are referable to the same species,

for which the oldest available specific name, “ duriuscula," should
be retained. The differences stated by Stephani in his published —

papers and in his manuscript notes are not always apparent and
seem hardly sufficient to separate the species. He states, for eX-

ample, that Z. Austin! has much smaller leaf-cells than Z Under- 4

woodii, and that the lobule of the latter species is inflated and has

an arched keel, whereas in Z. Austini the lobule is not inflated 3

and the keel is approximately straight or concavely arched. He
a

states also that the bracteole is free in Z. Underwoodii and slightly

connate on one side in Z. Austini. The first of these differences —

p

does not hold, since the cells of Z. Austini average fully as large ;

as those of L. Underwoodii. The second difference is unreliable,

because convexly arched, straight, and concavely arched keels may |
often be found on a single stem, and the degree of inflation de- —

pends largely upon the degree of development. The third differ- 5
ence is at best inconstant as the bracteole is uniformly free in the - |
specimens of L. Austini examined by the writer. Stephani has -
published no descriptive notes on Euosmolejeunea duriuscula but —
has kindly furnished a drawing of this species made from the type 1
material. This agrees essentially with the specimens distributed a
by Spruce. In reducing L. Austini and L. Underwoodii to synon- |
ymy the writer regrets differing from so eminent an authority aS ，

EUOSMOLEJEUNEA OPACA 139

Stephani but feels compelled to follow the evidence drawn from the
specimens themselves.

Opinions have varied with regard to the systematic position of
E. duriuscula. By Spruce it was first placed in Chezlo-L ejeunea,
a disposition now accepted by Stephani and by Schiffner. Spruce
afterwards transferred it to Zuosmolejeunca, thereby agreeing with
an earlier view of Stephani. The plant agrees with Enosmolejeunea
in its color and general habit, in its leaves and cell-structure and
in its perianth. Its underleaves, however, are smaller than is
usual in the genus and lack the basal cordations which normally
occur. Since several species of Zuosmolejeunea with equally small
underleaves have been described, this difference by itself should
hardly suffice to exclude Æ. duriuscula. The species certainly has
less in common with either Chetlolejeunea or Lejeunea proper than
it has with Zuosmolejeunea. :

E. duriuscula seems to be limited with us to the States border-
ing the Gulf of Mexico. The specimens from Kirkville, New York,
quoted by\Underwood as ZL. Austini * are referable to 0
cavifolia. These specimens were determined by Spruce, who had
apparently been misled by plants incorrectly referred to Z. Austini.
In addition to Z. cavifolia, no fewer than four of our other Le-
jeuneae have been found in herbaria incorrectly determined as Z.
Austini. Our only species, however, which it at all closely re-
sembles is Euosmolejeunea opaca, under which the differential char-
acters will be noted.

8. EUOSMOLEJEUNEA OPACA (Gottsche) Steph.
PLATE 19, Fiss. 1-11

Lejeunea opaca Gottsche ; G. L. & N. Syn. Hep. 362. 1845.

Lejeunea (Omphalanthus ?) Mohri Aust. Bull. Torrey Club,
052001875
i Lejeunea (Euosmo- Lejeunea) laxiuscula ne Hep. Amaz. et
And. 244. 1884.

Lejeunea (Euosmo-Lejeunea) opaca Steph. Hedwigia, 29: 79.
1890.

Euosmolejeunea opaca Steph. Bihang Svenska. Vet.-Akad.
Handl. 23°: 23. Bon hs d ur e pO I
TU mE Torrey Clob, 19: 299. 1892. :

140 LEJEUNEAE OF THE UNITED STATES AND CANADA

Yellowish-green, growing in depressed tufts: stems prostrate,
loosely appressed to substratum, 0.1 mm. in diameter, sparingly
and irregularly branched: rhizoids scanty: leaves loosely imbri-
cated, the lobe convex, obliquely spreading, broadly ovate, 0.5
mm. long, 0.4 mm. wide, rounded at the apex, margin entire or
sparingly crenulate from projecting cells, antical margin arching
across or just beyond axis, rounded at the base, postical margin
forming an angle a little greater than a right angle with keel;
lobule strongly inflated, triangular-ovate, 0.15 mm. long, 0.1 mm.
wide, keel slightly arched, free margin strongly involute to beyond
the subacute apex, then obliquely truncate or lunulate to end of keel;
cells of lobe plane or convex, with somewhat thickened walls and
large, triangular trigones, averaging 13 ¡1 in diameter at edge of lobe,
18 # in the middle and 25 x 18 p at the base: underleaves con-
tiguous or imbricated, orbicular, 0.35 mm. long, bifid about one
third with obtuse or acute, erect or subcontiguous lobes separated
by an obtuse or acute sinus, rounded or cordate at base and attached
by a long, curved line of insertion, margins entire : inflorescence
dioicous: 9 inflorescence borne on a short branch, innovating on
one side with a short, simple innovation, very rarely without in-
novation ; bracts complicate, unequally bifid, the lobe erect-spread-
ing, unsymmetrically ovate, usually rounded at the apex, but
varying to apiculate or acute, 0.6 mm. long, 0.35 mm. wide, lobule
variable, mostly oblong and rounded at the apex, 0.17 mm. long, -
0.07 mm. wide ; bracteole free, orbicular to oblong, 0.45 mm. long,
0.4 mm. wide, bifid about one third with acute or subacute, erect
lobes and narrow sinüs; perianth about half exserted, oblong-ob-
ovoid, 0.9 mm. long, 0.45 mm. wide, gradually narrowed toward
base, rounded at the apex and with a short beak, lateral keels
sharp, postical keel about as long as perianth, broad and sharply
two-angled, antical keel short, in upper part of perianth: 4 spike
borne on a short branch, sometimes proliferating ; bracts and
bracteoles similar to those of E. duriuscula, but the lobes of the
bracts not crenulate at apex,

` Type-locality, Surinam (Splitgerber).

On rotten logs. Florida (J. D. Smith, Lighthipe, Underwood,
Faxon) ; Alabama (Mohr). Widely distributed in tropical America.

Exsıc.: Hep. Amer. 1 34 (as Lejeunea ( Euosmo- Lej.) trifaria).
: The abundant specimens of E. opaca, collected by Underwood
in Florida, were at first confused with Æ trifarta (Nees) Schiffn.
This species 1s even more widely distributed, being found in the
tropics of both hemispheres, and وز‎ certainly to be expected in the
southern parts of Florida. It is a more robust plant than Æ. opaca, —

CHEILOLEJEUNEA POLYANTHA 141

and differs from it in its autoicous inflorescence and in its leaf-cells,
which are more convex on the free surface of the lobe and show
much more conspicuous trigones.

The type-specimens of Zejeunea Mohrii in the Austin herba-
rium are apparently referable to Æ. opaca, although sterile and in
rather poor condition. They do not agree fully with Austin’s de-
scription because they show bifid underleaves, but this is a charac-
ter which might have been overlooked.

E. opaca differs from E. duriuscula in its much larger under-
leaves, distinctly rounded or cordate at the base, and in the posi-
tion of its sexual organs, which are borne on short branches rather
than on leading branches. Occasionally one meets with specimens
of E. opaca in which the underleaves are smaller than is normal, but
in such cases a careful search will commonly show underleaves of
the usual size on some of the axes.

At one time Spruce * threw doubt on the validity of Æ. opaca,
suspecting that it was synonymous with Zgjeunea clausa Mont.
Later he apparently changed his mind, for he distributed speci-
mens under both of these names in his Hepaticae Spruceanae.

9. Cheilolejeunea polyantha sp. nov.
PLATE 19, FIGS. 12-21

| green or brownish, growing in depressed tufts: stems
` prostrate, loosely appressed to substratum, 0.1 mm. in diameter,
copiously and irregularly pinnately branched, sometimes dichot-
omous: rhizoids scanty: leaves densely imbricated, the lobe
convex, obliquely spreading, orbicular, 0.4 mm. long, rounded at
the apex, margin entire or slightly and irregularly angular-sinuate,
antical margin arching across axis and rounded at base, postical
margin usually rounded, forming an acute or right angle with
keel; lobule strongly inflated, ovoid-cylindrical, 0.12 mm. long,
0.08 mm. wide, keel slightly arched, free margin strongly involute,
as far as the bluntly pointed apex, then abruptly or obliquely
truncate to end of keel; cells of lobe slightly convex, with
somewhat thickened walls and distinct, triangular trigones, aver-
aging 12 p in diameter at edge of lobe, 21 pin the middle and
23 4 at the base: underleaves imbricated, sometimes densely so,
rapidly increasing in size toward a 9 inflorescence, plane, broadly
orbicular to reniform, 0.25 mm. long, 0.35 wide, bifid about one

* Hep. Amaz. et And. 245, 1884.

142 LEJEUNEAE OF THE UNITED STATES AND CANADA

half the distance from apex to the strongly curved line of inser-
tion, with broad, erect, triangular, obtuse to acute or apiculate
lobes and obtuse to acute sinus, rounded to cordate at base,
margin entire: inflorescence dioicous : $ inflorescence sometimes
on a leading branch, sometimes on a short branch, innovating on
one or occasionally on both sides, the innovations usually florifer-
ous and sometimes repeatedly so, occasionally sterile ; bracts
complicate, deeply and unequally bifid, the lobe obliquely spread-
ing, ovate, rounded or sometimes apiculate at apex, margins as In
leaves, 0.7 mm. long, 0.4 mm. wide, lobule ovate to lanceolate,
rounded to acuminate at the apex, 0.25 mm. long, 0.15 mm. wide:
bracteole free, oblong-elliptical to oval, convex (postically), 0.7
mm. long, 0.5 mm. wide, bifid one fourth to one third with tri-
angular, acute or apiculate lobes and broad obtuse to acute sinus:
perianth about half exserted, obovoid, 0.8 5 mm. long, 0.55 mm.
wide, cuneate toward base, gradually narrowed toward the
rounded apex and with a short beak, antical surface plane or with
a low and indistinct keel near apex, postical keel long, broad and
bluntly two-angled, lateral keels slightly compressed: 4 inflo-
rescence not seen: spores measuring 16 u in short diameter, elon-
gated, irregularly angular, thick-walled, with small ridge-like thick-
n ngs, arranged radiately in circles.

On bark. Eustis, Florida (Underwood, 1380 /.2.), the type-
locality.

Cheilolejeunea polyantha resembles to a considerable extent the
South American Pycnolejeunea papulosa Steph.,* a species described
by Spruce under the name Lejeunea (Pyeno- Lejeunea) contigua, vat.
latifolia.+ The new species, however, is much less robust than
P. papulosa, its leaves are less convex and its leaf-cells have thin-
ner walls and less conspicuous trigones; its underleaves are rela-
tively longer, and its perianth has an obsolete antical keel and a
longer beak.

The closely imbricated leaves and underleaves of C. polyantha
will at once distinguish it from any of our other Lejeuneae Schizo-
stipae, although in texture and color it somewhat resembles Zuos-
molejeunea opaca and E. duriuscula. From the first of these, it
differs also in its more widely spreading and narrower lobules, in
its broader underleaves, and in its broader perianth and perichaetial
ciabatta cima eer N cii i E

* Hedwigia, 35: 126. 1
H

896.
ep. Amaz. et And. 248. 1884. Afterwards distributed in Hep. Spruceanae
as a distinct species under a manuscript name,

CHEILOLEJEUNEA PHYLLOBOLA 143

bracts. From the second it differs more particularly in its larger
size and in its much larger underleaves, which are rounded or sub-
cordate at the base.

IO. CHEILOLEJEUNEA PHYLLOBOLA (Nees & Mont.) Schiffn.
PLATE 20, Fics. 1-13

Lejeunea phyllobola Nees & Mont.; Ramon de la Sagra, Hist.
fis. pol. y natur. Cuba, 9: 281. ras: Ax L N. Syn. Hep.
369. 1845.

Lejeunea (Hygro-Lejeunea) phyllobola Steph. Hedwigia, 29:
81. 18090.

Chetlolejcunea phyllobola Schiffn. Engler's Bot. Jahrb. 23: 591.
1897.

Hygrolejeunea phyllobola Massal. Mem. dell’ Accad. di Verona,

73:—(33). Pl. 9. f. 12. 1897.

ale green, growing in depressed mats: stems prostrate,
closely appressed to substratum, 0.07 mm. in diameter, irregularly
and rather densely pinnate, the branches sometimes upright and
flagelliform, leafless except at apex, but with well-developed un-
derleaves: rhizoids somewhat abundant: leaves loosely imbri-
cated, the lobe obliquely to widely spreading, orbicular-ovate,
0.5 mm. long, 0.45 mm. wide, rounded at the apex, margin entire
or very slightly and irregularly sinuate from projecting cells,
antical margin arching across axis, but almost straight near base,
postical margin straight or slightly curved, forming a very obtuse
angle with keel; lobule inflated, ovate, 0.1 mm. long, 0.07 mm.
wide, keel almost straight, free margin involute to or just beyond
apex, then obliquely truncate to end of keel, apex usually tipped
with a single rounded cell, sometimes acute or even ending in a row
of two or three cells, lobule often rudimentary and explanate ; cells
of lobe slightly convex, thin-walled, but with distinct trigones and
occasional intermediate thickenings, the latter sometimes very

crenulate, rarely with an angular tooth near the middle of one or
both sides: inflorescence autoicous : 9 inflorescence usually borne

144 _LEJEUNEAE OF THE UNITED STATES AND CANADA

on a short branch, rarely on a leading branch, innovating on one
side, the short innovation sometimes sterile, sometimes bearing
another 9 or a 8 inflorescence, rarely absent; bracts unequally
bifid, complicate, the lobe widely spreading, 0.45 mm. long, 0.35
mm. wide, ovate to orbicular, more or less falcate, rounded at
apex, margin as in leaves, postical margin strongly curved, lobule
ovate to lanceolate, rounded to acuminate at the apex, 0.3 mm.
long, 0.1 mm. wide; bracteole free, ovate, 0.35 mm. long,
0.25 mm. wide, bifid about one third with erect, acute di-
visions and acute sinus, margins indistinctly angular-crenulate ;
perianth about half exserted, obovoid but distinctly cuneate toward
base with almost straight sides, 0.6 mm. long, 0.4 mm. wide,
truncate or slightly retuse at the broad apex and with a short

Type-locality, Cuba (Herb. Montagne),

On trees. Florida (Underwood). The species is now known
from several of the West Indian Islands, from Mexico and from
Costa Rica. It has also lately been reported from China.

The determination of the present species is based on the speci- —
mens in Wright's Hepaticae Cubenses. These were named by ~
Gottsche, and the determination was afterwards confirmed by Schiff >
ner. The type specimens have not been consulted by the writer.

C. phyllobola is remarkable for the insecure way in which its
side-leaves are attached to their axes, the slightest force being
sometimes sufficient to break them away. As a result leafless
branches are frequent, although these are much less characteristic
than in C. versifolia. In its leaf-cells the species exhibits consid-
erable variation, the trigones being sometimes conspicuous and
sometimes difficult to demonstrate, while the intermediate thicken- _
Ings are sometimes large and abundant, sometimes smaller and E
more scattered and sometimes wholly absent. In all the various —
forms the cell-measurements remain fairly constant. à

C. phyllobola has been confused with Lejeunea cavifolia, with
Microlejeunea lucens and even with Zuosmolejeunea duriuscula.

CHEILOLEJEUNEA VERSIFOLIA 145

From the first of these species it differs in the shape of its leaf-
lobe, in its smaller lobule, in its smaller and more deeply cleft un-
derleaves and in its more flattened perianth, the antical face being
plane or nearly so. It is a much larger plant than Microlejeunea
lucens, and differs also in its autoicous inflorescence, in the shape
of its leaves, underleaves and perichaetial bracts and in its slightly
larger leaf-cells, usually with more conspicuous thickenings. From
Euosmolejeunca duriuscula, it differs in its paler color and more
delicate texture, in its autoicous inflorescence, in its slightly larger
leaf-cells and in the shape of its leaves and underleaves.

The species is one of those interesting links between genera
which are so frequent inthe Lejeuneae. Some of the reasons for
placing it in Chedlolejeunea rather than in Hygrolejeunea are stated
by Schiffner,* who maintains that, although the leaves and the leaf-
cells show an affinity with Hygrolejeunea, the underleaves and
perianths are more like those of CheZolejeumea. Apparently
Spruce was also a supporter of this view, for he described a South
American hepatic under the name Z. (Cheilo-Lejeunea) phyllobola.t
Unfortunately the plant which he studied was not Lejeunea phyllo-
bola Nees & Mont. but an undescribed species, afterwards distributed
under a manuscript name in Hepaticae Spruceanae. There seems
to be in fact no authentic record of C. phyllobola from South America.
Specimens from Brazil are quoted in the Synopsis, but these, ac-
cording to Stephani,t are referable to Z. oxyloba Lindenb. &
Gottsche.

11. CHEILOLEJEUNEA VERSIFOLIA (Gottsche) Schifin.

Lejeunea versifolia Gottsche; Wright, Hep. Cubenses (without
description) ; Schiffner, Engler's Bot. Jahrb. 23: 597. 1897 (as
synonym).

Cheilolejeunea versifolia Schiffin. l. c. pl. 5. f. 1-7:

Pale olive-green, growing in depressed tufts, often in company
with C. phvllobola : stems prostrate, 0.04 mm. in diameter, intricately
mixed together, copiously bipinnately branched, some of the
branches spreading from substratum, flagelliform, leafless except
for two or three rudimentary leaves near apex, but with densely

* Engler's Bot. Jahrb. 23: 591. 1897.
T Hep. Amaz. et And. 259. 1884.
j Hedwigia, 29: 81. 1390.

146 _LEJEUNEAE oF THE UNITED STATES AND CANADA

imbricated underleaves: rhizoids rather abundant: leaves imbri-
cated, very fragile, distinctly smaller toward the base of the
branches, the lobe subplane, widely spreading, obliquely ovate,
0.45 mm. long, 0.3 mm. wide, rounded at the apex, margin entire
or subrepand, antical margin arching across but scarcely beyond
axis, rounded at base, postical margin straight or nearly so, form-
ing an obtuse angle with keel; lobule inflated, ovoid-cylindrical,
0.11 mm. long, 0.04 mm. wide, slightly constricted near apex, keel
slightly arched, free margin involute to or beyond apex then trun-
cate to end of keel, apex tipped with a long, unicellular outwardly
curved tooth ; cells of lobe with uniformly thickened walls, aver-
aging 10 in diameter, basal ocelli 40 x 20 رم‎ consisting of two
to six empty inflated cells: underleaves on ordinary axes contig-
uous or subremote, on flagelliform branches densely imbricated,
plane, rotund-suboblate, often subangulate, 0.16 mm. long, 0.19
mm. wide, scarcely decurrent at the base, bifid almost to the mid-
dle with acute sinus and acute subconnivent lobes: inflorescence
dioicous: 9 inflorescence sometimes with a simple innovation;
bracts unequally bifid, the lobe obliquely ovate, rounded at apex, -
entire, 0.47 mm. long, 0.42 mm. wide, lobule rhomboid-lingulate,

apex obtuse or more rarely apiculate, 0.3 mm. long, 0.15 mm.

wide ; bracteole plane, broadly ovate, 0.42 mm. long, 0.35 mm. |
wide, slightly connate on one side with bract, shortly emarginate,
bifid at the apex with rounded divisions and acute sinus, sometimes —
simply emarginate: perianth unknown: 4 spikes occupying short
branches or intercalary on leading branches, bracts in few pairs,
large, subglobose, equally lobed, surface slightly roughened from :
projecting cells. 1

Type-locality, Cuba (Wright).

On trees. Lake Worth, Florida (Underwood).

Schiffner has described and figured this peculiar little species S0.
fully and so accurately that the description given above is mostly
compiled from his. The specimens from Florida agree closely ;
with those distributed in Hepaticae Cubenses,

The leafless flagelliform branches with closely imbricated un-
derleaves are not invariably found in C versifolia and yet they form
its most striking peculiarity, They resemble in a marked degree |
the flagelliform branches of Frullania Bolanderi,* a Pacific Coast
species found in California and northward. - They are much more
highly specialized than the leafless branches of the related C
Phyllobola, where the underleaves are no closer than on ordinary ，

* Evans, Trans. Conn. Acad. I0: pl, 2. f. 4, 1897.

CHEILOLEJEUNEA PILILOBA 147

branches. C. versifolia also differs from C. phyllobola in its
smaller size, in its much smaller leaf-cells which scarcely vary
in size throughout the lobe, in its basal ocelli, in its more sharply
pointed lobules and in its relatively larger underleaves.

Another West Indian species closely related to C. versifolia is
C. emarginuliffera Schiffn.,* a Cuban plant likewise collected and
distributed by Wright. According to its author this species has a
more graceful habit than C. versifolia, its branches are less numer-
ous and are never flagelliform, its leaves are more distant and their
lobules are less conspicuously toothed at the apex, its basal ocelli
are either wanting or very inconspicuous, its underleaves are smaller,
more distant and with a broader sinus, and its female plants are
repeatedly floriferous. The fragmentary material distributed in
Hepaticae Cubenses shows that these slight differences certainly
exist, but it would require the examination of many more specimens
before it could be decided whether they were constant enough to
be considered specific. In fact Schiffner himself throws doubt on
the validity of the species.

12. Cheilolejeunea pililoba (Spruce)

Lejeunea (Eulejeunea) pililoba Spruce, Jour. Linn. Soc. Bot.
30: 346. M. 23. f. 6-8. 1894. :

Pale to dark green, caespitose : stems prostrate, 0.07 mm. in
diameter, irregularly branched : rhizoids few : leaves contiguous to
imbricated, the lobe obliquely spreading, plane, broadly ovate to or-
bicular, 0.4 mm. long, 0.35 mm. wide, rounded to very obtuse at
apex, margin crenulate from projecting cells or subentire, antical
margin arching across or just beyond axis, rounded or sometimes
abruptly contracted at base; lobule filiform from a minute, tri-
angular base, 0.17 mm. long, 0.01 mm. wide, consisting of two
to twelve cells in a single row, rarely two or three cells wide at
or near base ; cells of lobe slightly convex, thin-walled, trigones
minute, often scarcely evident, intermediate thickenings wanting,
cells averaging 14 at edge of lobe, ۸ in the middle and at
base : underleaves distant or contiguous, ovate or narrowly ovate,
0.25 mm. long, 0.13 mm. wide, narrowed toward base, bifid to or
beyond the middle with erect, lanceolate, acuminate lobes and obtuse
to lunulate sinus, margin as in leaves : inflorescence autoicous : 9
inflorescence borne on a short branch or on a leading branch,
تست‎ E E a

سسس

* Engler's Bot. Jahrb. 23: 585. 1897.

i

148 — LEJEUNEAE OF THE UNITED STATES AND CANADA

innovating on one side, the innovation simple and sterile ; bracts 3
complicate, unequally bifid, the lobe widely spreading, ovate to ob-
ovate, rounded at apex, 0.55 mm. long, 0.3 mm. wide, lobule with
a narrowly rhomboidal base connate with lobe, running out into a |
subulate, acuminate point ending in a row of three or more cells,
0.25 mm. long, base 0.05 mm. wide; bracteole oblong to obovate |
from a narrow base, 0.35 mm. long, 0.17 mm. wide, bifid one third 2
or less with triangular, acute, contiguous lobes and narrow sinus, _
perianth about half exserted, obovoid, somewhat compressed, 0.5
mm. long, 0.35 mm. wide, narrowed toward base, broad and trun-
cate above and with a very short, broad beak, antical face plane or
with a very indistinct, low keel near apex, postical keel about half
as long as perianth, broad, obtusely two-angled, lateral keels blunt,
keels more or less roughened from projecting cells: 5 spike occu- —
pying a short branch, globose, bracts in about two pairs, inflated, —
keel strongly arched, shortly and subequally bifid with rounded |
divisions ; bracteole at base of spike small, shortly bifid with acute — |
divisions and narrow sinus.

Type-locality, Dominica ( Elliott). ۰

On bark or on the ground. Indian River and Cedar Keys, a
Florida (Herb: Austin). Also known from Cuba (Underwood).

C. pililoba presents so remarkable a character in its filiform
lobule that there is little danger of confusing it with any of our a
other Lejeuneae. Among tropical species Zejeunea (En-Lejeunea) ——
setiloba Spruce * is said to have a similar but shorter lobule. E
Through the kindness of M. B. Slater, the writer has had the —
privilege of examining the type-material of this species from the
Spruce herbarium. These specimens show that the filiform con-
dition of the lobule is by no means a constant feature. On some
of the stems one occasionally finds a small, broadly ovate, inflated
lobule of the ordinary type with the apex consisting of one or tw0 —.
projecting cells, and there are all gradations between such a lobule “
and the form described by Spruce. These specimens in fact give — |
‚one the impression that the filiform lobule of L. setiloba is really an —
abnormal character. Aside from these differences in the lobule, the
underleaves of L, setiloba are broader and usually less deeply cleft |
than in C. pililoba and are occasionally angular-unidentate on the =.
sides. The perianth also is obovoid and scarcely compressed, and —

its sharp antical keel indicates that the species is a true Lejeune. —
A tit eia — AS

* Hep. Amaz. et And. 281. 1884.

LEJEUNEA PATENS 149

According to Spruce C. pililoba should also be placed in his sub-
genus Zulejeunea, but it seems best to transfer it to Chetlolejeunea on
account of its compressed perianth with obsolete antical keel.

According to the original description, the type-specimens of C.
۱/0004 were entirely sterile, and yet on this negative evidence
the species was said to be dioicous. The specimens in the British
Museum, from which the published figures were drawn, show, how-
ever, flowers of both sexes on the same stem. None of the arche-
gonia are fertilized, and the plants look as if they had grown in an
unfavorable locality, where they were unable to develop normally.
The specimens are not quite so large as those described by Spruce,
and the filiform lobule is usually less than twelve cells long, al-
though this length is given as one of the specific characters. The
specimens from Florida and also those from Cuba (Matanzas,
mixed with Hep. Amer. 145) agree closely with the British Mu-
seum specimens. Both are distinctly autoicous and the Cuban
plants show well-developed perianths.

13. LEJEUNEA PATENS Lindb.

Jungermannia serpyllifolia Dicks. Fasc. Pl. Crypt. Brit. 4: 19.
1801. :

Pandulphinius serpyllifolius S. F. Gray, Nat. Arr. Brit. Plants,
1: 089. 1821. :

Lejeunea patens Lindb. Acta Soc. Sci. Fenn. 10: 482. 1875.

Pale or sometimes darker green, slightly glossy when dry,
caespitose : stems prostrate or slightly ascending, 0.08 mm. in
diameter, copiously and irregularly branched: rhizoids few:
leaves imbricated, the lobe ovate, widely spreading, falcate,
strongly convex, 0.5 mm. long, 0.35 mm. wide, rounded and
decurved at apex, margin distinctly crenulate from projecting cells,
antical margin rounded or sometimes contracted at base, arching
across or a little beyond axis, postical margin forming an angle oı
90” or less with keel, angle in poorly developed leaves often
obtuse ; lobule strongly inflated, ovate, 0.17 mm. long, 0.1 mm.
wide, keel arched, mostly crenulate, free margin curved and invo-
lute to beyond apex, then lunately truncate to end of keel, apex
tipped with a single blunt, projecting cell; cells of lobe convex,
thin-walled, but usually with conspicuous trigones and intermediate
thickenings, walls sometimes thin throughout, cells averaging at
edge of lobe 14 y in diameter, in the middle 19 y, at the base

150 LEJEUNEAE OF THE UNITED STATES AND CANADA

28 x 23: underleaves distant, about the size of the lobules or a D
little smaller, orbicular, 0.15 mm. long, strongly convex postically,

gradually narrowed toward base, bifid about one half with obtuse
to acute lobes separated by a sinus varying from narrow and acute
to broader and obtuse, margin crenulate from projecting cells:

inflorescence autoicous: 2 inflorescence sometimes on a leading —

branch, sometimes on a short branch, innovating on one side,
the innovation often but not always floriferous and branched ;
bracts complicate, unequally bifid, the lobe somewhat spread-
ing, oblong, rounded at the apex, o.5 mm. long, o.3 mm. wide,
margin crenulate, lobule narrowly oblong, rounded to obtuse at
apex, 0.25 mm. long, 0.12 mm. wide; bracteole free or nearly
so, ovate-oblong, 0.4 mm. long, 0.25 mm. wide, bifid about one
fourth with acute lobes and sinus, margin crenulate; perianth
about half exserted, oblong-obovoid, 0.9 mm. long, 0.5 mm. wide,

. gradually narrowed toward base, rounded above and with a short
beak, terete below, sharply five-keeled in upper half or third, the
keels crenulate from projecting cells, becoming blunter with age:

$ spike occupying a short branch, bracts in two to four pairs,
strongly inflated and with strongly arched keel, subequally bifid
about one fifth with rounded lobes, cells plane ; bracteoles limited

to base of spike, similar to the underleaves, but smaller: antheridia ，

in pairs.
Type-locality, Ireland (Lindberg).

On rocks, Newfoundland (Waghorne); Nova Scotia (Ma-
coun). The species is also known from Great Britain and from —

Norway.

Lejeunea patens and the two species which follow are so nearly 3

allied to one another that they have given rise to much confusion. :

The elucidation of their synonymy is due largely to Lindberg, who

was the first to distinguish clearly between Z. serpyllifolia Lib.

and L. patens. Although he described the latter as a distinct spe-

cies, he threw doubts on its validity, recognizing the close relation- -

ship which exists between the two species. At the time Z. patens
was published, it was taken for granted that Z. serpyllifolia Lib.

was the same as Jungermannia serpyllifolia Dicks.; but in 1879

; ve

Lindberg discovered that the type-specimen of Dickson's spe
cies in MIN with his Zejeunea patens. The name “ Lejeunea
serpyllifolia,” however, cannot now be taken up for Z. patens, ON

account of the older Z. serpyllifolia of Libert. At the same time -

e A AA
* Musc. Scand. 2, footnote, 1879.

LEJEUNEA CAVIFOLIA 151

we can avoid any confusion which might arise from the use of Li-
bert's name by restoring to Libert's species the still older name
“ cavifolta," of Ehrhart, a course which Lindberg and others have
already adopted.

L. patens has not before been recorded for America. Its dis-
covery in Newfoundland and in the neighboring island of Cape
Breton shows quite conclusively that its range is northern, and
that it does not belong to that group of subtropical hepaticae
which are characteristic of the southwestern corner of Ireland.

The species has been illustrated by Moore * and also by Pear-
son.f Moore's figures bring out clearly the crenulations on the
leaves and underleaves, an important character which Pearson's
figures do not show.

14. LEJEUNEA CAVIFOLIA (Ehrh.) Lindb.
Jungermannia cavifolia Ehrh. Beitr. 4: 45. 1789.
Jungermannia clavaeflora Nees; Martius, Fl. Crypt. Erl. 137.
Jd. f. do. 1815.

Lejeunea serpyllifolia Lib. Ann. Gén. Sc. Phys. 6: 374. pl.
QU Fu 820 GLEN Syn Hep. 324. 1845

Lejeunea cavifolia Lindb. Acta Soc. Sci. Fenn. 10: 43. 1871.
Not Stephani.

Lejeunea serpyllifolia, var. B cavifolia Lindb. /. c. 10: 485.
1875.

Lejeunea serpyllifolia, var. y Americana Lindb. p. p., l. c. 10:
486. 1875.

Lejeunea (Eu-Lejeunea) serpyllifolia Spruce, Hep. Amaz. et
And. 262. 1884.

Eulejeunea serpyllifolia Schiffn.; Engler € Prantl, Nat. Pflanzen-
fam. 1°: 122, . 1893

Pale to dark green, dull or glossy when dry, growing in depressed
tufts or creeping over other bryophytes: stems prostrate, 0.07 mm.
in diameter, irregularly branched, the branches widely spreading,
sometimes few but usually abundant : rhizoids scanty : leaves imbri-
cated, the lobes ovate, obliquely preading, plane to convex, 0.6 mm.
long, 0.35 mm. wide, rounded to very obtuse at the apex, margin en-

* Proc. Roy. Irish Acad. II. 2: 加 43. 1877.
T Hep. Brit. Isles, pl. rz, 1890.
1 Bol. da Soc. Brot. 4: 171. pl. 7. f. 5-7. 1886.

152 LEJEUNEAE OF THE UNITED STATES AND CANADA

tire or nearly so, antical margin arching across or just beyond axis,
slightly rounded at base, postical margin usually forming an angle of
120° or more with keel, angle rarely as small as 90°; lobule strongly
inflated, ovate, 0.2 mm. long, 0.14 mm. wide, keel strongly arched,
scarcely or not at all crenulate, free margin more or less curved
and involute as far as apex, then passing by a shallow lunulate
sinus to end of keel, apex tipped with a single, blunt, projecting
cell ; cells of lobe almost plane, thin-walled, trigones and scattered
intermediate thickenings sometimes distinct, sometimes scarcely

evident, cells at edge of leaf averaging 17 »# in diameter, in the —
middle 24 رم‎ at the base 30 x 21 p: underleaves distant, as large m

as or a little larger than lobule, ovate-orbicular, 0.2 mm. long, 0.18.
mm. wide, plane or slightly convex postically, gradually narrowed
toward base and neither decurrent nor rounded, bifid about one
half with rounded to acute lobes and obtuse to acute sinus, mar-

gin entire or nearly so: inflorescence autoicous: © inflorescence |

and sterile, sometimes floriferous ; bracts complicate, unequally
bifid, the lobe somewhat spreading, narrowly to broadly ovate, |
rounded to obtuse at the apex, entire, 0.5 mm. long, 0.3 mm wide, -

bracteole free, oval-obovate, a

0.4 mm. long, 0.2 mm. wide, bifid one fourth to one third with —

slender beak,

keels smooth, becoming blunter with age: d spike as in L. patens-

; Massachusetts (Stone) ; Connecticut (Eaton, Hall, Evans);
Britton) ; Pennsylvania (Bar-

bour); Wisconsin (Cheney) ; Minnesota (Holzinger). Widely distri-

y reported from Japan.

Exsic. : Hep. Bor.-Amer. 97 P. p. Hep. Amer. 8 (as Zejeumea

serpyllifolia).
Lejeunea cavifolia is the commone

and cannot always be determined,

terete below, sharply five-keeled in upper part, the

st representative of the Le-

Pag

zo fs

E
:

۱۵ bins

e O‏ ی

MC SS it E TS

|
|

LEJEUNEA CAVIFOLIA 158

Lejeunea patens is a paler plant than ۲۰ cavıfolia ; the lobes of
its leaves spread more abruptly from the lobules and are more
convex, especially when dry; the underleaves are more nearly
orbicular and are smaller than the lobules, while the margins of
both leaves and underleaves are distinctly crenulate from project-
ing cells. The leaf-cells also, as a rule, are more strongly convex
and show more conspicuous trigones and intermediate thickenings.
Unfortunately the abrupt spreading of the lobe from the lobule is
not always evident in Z. patens, especially on the leaves of the
smaller branches. Neither is this character by itself to be wholly
relied upon in separating the species, for in Z. 0 the lobe
sometimes spreads widely also. The figures recently published
by Pearson * show how variable both species are in this respect.
It is only when the spreading lobe is found in connection with the
other peculiarities of Z. patens that it becomes of value as a differ-
ential character.

Lindberg's variety planiuscula of L. cavıfolia, first distinguished
simply as a form, hardly seems worthy of varietal rank. Ac-
cording to the description, the plants of this variety are more
elongated, less branched and broader than in the type, the leaves are
less imbricated and less decurved, the lobules are relatively smaller
and the cells have thinner walls and scarcely evident trigones.
Faxon's specimens from Crawford Notch, New Hempshire, and
Underwood's from Manlius, New York, agree with this descrip-
tion and also with the specimens which Lindberg quotes, but they
are connected with typical Z. cavifolia by intermediate conditions.
Apparently their peculiarities are due to a moister habitat, which
would naturally produce a laxer growth and lead toa diminution
in the development of the lobules and of the cell-walls.

Two specimens of Zejeunea distributed in Hepaticae Americanae
should perhaps be noted here. One of them is the curious sub-
merged ** Kantia aquatica” Underw., distributed as no. 107. As
Underwood | himself has shown, this plant isa true Lejeunea,
apparently related to L. cavifolia. Asa result of its submerged

* Hep. Brit. Isles, ۸7۰ ro, zz. 1899.

T Zejeunea serpylifolia, forma a planiuscula Lindb. Acta Soc. Sci. Fenn, 10:
484. 1875. Lejeunea cavifolia, var. B planiuscula Lindb. Muse. Scand. 2. 1879.
+ Bull. Torrey Club, 23: 383. 1896 (footnote).

154 LEJEUNEAE OF THE UNITED STATES AND CANADA

habit it has lost its lobules, and, as the specimens have failed to
develop perianths, it is impossible to refer them definitely. :
The other specimen was distributed under no. 178, as Lejeune ur
(Eu-Lej.) serpyllifolia. The plants seem to be perfectly ste t
the lobes are crenulate on the margins and both lobules and 1
underleaves are very small, the latter being bifid beyond the -
middle with broad, obtuse lobes and sinus. Evidently the plant —
is not L. cavifolia, but itis in too poor a condition to be deter-
mined. Both of these specimens were collected in Florida.

15. Lejeunea Americana (Lindb.)

PLATE 20, Fics. 14-26 : 3

Jungermannia. serpyllifolia Nils. & Hook. p. P.; Drummond, |

Musc. Amer. St. Merid. 172. 1841. Not Dickson. ES

Lejeunea serpyllifolia Sull. Musc. Alleg. 272. 1845. Note

Libert. m

Lejeunea cavifolia Aust. p. p. Hep. Bor.-Amer. 97. 1873. Not. 5,
Lindb.

Lejeunea serpyllifolia, var, y Americana Lindb. p. p. Acta. Soc. »
Sci. Fenn. 10: 486. 1875.

Lejeunea (Eu-Lejeunea) flava, subsp. albida Spruce, Hep. A
1884. 8

29 x 23 pat the base:
a little larger than lobul
so, rounded or subcord

LEJEUNEA AMERICANA 155

usually acute or apiculate, occasionally obtuse, rarely rounded,
sinus acute to obtuse, margin entire or irregularly sinuate, rarely
with an indistinct, blunt tooth at about the middle of one or both
sides: inflorescence autoicous: $ inflorescence sometimes on a
leading branch, sometimes on a short branch, innovating on one or
rarely on both sides, the innovations often floriferous, bracts com-
plicate, unequally bifid, the lobe obliquely spreading, varying from
oblong to obovate, rounded to subacute at the apex, 0.6 mm. long,
0.3 mm. wide, lobule mostly oblong and rounded or truncate at the
apex, rarely indistinctly toothed or lobed with rounded divisions,
0.35 mm. long, 0.18 mm. wide; bracteole free, ovate, 0.55 mm.
long, 0.35 mm. wide, somewhat narrowed at base, bifid about one
half with erect, obtuse to acute divisions and narrow sinus ; perianth
obovoid, often distinctly dilated above middle, 0.7 mm. long, 0.35
mm. wide, gradually narrowed toward base, broad and truncate
above, and with a short beak, terete below, sharply five-keeled in
upper part, the keels smooth sometimes showing vague traces of
wings : 3 inflorescence usually occupying a Short branch, rarely
borne on a longer branch, bracts in two to eight pairs, similar to
those of Z. patens : spores greenish, angular, thick-walled, minutely
verruculose, averaging 14 # in short diameter.

Type-locality, Southern States (Austin) ?

On trees. North Carolina (Johnson); South Carolina (Sulli-
vant, Miss DuBois); Georgia (Small); Florida (Farlow, J. D.
Smith, Underwood, Straub); Alabama (Underwood) ; Mississippi
(Langlois, Lloyd and Tracy); Louisiana (Drumm ond, Langlois) ;
Texas (Hall). Also known from various parts of tropical Amer-
ica.

Exsıc.: Musc. Amer. St. Merid. 171 f. f. (as Jungermannia ser-
Pyllifolia). Musc. Alleg. 272 (as Lejeunea serpyllifolia). Hep.
Bor.-Amer. 97 f. p. (as Lejeunea cavifolia). Hep. Amer. 98 (as
Lejeunea (Micro-Lej.) Austin). Hep. Amer. 7 (as 6
(Micro- Lej.) lucens).

In proposing the present species as a variety of Lejeunea ser-
pyllifolia Lib., Lindberg referred to it all the American speci-
mens in his herbarium which presumably belonged to that species.
His dictum has been followed, almost without exception, by sub-
sequent writers on American hepaticae, who have apparently taken
it for granted that typical /. serpyllifolia Lib. did not occur in
America. Judging, however, from the specimens which Lindberg
quotes, his variety was an aggregate, made up of L. Americana, as

156 LEJEUNEAE OF THE UNITED STATES AND CANADA

described above, and LZ. cavifolia. He quotes three ات‎ ۱
Hep. Bor.-Amer. 97, Musc. Amer. St. Merid. 171 ۸.۰۸۰ (caespites

pallidi) and Musc. Alleg. 272; he also quotes two plants not dis-

tributed in exsiccatae, Cleve's specimens from Catskill, New M :
and Macoun's from Belleville, Canada. Of the first-mention ad

specimen, Austin gives as a habitat, “on rocks and Me
trees; chiefly on mountains." He does not tell, therefore, in w
part of the country they were collected. The writer has
ined this number in two sets of Hep. Bor.-Amer., and finds in E
cases pale tufts, which are clearly Z. "Americana, and pese tu =
which agree perfectly with the European specimens of Z. p
quoted by Lindberg, as, for example, Hep. Eur. 435. Sullivan

specimens from Charleston, South Carolina, are a poorly derum :
form of Z. Americana. They show characteristic perianths, s

1 1 ri i lo-
their leaves and underleaves indicate that they grew in a moist

cality, and their thin-walled leaf-cells, with small and often scarcely

1

i i i ond's .

evident trigones, point to the same conclusion. Drumm: N
. > ۳ y 100 An
Louisiana specimens are also a poorly developed Z. Amer

i i ut
Cleve's specimens have not been examined by the writer, b

specimens in the Underwood herbarium from Belleville, collected E |
by Macoun, are typical Z. cavifolia. It is a little difficult to say |

" , i d
what should be considered the type of Lindberg's variety, and

consequently of the present species. The writer would e 5
however, that the pale tufts in Hep. Bor.-Amer. 72 be so regarded: —

se
Austin's plants are the first ones quoted by Lindberg, and the

pale specimens, except for certain characters drawn from the le d

cells, agree closely with the original description.

i : 1 conc x

Few of our species have been the cause of quite so much. d :

It has been referred to L. id i E

7 ; 1 2 7 n fy
Euosmolejeunea duriuscula and also to Microlejeunea lucens a

fusion as /. Americana.

to be found under one or more of these various names (or theif
equivalents ) in most of our large herbaria.

The characters which se
slight, are sufficiently distinct whe
examined. Z., Americana isa
the lobes of its leaves spread
are relatively smaller, and the
sinuous on their margin and

more widely from the axis, the lobules

tending to be rounded or subcordate

ia ie
ee 2

parate it from ZL. cavifolia, ahoi a
n well developed specimens è ۳
paler plant than the northern species, —

less
underleaves are broader, more or

p ud

MICROLEJEUNEA LUCENS 157

at the base. The perianth, however, offers the best points of dif-
ference ; in the new species this is much broader in the upper part
than at the base and frequently flares out abruptly above the mid-
dle, its keels also are sharper than in 7۰ cavifolia and sometimes
shows rudimentary wings. In the United States Z. Americana
seems to be confined to the bark of trees, while Z. cavıfolia is more
at home on rocks, at the roots of trees, and creeping over other
hepatics.

L. Americana is a much paler and more delicate plant than
Euosmolejeunea duriuscula. Its leaves spread more widely from
the axis, and its lobules are relatively larger with their free mar-
gins more involute, so that the apical teeth cannot usually be seen
without dissection ; its underleaves also are larger and not cuneate
at the base. In LZ. Americana, moreover, on account of its autoi-
cous inflorescence, perianths are frequently developed, whereas in
the other species they are extremely rare.

L. Americana should hardly be confused with ۵
lucens. It is in the first place a much larger plant and has an
autoicous inflorescence. It differs markedly also in the more
conspicuous thickenings of its cell-walls and in its larger and
broader underleaves, which are never sharply unidentate. on their
lateral margins.

Spruce's Z. fava, subsp. albida should apparently be referred to
the present species, although his specimens are a little larger and
were found on living leaves instead of on bark. Spruce afterwards
recognized that the plant was specifically distinct from /. fava and
distributed it in Hep. Spruceanae as a manuscript species. The
writer has carefully compared these specimens with typical plants
of L. Americana and finds no essential differences between them.

16. Microlejeunea lucens (Tayl.)
PLATE و21‎ Fics. 1۲-0
Lejeunea cucullata Sull. Musc. Alleg. No. 274. 1845. Not
ees,
Lejeunea lucens Tayl. Lond. Jour. Bot. 5: 399. 1846. G. L.
& N. Syn. Hep. 764. 1847.
Lejeunea (Micro-Lejeunea) lucens Spruce, Hep. Amaz. et And.
288. 1884.

158 | LEJEUNEAE OF THE UNITED STATES AND CANADA

Lejeunea (Eu-Lejeunea) lucens Steph. Hedwigia, 29 : 84. 1890.

Pale green, scattered or growing in depressed tufts: stems
prostrate, 0.05 mm. in diameter, sparingly and irregularly branched,
the branches sometimes microphyllous: rhizoids scanty: leaves |
distant to imbricated, the lobe obliquely spreading, ovate, on well- |
developed axes 0.25 mm. long, 0.17 mm. wide, plane, rounded at
the apex or sometimes varying to obtuse, margin entire or sub-
crenulate from projecting cells, antical margin arching partially
across axis, slightly rounded or almost straight at base, postical
margin forming a continuous line or a very obtuse angle with keel;
lobule on well-developed axes about half as long as lobe, on mi-
crophyllous branches sometimes as long as lobe, strongly inflated,

ovoid, 0.01 mm. long, 0.09 mm. wide, keel strongly arched, free .—

ry rarely without innovation; bracts

complicate, subequally or somewhat unequally bifid, the lobe sub-
erect, oblong-obovate, rounded or very obtuse at the apex, margin —
entire or subcrenulate, 0.4 mm. long, 0.2 mm. wide, lobule rounded

to acuminate at the apex, 0.4 mm. long, 0.2 mm. wide, sometimes
narrower ; bracteole free or nearly so, ovate 0.45 mm. long, 0.25 p
= wide, bifid about one third with acute lobes and obtuse sinus, -
: Nate en bearing a blunt lobe-like tooth on one ，

"nat trees or moist rocks: Virginia (Sullivant); Florida (Miss
"c ecome, J. D. Smith, Underwood); Mississippi (Lloyd and

MICROLEJEUNEA LUCENS 159

Tracy), Louisiana (Langlois). Widely distributed in tropical
America.*

Exsic.: Musc. Alleg. 274 (as Lejeunea cucullata). Hep. Bor.-
Amer. 98 (as Lejeunea cucullata).

There has been a great deal of confusion about this little
species and many specimens have been referred to it which really
belong elsewhere. The writer ventures, therefore, to relate its
history in some detail.

The specimens in the Taylor herbarium, which are presumably
the type, consist of a few fragmentary stems. In the small en-
velope which contains them there is written in Wilson’s hand-
writing: “ Jung. cucullata Nees? Ohio, Sullivant, 147. I have
very little of it and all barren. W.” On the sheet upon which
this envelope is pasted Taylor has written: “ Jungermannia lucens
Tayl. Mss. No. 147. Sullivant, Ohio. Jung. cucullata Nees? No.
25 Jan.? 1846. W. Wilson.” The specimens are not only
fragmentary and sterile but on most of the leaves show poorly
developed lobules, indicating that they grew in an unfavorable
locality. From the notes it would appear that these specimens
were collected somewhere in Ohio, and Taylor, in his original
description of Zejeunea lucens, states that the species came from the
vicinity of Cincinnati. The evidence, however, is against this view,
as the following facts will show. The more abundant specimens
distributed by Sullivant in Musc. Alleg. agree perfectly with those
in the Taylor herbarium, and the same may be said of Austin’s
specimens in Hep. Bor.-Amer. Sullivant’s specimens came from
Cheat Mountain, Virginia, and Austin’s are labeled, “On moist
rocks in the Alleghany Mountains, Sullivant." In the second
edition of Gray's Manual, Sullivant limits the range of his “Z.
cucullata” to the Alleghany Mountains and quotes Taylor's Z.
lucens asa synonym. Apparently then no one has collected the
species in the northern United States with the exception of Sulli-
vant, and, as all of his dicus came from the Alleghany

* Lejeunea cucullata is listed as a Massachusetts plant by Tuckerman and Frost in
their ** Catalogue of Plants growing without cultivation within thirty miles of Amherst
College?’ (1875, p. 53). Unfortunately the specimens are now inaccessible, so that it
is impossible to say whether they represented Microlejeunea lucens or merely a small
form of Lejeunea cavifolia, the latter of course being much more probable.

160 LEJEUNEAE OF THE UNITED STATES AND CANADA

Mountains, those in the Taylor herbarium could not have come
from Ohio. |
All of these specimens from the Alleghany Mountains are per- "
fectly sterile and many different conclusions have been drawn from
them. As we have already seen they were referred by Sullivant E
to Z. cucullata Nees and by Taylor to his Z. lucens. Spruce -
thought at first that they were the same as the Mexican L. diversi- ۲
folia Gottsche * and afterwards transferred them to his Z. diversi —
loba.t Still later he referred them with some hesitation to L.
lucens.{ Carrington suggested that they were probably the same
as L. crenulifolia Gottsche Ms. from the island of Trinidad.$ |
If these specimens, which are so doubtful, are to be regarded -
as the type-specimens, the species is certainly unfortunate. Spruce, i
however, considered that the type of Lejeunea lucens was a speci- ۱
men which Taylor had deposited in the Kew herbarium. || This >
specimen came from Brazil and was labeled: “Pará, in cortice —
(T. L. R.) in Herb. Hook.” In his original description Taylor —
states that his specimens came from the Hooker herbarium, and
this fact would of course Support Spruce’s view. Spruce was able
to identify with this Kew specimen a fertile plant which he himself a :
had collected near Para and in this way gave the species a firm |
footing. The specimens from Florida and Louisiana agreeing —
closely with this South American plant, perhaps it would be wisest
to follow Spruce and consider the Para plant at Kew as the type —
of Microlejeunca lucens. We can then refer tentatively to this ۱
species the doubtful Alleghany Mountain specimens, until some- —
thing more definite is learned about them. =
The generic position of the present species is not altogether ，
definite. By Spruce it was first placed in the subgenus Miete co

* Mex. Leverm. 227:
198, footnote. 1876)
Alleghany Mountai I
the type of his Z. diversiloba,
that the Mexican plant was real

1863. Not Z. diversifolia Mitt. Spruce (Jour. Bot 53 ۰

at first referred to this Mexican species not only Sullivant # X
i but also the mi plant from Ireland which later became a
He decided afterwards that all three were distinct and
ly the same as his Z. erectifolia from South America

e 13 : 469. 1879. Microlejeunea crenulifolia Steph
14.

1 .
| Hep. Amaz. et And. 288, 1884, Jour. Bot, 25: 37. 1887.

MICROLEJEUNEA RUTHII 161

Lejeunea, while Stephani * considered it a Zu-Lejeunea. Both of

these opinions were afterwards reversed, Spruce f placing it in

Eu-Lejeunea and Stephani t in Micro-Lejeunea. As a matter of

fact it is intermediate between the two genera, and it is here placed

in Microlejeunea because it seems to have more in common with

our other species of this genus that it does with our species os

Lejeunea proper. Its differential characters will be discussed in ۰
connection with the remaining species of Microlejeunea.

17. Microlejeunea Ruthii sp. nov.
PLATE 21, Fics. 11-19

Pale or dull green, scattered among other hepatics or loosely
depressed-caespitose: stems prostrate, 0.04 mm. in diameter,
sparingly and irregularly branched, the branches widely spread-
ing : rhizoids scanty : leaves distant to loosely imbricated, the lobe .
obliquely spreading to suberect, ovate or broadly ovate, 0.35 mm.
long, 0.2 m. wide, rounded at the apex, margin subentire
or slightly angular-sinuate, antical margin extending partially
across axis, almost straight or slightly rounded at base, often a
little contracted just above insertion, postical margin almost straight,
forming a continuous line or sometimes on robust axes a very
obtuse angle with keel ; lobule about half as long as lobe, strongly
inflated, ovoid, 0.17 mm. long, 0.13 mm. wide, keel more or less
arched, free margin curved, strongly involute to apex, then passing
by an oblique and lunulate sinus to end of keel, apex tipped with a
single projecting, sometimes outwardly curved cell ; cells of lobe
plane or slightly convex, walls somewhat thickened with indistinct
trigones and occasional intermediate thickenings, averaging 12 ^ in
diameter at the edge of lobe, 18 x in the middle and at the base:
underleaves distant, orbicular, o. 14 mm. long, narrowed toward base
and neither rounded nor decurrent, bifid to about the middle with
broad, suberect, triangular lobes and obtuse sinus, lobes subacute
or acute, ending in a single cell or in a row of two cells, or robust
stems usually four cells long and four cells wide at base, margin
entire or subcrenulate from projecting cells, lateral margins
rounded, never toothed: inflorescence dioicous: ¥ inflorescence
borne on a leading branch, innovating on one side, the innovation
usually long and sterile ; bracts complicate, deeply and unequally
bifid, the lobe obliquely spreading, ovate to oblong, o.5 mm. long,

* Hedwigia, 29: 84. 1890
T Jour. Linn. Soc, Bot. 30:
t Bot. Gazette, 17: 171. 1

348. 1894.
892.

162 LEJEUNEAE OF THE UNITED STATES AND CANADA

0.25 mm. wide, rounded or very obtuse at the apex, margin as in ^s
leaves but usually more sinuate, lobule mostly ligulate, rounded
at the apex, varying to subacute, 0.3 mm. long, 0.1 mm. wide,
keel wingless or nearly so; bracteole free, ovate from an ab-
ruptly contracted base, 0.5 mm. long, 0.35 mm. wide, bifid about "
one third with sharp, more or less contorted segments and narrow ;
sinus, margin as in bracts : perianth and 4 inflorescence unknown. _
. On bark. Big Frog Mountain, Tennessee (A. Ruth), the
type-locality. Sm
Microlejeunea Ruthii is closely related to the European M.

ulicina (Tayl), agreeing with it in its general habit and dioicous
inflorescence. In M. ulicina, however, the lobes of the leaves
are more convex and tend to be more or less pointed, the leaf-
cells have thinner walls, and the underleaves are more deeply 3
divided, the narrower divisions being as a rule only two cells wide :
at the base. The bracts also are more distinctly winged at the 3
keel, their lobes are more pointed and are often denticulate on the T
margin, while the bracteole is narrower. a
Although firmer in texture than M. /ucens and darker in color, p.

M. Ruthii resembles this species in several respects. It maybe
distinguished by the less widely spreading lobes of its leaves, by |
its slightly smaller cells with thicker walls, and by its broader
underleaves, which have broader segments and are never uniden-
tate on the sides. The ¢ inflorescence also is borne on a leading |
branch instead of on a short branch as is usual in M. ۵۰

18. Microlejeunea Cardoti (Steph.) 5
Lejeunea (Micro-Lejeunea) Cardoti Steph. Bot. Gazette, 17:
1892 |

Dark green, loosel

0.03 mm. in diameter, copiously and irregularly branched, the .
branches widely spreading :

loosely imbricated, the lobe obli
long, 0.17 mm. wide, plane, apex broad and rounded

volute to apex, then passi i ۱
apex tipped with a single projecting, straight or outwardly curved

MICROLEJEUNEA CARDOTI 163

cell; lobule often poorly developed ; cells of lobe plane or nearly
so, thin-walled throughout or with very indistinct trigones, aver-
aging 13 بر‎ in diameter, ocelli at base of lobe usually present
in robust leaves, consisting of one to three enlarged cells, 25 ¢
in diameter: underleaves distant, ovate, 0.07 mm. long, 0.05
mm. wide, narrowed toward base and neither rounded nor decur-
rent, bifid beyond middle with narrow, lanceolate divisions and
acute sinus, the divisions about two cells broad at base and end-
ing in a row of two to four cells: inflorescence dioicous: 9 in-
florescence borne on a leading branch, innovating on one side, the
innovation mostly simple and sterile; bracts complicate, some-
what unequally bifid, the lobe obliquely spreading, oblong to ob-
ovate, 0.35 mm. long, 0.15 mm. wide, rounded to very obtuse at
apex, margin entire, lobule ovate, obtuse to subacute, 0.3 mm. lon
0.1 mm. wide, keel wingless or with a very narrow and indistinct
interrupted wing ; bracteole free, ovate, 0.25 mm. long, 0.12 mm.
wide, bifid about one third with subacute lobes and sinus ; perianth
somewhat exserted, pyriform from a narrowed base, inflated, 0.5
mm. long, 0.35 mm. wide, rounded at apex and with a short beak,
obtusely five-keeled: 4 spike occupying a short branch, some-
times proliferating and giving off short branches beyond inflo-
rescence ; bracts in about two pairs, inflated, subequally bilobed
with obtuse divisions: bracteole at base of spike similar to under-
leaves: spores greenish, with a thickened minutely verruculose
wall, averaging 23 in short diameter, varying in length.

On bark of willow. Pointe a la Hache, Louisiana (Langlois),
the type-locality ; also reported from Mexico.

The specimens of M. Cardoti at the writer's disposal are so
poorly developed that it is impossible to gain from them a clear
idea of the species. On this account, no figures are given. The
plants from Louisiana, collected by Langlois, are presumably the
type specimens, as they are the ones first quoted by Stephani.
They are perfectly sterile and all their lobules are apparently
rudimentary. They seem to lack also the basal ocelli which are
ascribed to the species. In the plants from Mexico (Hep. Amer.
136), a few of the stems bear perianths and some of the lobes show
ocelli, but even here a well developed lobule is rare. The original
description was probably drawn from the Mexican specimens.

Stephani compares M. Cardoti with M. ulicina, M. bullata and
M. lucens ; it is also related to M. Ruthit. It can be distinguished
from M. ulicina by its relatively smaller lobule, by its slightly
smaller leaf-cells, and by the rounded lobes and entire margins of

164 LEJEUNEAE OF THE UNITED STATES AND CANADA

its perichaetial bracts. The same peculiarities of the bracts will i
also serve to separate it from M. bullata, which is a much smaller :
plant with more erect lobes and relatively larger lobules. It dif- t
fers from M. lucens in its darker color, in its smaller leaf-cells, and
in its smaller and differently shaped underleaves. Its underleaves x
will also distinguish it from M. Ruthii, a larger plant, the leaf-cells 5
of which are also larger and have slightly thickened walls. Y

19. Microlejeunea bullata (Tayl.)
PLATE 21, Fics. 20-29

Lejeunea bullata Tayl. Lond. Jour. Bot. 5: 398. 1846.

Lejeunea (Micro-Lejeunea) bullata Spruce, Hep. Amaz. et And.
289. 1846. 3

Pale or bright green, scattered or loosely depressed-caespitose:
stems prostrate, closely appressed to substratum, 0.03 mm. I
diameter, sparingly and irregularly branched, the branches widely |
spreading : rhizoids scanty: leaves distant, the lobe erect OF —
slightly spreading, ovate, 0.17 mm. long, 0.1 mm. wide, rounded =
to obtuse at the apex, margin entire or subcrenulate from project- -
ing cells, antical margin variable, sometimes arching across axis '
and rounded at base, sometimes almost straight and reaching just x
beyond edge, postical margin forming a continuous line with keel; on
lobule strongly inflated, ovoid, 0.14 mm. long, 0.08, mm. wide, -
keel strongly arched, free margin entire, plane or usually involute :
to apex, then obliquely lunulate to end of keel, apex tipped uk
a single, straight or outwardly curved projecting cell; cells of lobe |
more or less convex, thin-walled throughout or sometimes W
very minute trigones, averaging 12 p in diameter: underleaves
distant, ovate to orbicular, 0.06 mm. 1 ng, 0.05 mm. wide, nar”

lobes, three to five cells long, two cells wide at base and ending M
a row of two cells, margin entire: inflorescence dioicous: 2 IM
florescence on a leading branch with a single, sterile and simple |
innovation ; bracts complicate, subequally bifid, the keel wingless
or with a narrow, crenulate wing, lobe obliquely spreading, oblong,

tooth near apex ; bracteole slightly connate on both sides, oblong,
0.3 mm. long, 0.15 mm. wide, bifid about two fifths, sinus variable,

MICROLEJEUNEA BULLATA 165

late in upper part; perianth pyriform, 0.4 mm. long, 0.2 mm. wide,
rostellate, sharply five-keeled, smooth: ¢ inflorescence occupying
a short branch, often proliferating at the end; bracts in two to
four pairs, larger than the leaves, imbricated, strongly inflated,
subequally bifid with a very shallow sinus, the lobes rounded at
the apex and entire; bracteoles similar to underleaves, present
only at base of spike.

Type-locality, St. Vincent (Guilding).

On bark: South Carolina (Sullivant); Florida (J. D. Smith,
Mrs. Russell, Underwood, Straub). Found also at various places
in tropical America.

Exsic.: Hep. Amer. 132 f. p. (as Lejeunea (Colo-Lej.) minu-
tissima).

Microlejeunea bullata is our smallest representative of the
Lejeuneae. It has not before been recorded for the United States,
but seems to be common in Florida. When growing by itself its
minute size makes it almost invisible, but as it often grows in com-
pany with other species it has been collected several times. The
type-specimens agree closely with those from Florida.

According to Lindberg, M. ulicina occurs in Sullivant's Musc.
Alleg. 272, mixed with Zejeunea Americana. In one example of
this exsiccata examined by the writer, there seems to be no ad-
mixture whatever; in another example there is a single sterile
stem of M. bullata, and it is probable that Lindberg confused this
species with M. ulicina. At all events it seems advisable to omit
the latter species from our lists until its presence is more definitely
established.

M. ulicina is about twice as large as M. bullata, its leaves spread
a little more from the axis and tend to be more pointed, the
cells of the lobe are considerably larger, averaging 18 ۸ in diame-
ter, the cell-walls are thicker and usually show minute trigones,
while the lobules are relatively smaller. The perichaetial bracts
of the two species, except for the difference in size, resemble each
other very closely, but the bracteole of M. ulicina is entire or
nearly so, while that of M. bullata is crenulate or denticulate.

M. bullata has also been confused with Cololejeunea minutissima,
with which it often grows. The presence of underleaves will at
once distinguish it from this species, and there is little danger of
confusing it with any of our other Lejeuneae.

166 LEJEUNEAE OF THE UNITED STATES AND CANADA

20. COLOLEJEUNEA MINUTISSIMA (Smith) Schiffn.
Jungermannia. minutissima Smith, Eng. Bot. pl. 7633. 1806. |
Jungermannia inconspicua Raddi, Mem. Mat. e Fis. Soc. Ital. *
Sci. Mod. 18: 34. pl. 5. f. 2. 1820. 1

Lejeunea minutissima Dumort, Comm. Bot. 111. 1822. G. E
L. & N. Syn. Hep. 387. 1845.

Lejeunea minutissima B inconspicua Nees, Naturgesch. Eur.
Leberm. 3: 279. 1838.

Jungermannia parvula Wils. & Hook.; Drummond, Muse.
Amer. St. Mend 172. 1841. ;

Lejeunea Taylori Spruce, Ann. Nat. Hist. II. 4: 116 (foot-
note). 1849. :

Lejeunea inconspicua De Not. Mem. Acc. Tor. IL 22: 386 —
fe 7.7.27, 180%

Lejeunea parvula Aust. ; Lindb. Acta Soc. Sci. Fenn. 10: 481.
1875. T
Lejeunea (Colo- Lejeunea) minutissima Spruce, Hep. Amaz. et :
And. 293. 1884. px
Leeunea (Colo-Lejeunea) parvula Steph. Bot. Gazette, 17: i
171. 1892 m

Cololejeunea minutissima Schifin. ; Engler & Prantl, Nat. Pflan-
zenfam. 1?: 122. 1893.

reen, vatying to yellowish, scattered or loosely de- :
pressed-caespitose : stems prostrate, geniculate at nodes, 0.05 MM: ۰
in diameter, irregularly branched : rhizoids somewhat abundant:
leaves distant to subimbricated, the lobe obliquely to widely
spreading, convex, ovate to subrotund, 0.22 mm. long, 0.19 mM.

wide, rounded or very obtuse at the apex, margin crenulate from |.
projecting cells, antical margin arching wholly or partially across
axis and slightly rounded at the very short base, postical margin
forming a continuous line with keel; lobule strongly inflated,
ovoid, 9.19 mm. long, o.12 mm. wide, keel strongly arched, free
margin curved, involute about half its length, bearing an obtuse OF
rounded tooth just beyond middle, then obliquely truncate or lunu- |
late with a tooth composed of two rounded cells midway between

the obtuse tooth and end of keel; lobule often imperfectly de- x
veloped, in such cases sometimes plane or nearly so, with an es ۱
most straight keel and one or both of the marginal teeth obsolete;
stylus reduced to a single papilla, sometimes borne on the b "s
marginal cell of the lobule, early obsolete ; cells of lobe convex o

COLOLEJEUNEA MINUTISSIMA 167

both surfaces, thin-walled and without trigones, averaging 18 p in
diameter : cres EI: autoicous: 9 inflorescence on a leading

tions; bracts somewhat complicate, unequally or subequally bifid,
the sinus sometimes very shallow, lobe obliquely spreading, nar-
rowly oblong, 0.4 mm. long, 0.17 mm. wide, rounded at the apex,
lobule similar to lobe but sometimes obtuse at the apex, 0.35
mm. long, 0.12 mm. wide, margin crenulate from projecting cells :
perianth obovoid, well exserted, 0.5 mm. long, 0.3 mm. wide,
cuneate toward base, truncate or rounded at the apex and with a
short broad beak, sharply five-keeled in upper part, surface papil-
lose from convex cells: 4 spike sometimes on a short branch, some-
times on a longer branch; bracts in two to eight pairs, imbricated,
very concave, subequally bilobed, the lobes rounded or very obtuse,
crenulate, postical lobe without marginal teeth; antheridia borne
singly or in pairs: spores greenish with a thickened, minutely ver-
ruculose wall, angular, averaging 17 ۸۸ in diameter.

Type-locality, England.

On trees. South Carolina (Ravenel); Florida (J. D. Smith,
Mrs. Russell, Underwood, Straub); Alabama (Sullivant) ; Missis-
sippi (Lloyd and Tracy) ; Louisiana (Drummond, Langlois) ; Texas
(Wright). Widely distributed in southern and western Europe.

Exsic.: Musc. Amer. St. Merid. 172 (as /ungermannia par-
viia) Musc. Alleg. 276 (as Lejeunea minutissima). Hep. Amer.
132 p. p. (as Lejeunea (Colo-Lej.) minutissima).

The complicated synonymy of Cololejeunea minutissima is due
largely to the fact that for many years it was not clearly distin-
guished in Europe from Microlejeunea ulicina. It is probable in-
deed that the original Jungermannia minutissima was a mixture of
these two species, which frequently grow together and bear much
superficial resemblance to each other. In this country C. minu-
tissima has been similarly confused with Microlejeunea bullata. It
can of course be readily distinguished from both these species by
its autoicous inflorescence, geniculate stems and particularly by
the entire absence of underleaves.

The species is variable, and it is not unusual to find specimens
in which the lobules are poorly developed. This iseven true of
fruiting plants, where the stems bearing perianths tend to be
crowded together in the middle of a tuft. In such cases the
sterile stems around the edges are likely to be more characteristic.
Of the published figures of this species, the only one which brings

168 LEJEUNEAE OF THE UNITED STATES AND CANADA

out clearly the marginal teeth of the lobule is that of Stephani,
illustrating a luxuriant form from Portugal. All the other figures
seem to have been drawn from imperfectly developed material. ۱
. Although Lindberg recognized Zejeunea parvula as a distinct
species he failed to give a complete description of it. He stated
merely that it agreed with Z. minutissima in being autoicous and in
having no underleaves but that it differed in its spreading, semioval
and plane lobules. These slight differences by themselves in 0
variable an organ as the lobule are hardly sufficient to separate the
two species. Lindberg referred to / . parvula the specimens distrib- —
uted by Drummond as Jungermannia parvula and those distributed
by Sullivant as Z. minutissima. . The writer has examined both and
finds that the material is not only fragmentary but badly weather- ，
worn. It is sufficient, however, to indicate that the specimens are
the same as that rather common species of the Southern States,
which subsequent writers have referred, without question, to Z.
minutissima. There seems to be no reason, therefore, for 8 E
to hold Z. parvula distinct. It should perhaps be noted that x
sometimes, in American specimens, the lobe is more narrowed at
the apex and less crenulate than in European plants. Both
these very slight differences, however, are inconstant. American
specimens in which the lobes are just as blunt and just as crenu- — |
late as in normal European forms are also frequent, and intermediate
conditions exist between the two extremes.
c minutissima is our smallest Cololejeunea. It can easily be
distinguished from our other three species by its relatively larger s
lobule and extremely rudimentary stylus. From C. Jooriama, .
with which it sometimes grows, it differs also in its autoicous in- - .
florescence, in its total lack of hyaline cells, in the shape and iñ —
the distinctly crenulate margin of its leaf-lobes. |

EEE NL

EE 4

Re

21. Cololejeunea Biddlecomiae (Aust.) EC
Lejeunea calcarea Sull. Muse. Alleg. 275. 1845. Not Libert.
Lejeunea echinata Aust. Hep. Bor.-Amer. 99. 1873. NOE.
Taylor.
E Lejeunea Biddlecomiae Aust.; Pearson, List of Canadian Hepa-
ticae, 5. pl. 5. 1890, oe

e ET MEE EU
* Hedwigia, 26 : Pl. 1. f. 2. 1887.

COLOLEJEUNEA BIDDLECOMIAE 169

Pale or bright green, scattered or depressed-caespitose : stems
prostrate, 0.04 mm. in diameter, irregularly pinnately branched :
rhizoids somewhat abundant: leaves distant to imbricated, the lobe
obliquely to widely spreading, plane or usually convex, ovate, 0.35
mm. long, 0.2 mm. wide, apex varying from rounded to acute,
usually obtuse, margin crenulate or denticulate from projecting
cells, antical margin arching just across axis, slightly rounded at
the very short base, postical margin almost straight, forming a con-
tinuous line with keel: lobule inflated, at least near keel, ovoid to
globose, 0.17 mm. long, 0.15 mm. wide, keel strongly arched, free
margin curved, involute about half its length, bearing an obtuse or
rounded, sometimes indistinct tooth just beyond the middle, then
obliquely truncate to end of keel, with a tooth usually composed
of two rounded cells midway between the blunt tooth and the end
of keel; stylus mostly conspicuous, usually composed of two to
ten cells in a single row, sometimes two cells broad a part of its
length, in rare cases reduced to a single papilla; cells of lobe thin-
walled, usually with minute trigones, averaging 10 p in diameter at
edge of lobe, 13 بر‎ in the middle and 21 X 13 p at base, walls on
outer surface of lobe convex to conical, sometimes with a distinct
thickening or wart in the middle; cells of lobule plane: inflores-
cence dioicous or autoicous: 2 inflorescence borne on a leading
branch, innovating on one side; bracts unequally bifid, the lobe
obliquely spreading, broadly ovate, 0.35 mm. long, 0.2 mm. wide,
obtuse to acute at the apex, lobule orbicular, 0.17 mm. long,
margin denticulate and outer surfaces of both lobe and lobule pa-
pillose from projecting cells; perianth partly exserted, obovoid to
oblong, 0.6 mm. long, 0.35 mm. wide, not compressed, rounded to
truncate at the apex, with a short beak, sharply five-keeled in
upper part, surface roughened from projecting cells except near
the base: 3 spikes usually borne on leading branches ; bracts in
two to five pairs, imbricated, similar to the leaves, but with a pro-
portionately larger lobule ; antheridia borne singly or in pairs.

Type-locality, Canada (Macoun).

On trees or rocks : Ontario (Macoun, Dearness); Massachusetts
(Stone) ; Connecticut (Eaton, Underwood, Evans) ; New York
(Underwood); New Jersey (Rau); District of Columbia (Coville,
Maxon); Ohio (Miss Biddlecome, Beardslee, Wilcox, Werner) ;
Florida (Miss Biddlecome) ; Alabama (Sullivant).

Exsic.: Musc. Alleg. 275 (as Zejeunea calcarea). Hep. Bor.-
Amer. gg (as Lejeunca echinata). Hep. Amer. 51 (as Lejeunea
(Colo-Lej.) calcarea. Can. Hep. 13 (as Lejeunea calcarea).

Cololejeunea Biddlecomiae is closely related to the European

170 LEJEUNEAE OF THE UNITED STATES AND CANADA

C. calcarea (Lib.) Schiffa. In the latter species the lobes of the
leaves are narrower than in C. Biddlecomiae and much more
pointed, varying from acute to acuminate ; they spread more widely
from the stem and the postical margin is incurved beyond the keel;
the lobule is more inflated with its free margin strongly involute,
the roughness of the leaves is more pronounced, and the stylus is
commonly shorter. €
Although Spruce at one time tentatively accepted C. Biddle-
comiae as a species, he afterwards regarded it as a luxuriant form ma
of C. calcarea. This opinion was based on the fact that it is some-
times difficult to distinguish certain specimens of the American
species from certain specimens of the European. The forms which
resemble each other so closely, however, seem to be always im-

When robust and well developed specimens are examined, there is —
little difficulty in distinguishing between them. C. calcarea grows” :
preferably on limestone rocks, it seems to be much rarer we
siliceous rocks and there are few records of its occurence on trees. =
C. Biddlecomiae, on the contrary, attains its best development om m
trees, particularly in cedar-swamps, and although it is not infre-
quent on siliceous rocks, it is often less robust in such localities.
Whether it is abundant on calcareous rocks does not appear from
the specimens examined. The poorly developed specimens, grow-
ing on rocks, are the ones which can scarcely be distinguished
from rudimentary plants of C. ca/carea, as for example the spec —
mens distributed by Gottsche and Rabenborst in Hep. Eur. 323: —
Since however no typical specimens of C. calcarea have as Yt —|
been collected in America, it seems wisest to refer these doubtful
rock-forms to C. Biddlecomiae. a

papilla can often be detected somewhere near its apex. Rega
us, opinions have varied. New
ce looked upon it as one of =

COLOLEJEUNEA MACOUNII 171

divisions of a bifid underleaf, the other having become obsolete.
Stephani calls attention to the improbability of this view. Asa
matter of fact the postical segments cut off from the apical cell
bear no appendages whatever except the rhizoids. The stylus, on
the other hand, arises from a lateral segment which normally
bears a leaf, and there seems to be no reason why it should not be
considered homologous with the stylus of /rw//ania, which is ac-
knowledged to be a part of the leaf.

In a note sent by Spruce to Underwood, it is stated that a few
stems of the European Cololejeunea Rosettiana (Massal.) Schiffn.
occur mixed with Hep. Amer. 51. Apparently on the strength
of this statement, Underwood has listed the species as an Amer-
ican plant. In the two sets of Hep. Amer. examined by the
writer, no. 51, which grew on trees, is made up entirely of C.
Biddlecomiae ; and since C. Rossettiana has not been collected in
Europe except on calcareous rocks, perhaps it would be well to
omit it from our lists until its presence is more definitely proved.
C. Rossettiana was first recoganized in Italy, but has since been
detected on the British Isles and i western France. It often
grows in company with C. calcarea, with which it was long con-
fused. It differs from this species in its broader and even rougher
lobe and in its plane lobule, which is denticulate on the margin
and rough on the postical surface. Its stylus also is reduced to a
unicellular papilla, which is often difficult to demonstrate.

22. Cololejeunea Macounii (Spruce)
PLATE 22, Fics. 1-8

Lejeunea (Cololejeunea) Macounii Spruce; Underwood, Bull.
Torrey Club, 17: 259. 1890

Pale green or yellowish, scattered or loosely depressed caespi-
tose : stems prostrate, 0.08 mm. in diameter, irregularly pinnately
branched : rhizoids not abundant : leaves imbricated, the lobes sub-
erect to obliquely spreading, straight or slightly falcate, plane, ob-
long-obovate, on robust plants 0.7 mm. long and 0.45 mm. wide,

long, 0.2 mm. wide, keel arched, free margin usually plane and

172 LEJEUNEAE OF THE UNITED STATES AND CANADA

appressed to lobe, bearing an acute tooth beyond the middle, then
obliquely truncate to end of keel with a tooth composed of two
rounded cells midway between the acute tooth and end of keel;
stylus inconspicuous and soon obsolete, composed of two or rarely
of three cells in a row, sometimes reduced to a one-celled papilla;
cells of lobe thin-walled with small but distinct trigones and occa-
sional intermediate thickenings averaging 10 in diameter at edge
of lobe, 20 in the middle and 25 x 20 at the base, walls on
outer surface of lobe slightly convex, each cell, except those near
the base, bearing a single, large, short-cylindrical, rounded pro-
jection, representing a local thickening of the wall: inflorescence
dioicous: $ inflorescence borne on a leading branch innovating |
on one side; bracts complicate, unequally bifid, the lobe similar to —
the leaf-lobes but narrower, 0.6 mm. long, 0.3 mm. wide, lobule
ovate, 0.35 mm. long, 0.17 mm. wide, apiculate or’ subacuminate
at apex, margin and cells of lobe as in leaves ; perianth exserted,
obconic in lower part, rectangular-oblong above, o 75 mm. long,
0.3 mm. wide, slightly compressed, subtruncate at apex, sharply —
five-keeled, the entire surface except near the base roughened as —
in leaves: 3 inflorescence unknown. ee
On maple bark. British Columbia (Macoun), the type-locality.
Exsic.: Hep. Amer. 177 (as Lejeunea (Colo-Lej.) Macounú).
Spruce's original description of C. Macounii, which was pub- — —
lished by Underwood, is so complete that very little has been e
added to it. In fact the characters of the perianth as stated above
are taken wholly from this description, the perianths in the speci-
mens examined by the writer being either too young or too old to — P
show these characters clearly.
In its rough leaves C. Macounii resembles C. Biddlecomiae, but
in the latter species the roughness is due mainly to the very con-
vex or conical cell walls, while in C. Macounii it is due almost en-
edes the globose warts on the cells. In C. Macouni also the — |
lobe is rounded at the apex with its broadest part above the
middle, while in C. Bidalecomiae the lobe is more or less sharply
pointed with its broadest part below the middle. There are also
differences in the marginal teeth of the lobule. ex
in " interesting note to Underwood, Spruce compares Ce r
~. with two autoicous species of tropical America—C. car
1 P a (Mont) and C. platyneura (Spruce). Both of these |
species differ in their compressed perianth with a low, bluntly two-
angled, postical keel, which is almost obliterated at maturity. €

COLOLEJEUNEA JOORIANA 173

cardiocarpa differs also in the hyaline apex of its lobe and in its
muticous lobule, while C. platyneura shows a very small lobule
and a false nerve of elongated cells in the lobe.

23. Cololejeunea Jooriana (Aust.)
PLATE 22, Fics. 9-20

Lejeunea Jooriana Aust. Bull. Torrey Club, 6: 20. 1875.
Lejeunea (Colo-Lejeunea) Jooriana Steph. Bot. Gazette, 17:
171, = - 1862:
ellowish-green or whitish, darkening somewhat with age,
scattered or loosely depressed-caespitose : stems prostrate, 0.05
mm. in diameter, irregularly pinnately branched: rhizoids some-
what abundant: leaves imbricated, the lobe widely spreading,
plane, ovate, 0.5 mm. long, 0.35 mm. wide, gradually narrowed
from just below the middle to the rounded, obtuse or subacute apex,
margin entire or subcrenulate from projecting cells, antical margin
arching across axis and rounded at the very narrow base, postical
margin forming a continuous line with keel, almost straight from
end of keel to apex; lobule inflated, ovoid, 0.25 mm. long, 0.15
mm. wide, keel arched, free margin involute about half its length,
bearing an obtuse tooth just beyond the middle, then obliquely
truncate or lunulate to end of keel, with a tooth composed of two
rounded cells midway between the obtuse tooth and the end of keel:
stylus inconspicuous and soon obsolete, composed of two cells in a
row or reduced to a one-celled papilla; cells of lobe scarcely con-
vex, with slightly thickened walls, very inconspicuous trigones and
occasional, minute intermediate thickenings, averaging 14 p in di-
ameter at edge of leaf, 18 » in the middle and 25 x 18 y at base;
hyaline cells at apex one to ten in number, radially elongated,
usually in a single row, united at their basal ends but often with
rounded, more or less spreading, free extremities, hyaline cells
sometimes present also along a part of antical margin, sometimes
absent altogether, usually becoming very indistinct with age: in-
florescence synoicous and sometimes paroicous: 9 inflorescence
usually borne on a leading branch, more rarely on a short branch,
innovating on one side, the innovation sometimes floriferous ; bracts
similar to the leaves, unequally bifid, with more or less inflated
lobule and arched keel, Jobe narrowly ovate, 0.5 mm. long, 0.25
mm. wide, margin slightly crenulate from projecting cells, usually
with a small group of hyaline cells at apex and occasionally on
antical margin, lobule ovate, 0.35 mm. long, 0.2 mm. wide, crenu-
late or denticulate in upper part from projecting cells and sometimes
bearing one to three hyaline cells at apex, stylus sometimes three

174 _LEJEUNEAE OF THE UNITED STATES AND CANADA

or four cells long; perianth about half exserted, broadly ovoid to
obovoid, 0.55 mm. long, 0.4 mm. wide, rounded at the base and -
rounded, truncate or very slightly retuse at apex, with an extremely — —
short and indistinct beak, somewhat flattened, antical face plane | ۱
or with a broad low keel, lateral keels sharp to blunt, postical keel —
broad, obtusely two-angled, cells of perianth with thicker walls —.
than the leaf-cells, slightly convex, especially along keels: anthe- | 3
ridia borne singly or in pairs in the axils of the 9 bracts and some
times in the unmodified leaf below the innovation: spores more |
or less elongated, angular, greenish, with a thickened, minutely |
verruculose wall, averaging 14 » in short diameter.
Type-locality, Louisiana (Joor). oe
On bark and on reeds. North Carolina (Johnson) ; Florida *
(J. D. Smith, Underwood). Also known from Bermuda (Howe). 3
Austin's original description of this interesting speciesisincom- ——
plete and omits several of its most essential characters. For this E
reason the species has not been understood, and although speci-
mens of it have been collected several times and at various locali-
ties, they have most of them been referred to C. minutissima, with ;
which they really have less in common than with either of our 2
other species. Apparently no one except Austin studied the type ，
material. Underwood's description, the only one since published, i:
is simply a translation of the original description, and Stephani — :
States distinctly that the species is unknown to him. The original a ۱
specimens were collected on reeds, but the plant grows preferably — |
aeu
i

ME una mo S OT ata a ay AE O ET SC

uds d, SEPT‏ کل روم وت

on bark. M
The most remarkable feature of C. Jooriana is its synoicous =
inflorescence. A paroicous inflorescence occurs in the European
C. microscopica (Tayl.) Schiffn. and in a few of the larger Lejeuneae
and Frullaniae, but a synoicous inflorescence has not before been
recorded for any member of the Jubuloideae. In fact the only
— of the leafy hepaticae in which it has been observed are D
the widely removed Marsupella and Gymnomitrium, belonging t a
the Epigoniantheae. In order to see the antheridia of C. Jooriama, —
an unfertilized flower must be examined or one in which & i
penanth is still very immature. As the perianth develops the an-
theridia, having carried out their function, shrivel up and are soon
difficult to demonstrate. In most flowers antheridia occur only in |
the axils of the perichaetial bracts, their occurrence in the axil a
the subbracteal leaf below the innovation being somewnat unusual

EXCLUDED SPECIES 175

It is interesting to note that the bracts or at any rate those with-
out innovations exhibit inflated lobules, an indication of their peri-
gonial character.

he hyaline leaf-cells of C. Jooriana are also a unique char-
acter among our Lejeuneae. On old leaves these cells soon be-
come indistinct, and on poorly developed leaves they may be scanty
or wanting, but it is rare to find a plant from which they are alto-
gether absent. Hyaline marginal cells are known in a number of
tropical Lejeuneae,* the majority of them being species of Colole-
jeunea. The one most closely allied to C. Jooriana is perhaps the
autoicous C. cardiocarpa, in which the 4 spike bears six to eight
pairs of bracts. In this species also, according to Spruce, the
lobule is plane, sublinguiform in shape, obtuse at the apex and
almost parallel with the axis.

Of our own species, C. Jooriana bears some resemblance to C.
Biddlecomiae and C. Macounii. Aside from its inflorescence and
hyaline cells, it differs from both these species in its ordinary leaf-
cells which are plane or very slightly convex. It differs also from

. C. Biddlecomiae in its much shorter stylus and from C. Macounü
in the very different shape of its leaf-lobes.

EXCLUDED SPECIES
The species in the following list have all been recorded from
the United States. They are now excluded either because they
are imperfectly known or because there is insufficient evidence that
they occur within our limits. All of these species are referred to
more fully in the preceding text.

. Neurolejeunea catenulata (Nees) Schiffn.
Platylejeunea transversalis (Swartz) Schiffn.

. Archilejeunea Auberiana (Mont.) Steph.

A. xanthocarpa (Lehm. € Lindenb.) Steph.
Lopholejeunea Sagracana (Mont.) Schiffn.
Acrolejeunea torulosa (Lehm. € Lindenb.) Schiffn.
Euosmolejeunea trifaria (Nees) Schiffn.
Pycnolejeunea macroloba (Mont.) Schiffn .
Lejeunea Caroliniana Aust.

o ON بپ ج مه‎ No

* Cf. Stephani, Hedwigia, 27: 288. 1888.

a

176 LEJEUNEAE OF THE UNITED STATES AND CANADA

10. L. diversiloba Spruce.

11. L. fava Swartz.

12. £. laete-fusca Aust.

13. L. /acte-virens Mont.

14. L. Ravenela Aust.

15. Z. Wrightii Gottsche.

16. Mierolejeunea crenulifolia Steph.

17. M. cucullata (R. Bl. & N.) Jack & Steph.
18. M. erectifolia (Spruce) Steph.

19. M. ulicina (Tayl.) Evans.

20. Cololejeunea calcarca (Lib.) Schiffn.
21. C. Rossettiana (Massal.) Schiffn.

YALE UNIVERSITY. AR

>

0

BIBLIOGRAPHY

Only the more important papers dealing with our Lejeuneae are 3
mentioned in the following list. It does not include, therefore, local —
catalogues in which species are merely enumerated. E.
Austin, C. F. Characters of some new Hepaticae (mostly North - i$

American) together with Notes on a few imperfectly described vM

Species. Proc. Acad. Nat. Sci. Philad. 1869: 218-234. F
Includes description of ? Zejeunea biseriata, from Georgia, afterwards

referred to the moss-genus Erpodium. Cf. Bot. Gazette, 2: 142 —

1877.

New Hepaticae. Bull. Torrey Club, 3: 9-18. 1872. i
Hepaticae Boreali-Americanae exsiccatae. Nos. 1-150. 1872 a
Quoted as Hep. Bor.-Amer. a
New Hepaticae. Bull. Torrey Club, 6: 17-21. 1875. ae
Notes on Hepaticology. Bot. Bulletin (now Bot. Gazette), 1: - :
35, 36. 1876. x
Drummond, T. Musci Americana, or Dried Specimens of 180 Species
of Mosses, Jungermanniae, etc., collected by T. Drummond in the

Southern States of North America, arranged and named by Wilson

c Hooker. 1841. Quoted as Musc. Amer. St. Merid.

Gottsche, C. M., Lindenberg, J. B. G., & Nees ab Esenbeck, C. 8°

Synopsis Hepaticarum. 8vo., 1-834. 1844-1847. E
Lehmann, J.G.C. Muscorum Hepaticarum Species Novae. Pugillus -

T: 15514. 1838 od

BIBLIOGRAPHY 177

Lindberg, S. 0. Hepaticae in Hibernia mense Julii 187 3lectae. Acta
Soc. Sci. Fenn. 10: 465-559. 1875.

Pearson, W. H. List of Canadian Hepaticae. 8vo. 1-31. fl. 1-12.
189o.

Schiffner, V. Hepaticae. Engler & Prantl, Nat. Pflanzenfam. 1°:
3-141. 7. 1-77. 1893.

Revision der Gattungen Omplalanthus und Lejeunea im Her-
barium des Berliner Museums. Engler’s Bot. Jahrb. 23: 578-600.
PE. 1001,

Schweinitz, L. D. de. Specimen Florae Americanae Septentrionalis
Cryptogamicae ; sistens: Muscos Hepaticos huc usque in America
septentrionali observatos. 8vo. 1-27. 1821.

Small, J. K., & Vail, A. M. Report on the botanical Exploration of
Southwestern Virginia during the Season of 1892. Mem. Torrey
Club, 4: 93-201. A. 75-82. 1894.

Spruce, R. Hepaticae of the Amazon and of the Andes of Peru and
Ecuador. Trans. and Proc. Bot. Soc. Edinb. 15: i-ix, 1-588.
pl. 7-22. 1884, 1885.

Hepaticae Elliottinae insulis Antillanis S" Vincentii et Domin-

ica a clar. W. R. Elliott, annis 1891-92, lectae. Jour. Linn. Soc.

Bot. 30: 331—372. pl. 20-30. 1894.

Lejeunea Holtii, a new Hepatic from Killarney. Jour. Bot.
25: 33-39; 72-82. pl. 272. 1887.

Stephani, F. Die Gattung Zeyewnea im Herbarium Lindenberg.
Hedwigia, 29: 1-23; 68-99; 133-142. 1890.

The North American Lejeuneae. Bot. Gazette, I7: 170-172.

1892. :

Hepaticae africanae. Englers Bot. Jahrb. 22: 299-321.
1895.

Sullivant, W. S. Musci Alleghanienses sive spicelegia Muscorum
atque Hepaticarum quas in Itinere a Marylandia usque ad Georgiam
per tractus montium, A.D. 1843, decerpserunt Asa Gray et W. ۰
Sullivant [Exsiccatae]. Nos. 1-292. 1845. Quoted as Musc. Alleg.

Hepaticae. A. Gray, Man. Bot., Ed. 1. 678-696. 1848 ;
Ed. 2: 682-702. pl. 6-8. 1856.

Taylor, T. New orr Lond. Jour. Bot. 5: 258-284; 365-
417. 1846

Underwood, L. M. North American Hepaticae. Bot. Gazette, 7:
18—21. 1882.

178 LEJEUNEAE OF THE UNITED STATES AND CANADA

Descriptive Catalogue of the North American Hepaticae, north
of Mexico. Bull. Ill. State Lab. Nat. Hist. 2: 1-133. 1884.
Notes on our Hepaticae. I. Northern Species. Bot. Gazette,
14: 191-198. 1889.

Hepaticae. A. Gray, Man. Bot , Ed. 6. 702-732. pl. 22-25.
1890.

A new North American Zejeunea. Bull. Torrey Club, 17: 258,
259. 1890.

A few Additions to the Hepatic Flora of the Manual Region.
Bull. Torrey Club, 19: 299-301. ۰

A preliminary Comparison of the hepatic Flora of boreal and
sub-boreal Regions. Bot. Gazette, 17: 305-312. 1892.

Underwood, L. M., & Cook, 0. F. Hepaticae Americanae [Exsiccatae].
Nos. 1-200. 1887-1899. Incomplete. Quoted as Hep. Amer.

Explanation of Plates
PLATE 16

Mio a ere clypeata (Schwein. ) Schiffn. 1. Stem with perianth, postical view,
e stem, postical view, X 27. 3. Vor antical view, X 27-

from middle d lobe, X 350. 5. Cells from antical margin of lobe, X 250. 6,7.
Apices of lobules, X 250, 8, 9. Bracts, X 40. 10. Bracteole, X 40. Il. Tray
verse section of perianth in upper third, X 40. The Mus are all drawn from speci-
mens collected by the writer at Woodbridge, Connecticut »

PEA Sellowiana Steph. 12. Stem with two perianths and d inflores- ;
cence, postical view, X 16. 13. Leaf, antical view, X 16. 14. Cells from middle of
lobe, X 350.
X 250. 17, 18. Bracts, X 40. 19. Bracteole, X 40. 20. Transverse section
perianth at about the middle, X 40. The Sida are all drawn from specimens col-
lected by A. Commons, at Laurel, Delaware (no. 19)

PLATE 17 E
rn conchifolia Evans, 1. Stem with perianth, postical view, X 27-
etin tical view, I 3. Cells from middle of lobe, X 350. 4. Cells from
antical margin of lobe, X 250. 5. Apex of-lobule, X 250. 6, 7. Bracts, X 40
Transverse section of ionem at about the middle, X ۰ The
drawings are all made from the type s a

Mastigolejeunea auriculata (Wils. & tia ) Schiffn. ro. Stem with periant 5 7
postical view. „X 11

by L. nderwood in Florida; 11 and 15
from Orange Bend ( ۳ 1583) and the remainder from Blandton (no. 1988)-

15. Cells from antical margin of lobe, X 250. 16. Apex of lobule, Hs

EXPLANATION OF PLATES 179

PLATE 18
Brachiolejeunea corticalis (Lehm. & Lindenb.) Schifín. 1, Stem with perianth,
postical view, X 31. 2. Leaf, postical view, X 35. 3. Cells from middle of lobe,
310. 4. Cells from antical margin of lobe, X 220. 5. argin of lobule,
inner portion, X 220. 6. Free margin of lobule, outer portion, Run, both from
leat 2; 55. Bici 5° = 9. Bracteole, X 35. 10, 11. Transverse sections of
pean, 10 from upper third, 11 from about the middle, d 35. The Ay gures are all
drawn from specimens collected at Lake Worth, Florida, by L. M. Underwood
(no. 293).

Euosmolejeunea duriuscula (Nees) Evans. 12. Stem with 9 senec pos-
tical view, X 35. 13, I4. Sterile stems, postical view, X 35. 15, 16. Cells from mid-
dle of lobe, X 310. 17. Apex of lobule, X 220. 18, 19. Bracts, X 35. 20. 0
teole samê e X 35. 21-23. Bracts and bracteole from another involucre,
Xa , 13, 15, and 17-20 are drawn from Hep. Bor.-Amer. no. 96; 14, 16, and
21-23, ia Hep pr no. 97.

Euosmolejeunea opaca (Gottsche) Steph. 1. Sterile stem, postical view, X 40. .
2, Branch with perianth, postical view, X 40. 3. Cells from middle of lobe, X 350.
. Cells from anticál hes of lobe, X 250. 5. Apex fe dae x 250. 6. Apex of
veer x 250 7. Apex of underleaf-segment, X 2 8, 9. Bracts, X 40. IO.
sole, X 40 11. Tra ansverse section of perianth in vie third, < 40. The figures
are ae drawn from are collected at qne ei Florida, by E. Faxon,

15. Cells from middle of lobe, X 350. 16. Apex of la Xx 250. 17. Apex
underleaf-segment, X 250. 18, 19. Bracts, X 40. 20. Bracteole, X 40. 21. Trans-
verse section of perianth in upper third, X 40. The figures are all drawn from the
type-specimens.

PLATE 20
Chetlolejeunea ae (Nees & Mont.) Schiffn. I. Stem with perianth and Y
inflorescence, X 40. 2. Cells from middle of lobe, X 350. 3. Cells from antical

margin of bea, ox n 4. Apex of lobule, X 250. 5. Underleaf, X 60. 6-8. De-
tails of ped wor 6 representing one side of base, 7 and 8 apices of segments,
x ts, «40. 11, 12. 2. Bracteoles, X 40. 13. Transverse section of
perianth at pee pum middle, 8 peci lected
at Lake Worth, Florida, by L. M. Und erwood (no. 301 ۰ (۰
Lejeunea Paene (Lindb.) Evans. 14. Stem with perianth and & inflores-
Neg. postical view, X 40. 15. Sterile stem (a slender form), postical view, X 40.
lls from middle of lobe, X 350. 17. Apex of lobule, X 250. 18. Base of
iud (one side), X 250. 19-22. Apices of underleaf-segments, X 250. 23, 24.
Bracts, X 40. 25. Bracteole, x 40. 26. Transverse section of perianth in upper third,
X40. 2 is drawn from specimens collected by A. B. Langlois near St. Martinsville ;
the other figures, from heran collected at Eustis, Florida, by L. M. Underwood
(no. 1

PLATE 21
guión rg lucens (Tayl. Evans. 1. Stem with 9 PUREE postical
view, X 3 2. Sterile stem, postical view, X 35. 3. Cells from middle of lobe,

310. s: Abiit base of lobe, X 220, 5. Apex of lobe, X 220. 3 و‎ of

180 LEJEUNEAE OF THE UNITED STATES AND CANADA

lobule, X 2 7. Stem ید‎ X 220. 8. Branch underleaf, X 220. 9. Bract,
RAR. 0: ee X35. 2isdrawn from the doubtful type-specimen in the Taylor
herbarium (no. 147); the a are all drawn from specimens collected at Manatee,
Florida, by L. M. Underwood (no. 305). E

Microlejeunea Ruthii Evans. 11. Stem with 9 inflorescence, postical view, X 35. —
12. Sterile stem, antical view, X 35. 13. Cells from middle of lobe, X 310. 14. . 1
Apex of lobe, X 220. 15. Apex of lobule, X 220. 16. Underleaf, X 220. 17, 18. T
Bracts, X 50. 19. Bracteole, X 50. The figures are all drawn from the type-speci- — — 5

bullata (Tayl.) Evans. 20. Stem with Q inflorescence, postical‏ و
view, X 50. 21. Sterile e stem, antical view, < 50. 22. Branch of 21, antical view,‏

DC 502 O.
25. Apex of lobule, < 220. 26-28. Underleaves, < 220. 29. Bracts and bracte-
ole, X 50. 20 is drawn from a specimen collected in Florida by J. D. Smith; the
remaining vali are drawn from specimens collected at Manatee, Florida, by L.
Underwood (no. 308).

PrATE 22

Cololejeunea Macounii (Spruce e) Evans. 1. Stem with very young perianth, $
tical view, 40. 2. Sterile stem, postical view, X 40. 3. Cells from middle
۱ lobe, X 350. 4. Similar cells in cross- Pag Us gu. 5. Cells from margin of d
X250. 6. Apex of lobule, X 250. 7, 8, x af The figures are all drawn
from specimens collected in British Columbia Kaa er

Colo ejeunea Jooriana oe ) Evans. 9. Stem with pore and bisexual inflor-
escence, d view, X 40. . Stem with perianth on a short branch, postical view,
X 40. . Two leaves, 0 view, X 40. 12. Cells from middle of lobe, X 350.
£3. Cell En antical margin of lobe, X 250. 14-16. Apices of lobes, X 250. ۰
Apex of lobule, X 250. 18, 19. Bracts, X 40. 20. Transverse section of perianth
In upper third, X 40. The figures are all drawn from specimens collected at Lisbon
Florida, by L. M. Underwood (no, 100).

INDEX TO SPECIES

Species described are in SMALL CAPITALS, the page at which the description occurs
being in full face type; species cited without description in Roman type; synonyms in

Italics.

ee [pose Schiffn.

sa ( Lehm. & Lindenb. ) Schiffn.,
ns 175
E ee ای‎ aad Free
— a (Mon Parag y eum s
CLYP A (Schwein. ) Schiffn.,
119, 1 2,

hie ER 128, 178
"n LLOWIANA Steph., 116, 119, 125,

ate carpa (Lehm. & Lindenb.)
Steph., 117, 127, 175

pine gn (Spruce) Schiffn.
or (Nees) Schiffn., 133
rosa (Lehm. & Lindenb.)
chiffn., 131, 179

oes (Spruce) Schiffn.‏ مرج و
hiffn, 一 Euosmolejeunea‏ ی duri‏

(Gottsche) Schiffn.,‏ وه

E ees 8 Mont. ) Schiffn.,‏ ی
,143

LOA o “a e) ee 147, 148

POLYANTHA Evans, 141,

VERSIFOLIA (Gottsche) Schifin., 144,

145

COLOLEJEUNEA (Spruce) Schiffn.

BIDDLECOMIAE (Aust.) Evans, 115,
119, 168, 169, 170, 171, 172,

175
calcarea (Lib.) Schiffn., 115, 170,
cardiocarpa (Mont. ) Evans, 172, 173,
1

75
JOORIANA (Aust.) Evans, 116, 120,
80

4
.—n (Smith) و114 و‎

117, 165, 166, 167, 1

platyneura (Spruce) Evans, 172,
3
Rossettiana (Massal.) Schiffn., 117,
171, 176

— (Spruce) Schiffn. = e
serpyllifolia Schiffn. — Lejeu
cavifolia
EUOSMOLEJEUNEA (Spruce) Schiffn.
DURIUSCULA (Nees) Evans, 114, 116,
117, 118, 135, 142, 144, 145, 156,
157, 179
OPACA eae Steph., 116, 118,
139,
نم‎ (Nees): Schiffn., 115, 118, 140,

| Frullania Raddi
Bolanderi Aust., 146
— uri: Mont.— Brachiolejeunea

rticalis
Virginica Gottsche, 114

iti gie ote Muse. Schiffn.

ATA ) Hook. ) Schiffn., 116, 119, 133
Hygrlejeunen (Spee Schiffn.
phyllo » Me Massal. — Cheilolejeunea
phyll
Jungermannia

auriculata Win & mene — Mastigo-
— auriculat
— brh. = و‎ cavifolia
avat, sr Vees = Lejeunea cav مج‎
e Schwein. — Archilejeunea

crtcais Leh Lehm. E I = Brach- '
Meas == Microlejeu-
Raddi = Cololejeunea

Ta y == eg san سم‎
ione tes Smith = Cololej

nutissima

paronia ee id Hook. — Cololejeu-

serpyliifolia a Dicks. = Lejeunea Bes
here osas Schwein, — Archilej
clypeata :
transversal Sw. = Platylejeunea
nsversali

rsalis
بوصم‎ Lehm. & Lindenb. =
Archilejeunea xanthocarpa

(181)

182

Kantia S. F

i . Gray
aquatica Underw., I 53

LEJEUNEA

Lib.
AMERICANA (Lindb.) Evans, P4
و118‎ 120, 152, 154, 155, 156, 157,

179
auriculata Sull.—Mastigolejeunea au-
riculata

Austini Lindb. = Euosmolejeunea
a |

Biddlecomiae Aust. — Cololejeunea | |

Biddlecomiae

biseriata Aust. =a moss |

bullata Eva icrol bullayt |

calcarea Lib. = Co lolejeunea calcaraa |

calyculata Tl s — Archilejeunea cle- |

eata

|

Ca: doti Steph. = Microlejeunea Car- |

Carolinensis Mont, — Archilejeunea
cl m

Caroliniana Aust., 116,

catenulata Nee یه‎ RR ca-

tenulat

CAVIDOLIA (Ehrh.) Lindb., 114, 119,
139, 144, 151, a 153 156, 157

clausa Nees & 41

don Sull. — nee و‎ it clype-

ntigua Nees
cortical "fepe ca Brachiolejeunea

"nux Gottsche — Microlejeunea‏ تست
crenulifolia‏
Nees = Microlejeunea cu-‏ 2

aeri Gottsche — Microlejeunea

deii TON 160, 176

Dorotheae Lehm. 一 "Archilejeunea
ein

rca m Euosmolejeunea

Cololejeunea cal-‏ = ی

cecilia Spruce = Microlejeunea
olia

flava Sw Se و.‎ 118, 137, 176

Hexuosa -— nb. — Euosmolejeunea

inconspicua DeNot. = Cololejeunea

ne Aust. — Cololejeunea Joori-
an

laete-fusca Aust., 116, die
laete-virens Mont., 116, 176
a Ta yl. =. Brachiolejeunea

Tayl. — Euosmolejeunea |

INDEX TO SPECIES

LEJEUNEA Lib.

i A Vn en
e

MictoLejsU SEA Spruce) Jack k & Steph. de
sux: A (Ta 1 3 Evans, 164, me i

lucens "Tayl. — Microlejeunea lu
Macounii Spruce = 23 را لو‎ 本 Ma.
cou
Mohrii €
Molleri Steph. = Harpalejeu ode
sare cee oe umort. = = Coll ejeun
uti
ca Gotts nak == Euosmolejeunea
opaca
ovata Tayl. — H y gone ovata
و‎ Lindenb. & Gottsch 145
pu Aust. = Cololej eunea minu-
یه‎
PATENS Lindb., 119, 120, 149, 150,
pryllobola "as & ا‎ = Cheilole-
jeunea phyllo
pili E Brace se = Cheilolejeunsa pili-
0
golyphylla Tayl. — Acrolejeunea toru-
losa

pulvinata Lehm, & Lindenb., 137

Rav mr qe agn CEPOL?

Massal. Coal)‏ رم

T olin Tib. e gal cavi-

sel e a Spruce

0 Rio — Euosmolejeunea
duriuscu

Sullivan Gottsche, HL 137

Tay

lejeunea minu-

cose ssim
ami, "gegen m — flava
Sw. — Platylejeunea
ini ; E
trifaria pr == Euosmolejeunea tri-
fi
ulicina Tayl. — Microlejeunea ulicina
ی‎ Steph. = Euosmolejeu-
nea duriuscula i Sel-
reae Gottsche — Archilejeunea
wiana

veria _ orsi — Cheilolejeunea
versifo

Wrightii ی‎ 118, 6

m.‏ مر

lejeunea xanthocarpa

Schiffn.

Mont.) Schiffn., 115

| ی er,‏ تاش ویو
00

humilis (Goxsche) Schiffn., =
epleta (Tayl.) Evans, 129, I

ee opaca

en ی‎ re

INDEX TO SPECIES

MICROLEJEUNEA (Spruce) Jack, ct Steph.
CARDOTI (Steph.) Evans, 118, 162,
16

sei Steph., 160,‏ نها
cullata (R. Bl. & Mees) Jack &‏

156, 157, 159, 162, 164,
RUTHII Evans, IIQ, re xd 164,

ulicina ( Tayl.) Evans, 117, 162, 165,

167, 17
Neurolejeunea (Spruce)
a lata ( Nees) سیب‎ i 115, 127,

Lejeunea

Panda phinin S. F: Gray ==
Lejeunea

en S. T Gray —

aten
Phragmicoma Dumort. = Marches
ipm Le ved x er جوم‎ nb. as Bes
olejeunea cortica
ons Nees er clype-

cyclostipa i aor = Archilejeunea Au-

beriana +Lopholejeunea Sagraeana |
Leprieurü = = Brachiolejeunea
corticalis
ligulata m, & Lindenb. E a

keine ‘igulata Schiffn., 129

183

Phragmicoma Dumort
melanophloea
chiolejeunea re:
Tayl. gc re-

= Murchesinia
*d Nees — Bra-

teretiuscula Lindenb. & Gottsche 7. û
— Mastigolejeunea auriculata
pice Tay ie — Pycnolejeunea ma-

iba Lehm. « سس و جوز‎
perti print auriculat
xanthocarpa Aust. = ER
Sellowiana
Platylejeunea (Spruce) Schiffn.
transversalis (Sw. ) Schiffn., 114, 115,

Ptychocoleus Trev
auriculatus ries. — Mastigolejeunea
auriculata
تس وهی‎ Trevis. = Brachiolejeunea

Pyenoejennes (Spruce) S
acroloba (Mont. ) thin. ; HI

175
papulosa Steph., 142

— culatum Trevis == Archilejeunea
lypeata

: 00

‘TORREY CLUB,

Mem. 1

1۹
2
站

O
>
m
uz
LÀ
o
x
z
a

M

8

—

TORREY CLUB,

وود

MEM.

Torrey CLUB, $

Mem.

EJEUNEA MACOUNII (Spruce) Evans |

Memoirs of the Torrey Botanical Club

The MEMOIRS are published at irregular intervals and are sold at a uniform
price of THREE DOLLARS a volume. Most of the articles can be furnished
separately at the prices indicated below. The price includes postage.

Volume 1.

No. 1.—Bailey, L. H. Studies of the Types of various Species of the Genus
Carez, ue 1-85. 1889.

No. —Martindale, I.C. Marine Algae of a = Jersey Coast and adjacent
Waters x Staten Bast Pp. 87-109. 1889. Pri cents,

No. 3.—Spruce, R. An Enumeration +: the Hepat ea by Dr. H.H.
> in pee — Pp. 113-140. 1890. 75

0. 4.— ant, E. L. On seedless E Pp. سروس‎ 1890. Price,

75 ae

Volume 2.

No. 1.—Halsted, B. D. On reserve Food Materials in Buds and surrounding
Parts. Pp. 1-26. Pl. 1, 2. 1890. Price, 50 cents.
No. 2.—Vail, A. M. and eg A. Contributions to the Botany of Virginia.
Pp. 27-56. Pl. 3, 4. 1890. Price, 75 centa.

No. 3.—Holm, T. as to wem Knowledge of the Germination of some
gr mecs Plants. Pp. 57-108. Pl. 5-19. 1891. (Notfurnished separately. )

—Wheelock, W. E. A Monograph of the North American Species of

9 ee Polygala. Pp. 109-152. 1891. Price, 75 cents.

Volume 3. t
No. 1.—Small, J. K. and Heller, A. A. On the Flora of a2. North u
lina and contiguous Territory. -39. 1 2:
No. 2.—Morong, T. A Revision of the North American indecens with Tilus-
trations of all the Species. Pp. 1-65. Pl. 20-74. 1893. Price, $2.00.
No. 3.—Rusby, H. H. An Enumeration of the Plants collected in Bolivia S
Miguel Bang. Pp.1-67. 1893. Price, 50 cents.

Volume 4. vo
Underwood, L. M. Index Hepaticarum. Part 1, Bibliography. Pp mem
TR ss Price, 75 cents. ed

No. 2.—Vail, A. M. and Small, J. K. Report on the Botanical Explonion dl 2
E^ oo the Season 1892. Pp. 92-202. Pl. 75-82. -e Price, 50

No. 3.—Rusby, H. H. An Enumeration of the Plants collected in Bolivia by

Miguel Bang—II. Pp, 203-274. 1895. Price, 50 cents. Rt.
No. 4.—Pettit, A. S. Arachis hypogaea L. > 215-996. 1895. Price 50

cents, du a

a E d. Monograph of Phy lis and relat Gener : row

Om) aT = Columbia University, 2 New York.

Memoirs of the Torrey Botanical Club

The MEMOIRS are published at irregular intervals and are sold at a uniform - E
ice of THREE DOLLARS a volume. Most of the articles can be furnished
separately at the prices indicated below. The price includes postage.

Volume 5.
List of Pteridophyta and Spermatophyta growing without Cultivation in North-
‘eastern North America. Pp. 1-377. 1893-1894. Price, $3.00.

Volume 6. :
1.—Rusby, H.H. An en. of the Plants collected in Bolivia by
Miguel Bang—III. Pp. 1-130. 1896. e, $1.25 |
o. 2.—Grout, A. J. Revision of نگ‎ North American Isotheciaceae a
T سس یرجم‎ Pp. 131-210. 1897. Price, 50 cen 2r
TT 3.—Hazen, T. E. The Life History of Sphacrlla gen سرت‎ =
© en Pp. 211-244. Pl. 86, 87 (colored). 1899, , 50 ce:
No. 2 Pape usen L.M. A Review of the Genera of Pii props pile
to 1832. Price, 50 cen
- No.5.—Pink, B. Notes o on the Lichen Distribution in the Upper Mississippi
Valley. Price, 25 ce

qe

-

Volume 7.
Howe, M. A. The Hepaticae and Anthocerotes of California. Pp. 1-208. Pl.
| 88-122. 1899. Price, $3.00.
Volume 8.

ox ML —Lloyd, P.E. The comparative Embryology of the Rubiaceae.
OL Pp. 1-26. PLI. 1899. Price, 50 cents... Part II. Pp. 27-111. Pl. 5-15. 1902.
Price, .25.

- .. Ne 2.—Hvans, A. W. The e of the United States and Canada. Pp.
3-183, PL. 16-22. 1902, Price, i
T No. 3.—Britton, E.G. & scia A. The Life History of Vittaria lineata. _
3 PL 23-31, 1902, Price, 75 cents,

Volume 9.

- Salmon, B. = A Monograph of the Erysiphaceae. Pp. 1-292. Pl. 1-9

1900. e 00. i
Volume 10.

Burgess, E. S. Studies in the Genus Aster. Part I. In Press.

Volume ii ;

1 ann D. The North American Sordariaceae. Pp. 1-134. Pl. *

E 2 Hate, T. B. The TT Trinthet i oe = ao North Amer
In Press.

É-

Address, TORREY BOTANICAL CLUB,
Columbia University, New 3 Yor

MEMOIRS |

OF THE ©

The Life History of Vittaria lineata
By ELIZABETH G. BRITTON AND ALEXANDRINA TAYLOR
(PLATES 23-31)

At the conclusion of our studies on Schisaea pusilla, Mr.
Tracy Hazen called our attention to the fact that Francis Wolle *
had figured a filamentous protonema, which he suspected to be-
long to. a fern, but had previously described as Poterophora Don-
nellit.+ It was collected on trees in Florida by John Donnell
Smith who loaned us the type from which drawings have been
made, as the figures given by Wolle were found to be incorrect.
The urns described by Wolle are podetia and produce spherical
or cylindric gemmae which Wolle did not figure. This curious
method of vegetative propagation has been figured by Goebel } in
a species of Trichomanes, but as the sporophyte is still unknown
in Poterophora and we knew of no epiphytal species of Tricho-
manes in Florida, suspicion was aroused regarding Vittaria lineata.
Living plants in all stages were secured from E. W. Reasoner, of
Oneco, Florida, and also from Miss Abbie M. Small at Jackson-
ville, Florida. We have kept these plants growing in the labora-
tory for more than a year, and watched the stages of their develop.
ment. As far as we can learn, the gametophyte of this fern has
never been described ; in fact the only reference to it is in a foot-
note in Eaton’s Ferns of North America, which states that “ J.
Donnell Smith and C. F. Austin found the prothalline growth
of this fern very abundant on palmettos along the Caloosa-

hatchie River, Florida." The peculiarities of its structure are

remarkable, agreeing in essential particulars with those of Vittaria
elongata, described by Goebel from specimens studied at the
Botanical Gardens at Buitenzorg, Java.$

* F. Wolle, F. W. Algae of N. A. 120. ۰ 106. 1887.

f Bull. Torrey Club, 6: 177, 188. 1877.

f Goebel, Ann. Jard. Buit 7: 78-87. A. 10-13. 1888.
¿ Goebel, Ann. Jard. Buit. 7: 95. P/. 12, f. 60-61.
Memoirs Torrey Botanical Club, Volume VIII.
[No. 3 issued 30 August, 1902.] en

186 THE Lire HISTORY OF

The systematic records of Vittaria lineata were also found to
be very imperfect ; specimens were much mixed in our collections
and several other closely allied species had been mistaken for this
species. The types of several of these need to be restudied bua
Scopically before we can reach a correct knowledge of the limits
and distribution of the American species. The material in Ameri-
can collections is necessarily incomplete, and the types are scattered
through European herbaria. However, with what specimens and
literature are available, we believe it will be helpful to give a brief
survey of the history of Vittaria lineata, and to describe and hs
the gametophyte of our specimens from Florida. It would be stil
more helpful if collectors in the West Indies would make it a point
to collect the prothallia and young stages of Vittaria, and preserve
them in condition for study.

The genus V aria is in need of revision, as there has been
mone made since that of Fée,* and subsequent treatment in the
Synopsis Filicum has largely disregarded the results attained in
his memoir. Kunze and Fée were the first to recognize the: value
‘of the spores and sporangiasters in specific distinctions, and Fée also
figured the scales and sporangia, although in many spécies the lip
cells were omitted. The venation is obscure in many species, and
in consequence has been wrongly described. Taenitis has bere
combined with Vittaria by some authors which tends to complicate

the synonymy, and Taentopsis was founded on the type species
of Vittaria.t

|

The most recent list given of the species is that of Salomon f
from which the following has been compiled, but it will be found
that a number of corrections have been made after consulting the
original place of publication.

CHRONOLOGICAL List OF THE AMERICAN SPECIES
1753. Pteris lineata L.. Sp. PL. 1073.
1801. Vittaria filiformis Cav. Praelect. 270 (671).
1803. V. angustifrons Michx. Flor. bor. Am. 2: 261.
1819. V. Schkuhrit Raddi, Fil. Bras. 51.

* Fée, Mem. des Foug. 3: 1-24. 1851.
tJ. Smith, Hook. Journ, Bot. 4 : 67. 1842,
$ Salomon, Nom. Gefäss.-Krypt. 1883.

VITTARIA LINEATA 187

1824. V. graminifolia Kaulf. Enum. Fil. 192.

1837. V. stipitata Kze. Analect. pteridogr. 28. pl. 18. f. ۰
1851. V. Gardneriana Fée, Mem. des Foug. 3: 15. Pl. 5. f. 1.
1851. V. filifolia Fée, Mem. des Foug. 3: 20. pl. 3. f. 6.
1851. V. Ruisiana Fée, Mem. des Foug. 3: 16. pl. 3. f. 3.
1851. V. scabrida Klotsch ; Fée, Mem. des Foug. 3: 20.
1854. V. remota Fée, Mem. des Foug. 7: 26. pl. 20. f. r.
1854. V. curvidentata CM. Bot. Zeitung, 12: 546. pl. 13. f. 1.
1854. V. dimorpha CM. 1. c. 547. pl. r3. f. 3.

1854. V. pachydictyon CM. L. c. 547. pl. 13. f. 4.

1854. V. Deppeana CM. 4. c. 547. pl. 13. f. 6.

1864. V. Karsteniana Mett. Ann. Sc. Nat. V. 2: 206.

1864. V. Moritziana Mett. l. c., 207. .

1869. V. Orbignyana Mett. Linnaea, 36 : 67.

1869. V. gracilis Moritz, Linnaea, 36 : 67.

1877. V. intramarginalis Baker, Journ. Bot. 15: 266.

1893. V. longipes Sodiro, Crypt. Vasc. Quit. 423.

Of these V. filiformis, V. angustifrons, V. Schkuhrit and ۰
curvidentata have been referred to V. lineata as synonyms, and V.
graminifolia has been reduced to a variety of both V. Zneata and
of V. curvidentata. _Mettenius* expressed the opinion that V.
graminifolia and V. filifolia were the same species, and figured ۰
filifolia with a cross-section of the leaf. The type of V. gramini-
folia should be found either at the University of Leipzig or in the
herbarium of Count du Roemer at Dresden and the type of V.
filifolia in the Cosson herbarium at Paris. Fée says it is remark-
able for its long stipe. V. Gardneriana and V. remota are too
closely allied and need more study; V. Ruisiana is said to be a
synonym of V. Moritsiana by Mettenius and if so has priority
over that species; V. scabrida has not been figured, and is ap-
parently closely related to V. dimorpha in size; F. Deppeana has
been referred to V. filifolia, by Mettenius, but is apparently distinct ;
V. gracilis and V. intramarginalis should be compared with Y.
filifolia ; and V. longipes should be compared with V. Orbignyana.

Karl Müller f has shown that the character of the scales sur-
rounding the base of the leaves vary in different species, and may

* Mett. Fil. Hort. Bot. Lips. 25-27. .از‎ 21, 22. 1856.
+ Bot. Zeitung, 12: 537-548. Al. 13. 1854.

188 THE Lire History OF

be either entirely parenchymatic, or the cell walls may be thick-
ened evenly, or even papillose. This character he made use of
in describing six new species, but he ignored the spores and para-
physes, so that it is hard to compare his species with those >
scribed and figured by Fee. Luerssen * followed Mettenius in
reducing V. Deppeana to V. filifolia, in spite of the differences of
the scales as figured by Müller, and maintained V. curvidentata to
which he referred V. graminifolia as a variety.

GEOGRAPHICAL DISTRIBUTION

In the following list the type locality is given first, and then
the distribution as cited by Fée, the authors of the species and sub- .
sequent lists, without any attempt to verify the determinations :

V. lineata, Haiti; West Indies, North and South America,
East Indies (?).

V. graminifolia, Brazil; Peru, Mexico.

V. stipitata, Peru; West Indies and South America.

V. Gardneriana, Brazil; South America and Mexico.

V. filifolia, Mexico; Central and South America, West Indies.

V. Ruiziana, Peru and South America. (V. Moritziana, Peru.)

V. scabrida, Brazil; Mexico.

V. remota, New Granada, South America and West Indies.

V. Karsteniana, United States of Colombia.

V. gracilis, United States of Colombia.

V. Orbignyana, Bolivia.

V. longipes, Ecuador.

V. intramarginalis, Jamaica.

V. dimorpha, Mexico.

V. curvidentata, Brazil.

V. pachydictyon, South America, Surinam.

V. Deppeana Mexico.

Leaving out of consideration the South American species, it

from Mexico and the West Indies. Of these the Mexican species
are most interesting and need revision. V. scabrida and V.
Deppeana have not been seen in any American herbarium. In V. 3
scabrida Fée t gives the length as 90 cm.; this is probably a mis- 4
* Luerssen, Fil. Graeff. 92-94. 1871. 5 s 3
T Fée, Mem. des Foug. 3:20. 1852.

VITTARIA LINEATA 189

take for he says it is a small species with short fronds, and it is
described in the Synopsis Filicum as two inches long, and
Sodiro + says the leaves are 2-6 inches. Specimens labeled V.
dimorpha collected by F. Müller in 1848-1855, at Orizaba, are in
the Columbia University herbarium, but do not answer to the de-
scription of this species. Fournier [ credits V. Zineata, V. grami-
folia and V. scabrida to Mexico, and Fée $ lists V. filifolia from
several Mexican localities and collectors.

We have not studied the Central American species, but we
find among the specimens distributed by J. Donnell Smith as V.
lineata, that no. 811, collected by Turckheim in Guatemala is ۰
Jilifoliaand that no. 7489 collected by H. Pittier in Costa Rica, is
probably V. graminifolia.

Grisebach || records only one species, V. /ineata, from the
West Indies, which he says occurs in Jamaica, St. Vincent and
Barbadoes. Charles Wright collected in Cuba as no. 865, in 1856,
two species, V. filifolia and V. lineata, which were distributed as
one. No. 865, of his collection of 1865, is V. stipitata. Kunze Y
records V. graminifolia from Cuba.

V. Gardneriana has been collected in Puerto Rico by Blauner
and Sintenis, and V. remota has also been credited to Puerto Rico.
Fée** has reported V. filifolia from that Island and V. lineata was
collected by Sintenis.

Jenman ++ has described V. Zineata, V. stipitata, V. remota and
V. intramarginalis from Jamaica: the last a new species closely
allied to V. filifolia, judging by the description and specimens
received from B. D. Gilbert; it has triangular spores, but the
paraphyses have not been seen. V. remota is described as having
the venation anastomosing once besides the normal marginal one;
this does not agree with Fée's |] figures, in which the venation
is oblique as in V. Gardneriana.

T Sodiro, Crypt. Vasc. Quit. 423. 1893.

1 Fournier, Miss, Scient. au Mex. et Am. Cent. 114. 1872.

2 Fee, Mem. des Foug. 3: 15. 1852; 11, 13. 1866.

|| Grisebach, Flora of the West Indies, 674. 1864.

Y Kunze, Syn. Plant. Crypt. Poepp., Linnaea, 9: 76, 77. 1833-

** Fée, Mem. 11: 13. 1866.

tf Jenman, Syn. List of Jamaica Ferns. Bull, Bot. Dept. Jam. 4: 212. 1897.
tt Fée, Mem. 7: 26. pl. jo. f. 1.

anon ne)
— شس سس سسس س‎

190 Tue Lire HISTORY or

Haiti is the type locality for V. Zineata, but we have never
seen a specimen of Vittaria, from Haiti or Santo Domingo. Spe-
cimens were distributed from Berlin, determined by Kuhn as Z
graminifolia var. filifolia collected by Hahn on Martinique.
V. filifolia has been credited by Fée to Guadeloupe collected by
L'Herminier, and Y. /ineata was collected on Trinidad by Fendler.
A dwarf form of V. lineata also occurs in the Bahamas. Mr.
Britton has collected sterile specimens of Y. Jilifolia on St. Kitts.

HisrORY OF THE GAMETOPHYTE

The statement has been made that the gametophyte of Vittaria
lineata has not been described, and so far as we knew, the game-
tophyte was unknown in all the other American species. Professor
Underwood has called our attention to a copy of Lindenberg's mon-
ograph of the Ricciaccae * which formerly belonged to Austin, in
which Austin had written opposite the description of Riccia
reticulata, “a species of Vittaria.” Lindenberg had classed it
among the “Species dubiae.” Riccia reticulata was described
by Swartz + as having the fronds laciniate, glabrous and re-
ticulate. He referred to it specimens from the Island of
[New ?] Providence described by Dillenius | as Zichenoides gelati-
nosus tenue reticulatum. Swartz§ stated that it grew on trees
covered with mosses in the mountains
it as follows :

RICCIA ? RETICULATA. “ Frons membranacea, plana, difformis,
lobata, laciniata, latitudine unciales, laciniis linearibus, saepe bipar-
titis, pollicaribus, glabra, obscure viridis, tenuissi ticulat

bs. Ob texturam teneriorem
Lichenibus.”

of Jamaica, and described

Hepaticis conjungere malui quam
The illustration given by Dillenius is simply an
outline, but the figure and description given by Lindenberg agree
exactly with no. 327 from St. Kitts, collected by Mr. Britton in
1901, which he also took for a hepatic. It proves to be the
gametophyte of a species of Vittaria, probably V. filifolia, the
leaves of which were collected in another ravine but were not
*J. B.W. Lio i AN
CITUR Me nt

+ Dillenius, Hist. Muse. r: 138. pl. r
, ae * P^ 19. f. 21. 1741.
¢ Sw. Fl. Ind. Occ. 3: 1883. 1806. =

VITTARIA LINEATA 191

found with or near no. 327. This implies an independent exist-
ence for the gametophyte, which agrees with that of V. ۵ in
being proliferous by its septate gemmae and in being lobed, but it
differs in the shape and size of the thallus, the shape of the lobes,
and the cellular structure. However, the specimens were dried
when examined, as were Lindenberg's, whose description and
figures were drawn from the type of Riccia reticulata, belonging
to Swartz, which had been collected thirty years at the time the
drawings were made. The cell contents had shrunk, forming a
fine mesh of the strings of protoplasm, radiating from the nucleus
of each cell, which gave the dotted and reticulated appearance
described by Swartz as visible with a hand lens. The gemmae
also must have disappeared, for they are neither figured nor
described. In the specimen from St. Kitts, they are most abundant
on the truncate ends of the long, narrow lobes.

That the gametophyte of Vittaria maintains an indefinite and
independent existence, without forming the sporophyte, seems to
be an established fact. It is autoicous, the archegonia being
borne mostly on the thallus, and the antheridia on the gemmae.
In a climate where violent winds and heavy rains are frequent
the gemmae would be liable to be detached without fertilization -
having taken place. Dr. Campbell reports having seen in Jamaica,
enormous sterile prothallia, forming circular mats at least three
inches in diameter on the trunks of palms, and resembling a liver:
wort. Dr. Campbell * has also described and figured a prothallium
collected in the Sandwich Islands, similar to that of Vittaria,
which he has referred to a species of Hymenophyllum. He
states that they lived for two or three years in his laboratory.
He also states that ‘Goebel has described in Vittaria " en
prothallium recalling that of Hymenophyllum, both in its large
size and extensive branching.” From our present knowledge
we are inclined to believe that Campbell has mistaken the genus,
and that these prothallia should be referred to Vittaria, probably

to V. elongata.
en DCN

* Campbell, Mosses and Ferns, 362. f. 187, 158. 1895.

192 THE Lire HISTORY OF

Synopsis of the North American Species of Vittaria*
Stipes green; 1-2 cm. long, margins recurved when d
vus arg paraphyses enteriform ; scales papillose, serrate.

en cm. long, 2-3 mm. wide, pendent. V. lineata.
0-15 cm., narrowly Mosi; erect. V. graminifolia,
"etn bein ; scales ei . V. intramarginalis,
Stipes green or yellow, 10-15 cm. long ; margins revolut

Spores lea ed mastoid- claviform, dits smooth, entire.

V. filifolia,
Spores tetrahedral? ; paraphyses enterifori?: ; scales papillose, aristate, serrate.

Deppeana,
Stipes green, short; leaves 4-16 c

Spores tetrahedral ; en سوت‎ ; ; scales dentate. V. scabrida.
Spores ; paraphyses ?; scales aristate, oes entire.
^ dimorpha.

Stipes brown or black, 2-9 cm. long; margins flat; spores eee
ves 20-30 cm. long; paraphyses cyathiform.
Leaves 7-9 mm. wide; rhizome creeping. V. Gardneriana.
V. remota.
Leaves I meter long, 3-5 mm. wide ; ; paraphyses pyriform. V. stipitata.

VITTARIA LINEATA (L.) J. E. Smith.

Lingua cervina longissimus & angustissimis foliis. Plumier
Fil. Am. 123. y. 143. 1705.

Phyllitis lineata, graminis folio longissimo. Pet. Fil. 126. ۰
14. 1913,

Pteris lineata L. Sp. Pl. 1073. 1 E,

Vittaria lineata J. E. Smith, Mem. Acad. Roy. Sci. Turin. 5:
413. 1793.

V. lineata Swartz, Schrad. Journ. 2: 72. 1800; Syn. Fil.
109. 1806.

V. filiformis Cav. Praelect. Bot. 270, 671. 1801; Willd. Sp.
Pl. 5: 405. 1801.

V. angustifrons Michx. Flor. Bor. Am. 2: 26
Bory. Itin. 2: 325. 1804.

V. Schkuhrü Raddi, Fil. Bras. 5I.

Taeniopsis lineata J. Smith, H

I. 1803; non

1819.

ook. Journ. Bot. 4: 67. 1842.
V. lineata Sw. Fée, Mem. Fam. Foug. 3: 17. 1851; Hook. Sp.
Fil. 5: 180. 1864; Hook. & Baker, Syn. Fil. 396. 1874; in part;
— rA of N. Am. 1: 289, 1879; Sodiro, Crypt Vasc.

* This synopsis is necessarily į
y incomplete, o us
and lack of knowledge of the types. en dienen emere

VITTARIA LINEATA 193

Quit. 420. 1893; Jenman, Bull. Bot. Dept. Jamaica, 4: 251.
1897.

V. curvidentata CM.? Bot. Zeitung, I2: 546. 1854.

Linnaeus founded this species on the plate and description given
by Plumier, cited Santo Domingo for the type locality, and adopted
the specific name from Petiver's description, according to citations.
Plumier figured the species very clearly on a large quarto plate,
and described it in French and Latin; as this work is rare in
America, it seems desirable to quote his description.

PLANCHE CXLIII.

ds Langue de Cerf à feuilles tres longues et tres étroites. La
farine de cette Langue de Cerf n'est proprement qu'un amas de
plusieurs pédicules pourris mélez de plusieurs petites fibres roussa-
tres, & toutes couvertes de mousses : elle jette quantité de feuilles
tres longues & tres étroites, car elles n'ont pas plus d'une ligne &
demie de large, sur environ trois pieds de longueur: elles sont
fort pointués, solides, épaisses comme du parchemin, unies, & d'un
raversées de biais en

vert assez gay par devant, mais toutes t
proches les unes des

derriére par quantité de petites nervüres tres
autres : outre ce elles ont aussi le bord replié en dedans, & couv-
rent tout le long de leur pli un nombre infini de tres petites ves-
sies noiratres.

J'ay rencontré cette Pl
l'Amérique, particuliérement à la Sou
vers le quartier du Port de Paix, dans l'Isle Saint Domingue."

Plumier's description agrees very well with the general accep-
tance of the characters of the species. It must be taken as the
type of the genus, for J. E. Smith founded the genus on Vittaria
lineata; theré is no reason for taking up Taeniopsis of J.
Smith, nor for citing Swartz as the authority for the species. It
has been found that the sporophyte only has been described, and
in the case of some of the synonyms, in the briefest and most
superficial manner, giving only macroscopic characters. It fol-
lows that further study may reinstate some of these as species.
This may be true of V. filiformis Cav. « foliis filiformibus, tripedali-
bus, glaucis." Fructification as in V. lineata ; habitat Peru, or of

V. angustifrons Michx. “ frondibus simplicibus, integerrimis angus-

ante en plusieurs endroits des Isles de
rce de la Riviére Froide,

194 ; THE Lire HISTORY OF

tissime longissimeque linearibus, debilibus. ‚ Fructification -
continua sub replicato frondis margine, ab istius sinu distincta.
Hab. in Florida juxta ammem Aisa-hatcha.” "
It is not likely that the Florida species will be found to be is
tinct from the West Indian, but until the gametophyte of both vi
known there is still this possibility. Fée has suggested it in his
memoir, as he states that he has seen in Willdenow's herbarium ۳
specimen of V. Zineata collected in Jamaica by Swartz, which :
bright amber-colored sporangiasters, with the terminal cell gs ve
or elongated pyriform, and that he also had a specimen co ger
in Florida from Richard which has ribbon-like sporangiasters. e
believed the two plants were distinct species. Yet he contradicts
himself in his description of V. lineata which he attributes to
Swartz and states that the sporangiasters are vittate, with dilated
colored contorted apex. He also states, that although V. 0
i 3 e exists the greatest vagueness and un-
certainty in the descriptions and that he considers it to be exclu-
sively American. Unfortunately, he did not figure V. lineata, only
the variety graminifolia Kaulf, without stating whether he had
seen the types. He also took up Taeniopsis, as a subgenus, and
referred V. lineata to it; we have shown that this position is en
tenable. In this he was followed by the authors of the Synopsis
Filicum, who also reduced Y. filifolia and several Asiatic species
to V. lineata. We have already shown in the key, that V. filifolia

has the characters of à good species, but we have not studied the
exotic species,

looked

plate. Jenman has descri

of V. lineata from which it i
small barren fronds
As none of the descriptio:
we have framed one from fres
obliged to limit it to F lorida
ERI GS Ber aes

Ld

which is epiphytic on trees and much subject to t

VITTARIA LINEATA 195

Gametophyte capable of indefinite growth and duration, branch-
ing irregularly or bifurcating into numerous lobes, which are pro-
liferous by gemmae borne from podetia, usually on marginal cells ;
serving to propagate the plant asexually, and also rendering it au-
toicous by bearing antheridia ; archegonia borne on the under side
of the thallus.

Sporophyte forming rudimentary leaves, 1-4 cm. long, spathu-
late, with a single central vein ending below the apex or forming
a loop in the rounded apex ; rhizome short, divided, creeping, spar-
ingly, densely covered with lanceolate, serrate, filiform-pointed
scales, with papillose walled cells; leaves dark green or glaucous,
two-ranked, crowded in dense compact pendent tufts, often 2-3 ft.
long, 2-3 mm. wide, coriaceous, three-veined, the midvein pinnate
and anastomosing with the lateral veins, with a deep sporangial
groove over each lateral vein, in which the stomata are borne,
paraphyses branching twice, usually forking, cylindric, the end cells
curved and larger, brown; spores reniform, smooth, .027 X .054

m

Florida, as far north as St. Augustine ; also credited to South
Carolina by Fée and Hooker, but probably by mistake. Pre-
sumably the same as the West Indian species, of which the game-
tophyte is still undescribed, and the microscopic structure imper-
fectly known.

GAMETOPHYTE *

PROTHALLIUM.—(F. 1-38). The gametophyte of Vittaria
lineata is an irregularly branched prothallium (Y. 1-4), consist-
ing of a single layer of distended cells containing chlorophyl (f.
14, 15). New prothallia may arise from any one of the marginal
cells, or from two or more of the lamellar cells of the prothallium
(f. 5-8). This capacity for vegetative reproduction by prolif-

eration of the prothallus must be of great service to this species
he attack of small

animals. Almost any fragment of the prothallus might continue

growth independently.

The younger portions of the prothallia assume à horizontal
position at right angles to the trunk of the tree to which they are
attached, extending to a distance of 9 mm. in some instances. As
the formation and growth of rhizoids proceeds successively in

based was collected by Miss
he bark of the palmetto.

* The material on which the anatomical study is
Abbie M. Small at St. Augustine, Florida, 22 Feb. 1901, ont

.*

196 THE Lire HISTORY OF

these younger regions, however, the rhizoids penetrate the bark of
the tree and then draw the prothallus down against it. The
rhizoids may be formed in clusters (f. 7) from an undifferentiated
cell of the prothallus, or from a proliferation consisting of a single
cell. Rarely more than one rhizoid originates from a single cell.
The rhizoids are long colorless structures rarely branching near
the apex, and sometimes distended at the base (/ 73).

The prothallium is occasionally monoicous-autoicous, but
more frequently pseudo-dioicous with the antheridia and arche-
gonia produced indirectly from the same plant. In such instances
the archegonia are borne on the main prothallium, which also pro-
duces gemmae from which antheridia originate.

Barrel-form or spherical sterigmata containing chlorophyl
arise from the margin, apex and occasionally from the middle of
the prothallus bearing gemmae in great numbers. A sterigma
may give rise to one or several gemmae, and these gemmae may
divide repeatedly before falling off. The gemmae may either lie
on the surface of the prothallus or stand out from it after the
manner described by Goebel, who found the first position assumed
in Vittaria elongata and the second in Monogramma paradoxa
(f. 19-25)* A sterigma may give rise to several sterigmata
before the gemmae are produced (f. 22). The gemmae are very
similar to those of Vittaria elongata as described by Goebel and are
in the form of spindles composed of § to 11 moniliform cells (f.
25, 29). The terminal cells may give rise to rhizoids, but these
and the antheridia may be produced from any one of the cells of
the spindle (f. 30-33). The gemmae produce prothallia in va-
rious ways, and also may develop other gemmae, and antheridia
(f. 33-37).

ANTHERIDIA.— The antheridia are produced either directly
m one or more cells of the gemmae (f. 39, 40), or on the
ower surface of the prothallium, the margin and occasionally on
iam upper surface. A prothallium sometimes bears both antheridia
ws سس‎ 7 36). The development of the antheridia is

È in any one of the positions named sends out a
protuberance containing chlorophyl which becomes club-shaped

“Gobel X Mali dal Das Gon. a i
و‎ os pholog e and Biologische Studien, Ann. Jard. Bot. Bui-

VITTARIA LINEATA 197

and then cuts off the terminal portion by a partition wall, making
two cells of unequal size, the lower being much larger (/.
400, 41). The apical cell becomes almost spherical, and the next,
wall formed in it is funnel-shaped, connecting with the wall
previously formed in such manner as to separate a central cell. A
third wall is now formed: in the apical portion of the peripheral
cell cutting off a cap cell shaped like a watch glass (f. 43-45).
The antheridium now consists of four cells, a pedestal cell, a hemi-
spherical central cell enclosed by a cylindrical peripheral cell, and
crowned by a cap cell (f. 42). Further division occurs in the
peripheral cell in rare instances. The first division wall in the
central cell is generally longitudinal, and the second one is at
right angles to it, and this sequence is followed during the forma-
tion of an indefinite number of mother cells (f. 42, 44).

An antherozoid is produced in each mother cell The con-
necting walls of the mother cells dissolve and the antheridia lie free
in the antheridium (£ 48a). When the antheridium is ripe the
antherozoids are ejected by the swelling of the peripheral cell and
of the upper contiguous portion of the pedicel (f. 49) which is
pushed up into the antheridial cavity; the cap cell is ruptured
either at the center or at one side. When the antherozoids escape
from the antheridium they are enclosed by a thin membrane, which
is thrown off shortly after their liberation. The antherozoids un-
coil with a whip-like motion, curling and uncurling very rapidly
during a few seconds, the movements of the cilia increasing in
rapidity. The action of the cilia propels the antherozoid at a com-
paratively great speed, and they remain active for some time
(Æ 50).

Four antheridia were seen in the act of discharging the an-
therozoids: the latter escaped in rapid succession, although two
or three came out together in some instances. It was not difficult
to count the emerging antherozoids, and two antheridia were found
to contain twelve and another twenty-four. One antheridium was
emptied in a little more than a minute, and the antherozoids burst
the enclosing membrane in forty seconds or less after emergence.
The entire number (twelve) had freed themselves in this manner in
less than two minutes from the time action began. A portion of
the mother cell in the form of a hyaline sphere with a granulated

198 Tue Lire History OF

center was carried by the antherozoids to which it was still at-
tached forty-five minutes later.

Archegonia (f. 57-60).—The few archegonia that were seen
occupied positions similar to those of Vittaria elongata as described
by Goebel, being borne on the under surface of the prothallium, and
also on the small lobes among the rhizoids. The archegonia have
a slightly curved neck, although the number of cells shown on the
concave and convex surfaces are the same, being four or five in
number.

THE SPOROPHYTE

The sporophyte generally remains attached to the gametophyte
for some time after its formation (/ 89-97).

RuizowE.—The main axis of the sporophyte consists of a dor-
siventral rhizome with short internodes bearing the tufted leaves
in two ranks, the roots are very numerous and completely clothe
the lower surface of the rhizome. The growing portion of the
rhizome appears as a fleshy green bulbous formation with a con-
ical apex, completely enveloped in long brownish scales ) 56,
39). Glands, similar to those of the leaves and hairs, resembling
those of the root are also present. The apical cell is wedge-
shaped and the differentiation of the tissues follows quite closely
the advance of the apical meristem ) 67, 62).

In the earlier stages of development of the stelar tissues they
appear as a cylinder of thin-walled elements (7. 63) in which the
scattered groups of protoxylem soon appear (f. 64d). The
metaxylem next appears as a complete cylinder of scalariform
tracheids (f. 65). The metaphloém is much more highly devel-
oped externally to the cylinder and consists of numbers of sieve
cells separated from the metaphloém and from the pericycle by
parenchymatous elements. The pericycle is well defined (f. 654).
= مهم‎ u traced upward along the stem, gaps appear

leat (f. 66 oues of the fibrovascular ring is sent out into
ar íi Wi eatin see pe which is composed of ex-
somewhat thickened he ng et ber
mis does not participate en re organo. Thé spin

n indeed i‏ و ردو
zomes opposite the eee ie v mbers from the 8‏

s and are usually quite advanced in devel- -

VITTARIA LINEATA 199

opment as soon as visible. Hairs are produced profusely, origi-
nating .5 mm. from the tip. The roots quickly penetrate the bark
of the palmetto tree to which the plant is attached, and as they are
exposed to the light in various places form chlorophyl.

The root hairs measure from 0.6-4.5 mm. in length and 5
.0299 in width. The hairs seldom form cross walls but occa-
sionally branch near the apex, although the branches do not
attain any considerable length. The roots branch profusely:
one 13 mm. long produced 6 branches and 1 branch 7 mm. long
produced 3 branches (f. $8) The central cylinder of the root
is apparently diarchic (f. 86, 87) and the endodermis is plainly
differentiated. The cortex consists of 4-6 layers of large cells
with strongly differentiated walls; the outer rows show this stri-
ation less plainly however. The outer layer immediately in-
ternal to the epidermis contains large numbers of crystalline
bodies connected with their inner walls (f. 87, 88). The epi-
dermal cells are large, the thin outer walls being pressed inwardly
in many instances.

STERILE Leaves. — The leaves of Vittaria lineata show pro-
gressive variation in form and venation (f. 89-95). Those which
arise earliest are spatulate and have a single central vein, later
they show a branching venation ) $9, 93). The laminar por-
tion of the first leaves is flattened and the cross section does not
show such complete dorsiventral differentiation as may be found
in the leaves arising later. They are but slightly curved when
they emerge from the scales, and the curvature soon disappears.
The epidermal cells are somewhat longer than broad with wavy
walls (f. 97). Among these cells are a few elements with
pointed ends and wavy walls (/. 98a), which are most abun-
dant near the fibrovascular bundles. The epidermal cells gener-
ally contain chlorophyl with the exception of the hyaline border.
The latter is composed of elongated elements with oblique end
walls (f. 96a). The epidermal cells of the petioles are pros-
enchymatous and rarely have wavy walls. Stomata occur on the
under side of the lamina but do not occur immediately external
to the veins. Numbers of club-shaped bodies, composed of three
cells (f. 97a), resembling glands, are to be found on both sur-
faces. Scales arise on the basal portion of the earlier leaves di-

200 THE Lire History OF

rectly external to the bundles, while these formations occur more
abundantly on leaves developing later yet they are rarely over a
bundle ( f. 99).

Leaves formed from the rhizome at a later stage of growth
show a variation in form from oblanceolate to linear ) 94, 95).
Such leaves have a vein extending through the middle and one
along each margin with numerous anastomosing branches. Glands
and scales are somewhat more numerous and the epidermis con-
tains but little chlorophyl.

The latest leaves formed show the greatest length and are
ridged over the nerves on the upper surface, bearing scales between
the central and marginal veins (f. 709). These scales, while no
more numerous on the upper surface than on the lower, are fre-
quently larger, measuring as much as 1.9 mm. in length. It is
found that the regularity of arrangement of the stomata and the
increase of the ridges of the leaves increases as a form approxi-
mate to that of the adult is reached ( f. 704).

SPOROPHYLLS.— The sporophylls are long linear coriaceous
leaves with a short petiole exhibiting the epinastic growth of so
many ferns. The numerous scales which doubtless serve as a
protection to the unrolling leaves persist in mature forms (f. 709-
III) The dorsal surface is marked by two deep grooves imme-
diately external to the lateral veins, and these grooves contain the
stomata, glands, sporangia and accompanying paraphyses ( f. 104-
709). The two fascicles of stelar elements given off from the
edges of the incomplete cylinder in the foliar gap in the rhizome do
not fuse until some distance in the petiole has been traversed, and
the lateral branches are given off before the fusion is complete
(f. 102, 103, 1, 6, 7). The lateral veins and the central vein are
connected by various anastomosing branches (f. 102-100). The
central vein runs quite to the apex of the leaf (f. 94-95), while
the lateral veins end free a short distance below the tip or termi-
nate in an anastomosing branch (f. 103-109). The vascular ele-
ments of the leaves are the same as the rhizome in composition,
and the bundles are enclosed in a sheath of two to six layers of cells
iin SR walls ) 108d, f). The mesophyll of the upper
anions peius closely packed and contains a large amount of

; giving this surface a dark green color (f. 774). The

VITTARIA LINEATA 201 .

mesophyll of the lower side is more loosely arranged and contains
less chlorophyll (f. 776).

The epidermis is composed of elements with very thick walls
except in the grooves (f. 714-119). A reduction in size is
noticeable over the middle and lateral veins and at the margins of
the leaves. Chlorophyl is to be found in the epidermal cells in
the grooves but seldom elsewhere. The epidermis of the grooves
is uniformly thin-walled and gives rise to stomata and glands ; the
latter are also found over the entire surface of the sporophyll. A
narrow band of two to six rows of cells immediately external to
the bundle underlying the groove bears the sporangia and para-
physes (f. 723e). One to four layers of prosenchymatous cells
are to be seen underneath the epidermis of the upper side of
the leaf (f. 7746). Similar elements are to be found on the
lower side of the leaf between the grooves, and these cells
are somewhat larger than those of the upper surface (f. 71160).
These hypodermal prosenchymatous elements are generally free
from chlorophyl and are separated by numerous large air-spaces,
giving the region a light gray color. The absence of these ele-
ments in some instances cannot be accounted for.

SCALES. — Scales are most abundant on rhizomes and least so
on the leaves produced earliest (f. 72-74). The first leaf formed
may bear but one scale while the later leaves may be covered
with them. The scales are never more than one cell in thick-
ness and nine or ten in width, attaining a length of 5 mm. and a
breadth of .4-1.1 mm. The growth of these structures is long-
continued and the basal cells may be in a state of division while
the apical portion may be withered and brown (f. 75-82). The
thin walls of the younger cells at the base measure 1.9 a in
thickness which later increases to as much as 19.2 u, which is a
most noticeable multiplication. This thickening is confined to
the cross walls, the outer walls not participating in this increase
(f. 76a, 81, 82a). The thickened walls show great numbers of
dots or minute papillae, which are most numerous in scales borne
on the stems and are not easily found in scales of the earlier leaves.
Both the thickening of the walls and the formation of papillae are
less marked in the marginal region of the scales.

STOMATA. —Stomata are present only in the grooves of the

202 THE Lire HISTORY or

leaves. In the formation of these organs the upper portion of an
epidermal cell is cut off by a wall having a horse-shoe shape in
surface view (/. 7245). The newly formed cell appears to con-
tain chlorophyl and the nucleus increases in size while the
protoplasm becomes granular (f. 725c). This mother cell
divides along a line parallel to the long axis of the epidermal
cell from which it arose (f. r26e). The later stages of the splitting
of the wall and the development of the guard cells follows in the
usual manner (f. 727). The original epidermal cell from which
the stomatal mother cell was cut off undergoes some alterations
and it is difficult to distinguish it from its neighbors (f. 121, 122).

GLanDs. — The glands are club-shaped, measuring about .08
mm. in length and consist of three cells. These organs originate
from the upper or apical end of the epidermal cells on both sur-
faces of the leaves and are inclined toward the apex of the leaf.
The cross walls become suberized in a few cases and turn a rich
golden brown (#97, £10; 112, 128).

SPORANGIA. — The sporangia and the attendant paraphyses are
borne on the epidermal cells in the bottom of the two grooves on
the under sides of the sporophyll ) 723, 129). The margin
of the leaf and the walls of the groove fold over and cover the
sporangia and form a false indusium. An examination of any
single sporophyll will show the presence of sporangia in all
stages of development and these bodies are produced in such
great profusion that many are not brought to full development.
The principal features in the development of the sporangia are
as follows : An epidermal cell sends out a tubular extension
m which a wall soon appears parallel with the outer wall of
the epidermis (f. 130

VITTARIA LINEATA i 208

and a fifth central tetrahedral spore mother cell (f. 130, 136d).
The cells which are to form the annulus divide by parallel walls
into 13-15 cells (f. 737) which extend from the stalk at the base
of the sporangium over the back a short distance down the other
side, where they join with the thin-walled cells which connect
with the lip cells (f. 7336). When the sporangium is young,
these cells have thin walls and become inflated and raised above
the other cells (f. 733a). These cells contain chlorophyl which
is often retained until the spores are fully formed.

As the sporangium ripens, the walls of the annulus thicken
considerably with the exception of the dorsal wall, becoming
cuticularized (f. 7344) and take on a golden brown color
which later turns to a dull brown. The dorsal walls turn a dull
yellow, and frequently bend back toward the inner walls of the
cells, so that the cross walls form ridges along the annulus of the
old sporangium. The other cells of the sporangium also divide.
The frontal cells are generally eight in number and extend from
the annulus to the pedicel at the base of the sporangium. They
are composed of two kinds of cells, those whose walls do not
thicken to any extent—called connective cells—and those whose
walls thicken considerably forming the lips. There are four
lip cells and they occupy the center of the row of seven or
eight, occasionally nine (f. 233c, I 34c) These lip cells differ
greatly from the other cells of the sporangium, having a short
longitudinal diameter and a very long radial one. They are con-
nected with the annulus by one to three thinner walled cells,
and with the pedicel by two (/. 13366, 13400). The two lip
cells, which occupy the center of the group of four, frequently
thicken their walls to a greater extent than the other two, and it
is between these two cells that the sporangium opens when the
spores are ripe ( f. 754), the slit extending back generally in a
straight line through the thin walls of the other cells almost to
the annulus. The central tetrahedral cell of the sporangium
divides by oblique walls until another tetrahedral cell has been
formed inside of it (f. 730 gf), this inner cell dividing up into
the spore mother cells and the cells between this cell and those
constituting the sporangial wall form the tapetal cells ) 130,
98), which dissolve after a time so that the spore mother cells

204 Tue Lire HISTORY OF

lie free in the sporangium. Each mother cell divides into four
spores. These ripe spores are reniform with a smooth surface.

PEDICEL. 一 The pedicel is long and consists of three or more
cells, though only one cell wide until near the sporangium, where
the first longitudinal walls are formed (f. 739, 9, 133€, 1346
135, 7, 2); here it is generally two cells in width, less often three,
though at the base of the sporangium there are almost always
three cells. These small cells are formed at or near the front
of the sporangium (f. 733, 734). As the sporangium ripens
all the cells of the upper part of the pedicel become inflated and
throw the sporangium back ( f. 134), so that when the split occurs,
' and the lip cells open by the everting of the annulus, the spores
have a less interrupted access to the opening of the groove, their
original position being such that did they keep it during the dis-
charge of the spores many of them would become entangled
among the branches of paraphyses from which they would be
powerless to escape.

PARAPHYSES. — The paraphyses develop before and with the
sporangia and frequently attain a large size while the sporangia are
still very immature. They are always multicellular and branched,
except when young (f. 129) ; they grow to some length as a single
cell before they branch into two or three divisions ( 729, 136-
139), the first cell of each forming a stalk. The round branches
generally branch in their turn. The cells are two or three times
longer than wide, each containing a good-sized nucleus and some
chlorophyl ; as they grow older they become discolored and
shrivel up. The branches terminate in cells which are always
ga long, much wider at their apex and generally curved ) 73 6,

SUMMARY
e — ag of Vittaria lineata is an irregularly branched,
fe s prothallium, bearing the archegonia and giving
gemmae which bear the ant

idi new
prothallia. heridia and also produce

pinea — are produced in great numbers, rarely on the
four pata — with the archegonia. From twelve to twenty-
ntherozoids are produced in one antheridium and the open-

in i
8 occurs in the center of the cap cell, or at its junction with the

VITTARIA LINEATA 205

peripheral cells. It ruptures by the swelling of the peripheral
cell or cells, and the portion of the pedicel immediately under the
antheridium ; this drives the wall that separates them up into the
antheridial cavity, and ejects the antherozoids.

The archegonia occur among the rhizoids near the margin, on
the under side of the older portions of the prothallium ; also on
specialized branches which are formed at the margin of the main
prothallium. They have slightly curved necks, but the rows of
cells in both the anterior and posterior sides are equal in number.

The rhizoids are unicellular, simple or branched, structures
and are produced either directly from a cell of the prothallium or
by the formation of an initial cell.

The sporophyte consists of a creeping, dorsiventral rhizome,
bearing its leaves ordinarily in two ranks. It frequently branches
and the internodes of both stem and branch are so short that the
leaves appear crowded together. The growing point appears as
a fleshy green bulb which is densely covered by brown scales
having a structure peculiar to this genus. The rhizome has a
wedge-shaped apical cell and a concentric, tubular bundle, with
foliar-gaps associated with the leaf- bundles.

The roots are produced near the growing tip and are usually
twice as numerous as the leaves. They have a normal root-cap
and an axial diarch bundle, the two groups of xylem are gener-
ally united after a time by large tracheids. The walls of the
parenchymatous cells are strongly striated and in the young roots
there are crystalline substances formed against the inner walls of
the cells immediately under the epidermis. The roots branch fre-
quently and the branching appears to bear a definite relation to
the original xylem groups of the diarch bundle.

There are three kinds of /eaves; first, the rudimentary ones
which are spathulate and long-petioled, usually with a single
central vein; second, the young sterile leaves, which are short-
petioled, oblanceolate to linear, and show all gradations of vena-
tion, from the simple central vein to the three veins which
characterize the sporophyll. The sporophylls are short-petioled
and linear.

The venation of the leaf arises from two primary bundles, each
giving rise to one smaller one which become the lateral bundles

206 Tue Lire HISTORY OF

of the leaf. The larger portion of the two primary bundles unite
to form the central bundle; these three bundles run parallel with
the long axis of the leaf and are united with each other at intervals
by small strands taking an oblique apical course from the central
one to each lateral one.

The sporophyll has two deep dorsal grooves, running parallel
with the lateral bundles and just over them ; in these grooves occur
' the stomata and glands, and in them are borne the sporangia and
paraphyses. All the leaves bear scales which show a higher order
of development as the leaves approach the type of the sporophyll.

The sporangia are long-stalked and spherical with an annulus
of about fourteen cells, four lip cells and four connectives. The
spores are reniform with a smooth outer coat. The smaller cells
on the pedicel at the base of the sporangium become very much
inflated and when the sporangium opens they tilt it back, thus
aiding in the escape of the spores. The paraphyses are long-
stalked, branched and multicellular; the terminal cells being
longer, curved, and enlarged at the apex.

Explanation of Piates

The figures were drawn with a magnification twice that indicated and the resulting
figures were reduced one half.

PLATE 23

FiG. 1. Under side of a prothallium, showing its irregular branching : a, rhizoids ;

4, ee 2T.
. 2. A young prothallium showing sterigmata along margin. X 7.
Fic. 3. Prothallium showing manner of branching at apex. ۰
F
IG. 4. An older piece with the archegonia just forming. 4, rhizoids ; 4, arche-

gonia ; c, ¿ei d, young sporophyte. X 9.

Fic. 5. M
ore 5. Margin of prothallium at apex, showing branching and 4, sterigmata.

Figs. 6, 7. oe prothallia starting from margin of old ones. >< 120.

Fic. 8. A you g prothallium, a,
single oe of an old, edel x32
Fic.

with three growing points starting from a
. F ; i
9. Four marginal cells of the prothallium, each one giving rise to a rhizoid.

. 10. Same as Fig. 9. X 87.

I. A rhizoid formed from an initial cell. X 87.
. 12. A rhizoid formed from two cells. % 120,
> 13. Rhizoid with inflated. cell. 87,

oo
mar
۳ Oo n

VITTARIA LINEATA 207

Fic. 14. Side view of portion of prothallium nor inflated cells; a, rhizoids ;
à, sterigmata with gemmae and also curved apex.

Fic. 15. Portion of prothallium with apex recurved ; gemmae are shown at the
margin. a 25.
Fic. 16. Apex of prothallium, 2; 4, sterigma and gemmae ; d, filament growing
from the margin ; e, nd f, young prothallium with a sterigma and gemma ; g,
growing end of filam 87.

1۲16 17. A ER showing eleven growing points.

Fic. 18. A filament from a margin of a prothallium. X 45.

PLATE 24
Fics. 19-22. Apex of prothallia showing sterigmata and gemmae. EB:
Fic. 23. Marginal cell, sterigmata and gemmae. X 120.
Fics. 24-26. Prothallia showing many-celled gemmae. X 25.
Fic. 27. Gemmae. X 87.
Fic. 28. Gemmae of six cells : cells a, a forming rhizoids. X 120.
Fic. 29. Same as Fig. 28, only older. X 87.
Fic. 30. A gemma of which one cell has sent out a filament of three cells and an
antheridium. X 87
Fics. 31, 32. Gemmae with rhizoids. X 35.
FIG. 33. Gemma, with an antheridium, 4, a rhizoid at each end and small apro-
thallium originating from two cells. X 45
Fic. 34. Gemma, with rhizoids and prothallium. X 35.
Fic. 35. Gemma with a young prothallium originating from a single cell. X 87.
Fic. 36. A gemma with two withered antheridia, a, a, and a young prothallium,
from the margin of which have arisen three prothallia ۰
IG. ۰ emma of five cells, bearing two rhizoids and two prothallia, the older
one وت‎ sterigmata and branching in two places.
G. 38. Portion of the apex of the parent prothallium with two filaments: one
pr growing and bearing an antheridium, a, and a rhizoid, Û, and the other ا‎
(broken) and bearing two rhizoids.

PLATE 25

. 39. Gemma showing an antheridium, 2, asa small protuberance from one of
the ai; the next cell has given rise to a filament of two cells, ^; c, c, rhizoids.
x 87.

IG. Gemma with an antheridium, a, showing first wall, 4; c, nuclei; d,
Fein het EV 120. In places the gemma is more than one cell wide.
` Fics. 41-49. Stages in the development of the antheridia and antherozoids.
Fic. 41. a, Nuclei. X 210.
G. 42. a, Peripheral cell; 5, cap cell; <, central cell dividing into the mother
cells d the neam X 210.

Fics. 43-45. is tg the different ways the walls may be formed, in tiers or ob-
liquely ; a d X 120.

208 THE Lire HISTORY OF
Fics. 46, 47. Looking down on the antheridium, showing the mother cells, @
10. ;

Fic. 48. A ripe antheridium with the mother cells, æ, lying free, each one con-
taining an antherozoid.

Fic. 49. An antheridium with two antherozoids still inside ; a, ruptured cap cell;
the antheridium also showing the inflated peripheral cell, and inflated upper portion of
the pedice O.

Fic. 50. Three antherozoids, a; 4, cilia; c, portion of mother cell. X 600.

Fics. 51, 52. Two antheridia consisting of more than one peripheral cell. >< 210.

IG. 53. Gemma with three antheridia, one in the act of discharging the anther;
ozoids; a, rhizoid. X 88

Fics. 54, 55. Gemmae with antheridia, a, and rhizoids, 6. \ 120.

Fic. 56. Anantheridium from near margin of parent prothallium ; a, antheridium;
û, pedicel. I

Fic. 57. Under side of wing of prothallium showing archegonia, a, 4, the four
stigmatic cells. X 87.
Fic. 58. A ge with young sporophyte ; a, prothallium ; 4, archegonia ; c-
first leaf of young plant; ¢, root; e, root hairs ; Jj foot, y, scales around the growing
tip of rhizome and next leaf AE

Fic. 59. Wing of a prothallium with three young archegonia, a, and sporophyte,
X ۰

Fic. 60. a of Fig. 59 magnified 35 times, to show young archegonia, 4; ¢, foot
imbedded in the calyptra ; d, rhizome ; 5 €, root. 1

PLATE 6

I 61. Longitudinal section of rhizome near apex, a; 5, young leaf; c, scales.

d Fic. 62. A portion of the same showing the differentiating tissues: a, apical por-
ion ; Û, epidermis; c, cortical êr d, vascular bundle; e, medullary paren-
PEM Í, young scales,

Fic. 63. Cross section of rhizome near tip, a, cortex ; 4, meristem of bundle; %
medullary parenchyma. 0.
à IG. 64. The same further from tip, a, cortex; 5, medullary Be 5
phloém ; 4, xylem, the walls of the tracheids thickening. x 120.

Fic. 65. Older portion of rhizome. Marking as in Fig. 64. X 120.
A T - Cross section of the tubular bundle, a ; showing two root bundles, 4 and
: and the two bundles of a leaf, g, d; with a “foliar gap," e; f, medullary paren-
cayma ; g, cortical parenchyma. >< 45.

Fı =
ms iı 67-71. showing the different aspects of the tubular bundle m

on, a, leaf e. ^, **foliar gap” of leaf above ; ; c, root bundles. X 11
Fics. 72, 73. Outlines of scales. X 13.

F
1G. 74. The same, magnified ten times to show the cells.

VITTARIA LINEATA 209

PLATE 27.

Fic. 75. Basal portion of scale showing the thin-walled cells. X 87.
Fic. 76. s from the margin half way up th le showing the thin outer wall,
«t, and the papillae on the thickened inner walls, 6. >< 120.

Fic. 77. Portion of scale showing the framework of suberized walls, a. 120
Fic. 78. The characteristic twist of the thin portion of the cross walls at margin.
YX 210. .
Fic. 79. View of the back of the scale at base showing bend. > ۰
Fic. 80. Base of scale showing twist. X 25.
Fic. 81, 82. Apex of scales showing the wrinkling of the thin outer wall, a.
X 875.
3. Longitudinal section through tip of root : a, root cap; 4, apical portion ;
"A RE AREN cells ; d, meristem of axial bundle. 45.
. Tip of root showing how the cap cells often curl before being thrown
off — a; 4, tienes cap cells still remaining attached to roo
Fic. 85. Portion of rhizome bundle showing connection of root bundle ; «a, endo-
‘dermis ; 4, phlo&m ; c, xylem; d, cortex ; c, medullary parenchyma. x 120.
Fic. 86. Diarch bundle of young root: a, cortex ; 4, endodermis ; ¢, c, phloém ;
.d, d, two primary xylem groups. X 292.
Fic. 87. Cross section of young root: a, rhizoids ; رم‎ cells containing crystalline
‘substances ; c, endodermis ; 4, the two phloém groups ; e, xylem. 8
Fic. 88. Cross section through root and two of its branches, and a longitudinal
‘section through a third branch at its point of attachment with the first root: 2 w
given off first ; then 3; and then 4. Marking as in Fig. 87. X ۰

PLATE 28

Fics. 89, 90. Prothallia and young sporophytes growing on the bark of the pal-
metto X 4.

Fic. 91. A wing a, of the prothallium showing young sporophyte ; 4, first leaf ;
€, root. X ۰

Fics. 92, 93. Rudimentary primary leaves showing variation. x 2
Fic. 94. Variation of secondary leaf.
95. Mature sterile leaf, the sporophyll shows the same variation with the ex-

d Es the two primary veins, 4, are apt to run a longer distance before anasto-
mosing ; 2 is basal portion of 1, they join at æ ; 5 primary veins. ;

Fic. 96. Apex of first leaf showing epidermis of under side of hyaline border.
X 87%.

Fic. 97. Portion of epidermis of same under side: a, gland ; b, stoma. X 210.

Fic. 98. Portion of epidermis showing prosenchymatic cella. X 210.

Fic. 99. Scale from upper side of first leaf. X 87.

Fic. 100. Cross section of petiole showing vascular bundle with two xylem
groups, b, in primary leaves.

210 THE Lire History or

PLATE 29

Fic. 101. Surface view of thizome, with the leaves removed to show the two-
ranked arrangement, and the two primary bundles, at.
Fic. 102 Diagram of leaf bundles : a, bundle of rhizome; 4, the two primary
bundles ; c, first lateral bundle ; d, second lateral bundle.
FiG. 103. Nine diagrams showing different cross sections of leaves with bundles.

the lateral bundles have been given off; 9 shows the apex of a leaf with one lateral
bundle ending free,

Fic. 104. Diagram of middle portion of leaf: a, central bund] posed of p
of the two primary bundles ; 6, thick-walled cells surrounding bundle ; c, mesophyll
4, band of thick-walled cells and epidermis ; €, groove (under side of leaf) ; /, lateral
bundle; c, denser mesophyll of upper side of leaf. XE.

arta

Fics. 1 ross sections of petioles showing different degrees of ی‎
of the outer parenchymatous cells of the cortex. Fig. 105 is more often found in :
Sporophylls. 1, upper Side; a, primary bundles, In Fig. 106 one lateral bundle is
shown, 4, X 25. :

‘Fic. 107. Cross section of centra] bundle of leaf; a, phloém; 4, xylem ; e, endo-
dermis; q, thick-walled cells of the ground tissue surrounding bundle.

Fic. 108. Cross section of one of the lateral bundles. Marking as in Fig ge
f, mesophyll of under side; f thin-walled epidermal layer which gives rise to
sporangia. X 120, ; ;

Fic. 109. Curled apex of young sporophyll showing scales, a, and cut so as to y
the bundles, ۸۰ and ۱۳۳ ۲0۳9 6. W te.

Fic. 110, Side view of apex of young leaf showing curl and glands. X 13.

Fic. ۰ Front view of same, 3

Fic. 112, Portion of epidermis from under side of young leaf; a, scale; 4, gland;
€, stoma just forming, X.. ۱

PLATE 30
Fic. 113. Apex of leaf. Dotted line indicates the area of the dense mesophyll -
cells, containing chlorophyll, under epidermis. X 87,

IG. 114. Portion of upper side of leaf, Over bundle, seen in cross section ; $
epidermis; 4, thick-walled cells, generally Prosenchymatic, often becoming bast-like+
^ mesophyll cells containing chlorophy], X 87.

Fic. 115, Portion of epidermi
stoma, d. Marking as in Fig, 114.
Fic. 116. Rid

S, 4, from under side of young frond showing i

Sê over bundle on under side, مد‎ 87. in
- Inner mars; e ; 4 3 rking as
Nie vis X 8. rgin of groove extending from ridge to ridge. Ma : |
سب‎ ML folding of the epidermal walls. Fig. 119 has um
Mn with potash to show the Way they fit into each other. —
^ 120, of surface of under side of leaf showing the ridge and
grooves, a, the stomata and glands, X 13.

Fic. 121. Epidermis from groove; q, stoma ; 4, gland. X87.

VITTARIA LINEATA 211

IG. . A, represents width of groove from edge to the sporangial layer; Û,
stoma ; c, initial cell; g, original epidermal cell

Fic. 123. Cross section of groove: a, stoma; 4, epidermal cell; v, air space; d,
thick- ated cells around bundle; e, thin-walled epidermal layer giving rise to the
sporangia, /, and paraphyses, g. 20

Fics. 124-126. Showing origin of the stoma: a, original epidermal cell; 5, in-
itial cell of the stoma; c, the mother cell of the stoma; 4, longitudinal wall Ars
mother cell, in Fig. 126, dividing it into the two guard cells, e, e; f, nucleus; g,
chlorophyl. X 294.

FIG. 127. Mature stoma. Dotted line, 4, represents wall of mother cell, the two
guard cells being raised somewhat above the epidermis as in the cross section (Fig.
123). Marking as in Figs. 124-126. X 310

Fic. 128. Gland. X 294.

PLATE 31

Fic. 129. A diagram of the longitudinal section of the leaf passing through center
of groove : a, upper epidermis and band of thick- و‎ cells ; 4, mesophyll ; c, c, band
of thick-walled cells surrounding bundle; 4, lateral bundle; e, thin-walled layer of
epidermis giving rise to the sporangia ; /, sporangia here shown in all stages of devel-
«ad, g, the paraphyses. X 25.

130. ent sections showing the development of the sporangium : a, or iginal

cell x pd sporangium; 4, first oblique wall; c, second wall; 4, first tetrahedral cell ; e,
CE cell dividing to form the sporangial wall. No. 9 shows the divisions of T

trahedral cell, which form the mother cells of the spores and the tapetal cells, £,
pelis No. 5, surface view of very young sporangium. X 120.

FIG. 13 urface views of older sporangia: «a, annulus; 4, lip
cells just forming ; c, cells of pedicel at base of sporangium. X 120.

FIG. 132. Mother cells from a yourlg sporangium before dividing to form the spores.

10,

133. Young و‎ all the cells containing chlorophyl, except the lip
ce ۱ a, دس‎ us; 4, 4, four connective cells; <, the four lip cells; 4, thin-walled
cells of ag at wall; e, prin x 120
G. 134. Ripe en showing the way in which it opens : f, spores.
ws as in Fig. 133. X 1
135. Different ways in which cells of divide ; part of the basal por-
tion xs is sporangium is shown at a in No. 1. 1, X 87. No. 2, X 120.

Fic. 136. Four cells SF epidermal layer, a, one of which has given rise to a
T 1 û c, first tetrahedral cell in process of dividing (this stage is be-
tween Nos. 8 and 9 of Fig. 3 IE ^ To wall; e, pedicel ; /, young paraphy-
sis from adjoining cell containing a nucleus and some chlorophyl. X 120

Fics. 137, 138. Young 9۵ ru 120.

Fic. hi Seven diagrams showing the branching of the paraphyses.

Fic. 1 . Young sporophyte, 4, growing from mutilated end of rhizome, the
growing ae id rhizome having been removed ; a, base of two old sporophytes.

Mark-

FL

“Mem. Torr. Bor. CLUB, 8.

C

> Us

IR ;
ی‎ TY gic 2
۸ Ba?

PL.23 —
LAIR
LS CH

ی

d
B
2
en
2
<
E
=

EM. TORREY CLUB, 8

M

Mem. TORREY CLUB, 8 ; PL. 3

> 0. VITTARIA LINEATA —

PL. 35

8

Mem. Torrey CLUB,

b,

IL
ies.

Lao

WE AÑ

VITTARIA LINEATA

PL. 26 ;

Nr

Y

HE
IO

al

J
wp JES

N :
\ >

> VITTARIA LINEATA

J/

0

i H

HE y:

A

un
;

$

È

v M n
Wi y

Bun;

Mem. Torrey CLUB, 8

VE e
- 1
PAR
ZIT

KH]
0:
AY
to

e

5.
=

SIT 7
EEE

RE

va

Mem. Torrey CLUB, 8

4
E
A
2
中
<
-
:
-

PL. 28

Mem. Torrey CLUB, 8

x

a...

Ó

O

E:
E
<
B
á
2
a
—
E
E
E
Lan!
-

MEM. Torrey CLUB, 8

Dy
"2

VITTARIA LINEATA

Mem. Torrey CLUB, 8

H

INN

VITTARIA LINEATA

n

VITTARIA LINEATA

8

,

a 27 +
(o SZ
Seria’

. 023 34758 Ye
NES Vex

[LTD Aes see E

۳
v
6

€

Mem. TonREY CLUB

Memoirs of the Torrey Botanical Club

The MEMOIRS are published at irregular intervals and are sold at a uniform
price of THREE DOLLARS a volume. Most of the articles can be furnished
separately at the prices indicated below. The price includes postage.

Volume 1.
No. 1.—Bailey, L. H. Studies of the Types of various Species of the Genus
Cares. 1-85. 1889. Price, $1.00. i
No. Martindale, I. C. Marine Algae of the New Jersey Coast and adjacent
Waters s Staten Toni. Pp. 87-109. 1889. Price 50 cen
o. 3.—Spruce, R. An Mire of the Hepaticae e by Dr. H. H.
Rusby ? South America. Pp. 113-140. 1890. Price, 75 ce i
No. turtevant, E. L. 2 seedless Fruit. Pp. دیدن‎ 1890. Prio; -
75 an
Volume 2.
. No. 1.—Halsted, B. D. On reserve Food en in Buds and surrounding
Parts. Pp. 1-26. Pl. 1,2. 1890. Price, 50 e
No, 2.—Vail, A. M. and Hollick, A. ne the noter of Virginia.
Pp. 27-56. Pl. 3, > 1890. | Priee, 78. centa,
No Ao مج‎ pedes ع‎ A
North American a 2 nd Pl. 5-19. 1891. (Not furnished separately. ) Be
No. 4.—Wheelock, W. E. onograph of the rm American PINE of 2
the Ginu Polygala. Pp. PRESS a Price, 75 ce

ae

Volume 3.
— Small, J. K. and Heller, A. A. On the Flora of Western North I T

No. 2,—Mor
trations of all the Species. Pp. 1-65. PI. 20-74. 1893. Price, SES
No. 3.—Rusby, H. H. An Enumeration of the Plants collected in Bolivia da p
Miguel Bang. Pp. 1-67. 1893. Prios, 50 cente. - :

Volume 4. — — vex.
No. 1.—Underwood, L. M. Index Hepaticarom. Part 1, Bibliography. =
1-91. 1893. Price, 75 cents. —
No. 2.—Vail, A. M. and Small, J. K. Report on th he Botanical Explora
Virginia during the Season 1892, PR 92-202, Pl. 75-82. REIN Price,
cents.

ation

No. 3.—Rusby, H. H. An Enumeration of the Plants collected in Bolivia by
Miguel et = RER 1898. Price, 50 cents, E
is hypogaea E Pp. 275-206. 1895. Price

5 Memoirs of the Torrey Botanical Club

` The MEMOIRS are published at irregular intervals and are sold at a uniform
` price of THREE DOLLARS a volume. Most of the articles can be furnished
T — at the prices indicated below. The price includes postage. :

Volume 6.
o. 1.—Rusby, H. H. An Enumeration of the Plants collected in Bolivia by
p ASTIN Pp. 1-130. 1896. Price, $1.25.
A. J. Revision of the aa American Isotheciaceae and
adi Pp. a 1897. Price, 50 cents.
No. 3.—Hazen, T. E. e Lite History of a lacustris (Haematococcus
— Pp. 211-244. Eg: 86, 87 (colored). 1899. Price, 50 cents.
—Underwood, L. M. A sass s the Genera of Ferns proposed prior
Ad Sr Pp. 247-283. 1899. Price, 50 ce
No, 5.—Fink, B. or on the Lichen bibas in the Upper Mile
Valley. Price, 25 cen

an Volume 7.
. Howe, M. A. The Hepati d Anthocerotes of California. Pp. 1-208. Pl.
88-122. 1899. Price, $3.00.

a Volume 8. |

. No. 1.—Lloyd, F. E. The comparative Embryology of the Rubiaceae. Part
di Pp. 1-26. PL t4. 1899. Price, 50 cents. Part II. Pp. 27-111. Pl. 2-15. 1902.
$1259

: The Lejeuneae ot the United States and Canada. Pp- 1
Tu PL E a rk $1.00. oe
No. 3.— E.G. & aylor, ~ The Life History of Vittaria lineata.
185-211. PL 89-81, e, Price, 75 cents, E

mr E ون‎

on, M A رخ و‎ ot the Erysiphaeeae, ten 1-292. Pl. 14
jus ۰۶ ۰: ۰. Volume 10.
or ie , E. S. Studies in the Genus Aster. Part L Pp. 1-284.

a‏ | : ده
D. The North 3 Sordariaceae. Pp. | de Pl. r‏

e T The Ulotbileadius end اند‎ ot the Uni
= ger ۳ er 1902, و‎